{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.423542989036353,
  "eval_steps": 3000,
  "global_step": 21000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 3.7323212385177613,
      "epoch": 0.0005770340450086555,
      "grad_norm": 34.0,
      "learning_rate": 2e-06,
      "loss": 13.6143,
      "mean_token_accuracy": 0.042970035597682,
      "num_tokens": 13337.0,
      "step": 5
    },
    {
      "entropy": 3.6967998266220095,
      "epoch": 0.001154068090017311,
      "grad_norm": 32.5,
      "learning_rate": 4.5e-06,
      "loss": 13.5571,
      "mean_token_accuracy": 0.04257573280483484,
      "num_tokens": 26339.0,
      "step": 10
    },
    {
      "entropy": 4.161470651626587,
      "epoch": 0.0017311021350259665,
      "grad_norm": 32.5,
      "learning_rate": 7e-06,
      "loss": 13.2735,
      "mean_token_accuracy": 0.030668590404093265,
      "num_tokens": 39118.0,
      "step": 15
    },
    {
      "entropy": 5.517871713638305,
      "epoch": 0.002308136180034622,
      "grad_norm": 36.25,
      "learning_rate": 9.5e-06,
      "loss": 12.3718,
      "mean_token_accuracy": 0.020669543370604514,
      "num_tokens": 51367.0,
      "step": 20
    },
    {
      "entropy": 9.014927005767822,
      "epoch": 0.0028851702250432777,
      "grad_norm": 14.375,
      "learning_rate": 1.2e-05,
      "loss": 11.0046,
      "mean_token_accuracy": 0.017831112071871758,
      "num_tokens": 64639.0,
      "step": 25
    },
    {
      "entropy": 10.463701057434083,
      "epoch": 0.003462204270051933,
      "grad_norm": 4.9375,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 10.5931,
      "mean_token_accuracy": 0.01491671558469534,
      "num_tokens": 76244.0,
      "step": 30
    },
    {
      "entropy": 10.491576671600342,
      "epoch": 0.004039238315060588,
      "grad_norm": 2.96875,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 10.3479,
      "mean_token_accuracy": 0.02280671913176775,
      "num_tokens": 88490.0,
      "step": 35
    },
    {
      "entropy": 10.193946361541748,
      "epoch": 0.004616272360069244,
      "grad_norm": 2.515625,
      "learning_rate": 1.95e-05,
      "loss": 10.0133,
      "mean_token_accuracy": 0.048479363694787024,
      "num_tokens": 102022.0,
      "step": 40
    },
    {
      "entropy": 10.412507057189941,
      "epoch": 0.0051933064050779,
      "grad_norm": 2.046875,
      "learning_rate": 2.2e-05,
      "loss": 9.8157,
      "mean_token_accuracy": 0.06318248882889747,
      "num_tokens": 114442.0,
      "step": 45
    },
    {
      "entropy": 10.123257160186768,
      "epoch": 0.005770340450086555,
      "grad_norm": 2.015625,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 9.7698,
      "mean_token_accuracy": 0.06782412566244603,
      "num_tokens": 128044.0,
      "step": 50
    },
    {
      "entropy": 10.276771545410156,
      "epoch": 0.006347374495095211,
      "grad_norm": 1.7265625,
      "learning_rate": 2.7e-05,
      "loss": 9.7272,
      "mean_token_accuracy": 0.062359335273504256,
      "num_tokens": 141818.0,
      "step": 55
    },
    {
      "entropy": 10.322304344177246,
      "epoch": 0.006924408540103866,
      "grad_norm": 1.75,
      "learning_rate": 2.95e-05,
      "loss": 9.5725,
      "mean_token_accuracy": 0.06548289433121682,
      "num_tokens": 155109.0,
      "step": 60
    },
    {
      "entropy": 10.235727787017822,
      "epoch": 0.0075014425851125215,
      "grad_norm": 1.59375,
      "learning_rate": 3.2e-05,
      "loss": 9.5017,
      "mean_token_accuracy": 0.07019429132342339,
      "num_tokens": 167036.0,
      "step": 65
    },
    {
      "entropy": 10.236877536773681,
      "epoch": 0.008078476630121177,
      "grad_norm": 1.3046875,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.5416,
      "mean_token_accuracy": 0.0642974752932787,
      "num_tokens": 181398.0,
      "step": 70
    },
    {
      "entropy": 10.093268299102784,
      "epoch": 0.008655510675129832,
      "grad_norm": 1.5546875,
      "learning_rate": 3.7e-05,
      "loss": 9.3665,
      "mean_token_accuracy": 0.07238649874925614,
      "num_tokens": 193242.0,
      "step": 75
    },
    {
      "entropy": 10.19960355758667,
      "epoch": 0.009232544720138488,
      "grad_norm": 1.4609375,
      "learning_rate": 3.95e-05,
      "loss": 9.3623,
      "mean_token_accuracy": 0.07175844348967075,
      "num_tokens": 206744.0,
      "step": 80
    },
    {
      "entropy": 10.094919395446777,
      "epoch": 0.009809578765147143,
      "grad_norm": 1.5,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.2726,
      "mean_token_accuracy": 0.07449138984084129,
      "num_tokens": 219120.0,
      "step": 85
    },
    {
      "entropy": 10.15459098815918,
      "epoch": 0.0103866128101558,
      "grad_norm": 1.8828125,
      "learning_rate": 4.45e-05,
      "loss": 9.2401,
      "mean_token_accuracy": 0.07340459935367108,
      "num_tokens": 231942.0,
      "step": 90
    },
    {
      "entropy": 9.965401363372802,
      "epoch": 0.010963646855164455,
      "grad_norm": 1.6953125,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 9.1966,
      "mean_token_accuracy": 0.07440538741648198,
      "num_tokens": 244786.0,
      "step": 95
    },
    {
      "entropy": 10.01766061782837,
      "epoch": 0.01154068090017311,
      "grad_norm": 1.9921875,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 9.036,
      "mean_token_accuracy": 0.08044340573251248,
      "num_tokens": 257503.0,
      "step": 100
    },
    {
      "entropy": 9.868921279907227,
      "epoch": 0.012117714945181766,
      "grad_norm": 1.4921875,
      "learning_rate": 5.2e-05,
      "loss": 8.9711,
      "mean_token_accuracy": 0.08332463353872299,
      "num_tokens": 270330.0,
      "step": 105
    },
    {
      "entropy": 9.76955509185791,
      "epoch": 0.012694748990190421,
      "grad_norm": 2.09375,
      "learning_rate": 5.45e-05,
      "loss": 9.0165,
      "mean_token_accuracy": 0.08288254365324974,
      "num_tokens": 283866.0,
      "step": 110
    },
    {
      "entropy": 9.99423532485962,
      "epoch": 0.013271783035199077,
      "grad_norm": 1.3984375,
      "learning_rate": 5.7e-05,
      "loss": 8.9324,
      "mean_token_accuracy": 0.07315073981881141,
      "num_tokens": 296715.0,
      "step": 115
    },
    {
      "entropy": 9.858692646026611,
      "epoch": 0.013848817080207732,
      "grad_norm": 1.2265625,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.8353,
      "mean_token_accuracy": 0.07335402593016624,
      "num_tokens": 309268.0,
      "step": 120
    },
    {
      "entropy": 9.587060260772706,
      "epoch": 0.014425851125216388,
      "grad_norm": 1.4609375,
      "learning_rate": 6.2e-05,
      "loss": 8.7874,
      "mean_token_accuracy": 0.08877900540828705,
      "num_tokens": 322471.0,
      "step": 125
    },
    {
      "entropy": 9.701264572143554,
      "epoch": 0.015002885170225043,
      "grad_norm": 1.484375,
      "learning_rate": 6.450000000000001e-05,
      "loss": 8.7586,
      "mean_token_accuracy": 0.0769478652626276,
      "num_tokens": 335728.0,
      "step": 130
    },
    {
      "entropy": 9.516640281677246,
      "epoch": 0.015579919215233698,
      "grad_norm": 1.140625,
      "learning_rate": 6.7e-05,
      "loss": 8.6412,
      "mean_token_accuracy": 0.07981666177511215,
      "num_tokens": 347861.0,
      "step": 135
    },
    {
      "entropy": 9.311244106292724,
      "epoch": 0.016156953260242354,
      "grad_norm": 1.7109375,
      "learning_rate": 6.950000000000001e-05,
      "loss": 8.6347,
      "mean_token_accuracy": 0.08614776656031609,
      "num_tokens": 360275.0,
      "step": 140
    },
    {
      "entropy": 9.333715534210205,
      "epoch": 0.01673398730525101,
      "grad_norm": 1.0078125,
      "learning_rate": 7.2e-05,
      "loss": 8.571,
      "mean_token_accuracy": 0.07833859249949456,
      "num_tokens": 373948.0,
      "step": 145
    },
    {
      "entropy": 9.198777198791504,
      "epoch": 0.017311021350259664,
      "grad_norm": 0.90625,
      "learning_rate": 7.45e-05,
      "loss": 8.5338,
      "mean_token_accuracy": 0.07528047934174538,
      "num_tokens": 386267.0,
      "step": 150
    },
    {
      "entropy": 9.142632865905762,
      "epoch": 0.01788805539526832,
      "grad_norm": 1.078125,
      "learning_rate": 7.7e-05,
      "loss": 8.466,
      "mean_token_accuracy": 0.07569638900458812,
      "num_tokens": 398595.0,
      "step": 155
    },
    {
      "entropy": 9.001069831848145,
      "epoch": 0.018465089440276975,
      "grad_norm": 1.2109375,
      "learning_rate": 7.950000000000001e-05,
      "loss": 8.4415,
      "mean_token_accuracy": 0.07843919023871422,
      "num_tokens": 410626.0,
      "step": 160
    },
    {
      "entropy": 8.904146099090577,
      "epoch": 0.019042123485285632,
      "grad_norm": 0.875,
      "learning_rate": 8.2e-05,
      "loss": 8.4353,
      "mean_token_accuracy": 0.0845392294228077,
      "num_tokens": 423176.0,
      "step": 165
    },
    {
      "entropy": 8.805705642700195,
      "epoch": 0.019619157530294286,
      "grad_norm": 1.0703125,
      "learning_rate": 8.450000000000001e-05,
      "loss": 8.393,
      "mean_token_accuracy": 0.08622003495693206,
      "num_tokens": 435997.0,
      "step": 170
    },
    {
      "entropy": 8.710006427764892,
      "epoch": 0.020196191575302943,
      "grad_norm": 0.953125,
      "learning_rate": 8.7e-05,
      "loss": 8.3767,
      "mean_token_accuracy": 0.0884266272187233,
      "num_tokens": 449295.0,
      "step": 175
    },
    {
      "entropy": 8.714795589447021,
      "epoch": 0.0207732256203116,
      "grad_norm": 0.90234375,
      "learning_rate": 8.95e-05,
      "loss": 8.4022,
      "mean_token_accuracy": 0.08629754856228829,
      "num_tokens": 463983.0,
      "step": 180
    },
    {
      "entropy": 8.6787428855896,
      "epoch": 0.021350259665320254,
      "grad_norm": 0.9453125,
      "learning_rate": 9.2e-05,
      "loss": 8.4184,
      "mean_token_accuracy": 0.08215429335832596,
      "num_tokens": 476849.0,
      "step": 185
    },
    {
      "entropy": 8.687536239624023,
      "epoch": 0.02192729371032891,
      "grad_norm": 1.296875,
      "learning_rate": 9.45e-05,
      "loss": 8.3669,
      "mean_token_accuracy": 0.0866314373910427,
      "num_tokens": 489120.0,
      "step": 190
    },
    {
      "entropy": 8.578043460845947,
      "epoch": 0.022504327755337564,
      "grad_norm": 1.0703125,
      "learning_rate": 9.7e-05,
      "loss": 8.2956,
      "mean_token_accuracy": 0.08582662940025329,
      "num_tokens": 501085.0,
      "step": 195
    },
    {
      "entropy": 8.653344535827637,
      "epoch": 0.02308136180034622,
      "grad_norm": 0.81640625,
      "learning_rate": 9.95e-05,
      "loss": 8.363,
      "mean_token_accuracy": 0.08150185570120812,
      "num_tokens": 513207.0,
      "step": 200
    },
    {
      "entropy": 8.69477891921997,
      "epoch": 0.023658395845354875,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000102,
      "loss": 8.2959,
      "mean_token_accuracy": 0.08695114329457283,
      "num_tokens": 524809.0,
      "step": 205
    },
    {
      "entropy": 8.637209606170654,
      "epoch": 0.024235429890363532,
      "grad_norm": 1.125,
      "learning_rate": 0.00010449999999999999,
      "loss": 8.333,
      "mean_token_accuracy": 0.08224210143089294,
      "num_tokens": 538950.0,
      "step": 210
    },
    {
      "entropy": 8.533492851257325,
      "epoch": 0.024812463935372186,
      "grad_norm": 1.25,
      "learning_rate": 0.000107,
      "loss": 8.3069,
      "mean_token_accuracy": 0.09452523067593574,
      "num_tokens": 551199.0,
      "step": 215
    },
    {
      "entropy": 8.584404468536377,
      "epoch": 0.025389497980380843,
      "grad_norm": 1.265625,
      "learning_rate": 0.0001095,
      "loss": 8.2458,
      "mean_token_accuracy": 0.09130036383867264,
      "num_tokens": 563202.0,
      "step": 220
    },
    {
      "entropy": 8.542673492431641,
      "epoch": 0.025966532025389497,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000112,
      "loss": 8.2308,
      "mean_token_accuracy": 0.09193311482667924,
      "num_tokens": 576716.0,
      "step": 225
    },
    {
      "entropy": 8.472547721862792,
      "epoch": 0.026543566070398154,
      "grad_norm": 0.984375,
      "learning_rate": 0.0001145,
      "loss": 8.2372,
      "mean_token_accuracy": 0.09359257072210311,
      "num_tokens": 588413.0,
      "step": 230
    },
    {
      "entropy": 8.6172194480896,
      "epoch": 0.027120600115406807,
      "grad_norm": 0.984375,
      "learning_rate": 0.00011700000000000001,
      "loss": 8.2871,
      "mean_token_accuracy": 0.08904579505324364,
      "num_tokens": 602052.0,
      "step": 235
    },
    {
      "entropy": 8.518135070800781,
      "epoch": 0.027697634160415464,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00011949999999999999,
      "loss": 8.2266,
      "mean_token_accuracy": 0.09081303253769875,
      "num_tokens": 614508.0,
      "step": 240
    },
    {
      "entropy": 8.339745807647706,
      "epoch": 0.02827466820542412,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000122,
      "loss": 8.2549,
      "mean_token_accuracy": 0.09370925053954124,
      "num_tokens": 628758.0,
      "step": 245
    },
    {
      "entropy": 8.498612117767333,
      "epoch": 0.028851702250432775,
      "grad_norm": 1.21875,
      "learning_rate": 0.0001245,
      "loss": 8.1684,
      "mean_token_accuracy": 0.09330120757222175,
      "num_tokens": 641046.0,
      "step": 250
    },
    {
      "entropy": 8.45691146850586,
      "epoch": 0.029428736295441432,
      "grad_norm": 0.953125,
      "learning_rate": 0.000127,
      "loss": 8.1807,
      "mean_token_accuracy": 0.09563048109412194,
      "num_tokens": 653910.0,
      "step": 255
    },
    {
      "entropy": 8.430898475646973,
      "epoch": 0.030005770340450086,
      "grad_norm": 1.09375,
      "learning_rate": 0.0001295,
      "loss": 8.1546,
      "mean_token_accuracy": 0.10104967132210732,
      "num_tokens": 665862.0,
      "step": 260
    },
    {
      "entropy": 8.375243663787842,
      "epoch": 0.030582804385458743,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000132,
      "loss": 8.1507,
      "mean_token_accuracy": 0.10190046802163125,
      "num_tokens": 678854.0,
      "step": 265
    },
    {
      "entropy": 8.442987442016602,
      "epoch": 0.031159838430467397,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00013450000000000002,
      "loss": 8.153,
      "mean_token_accuracy": 0.0963923156261444,
      "num_tokens": 690188.0,
      "step": 270
    },
    {
      "entropy": 8.430098915100098,
      "epoch": 0.031736872475476054,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00013700000000000002,
      "loss": 8.1145,
      "mean_token_accuracy": 0.10212414339184761,
      "num_tokens": 702041.0,
      "step": 275
    },
    {
      "entropy": 8.371403026580811,
      "epoch": 0.03231390652048471,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0001395,
      "loss": 8.1336,
      "mean_token_accuracy": 0.104609215259552,
      "num_tokens": 714109.0,
      "step": 280
    },
    {
      "entropy": 8.367766094207763,
      "epoch": 0.03289094056549336,
      "grad_norm": 1.125,
      "learning_rate": 0.00014199999999999998,
      "loss": 8.1545,
      "mean_token_accuracy": 0.09640730917453766,
      "num_tokens": 726120.0,
      "step": 285
    },
    {
      "entropy": 8.413183784484863,
      "epoch": 0.03346797461050202,
      "grad_norm": 1.265625,
      "learning_rate": 0.0001445,
      "loss": 8.1028,
      "mean_token_accuracy": 0.1032490886747837,
      "num_tokens": 737868.0,
      "step": 290
    },
    {
      "entropy": 8.353041744232177,
      "epoch": 0.034045008655510675,
      "grad_norm": 1.171875,
      "learning_rate": 0.000147,
      "loss": 8.193,
      "mean_token_accuracy": 0.10336777493357659,
      "num_tokens": 751829.0,
      "step": 295
    },
    {
      "entropy": 8.405898094177246,
      "epoch": 0.03462204270051933,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0001495,
      "loss": 8.1197,
      "mean_token_accuracy": 0.10570979490876198,
      "num_tokens": 764457.0,
      "step": 300
    },
    {
      "entropy": 8.260132122039796,
      "epoch": 0.03519907674552799,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000152,
      "loss": 8.1569,
      "mean_token_accuracy": 0.10863806083798408,
      "num_tokens": 776887.0,
      "step": 305
    },
    {
      "entropy": 8.448872661590576,
      "epoch": 0.03577611079053664,
      "grad_norm": 1.265625,
      "learning_rate": 0.00015450000000000001,
      "loss": 8.1229,
      "mean_token_accuracy": 0.10281555876135826,
      "num_tokens": 789001.0,
      "step": 310
    },
    {
      "entropy": 8.287611865997315,
      "epoch": 0.0363531448355453,
      "grad_norm": 1.171875,
      "learning_rate": 0.000157,
      "loss": 8.1195,
      "mean_token_accuracy": 0.1042421966791153,
      "num_tokens": 802719.0,
      "step": 315
    },
    {
      "entropy": 8.269627094268799,
      "epoch": 0.03693017888055395,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0001595,
      "loss": 8.0398,
      "mean_token_accuracy": 0.10676939561963081,
      "num_tokens": 815064.0,
      "step": 320
    },
    {
      "entropy": 8.33767261505127,
      "epoch": 0.03750721292556261,
      "grad_norm": 1.8203125,
      "learning_rate": 0.000162,
      "loss": 8.0822,
      "mean_token_accuracy": 0.11563392803072929,
      "num_tokens": 827778.0,
      "step": 325
    },
    {
      "entropy": 8.284006214141845,
      "epoch": 0.038084246970571264,
      "grad_norm": 1.03125,
      "learning_rate": 0.00016450000000000001,
      "loss": 8.0403,
      "mean_token_accuracy": 0.10939318239688874,
      "num_tokens": 840691.0,
      "step": 330
    },
    {
      "entropy": 8.257933616638184,
      "epoch": 0.03866128101557992,
      "grad_norm": 1.078125,
      "learning_rate": 0.00016700000000000002,
      "loss": 8.0932,
      "mean_token_accuracy": 0.0974455177783966,
      "num_tokens": 852818.0,
      "step": 335
    },
    {
      "entropy": 8.325709438323974,
      "epoch": 0.03923831506058857,
      "grad_norm": 1.3125,
      "learning_rate": 0.00016950000000000003,
      "loss": 8.0914,
      "mean_token_accuracy": 0.11402691900730133,
      "num_tokens": 866854.0,
      "step": 340
    },
    {
      "entropy": 8.225966930389404,
      "epoch": 0.03981534910559723,
      "grad_norm": 1.34375,
      "learning_rate": 0.00017199999999999998,
      "loss": 8.0414,
      "mean_token_accuracy": 0.10918436199426651,
      "num_tokens": 879234.0,
      "step": 345
    },
    {
      "entropy": 8.31819543838501,
      "epoch": 0.040392383150605886,
      "grad_norm": 1.078125,
      "learning_rate": 0.00017449999999999999,
      "loss": 7.9951,
      "mean_token_accuracy": 0.10175195261836052,
      "num_tokens": 890927.0,
      "step": 350
    },
    {
      "entropy": 8.258775520324708,
      "epoch": 0.04096941719561454,
      "grad_norm": 1.203125,
      "learning_rate": 0.000177,
      "loss": 8.0307,
      "mean_token_accuracy": 0.11141346469521522,
      "num_tokens": 902965.0,
      "step": 355
    },
    {
      "entropy": 8.264652919769286,
      "epoch": 0.0415464512406232,
      "grad_norm": 1.203125,
      "learning_rate": 0.0001795,
      "loss": 7.9938,
      "mean_token_accuracy": 0.11707152500748634,
      "num_tokens": 916384.0,
      "step": 360
    },
    {
      "entropy": 8.206478118896484,
      "epoch": 0.042123485285631854,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000182,
      "loss": 7.8976,
      "mean_token_accuracy": 0.12026941180229186,
      "num_tokens": 928618.0,
      "step": 365
    },
    {
      "entropy": 8.14428939819336,
      "epoch": 0.04270051933064051,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0001845,
      "loss": 8.0097,
      "mean_token_accuracy": 0.10829245820641517,
      "num_tokens": 940087.0,
      "step": 370
    },
    {
      "entropy": 8.182055187225341,
      "epoch": 0.04327755337564916,
      "grad_norm": 1.6484375,
      "learning_rate": 0.000187,
      "loss": 7.9847,
      "mean_token_accuracy": 0.11593637242913246,
      "num_tokens": 953322.0,
      "step": 375
    },
    {
      "entropy": 8.304078960418702,
      "epoch": 0.04385458742065782,
      "grad_norm": 1.34375,
      "learning_rate": 0.0001895,
      "loss": 8.0352,
      "mean_token_accuracy": 0.11048125922679901,
      "num_tokens": 965815.0,
      "step": 380
    },
    {
      "entropy": 8.090431594848633,
      "epoch": 0.044431621465666475,
      "grad_norm": 1.140625,
      "learning_rate": 0.000192,
      "loss": 7.9867,
      "mean_token_accuracy": 0.11254308521747589,
      "num_tokens": 978731.0,
      "step": 385
    },
    {
      "entropy": 8.222192001342773,
      "epoch": 0.04500865551067513,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0001945,
      "loss": 7.8833,
      "mean_token_accuracy": 0.11793898567557334,
      "num_tokens": 989505.0,
      "step": 390
    },
    {
      "entropy": 8.180580139160156,
      "epoch": 0.04558568955568378,
      "grad_norm": 1.40625,
      "learning_rate": 0.00019700000000000002,
      "loss": 8.0039,
      "mean_token_accuracy": 0.11668088957667351,
      "num_tokens": 1001888.0,
      "step": 395
    },
    {
      "entropy": 8.087317895889282,
      "epoch": 0.04616272360069244,
      "grad_norm": 1.03125,
      "learning_rate": 0.00019950000000000002,
      "loss": 7.9376,
      "mean_token_accuracy": 0.12477797642350197,
      "num_tokens": 1015362.0,
      "step": 400
    },
    {
      "entropy": 8.24937629699707,
      "epoch": 0.0467397576457011,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000202,
      "loss": 8.0003,
      "mean_token_accuracy": 0.11018196940422058,
      "num_tokens": 1028700.0,
      "step": 405
    },
    {
      "entropy": 8.146645069122314,
      "epoch": 0.04731679169070975,
      "grad_norm": 1.046875,
      "learning_rate": 0.00020449999999999998,
      "loss": 8.0355,
      "mean_token_accuracy": 0.1107394464313984,
      "num_tokens": 1040766.0,
      "step": 410
    },
    {
      "entropy": 8.195110607147218,
      "epoch": 0.04789382573571841,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000207,
      "loss": 7.8899,
      "mean_token_accuracy": 0.11536648571491241,
      "num_tokens": 1054321.0,
      "step": 415
    },
    {
      "entropy": 8.083704519271851,
      "epoch": 0.048470859780727064,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002095,
      "loss": 7.9538,
      "mean_token_accuracy": 0.11945511400699615,
      "num_tokens": 1065410.0,
      "step": 420
    },
    {
      "entropy": 8.268659496307373,
      "epoch": 0.04904789382573572,
      "grad_norm": 1.15625,
      "learning_rate": 0.000212,
      "loss": 7.9799,
      "mean_token_accuracy": 0.10876310393214225,
      "num_tokens": 1077369.0,
      "step": 425
    },
    {
      "entropy": 8.160646915435791,
      "epoch": 0.04962492787074437,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002145,
      "loss": 7.9156,
      "mean_token_accuracy": 0.11709627509117126,
      "num_tokens": 1089065.0,
      "step": 430
    },
    {
      "entropy": 8.05447473526001,
      "epoch": 0.05020196191575303,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00021700000000000002,
      "loss": 7.8324,
      "mean_token_accuracy": 0.12342519015073776,
      "num_tokens": 1101285.0,
      "step": 435
    },
    {
      "entropy": 8.127099323272706,
      "epoch": 0.050778995960761686,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002195,
      "loss": 7.8574,
      "mean_token_accuracy": 0.11566452980041504,
      "num_tokens": 1114001.0,
      "step": 440
    },
    {
      "entropy": 8.098159694671631,
      "epoch": 0.05135603000577034,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000222,
      "loss": 7.931,
      "mean_token_accuracy": 0.11353125795722008,
      "num_tokens": 1126751.0,
      "step": 445
    },
    {
      "entropy": 8.043973064422607,
      "epoch": 0.05193306405077899,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002245,
      "loss": 7.9191,
      "mean_token_accuracy": 0.1132647417485714,
      "num_tokens": 1140737.0,
      "step": 450
    },
    {
      "entropy": 8.041358852386475,
      "epoch": 0.052510098095787654,
      "grad_norm": 1.203125,
      "learning_rate": 0.00022700000000000002,
      "loss": 7.8619,
      "mean_token_accuracy": 0.1132751502096653,
      "num_tokens": 1152569.0,
      "step": 455
    },
    {
      "entropy": 8.193452262878418,
      "epoch": 0.05308713214079631,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00022950000000000002,
      "loss": 7.8861,
      "mean_token_accuracy": 0.11552667319774627,
      "num_tokens": 1166163.0,
      "step": 460
    },
    {
      "entropy": 8.04149374961853,
      "epoch": 0.05366416618580496,
      "grad_norm": 1.140625,
      "learning_rate": 0.00023200000000000003,
      "loss": 7.8398,
      "mean_token_accuracy": 0.11528105437755584,
      "num_tokens": 1177647.0,
      "step": 465
    },
    {
      "entropy": 8.003826427459718,
      "epoch": 0.054241200230813615,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023449999999999998,
      "loss": 7.8806,
      "mean_token_accuracy": 0.11894029006361961,
      "num_tokens": 1190884.0,
      "step": 470
    },
    {
      "entropy": 8.122087001800537,
      "epoch": 0.054818234275822275,
      "grad_norm": 1.140625,
      "learning_rate": 0.000237,
      "loss": 7.9046,
      "mean_token_accuracy": 0.11480232924222947,
      "num_tokens": 1203276.0,
      "step": 475
    },
    {
      "entropy": 8.040111637115478,
      "epoch": 0.05539526832083093,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002395,
      "loss": 7.7945,
      "mean_token_accuracy": 0.12931760177016258,
      "num_tokens": 1216001.0,
      "step": 480
    },
    {
      "entropy": 8.027564573287965,
      "epoch": 0.05597230236583958,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000242,
      "loss": 7.9184,
      "mean_token_accuracy": 0.1165414072573185,
      "num_tokens": 1228724.0,
      "step": 485
    },
    {
      "entropy": 8.061245822906494,
      "epoch": 0.05654933641084824,
      "grad_norm": 1.53125,
      "learning_rate": 0.0002445,
      "loss": 7.9177,
      "mean_token_accuracy": 0.12267720922827721,
      "num_tokens": 1242731.0,
      "step": 490
    },
    {
      "entropy": 7.968474864959717,
      "epoch": 0.0571263704558569,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000247,
      "loss": 7.8711,
      "mean_token_accuracy": 0.11604229137301444,
      "num_tokens": 1255275.0,
      "step": 495
    },
    {
      "entropy": 8.032691955566406,
      "epoch": 0.05770340450086555,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002495,
      "loss": 7.8464,
      "mean_token_accuracy": 0.12695102244615555,
      "num_tokens": 1267719.0,
      "step": 500
    },
    {
      "entropy": 8.0850115776062,
      "epoch": 0.058280438545874204,
      "grad_norm": 1.046875,
      "learning_rate": 0.000252,
      "loss": 7.8315,
      "mean_token_accuracy": 0.11815451011061669,
      "num_tokens": 1279960.0,
      "step": 505
    },
    {
      "entropy": 7.9902856826782225,
      "epoch": 0.058857472590882864,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002545,
      "loss": 7.8152,
      "mean_token_accuracy": 0.1192896917462349,
      "num_tokens": 1292570.0,
      "step": 510
    },
    {
      "entropy": 8.022426271438599,
      "epoch": 0.05943450663589152,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000257,
      "loss": 7.8112,
      "mean_token_accuracy": 0.11861110255122184,
      "num_tokens": 1305548.0,
      "step": 515
    },
    {
      "entropy": 7.9653082370758055,
      "epoch": 0.06001154068090017,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002595,
      "loss": 7.7512,
      "mean_token_accuracy": 0.12333761304616928,
      "num_tokens": 1318904.0,
      "step": 520
    },
    {
      "entropy": 7.999736309051514,
      "epoch": 0.060588574725908825,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000262,
      "loss": 7.8275,
      "mean_token_accuracy": 0.11803121715784073,
      "num_tokens": 1332841.0,
      "step": 525
    },
    {
      "entropy": 7.946177673339844,
      "epoch": 0.061165608770917486,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00026450000000000003,
      "loss": 7.828,
      "mean_token_accuracy": 0.11700976938009262,
      "num_tokens": 1345708.0,
      "step": 530
    },
    {
      "entropy": 8.01097912788391,
      "epoch": 0.06174264281592614,
      "grad_norm": 1.453125,
      "learning_rate": 0.00026700000000000004,
      "loss": 7.7805,
      "mean_token_accuracy": 0.12158992290496826,
      "num_tokens": 1358611.0,
      "step": 535
    },
    {
      "entropy": 7.956462049484253,
      "epoch": 0.06231967686093479,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00026950000000000005,
      "loss": 7.8431,
      "mean_token_accuracy": 0.1208947941660881,
      "num_tokens": 1372753.0,
      "step": 540
    },
    {
      "entropy": 7.846506118774414,
      "epoch": 0.06289671090594345,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00027200000000000005,
      "loss": 7.7475,
      "mean_token_accuracy": 0.13231490254402162,
      "num_tokens": 1383811.0,
      "step": 545
    },
    {
      "entropy": 7.998351097106934,
      "epoch": 0.06347374495095211,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002745,
      "loss": 7.8154,
      "mean_token_accuracy": 0.12150197923183441,
      "num_tokens": 1395228.0,
      "step": 550
    },
    {
      "entropy": 7.844149923324585,
      "epoch": 0.06405077899596076,
      "grad_norm": 1.4375,
      "learning_rate": 0.000277,
      "loss": 7.8298,
      "mean_token_accuracy": 0.11782822757959366,
      "num_tokens": 1408869.0,
      "step": 555
    },
    {
      "entropy": 7.925115871429443,
      "epoch": 0.06462781304096941,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002795,
      "loss": 7.7363,
      "mean_token_accuracy": 0.126988285779953,
      "num_tokens": 1420926.0,
      "step": 560
    },
    {
      "entropy": 7.919132804870605,
      "epoch": 0.06520484708597807,
      "grad_norm": 1.125,
      "learning_rate": 0.00028199999999999997,
      "loss": 7.7324,
      "mean_token_accuracy": 0.1244761623442173,
      "num_tokens": 1432435.0,
      "step": 565
    },
    {
      "entropy": 7.929744482040405,
      "epoch": 0.06578188113098672,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002845,
      "loss": 7.8638,
      "mean_token_accuracy": 0.11697117760777473,
      "num_tokens": 1444716.0,
      "step": 570
    },
    {
      "entropy": 7.861667013168335,
      "epoch": 0.06635891517599539,
      "grad_norm": 1.203125,
      "learning_rate": 0.000287,
      "loss": 7.7398,
      "mean_token_accuracy": 0.12333890348672867,
      "num_tokens": 1458129.0,
      "step": 575
    },
    {
      "entropy": 7.875370264053345,
      "epoch": 0.06693594922100404,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002895,
      "loss": 7.7118,
      "mean_token_accuracy": 0.1261540025472641,
      "num_tokens": 1472076.0,
      "step": 580
    },
    {
      "entropy": 7.806885862350464,
      "epoch": 0.0675129832660127,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000292,
      "loss": 7.7086,
      "mean_token_accuracy": 0.1355113133788109,
      "num_tokens": 1486026.0,
      "step": 585
    },
    {
      "entropy": 7.809766578674316,
      "epoch": 0.06809001731102135,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002945,
      "loss": 7.704,
      "mean_token_accuracy": 0.13341286033391953,
      "num_tokens": 1498735.0,
      "step": 590
    },
    {
      "entropy": 7.84214186668396,
      "epoch": 0.06866705135603,
      "grad_norm": 1.046875,
      "learning_rate": 0.000297,
      "loss": 7.6793,
      "mean_token_accuracy": 0.12912188470363617,
      "num_tokens": 1510755.0,
      "step": 595
    },
    {
      "entropy": 7.897479295730591,
      "epoch": 0.06924408540103866,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002995,
      "loss": 7.7216,
      "mean_token_accuracy": 0.12750375717878343,
      "num_tokens": 1522386.0,
      "step": 600
    },
    {
      "entropy": 7.845736980438232,
      "epoch": 0.06982111944604731,
      "grad_norm": 1.03125,
      "learning_rate": 0.000302,
      "loss": 7.742,
      "mean_token_accuracy": 0.12660543844103814,
      "num_tokens": 1534757.0,
      "step": 605
    },
    {
      "entropy": 7.81597785949707,
      "epoch": 0.07039815349105598,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003045,
      "loss": 7.724,
      "mean_token_accuracy": 0.12527383714914322,
      "num_tokens": 1547684.0,
      "step": 610
    },
    {
      "entropy": 7.850346899032592,
      "epoch": 0.07097518753606463,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000307,
      "loss": 7.7303,
      "mean_token_accuracy": 0.11933122873306275,
      "num_tokens": 1560583.0,
      "step": 615
    },
    {
      "entropy": 7.878012609481812,
      "epoch": 0.07155222158107329,
      "grad_norm": 0.875,
      "learning_rate": 0.0003095,
      "loss": 7.7424,
      "mean_token_accuracy": 0.12126323059201241,
      "num_tokens": 1573174.0,
      "step": 620
    },
    {
      "entropy": 7.815459823608398,
      "epoch": 0.07212925562608194,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000312,
      "loss": 7.7149,
      "mean_token_accuracy": 0.12385709285736084,
      "num_tokens": 1585746.0,
      "step": 625
    },
    {
      "entropy": 7.842224073410034,
      "epoch": 0.0727062896710906,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003145,
      "loss": 7.695,
      "mean_token_accuracy": 0.11776623651385307,
      "num_tokens": 1598538.0,
      "step": 630
    },
    {
      "entropy": 7.825200510025025,
      "epoch": 0.07328332371609925,
      "grad_norm": 1.125,
      "learning_rate": 0.000317,
      "loss": 7.674,
      "mean_token_accuracy": 0.1275767132639885,
      "num_tokens": 1612598.0,
      "step": 635
    },
    {
      "entropy": 7.780414009094239,
      "epoch": 0.0738603577611079,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003195,
      "loss": 7.6161,
      "mean_token_accuracy": 0.13024978637695311,
      "num_tokens": 1624418.0,
      "step": 640
    },
    {
      "entropy": 7.701315689086914,
      "epoch": 0.07443739180611655,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000322,
      "loss": 7.7056,
      "mean_token_accuracy": 0.13014591559767724,
      "num_tokens": 1637512.0,
      "step": 645
    },
    {
      "entropy": 7.68240270614624,
      "epoch": 0.07501442585112522,
      "grad_norm": 1.140625,
      "learning_rate": 0.00032450000000000003,
      "loss": 7.6419,
      "mean_token_accuracy": 0.1289619944989681,
      "num_tokens": 1650413.0,
      "step": 650
    },
    {
      "entropy": 7.799037647247315,
      "epoch": 0.07559145989613388,
      "grad_norm": 1.03125,
      "learning_rate": 0.00032700000000000003,
      "loss": 7.6123,
      "mean_token_accuracy": 0.13126565814018248,
      "num_tokens": 1662225.0,
      "step": 655
    },
    {
      "entropy": 7.818531084060669,
      "epoch": 0.07616849394114253,
      "grad_norm": 1.1875,
      "learning_rate": 0.00032950000000000004,
      "loss": 7.6901,
      "mean_token_accuracy": 0.12705290466547012,
      "num_tokens": 1674458.0,
      "step": 660
    },
    {
      "entropy": 7.750899171829223,
      "epoch": 0.07674552798615118,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00033200000000000005,
      "loss": 7.701,
      "mean_token_accuracy": 0.12313887178897857,
      "num_tokens": 1686366.0,
      "step": 665
    },
    {
      "entropy": 7.7861656665802,
      "epoch": 0.07732256203115984,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033450000000000005,
      "loss": 7.6882,
      "mean_token_accuracy": 0.12421951442956924,
      "num_tokens": 1699654.0,
      "step": 670
    },
    {
      "entropy": 7.768671703338623,
      "epoch": 0.07789959607616849,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000337,
      "loss": 7.6462,
      "mean_token_accuracy": 0.12908095195889474,
      "num_tokens": 1711923.0,
      "step": 675
    },
    {
      "entropy": 7.797735977172851,
      "epoch": 0.07847663012117714,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003395,
      "loss": 7.6328,
      "mean_token_accuracy": 0.12395294979214669,
      "num_tokens": 1723413.0,
      "step": 680
    },
    {
      "entropy": 7.680013036727905,
      "epoch": 0.07905366416618581,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000342,
      "loss": 7.6006,
      "mean_token_accuracy": 0.13139746859669685,
      "num_tokens": 1735975.0,
      "step": 685
    },
    {
      "entropy": 7.710196876525879,
      "epoch": 0.07963069821119446,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034449999999999997,
      "loss": 7.6149,
      "mean_token_accuracy": 0.1285354495048523,
      "num_tokens": 1748563.0,
      "step": 690
    },
    {
      "entropy": 7.633371353149414,
      "epoch": 0.08020773225620312,
      "grad_norm": 1.15625,
      "learning_rate": 0.000347,
      "loss": 7.6288,
      "mean_token_accuracy": 0.12319590821862221,
      "num_tokens": 1762914.0,
      "step": 695
    },
    {
      "entropy": 7.7966938495635985,
      "epoch": 0.08078476630121177,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003495,
      "loss": 7.5953,
      "mean_token_accuracy": 0.13221077919006347,
      "num_tokens": 1776084.0,
      "step": 700
    },
    {
      "entropy": 7.640158843994141,
      "epoch": 0.08136180034622043,
      "grad_norm": 1.171875,
      "learning_rate": 0.000352,
      "loss": 7.5445,
      "mean_token_accuracy": 0.1401802346110344,
      "num_tokens": 1789316.0,
      "step": 705
    },
    {
      "entropy": 7.714691829681397,
      "epoch": 0.08193883439122908,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003545,
      "loss": 7.637,
      "mean_token_accuracy": 0.1277747370302677,
      "num_tokens": 1801576.0,
      "step": 710
    },
    {
      "entropy": 7.588133955001831,
      "epoch": 0.08251586843623773,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000357,
      "loss": 7.5861,
      "mean_token_accuracy": 0.13175102695822716,
      "num_tokens": 1814039.0,
      "step": 715
    },
    {
      "entropy": 7.677944850921631,
      "epoch": 0.0830929024812464,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003595,
      "loss": 7.53,
      "mean_token_accuracy": 0.131522336602211,
      "num_tokens": 1825193.0,
      "step": 720
    },
    {
      "entropy": 7.593373250961304,
      "epoch": 0.08366993652625505,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000362,
      "loss": 7.6051,
      "mean_token_accuracy": 0.12541260421276093,
      "num_tokens": 1837983.0,
      "step": 725
    },
    {
      "entropy": 7.725003337860107,
      "epoch": 0.08424697057126371,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003645,
      "loss": 7.5148,
      "mean_token_accuracy": 0.13810623735189437,
      "num_tokens": 1849830.0,
      "step": 730
    },
    {
      "entropy": 7.61106128692627,
      "epoch": 0.08482400461627236,
      "grad_norm": 1.40625,
      "learning_rate": 0.000367,
      "loss": 7.5923,
      "mean_token_accuracy": 0.12345706224441529,
      "num_tokens": 1862577.0,
      "step": 735
    },
    {
      "entropy": 7.646603107452393,
      "epoch": 0.08540103866128101,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003695,
      "loss": 7.603,
      "mean_token_accuracy": 0.13642554581165314,
      "num_tokens": 1874888.0,
      "step": 740
    },
    {
      "entropy": 7.624688339233399,
      "epoch": 0.08597807270628967,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000372,
      "loss": 7.5309,
      "mean_token_accuracy": 0.1350684270262718,
      "num_tokens": 1888400.0,
      "step": 745
    },
    {
      "entropy": 7.609202241897583,
      "epoch": 0.08655510675129832,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003745,
      "loss": 7.6054,
      "mean_token_accuracy": 0.12247811630368233,
      "num_tokens": 1900961.0,
      "step": 750
    },
    {
      "entropy": 7.590191173553467,
      "epoch": 0.08713214079630698,
      "grad_norm": 0.96875,
      "learning_rate": 0.000377,
      "loss": 7.5243,
      "mean_token_accuracy": 0.13634067475795747,
      "num_tokens": 1913410.0,
      "step": 755
    },
    {
      "entropy": 7.664303922653199,
      "epoch": 0.08770917484131564,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003795,
      "loss": 7.5827,
      "mean_token_accuracy": 0.12373366355895996,
      "num_tokens": 1926602.0,
      "step": 760
    },
    {
      "entropy": 7.6440318584442135,
      "epoch": 0.0882862088863243,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000382,
      "loss": 7.5379,
      "mean_token_accuracy": 0.12371056452393532,
      "num_tokens": 1938904.0,
      "step": 765
    },
    {
      "entropy": 7.585968399047852,
      "epoch": 0.08886324293133295,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003845,
      "loss": 7.577,
      "mean_token_accuracy": 0.12310778871178626,
      "num_tokens": 1952827.0,
      "step": 770
    },
    {
      "entropy": 7.638195276260376,
      "epoch": 0.0894402769763416,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038700000000000003,
      "loss": 7.548,
      "mean_token_accuracy": 0.13129694759845734,
      "num_tokens": 1965182.0,
      "step": 775
    },
    {
      "entropy": 7.510479879379273,
      "epoch": 0.09001731102135026,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038950000000000003,
      "loss": 7.5029,
      "mean_token_accuracy": 0.12708307728171347,
      "num_tokens": 1976953.0,
      "step": 780
    },
    {
      "entropy": 7.737362003326416,
      "epoch": 0.09059434506635891,
      "grad_norm": 0.9375,
      "learning_rate": 0.00039200000000000004,
      "loss": 7.6022,
      "mean_token_accuracy": 0.12702567875385284,
      "num_tokens": 1990019.0,
      "step": 785
    },
    {
      "entropy": 7.611118268966675,
      "epoch": 0.09117137911136756,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039450000000000005,
      "loss": 7.5488,
      "mean_token_accuracy": 0.1295396715402603,
      "num_tokens": 2002976.0,
      "step": 790
    },
    {
      "entropy": 7.59055905342102,
      "epoch": 0.09174841315637623,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00039700000000000005,
      "loss": 7.5906,
      "mean_token_accuracy": 0.1250334121286869,
      "num_tokens": 2017368.0,
      "step": 795
    },
    {
      "entropy": 7.517406940460205,
      "epoch": 0.09232544720138489,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003995,
      "loss": 7.4334,
      "mean_token_accuracy": 0.13203005865216255,
      "num_tokens": 2029838.0,
      "step": 800
    },
    {
      "entropy": 7.63545331954956,
      "epoch": 0.09290248124639354,
      "grad_norm": 1.125,
      "learning_rate": 0.000402,
      "loss": 7.5516,
      "mean_token_accuracy": 0.12461932525038719,
      "num_tokens": 2042085.0,
      "step": 805
    },
    {
      "entropy": 7.604149198532104,
      "epoch": 0.0934795152914022,
      "grad_norm": 1.0,
      "learning_rate": 0.0004045,
      "loss": 7.5652,
      "mean_token_accuracy": 0.12991983070969582,
      "num_tokens": 2053268.0,
      "step": 810
    },
    {
      "entropy": 7.614420986175537,
      "epoch": 0.09405654933641085,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040699999999999997,
      "loss": 7.5398,
      "mean_token_accuracy": 0.12179678082466125,
      "num_tokens": 2066621.0,
      "step": 815
    },
    {
      "entropy": 7.462197637557983,
      "epoch": 0.0946335833814195,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004095,
      "loss": 7.4241,
      "mean_token_accuracy": 0.13013325929641723,
      "num_tokens": 2079590.0,
      "step": 820
    },
    {
      "entropy": 7.571118640899658,
      "epoch": 0.09521061742642815,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000412,
      "loss": 7.5334,
      "mean_token_accuracy": 0.12693382874131204,
      "num_tokens": 2091729.0,
      "step": 825
    },
    {
      "entropy": 7.510286808013916,
      "epoch": 0.09578765147143682,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004145,
      "loss": 7.4885,
      "mean_token_accuracy": 0.13465983644127846,
      "num_tokens": 2104254.0,
      "step": 830
    },
    {
      "entropy": 7.585655641555786,
      "epoch": 0.09636468551644548,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000417,
      "loss": 7.5596,
      "mean_token_accuracy": 0.12383549585938454,
      "num_tokens": 2116327.0,
      "step": 835
    },
    {
      "entropy": 7.621681976318359,
      "epoch": 0.09694171956145413,
      "grad_norm": 1.0,
      "learning_rate": 0.0004195,
      "loss": 7.5559,
      "mean_token_accuracy": 0.1234694242477417,
      "num_tokens": 2129153.0,
      "step": 840
    },
    {
      "entropy": 7.49256067276001,
      "epoch": 0.09751875360646278,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000422,
      "loss": 7.508,
      "mean_token_accuracy": 0.136223391443491,
      "num_tokens": 2141705.0,
      "step": 845
    },
    {
      "entropy": 7.467943477630615,
      "epoch": 0.09809578765147144,
      "grad_norm": 1.0,
      "learning_rate": 0.0004245,
      "loss": 7.4178,
      "mean_token_accuracy": 0.1418270103633404,
      "num_tokens": 2153760.0,
      "step": 850
    },
    {
      "entropy": 7.600914192199707,
      "epoch": 0.09867282169648009,
      "grad_norm": 0.96875,
      "learning_rate": 0.000427,
      "loss": 7.5065,
      "mean_token_accuracy": 0.13413666784763337,
      "num_tokens": 2168206.0,
      "step": 855
    },
    {
      "entropy": 7.570523405075074,
      "epoch": 0.09924985574148874,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004295,
      "loss": 7.5117,
      "mean_token_accuracy": 0.13071473613381385,
      "num_tokens": 2180779.0,
      "step": 860
    },
    {
      "entropy": 7.412028741836548,
      "epoch": 0.0998268897864974,
      "grad_norm": 1.125,
      "learning_rate": 0.000432,
      "loss": 7.4373,
      "mean_token_accuracy": 0.13453905433416366,
      "num_tokens": 2194075.0,
      "step": 865
    },
    {
      "entropy": 7.537879610061646,
      "epoch": 0.10040392383150606,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004345,
      "loss": 7.4232,
      "mean_token_accuracy": 0.13529622703790664,
      "num_tokens": 2206424.0,
      "step": 870
    },
    {
      "entropy": 7.555378055572509,
      "epoch": 0.10098095787651472,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000437,
      "loss": 7.4432,
      "mean_token_accuracy": 0.1277135990560055,
      "num_tokens": 2218550.0,
      "step": 875
    },
    {
      "entropy": 7.4310328483581545,
      "epoch": 0.10155799192152337,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004395,
      "loss": 7.4761,
      "mean_token_accuracy": 0.1309943303465843,
      "num_tokens": 2231857.0,
      "step": 880
    },
    {
      "entropy": 7.488246202468872,
      "epoch": 0.10213502596653203,
      "grad_norm": 1.109375,
      "learning_rate": 0.000442,
      "loss": 7.4876,
      "mean_token_accuracy": 0.13225487917661666,
      "num_tokens": 2244057.0,
      "step": 885
    },
    {
      "entropy": 7.613044691085816,
      "epoch": 0.10271206001154068,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004445,
      "loss": 7.5175,
      "mean_token_accuracy": 0.12458534613251686,
      "num_tokens": 2257361.0,
      "step": 890
    },
    {
      "entropy": 7.529584169387817,
      "epoch": 0.10328909405654933,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000447,
      "loss": 7.4967,
      "mean_token_accuracy": 0.12664385214447976,
      "num_tokens": 2269599.0,
      "step": 895
    },
    {
      "entropy": 7.492797708511352,
      "epoch": 0.10386612810155799,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044950000000000003,
      "loss": 7.5168,
      "mean_token_accuracy": 0.12740001678466797,
      "num_tokens": 2282791.0,
      "step": 900
    },
    {
      "entropy": 7.53049669265747,
      "epoch": 0.10444316214656665,
      "grad_norm": 1.125,
      "learning_rate": 0.00045200000000000004,
      "loss": 7.4807,
      "mean_token_accuracy": 0.13134395331144333,
      "num_tokens": 2295336.0,
      "step": 905
    },
    {
      "entropy": 7.494025135040284,
      "epoch": 0.10502019619157531,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045450000000000004,
      "loss": 7.4828,
      "mean_token_accuracy": 0.12917752116918563,
      "num_tokens": 2308813.0,
      "step": 910
    },
    {
      "entropy": 7.475014638900757,
      "epoch": 0.10559723023658396,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045700000000000005,
      "loss": 7.4671,
      "mean_token_accuracy": 0.12653358280658722,
      "num_tokens": 2321374.0,
      "step": 915
    },
    {
      "entropy": 7.4855108737945555,
      "epoch": 0.10617426428159261,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00045950000000000006,
      "loss": 7.4787,
      "mean_token_accuracy": 0.12270581796765327,
      "num_tokens": 2333698.0,
      "step": 920
    },
    {
      "entropy": 7.527695846557617,
      "epoch": 0.10675129832660127,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000462,
      "loss": 7.4649,
      "mean_token_accuracy": 0.13576419949531554,
      "num_tokens": 2346557.0,
      "step": 925
    },
    {
      "entropy": 7.328889083862305,
      "epoch": 0.10732833237160992,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004645,
      "loss": 7.3053,
      "mean_token_accuracy": 0.1375327415764332,
      "num_tokens": 2360003.0,
      "step": 930
    },
    {
      "entropy": 7.464563989639283,
      "epoch": 0.10790536641661858,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000467,
      "loss": 7.3543,
      "mean_token_accuracy": 0.13607896789908408,
      "num_tokens": 2372765.0,
      "step": 935
    },
    {
      "entropy": 7.372702598571777,
      "epoch": 0.10848240046162723,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004695,
      "loss": 7.3561,
      "mean_token_accuracy": 0.13873768970370293,
      "num_tokens": 2385154.0,
      "step": 940
    },
    {
      "entropy": 7.381151580810547,
      "epoch": 0.1090594345066359,
      "grad_norm": 1.09375,
      "learning_rate": 0.000472,
      "loss": 7.3805,
      "mean_token_accuracy": 0.13930853679776192,
      "num_tokens": 2397264.0,
      "step": 945
    },
    {
      "entropy": 7.4766237258911135,
      "epoch": 0.10963646855164455,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004745,
      "loss": 7.4843,
      "mean_token_accuracy": 0.13500367030501365,
      "num_tokens": 2408652.0,
      "step": 950
    },
    {
      "entropy": 7.411814785003662,
      "epoch": 0.1102135025966532,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000477,
      "loss": 7.4289,
      "mean_token_accuracy": 0.12838349416851996,
      "num_tokens": 2421683.0,
      "step": 955
    },
    {
      "entropy": 7.380972909927368,
      "epoch": 0.11079053664166186,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004795,
      "loss": 7.3453,
      "mean_token_accuracy": 0.1313640892505646,
      "num_tokens": 2434622.0,
      "step": 960
    },
    {
      "entropy": 7.420246410369873,
      "epoch": 0.11136757068667051,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000482,
      "loss": 7.3595,
      "mean_token_accuracy": 0.14137982428073884,
      "num_tokens": 2448157.0,
      "step": 965
    },
    {
      "entropy": 7.482502222061157,
      "epoch": 0.11194460473167916,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004845,
      "loss": 7.4034,
      "mean_token_accuracy": 0.1286494918167591,
      "num_tokens": 2460284.0,
      "step": 970
    },
    {
      "entropy": 7.2979847431182865,
      "epoch": 0.11252163877668782,
      "grad_norm": 1.0,
      "learning_rate": 0.000487,
      "loss": 7.4166,
      "mean_token_accuracy": 0.13066136166453363,
      "num_tokens": 2473031.0,
      "step": 975
    },
    {
      "entropy": 7.43588399887085,
      "epoch": 0.11309867282169649,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004895,
      "loss": 7.429,
      "mean_token_accuracy": 0.1351068213582039,
      "num_tokens": 2486870.0,
      "step": 980
    },
    {
      "entropy": 7.407749509811401,
      "epoch": 0.11367570686670514,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000492,
      "loss": 7.3198,
      "mean_token_accuracy": 0.1358391098678112,
      "num_tokens": 2498921.0,
      "step": 985
    },
    {
      "entropy": 7.458673715591431,
      "epoch": 0.1142527409117138,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004945,
      "loss": 7.3969,
      "mean_token_accuracy": 0.13049472868442535,
      "num_tokens": 2511122.0,
      "step": 990
    },
    {
      "entropy": 7.389868688583374,
      "epoch": 0.11482977495672245,
      "grad_norm": 1.046875,
      "learning_rate": 0.000497,
      "loss": 7.4051,
      "mean_token_accuracy": 0.1294924721121788,
      "num_tokens": 2523726.0,
      "step": 995
    },
    {
      "entropy": 7.3466997146606445,
      "epoch": 0.1154068090017311,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995,
      "loss": 7.3137,
      "mean_token_accuracy": 0.13630730137228966,
      "num_tokens": 2536192.0,
      "step": 1000
    },
    {
      "entropy": 7.343966245651245,
      "epoch": 0.11598384304673975,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999999975783157,
      "loss": 7.3879,
      "mean_token_accuracy": 0.13277141600847245,
      "num_tokens": 2550200.0,
      "step": 1005
    },
    {
      "entropy": 7.345440340042114,
      "epoch": 0.11656087709174841,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999999877402236,
      "loss": 7.3285,
      "mean_token_accuracy": 0.137277489900589,
      "num_tokens": 2564045.0,
      "step": 1010
    },
    {
      "entropy": 7.331404352188111,
      "epoch": 0.11713791113675708,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999999703343686,
      "loss": 7.3359,
      "mean_token_accuracy": 0.13079950734972953,
      "num_tokens": 2576210.0,
      "step": 1015
    },
    {
      "entropy": 7.2891435623168945,
      "epoch": 0.11771494518176573,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999999453607515,
      "loss": 7.3348,
      "mean_token_accuracy": 0.13124105408787728,
      "num_tokens": 2589325.0,
      "step": 1020
    },
    {
      "entropy": 7.409384393692017,
      "epoch": 0.11829197922677438,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999999128193729,
      "loss": 7.4309,
      "mean_token_accuracy": 0.13288950622081758,
      "num_tokens": 2601522.0,
      "step": 1025
    },
    {
      "entropy": 7.3894264698028564,
      "epoch": 0.11886901327178304,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499999872710234,
      "loss": 7.3974,
      "mean_token_accuracy": 0.1278626799583435,
      "num_tokens": 2615364.0,
      "step": 1030
    },
    {
      "entropy": 7.375164937973023,
      "epoch": 0.11944604731679169,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999998250333361,
      "loss": 7.3142,
      "mean_token_accuracy": 0.1343100145459175,
      "num_tokens": 2628528.0,
      "step": 1035
    },
    {
      "entropy": 7.282582569122314,
      "epoch": 0.12002308136180034,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499999769788681,
      "loss": 7.3058,
      "mean_token_accuracy": 0.1310929149389267,
      "num_tokens": 2641877.0,
      "step": 1040
    },
    {
      "entropy": 7.369835758209229,
      "epoch": 0.120600115406809,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999997069762703,
      "loss": 7.3171,
      "mean_token_accuracy": 0.13292421400547028,
      "num_tokens": 2655419.0,
      "step": 1045
    },
    {
      "entropy": 7.171972227096558,
      "epoch": 0.12117714945181765,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999996365961062,
      "loss": 7.2777,
      "mean_token_accuracy": 0.13511810302734376,
      "num_tokens": 2667551.0,
      "step": 1050
    },
    {
      "entropy": 7.34223484992981,
      "epoch": 0.12175418349682632,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999995586481912,
      "loss": 7.3047,
      "mean_token_accuracy": 0.13582876622676848,
      "num_tokens": 2679948.0,
      "step": 1055
    },
    {
      "entropy": 7.260743761062622,
      "epoch": 0.12233121754183497,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999994731325276,
      "loss": 7.3492,
      "mean_token_accuracy": 0.1337125964462757,
      "num_tokens": 2691501.0,
      "step": 1060
    },
    {
      "entropy": 7.260812520980835,
      "epoch": 0.12290825158684363,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999993800491187,
      "loss": 7.2664,
      "mean_token_accuracy": 0.13844716772437096,
      "num_tokens": 2702689.0,
      "step": 1065
    },
    {
      "entropy": 7.332363796234131,
      "epoch": 0.12348528563185228,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999992793979672,
      "loss": 7.2764,
      "mean_token_accuracy": 0.13439239487051963,
      "num_tokens": 2715023.0,
      "step": 1070
    },
    {
      "entropy": 7.202269220352173,
      "epoch": 0.12406231967686093,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999991711790769,
      "loss": 7.2274,
      "mean_token_accuracy": 0.13450795337557792,
      "num_tokens": 2729517.0,
      "step": 1075
    },
    {
      "entropy": 7.301486778259277,
      "epoch": 0.12463935372186959,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999990553924511,
      "loss": 7.3108,
      "mean_token_accuracy": 0.13219440579414368,
      "num_tokens": 2740984.0,
      "step": 1080
    },
    {
      "entropy": 7.299670457839966,
      "epoch": 0.12521638776687824,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999989320380939,
      "loss": 7.252,
      "mean_token_accuracy": 0.13629381954669953,
      "num_tokens": 2752943.0,
      "step": 1085
    },
    {
      "entropy": 7.296296072006226,
      "epoch": 0.1257934218118869,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999988011160094,
      "loss": 7.3008,
      "mean_token_accuracy": 0.13499366268515586,
      "num_tokens": 2765613.0,
      "step": 1090
    },
    {
      "entropy": 7.2634941101074215,
      "epoch": 0.12637045585689555,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499998662626202,
      "loss": 7.2476,
      "mean_token_accuracy": 0.13409144133329393,
      "num_tokens": 2776820.0,
      "step": 1095
    },
    {
      "entropy": 7.248972606658936,
      "epoch": 0.12694748990190421,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999985165686764,
      "loss": 7.2846,
      "mean_token_accuracy": 0.13224128484725953,
      "num_tokens": 2788981.0,
      "step": 1100
    },
    {
      "entropy": 7.163676500320435,
      "epoch": 0.12752452394691285,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004999983629434373,
      "loss": 7.1433,
      "mean_token_accuracy": 0.14286725521087645,
      "num_tokens": 2801975.0,
      "step": 1105
    },
    {
      "entropy": 7.309284687042236,
      "epoch": 0.12810155799192152,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999982017504901,
      "loss": 7.2511,
      "mean_token_accuracy": 0.12998932003974914,
      "num_tokens": 2814708.0,
      "step": 1110
    },
    {
      "entropy": 7.249689722061158,
      "epoch": 0.1286785920369302,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999980329898401,
      "loss": 7.2684,
      "mean_token_accuracy": 0.13964115902781488,
      "num_tokens": 2827903.0,
      "step": 1115
    },
    {
      "entropy": 7.2154786586761475,
      "epoch": 0.12925562608193883,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499997856661493,
      "loss": 7.2893,
      "mean_token_accuracy": 0.1352258048951626,
      "num_tokens": 2841077.0,
      "step": 1120
    },
    {
      "entropy": 7.290196561813355,
      "epoch": 0.1298326601269475,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499997672765455,
      "loss": 7.2744,
      "mean_token_accuracy": 0.13311855792999266,
      "num_tokens": 2853235.0,
      "step": 1125
    },
    {
      "entropy": 7.231596946716309,
      "epoch": 0.13040969417195614,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999974813017318,
      "loss": 7.2546,
      "mean_token_accuracy": 0.1340233527123928,
      "num_tokens": 2864937.0,
      "step": 1130
    },
    {
      "entropy": 7.187311124801636,
      "epoch": 0.1309867282169648,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999972822703301,
      "loss": 7.1922,
      "mean_token_accuracy": 0.148747056722641,
      "num_tokens": 2877791.0,
      "step": 1135
    },
    {
      "entropy": 7.157298278808594,
      "epoch": 0.13156376226197344,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999970756712567,
      "loss": 7.1925,
      "mean_token_accuracy": 0.13003327026963235,
      "num_tokens": 2889897.0,
      "step": 1140
    },
    {
      "entropy": 7.245456647872925,
      "epoch": 0.1321407963069821,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999968615045183,
      "loss": 7.2011,
      "mean_token_accuracy": 0.13551229760050773,
      "num_tokens": 2902806.0,
      "step": 1145
    },
    {
      "entropy": 7.156963729858399,
      "epoch": 0.13271783035199078,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999966397701222,
      "loss": 7.109,
      "mean_token_accuracy": 0.14400240257382393,
      "num_tokens": 2915703.0,
      "step": 1150
    },
    {
      "entropy": 7.207495975494385,
      "epoch": 0.13329486439699942,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999964104680759,
      "loss": 7.316,
      "mean_token_accuracy": 0.12647873908281326,
      "num_tokens": 2928171.0,
      "step": 1155
    },
    {
      "entropy": 7.237271070480347,
      "epoch": 0.13387189844200809,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999961735983871,
      "loss": 7.1198,
      "mean_token_accuracy": 0.14833424836397172,
      "num_tokens": 2939909.0,
      "step": 1160
    },
    {
      "entropy": 7.277281904220581,
      "epoch": 0.13444893248701673,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999959291610639,
      "loss": 7.2618,
      "mean_token_accuracy": 0.13249962106347085,
      "num_tokens": 2951510.0,
      "step": 1165
    },
    {
      "entropy": 7.115877532958985,
      "epoch": 0.1350259665320254,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999956771561143,
      "loss": 7.1946,
      "mean_token_accuracy": 0.13074965700507163,
      "num_tokens": 2963326.0,
      "step": 1170
    },
    {
      "entropy": 7.202230167388916,
      "epoch": 0.13560300057703403,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999954175835469,
      "loss": 7.2159,
      "mean_token_accuracy": 0.14419033750891685,
      "num_tokens": 2976226.0,
      "step": 1175
    },
    {
      "entropy": 7.2409210205078125,
      "epoch": 0.1361800346220427,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999951504433703,
      "loss": 7.2179,
      "mean_token_accuracy": 0.13466040119528772,
      "num_tokens": 2988180.0,
      "step": 1180
    },
    {
      "entropy": 7.1283531665802,
      "epoch": 0.13675706866705137,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999948757355935,
      "loss": 7.1966,
      "mean_token_accuracy": 0.1367337226867676,
      "num_tokens": 3000215.0,
      "step": 1185
    },
    {
      "entropy": 7.235973787307739,
      "epoch": 0.13733410271206,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499994593460226,
      "loss": 7.2033,
      "mean_token_accuracy": 0.13479142561554908,
      "num_tokens": 3012516.0,
      "step": 1190
    },
    {
      "entropy": 7.171559572219849,
      "epoch": 0.13791113675706868,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999943036172771,
      "loss": 7.1937,
      "mean_token_accuracy": 0.13718242719769477,
      "num_tokens": 3024786.0,
      "step": 1195
    },
    {
      "entropy": 7.207344961166382,
      "epoch": 0.13848817080207732,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999940062067565,
      "loss": 7.1336,
      "mean_token_accuracy": 0.14108351469039918,
      "num_tokens": 3037179.0,
      "step": 1200
    },
    {
      "entropy": 7.102613258361816,
      "epoch": 0.13906520484708598,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999937012286742,
      "loss": 7.2248,
      "mean_token_accuracy": 0.13997655510902404,
      "num_tokens": 3049443.0,
      "step": 1205
    },
    {
      "entropy": 7.126314640045166,
      "epoch": 0.13964223889209462,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999933886830405,
      "loss": 7.1425,
      "mean_token_accuracy": 0.13395074158906936,
      "num_tokens": 3061857.0,
      "step": 1210
    },
    {
      "entropy": 7.17795877456665,
      "epoch": 0.1402192729371033,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499993068569866,
      "loss": 7.1606,
      "mean_token_accuracy": 0.14002601727843283,
      "num_tokens": 3075185.0,
      "step": 1215
    },
    {
      "entropy": 7.182628631591797,
      "epoch": 0.14079630698211196,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999927408891614,
      "loss": 7.1809,
      "mean_token_accuracy": 0.1401163399219513,
      "num_tokens": 3087645.0,
      "step": 1220
    },
    {
      "entropy": 7.178424263000489,
      "epoch": 0.1413733410271206,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999924056409378,
      "loss": 7.1728,
      "mean_token_accuracy": 0.13846464678645135,
      "num_tokens": 3101185.0,
      "step": 1225
    },
    {
      "entropy": 7.171338224411011,
      "epoch": 0.14195037507212926,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999920628252063,
      "loss": 7.1424,
      "mean_token_accuracy": 0.1404540069401264,
      "num_tokens": 3113694.0,
      "step": 1230
    },
    {
      "entropy": 7.138214826583862,
      "epoch": 0.1425274091171379,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999917124419785,
      "loss": 7.2063,
      "mean_token_accuracy": 0.13672061413526534,
      "num_tokens": 3128225.0,
      "step": 1235
    },
    {
      "entropy": 7.167185831069946,
      "epoch": 0.14310444316214657,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999913544912664,
      "loss": 7.1181,
      "mean_token_accuracy": 0.14096531718969346,
      "num_tokens": 3140118.0,
      "step": 1240
    },
    {
      "entropy": 7.1235737800598145,
      "epoch": 0.1436814772071552,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999909889730817,
      "loss": 7.1932,
      "mean_token_accuracy": 0.13819051831960677,
      "num_tokens": 3153479.0,
      "step": 1245
    },
    {
      "entropy": 7.166897106170654,
      "epoch": 0.14425851125216388,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999906158874368,
      "loss": 7.2083,
      "mean_token_accuracy": 0.1376192092895508,
      "num_tokens": 3167062.0,
      "step": 1250
    },
    {
      "entropy": 7.201770496368408,
      "epoch": 0.14483554529717255,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999902352343444,
      "loss": 7.1495,
      "mean_token_accuracy": 0.13775450959801674,
      "num_tokens": 3179061.0,
      "step": 1255
    },
    {
      "entropy": 7.051590633392334,
      "epoch": 0.1454125793421812,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999898470138171,
      "loss": 7.0898,
      "mean_token_accuracy": 0.13835083171725274,
      "num_tokens": 3191333.0,
      "step": 1260
    },
    {
      "entropy": 7.196379327774048,
      "epoch": 0.14598961338718985,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999894512258681,
      "loss": 7.2138,
      "mean_token_accuracy": 0.13437894061207772,
      "num_tokens": 3204487.0,
      "step": 1265
    },
    {
      "entropy": 7.099405574798584,
      "epoch": 0.1465666474321985,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999890478705107,
      "loss": 7.1498,
      "mean_token_accuracy": 0.13982188031077386,
      "num_tokens": 3216733.0,
      "step": 1270
    },
    {
      "entropy": 7.126603460311889,
      "epoch": 0.14714368147720716,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999886369477585,
      "loss": 7.1008,
      "mean_token_accuracy": 0.14624265134334563,
      "num_tokens": 3229605.0,
      "step": 1275
    },
    {
      "entropy": 7.108373546600342,
      "epoch": 0.1477207155222158,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999882184576251,
      "loss": 7.1291,
      "mean_token_accuracy": 0.14253087043762208,
      "num_tokens": 3241023.0,
      "step": 1280
    },
    {
      "entropy": 7.071506214141846,
      "epoch": 0.14829774956722447,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999877924001248,
      "loss": 7.1085,
      "mean_token_accuracy": 0.140438162535429,
      "num_tokens": 3252435.0,
      "step": 1285
    },
    {
      "entropy": 7.088503217697143,
      "epoch": 0.1488747836122331,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999873587752718,
      "loss": 7.1087,
      "mean_token_accuracy": 0.14706358760595323,
      "num_tokens": 3265663.0,
      "step": 1290
    },
    {
      "entropy": 7.13412880897522,
      "epoch": 0.14945181765724178,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999869175830808,
      "loss": 7.1706,
      "mean_token_accuracy": 0.13438717648386955,
      "num_tokens": 3278309.0,
      "step": 1295
    },
    {
      "entropy": 7.13091197013855,
      "epoch": 0.15002885170225044,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999864688235666,
      "loss": 7.1685,
      "mean_token_accuracy": 0.13528112545609475,
      "num_tokens": 3290899.0,
      "step": 1300
    },
    {
      "entropy": 7.147723531723022,
      "epoch": 0.15060588574725908,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999860124967442,
      "loss": 7.1177,
      "mean_token_accuracy": 0.1383133314549923,
      "num_tokens": 3304383.0,
      "step": 1305
    },
    {
      "entropy": 7.104861927032471,
      "epoch": 0.15118291979226775,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999855486026291,
      "loss": 7.1012,
      "mean_token_accuracy": 0.13756937012076378,
      "num_tokens": 3316310.0,
      "step": 1310
    },
    {
      "entropy": 7.1159440040588375,
      "epoch": 0.1517599538372764,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999850771412369,
      "loss": 7.2206,
      "mean_token_accuracy": 0.1360413908958435,
      "num_tokens": 3330018.0,
      "step": 1315
    },
    {
      "entropy": 7.2038905143737795,
      "epoch": 0.15233698788228506,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999845981125832,
      "loss": 7.2703,
      "mean_token_accuracy": 0.13766002431511878,
      "num_tokens": 3343134.0,
      "step": 1320
    },
    {
      "entropy": 6.998542404174804,
      "epoch": 0.1529140219272937,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999841115166844,
      "loss": 7.0589,
      "mean_token_accuracy": 0.14428165778517724,
      "num_tokens": 3356262.0,
      "step": 1325
    },
    {
      "entropy": 7.208941078186035,
      "epoch": 0.15349105597230236,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999836173535568,
      "loss": 7.2169,
      "mean_token_accuracy": 0.1266804538667202,
      "num_tokens": 3369672.0,
      "step": 1330
    },
    {
      "entropy": 7.099120426177978,
      "epoch": 0.15406809001731103,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999831156232169,
      "loss": 7.1171,
      "mean_token_accuracy": 0.14152806103229523,
      "num_tokens": 3381677.0,
      "step": 1335
    },
    {
      "entropy": 7.098885917663575,
      "epoch": 0.15464512406231967,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999826063256818,
      "loss": 7.0637,
      "mean_token_accuracy": 0.14230246171355249,
      "num_tokens": 3394662.0,
      "step": 1340
    },
    {
      "entropy": 7.04814887046814,
      "epoch": 0.15522215810732834,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999820894609683,
      "loss": 7.0554,
      "mean_token_accuracy": 0.1400335341691971,
      "num_tokens": 3407286.0,
      "step": 1345
    },
    {
      "entropy": 7.218550014495849,
      "epoch": 0.15579919215233698,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999815650290941,
      "loss": 7.1649,
      "mean_token_accuracy": 0.13119087666273116,
      "num_tokens": 3419478.0,
      "step": 1350
    },
    {
      "entropy": 7.062303972244263,
      "epoch": 0.15637622619734565,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999810330300766,
      "loss": 7.1215,
      "mean_token_accuracy": 0.13765157833695413,
      "num_tokens": 3433188.0,
      "step": 1355
    },
    {
      "entropy": 7.134823131561279,
      "epoch": 0.1569532602423543,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499980493463934,
      "loss": 7.055,
      "mean_token_accuracy": 0.13748886287212372,
      "num_tokens": 3445806.0,
      "step": 1360
    },
    {
      "entropy": 7.075977420806884,
      "epoch": 0.15753029428736295,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999799463306841,
      "loss": 7.1166,
      "mean_token_accuracy": 0.1390516623854637,
      "num_tokens": 3458931.0,
      "step": 1365
    },
    {
      "entropy": 7.105384302139282,
      "epoch": 0.15810732833237162,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004999793916303455,
      "loss": 7.1395,
      "mean_token_accuracy": 0.13716089800000192,
      "num_tokens": 3471338.0,
      "step": 1370
    },
    {
      "entropy": 7.144868850708008,
      "epoch": 0.15868436237738026,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999788293629368,
      "loss": 7.1956,
      "mean_token_accuracy": 0.13625281676650047,
      "num_tokens": 3484865.0,
      "step": 1375
    },
    {
      "entropy": 7.1175049304962155,
      "epoch": 0.15926139642238893,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499978259528477,
      "loss": 7.027,
      "mean_token_accuracy": 0.14556030482053756,
      "num_tokens": 3496676.0,
      "step": 1380
    },
    {
      "entropy": 6.959357500076294,
      "epoch": 0.15983843046739757,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999776821269852,
      "loss": 7.0128,
      "mean_token_accuracy": 0.14267347529530525,
      "num_tokens": 3508748.0,
      "step": 1385
    },
    {
      "entropy": 7.115108299255371,
      "epoch": 0.16041546451240624,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999770971584807,
      "loss": 7.1525,
      "mean_token_accuracy": 0.13423006609082222,
      "num_tokens": 3521490.0,
      "step": 1390
    },
    {
      "entropy": 7.066515588760376,
      "epoch": 0.16099249855741488,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999765046229834,
      "loss": 7.1036,
      "mean_token_accuracy": 0.13900410309433936,
      "num_tokens": 3534092.0,
      "step": 1395
    },
    {
      "entropy": 7.138736629486084,
      "epoch": 0.16156953260242354,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499975904520513,
      "loss": 7.0817,
      "mean_token_accuracy": 0.13845248967409135,
      "num_tokens": 3546583.0,
      "step": 1400
    },
    {
      "entropy": 7.003001499176025,
      "epoch": 0.1621465666474322,
      "grad_norm": 0.9375,
      "learning_rate": 0.00049997529685109,
      "loss": 7.062,
      "mean_token_accuracy": 0.13950492888689042,
      "num_tokens": 3559409.0,
      "step": 1405
    },
    {
      "entropy": 7.0664629459381105,
      "epoch": 0.16272360069244085,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999746816147344,
      "loss": 7.0779,
      "mean_token_accuracy": 0.1423931375145912,
      "num_tokens": 3572362.0,
      "step": 1410
    },
    {
      "entropy": 6.987629556655884,
      "epoch": 0.16330063473744952,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999740588114673,
      "loss": 7.0577,
      "mean_token_accuracy": 0.1511901132762432,
      "num_tokens": 3585505.0,
      "step": 1415
    },
    {
      "entropy": 7.129745292663574,
      "epoch": 0.16387766878245816,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999734284413095,
      "loss": 7.089,
      "mean_token_accuracy": 0.14135312885046006,
      "num_tokens": 3597322.0,
      "step": 1420
    },
    {
      "entropy": 7.032627248764038,
      "epoch": 0.16445470282746683,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499972790504282,
      "loss": 7.08,
      "mean_token_accuracy": 0.13861621245741845,
      "num_tokens": 3609941.0,
      "step": 1425
    },
    {
      "entropy": 7.072962760925293,
      "epoch": 0.16503173687247547,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999721450004067,
      "loss": 7.1308,
      "mean_token_accuracy": 0.1348639465868473,
      "num_tokens": 3621751.0,
      "step": 1430
    },
    {
      "entropy": 7.07695050239563,
      "epoch": 0.16560877091748413,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999714919297049,
      "loss": 7.0637,
      "mean_token_accuracy": 0.13544038236141204,
      "num_tokens": 3633070.0,
      "step": 1435
    },
    {
      "entropy": 7.08627667427063,
      "epoch": 0.1661858049624928,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999708312921987,
      "loss": 7.062,
      "mean_token_accuracy": 0.1404960945248604,
      "num_tokens": 3646730.0,
      "step": 1440
    },
    {
      "entropy": 7.044391202926636,
      "epoch": 0.16676283900750144,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999701630879103,
      "loss": 7.0817,
      "mean_token_accuracy": 0.13194262310862542,
      "num_tokens": 3659976.0,
      "step": 1445
    },
    {
      "entropy": 7.109149980545044,
      "epoch": 0.1673398730525101,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999694873168623,
      "loss": 7.0424,
      "mean_token_accuracy": 0.13931129574775697,
      "num_tokens": 3672607.0,
      "step": 1450
    },
    {
      "entropy": 6.845404720306396,
      "epoch": 0.16791690709751875,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999688039790773,
      "loss": 6.9822,
      "mean_token_accuracy": 0.1478599414229393,
      "num_tokens": 3686338.0,
      "step": 1455
    },
    {
      "entropy": 7.102566385269165,
      "epoch": 0.16849394114252741,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999681130745784,
      "loss": 6.9808,
      "mean_token_accuracy": 0.14532144814729692,
      "num_tokens": 3698986.0,
      "step": 1460
    },
    {
      "entropy": 6.975255870819092,
      "epoch": 0.16907097518753605,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999674146033888,
      "loss": 7.0262,
      "mean_token_accuracy": 0.14121559485793114,
      "num_tokens": 3712453.0,
      "step": 1465
    },
    {
      "entropy": 6.975798940658569,
      "epoch": 0.16964800923254472,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999667085655318,
      "loss": 7.0204,
      "mean_token_accuracy": 0.13978162109851838,
      "num_tokens": 3724769.0,
      "step": 1470
    },
    {
      "entropy": 7.082622957229614,
      "epoch": 0.1702250432775534,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999659949610313,
      "loss": 7.0893,
      "mean_token_accuracy": 0.1329570822417736,
      "num_tokens": 3738052.0,
      "step": 1475
    },
    {
      "entropy": 7.094326066970825,
      "epoch": 0.17080207732256203,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999652737899114,
      "loss": 7.069,
      "mean_token_accuracy": 0.1364828757941723,
      "num_tokens": 3751279.0,
      "step": 1480
    },
    {
      "entropy": 7.166085195541382,
      "epoch": 0.1713791113675707,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004999645450521963,
      "loss": 7.103,
      "mean_token_accuracy": 0.13400057479739189,
      "num_tokens": 3764688.0,
      "step": 1485
    },
    {
      "entropy": 6.996908283233642,
      "epoch": 0.17195614541257934,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999638087479104,
      "loss": 7.0783,
      "mean_token_accuracy": 0.13772108033299446,
      "num_tokens": 3776892.0,
      "step": 1490
    },
    {
      "entropy": 7.068263149261474,
      "epoch": 0.172533179457588,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999630648770786,
      "loss": 6.9956,
      "mean_token_accuracy": 0.14322299733757973,
      "num_tokens": 3789287.0,
      "step": 1495
    },
    {
      "entropy": 7.010704708099365,
      "epoch": 0.17311021350259664,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999623134397257,
      "loss": 6.9804,
      "mean_token_accuracy": 0.14074287414550782,
      "num_tokens": 3801467.0,
      "step": 1500
    },
    {
      "entropy": 7.005624055862427,
      "epoch": 0.1736872475476053,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999615544358774,
      "loss": 7.0687,
      "mean_token_accuracy": 0.1379397466778755,
      "num_tokens": 3814641.0,
      "step": 1505
    },
    {
      "entropy": 7.096045875549317,
      "epoch": 0.17426428159261395,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999607878655587,
      "loss": 7.1202,
      "mean_token_accuracy": 0.1285720944404602,
      "num_tokens": 3827187.0,
      "step": 1510
    },
    {
      "entropy": 6.925661325454712,
      "epoch": 0.17484131563762262,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999600137287958,
      "loss": 6.9458,
      "mean_token_accuracy": 0.14802878797054292,
      "num_tokens": 3839832.0,
      "step": 1515
    },
    {
      "entropy": 6.925600910186768,
      "epoch": 0.17541834968263129,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999592320256146,
      "loss": 6.9369,
      "mean_token_accuracy": 0.14163559377193452,
      "num_tokens": 3852026.0,
      "step": 1520
    },
    {
      "entropy": 7.0402978420257565,
      "epoch": 0.17599538372763993,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999584427560412,
      "loss": 6.9887,
      "mean_token_accuracy": 0.14230965897440911,
      "num_tokens": 3864481.0,
      "step": 1525
    },
    {
      "entropy": 6.93634524345398,
      "epoch": 0.1765724177726486,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999576459201024,
      "loss": 6.9934,
      "mean_token_accuracy": 0.13747925013303758,
      "num_tokens": 3876896.0,
      "step": 1530
    },
    {
      "entropy": 6.970355272293091,
      "epoch": 0.17714945181765723,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499956841517825,
      "loss": 6.9343,
      "mean_token_accuracy": 0.14277037978172302,
      "num_tokens": 3890230.0,
      "step": 1535
    },
    {
      "entropy": 7.0250482082366945,
      "epoch": 0.1777264858626659,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499956029549236,
      "loss": 7.0565,
      "mean_token_accuracy": 0.13747095838189125,
      "num_tokens": 3903138.0,
      "step": 1540
    },
    {
      "entropy": 6.930648469924927,
      "epoch": 0.17830351990767454,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999552100143625,
      "loss": 6.9767,
      "mean_token_accuracy": 0.14407200664281844,
      "num_tokens": 3915709.0,
      "step": 1545
    },
    {
      "entropy": 6.993716144561768,
      "epoch": 0.1788805539526832,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004999543829132324,
      "loss": 7.0377,
      "mean_token_accuracy": 0.14300602152943612,
      "num_tokens": 3929265.0,
      "step": 1550
    },
    {
      "entropy": 6.9691534519195555,
      "epoch": 0.17945758799769188,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999535482458734,
      "loss": 7.0303,
      "mean_token_accuracy": 0.14471401274204254,
      "num_tokens": 3940416.0,
      "step": 1555
    },
    {
      "entropy": 7.017536211013794,
      "epoch": 0.18003462204270052,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999527060123135,
      "loss": 7.0353,
      "mean_token_accuracy": 0.13983824625611305,
      "num_tokens": 3952999.0,
      "step": 1560
    },
    {
      "entropy": 6.961780738830567,
      "epoch": 0.18061165608770918,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999518562125811,
      "loss": 7.0434,
      "mean_token_accuracy": 0.1446625217795372,
      "num_tokens": 3966852.0,
      "step": 1565
    },
    {
      "entropy": 7.034871435165405,
      "epoch": 0.18118869013271782,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999509988467047,
      "loss": 6.9454,
      "mean_token_accuracy": 0.1454066276550293,
      "num_tokens": 3978242.0,
      "step": 1570
    },
    {
      "entropy": 6.999048185348511,
      "epoch": 0.1817657241777265,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999501339147132,
      "loss": 7.0631,
      "mean_token_accuracy": 0.1318775437772274,
      "num_tokens": 3993122.0,
      "step": 1575
    },
    {
      "entropy": 7.070727872848511,
      "epoch": 0.18234275822273513,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999492614166357,
      "loss": 6.9982,
      "mean_token_accuracy": 0.13728258460760118,
      "num_tokens": 4005230.0,
      "step": 1580
    },
    {
      "entropy": 6.8208441734313965,
      "epoch": 0.1829197922677438,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999483813525014,
      "loss": 6.9562,
      "mean_token_accuracy": 0.1456133134663105,
      "num_tokens": 4017413.0,
      "step": 1585
    },
    {
      "entropy": 7.0011646270751955,
      "epoch": 0.18349682631275246,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999474937223403,
      "loss": 6.9658,
      "mean_token_accuracy": 0.13678978830575944,
      "num_tokens": 4029679.0,
      "step": 1590
    },
    {
      "entropy": 6.959320497512818,
      "epoch": 0.1840738603577611,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999465985261818,
      "loss": 6.9334,
      "mean_token_accuracy": 0.13753121122717857,
      "num_tokens": 4041857.0,
      "step": 1595
    },
    {
      "entropy": 6.92169075012207,
      "epoch": 0.18465089440276977,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999456957640562,
      "loss": 7.0058,
      "mean_token_accuracy": 0.13973769545555115,
      "num_tokens": 4054743.0,
      "step": 1600
    },
    {
      "entropy": 6.9015570163726805,
      "epoch": 0.1852279284477784,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999447854359939,
      "loss": 6.9409,
      "mean_token_accuracy": 0.14501563981175422,
      "num_tokens": 4067554.0,
      "step": 1605
    },
    {
      "entropy": 6.993334484100342,
      "epoch": 0.18580496249278708,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999438675420255,
      "loss": 6.9275,
      "mean_token_accuracy": 0.1423199862241745,
      "num_tokens": 4079912.0,
      "step": 1610
    },
    {
      "entropy": 6.959068775177002,
      "epoch": 0.18638199653779572,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999429420821818,
      "loss": 7.0423,
      "mean_token_accuracy": 0.13619612157344818,
      "num_tokens": 4093217.0,
      "step": 1615
    },
    {
      "entropy": 6.929660987854004,
      "epoch": 0.1869590305828044,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499942009056494,
      "loss": 6.9966,
      "mean_token_accuracy": 0.13814261704683303,
      "num_tokens": 4104328.0,
      "step": 1620
    },
    {
      "entropy": 7.034404706954956,
      "epoch": 0.18753606462781305,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999410684649935,
      "loss": 6.9616,
      "mean_token_accuracy": 0.13907690942287446,
      "num_tokens": 4116711.0,
      "step": 1625
    },
    {
      "entropy": 6.979753398895264,
      "epoch": 0.1881130986728217,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999401203077119,
      "loss": 6.9733,
      "mean_token_accuracy": 0.14133966118097305,
      "num_tokens": 4129066.0,
      "step": 1630
    },
    {
      "entropy": 6.9114867687225345,
      "epoch": 0.18869013271783036,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000499939164584681,
      "loss": 6.8953,
      "mean_token_accuracy": 0.14570464044809342,
      "num_tokens": 4141312.0,
      "step": 1635
    },
    {
      "entropy": 6.891939115524292,
      "epoch": 0.189267166762839,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999382012959331,
      "loss": 6.9311,
      "mean_token_accuracy": 0.14414507225155831,
      "num_tokens": 4154687.0,
      "step": 1640
    },
    {
      "entropy": 7.044634866714477,
      "epoch": 0.18984420080784767,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999372304415005,
      "loss": 7.0448,
      "mean_token_accuracy": 0.13355328813195227,
      "num_tokens": 4168401.0,
      "step": 1645
    },
    {
      "entropy": 6.9447554588317875,
      "epoch": 0.1904212348528563,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999362520214159,
      "loss": 6.949,
      "mean_token_accuracy": 0.14309759065508842,
      "num_tokens": 4180707.0,
      "step": 1650
    },
    {
      "entropy": 6.960445737838745,
      "epoch": 0.19099826889786498,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999352660357122,
      "loss": 6.8926,
      "mean_token_accuracy": 0.145469331741333,
      "num_tokens": 4192939.0,
      "step": 1655
    },
    {
      "entropy": 6.897533369064331,
      "epoch": 0.19157530294287364,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999342724844226,
      "loss": 6.9289,
      "mean_token_accuracy": 0.14193087741732596,
      "num_tokens": 4206277.0,
      "step": 1660
    },
    {
      "entropy": 6.960471534729004,
      "epoch": 0.19215233698788228,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999332713675804,
      "loss": 6.9495,
      "mean_token_accuracy": 0.1454074963927269,
      "num_tokens": 4218672.0,
      "step": 1665
    },
    {
      "entropy": 6.909567642211914,
      "epoch": 0.19272937103289095,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999322626852193,
      "loss": 6.9078,
      "mean_token_accuracy": 0.15034415572881699,
      "num_tokens": 4229002.0,
      "step": 1670
    },
    {
      "entropy": 6.856534433364868,
      "epoch": 0.1933064050778996,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999312464373734,
      "loss": 6.822,
      "mean_token_accuracy": 0.15208742171525955,
      "num_tokens": 4241986.0,
      "step": 1675
    },
    {
      "entropy": 6.847362995147705,
      "epoch": 0.19388343912290826,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999302226240767,
      "loss": 6.9334,
      "mean_token_accuracy": 0.14470148533582688,
      "num_tokens": 4254670.0,
      "step": 1680
    },
    {
      "entropy": 6.94252290725708,
      "epoch": 0.1944604731679169,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999291912453637,
      "loss": 6.9721,
      "mean_token_accuracy": 0.14320756793022155,
      "num_tokens": 4267561.0,
      "step": 1685
    },
    {
      "entropy": 6.993172025680542,
      "epoch": 0.19503750721292556,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999281523012691,
      "loss": 6.8919,
      "mean_token_accuracy": 0.14741230085492135,
      "num_tokens": 4280145.0,
      "step": 1690
    },
    {
      "entropy": 6.8782483577728275,
      "epoch": 0.1956145412579342,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999271057918278,
      "loss": 6.9383,
      "mean_token_accuracy": 0.1460707038640976,
      "num_tokens": 4294089.0,
      "step": 1695
    },
    {
      "entropy": 6.885042476654053,
      "epoch": 0.19619157530294287,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999260517170751,
      "loss": 6.8834,
      "mean_token_accuracy": 0.14247703105211257,
      "num_tokens": 4305926.0,
      "step": 1700
    },
    {
      "entropy": 6.926484394073486,
      "epoch": 0.19676860934795154,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999249900770463,
      "loss": 6.9204,
      "mean_token_accuracy": 0.1449449948966503,
      "num_tokens": 4319319.0,
      "step": 1705
    },
    {
      "entropy": 6.942605447769165,
      "epoch": 0.19734564339296018,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999239208717772,
      "loss": 6.9145,
      "mean_token_accuracy": 0.14836193472146988,
      "num_tokens": 4331650.0,
      "step": 1710
    },
    {
      "entropy": 6.88254075050354,
      "epoch": 0.19792267743796885,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999228441013037,
      "loss": 6.9039,
      "mean_token_accuracy": 0.14810601770877838,
      "num_tokens": 4344289.0,
      "step": 1715
    },
    {
      "entropy": 6.884267568588257,
      "epoch": 0.1984997114829775,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499921759765662,
      "loss": 6.8864,
      "mean_token_accuracy": 0.14529623389244078,
      "num_tokens": 4357316.0,
      "step": 1720
    },
    {
      "entropy": 6.860304594039917,
      "epoch": 0.19907674552798615,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999206678648888,
      "loss": 6.8982,
      "mean_token_accuracy": 0.1503975972533226,
      "num_tokens": 4369383.0,
      "step": 1725
    },
    {
      "entropy": 6.8668560028076175,
      "epoch": 0.1996537795729948,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999195683990206,
      "loss": 6.9096,
      "mean_token_accuracy": 0.1409059315919876,
      "num_tokens": 4383509.0,
      "step": 1730
    },
    {
      "entropy": 6.973160696029663,
      "epoch": 0.20023081361800346,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999184613680945,
      "loss": 6.8692,
      "mean_token_accuracy": 0.14564202576875687,
      "num_tokens": 4395105.0,
      "step": 1735
    },
    {
      "entropy": 6.832128429412842,
      "epoch": 0.20080784766301213,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999173467721476,
      "loss": 6.872,
      "mean_token_accuracy": 0.15373900681734085,
      "num_tokens": 4407473.0,
      "step": 1740
    },
    {
      "entropy": 7.00480284690857,
      "epoch": 0.20138488170802077,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999162246112175,
      "loss": 6.9639,
      "mean_token_accuracy": 0.13571243211627007,
      "num_tokens": 4420186.0,
      "step": 1745
    },
    {
      "entropy": 6.829422235488892,
      "epoch": 0.20196191575302944,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999150948853419,
      "loss": 6.8759,
      "mean_token_accuracy": 0.14562891870737077,
      "num_tokens": 4432944.0,
      "step": 1750
    },
    {
      "entropy": 6.9459367275238035,
      "epoch": 0.20253894979803808,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999139575945587,
      "loss": 6.9292,
      "mean_token_accuracy": 0.14216466695070268,
      "num_tokens": 4445784.0,
      "step": 1755
    },
    {
      "entropy": 6.793893766403198,
      "epoch": 0.20311598384304674,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999128127389065,
      "loss": 6.8261,
      "mean_token_accuracy": 0.14876734018325805,
      "num_tokens": 4458338.0,
      "step": 1760
    },
    {
      "entropy": 6.883917713165284,
      "epoch": 0.20369301788805538,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999116603184233,
      "loss": 6.8566,
      "mean_token_accuracy": 0.14208449944853782,
      "num_tokens": 4471058.0,
      "step": 1765
    },
    {
      "entropy": 6.897007083892822,
      "epoch": 0.20427005193306405,
      "grad_norm": 1.5,
      "learning_rate": 0.0004999105003331482,
      "loss": 6.8803,
      "mean_token_accuracy": 0.14604100286960603,
      "num_tokens": 4485561.0,
      "step": 1770
    },
    {
      "entropy": 6.783369731903076,
      "epoch": 0.20484708597807272,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999093327831202,
      "loss": 6.899,
      "mean_token_accuracy": 0.14240699410438537,
      "num_tokens": 4497865.0,
      "step": 1775
    },
    {
      "entropy": 6.795016145706176,
      "epoch": 0.20542412002308136,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999081576683784,
      "loss": 6.7786,
      "mean_token_accuracy": 0.1580390974879265,
      "num_tokens": 4510291.0,
      "step": 1780
    },
    {
      "entropy": 6.883041477203369,
      "epoch": 0.20600115406809003,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999069749889626,
      "loss": 6.857,
      "mean_token_accuracy": 0.1457240477204323,
      "num_tokens": 4521855.0,
      "step": 1785
    },
    {
      "entropy": 6.8854797840118405,
      "epoch": 0.20657818811309867,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999057847449123,
      "loss": 6.971,
      "mean_token_accuracy": 0.14121209233999252,
      "num_tokens": 4534123.0,
      "step": 1790
    },
    {
      "entropy": 6.984848356246948,
      "epoch": 0.20715522215810733,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999045869362678,
      "loss": 6.9267,
      "mean_token_accuracy": 0.13553888350725174,
      "num_tokens": 4547298.0,
      "step": 1795
    },
    {
      "entropy": 6.798803997039795,
      "epoch": 0.20773225620311597,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499903381563069,
      "loss": 6.8454,
      "mean_token_accuracy": 0.14686679393053054,
      "num_tokens": 4560945.0,
      "step": 1800
    },
    {
      "entropy": 6.964410018920899,
      "epoch": 0.20830929024812464,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999021686253568,
      "loss": 6.9114,
      "mean_token_accuracy": 0.14516634494066238,
      "num_tokens": 4572696.0,
      "step": 1805
    },
    {
      "entropy": 6.751922273635865,
      "epoch": 0.2088863242931333,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999009481231719,
      "loss": 6.8756,
      "mean_token_accuracy": 0.14628861993551254,
      "num_tokens": 4585670.0,
      "step": 1810
    },
    {
      "entropy": 6.973953628540039,
      "epoch": 0.20946335833814195,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004998997200565553,
      "loss": 6.8579,
      "mean_token_accuracy": 0.1476499393582344,
      "num_tokens": 4597694.0,
      "step": 1815
    },
    {
      "entropy": 6.862924671173095,
      "epoch": 0.21004039238315061,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998984844255483,
      "loss": 6.8584,
      "mean_token_accuracy": 0.145158788561821,
      "num_tokens": 4609924.0,
      "step": 1820
    },
    {
      "entropy": 6.824046230316162,
      "epoch": 0.21061742642815925,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998972412301925,
      "loss": 6.8614,
      "mean_token_accuracy": 0.14629337787628174,
      "num_tokens": 4621668.0,
      "step": 1825
    },
    {
      "entropy": 6.91977014541626,
      "epoch": 0.21119446047316792,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004998959904705297,
      "loss": 6.925,
      "mean_token_accuracy": 0.14481958076357843,
      "num_tokens": 4635970.0,
      "step": 1830
    },
    {
      "entropy": 6.877291536331176,
      "epoch": 0.21177149451817656,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998947321466021,
      "loss": 6.9056,
      "mean_token_accuracy": 0.14075979888439177,
      "num_tokens": 4649557.0,
      "step": 1835
    },
    {
      "entropy": 6.868559741973877,
      "epoch": 0.21234852856318523,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998934662584518,
      "loss": 6.8864,
      "mean_token_accuracy": 0.1476551428437233,
      "num_tokens": 4662572.0,
      "step": 1840
    },
    {
      "entropy": 6.886713933944702,
      "epoch": 0.2129255626081939,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998921928061214,
      "loss": 6.8409,
      "mean_token_accuracy": 0.13886594846844674,
      "num_tokens": 4674783.0,
      "step": 1845
    },
    {
      "entropy": 6.78210129737854,
      "epoch": 0.21350259665320254,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998909117896539,
      "loss": 6.8551,
      "mean_token_accuracy": 0.150478532910347,
      "num_tokens": 4687543.0,
      "step": 1850
    },
    {
      "entropy": 6.975531578063965,
      "epoch": 0.2140796306982112,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998896232090922,
      "loss": 6.8099,
      "mean_token_accuracy": 0.14673146605491638,
      "num_tokens": 4700225.0,
      "step": 1855
    },
    {
      "entropy": 6.6987810134887695,
      "epoch": 0.21465666474321984,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004998883270644798,
      "loss": 6.8708,
      "mean_token_accuracy": 0.14261882826685907,
      "num_tokens": 4712950.0,
      "step": 1860
    },
    {
      "entropy": 6.979980373382569,
      "epoch": 0.2152336987882285,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998870233558602,
      "loss": 6.9684,
      "mean_token_accuracy": 0.14130085632205008,
      "num_tokens": 4725563.0,
      "step": 1865
    },
    {
      "entropy": 6.872949504852295,
      "epoch": 0.21581073283323715,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998857120832774,
      "loss": 6.8768,
      "mean_token_accuracy": 0.14834909737110138,
      "num_tokens": 4739057.0,
      "step": 1870
    },
    {
      "entropy": 6.829817581176758,
      "epoch": 0.21638776687824582,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998843932467751,
      "loss": 6.8336,
      "mean_token_accuracy": 0.15014879405498505,
      "num_tokens": 4752782.0,
      "step": 1875
    },
    {
      "entropy": 6.834365272521973,
      "epoch": 0.21696480092325446,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998830668463982,
      "loss": 6.7885,
      "mean_token_accuracy": 0.1466831497848034,
      "num_tokens": 4765740.0,
      "step": 1880
    },
    {
      "entropy": 6.809153747558594,
      "epoch": 0.21754183496826313,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998817328821909,
      "loss": 6.8593,
      "mean_token_accuracy": 0.1415070801973343,
      "num_tokens": 4778711.0,
      "step": 1885
    },
    {
      "entropy": 6.89850492477417,
      "epoch": 0.2181188690132718,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998803913541983,
      "loss": 6.8493,
      "mean_token_accuracy": 0.143658284842968,
      "num_tokens": 4791402.0,
      "step": 1890
    },
    {
      "entropy": 6.818964958190918,
      "epoch": 0.21869590305828043,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998790422624654,
      "loss": 6.8456,
      "mean_token_accuracy": 0.1407043866813183,
      "num_tokens": 4804256.0,
      "step": 1895
    },
    {
      "entropy": 6.819336557388306,
      "epoch": 0.2192729371032891,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998776856070376,
      "loss": 6.789,
      "mean_token_accuracy": 0.14376113265752793,
      "num_tokens": 4817424.0,
      "step": 1900
    },
    {
      "entropy": 6.805288505554199,
      "epoch": 0.21984997114829774,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004998763213879606,
      "loss": 6.8274,
      "mean_token_accuracy": 0.1412200078368187,
      "num_tokens": 4829222.0,
      "step": 1905
    },
    {
      "entropy": 6.7459861755371096,
      "epoch": 0.2204270051933064,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998749496052803,
      "loss": 6.8126,
      "mean_token_accuracy": 0.14899692088365554,
      "num_tokens": 4844011.0,
      "step": 1910
    },
    {
      "entropy": 6.881348991394043,
      "epoch": 0.22100403923831505,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998735702590426,
      "loss": 6.8511,
      "mean_token_accuracy": 0.1448480397462845,
      "num_tokens": 4857116.0,
      "step": 1915
    },
    {
      "entropy": 6.781710624694824,
      "epoch": 0.22158107328332372,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998721833492942,
      "loss": 6.7721,
      "mean_token_accuracy": 0.15325645804405214,
      "num_tokens": 4867680.0,
      "step": 1920
    },
    {
      "entropy": 6.845559120178223,
      "epoch": 0.22215810732833238,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004998707888760816,
      "loss": 6.8473,
      "mean_token_accuracy": 0.1444242402911186,
      "num_tokens": 4880553.0,
      "step": 1925
    },
    {
      "entropy": 6.795937204360962,
      "epoch": 0.22273514137334102,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998693868394516,
      "loss": 6.8314,
      "mean_token_accuracy": 0.14200094863772392,
      "num_tokens": 4893706.0,
      "step": 1930
    },
    {
      "entropy": 6.812420654296875,
      "epoch": 0.2233121754183497,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998679772394516,
      "loss": 6.884,
      "mean_token_accuracy": 0.14432224184274672,
      "num_tokens": 4906016.0,
      "step": 1935
    },
    {
      "entropy": 6.793758773803711,
      "epoch": 0.22388920946335833,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998665600761289,
      "loss": 6.7693,
      "mean_token_accuracy": 0.15037692189216614,
      "num_tokens": 4918628.0,
      "step": 1940
    },
    {
      "entropy": 6.891831541061402,
      "epoch": 0.224466243508367,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499865135349531,
      "loss": 6.85,
      "mean_token_accuracy": 0.1410201907157898,
      "num_tokens": 4930525.0,
      "step": 1945
    },
    {
      "entropy": 6.862972545623779,
      "epoch": 0.22504327755337564,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998637030597061,
      "loss": 6.8579,
      "mean_token_accuracy": 0.14448119103908538,
      "num_tokens": 4942510.0,
      "step": 1950
    },
    {
      "entropy": 6.852001047134399,
      "epoch": 0.2256203115983843,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998622632067022,
      "loss": 6.8322,
      "mean_token_accuracy": 0.142857813090086,
      "num_tokens": 4954865.0,
      "step": 1955
    },
    {
      "entropy": 6.849268865585327,
      "epoch": 0.22619734564339297,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004998608157905678,
      "loss": 6.9068,
      "mean_token_accuracy": 0.13810089901089667,
      "num_tokens": 4968078.0,
      "step": 1960
    },
    {
      "entropy": 6.828215026855469,
      "epoch": 0.2267743796884016,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004998593608113515,
      "loss": 6.7723,
      "mean_token_accuracy": 0.14653953462839125,
      "num_tokens": 4980273.0,
      "step": 1965
    },
    {
      "entropy": 6.837485408782959,
      "epoch": 0.22735141373341028,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998578982691024,
      "loss": 6.8295,
      "mean_token_accuracy": 0.1414664812386036,
      "num_tokens": 4992372.0,
      "step": 1970
    },
    {
      "entropy": 6.791807365417481,
      "epoch": 0.22792844777841892,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998564281638694,
      "loss": 6.8131,
      "mean_token_accuracy": 0.14089381247758864,
      "num_tokens": 5004766.0,
      "step": 1975
    },
    {
      "entropy": 6.786174869537353,
      "epoch": 0.2285054818234276,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998549504957023,
      "loss": 6.7601,
      "mean_token_accuracy": 0.14380791336297988,
      "num_tokens": 5017007.0,
      "step": 1980
    },
    {
      "entropy": 6.821545076370239,
      "epoch": 0.22908251586843623,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998534652646506,
      "loss": 6.8221,
      "mean_token_accuracy": 0.14419303238391876,
      "num_tokens": 5029506.0,
      "step": 1985
    },
    {
      "entropy": 6.815482902526855,
      "epoch": 0.2296595499134449,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998519724707642,
      "loss": 6.8091,
      "mean_token_accuracy": 0.1457746997475624,
      "num_tokens": 5042209.0,
      "step": 1990
    },
    {
      "entropy": 6.8471766948699955,
      "epoch": 0.23023658395845356,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998504721140934,
      "loss": 6.8541,
      "mean_token_accuracy": 0.1391491897404194,
      "num_tokens": 5056026.0,
      "step": 1995
    },
    {
      "entropy": 6.837068510055542,
      "epoch": 0.2308136180034622,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998489641946887,
      "loss": 6.8116,
      "mean_token_accuracy": 0.14428771734237672,
      "num_tokens": 5068194.0,
      "step": 2000
    },
    {
      "entropy": 6.664884614944458,
      "epoch": 0.23139065204847087,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998474487126009,
      "loss": 6.7351,
      "mean_token_accuracy": 0.1473590686917305,
      "num_tokens": 5080468.0,
      "step": 2005
    },
    {
      "entropy": 6.884929609298706,
      "epoch": 0.2319676860934795,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004998459256678806,
      "loss": 6.9369,
      "mean_token_accuracy": 0.1359821602702141,
      "num_tokens": 5093104.0,
      "step": 2010
    },
    {
      "entropy": 6.826117992401123,
      "epoch": 0.23254472013848818,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004998443950605796,
      "loss": 6.7957,
      "mean_token_accuracy": 0.14508781284093858,
      "num_tokens": 5105710.0,
      "step": 2015
    },
    {
      "entropy": 6.905029249191284,
      "epoch": 0.23312175418349682,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998428568907488,
      "loss": 6.8745,
      "mean_token_accuracy": 0.14229415580630303,
      "num_tokens": 5118946.0,
      "step": 2020
    },
    {
      "entropy": 6.728935241699219,
      "epoch": 0.23369878822850548,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998413111584403,
      "loss": 6.7202,
      "mean_token_accuracy": 0.15764200091362,
      "num_tokens": 5132567.0,
      "step": 2025
    },
    {
      "entropy": 6.694896554946899,
      "epoch": 0.23427582227351415,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998397578637059,
      "loss": 6.7779,
      "mean_token_accuracy": 0.15084137320518493,
      "num_tokens": 5145669.0,
      "step": 2030
    },
    {
      "entropy": 6.758299112319946,
      "epoch": 0.2348528563185228,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000499838197006598,
      "loss": 6.7169,
      "mean_token_accuracy": 0.1479812353849411,
      "num_tokens": 5157753.0,
      "step": 2035
    },
    {
      "entropy": 6.72536416053772,
      "epoch": 0.23542989036353146,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499836628587169,
      "loss": 6.738,
      "mean_token_accuracy": 0.1493836060166359,
      "num_tokens": 5170167.0,
      "step": 2040
    },
    {
      "entropy": 6.763384819030762,
      "epoch": 0.2360069244085401,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998350526054716,
      "loss": 6.7365,
      "mean_token_accuracy": 0.14397142082452774,
      "num_tokens": 5182143.0,
      "step": 2045
    },
    {
      "entropy": 6.825057792663574,
      "epoch": 0.23658395845354876,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998334690615591,
      "loss": 6.7837,
      "mean_token_accuracy": 0.1431373730301857,
      "num_tokens": 5195586.0,
      "step": 2050
    },
    {
      "entropy": 6.68499903678894,
      "epoch": 0.2371609924985574,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998318779554844,
      "loss": 6.7554,
      "mean_token_accuracy": 0.14732947051525117,
      "num_tokens": 5207774.0,
      "step": 2055
    },
    {
      "entropy": 6.71256422996521,
      "epoch": 0.23773802654356607,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998302792873012,
      "loss": 6.7794,
      "mean_token_accuracy": 0.14680352061986923,
      "num_tokens": 5219330.0,
      "step": 2060
    },
    {
      "entropy": 6.830109930038452,
      "epoch": 0.23831506058857474,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004998286730570631,
      "loss": 6.8465,
      "mean_token_accuracy": 0.1373237304389477,
      "num_tokens": 5233817.0,
      "step": 2065
    },
    {
      "entropy": 6.789051580429077,
      "epoch": 0.23889209463358338,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004998270592648245,
      "loss": 6.8371,
      "mean_token_accuracy": 0.14460220336914062,
      "num_tokens": 5246483.0,
      "step": 2070
    },
    {
      "entropy": 6.836073017120361,
      "epoch": 0.23946912867859205,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998254379106394,
      "loss": 6.829,
      "mean_token_accuracy": 0.14329571947455405,
      "num_tokens": 5258695.0,
      "step": 2075
    },
    {
      "entropy": 6.749055814743042,
      "epoch": 0.2400461627236007,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998238089945622,
      "loss": 6.7726,
      "mean_token_accuracy": 0.14508575648069383,
      "num_tokens": 5271355.0,
      "step": 2080
    },
    {
      "entropy": 6.937589025497436,
      "epoch": 0.24062319676860935,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004998221725166479,
      "loss": 6.8964,
      "mean_token_accuracy": 0.1389458566904068,
      "num_tokens": 5284905.0,
      "step": 2085
    },
    {
      "entropy": 6.719331741333008,
      "epoch": 0.241200230813618,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998205284769516,
      "loss": 6.8109,
      "mean_token_accuracy": 0.14349860548973084,
      "num_tokens": 5298266.0,
      "step": 2090
    },
    {
      "entropy": 6.828605842590332,
      "epoch": 0.24177726485862666,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998188768755285,
      "loss": 6.8694,
      "mean_token_accuracy": 0.1355933129787445,
      "num_tokens": 5311148.0,
      "step": 2095
    },
    {
      "entropy": 6.83973069190979,
      "epoch": 0.2423542989036353,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998172177124341,
      "loss": 6.7748,
      "mean_token_accuracy": 0.14382925629615784,
      "num_tokens": 5322638.0,
      "step": 2100
    },
    {
      "entropy": 6.663211250305176,
      "epoch": 0.24293133294864397,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004998155509877242,
      "loss": 6.6548,
      "mean_token_accuracy": 0.15735870003700256,
      "num_tokens": 5336116.0,
      "step": 2105
    },
    {
      "entropy": 6.7159160614013675,
      "epoch": 0.24350836699365264,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499813876701455,
      "loss": 6.7707,
      "mean_token_accuracy": 0.14908685460686683,
      "num_tokens": 5347561.0,
      "step": 2110
    },
    {
      "entropy": 6.844922876358032,
      "epoch": 0.24408540103866128,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998121948536828,
      "loss": 6.8154,
      "mean_token_accuracy": 0.14128955006599425,
      "num_tokens": 5360055.0,
      "step": 2115
    },
    {
      "entropy": 6.773591470718384,
      "epoch": 0.24466243508366994,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004998105054444639,
      "loss": 6.7771,
      "mean_token_accuracy": 0.15177889317274093,
      "num_tokens": 5372722.0,
      "step": 2120
    },
    {
      "entropy": 6.782438707351685,
      "epoch": 0.24523946912867858,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004998088084738555,
      "loss": 6.7404,
      "mean_token_accuracy": 0.14531630873680115,
      "num_tokens": 5385586.0,
      "step": 2125
    },
    {
      "entropy": 6.681029987335205,
      "epoch": 0.24581650317368725,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004998071039419144,
      "loss": 6.7641,
      "mean_token_accuracy": 0.14682294577360153,
      "num_tokens": 5399266.0,
      "step": 2130
    },
    {
      "entropy": 6.774657392501831,
      "epoch": 0.2463935372186959,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499805391848698,
      "loss": 6.7644,
      "mean_token_accuracy": 0.14703176617622377,
      "num_tokens": 5412093.0,
      "step": 2135
    },
    {
      "entropy": 6.70880913734436,
      "epoch": 0.24697057126370456,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499803672194264,
      "loss": 6.7833,
      "mean_token_accuracy": 0.1430216610431671,
      "num_tokens": 5425181.0,
      "step": 2140
    },
    {
      "entropy": 6.749992370605469,
      "epoch": 0.24754760530871323,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998019449786701,
      "loss": 6.6795,
      "mean_token_accuracy": 0.15330939292907714,
      "num_tokens": 5437392.0,
      "step": 2145
    },
    {
      "entropy": 6.751720428466797,
      "epoch": 0.24812463935372187,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998002102019744,
      "loss": 6.7483,
      "mean_token_accuracy": 0.15037278309464455,
      "num_tokens": 5449455.0,
      "step": 2150
    },
    {
      "entropy": 6.753234720230102,
      "epoch": 0.24870167339873053,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004997984678642354,
      "loss": 6.7739,
      "mean_token_accuracy": 0.14836794808506965,
      "num_tokens": 5463095.0,
      "step": 2155
    },
    {
      "entropy": 6.702136754989624,
      "epoch": 0.24927870744373917,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997967179655115,
      "loss": 6.7224,
      "mean_token_accuracy": 0.15014025568962097,
      "num_tokens": 5474440.0,
      "step": 2160
    },
    {
      "entropy": 6.758479595184326,
      "epoch": 0.24985574148874784,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997949605058617,
      "loss": 6.7367,
      "mean_token_accuracy": 0.15418562889099122,
      "num_tokens": 5487365.0,
      "step": 2165
    },
    {
      "entropy": 6.72683892250061,
      "epoch": 0.2504327755337565,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004997931954853451,
      "loss": 6.8064,
      "mean_token_accuracy": 0.14088582396507263,
      "num_tokens": 5502356.0,
      "step": 2170
    },
    {
      "entropy": 6.764380550384521,
      "epoch": 0.2510098095787651,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499791422904021,
      "loss": 6.8133,
      "mean_token_accuracy": 0.14515454694628716,
      "num_tokens": 5514504.0,
      "step": 2175
    },
    {
      "entropy": 6.811923265457153,
      "epoch": 0.2515868436237738,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997896427619491,
      "loss": 6.8184,
      "mean_token_accuracy": 0.14146610125899314,
      "num_tokens": 5526415.0,
      "step": 2180
    },
    {
      "entropy": 6.721636390686035,
      "epoch": 0.25216387766878245,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997878550591892,
      "loss": 6.6809,
      "mean_token_accuracy": 0.14960418790578842,
      "num_tokens": 5538484.0,
      "step": 2185
    },
    {
      "entropy": 6.822943162918091,
      "epoch": 0.2527409117137911,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004997860597958013,
      "loss": 6.7458,
      "mean_token_accuracy": 0.15074630677700043,
      "num_tokens": 5551315.0,
      "step": 2190
    },
    {
      "entropy": 6.600991916656494,
      "epoch": 0.2533179457587998,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004997842569718461,
      "loss": 6.6448,
      "mean_token_accuracy": 0.14987451508641242,
      "num_tokens": 5564179.0,
      "step": 2195
    },
    {
      "entropy": 6.7274760723114015,
      "epoch": 0.25389497980380843,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499782446587384,
      "loss": 6.7716,
      "mean_token_accuracy": 0.13751927390694618,
      "num_tokens": 5577060.0,
      "step": 2200
    },
    {
      "entropy": 6.6231166362762455,
      "epoch": 0.25447201384881707,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499780628642476,
      "loss": 6.5854,
      "mean_token_accuracy": 0.16829559057950974,
      "num_tokens": 5589170.0,
      "step": 2205
    },
    {
      "entropy": 6.710141897201538,
      "epoch": 0.2550490478938257,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997788031371834,
      "loss": 6.7737,
      "mean_token_accuracy": 0.14554792270064354,
      "num_tokens": 5601747.0,
      "step": 2210
    },
    {
      "entropy": 6.7795628070831295,
      "epoch": 0.2556260819388344,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004997769700715672,
      "loss": 6.7876,
      "mean_token_accuracy": 0.13844683319330214,
      "num_tokens": 5614308.0,
      "step": 2215
    },
    {
      "entropy": 6.6915586471557615,
      "epoch": 0.25620311598384304,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004997751294456892,
      "loss": 6.7544,
      "mean_token_accuracy": 0.14007925242185593,
      "num_tokens": 5626351.0,
      "step": 2220
    },
    {
      "entropy": 6.810964775085449,
      "epoch": 0.2567801500288517,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004997732812596114,
      "loss": 6.7285,
      "mean_token_accuracy": 0.15032831579446793,
      "num_tokens": 5638339.0,
      "step": 2225
    },
    {
      "entropy": 6.73750147819519,
      "epoch": 0.2573571840738604,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997714255133961,
      "loss": 6.7858,
      "mean_token_accuracy": 0.14408532232046128,
      "num_tokens": 5651210.0,
      "step": 2230
    },
    {
      "entropy": 6.72407546043396,
      "epoch": 0.257934218118869,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997695622071054,
      "loss": 6.7764,
      "mean_token_accuracy": 0.148297780752182,
      "num_tokens": 5664150.0,
      "step": 2235
    },
    {
      "entropy": 6.773475217819214,
      "epoch": 0.25851125216387766,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004997676913408021,
      "loss": 6.6947,
      "mean_token_accuracy": 0.15271911919116973,
      "num_tokens": 5676009.0,
      "step": 2240
    },
    {
      "entropy": 6.8198034286499025,
      "epoch": 0.2590882862088863,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004997658129145493,
      "loss": 6.8775,
      "mean_token_accuracy": 0.14045544862747192,
      "num_tokens": 5688557.0,
      "step": 2245
    },
    {
      "entropy": 6.7614494323730465,
      "epoch": 0.259665320253895,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004997639269284099,
      "loss": 6.8247,
      "mean_token_accuracy": 0.14492859840393066,
      "num_tokens": 5702032.0,
      "step": 2250
    },
    {
      "entropy": 6.77425799369812,
      "epoch": 0.26024235429890363,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997620333824476,
      "loss": 6.6499,
      "mean_token_accuracy": 0.1555098697543144,
      "num_tokens": 5714623.0,
      "step": 2255
    },
    {
      "entropy": 6.668704319000244,
      "epoch": 0.2608193883439123,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997601322767258,
      "loss": 6.6668,
      "mean_token_accuracy": 0.1488153889775276,
      "num_tokens": 5727415.0,
      "step": 2260
    },
    {
      "entropy": 6.739748096466064,
      "epoch": 0.26139642238892097,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997582236113087,
      "loss": 6.7189,
      "mean_token_accuracy": 0.14608575105667115,
      "num_tokens": 5740129.0,
      "step": 2265
    },
    {
      "entropy": 6.70829586982727,
      "epoch": 0.2619734564339296,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997563073862603,
      "loss": 6.6847,
      "mean_token_accuracy": 0.15272213518619537,
      "num_tokens": 5752469.0,
      "step": 2270
    },
    {
      "entropy": 6.702632331848145,
      "epoch": 0.26255049047893825,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997543836016453,
      "loss": 6.782,
      "mean_token_accuracy": 0.15359314829111098,
      "num_tokens": 5765408.0,
      "step": 2275
    },
    {
      "entropy": 6.852825546264649,
      "epoch": 0.2631275245239469,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997524522575281,
      "loss": 6.712,
      "mean_token_accuracy": 0.14011794328689575,
      "num_tokens": 5777450.0,
      "step": 2280
    },
    {
      "entropy": 6.71787576675415,
      "epoch": 0.2637045585689556,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004997505133539738,
      "loss": 6.7499,
      "mean_token_accuracy": 0.14179244339466096,
      "num_tokens": 5790435.0,
      "step": 2285
    },
    {
      "entropy": 6.795726108551025,
      "epoch": 0.2642815926139642,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997485668910476,
      "loss": 6.7392,
      "mean_token_accuracy": 0.14688139259815217,
      "num_tokens": 5803517.0,
      "step": 2290
    },
    {
      "entropy": 6.632097005844116,
      "epoch": 0.26485862665897286,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997466128688151,
      "loss": 6.6931,
      "mean_token_accuracy": 0.1566813513636589,
      "num_tokens": 5815802.0,
      "step": 2295
    },
    {
      "entropy": 6.698785448074341,
      "epoch": 0.26543566070398156,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499744651287342,
      "loss": 6.72,
      "mean_token_accuracy": 0.14845843315124513,
      "num_tokens": 5827807.0,
      "step": 2300
    },
    {
      "entropy": 6.728725481033325,
      "epoch": 0.2660126947489902,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499742682146694,
      "loss": 6.7315,
      "mean_token_accuracy": 0.1497019961476326,
      "num_tokens": 5840759.0,
      "step": 2305
    },
    {
      "entropy": 6.780129289627075,
      "epoch": 0.26658972879399884,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004997407054469377,
      "loss": 6.7611,
      "mean_token_accuracy": 0.14581131488084792,
      "num_tokens": 5853117.0,
      "step": 2310
    },
    {
      "entropy": 6.565095138549805,
      "epoch": 0.2671667628390075,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004997387211881392,
      "loss": 6.6753,
      "mean_token_accuracy": 0.1566202774643898,
      "num_tokens": 5866717.0,
      "step": 2315
    },
    {
      "entropy": 6.743575382232666,
      "epoch": 0.26774379688401617,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997367293703656,
      "loss": 6.7424,
      "mean_token_accuracy": 0.14907404631376267,
      "num_tokens": 5879017.0,
      "step": 2320
    },
    {
      "entropy": 6.713337516784668,
      "epoch": 0.2683208309290248,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997347299936837,
      "loss": 6.7133,
      "mean_token_accuracy": 0.14949267506599426,
      "num_tokens": 5891481.0,
      "step": 2325
    },
    {
      "entropy": 6.698743152618408,
      "epoch": 0.26889786497403345,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997327230581608,
      "loss": 6.7359,
      "mean_token_accuracy": 0.14829655289649962,
      "num_tokens": 5903389.0,
      "step": 2330
    },
    {
      "entropy": 6.676599073410034,
      "epoch": 0.26947489901904215,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997307085638644,
      "loss": 6.629,
      "mean_token_accuracy": 0.14779743552207947,
      "num_tokens": 5916084.0,
      "step": 2335
    },
    {
      "entropy": 6.671857261657715,
      "epoch": 0.2700519330640508,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004997286865108621,
      "loss": 6.6841,
      "mean_token_accuracy": 0.15207995921373368,
      "num_tokens": 5927797.0,
      "step": 2340
    },
    {
      "entropy": 6.579140996932983,
      "epoch": 0.2706289671090594,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997266568992222,
      "loss": 6.609,
      "mean_token_accuracy": 0.15099724531173705,
      "num_tokens": 5941886.0,
      "step": 2345
    },
    {
      "entropy": 6.694521522521972,
      "epoch": 0.27120600115406807,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004997246197290128,
      "loss": 6.7122,
      "mean_token_accuracy": 0.15124790519475936,
      "num_tokens": 5953655.0,
      "step": 2350
    },
    {
      "entropy": 6.872830104827881,
      "epoch": 0.27178303519907676,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997225750003024,
      "loss": 6.7348,
      "mean_token_accuracy": 0.14656617790460585,
      "num_tokens": 5967818.0,
      "step": 2355
    },
    {
      "entropy": 6.621160650253296,
      "epoch": 0.2723600692440854,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997205227131599,
      "loss": 6.6683,
      "mean_token_accuracy": 0.15280671566724777,
      "num_tokens": 5980742.0,
      "step": 2360
    },
    {
      "entropy": 6.720431566238403,
      "epoch": 0.27293710328909404,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997184628676542,
      "loss": 6.6961,
      "mean_token_accuracy": 0.14359625577926635,
      "num_tokens": 5993263.0,
      "step": 2365
    },
    {
      "entropy": 6.768740940093994,
      "epoch": 0.27351413733410274,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997163954638546,
      "loss": 6.7821,
      "mean_token_accuracy": 0.13767419382929802,
      "num_tokens": 6007820.0,
      "step": 2370
    },
    {
      "entropy": 6.749808263778687,
      "epoch": 0.2740911713791114,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997143205018306,
      "loss": 6.7393,
      "mean_token_accuracy": 0.14474162012338637,
      "num_tokens": 6020665.0,
      "step": 2375
    },
    {
      "entropy": 6.737027406692505,
      "epoch": 0.27466820542412,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997122379816523,
      "loss": 6.6633,
      "mean_token_accuracy": 0.15110542327165605,
      "num_tokens": 6032847.0,
      "step": 2380
    },
    {
      "entropy": 6.652834367752075,
      "epoch": 0.27524523946912866,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997101479033893,
      "loss": 6.7217,
      "mean_token_accuracy": 0.14369864910840988,
      "num_tokens": 6045291.0,
      "step": 2385
    },
    {
      "entropy": 6.695873260498047,
      "epoch": 0.27582227351413735,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997080502671122,
      "loss": 6.6998,
      "mean_token_accuracy": 0.1474764734506607,
      "num_tokens": 6057872.0,
      "step": 2390
    },
    {
      "entropy": 6.653308582305908,
      "epoch": 0.276399307559146,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004997059450728913,
      "loss": 6.6736,
      "mean_token_accuracy": 0.1510896533727646,
      "num_tokens": 6070252.0,
      "step": 2395
    },
    {
      "entropy": 6.759108924865723,
      "epoch": 0.27697634160415463,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997038323207977,
      "loss": 6.7658,
      "mean_token_accuracy": 0.14066046848893166,
      "num_tokens": 6083728.0,
      "step": 2400
    },
    {
      "entropy": 6.719287300109864,
      "epoch": 0.2775533756491633,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004997017120109022,
      "loss": 6.7103,
      "mean_token_accuracy": 0.14144404977560043,
      "num_tokens": 6097140.0,
      "step": 2405
    },
    {
      "entropy": 6.596685552597046,
      "epoch": 0.27813040969417197,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004996995841432762,
      "loss": 6.6862,
      "mean_token_accuracy": 0.14334119111299515,
      "num_tokens": 6109518.0,
      "step": 2410
    },
    {
      "entropy": 6.788815879821778,
      "epoch": 0.2787074437391806,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996974487179915,
      "loss": 6.6762,
      "mean_token_accuracy": 0.1471431717276573,
      "num_tokens": 6122160.0,
      "step": 2415
    },
    {
      "entropy": 6.597071027755737,
      "epoch": 0.27928447778418924,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996953057351195,
      "loss": 6.7212,
      "mean_token_accuracy": 0.1476117990911007,
      "num_tokens": 6134638.0,
      "step": 2420
    },
    {
      "entropy": 6.760407972335815,
      "epoch": 0.27986151182919794,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004996931551947327,
      "loss": 6.7589,
      "mean_token_accuracy": 0.14486420899629593,
      "num_tokens": 6147142.0,
      "step": 2425
    },
    {
      "entropy": 6.794008064270019,
      "epoch": 0.2804385458742066,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004996909970969031,
      "loss": 6.744,
      "mean_token_accuracy": 0.1503509297966957,
      "num_tokens": 6159137.0,
      "step": 2430
    },
    {
      "entropy": 6.6532281875610355,
      "epoch": 0.2810155799192152,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004996888314417034,
      "loss": 6.7328,
      "mean_token_accuracy": 0.14441587403416634,
      "num_tokens": 6171663.0,
      "step": 2435
    },
    {
      "entropy": 6.826012897491455,
      "epoch": 0.2815926139642239,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996866582292067,
      "loss": 6.7408,
      "mean_token_accuracy": 0.14965420067310334,
      "num_tokens": 6184282.0,
      "step": 2440
    },
    {
      "entropy": 6.679483222961426,
      "epoch": 0.28216964800923255,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996844774594856,
      "loss": 6.7343,
      "mean_token_accuracy": 0.14577517807483673,
      "num_tokens": 6197243.0,
      "step": 2445
    },
    {
      "entropy": 6.663284683227539,
      "epoch": 0.2827466820542412,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004996822891326138,
      "loss": 6.6567,
      "mean_token_accuracy": 0.1508615292608738,
      "num_tokens": 6209061.0,
      "step": 2450
    },
    {
      "entropy": 6.653768301010132,
      "epoch": 0.28332371609924983,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004996800932486648,
      "loss": 6.6216,
      "mean_token_accuracy": 0.15076174587011337,
      "num_tokens": 6223497.0,
      "step": 2455
    },
    {
      "entropy": 6.671743392944336,
      "epoch": 0.28390075014425853,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996778898077126,
      "loss": 6.6509,
      "mean_token_accuracy": 0.15512303858995438,
      "num_tokens": 6236395.0,
      "step": 2460
    },
    {
      "entropy": 6.668979549407959,
      "epoch": 0.28447778418926717,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996756788098309,
      "loss": 6.6199,
      "mean_token_accuracy": 0.15135869532823562,
      "num_tokens": 6248700.0,
      "step": 2465
    },
    {
      "entropy": 6.690337991714477,
      "epoch": 0.2850548182342758,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004996734602550945,
      "loss": 6.6867,
      "mean_token_accuracy": 0.14691711217164993,
      "num_tokens": 6262728.0,
      "step": 2470
    },
    {
      "entropy": 6.60182819366455,
      "epoch": 0.2856318522792845,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996712341435779,
      "loss": 6.6504,
      "mean_token_accuracy": 0.14799974262714385,
      "num_tokens": 6276007.0,
      "step": 2475
    },
    {
      "entropy": 6.800230884552002,
      "epoch": 0.28620888632429314,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004996690004753559,
      "loss": 6.7706,
      "mean_token_accuracy": 0.14552113711833953,
      "num_tokens": 6289033.0,
      "step": 2480
    },
    {
      "entropy": 6.621489667892456,
      "epoch": 0.2867859203693018,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996667592505037,
      "loss": 6.7187,
      "mean_token_accuracy": 0.139874604344368,
      "num_tokens": 6300994.0,
      "step": 2485
    },
    {
      "entropy": 6.712352705001831,
      "epoch": 0.2873629544143104,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996645104690967,
      "loss": 6.6761,
      "mean_token_accuracy": 0.14659769386053084,
      "num_tokens": 6313447.0,
      "step": 2490
    },
    {
      "entropy": 6.69777102470398,
      "epoch": 0.2879399884593191,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996622541312103,
      "loss": 6.6714,
      "mean_token_accuracy": 0.1478801764547825,
      "num_tokens": 6326622.0,
      "step": 2495
    },
    {
      "entropy": 6.648268127441407,
      "epoch": 0.28851702250432776,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996599902369207,
      "loss": 6.6896,
      "mean_token_accuracy": 0.14757568538188934,
      "num_tokens": 6339286.0,
      "step": 2500
    },
    {
      "entropy": 6.6651811599731445,
      "epoch": 0.2890940565493364,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004996577187863039,
      "loss": 6.6847,
      "mean_token_accuracy": 0.14420827105641365,
      "num_tokens": 6351869.0,
      "step": 2505
    },
    {
      "entropy": 6.638708543777466,
      "epoch": 0.2896710905943451,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996554397794364,
      "loss": 6.6235,
      "mean_token_accuracy": 0.15295252203941345,
      "num_tokens": 6363974.0,
      "step": 2510
    },
    {
      "entropy": 6.6614098072052,
      "epoch": 0.29024812463935373,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996531532163947,
      "loss": 6.6232,
      "mean_token_accuracy": 0.15056563541293144,
      "num_tokens": 6376569.0,
      "step": 2515
    },
    {
      "entropy": 6.636733722686768,
      "epoch": 0.2908251586843624,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996508590972558,
      "loss": 6.6341,
      "mean_token_accuracy": 0.1568033814430237,
      "num_tokens": 6388830.0,
      "step": 2520
    },
    {
      "entropy": 6.627635335922241,
      "epoch": 0.291402192729371,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996485574220969,
      "loss": 6.6377,
      "mean_token_accuracy": 0.1470814898610115,
      "num_tokens": 6401542.0,
      "step": 2525
    },
    {
      "entropy": 6.600588321685791,
      "epoch": 0.2919792267743797,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996462481909953,
      "loss": 6.6167,
      "mean_token_accuracy": 0.15283410847187043,
      "num_tokens": 6414123.0,
      "step": 2530
    },
    {
      "entropy": 6.704501485824585,
      "epoch": 0.29255626081938835,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004996439314040287,
      "loss": 6.6938,
      "mean_token_accuracy": 0.14676552414894103,
      "num_tokens": 6427269.0,
      "step": 2535
    },
    {
      "entropy": 6.631581497192383,
      "epoch": 0.293133294864397,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499641607061275,
      "loss": 6.7408,
      "mean_token_accuracy": 0.14724025428295134,
      "num_tokens": 6442296.0,
      "step": 2540
    },
    {
      "entropy": 6.733328866958618,
      "epoch": 0.2937103289094057,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004996392751628126,
      "loss": 6.7476,
      "mean_token_accuracy": 0.14128965958952905,
      "num_tokens": 6456954.0,
      "step": 2545
    },
    {
      "entropy": 6.6841991424560545,
      "epoch": 0.2942873629544143,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004996369357087196,
      "loss": 6.7437,
      "mean_token_accuracy": 0.14420021548867226,
      "num_tokens": 6470139.0,
      "step": 2550
    },
    {
      "entropy": 6.810974264144898,
      "epoch": 0.29486439699942296,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499634588699075,
      "loss": 6.6532,
      "mean_token_accuracy": 0.15219533294439316,
      "num_tokens": 6483106.0,
      "step": 2555
    },
    {
      "entropy": 6.595229625701904,
      "epoch": 0.2954414310444316,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004996322341339575,
      "loss": 6.6908,
      "mean_token_accuracy": 0.1466153398156166,
      "num_tokens": 6495921.0,
      "step": 2560
    },
    {
      "entropy": 6.637534332275391,
      "epoch": 0.2960184650894403,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996298720134465,
      "loss": 6.5823,
      "mean_token_accuracy": 0.1488182932138443,
      "num_tokens": 6508408.0,
      "step": 2565
    },
    {
      "entropy": 6.580471229553223,
      "epoch": 0.29659549913444894,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004996275023376213,
      "loss": 6.6766,
      "mean_token_accuracy": 0.1433579817414284,
      "num_tokens": 6522081.0,
      "step": 2570
    },
    {
      "entropy": 6.743631982803345,
      "epoch": 0.2971725331794576,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996251251065615,
      "loss": 6.6524,
      "mean_token_accuracy": 0.14642552882432938,
      "num_tokens": 6535065.0,
      "step": 2575
    },
    {
      "entropy": 6.65220079421997,
      "epoch": 0.2977495672244662,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996227403203473,
      "loss": 6.6288,
      "mean_token_accuracy": 0.15091437101364136,
      "num_tokens": 6547972.0,
      "step": 2580
    },
    {
      "entropy": 6.619907283782959,
      "epoch": 0.2983266012694749,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004996203479790589,
      "loss": 6.5948,
      "mean_token_accuracy": 0.1509481832385063,
      "num_tokens": 6559935.0,
      "step": 2585
    },
    {
      "entropy": 6.593966054916382,
      "epoch": 0.29890363531448355,
      "grad_norm": 0.875,
      "learning_rate": 0.0004996179480827764,
      "loss": 6.6033,
      "mean_token_accuracy": 0.15155731663107871,
      "num_tokens": 6572373.0,
      "step": 2590
    },
    {
      "entropy": 6.682660484313965,
      "epoch": 0.2994806693594922,
      "grad_norm": 0.859375,
      "learning_rate": 0.000499615540631581,
      "loss": 6.677,
      "mean_token_accuracy": 0.14595670402050018,
      "num_tokens": 6585846.0,
      "step": 2595
    },
    {
      "entropy": 6.691784238815307,
      "epoch": 0.3000577034045009,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996131256255534,
      "loss": 6.7495,
      "mean_token_accuracy": 0.14251318722963333,
      "num_tokens": 6600387.0,
      "step": 2600
    },
    {
      "entropy": 6.625003719329834,
      "epoch": 0.3006347374495095,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004996107030647749,
      "loss": 6.667,
      "mean_token_accuracy": 0.14578246474266052,
      "num_tokens": 6613142.0,
      "step": 2605
    },
    {
      "entropy": 6.596926593780518,
      "epoch": 0.30121177149451817,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004996082729493269,
      "loss": 6.5739,
      "mean_token_accuracy": 0.15168050229549407,
      "num_tokens": 6626273.0,
      "step": 2610
    },
    {
      "entropy": 6.659254646301269,
      "epoch": 0.3017888055395268,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004996058352792913,
      "loss": 6.6539,
      "mean_token_accuracy": 0.14640988335013388,
      "num_tokens": 6639831.0,
      "step": 2615
    },
    {
      "entropy": 6.739427900314331,
      "epoch": 0.3023658395845355,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00049960339005475,
      "loss": 6.7259,
      "mean_token_accuracy": 0.1433930814266205,
      "num_tokens": 6652776.0,
      "step": 2620
    },
    {
      "entropy": 6.713594102859497,
      "epoch": 0.30294287362954414,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004996009372757852,
      "loss": 6.6827,
      "mean_token_accuracy": 0.14745238050818443,
      "num_tokens": 6665465.0,
      "step": 2625
    },
    {
      "entropy": 6.552315902709961,
      "epoch": 0.3035199076745528,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004995984769424795,
      "loss": 6.5667,
      "mean_token_accuracy": 0.1607323855161667,
      "num_tokens": 6677472.0,
      "step": 2630
    },
    {
      "entropy": 6.593226766586303,
      "epoch": 0.3040969417195615,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004995960090549155,
      "loss": 6.5878,
      "mean_token_accuracy": 0.1506764367222786,
      "num_tokens": 6689145.0,
      "step": 2635
    },
    {
      "entropy": 6.589013576507568,
      "epoch": 0.3046739757645701,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995935336131764,
      "loss": 6.5832,
      "mean_token_accuracy": 0.15121154189109803,
      "num_tokens": 6701520.0,
      "step": 2640
    },
    {
      "entropy": 6.666328239440918,
      "epoch": 0.30525100980957875,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004995910506173452,
      "loss": 6.6851,
      "mean_token_accuracy": 0.14975427985191345,
      "num_tokens": 6715615.0,
      "step": 2645
    },
    {
      "entropy": 6.676078271865845,
      "epoch": 0.3058280438545874,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004995885600675059,
      "loss": 6.6783,
      "mean_token_accuracy": 0.14618042930960656,
      "num_tokens": 6728772.0,
      "step": 2650
    },
    {
      "entropy": 6.670553731918335,
      "epoch": 0.3064050778995961,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995860619637415,
      "loss": 6.637,
      "mean_token_accuracy": 0.14901359081268312,
      "num_tokens": 6741454.0,
      "step": 2655
    },
    {
      "entropy": 6.599151611328125,
      "epoch": 0.30698211194460473,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004995835563061367,
      "loss": 6.6555,
      "mean_token_accuracy": 0.1510484904050827,
      "num_tokens": 6753983.0,
      "step": 2660
    },
    {
      "entropy": 6.74462776184082,
      "epoch": 0.30755914598961337,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004995810430947756,
      "loss": 6.699,
      "mean_token_accuracy": 0.1489512011408806,
      "num_tokens": 6767377.0,
      "step": 2665
    },
    {
      "entropy": 6.649488878250122,
      "epoch": 0.30813618003462206,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004995785223297426,
      "loss": 6.6467,
      "mean_token_accuracy": 0.15547565817832948,
      "num_tokens": 6779220.0,
      "step": 2670
    },
    {
      "entropy": 6.671740865707397,
      "epoch": 0.3087132140796307,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004995759940111225,
      "loss": 6.626,
      "mean_token_accuracy": 0.15176912397146225,
      "num_tokens": 6792202.0,
      "step": 2675
    },
    {
      "entropy": 6.538708877563477,
      "epoch": 0.30929024812463934,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995734581390005,
      "loss": 6.623,
      "mean_token_accuracy": 0.15197737216949464,
      "num_tokens": 6804988.0,
      "step": 2680
    },
    {
      "entropy": 6.68460521697998,
      "epoch": 0.309867282169648,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004995709147134617,
      "loss": 6.6537,
      "mean_token_accuracy": 0.1486986458301544,
      "num_tokens": 6817311.0,
      "step": 2685
    },
    {
      "entropy": 6.667502927780151,
      "epoch": 0.3104443162146567,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004995683637345919,
      "loss": 6.6367,
      "mean_token_accuracy": 0.14956765994429588,
      "num_tokens": 6829875.0,
      "step": 2690
    },
    {
      "entropy": 6.645153999328613,
      "epoch": 0.3110213502596653,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004995658052024765,
      "loss": 6.6112,
      "mean_token_accuracy": 0.155751770734787,
      "num_tokens": 6843080.0,
      "step": 2695
    },
    {
      "entropy": 6.594676733016968,
      "epoch": 0.31159838430467396,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004995632391172019,
      "loss": 6.5964,
      "mean_token_accuracy": 0.15243575125932693,
      "num_tokens": 6857202.0,
      "step": 2700
    },
    {
      "entropy": 6.682533931732178,
      "epoch": 0.31217541834968265,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004995606654788543,
      "loss": 6.7016,
      "mean_token_accuracy": 0.1450270637869835,
      "num_tokens": 6870801.0,
      "step": 2705
    },
    {
      "entropy": 6.654961776733399,
      "epoch": 0.3127524523946913,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004995580842875202,
      "loss": 6.6479,
      "mean_token_accuracy": 0.1479589194059372,
      "num_tokens": 6883084.0,
      "step": 2710
    },
    {
      "entropy": 6.708399629592895,
      "epoch": 0.31332948643969993,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995554955432866,
      "loss": 6.6756,
      "mean_token_accuracy": 0.1470663696527481,
      "num_tokens": 6895309.0,
      "step": 2715
    },
    {
      "entropy": 6.572336673736572,
      "epoch": 0.3139065204847086,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995528992462404,
      "loss": 6.6248,
      "mean_token_accuracy": 0.15860298946499823,
      "num_tokens": 6907646.0,
      "step": 2720
    },
    {
      "entropy": 6.6901154041290285,
      "epoch": 0.31448355452971727,
      "grad_norm": 1.0,
      "learning_rate": 0.000499550295396469,
      "loss": 6.6142,
      "mean_token_accuracy": 0.15778794139623642,
      "num_tokens": 6919769.0,
      "step": 2725
    },
    {
      "entropy": 6.4766926765441895,
      "epoch": 0.3150605885747259,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004995476839940598,
      "loss": 6.6201,
      "mean_token_accuracy": 0.15026881247758866,
      "num_tokens": 6933348.0,
      "step": 2730
    },
    {
      "entropy": 6.714725732803345,
      "epoch": 0.31563762261973455,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499545065039101,
      "loss": 6.6743,
      "mean_token_accuracy": 0.14582089185714722,
      "num_tokens": 6945105.0,
      "step": 2735
    },
    {
      "entropy": 6.545088768005371,
      "epoch": 0.31621465666474324,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004995424385316803,
      "loss": 6.6618,
      "mean_token_accuracy": 0.1445609971880913,
      "num_tokens": 6958009.0,
      "step": 2740
    },
    {
      "entropy": 6.6502604484558105,
      "epoch": 0.3167916907097519,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995398044718863,
      "loss": 6.5485,
      "mean_token_accuracy": 0.15240858346223832,
      "num_tokens": 6970156.0,
      "step": 2745
    },
    {
      "entropy": 6.583297348022461,
      "epoch": 0.3173687247547605,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004995371628598074,
      "loss": 6.6473,
      "mean_token_accuracy": 0.151313416659832,
      "num_tokens": 6983431.0,
      "step": 2750
    },
    {
      "entropy": 6.705528640747071,
      "epoch": 0.31794575879976916,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004995345136955328,
      "loss": 6.6931,
      "mean_token_accuracy": 0.14359337761998175,
      "num_tokens": 6997470.0,
      "step": 2755
    },
    {
      "entropy": 6.686682987213135,
      "epoch": 0.31852279284477786,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004995318569791512,
      "loss": 6.6505,
      "mean_token_accuracy": 0.15525564998388292,
      "num_tokens": 7010337.0,
      "step": 2760
    },
    {
      "entropy": 6.64365782737732,
      "epoch": 0.3190998268897865,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004995291927107522,
      "loss": 6.6628,
      "mean_token_accuracy": 0.14557797089219093,
      "num_tokens": 7023950.0,
      "step": 2765
    },
    {
      "entropy": 6.638036775588989,
      "epoch": 0.31967686093479514,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995265208904252,
      "loss": 6.622,
      "mean_token_accuracy": 0.15690350979566575,
      "num_tokens": 7036409.0,
      "step": 2770
    },
    {
      "entropy": 6.616341543197632,
      "epoch": 0.32025389497980383,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004995238415182603,
      "loss": 6.6326,
      "mean_token_accuracy": 0.14718683063983917,
      "num_tokens": 7049409.0,
      "step": 2775
    },
    {
      "entropy": 6.688148880004883,
      "epoch": 0.3208309290248125,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004995211545943476,
      "loss": 6.6907,
      "mean_token_accuracy": 0.14982038736343384,
      "num_tokens": 7062120.0,
      "step": 2780
    },
    {
      "entropy": 6.680778884887696,
      "epoch": 0.3214079630698211,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995184601187772,
      "loss": 6.668,
      "mean_token_accuracy": 0.14194481670856476,
      "num_tokens": 7073150.0,
      "step": 2785
    },
    {
      "entropy": 6.626657867431641,
      "epoch": 0.32198499711482975,
      "grad_norm": 0.953125,
      "learning_rate": 0.00049951575809164,
      "loss": 6.6001,
      "mean_token_accuracy": 0.142817410081625,
      "num_tokens": 7085927.0,
      "step": 2790
    },
    {
      "entropy": 6.64122838973999,
      "epoch": 0.32256203115983845,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995130485130269,
      "loss": 6.6837,
      "mean_token_accuracy": 0.14988088607788086,
      "num_tokens": 7097607.0,
      "step": 2795
    },
    {
      "entropy": 6.641415691375732,
      "epoch": 0.3231390652048471,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995103313830289,
      "loss": 6.5998,
      "mean_token_accuracy": 0.15160279124975204,
      "num_tokens": 7109771.0,
      "step": 2800
    },
    {
      "entropy": 6.694540548324585,
      "epoch": 0.3237160992498557,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004995076067017373,
      "loss": 6.6737,
      "mean_token_accuracy": 0.14511681348085403,
      "num_tokens": 7122757.0,
      "step": 2805
    },
    {
      "entropy": 6.624036026000977,
      "epoch": 0.3242931332948644,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004995048744692439,
      "loss": 6.6436,
      "mean_token_accuracy": 0.1493765577673912,
      "num_tokens": 7135516.0,
      "step": 2810
    },
    {
      "entropy": 6.639794206619262,
      "epoch": 0.32487016733987306,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995021346856405,
      "loss": 6.7264,
      "mean_token_accuracy": 0.14374846667051316,
      "num_tokens": 7149125.0,
      "step": 2815
    },
    {
      "entropy": 6.706334114074707,
      "epoch": 0.3254472013848817,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994993873510196,
      "loss": 6.6653,
      "mean_token_accuracy": 0.14253946542739868,
      "num_tokens": 7160723.0,
      "step": 2820
    },
    {
      "entropy": 6.645235395431518,
      "epoch": 0.32602423542989034,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004994966324654731,
      "loss": 6.6433,
      "mean_token_accuracy": 0.144842017441988,
      "num_tokens": 7174551.0,
      "step": 2825
    },
    {
      "entropy": 6.650976371765137,
      "epoch": 0.32660126947489904,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499493870029094,
      "loss": 6.6189,
      "mean_token_accuracy": 0.14779492169618608,
      "num_tokens": 7186764.0,
      "step": 2830
    },
    {
      "entropy": 6.611865520477295,
      "epoch": 0.3271783035199077,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004994911000419749,
      "loss": 6.6143,
      "mean_token_accuracy": 0.15066873282194138,
      "num_tokens": 7199645.0,
      "step": 2835
    },
    {
      "entropy": 6.6020160675048825,
      "epoch": 0.3277553375649163,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994883225042093,
      "loss": 6.54,
      "mean_token_accuracy": 0.1572998732328415,
      "num_tokens": 7212579.0,
      "step": 2840
    },
    {
      "entropy": 6.586337566375732,
      "epoch": 0.328332371609925,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994855374158905,
      "loss": 6.571,
      "mean_token_accuracy": 0.15372254252433776,
      "num_tokens": 7225139.0,
      "step": 2845
    },
    {
      "entropy": 6.710922813415527,
      "epoch": 0.32890940565493365,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499482744777112,
      "loss": 6.7311,
      "mean_token_accuracy": 0.14069035798311233,
      "num_tokens": 7239220.0,
      "step": 2850
    },
    {
      "entropy": 6.675474834442139,
      "epoch": 0.3294864396999423,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000499479944587968,
      "loss": 6.7533,
      "mean_token_accuracy": 0.1474220335483551,
      "num_tokens": 7252613.0,
      "step": 2855
    },
    {
      "entropy": 6.580682802200317,
      "epoch": 0.33006347374495093,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004994771368485526,
      "loss": 6.5772,
      "mean_token_accuracy": 0.15456188917160035,
      "num_tokens": 7263658.0,
      "step": 2860
    },
    {
      "entropy": 6.643721866607666,
      "epoch": 0.3306405077899596,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004994743215589602,
      "loss": 6.6064,
      "mean_token_accuracy": 0.1567901760339737,
      "num_tokens": 7275782.0,
      "step": 2865
    },
    {
      "entropy": 6.544418811798096,
      "epoch": 0.33121754183496827,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004994714987192854,
      "loss": 6.6359,
      "mean_token_accuracy": 0.15609208047389983,
      "num_tokens": 7288536.0,
      "step": 2870
    },
    {
      "entropy": 6.609630966186524,
      "epoch": 0.3317945758799769,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004994686683296234,
      "loss": 6.6019,
      "mean_token_accuracy": 0.15594898611307145,
      "num_tokens": 7300480.0,
      "step": 2875
    },
    {
      "entropy": 6.634557008743286,
      "epoch": 0.3323716099249856,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004994658303900691,
      "loss": 6.6475,
      "mean_token_accuracy": 0.15181711912155152,
      "num_tokens": 7313168.0,
      "step": 2880
    },
    {
      "entropy": 6.5809680938720705,
      "epoch": 0.33294864396999424,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004994629849007182,
      "loss": 6.6147,
      "mean_token_accuracy": 0.15636203438043594,
      "num_tokens": 7327347.0,
      "step": 2885
    },
    {
      "entropy": 6.733543872833252,
      "epoch": 0.3335256780150029,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004994601318616663,
      "loss": 6.6466,
      "mean_token_accuracy": 0.1510158471763134,
      "num_tokens": 7339955.0,
      "step": 2890
    },
    {
      "entropy": 6.584178400039673,
      "epoch": 0.3341027120600115,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994572712730092,
      "loss": 6.6188,
      "mean_token_accuracy": 0.14996251463890076,
      "num_tokens": 7353586.0,
      "step": 2895
    },
    {
      "entropy": 6.593496084213257,
      "epoch": 0.3346797461050202,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994544031348434,
      "loss": 6.6033,
      "mean_token_accuracy": 0.14958511143922806,
      "num_tokens": 7365802.0,
      "step": 2900
    },
    {
      "entropy": 6.634331750869751,
      "epoch": 0.33525678015002885,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994515274472653,
      "loss": 6.5569,
      "mean_token_accuracy": 0.14671666771173478,
      "num_tokens": 7377864.0,
      "step": 2905
    },
    {
      "entropy": 6.52061505317688,
      "epoch": 0.3358338141950375,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004994486442103715,
      "loss": 6.5051,
      "mean_token_accuracy": 0.1553037017583847,
      "num_tokens": 7389921.0,
      "step": 2910
    },
    {
      "entropy": 6.606663084030151,
      "epoch": 0.3364108482400462,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499445753424259,
      "loss": 6.585,
      "mean_token_accuracy": 0.14963840320706367,
      "num_tokens": 7402765.0,
      "step": 2915
    },
    {
      "entropy": 6.612658452987671,
      "epoch": 0.33698788228505483,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004994428550890251,
      "loss": 6.588,
      "mean_token_accuracy": 0.14756084382534027,
      "num_tokens": 7416330.0,
      "step": 2920
    },
    {
      "entropy": 6.546629619598389,
      "epoch": 0.33756491633006347,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004994399492047672,
      "loss": 6.5569,
      "mean_token_accuracy": 0.1625090643763542,
      "num_tokens": 7428701.0,
      "step": 2925
    },
    {
      "entropy": 6.616139030456543,
      "epoch": 0.3381419503750721,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004994370357715832,
      "loss": 6.5609,
      "mean_token_accuracy": 0.1539958417415619,
      "num_tokens": 7440687.0,
      "step": 2930
    },
    {
      "entropy": 6.554273843765259,
      "epoch": 0.3387189844200808,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499434114789571,
      "loss": 6.65,
      "mean_token_accuracy": 0.14359127953648568,
      "num_tokens": 7454448.0,
      "step": 2935
    },
    {
      "entropy": 6.695260524749756,
      "epoch": 0.33929601846508944,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004994311862588286,
      "loss": 6.5699,
      "mean_token_accuracy": 0.15439818054437637,
      "num_tokens": 7466722.0,
      "step": 2940
    },
    {
      "entropy": 6.530065727233887,
      "epoch": 0.3398730525100981,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994282501794551,
      "loss": 6.5843,
      "mean_token_accuracy": 0.15763369649648667,
      "num_tokens": 7478574.0,
      "step": 2945
    },
    {
      "entropy": 6.518723726272583,
      "epoch": 0.3404500865551068,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004994253065515486,
      "loss": 6.4807,
      "mean_token_accuracy": 0.15764836221933365,
      "num_tokens": 7490424.0,
      "step": 2950
    },
    {
      "entropy": 6.633134555816651,
      "epoch": 0.3410271206001154,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994223553752084,
      "loss": 6.6256,
      "mean_token_accuracy": 0.15402503609657286,
      "num_tokens": 7502832.0,
      "step": 2955
    },
    {
      "entropy": 6.511202335357666,
      "epoch": 0.34160415464512406,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004994193966505339,
      "loss": 6.5529,
      "mean_token_accuracy": 0.15395866483449935,
      "num_tokens": 7514151.0,
      "step": 2960
    },
    {
      "entropy": 6.612486219406128,
      "epoch": 0.3421811886901327,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004994164303776244,
      "loss": 6.5929,
      "mean_token_accuracy": 0.1546282023191452,
      "num_tokens": 7526877.0,
      "step": 2965
    },
    {
      "entropy": 6.601590824127197,
      "epoch": 0.3427582227351414,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994134565565797,
      "loss": 6.556,
      "mean_token_accuracy": 0.15401398837566377,
      "num_tokens": 7538858.0,
      "step": 2970
    },
    {
      "entropy": 6.674258995056152,
      "epoch": 0.34333525678015003,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004994104751875,
      "loss": 6.5985,
      "mean_token_accuracy": 0.15456424802541732,
      "num_tokens": 7550804.0,
      "step": 2975
    },
    {
      "entropy": 6.490966129302978,
      "epoch": 0.3439122908251587,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994074862704854,
      "loss": 6.5304,
      "mean_token_accuracy": 0.15696495920419692,
      "num_tokens": 7562269.0,
      "step": 2980
    },
    {
      "entropy": 6.60324239730835,
      "epoch": 0.3444893248701673,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994044898056363,
      "loss": 6.6167,
      "mean_token_accuracy": 0.14663480520248412,
      "num_tokens": 7573695.0,
      "step": 2985
    },
    {
      "entropy": 6.655906391143799,
      "epoch": 0.345066358915176,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004994014857930538,
      "loss": 6.6318,
      "mean_token_accuracy": 0.15162624269723893,
      "num_tokens": 7585943.0,
      "step": 2990
    },
    {
      "entropy": 6.56010856628418,
      "epoch": 0.34564339296018465,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004993984742328387,
      "loss": 6.6145,
      "mean_token_accuracy": 0.1463395446538925,
      "num_tokens": 7598326.0,
      "step": 2995
    },
    {
      "entropy": 6.628453969955444,
      "epoch": 0.3462204270051933,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004993954551250925,
      "loss": 6.6109,
      "mean_token_accuracy": 0.14573243409395217,
      "num_tokens": 7609950.0,
      "step": 3000
    },
    {
      "epoch": 0.3462204270051933,
      "eval_entropy": 6.495445118866964,
      "eval_loss": 6.647010326385498,
      "eval_mean_token_accuracy": 0.15230962459314673,
      "eval_num_tokens": 7609950.0,
      "eval_runtime": 22.9056,
      "eval_samples_per_second": 1228.347,
      "eval_steps_per_second": 153.543,
      "step": 3000
    },
    {
      "entropy": 6.591029357910156,
      "epoch": 0.346797461050202,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004993924284699165,
      "loss": 6.5722,
      "mean_token_accuracy": 0.15271425247192383,
      "num_tokens": 7621987.0,
      "step": 3005
    },
    {
      "entropy": 6.588593149185181,
      "epoch": 0.3473744950952106,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004993893942674127,
      "loss": 6.6327,
      "mean_token_accuracy": 0.14458775967359544,
      "num_tokens": 7634801.0,
      "step": 3010
    },
    {
      "entropy": 6.676081657409668,
      "epoch": 0.34795152914021926,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499386352517683,
      "loss": 6.5669,
      "mean_token_accuracy": 0.1588612362742424,
      "num_tokens": 7647942.0,
      "step": 3015
    },
    {
      "entropy": 6.5854696273803714,
      "epoch": 0.3485285631852279,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00049938330322083,
      "loss": 6.5803,
      "mean_token_accuracy": 0.15369187146425248,
      "num_tokens": 7662361.0,
      "step": 3020
    },
    {
      "entropy": 6.562425899505615,
      "epoch": 0.3491055972302366,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993802463769558,
      "loss": 6.5929,
      "mean_token_accuracy": 0.14925703704357146,
      "num_tokens": 7674161.0,
      "step": 3025
    },
    {
      "entropy": 6.620813989639283,
      "epoch": 0.34968263127524524,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993771819861636,
      "loss": 6.5787,
      "mean_token_accuracy": 0.1493277445435524,
      "num_tokens": 7686957.0,
      "step": 3030
    },
    {
      "entropy": 6.521798992156983,
      "epoch": 0.3502596653202539,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004993741100485564,
      "loss": 6.5618,
      "mean_token_accuracy": 0.15716940760612488,
      "num_tokens": 7699436.0,
      "step": 3035
    },
    {
      "entropy": 6.517621612548828,
      "epoch": 0.35083669936526257,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004993710305642374,
      "loss": 6.5573,
      "mean_token_accuracy": 0.15328558534383774,
      "num_tokens": 7711794.0,
      "step": 3040
    },
    {
      "entropy": 6.611809253692627,
      "epoch": 0.3514137334102712,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004993679435333102,
      "loss": 6.5114,
      "mean_token_accuracy": 0.15407811254262924,
      "num_tokens": 7723496.0,
      "step": 3045
    },
    {
      "entropy": 6.512660932540894,
      "epoch": 0.35199076745527985,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004993648489558788,
      "loss": 6.543,
      "mean_token_accuracy": 0.1530572846531868,
      "num_tokens": 7735980.0,
      "step": 3050
    },
    {
      "entropy": 6.637914991378784,
      "epoch": 0.3525678015002885,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004993617468320471,
      "loss": 6.6016,
      "mean_token_accuracy": 0.15019816905260086,
      "num_tokens": 7750145.0,
      "step": 3055
    },
    {
      "entropy": 6.551783275604248,
      "epoch": 0.3531448355452972,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004993586371619193,
      "loss": 6.6024,
      "mean_token_accuracy": 0.15270411223173141,
      "num_tokens": 7762284.0,
      "step": 3060
    },
    {
      "entropy": 6.54249119758606,
      "epoch": 0.3537218695903058,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004993555199456004,
      "loss": 6.5582,
      "mean_token_accuracy": 0.16024676263332366,
      "num_tokens": 7776225.0,
      "step": 3065
    },
    {
      "entropy": 6.536019992828369,
      "epoch": 0.35429890363531447,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004993523951831949,
      "loss": 6.4886,
      "mean_token_accuracy": 0.1622866541147232,
      "num_tokens": 7789408.0,
      "step": 3070
    },
    {
      "entropy": 6.610622406005859,
      "epoch": 0.35487593768032316,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993492628748081,
      "loss": 6.5633,
      "mean_token_accuracy": 0.14859718829393387,
      "num_tokens": 7801824.0,
      "step": 3075
    },
    {
      "entropy": 6.5355761528015135,
      "epoch": 0.3554529717253318,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004993461230205453,
      "loss": 6.5596,
      "mean_token_accuracy": 0.1570269212126732,
      "num_tokens": 7814167.0,
      "step": 3080
    },
    {
      "entropy": 6.599206972122192,
      "epoch": 0.35603000577034044,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004993429756205121,
      "loss": 6.5808,
      "mean_token_accuracy": 0.15419002175331115,
      "num_tokens": 7826736.0,
      "step": 3085
    },
    {
      "entropy": 6.536418008804321,
      "epoch": 0.3566070398153491,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004993398206748142,
      "loss": 6.5655,
      "mean_token_accuracy": 0.15275473296642303,
      "num_tokens": 7838626.0,
      "step": 3090
    },
    {
      "entropy": 6.610546350479126,
      "epoch": 0.3571840738603578,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004993366581835581,
      "loss": 6.5528,
      "mean_token_accuracy": 0.15237838327884673,
      "num_tokens": 7850608.0,
      "step": 3095
    },
    {
      "entropy": 6.549269485473633,
      "epoch": 0.3577611079053664,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004993334881468498,
      "loss": 6.5346,
      "mean_token_accuracy": 0.15190107375383377,
      "num_tokens": 7862800.0,
      "step": 3100
    },
    {
      "entropy": 6.628948211669922,
      "epoch": 0.35833814195037506,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004993303105647961,
      "loss": 6.612,
      "mean_token_accuracy": 0.15180804431438447,
      "num_tokens": 7876198.0,
      "step": 3105
    },
    {
      "entropy": 6.586100435256958,
      "epoch": 0.35891517599538375,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004993271254375038,
      "loss": 6.6374,
      "mean_token_accuracy": 0.14822730123996736,
      "num_tokens": 7888129.0,
      "step": 3110
    },
    {
      "entropy": 6.564333963394165,
      "epoch": 0.3594922100403924,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00049932393276508,
      "loss": 6.5923,
      "mean_token_accuracy": 0.14657284393906594,
      "num_tokens": 7900453.0,
      "step": 3115
    },
    {
      "entropy": 6.639041471481323,
      "epoch": 0.36006924408540103,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004993207325476323,
      "loss": 6.6062,
      "mean_token_accuracy": 0.1456374153494835,
      "num_tokens": 7913538.0,
      "step": 3120
    },
    {
      "entropy": 6.564752388000488,
      "epoch": 0.36064627813040967,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993175247852681,
      "loss": 6.5663,
      "mean_token_accuracy": 0.15041694641113282,
      "num_tokens": 7927744.0,
      "step": 3125
    },
    {
      "entropy": 6.5582201957702635,
      "epoch": 0.36122331217541837,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004993143094780954,
      "loss": 6.5274,
      "mean_token_accuracy": 0.15242796540260314,
      "num_tokens": 7939975.0,
      "step": 3130
    },
    {
      "entropy": 6.606953287124634,
      "epoch": 0.361800346220427,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004993110866262224,
      "loss": 6.567,
      "mean_token_accuracy": 0.1520855814218521,
      "num_tokens": 7951692.0,
      "step": 3135
    },
    {
      "entropy": 6.509029150009155,
      "epoch": 0.36237738026543564,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993078562297573,
      "loss": 6.5709,
      "mean_token_accuracy": 0.15134818702936173,
      "num_tokens": 7964887.0,
      "step": 3140
    },
    {
      "entropy": 6.622086668014527,
      "epoch": 0.36295441431044434,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004993046182888089,
      "loss": 6.5962,
      "mean_token_accuracy": 0.1504879489541054,
      "num_tokens": 7977608.0,
      "step": 3145
    },
    {
      "entropy": 6.4911338806152346,
      "epoch": 0.363531448355453,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004993013728034861,
      "loss": 6.506,
      "mean_token_accuracy": 0.15199390202760696,
      "num_tokens": 7990094.0,
      "step": 3150
    },
    {
      "entropy": 6.63452377319336,
      "epoch": 0.3641084824004616,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499298119773898,
      "loss": 6.5799,
      "mean_token_accuracy": 0.1515541434288025,
      "num_tokens": 8003477.0,
      "step": 3155
    },
    {
      "entropy": 6.549254274368286,
      "epoch": 0.36468551644547026,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992948592001541,
      "loss": 6.5276,
      "mean_token_accuracy": 0.15579250305891038,
      "num_tokens": 8016241.0,
      "step": 3160
    },
    {
      "entropy": 6.465305948257447,
      "epoch": 0.36526255049047895,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499291591082364,
      "loss": 6.4944,
      "mean_token_accuracy": 0.16084357649087905,
      "num_tokens": 8028352.0,
      "step": 3165
    },
    {
      "entropy": 6.608652591705322,
      "epoch": 0.3658395845354876,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004992883154206377,
      "loss": 6.5741,
      "mean_token_accuracy": 0.15175243765115737,
      "num_tokens": 8040215.0,
      "step": 3170
    },
    {
      "entropy": 6.542947626113891,
      "epoch": 0.36641661858049623,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004992850322150853,
      "loss": 6.5156,
      "mean_token_accuracy": 0.15719302594661713,
      "num_tokens": 8051633.0,
      "step": 3175
    },
    {
      "entropy": 6.570356178283691,
      "epoch": 0.36699365262550493,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004992817414658172,
      "loss": 6.581,
      "mean_token_accuracy": 0.1489927276968956,
      "num_tokens": 8063390.0,
      "step": 3180
    },
    {
      "entropy": 6.5481853008270265,
      "epoch": 0.36757068667051357,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004992784431729442,
      "loss": 6.5419,
      "mean_token_accuracy": 0.15093412175774573,
      "num_tokens": 8077561.0,
      "step": 3185
    },
    {
      "entropy": 6.542624521255493,
      "epoch": 0.3681477207155222,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004992751373365771,
      "loss": 6.6039,
      "mean_token_accuracy": 0.15380253195762633,
      "num_tokens": 8091345.0,
      "step": 3190
    },
    {
      "entropy": 6.52802619934082,
      "epoch": 0.36872475476053085,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004992718239568273,
      "loss": 6.4689,
      "mean_token_accuracy": 0.15757767856121063,
      "num_tokens": 8104179.0,
      "step": 3195
    },
    {
      "entropy": 6.5984804153442385,
      "epoch": 0.36930178880553954,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499268503033806,
      "loss": 6.606,
      "mean_token_accuracy": 0.14631188064813613,
      "num_tokens": 8115831.0,
      "step": 3200
    },
    {
      "entropy": 6.552531099319458,
      "epoch": 0.3698788228505482,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004992651745676249,
      "loss": 6.5472,
      "mean_token_accuracy": 0.15520148873329162,
      "num_tokens": 8129048.0,
      "step": 3205
    },
    {
      "entropy": 6.570079278945923,
      "epoch": 0.3704558568955568,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004992618385583962,
      "loss": 6.5624,
      "mean_token_accuracy": 0.14829408824443818,
      "num_tokens": 8141679.0,
      "step": 3210
    },
    {
      "entropy": 6.546512317657471,
      "epoch": 0.3710328909405655,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992584950062319,
      "loss": 6.5338,
      "mean_token_accuracy": 0.15072018280625343,
      "num_tokens": 8155004.0,
      "step": 3215
    },
    {
      "entropy": 6.447714996337891,
      "epoch": 0.37160992498557416,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004992551439112446,
      "loss": 6.4848,
      "mean_token_accuracy": 0.1548818290233612,
      "num_tokens": 8166718.0,
      "step": 3220
    },
    {
      "entropy": 6.644821548461914,
      "epoch": 0.3721869590305828,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992517852735469,
      "loss": 6.6559,
      "mean_token_accuracy": 0.1469529762864113,
      "num_tokens": 8180011.0,
      "step": 3225
    },
    {
      "entropy": 6.549525928497315,
      "epoch": 0.37276399307559144,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004992484190932517,
      "loss": 6.4946,
      "mean_token_accuracy": 0.15516663938760758,
      "num_tokens": 8194800.0,
      "step": 3230
    },
    {
      "entropy": 6.580848121643067,
      "epoch": 0.37334102712060013,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004992450453704723,
      "loss": 6.5921,
      "mean_token_accuracy": 0.15195617005228995,
      "num_tokens": 8207153.0,
      "step": 3235
    },
    {
      "entropy": 6.61992621421814,
      "epoch": 0.3739180611656088,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004992416641053223,
      "loss": 6.5844,
      "mean_token_accuracy": 0.14998508989810944,
      "num_tokens": 8221135.0,
      "step": 3240
    },
    {
      "entropy": 6.415407657623291,
      "epoch": 0.3744950952106174,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004992382752979153,
      "loss": 6.456,
      "mean_token_accuracy": 0.1594781830906868,
      "num_tokens": 8234271.0,
      "step": 3245
    },
    {
      "entropy": 6.60991153717041,
      "epoch": 0.3750721292556261,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004992348789483651,
      "loss": 6.6003,
      "mean_token_accuracy": 0.147911337018013,
      "num_tokens": 8246710.0,
      "step": 3250
    },
    {
      "entropy": 6.593240404129029,
      "epoch": 0.37564916330063475,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004992314750567864,
      "loss": 6.6365,
      "mean_token_accuracy": 0.15298002660274507,
      "num_tokens": 8259817.0,
      "step": 3255
    },
    {
      "entropy": 6.550424814224243,
      "epoch": 0.3762261973456434,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004992280636232933,
      "loss": 6.5679,
      "mean_token_accuracy": 0.16052095741033554,
      "num_tokens": 8274359.0,
      "step": 3260
    },
    {
      "entropy": 6.579660654067993,
      "epoch": 0.376803231390652,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004992246446480007,
      "loss": 6.5213,
      "mean_token_accuracy": 0.15295627862215042,
      "num_tokens": 8286355.0,
      "step": 3265
    },
    {
      "entropy": 6.581032180786133,
      "epoch": 0.3773802654356607,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004992212181310235,
      "loss": 6.5394,
      "mean_token_accuracy": 0.14936599284410476,
      "num_tokens": 8298338.0,
      "step": 3270
    },
    {
      "entropy": 6.411511325836182,
      "epoch": 0.37795729948066936,
      "grad_norm": 1.0,
      "learning_rate": 0.000499217784072477,
      "loss": 6.5144,
      "mean_token_accuracy": 0.15907332748174668,
      "num_tokens": 8312367.0,
      "step": 3275
    },
    {
      "entropy": 6.608388662338257,
      "epoch": 0.378534333525678,
      "grad_norm": 0.875,
      "learning_rate": 0.0004992143424724768,
      "loss": 6.6026,
      "mean_token_accuracy": 0.1458575025200844,
      "num_tokens": 8325577.0,
      "step": 3280
    },
    {
      "entropy": 6.583384847640991,
      "epoch": 0.3791113675706867,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004992108933311385,
      "loss": 6.6091,
      "mean_token_accuracy": 0.1506508022546768,
      "num_tokens": 8338965.0,
      "step": 3285
    },
    {
      "entropy": 6.676050329208374,
      "epoch": 0.37968840161569534,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004992074366485782,
      "loss": 6.5431,
      "mean_token_accuracy": 0.14468644559383392,
      "num_tokens": 8351669.0,
      "step": 3290
    },
    {
      "entropy": 6.50412654876709,
      "epoch": 0.380265435660704,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499203972424912,
      "loss": 6.6094,
      "mean_token_accuracy": 0.1484062507748604,
      "num_tokens": 8364194.0,
      "step": 3295
    },
    {
      "entropy": 6.55182638168335,
      "epoch": 0.3808424697057126,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004992005006602567,
      "loss": 6.6111,
      "mean_token_accuracy": 0.1497878022491932,
      "num_tokens": 8377001.0,
      "step": 3300
    },
    {
      "entropy": 6.5773975372314455,
      "epoch": 0.3814195037507213,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004991970213547289,
      "loss": 6.6088,
      "mean_token_accuracy": 0.14910247027873993,
      "num_tokens": 8389247.0,
      "step": 3305
    },
    {
      "entropy": 6.583339834213257,
      "epoch": 0.38199653779572995,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004991935345084457,
      "loss": 6.4987,
      "mean_token_accuracy": 0.15111185982823372,
      "num_tokens": 8402915.0,
      "step": 3310
    },
    {
      "entropy": 6.585533714294433,
      "epoch": 0.3825735718407386,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004991900401215243,
      "loss": 6.6005,
      "mean_token_accuracy": 0.15311083793640137,
      "num_tokens": 8414590.0,
      "step": 3315
    },
    {
      "entropy": 6.6513701438903805,
      "epoch": 0.3831506058857473,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004991865381940822,
      "loss": 6.584,
      "mean_token_accuracy": 0.15581532269716264,
      "num_tokens": 8427337.0,
      "step": 3320
    },
    {
      "entropy": 6.428265810012817,
      "epoch": 0.3837276399307559,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004991830287262374,
      "loss": 6.4612,
      "mean_token_accuracy": 0.1581800550222397,
      "num_tokens": 8440095.0,
      "step": 3325
    },
    {
      "entropy": 6.53554334640503,
      "epoch": 0.38430467397576457,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004991795117181077,
      "loss": 6.477,
      "mean_token_accuracy": 0.15249311774969102,
      "num_tokens": 8452448.0,
      "step": 3330
    },
    {
      "entropy": 6.591722297668457,
      "epoch": 0.3848817080207732,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004991759871698113,
      "loss": 6.5265,
      "mean_token_accuracy": 0.15830566138029098,
      "num_tokens": 8465289.0,
      "step": 3335
    },
    {
      "entropy": 6.497562980651855,
      "epoch": 0.3854587420657819,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004991724550814671,
      "loss": 6.5512,
      "mean_token_accuracy": 0.15357236713171005,
      "num_tokens": 8477365.0,
      "step": 3340
    },
    {
      "entropy": 6.607602119445801,
      "epoch": 0.38603577611079054,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004991689154531937,
      "loss": 6.6112,
      "mean_token_accuracy": 0.15152789950370787,
      "num_tokens": 8489104.0,
      "step": 3345
    },
    {
      "entropy": 6.519052219390869,
      "epoch": 0.3866128101557992,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004991653682851102,
      "loss": 6.5022,
      "mean_token_accuracy": 0.1561011031270027,
      "num_tokens": 8501509.0,
      "step": 3350
    },
    {
      "entropy": 6.54585828781128,
      "epoch": 0.3871898442008078,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499161813577336,
      "loss": 6.536,
      "mean_token_accuracy": 0.15066452771425248,
      "num_tokens": 8514224.0,
      "step": 3355
    },
    {
      "entropy": 6.612550258636475,
      "epoch": 0.3877668782458165,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004991582513299903,
      "loss": 6.6069,
      "mean_token_accuracy": 0.1464258462190628,
      "num_tokens": 8526689.0,
      "step": 3360
    },
    {
      "entropy": 6.511368894577027,
      "epoch": 0.38834391229082516,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991546815431932,
      "loss": 6.6481,
      "mean_token_accuracy": 0.14692416340112685,
      "num_tokens": 8539952.0,
      "step": 3365
    },
    {
      "entropy": 6.696336126327514,
      "epoch": 0.3889209463358338,
      "grad_norm": 0.890625,
      "learning_rate": 0.000499151104217065,
      "loss": 6.6221,
      "mean_token_accuracy": 0.14956214874982834,
      "num_tokens": 8553428.0,
      "step": 3370
    },
    {
      "entropy": 6.528138971328735,
      "epoch": 0.3894979803808425,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991475193517254,
      "loss": 6.4912,
      "mean_token_accuracy": 0.1579304374754429,
      "num_tokens": 8566066.0,
      "step": 3375
    },
    {
      "entropy": 6.488394165039063,
      "epoch": 0.39007501442585113,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004991439269472956,
      "loss": 6.4463,
      "mean_token_accuracy": 0.15376436561346055,
      "num_tokens": 8579211.0,
      "step": 3380
    },
    {
      "entropy": 6.568374729156494,
      "epoch": 0.39065204847085977,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004991403270038961,
      "loss": 6.5617,
      "mean_token_accuracy": 0.14594289362430574,
      "num_tokens": 8591647.0,
      "step": 3385
    },
    {
      "entropy": 6.454288911819458,
      "epoch": 0.3912290825158684,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499136719521648,
      "loss": 6.5486,
      "mean_token_accuracy": 0.14656399562954903,
      "num_tokens": 8604279.0,
      "step": 3390
    },
    {
      "entropy": 6.582765245437622,
      "epoch": 0.3918061165608771,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004991331045006726,
      "loss": 6.4704,
      "mean_token_accuracy": 0.15898291915655136,
      "num_tokens": 8617722.0,
      "step": 3395
    },
    {
      "entropy": 6.489694404602051,
      "epoch": 0.39238315060588574,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004991294819410916,
      "loss": 6.4867,
      "mean_token_accuracy": 0.1653321497142315,
      "num_tokens": 8630250.0,
      "step": 3400
    },
    {
      "entropy": 6.405242109298706,
      "epoch": 0.3929601846508944,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004991258518430268,
      "loss": 6.3755,
      "mean_token_accuracy": 0.15958154797554017,
      "num_tokens": 8641900.0,
      "step": 3405
    },
    {
      "entropy": 6.507018613815307,
      "epoch": 0.3935372186959031,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004991222142066003,
      "loss": 6.5126,
      "mean_token_accuracy": 0.15494983345270158,
      "num_tokens": 8654588.0,
      "step": 3410
    },
    {
      "entropy": 6.548699235916137,
      "epoch": 0.3941142527409117,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991185690319345,
      "loss": 6.5202,
      "mean_token_accuracy": 0.14699627980589866,
      "num_tokens": 8668111.0,
      "step": 3415
    },
    {
      "entropy": 6.519062423706055,
      "epoch": 0.39469128678592036,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000499114916319152,
      "loss": 6.5989,
      "mean_token_accuracy": 0.15181139409542083,
      "num_tokens": 8681461.0,
      "step": 3420
    },
    {
      "entropy": 6.594384002685547,
      "epoch": 0.395268320830929,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004991112560683755,
      "loss": 6.5317,
      "mean_token_accuracy": 0.15306874513626098,
      "num_tokens": 8692893.0,
      "step": 3425
    },
    {
      "entropy": 6.477481889724731,
      "epoch": 0.3958453548759377,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004991075882797283,
      "loss": 6.4166,
      "mean_token_accuracy": 0.16273681819438934,
      "num_tokens": 8704704.0,
      "step": 3430
    },
    {
      "entropy": 6.55779709815979,
      "epoch": 0.39642238892094633,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991039129533337,
      "loss": 6.5896,
      "mean_token_accuracy": 0.14928019493818284,
      "num_tokens": 8717595.0,
      "step": 3435
    },
    {
      "entropy": 6.431856870651245,
      "epoch": 0.396999422965955,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991002300893152,
      "loss": 6.5343,
      "mean_token_accuracy": 0.1473251223564148,
      "num_tokens": 8729712.0,
      "step": 3440
    },
    {
      "entropy": 6.550535678863525,
      "epoch": 0.39757645701096367,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004990965396877969,
      "loss": 6.4622,
      "mean_token_accuracy": 0.15027807056903839,
      "num_tokens": 8742668.0,
      "step": 3445
    },
    {
      "entropy": 6.577229022979736,
      "epoch": 0.3981534910559723,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004990928417489027,
      "loss": 6.5481,
      "mean_token_accuracy": 0.16001597195863723,
      "num_tokens": 8754755.0,
      "step": 3450
    },
    {
      "entropy": 6.500936794281006,
      "epoch": 0.39873052510098095,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004990891362727572,
      "loss": 6.4758,
      "mean_token_accuracy": 0.15968786627054216,
      "num_tokens": 8766063.0,
      "step": 3455
    },
    {
      "entropy": 6.566354894638062,
      "epoch": 0.3993075591459896,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990854232594848,
      "loss": 6.5367,
      "mean_token_accuracy": 0.15854692906141282,
      "num_tokens": 8779226.0,
      "step": 3460
    },
    {
      "entropy": 6.508341217041016,
      "epoch": 0.3998845931909983,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004990817027092106,
      "loss": 6.5593,
      "mean_token_accuracy": 0.15566189736127853,
      "num_tokens": 8791404.0,
      "step": 3465
    },
    {
      "entropy": 6.57790846824646,
      "epoch": 0.4004616272360069,
      "grad_norm": 0.875,
      "learning_rate": 0.0004990779746220595,
      "loss": 6.5821,
      "mean_token_accuracy": 0.1489529937505722,
      "num_tokens": 8805306.0,
      "step": 3470
    },
    {
      "entropy": 6.529441833496094,
      "epoch": 0.40103866128101556,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004990742389981572,
      "loss": 6.4577,
      "mean_token_accuracy": 0.15899457931518554,
      "num_tokens": 8817850.0,
      "step": 3475
    },
    {
      "entropy": 6.46161298751831,
      "epoch": 0.40161569532602426,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499070495837629,
      "loss": 6.5026,
      "mean_token_accuracy": 0.1616385281085968,
      "num_tokens": 8830765.0,
      "step": 3480
    },
    {
      "entropy": 6.586075305938721,
      "epoch": 0.4021927293710329,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004990667451406012,
      "loss": 6.5211,
      "mean_token_accuracy": 0.1506490483880043,
      "num_tokens": 8843329.0,
      "step": 3485
    },
    {
      "entropy": 6.627585506439209,
      "epoch": 0.40276976341604154,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004990629869071995,
      "loss": 6.5412,
      "mean_token_accuracy": 0.15043091773986816,
      "num_tokens": 8856712.0,
      "step": 3490
    },
    {
      "entropy": 6.413798761367798,
      "epoch": 0.4033467974610502,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004990592211375506,
      "loss": 6.4955,
      "mean_token_accuracy": 0.15782689303159714,
      "num_tokens": 8868364.0,
      "step": 3495
    },
    {
      "entropy": 6.487509155273438,
      "epoch": 0.4039238315060589,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004990554478317811,
      "loss": 6.4598,
      "mean_token_accuracy": 0.15257180631160736,
      "num_tokens": 8881335.0,
      "step": 3500
    },
    {
      "entropy": 6.470539999008179,
      "epoch": 0.4045008655510675,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990516669900179,
      "loss": 6.5372,
      "mean_token_accuracy": 0.15584255158901214,
      "num_tokens": 8894049.0,
      "step": 3505
    },
    {
      "entropy": 6.5064982891082765,
      "epoch": 0.40507789959607615,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004990478786123882,
      "loss": 6.4361,
      "mean_token_accuracy": 0.15555315762758254,
      "num_tokens": 8906056.0,
      "step": 3510
    },
    {
      "entropy": 6.553112125396728,
      "epoch": 0.40565493364108485,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004990440826990193,
      "loss": 6.5165,
      "mean_token_accuracy": 0.15495406240224838,
      "num_tokens": 8919606.0,
      "step": 3515
    },
    {
      "entropy": 6.522773551940918,
      "epoch": 0.4062319676860935,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004990402792500392,
      "loss": 6.5123,
      "mean_token_accuracy": 0.15581716746091842,
      "num_tokens": 8934733.0,
      "step": 3520
    },
    {
      "entropy": 6.408885335922241,
      "epoch": 0.4068090017311021,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004990364682655754,
      "loss": 6.5339,
      "mean_token_accuracy": 0.14826919585466386,
      "num_tokens": 8948331.0,
      "step": 3525
    },
    {
      "entropy": 6.590506315231323,
      "epoch": 0.40738603577611077,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004990326497457564,
      "loss": 6.5164,
      "mean_token_accuracy": 0.15685458332300187,
      "num_tokens": 8961171.0,
      "step": 3530
    },
    {
      "entropy": 6.5318520069122314,
      "epoch": 0.40796306982111946,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004990288236907104,
      "loss": 6.5597,
      "mean_token_accuracy": 0.15040131211280822,
      "num_tokens": 8973164.0,
      "step": 3535
    },
    {
      "entropy": 6.631446075439453,
      "epoch": 0.4085401038661281,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004990249901005661,
      "loss": 6.6047,
      "mean_token_accuracy": 0.14263734072446824,
      "num_tokens": 8985981.0,
      "step": 3540
    },
    {
      "entropy": 6.585731554031372,
      "epoch": 0.40911713791113674,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004990211489754527,
      "loss": 6.4953,
      "mean_token_accuracy": 0.15466451197862624,
      "num_tokens": 8999473.0,
      "step": 3545
    },
    {
      "entropy": 6.426634979248047,
      "epoch": 0.40969417195614544,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004990173003154993,
      "loss": 6.4236,
      "mean_token_accuracy": 0.16166132539510727,
      "num_tokens": 9010876.0,
      "step": 3550
    },
    {
      "entropy": 6.612064552307129,
      "epoch": 0.4102712060011541,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004990134441208351,
      "loss": 6.5338,
      "mean_token_accuracy": 0.152121452242136,
      "num_tokens": 9024704.0,
      "step": 3555
    },
    {
      "entropy": 6.4920995235443115,
      "epoch": 0.4108482400461627,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004990095803915901,
      "loss": 6.5204,
      "mean_token_accuracy": 0.15348053574562073,
      "num_tokens": 9037006.0,
      "step": 3560
    },
    {
      "entropy": 6.612022972106933,
      "epoch": 0.41142527409117136,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004990057091278942,
      "loss": 6.5927,
      "mean_token_accuracy": 0.14343018233776092,
      "num_tokens": 9050205.0,
      "step": 3565
    },
    {
      "entropy": 6.590859603881836,
      "epoch": 0.41200230813618005,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004990018303298773,
      "loss": 6.4965,
      "mean_token_accuracy": 0.149009857326746,
      "num_tokens": 9062900.0,
      "step": 3570
    },
    {
      "entropy": 6.346018934249878,
      "epoch": 0.4125793421811887,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004989979439976703,
      "loss": 6.4456,
      "mean_token_accuracy": 0.15227758884429932,
      "num_tokens": 9074767.0,
      "step": 3575
    },
    {
      "entropy": 6.624157190322876,
      "epoch": 0.41315637622619733,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004989940501314037,
      "loss": 6.553,
      "mean_token_accuracy": 0.14820340275764465,
      "num_tokens": 9087314.0,
      "step": 3580
    },
    {
      "entropy": 6.481740808486938,
      "epoch": 0.413733410271206,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004989901487312084,
      "loss": 6.5113,
      "mean_token_accuracy": 0.1483243942260742,
      "num_tokens": 9099543.0,
      "step": 3585
    },
    {
      "entropy": 6.548365497589112,
      "epoch": 0.41431044431621467,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989862397972157,
      "loss": 6.4614,
      "mean_token_accuracy": 0.15363820791244506,
      "num_tokens": 9112262.0,
      "step": 3590
    },
    {
      "entropy": 6.383531141281128,
      "epoch": 0.4148874783612233,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989823233295572,
      "loss": 6.3875,
      "mean_token_accuracy": 0.157829087972641,
      "num_tokens": 9124444.0,
      "step": 3595
    },
    {
      "entropy": 6.385137987136841,
      "epoch": 0.41546451240623195,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004989783993283644,
      "loss": 6.4434,
      "mean_token_accuracy": 0.16091854870319366,
      "num_tokens": 9137848.0,
      "step": 3600
    },
    {
      "entropy": 6.5218345642089846,
      "epoch": 0.41604154645124064,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004989744677937694,
      "loss": 6.46,
      "mean_token_accuracy": 0.15344048589468,
      "num_tokens": 9148985.0,
      "step": 3605
    },
    {
      "entropy": 6.504273557662964,
      "epoch": 0.4166185804962493,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004989705287259046,
      "loss": 6.5107,
      "mean_token_accuracy": 0.15487662255764006,
      "num_tokens": 9161313.0,
      "step": 3610
    },
    {
      "entropy": 6.51858696937561,
      "epoch": 0.4171956145412579,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004989665821249021,
      "loss": 6.493,
      "mean_token_accuracy": 0.15578579902648926,
      "num_tokens": 9173616.0,
      "step": 3615
    },
    {
      "entropy": 6.535366773605347,
      "epoch": 0.4177726485862666,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000498962627990895,
      "loss": 6.5551,
      "mean_token_accuracy": 0.1463634192943573,
      "num_tokens": 9185951.0,
      "step": 3620
    },
    {
      "entropy": 6.540111637115478,
      "epoch": 0.41834968263127525,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989586663240161,
      "loss": 6.5724,
      "mean_token_accuracy": 0.15049021393060685,
      "num_tokens": 9199844.0,
      "step": 3625
    },
    {
      "entropy": 6.431955146789551,
      "epoch": 0.4189267166762839,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004989546971243988,
      "loss": 6.4512,
      "mean_token_accuracy": 0.16481505632400512,
      "num_tokens": 9213034.0,
      "step": 3630
    },
    {
      "entropy": 6.481080818176269,
      "epoch": 0.41950375072129253,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004989507203921763,
      "loss": 6.4543,
      "mean_token_accuracy": 0.15640199631452562,
      "num_tokens": 9224885.0,
      "step": 3635
    },
    {
      "entropy": 6.5513505935668945,
      "epoch": 0.42008078476630123,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989467361274828,
      "loss": 6.5258,
      "mean_token_accuracy": 0.14576067179441451,
      "num_tokens": 9238960.0,
      "step": 3640
    },
    {
      "entropy": 6.5696861743927,
      "epoch": 0.42065781881130987,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004989427443304519,
      "loss": 6.5289,
      "mean_token_accuracy": 0.14842990189790725,
      "num_tokens": 9253075.0,
      "step": 3645
    },
    {
      "entropy": 6.415720844268799,
      "epoch": 0.4212348528563185,
      "grad_norm": 0.90625,
      "learning_rate": 0.000498938745001218,
      "loss": 6.4598,
      "mean_token_accuracy": 0.15686085671186448,
      "num_tokens": 9266865.0,
      "step": 3650
    },
    {
      "entropy": 6.569009399414062,
      "epoch": 0.4218118869013272,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004989347381399158,
      "loss": 6.493,
      "mean_token_accuracy": 0.15215641111135483,
      "num_tokens": 9278046.0,
      "step": 3655
    },
    {
      "entropy": 6.42589693069458,
      "epoch": 0.42238892094633584,
      "grad_norm": 1.125,
      "learning_rate": 0.0004989307237466799,
      "loss": 6.4037,
      "mean_token_accuracy": 0.16045829504728318,
      "num_tokens": 9291998.0,
      "step": 3660
    },
    {
      "entropy": 6.429590511322021,
      "epoch": 0.4229659549913445,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004989267018216453,
      "loss": 6.4819,
      "mean_token_accuracy": 0.16067093014717101,
      "num_tokens": 9305023.0,
      "step": 3665
    },
    {
      "entropy": 6.465304327011109,
      "epoch": 0.4235429890363531,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004989226723649473,
      "loss": 6.4424,
      "mean_token_accuracy": 0.15339999943971633,
      "num_tokens": 9316635.0,
      "step": 3670
    },
    {
      "entropy": 6.5440758228302,
      "epoch": 0.4241200230813618,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004989186353767214,
      "loss": 6.488,
      "mean_token_accuracy": 0.1488993614912033,
      "num_tokens": 9329671.0,
      "step": 3675
    },
    {
      "entropy": 6.511537265777588,
      "epoch": 0.42469705712637046,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004989145908571035,
      "loss": 6.523,
      "mean_token_accuracy": 0.1477010205388069,
      "num_tokens": 9342634.0,
      "step": 3680
    },
    {
      "entropy": 6.601963520050049,
      "epoch": 0.4252740911713791,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004989105388062295,
      "loss": 6.5018,
      "mean_token_accuracy": 0.15283710807561873,
      "num_tokens": 9355096.0,
      "step": 3685
    },
    {
      "entropy": 6.392455291748047,
      "epoch": 0.4258511252163878,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004989064792242358,
      "loss": 6.4947,
      "mean_token_accuracy": 0.15217999070882798,
      "num_tokens": 9367722.0,
      "step": 3690
    },
    {
      "entropy": 6.576960325241089,
      "epoch": 0.42642815926139643,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004989024121112589,
      "loss": 6.5426,
      "mean_token_accuracy": 0.14412356466054915,
      "num_tokens": 9380247.0,
      "step": 3695
    },
    {
      "entropy": 6.461992454528809,
      "epoch": 0.4270051933064051,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004988983374674356,
      "loss": 6.3924,
      "mean_token_accuracy": 0.16220575124025344,
      "num_tokens": 9392216.0,
      "step": 3700
    },
    {
      "entropy": 6.481812000274658,
      "epoch": 0.4275822273514137,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004988942552929029,
      "loss": 6.5214,
      "mean_token_accuracy": 0.15055936574935913,
      "num_tokens": 9405215.0,
      "step": 3705
    },
    {
      "entropy": 6.48148832321167,
      "epoch": 0.4281592613964224,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004988901655877981,
      "loss": 6.4872,
      "mean_token_accuracy": 0.1521611511707306,
      "num_tokens": 9417582.0,
      "step": 3710
    },
    {
      "entropy": 6.498448276519776,
      "epoch": 0.42873629544143105,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004988860683522589,
      "loss": 6.4736,
      "mean_token_accuracy": 0.15102622658014297,
      "num_tokens": 9428680.0,
      "step": 3715
    },
    {
      "entropy": 6.522154855728149,
      "epoch": 0.4293133294864397,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000498881963586423,
      "loss": 6.3747,
      "mean_token_accuracy": 0.16042813509702683,
      "num_tokens": 9441669.0,
      "step": 3720
    },
    {
      "entropy": 6.450927782058716,
      "epoch": 0.4298903635314484,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004988778512904282,
      "loss": 6.5265,
      "mean_token_accuracy": 0.15414160937070848,
      "num_tokens": 9455214.0,
      "step": 3725
    },
    {
      "entropy": 6.521457576751709,
      "epoch": 0.430467397576457,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004988737314644134,
      "loss": 6.5081,
      "mean_token_accuracy": 0.15382953882217407,
      "num_tokens": 9468180.0,
      "step": 3730
    },
    {
      "entropy": 6.491891288757325,
      "epoch": 0.43104443162146566,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004988696041085168,
      "loss": 6.4818,
      "mean_token_accuracy": 0.16082691997289658,
      "num_tokens": 9481243.0,
      "step": 3735
    },
    {
      "entropy": 6.431043529510498,
      "epoch": 0.4316214656664743,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004988654692228772,
      "loss": 6.3971,
      "mean_token_accuracy": 0.16716998517513276,
      "num_tokens": 9493511.0,
      "step": 3740
    },
    {
      "entropy": 6.491238784790039,
      "epoch": 0.432198499711483,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004988613268076335,
      "loss": 6.4489,
      "mean_token_accuracy": 0.16466091573238373,
      "num_tokens": 9505450.0,
      "step": 3745
    },
    {
      "entropy": 6.50264720916748,
      "epoch": 0.43277553375649164,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004988571768629257,
      "loss": 6.4954,
      "mean_token_accuracy": 0.1550997629761696,
      "num_tokens": 9518684.0,
      "step": 3750
    },
    {
      "entropy": 6.456662368774414,
      "epoch": 0.4333525678015003,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004988530193888927,
      "loss": 6.4644,
      "mean_token_accuracy": 0.1511946514248848,
      "num_tokens": 9532064.0,
      "step": 3755
    },
    {
      "entropy": 6.490679693222046,
      "epoch": 0.4339296018465089,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988488543856747,
      "loss": 6.4161,
      "mean_token_accuracy": 0.16238633692264556,
      "num_tokens": 9543810.0,
      "step": 3760
    },
    {
      "entropy": 6.5098954200744625,
      "epoch": 0.4345066358915176,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004988446818534115,
      "loss": 6.556,
      "mean_token_accuracy": 0.1430600553750992,
      "num_tokens": 9556261.0,
      "step": 3765
    },
    {
      "entropy": 6.468726873397827,
      "epoch": 0.43508366993652625,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004988405017922438,
      "loss": 6.3753,
      "mean_token_accuracy": 0.15887598246335982,
      "num_tokens": 9569627.0,
      "step": 3770
    },
    {
      "entropy": 6.479031372070312,
      "epoch": 0.4356607039815349,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000498836314202312,
      "loss": 6.4952,
      "mean_token_accuracy": 0.15334070101380348,
      "num_tokens": 9582263.0,
      "step": 3775
    },
    {
      "entropy": 6.462040567398072,
      "epoch": 0.4362377380265436,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004988321190837568,
      "loss": 6.373,
      "mean_token_accuracy": 0.166064028441906,
      "num_tokens": 9594282.0,
      "step": 3780
    },
    {
      "entropy": 6.462106943130493,
      "epoch": 0.4368147720715522,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004988279164367196,
      "loss": 6.4191,
      "mean_token_accuracy": 0.16245402097702027,
      "num_tokens": 9607159.0,
      "step": 3785
    },
    {
      "entropy": 6.424819564819336,
      "epoch": 0.43739180611656087,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004988237062613415,
      "loss": 6.4386,
      "mean_token_accuracy": 0.1614897295832634,
      "num_tokens": 9619221.0,
      "step": 3790
    },
    {
      "entropy": 6.459392023086548,
      "epoch": 0.4379688401615695,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004988194885577644,
      "loss": 6.4763,
      "mean_token_accuracy": 0.15304807126522063,
      "num_tokens": 9631956.0,
      "step": 3795
    },
    {
      "entropy": 6.48423547744751,
      "epoch": 0.4385458742065782,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004988152633261299,
      "loss": 6.4867,
      "mean_token_accuracy": 0.15261341035366058,
      "num_tokens": 9643557.0,
      "step": 3800
    },
    {
      "entropy": 6.482478094100952,
      "epoch": 0.43912290825158684,
      "grad_norm": 0.984375,
      "learning_rate": 0.00049881103056658,
      "loss": 6.4409,
      "mean_token_accuracy": 0.15607529878616333,
      "num_tokens": 9655422.0,
      "step": 3805
    },
    {
      "entropy": 6.5474756240844725,
      "epoch": 0.4396999422965955,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004988067902792575,
      "loss": 6.5229,
      "mean_token_accuracy": 0.15297043174505234,
      "num_tokens": 9669522.0,
      "step": 3810
    },
    {
      "entropy": 6.483385181427002,
      "epoch": 0.4402769763416042,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004988025424643047,
      "loss": 6.5317,
      "mean_token_accuracy": 0.15213618725538253,
      "num_tokens": 9681301.0,
      "step": 3815
    },
    {
      "entropy": 6.465461111068725,
      "epoch": 0.4408540103866128,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004987982871218645,
      "loss": 6.4304,
      "mean_token_accuracy": 0.15666209310293197,
      "num_tokens": 9693894.0,
      "step": 3820
    },
    {
      "entropy": 6.560495281219483,
      "epoch": 0.44143104443162146,
      "grad_norm": 0.875,
      "learning_rate": 0.0004987940242520802,
      "loss": 6.5424,
      "mean_token_accuracy": 0.14954719096422195,
      "num_tokens": 9706499.0,
      "step": 3825
    },
    {
      "entropy": 6.406473302841187,
      "epoch": 0.4420080784766301,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000498789753855095,
      "loss": 6.3915,
      "mean_token_accuracy": 0.1605202242732048,
      "num_tokens": 9718632.0,
      "step": 3830
    },
    {
      "entropy": 6.487658500671387,
      "epoch": 0.4425851125216388,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004987854759310526,
      "loss": 6.4642,
      "mean_token_accuracy": 0.15463823229074478,
      "num_tokens": 9730451.0,
      "step": 3835
    },
    {
      "entropy": 6.549819898605347,
      "epoch": 0.44316214656664743,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004987811904800968,
      "loss": 6.4982,
      "mean_token_accuracy": 0.15468267053365709,
      "num_tokens": 9742693.0,
      "step": 3840
    },
    {
      "entropy": 6.477394723892212,
      "epoch": 0.44373918061165607,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004987768975023719,
      "loss": 6.4786,
      "mean_token_accuracy": 0.15444377213716506,
      "num_tokens": 9755642.0,
      "step": 3845
    },
    {
      "entropy": 6.509245300292969,
      "epoch": 0.44431621465666477,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004987725969980222,
      "loss": 6.5027,
      "mean_token_accuracy": 0.15192893296480178,
      "num_tokens": 9767674.0,
      "step": 3850
    },
    {
      "entropy": 6.4525409698486325,
      "epoch": 0.4448932487016734,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004987682889671923,
      "loss": 6.4225,
      "mean_token_accuracy": 0.16151144802570344,
      "num_tokens": 9781018.0,
      "step": 3855
    },
    {
      "entropy": 6.47997407913208,
      "epoch": 0.44547028274668204,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004987639734100272,
      "loss": 6.4968,
      "mean_token_accuracy": 0.15412253737449647,
      "num_tokens": 9792521.0,
      "step": 3860
    },
    {
      "entropy": 6.500252342224121,
      "epoch": 0.4460473167916907,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004987596503266721,
      "loss": 6.466,
      "mean_token_accuracy": 0.1622908443212509,
      "num_tokens": 9805356.0,
      "step": 3865
    },
    {
      "entropy": 6.449088191986084,
      "epoch": 0.4466243508366994,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004987553197172722,
      "loss": 6.4131,
      "mean_token_accuracy": 0.1637939691543579,
      "num_tokens": 9818779.0,
      "step": 3870
    },
    {
      "entropy": 6.551122665405273,
      "epoch": 0.447201384881708,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004987509815819732,
      "loss": 6.5006,
      "mean_token_accuracy": 0.1476660117506981,
      "num_tokens": 9832328.0,
      "step": 3875
    },
    {
      "entropy": 6.439172458648682,
      "epoch": 0.44777841892671666,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004987466359209213,
      "loss": 6.4513,
      "mean_token_accuracy": 0.15210386365652084,
      "num_tokens": 9846281.0,
      "step": 3880
    },
    {
      "entropy": 6.461850547790528,
      "epoch": 0.44835545297172535,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004987422827342622,
      "loss": 6.5029,
      "mean_token_accuracy": 0.15145214796066284,
      "num_tokens": 9858737.0,
      "step": 3885
    },
    {
      "entropy": 6.518165206909179,
      "epoch": 0.448932487016734,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004987379220221426,
      "loss": 6.5224,
      "mean_token_accuracy": 0.15191510170698166,
      "num_tokens": 9872852.0,
      "step": 3890
    },
    {
      "entropy": 6.493534088134766,
      "epoch": 0.44950952106174263,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004987335537847092,
      "loss": 6.4254,
      "mean_token_accuracy": 0.16384721994400026,
      "num_tokens": 9886339.0,
      "step": 3895
    },
    {
      "entropy": 6.534852457046509,
      "epoch": 0.4500865551067513,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004987291780221088,
      "loss": 6.4292,
      "mean_token_accuracy": 0.15550620704889298,
      "num_tokens": 9897525.0,
      "step": 3900
    },
    {
      "entropy": 6.390138387680054,
      "epoch": 0.45066358915175997,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004987247947344887,
      "loss": 6.4477,
      "mean_token_accuracy": 0.16208996921777724,
      "num_tokens": 9909733.0,
      "step": 3905
    },
    {
      "entropy": 6.469261932373047,
      "epoch": 0.4512406231967686,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004987204039219962,
      "loss": 6.4467,
      "mean_token_accuracy": 0.1487193763256073,
      "num_tokens": 9922503.0,
      "step": 3910
    },
    {
      "entropy": 6.543971729278565,
      "epoch": 0.45181765724177725,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004987160055847791,
      "loss": 6.4757,
      "mean_token_accuracy": 0.1538498356938362,
      "num_tokens": 9935517.0,
      "step": 3915
    },
    {
      "entropy": 6.493841934204101,
      "epoch": 0.45239469128678594,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004987115997229852,
      "loss": 6.525,
      "mean_token_accuracy": 0.1527313768863678,
      "num_tokens": 9947663.0,
      "step": 3920
    },
    {
      "entropy": 6.394566059112549,
      "epoch": 0.4529717253317946,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004987071863367629,
      "loss": 6.3061,
      "mean_token_accuracy": 0.1607285737991333,
      "num_tokens": 9960924.0,
      "step": 3925
    },
    {
      "entropy": 6.4464010238647464,
      "epoch": 0.4535487593768032,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004987027654262604,
      "loss": 6.4624,
      "mean_token_accuracy": 0.15592009723186492,
      "num_tokens": 9973506.0,
      "step": 3930
    },
    {
      "entropy": 6.472143888473511,
      "epoch": 0.45412579342181186,
      "grad_norm": 0.875,
      "learning_rate": 0.0004986983369916264,
      "loss": 6.4346,
      "mean_token_accuracy": 0.16311472207307814,
      "num_tokens": 9988334.0,
      "step": 3935
    },
    {
      "entropy": 6.535774040222168,
      "epoch": 0.45470282746682056,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004986939010330102,
      "loss": 6.52,
      "mean_token_accuracy": 0.15344877243041993,
      "num_tokens": 10001295.0,
      "step": 3940
    },
    {
      "entropy": 6.493282842636108,
      "epoch": 0.4552798615118292,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004986894575505606,
      "loss": 6.4134,
      "mean_token_accuracy": 0.15587726086378098,
      "num_tokens": 10012396.0,
      "step": 3945
    },
    {
      "entropy": 6.461516666412353,
      "epoch": 0.45585689555683784,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004986850065444272,
      "loss": 6.4693,
      "mean_token_accuracy": 0.1560237690806389,
      "num_tokens": 10025918.0,
      "step": 3950
    },
    {
      "entropy": 6.513438034057617,
      "epoch": 0.45643392960184653,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004986805480147598,
      "loss": 6.4819,
      "mean_token_accuracy": 0.1522982433438301,
      "num_tokens": 10038674.0,
      "step": 3955
    },
    {
      "entropy": 6.448969125747681,
      "epoch": 0.4570109636468552,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004986760819617082,
      "loss": 6.4441,
      "mean_token_accuracy": 0.1599159747362137,
      "num_tokens": 10051623.0,
      "step": 3960
    },
    {
      "entropy": 6.485679864883423,
      "epoch": 0.4575879976918638,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004986716083854228,
      "loss": 6.5285,
      "mean_token_accuracy": 0.15235468298196791,
      "num_tokens": 10063731.0,
      "step": 3965
    },
    {
      "entropy": 6.5528758525848385,
      "epoch": 0.45816503173687245,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004986671272860538,
      "loss": 6.5196,
      "mean_token_accuracy": 0.154839688539505,
      "num_tokens": 10075897.0,
      "step": 3970
    },
    {
      "entropy": 6.470433712005615,
      "epoch": 0.45874206578188115,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004986626386637521,
      "loss": 6.4832,
      "mean_token_accuracy": 0.14957205057144166,
      "num_tokens": 10088605.0,
      "step": 3975
    },
    {
      "entropy": 6.495354461669922,
      "epoch": 0.4593190998268898,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004986581425186688,
      "loss": 6.4681,
      "mean_token_accuracy": 0.15832031965255738,
      "num_tokens": 10101758.0,
      "step": 3980
    },
    {
      "entropy": 6.462432050704956,
      "epoch": 0.4598961338718984,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004986536388509548,
      "loss": 6.5056,
      "mean_token_accuracy": 0.14674035161733628,
      "num_tokens": 10113838.0,
      "step": 3985
    },
    {
      "entropy": 6.44532413482666,
      "epoch": 0.4604731679169071,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004986491276607618,
      "loss": 6.4418,
      "mean_token_accuracy": 0.15224733501672744,
      "num_tokens": 10127676.0,
      "step": 3990
    },
    {
      "entropy": 6.419986581802368,
      "epoch": 0.46105020196191576,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004986446089482416,
      "loss": 6.3945,
      "mean_token_accuracy": 0.1542766273021698,
      "num_tokens": 10139734.0,
      "step": 3995
    },
    {
      "entropy": 6.494247627258301,
      "epoch": 0.4616272360069244,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004986400827135459,
      "loss": 6.497,
      "mean_token_accuracy": 0.15193507373332976,
      "num_tokens": 10151842.0,
      "step": 4000
    },
    {
      "entropy": 6.504134607315064,
      "epoch": 0.46220427005193304,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004986355489568272,
      "loss": 6.5305,
      "mean_token_accuracy": 0.1533181682229042,
      "num_tokens": 10163539.0,
      "step": 4005
    },
    {
      "entropy": 6.490522575378418,
      "epoch": 0.46278130409694174,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004986310076782379,
      "loss": 6.4898,
      "mean_token_accuracy": 0.1537608966231346,
      "num_tokens": 10174760.0,
      "step": 4010
    },
    {
      "entropy": 6.499795389175415,
      "epoch": 0.4633583381419504,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004986264588779307,
      "loss": 6.5061,
      "mean_token_accuracy": 0.1535938248038292,
      "num_tokens": 10186627.0,
      "step": 4015
    },
    {
      "entropy": 6.602530479431152,
      "epoch": 0.463935372186959,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004986219025560586,
      "loss": 6.4011,
      "mean_token_accuracy": 0.16631423234939574,
      "num_tokens": 10200084.0,
      "step": 4020
    },
    {
      "entropy": 6.425627183914185,
      "epoch": 0.4645124062319677,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000498617338712775,
      "loss": 6.4761,
      "mean_token_accuracy": 0.15592840760946275,
      "num_tokens": 10212121.0,
      "step": 4025
    },
    {
      "entropy": 6.478949880599975,
      "epoch": 0.46508944027697635,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004986127673482332,
      "loss": 6.4587,
      "mean_token_accuracy": 0.1573188602924347,
      "num_tokens": 10225138.0,
      "step": 4030
    },
    {
      "entropy": 6.532035636901855,
      "epoch": 0.465666474321985,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004986081884625871,
      "loss": 6.4765,
      "mean_token_accuracy": 0.15413720309734344,
      "num_tokens": 10238309.0,
      "step": 4035
    },
    {
      "entropy": 6.493792390823364,
      "epoch": 0.46624350836699363,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004986036020559906,
      "loss": 6.4496,
      "mean_token_accuracy": 0.15392484962940217,
      "num_tokens": 10251405.0,
      "step": 4040
    },
    {
      "entropy": 6.491559934616089,
      "epoch": 0.4668205424120023,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000498599008128598,
      "loss": 6.4092,
      "mean_token_accuracy": 0.16049956083297728,
      "num_tokens": 10262870.0,
      "step": 4045
    },
    {
      "entropy": 6.337639427185058,
      "epoch": 0.46739757645701097,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004985944066805639,
      "loss": 6.4816,
      "mean_token_accuracy": 0.14968820363283158,
      "num_tokens": 10274283.0,
      "step": 4050
    },
    {
      "entropy": 6.5992334365844725,
      "epoch": 0.4679746105020196,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000498589797712043,
      "loss": 6.4818,
      "mean_token_accuracy": 0.15302202925086023,
      "num_tokens": 10287581.0,
      "step": 4055
    },
    {
      "entropy": 6.498902034759522,
      "epoch": 0.4685516445470283,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004985851812231903,
      "loss": 6.4056,
      "mean_token_accuracy": 0.15904517769813536,
      "num_tokens": 10299654.0,
      "step": 4060
    },
    {
      "entropy": 6.376985263824463,
      "epoch": 0.46912867859203694,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004985805572141611,
      "loss": 6.5082,
      "mean_token_accuracy": 0.14863618314266205,
      "num_tokens": 10312323.0,
      "step": 4065
    },
    {
      "entropy": 6.533452844619751,
      "epoch": 0.4697057126370456,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004985759256851108,
      "loss": 6.3638,
      "mean_token_accuracy": 0.16427833288908006,
      "num_tokens": 10325078.0,
      "step": 4070
    },
    {
      "entropy": 6.430614852905274,
      "epoch": 0.4702827466820542,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004985712866361954,
      "loss": 6.4746,
      "mean_token_accuracy": 0.1546455442905426,
      "num_tokens": 10339938.0,
      "step": 4075
    },
    {
      "entropy": 6.452829122543335,
      "epoch": 0.4708597807270629,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004985666400675709,
      "loss": 6.5019,
      "mean_token_accuracy": 0.15196197479963303,
      "num_tokens": 10352024.0,
      "step": 4080
    },
    {
      "entropy": 6.40549521446228,
      "epoch": 0.47143681477207156,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004985619859793934,
      "loss": 6.292,
      "mean_token_accuracy": 0.17043348848819734,
      "num_tokens": 10365919.0,
      "step": 4085
    },
    {
      "entropy": 6.3506697654724125,
      "epoch": 0.4720138488170802,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004985573243718195,
      "loss": 6.4038,
      "mean_token_accuracy": 0.1600800409913063,
      "num_tokens": 10377488.0,
      "step": 4090
    },
    {
      "entropy": 6.446927165985107,
      "epoch": 0.4725908828620889,
      "grad_norm": 0.875,
      "learning_rate": 0.0004985526552450061,
      "loss": 6.4132,
      "mean_token_accuracy": 0.16118346750736237,
      "num_tokens": 10391187.0,
      "step": 4095
    },
    {
      "entropy": 6.396145057678223,
      "epoch": 0.47316791690709753,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004985479785991103,
      "loss": 6.3763,
      "mean_token_accuracy": 0.1608425721526146,
      "num_tokens": 10403458.0,
      "step": 4100
    },
    {
      "entropy": 6.575830936431885,
      "epoch": 0.47374495095210617,
      "grad_norm": 1.0,
      "learning_rate": 0.0004985432944342892,
      "loss": 6.565,
      "mean_token_accuracy": 0.14376562014222144,
      "num_tokens": 10416533.0,
      "step": 4105
    },
    {
      "entropy": 6.531299591064453,
      "epoch": 0.4743219849971148,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004985386027507003,
      "loss": 6.4247,
      "mean_token_accuracy": 0.1552041932940483,
      "num_tokens": 10429593.0,
      "step": 4110
    },
    {
      "entropy": 6.37828631401062,
      "epoch": 0.4748990190421235,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004985339035485018,
      "loss": 6.4573,
      "mean_token_accuracy": 0.15608247965574265,
      "num_tokens": 10440520.0,
      "step": 4115
    },
    {
      "entropy": 6.4838886737823485,
      "epoch": 0.47547605308713214,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004985291968278513,
      "loss": 6.435,
      "mean_token_accuracy": 0.15583639293909074,
      "num_tokens": 10453971.0,
      "step": 4120
    },
    {
      "entropy": 6.435753011703492,
      "epoch": 0.4760530871321408,
      "grad_norm": 0.875,
      "learning_rate": 0.0004985244825889074,
      "loss": 6.4502,
      "mean_token_accuracy": 0.15667429938912392,
      "num_tokens": 10466033.0,
      "step": 4125
    },
    {
      "entropy": 6.469525909423828,
      "epoch": 0.4766301211771495,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004985197608318284,
      "loss": 6.4303,
      "mean_token_accuracy": 0.16232919096946716,
      "num_tokens": 10478608.0,
      "step": 4130
    },
    {
      "entropy": 6.539208793640137,
      "epoch": 0.4772071552221581,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004985150315567735,
      "loss": 6.5422,
      "mean_token_accuracy": 0.16004473865032195,
      "num_tokens": 10491223.0,
      "step": 4135
    },
    {
      "entropy": 6.573675584793091,
      "epoch": 0.47778418926716676,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004985102947639013,
      "loss": 6.5438,
      "mean_token_accuracy": 0.14932569712400437,
      "num_tokens": 10504398.0,
      "step": 4140
    },
    {
      "entropy": 6.420360708236695,
      "epoch": 0.4783612233121754,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004985055504533715,
      "loss": 6.3827,
      "mean_token_accuracy": 0.15640563815832137,
      "num_tokens": 10516998.0,
      "step": 4145
    },
    {
      "entropy": 6.416196393966675,
      "epoch": 0.4789382573571841,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004985007986253435,
      "loss": 6.3796,
      "mean_token_accuracy": 0.15376745015382767,
      "num_tokens": 10529632.0,
      "step": 4150
    },
    {
      "entropy": 6.343784666061401,
      "epoch": 0.47951529140219273,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004984960392799772,
      "loss": 6.3887,
      "mean_token_accuracy": 0.16021796613931655,
      "num_tokens": 10542479.0,
      "step": 4155
    },
    {
      "entropy": 6.387730264663697,
      "epoch": 0.4800923254472014,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004984912724174326,
      "loss": 6.3099,
      "mean_token_accuracy": 0.16574347615242005,
      "num_tokens": 10555322.0,
      "step": 4160
    },
    {
      "entropy": 6.508089923858643,
      "epoch": 0.48066935949221,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004984864980378701,
      "loss": 6.4607,
      "mean_token_accuracy": 0.15309241712093352,
      "num_tokens": 10567757.0,
      "step": 4165
    },
    {
      "entropy": 6.5475188255310055,
      "epoch": 0.4812463935372187,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004984817161414502,
      "loss": 6.4896,
      "mean_token_accuracy": 0.1458849310874939,
      "num_tokens": 10580820.0,
      "step": 4170
    },
    {
      "entropy": 6.495856809616089,
      "epoch": 0.48182342758222735,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004984769267283338,
      "loss": 6.4946,
      "mean_token_accuracy": 0.1546245776116848,
      "num_tokens": 10595690.0,
      "step": 4175
    },
    {
      "entropy": 6.441578054428101,
      "epoch": 0.482400461627236,
      "grad_norm": 1.0,
      "learning_rate": 0.0004984721297986819,
      "loss": 6.4076,
      "mean_token_accuracy": 0.151973195374012,
      "num_tokens": 10608393.0,
      "step": 4180
    },
    {
      "entropy": 6.5456239700317385,
      "epoch": 0.4829774956722447,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004984673253526561,
      "loss": 6.4881,
      "mean_token_accuracy": 0.1548421025276184,
      "num_tokens": 10621073.0,
      "step": 4185
    },
    {
      "entropy": 6.380731630325317,
      "epoch": 0.4835545297172533,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004984625133904176,
      "loss": 6.3552,
      "mean_token_accuracy": 0.16619164645671844,
      "num_tokens": 10633945.0,
      "step": 4190
    },
    {
      "entropy": 6.433036184310913,
      "epoch": 0.48413156376226196,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004984576939121286,
      "loss": 6.4654,
      "mean_token_accuracy": 0.15143086314201354,
      "num_tokens": 10645727.0,
      "step": 4195
    },
    {
      "entropy": 6.486042499542236,
      "epoch": 0.4847085978072706,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004984528669179511,
      "loss": 6.3347,
      "mean_token_accuracy": 0.15578174144029616,
      "num_tokens": 10658488.0,
      "step": 4200
    },
    {
      "entropy": 6.4260471820831295,
      "epoch": 0.4852856318522793,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004984480324080472,
      "loss": 6.4244,
      "mean_token_accuracy": 0.15568319112062454,
      "num_tokens": 10671361.0,
      "step": 4205
    },
    {
      "entropy": 6.479542827606201,
      "epoch": 0.48586266589728794,
      "grad_norm": 1.0,
      "learning_rate": 0.00049844319038258,
      "loss": 6.3959,
      "mean_token_accuracy": 0.16389940679073334,
      "num_tokens": 10684825.0,
      "step": 4210
    },
    {
      "entropy": 6.471680784225464,
      "epoch": 0.4864396999422966,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004984383408417119,
      "loss": 6.495,
      "mean_token_accuracy": 0.156496562063694,
      "num_tokens": 10698627.0,
      "step": 4215
    },
    {
      "entropy": 6.47646279335022,
      "epoch": 0.4870167339873053,
      "grad_norm": 1.125,
      "learning_rate": 0.0004984334837856064,
      "loss": 6.4108,
      "mean_token_accuracy": 0.1566810429096222,
      "num_tokens": 10712124.0,
      "step": 4220
    },
    {
      "entropy": 6.448310136795044,
      "epoch": 0.4875937680323139,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004984286192144263,
      "loss": 6.4294,
      "mean_token_accuracy": 0.150185289978981,
      "num_tokens": 10723939.0,
      "step": 4225
    },
    {
      "entropy": 6.461872816085815,
      "epoch": 0.48817080207732255,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004984237471283359,
      "loss": 6.4392,
      "mean_token_accuracy": 0.15271908789873123,
      "num_tokens": 10736795.0,
      "step": 4230
    },
    {
      "entropy": 6.367505502700806,
      "epoch": 0.4887478361223312,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004984188675274986,
      "loss": 6.3276,
      "mean_token_accuracy": 0.16758141219615935,
      "num_tokens": 10749905.0,
      "step": 4235
    },
    {
      "entropy": 6.463683032989502,
      "epoch": 0.4893248701673399,
      "grad_norm": 0.875,
      "learning_rate": 0.0004984139804120787,
      "loss": 6.4591,
      "mean_token_accuracy": 0.15041454136371613,
      "num_tokens": 10761852.0,
      "step": 4240
    },
    {
      "entropy": 6.517010068893432,
      "epoch": 0.4899019042123485,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004984090857822406,
      "loss": 6.4172,
      "mean_token_accuracy": 0.1570562407374382,
      "num_tokens": 10774673.0,
      "step": 4245
    },
    {
      "entropy": 6.37641396522522,
      "epoch": 0.49047893825735717,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004984041836381488,
      "loss": 6.3817,
      "mean_token_accuracy": 0.15902095884084702,
      "num_tokens": 10786892.0,
      "step": 4250
    },
    {
      "entropy": 6.415896892547607,
      "epoch": 0.49105597230236586,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004983992739799682,
      "loss": 6.4318,
      "mean_token_accuracy": 0.16172775477170945,
      "num_tokens": 10798846.0,
      "step": 4255
    },
    {
      "entropy": 6.483748197555542,
      "epoch": 0.4916330063473745,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004983943568078642,
      "loss": 6.3364,
      "mean_token_accuracy": 0.1619258478283882,
      "num_tokens": 10811893.0,
      "step": 4260
    },
    {
      "entropy": 6.423686265945435,
      "epoch": 0.49221004039238314,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004983894321220017,
      "loss": 6.3847,
      "mean_token_accuracy": 0.16486125141382219,
      "num_tokens": 10825595.0,
      "step": 4265
    },
    {
      "entropy": 6.3727141380310055,
      "epoch": 0.4927870744373918,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004983844999225468,
      "loss": 6.3654,
      "mean_token_accuracy": 0.16543812602758406,
      "num_tokens": 10837546.0,
      "step": 4270
    },
    {
      "entropy": 6.466187381744385,
      "epoch": 0.4933641084824005,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004983795602096651,
      "loss": 6.47,
      "mean_token_accuracy": 0.14846401363611222,
      "num_tokens": 10851234.0,
      "step": 4275
    },
    {
      "entropy": 6.429861164093017,
      "epoch": 0.4939411425274091,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000498374612983523,
      "loss": 6.4186,
      "mean_token_accuracy": 0.15042614340782165,
      "num_tokens": 10864397.0,
      "step": 4280
    },
    {
      "entropy": 6.525364637374878,
      "epoch": 0.49451817657241776,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004983696582442866,
      "loss": 6.4138,
      "mean_token_accuracy": 0.15194485038518907,
      "num_tokens": 10877230.0,
      "step": 4285
    },
    {
      "entropy": 6.38671498298645,
      "epoch": 0.49509521061742645,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004983646959921227,
      "loss": 6.4069,
      "mean_token_accuracy": 0.15624967962503433,
      "num_tokens": 10890847.0,
      "step": 4290
    },
    {
      "entropy": 6.403861570358276,
      "epoch": 0.4956722446624351,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004983597262271984,
      "loss": 6.3446,
      "mean_token_accuracy": 0.164557047188282,
      "num_tokens": 10903488.0,
      "step": 4295
    },
    {
      "entropy": 6.514218997955322,
      "epoch": 0.49624927870744373,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004983547489496804,
      "loss": 6.4404,
      "mean_token_accuracy": 0.15111730098724366,
      "num_tokens": 10914744.0,
      "step": 4300
    },
    {
      "entropy": 6.411670064926147,
      "epoch": 0.49682631275245237,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004983497641597365,
      "loss": 6.4352,
      "mean_token_accuracy": 0.15009788274765015,
      "num_tokens": 10928307.0,
      "step": 4305
    },
    {
      "entropy": 6.440000438690186,
      "epoch": 0.49740334679746107,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004983447718575342,
      "loss": 6.4617,
      "mean_token_accuracy": 0.158953058719635,
      "num_tokens": 10939876.0,
      "step": 4310
    },
    {
      "entropy": 6.422386884689331,
      "epoch": 0.4979803808424697,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004983397720432414,
      "loss": 6.3553,
      "mean_token_accuracy": 0.1608654499053955,
      "num_tokens": 10952123.0,
      "step": 4315
    },
    {
      "entropy": 6.462823438644409,
      "epoch": 0.49855741488747835,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004983347647170264,
      "loss": 6.4282,
      "mean_token_accuracy": 0.1512964189052582,
      "num_tokens": 10967374.0,
      "step": 4320
    },
    {
      "entropy": 6.469349813461304,
      "epoch": 0.49913444893248704,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004983297498790574,
      "loss": 6.4819,
      "mean_token_accuracy": 0.1485070012509823,
      "num_tokens": 10980761.0,
      "step": 4325
    },
    {
      "entropy": 6.5225193977355955,
      "epoch": 0.4997114829774957,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004983247275295034,
      "loss": 6.4154,
      "mean_token_accuracy": 0.15292058289051055,
      "num_tokens": 10994538.0,
      "step": 4330
    },
    {
      "entropy": 6.410770750045776,
      "epoch": 0.5002885170225043,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004983196976685329,
      "loss": 6.46,
      "mean_token_accuracy": 0.15204170644283294,
      "num_tokens": 11006956.0,
      "step": 4335
    },
    {
      "entropy": 6.472179841995239,
      "epoch": 0.500865551067513,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004983146602963155,
      "loss": 6.4308,
      "mean_token_accuracy": 0.16204098463058472,
      "num_tokens": 11019202.0,
      "step": 4340
    },
    {
      "entropy": 6.4863739013671875,
      "epoch": 0.5014425851125216,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004983096154130203,
      "loss": 6.4012,
      "mean_token_accuracy": 0.16002618819475173,
      "num_tokens": 11031560.0,
      "step": 4345
    },
    {
      "entropy": 6.441300630569458,
      "epoch": 0.5020196191575302,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004983045630188171,
      "loss": 6.5179,
      "mean_token_accuracy": 0.15357963740825653,
      "num_tokens": 11044537.0,
      "step": 4350
    },
    {
      "entropy": 6.443005847930908,
      "epoch": 0.502596653202539,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004982995031138759,
      "loss": 6.3569,
      "mean_token_accuracy": 0.15859626233577728,
      "num_tokens": 11056642.0,
      "step": 4355
    },
    {
      "entropy": 6.424235916137695,
      "epoch": 0.5031736872475476,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004982944356983666,
      "loss": 6.3515,
      "mean_token_accuracy": 0.1586825057864189,
      "num_tokens": 11068642.0,
      "step": 4360
    },
    {
      "entropy": 6.446497058868408,
      "epoch": 0.5037507212925563,
      "grad_norm": 4.0,
      "learning_rate": 0.0004982893607724602,
      "loss": 6.7038,
      "mean_token_accuracy": 0.15085916221141815,
      "num_tokens": 11082056.0,
      "step": 4365
    },
    {
      "entropy": 6.511081314086914,
      "epoch": 0.5043277553375649,
      "grad_norm": 1.015625,
      "learning_rate": 0.000498284278336327,
      "loss": 6.4539,
      "mean_token_accuracy": 0.15035639703273773,
      "num_tokens": 11093722.0,
      "step": 4370
    },
    {
      "entropy": 6.434291791915894,
      "epoch": 0.5049047893825735,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004982791883901379,
      "loss": 6.4472,
      "mean_token_accuracy": 0.14692117869853974,
      "num_tokens": 11105949.0,
      "step": 4375
    },
    {
      "entropy": 6.393206596374512,
      "epoch": 0.5054818234275822,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004982740909340643,
      "loss": 6.3449,
      "mean_token_accuracy": 0.16004691272974014,
      "num_tokens": 11118074.0,
      "step": 4380
    },
    {
      "entropy": 6.51700382232666,
      "epoch": 0.5060588574725908,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004982689859682775,
      "loss": 6.414,
      "mean_token_accuracy": 0.15509128123521804,
      "num_tokens": 11130917.0,
      "step": 4385
    },
    {
      "entropy": 6.4009030818939205,
      "epoch": 0.5066358915175996,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004982638734929492,
      "loss": 6.4533,
      "mean_token_accuracy": 0.15541249811649321,
      "num_tokens": 11142829.0,
      "step": 4390
    },
    {
      "entropy": 6.477549314498901,
      "epoch": 0.5072129255626082,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004982587535082515,
      "loss": 6.3917,
      "mean_token_accuracy": 0.16011728048324586,
      "num_tokens": 11154606.0,
      "step": 4395
    },
    {
      "entropy": 6.471315193176269,
      "epoch": 0.5077899596076169,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004982536260143565,
      "loss": 6.4023,
      "mean_token_accuracy": 0.15557657480239867,
      "num_tokens": 11167160.0,
      "step": 4400
    },
    {
      "entropy": 6.333266258239746,
      "epoch": 0.5083669936526255,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004982484910114367,
      "loss": 6.3859,
      "mean_token_accuracy": 0.16411517858505248,
      "num_tokens": 11178813.0,
      "step": 4405
    },
    {
      "entropy": 6.429812431335449,
      "epoch": 0.5089440276976341,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004982433484996648,
      "loss": 6.3447,
      "mean_token_accuracy": 0.15951201021671296,
      "num_tokens": 11192648.0,
      "step": 4410
    },
    {
      "entropy": 6.530735874176026,
      "epoch": 0.5095210617426428,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004982381984792138,
      "loss": 6.4631,
      "mean_token_accuracy": 0.14925868958234786,
      "num_tokens": 11205519.0,
      "step": 4415
    },
    {
      "entropy": 6.410419988632202,
      "epoch": 0.5100980957876514,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004982330409502568,
      "loss": 6.4389,
      "mean_token_accuracy": 0.15447123795747758,
      "num_tokens": 11218362.0,
      "step": 4420
    },
    {
      "entropy": 6.434107637405395,
      "epoch": 0.5106751298326602,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004982278759129674,
      "loss": 6.4289,
      "mean_token_accuracy": 0.15314491987228393,
      "num_tokens": 11231821.0,
      "step": 4425
    },
    {
      "entropy": 6.496310567855835,
      "epoch": 0.5112521638776688,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004982227033675194,
      "loss": 6.467,
      "mean_token_accuracy": 0.14909724295139312,
      "num_tokens": 11246300.0,
      "step": 4430
    },
    {
      "entropy": 6.450144529342651,
      "epoch": 0.5118291979226774,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004982175233140865,
      "loss": 6.3611,
      "mean_token_accuracy": 0.1621922507882118,
      "num_tokens": 11258561.0,
      "step": 4435
    },
    {
      "entropy": 6.45408444404602,
      "epoch": 0.5124062319676861,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004982123357528431,
      "loss": 6.3501,
      "mean_token_accuracy": 0.16058319658041,
      "num_tokens": 11271719.0,
      "step": 4440
    },
    {
      "entropy": 6.260220098495483,
      "epoch": 0.5129832660126947,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004982071406839636,
      "loss": 6.3721,
      "mean_token_accuracy": 0.16080536022782327,
      "num_tokens": 11284166.0,
      "step": 4445
    },
    {
      "entropy": 6.496598720550537,
      "epoch": 0.5135603000577034,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004982019381076229,
      "loss": 6.4275,
      "mean_token_accuracy": 0.15971364378929137,
      "num_tokens": 11296140.0,
      "step": 4450
    },
    {
      "entropy": 6.493388223648071,
      "epoch": 0.514137334102712,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004981967280239958,
      "loss": 6.4452,
      "mean_token_accuracy": 0.1555894523859024,
      "num_tokens": 11308680.0,
      "step": 4455
    },
    {
      "entropy": 6.435449314117432,
      "epoch": 0.5147143681477208,
      "grad_norm": 1.0,
      "learning_rate": 0.0004981915104332578,
      "loss": 6.421,
      "mean_token_accuracy": 0.15415364056825637,
      "num_tokens": 11320887.0,
      "step": 4460
    },
    {
      "entropy": 6.405152940750122,
      "epoch": 0.5152914021927294,
      "grad_norm": 0.890625,
      "learning_rate": 0.000498186285335584,
      "loss": 6.4172,
      "mean_token_accuracy": 0.15710679441690445,
      "num_tokens": 11333873.0,
      "step": 4465
    },
    {
      "entropy": 6.491097497940063,
      "epoch": 0.515868436237738,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004981810527311505,
      "loss": 6.5022,
      "mean_token_accuracy": 0.15293778330087662,
      "num_tokens": 11345526.0,
      "step": 4470
    },
    {
      "entropy": 6.469247579574585,
      "epoch": 0.5164454702827467,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000498175812620133,
      "loss": 6.3848,
      "mean_token_accuracy": 0.15478820949792862,
      "num_tokens": 11356840.0,
      "step": 4475
    },
    {
      "entropy": 6.430768203735352,
      "epoch": 0.5170225043277553,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004981705650027081,
      "loss": 6.4581,
      "mean_token_accuracy": 0.15306234657764434,
      "num_tokens": 11369146.0,
      "step": 4480
    },
    {
      "entropy": 6.459708118438721,
      "epoch": 0.517599538372764,
      "grad_norm": 0.953125,
      "learning_rate": 0.000498165309879052,
      "loss": 6.3846,
      "mean_token_accuracy": 0.155629900097847,
      "num_tokens": 11382399.0,
      "step": 4485
    },
    {
      "entropy": 6.392365121841431,
      "epoch": 0.5181765724177726,
      "grad_norm": 0.875,
      "learning_rate": 0.0004981600472493415,
      "loss": 6.3096,
      "mean_token_accuracy": 0.1627783253788948,
      "num_tokens": 11395602.0,
      "step": 4490
    },
    {
      "entropy": 6.396939373016357,
      "epoch": 0.5187536064627813,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004981547771137537,
      "loss": 6.4111,
      "mean_token_accuracy": 0.15267581939697267,
      "num_tokens": 11408530.0,
      "step": 4495
    },
    {
      "entropy": 6.43686351776123,
      "epoch": 0.51933064050779,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004981494994724658,
      "loss": 6.4165,
      "mean_token_accuracy": 0.157474847137928,
      "num_tokens": 11420150.0,
      "step": 4500
    },
    {
      "entropy": 6.3771324634552,
      "epoch": 0.5199076745527986,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004981442143256554,
      "loss": 6.4622,
      "mean_token_accuracy": 0.14853575974702835,
      "num_tokens": 11432318.0,
      "step": 4505
    },
    {
      "entropy": 6.538793992996216,
      "epoch": 0.5204847085978073,
      "grad_norm": 2.3125,
      "learning_rate": 0.0004981389216735001,
      "loss": 6.505,
      "mean_token_accuracy": 0.1544039450585842,
      "num_tokens": 11444143.0,
      "step": 4510
    },
    {
      "entropy": 6.399224472045899,
      "epoch": 0.5210617426428159,
      "grad_norm": 0.921875,
      "learning_rate": 0.000498133621516178,
      "loss": 6.3804,
      "mean_token_accuracy": 0.15825798958539963,
      "num_tokens": 11455997.0,
      "step": 4515
    },
    {
      "entropy": 6.511293268203735,
      "epoch": 0.5216387766878245,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004981283138538675,
      "loss": 6.4836,
      "mean_token_accuracy": 0.14820267930626868,
      "num_tokens": 11468271.0,
      "step": 4520
    },
    {
      "entropy": 6.473857879638672,
      "epoch": 0.5222158107328332,
      "grad_norm": 0.953125,
      "learning_rate": 0.000498122998686747,
      "loss": 6.4021,
      "mean_token_accuracy": 0.15662144273519515,
      "num_tokens": 11479729.0,
      "step": 4525
    },
    {
      "entropy": 6.288805770874023,
      "epoch": 0.5227928447778419,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004981176760149951,
      "loss": 6.3315,
      "mean_token_accuracy": 0.15843190997838974,
      "num_tokens": 11492261.0,
      "step": 4530
    },
    {
      "entropy": 6.433729362487793,
      "epoch": 0.5233698788228506,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004981123458387911,
      "loss": 6.3441,
      "mean_token_accuracy": 0.15808490216732024,
      "num_tokens": 11504802.0,
      "step": 4535
    },
    {
      "entropy": 6.471891021728515,
      "epoch": 0.5239469128678592,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004981070081583142,
      "loss": 6.4314,
      "mean_token_accuracy": 0.1515705719590187,
      "num_tokens": 11518105.0,
      "step": 4540
    },
    {
      "entropy": 6.413583993911743,
      "epoch": 0.5245239469128679,
      "grad_norm": 0.875,
      "learning_rate": 0.000498101662973744,
      "loss": 6.3827,
      "mean_token_accuracy": 0.15620462894439696,
      "num_tokens": 11532085.0,
      "step": 4545
    },
    {
      "entropy": 6.428698539733887,
      "epoch": 0.5251009809578765,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00049809631028526,
      "loss": 6.3411,
      "mean_token_accuracy": 0.16162063479423522,
      "num_tokens": 11544328.0,
      "step": 4550
    },
    {
      "entropy": 6.388873100280762,
      "epoch": 0.5256780150028851,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004980909500930424,
      "loss": 6.4345,
      "mean_token_accuracy": 0.16083729565143584,
      "num_tokens": 11556496.0,
      "step": 4555
    },
    {
      "entropy": 6.50629677772522,
      "epoch": 0.5262550490478938,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004980855823972717,
      "loss": 6.4712,
      "mean_token_accuracy": 0.15554953515529632,
      "num_tokens": 11569330.0,
      "step": 4560
    },
    {
      "entropy": 6.417497014999389,
      "epoch": 0.5268320830929025,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000498080207198128,
      "loss": 6.377,
      "mean_token_accuracy": 0.15880116671323777,
      "num_tokens": 11581335.0,
      "step": 4565
    },
    {
      "entropy": 6.397104930877686,
      "epoch": 0.5274091171379112,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004980748244957925,
      "loss": 6.3423,
      "mean_token_accuracy": 0.1576380878686905,
      "num_tokens": 11593130.0,
      "step": 4570
    },
    {
      "entropy": 6.324112606048584,
      "epoch": 0.5279861511829198,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004980694342904461,
      "loss": 6.3545,
      "mean_token_accuracy": 0.16386826783418657,
      "num_tokens": 11605140.0,
      "step": 4575
    },
    {
      "entropy": 6.415776777267456,
      "epoch": 0.5285631852279284,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004980640365822701,
      "loss": 6.373,
      "mean_token_accuracy": 0.15446255803108216,
      "num_tokens": 11617361.0,
      "step": 4580
    },
    {
      "entropy": 6.354536342620849,
      "epoch": 0.5291402192729371,
      "grad_norm": 1.0625,
      "learning_rate": 0.000498058631371446,
      "loss": 6.3891,
      "mean_token_accuracy": 0.15718972086906433,
      "num_tokens": 11630780.0,
      "step": 4585
    },
    {
      "entropy": 6.348697900772095,
      "epoch": 0.5297172533179457,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004980532186581555,
      "loss": 6.3447,
      "mean_token_accuracy": 0.16130129098892212,
      "num_tokens": 11643120.0,
      "step": 4590
    },
    {
      "entropy": 6.4900679111480715,
      "epoch": 0.5302942873629544,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000498047798442581,
      "loss": 6.3807,
      "mean_token_accuracy": 0.15732213705778123,
      "num_tokens": 11655849.0,
      "step": 4595
    },
    {
      "entropy": 6.353942251205444,
      "epoch": 0.5308713214079631,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004980423707249044,
      "loss": 6.362,
      "mean_token_accuracy": 0.16127122938632965,
      "num_tokens": 11669072.0,
      "step": 4600
    },
    {
      "entropy": 6.3463681697845455,
      "epoch": 0.5314483554529718,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004980369355053086,
      "loss": 6.3364,
      "mean_token_accuracy": 0.16097496449947357,
      "num_tokens": 11680253.0,
      "step": 4605
    },
    {
      "entropy": 6.38175482749939,
      "epoch": 0.5320253894979804,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004980314927839763,
      "loss": 6.3031,
      "mean_token_accuracy": 0.15603722333908082,
      "num_tokens": 11691465.0,
      "step": 4610
    },
    {
      "entropy": 6.375001859664917,
      "epoch": 0.532602423542989,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004980260425610903,
      "loss": 6.4634,
      "mean_token_accuracy": 0.15667866170406342,
      "num_tokens": 11703679.0,
      "step": 4615
    },
    {
      "entropy": 6.43471360206604,
      "epoch": 0.5331794575879977,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004980205848368343,
      "loss": 6.4089,
      "mean_token_accuracy": 0.15696175545454025,
      "num_tokens": 11716204.0,
      "step": 4620
    },
    {
      "entropy": 6.4551807880401615,
      "epoch": 0.5337564916330063,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004980151196113917,
      "loss": 6.3861,
      "mean_token_accuracy": 0.15830003768205642,
      "num_tokens": 11728404.0,
      "step": 4625
    },
    {
      "entropy": 6.359253740310669,
      "epoch": 0.534333525678015,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004980096468849464,
      "loss": 6.3273,
      "mean_token_accuracy": 0.1569062277674675,
      "num_tokens": 11740170.0,
      "step": 4630
    },
    {
      "entropy": 6.334492588043213,
      "epoch": 0.5349105597230237,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004980041666576823,
      "loss": 6.3046,
      "mean_token_accuracy": 0.15944114625453948,
      "num_tokens": 11753062.0,
      "step": 4635
    },
    {
      "entropy": 6.40184760093689,
      "epoch": 0.5354875937680323,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004979986789297837,
      "loss": 6.3171,
      "mean_token_accuracy": 0.16188960522413254,
      "num_tokens": 11765588.0,
      "step": 4640
    },
    {
      "entropy": 6.4119038581848145,
      "epoch": 0.536064627813041,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004979931837014355,
      "loss": 6.4281,
      "mean_token_accuracy": 0.15921278595924376,
      "num_tokens": 11777358.0,
      "step": 4645
    },
    {
      "entropy": 6.4223737716674805,
      "epoch": 0.5366416618580496,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004979876809728223,
      "loss": 6.4563,
      "mean_token_accuracy": 0.1526797816157341,
      "num_tokens": 11791572.0,
      "step": 4650
    },
    {
      "entropy": 6.453306102752686,
      "epoch": 0.5372186959030583,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004979821707441292,
      "loss": 6.4828,
      "mean_token_accuracy": 0.15242306441068648,
      "num_tokens": 11805198.0,
      "step": 4655
    },
    {
      "entropy": 6.475149869918823,
      "epoch": 0.5377957299480669,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004979766530155416,
      "loss": 6.4014,
      "mean_token_accuracy": 0.14970184713602067,
      "num_tokens": 11817882.0,
      "step": 4660
    },
    {
      "entropy": 6.40512523651123,
      "epoch": 0.5383727639930755,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004979711277872449,
      "loss": 6.2937,
      "mean_token_accuracy": 0.16547144055366517,
      "num_tokens": 11830334.0,
      "step": 4665
    },
    {
      "entropy": 6.331349182128906,
      "epoch": 0.5389497980380843,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004979655950594252,
      "loss": 6.3778,
      "mean_token_accuracy": 0.1643234074115753,
      "num_tokens": 11843057.0,
      "step": 4670
    },
    {
      "entropy": 6.47933030128479,
      "epoch": 0.5395268320830929,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004979600548322684,
      "loss": 6.4906,
      "mean_token_accuracy": 0.1458774909377098,
      "num_tokens": 11856086.0,
      "step": 4675
    },
    {
      "entropy": 6.442249345779419,
      "epoch": 0.5401038661281016,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000497954507105961,
      "loss": 6.4089,
      "mean_token_accuracy": 0.1492507502436638,
      "num_tokens": 11867799.0,
      "step": 4680
    },
    {
      "entropy": 6.347359037399292,
      "epoch": 0.5406809001731102,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004979489518806893,
      "loss": 6.3288,
      "mean_token_accuracy": 0.16592927724123002,
      "num_tokens": 11881488.0,
      "step": 4685
    },
    {
      "entropy": 6.452404975891113,
      "epoch": 0.5412579342181189,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004979433891566405,
      "loss": 6.4137,
      "mean_token_accuracy": 0.15162552148103714,
      "num_tokens": 11893495.0,
      "step": 4690
    },
    {
      "entropy": 6.481576728820801,
      "epoch": 0.5418349682631275,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004979378189340015,
      "loss": 6.36,
      "mean_token_accuracy": 0.15673246532678603,
      "num_tokens": 11905456.0,
      "step": 4695
    },
    {
      "entropy": 6.353079605102539,
      "epoch": 0.5424120023081361,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004979322412129597,
      "loss": 6.4098,
      "mean_token_accuracy": 0.15579713732004166,
      "num_tokens": 11917987.0,
      "step": 4700
    },
    {
      "entropy": 6.447024393081665,
      "epoch": 0.5429890363531449,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004979266559937028,
      "loss": 6.4015,
      "mean_token_accuracy": 0.1588173270225525,
      "num_tokens": 11930104.0,
      "step": 4705
    },
    {
      "entropy": 6.415519332885742,
      "epoch": 0.5435660703981535,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004979210632764185,
      "loss": 6.4008,
      "mean_token_accuracy": 0.14698654785752296,
      "num_tokens": 11942841.0,
      "step": 4710
    },
    {
      "entropy": 6.483328008651734,
      "epoch": 0.5441431044431622,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004979154630612949,
      "loss": 6.388,
      "mean_token_accuracy": 0.15481396466493608,
      "num_tokens": 11955509.0,
      "step": 4715
    },
    {
      "entropy": 6.379427814483643,
      "epoch": 0.5447201384881708,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004979098553485205,
      "loss": 6.3866,
      "mean_token_accuracy": 0.15711761713027955,
      "num_tokens": 11967389.0,
      "step": 4720
    },
    {
      "entropy": 6.454885482788086,
      "epoch": 0.5452971725331794,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004979042401382838,
      "loss": 6.4377,
      "mean_token_accuracy": 0.16024379879236222,
      "num_tokens": 11980173.0,
      "step": 4725
    },
    {
      "entropy": 6.444041919708252,
      "epoch": 0.5458742065781881,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004978986174307737,
      "loss": 6.3829,
      "mean_token_accuracy": 0.15860662013292312,
      "num_tokens": 11993039.0,
      "step": 4730
    },
    {
      "entropy": 6.312199211120605,
      "epoch": 0.5464512406231967,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004978929872261794,
      "loss": 6.2768,
      "mean_token_accuracy": 0.1646609917283058,
      "num_tokens": 12005861.0,
      "step": 4735
    },
    {
      "entropy": 6.322019433975219,
      "epoch": 0.5470282746682055,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004978873495246901,
      "loss": 6.2311,
      "mean_token_accuracy": 0.16512125879526138,
      "num_tokens": 12018137.0,
      "step": 4740
    },
    {
      "entropy": 6.4125762462615965,
      "epoch": 0.5476053087132141,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004978817043264955,
      "loss": 6.4044,
      "mean_token_accuracy": 0.15462562143802644,
      "num_tokens": 12030086.0,
      "step": 4745
    },
    {
      "entropy": 6.367450761795044,
      "epoch": 0.5481823427582228,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004978760516317856,
      "loss": 6.373,
      "mean_token_accuracy": 0.16457190215587617,
      "num_tokens": 12043166.0,
      "step": 4750
    },
    {
      "entropy": 6.366231203079224,
      "epoch": 0.5487593768032314,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004978703914407503,
      "loss": 6.3196,
      "mean_token_accuracy": 0.1610228791832924,
      "num_tokens": 12055373.0,
      "step": 4755
    },
    {
      "entropy": 6.3848343849182125,
      "epoch": 0.54933641084824,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004978647237535802,
      "loss": 6.3157,
      "mean_token_accuracy": 0.1572040781378746,
      "num_tokens": 12066369.0,
      "step": 4760
    },
    {
      "entropy": 6.374715852737427,
      "epoch": 0.5499134448932487,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004978590485704657,
      "loss": 6.2515,
      "mean_token_accuracy": 0.16094694957137107,
      "num_tokens": 12077939.0,
      "step": 4765
    },
    {
      "entropy": 6.389651966094971,
      "epoch": 0.5504904789382573,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004978533658915979,
      "loss": 6.4532,
      "mean_token_accuracy": 0.15065887272357942,
      "num_tokens": 12091102.0,
      "step": 4770
    },
    {
      "entropy": 6.478321647644043,
      "epoch": 0.5510675129832661,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004978476757171678,
      "loss": 6.3739,
      "mean_token_accuracy": 0.15410165637731552,
      "num_tokens": 12103144.0,
      "step": 4775
    },
    {
      "entropy": 6.333250713348389,
      "epoch": 0.5516445470282747,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004978419780473668,
      "loss": 6.4483,
      "mean_token_accuracy": 0.15764946788549422,
      "num_tokens": 12116664.0,
      "step": 4780
    },
    {
      "entropy": 6.436701679229737,
      "epoch": 0.5522215810732833,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004978362728823865,
      "loss": 6.3813,
      "mean_token_accuracy": 0.15632593482732773,
      "num_tokens": 12129656.0,
      "step": 4785
    },
    {
      "entropy": 6.403853178024292,
      "epoch": 0.552798615118292,
      "grad_norm": 1.0,
      "learning_rate": 0.0004978305602224189,
      "loss": 6.3414,
      "mean_token_accuracy": 0.15877176821231842,
      "num_tokens": 12143221.0,
      "step": 4790
    },
    {
      "entropy": 6.362843656539917,
      "epoch": 0.5533756491633006,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004978248400676562,
      "loss": 6.2979,
      "mean_token_accuracy": 0.16119197756052017,
      "num_tokens": 12156409.0,
      "step": 4795
    },
    {
      "entropy": 6.28771858215332,
      "epoch": 0.5539526832083093,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004978191124182905,
      "loss": 6.2273,
      "mean_token_accuracy": 0.17134579718112947,
      "num_tokens": 12169589.0,
      "step": 4800
    },
    {
      "entropy": 6.354664611816406,
      "epoch": 0.5545297172533179,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004978133772745149,
      "loss": 6.3729,
      "mean_token_accuracy": 0.15380450785160066,
      "num_tokens": 12182692.0,
      "step": 4805
    },
    {
      "entropy": 6.3929492950439455,
      "epoch": 0.5551067512983267,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004978076346365218,
      "loss": 6.3419,
      "mean_token_accuracy": 0.15942178666591644,
      "num_tokens": 12195892.0,
      "step": 4810
    },
    {
      "entropy": 6.423626279830932,
      "epoch": 0.5556837853433353,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004978018845045047,
      "loss": 6.3478,
      "mean_token_accuracy": 0.15422828048467635,
      "num_tokens": 12208974.0,
      "step": 4815
    },
    {
      "entropy": 6.394153594970703,
      "epoch": 0.5562608193883439,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004977961268786568,
      "loss": 6.4159,
      "mean_token_accuracy": 0.1570199027657509,
      "num_tokens": 12221342.0,
      "step": 4820
    },
    {
      "entropy": 6.338186359405517,
      "epoch": 0.5568378534333526,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004977903617591719,
      "loss": 6.397,
      "mean_token_accuracy": 0.15388095527887344,
      "num_tokens": 12233778.0,
      "step": 4825
    },
    {
      "entropy": 6.382364320755005,
      "epoch": 0.5574148874783612,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004977845891462439,
      "loss": 6.2511,
      "mean_token_accuracy": 0.163596773147583,
      "num_tokens": 12246001.0,
      "step": 4830
    },
    {
      "entropy": 6.322884130477905,
      "epoch": 0.5579919215233698,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004977788090400668,
      "loss": 6.2718,
      "mean_token_accuracy": 0.15886229425668716,
      "num_tokens": 12258677.0,
      "step": 4835
    },
    {
      "entropy": 6.393141174316407,
      "epoch": 0.5585689555683785,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004977730214408352,
      "loss": 6.377,
      "mean_token_accuracy": 0.157019679248333,
      "num_tokens": 12270786.0,
      "step": 4840
    },
    {
      "entropy": 6.347432708740234,
      "epoch": 0.5591459896133872,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004977672263487435,
      "loss": 6.3967,
      "mean_token_accuracy": 0.15893087387084961,
      "num_tokens": 12284439.0,
      "step": 4845
    },
    {
      "entropy": 6.405143547058105,
      "epoch": 0.5597230236583959,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000497761423763987,
      "loss": 6.3776,
      "mean_token_accuracy": 0.15470390766859055,
      "num_tokens": 12297005.0,
      "step": 4850
    },
    {
      "entropy": 6.419815921783448,
      "epoch": 0.5603000577034045,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004977556136867606,
      "loss": 6.3953,
      "mean_token_accuracy": 0.15327109545469284,
      "num_tokens": 12308855.0,
      "step": 4855
    },
    {
      "entropy": 6.490292358398437,
      "epoch": 0.5608770917484132,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004977497961172598,
      "loss": 6.3448,
      "mean_token_accuracy": 0.15836585462093353,
      "num_tokens": 12321120.0,
      "step": 4860
    },
    {
      "entropy": 6.357276201248169,
      "epoch": 0.5614541257934218,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004977439710556802,
      "loss": 6.4136,
      "mean_token_accuracy": 0.1530259892344475,
      "num_tokens": 12333793.0,
      "step": 4865
    },
    {
      "entropy": 6.390624952316284,
      "epoch": 0.5620311598384304,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004977381385022179,
      "loss": 6.2873,
      "mean_token_accuracy": 0.16176307946443558,
      "num_tokens": 12345998.0,
      "step": 4870
    },
    {
      "entropy": 6.390095329284668,
      "epoch": 0.5626081938834391,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004977322984570688,
      "loss": 6.3457,
      "mean_token_accuracy": 0.16212226152420045,
      "num_tokens": 12358580.0,
      "step": 4875
    },
    {
      "entropy": 6.469683980941772,
      "epoch": 0.5631852279284478,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004977264509204296,
      "loss": 6.4172,
      "mean_token_accuracy": 0.15353054255247117,
      "num_tokens": 12370867.0,
      "step": 4880
    },
    {
      "entropy": 6.420956182479858,
      "epoch": 0.5637622619734565,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004977205958924967,
      "loss": 6.4102,
      "mean_token_accuracy": 0.1578800857067108,
      "num_tokens": 12384086.0,
      "step": 4885
    },
    {
      "entropy": 6.288908243179321,
      "epoch": 0.5643392960184651,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004977147333734673,
      "loss": 6.2587,
      "mean_token_accuracy": 0.1637159377336502,
      "num_tokens": 12396852.0,
      "step": 4890
    },
    {
      "entropy": 6.4398151397705075,
      "epoch": 0.5649163300634737,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004977088633635385,
      "loss": 6.3255,
      "mean_token_accuracy": 0.16203283667564392,
      "num_tokens": 12409924.0,
      "step": 4895
    },
    {
      "entropy": 6.229389286041259,
      "epoch": 0.5654933641084824,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004977029858629076,
      "loss": 6.3156,
      "mean_token_accuracy": 0.1679264113306999,
      "num_tokens": 12422218.0,
      "step": 4900
    },
    {
      "entropy": 6.400584363937378,
      "epoch": 0.566070398153491,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004976971008717726,
      "loss": 6.3317,
      "mean_token_accuracy": 0.16040289402008057,
      "num_tokens": 12433522.0,
      "step": 4905
    },
    {
      "entropy": 6.611154556274414,
      "epoch": 0.5666474321984997,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004976912083903309,
      "loss": 6.4105,
      "mean_token_accuracy": 0.151394285261631,
      "num_tokens": 12445526.0,
      "step": 4910
    },
    {
      "entropy": 6.3067789554595945,
      "epoch": 0.5672244662435084,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004976853084187814,
      "loss": 6.3594,
      "mean_token_accuracy": 0.16232792884111405,
      "num_tokens": 12458403.0,
      "step": 4915
    },
    {
      "entropy": 6.458930969238281,
      "epoch": 0.5678015002885171,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004976794009573219,
      "loss": 6.4035,
      "mean_token_accuracy": 0.15569291710853578,
      "num_tokens": 12470437.0,
      "step": 4920
    },
    {
      "entropy": 6.44275689125061,
      "epoch": 0.5683785343335257,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004976734860061515,
      "loss": 6.3432,
      "mean_token_accuracy": 0.16315540671348572,
      "num_tokens": 12483345.0,
      "step": 4925
    },
    {
      "entropy": 6.269897222518921,
      "epoch": 0.5689555683785343,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004976675635654688,
      "loss": 6.3438,
      "mean_token_accuracy": 0.15834786742925644,
      "num_tokens": 12495675.0,
      "step": 4930
    },
    {
      "entropy": 6.415697908401489,
      "epoch": 0.569532602423543,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004976616336354733,
      "loss": 6.3663,
      "mean_token_accuracy": 0.15901082754135132,
      "num_tokens": 12507569.0,
      "step": 4935
    },
    {
      "entropy": 6.452373361587524,
      "epoch": 0.5701096364685516,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004976556962163645,
      "loss": 6.4099,
      "mean_token_accuracy": 0.1575494185090065,
      "num_tokens": 12521198.0,
      "step": 4940
    },
    {
      "entropy": 6.333392381668091,
      "epoch": 0.5706866705135603,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004976497513083419,
      "loss": 6.3056,
      "mean_token_accuracy": 0.15893585830926896,
      "num_tokens": 12533242.0,
      "step": 4945
    },
    {
      "entropy": 6.439319181442261,
      "epoch": 0.571263704558569,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004976437989116055,
      "loss": 6.4368,
      "mean_token_accuracy": 0.15476369261741638,
      "num_tokens": 12545918.0,
      "step": 4950
    },
    {
      "entropy": 6.4090534210205075,
      "epoch": 0.5718407386035776,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004976378390263554,
      "loss": 6.3808,
      "mean_token_accuracy": 0.15577974021434784,
      "num_tokens": 12558090.0,
      "step": 4955
    },
    {
      "entropy": 6.451124382019043,
      "epoch": 0.5724177726485863,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004976318716527924,
      "loss": 6.3967,
      "mean_token_accuracy": 0.1588565841317177,
      "num_tokens": 12571453.0,
      "step": 4960
    },
    {
      "entropy": 6.405358076095581,
      "epoch": 0.5729948066935949,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004976258967911168,
      "loss": 6.3444,
      "mean_token_accuracy": 0.15529514700174332,
      "num_tokens": 12584762.0,
      "step": 4965
    },
    {
      "entropy": 6.396466255187988,
      "epoch": 0.5735718407386036,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004976199144415298,
      "loss": 6.3481,
      "mean_token_accuracy": 0.15398240387439727,
      "num_tokens": 12597224.0,
      "step": 4970
    },
    {
      "entropy": 6.371331834793091,
      "epoch": 0.5741488747836122,
      "grad_norm": 0.875,
      "learning_rate": 0.0004976139246042325,
      "loss": 6.4026,
      "mean_token_accuracy": 0.15853663980960847,
      "num_tokens": 12609837.0,
      "step": 4975
    },
    {
      "entropy": 6.444687557220459,
      "epoch": 0.5747259088286208,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004976079272794265,
      "loss": 6.447,
      "mean_token_accuracy": 0.14801184684038163,
      "num_tokens": 12622596.0,
      "step": 4980
    },
    {
      "entropy": 6.411276197433471,
      "epoch": 0.5753029428736296,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004976019224673134,
      "loss": 6.3408,
      "mean_token_accuracy": 0.16246938481926917,
      "num_tokens": 12634721.0,
      "step": 4985
    },
    {
      "entropy": 6.441895008087158,
      "epoch": 0.5758799769186382,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004975959101680953,
      "loss": 6.358,
      "mean_token_accuracy": 0.15530874878168105,
      "num_tokens": 12646389.0,
      "step": 4990
    },
    {
      "entropy": 6.423140668869019,
      "epoch": 0.5764570109636469,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004975898903819742,
      "loss": 6.4604,
      "mean_token_accuracy": 0.14599337428808212,
      "num_tokens": 12661081.0,
      "step": 4995
    },
    {
      "entropy": 6.567923545837402,
      "epoch": 0.5770340450086555,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004975838631091527,
      "loss": 6.4175,
      "mean_token_accuracy": 0.14941588342189788,
      "num_tokens": 12672505.0,
      "step": 5000
    },
    {
      "entropy": 6.446028232574463,
      "epoch": 0.5776110790536642,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004975778283498336,
      "loss": 6.4353,
      "mean_token_accuracy": 0.15279321372509003,
      "num_tokens": 12684805.0,
      "step": 5005
    },
    {
      "entropy": 6.325024652481079,
      "epoch": 0.5781881130986728,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004975717861042198,
      "loss": 6.358,
      "mean_token_accuracy": 0.15606274008750914,
      "num_tokens": 12698557.0,
      "step": 5010
    },
    {
      "entropy": 6.471566247940063,
      "epoch": 0.5787651471436814,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004975657363725146,
      "loss": 6.3822,
      "mean_token_accuracy": 0.1520364172756672,
      "num_tokens": 12712472.0,
      "step": 5015
    },
    {
      "entropy": 6.402759695053101,
      "epoch": 0.5793421811886902,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004975596791549213,
      "loss": 6.2544,
      "mean_token_accuracy": 0.16502293348312377,
      "num_tokens": 12725698.0,
      "step": 5020
    },
    {
      "entropy": 6.2927028179168705,
      "epoch": 0.5799192152336988,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004975536144516437,
      "loss": 6.3588,
      "mean_token_accuracy": 0.15943397134542464,
      "num_tokens": 12738126.0,
      "step": 5025
    },
    {
      "entropy": 6.465342044830322,
      "epoch": 0.5804962492787075,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000497547542262886,
      "loss": 6.3565,
      "mean_token_accuracy": 0.1686787113547325,
      "num_tokens": 12751402.0,
      "step": 5030
    },
    {
      "entropy": 6.426917457580567,
      "epoch": 0.5810732833237161,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004975414625888521,
      "loss": 6.3787,
      "mean_token_accuracy": 0.1541376531124115,
      "num_tokens": 12763957.0,
      "step": 5035
    },
    {
      "entropy": 6.412431812286377,
      "epoch": 0.5816503173687247,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004975353754297468,
      "loss": 6.3552,
      "mean_token_accuracy": 0.15923605114221573,
      "num_tokens": 12776498.0,
      "step": 5040
    },
    {
      "entropy": 6.455137538909912,
      "epoch": 0.5822273514137334,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004975292807857745,
      "loss": 6.341,
      "mean_token_accuracy": 0.16326855272054672,
      "num_tokens": 12788498.0,
      "step": 5045
    },
    {
      "entropy": 6.261941957473755,
      "epoch": 0.582804385458742,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004975231786571406,
      "loss": 6.269,
      "mean_token_accuracy": 0.15788996666669847,
      "num_tokens": 12799907.0,
      "step": 5050
    },
    {
      "entropy": 6.40986852645874,
      "epoch": 0.5833814195037508,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004975170690440499,
      "loss": 6.3896,
      "mean_token_accuracy": 0.1579787701368332,
      "num_tokens": 12812610.0,
      "step": 5055
    },
    {
      "entropy": 6.45856614112854,
      "epoch": 0.5839584535487594,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004975109519467083,
      "loss": 6.3885,
      "mean_token_accuracy": 0.1525486320257187,
      "num_tokens": 12825082.0,
      "step": 5060
    },
    {
      "entropy": 6.317858934402466,
      "epoch": 0.584535487593768,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004975048273653212,
      "loss": 6.3103,
      "mean_token_accuracy": 0.16435761600732804,
      "num_tokens": 12837999.0,
      "step": 5065
    },
    {
      "entropy": 6.357636547088623,
      "epoch": 0.5851125216387767,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004974986953000948,
      "loss": 6.2458,
      "mean_token_accuracy": 0.16461012214422227,
      "num_tokens": 12850020.0,
      "step": 5070
    },
    {
      "entropy": 6.361400747299195,
      "epoch": 0.5856895556837853,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004974925557512354,
      "loss": 6.353,
      "mean_token_accuracy": 0.1577102944254875,
      "num_tokens": 12863818.0,
      "step": 5075
    },
    {
      "entropy": 6.361227655410767,
      "epoch": 0.586266589728794,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004974864087189493,
      "loss": 6.3623,
      "mean_token_accuracy": 0.15480730086565017,
      "num_tokens": 12876959.0,
      "step": 5080
    },
    {
      "entropy": 6.409631156921387,
      "epoch": 0.5868436237738026,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004974802542034434,
      "loss": 6.3404,
      "mean_token_accuracy": 0.15573345720767975,
      "num_tokens": 12889817.0,
      "step": 5085
    },
    {
      "entropy": 6.360256958007812,
      "epoch": 0.5874206578188114,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004974740922049246,
      "loss": 6.3216,
      "mean_token_accuracy": 0.16168117076158522,
      "num_tokens": 12901438.0,
      "step": 5090
    },
    {
      "entropy": 6.416221952438354,
      "epoch": 0.58799769186382,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004974679227236004,
      "loss": 6.2535,
      "mean_token_accuracy": 0.1679604932665825,
      "num_tokens": 12914115.0,
      "step": 5095
    },
    {
      "entropy": 6.317728328704834,
      "epoch": 0.5885747259088286,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004974617457596779,
      "loss": 6.3141,
      "mean_token_accuracy": 0.1633978709578514,
      "num_tokens": 12926822.0,
      "step": 5100
    },
    {
      "entropy": 6.375456619262695,
      "epoch": 0.5891517599538373,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004974555613133652,
      "loss": 6.3422,
      "mean_token_accuracy": 0.15513903349637986,
      "num_tokens": 12939515.0,
      "step": 5105
    },
    {
      "entropy": 6.237862253189087,
      "epoch": 0.5897287939988459,
      "grad_norm": 1.0,
      "learning_rate": 0.0004974493693848702,
      "loss": 6.2701,
      "mean_token_accuracy": 0.17829071134328842,
      "num_tokens": 12952195.0,
      "step": 5110
    },
    {
      "entropy": 6.428456497192383,
      "epoch": 0.5903058280438546,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004974431699744011,
      "loss": 6.2987,
      "mean_token_accuracy": 0.15931333899497985,
      "num_tokens": 12964563.0,
      "step": 5115
    },
    {
      "entropy": 6.326723623275757,
      "epoch": 0.5908828620888632,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004974369630821665,
      "loss": 6.2308,
      "mean_token_accuracy": 0.16145689487457277,
      "num_tokens": 12976564.0,
      "step": 5120
    },
    {
      "entropy": 6.172460556030273,
      "epoch": 0.5914598961338718,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004974307487083752,
      "loss": 6.1864,
      "mean_token_accuracy": 0.16396143138408661,
      "num_tokens": 12989757.0,
      "step": 5125
    },
    {
      "entropy": 6.438552045822144,
      "epoch": 0.5920369301788806,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004974245268532361,
      "loss": 6.4279,
      "mean_token_accuracy": 0.15544405281543733,
      "num_tokens": 13003846.0,
      "step": 5130
    },
    {
      "entropy": 6.4779815673828125,
      "epoch": 0.5926139642238892,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004974182975169585,
      "loss": 6.3898,
      "mean_token_accuracy": 0.1511077642440796,
      "num_tokens": 13018389.0,
      "step": 5135
    },
    {
      "entropy": 6.404718494415283,
      "epoch": 0.5931909982688979,
      "grad_norm": 0.90625,
      "learning_rate": 0.000497412060699752,
      "loss": 6.4045,
      "mean_token_accuracy": 0.15122631639242173,
      "num_tokens": 13031893.0,
      "step": 5140
    },
    {
      "entropy": 6.383275890350342,
      "epoch": 0.5937680323139065,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004974058164018263,
      "loss": 6.3507,
      "mean_token_accuracy": 0.16599466204643248,
      "num_tokens": 13045916.0,
      "step": 5145
    },
    {
      "entropy": 6.448142147064209,
      "epoch": 0.5943450663589152,
      "grad_norm": 0.875,
      "learning_rate": 0.0004973995646233914,
      "loss": 6.3616,
      "mean_token_accuracy": 0.15640936940908431,
      "num_tokens": 13058468.0,
      "step": 5150
    },
    {
      "entropy": 6.3777610778808596,
      "epoch": 0.5949221004039238,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004973933053646576,
      "loss": 6.389,
      "mean_token_accuracy": 0.15138502717018126,
      "num_tokens": 13070778.0,
      "step": 5155
    },
    {
      "entropy": 6.37696099281311,
      "epoch": 0.5954991344489324,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004973870386258355,
      "loss": 6.2634,
      "mean_token_accuracy": 0.16085633337497712,
      "num_tokens": 13082287.0,
      "step": 5160
    },
    {
      "entropy": 6.280758380889893,
      "epoch": 0.5960761684939412,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004973807644071357,
      "loss": 6.284,
      "mean_token_accuracy": 0.16869453936815262,
      "num_tokens": 13095505.0,
      "step": 5165
    },
    {
      "entropy": 6.321608257293701,
      "epoch": 0.5966532025389498,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004973744827087695,
      "loss": 6.3053,
      "mean_token_accuracy": 0.15722583681344987,
      "num_tokens": 13108351.0,
      "step": 5170
    },
    {
      "entropy": 6.45554256439209,
      "epoch": 0.5972302365839585,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000497368193530948,
      "loss": 6.3585,
      "mean_token_accuracy": 0.1610609695315361,
      "num_tokens": 13120941.0,
      "step": 5175
    },
    {
      "entropy": 6.376472330093383,
      "epoch": 0.5978072706289671,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004973618968738828,
      "loss": 6.3005,
      "mean_token_accuracy": 0.16430403739213945,
      "num_tokens": 13134217.0,
      "step": 5180
    },
    {
      "entropy": 6.34185757637024,
      "epoch": 0.5983843046739757,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004973555927377856,
      "loss": 6.3366,
      "mean_token_accuracy": 0.16254360973834991,
      "num_tokens": 13146245.0,
      "step": 5185
    },
    {
      "entropy": 6.4101691246032715,
      "epoch": 0.5989613387189844,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004973492811228685,
      "loss": 6.341,
      "mean_token_accuracy": 0.15987956374883652,
      "num_tokens": 13158498.0,
      "step": 5190
    },
    {
      "entropy": 6.381132555007935,
      "epoch": 0.599538372763993,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004973429620293438,
      "loss": 6.3767,
      "mean_token_accuracy": 0.15625806748867035,
      "num_tokens": 13170464.0,
      "step": 5195
    },
    {
      "entropy": 6.4172666549682615,
      "epoch": 0.6001154068090018,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004973366354574239,
      "loss": 6.4276,
      "mean_token_accuracy": 0.15729153752326966,
      "num_tokens": 13183315.0,
      "step": 5200
    },
    {
      "entropy": 6.4859534740448,
      "epoch": 0.6006924408540104,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004973303014073219,
      "loss": 6.4391,
      "mean_token_accuracy": 0.1567590430378914,
      "num_tokens": 13196536.0,
      "step": 5205
    },
    {
      "entropy": 6.3746849536895756,
      "epoch": 0.601269474899019,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004973239598792504,
      "loss": 6.316,
      "mean_token_accuracy": 0.15811018496751786,
      "num_tokens": 13209159.0,
      "step": 5210
    },
    {
      "entropy": 6.417537641525269,
      "epoch": 0.6018465089440277,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004973176108734232,
      "loss": 6.3353,
      "mean_token_accuracy": 0.1621304765343666,
      "num_tokens": 13221563.0,
      "step": 5215
    },
    {
      "entropy": 6.412879133224488,
      "epoch": 0.6024235429890363,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004973112543900535,
      "loss": 6.4159,
      "mean_token_accuracy": 0.15605517327785492,
      "num_tokens": 13234209.0,
      "step": 5220
    },
    {
      "entropy": 6.443006086349487,
      "epoch": 0.603000577034045,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004973048904293551,
      "loss": 6.315,
      "mean_token_accuracy": 0.15259450376033784,
      "num_tokens": 13246198.0,
      "step": 5225
    },
    {
      "entropy": 6.339201545715332,
      "epoch": 0.6035776110790536,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004972985189915422,
      "loss": 6.3209,
      "mean_token_accuracy": 0.1625668078660965,
      "num_tokens": 13259149.0,
      "step": 5230
    },
    {
      "entropy": 6.3535558700561525,
      "epoch": 0.6041546451240624,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000497292140076829,
      "loss": 6.3331,
      "mean_token_accuracy": 0.1545357458293438,
      "num_tokens": 13271138.0,
      "step": 5235
    },
    {
      "entropy": 6.374079895019531,
      "epoch": 0.604731679169071,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004972857536854301,
      "loss": 6.3338,
      "mean_token_accuracy": 0.15137291848659515,
      "num_tokens": 13283639.0,
      "step": 5240
    },
    {
      "entropy": 6.341068077087402,
      "epoch": 0.6053087132140796,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004972793598175603,
      "loss": 6.366,
      "mean_token_accuracy": 0.15684750229120253,
      "num_tokens": 13295322.0,
      "step": 5245
    },
    {
      "entropy": 6.383493041992187,
      "epoch": 0.6058857472590883,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004972729584734347,
      "loss": 6.3259,
      "mean_token_accuracy": 0.16073982119560243,
      "num_tokens": 13308431.0,
      "step": 5250
    },
    {
      "entropy": 6.338928651809693,
      "epoch": 0.6064627813040969,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004972665496532685,
      "loss": 6.2627,
      "mean_token_accuracy": 0.16160609871149062,
      "num_tokens": 13321878.0,
      "step": 5255
    },
    {
      "entropy": 6.295150327682495,
      "epoch": 0.6070398153491056,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004972601333572774,
      "loss": 6.3406,
      "mean_token_accuracy": 0.15859735310077666,
      "num_tokens": 13334166.0,
      "step": 5260
    },
    {
      "entropy": 6.505914926528931,
      "epoch": 0.6076168493941142,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004972537095856769,
      "loss": 6.4029,
      "mean_token_accuracy": 0.16045023798942565,
      "num_tokens": 13347704.0,
      "step": 5265
    },
    {
      "entropy": 6.322386932373047,
      "epoch": 0.608193883439123,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004972472783386834,
      "loss": 6.2705,
      "mean_token_accuracy": 0.16216420233249665,
      "num_tokens": 13359659.0,
      "step": 5270
    },
    {
      "entropy": 6.275593280792236,
      "epoch": 0.6087709174841316,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004972408396165132,
      "loss": 6.3174,
      "mean_token_accuracy": 0.15916675925254822,
      "num_tokens": 13371952.0,
      "step": 5275
    },
    {
      "entropy": 6.482452058792115,
      "epoch": 0.6093479515291402,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004972343934193826,
      "loss": 6.3249,
      "mean_token_accuracy": 0.1610620617866516,
      "num_tokens": 13385757.0,
      "step": 5280
    },
    {
      "entropy": 6.33489408493042,
      "epoch": 0.6099249855741489,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004972279397475086,
      "loss": 6.297,
      "mean_token_accuracy": 0.15858799815177918,
      "num_tokens": 13397657.0,
      "step": 5285
    },
    {
      "entropy": 6.291019105911255,
      "epoch": 0.6105020196191575,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004972214786011083,
      "loss": 6.3192,
      "mean_token_accuracy": 0.16493454426527024,
      "num_tokens": 13410882.0,
      "step": 5290
    },
    {
      "entropy": 6.298425388336182,
      "epoch": 0.6110790536641661,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000497215009980399,
      "loss": 6.3537,
      "mean_token_accuracy": 0.15681492537260056,
      "num_tokens": 13423310.0,
      "step": 5295
    },
    {
      "entropy": 6.420735311508179,
      "epoch": 0.6116560877091748,
      "grad_norm": 0.953125,
      "learning_rate": 0.000497208533885598,
      "loss": 6.3479,
      "mean_token_accuracy": 0.15740713924169542,
      "num_tokens": 13437837.0,
      "step": 5300
    },
    {
      "entropy": 6.49090895652771,
      "epoch": 0.6122331217541835,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004972020503169235,
      "loss": 6.3385,
      "mean_token_accuracy": 0.1601913809776306,
      "num_tokens": 13450637.0,
      "step": 5305
    },
    {
      "entropy": 6.352234268188477,
      "epoch": 0.6128101557991922,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004971955592745934,
      "loss": 6.4016,
      "mean_token_accuracy": 0.15036924481391906,
      "num_tokens": 13464429.0,
      "step": 5310
    },
    {
      "entropy": 6.36677737236023,
      "epoch": 0.6133871898442008,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000497189060758826,
      "loss": 6.2687,
      "mean_token_accuracy": 0.16355304121971131,
      "num_tokens": 13476710.0,
      "step": 5315
    },
    {
      "entropy": 6.316545057296753,
      "epoch": 0.6139642238892095,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00049718255476984,
      "loss": 6.3198,
      "mean_token_accuracy": 0.15957548469305038,
      "num_tokens": 13489828.0,
      "step": 5320
    },
    {
      "entropy": 6.383925819396973,
      "epoch": 0.6145412579342181,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004971760413078539,
      "loss": 6.3266,
      "mean_token_accuracy": 0.16073977798223496,
      "num_tokens": 13503584.0,
      "step": 5325
    },
    {
      "entropy": 6.3466229915618895,
      "epoch": 0.6151182919792267,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004971695203730872,
      "loss": 6.2913,
      "mean_token_accuracy": 0.1634427398443222,
      "num_tokens": 13514658.0,
      "step": 5330
    },
    {
      "entropy": 6.346750402450562,
      "epoch": 0.6156953260242354,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000497162991965759,
      "loss": 6.3407,
      "mean_token_accuracy": 0.1678822249174118,
      "num_tokens": 13526424.0,
      "step": 5335
    },
    {
      "entropy": 6.340521764755249,
      "epoch": 0.6162723600692441,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004971564560860889,
      "loss": 6.283,
      "mean_token_accuracy": 0.1599597528576851,
      "num_tokens": 13540515.0,
      "step": 5340
    },
    {
      "entropy": 6.410128021240235,
      "epoch": 0.6168493941142528,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004971499127342968,
      "loss": 6.3893,
      "mean_token_accuracy": 0.15507052838802338,
      "num_tokens": 13553634.0,
      "step": 5345
    },
    {
      "entropy": 6.42756404876709,
      "epoch": 0.6174264281592614,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004971433619106027,
      "loss": 6.4008,
      "mean_token_accuracy": 0.15787786096334458,
      "num_tokens": 13565168.0,
      "step": 5350
    },
    {
      "entropy": 6.352780103683472,
      "epoch": 0.61800346220427,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000497136803615227,
      "loss": 6.335,
      "mean_token_accuracy": 0.15886600613594054,
      "num_tokens": 13578511.0,
      "step": 5355
    },
    {
      "entropy": 6.345577764511108,
      "epoch": 0.6185804962492787,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004971302378483902,
      "loss": 6.2896,
      "mean_token_accuracy": 0.16366638839244843,
      "num_tokens": 13590535.0,
      "step": 5360
    },
    {
      "entropy": 6.387300443649292,
      "epoch": 0.6191575302942873,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004971236646103132,
      "loss": 6.3505,
      "mean_token_accuracy": 0.15835938602685928,
      "num_tokens": 13604549.0,
      "step": 5365
    },
    {
      "entropy": 6.357629632949829,
      "epoch": 0.619734564339296,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004971170839012171,
      "loss": 6.3247,
      "mean_token_accuracy": 0.15468093603849412,
      "num_tokens": 13617658.0,
      "step": 5370
    },
    {
      "entropy": 6.376674747467041,
      "epoch": 0.6203115983843047,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004971104957213233,
      "loss": 6.2588,
      "mean_token_accuracy": 0.16765327602624894,
      "num_tokens": 13630559.0,
      "step": 5375
    },
    {
      "entropy": 6.319404983520508,
      "epoch": 0.6208886324293134,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004971039000708531,
      "loss": 6.3043,
      "mean_token_accuracy": 0.16685150116682052,
      "num_tokens": 13643018.0,
      "step": 5380
    },
    {
      "entropy": 6.375686836242676,
      "epoch": 0.621465666474322,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004970972969500286,
      "loss": 6.2984,
      "mean_token_accuracy": 0.1589001879096031,
      "num_tokens": 13655053.0,
      "step": 5385
    },
    {
      "entropy": 6.350291776657104,
      "epoch": 0.6220427005193306,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000497090686359072,
      "loss": 6.2662,
      "mean_token_accuracy": 0.16407257318496704,
      "num_tokens": 13667420.0,
      "step": 5390
    },
    {
      "entropy": 6.299328756332398,
      "epoch": 0.6226197345643393,
      "grad_norm": 0.875,
      "learning_rate": 0.0004970840682982054,
      "loss": 6.2274,
      "mean_token_accuracy": 0.16318354606628419,
      "num_tokens": 13680662.0,
      "step": 5395
    },
    {
      "entropy": 6.346205329895019,
      "epoch": 0.6231967686093479,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004970774427676514,
      "loss": 6.3367,
      "mean_token_accuracy": 0.15316741168498993,
      "num_tokens": 13694720.0,
      "step": 5400
    },
    {
      "entropy": 6.360504245758056,
      "epoch": 0.6237738026543566,
      "grad_norm": 1.0,
      "learning_rate": 0.000497070809767633,
      "loss": 6.2742,
      "mean_token_accuracy": 0.15734955817461013,
      "num_tokens": 13706826.0,
      "step": 5405
    },
    {
      "entropy": 6.322917318344116,
      "epoch": 0.6243508366993653,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004970641692983731,
      "loss": 6.3289,
      "mean_token_accuracy": 0.15646012127399445,
      "num_tokens": 13719526.0,
      "step": 5410
    },
    {
      "entropy": 6.348106336593628,
      "epoch": 0.624927870744374,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004970575213600954,
      "loss": 6.2264,
      "mean_token_accuracy": 0.16836631298065186,
      "num_tokens": 13731583.0,
      "step": 5415
    },
    {
      "entropy": 6.3745146751403805,
      "epoch": 0.6255049047893826,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004970508659530231,
      "loss": 6.3378,
      "mean_token_accuracy": 0.16159322559833528,
      "num_tokens": 13743690.0,
      "step": 5420
    },
    {
      "entropy": 6.348368406295776,
      "epoch": 0.6260819388343912,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004970442030773802,
      "loss": 6.2932,
      "mean_token_accuracy": 0.15742024034261703,
      "num_tokens": 13756264.0,
      "step": 5425
    },
    {
      "entropy": 6.309341907501221,
      "epoch": 0.6266589728793999,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004970375327333909,
      "loss": 6.3165,
      "mean_token_accuracy": 0.15393718481063842,
      "num_tokens": 13768089.0,
      "step": 5430
    },
    {
      "entropy": 6.4129860401153564,
      "epoch": 0.6272360069244085,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004970308549212796,
      "loss": 6.3376,
      "mean_token_accuracy": 0.15792135149240494,
      "num_tokens": 13780671.0,
      "step": 5435
    },
    {
      "entropy": 6.446494483947754,
      "epoch": 0.6278130409694171,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004970241696412705,
      "loss": 6.3463,
      "mean_token_accuracy": 0.15591817051172258,
      "num_tokens": 13793678.0,
      "step": 5440
    },
    {
      "entropy": 6.344524288177491,
      "epoch": 0.6283900750144259,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000497017476893589,
      "loss": 6.2853,
      "mean_token_accuracy": 0.15855076313018798,
      "num_tokens": 13807001.0,
      "step": 5445
    },
    {
      "entropy": 6.336147451400757,
      "epoch": 0.6289671090594345,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004970107766784598,
      "loss": 6.2561,
      "mean_token_accuracy": 0.16313763558864594,
      "num_tokens": 13820227.0,
      "step": 5450
    },
    {
      "entropy": 6.3735105991363525,
      "epoch": 0.6295441431044432,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004970040689961084,
      "loss": 6.351,
      "mean_token_accuracy": 0.1633874371647835,
      "num_tokens": 13831663.0,
      "step": 5455
    },
    {
      "entropy": 6.339467096328735,
      "epoch": 0.6301211771494518,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004969973538467605,
      "loss": 6.2552,
      "mean_token_accuracy": 0.16474179476499556,
      "num_tokens": 13845173.0,
      "step": 5460
    },
    {
      "entropy": 6.36610860824585,
      "epoch": 0.6306982111944605,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004969906312306419,
      "loss": 6.2917,
      "mean_token_accuracy": 0.15667524933815002,
      "num_tokens": 13857551.0,
      "step": 5465
    },
    {
      "entropy": 6.200679540634155,
      "epoch": 0.6312752452394691,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004969839011479786,
      "loss": 6.2073,
      "mean_token_accuracy": 0.16565943658351898,
      "num_tokens": 13870238.0,
      "step": 5470
    },
    {
      "entropy": 6.311519002914428,
      "epoch": 0.6318522792844777,
      "grad_norm": 0.953125,
      "learning_rate": 0.000496977163598997,
      "loss": 6.2938,
      "mean_token_accuracy": 0.1564420208334923,
      "num_tokens": 13883145.0,
      "step": 5475
    },
    {
      "entropy": 6.425129699707031,
      "epoch": 0.6324293133294865,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004969704185839239,
      "loss": 6.349,
      "mean_token_accuracy": 0.15295332670211792,
      "num_tokens": 13895777.0,
      "step": 5480
    },
    {
      "entropy": 6.326649188995361,
      "epoch": 0.6330063473744951,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004969636661029859,
      "loss": 6.2376,
      "mean_token_accuracy": 0.17034487128257753,
      "num_tokens": 13907641.0,
      "step": 5485
    },
    {
      "entropy": 6.359035539627075,
      "epoch": 0.6335833814195038,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004969569061564103,
      "loss": 6.3502,
      "mean_token_accuracy": 0.15359617322683333,
      "num_tokens": 13921576.0,
      "step": 5490
    },
    {
      "entropy": 6.381617069244385,
      "epoch": 0.6341604154645124,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004969501387444245,
      "loss": 6.2842,
      "mean_token_accuracy": 0.15700178891420363,
      "num_tokens": 13935158.0,
      "step": 5495
    },
    {
      "entropy": 6.271814680099487,
      "epoch": 0.634737449509521,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004969433638672559,
      "loss": 6.2346,
      "mean_token_accuracy": 0.16232349276542662,
      "num_tokens": 13947803.0,
      "step": 5500
    },
    {
      "entropy": 6.373715782165528,
      "epoch": 0.6353144835545297,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004969365815251325,
      "loss": 6.3264,
      "mean_token_accuracy": 0.16024259328842164,
      "num_tokens": 13960573.0,
      "step": 5505
    },
    {
      "entropy": 6.345321321487427,
      "epoch": 0.6358915175995383,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004969297917182825,
      "loss": 6.1923,
      "mean_token_accuracy": 0.16617179661989212,
      "num_tokens": 13970994.0,
      "step": 5510
    },
    {
      "entropy": 6.37383542060852,
      "epoch": 0.6364685516445471,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004969229944469342,
      "loss": 6.41,
      "mean_token_accuracy": 0.15260985046625136,
      "num_tokens": 13983617.0,
      "step": 5515
    },
    {
      "entropy": 6.4177939891815186,
      "epoch": 0.6370455856895557,
      "grad_norm": 1.0,
      "learning_rate": 0.0004969161897113162,
      "loss": 6.3077,
      "mean_token_accuracy": 0.161920066177845,
      "num_tokens": 13995779.0,
      "step": 5520
    },
    {
      "entropy": 6.3209868431091305,
      "epoch": 0.6376226197345644,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004969093775116574,
      "loss": 6.3421,
      "mean_token_accuracy": 0.1608972117304802,
      "num_tokens": 14008650.0,
      "step": 5525
    },
    {
      "entropy": 6.283117008209229,
      "epoch": 0.638199653779573,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004969025578481869,
      "loss": 6.2479,
      "mean_token_accuracy": 0.16362528949975969,
      "num_tokens": 14021787.0,
      "step": 5530
    },
    {
      "entropy": 6.4909834384918215,
      "epoch": 0.6387766878245816,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000496895730721134,
      "loss": 6.3977,
      "mean_token_accuracy": 0.14869729578495025,
      "num_tokens": 14035036.0,
      "step": 5535
    },
    {
      "entropy": 6.414866399765015,
      "epoch": 0.6393537218695903,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004968888961307286,
      "loss": 6.3352,
      "mean_token_accuracy": 0.15972763746976854,
      "num_tokens": 14048589.0,
      "step": 5540
    },
    {
      "entropy": 6.356158113479614,
      "epoch": 0.6399307559145989,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004968820540772003,
      "loss": 6.2264,
      "mean_token_accuracy": 0.18150741159915923,
      "num_tokens": 14061007.0,
      "step": 5545
    },
    {
      "entropy": 6.408500719070434,
      "epoch": 0.6405077899596077,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004968752045607794,
      "loss": 6.3574,
      "mean_token_accuracy": 0.15625543296337127,
      "num_tokens": 14074165.0,
      "step": 5550
    },
    {
      "entropy": 6.507462358474731,
      "epoch": 0.6410848240046163,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004968683475816961,
      "loss": 6.4066,
      "mean_token_accuracy": 0.1469891130924225,
      "num_tokens": 14086261.0,
      "step": 5555
    },
    {
      "entropy": 6.291381883621216,
      "epoch": 0.641661858049625,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004968614831401811,
      "loss": 6.2536,
      "mean_token_accuracy": 0.17122525423765184,
      "num_tokens": 14098458.0,
      "step": 5560
    },
    {
      "entropy": 6.282268095016479,
      "epoch": 0.6422388920946336,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004968546112364654,
      "loss": 6.4191,
      "mean_token_accuracy": 0.15233878493309022,
      "num_tokens": 14110754.0,
      "step": 5565
    },
    {
      "entropy": 6.441370582580566,
      "epoch": 0.6428159261396422,
      "grad_norm": 1.03125,
      "learning_rate": 0.00049684773187078,
      "loss": 6.2984,
      "mean_token_accuracy": 0.15455365926027298,
      "num_tokens": 14123361.0,
      "step": 5570
    },
    {
      "entropy": 6.333598804473877,
      "epoch": 0.6433929601846509,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004968408450433565,
      "loss": 6.2576,
      "mean_token_accuracy": 0.1661595180630684,
      "num_tokens": 14135576.0,
      "step": 5575
    },
    {
      "entropy": 6.349913930892944,
      "epoch": 0.6439699942296595,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004968339507544263,
      "loss": 6.327,
      "mean_token_accuracy": 0.16096770465373994,
      "num_tokens": 14147879.0,
      "step": 5580
    },
    {
      "entropy": 6.413717365264892,
      "epoch": 0.6445470282746683,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004968270490042212,
      "loss": 6.369,
      "mean_token_accuracy": 0.15601970106363297,
      "num_tokens": 14159449.0,
      "step": 5585
    },
    {
      "entropy": 6.285684204101562,
      "epoch": 0.6451240623196769,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004968201397929737,
      "loss": 6.2344,
      "mean_token_accuracy": 0.16120658218860626,
      "num_tokens": 14172547.0,
      "step": 5590
    },
    {
      "entropy": 6.374817132949829,
      "epoch": 0.6457010963646855,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004968132231209158,
      "loss": 6.2949,
      "mean_token_accuracy": 0.16361537724733352,
      "num_tokens": 14184670.0,
      "step": 5595
    },
    {
      "entropy": 6.375036382675171,
      "epoch": 0.6462781304096942,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004968062989882803,
      "loss": 6.2837,
      "mean_token_accuracy": 0.16073961257934571,
      "num_tokens": 14198048.0,
      "step": 5600
    },
    {
      "entropy": 6.327577066421509,
      "epoch": 0.6468551644547028,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004967993673953003,
      "loss": 6.2505,
      "mean_token_accuracy": 0.16245031356811523,
      "num_tokens": 14209470.0,
      "step": 5605
    },
    {
      "entropy": 6.348019647598266,
      "epoch": 0.6474321984997115,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004967924283422087,
      "loss": 6.2622,
      "mean_token_accuracy": 0.1630554586648941,
      "num_tokens": 14221731.0,
      "step": 5610
    },
    {
      "entropy": 6.4256274700164795,
      "epoch": 0.6480092325447201,
      "grad_norm": 1.0,
      "learning_rate": 0.000496785481829239,
      "loss": 6.3198,
      "mean_token_accuracy": 0.15379041880369188,
      "num_tokens": 14235125.0,
      "step": 5615
    },
    {
      "entropy": 6.327840042114258,
      "epoch": 0.6485862665897288,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004967785278566245,
      "loss": 6.2797,
      "mean_token_accuracy": 0.164201720058918,
      "num_tokens": 14248941.0,
      "step": 5620
    },
    {
      "entropy": 6.322573518753051,
      "epoch": 0.6491633006347375,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004967715664245997,
      "loss": 6.2022,
      "mean_token_accuracy": 0.16699752658605577,
      "num_tokens": 14260955.0,
      "step": 5625
    },
    {
      "entropy": 6.285229063034057,
      "epoch": 0.6497403346797461,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004967645975333983,
      "loss": 6.2184,
      "mean_token_accuracy": 0.17027618139982223,
      "num_tokens": 14274123.0,
      "step": 5630
    },
    {
      "entropy": 6.371972560882568,
      "epoch": 0.6503173687247548,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004967576211832548,
      "loss": 6.2849,
      "mean_token_accuracy": 0.1675660490989685,
      "num_tokens": 14286920.0,
      "step": 5635
    },
    {
      "entropy": 6.358508539199829,
      "epoch": 0.6508944027697634,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004967506373744039,
      "loss": 6.2856,
      "mean_token_accuracy": 0.15663084164261817,
      "num_tokens": 14299297.0,
      "step": 5640
    },
    {
      "entropy": 6.378447341918945,
      "epoch": 0.651471436814772,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004967436461070805,
      "loss": 6.2818,
      "mean_token_accuracy": 0.16298537999391555,
      "num_tokens": 14312107.0,
      "step": 5645
    },
    {
      "entropy": 6.32569990158081,
      "epoch": 0.6520484708597807,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004967366473815196,
      "loss": 6.2829,
      "mean_token_accuracy": 0.15967210233211518,
      "num_tokens": 14325777.0,
      "step": 5650
    },
    {
      "entropy": 6.362534809112549,
      "epoch": 0.6526255049047894,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004967296411979568,
      "loss": 6.2874,
      "mean_token_accuracy": 0.15898593962192537,
      "num_tokens": 14337617.0,
      "step": 5655
    },
    {
      "entropy": 6.287278366088867,
      "epoch": 0.6532025389497981,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004967226275566275,
      "loss": 6.2351,
      "mean_token_accuracy": 0.16494760364294053,
      "num_tokens": 14349756.0,
      "step": 5660
    },
    {
      "entropy": 6.388342428207397,
      "epoch": 0.6537795729948067,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000496715606457768,
      "loss": 6.3315,
      "mean_token_accuracy": 0.1591486856341362,
      "num_tokens": 14363358.0,
      "step": 5665
    },
    {
      "entropy": 6.322444152832031,
      "epoch": 0.6543566070398154,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004967085779016142,
      "loss": 6.2908,
      "mean_token_accuracy": 0.16137947589159013,
      "num_tokens": 14375510.0,
      "step": 5670
    },
    {
      "entropy": 6.3729493618011475,
      "epoch": 0.654933641084824,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004967015418884022,
      "loss": 6.2923,
      "mean_token_accuracy": 0.16366371214389802,
      "num_tokens": 14389523.0,
      "step": 5675
    },
    {
      "entropy": 6.3052726745605465,
      "epoch": 0.6555106751298326,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004966944984183692,
      "loss": 6.2971,
      "mean_token_accuracy": 0.16051067858934404,
      "num_tokens": 14400457.0,
      "step": 5680
    },
    {
      "entropy": 6.246604776382446,
      "epoch": 0.6560877091748413,
      "grad_norm": 1.0,
      "learning_rate": 0.0004966874474917518,
      "loss": 6.2438,
      "mean_token_accuracy": 0.1678729847073555,
      "num_tokens": 14413088.0,
      "step": 5685
    },
    {
      "entropy": 6.45229139328003,
      "epoch": 0.65666474321985,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004966803891087873,
      "loss": 6.2931,
      "mean_token_accuracy": 0.15690880119800568,
      "num_tokens": 14425816.0,
      "step": 5690
    },
    {
      "entropy": 6.211046266555786,
      "epoch": 0.6572417772648587,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000496673323269713,
      "loss": 6.1869,
      "mean_token_accuracy": 0.1689558207988739,
      "num_tokens": 14438739.0,
      "step": 5695
    },
    {
      "entropy": 6.314014482498169,
      "epoch": 0.6578188113098673,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004966662499747666,
      "loss": 6.3215,
      "mean_token_accuracy": 0.15810210704803468,
      "num_tokens": 14451543.0,
      "step": 5700
    },
    {
      "entropy": 6.454266500473023,
      "epoch": 0.6583958453548759,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004966591692241859,
      "loss": 6.3404,
      "mean_token_accuracy": 0.15735524743795395,
      "num_tokens": 14464969.0,
      "step": 5705
    },
    {
      "entropy": 6.323182249069214,
      "epoch": 0.6589728793998846,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004966520810182092,
      "loss": 6.3107,
      "mean_token_accuracy": 0.15808846056461334,
      "num_tokens": 14477622.0,
      "step": 5710
    },
    {
      "entropy": 6.42166223526001,
      "epoch": 0.6595499134448932,
      "grad_norm": 1.0,
      "learning_rate": 0.0004966449853570749,
      "loss": 6.3078,
      "mean_token_accuracy": 0.15966509729623796,
      "num_tokens": 14490306.0,
      "step": 5715
    },
    {
      "entropy": 6.2887489795684814,
      "epoch": 0.6601269474899019,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004966378822410215,
      "loss": 6.1672,
      "mean_token_accuracy": 0.1797989562153816,
      "num_tokens": 14501701.0,
      "step": 5720
    },
    {
      "entropy": 6.269540071487427,
      "epoch": 0.6607039815349106,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004966307716702881,
      "loss": 6.2965,
      "mean_token_accuracy": 0.16270702928304673,
      "num_tokens": 14515700.0,
      "step": 5725
    },
    {
      "entropy": 6.3525707721710205,
      "epoch": 0.6612810155799193,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004966236536451138,
      "loss": 6.2695,
      "mean_token_accuracy": 0.15902794599533082,
      "num_tokens": 14528583.0,
      "step": 5730
    },
    {
      "entropy": 6.279529047012329,
      "epoch": 0.6618580496249279,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000496616528165738,
      "loss": 6.2428,
      "mean_token_accuracy": 0.16624296307563782,
      "num_tokens": 14541896.0,
      "step": 5735
    },
    {
      "entropy": 6.268627738952636,
      "epoch": 0.6624350836699365,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004966093952324003,
      "loss": 6.1636,
      "mean_token_accuracy": 0.16777980476617813,
      "num_tokens": 14556948.0,
      "step": 5740
    },
    {
      "entropy": 6.322888135910034,
      "epoch": 0.6630121177149452,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004966022548453406,
      "loss": 6.3422,
      "mean_token_accuracy": 0.15742336809635163,
      "num_tokens": 14568802.0,
      "step": 5745
    },
    {
      "entropy": 6.374240493774414,
      "epoch": 0.6635891517599538,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004965951070047993,
      "loss": 6.2803,
      "mean_token_accuracy": 0.15682113766670228,
      "num_tokens": 14582664.0,
      "step": 5750
    },
    {
      "entropy": 6.322626781463623,
      "epoch": 0.6641661858049625,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004965879517110166,
      "loss": 6.3689,
      "mean_token_accuracy": 0.15610907673835756,
      "num_tokens": 14596805.0,
      "step": 5755
    },
    {
      "entropy": 6.333931922912598,
      "epoch": 0.6647432198499712,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004965807889642333,
      "loss": 6.3037,
      "mean_token_accuracy": 0.15626824349164964,
      "num_tokens": 14610532.0,
      "step": 5760
    },
    {
      "entropy": 6.3743157386779785,
      "epoch": 0.6653202538949798,
      "grad_norm": 0.984375,
      "learning_rate": 0.00049657361876469,
      "loss": 6.304,
      "mean_token_accuracy": 0.1579776421189308,
      "num_tokens": 14622928.0,
      "step": 5765
    },
    {
      "entropy": 6.2759346008300785,
      "epoch": 0.6658972879399885,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004965664411126282,
      "loss": 6.1941,
      "mean_token_accuracy": 0.16473535001277922,
      "num_tokens": 14634410.0,
      "step": 5770
    },
    {
      "entropy": 6.2235535144805905,
      "epoch": 0.6664743219849971,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004965592560082894,
      "loss": 6.2146,
      "mean_token_accuracy": 0.1643059641122818,
      "num_tokens": 14647571.0,
      "step": 5775
    },
    {
      "entropy": 6.275019311904908,
      "epoch": 0.6670513560300058,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004965520634519149,
      "loss": 6.1087,
      "mean_token_accuracy": 0.18141666501760484,
      "num_tokens": 14660656.0,
      "step": 5780
    },
    {
      "entropy": 6.449819564819336,
      "epoch": 0.6676283900750144,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004965448634437468,
      "loss": 6.3278,
      "mean_token_accuracy": 0.15872610807418824,
      "num_tokens": 14672267.0,
      "step": 5785
    },
    {
      "entropy": 6.276329278945923,
      "epoch": 0.668205424120023,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004965376559840272,
      "loss": 6.3068,
      "mean_token_accuracy": 0.157090200483799,
      "num_tokens": 14684590.0,
      "step": 5790
    },
    {
      "entropy": 6.366933584213257,
      "epoch": 0.6687824581650318,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004965304410729986,
      "loss": 6.3106,
      "mean_token_accuracy": 0.1580670401453972,
      "num_tokens": 14697299.0,
      "step": 5795
    },
    {
      "entropy": 6.33537712097168,
      "epoch": 0.6693594922100404,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004965232187109037,
      "loss": 6.2802,
      "mean_token_accuracy": 0.16264077872037888,
      "num_tokens": 14709328.0,
      "step": 5800
    },
    {
      "entropy": 6.304199457168579,
      "epoch": 0.6699365262550491,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004965159888979854,
      "loss": 6.2574,
      "mean_token_accuracy": 0.164343424141407,
      "num_tokens": 14721786.0,
      "step": 5805
    },
    {
      "entropy": 6.37620587348938,
      "epoch": 0.6705135603000577,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004965087516344869,
      "loss": 6.34,
      "mean_token_accuracy": 0.15894667953252792,
      "num_tokens": 14734120.0,
      "step": 5810
    },
    {
      "entropy": 6.348331451416016,
      "epoch": 0.6710905943450663,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004965015069206515,
      "loss": 6.3474,
      "mean_token_accuracy": 0.15966461598873138,
      "num_tokens": 14746159.0,
      "step": 5815
    },
    {
      "entropy": 6.4267627716064455,
      "epoch": 0.671667628390075,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000496494254756723,
      "loss": 6.2811,
      "mean_token_accuracy": 0.16151547953486442,
      "num_tokens": 14758132.0,
      "step": 5820
    },
    {
      "entropy": 6.2220992088317875,
      "epoch": 0.6722446624350836,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004964869951429451,
      "loss": 6.1678,
      "mean_token_accuracy": 0.16763481944799424,
      "num_tokens": 14770934.0,
      "step": 5825
    },
    {
      "entropy": 6.302060794830322,
      "epoch": 0.6728216964800924,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004964797280795622,
      "loss": 6.3176,
      "mean_token_accuracy": 0.15710745006799698,
      "num_tokens": 14785084.0,
      "step": 5830
    },
    {
      "entropy": 6.400588703155518,
      "epoch": 0.673398730525101,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004964724535668188,
      "loss": 6.3686,
      "mean_token_accuracy": 0.16307279765605925,
      "num_tokens": 14798553.0,
      "step": 5835
    },
    {
      "entropy": 6.395945835113525,
      "epoch": 0.6739757645701097,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004964651716049593,
      "loss": 6.321,
      "mean_token_accuracy": 0.15653637498617173,
      "num_tokens": 14810239.0,
      "step": 5840
    },
    {
      "entropy": 6.374822330474854,
      "epoch": 0.6745527986151183,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004964578821942288,
      "loss": 6.2856,
      "mean_token_accuracy": 0.16032001823186875,
      "num_tokens": 14822704.0,
      "step": 5845
    },
    {
      "entropy": 6.336796617507934,
      "epoch": 0.6751298326601269,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004964505853348724,
      "loss": 6.3266,
      "mean_token_accuracy": 0.16033909022808074,
      "num_tokens": 14836913.0,
      "step": 5850
    },
    {
      "entropy": 6.325709247589112,
      "epoch": 0.6757068667051356,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004964432810271356,
      "loss": 6.2718,
      "mean_token_accuracy": 0.16571087688207625,
      "num_tokens": 14849697.0,
      "step": 5855
    },
    {
      "entropy": 6.2545506954193115,
      "epoch": 0.6762839007501442,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004964359692712641,
      "loss": 6.2446,
      "mean_token_accuracy": 0.1606735423207283,
      "num_tokens": 14861719.0,
      "step": 5860
    },
    {
      "entropy": 6.346668863296509,
      "epoch": 0.676860934795153,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004964286500675037,
      "loss": 6.2715,
      "mean_token_accuracy": 0.16198989748954773,
      "num_tokens": 14873786.0,
      "step": 5865
    },
    {
      "entropy": 6.363369035720825,
      "epoch": 0.6774379688401616,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004964213234161007,
      "loss": 6.2047,
      "mean_token_accuracy": 0.16385551393032075,
      "num_tokens": 14888952.0,
      "step": 5870
    },
    {
      "entropy": 6.399168872833252,
      "epoch": 0.6780150028851702,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004964139893173014,
      "loss": 6.3668,
      "mean_token_accuracy": 0.15990081429481506,
      "num_tokens": 14903039.0,
      "step": 5875
    },
    {
      "entropy": 6.311770677566528,
      "epoch": 0.6785920369301789,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004964066477713525,
      "loss": 6.3108,
      "mean_token_accuracy": 0.15290780514478683,
      "num_tokens": 14915949.0,
      "step": 5880
    },
    {
      "entropy": 6.3547607421875,
      "epoch": 0.6791690709751875,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004963992987785011,
      "loss": 6.366,
      "mean_token_accuracy": 0.1586316168308258,
      "num_tokens": 14928724.0,
      "step": 5885
    },
    {
      "entropy": 6.392857503890991,
      "epoch": 0.6797461050201962,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004963919423389942,
      "loss": 6.2635,
      "mean_token_accuracy": 0.16089897006750106,
      "num_tokens": 14941759.0,
      "step": 5890
    },
    {
      "entropy": 6.262146615982056,
      "epoch": 0.6803231390652048,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004963845784530794,
      "loss": 6.3146,
      "mean_token_accuracy": 0.15942233800888062,
      "num_tokens": 14954373.0,
      "step": 5895
    },
    {
      "entropy": 6.341342258453369,
      "epoch": 0.6809001731102136,
      "grad_norm": 0.921875,
      "learning_rate": 0.000496377207121004,
      "loss": 6.1866,
      "mean_token_accuracy": 0.17055265456438065,
      "num_tokens": 14966195.0,
      "step": 5900
    },
    {
      "entropy": 6.316828107833862,
      "epoch": 0.6814772071552222,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004963698283430164,
      "loss": 6.2748,
      "mean_token_accuracy": 0.1601161077618599,
      "num_tokens": 14979594.0,
      "step": 5905
    },
    {
      "entropy": 6.359366750717163,
      "epoch": 0.6820542412002308,
      "grad_norm": 1.0,
      "learning_rate": 0.0004963624421193646,
      "loss": 6.3482,
      "mean_token_accuracy": 0.15728830099105834,
      "num_tokens": 14993215.0,
      "step": 5910
    },
    {
      "entropy": 6.438364315032959,
      "epoch": 0.6826312752452395,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000496355048450297,
      "loss": 6.3213,
      "mean_token_accuracy": 0.15949019491672517,
      "num_tokens": 15006241.0,
      "step": 5915
    },
    {
      "entropy": 6.289816570281983,
      "epoch": 0.6832083092902481,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004963476473360623,
      "loss": 6.2857,
      "mean_token_accuracy": 0.15822021663188934,
      "num_tokens": 15018750.0,
      "step": 5920
    },
    {
      "entropy": 6.3044578552246096,
      "epoch": 0.6837853433352568,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004963402387769094,
      "loss": 6.2432,
      "mean_token_accuracy": 0.16088712811470032,
      "num_tokens": 15031065.0,
      "step": 5925
    },
    {
      "entropy": 6.382319450378418,
      "epoch": 0.6843623773802654,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004963328227730876,
      "loss": 6.2538,
      "mean_token_accuracy": 0.16127691864967347,
      "num_tokens": 15043129.0,
      "step": 5930
    },
    {
      "entropy": 6.313252925872803,
      "epoch": 0.684939411425274,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004963253993248461,
      "loss": 6.2608,
      "mean_token_accuracy": 0.16144487857818604,
      "num_tokens": 15057531.0,
      "step": 5935
    },
    {
      "entropy": 6.3104087829589846,
      "epoch": 0.6855164454702828,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004963179684324349,
      "loss": 6.3213,
      "mean_token_accuracy": 0.15617611408233642,
      "num_tokens": 15070105.0,
      "step": 5940
    },
    {
      "entropy": 6.379939031600952,
      "epoch": 0.6860934795152914,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004963105300961036,
      "loss": 6.3039,
      "mean_token_accuracy": 0.1571717992424965,
      "num_tokens": 15082482.0,
      "step": 5945
    },
    {
      "entropy": 6.342560482025147,
      "epoch": 0.6866705135603001,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004963030843161026,
      "loss": 6.3272,
      "mean_token_accuracy": 0.1526414841413498,
      "num_tokens": 15095564.0,
      "step": 5950
    },
    {
      "entropy": 6.442060327529907,
      "epoch": 0.6872475476053087,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004962956310926823,
      "loss": 6.3231,
      "mean_token_accuracy": 0.1559424176812172,
      "num_tokens": 15108948.0,
      "step": 5955
    },
    {
      "entropy": 6.272290563583374,
      "epoch": 0.6878245816503173,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004962881704260934,
      "loss": 6.2254,
      "mean_token_accuracy": 0.16100811660289766,
      "num_tokens": 15121537.0,
      "step": 5960
    },
    {
      "entropy": 6.251233100891113,
      "epoch": 0.688401615695326,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004962807023165868,
      "loss": 6.2425,
      "mean_token_accuracy": 0.16593978255987168,
      "num_tokens": 15134657.0,
      "step": 5965
    },
    {
      "entropy": 6.326472616195678,
      "epoch": 0.6889786497403346,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004962732267644138,
      "loss": 6.2095,
      "mean_token_accuracy": 0.16449764519929885,
      "num_tokens": 15146301.0,
      "step": 5970
    },
    {
      "entropy": 6.32165675163269,
      "epoch": 0.6895556837853434,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004962657437698254,
      "loss": 6.3095,
      "mean_token_accuracy": 0.15792420208454133,
      "num_tokens": 15157886.0,
      "step": 5975
    },
    {
      "entropy": 6.397644996643066,
      "epoch": 0.690132717830352,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004962582533330739,
      "loss": 6.2621,
      "mean_token_accuracy": 0.16016658246517182,
      "num_tokens": 15170457.0,
      "step": 5980
    },
    {
      "entropy": 6.30787353515625,
      "epoch": 0.6907097518753607,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004962507554544109,
      "loss": 6.3003,
      "mean_token_accuracy": 0.15999116152524948,
      "num_tokens": 15182504.0,
      "step": 5985
    },
    {
      "entropy": 6.354946184158325,
      "epoch": 0.6912867859203693,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004962432501340886,
      "loss": 6.2951,
      "mean_token_accuracy": 0.1576322704553604,
      "num_tokens": 15194809.0,
      "step": 5990
    },
    {
      "entropy": 6.393294954299927,
      "epoch": 0.6918638199653779,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004962357373723596,
      "loss": 6.2499,
      "mean_token_accuracy": 0.15967389643192292,
      "num_tokens": 15207185.0,
      "step": 5995
    },
    {
      "entropy": 6.320103645324707,
      "epoch": 0.6924408540103866,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004962282171694763,
      "loss": 6.2767,
      "mean_token_accuracy": 0.16086386442184447,
      "num_tokens": 15220406.0,
      "step": 6000
    },
    {
      "epoch": 0.6924408540103866,
      "eval_entropy": 6.151910295329176,
      "eval_loss": 6.343685626983643,
      "eval_mean_token_accuracy": 0.16159232160543088,
      "eval_num_tokens": 15220406.0,
      "eval_runtime": 22.7328,
      "eval_samples_per_second": 1237.684,
      "eval_steps_per_second": 154.71,
      "step": 6000
    },
    {
      "entropy": 6.3117852210998535,
      "epoch": 0.6930178880553952,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004962206895256919,
      "loss": 6.3232,
      "mean_token_accuracy": 0.15651306807994841,
      "num_tokens": 15231698.0,
      "step": 6005
    },
    {
      "entropy": 6.413022518157959,
      "epoch": 0.693594922100404,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004962131544412595,
      "loss": 6.3405,
      "mean_token_accuracy": 0.1551682397723198,
      "num_tokens": 15243462.0,
      "step": 6010
    },
    {
      "entropy": 6.42551736831665,
      "epoch": 0.6941719561454126,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004962056119164325,
      "loss": 6.3376,
      "mean_token_accuracy": 0.15194634348154068,
      "num_tokens": 15257311.0,
      "step": 6015
    },
    {
      "entropy": 6.30859055519104,
      "epoch": 0.6947489901904212,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004961980619514646,
      "loss": 6.3275,
      "mean_token_accuracy": 0.15747295171022416,
      "num_tokens": 15270186.0,
      "step": 6020
    },
    {
      "entropy": 6.320661115646362,
      "epoch": 0.6953260242354299,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004961905045466098,
      "loss": 6.2651,
      "mean_token_accuracy": 0.15744795501232148,
      "num_tokens": 15282573.0,
      "step": 6025
    },
    {
      "entropy": 6.3932088851928714,
      "epoch": 0.6959030582804385,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004961829397021222,
      "loss": 6.3098,
      "mean_token_accuracy": 0.1589531883597374,
      "num_tokens": 15296721.0,
      "step": 6030
    },
    {
      "entropy": 6.350703859329224,
      "epoch": 0.6964800923254472,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004961753674182564,
      "loss": 6.2821,
      "mean_token_accuracy": 0.1661172926425934,
      "num_tokens": 15308458.0,
      "step": 6035
    },
    {
      "entropy": 6.310303258895874,
      "epoch": 0.6970571263704558,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004961677876952669,
      "loss": 6.29,
      "mean_token_accuracy": 0.15591613054275513,
      "num_tokens": 15320402.0,
      "step": 6040
    },
    {
      "entropy": 6.392130136489868,
      "epoch": 0.6976341604154646,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004961602005334087,
      "loss": 6.3308,
      "mean_token_accuracy": 0.15934121757745742,
      "num_tokens": 15333196.0,
      "step": 6045
    },
    {
      "entropy": 6.262461423873901,
      "epoch": 0.6982111944604732,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000496152605932937,
      "loss": 6.273,
      "mean_token_accuracy": 0.1596505254507065,
      "num_tokens": 15345216.0,
      "step": 6050
    },
    {
      "entropy": 6.375197076797486,
      "epoch": 0.6987882285054818,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004961450038941074,
      "loss": 6.328,
      "mean_token_accuracy": 0.15896058529615403,
      "num_tokens": 15357772.0,
      "step": 6055
    },
    {
      "entropy": 6.341615486145019,
      "epoch": 0.6993652625504905,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004961373944171754,
      "loss": 6.2258,
      "mean_token_accuracy": 0.16276210695505142,
      "num_tokens": 15371110.0,
      "step": 6060
    },
    {
      "entropy": 6.201347017288208,
      "epoch": 0.6999422965954991,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004961297775023968,
      "loss": 6.2187,
      "mean_token_accuracy": 0.16800244599580766,
      "num_tokens": 15383731.0,
      "step": 6065
    },
    {
      "entropy": 6.360866641998291,
      "epoch": 0.7005193306405078,
      "grad_norm": 1.015625,
      "learning_rate": 0.000496122153150028,
      "loss": 6.2769,
      "mean_token_accuracy": 0.15956784337759017,
      "num_tokens": 15395849.0,
      "step": 6070
    },
    {
      "entropy": 6.349894857406616,
      "epoch": 0.7010963646855164,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004961145213603255,
      "loss": 6.2268,
      "mean_token_accuracy": 0.16650779247283937,
      "num_tokens": 15408552.0,
      "step": 6075
    },
    {
      "entropy": 6.302218294143676,
      "epoch": 0.7016733987305251,
      "grad_norm": 1.546875,
      "learning_rate": 0.000496106882133546,
      "loss": 6.3159,
      "mean_token_accuracy": 0.16129084676504135,
      "num_tokens": 15421940.0,
      "step": 6080
    },
    {
      "entropy": 6.302323484420777,
      "epoch": 0.7022504327755338,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004960992354699463,
      "loss": 6.2773,
      "mean_token_accuracy": 0.15871342569589614,
      "num_tokens": 15434232.0,
      "step": 6085
    },
    {
      "entropy": 6.385793209075928,
      "epoch": 0.7028274668205424,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004960915813697836,
      "loss": 6.2702,
      "mean_token_accuracy": 0.15371712446212768,
      "num_tokens": 15448052.0,
      "step": 6090
    },
    {
      "entropy": 6.352459335327149,
      "epoch": 0.7034045008655511,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004960839198333154,
      "loss": 6.18,
      "mean_token_accuracy": 0.16577911525964736,
      "num_tokens": 15459141.0,
      "step": 6095
    },
    {
      "entropy": 6.196254682540894,
      "epoch": 0.7039815349105597,
      "grad_norm": 1.0,
      "learning_rate": 0.0004960762508607994,
      "loss": 6.1413,
      "mean_token_accuracy": 0.17136777937412262,
      "num_tokens": 15471705.0,
      "step": 6100
    },
    {
      "entropy": 6.254960727691651,
      "epoch": 0.7045585689555683,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004960685744524934,
      "loss": 6.2446,
      "mean_token_accuracy": 0.16158175617456436,
      "num_tokens": 15484821.0,
      "step": 6105
    },
    {
      "entropy": 6.368754720687866,
      "epoch": 0.705135603000577,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004960608906086558,
      "loss": 6.2741,
      "mean_token_accuracy": 0.1557539790868759,
      "num_tokens": 15497157.0,
      "step": 6110
    },
    {
      "entropy": 6.336442375183106,
      "epoch": 0.7057126370455857,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000496053199329545,
      "loss": 6.2906,
      "mean_token_accuracy": 0.1631980761885643,
      "num_tokens": 15508919.0,
      "step": 6115
    },
    {
      "entropy": 6.362679052352905,
      "epoch": 0.7062896710905944,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004960455006154196,
      "loss": 6.3171,
      "mean_token_accuracy": 0.15459487587213516,
      "num_tokens": 15521235.0,
      "step": 6120
    },
    {
      "entropy": 6.331986808776856,
      "epoch": 0.706866705135603,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004960377944665386,
      "loss": 6.2839,
      "mean_token_accuracy": 0.1561708405613899,
      "num_tokens": 15534278.0,
      "step": 6125
    },
    {
      "entropy": 6.272089147567749,
      "epoch": 0.7074437391806117,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004960300808831611,
      "loss": 6.2455,
      "mean_token_accuracy": 0.16454179733991622,
      "num_tokens": 15548034.0,
      "step": 6130
    },
    {
      "entropy": 6.402075338363647,
      "epoch": 0.7080207732256203,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004960223598655467,
      "loss": 6.2189,
      "mean_token_accuracy": 0.16663224399089813,
      "num_tokens": 15559425.0,
      "step": 6135
    },
    {
      "entropy": 6.374712610244751,
      "epoch": 0.7085978072706289,
      "grad_norm": 1.0,
      "learning_rate": 0.000496014631413955,
      "loss": 6.3215,
      "mean_token_accuracy": 0.16596050709486007,
      "num_tokens": 15570634.0,
      "step": 6140
    },
    {
      "entropy": 6.350189828872681,
      "epoch": 0.7091748413156376,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004960068955286459,
      "loss": 6.356,
      "mean_token_accuracy": 0.1646140307188034,
      "num_tokens": 15582710.0,
      "step": 6145
    },
    {
      "entropy": 6.373179197311401,
      "epoch": 0.7097518753606463,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004959991522098797,
      "loss": 6.2712,
      "mean_token_accuracy": 0.15503247380256652,
      "num_tokens": 15594798.0,
      "step": 6150
    },
    {
      "entropy": 6.347809934616089,
      "epoch": 0.710328909405655,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004959914014579168,
      "loss": 6.2034,
      "mean_token_accuracy": 0.16503212749958038,
      "num_tokens": 15606482.0,
      "step": 6155
    },
    {
      "entropy": 6.3105134010314945,
      "epoch": 0.7109059434506636,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004959836432730178,
      "loss": 6.2485,
      "mean_token_accuracy": 0.1680520221590996,
      "num_tokens": 15620178.0,
      "step": 6160
    },
    {
      "entropy": 6.24373574256897,
      "epoch": 0.7114829774956722,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004959758776554438,
      "loss": 6.268,
      "mean_token_accuracy": 0.15886785089969635,
      "num_tokens": 15632469.0,
      "step": 6165
    },
    {
      "entropy": 6.387234163284302,
      "epoch": 0.7120600115406809,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000495968104605456,
      "loss": 6.2929,
      "mean_token_accuracy": 0.15935038924217224,
      "num_tokens": 15645610.0,
      "step": 6170
    },
    {
      "entropy": 6.394750118255615,
      "epoch": 0.7126370455856895,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004959603241233157,
      "loss": 6.2758,
      "mean_token_accuracy": 0.16319217681884765,
      "num_tokens": 15657716.0,
      "step": 6175
    },
    {
      "entropy": 6.316469669342041,
      "epoch": 0.7132140796306982,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004959525362092846,
      "loss": 6.2672,
      "mean_token_accuracy": 0.15528870224952698,
      "num_tokens": 15670258.0,
      "step": 6180
    },
    {
      "entropy": 6.305664682388306,
      "epoch": 0.7137911136757069,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004959447408636247,
      "loss": 6.2207,
      "mean_token_accuracy": 0.16739086508750917,
      "num_tokens": 15681196.0,
      "step": 6185
    },
    {
      "entropy": 6.392039203643799,
      "epoch": 0.7143681477207156,
      "grad_norm": 0.875,
      "learning_rate": 0.0004959369380865983,
      "loss": 6.3365,
      "mean_token_accuracy": 0.15902895629405975,
      "num_tokens": 15694591.0,
      "step": 6190
    },
    {
      "entropy": 6.305923986434936,
      "epoch": 0.7149451817657242,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004959291278784676,
      "loss": 6.1828,
      "mean_token_accuracy": 0.16872139424085617,
      "num_tokens": 15708272.0,
      "step": 6195
    },
    {
      "entropy": 6.294431447982788,
      "epoch": 0.7155222158107328,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004959213102394954,
      "loss": 6.2348,
      "mean_token_accuracy": 0.169078066945076,
      "num_tokens": 15720136.0,
      "step": 6200
    },
    {
      "entropy": 6.362851905822754,
      "epoch": 0.7160992498557415,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004959134851699449,
      "loss": 6.2375,
      "mean_token_accuracy": 0.16505309492349624,
      "num_tokens": 15732043.0,
      "step": 6205
    },
    {
      "entropy": 6.3062849044799805,
      "epoch": 0.7166762839007501,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004959056526700788,
      "loss": 6.2709,
      "mean_token_accuracy": 0.16027791649103165,
      "num_tokens": 15743471.0,
      "step": 6210
    },
    {
      "entropy": 6.358971881866455,
      "epoch": 0.7172533179457588,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004958978127401609,
      "loss": 6.2843,
      "mean_token_accuracy": 0.15638664811849595,
      "num_tokens": 15756846.0,
      "step": 6215
    },
    {
      "entropy": 6.326319360733033,
      "epoch": 0.7178303519907675,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004958899653804548,
      "loss": 6.3084,
      "mean_token_accuracy": 0.15958495140075685,
      "num_tokens": 15769993.0,
      "step": 6220
    },
    {
      "entropy": 6.2971519947052,
      "epoch": 0.7184073860357761,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004958821105912246,
      "loss": 6.2818,
      "mean_token_accuracy": 0.15774881839752197,
      "num_tokens": 15782081.0,
      "step": 6225
    },
    {
      "entropy": 6.3659319400787355,
      "epoch": 0.7189844200807848,
      "grad_norm": 0.9375,
      "learning_rate": 0.000495874248372734,
      "loss": 6.2956,
      "mean_token_accuracy": 0.15927491039037706,
      "num_tokens": 15793987.0,
      "step": 6230
    },
    {
      "entropy": 6.306554555892944,
      "epoch": 0.7195614541257934,
      "grad_norm": 0.921875,
      "learning_rate": 0.000495866378725248,
      "loss": 6.2341,
      "mean_token_accuracy": 0.16659074723720552,
      "num_tokens": 15805749.0,
      "step": 6235
    },
    {
      "entropy": 6.242205476760864,
      "epoch": 0.7201384881708021,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000495858501649031,
      "loss": 6.1824,
      "mean_token_accuracy": 0.16428372114896775,
      "num_tokens": 15818402.0,
      "step": 6240
    },
    {
      "entropy": 6.255460739135742,
      "epoch": 0.7207155222158107,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000495850617144348,
      "loss": 6.181,
      "mean_token_accuracy": 0.17008669227361678,
      "num_tokens": 15830783.0,
      "step": 6245
    },
    {
      "entropy": 6.307915115356446,
      "epoch": 0.7212925562608193,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004958427252114643,
      "loss": 6.1784,
      "mean_token_accuracy": 0.16647693812847136,
      "num_tokens": 15843231.0,
      "step": 6250
    },
    {
      "entropy": 6.208975172042846,
      "epoch": 0.7218695903058281,
      "grad_norm": 0.921875,
      "learning_rate": 0.000495834825850645,
      "loss": 6.1986,
      "mean_token_accuracy": 0.1770086407661438,
      "num_tokens": 15855770.0,
      "step": 6255
    },
    {
      "entropy": 6.363209676742554,
      "epoch": 0.7224466243508367,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004958269190621562,
      "loss": 6.2768,
      "mean_token_accuracy": 0.1566353589296341,
      "num_tokens": 15867979.0,
      "step": 6260
    },
    {
      "entropy": 6.370323896408081,
      "epoch": 0.7230236583958454,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004958190048462637,
      "loss": 6.312,
      "mean_token_accuracy": 0.15921126455068588,
      "num_tokens": 15880878.0,
      "step": 6265
    },
    {
      "entropy": 6.2488103866577145,
      "epoch": 0.723600692440854,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004958110832032336,
      "loss": 6.2493,
      "mean_token_accuracy": 0.16388887912034988,
      "num_tokens": 15893599.0,
      "step": 6270
    },
    {
      "entropy": 6.2757346630096436,
      "epoch": 0.7241777264858626,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004958031541333322,
      "loss": 6.2163,
      "mean_token_accuracy": 0.16289129704236985,
      "num_tokens": 15906566.0,
      "step": 6275
    },
    {
      "entropy": 6.372100734710694,
      "epoch": 0.7247547605308713,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004957952176368267,
      "loss": 6.2278,
      "mean_token_accuracy": 0.1621674656867981,
      "num_tokens": 15920269.0,
      "step": 6280
    },
    {
      "entropy": 6.28837022781372,
      "epoch": 0.7253317945758799,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004957872737139836,
      "loss": 6.2808,
      "mean_token_accuracy": 0.15888100117444992,
      "num_tokens": 15932559.0,
      "step": 6285
    },
    {
      "entropy": 6.279818773269653,
      "epoch": 0.7259088286208887,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004957793223650702,
      "loss": 6.2223,
      "mean_token_accuracy": 0.15796030014753343,
      "num_tokens": 15945554.0,
      "step": 6290
    },
    {
      "entropy": 6.26359715461731,
      "epoch": 0.7264858626658973,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000495771363590354,
      "loss": 6.1351,
      "mean_token_accuracy": 0.17019696682691574,
      "num_tokens": 15958012.0,
      "step": 6295
    },
    {
      "entropy": 6.314996814727783,
      "epoch": 0.727062896710906,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004957633973901027,
      "loss": 6.2463,
      "mean_token_accuracy": 0.15923839658498765,
      "num_tokens": 15971508.0,
      "step": 6300
    },
    {
      "entropy": 6.3264491081237795,
      "epoch": 0.7276399307559146,
      "grad_norm": 1.0,
      "learning_rate": 0.0004957554237645841,
      "loss": 6.2138,
      "mean_token_accuracy": 0.1610023185610771,
      "num_tokens": 15983445.0,
      "step": 6305
    },
    {
      "entropy": 6.30993881225586,
      "epoch": 0.7282169648009232,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004957474427140665,
      "loss": 6.2428,
      "mean_token_accuracy": 0.1632042720913887,
      "num_tokens": 15995936.0,
      "step": 6310
    },
    {
      "entropy": 6.193361854553222,
      "epoch": 0.7287939988459319,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004957394542388182,
      "loss": 6.2117,
      "mean_token_accuracy": 0.16288071870803833,
      "num_tokens": 16008192.0,
      "step": 6315
    },
    {
      "entropy": 6.375859880447388,
      "epoch": 0.7293710328909405,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004957314583391082,
      "loss": 6.2872,
      "mean_token_accuracy": 0.15950585305690765,
      "num_tokens": 16020585.0,
      "step": 6320
    },
    {
      "entropy": 6.413668632507324,
      "epoch": 0.7299480669359493,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004957234550152052,
      "loss": 6.3246,
      "mean_token_accuracy": 0.15500302165746688,
      "num_tokens": 16032627.0,
      "step": 6325
    },
    {
      "entropy": 6.3167119979858395,
      "epoch": 0.7305251009809579,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004957154442673784,
      "loss": 6.2688,
      "mean_token_accuracy": 0.16070621758699416,
      "num_tokens": 16047196.0,
      "step": 6330
    },
    {
      "entropy": 6.245927858352661,
      "epoch": 0.7311021350259665,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004957074260958972,
      "loss": 6.2062,
      "mean_token_accuracy": 0.165318962931633,
      "num_tokens": 16059875.0,
      "step": 6335
    },
    {
      "entropy": 6.360879945755005,
      "epoch": 0.7316791690709752,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004956994005010315,
      "loss": 6.2673,
      "mean_token_accuracy": 0.16559662967920302,
      "num_tokens": 16071991.0,
      "step": 6340
    },
    {
      "entropy": 6.344012689590454,
      "epoch": 0.7322562031159838,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000495691367483051,
      "loss": 6.3267,
      "mean_token_accuracy": 0.15291294902563096,
      "num_tokens": 16085628.0,
      "step": 6345
    },
    {
      "entropy": 6.344559192657471,
      "epoch": 0.7328332371609925,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004956833270422259,
      "loss": 6.2882,
      "mean_token_accuracy": 0.16019604802131654,
      "num_tokens": 16098020.0,
      "step": 6350
    },
    {
      "entropy": 6.256140899658203,
      "epoch": 0.7334102712060011,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004956752791788269,
      "loss": 6.2043,
      "mean_token_accuracy": 0.16242919713258744,
      "num_tokens": 16109965.0,
      "step": 6355
    },
    {
      "entropy": 6.2727375507354735,
      "epoch": 0.7339873052510099,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004956672238931244,
      "loss": 6.2866,
      "mean_token_accuracy": 0.15539705604314805,
      "num_tokens": 16121891.0,
      "step": 6360
    },
    {
      "entropy": 6.390357112884521,
      "epoch": 0.7345643392960185,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004956591611853894,
      "loss": 6.2203,
      "mean_token_accuracy": 0.1597520515322685,
      "num_tokens": 16134181.0,
      "step": 6365
    },
    {
      "entropy": 6.298908519744873,
      "epoch": 0.7351413733410271,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004956510910558931,
      "loss": 6.2901,
      "mean_token_accuracy": 0.15978810638189317,
      "num_tokens": 16146685.0,
      "step": 6370
    },
    {
      "entropy": 6.246630859375,
      "epoch": 0.7357184073860358,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004956430135049071,
      "loss": 6.2535,
      "mean_token_accuracy": 0.15794457942247392,
      "num_tokens": 16159618.0,
      "step": 6375
    },
    {
      "entropy": 6.380770683288574,
      "epoch": 0.7362954414310444,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004956349285327028,
      "loss": 6.1689,
      "mean_token_accuracy": 0.16803180128335954,
      "num_tokens": 16172162.0,
      "step": 6380
    },
    {
      "entropy": 6.246574258804321,
      "epoch": 0.7368724754760531,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004956268361395523,
      "loss": 6.2672,
      "mean_token_accuracy": 0.15536082834005355,
      "num_tokens": 16184663.0,
      "step": 6385
    },
    {
      "entropy": 6.294555282592773,
      "epoch": 0.7374495095210617,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004956187363257278,
      "loss": 6.251,
      "mean_token_accuracy": 0.1647225648164749,
      "num_tokens": 16196459.0,
      "step": 6390
    },
    {
      "entropy": 6.396480321884155,
      "epoch": 0.7380265435660704,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004956106290915017,
      "loss": 6.2753,
      "mean_token_accuracy": 0.1614134669303894,
      "num_tokens": 16208008.0,
      "step": 6395
    },
    {
      "entropy": 6.305571746826172,
      "epoch": 0.7386035776110791,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004956025144371467,
      "loss": 6.2945,
      "mean_token_accuracy": 0.16187050193548203,
      "num_tokens": 16221844.0,
      "step": 6400
    },
    {
      "entropy": 6.335601472854615,
      "epoch": 0.7391806116560877,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004955943923629356,
      "loss": 6.3092,
      "mean_token_accuracy": 0.16016998887062073,
      "num_tokens": 16234773.0,
      "step": 6405
    },
    {
      "entropy": 6.2681526184082035,
      "epoch": 0.7397576457010964,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004955862628691417,
      "loss": 6.1818,
      "mean_token_accuracy": 0.16729002445936203,
      "num_tokens": 16247464.0,
      "step": 6410
    },
    {
      "entropy": 6.377269506454468,
      "epoch": 0.740334679746105,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004955781259560386,
      "loss": 6.2624,
      "mean_token_accuracy": 0.15353865921497345,
      "num_tokens": 16259947.0,
      "step": 6415
    },
    {
      "entropy": 6.330693244934082,
      "epoch": 0.7409117137911136,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004955699816238995,
      "loss": 6.3696,
      "mean_token_accuracy": 0.15626734644174575,
      "num_tokens": 16272280.0,
      "step": 6420
    },
    {
      "entropy": 6.390060138702393,
      "epoch": 0.7414887478361223,
      "grad_norm": 1.0,
      "learning_rate": 0.0004955618298729988,
      "loss": 6.2601,
      "mean_token_accuracy": 0.15821194499731064,
      "num_tokens": 16284476.0,
      "step": 6425
    },
    {
      "entropy": 6.298096799850464,
      "epoch": 0.742065781881131,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004955536707036105,
      "loss": 6.1789,
      "mean_token_accuracy": 0.16258774995803832,
      "num_tokens": 16296877.0,
      "step": 6430
    },
    {
      "entropy": 6.176366090774536,
      "epoch": 0.7426428159261397,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000495545504116009,
      "loss": 6.1854,
      "mean_token_accuracy": 0.16784305721521378,
      "num_tokens": 16310360.0,
      "step": 6435
    },
    {
      "entropy": 6.393213129043579,
      "epoch": 0.7432198499711483,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004955373301104691,
      "loss": 6.3361,
      "mean_token_accuracy": 0.1535481706261635,
      "num_tokens": 16322965.0,
      "step": 6440
    },
    {
      "entropy": 6.345073318481445,
      "epoch": 0.743796884016157,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004955291486872657,
      "loss": 6.2822,
      "mean_token_accuracy": 0.15675753802061082,
      "num_tokens": 16337756.0,
      "step": 6445
    },
    {
      "entropy": 6.355855798721313,
      "epoch": 0.7443739180611656,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004955209598466738,
      "loss": 6.2384,
      "mean_token_accuracy": 0.1589888334274292,
      "num_tokens": 16350400.0,
      "step": 6450
    },
    {
      "entropy": 6.326137495040894,
      "epoch": 0.7449509521061742,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000495512763588969,
      "loss": 6.3259,
      "mean_token_accuracy": 0.15499866157770156,
      "num_tokens": 16363655.0,
      "step": 6455
    },
    {
      "entropy": 6.340245914459229,
      "epoch": 0.7455279861511829,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004955045599144269,
      "loss": 6.22,
      "mean_token_accuracy": 0.1656513459980488,
      "num_tokens": 16375580.0,
      "step": 6460
    },
    {
      "entropy": 6.31332745552063,
      "epoch": 0.7461050201961916,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004954963488233235,
      "loss": 6.2822,
      "mean_token_accuracy": 0.16308767050504686,
      "num_tokens": 16388447.0,
      "step": 6465
    },
    {
      "entropy": 6.37180380821228,
      "epoch": 0.7466820542412003,
      "grad_norm": 1.15625,
      "learning_rate": 0.000495488130315935,
      "loss": 6.1771,
      "mean_token_accuracy": 0.16913487613201142,
      "num_tokens": 16401082.0,
      "step": 6470
    },
    {
      "entropy": 6.183729076385498,
      "epoch": 0.7472590882862089,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004954799043925377,
      "loss": 6.2069,
      "mean_token_accuracy": 0.15926148295402526,
      "num_tokens": 16415108.0,
      "step": 6475
    },
    {
      "entropy": 6.2923017024993895,
      "epoch": 0.7478361223312175,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004954716710534082,
      "loss": 6.1808,
      "mean_token_accuracy": 0.16519278585910796,
      "num_tokens": 16427562.0,
      "step": 6480
    },
    {
      "entropy": 6.310155820846558,
      "epoch": 0.7484131563762262,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004954634302988237,
      "loss": 6.2101,
      "mean_token_accuracy": 0.1652577206492424,
      "num_tokens": 16441282.0,
      "step": 6485
    },
    {
      "entropy": 6.249753093719482,
      "epoch": 0.7489901904212348,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004954551821290612,
      "loss": 6.2405,
      "mean_token_accuracy": 0.16182129085063934,
      "num_tokens": 16453433.0,
      "step": 6490
    },
    {
      "entropy": 6.318871402740479,
      "epoch": 0.7495672244662435,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004954469265443981,
      "loss": 6.2562,
      "mean_token_accuracy": 0.1590650722384453,
      "num_tokens": 16466216.0,
      "step": 6495
    },
    {
      "entropy": 6.2545201778411865,
      "epoch": 0.7501442585112522,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004954386635451123,
      "loss": 6.2496,
      "mean_token_accuracy": 0.1611919954419136,
      "num_tokens": 16479046.0,
      "step": 6500
    },
    {
      "entropy": 6.278717374801635,
      "epoch": 0.7507212925562609,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004954303931314814,
      "loss": 6.2225,
      "mean_token_accuracy": 0.16637153327465057,
      "num_tokens": 16491407.0,
      "step": 6505
    },
    {
      "entropy": 6.3345311164855955,
      "epoch": 0.7512983266012695,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004954221153037837,
      "loss": 6.2123,
      "mean_token_accuracy": 0.15814678370952606,
      "num_tokens": 16504726.0,
      "step": 6510
    },
    {
      "entropy": 6.374265193939209,
      "epoch": 0.7518753606462781,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004954138300622978,
      "loss": 6.2813,
      "mean_token_accuracy": 0.16176176816225052,
      "num_tokens": 16517835.0,
      "step": 6515
    },
    {
      "entropy": 6.31195559501648,
      "epoch": 0.7524523946912868,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000495405537407302,
      "loss": 6.2835,
      "mean_token_accuracy": 0.15772045999765397,
      "num_tokens": 16532018.0,
      "step": 6520
    },
    {
      "entropy": 6.28839077949524,
      "epoch": 0.7530294287362954,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004953972373390755,
      "loss": 6.2304,
      "mean_token_accuracy": 0.16343475133180618,
      "num_tokens": 16546211.0,
      "step": 6525
    },
    {
      "entropy": 6.2691858291625975,
      "epoch": 0.753606462781304,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004953889298578975,
      "loss": 6.2951,
      "mean_token_accuracy": 0.15629006847739219,
      "num_tokens": 16559788.0,
      "step": 6530
    },
    {
      "entropy": 6.227923965454101,
      "epoch": 0.7541834968263128,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004953806149640473,
      "loss": 6.1513,
      "mean_token_accuracy": 0.16980354338884354,
      "num_tokens": 16573694.0,
      "step": 6535
    },
    {
      "entropy": 6.3080932140350345,
      "epoch": 0.7547605308713214,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004953722926578045,
      "loss": 6.2381,
      "mean_token_accuracy": 0.16089166104793548,
      "num_tokens": 16585506.0,
      "step": 6540
    },
    {
      "entropy": 6.289967107772827,
      "epoch": 0.7553375649163301,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004953639629394492,
      "loss": 6.2907,
      "mean_token_accuracy": 0.16097813844680786,
      "num_tokens": 16598991.0,
      "step": 6545
    },
    {
      "entropy": 6.374682903289795,
      "epoch": 0.7559145989613387,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004953556258092613,
      "loss": 6.2935,
      "mean_token_accuracy": 0.15956879705190657,
      "num_tokens": 16612025.0,
      "step": 6550
    },
    {
      "entropy": 6.282010126113891,
      "epoch": 0.7564916330063474,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004953472812675216,
      "loss": 6.1921,
      "mean_token_accuracy": 0.16668391227722168,
      "num_tokens": 16624509.0,
      "step": 6555
    },
    {
      "entropy": 6.223297929763794,
      "epoch": 0.757068667051356,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004953389293145104,
      "loss": 6.2349,
      "mean_token_accuracy": 0.15582622289657594,
      "num_tokens": 16636920.0,
      "step": 6560
    },
    {
      "entropy": 6.282078266143799,
      "epoch": 0.7576457010963646,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004953305699505088,
      "loss": 6.152,
      "mean_token_accuracy": 0.16699379980564116,
      "num_tokens": 16649381.0,
      "step": 6565
    },
    {
      "entropy": 6.377573394775391,
      "epoch": 0.7582227351413734,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004953222031757979,
      "loss": 6.2647,
      "mean_token_accuracy": 0.15163747519254683,
      "num_tokens": 16662273.0,
      "step": 6570
    },
    {
      "entropy": 6.3311680316925045,
      "epoch": 0.758799769186382,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004953138289906592,
      "loss": 6.2617,
      "mean_token_accuracy": 0.16328519582748413,
      "num_tokens": 16675184.0,
      "step": 6575
    },
    {
      "entropy": 6.2730225086212155,
      "epoch": 0.7593768032313907,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004953054473953742,
      "loss": 6.2299,
      "mean_token_accuracy": 0.16334724575281143,
      "num_tokens": 16687658.0,
      "step": 6580
    },
    {
      "entropy": 6.305401706695557,
      "epoch": 0.7599538372763993,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004952970583902251,
      "loss": 6.1994,
      "mean_token_accuracy": 0.16126615852117537,
      "num_tokens": 16699806.0,
      "step": 6585
    },
    {
      "entropy": 6.312052536010742,
      "epoch": 0.760530871321408,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004952886619754937,
      "loss": 6.2493,
      "mean_token_accuracy": 0.16245510876178743,
      "num_tokens": 16712647.0,
      "step": 6590
    },
    {
      "entropy": 6.313289022445678,
      "epoch": 0.7611079053664166,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004952802581514625,
      "loss": 6.2134,
      "mean_token_accuracy": 0.16226628422737122,
      "num_tokens": 16725438.0,
      "step": 6595
    },
    {
      "entropy": 6.26066517829895,
      "epoch": 0.7616849394114252,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004952718469184144,
      "loss": 6.2332,
      "mean_token_accuracy": 0.16719795763492584,
      "num_tokens": 16737346.0,
      "step": 6600
    },
    {
      "entropy": 6.306323671340943,
      "epoch": 0.762261973456434,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004952634282766322,
      "loss": 6.2212,
      "mean_token_accuracy": 0.16295539289712907,
      "num_tokens": 16749327.0,
      "step": 6605
    },
    {
      "entropy": 6.2595268249511715,
      "epoch": 0.7628390075014426,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004952550022263988,
      "loss": 6.1536,
      "mean_token_accuracy": 0.1651233419775963,
      "num_tokens": 16761323.0,
      "step": 6610
    },
    {
      "entropy": 6.243611764907837,
      "epoch": 0.7634160415464513,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004952465687679979,
      "loss": 6.3256,
      "mean_token_accuracy": 0.15572553426027297,
      "num_tokens": 16774688.0,
      "step": 6615
    },
    {
      "entropy": 6.408269548416138,
      "epoch": 0.7639930755914599,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000495238127901713,
      "loss": 6.312,
      "mean_token_accuracy": 0.15856917053461075,
      "num_tokens": 16787538.0,
      "step": 6620
    },
    {
      "entropy": 6.28782320022583,
      "epoch": 0.7645701096364685,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004952296796278282,
      "loss": 6.2161,
      "mean_token_accuracy": 0.1665654569864273,
      "num_tokens": 16798642.0,
      "step": 6625
    },
    {
      "entropy": 6.294420623779297,
      "epoch": 0.7651471436814772,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004952212239466274,
      "loss": 6.3312,
      "mean_token_accuracy": 0.16053057312965394,
      "num_tokens": 16810785.0,
      "step": 6630
    },
    {
      "entropy": 6.410816287994384,
      "epoch": 0.7657241777264858,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004952127608583953,
      "loss": 6.2946,
      "mean_token_accuracy": 0.15736103802919388,
      "num_tokens": 16823978.0,
      "step": 6635
    },
    {
      "entropy": 6.293875408172608,
      "epoch": 0.7663012117714946,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004952042903634162,
      "loss": 6.2263,
      "mean_token_accuracy": 0.1684988558292389,
      "num_tokens": 16836878.0,
      "step": 6640
    },
    {
      "entropy": 6.262284994125366,
      "epoch": 0.7668782458165032,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004951958124619752,
      "loss": 6.2512,
      "mean_token_accuracy": 0.16488370150327683,
      "num_tokens": 16849849.0,
      "step": 6645
    },
    {
      "entropy": 6.285416841506958,
      "epoch": 0.7674552798615119,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004951873271543573,
      "loss": 6.2535,
      "mean_token_accuracy": 0.16183283179998398,
      "num_tokens": 16862022.0,
      "step": 6650
    },
    {
      "entropy": 6.337986326217651,
      "epoch": 0.7680323139065205,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004951788344408483,
      "loss": 6.2974,
      "mean_token_accuracy": 0.16393122375011443,
      "num_tokens": 16875421.0,
      "step": 6655
    },
    {
      "entropy": 6.299657344818115,
      "epoch": 0.7686093479515291,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004951703343217335,
      "loss": 6.1897,
      "mean_token_accuracy": 0.17819436639547348,
      "num_tokens": 16889520.0,
      "step": 6660
    },
    {
      "entropy": 6.263566303253174,
      "epoch": 0.7691863819965378,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004951618267972987,
      "loss": 6.2152,
      "mean_token_accuracy": 0.16138704270124435,
      "num_tokens": 16901449.0,
      "step": 6665
    },
    {
      "entropy": 6.354003810882569,
      "epoch": 0.7697634160415464,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004951533118678304,
      "loss": 6.2737,
      "mean_token_accuracy": 0.16266804486513137,
      "num_tokens": 16914742.0,
      "step": 6670
    },
    {
      "entropy": 6.374794387817383,
      "epoch": 0.7703404500865552,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004951447895336147,
      "loss": 6.2476,
      "mean_token_accuracy": 0.1604885146021843,
      "num_tokens": 16926645.0,
      "step": 6675
    },
    {
      "entropy": 6.282076501846314,
      "epoch": 0.7709174841315638,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004951362597949384,
      "loss": 6.2091,
      "mean_token_accuracy": 0.1674415647983551,
      "num_tokens": 16939499.0,
      "step": 6680
    },
    {
      "entropy": 6.27974910736084,
      "epoch": 0.7714945181765724,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004951277226520883,
      "loss": 6.2155,
      "mean_token_accuracy": 0.1670735940337181,
      "num_tokens": 16951811.0,
      "step": 6685
    },
    {
      "entropy": 6.277884769439697,
      "epoch": 0.7720715522215811,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004951191781053517,
      "loss": 6.2155,
      "mean_token_accuracy": 0.16209737211465836,
      "num_tokens": 16963126.0,
      "step": 6690
    },
    {
      "entropy": 6.318505191802979,
      "epoch": 0.7726485862665897,
      "grad_norm": 1.0,
      "learning_rate": 0.0004951106261550157,
      "loss": 6.2226,
      "mean_token_accuracy": 0.16463182121515274,
      "num_tokens": 16976853.0,
      "step": 6695
    },
    {
      "entropy": 6.239604473114014,
      "epoch": 0.7732256203115984,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004951020668013683,
      "loss": 6.2625,
      "mean_token_accuracy": 0.16043406426906587,
      "num_tokens": 16988720.0,
      "step": 6700
    },
    {
      "entropy": 6.234865522384643,
      "epoch": 0.773802654356607,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004950935000446972,
      "loss": 6.1635,
      "mean_token_accuracy": 0.1725506454706192,
      "num_tokens": 17000880.0,
      "step": 6705
    },
    {
      "entropy": 6.162816905975342,
      "epoch": 0.7743796884016156,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004950849258852905,
      "loss": 6.0462,
      "mean_token_accuracy": 0.17703194320201873,
      "num_tokens": 17012540.0,
      "step": 6710
    },
    {
      "entropy": 6.234879636764527,
      "epoch": 0.7749567224466244,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004950763443234366,
      "loss": 6.1816,
      "mean_token_accuracy": 0.1716555267572403,
      "num_tokens": 17024163.0,
      "step": 6715
    },
    {
      "entropy": 6.338557767868042,
      "epoch": 0.775533756491633,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004950677553594243,
      "loss": 6.225,
      "mean_token_accuracy": 0.15775014460086823,
      "num_tokens": 17037020.0,
      "step": 6720
    },
    {
      "entropy": 6.312491703033447,
      "epoch": 0.7761107905366417,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004950591589935422,
      "loss": 6.2651,
      "mean_token_accuracy": 0.1570696786046028,
      "num_tokens": 17050132.0,
      "step": 6725
    },
    {
      "entropy": 6.300922155380249,
      "epoch": 0.7766878245816503,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004950505552260798,
      "loss": 6.2119,
      "mean_token_accuracy": 0.16404270231723786,
      "num_tokens": 17063299.0,
      "step": 6730
    },
    {
      "entropy": 6.281912469863892,
      "epoch": 0.777264858626659,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004950419440573261,
      "loss": 6.2455,
      "mean_token_accuracy": 0.1641908273100853,
      "num_tokens": 17076312.0,
      "step": 6735
    },
    {
      "entropy": 6.21780743598938,
      "epoch": 0.7778418926716676,
      "grad_norm": 0.953125,
      "learning_rate": 0.000495033325487571,
      "loss": 6.2342,
      "mean_token_accuracy": 0.1627377912402153,
      "num_tokens": 17089040.0,
      "step": 6740
    },
    {
      "entropy": 6.364416837692261,
      "epoch": 0.7784189267166762,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004950246995171042,
      "loss": 6.3155,
      "mean_token_accuracy": 0.16170232594013215,
      "num_tokens": 17102806.0,
      "step": 6745
    },
    {
      "entropy": 6.2666021347045895,
      "epoch": 0.778995960761685,
      "grad_norm": 0.953125,
      "learning_rate": 0.000495016066146216,
      "loss": 6.1634,
      "mean_token_accuracy": 0.1747506693005562,
      "num_tokens": 17114782.0,
      "step": 6750
    },
    {
      "entropy": 6.217280006408691,
      "epoch": 0.7795729948066936,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004950074253751968,
      "loss": 6.221,
      "mean_token_accuracy": 0.16827250719070436,
      "num_tokens": 17127386.0,
      "step": 6755
    },
    {
      "entropy": 6.279114580154419,
      "epoch": 0.7801500288517023,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000494998777204337,
      "loss": 6.1765,
      "mean_token_accuracy": 0.16694536656141282,
      "num_tokens": 17139244.0,
      "step": 6760
    },
    {
      "entropy": 6.366572666168213,
      "epoch": 0.7807270628967109,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004949901216339276,
      "loss": 6.3086,
      "mean_token_accuracy": 0.15615349560976027,
      "num_tokens": 17153136.0,
      "step": 6765
    },
    {
      "entropy": 6.264227628707886,
      "epoch": 0.7813040969417195,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004949814586642598,
      "loss": 6.1751,
      "mean_token_accuracy": 0.16334059238433837,
      "num_tokens": 17166420.0,
      "step": 6770
    },
    {
      "entropy": 6.321720314025879,
      "epoch": 0.7818811309867282,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000494972788295625,
      "loss": 6.317,
      "mean_token_accuracy": 0.1529185950756073,
      "num_tokens": 17180796.0,
      "step": 6775
    },
    {
      "entropy": 6.364991521835327,
      "epoch": 0.7824581650317368,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004949641105283144,
      "loss": 6.2722,
      "mean_token_accuracy": 0.15877276957035064,
      "num_tokens": 17192275.0,
      "step": 6780
    },
    {
      "entropy": 6.257094860076904,
      "epoch": 0.7830351990767456,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004949554253626205,
      "loss": 6.2202,
      "mean_token_accuracy": 0.1650067463517189,
      "num_tokens": 17203651.0,
      "step": 6785
    },
    {
      "entropy": 6.342705535888672,
      "epoch": 0.7836122331217542,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004949467327988351,
      "loss": 6.2845,
      "mean_token_accuracy": 0.16562952995300292,
      "num_tokens": 17216521.0,
      "step": 6790
    },
    {
      "entropy": 6.370225095748902,
      "epoch": 0.7841892671667628,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004949380328372505,
      "loss": 6.2092,
      "mean_token_accuracy": 0.16813843101263046,
      "num_tokens": 17228435.0,
      "step": 6795
    },
    {
      "entropy": 6.2661339282989506,
      "epoch": 0.7847663012117715,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004949293254781595,
      "loss": 6.137,
      "mean_token_accuracy": 0.16679165065288543,
      "num_tokens": 17240746.0,
      "step": 6800
    },
    {
      "entropy": 6.094602251052857,
      "epoch": 0.7853433352567801,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004949206107218547,
      "loss": 6.1622,
      "mean_token_accuracy": 0.17144323885440826,
      "num_tokens": 17254089.0,
      "step": 6805
    },
    {
      "entropy": 6.24790587425232,
      "epoch": 0.7859203693017888,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004949118885686296,
      "loss": 6.1857,
      "mean_token_accuracy": 0.16202608942985536,
      "num_tokens": 17266389.0,
      "step": 6810
    },
    {
      "entropy": 6.317455196380616,
      "epoch": 0.7864974033467974,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004949031590187774,
      "loss": 6.2084,
      "mean_token_accuracy": 0.1603664830327034,
      "num_tokens": 17279049.0,
      "step": 6815
    },
    {
      "entropy": 6.294656372070312,
      "epoch": 0.7870744373918062,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004948944220725915,
      "loss": 6.2621,
      "mean_token_accuracy": 0.15909883230924607,
      "num_tokens": 17291134.0,
      "step": 6820
    },
    {
      "entropy": 6.273614072799683,
      "epoch": 0.7876514714368148,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000494885677730366,
      "loss": 6.1548,
      "mean_token_accuracy": 0.1717804864048958,
      "num_tokens": 17304728.0,
      "step": 6825
    },
    {
      "entropy": 6.30904541015625,
      "epoch": 0.7882285054818234,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000494876925992395,
      "loss": 6.2485,
      "mean_token_accuracy": 0.1600194126367569,
      "num_tokens": 17317392.0,
      "step": 6830
    },
    {
      "entropy": 6.250168180465698,
      "epoch": 0.7888055395268321,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004948681668589728,
      "loss": 6.2049,
      "mean_token_accuracy": 0.16747011691331865,
      "num_tokens": 17329639.0,
      "step": 6835
    },
    {
      "entropy": 6.242401552200318,
      "epoch": 0.7893825735718407,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000494859400330394,
      "loss": 6.2331,
      "mean_token_accuracy": 0.1605785980820656,
      "num_tokens": 17341246.0,
      "step": 6840
    },
    {
      "entropy": 6.277280378341675,
      "epoch": 0.7899596076168494,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004948506264069535,
      "loss": 6.2397,
      "mean_token_accuracy": 0.1591868743300438,
      "num_tokens": 17353699.0,
      "step": 6845
    },
    {
      "entropy": 6.30746898651123,
      "epoch": 0.790536641661858,
      "grad_norm": 1.0,
      "learning_rate": 0.0004948418450889464,
      "loss": 6.2194,
      "mean_token_accuracy": 0.16685952246189117,
      "num_tokens": 17365950.0,
      "step": 6850
    },
    {
      "entropy": 6.171130704879761,
      "epoch": 0.7911136757068667,
      "grad_norm": 1.03125,
      "learning_rate": 0.000494833056376668,
      "loss": 6.1431,
      "mean_token_accuracy": 0.16583137810230256,
      "num_tokens": 17378654.0,
      "step": 6855
    },
    {
      "entropy": 6.271834039688111,
      "epoch": 0.7916907097518754,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004948242602704139,
      "loss": 6.2331,
      "mean_token_accuracy": 0.16291293054819106,
      "num_tokens": 17391354.0,
      "step": 6860
    },
    {
      "entropy": 6.308564567565918,
      "epoch": 0.792267743796884,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004948154567704801,
      "loss": 6.2183,
      "mean_token_accuracy": 0.16628915667533875,
      "num_tokens": 17403019.0,
      "step": 6865
    },
    {
      "entropy": 6.252381896972656,
      "epoch": 0.7928447778418927,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004948066458771625,
      "loss": 6.24,
      "mean_token_accuracy": 0.16437227725982667,
      "num_tokens": 17415886.0,
      "step": 6870
    },
    {
      "entropy": 6.328334617614746,
      "epoch": 0.7934218118869013,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004947978275907577,
      "loss": 6.3518,
      "mean_token_accuracy": 0.15236463844776155,
      "num_tokens": 17429734.0,
      "step": 6875
    },
    {
      "entropy": 6.31690092086792,
      "epoch": 0.79399884593191,
      "grad_norm": 1.34375,
      "learning_rate": 0.000494789001911562,
      "loss": 6.1656,
      "mean_token_accuracy": 0.16355670094490052,
      "num_tokens": 17441749.0,
      "step": 6880
    },
    {
      "entropy": 6.362261867523193,
      "epoch": 0.7945758799769186,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004947801688398725,
      "loss": 6.256,
      "mean_token_accuracy": 0.15867505967617035,
      "num_tokens": 17453626.0,
      "step": 6885
    },
    {
      "entropy": 6.225351238250733,
      "epoch": 0.7951529140219273,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004947713283759862,
      "loss": 6.2227,
      "mean_token_accuracy": 0.16616876274347306,
      "num_tokens": 17465038.0,
      "step": 6890
    },
    {
      "entropy": 6.2706618309021,
      "epoch": 0.795729948066936,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004947624805202003,
      "loss": 6.2889,
      "mean_token_accuracy": 0.15150156915187835,
      "num_tokens": 17478319.0,
      "step": 6895
    },
    {
      "entropy": 6.311843776702881,
      "epoch": 0.7963069821119446,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004947536252728126,
      "loss": 6.1522,
      "mean_token_accuracy": 0.1657789647579193,
      "num_tokens": 17491534.0,
      "step": 6900
    },
    {
      "entropy": 6.244232034683227,
      "epoch": 0.7968840161569533,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004947447626341209,
      "loss": 6.1747,
      "mean_token_accuracy": 0.1646965265274048,
      "num_tokens": 17504957.0,
      "step": 6905
    },
    {
      "entropy": 6.35973048210144,
      "epoch": 0.7974610502019619,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004947358926044232,
      "loss": 6.2018,
      "mean_token_accuracy": 0.16672317534685135,
      "num_tokens": 17518265.0,
      "step": 6910
    },
    {
      "entropy": 6.216989517211914,
      "epoch": 0.7980380842469705,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004947270151840179,
      "loss": 6.1764,
      "mean_token_accuracy": 0.1739889606833458,
      "num_tokens": 17530588.0,
      "step": 6915
    },
    {
      "entropy": 6.151643896102906,
      "epoch": 0.7986151182919792,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004947181303732038,
      "loss": 6.1417,
      "mean_token_accuracy": 0.16983063966035844,
      "num_tokens": 17545429.0,
      "step": 6920
    },
    {
      "entropy": 6.33322057723999,
      "epoch": 0.7991921523369879,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004947092381722793,
      "loss": 6.2453,
      "mean_token_accuracy": 0.16397999078035355,
      "num_tokens": 17558437.0,
      "step": 6925
    },
    {
      "entropy": 6.222066497802734,
      "epoch": 0.7997691863819966,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004947003385815438,
      "loss": 6.1895,
      "mean_token_accuracy": 0.15973960906267165,
      "num_tokens": 17569863.0,
      "step": 6930
    },
    {
      "entropy": 6.311473321914673,
      "epoch": 0.8003462204270052,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004946914316012964,
      "loss": 6.2621,
      "mean_token_accuracy": 0.16151933521032333,
      "num_tokens": 17581987.0,
      "step": 6935
    },
    {
      "entropy": 6.27441291809082,
      "epoch": 0.8009232544720138,
      "grad_norm": 0.84375,
      "learning_rate": 0.000494682517231837,
      "loss": 6.2645,
      "mean_token_accuracy": 0.1605035811662674,
      "num_tokens": 17594064.0,
      "step": 6940
    },
    {
      "entropy": 6.242759466171265,
      "epoch": 0.8015002885170225,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004946735954734652,
      "loss": 6.1514,
      "mean_token_accuracy": 0.1639345332980156,
      "num_tokens": 17606557.0,
      "step": 6945
    },
    {
      "entropy": 6.3339780330657955,
      "epoch": 0.8020773225620311,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004946646663264811,
      "loss": 6.2543,
      "mean_token_accuracy": 0.167417311668396,
      "num_tokens": 17620095.0,
      "step": 6950
    },
    {
      "entropy": 6.2512797832489015,
      "epoch": 0.8026543566070398,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004946557297911852,
      "loss": 6.1282,
      "mean_token_accuracy": 0.1676596835255623,
      "num_tokens": 17633006.0,
      "step": 6955
    },
    {
      "entropy": 6.24902753829956,
      "epoch": 0.8032313906520485,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004946467858678777,
      "loss": 6.2063,
      "mean_token_accuracy": 0.15968579351902007,
      "num_tokens": 17646863.0,
      "step": 6960
    },
    {
      "entropy": 6.321612787246704,
      "epoch": 0.8038084246970572,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004946378345568597,
      "loss": 6.1997,
      "mean_token_accuracy": 0.1605960473418236,
      "num_tokens": 17659917.0,
      "step": 6965
    },
    {
      "entropy": 6.2613342761993405,
      "epoch": 0.8043854587420658,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004946288758584324,
      "loss": 6.2499,
      "mean_token_accuracy": 0.15809816420078276,
      "num_tokens": 17672867.0,
      "step": 6970
    },
    {
      "entropy": 6.2937806129455565,
      "epoch": 0.8049624927870744,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004946199097728968,
      "loss": 6.24,
      "mean_token_accuracy": 0.16138594448566437,
      "num_tokens": 17684600.0,
      "step": 6975
    },
    {
      "entropy": 6.2805427551269535,
      "epoch": 0.8055395268320831,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004946109363005548,
      "loss": 6.1743,
      "mean_token_accuracy": 0.16461528092622757,
      "num_tokens": 17697023.0,
      "step": 6980
    },
    {
      "entropy": 6.229381465911866,
      "epoch": 0.8061165608770917,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000494601955441708,
      "loss": 6.2163,
      "mean_token_accuracy": 0.164045812189579,
      "num_tokens": 17711592.0,
      "step": 6985
    },
    {
      "entropy": 6.312086486816407,
      "epoch": 0.8066935949221004,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004945929671966585,
      "loss": 6.2278,
      "mean_token_accuracy": 0.1617693930864334,
      "num_tokens": 17723039.0,
      "step": 6990
    },
    {
      "entropy": 6.235287380218506,
      "epoch": 0.8072706289671091,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004945839715657085,
      "loss": 6.1852,
      "mean_token_accuracy": 0.16383668333292006,
      "num_tokens": 17736851.0,
      "step": 6995
    },
    {
      "entropy": 6.365895748138428,
      "epoch": 0.8078476630121177,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004945749685491607,
      "loss": 6.2312,
      "mean_token_accuracy": 0.1618833914399147,
      "num_tokens": 17749483.0,
      "step": 7000
    },
    {
      "entropy": 6.284995698928833,
      "epoch": 0.8084246970571264,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004945659581473181,
      "loss": 6.2353,
      "mean_token_accuracy": 0.15903590619564056,
      "num_tokens": 17760463.0,
      "step": 7005
    },
    {
      "entropy": 6.370028114318847,
      "epoch": 0.809001731102135,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004945569403604833,
      "loss": 6.3115,
      "mean_token_accuracy": 0.15611679702997208,
      "num_tokens": 17772557.0,
      "step": 7010
    },
    {
      "entropy": 6.335330581665039,
      "epoch": 0.8095787651471437,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004945479151889601,
      "loss": 6.2404,
      "mean_token_accuracy": 0.16521068662405014,
      "num_tokens": 17785399.0,
      "step": 7015
    },
    {
      "entropy": 6.320019245147705,
      "epoch": 0.8101557991921523,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004945388826330517,
      "loss": 6.2265,
      "mean_token_accuracy": 0.16256107836961747,
      "num_tokens": 17798766.0,
      "step": 7020
    },
    {
      "entropy": 6.290906858444214,
      "epoch": 0.8107328332371609,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004945298426930621,
      "loss": 6.2403,
      "mean_token_accuracy": 0.1617645040154457,
      "num_tokens": 17811726.0,
      "step": 7025
    },
    {
      "entropy": 6.237924432754516,
      "epoch": 0.8113098672821697,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004945207953692952,
      "loss": 6.1562,
      "mean_token_accuracy": 0.1691751718521118,
      "num_tokens": 17824142.0,
      "step": 7030
    },
    {
      "entropy": 6.291462850570679,
      "epoch": 0.8118869013271783,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004945117406620556,
      "loss": 6.1784,
      "mean_token_accuracy": 0.16700648367404938,
      "num_tokens": 17837497.0,
      "step": 7035
    },
    {
      "entropy": 6.235440921783447,
      "epoch": 0.812463935372187,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004945026785716474,
      "loss": 6.2266,
      "mean_token_accuracy": 0.1628534495830536,
      "num_tokens": 17849186.0,
      "step": 7040
    },
    {
      "entropy": 6.2957484245300295,
      "epoch": 0.8130409694171956,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004944936090983757,
      "loss": 6.2563,
      "mean_token_accuracy": 0.16264180094003677,
      "num_tokens": 17861460.0,
      "step": 7045
    },
    {
      "entropy": 6.362234973907471,
      "epoch": 0.8136180034622043,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004944845322425455,
      "loss": 6.1639,
      "mean_token_accuracy": 0.1692268431186676,
      "num_tokens": 17874031.0,
      "step": 7050
    },
    {
      "entropy": 6.1800144672393795,
      "epoch": 0.8141950375072129,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004944754480044621,
      "loss": 6.193,
      "mean_token_accuracy": 0.16117312461137773,
      "num_tokens": 17886623.0,
      "step": 7055
    },
    {
      "entropy": 6.26304178237915,
      "epoch": 0.8147720715522215,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004944663563844311,
      "loss": 6.1869,
      "mean_token_accuracy": 0.15768052786588668,
      "num_tokens": 17899888.0,
      "step": 7060
    },
    {
      "entropy": 6.285961055755616,
      "epoch": 0.8153491055972303,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004944572573827581,
      "loss": 6.2177,
      "mean_token_accuracy": 0.1622305005788803,
      "num_tokens": 17912089.0,
      "step": 7065
    },
    {
      "entropy": 6.300604152679443,
      "epoch": 0.8159261396422389,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004944481509997494,
      "loss": 6.2744,
      "mean_token_accuracy": 0.1614094242453575,
      "num_tokens": 17923790.0,
      "step": 7070
    },
    {
      "entropy": 6.389362001419068,
      "epoch": 0.8165031736872476,
      "grad_norm": 0.9375,
      "learning_rate": 0.000494439037235711,
      "loss": 6.2016,
      "mean_token_accuracy": 0.16226660311222077,
      "num_tokens": 17936537.0,
      "step": 7075
    },
    {
      "entropy": 6.1930914402008055,
      "epoch": 0.8170802077322562,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004944299160909495,
      "loss": 6.1823,
      "mean_token_accuracy": 0.16115268766880037,
      "num_tokens": 17948230.0,
      "step": 7080
    },
    {
      "entropy": 6.288032531738281,
      "epoch": 0.8176572417772648,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000494420787565772,
      "loss": 6.1793,
      "mean_token_accuracy": 0.1660292848944664,
      "num_tokens": 17960927.0,
      "step": 7085
    },
    {
      "entropy": 6.242203283309936,
      "epoch": 0.8182342758222735,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004944116516604852,
      "loss": 6.1479,
      "mean_token_accuracy": 0.17034452110528947,
      "num_tokens": 17974392.0,
      "step": 7090
    },
    {
      "entropy": 6.2045286178588865,
      "epoch": 0.8188113098672821,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004944025083753965,
      "loss": 6.1146,
      "mean_token_accuracy": 0.1729827955365181,
      "num_tokens": 17987845.0,
      "step": 7095
    },
    {
      "entropy": 6.256116485595703,
      "epoch": 0.8193883439122909,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004943933577108133,
      "loss": 6.2424,
      "mean_token_accuracy": 0.16049776524305343,
      "num_tokens": 18000610.0,
      "step": 7100
    },
    {
      "entropy": 6.210866498947143,
      "epoch": 0.8199653779572995,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004943841996670436,
      "loss": 6.1684,
      "mean_token_accuracy": 0.16325817555189132,
      "num_tokens": 18014322.0,
      "step": 7105
    },
    {
      "entropy": 6.251032924652099,
      "epoch": 0.8205424120023082,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004943750342443953,
      "loss": 6.2093,
      "mean_token_accuracy": 0.16301546841859818,
      "num_tokens": 18028072.0,
      "step": 7110
    },
    {
      "entropy": 6.279691410064697,
      "epoch": 0.8211194460473168,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004943658614431767,
      "loss": 6.1845,
      "mean_token_accuracy": 0.16464999467134475,
      "num_tokens": 18039629.0,
      "step": 7115
    },
    {
      "entropy": 6.348135852813721,
      "epoch": 0.8216964800923254,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004943566812636963,
      "loss": 6.2289,
      "mean_token_accuracy": 0.16756125837564467,
      "num_tokens": 18052879.0,
      "step": 7120
    },
    {
      "entropy": 6.287117338180542,
      "epoch": 0.8222735141373341,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000494347493706263,
      "loss": 6.2403,
      "mean_token_accuracy": 0.16453346908092498,
      "num_tokens": 18065037.0,
      "step": 7125
    },
    {
      "entropy": 6.246681213378906,
      "epoch": 0.8228505481823427,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004943382987711856,
      "loss": 6.1765,
      "mean_token_accuracy": 0.16766744256019592,
      "num_tokens": 18078595.0,
      "step": 7130
    },
    {
      "entropy": 6.179457139968872,
      "epoch": 0.8234275822273515,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004943290964587734,
      "loss": 6.0461,
      "mean_token_accuracy": 0.17528246641159057,
      "num_tokens": 18089790.0,
      "step": 7135
    },
    {
      "entropy": 6.160900354385376,
      "epoch": 0.8240046162723601,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004943198867693361,
      "loss": 6.1232,
      "mean_token_accuracy": 0.174264357984066,
      "num_tokens": 18101343.0,
      "step": 7140
    },
    {
      "entropy": 6.296756839752197,
      "epoch": 0.8245816503173687,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004943106697031833,
      "loss": 6.2519,
      "mean_token_accuracy": 0.1631760612130165,
      "num_tokens": 18113124.0,
      "step": 7145
    },
    {
      "entropy": 6.316248369216919,
      "epoch": 0.8251586843623774,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004943014452606251,
      "loss": 6.2102,
      "mean_token_accuracy": 0.1668697863817215,
      "num_tokens": 18125322.0,
      "step": 7150
    },
    {
      "entropy": 6.35999026298523,
      "epoch": 0.825735718407386,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004942922134419717,
      "loss": 6.2612,
      "mean_token_accuracy": 0.15659618824720384,
      "num_tokens": 18137975.0,
      "step": 7155
    },
    {
      "entropy": 6.344512605667115,
      "epoch": 0.8263127524523947,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004942829742475336,
      "loss": 6.2118,
      "mean_token_accuracy": 0.15969525873661042,
      "num_tokens": 18151149.0,
      "step": 7160
    },
    {
      "entropy": 6.2811548709869385,
      "epoch": 0.8268897864974033,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004942737276776217,
      "loss": 6.2169,
      "mean_token_accuracy": 0.1525534301996231,
      "num_tokens": 18164451.0,
      "step": 7165
    },
    {
      "entropy": 6.2008974075317385,
      "epoch": 0.827466820542412,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004942644737325468,
      "loss": 6.1717,
      "mean_token_accuracy": 0.16277583539485932,
      "num_tokens": 18177148.0,
      "step": 7170
    },
    {
      "entropy": 6.272535705566407,
      "epoch": 0.8280438545874207,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004942552124126202,
      "loss": 6.1864,
      "mean_token_accuracy": 0.16461812555789948,
      "num_tokens": 18189428.0,
      "step": 7175
    },
    {
      "entropy": 6.2956990718841555,
      "epoch": 0.8286208886324293,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004942459437181535,
      "loss": 6.1816,
      "mean_token_accuracy": 0.16193560808897017,
      "num_tokens": 18201644.0,
      "step": 7180
    },
    {
      "entropy": 6.250952911376953,
      "epoch": 0.829197922677438,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004942366676494584,
      "loss": 6.2076,
      "mean_token_accuracy": 0.16112402975559234,
      "num_tokens": 18213366.0,
      "step": 7185
    },
    {
      "entropy": 6.233986902236938,
      "epoch": 0.8297749567224466,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004942273842068469,
      "loss": 6.2147,
      "mean_token_accuracy": 0.16751194298267363,
      "num_tokens": 18226767.0,
      "step": 7190
    },
    {
      "entropy": 6.25657548904419,
      "epoch": 0.8303519907674553,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004942180933906311,
      "loss": 6.1392,
      "mean_token_accuracy": 0.16658232510089874,
      "num_tokens": 18239250.0,
      "step": 7195
    },
    {
      "entropy": 6.193741655349731,
      "epoch": 0.8309290248124639,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004942087952011237,
      "loss": 6.194,
      "mean_token_accuracy": 0.1681028738617897,
      "num_tokens": 18252053.0,
      "step": 7200
    },
    {
      "entropy": 6.285857248306274,
      "epoch": 0.8315060588574726,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004941994896386374,
      "loss": 6.2336,
      "mean_token_accuracy": 0.15824792087078093,
      "num_tokens": 18264769.0,
      "step": 7205
    },
    {
      "entropy": 6.349163341522217,
      "epoch": 0.8320830929024813,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004941901767034851,
      "loss": 6.2638,
      "mean_token_accuracy": 0.15891654938459396,
      "num_tokens": 18278185.0,
      "step": 7210
    },
    {
      "entropy": 6.215603828430176,
      "epoch": 0.8326601269474899,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004941808563959802,
      "loss": 6.1388,
      "mean_token_accuracy": 0.16748507022857667,
      "num_tokens": 18290997.0,
      "step": 7215
    },
    {
      "entropy": 6.169051933288574,
      "epoch": 0.8332371609924986,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004941715287164359,
      "loss": 6.1372,
      "mean_token_accuracy": 0.1674881771206856,
      "num_tokens": 18303735.0,
      "step": 7220
    },
    {
      "entropy": 6.165251302719116,
      "epoch": 0.8338141950375072,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004941621936651661,
      "loss": 6.0948,
      "mean_token_accuracy": 0.1715788647532463,
      "num_tokens": 18316061.0,
      "step": 7225
    },
    {
      "entropy": 6.316934299468994,
      "epoch": 0.8343912290825158,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004941528512424849,
      "loss": 6.1622,
      "mean_token_accuracy": 0.1689636930823326,
      "num_tokens": 18329478.0,
      "step": 7230
    },
    {
      "entropy": 6.305386638641357,
      "epoch": 0.8349682631275245,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004941435014487064,
      "loss": 6.1551,
      "mean_token_accuracy": 0.16330277919769287,
      "num_tokens": 18341790.0,
      "step": 7235
    },
    {
      "entropy": 6.25369668006897,
      "epoch": 0.8355452971725332,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000494134144284145,
      "loss": 6.126,
      "mean_token_accuracy": 0.17124321311712265,
      "num_tokens": 18353807.0,
      "step": 7240
    },
    {
      "entropy": 6.257370948791504,
      "epoch": 0.8361223312175419,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004941247797491156,
      "loss": 6.2095,
      "mean_token_accuracy": 0.15948031544685365,
      "num_tokens": 18366118.0,
      "step": 7245
    },
    {
      "entropy": 6.194285345077515,
      "epoch": 0.8366993652625505,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004941154078439329,
      "loss": 6.1636,
      "mean_token_accuracy": 0.1731121778488159,
      "num_tokens": 18380542.0,
      "step": 7250
    },
    {
      "entropy": 6.335092067718506,
      "epoch": 0.8372763993075591,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004941060285689126,
      "loss": 6.2347,
      "mean_token_accuracy": 0.15947120040655136,
      "num_tokens": 18393976.0,
      "step": 7255
    },
    {
      "entropy": 6.240130519866943,
      "epoch": 0.8378534333525678,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004940966419243695,
      "loss": 6.1869,
      "mean_token_accuracy": 0.16360327899456023,
      "num_tokens": 18405770.0,
      "step": 7260
    },
    {
      "entropy": 6.286680459976196,
      "epoch": 0.8384304673975764,
      "grad_norm": 0.96875,
      "learning_rate": 0.00049408724791062,
      "loss": 6.2777,
      "mean_token_accuracy": 0.16267006397247313,
      "num_tokens": 18417980.0,
      "step": 7265
    },
    {
      "entropy": 6.340583944320679,
      "epoch": 0.8390075014425851,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004940778465279797,
      "loss": 6.1462,
      "mean_token_accuracy": 0.1701750338077545,
      "num_tokens": 18429511.0,
      "step": 7270
    },
    {
      "entropy": 6.227135133743286,
      "epoch": 0.8395845354875938,
      "grad_norm": 1.0,
      "learning_rate": 0.0004940684377767647,
      "loss": 6.2203,
      "mean_token_accuracy": 0.16073543578386307,
      "num_tokens": 18443176.0,
      "step": 7275
    },
    {
      "entropy": 6.32651653289795,
      "epoch": 0.8401615695326025,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004940590216572916,
      "loss": 6.1758,
      "mean_token_accuracy": 0.17092685550451278,
      "num_tokens": 18454664.0,
      "step": 7280
    },
    {
      "entropy": 6.199286365509034,
      "epoch": 0.8407386035776111,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004940495981698772,
      "loss": 6.1573,
      "mean_token_accuracy": 0.16465185284614564,
      "num_tokens": 18465919.0,
      "step": 7285
    },
    {
      "entropy": 6.190265893936157,
      "epoch": 0.8413156376226197,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004940401673148384,
      "loss": 6.0785,
      "mean_token_accuracy": 0.16842346340417863,
      "num_tokens": 18478847.0,
      "step": 7290
    },
    {
      "entropy": 6.308824157714843,
      "epoch": 0.8418926716676284,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004940307290924923,
      "loss": 6.2224,
      "mean_token_accuracy": 0.1603875145316124,
      "num_tokens": 18490932.0,
      "step": 7295
    },
    {
      "entropy": 6.292608785629272,
      "epoch": 0.842469705712637,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004940212835031565,
      "loss": 6.2083,
      "mean_token_accuracy": 0.15687716901302337,
      "num_tokens": 18504333.0,
      "step": 7300
    },
    {
      "entropy": 6.273757696151733,
      "epoch": 0.8430467397576457,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004940118305471486,
      "loss": 6.196,
      "mean_token_accuracy": 0.16300477981567382,
      "num_tokens": 18516021.0,
      "step": 7305
    },
    {
      "entropy": 6.172683048248291,
      "epoch": 0.8436237738026544,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004940023702247866,
      "loss": 6.0876,
      "mean_token_accuracy": 0.17308575958013533,
      "num_tokens": 18529281.0,
      "step": 7310
    },
    {
      "entropy": 6.2051478862762455,
      "epoch": 0.844200807847663,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004939929025363886,
      "loss": 6.1656,
      "mean_token_accuracy": 0.16928713619709015,
      "num_tokens": 18541605.0,
      "step": 7315
    },
    {
      "entropy": 6.328506088256836,
      "epoch": 0.8447778418926717,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004939834274822731,
      "loss": 6.2014,
      "mean_token_accuracy": 0.1676449403166771,
      "num_tokens": 18553171.0,
      "step": 7320
    },
    {
      "entropy": 6.282281589508057,
      "epoch": 0.8453548759376803,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004939739450627588,
      "loss": 6.0708,
      "mean_token_accuracy": 0.16659154295921325,
      "num_tokens": 18565347.0,
      "step": 7325
    },
    {
      "entropy": 6.174135971069336,
      "epoch": 0.845931909982689,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004939644552781647,
      "loss": 6.2313,
      "mean_token_accuracy": 0.16193186789751052,
      "num_tokens": 18578102.0,
      "step": 7330
    },
    {
      "entropy": 6.2800623893737795,
      "epoch": 0.8465089440276976,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004939549581288099,
      "loss": 6.1935,
      "mean_token_accuracy": 0.16096053570508956,
      "num_tokens": 18590184.0,
      "step": 7335
    },
    {
      "entropy": 6.299115943908691,
      "epoch": 0.8470859780727062,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004939454536150138,
      "loss": 6.2224,
      "mean_token_accuracy": 0.16214278042316438,
      "num_tokens": 18603026.0,
      "step": 7340
    },
    {
      "entropy": 6.23553318977356,
      "epoch": 0.847663012117715,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000493935941737096,
      "loss": 6.2067,
      "mean_token_accuracy": 0.15990966856479644,
      "num_tokens": 18616841.0,
      "step": 7345
    },
    {
      "entropy": 6.27131986618042,
      "epoch": 0.8482400461627236,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004939264224953768,
      "loss": 6.1487,
      "mean_token_accuracy": 0.17177173346281052,
      "num_tokens": 18628777.0,
      "step": 7350
    },
    {
      "entropy": 6.308268356323242,
      "epoch": 0.8488170802077323,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000493916895890176,
      "loss": 6.1974,
      "mean_token_accuracy": 0.15837332606315613,
      "num_tokens": 18641626.0,
      "step": 7355
    },
    {
      "entropy": 6.3466332912445065,
      "epoch": 0.8493941142527409,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000493907361921814,
      "loss": 6.1786,
      "mean_token_accuracy": 0.16663186699151994,
      "num_tokens": 18653576.0,
      "step": 7360
    },
    {
      "entropy": 6.185823631286621,
      "epoch": 0.8499711482977496,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004938978205906118,
      "loss": 6.2137,
      "mean_token_accuracy": 0.16485827267169953,
      "num_tokens": 18666134.0,
      "step": 7365
    },
    {
      "entropy": 6.259935998916626,
      "epoch": 0.8505481823427582,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004938882718968901,
      "loss": 6.1803,
      "mean_token_accuracy": 0.15593712627887726,
      "num_tokens": 18680029.0,
      "step": 7370
    },
    {
      "entropy": 6.341277742385865,
      "epoch": 0.8511252163877668,
      "grad_norm": 1.0,
      "learning_rate": 0.0004938787158409702,
      "loss": 6.184,
      "mean_token_accuracy": 0.16573450118303298,
      "num_tokens": 18693136.0,
      "step": 7375
    },
    {
      "entropy": 6.18972110748291,
      "epoch": 0.8517022504327756,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004938691524231733,
      "loss": 6.2416,
      "mean_token_accuracy": 0.1579619526863098,
      "num_tokens": 18706613.0,
      "step": 7380
    },
    {
      "entropy": 6.2843207836151125,
      "epoch": 0.8522792844777842,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004938595816438212,
      "loss": 6.2177,
      "mean_token_accuracy": 0.15311624854803085,
      "num_tokens": 18719541.0,
      "step": 7385
    },
    {
      "entropy": 6.303366565704346,
      "epoch": 0.8528563185227929,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004938500035032358,
      "loss": 6.2071,
      "mean_token_accuracy": 0.16861853003501892,
      "num_tokens": 18730992.0,
      "step": 7390
    },
    {
      "entropy": 6.251425552368164,
      "epoch": 0.8534333525678015,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004938404180017392,
      "loss": 6.1304,
      "mean_token_accuracy": 0.16327824145555497,
      "num_tokens": 18743945.0,
      "step": 7395
    },
    {
      "entropy": 6.181228446960449,
      "epoch": 0.8540103866128101,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004938308251396539,
      "loss": 6.224,
      "mean_token_accuracy": 0.16805968433618546,
      "num_tokens": 18756656.0,
      "step": 7400
    },
    {
      "entropy": 6.279729652404785,
      "epoch": 0.8545874206578188,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004938212249173024,
      "loss": 6.1528,
      "mean_token_accuracy": 0.17048481553792955,
      "num_tokens": 18769224.0,
      "step": 7405
    },
    {
      "entropy": 6.28008451461792,
      "epoch": 0.8551644547028274,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004938116173350078,
      "loss": 6.1611,
      "mean_token_accuracy": 0.17299398332834243,
      "num_tokens": 18782162.0,
      "step": 7410
    },
    {
      "entropy": 6.245051431655884,
      "epoch": 0.8557414887478362,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004938020023930932,
      "loss": 6.2297,
      "mean_token_accuracy": 0.1609863668680191,
      "num_tokens": 18793014.0,
      "step": 7415
    },
    {
      "entropy": 6.323724889755249,
      "epoch": 0.8563185227928448,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004937923800918817,
      "loss": 6.265,
      "mean_token_accuracy": 0.16229446828365326,
      "num_tokens": 18804879.0,
      "step": 7420
    },
    {
      "entropy": 6.287072467803955,
      "epoch": 0.8568955568378535,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004937827504316974,
      "loss": 6.1358,
      "mean_token_accuracy": 0.1653107985854149,
      "num_tokens": 18817301.0,
      "step": 7425
    },
    {
      "entropy": 6.187091636657715,
      "epoch": 0.8574725908828621,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004937731134128639,
      "loss": 6.1667,
      "mean_token_accuracy": 0.1633290708065033,
      "num_tokens": 18828424.0,
      "step": 7430
    },
    {
      "entropy": 6.227798700332642,
      "epoch": 0.8580496249278707,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004937634690357054,
      "loss": 6.1888,
      "mean_token_accuracy": 0.16297754198312758,
      "num_tokens": 18840526.0,
      "step": 7435
    },
    {
      "entropy": 6.337333679199219,
      "epoch": 0.8586266589728794,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004937538173005462,
      "loss": 6.2682,
      "mean_token_accuracy": 0.1581204354763031,
      "num_tokens": 18854557.0,
      "step": 7440
    },
    {
      "entropy": 6.2702559471130375,
      "epoch": 0.859203693017888,
      "grad_norm": 1.0,
      "learning_rate": 0.0004937441582077111,
      "loss": 6.2139,
      "mean_token_accuracy": 0.1641816481947899,
      "num_tokens": 18868469.0,
      "step": 7445
    },
    {
      "entropy": 6.280657243728638,
      "epoch": 0.8597807270628968,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004937344917575249,
      "loss": 6.1685,
      "mean_token_accuracy": 0.16400496810674667,
      "num_tokens": 18880507.0,
      "step": 7450
    },
    {
      "entropy": 6.350762271881104,
      "epoch": 0.8603577611079054,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004937248179503127,
      "loss": 6.2474,
      "mean_token_accuracy": 0.15723391324281694,
      "num_tokens": 18892666.0,
      "step": 7455
    },
    {
      "entropy": 6.2994081497192385,
      "epoch": 0.860934795152914,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004937151367863998,
      "loss": 6.2255,
      "mean_token_accuracy": 0.1623070165514946,
      "num_tokens": 18905022.0,
      "step": 7460
    },
    {
      "entropy": 6.271572542190552,
      "epoch": 0.8615118291979227,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000493705448266112,
      "loss": 6.2561,
      "mean_token_accuracy": 0.1611558198928833,
      "num_tokens": 18918279.0,
      "step": 7465
    },
    {
      "entropy": 6.278350019454956,
      "epoch": 0.8620888632429313,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004936957523897752,
      "loss": 6.137,
      "mean_token_accuracy": 0.16462632715702058,
      "num_tokens": 18930494.0,
      "step": 7470
    },
    {
      "entropy": 6.234484338760376,
      "epoch": 0.86266589728794,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004936860491577153,
      "loss": 6.1079,
      "mean_token_accuracy": 0.16864698380231857,
      "num_tokens": 18943918.0,
      "step": 7475
    },
    {
      "entropy": 6.3203222274780275,
      "epoch": 0.8632429313329486,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004936763385702588,
      "loss": 6.2284,
      "mean_token_accuracy": 0.16031141877174376,
      "num_tokens": 18957170.0,
      "step": 7480
    },
    {
      "entropy": 6.225881099700928,
      "epoch": 0.8638199653779572,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004936666206277322,
      "loss": 6.1369,
      "mean_token_accuracy": 0.15548645108938217,
      "num_tokens": 18970106.0,
      "step": 7485
    },
    {
      "entropy": 6.340928936004639,
      "epoch": 0.864396999422966,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004936568953304624,
      "loss": 6.2257,
      "mean_token_accuracy": 0.1618570491671562,
      "num_tokens": 18982797.0,
      "step": 7490
    },
    {
      "entropy": 6.334622764587403,
      "epoch": 0.8649740334679746,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004936471626787766,
      "loss": 6.2881,
      "mean_token_accuracy": 0.1659012496471405,
      "num_tokens": 18997122.0,
      "step": 7495
    },
    {
      "entropy": 6.311051559448242,
      "epoch": 0.8655510675129833,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004936374226730022,
      "loss": 6.2472,
      "mean_token_accuracy": 0.15797540843486785,
      "num_tokens": 19009345.0,
      "step": 7500
    },
    {
      "entropy": 6.3252246379852295,
      "epoch": 0.8661281015579919,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004936276753134666,
      "loss": 6.2369,
      "mean_token_accuracy": 0.1617169827222824,
      "num_tokens": 19021585.0,
      "step": 7505
    },
    {
      "entropy": 6.257892799377442,
      "epoch": 0.8667051356030006,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004936179206004976,
      "loss": 6.2072,
      "mean_token_accuracy": 0.15795013010501863,
      "num_tokens": 19034701.0,
      "step": 7510
    },
    {
      "entropy": 6.239010572433472,
      "epoch": 0.8672821696480092,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004936081585344236,
      "loss": 6.1975,
      "mean_token_accuracy": 0.16772165298461914,
      "num_tokens": 19047810.0,
      "step": 7515
    },
    {
      "entropy": 6.275329446792602,
      "epoch": 0.8678592036930178,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004935983891155727,
      "loss": 6.1711,
      "mean_token_accuracy": 0.1709216982126236,
      "num_tokens": 19059546.0,
      "step": 7520
    },
    {
      "entropy": 6.3266302108764645,
      "epoch": 0.8684362377380266,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004935886123442737,
      "loss": 6.2395,
      "mean_token_accuracy": 0.16143498867750167,
      "num_tokens": 19072302.0,
      "step": 7525
    },
    {
      "entropy": 6.354724788665772,
      "epoch": 0.8690132717830352,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004935788282208551,
      "loss": 6.2354,
      "mean_token_accuracy": 0.16271546334028245,
      "num_tokens": 19086283.0,
      "step": 7530
    },
    {
      "entropy": 6.258589458465576,
      "epoch": 0.8695903058280439,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004935690367456464,
      "loss": 6.1464,
      "mean_token_accuracy": 0.16817453801631926,
      "num_tokens": 19098586.0,
      "step": 7535
    },
    {
      "entropy": 6.18815016746521,
      "epoch": 0.8701673398730525,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004935592379189766,
      "loss": 6.139,
      "mean_token_accuracy": 0.16720104068517685,
      "num_tokens": 19110795.0,
      "step": 7540
    },
    {
      "entropy": 6.295345783233643,
      "epoch": 0.8707443739180611,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004935494317411754,
      "loss": 6.1126,
      "mean_token_accuracy": 0.16858818680047988,
      "num_tokens": 19123391.0,
      "step": 7545
    },
    {
      "entropy": 6.228914213180542,
      "epoch": 0.8713214079630698,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004935396182125726,
      "loss": 6.1767,
      "mean_token_accuracy": 0.16556897908449172,
      "num_tokens": 19135436.0,
      "step": 7550
    },
    {
      "entropy": 6.255138540267945,
      "epoch": 0.8718984420080784,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004935297973334983,
      "loss": 6.143,
      "mean_token_accuracy": 0.16952496469020845,
      "num_tokens": 19147210.0,
      "step": 7555
    },
    {
      "entropy": 6.286662626266479,
      "epoch": 0.8724754760530872,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004935199691042828,
      "loss": 6.2877,
      "mean_token_accuracy": 0.16529511064291,
      "num_tokens": 19161309.0,
      "step": 7560
    },
    {
      "entropy": 6.318764019012451,
      "epoch": 0.8730525100980958,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004935101335252568,
      "loss": 6.2291,
      "mean_token_accuracy": 0.16191016063094138,
      "num_tokens": 19174073.0,
      "step": 7565
    },
    {
      "entropy": 6.27949743270874,
      "epoch": 0.8736295441431045,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004935002905967509,
      "loss": 6.1201,
      "mean_token_accuracy": 0.1638078808784485,
      "num_tokens": 19187120.0,
      "step": 7570
    },
    {
      "entropy": 6.254374551773071,
      "epoch": 0.8742065781881131,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004934904403190963,
      "loss": 6.0853,
      "mean_token_accuracy": 0.1709825038909912,
      "num_tokens": 19201195.0,
      "step": 7575
    },
    {
      "entropy": 6.245515060424805,
      "epoch": 0.8747836122331217,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004934805826926242,
      "loss": 6.223,
      "mean_token_accuracy": 0.15933709740638732,
      "num_tokens": 19214050.0,
      "step": 7580
    },
    {
      "entropy": 6.328765058517456,
      "epoch": 0.8753606462781304,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004934707177176663,
      "loss": 6.2157,
      "mean_token_accuracy": 0.16172537207603455,
      "num_tokens": 19226678.0,
      "step": 7585
    },
    {
      "entropy": 6.320084714889527,
      "epoch": 0.875937680323139,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004934608453945543,
      "loss": 6.2491,
      "mean_token_accuracy": 0.1552565038204193,
      "num_tokens": 19239038.0,
      "step": 7590
    },
    {
      "entropy": 6.058368110656739,
      "epoch": 0.8765147143681478,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004934509657236203,
      "loss": 6.1323,
      "mean_token_accuracy": 0.16718279719352722,
      "num_tokens": 19252934.0,
      "step": 7595
    },
    {
      "entropy": 6.346228551864624,
      "epoch": 0.8770917484131564,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004934410787051965,
      "loss": 6.1857,
      "mean_token_accuracy": 0.16360338181257247,
      "num_tokens": 19267646.0,
      "step": 7600
    },
    {
      "entropy": 6.345105838775635,
      "epoch": 0.877668782458165,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004934311843396157,
      "loss": 6.1817,
      "mean_token_accuracy": 0.1632274717092514,
      "num_tokens": 19279529.0,
      "step": 7605
    },
    {
      "entropy": 6.276775598526001,
      "epoch": 0.8782458165031737,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004934212826272104,
      "loss": 6.2216,
      "mean_token_accuracy": 0.1661163493990898,
      "num_tokens": 19291315.0,
      "step": 7610
    },
    {
      "entropy": 6.220012092590332,
      "epoch": 0.8788228505481823,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004934113735683137,
      "loss": 6.0703,
      "mean_token_accuracy": 0.17227137684822083,
      "num_tokens": 19303763.0,
      "step": 7615
    },
    {
      "entropy": 6.244149112701416,
      "epoch": 0.879399884593191,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000493401457163259,
      "loss": 6.2251,
      "mean_token_accuracy": 0.15755569338798522,
      "num_tokens": 19316400.0,
      "step": 7620
    },
    {
      "entropy": 6.280155801773072,
      "epoch": 0.8799769186381996,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004933915334123798,
      "loss": 6.1195,
      "mean_token_accuracy": 0.17527248710393906,
      "num_tokens": 19328227.0,
      "step": 7625
    },
    {
      "entropy": 6.170809268951416,
      "epoch": 0.8805539526832084,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004933816023160099,
      "loss": 6.0741,
      "mean_token_accuracy": 0.17317016869783403,
      "num_tokens": 19341211.0,
      "step": 7630
    },
    {
      "entropy": 6.2988762855529785,
      "epoch": 0.881130986728217,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004933716638744832,
      "loss": 6.2142,
      "mean_token_accuracy": 0.16570787876844406,
      "num_tokens": 19353165.0,
      "step": 7635
    },
    {
      "entropy": 6.259564447402954,
      "epoch": 0.8817080207732256,
      "grad_norm": 1.6640625,
      "learning_rate": 0.000493361718088134,
      "loss": 6.3295,
      "mean_token_accuracy": 0.15990637391805648,
      "num_tokens": 19365601.0,
      "step": 7640
    },
    {
      "entropy": 6.335362768173217,
      "epoch": 0.8822850548182343,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000493351764957297,
      "loss": 6.1819,
      "mean_token_accuracy": 0.16016467064619064,
      "num_tokens": 19377828.0,
      "step": 7645
    },
    {
      "entropy": 6.357305288314819,
      "epoch": 0.8828620888632429,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004933418044823068,
      "loss": 6.195,
      "mean_token_accuracy": 0.16444187760353088,
      "num_tokens": 19390019.0,
      "step": 7650
    },
    {
      "entropy": 6.219047880172729,
      "epoch": 0.8834391229082516,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004933318366634984,
      "loss": 6.1609,
      "mean_token_accuracy": 0.16621898859739304,
      "num_tokens": 19402179.0,
      "step": 7655
    },
    {
      "entropy": 6.24358491897583,
      "epoch": 0.8840161569532602,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004933218615012072,
      "loss": 6.1299,
      "mean_token_accuracy": 0.16143215745687484,
      "num_tokens": 19412794.0,
      "step": 7660
    },
    {
      "entropy": 6.246716451644898,
      "epoch": 0.8845931909982689,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004933118789957687,
      "loss": 6.2116,
      "mean_token_accuracy": 0.15541283041238785,
      "num_tokens": 19425932.0,
      "step": 7665
    },
    {
      "entropy": 6.297798824310303,
      "epoch": 0.8851702250432776,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004933018891475186,
      "loss": 6.1336,
      "mean_token_accuracy": 0.1639110490679741,
      "num_tokens": 19437543.0,
      "step": 7670
    },
    {
      "entropy": 6.267187070846558,
      "epoch": 0.8857472590882862,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004932918919567927,
      "loss": 6.1196,
      "mean_token_accuracy": 0.17057135850191116,
      "num_tokens": 19450122.0,
      "step": 7675
    },
    {
      "entropy": 6.120025968551635,
      "epoch": 0.8863242931332949,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004932818874239275,
      "loss": 6.1427,
      "mean_token_accuracy": 0.16066610366106032,
      "num_tokens": 19464553.0,
      "step": 7680
    },
    {
      "entropy": 6.179195547103882,
      "epoch": 0.8869013271783035,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004932718755492595,
      "loss": 6.1007,
      "mean_token_accuracy": 0.16897430121898652,
      "num_tokens": 19477956.0,
      "step": 7685
    },
    {
      "entropy": 6.259118700027466,
      "epoch": 0.8874783612233121,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004932618563331254,
      "loss": 6.1514,
      "mean_token_accuracy": 0.16030027568340302,
      "num_tokens": 19491610.0,
      "step": 7690
    },
    {
      "entropy": 6.208391571044922,
      "epoch": 0.8880553952683208,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004932518297758622,
      "loss": 6.0502,
      "mean_token_accuracy": 0.1690255731344223,
      "num_tokens": 19503612.0,
      "step": 7695
    },
    {
      "entropy": 6.168671274185181,
      "epoch": 0.8886324293133295,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004932417958778071,
      "loss": 6.1344,
      "mean_token_accuracy": 0.16841337829828262,
      "num_tokens": 19516956.0,
      "step": 7700
    },
    {
      "entropy": 6.1559232711792,
      "epoch": 0.8892094633583382,
      "grad_norm": 1.0,
      "learning_rate": 0.0004932317546392976,
      "loss": 6.1254,
      "mean_token_accuracy": 0.166995607316494,
      "num_tokens": 19528545.0,
      "step": 7705
    },
    {
      "entropy": 6.2624321460723875,
      "epoch": 0.8897864974033468,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004932217060606714,
      "loss": 6.0518,
      "mean_token_accuracy": 0.17175974249839782,
      "num_tokens": 19539920.0,
      "step": 7710
    },
    {
      "entropy": 6.230217599868775,
      "epoch": 0.8903635314483554,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004932116501422665,
      "loss": 6.1678,
      "mean_token_accuracy": 0.1643459051847458,
      "num_tokens": 19551975.0,
      "step": 7715
    },
    {
      "entropy": 6.243873405456543,
      "epoch": 0.8909405654933641,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004932015868844211,
      "loss": 6.1765,
      "mean_token_accuracy": 0.16330357640981674,
      "num_tokens": 19565952.0,
      "step": 7720
    },
    {
      "entropy": 6.218496561050415,
      "epoch": 0.8915175995383727,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004931915162874738,
      "loss": 6.1334,
      "mean_token_accuracy": 0.16783367693424225,
      "num_tokens": 19578400.0,
      "step": 7725
    },
    {
      "entropy": 6.194408941268921,
      "epoch": 0.8920946335833814,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004931814383517632,
      "loss": 6.1063,
      "mean_token_accuracy": 0.1704055592417717,
      "num_tokens": 19592237.0,
      "step": 7730
    },
    {
      "entropy": 6.316039419174194,
      "epoch": 0.8926716676283901,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004931713530776284,
      "loss": 6.2748,
      "mean_token_accuracy": 0.16063932180404664,
      "num_tokens": 19606172.0,
      "step": 7735
    },
    {
      "entropy": 6.264684247970581,
      "epoch": 0.8932487016733988,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004931612604654083,
      "loss": 6.2387,
      "mean_token_accuracy": 0.16038593649864197,
      "num_tokens": 19618245.0,
      "step": 7740
    },
    {
      "entropy": 6.32139253616333,
      "epoch": 0.8938257357184074,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004931511605154428,
      "loss": 6.1754,
      "mean_token_accuracy": 0.16571465879678726,
      "num_tokens": 19630921.0,
      "step": 7745
    },
    {
      "entropy": 6.262192964553833,
      "epoch": 0.894402769763416,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004931410532280711,
      "loss": 6.1781,
      "mean_token_accuracy": 0.16053754091262817,
      "num_tokens": 19643869.0,
      "step": 7750
    },
    {
      "entropy": 6.268033361434936,
      "epoch": 0.8949798038084247,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004931309386036337,
      "loss": 6.167,
      "mean_token_accuracy": 0.16265616416931153,
      "num_tokens": 19657819.0,
      "step": 7755
    },
    {
      "entropy": 6.209354496002197,
      "epoch": 0.8955568378534333,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004931208166424704,
      "loss": 6.2356,
      "mean_token_accuracy": 0.16021609604358672,
      "num_tokens": 19670847.0,
      "step": 7760
    },
    {
      "entropy": 6.2668853282928465,
      "epoch": 0.896133871898442,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004931106873449219,
      "loss": 6.146,
      "mean_token_accuracy": 0.1613035649061203,
      "num_tokens": 19684055.0,
      "step": 7765
    },
    {
      "entropy": 6.176290988922119,
      "epoch": 0.8967109059434507,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004931005507113285,
      "loss": 6.0642,
      "mean_token_accuracy": 0.1686693862080574,
      "num_tokens": 19696731.0,
      "step": 7770
    },
    {
      "entropy": 6.292421436309814,
      "epoch": 0.8972879399884593,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004930904067420317,
      "loss": 6.205,
      "mean_token_accuracy": 0.16288371682167052,
      "num_tokens": 19708787.0,
      "step": 7775
    },
    {
      "entropy": 6.2655213356018065,
      "epoch": 0.897864974033468,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004930802554373723,
      "loss": 6.2085,
      "mean_token_accuracy": 0.1596529468894005,
      "num_tokens": 19722570.0,
      "step": 7780
    },
    {
      "entropy": 6.331230306625367,
      "epoch": 0.8984420080784766,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004930700967976918,
      "loss": 6.2074,
      "mean_token_accuracy": 0.16459772139787673,
      "num_tokens": 19734690.0,
      "step": 7785
    },
    {
      "entropy": 6.2219703674316404,
      "epoch": 0.8990190421234853,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000493059930823332,
      "loss": 6.2014,
      "mean_token_accuracy": 0.161789208650589,
      "num_tokens": 19747907.0,
      "step": 7790
    },
    {
      "entropy": 6.223333263397217,
      "epoch": 0.8995960761684939,
      "grad_norm": 1.0,
      "learning_rate": 0.0004930497575146346,
      "loss": 6.1666,
      "mean_token_accuracy": 0.1622400999069214,
      "num_tokens": 19760869.0,
      "step": 7795
    },
    {
      "entropy": 6.236398792266845,
      "epoch": 0.9001731102135025,
      "grad_norm": 1.0,
      "learning_rate": 0.0004930395768719421,
      "loss": 6.0685,
      "mean_token_accuracy": 0.17106165438890458,
      "num_tokens": 19773097.0,
      "step": 7800
    },
    {
      "entropy": 6.196551179885864,
      "epoch": 0.9007501442585113,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004930293888955966,
      "loss": 6.1912,
      "mean_token_accuracy": 0.15834959149360656,
      "num_tokens": 19786070.0,
      "step": 7805
    },
    {
      "entropy": 6.301648902893066,
      "epoch": 0.9013271783035199,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000493019193585941,
      "loss": 6.1699,
      "mean_token_accuracy": 0.16527480632066727,
      "num_tokens": 19799244.0,
      "step": 7810
    },
    {
      "entropy": 6.230715274810791,
      "epoch": 0.9019042123485286,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000493008990943318,
      "loss": 6.2074,
      "mean_token_accuracy": 0.1580963686108589,
      "num_tokens": 19811965.0,
      "step": 7815
    },
    {
      "entropy": 6.2934356212615965,
      "epoch": 0.9024812463935372,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004929987809680709,
      "loss": 6.1913,
      "mean_token_accuracy": 0.16129925698041916,
      "num_tokens": 19825531.0,
      "step": 7820
    },
    {
      "entropy": 6.274839973449707,
      "epoch": 0.9030582804385459,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004929885636605432,
      "loss": 6.1262,
      "mean_token_accuracy": 0.16835285872220992,
      "num_tokens": 19839166.0,
      "step": 7825
    },
    {
      "entropy": 6.204418516159057,
      "epoch": 0.9036353144835545,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004929783390210784,
      "loss": 6.1644,
      "mean_token_accuracy": 0.16296615898609162,
      "num_tokens": 19852187.0,
      "step": 7830
    },
    {
      "entropy": 6.249357986450195,
      "epoch": 0.9042123485285631,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004929681070500204,
      "loss": 6.2238,
      "mean_token_accuracy": 0.15767208635807037,
      "num_tokens": 19864309.0,
      "step": 7835
    },
    {
      "entropy": 6.309257173538208,
      "epoch": 0.9047893825735719,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004929578677477135,
      "loss": 6.2144,
      "mean_token_accuracy": 0.15406629145145417,
      "num_tokens": 19877475.0,
      "step": 7840
    },
    {
      "entropy": 6.312673759460449,
      "epoch": 0.9053664166185805,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004929476211145019,
      "loss": 6.1561,
      "mean_token_accuracy": 0.15989240407943725,
      "num_tokens": 19889984.0,
      "step": 7845
    },
    {
      "entropy": 6.212323331832886,
      "epoch": 0.9059434506635892,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004929373671507303,
      "loss": 6.106,
      "mean_token_accuracy": 0.1751035049557686,
      "num_tokens": 19902135.0,
      "step": 7850
    },
    {
      "entropy": 6.204410600662231,
      "epoch": 0.9065204847085978,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004929271058567436,
      "loss": 6.1522,
      "mean_token_accuracy": 0.16679489761590957,
      "num_tokens": 19915214.0,
      "step": 7855
    },
    {
      "entropy": 6.247037124633789,
      "epoch": 0.9070975187536064,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000492916837232887,
      "loss": 6.1772,
      "mean_token_accuracy": 0.15965988636016845,
      "num_tokens": 19929113.0,
      "step": 7860
    },
    {
      "entropy": 6.3158618927001955,
      "epoch": 0.9076745527986151,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004929065612795058,
      "loss": 6.2175,
      "mean_token_accuracy": 0.1604028031229973,
      "num_tokens": 19942727.0,
      "step": 7865
    },
    {
      "entropy": 6.240792608261108,
      "epoch": 0.9082515868436237,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004928962779969456,
      "loss": 6.1234,
      "mean_token_accuracy": 0.16545403003692627,
      "num_tokens": 19954863.0,
      "step": 7870
    },
    {
      "entropy": 6.164105224609375,
      "epoch": 0.9088286208886325,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004928859873855524,
      "loss": 6.0879,
      "mean_token_accuracy": 0.17442544400691987,
      "num_tokens": 19967555.0,
      "step": 7875
    },
    {
      "entropy": 6.332518196105957,
      "epoch": 0.9094056549336411,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004928756894456723,
      "loss": 6.2102,
      "mean_token_accuracy": 0.16057728379964828,
      "num_tokens": 19979992.0,
      "step": 7880
    },
    {
      "entropy": 6.301041126251221,
      "epoch": 0.9099826889786498,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004928653841776515,
      "loss": 6.238,
      "mean_token_accuracy": 0.15930857062339782,
      "num_tokens": 19993231.0,
      "step": 7885
    },
    {
      "entropy": 6.298246097564697,
      "epoch": 0.9105597230236584,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004928550715818368,
      "loss": 6.2143,
      "mean_token_accuracy": 0.15833520889282227,
      "num_tokens": 20005839.0,
      "step": 7890
    },
    {
      "entropy": 6.232977724075317,
      "epoch": 0.911136757068667,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004928447516585749,
      "loss": 6.1732,
      "mean_token_accuracy": 0.16366972774267197,
      "num_tokens": 20017423.0,
      "step": 7895
    },
    {
      "entropy": 6.221403169631958,
      "epoch": 0.9117137911136757,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000492834424408213,
      "loss": 6.2058,
      "mean_token_accuracy": 0.1618037447333336,
      "num_tokens": 20030989.0,
      "step": 7900
    },
    {
      "entropy": 6.335241270065308,
      "epoch": 0.9122908251586843,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004928240898310984,
      "loss": 6.2622,
      "mean_token_accuracy": 0.1539841316640377,
      "num_tokens": 20043763.0,
      "step": 7905
    },
    {
      "entropy": 6.299007415771484,
      "epoch": 0.9128678592036931,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004928137479275789,
      "loss": 6.109,
      "mean_token_accuracy": 0.17287227511405945,
      "num_tokens": 20057012.0,
      "step": 7910
    },
    {
      "entropy": 6.196064376831055,
      "epoch": 0.9134448932487017,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000492803398698002,
      "loss": 6.1467,
      "mean_token_accuracy": 0.16344195157289504,
      "num_tokens": 20069093.0,
      "step": 7915
    },
    {
      "entropy": 6.206642723083496,
      "epoch": 0.9140219272937103,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004927930421427161,
      "loss": 6.1289,
      "mean_token_accuracy": 0.16734737753868104,
      "num_tokens": 20081392.0,
      "step": 7920
    },
    {
      "entropy": 6.200254678726196,
      "epoch": 0.914598961338719,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004927826782620694,
      "loss": 6.1336,
      "mean_token_accuracy": 0.16513927578926085,
      "num_tokens": 20092621.0,
      "step": 7925
    },
    {
      "entropy": 6.2276640892028805,
      "epoch": 0.9151759953837276,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004927723070564104,
      "loss": 6.1594,
      "mean_token_accuracy": 0.16101684272289277,
      "num_tokens": 20105423.0,
      "step": 7930
    },
    {
      "entropy": 6.3159819602966305,
      "epoch": 0.9157530294287363,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004927619285260881,
      "loss": 6.1453,
      "mean_token_accuracy": 0.1588059961795807,
      "num_tokens": 20117835.0,
      "step": 7935
    },
    {
      "entropy": 6.15804443359375,
      "epoch": 0.9163300634737449,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004927515426714515,
      "loss": 6.0449,
      "mean_token_accuracy": 0.1740433931350708,
      "num_tokens": 20130563.0,
      "step": 7940
    },
    {
      "entropy": 6.079877805709839,
      "epoch": 0.9169070975187537,
      "grad_norm": 0.90625,
      "learning_rate": 0.00049274114949285,
      "loss": 6.1307,
      "mean_token_accuracy": 0.16842392086982727,
      "num_tokens": 20144507.0,
      "step": 7945
    },
    {
      "entropy": 6.287519216537476,
      "epoch": 0.9174841315637623,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000492730748990633,
      "loss": 6.143,
      "mean_token_accuracy": 0.16973081678152085,
      "num_tokens": 20156341.0,
      "step": 7950
    },
    {
      "entropy": 6.212476110458374,
      "epoch": 0.9180611656087709,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004927203411651504,
      "loss": 6.1863,
      "mean_token_accuracy": 0.16502414047718048,
      "num_tokens": 20169410.0,
      "step": 7955
    },
    {
      "entropy": 6.247562980651855,
      "epoch": 0.9186381996537796,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004927099260167523,
      "loss": 6.1849,
      "mean_token_accuracy": 0.16723521798849106,
      "num_tokens": 20181233.0,
      "step": 7960
    },
    {
      "entropy": 6.300972557067871,
      "epoch": 0.9192152336987882,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004926995035457889,
      "loss": 6.1888,
      "mean_token_accuracy": 0.16301036924123763,
      "num_tokens": 20193234.0,
      "step": 7965
    },
    {
      "entropy": 6.208461761474609,
      "epoch": 0.9197922677437969,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000492689073752611,
      "loss": 6.106,
      "mean_token_accuracy": 0.16746279150247573,
      "num_tokens": 20206784.0,
      "step": 7970
    },
    {
      "entropy": 6.145061016082764,
      "epoch": 0.9203693017888055,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004926786366375691,
      "loss": 6.0977,
      "mean_token_accuracy": 0.1713881641626358,
      "num_tokens": 20218597.0,
      "step": 7975
    },
    {
      "entropy": 6.1318779468536375,
      "epoch": 0.9209463358338142,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004926681922010145,
      "loss": 6.0553,
      "mean_token_accuracy": 0.17041801065206527,
      "num_tokens": 20232703.0,
      "step": 7980
    },
    {
      "entropy": 6.240469217300415,
      "epoch": 0.9215233698788229,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004926577404432982,
      "loss": 6.1428,
      "mean_token_accuracy": 0.1627808168530464,
      "num_tokens": 20245811.0,
      "step": 7985
    },
    {
      "entropy": 6.147562026977539,
      "epoch": 0.9221004039238315,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004926472813647721,
      "loss": 6.0427,
      "mean_token_accuracy": 0.16625858694314957,
      "num_tokens": 20258055.0,
      "step": 7990
    },
    {
      "entropy": 6.182660675048828,
      "epoch": 0.9226774379688402,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004926368149657876,
      "loss": 6.1337,
      "mean_token_accuracy": 0.16746827811002732,
      "num_tokens": 20271250.0,
      "step": 7995
    },
    {
      "entropy": 6.184801864624023,
      "epoch": 0.9232544720138488,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004926263412466972,
      "loss": 6.0799,
      "mean_token_accuracy": 0.17483011782169341,
      "num_tokens": 20284013.0,
      "step": 8000
    },
    {
      "entropy": 6.23457818031311,
      "epoch": 0.9238315060588574,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004926158602078527,
      "loss": 6.1689,
      "mean_token_accuracy": 0.1689364418387413,
      "num_tokens": 20297413.0,
      "step": 8005
    },
    {
      "entropy": 6.154371309280395,
      "epoch": 0.9244085401038661,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004926053718496069,
      "loss": 6.1109,
      "mean_token_accuracy": 0.16915992349386216,
      "num_tokens": 20310026.0,
      "step": 8010
    },
    {
      "entropy": 6.195466709136963,
      "epoch": 0.9249855741488748,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004925948761723126,
      "loss": 6.0992,
      "mean_token_accuracy": 0.1695803016424179,
      "num_tokens": 20322118.0,
      "step": 8015
    },
    {
      "entropy": 6.218642616271973,
      "epoch": 0.9255626081938835,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004925843731763226,
      "loss": 6.1692,
      "mean_token_accuracy": 0.1613602340221405,
      "num_tokens": 20335252.0,
      "step": 8020
    },
    {
      "entropy": 6.332771968841553,
      "epoch": 0.9261396422388921,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004925738628619904,
      "loss": 6.2061,
      "mean_token_accuracy": 0.16417815238237382,
      "num_tokens": 20348201.0,
      "step": 8025
    },
    {
      "entropy": 6.228202247619629,
      "epoch": 0.9267166762839008,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004925633452296693,
      "loss": 6.1304,
      "mean_token_accuracy": 0.16839785873889923,
      "num_tokens": 20360250.0,
      "step": 8030
    },
    {
      "entropy": 6.234412288665771,
      "epoch": 0.9272937103289094,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004925528202797131,
      "loss": 6.1121,
      "mean_token_accuracy": 0.1637949377298355,
      "num_tokens": 20372376.0,
      "step": 8035
    },
    {
      "entropy": 6.260030746459961,
      "epoch": 0.927870744373918,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004925422880124761,
      "loss": 6.1126,
      "mean_token_accuracy": 0.1594625473022461,
      "num_tokens": 20384484.0,
      "step": 8040
    },
    {
      "entropy": 6.215761423110962,
      "epoch": 0.9284477784189267,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004925317484283121,
      "loss": 6.1721,
      "mean_token_accuracy": 0.16395566761493682,
      "num_tokens": 20396519.0,
      "step": 8045
    },
    {
      "entropy": 6.184128761291504,
      "epoch": 0.9290248124639354,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004925212015275758,
      "loss": 6.1447,
      "mean_token_accuracy": 0.16434455364942552,
      "num_tokens": 20409108.0,
      "step": 8050
    },
    {
      "entropy": 6.253138256072998,
      "epoch": 0.9296018465089441,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000492510647310622,
      "loss": 6.1974,
      "mean_token_accuracy": 0.16477228701114655,
      "num_tokens": 20421929.0,
      "step": 8055
    },
    {
      "entropy": 6.186897325515747,
      "epoch": 0.9301788805539527,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004925000857778055,
      "loss": 6.1844,
      "mean_token_accuracy": 0.1656649813055992,
      "num_tokens": 20435453.0,
      "step": 8060
    },
    {
      "entropy": 6.221740627288819,
      "epoch": 0.9307559145989613,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004924895169294818,
      "loss": 6.1426,
      "mean_token_accuracy": 0.16689223200082778,
      "num_tokens": 20447867.0,
      "step": 8065
    },
    {
      "entropy": 6.296292066574097,
      "epoch": 0.93133294864397,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004924789407660062,
      "loss": 6.1889,
      "mean_token_accuracy": 0.17134887278079985,
      "num_tokens": 20461150.0,
      "step": 8070
    },
    {
      "entropy": 6.233514451980591,
      "epoch": 0.9319099826889786,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004924683572877345,
      "loss": 6.0993,
      "mean_token_accuracy": 0.16771567910909652,
      "num_tokens": 20474012.0,
      "step": 8075
    },
    {
      "entropy": 6.21320652961731,
      "epoch": 0.9324870167339873,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004924577664950225,
      "loss": 6.1425,
      "mean_token_accuracy": 0.158681820333004,
      "num_tokens": 20485833.0,
      "step": 8080
    },
    {
      "entropy": 6.278991270065307,
      "epoch": 0.933064050778996,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004924471683882266,
      "loss": 6.1697,
      "mean_token_accuracy": 0.16799163520336152,
      "num_tokens": 20499277.0,
      "step": 8085
    },
    {
      "entropy": 6.22080397605896,
      "epoch": 0.9336410848240047,
      "grad_norm": 1.0,
      "learning_rate": 0.0004924365629677031,
      "loss": 6.1534,
      "mean_token_accuracy": 0.16587530374526976,
      "num_tokens": 20510789.0,
      "step": 8090
    },
    {
      "entropy": 6.208702468872071,
      "epoch": 0.9342181188690133,
      "grad_norm": 1.109375,
      "learning_rate": 0.000492425950233809,
      "loss": 6.1383,
      "mean_token_accuracy": 0.17013276368379593,
      "num_tokens": 20523273.0,
      "step": 8095
    },
    {
      "entropy": 6.274311494827271,
      "epoch": 0.9347951529140219,
      "grad_norm": 0.9375,
      "learning_rate": 0.000492415330186901,
      "loss": 6.1954,
      "mean_token_accuracy": 0.16350119262933732,
      "num_tokens": 20535466.0,
      "step": 8100
    },
    {
      "entropy": 6.344478178024292,
      "epoch": 0.9353721869590306,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004924047028273364,
      "loss": 6.2375,
      "mean_token_accuracy": 0.16144156754016875,
      "num_tokens": 20548108.0,
      "step": 8105
    },
    {
      "entropy": 6.166890001296997,
      "epoch": 0.9359492210040392,
      "grad_norm": 1.0,
      "learning_rate": 0.0004923940681554725,
      "loss": 6.1023,
      "mean_token_accuracy": 0.1647922068834305,
      "num_tokens": 20559859.0,
      "step": 8110
    },
    {
      "entropy": 6.190105819702149,
      "epoch": 0.9365262550490479,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004923834261716672,
      "loss": 6.2392,
      "mean_token_accuracy": 0.16303139328956603,
      "num_tokens": 20572569.0,
      "step": 8115
    },
    {
      "entropy": 6.257817125320434,
      "epoch": 0.9371032890940566,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004923727768762782,
      "loss": 6.1467,
      "mean_token_accuracy": 0.1615183711051941,
      "num_tokens": 20585014.0,
      "step": 8120
    },
    {
      "entropy": 6.2554769039154055,
      "epoch": 0.9376803231390652,
      "grad_norm": 1.046875,
      "learning_rate": 0.000492362120269664,
      "loss": 6.197,
      "mean_token_accuracy": 0.16425224393606186,
      "num_tokens": 20597216.0,
      "step": 8125
    },
    {
      "entropy": 6.146339368820191,
      "epoch": 0.9382573571840739,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004923514563521827,
      "loss": 6.1193,
      "mean_token_accuracy": 0.1756697714328766,
      "num_tokens": 20609193.0,
      "step": 8130
    },
    {
      "entropy": 6.210788679122925,
      "epoch": 0.9388343912290825,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004923407851241933,
      "loss": 6.1332,
      "mean_token_accuracy": 0.16473837196826935,
      "num_tokens": 20620777.0,
      "step": 8135
    },
    {
      "entropy": 6.174554967880249,
      "epoch": 0.9394114252740912,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004923301065860545,
      "loss": 6.1754,
      "mean_token_accuracy": 0.1685077428817749,
      "num_tokens": 20633244.0,
      "step": 8140
    },
    {
      "entropy": 6.30948805809021,
      "epoch": 0.9399884593190998,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004923194207381254,
      "loss": 6.1431,
      "mean_token_accuracy": 0.17263807952404023,
      "num_tokens": 20645622.0,
      "step": 8145
    },
    {
      "entropy": 6.253878784179688,
      "epoch": 0.9405654933641084,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004923087275807656,
      "loss": 6.2323,
      "mean_token_accuracy": 0.1632181480526924,
      "num_tokens": 20658793.0,
      "step": 8150
    },
    {
      "entropy": 6.217817211151123,
      "epoch": 0.9411425274091172,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004922980271143347,
      "loss": 6.0484,
      "mean_token_accuracy": 0.17815047651529312,
      "num_tokens": 20673009.0,
      "step": 8155
    },
    {
      "entropy": 6.2129638671875,
      "epoch": 0.9417195614541258,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004922873193391927,
      "loss": 6.2529,
      "mean_token_accuracy": 0.1666855126619339,
      "num_tokens": 20685061.0,
      "step": 8160
    },
    {
      "entropy": 6.187131643295288,
      "epoch": 0.9422965954991345,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004922766042556994,
      "loss": 6.056,
      "mean_token_accuracy": 0.1666456550359726,
      "num_tokens": 20696854.0,
      "step": 8165
    },
    {
      "entropy": 6.192790985107422,
      "epoch": 0.9428736295441431,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004922658818642156,
      "loss": 6.1992,
      "mean_token_accuracy": 0.1551530748605728,
      "num_tokens": 20709259.0,
      "step": 8170
    },
    {
      "entropy": 6.209788370132446,
      "epoch": 0.9434506635891518,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004922551521651018,
      "loss": 5.9965,
      "mean_token_accuracy": 0.17598087042570115,
      "num_tokens": 20722382.0,
      "step": 8175
    },
    {
      "entropy": 6.131197071075439,
      "epoch": 0.9440276976341604,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004922444151587189,
      "loss": 6.0658,
      "mean_token_accuracy": 0.17491336464881896,
      "num_tokens": 20734809.0,
      "step": 8180
    },
    {
      "entropy": 6.12338604927063,
      "epoch": 0.944604731679169,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004922336708454279,
      "loss": 6.0458,
      "mean_token_accuracy": 0.16772150695323945,
      "num_tokens": 20747490.0,
      "step": 8185
    },
    {
      "entropy": 6.264825630187988,
      "epoch": 0.9451817657241778,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004922229192255903,
      "loss": 6.1195,
      "mean_token_accuracy": 0.16313235461711884,
      "num_tokens": 20759398.0,
      "step": 8190
    },
    {
      "entropy": 6.198694324493408,
      "epoch": 0.9457587997691864,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004922121602995678,
      "loss": 6.1404,
      "mean_token_accuracy": 0.1738876223564148,
      "num_tokens": 20773057.0,
      "step": 8195
    },
    {
      "entropy": 6.2284547805786135,
      "epoch": 0.9463358338141951,
      "grad_norm": 1.0,
      "learning_rate": 0.0004922013940677221,
      "loss": 6.1667,
      "mean_token_accuracy": 0.15642004907131196,
      "num_tokens": 20786084.0,
      "step": 8200
    },
    {
      "entropy": 6.21826024055481,
      "epoch": 0.9469128678592037,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004921906205304155,
      "loss": 6.0829,
      "mean_token_accuracy": 0.17130012363195418,
      "num_tokens": 20798459.0,
      "step": 8205
    },
    {
      "entropy": 5.988680744171143,
      "epoch": 0.9474899019042123,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004921798396880101,
      "loss": 5.915,
      "mean_token_accuracy": 0.17540648579597473,
      "num_tokens": 20811242.0,
      "step": 8210
    },
    {
      "entropy": 6.155648422241211,
      "epoch": 0.948066935949221,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004921690515408688,
      "loss": 6.1371,
      "mean_token_accuracy": 0.16307853013277054,
      "num_tokens": 20824757.0,
      "step": 8215
    },
    {
      "entropy": 6.329434871673584,
      "epoch": 0.9486439699942296,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004921582560893543,
      "loss": 6.1504,
      "mean_token_accuracy": 0.1616469219326973,
      "num_tokens": 20836794.0,
      "step": 8220
    },
    {
      "entropy": 6.220714092254639,
      "epoch": 0.9492210040392384,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004921474533338297,
      "loss": 6.1132,
      "mean_token_accuracy": 0.16348361372947692,
      "num_tokens": 20849240.0,
      "step": 8225
    },
    {
      "entropy": 6.1539782047271725,
      "epoch": 0.949798038084247,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004921366432746585,
      "loss": 6.1489,
      "mean_token_accuracy": 0.1658232182264328,
      "num_tokens": 20861362.0,
      "step": 8230
    },
    {
      "entropy": 6.236544179916382,
      "epoch": 0.9503750721292556,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000492125825912204,
      "loss": 6.1231,
      "mean_token_accuracy": 0.17040292620658876,
      "num_tokens": 20874925.0,
      "step": 8235
    },
    {
      "entropy": 6.174605655670166,
      "epoch": 0.9509521061742643,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004921150012468302,
      "loss": 6.1482,
      "mean_token_accuracy": 0.16474636346101762,
      "num_tokens": 20888098.0,
      "step": 8240
    },
    {
      "entropy": 6.141753625869751,
      "epoch": 0.9515291402192729,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004921041692789013,
      "loss": 6.1597,
      "mean_token_accuracy": 0.17203060388565064,
      "num_tokens": 20901125.0,
      "step": 8245
    },
    {
      "entropy": 6.308045053482056,
      "epoch": 0.9521061742642816,
      "grad_norm": 1.0,
      "learning_rate": 0.0004920933300087813,
      "loss": 6.1922,
      "mean_token_accuracy": 0.16417931914329528,
      "num_tokens": 20914815.0,
      "step": 8250
    },
    {
      "entropy": 6.096014881134034,
      "epoch": 0.9526832083092902,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004920824834368353,
      "loss": 6.0405,
      "mean_token_accuracy": 0.17727964073419572,
      "num_tokens": 20928153.0,
      "step": 8255
    },
    {
      "entropy": 6.338839864730835,
      "epoch": 0.953260242354299,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004920716295634275,
      "loss": 6.218,
      "mean_token_accuracy": 0.15484269857406616,
      "num_tokens": 20940601.0,
      "step": 8260
    },
    {
      "entropy": 6.267004776000976,
      "epoch": 0.9538372763993076,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004920607683889235,
      "loss": 6.2089,
      "mean_token_accuracy": 0.1607002094388008,
      "num_tokens": 20953158.0,
      "step": 8265
    },
    {
      "entropy": 6.24556884765625,
      "epoch": 0.9544143104443162,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004920498999136882,
      "loss": 6.1816,
      "mean_token_accuracy": 0.16401962339878082,
      "num_tokens": 20966572.0,
      "step": 8270
    },
    {
      "entropy": 6.3140825748443605,
      "epoch": 0.9549913444893249,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004920390241380874,
      "loss": 6.1546,
      "mean_token_accuracy": 0.16595411002635957,
      "num_tokens": 20978432.0,
      "step": 8275
    },
    {
      "entropy": 6.243206453323364,
      "epoch": 0.9555683785343335,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004920281410624868,
      "loss": 6.1968,
      "mean_token_accuracy": 0.16943021416664122,
      "num_tokens": 20991454.0,
      "step": 8280
    },
    {
      "entropy": 6.204116725921631,
      "epoch": 0.9561454125793422,
      "grad_norm": 1.0,
      "learning_rate": 0.0004920172506872525,
      "loss": 6.0875,
      "mean_token_accuracy": 0.16938930302858352,
      "num_tokens": 21004111.0,
      "step": 8285
    },
    {
      "entropy": 6.182796907424927,
      "epoch": 0.9567224466243508,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004920063530127508,
      "loss": 6.0797,
      "mean_token_accuracy": 0.15717935264110566,
      "num_tokens": 21017523.0,
      "step": 8290
    },
    {
      "entropy": 6.230190134048462,
      "epoch": 0.9572994806693594,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004919954480393482,
      "loss": 6.1403,
      "mean_token_accuracy": 0.16158571541309358,
      "num_tokens": 21028723.0,
      "step": 8295
    },
    {
      "entropy": 6.252023077011108,
      "epoch": 0.9578765147143682,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004919845357674114,
      "loss": 6.1465,
      "mean_token_accuracy": 0.16629870235919952,
      "num_tokens": 21042205.0,
      "step": 8300
    },
    {
      "entropy": 6.17732982635498,
      "epoch": 0.9584535487593768,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004919736161973076,
      "loss": 6.077,
      "mean_token_accuracy": 0.16344911605119705,
      "num_tokens": 21053798.0,
      "step": 8305
    },
    {
      "entropy": 6.202902746200562,
      "epoch": 0.9590305828043855,
      "grad_norm": 1.078125,
      "learning_rate": 0.000491962689329404,
      "loss": 6.2381,
      "mean_token_accuracy": 0.15746907889842987,
      "num_tokens": 21068333.0,
      "step": 8310
    },
    {
      "entropy": 6.204369926452637,
      "epoch": 0.9596076168493941,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004919517551640681,
      "loss": 6.0899,
      "mean_token_accuracy": 0.17300662845373155,
      "num_tokens": 21079484.0,
      "step": 8315
    },
    {
      "entropy": 6.149711179733276,
      "epoch": 0.9601846508944027,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004919408137016677,
      "loss": 6.1087,
      "mean_token_accuracy": 0.1708444282412529,
      "num_tokens": 21090940.0,
      "step": 8320
    },
    {
      "entropy": 6.310809135437012,
      "epoch": 0.9607616849394114,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004919298649425708,
      "loss": 6.1788,
      "mean_token_accuracy": 0.16207325607538223,
      "num_tokens": 21104351.0,
      "step": 8325
    },
    {
      "entropy": 6.241118431091309,
      "epoch": 0.96133871898442,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004919189088871456,
      "loss": 6.1309,
      "mean_token_accuracy": 0.16217202246189116,
      "num_tokens": 21116272.0,
      "step": 8330
    },
    {
      "entropy": 6.258427095413208,
      "epoch": 0.9619157530294288,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004919079455357608,
      "loss": 6.2099,
      "mean_token_accuracy": 0.15867260843515396,
      "num_tokens": 21129448.0,
      "step": 8335
    },
    {
      "entropy": 6.236162090301514,
      "epoch": 0.9624927870744374,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004918969748887847,
      "loss": 6.1869,
      "mean_token_accuracy": 0.15821556895971298,
      "num_tokens": 21143668.0,
      "step": 8340
    },
    {
      "entropy": 6.27112832069397,
      "epoch": 0.9630698211194461,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004918859969465868,
      "loss": 6.2007,
      "mean_token_accuracy": 0.16267903447151183,
      "num_tokens": 21156662.0,
      "step": 8345
    },
    {
      "entropy": 6.239499092102051,
      "epoch": 0.9636468551644547,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004918750117095361,
      "loss": 6.1254,
      "mean_token_accuracy": 0.16735378354787828,
      "num_tokens": 21170271.0,
      "step": 8350
    },
    {
      "entropy": 6.190868997573853,
      "epoch": 0.9642238892094633,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004918640191780021,
      "loss": 6.2036,
      "mean_token_accuracy": 0.15912721455097198,
      "num_tokens": 21183878.0,
      "step": 8355
    },
    {
      "entropy": 6.261455106735229,
      "epoch": 0.964800923254472,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004918530193523546,
      "loss": 6.0992,
      "mean_token_accuracy": 0.16800848841667176,
      "num_tokens": 21196088.0,
      "step": 8360
    },
    {
      "entropy": 6.228429746627808,
      "epoch": 0.9653779572994806,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004918420122329634,
      "loss": 6.2326,
      "mean_token_accuracy": 0.16445968747138978,
      "num_tokens": 21208640.0,
      "step": 8365
    },
    {
      "entropy": 6.294741201400757,
      "epoch": 0.9659549913444894,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004918309978201989,
      "loss": 6.0165,
      "mean_token_accuracy": 0.17763424962759017,
      "num_tokens": 21221246.0,
      "step": 8370
    },
    {
      "entropy": 6.154795694351196,
      "epoch": 0.966532025389498,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004918199761144315,
      "loss": 6.0911,
      "mean_token_accuracy": 0.16731215566396712,
      "num_tokens": 21233834.0,
      "step": 8375
    },
    {
      "entropy": 6.177775287628174,
      "epoch": 0.9671090594345066,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004918089471160318,
      "loss": 6.1213,
      "mean_token_accuracy": 0.1693127781152725,
      "num_tokens": 21247388.0,
      "step": 8380
    },
    {
      "entropy": 6.2543299198150635,
      "epoch": 0.9676860934795153,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004917979108253709,
      "loss": 6.1325,
      "mean_token_accuracy": 0.1634067639708519,
      "num_tokens": 21259800.0,
      "step": 8385
    },
    {
      "entropy": 6.22504940032959,
      "epoch": 0.9682631275245239,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00049178686724282,
      "loss": 6.1787,
      "mean_token_accuracy": 0.16314368098974227,
      "num_tokens": 21271776.0,
      "step": 8390
    },
    {
      "entropy": 6.186864757537842,
      "epoch": 0.9688401615695326,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004917758163687506,
      "loss": 6.0804,
      "mean_token_accuracy": 0.16572830826044083,
      "num_tokens": 21285368.0,
      "step": 8395
    },
    {
      "entropy": 6.107567024230957,
      "epoch": 0.9694171956145412,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004917647582035341,
      "loss": 6.0659,
      "mean_token_accuracy": 0.16884899735450745,
      "num_tokens": 21299077.0,
      "step": 8400
    },
    {
      "entropy": 6.221385669708252,
      "epoch": 0.96999422965955,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004917536927475428,
      "loss": 6.1536,
      "mean_token_accuracy": 0.16265100687742234,
      "num_tokens": 21312379.0,
      "step": 8405
    },
    {
      "entropy": 6.221223306655884,
      "epoch": 0.9705712637045586,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004917426200011486,
      "loss": 6.1111,
      "mean_token_accuracy": 0.17083193361759186,
      "num_tokens": 21325530.0,
      "step": 8410
    },
    {
      "entropy": 6.114801692962646,
      "epoch": 0.9711482977495672,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000491731539964724,
      "loss": 5.9902,
      "mean_token_accuracy": 0.175082466006279,
      "num_tokens": 21339200.0,
      "step": 8415
    },
    {
      "entropy": 6.242129802703857,
      "epoch": 0.9717253317945759,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004917204526386418,
      "loss": 6.0668,
      "mean_token_accuracy": 0.17476777285337447,
      "num_tokens": 21351072.0,
      "step": 8420
    },
    {
      "entropy": 6.1507315158844,
      "epoch": 0.9723023658395845,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004917093580232748,
      "loss": 6.1114,
      "mean_token_accuracy": 0.16849047988653182,
      "num_tokens": 21363691.0,
      "step": 8425
    },
    {
      "entropy": 6.24981722831726,
      "epoch": 0.9728793998845932,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004916982561189962,
      "loss": 6.1229,
      "mean_token_accuracy": 0.15974433571100236,
      "num_tokens": 21376896.0,
      "step": 8430
    },
    {
      "entropy": 6.235975933074951,
      "epoch": 0.9734564339296018,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004916871469261794,
      "loss": 6.1612,
      "mean_token_accuracy": 0.16906781792640685,
      "num_tokens": 21388590.0,
      "step": 8435
    },
    {
      "entropy": 6.259327507019043,
      "epoch": 0.9740334679746105,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000491676030445198,
      "loss": 6.1956,
      "mean_token_accuracy": 0.16068254709243773,
      "num_tokens": 21401912.0,
      "step": 8440
    },
    {
      "entropy": 6.269289112091064,
      "epoch": 0.9746105020196192,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004916649066764259,
      "loss": 6.1596,
      "mean_token_accuracy": 0.1662818118929863,
      "num_tokens": 21415472.0,
      "step": 8445
    },
    {
      "entropy": 6.223597717285156,
      "epoch": 0.9751875360646278,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004916537756202375,
      "loss": 6.1015,
      "mean_token_accuracy": 0.16913005262613295,
      "num_tokens": 21426758.0,
      "step": 8450
    },
    {
      "entropy": 6.184248971939087,
      "epoch": 0.9757645701096365,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004916426372770069,
      "loss": 6.1168,
      "mean_token_accuracy": 0.17063073068857193,
      "num_tokens": 21439347.0,
      "step": 8455
    },
    {
      "entropy": 6.276047611236573,
      "epoch": 0.9763416041546451,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004916314916471087,
      "loss": 6.2106,
      "mean_token_accuracy": 0.15706343948841095,
      "num_tokens": 21451359.0,
      "step": 8460
    },
    {
      "entropy": 6.300260019302368,
      "epoch": 0.9769186381996537,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004916203387309179,
      "loss": 6.1961,
      "mean_token_accuracy": 0.15965360552072524,
      "num_tokens": 21464926.0,
      "step": 8465
    },
    {
      "entropy": 6.196961307525635,
      "epoch": 0.9774956722446624,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004916091785288096,
      "loss": 6.1416,
      "mean_token_accuracy": 0.1629567489027977,
      "num_tokens": 21477234.0,
      "step": 8470
    },
    {
      "entropy": 6.124915599822998,
      "epoch": 0.9780727062896711,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004915980110411593,
      "loss": 5.9918,
      "mean_token_accuracy": 0.17941092997789382,
      "num_tokens": 21488751.0,
      "step": 8475
    },
    {
      "entropy": 6.230865001678467,
      "epoch": 0.9786497403346798,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004915868362683425,
      "loss": 6.1139,
      "mean_token_accuracy": 0.16764342188835143,
      "num_tokens": 21500532.0,
      "step": 8480
    },
    {
      "entropy": 6.164604663848877,
      "epoch": 0.9792267743796884,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004915756542107349,
      "loss": 6.1192,
      "mean_token_accuracy": 0.16191476583480835,
      "num_tokens": 21513725.0,
      "step": 8485
    },
    {
      "entropy": 6.061567544937134,
      "epoch": 0.979803808424697,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004915644648687127,
      "loss": 6.0896,
      "mean_token_accuracy": 0.17363775372505189,
      "num_tokens": 21526886.0,
      "step": 8490
    },
    {
      "entropy": 6.113343572616577,
      "epoch": 0.9803808424697057,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004915532682426524,
      "loss": 5.9938,
      "mean_token_accuracy": 0.1735152021050453,
      "num_tokens": 21539278.0,
      "step": 8495
    },
    {
      "entropy": 6.2418114185333256,
      "epoch": 0.9809578765147143,
      "grad_norm": 1.0,
      "learning_rate": 0.0004915420643329306,
      "loss": 6.2284,
      "mean_token_accuracy": 0.15796669721603393,
      "num_tokens": 21551864.0,
      "step": 8500
    },
    {
      "entropy": 6.267448663711548,
      "epoch": 0.981534910559723,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000491530853139924,
      "loss": 6.1565,
      "mean_token_accuracy": 0.16442210376262664,
      "num_tokens": 21564398.0,
      "step": 8505
    },
    {
      "entropy": 6.291948080062866,
      "epoch": 0.9821119446047317,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004915196346640095,
      "loss": 6.1551,
      "mean_token_accuracy": 0.1636106088757515,
      "num_tokens": 21575998.0,
      "step": 8510
    },
    {
      "entropy": 6.217896223068237,
      "epoch": 0.9826889786497404,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004915084089055648,
      "loss": 6.1438,
      "mean_token_accuracy": 0.16385289579629897,
      "num_tokens": 21589243.0,
      "step": 8515
    },
    {
      "entropy": 6.148995018005371,
      "epoch": 0.983266012694749,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004914971758649673,
      "loss": 6.079,
      "mean_token_accuracy": 0.17255953401327134,
      "num_tokens": 21601229.0,
      "step": 8520
    },
    {
      "entropy": 6.2313531875610355,
      "epoch": 0.9838430467397576,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004914859355425948,
      "loss": 6.1345,
      "mean_token_accuracy": 0.16398843228816987,
      "num_tokens": 21613270.0,
      "step": 8525
    },
    {
      "entropy": 6.151173210144043,
      "epoch": 0.9844200807847663,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004914746879388255,
      "loss": 6.0515,
      "mean_token_accuracy": 0.17331353574991226,
      "num_tokens": 21626463.0,
      "step": 8530
    },
    {
      "entropy": 6.21139497756958,
      "epoch": 0.9849971148297749,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004914634330540373,
      "loss": 6.1728,
      "mean_token_accuracy": 0.1673821434378624,
      "num_tokens": 21638172.0,
      "step": 8535
    },
    {
      "entropy": 6.228831768035889,
      "epoch": 0.9855741488747836,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004914521708886093,
      "loss": 6.0913,
      "mean_token_accuracy": 0.1711085930466652,
      "num_tokens": 21650015.0,
      "step": 8540
    },
    {
      "entropy": 6.262759256362915,
      "epoch": 0.9861511829197923,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004914409014429201,
      "loss": 6.1796,
      "mean_token_accuracy": 0.16166077554225922,
      "num_tokens": 21662502.0,
      "step": 8545
    },
    {
      "entropy": 6.188414144515991,
      "epoch": 0.986728216964801,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004914296247173486,
      "loss": 6.0821,
      "mean_token_accuracy": 0.17665434628725052,
      "num_tokens": 21675691.0,
      "step": 8550
    },
    {
      "entropy": 6.235756969451904,
      "epoch": 0.9873052510098096,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004914183407122741,
      "loss": 6.1496,
      "mean_token_accuracy": 0.15848990827798842,
      "num_tokens": 21687904.0,
      "step": 8555
    },
    {
      "entropy": 6.280721378326416,
      "epoch": 0.9878822850548182,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004914070494280763,
      "loss": 6.213,
      "mean_token_accuracy": 0.16239575743675233,
      "num_tokens": 21700371.0,
      "step": 8560
    },
    {
      "entropy": 6.1972327709198,
      "epoch": 0.9884593190998269,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004913957508651349,
      "loss": 6.107,
      "mean_token_accuracy": 0.17504676431417465,
      "num_tokens": 21715235.0,
      "step": 8565
    },
    {
      "entropy": 6.225125932693482,
      "epoch": 0.9890363531448355,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004913844450238299,
      "loss": 6.091,
      "mean_token_accuracy": 0.16794272810220717,
      "num_tokens": 21727994.0,
      "step": 8570
    },
    {
      "entropy": 6.174360609054565,
      "epoch": 0.9896133871898442,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004913731319045416,
      "loss": 6.0778,
      "mean_token_accuracy": 0.16596533209085465,
      "num_tokens": 21739894.0,
      "step": 8575
    },
    {
      "entropy": 6.238060760498047,
      "epoch": 0.9901904212348529,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004913618115076505,
      "loss": 6.2158,
      "mean_token_accuracy": 0.16331194043159486,
      "num_tokens": 21752957.0,
      "step": 8580
    },
    {
      "entropy": 6.218990612030029,
      "epoch": 0.9907674552798615,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004913504838335372,
      "loss": 6.0978,
      "mean_token_accuracy": 0.16436419039964675,
      "num_tokens": 21765547.0,
      "step": 8585
    },
    {
      "entropy": 6.233816385269165,
      "epoch": 0.9913444893248702,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004913391488825829,
      "loss": 6.1835,
      "mean_token_accuracy": 0.170366770029068,
      "num_tokens": 21778817.0,
      "step": 8590
    },
    {
      "entropy": 6.171450757980347,
      "epoch": 0.9919215233698788,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004913278066551689,
      "loss": 6.0753,
      "mean_token_accuracy": 0.16817353516817093,
      "num_tokens": 21791002.0,
      "step": 8595
    },
    {
      "entropy": 6.227829933166504,
      "epoch": 0.9924985574148875,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004913164571516765,
      "loss": 6.1637,
      "mean_token_accuracy": 0.16633816808462143,
      "num_tokens": 21803877.0,
      "step": 8600
    },
    {
      "entropy": 6.238053655624389,
      "epoch": 0.9930755914598961,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004913051003724876,
      "loss": 6.093,
      "mean_token_accuracy": 0.16335604339838028,
      "num_tokens": 21815884.0,
      "step": 8605
    },
    {
      "entropy": 6.196093225479126,
      "epoch": 0.9936526255049047,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004912937363179839,
      "loss": 6.1186,
      "mean_token_accuracy": 0.16944066882133485,
      "num_tokens": 21826990.0,
      "step": 8610
    },
    {
      "entropy": 6.152547025680542,
      "epoch": 0.9942296595499135,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000491282364988548,
      "loss": 6.1178,
      "mean_token_accuracy": 0.16797010153532027,
      "num_tokens": 21838835.0,
      "step": 8615
    },
    {
      "entropy": 6.294097375869751,
      "epoch": 0.9948066935949221,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004912709863845621,
      "loss": 6.2017,
      "mean_token_accuracy": 0.16497852355241777,
      "num_tokens": 21850924.0,
      "step": 8620
    },
    {
      "entropy": 6.237844610214234,
      "epoch": 0.9953837276399308,
      "grad_norm": 1.03125,
      "learning_rate": 0.000491259600506409,
      "loss": 6.1711,
      "mean_token_accuracy": 0.16704185158014298,
      "num_tokens": 21862720.0,
      "step": 8625
    },
    {
      "entropy": 6.2263233184814455,
      "epoch": 0.9959607616849394,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004912482073544716,
      "loss": 6.1559,
      "mean_token_accuracy": 0.16471816897392272,
      "num_tokens": 21874655.0,
      "step": 8630
    },
    {
      "entropy": 6.128899717330933,
      "epoch": 0.996537795729948,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004912368069291333,
      "loss": 6.0891,
      "mean_token_accuracy": 0.16942135542631148,
      "num_tokens": 21886911.0,
      "step": 8635
    },
    {
      "entropy": 6.12626428604126,
      "epoch": 0.9971148297749567,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004912253992307773,
      "loss": 6.064,
      "mean_token_accuracy": 0.1702150359749794,
      "num_tokens": 21899614.0,
      "step": 8640
    },
    {
      "entropy": 6.261716032028199,
      "epoch": 0.9976918638199653,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004912139842597875,
      "loss": 6.0962,
      "mean_token_accuracy": 0.16818332225084304,
      "num_tokens": 21912756.0,
      "step": 8645
    },
    {
      "entropy": 6.150550460815429,
      "epoch": 0.9982688978649741,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004912025620165477,
      "loss": 6.1167,
      "mean_token_accuracy": 0.1626631960272789,
      "num_tokens": 21924911.0,
      "step": 8650
    },
    {
      "entropy": 6.154950523376465,
      "epoch": 0.9988459319099827,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004911911325014421,
      "loss": 6.0569,
      "mean_token_accuracy": 0.16787694692611693,
      "num_tokens": 21938204.0,
      "step": 8655
    },
    {
      "entropy": 6.24123969078064,
      "epoch": 0.9994229659549914,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004911796957148552,
      "loss": 6.1276,
      "mean_token_accuracy": 0.1662372589111328,
      "num_tokens": 21951285.0,
      "step": 8660
    },
    {
      "entropy": 6.226286458969116,
      "epoch": 1.0,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004911682516571715,
      "loss": 6.163,
      "mean_token_accuracy": 0.16109730154275895,
      "num_tokens": 21963117.0,
      "step": 8665
    },
    {
      "entropy": 6.227232599258423,
      "epoch": 1.0005770340450086,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004911568003287761,
      "loss": 6.0354,
      "mean_token_accuracy": 0.16565361469984055,
      "num_tokens": 21976103.0,
      "step": 8670
    },
    {
      "entropy": 6.096490049362183,
      "epoch": 1.0011540680900173,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004911453417300541,
      "loss": 5.9189,
      "mean_token_accuracy": 0.18426355868577957,
      "num_tokens": 21987905.0,
      "step": 8675
    },
    {
      "entropy": 6.107545471191406,
      "epoch": 1.001731102135026,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004911338758613909,
      "loss": 5.9998,
      "mean_token_accuracy": 0.17205023020505905,
      "num_tokens": 22000872.0,
      "step": 8680
    },
    {
      "entropy": 6.173818063735962,
      "epoch": 1.0023081361800346,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004911224027231722,
      "loss": 6.0231,
      "mean_token_accuracy": 0.1655685558915138,
      "num_tokens": 22012725.0,
      "step": 8685
    },
    {
      "entropy": 6.166259765625,
      "epoch": 1.0028851702250432,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004911109223157838,
      "loss": 5.9681,
      "mean_token_accuracy": 0.17618713527917862,
      "num_tokens": 22024802.0,
      "step": 8690
    },
    {
      "entropy": 6.172052955627441,
      "epoch": 1.0034622042700518,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004910994346396118,
      "loss": 6.0584,
      "mean_token_accuracy": 0.1658712923526764,
      "num_tokens": 22037481.0,
      "step": 8695
    },
    {
      "entropy": 6.2186219692230225,
      "epoch": 1.0040392383150605,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004910879396950427,
      "loss": 6.0635,
      "mean_token_accuracy": 0.17320849150419235,
      "num_tokens": 22049387.0,
      "step": 8700
    },
    {
      "entropy": 6.259855270385742,
      "epoch": 1.0046162723600693,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000491076437482463,
      "loss": 6.0532,
      "mean_token_accuracy": 0.1628953218460083,
      "num_tokens": 22061323.0,
      "step": 8705
    },
    {
      "entropy": 6.160572195053101,
      "epoch": 1.005193306405078,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004910649280022599,
      "loss": 5.9899,
      "mean_token_accuracy": 0.17184107899665832,
      "num_tokens": 22073163.0,
      "step": 8710
    },
    {
      "entropy": 6.094312953948974,
      "epoch": 1.0057703404500866,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004910534112548201,
      "loss": 5.9694,
      "mean_token_accuracy": 0.17073361724615096,
      "num_tokens": 22085583.0,
      "step": 8715
    },
    {
      "entropy": 6.252714490890503,
      "epoch": 1.0063473744950953,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004910418872405312,
      "loss": 6.0702,
      "mean_token_accuracy": 0.16884851455688477,
      "num_tokens": 22098241.0,
      "step": 8720
    },
    {
      "entropy": 6.215780639648438,
      "epoch": 1.006924408540104,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004910303559597806,
      "loss": 5.9723,
      "mean_token_accuracy": 0.1671421378850937,
      "num_tokens": 22111626.0,
      "step": 8725
    },
    {
      "entropy": 6.110291814804077,
      "epoch": 1.0075014425851125,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004910188174129564,
      "loss": 5.9993,
      "mean_token_accuracy": 0.16508312821388244,
      "num_tokens": 22124653.0,
      "step": 8730
    },
    {
      "entropy": 6.221698331832886,
      "epoch": 1.0080784766301212,
      "grad_norm": 1.0,
      "learning_rate": 0.0004910072716004466,
      "loss": 5.9945,
      "mean_token_accuracy": 0.17572650015354158,
      "num_tokens": 22137154.0,
      "step": 8735
    },
    {
      "entropy": 6.214560508728027,
      "epoch": 1.0086555106751298,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004909957185226394,
      "loss": 6.1204,
      "mean_token_accuracy": 0.16073711812496186,
      "num_tokens": 22149374.0,
      "step": 8740
    },
    {
      "entropy": 6.1474018573760985,
      "epoch": 1.0092325447201385,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004909841581799236,
      "loss": 6.0745,
      "mean_token_accuracy": 0.17107634842395783,
      "num_tokens": 22161988.0,
      "step": 8745
    },
    {
      "entropy": 6.269350671768189,
      "epoch": 1.009809578765147,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004909725905726879,
      "loss": 6.1285,
      "mean_token_accuracy": 0.16379568725824356,
      "num_tokens": 22173490.0,
      "step": 8750
    },
    {
      "entropy": 6.1709850311279295,
      "epoch": 1.0103866128101557,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004909610157013214,
      "loss": 6.0641,
      "mean_token_accuracy": 0.16984837204217912,
      "num_tokens": 22185065.0,
      "step": 8755
    },
    {
      "entropy": 6.1411418437957765,
      "epoch": 1.0109636468551644,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004909494335662134,
      "loss": 6.0223,
      "mean_token_accuracy": 0.17490147948265075,
      "num_tokens": 22198325.0,
      "step": 8760
    },
    {
      "entropy": 6.068651723861694,
      "epoch": 1.011540680900173,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004909378441677535,
      "loss": 5.9463,
      "mean_token_accuracy": 0.17895367592573166,
      "num_tokens": 22211827.0,
      "step": 8765
    },
    {
      "entropy": 6.166070508956909,
      "epoch": 1.0121177149451817,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004909262475063314,
      "loss": 6.0042,
      "mean_token_accuracy": 0.1720341071486473,
      "num_tokens": 22223411.0,
      "step": 8770
    },
    {
      "entropy": 6.131281471252441,
      "epoch": 1.0126947489901905,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004909146435823373,
      "loss": 5.9973,
      "mean_token_accuracy": 0.16937244981527327,
      "num_tokens": 22237752.0,
      "step": 8775
    },
    {
      "entropy": 6.228354215621948,
      "epoch": 1.0132717830351992,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004909030323961613,
      "loss": 6.0726,
      "mean_token_accuracy": 0.1689805120229721,
      "num_tokens": 22249778.0,
      "step": 8780
    },
    {
      "entropy": 6.18192229270935,
      "epoch": 1.0138488170802078,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004908914139481942,
      "loss": 6.0441,
      "mean_token_accuracy": 0.16860453486442567,
      "num_tokens": 22262514.0,
      "step": 8785
    },
    {
      "entropy": 6.162627363204956,
      "epoch": 1.0144258511252164,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004908797882388265,
      "loss": 5.9892,
      "mean_token_accuracy": 0.1707977071404457,
      "num_tokens": 22273662.0,
      "step": 8790
    },
    {
      "entropy": 6.152282333374023,
      "epoch": 1.015002885170225,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004908681552684495,
      "loss": 6.1058,
      "mean_token_accuracy": 0.15829371213912963,
      "num_tokens": 22286213.0,
      "step": 8795
    },
    {
      "entropy": 6.109225416183472,
      "epoch": 1.0155799192152337,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004908565150374542,
      "loss": 5.8835,
      "mean_token_accuracy": 0.1747447818517685,
      "num_tokens": 22299191.0,
      "step": 8800
    },
    {
      "entropy": 6.0982176780700685,
      "epoch": 1.0161569532602424,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004908448675462323,
      "loss": 5.9792,
      "mean_token_accuracy": 0.17498499304056167,
      "num_tokens": 22312047.0,
      "step": 8805
    },
    {
      "entropy": 6.16484694480896,
      "epoch": 1.016733987305251,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004908332127951756,
      "loss": 6.0988,
      "mean_token_accuracy": 0.1634947806596756,
      "num_tokens": 22326069.0,
      "step": 8810
    },
    {
      "entropy": 6.250838279724121,
      "epoch": 1.0173110213502596,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004908215507846757,
      "loss": 6.0194,
      "mean_token_accuracy": 0.1671513795852661,
      "num_tokens": 22338593.0,
      "step": 8815
    },
    {
      "entropy": 6.152573919296264,
      "epoch": 1.0178880553952683,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004908098815151254,
      "loss": 5.9992,
      "mean_token_accuracy": 0.1735528826713562,
      "num_tokens": 22352286.0,
      "step": 8820
    },
    {
      "entropy": 6.190534877777099,
      "epoch": 1.018465089440277,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004907982049869168,
      "loss": 5.9947,
      "mean_token_accuracy": 0.16935011893510818,
      "num_tokens": 22364270.0,
      "step": 8825
    },
    {
      "entropy": 6.1137467384338375,
      "epoch": 1.0190421234852856,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004907865212004428,
      "loss": 6.0353,
      "mean_token_accuracy": 0.16441887617111206,
      "num_tokens": 22377799.0,
      "step": 8830
    },
    {
      "entropy": 6.2154974937438965,
      "epoch": 1.0196191575302942,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004907748301560963,
      "loss": 6.0375,
      "mean_token_accuracy": 0.17162969708442688,
      "num_tokens": 22389913.0,
      "step": 8835
    },
    {
      "entropy": 6.1877717018127445,
      "epoch": 1.0201961915753028,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004907631318542707,
      "loss": 6.0642,
      "mean_token_accuracy": 0.16658550649881362,
      "num_tokens": 22401545.0,
      "step": 8840
    },
    {
      "entropy": 6.083741950988769,
      "epoch": 1.0207732256203117,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004907514262953594,
      "loss": 5.9591,
      "mean_token_accuracy": 0.16832875609397888,
      "num_tokens": 22414084.0,
      "step": 8845
    },
    {
      "entropy": 6.101075029373169,
      "epoch": 1.0213502596653203,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004907397134797559,
      "loss": 5.9634,
      "mean_token_accuracy": 0.17093513607978822,
      "num_tokens": 22427406.0,
      "step": 8850
    },
    {
      "entropy": 6.240686225891113,
      "epoch": 1.021927293710329,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004907279934078543,
      "loss": 6.0635,
      "mean_token_accuracy": 0.1625946819782257,
      "num_tokens": 22441118.0,
      "step": 8855
    },
    {
      "entropy": 6.2215940952301025,
      "epoch": 1.0225043277553376,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004907162660800488,
      "loss": 6.1258,
      "mean_token_accuracy": 0.164428973197937,
      "num_tokens": 22453799.0,
      "step": 8860
    },
    {
      "entropy": 6.193825721740723,
      "epoch": 1.0230813618003463,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004907045314967338,
      "loss": 6.0773,
      "mean_token_accuracy": 0.16602628976106643,
      "num_tokens": 22466573.0,
      "step": 8865
    },
    {
      "entropy": 6.117796802520752,
      "epoch": 1.023658395845355,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004906927896583041,
      "loss": 5.9472,
      "mean_token_accuracy": 0.17610322386026384,
      "num_tokens": 22478968.0,
      "step": 8870
    },
    {
      "entropy": 6.134650421142578,
      "epoch": 1.0242354298903635,
      "grad_norm": 0.875,
      "learning_rate": 0.0004906810405651546,
      "loss": 5.9533,
      "mean_token_accuracy": 0.18308767080307006,
      "num_tokens": 22490669.0,
      "step": 8875
    },
    {
      "entropy": 6.179800796508789,
      "epoch": 1.0248124639353722,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004906692842176803,
      "loss": 6.0403,
      "mean_token_accuracy": 0.1677374169230461,
      "num_tokens": 22502235.0,
      "step": 8880
    },
    {
      "entropy": 6.194182205200195,
      "epoch": 1.0253894979803808,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004906575206162769,
      "loss": 6.0293,
      "mean_token_accuracy": 0.1661167785525322,
      "num_tokens": 22514535.0,
      "step": 8885
    },
    {
      "entropy": 6.119495820999146,
      "epoch": 1.0259665320253895,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004906457497613399,
      "loss": 6.0196,
      "mean_token_accuracy": 0.16883869618177413,
      "num_tokens": 22527773.0,
      "step": 8890
    },
    {
      "entropy": 6.133694362640381,
      "epoch": 1.026543566070398,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004906339716532651,
      "loss": 6.0208,
      "mean_token_accuracy": 0.1731777474284172,
      "num_tokens": 22541853.0,
      "step": 8895
    },
    {
      "entropy": 6.244993209838867,
      "epoch": 1.0271206001154067,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004906221862924488,
      "loss": 6.0774,
      "mean_token_accuracy": 0.1621861830353737,
      "num_tokens": 22554473.0,
      "step": 8900
    },
    {
      "entropy": 6.2071043968200685,
      "epoch": 1.0276976341604154,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004906103936792875,
      "loss": 6.0658,
      "mean_token_accuracy": 0.1669753909111023,
      "num_tokens": 22567107.0,
      "step": 8905
    },
    {
      "entropy": 6.203036117553711,
      "epoch": 1.028274668205424,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004905985938141777,
      "loss": 6.2166,
      "mean_token_accuracy": 0.15614844709634781,
      "num_tokens": 22579806.0,
      "step": 8910
    },
    {
      "entropy": 6.211770343780517,
      "epoch": 1.0288517022504329,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004905867866975163,
      "loss": 6.0207,
      "mean_token_accuracy": 0.16908140629529952,
      "num_tokens": 22593460.0,
      "step": 8915
    },
    {
      "entropy": 6.17006516456604,
      "epoch": 1.0294287362954415,
      "grad_norm": 1.0,
      "learning_rate": 0.0004905749723297003,
      "loss": 6.0625,
      "mean_token_accuracy": 0.1722796306014061,
      "num_tokens": 22607113.0,
      "step": 8920
    },
    {
      "entropy": 6.181873321533203,
      "epoch": 1.0300057703404502,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004905631507111271,
      "loss": 5.9728,
      "mean_token_accuracy": 0.1723117783665657,
      "num_tokens": 22619749.0,
      "step": 8925
    },
    {
      "entropy": 6.082556247711182,
      "epoch": 1.0305828043854588,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004905513218421946,
      "loss": 5.9512,
      "mean_token_accuracy": 0.1765432119369507,
      "num_tokens": 22632584.0,
      "step": 8930
    },
    {
      "entropy": 6.143266582489014,
      "epoch": 1.0311598384304674,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004905394857233004,
      "loss": 6.0434,
      "mean_token_accuracy": 0.1704893469810486,
      "num_tokens": 22645181.0,
      "step": 8935
    },
    {
      "entropy": 6.232394742965698,
      "epoch": 1.031736872475476,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004905276423548426,
      "loss": 6.0655,
      "mean_token_accuracy": 0.16858468651771547,
      "num_tokens": 22657666.0,
      "step": 8940
    },
    {
      "entropy": 6.227100372314453,
      "epoch": 1.0323139065204847,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004905157917372197,
      "loss": 6.1169,
      "mean_token_accuracy": 0.16404200047254563,
      "num_tokens": 22670404.0,
      "step": 8945
    },
    {
      "entropy": 6.129315137863159,
      "epoch": 1.0328909405654934,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004905039338708302,
      "loss": 6.1093,
      "mean_token_accuracy": 0.1599951133131981,
      "num_tokens": 22683558.0,
      "step": 8950
    },
    {
      "entropy": 6.216488027572632,
      "epoch": 1.033467974610502,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004904920687560728,
      "loss": 5.9685,
      "mean_token_accuracy": 0.17302214950323105,
      "num_tokens": 22694907.0,
      "step": 8955
    },
    {
      "entropy": 6.164621257781983,
      "epoch": 1.0340450086555106,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004904801963933469,
      "loss": 6.1454,
      "mean_token_accuracy": 0.16209454834461212,
      "num_tokens": 22707447.0,
      "step": 8960
    },
    {
      "entropy": 6.166344690322876,
      "epoch": 1.0346220427005193,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004904683167830515,
      "loss": 6.0127,
      "mean_token_accuracy": 0.17258192896842955,
      "num_tokens": 22719150.0,
      "step": 8965
    },
    {
      "entropy": 6.235004234313965,
      "epoch": 1.035199076745528,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004904564299255862,
      "loss": 6.0975,
      "mean_token_accuracy": 0.1636883094906807,
      "num_tokens": 22732489.0,
      "step": 8970
    },
    {
      "entropy": 6.06142225265503,
      "epoch": 1.0357761107905366,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004904445358213511,
      "loss": 5.9538,
      "mean_token_accuracy": 0.16832381188869477,
      "num_tokens": 22745543.0,
      "step": 8975
    },
    {
      "entropy": 6.195770740509033,
      "epoch": 1.0363531448355452,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004904326344707461,
      "loss": 6.0725,
      "mean_token_accuracy": 0.16245600134134291,
      "num_tokens": 22757643.0,
      "step": 8980
    },
    {
      "entropy": 6.214196348190308,
      "epoch": 1.036930178880554,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004904207258741712,
      "loss": 6.0278,
      "mean_token_accuracy": 0.1692509427666664,
      "num_tokens": 22770707.0,
      "step": 8985
    },
    {
      "entropy": 6.1111791133880615,
      "epoch": 1.0375072129255627,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004904088100320274,
      "loss": 6.0681,
      "mean_token_accuracy": 0.17080598771572114,
      "num_tokens": 22782630.0,
      "step": 8990
    },
    {
      "entropy": 6.1230621337890625,
      "epoch": 1.0380842469705713,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004903968869447151,
      "loss": 6.1124,
      "mean_token_accuracy": 0.16730864197015763,
      "num_tokens": 22795693.0,
      "step": 8995
    },
    {
      "entropy": 6.263205575942993,
      "epoch": 1.03866128101558,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004903849566126356,
      "loss": 6.136,
      "mean_token_accuracy": 0.16162792444229127,
      "num_tokens": 22807715.0,
      "step": 9000
    },
    {
      "epoch": 1.03866128101558,
      "eval_entropy": 6.081961028983317,
      "eval_loss": 6.183651447296143,
      "eval_mean_token_accuracy": 0.16796843689872454,
      "eval_num_tokens": 22807715.0,
      "eval_runtime": 22.7143,
      "eval_samples_per_second": 1238.689,
      "eval_steps_per_second": 154.836,
      "step": 9000
    },
    {
      "entropy": 6.165067481994629,
      "epoch": 1.0392383150605886,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004903730190361902,
      "loss": 6.0137,
      "mean_token_accuracy": 0.1704970955848694,
      "num_tokens": 22820133.0,
      "step": 9005
    },
    {
      "entropy": 6.098128080368042,
      "epoch": 1.0398153491055973,
      "grad_norm": 1.234375,
      "learning_rate": 0.00049036107421578,
      "loss": 6.0534,
      "mean_token_accuracy": 0.1685325101017952,
      "num_tokens": 22833928.0,
      "step": 9010
    },
    {
      "entropy": 6.146722030639649,
      "epoch": 1.040392383150606,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004903491221518073,
      "loss": 5.9598,
      "mean_token_accuracy": 0.17696574926376343,
      "num_tokens": 22847883.0,
      "step": 9015
    },
    {
      "entropy": 6.168456602096557,
      "epoch": 1.0409694171956145,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004903371628446737,
      "loss": 6.0677,
      "mean_token_accuracy": 0.16453583389520646,
      "num_tokens": 22861540.0,
      "step": 9020
    },
    {
      "entropy": 6.203205919265747,
      "epoch": 1.0415464512406232,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004903251962947817,
      "loss": 6.0229,
      "mean_token_accuracy": 0.16489732414484023,
      "num_tokens": 22874778.0,
      "step": 9025
    },
    {
      "entropy": 6.164117002487183,
      "epoch": 1.0421234852856318,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004903132225025335,
      "loss": 6.0805,
      "mean_token_accuracy": 0.1721145763993263,
      "num_tokens": 22886868.0,
      "step": 9030
    },
    {
      "entropy": 6.150497579574585,
      "epoch": 1.0427005193306405,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004903012414683322,
      "loss": 6.0314,
      "mean_token_accuracy": 0.16408417224884034,
      "num_tokens": 22899549.0,
      "step": 9035
    },
    {
      "entropy": 6.203818035125733,
      "epoch": 1.043277553375649,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004902892531925805,
      "loss": 6.0081,
      "mean_token_accuracy": 0.17528444677591323,
      "num_tokens": 22912301.0,
      "step": 9040
    },
    {
      "entropy": 6.104867219924927,
      "epoch": 1.0438545874206577,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004902772576756818,
      "loss": 6.0441,
      "mean_token_accuracy": 0.17140656858682632,
      "num_tokens": 22924784.0,
      "step": 9045
    },
    {
      "entropy": 6.226058387756348,
      "epoch": 1.0444316214656664,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004902652549180394,
      "loss": 6.0521,
      "mean_token_accuracy": 0.16776583343744278,
      "num_tokens": 22937095.0,
      "step": 9050
    },
    {
      "entropy": 6.163409423828125,
      "epoch": 1.0450086555106752,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004902532449200571,
      "loss": 5.9652,
      "mean_token_accuracy": 0.17916348129510878,
      "num_tokens": 22951385.0,
      "step": 9055
    },
    {
      "entropy": 6.188064670562744,
      "epoch": 1.0455856895556839,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004902412276821386,
      "loss": 6.0536,
      "mean_token_accuracy": 0.16483981013298035,
      "num_tokens": 22965228.0,
      "step": 9060
    },
    {
      "entropy": 6.143165445327758,
      "epoch": 1.0461627236006925,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004902292032046887,
      "loss": 6.0213,
      "mean_token_accuracy": 0.17677357345819472,
      "num_tokens": 22978376.0,
      "step": 9065
    },
    {
      "entropy": 6.089274597167969,
      "epoch": 1.0467397576457012,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004902171714881112,
      "loss": 6.0482,
      "mean_token_accuracy": 0.17167790681123735,
      "num_tokens": 22991572.0,
      "step": 9070
    },
    {
      "entropy": 6.194202041625976,
      "epoch": 1.0473167916907098,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004902051325328112,
      "loss": 6.0042,
      "mean_token_accuracy": 0.1699507489800453,
      "num_tokens": 23004001.0,
      "step": 9075
    },
    {
      "entropy": 6.082419109344483,
      "epoch": 1.0478938257357184,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004901930863391934,
      "loss": 5.978,
      "mean_token_accuracy": 0.1732077866792679,
      "num_tokens": 23016383.0,
      "step": 9080
    },
    {
      "entropy": 6.2099583625793455,
      "epoch": 1.048470859780727,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000490181032907663,
      "loss": 6.0696,
      "mean_token_accuracy": 0.16540407687425612,
      "num_tokens": 23027836.0,
      "step": 9085
    },
    {
      "entropy": 6.184118986129761,
      "epoch": 1.0490478938257357,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004901689722386255,
      "loss": 6.0031,
      "mean_token_accuracy": 0.16609950512647628,
      "num_tokens": 23040083.0,
      "step": 9090
    },
    {
      "entropy": 6.101901340484619,
      "epoch": 1.0496249278707444,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004901569043324864,
      "loss": 6.0233,
      "mean_token_accuracy": 0.1696107119321823,
      "num_tokens": 23051232.0,
      "step": 9095
    },
    {
      "entropy": 6.232674598693848,
      "epoch": 1.050201961915753,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004901448291896518,
      "loss": 6.1155,
      "mean_token_accuracy": 0.16371671557426454,
      "num_tokens": 23063355.0,
      "step": 9100
    },
    {
      "entropy": 6.020437860488892,
      "epoch": 1.0507789959607616,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004901327468105277,
      "loss": 5.8043,
      "mean_token_accuracy": 0.18257721066474913,
      "num_tokens": 23076268.0,
      "step": 9105
    },
    {
      "entropy": 6.082842826843262,
      "epoch": 1.0513560300057703,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004901206571955205,
      "loss": 5.9629,
      "mean_token_accuracy": 0.17205547243356706,
      "num_tokens": 23088753.0,
      "step": 9110
    },
    {
      "entropy": 6.1010113716125485,
      "epoch": 1.051933064050779,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004901085603450369,
      "loss": 5.9103,
      "mean_token_accuracy": 0.1768498182296753,
      "num_tokens": 23100452.0,
      "step": 9115
    },
    {
      "entropy": 6.162375640869141,
      "epoch": 1.0525100980957875,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004900964562594838,
      "loss": 6.0208,
      "mean_token_accuracy": 0.1692277193069458,
      "num_tokens": 23113368.0,
      "step": 9120
    },
    {
      "entropy": 6.081042814254761,
      "epoch": 1.0530871321407964,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000490084344939268,
      "loss": 5.9429,
      "mean_token_accuracy": 0.17976068705320358,
      "num_tokens": 23125940.0,
      "step": 9125
    },
    {
      "entropy": 6.0908098220825195,
      "epoch": 1.053664166185805,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004900722263847973,
      "loss": 5.9983,
      "mean_token_accuracy": 0.17391017228364944,
      "num_tokens": 23137282.0,
      "step": 9130
    },
    {
      "entropy": 6.210883378982544,
      "epoch": 1.0542412002308137,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004900601005964791,
      "loss": 6.0852,
      "mean_token_accuracy": 0.17008815556764603,
      "num_tokens": 23149569.0,
      "step": 9135
    },
    {
      "entropy": 6.157468986511231,
      "epoch": 1.0548182342758223,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004900479675747212,
      "loss": 6.0719,
      "mean_token_accuracy": 0.170349058508873,
      "num_tokens": 23163290.0,
      "step": 9140
    },
    {
      "entropy": 6.1695657730102536,
      "epoch": 1.055395268320831,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004900358273199316,
      "loss": 6.0204,
      "mean_token_accuracy": 0.18001676648855208,
      "num_tokens": 23175092.0,
      "step": 9145
    },
    {
      "entropy": 6.187047243118286,
      "epoch": 1.0559723023658396,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000490023679832519,
      "loss": 5.9901,
      "mean_token_accuracy": 0.17541519850492476,
      "num_tokens": 23189064.0,
      "step": 9150
    },
    {
      "entropy": 6.154259967803955,
      "epoch": 1.0565493364108482,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004900115251128916,
      "loss": 6.0647,
      "mean_token_accuracy": 0.16230656951665878,
      "num_tokens": 23202619.0,
      "step": 9155
    },
    {
      "entropy": 6.122792530059814,
      "epoch": 1.057126370455857,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004899993631614583,
      "loss": 5.9431,
      "mean_token_accuracy": 0.17329391241073608,
      "num_tokens": 23214575.0,
      "step": 9160
    },
    {
      "entropy": 6.227842044830322,
      "epoch": 1.0577034045008655,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004899871939786283,
      "loss": 6.0881,
      "mean_token_accuracy": 0.15922293663024903,
      "num_tokens": 23226526.0,
      "step": 9165
    },
    {
      "entropy": 6.120679903030395,
      "epoch": 1.0582804385458742,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004899750175648107,
      "loss": 6.0459,
      "mean_token_accuracy": 0.1716884359717369,
      "num_tokens": 23239942.0,
      "step": 9170
    },
    {
      "entropy": 6.118875169754029,
      "epoch": 1.0588574725908828,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004899628339204154,
      "loss": 5.9934,
      "mean_token_accuracy": 0.17025153189897538,
      "num_tokens": 23252798.0,
      "step": 9175
    },
    {
      "entropy": 6.115894222259522,
      "epoch": 1.0594345066358914,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004899506430458518,
      "loss": 5.9569,
      "mean_token_accuracy": 0.17560370564460753,
      "num_tokens": 23266195.0,
      "step": 9180
    },
    {
      "entropy": 6.068709945678711,
      "epoch": 1.0600115406809,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004899384449415302,
      "loss": 5.9251,
      "mean_token_accuracy": 0.18054825067520142,
      "num_tokens": 23279057.0,
      "step": 9185
    },
    {
      "entropy": 6.224994373321533,
      "epoch": 1.0605885747259087,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004899262396078606,
      "loss": 6.063,
      "mean_token_accuracy": 0.16715140789747238,
      "num_tokens": 23292011.0,
      "step": 9190
    },
    {
      "entropy": 6.095200157165527,
      "epoch": 1.0611656087709176,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004899140270452539,
      "loss": 5.962,
      "mean_token_accuracy": 0.18029213100671768,
      "num_tokens": 23303719.0,
      "step": 9195
    },
    {
      "entropy": 6.12547197341919,
      "epoch": 1.0617426428159262,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004899018072541204,
      "loss": 6.0781,
      "mean_token_accuracy": 0.16641788482666015,
      "num_tokens": 23317291.0,
      "step": 9200
    },
    {
      "entropy": 6.187509059906006,
      "epoch": 1.0623196768609349,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004898895802348715,
      "loss": 6.0275,
      "mean_token_accuracy": 0.17190539687871934,
      "num_tokens": 23329546.0,
      "step": 9205
    },
    {
      "entropy": 6.090209245681763,
      "epoch": 1.0628967109059435,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004898773459879183,
      "loss": 5.9601,
      "mean_token_accuracy": 0.17489005625247955,
      "num_tokens": 23342171.0,
      "step": 9210
    },
    {
      "entropy": 6.1125093460083,
      "epoch": 1.0634737449509521,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004898651045136724,
      "loss": 6.0187,
      "mean_token_accuracy": 0.16217389702796936,
      "num_tokens": 23355400.0,
      "step": 9215
    },
    {
      "entropy": 6.200381135940551,
      "epoch": 1.0640507789959608,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004898528558125453,
      "loss": 5.9865,
      "mean_token_accuracy": 0.17022657692432402,
      "num_tokens": 23368011.0,
      "step": 9220
    },
    {
      "entropy": 6.094261741638183,
      "epoch": 1.0646278130409694,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004898405998849492,
      "loss": 6.0414,
      "mean_token_accuracy": 0.17466471195220948,
      "num_tokens": 23380938.0,
      "step": 9225
    },
    {
      "entropy": 6.1515178203582765,
      "epoch": 1.065204847085978,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004898283367312962,
      "loss": 6.1145,
      "mean_token_accuracy": 0.16766797006130219,
      "num_tokens": 23392880.0,
      "step": 9230
    },
    {
      "entropy": 6.217365312576294,
      "epoch": 1.0657818811309867,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004898160663519988,
      "loss": 6.0131,
      "mean_token_accuracy": 0.17246604859828948,
      "num_tokens": 23407809.0,
      "step": 9235
    },
    {
      "entropy": 6.201076650619507,
      "epoch": 1.0663589151759953,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004898037887474697,
      "loss": 6.0513,
      "mean_token_accuracy": 0.16690814346075059,
      "num_tokens": 23419908.0,
      "step": 9240
    },
    {
      "entropy": 6.17437219619751,
      "epoch": 1.066935949221004,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004897915039181219,
      "loss": 6.1151,
      "mean_token_accuracy": 0.16591658890247346,
      "num_tokens": 23431743.0,
      "step": 9245
    },
    {
      "entropy": 6.20676555633545,
      "epoch": 1.0675129832660126,
      "grad_norm": 1.0,
      "learning_rate": 0.0004897792118643685,
      "loss": 6.0093,
      "mean_token_accuracy": 0.17017095685005187,
      "num_tokens": 23444234.0,
      "step": 9250
    },
    {
      "entropy": 6.1153007507324215,
      "epoch": 1.0680900173110213,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004897669125866231,
      "loss": 5.962,
      "mean_token_accuracy": 0.17086610049009324,
      "num_tokens": 23457393.0,
      "step": 9255
    },
    {
      "entropy": 6.10748324394226,
      "epoch": 1.06866705135603,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004897546060852993,
      "loss": 6.0847,
      "mean_token_accuracy": 0.15891457200050355,
      "num_tokens": 23471520.0,
      "step": 9260
    },
    {
      "entropy": 6.140754508972168,
      "epoch": 1.0692440854010385,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004897422923608108,
      "loss": 6.006,
      "mean_token_accuracy": 0.16981667578220366,
      "num_tokens": 23484211.0,
      "step": 9265
    },
    {
      "entropy": 6.154041814804077,
      "epoch": 1.0698211194460474,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004897299714135721,
      "loss": 6.0618,
      "mean_token_accuracy": 0.1673455998301506,
      "num_tokens": 23497110.0,
      "step": 9270
    },
    {
      "entropy": 6.155906009674072,
      "epoch": 1.070398153491056,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004897176432439974,
      "loss": 6.0246,
      "mean_token_accuracy": 0.17300473004579545,
      "num_tokens": 23508803.0,
      "step": 9275
    },
    {
      "entropy": 6.03808560371399,
      "epoch": 1.0709751875360647,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004897053078525016,
      "loss": 5.9544,
      "mean_token_accuracy": 0.17769061177968978,
      "num_tokens": 23522209.0,
      "step": 9280
    },
    {
      "entropy": 6.155244541168213,
      "epoch": 1.0715522215810733,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004896929652394993,
      "loss": 6.0141,
      "mean_token_accuracy": 0.16808729767799377,
      "num_tokens": 23534728.0,
      "step": 9285
    },
    {
      "entropy": 6.117278146743774,
      "epoch": 1.072129255626082,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004896806154054057,
      "loss": 6.0943,
      "mean_token_accuracy": 0.17026905566453934,
      "num_tokens": 23548389.0,
      "step": 9290
    },
    {
      "entropy": 6.151943063735962,
      "epoch": 1.0727062896710906,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004896682583506363,
      "loss": 5.9972,
      "mean_token_accuracy": 0.1696242719888687,
      "num_tokens": 23560260.0,
      "step": 9295
    },
    {
      "entropy": 6.14499192237854,
      "epoch": 1.0732833237160992,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004896558940756067,
      "loss": 6.0013,
      "mean_token_accuracy": 0.16969414949417114,
      "num_tokens": 23572771.0,
      "step": 9300
    },
    {
      "entropy": 6.142427682876587,
      "epoch": 1.0738603577611079,
      "grad_norm": 1.0,
      "learning_rate": 0.0004896435225807327,
      "loss": 6.0524,
      "mean_token_accuracy": 0.16460990607738496,
      "num_tokens": 23584769.0,
      "step": 9305
    },
    {
      "entropy": 6.139098167419434,
      "epoch": 1.0744373918061165,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004896311438664304,
      "loss": 5.9848,
      "mean_token_accuracy": 0.1744297832250595,
      "num_tokens": 23598912.0,
      "step": 9310
    },
    {
      "entropy": 6.149310207366943,
      "epoch": 1.0750144258511252,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004896187579331163,
      "loss": 6.0585,
      "mean_token_accuracy": 0.16703784465789795,
      "num_tokens": 23611781.0,
      "step": 9315
    },
    {
      "entropy": 6.142507696151734,
      "epoch": 1.0755914598961338,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004896063647812068,
      "loss": 6.0822,
      "mean_token_accuracy": 0.16931056082248688,
      "num_tokens": 23624116.0,
      "step": 9320
    },
    {
      "entropy": 6.170571947097779,
      "epoch": 1.0761684939411424,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004895939644111189,
      "loss": 6.0737,
      "mean_token_accuracy": 0.16502795070409776,
      "num_tokens": 23637456.0,
      "step": 9325
    },
    {
      "entropy": 6.1560173511505125,
      "epoch": 1.076745527986151,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004895815568232694,
      "loss": 6.0453,
      "mean_token_accuracy": 0.16622652262449264,
      "num_tokens": 23649572.0,
      "step": 9330
    },
    {
      "entropy": 6.130175495147705,
      "epoch": 1.07732256203116,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004895691420180759,
      "loss": 6.0039,
      "mean_token_accuracy": 0.17502406388521194,
      "num_tokens": 23662934.0,
      "step": 9335
    },
    {
      "entropy": 6.120934629440308,
      "epoch": 1.0778995960761686,
      "grad_norm": 0.921875,
      "learning_rate": 0.000489556719995956,
      "loss": 5.8914,
      "mean_token_accuracy": 0.17915162891149522,
      "num_tokens": 23676258.0,
      "step": 9340
    },
    {
      "entropy": 6.241963720321655,
      "epoch": 1.0784766301211772,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004895442907573274,
      "loss": 6.0457,
      "mean_token_accuracy": 0.16247303485870362,
      "num_tokens": 23689361.0,
      "step": 9345
    },
    {
      "entropy": 6.084423017501831,
      "epoch": 1.0790536641661859,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000489531854302608,
      "loss": 6.0023,
      "mean_token_accuracy": 0.17091086208820344,
      "num_tokens": 23702157.0,
      "step": 9350
    },
    {
      "entropy": 6.105298709869385,
      "epoch": 1.0796306982111945,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004895194106322164,
      "loss": 6.0338,
      "mean_token_accuracy": 0.17374980598688125,
      "num_tokens": 23713846.0,
      "step": 9355
    },
    {
      "entropy": 6.119005012512207,
      "epoch": 1.0802077322562031,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004895069597465709,
      "loss": 5.9378,
      "mean_token_accuracy": 0.17695297598838805,
      "num_tokens": 23727685.0,
      "step": 9360
    },
    {
      "entropy": 6.095948839187622,
      "epoch": 1.0807847663012118,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004894945016460904,
      "loss": 5.9731,
      "mean_token_accuracy": 0.17229541391134262,
      "num_tokens": 23740806.0,
      "step": 9365
    },
    {
      "entropy": 6.091950273513794,
      "epoch": 1.0813618003462204,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004894820363311938,
      "loss": 5.9594,
      "mean_token_accuracy": 0.17340441644191742,
      "num_tokens": 23752968.0,
      "step": 9370
    },
    {
      "entropy": 6.212357664108277,
      "epoch": 1.081938834391229,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004894695638023005,
      "loss": 6.0842,
      "mean_token_accuracy": 0.17258026599884033,
      "num_tokens": 23765141.0,
      "step": 9375
    },
    {
      "entropy": 6.108956670761108,
      "epoch": 1.0825158684362377,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048945708405983,
      "loss": 5.9911,
      "mean_token_accuracy": 0.16863977015018464,
      "num_tokens": 23778294.0,
      "step": 9380
    },
    {
      "entropy": 6.134217357635498,
      "epoch": 1.0830929024812463,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004894445971042019,
      "loss": 6.0044,
      "mean_token_accuracy": 0.1776563212275505,
      "num_tokens": 23790687.0,
      "step": 9385
    },
    {
      "entropy": 6.086469745635986,
      "epoch": 1.083669936526255,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004894321029358364,
      "loss": 5.9765,
      "mean_token_accuracy": 0.16945710331201552,
      "num_tokens": 23803569.0,
      "step": 9390
    },
    {
      "entropy": 6.10461573600769,
      "epoch": 1.0842469705712636,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004894196015551536,
      "loss": 6.0151,
      "mean_token_accuracy": 0.1653943568468094,
      "num_tokens": 23816693.0,
      "step": 9395
    },
    {
      "entropy": 6.194375944137573,
      "epoch": 1.0848240046162723,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000489407092962574,
      "loss": 6.0381,
      "mean_token_accuracy": 0.17516039311885834,
      "num_tokens": 23829139.0,
      "step": 9400
    },
    {
      "entropy": 6.158189868927002,
      "epoch": 1.085401038661281,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004893945771585183,
      "loss": 6.0557,
      "mean_token_accuracy": 0.17037415653467178,
      "num_tokens": 23842489.0,
      "step": 9405
    },
    {
      "entropy": 6.130629205703736,
      "epoch": 1.0859780727062898,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004893820541434075,
      "loss": 6.0047,
      "mean_token_accuracy": 0.17533402144908905,
      "num_tokens": 23854555.0,
      "step": 9410
    },
    {
      "entropy": 6.134566354751587,
      "epoch": 1.0865551067512984,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004893695239176629,
      "loss": 6.0494,
      "mean_token_accuracy": 0.1653420463204384,
      "num_tokens": 23868294.0,
      "step": 9415
    },
    {
      "entropy": 6.1336993217468265,
      "epoch": 1.087132140796307,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004893569864817057,
      "loss": 5.8956,
      "mean_token_accuracy": 0.18521217554807662,
      "num_tokens": 23880475.0,
      "step": 9420
    },
    {
      "entropy": 6.15223274230957,
      "epoch": 1.0877091748413157,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004893444418359579,
      "loss": 6.1148,
      "mean_token_accuracy": 0.16418624818325042,
      "num_tokens": 23893008.0,
      "step": 9425
    },
    {
      "entropy": 6.146541690826416,
      "epoch": 1.0882862088863243,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000489331889980841,
      "loss": 6.0266,
      "mean_token_accuracy": 0.16969874799251555,
      "num_tokens": 23906216.0,
      "step": 9430
    },
    {
      "entropy": 6.108345079421997,
      "epoch": 1.088863242931333,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004893193309167778,
      "loss": 5.9725,
      "mean_token_accuracy": 0.17177551090717316,
      "num_tokens": 23918978.0,
      "step": 9435
    },
    {
      "entropy": 6.177945470809936,
      "epoch": 1.0894402769763416,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004893067646441902,
      "loss": 6.0225,
      "mean_token_accuracy": 0.1699276313185692,
      "num_tokens": 23931536.0,
      "step": 9440
    },
    {
      "entropy": 6.155810737609864,
      "epoch": 1.0900173110213502,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000489294191163501,
      "loss": 6.045,
      "mean_token_accuracy": 0.1679830238223076,
      "num_tokens": 23945364.0,
      "step": 9445
    },
    {
      "entropy": 6.127289342880249,
      "epoch": 1.0905943450663589,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004892816104751331,
      "loss": 5.9965,
      "mean_token_accuracy": 0.1683461770415306,
      "num_tokens": 23956654.0,
      "step": 9450
    },
    {
      "entropy": 6.23105001449585,
      "epoch": 1.0911713791113675,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004892690225795096,
      "loss": 6.0662,
      "mean_token_accuracy": 0.16090586483478547,
      "num_tokens": 23970902.0,
      "step": 9455
    },
    {
      "entropy": 6.2295145988464355,
      "epoch": 1.0917484131563762,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004892564274770542,
      "loss": 6.1072,
      "mean_token_accuracy": 0.1628488004207611,
      "num_tokens": 23984312.0,
      "step": 9460
    },
    {
      "entropy": 6.139108800888062,
      "epoch": 1.0923254472013848,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004892438251681901,
      "loss": 6.0028,
      "mean_token_accuracy": 0.17114972472190856,
      "num_tokens": 23996040.0,
      "step": 9465
    },
    {
      "entropy": 6.050161457061767,
      "epoch": 1.0929024812463934,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004892312156533413,
      "loss": 5.9371,
      "mean_token_accuracy": 0.17475649267435073,
      "num_tokens": 24008737.0,
      "step": 9470
    },
    {
      "entropy": 6.098409843444824,
      "epoch": 1.0934795152914023,
      "grad_norm": 1.0,
      "learning_rate": 0.0004892185989329321,
      "loss": 5.9955,
      "mean_token_accuracy": 0.17283411771059037,
      "num_tokens": 24021735.0,
      "step": 9475
    },
    {
      "entropy": 6.185855865478516,
      "epoch": 1.094056549336411,
      "grad_norm": 0.875,
      "learning_rate": 0.0004892059750073868,
      "loss": 6.0008,
      "mean_token_accuracy": 0.16559487581253052,
      "num_tokens": 24034391.0,
      "step": 9480
    },
    {
      "entropy": 6.164184665679931,
      "epoch": 1.0946335833814196,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004891933438771299,
      "loss": 6.0361,
      "mean_token_accuracy": 0.17066800147294997,
      "num_tokens": 24047331.0,
      "step": 9485
    },
    {
      "entropy": 6.1660871505737305,
      "epoch": 1.0952106174264282,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004891807055425862,
      "loss": 6.0201,
      "mean_token_accuracy": 0.17328795790672302,
      "num_tokens": 24060111.0,
      "step": 9490
    },
    {
      "entropy": 6.110420370101929,
      "epoch": 1.0957876514714369,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004891680600041809,
      "loss": 6.0481,
      "mean_token_accuracy": 0.16835678815841676,
      "num_tokens": 24072275.0,
      "step": 9495
    },
    {
      "entropy": 6.170684194564819,
      "epoch": 1.0963646855164455,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004891554072623392,
      "loss": 5.9982,
      "mean_token_accuracy": 0.16939997524023057,
      "num_tokens": 24084647.0,
      "step": 9500
    },
    {
      "entropy": 6.051642608642578,
      "epoch": 1.0969417195614541,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004891427473174869,
      "loss": 5.9005,
      "mean_token_accuracy": 0.18385812193155288,
      "num_tokens": 24097565.0,
      "step": 9505
    },
    {
      "entropy": 6.134908056259155,
      "epoch": 1.0975187536064628,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004891300801700496,
      "loss": 6.0107,
      "mean_token_accuracy": 0.17309229522943498,
      "num_tokens": 24111603.0,
      "step": 9510
    },
    {
      "entropy": 6.1023242473602295,
      "epoch": 1.0980957876514714,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004891174058204534,
      "loss": 5.9339,
      "mean_token_accuracy": 0.17808732837438584,
      "num_tokens": 24122763.0,
      "step": 9515
    },
    {
      "entropy": 6.073727703094482,
      "epoch": 1.09867282169648,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004891047242691246,
      "loss": 5.9701,
      "mean_token_accuracy": 0.16984187066555023,
      "num_tokens": 24134591.0,
      "step": 9520
    },
    {
      "entropy": 6.164652299880982,
      "epoch": 1.0992498557414887,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004890920355164897,
      "loss": 6.0795,
      "mean_token_accuracy": 0.1696622282266617,
      "num_tokens": 24147574.0,
      "step": 9525
    },
    {
      "entropy": 6.178083896636963,
      "epoch": 1.0998268897864973,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004890793395629756,
      "loss": 6.0213,
      "mean_token_accuracy": 0.1705864265561104,
      "num_tokens": 24159760.0,
      "step": 9530
    },
    {
      "entropy": 6.039093542098999,
      "epoch": 1.100403923831506,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004890666364090093,
      "loss": 5.9479,
      "mean_token_accuracy": 0.1791314274072647,
      "num_tokens": 24173577.0,
      "step": 9535
    },
    {
      "entropy": 6.151444387435913,
      "epoch": 1.1009809578765146,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004890539260550181,
      "loss": 5.9887,
      "mean_token_accuracy": 0.1732378125190735,
      "num_tokens": 24185180.0,
      "step": 9540
    },
    {
      "entropy": 6.12696008682251,
      "epoch": 1.1015579919215233,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004890412085014292,
      "loss": 5.9749,
      "mean_token_accuracy": 0.16669239699840546,
      "num_tokens": 24198967.0,
      "step": 9545
    },
    {
      "entropy": 6.141853094100952,
      "epoch": 1.1021350259665321,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004890284837486706,
      "loss": 6.0593,
      "mean_token_accuracy": 0.16915735602378845,
      "num_tokens": 24210770.0,
      "step": 9550
    },
    {
      "entropy": 6.15566840171814,
      "epoch": 1.1027120600115408,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004890157517971704,
      "loss": 6.0001,
      "mean_token_accuracy": 0.17189182192087174,
      "num_tokens": 24222970.0,
      "step": 9555
    },
    {
      "entropy": 6.1692249298095705,
      "epoch": 1.1032890940565494,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004890030126473566,
      "loss": 6.0511,
      "mean_token_accuracy": 0.16246043294668197,
      "num_tokens": 24235591.0,
      "step": 9560
    },
    {
      "entropy": 6.156991338729858,
      "epoch": 1.103866128101558,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004889902662996578,
      "loss": 5.9932,
      "mean_token_accuracy": 0.1733967199921608,
      "num_tokens": 24248844.0,
      "step": 9565
    },
    {
      "entropy": 6.132727479934692,
      "epoch": 1.1044431621465667,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004889775127545027,
      "loss": 5.9992,
      "mean_token_accuracy": 0.1738229751586914,
      "num_tokens": 24260267.0,
      "step": 9570
    },
    {
      "entropy": 6.169062805175781,
      "epoch": 1.1050201961915753,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004889647520123202,
      "loss": 6.0671,
      "mean_token_accuracy": 0.16595419645309448,
      "num_tokens": 24273019.0,
      "step": 9575
    },
    {
      "entropy": 6.082721519470215,
      "epoch": 1.105597230236584,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004889519840735396,
      "loss": 5.965,
      "mean_token_accuracy": 0.17550068944692612,
      "num_tokens": 24284805.0,
      "step": 9580
    },
    {
      "entropy": 6.123953866958618,
      "epoch": 1.1061742642815926,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004889392089385903,
      "loss": 6.0483,
      "mean_token_accuracy": 0.1621156945824623,
      "num_tokens": 24297080.0,
      "step": 9585
    },
    {
      "entropy": 6.147588586807251,
      "epoch": 1.1067512983266012,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004889264266079019,
      "loss": 6.0183,
      "mean_token_accuracy": 0.16932634711265565,
      "num_tokens": 24308237.0,
      "step": 9590
    },
    {
      "entropy": 6.138839960098267,
      "epoch": 1.1073283323716099,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004889136370819045,
      "loss": 6.018,
      "mean_token_accuracy": 0.16850509196519853,
      "num_tokens": 24322435.0,
      "step": 9595
    },
    {
      "entropy": 6.1553730964660645,
      "epoch": 1.1079053664166185,
      "grad_norm": 1.0,
      "learning_rate": 0.0004889008403610281,
      "loss": 6.0326,
      "mean_token_accuracy": 0.1633181557059288,
      "num_tokens": 24335372.0,
      "step": 9600
    },
    {
      "entropy": 6.1259418487548825,
      "epoch": 1.1084824004616272,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004888880364457033,
      "loss": 5.945,
      "mean_token_accuracy": 0.1769667759537697,
      "num_tokens": 24347956.0,
      "step": 9605
    },
    {
      "entropy": 6.143293809890747,
      "epoch": 1.1090594345066358,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004888752253363604,
      "loss": 5.9864,
      "mean_token_accuracy": 0.1717490866780281,
      "num_tokens": 24360378.0,
      "step": 9610
    },
    {
      "entropy": 6.126335287094117,
      "epoch": 1.1096364685516447,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004888624070334308,
      "loss": 6.0175,
      "mean_token_accuracy": 0.17199165225028992,
      "num_tokens": 24372492.0,
      "step": 9615
    },
    {
      "entropy": 6.186481761932373,
      "epoch": 1.1102135025966533,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004888495815373452,
      "loss": 6.0469,
      "mean_token_accuracy": 0.16621554791927337,
      "num_tokens": 24384634.0,
      "step": 9620
    },
    {
      "entropy": 6.154370498657227,
      "epoch": 1.110790536641662,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004888367488485351,
      "loss": 6.0587,
      "mean_token_accuracy": 0.16920675039291383,
      "num_tokens": 24397117.0,
      "step": 9625
    },
    {
      "entropy": 6.127230834960938,
      "epoch": 1.1113675706866706,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004888239089674323,
      "loss": 6.0039,
      "mean_token_accuracy": 0.17373434454202652,
      "num_tokens": 24409071.0,
      "step": 9630
    },
    {
      "entropy": 6.06903600692749,
      "epoch": 1.1119446047316792,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004888110618944686,
      "loss": 5.9878,
      "mean_token_accuracy": 0.16834839731454848,
      "num_tokens": 24422016.0,
      "step": 9635
    },
    {
      "entropy": 6.162847185134888,
      "epoch": 1.1125216387766879,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004887982076300759,
      "loss": 6.0319,
      "mean_token_accuracy": 0.1700489729642868,
      "num_tokens": 24433356.0,
      "step": 9640
    },
    {
      "entropy": 6.015257453918457,
      "epoch": 1.1130986728216965,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004887853461746867,
      "loss": 5.9455,
      "mean_token_accuracy": 0.18017970770597458,
      "num_tokens": 24449027.0,
      "step": 9645
    },
    {
      "entropy": 6.160163164138794,
      "epoch": 1.1136757068667051,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004887724775287336,
      "loss": 6.017,
      "mean_token_accuracy": 0.16833999305963515,
      "num_tokens": 24461592.0,
      "step": 9650
    },
    {
      "entropy": 6.141183567047119,
      "epoch": 1.1142527409117138,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004887596016926494,
      "loss": 6.0319,
      "mean_token_accuracy": 0.16653386503458023,
      "num_tokens": 24474003.0,
      "step": 9655
    },
    {
      "entropy": 6.138088941574097,
      "epoch": 1.1148297749567224,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004887467186668673,
      "loss": 5.9615,
      "mean_token_accuracy": 0.1718171179294586,
      "num_tokens": 24487096.0,
      "step": 9660
    },
    {
      "entropy": 6.115821647644043,
      "epoch": 1.115406809001731,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004887338284518204,
      "loss": 5.9003,
      "mean_token_accuracy": 0.18337261378765107,
      "num_tokens": 24499685.0,
      "step": 9665
    },
    {
      "entropy": 6.0980370998382565,
      "epoch": 1.1159838430467397,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004887209310479423,
      "loss": 6.0535,
      "mean_token_accuracy": 0.16609314531087876,
      "num_tokens": 24512372.0,
      "step": 9670
    },
    {
      "entropy": 5.989741563796997,
      "epoch": 1.1165608770917483,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004887080264556668,
      "loss": 5.7998,
      "mean_token_accuracy": 0.18902382701635362,
      "num_tokens": 24524694.0,
      "step": 9675
    },
    {
      "entropy": 6.095667028427124,
      "epoch": 1.117137911136757,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004886951146754282,
      "loss": 5.9344,
      "mean_token_accuracy": 0.17629175186157225,
      "num_tokens": 24537779.0,
      "step": 9680
    },
    {
      "entropy": 6.103324699401855,
      "epoch": 1.1177149451817656,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004886821957076603,
      "loss": 5.9453,
      "mean_token_accuracy": 0.17716581672430037,
      "num_tokens": 24551012.0,
      "step": 9685
    },
    {
      "entropy": 6.0966085433959964,
      "epoch": 1.1182919792267745,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000488669269552798,
      "loss": 5.9681,
      "mean_token_accuracy": 0.16760742366313935,
      "num_tokens": 24562464.0,
      "step": 9690
    },
    {
      "entropy": 6.185573768615723,
      "epoch": 1.1188690132717831,
      "grad_norm": 1.015625,
      "learning_rate": 0.000488656336211276,
      "loss": 5.9728,
      "mean_token_accuracy": 0.17604034543037414,
      "num_tokens": 24574223.0,
      "step": 9695
    },
    {
      "entropy": 6.1533203125,
      "epoch": 1.1194460473167918,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004886433956835292,
      "loss": 6.0023,
      "mean_token_accuracy": 0.1715945780277252,
      "num_tokens": 24586464.0,
      "step": 9700
    },
    {
      "entropy": 6.050113296508789,
      "epoch": 1.1200230813618004,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004886304479699929,
      "loss": 6.0083,
      "mean_token_accuracy": 0.16716951727867127,
      "num_tokens": 24599071.0,
      "step": 9705
    },
    {
      "entropy": 6.041171741485596,
      "epoch": 1.120600115406809,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004886174930711027,
      "loss": 5.8929,
      "mean_token_accuracy": 0.1855643182992935,
      "num_tokens": 24611302.0,
      "step": 9710
    },
    {
      "entropy": 6.109375619888306,
      "epoch": 1.1211771494518177,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004886045309872941,
      "loss": 6.0575,
      "mean_token_accuracy": 0.16989795863628387,
      "num_tokens": 24624725.0,
      "step": 9715
    },
    {
      "entropy": 6.160960006713867,
      "epoch": 1.1217541834968263,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004885915617190034,
      "loss": 6.0679,
      "mean_token_accuracy": 0.16822285503149031,
      "num_tokens": 24638797.0,
      "step": 9720
    },
    {
      "entropy": 6.200824403762818,
      "epoch": 1.122331217541835,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004885785852666664,
      "loss": 6.002,
      "mean_token_accuracy": 0.16909609735012054,
      "num_tokens": 24651462.0,
      "step": 9725
    },
    {
      "entropy": 6.068878698348999,
      "epoch": 1.1229082515868436,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004885656016307199,
      "loss": 5.9979,
      "mean_token_accuracy": 0.17408033162355424,
      "num_tokens": 24664351.0,
      "step": 9730
    },
    {
      "entropy": 6.155042219161987,
      "epoch": 1.1234852856318522,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004885526108116004,
      "loss": 6.0066,
      "mean_token_accuracy": 0.17068607807159425,
      "num_tokens": 24677332.0,
      "step": 9735
    },
    {
      "entropy": 6.14207124710083,
      "epoch": 1.1240623196768609,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000488539612809745,
      "loss": 6.0272,
      "mean_token_accuracy": 0.167741958796978,
      "num_tokens": 24690408.0,
      "step": 9740
    },
    {
      "entropy": 6.10364089012146,
      "epoch": 1.1246393537218695,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004885266076255907,
      "loss": 6.1357,
      "mean_token_accuracy": 0.16068277209997178,
      "num_tokens": 24703776.0,
      "step": 9745
    },
    {
      "entropy": 6.147567939758301,
      "epoch": 1.1252163877668782,
      "grad_norm": 1.859375,
      "learning_rate": 0.000488513595259575,
      "loss": 5.9106,
      "mean_token_accuracy": 0.1776602402329445,
      "num_tokens": 24716385.0,
      "step": 9750
    },
    {
      "entropy": 6.071480464935303,
      "epoch": 1.125793421811887,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004885005757121357,
      "loss": 5.9261,
      "mean_token_accuracy": 0.17253369987010955,
      "num_tokens": 24729055.0,
      "step": 9755
    },
    {
      "entropy": 6.101696443557739,
      "epoch": 1.1263704558568954,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004884875489837105,
      "loss": 5.9824,
      "mean_token_accuracy": 0.17328732907772065,
      "num_tokens": 24741168.0,
      "step": 9760
    },
    {
      "entropy": 6.190765237808227,
      "epoch": 1.1269474899019043,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004884745150747378,
      "loss": 5.9625,
      "mean_token_accuracy": 0.17811059653759004,
      "num_tokens": 24753682.0,
      "step": 9765
    },
    {
      "entropy": 6.010779094696045,
      "epoch": 1.127524523946913,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004884614739856556,
      "loss": 5.958,
      "mean_token_accuracy": 0.17995766550302505,
      "num_tokens": 24767528.0,
      "step": 9770
    },
    {
      "entropy": 6.110750675201416,
      "epoch": 1.1281015579919216,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004884484257169028,
      "loss": 6.047,
      "mean_token_accuracy": 0.1675013780593872,
      "num_tokens": 24780989.0,
      "step": 9775
    },
    {
      "entropy": 6.17375545501709,
      "epoch": 1.1286785920369302,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004884353702689183,
      "loss": 6.0322,
      "mean_token_accuracy": 0.165440109372139,
      "num_tokens": 24792595.0,
      "step": 9780
    },
    {
      "entropy": 6.083075046539307,
      "epoch": 1.1292556260819389,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004884223076421411,
      "loss": 5.9495,
      "mean_token_accuracy": 0.1709774300456047,
      "num_tokens": 24805665.0,
      "step": 9785
    },
    {
      "entropy": 6.11997652053833,
      "epoch": 1.1298326601269475,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004884092378370106,
      "loss": 5.9103,
      "mean_token_accuracy": 0.17966125458478927,
      "num_tokens": 24818276.0,
      "step": 9790
    },
    {
      "entropy": 6.092645502090454,
      "epoch": 1.1304096941719561,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004883961608539664,
      "loss": 5.9657,
      "mean_token_accuracy": 0.17110249996185303,
      "num_tokens": 24831828.0,
      "step": 9795
    },
    {
      "entropy": 6.066320753097534,
      "epoch": 1.1309867282169648,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004883830766934484,
      "loss": 5.9453,
      "mean_token_accuracy": 0.18093480467796325,
      "num_tokens": 24844597.0,
      "step": 9800
    },
    {
      "entropy": 6.094020938873291,
      "epoch": 1.1315637622619734,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004883699853558965,
      "loss": 5.9397,
      "mean_token_accuracy": 0.17109753489494323,
      "num_tokens": 24856136.0,
      "step": 9805
    },
    {
      "entropy": 6.190833854675293,
      "epoch": 1.132140796306982,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004883568868417511,
      "loss": 6.1154,
      "mean_token_accuracy": 0.16142057478427888,
      "num_tokens": 24870888.0,
      "step": 9810
    },
    {
      "entropy": 6.146378087997436,
      "epoch": 1.1327178303519907,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004883437811514528,
      "loss": 6.0301,
      "mean_token_accuracy": 0.17373637109994888,
      "num_tokens": 24884239.0,
      "step": 9815
    },
    {
      "entropy": 6.141600227355957,
      "epoch": 1.1332948643969993,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004883306682854424,
      "loss": 5.9696,
      "mean_token_accuracy": 0.18201631009578706,
      "num_tokens": 24896370.0,
      "step": 9820
    },
    {
      "entropy": 6.0363202571868895,
      "epoch": 1.133871898442008,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004883175482441611,
      "loss": 6.036,
      "mean_token_accuracy": 0.16015657782554626,
      "num_tokens": 24908868.0,
      "step": 9825
    },
    {
      "entropy": 6.104412841796875,
      "epoch": 1.1344489324870168,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004883044210280499,
      "loss": 6.0725,
      "mean_token_accuracy": 0.1708388537168503,
      "num_tokens": 24920630.0,
      "step": 9830
    },
    {
      "entropy": 6.045358276367187,
      "epoch": 1.1350259665320255,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004882912866375505,
      "loss": 5.8725,
      "mean_token_accuracy": 0.1796233534812927,
      "num_tokens": 24932787.0,
      "step": 9835
    },
    {
      "entropy": 6.1177649974823,
      "epoch": 1.1356030005770341,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048827814507310455,
      "loss": 5.9633,
      "mean_token_accuracy": 0.17206400632858276,
      "num_tokens": 24944436.0,
      "step": 9840
    },
    {
      "entropy": 6.140950679779053,
      "epoch": 1.1361800346220428,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048826499633515416,
      "loss": 5.9715,
      "mean_token_accuracy": 0.17313409894704818,
      "num_tokens": 24956944.0,
      "step": 9845
    },
    {
      "entropy": 6.183871793746948,
      "epoch": 1.1367570686670514,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048825184042414156,
      "loss": 6.0278,
      "mean_token_accuracy": 0.16846390217542648,
      "num_tokens": 24971540.0,
      "step": 9850
    },
    {
      "entropy": 6.122875022888183,
      "epoch": 1.13733410271206,
      "grad_norm": 1.0,
      "learning_rate": 0.0004882386773405092,
      "loss": 6.0572,
      "mean_token_accuracy": 0.16872498840093614,
      "num_tokens": 24984056.0,
      "step": 9855
    },
    {
      "entropy": 6.083562994003296,
      "epoch": 1.1379111367570687,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004882255070846999,
      "loss": 6.0692,
      "mean_token_accuracy": 0.1663650706410408,
      "num_tokens": 24995625.0,
      "step": 9860
    },
    {
      "entropy": 6.187272310256958,
      "epoch": 1.1384881708020773,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048821232965715663,
      "loss": 5.9302,
      "mean_token_accuracy": 0.1676005020737648,
      "num_tokens": 25007336.0,
      "step": 9865
    },
    {
      "entropy": 6.143199348449707,
      "epoch": 1.139065204847086,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004881991450583225,
      "loss": 6.021,
      "mean_token_accuracy": 0.17064233720302582,
      "num_tokens": 25019446.0,
      "step": 9870
    },
    {
      "entropy": 6.10345196723938,
      "epoch": 1.1396422388920946,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000488185953288641,
      "loss": 5.9614,
      "mean_token_accuracy": 0.17136456072330475,
      "num_tokens": 25031518.0,
      "step": 9875
    },
    {
      "entropy": 6.080751657485962,
      "epoch": 1.1402192729371032,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004881727543485559,
      "loss": 5.9022,
      "mean_token_accuracy": 0.1812594398856163,
      "num_tokens": 25043365.0,
      "step": 9880
    },
    {
      "entropy": 6.138740921020508,
      "epoch": 1.1407963069821119,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048815954823851107,
      "loss": 6.0706,
      "mean_token_accuracy": 0.1644753023982048,
      "num_tokens": 25055748.0,
      "step": 9885
    },
    {
      "entropy": 6.083611583709716,
      "epoch": 1.1413733410271205,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048814633495895067,
      "loss": 5.93,
      "mean_token_accuracy": 0.1717399090528488,
      "num_tokens": 25067904.0,
      "step": 9890
    },
    {
      "entropy": 6.06887469291687,
      "epoch": 1.1419503750721294,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004881331145103192,
      "loss": 5.9845,
      "mean_token_accuracy": 0.17358410954475403,
      "num_tokens": 25081584.0,
      "step": 9895
    },
    {
      "entropy": 6.116873836517334,
      "epoch": 1.1425274091171378,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004881198868930614,
      "loss": 6.0129,
      "mean_token_accuracy": 0.16618734300136567,
      "num_tokens": 25094555.0,
      "step": 9900
    },
    {
      "entropy": 6.172186708450317,
      "epoch": 1.1431044431621467,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048810665210762195,
      "loss": 6.0634,
      "mean_token_accuracy": 0.17028094530105592,
      "num_tokens": 25108405.0,
      "step": 9905
    },
    {
      "entropy": 6.150914287567138,
      "epoch": 1.1436814772071553,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048809341015444615,
      "loss": 6.0276,
      "mean_token_accuracy": 0.16898710429668426,
      "num_tokens": 25121404.0,
      "step": 9910
    },
    {
      "entropy": 6.149497175216675,
      "epoch": 1.144258511252164,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048808016103397934,
      "loss": 5.9617,
      "mean_token_accuracy": 0.1780024603009224,
      "num_tokens": 25134773.0,
      "step": 9915
    },
    {
      "entropy": 6.109142875671386,
      "epoch": 1.1448355452971726,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004880669047466671,
      "loss": 5.9399,
      "mean_token_accuracy": 0.17508725970983505,
      "num_tokens": 25147873.0,
      "step": 9920
    },
    {
      "entropy": 6.045895671844482,
      "epoch": 1.1454125793421812,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048805364129295554,
      "loss": 5.903,
      "mean_token_accuracy": 0.1838482975959778,
      "num_tokens": 25162081.0,
      "step": 9925
    },
    {
      "entropy": 6.183807039260865,
      "epoch": 1.1459896133871899,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048804037067329037,
      "loss": 6.083,
      "mean_token_accuracy": 0.1692851409316063,
      "num_tokens": 25174203.0,
      "step": 9930
    },
    {
      "entropy": 6.181288766860962,
      "epoch": 1.1465666474321985,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004880270928881183,
      "loss": 6.0812,
      "mean_token_accuracy": 0.17000937908887864,
      "num_tokens": 25187238.0,
      "step": 9935
    },
    {
      "entropy": 6.153555583953858,
      "epoch": 1.1471436814772071,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004880138079378857,
      "loss": 5.9243,
      "mean_token_accuracy": 0.18020797818899154,
      "num_tokens": 25199456.0,
      "step": 9940
    },
    {
      "entropy": 6.12066764831543,
      "epoch": 1.1477207155222158,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004880005158230395,
      "loss": 6.0125,
      "mean_token_accuracy": 0.16558817774057388,
      "num_tokens": 25215163.0,
      "step": 9945
    },
    {
      "entropy": 6.071428203582764,
      "epoch": 1.1482977495672244,
      "grad_norm": 1.0,
      "learning_rate": 0.0004879872165440268,
      "loss": 5.9779,
      "mean_token_accuracy": 0.17602927535772322,
      "num_tokens": 25226946.0,
      "step": 9950
    },
    {
      "entropy": 6.112587118148804,
      "epoch": 1.148874783612233,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004879739101012948,
      "loss": 5.9342,
      "mean_token_accuracy": 0.17014381438493728,
      "num_tokens": 25238691.0,
      "step": 9955
    },
    {
      "entropy": 6.080486536026001,
      "epoch": 1.1494518176572417,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004879605964952911,
      "loss": 5.9269,
      "mean_token_accuracy": 0.17008888572454453,
      "num_tokens": 25251513.0,
      "step": 9960
    },
    {
      "entropy": 6.087462854385376,
      "epoch": 1.1500288517022503,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048794727572646366,
      "loss": 5.8794,
      "mean_token_accuracy": 0.17589139491319655,
      "num_tokens": 25263513.0,
      "step": 9965
    },
    {
      "entropy": 6.056052923202515,
      "epoch": 1.1506058857472592,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004879339477952603,
      "loss": 6.0009,
      "mean_token_accuracy": 0.1690943330526352,
      "num_tokens": 25276989.0,
      "step": 9970
    },
    {
      "entropy": 6.146207427978515,
      "epoch": 1.1511829197922678,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048792061270212935,
      "loss": 6.0052,
      "mean_token_accuracy": 0.17554166465997695,
      "num_tokens": 25289474.0,
      "step": 9975
    },
    {
      "entropy": 6.180155324935913,
      "epoch": 1.1517599538372765,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004879072704475193,
      "loss": 5.9994,
      "mean_token_accuracy": 0.17543792128562927,
      "num_tokens": 25302102.0,
      "step": 9980
    },
    {
      "entropy": 6.067106246948242,
      "epoch": 1.1523369878822851,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048789392103187906,
      "loss": 5.969,
      "mean_token_accuracy": 0.17342827767133712,
      "num_tokens": 25315645.0,
      "step": 9985
    },
    {
      "entropy": 6.186014795303345,
      "epoch": 1.1529140219272938,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004878805644556575,
      "loss": 6.0534,
      "mean_token_accuracy": 0.16841434687376022,
      "num_tokens": 25328272.0,
      "step": 9990
    },
    {
      "entropy": 6.07266526222229,
      "epoch": 1.1534910559723024,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000487867200719304,
      "loss": 5.9849,
      "mean_token_accuracy": 0.17112089842557907,
      "num_tokens": 25340449.0,
      "step": 9995
    },
    {
      "entropy": 6.107190036773682,
      "epoch": 1.154068090017311,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004878538298232678,
      "loss": 5.9438,
      "mean_token_accuracy": 0.17522693127393724,
      "num_tokens": 25353232.0,
      "step": 10000
    },
    {
      "entropy": 6.10538387298584,
      "epoch": 1.1546451240623197,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004878404517679988,
      "loss": 5.9974,
      "mean_token_accuracy": 0.1720699042081833,
      "num_tokens": 25364575.0,
      "step": 10005
    },
    {
      "entropy": 6.01102204322815,
      "epoch": 1.1552221581073283,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048782706655394695,
      "loss": 5.9372,
      "mean_token_accuracy": 0.17419279366731644,
      "num_tokens": 25377992.0,
      "step": 10010
    },
    {
      "entropy": 6.181147480010987,
      "epoch": 1.155799192152337,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004878136741815624,
      "loss": 6.0637,
      "mean_token_accuracy": 0.16948312819004058,
      "num_tokens": 25391168.0,
      "step": 10015
    },
    {
      "entropy": 6.089163494110108,
      "epoch": 1.1563762261973456,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004878002746512956,
      "loss": 5.9222,
      "mean_token_accuracy": 0.17240351289510727,
      "num_tokens": 25405211.0,
      "step": 10020
    },
    {
      "entropy": 6.079294681549072,
      "epoch": 1.1569532602423542,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004877868679635974,
      "loss": 5.9846,
      "mean_token_accuracy": 0.17463910728693008,
      "num_tokens": 25416902.0,
      "step": 10025
    },
    {
      "entropy": 6.122024202346802,
      "epoch": 1.1575302942873629,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004877734541189185,
      "loss": 6.0462,
      "mean_token_accuracy": 0.16756020188331605,
      "num_tokens": 25429071.0,
      "step": 10030
    },
    {
      "entropy": 6.080726957321167,
      "epoch": 1.1581073283323717,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048776003311771013,
      "loss": 5.9464,
      "mean_token_accuracy": 0.16763327717781068,
      "num_tokens": 25441153.0,
      "step": 10035
    },
    {
      "entropy": 6.118249130249024,
      "epoch": 1.1586843623773802,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004877466049604238,
      "loss": 5.931,
      "mean_token_accuracy": 0.17346419245004654,
      "num_tokens": 25454302.0,
      "step": 10040
    },
    {
      "entropy": 6.151766061782837,
      "epoch": 1.159261396422389,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048773316964751106,
      "loss": 6.018,
      "mean_token_accuracy": 0.16664298474788666,
      "num_tokens": 25466507.0,
      "step": 10045
    },
    {
      "entropy": 6.1742949962615965,
      "epoch": 1.1598384304673977,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004877197271794239,
      "loss": 5.9981,
      "mean_token_accuracy": 0.1746511399745941,
      "num_tokens": 25479952.0,
      "step": 10050
    },
    {
      "entropy": 6.159850978851319,
      "epoch": 1.1604154645124063,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004877062775566142,
      "loss": 6.0819,
      "mean_token_accuracy": 0.16323066055774688,
      "num_tokens": 25491952.0,
      "step": 10055
    },
    {
      "entropy": 6.17249321937561,
      "epoch": 1.160992498557415,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048769282077953464,
      "loss": 5.977,
      "mean_token_accuracy": 0.17066630870103836,
      "num_tokens": 25505053.0,
      "step": 10060
    },
    {
      "entropy": 6.195417833328247,
      "epoch": 1.1615695326024236,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048767935684863775,
      "loss": 6.0504,
      "mean_token_accuracy": 0.16673124134540557,
      "num_tokens": 25517367.0,
      "step": 10065
    },
    {
      "entropy": 6.158107900619507,
      "epoch": 1.1621465666474322,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004876658857643762,
      "loss": 6.0131,
      "mean_token_accuracy": 0.1687976449728012,
      "num_tokens": 25528697.0,
      "step": 10070
    },
    {
      "entropy": 6.141379261016846,
      "epoch": 1.1627236006924409,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004876524075272034,
      "loss": 6.0119,
      "mean_token_accuracy": 0.1653753399848938,
      "num_tokens": 25541109.0,
      "step": 10075
    },
    {
      "entropy": 6.04984450340271,
      "epoch": 1.1633006347374495,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048763892213757234,
      "loss": 5.9524,
      "mean_token_accuracy": 0.171206097304821,
      "num_tokens": 25553870.0,
      "step": 10080
    },
    {
      "entropy": 6.139471387863159,
      "epoch": 1.1638776687824581,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048762542959593683,
      "loss": 6.0234,
      "mean_token_accuracy": 0.16759473830461502,
      "num_tokens": 25566048.0,
      "step": 10085
    },
    {
      "entropy": 6.131299448013306,
      "epoch": 1.1644547028274668,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004876119299027506,
      "loss": 5.9071,
      "mean_token_accuracy": 0.17831585705280303,
      "num_tokens": 25578282.0,
      "step": 10090
    },
    {
      "entropy": 6.064617490768432,
      "epoch": 1.1650317368724754,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048759842305846766,
      "loss": 5.912,
      "mean_token_accuracy": 0.17697295397520066,
      "num_tokens": 25590359.0,
      "step": 10095
    },
    {
      "entropy": 6.184170055389404,
      "epoch": 1.165608770917484,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004875849090635425,
      "loss": 6.0475,
      "mean_token_accuracy": 0.16841991543769835,
      "num_tokens": 25603256.0,
      "step": 10100
    },
    {
      "entropy": 6.1224284172058105,
      "epoch": 1.1661858049624927,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048757138791842943,
      "loss": 6.0344,
      "mean_token_accuracy": 0.1670784518122673,
      "num_tokens": 25616844.0,
      "step": 10105
    },
    {
      "entropy": 6.08445987701416,
      "epoch": 1.1667628390075016,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004875578596235832,
      "loss": 5.9025,
      "mean_token_accuracy": 0.18193238377571105,
      "num_tokens": 25630490.0,
      "step": 10110
    },
    {
      "entropy": 6.129186010360717,
      "epoch": 1.1673398730525102,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004875443241794591,
      "loss": 5.9851,
      "mean_token_accuracy": 0.16921836584806443,
      "num_tokens": 25645082.0,
      "step": 10115
    },
    {
      "entropy": 6.149543714523316,
      "epoch": 1.1679169070975188,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048753078158651227,
      "loss": 6.0386,
      "mean_token_accuracy": 0.16803991198539733,
      "num_tokens": 25658108.0,
      "step": 10120
    },
    {
      "entropy": 6.1145768642425535,
      "epoch": 1.1684939411425275,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000487517231845198,
      "loss": 6.0108,
      "mean_token_accuracy": 0.17040407359600068,
      "num_tokens": 25669883.0,
      "step": 10125
    },
    {
      "entropy": 6.145943021774292,
      "epoch": 1.169070975187536,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004875036749559724,
      "loss": 6.018,
      "mean_token_accuracy": 0.16489063054323197,
      "num_tokens": 25683393.0,
      "step": 10130
    },
    {
      "entropy": 6.211561393737793,
      "epoch": 1.1696480092325447,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048749011091929115,
      "loss": 6.0056,
      "mean_token_accuracy": 0.171510811150074,
      "num_tokens": 25696564.0,
      "step": 10135
    },
    {
      "entropy": 6.055604934692383,
      "epoch": 1.1702250432775534,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004874765397356105,
      "loss": 5.9167,
      "mean_token_accuracy": 0.1721386045217514,
      "num_tokens": 25709350.0,
      "step": 10140
    },
    {
      "entropy": 6.1124763011932375,
      "epoch": 1.170802077322562,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004874629614053871,
      "loss": 5.9908,
      "mean_token_accuracy": 0.16749476194381713,
      "num_tokens": 25721559.0,
      "step": 10145
    },
    {
      "entropy": 6.210343074798584,
      "epoch": 1.1713791113675707,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004874493759290775,
      "loss": 6.098,
      "mean_token_accuracy": 0.1640730544924736,
      "num_tokens": 25734185.0,
      "step": 10150
    },
    {
      "entropy": 6.10564923286438,
      "epoch": 1.1719561454125793,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048743578330713854,
      "loss": 5.9391,
      "mean_token_accuracy": 0.1765938386321068,
      "num_tokens": 25746456.0,
      "step": 10155
    },
    {
      "entropy": 6.058978843688965,
      "epoch": 1.172533179457588,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004874221835400277,
      "loss": 5.9507,
      "mean_token_accuracy": 0.1752479001879692,
      "num_tokens": 25759893.0,
      "step": 10160
    },
    {
      "entropy": 6.198644065856934,
      "epoch": 1.1731102135025966,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004874085766282022,
      "loss": 6.0786,
      "mean_token_accuracy": 0.16533454954624177,
      "num_tokens": 25772574.0,
      "step": 10165
    },
    {
      "entropy": 6.133393144607544,
      "epoch": 1.1736872475476052,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048739496257211966,
      "loss": 5.9424,
      "mean_token_accuracy": 0.1788910761475563,
      "num_tokens": 25785748.0,
      "step": 10170
    },
    {
      "entropy": 6.054536294937134,
      "epoch": 1.1742642815926139,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048738134137223815,
      "loss": 5.9675,
      "mean_token_accuracy": 0.1723702758550644,
      "num_tokens": 25799073.0,
      "step": 10175
    },
    {
      "entropy": 6.166691827774048,
      "epoch": 1.1748413156376225,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048736771302901566,
      "loss": 5.9959,
      "mean_token_accuracy": 0.166565702855587,
      "num_tokens": 25810896.0,
      "step": 10180
    },
    {
      "entropy": 6.147251844406128,
      "epoch": 1.1754183496826314,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048735407754291063,
      "loss": 5.9618,
      "mean_token_accuracy": 0.1683138757944107,
      "num_tokens": 25824728.0,
      "step": 10185
    },
    {
      "entropy": 6.098790693283081,
      "epoch": 1.17599538372764,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048734043491438175,
      "loss": 6.0155,
      "mean_token_accuracy": 0.1655132442712784,
      "num_tokens": 25837569.0,
      "step": 10190
    },
    {
      "entropy": 6.1793218612670895,
      "epoch": 1.1765724177726486,
      "grad_norm": 1.0,
      "learning_rate": 0.00048732678514388773,
      "loss": 6.0856,
      "mean_token_accuracy": 0.1650678649544716,
      "num_tokens": 25850232.0,
      "step": 10195
    },
    {
      "entropy": 6.135155105590821,
      "epoch": 1.1771494518176573,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004873131282318878,
      "loss": 6.0279,
      "mean_token_accuracy": 0.1735972598195076,
      "num_tokens": 25862038.0,
      "step": 10200
    },
    {
      "entropy": 6.189524555206299,
      "epoch": 1.177726485862666,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048729946417884126,
      "loss": 6.0928,
      "mean_token_accuracy": 0.16132732927799226,
      "num_tokens": 25874337.0,
      "step": 10205
    },
    {
      "entropy": 6.148887777328492,
      "epoch": 1.1783035199076746,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004872857929852076,
      "loss": 6.0115,
      "mean_token_accuracy": 0.17398639917373657,
      "num_tokens": 25888175.0,
      "step": 10210
    },
    {
      "entropy": 6.1385846614837645,
      "epoch": 1.1788805539526832,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004872721146514469,
      "loss": 5.9199,
      "mean_token_accuracy": 0.17255275696516037,
      "num_tokens": 25900747.0,
      "step": 10215
    },
    {
      "entropy": 6.026354122161865,
      "epoch": 1.1794575879976918,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000487258429178019,
      "loss": 5.9781,
      "mean_token_accuracy": 0.16793400198221206,
      "num_tokens": 25913010.0,
      "step": 10220
    },
    {
      "entropy": 6.112648582458496,
      "epoch": 1.1800346220427005,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048724473656538427,
      "loss": 5.9111,
      "mean_token_accuracy": 0.16987087577581406,
      "num_tokens": 25925757.0,
      "step": 10225
    },
    {
      "entropy": 6.163606929779053,
      "epoch": 1.1806116560877091,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004872310368140032,
      "loss": 5.9977,
      "mean_token_accuracy": 0.16646733582019807,
      "num_tokens": 25939574.0,
      "step": 10230
    },
    {
      "entropy": 6.114997529983521,
      "epoch": 1.1811886901327178,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004872173299243368,
      "loss": 6.01,
      "mean_token_accuracy": 0.1681802049279213,
      "num_tokens": 25951142.0,
      "step": 10235
    },
    {
      "entropy": 6.117448472976685,
      "epoch": 1.1817657241777264,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048720361589684575,
      "loss": 5.9948,
      "mean_token_accuracy": 0.1701868936419487,
      "num_tokens": 25962575.0,
      "step": 10240
    },
    {
      "entropy": 6.111576509475708,
      "epoch": 1.182342758222735,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048718989473199147,
      "loss": 5.9942,
      "mean_token_accuracy": 0.17484725415706634,
      "num_tokens": 25977012.0,
      "step": 10245
    },
    {
      "entropy": 6.108498764038086,
      "epoch": 1.182919792267744,
      "grad_norm": 1.0,
      "learning_rate": 0.0004871761664302356,
      "loss": 6.0236,
      "mean_token_accuracy": 0.17297681719064711,
      "num_tokens": 25991520.0,
      "step": 10250
    },
    {
      "entropy": 6.063127231597901,
      "epoch": 1.1834968263127525,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004871624309920397,
      "loss": 5.9156,
      "mean_token_accuracy": 0.17331574410200118,
      "num_tokens": 26004644.0,
      "step": 10255
    },
    {
      "entropy": 6.124304628372192,
      "epoch": 1.1840738603577612,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048714868841786586,
      "loss": 5.9488,
      "mean_token_accuracy": 0.17636215090751647,
      "num_tokens": 26016452.0,
      "step": 10260
    },
    {
      "entropy": 6.144187068939209,
      "epoch": 1.1846508944027698,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048713493870817626,
      "loss": 5.9595,
      "mean_token_accuracy": 0.17459812313318251,
      "num_tokens": 26028675.0,
      "step": 10265
    },
    {
      "entropy": 6.139688205718994,
      "epoch": 1.1852279284477785,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048712118186343336,
      "loss": 6.0275,
      "mean_token_accuracy": 0.16822566837072372,
      "num_tokens": 26041561.0,
      "step": 10270
    },
    {
      "entropy": 5.987717628479004,
      "epoch": 1.185804962492787,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048710741788409985,
      "loss": 5.8679,
      "mean_token_accuracy": 0.17252693474292755,
      "num_tokens": 26052808.0,
      "step": 10275
    },
    {
      "entropy": 6.136587142944336,
      "epoch": 1.1863819965377957,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004870936467706387,
      "loss": 5.9914,
      "mean_token_accuracy": 0.16863525658845901,
      "num_tokens": 26065624.0,
      "step": 10280
    },
    {
      "entropy": 6.068053150177002,
      "epoch": 1.1869590305828044,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004870798685235131,
      "loss": 5.9151,
      "mean_token_accuracy": 0.1792290285229683,
      "num_tokens": 26078075.0,
      "step": 10285
    },
    {
      "entropy": 6.133761358261109,
      "epoch": 1.187536064627813,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048706608314318654,
      "loss": 6.0436,
      "mean_token_accuracy": 0.17365920841693877,
      "num_tokens": 26090454.0,
      "step": 10290
    },
    {
      "entropy": 6.058785724639892,
      "epoch": 1.1881130986728217,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004870522906301225,
      "loss": 5.9187,
      "mean_token_accuracy": 0.17744332998991014,
      "num_tokens": 26103516.0,
      "step": 10295
    },
    {
      "entropy": 6.136025285720825,
      "epoch": 1.1886901327178303,
      "grad_norm": 0.921875,
      "learning_rate": 0.000487038490984785,
      "loss": 6.0242,
      "mean_token_accuracy": 0.1620248571038246,
      "num_tokens": 26116420.0,
      "step": 10300
    },
    {
      "entropy": 6.0921101570129395,
      "epoch": 1.189267166762839,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048702468420763826,
      "loss": 5.985,
      "mean_token_accuracy": 0.1730554595589638,
      "num_tokens": 26129685.0,
      "step": 10305
    },
    {
      "entropy": 6.199032735824585,
      "epoch": 1.1898442008078476,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048701087029914657,
      "loss": 6.0,
      "mean_token_accuracy": 0.1692521944642067,
      "num_tokens": 26142909.0,
      "step": 10310
    },
    {
      "entropy": 6.0937087059021,
      "epoch": 1.1904212348528562,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004869970492597745,
      "loss": 5.9901,
      "mean_token_accuracy": 0.1722691535949707,
      "num_tokens": 26155229.0,
      "step": 10315
    },
    {
      "entropy": 6.122260856628418,
      "epoch": 1.1909982688978649,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004869832210899871,
      "loss": 5.9998,
      "mean_token_accuracy": 0.17114177495241165,
      "num_tokens": 26168309.0,
      "step": 10320
    },
    {
      "entropy": 6.141047525405884,
      "epoch": 1.1915753029428737,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048696938579024927,
      "loss": 5.9736,
      "mean_token_accuracy": 0.1769928365945816,
      "num_tokens": 26180189.0,
      "step": 10325
    },
    {
      "entropy": 6.049961090087891,
      "epoch": 1.1921523369878824,
      "grad_norm": 1.0,
      "learning_rate": 0.00048695554336102644,
      "loss": 5.9434,
      "mean_token_accuracy": 0.17389908879995347,
      "num_tokens": 26193125.0,
      "step": 10330
    },
    {
      "entropy": 6.137027072906494,
      "epoch": 1.192729371032891,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048694169380278417,
      "loss": 6.0075,
      "mean_token_accuracy": 0.16961179822683334,
      "num_tokens": 26206179.0,
      "step": 10335
    },
    {
      "entropy": 6.163803815841675,
      "epoch": 1.1933064050778996,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004869278371159884,
      "loss": 5.9484,
      "mean_token_accuracy": 0.17317136973142624,
      "num_tokens": 26217887.0,
      "step": 10340
    },
    {
      "entropy": 6.135565567016601,
      "epoch": 1.1938834391229083,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048691397330110503,
      "loss": 5.9556,
      "mean_token_accuracy": 0.1751647710800171,
      "num_tokens": 26231280.0,
      "step": 10345
    },
    {
      "entropy": 6.066518020629883,
      "epoch": 1.194460473167917,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004869001023586005,
      "loss": 5.8433,
      "mean_token_accuracy": 0.18549991101026536,
      "num_tokens": 26245565.0,
      "step": 10350
    },
    {
      "entropy": 6.071656084060669,
      "epoch": 1.1950375072129256,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004868862242889413,
      "loss": 5.9441,
      "mean_token_accuracy": 0.17392434030771256,
      "num_tokens": 26257313.0,
      "step": 10355
    },
    {
      "entropy": 6.098978757858276,
      "epoch": 1.1956145412579342,
      "grad_norm": 0.921875,
      "learning_rate": 0.000486872339092594,
      "loss": 5.9322,
      "mean_token_accuracy": 0.17885496765375136,
      "num_tokens": 26270566.0,
      "step": 10360
    },
    {
      "entropy": 6.141296482086181,
      "epoch": 1.1961915753029428,
      "grad_norm": 1.0,
      "learning_rate": 0.000486858446770026,
      "loss": 6.0603,
      "mean_token_accuracy": 0.16620281785726548,
      "num_tokens": 26283490.0,
      "step": 10365
    },
    {
      "entropy": 6.165494918823242,
      "epoch": 1.1967686093479515,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004868445473217043,
      "loss": 6.0029,
      "mean_token_accuracy": 0.167890702188015,
      "num_tokens": 26297702.0,
      "step": 10370
    },
    {
      "entropy": 6.111449718475342,
      "epoch": 1.1973456433929601,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004868306407480965,
      "loss": 5.9253,
      "mean_token_accuracy": 0.1748066872358322,
      "num_tokens": 26309543.0,
      "step": 10375
    },
    {
      "entropy": 6.156530475616455,
      "epoch": 1.1979226774379688,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048681672704967036,
      "loss": 6.0665,
      "mean_token_accuracy": 0.16524184942245485,
      "num_tokens": 26322259.0,
      "step": 10380
    },
    {
      "entropy": 6.063548278808594,
      "epoch": 1.1984997114829774,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004868028062268938,
      "loss": 5.9465,
      "mean_token_accuracy": 0.1727142333984375,
      "num_tokens": 26334985.0,
      "step": 10385
    },
    {
      "entropy": 5.987543725967408,
      "epoch": 1.1990767455279863,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048678887828023504,
      "loss": 5.7952,
      "mean_token_accuracy": 0.18351368308067323,
      "num_tokens": 26348441.0,
      "step": 10390
    },
    {
      "entropy": 6.138420295715332,
      "epoch": 1.1996537795729947,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004867749432101626,
      "loss": 5.9963,
      "mean_token_accuracy": 0.16941446363925933,
      "num_tokens": 26360848.0,
      "step": 10395
    },
    {
      "entropy": 6.087491035461426,
      "epoch": 1.2002308136180035,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004867610010171451,
      "loss": 5.9097,
      "mean_token_accuracy": 0.18984321504831314,
      "num_tokens": 26373704.0,
      "step": 10400
    },
    {
      "entropy": 6.091511058807373,
      "epoch": 1.2008078476630122,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048674705170165147,
      "loss": 5.9637,
      "mean_token_accuracy": 0.17630596160888673,
      "num_tokens": 26387035.0,
      "step": 10405
    },
    {
      "entropy": 6.074842119216919,
      "epoch": 1.2013848817080208,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000486733095264151,
      "loss": 5.8621,
      "mean_token_accuracy": 0.18056329041719438,
      "num_tokens": 26399064.0,
      "step": 10410
    },
    {
      "entropy": 6.083167934417725,
      "epoch": 1.2019619157530295,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048671913170511306,
      "loss": 5.9939,
      "mean_token_accuracy": 0.1720001295208931,
      "num_tokens": 26411930.0,
      "step": 10415
    },
    {
      "entropy": 6.046968412399292,
      "epoch": 1.202538949798038,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004867051610250073,
      "loss": 5.9365,
      "mean_token_accuracy": 0.16810204237699508,
      "num_tokens": 26425855.0,
      "step": 10420
    },
    {
      "entropy": 6.147312355041504,
      "epoch": 1.2031159838430467,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004866911832243035,
      "loss": 5.9977,
      "mean_token_accuracy": 0.1728080317378044,
      "num_tokens": 26437649.0,
      "step": 10425
    },
    {
      "entropy": 6.127640342712402,
      "epoch": 1.2036930178880554,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048667719830347203,
      "loss": 5.926,
      "mean_token_accuracy": 0.1685586541891098,
      "num_tokens": 26449503.0,
      "step": 10430
    },
    {
      "entropy": 6.106330299377442,
      "epoch": 1.204270051933064,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048666320626298307,
      "loss": 5.9622,
      "mean_token_accuracy": 0.17703323662281037,
      "num_tokens": 26462199.0,
      "step": 10435
    },
    {
      "entropy": 6.090756559371949,
      "epoch": 1.2048470859780727,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048664920710330735,
      "loss": 5.9878,
      "mean_token_accuracy": 0.17157117277383804,
      "num_tokens": 26475295.0,
      "step": 10440
    },
    {
      "entropy": 6.162950325012207,
      "epoch": 1.2054241200230813,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048663520082491564,
      "loss": 5.9743,
      "mean_token_accuracy": 0.1652437463402748,
      "num_tokens": 26488846.0,
      "step": 10445
    },
    {
      "entropy": 6.170254945755005,
      "epoch": 1.20600115406809,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004866211874282791,
      "loss": 6.0101,
      "mean_token_accuracy": 0.17079301923513412,
      "num_tokens": 26501645.0,
      "step": 10450
    },
    {
      "entropy": 6.095542669296265,
      "epoch": 1.2065781881130986,
      "grad_norm": 1.0,
      "learning_rate": 0.000486607166913869,
      "loss": 5.947,
      "mean_token_accuracy": 0.17275329679250717,
      "num_tokens": 26513494.0,
      "step": 10455
    },
    {
      "entropy": 6.156787729263305,
      "epoch": 1.2071552221581072,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048659313928215705,
      "loss": 5.9533,
      "mean_token_accuracy": 0.1752861961722374,
      "num_tokens": 26525931.0,
      "step": 10460
    },
    {
      "entropy": 6.097639036178589,
      "epoch": 1.207732256203116,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004865791045336149,
      "loss": 6.0178,
      "mean_token_accuracy": 0.16715566515922547,
      "num_tokens": 26539445.0,
      "step": 10465
    },
    {
      "entropy": 6.14093713760376,
      "epoch": 1.2083092902481247,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004865650626687146,
      "loss": 6.016,
      "mean_token_accuracy": 0.168517829477787,
      "num_tokens": 26551758.0,
      "step": 10470
    },
    {
      "entropy": 6.1033727645874025,
      "epoch": 1.2088863242931334,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048655101368792866,
      "loss": 5.9637,
      "mean_token_accuracy": 0.17112676054239273,
      "num_tokens": 26564756.0,
      "step": 10475
    },
    {
      "entropy": 6.064315557479858,
      "epoch": 1.209463358338142,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004865369575917293,
      "loss": 5.9252,
      "mean_token_accuracy": 0.1811230793595314,
      "num_tokens": 26579152.0,
      "step": 10480
    },
    {
      "entropy": 6.086688470840454,
      "epoch": 1.2100403923831506,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048652289438058953,
      "loss": 5.942,
      "mean_token_accuracy": 0.17501748204231263,
      "num_tokens": 26590749.0,
      "step": 10485
    },
    {
      "entropy": 6.214601373672485,
      "epoch": 1.2106174264281593,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048650882405498226,
      "loss": 6.0472,
      "mean_token_accuracy": 0.16946351975202562,
      "num_tokens": 26603859.0,
      "step": 10490
    },
    {
      "entropy": 6.024354410171509,
      "epoch": 1.211194460473168,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004864947466153808,
      "loss": 5.8736,
      "mean_token_accuracy": 0.18842062205076218,
      "num_tokens": 26616002.0,
      "step": 10495
    },
    {
      "entropy": 6.0380940437316895,
      "epoch": 1.2117714945181766,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004864806620622585,
      "loss": 6.007,
      "mean_token_accuracy": 0.17046746015548705,
      "num_tokens": 26629092.0,
      "step": 10500
    },
    {
      "entropy": 6.165492868423462,
      "epoch": 1.2123485285631852,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004864665703960892,
      "loss": 5.9751,
      "mean_token_accuracy": 0.1697184905409813,
      "num_tokens": 26642830.0,
      "step": 10505
    },
    {
      "entropy": 6.1605650901794435,
      "epoch": 1.2129255626081938,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004864524716173469,
      "loss": 6.1361,
      "mean_token_accuracy": 0.15956175401806832,
      "num_tokens": 26656421.0,
      "step": 10510
    },
    {
      "entropy": 6.006120395660401,
      "epoch": 1.2135025966532025,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004864383657265058,
      "loss": 5.8994,
      "mean_token_accuracy": 0.17321467101573945,
      "num_tokens": 26669267.0,
      "step": 10515
    },
    {
      "entropy": 6.1963605880737305,
      "epoch": 1.2140796306982111,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048642425272404016,
      "loss": 5.9407,
      "mean_token_accuracy": 0.17217683047056198,
      "num_tokens": 26681074.0,
      "step": 10520
    },
    {
      "entropy": 6.063516855239868,
      "epoch": 1.2146566647432198,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004864101326104248,
      "loss": 5.9181,
      "mean_token_accuracy": 0.1792256161570549,
      "num_tokens": 26694772.0,
      "step": 10525
    },
    {
      "entropy": 6.084238433837891,
      "epoch": 1.2152336987882286,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004863960053861348,
      "loss": 5.9959,
      "mean_token_accuracy": 0.17432879358530046,
      "num_tokens": 26706173.0,
      "step": 10530
    },
    {
      "entropy": 6.147999382019043,
      "epoch": 1.215810732833237,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048638187105164507,
      "loss": 6.0511,
      "mean_token_accuracy": 0.16493165045976638,
      "num_tokens": 26718273.0,
      "step": 10535
    },
    {
      "entropy": 6.15310378074646,
      "epoch": 1.216387766878246,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004863677296074311,
      "loss": 6.0333,
      "mean_token_accuracy": 0.17173090875148772,
      "num_tokens": 26730781.0,
      "step": 10540
    },
    {
      "entropy": 6.057547950744629,
      "epoch": 1.2169648009232545,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004863535810539686,
      "loss": 5.9854,
      "mean_token_accuracy": 0.17228724509477616,
      "num_tokens": 26743661.0,
      "step": 10545
    },
    {
      "entropy": 6.073516607284546,
      "epoch": 1.2175418349682632,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048633942539173325,
      "loss": 5.9237,
      "mean_token_accuracy": 0.1753203898668289,
      "num_tokens": 26755307.0,
      "step": 10550
    },
    {
      "entropy": 6.090433645248413,
      "epoch": 1.2181188690132718,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048632526262120144,
      "loss": 5.9542,
      "mean_token_accuracy": 0.17189480364322662,
      "num_tokens": 26769679.0,
      "step": 10555
    },
    {
      "entropy": 6.07599401473999,
      "epoch": 1.2186959030582805,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004863110927428493,
      "loss": 5.8726,
      "mean_token_accuracy": 0.17687562555074693,
      "num_tokens": 26782141.0,
      "step": 10560
    },
    {
      "entropy": 6.115686464309692,
      "epoch": 1.219272937103289,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004862969157571536,
      "loss": 6.0046,
      "mean_token_accuracy": 0.17116649746894835,
      "num_tokens": 26794721.0,
      "step": 10565
    },
    {
      "entropy": 6.07307391166687,
      "epoch": 1.2198499711482977,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048628273166459105,
      "loss": 5.934,
      "mean_token_accuracy": 0.17897441387176513,
      "num_tokens": 26807844.0,
      "step": 10570
    },
    {
      "entropy": 6.138726711273193,
      "epoch": 1.2204270051933064,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048626854046563876,
      "loss": 5.994,
      "mean_token_accuracy": 0.1666283279657364,
      "num_tokens": 26820150.0,
      "step": 10575
    },
    {
      "entropy": 6.196956777572632,
      "epoch": 1.221004039238315,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048625434216077404,
      "loss": 6.015,
      "mean_token_accuracy": 0.16749742478132248,
      "num_tokens": 26833949.0,
      "step": 10580
    },
    {
      "entropy": 5.988670444488525,
      "epoch": 1.2215810732833237,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048624013675047453,
      "loss": 5.8821,
      "mean_token_accuracy": 0.17497682869434356,
      "num_tokens": 26847637.0,
      "step": 10585
    },
    {
      "entropy": 6.1034351825714115,
      "epoch": 1.2221581073283323,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048622592423521783,
      "loss": 5.9848,
      "mean_token_accuracy": 0.17601878345012664,
      "num_tokens": 26859173.0,
      "step": 10590
    },
    {
      "entropy": 6.108186292648315,
      "epoch": 1.222735141373341,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004862117046154822,
      "loss": 5.9838,
      "mean_token_accuracy": 0.17103430926799773,
      "num_tokens": 26870883.0,
      "step": 10595
    },
    {
      "entropy": 6.188717794418335,
      "epoch": 1.2233121754183496,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048619747789174577,
      "loss": 6.0168,
      "mean_token_accuracy": 0.1715303599834442,
      "num_tokens": 26882813.0,
      "step": 10600
    },
    {
      "entropy": 6.030700588226319,
      "epoch": 1.2238892094633584,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004861832440644871,
      "loss": 5.9086,
      "mean_token_accuracy": 0.17621286511421203,
      "num_tokens": 26893928.0,
      "step": 10605
    },
    {
      "entropy": 6.009361791610718,
      "epoch": 1.224466243508367,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048616900313418486,
      "loss": 5.9643,
      "mean_token_accuracy": 0.17730257511138917,
      "num_tokens": 26907431.0,
      "step": 10610
    },
    {
      "entropy": 6.194645690917969,
      "epoch": 1.2250432775533757,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048615475510131815,
      "loss": 6.0127,
      "mean_token_accuracy": 0.16326101124286652,
      "num_tokens": 26920278.0,
      "step": 10615
    },
    {
      "entropy": 6.1857822895050045,
      "epoch": 1.2256203115983844,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048614049996636614,
      "loss": 6.0771,
      "mean_token_accuracy": 0.16853276193141936,
      "num_tokens": 26933348.0,
      "step": 10620
    },
    {
      "entropy": 6.082099151611328,
      "epoch": 1.226197345643393,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048612623772980825,
      "loss": 6.0235,
      "mean_token_accuracy": 0.1712267816066742,
      "num_tokens": 26946017.0,
      "step": 10625
    },
    {
      "entropy": 6.043529415130616,
      "epoch": 1.2267743796884016,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048611196839212426,
      "loss": 5.8141,
      "mean_token_accuracy": 0.18340473622083664,
      "num_tokens": 26960230.0,
      "step": 10630
    },
    {
      "entropy": 6.145419549942017,
      "epoch": 1.2273514137334103,
      "grad_norm": 0.984375,
      "learning_rate": 0.000486097691953794,
      "loss": 6.0243,
      "mean_token_accuracy": 0.16695577949285506,
      "num_tokens": 26971811.0,
      "step": 10635
    },
    {
      "entropy": 6.124272203445434,
      "epoch": 1.227928447778419,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048608340841529784,
      "loss": 5.989,
      "mean_token_accuracy": 0.1742856413125992,
      "num_tokens": 26983098.0,
      "step": 10640
    },
    {
      "entropy": 6.068958711624146,
      "epoch": 1.2285054818234276,
      "grad_norm": 1.03125,
      "learning_rate": 0.000486069117777116,
      "loss": 6.0196,
      "mean_token_accuracy": 0.1602456346154213,
      "num_tokens": 26996410.0,
      "step": 10645
    },
    {
      "entropy": 6.227188587188721,
      "epoch": 1.2290825158684362,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004860548200397293,
      "loss": 5.9999,
      "mean_token_accuracy": 0.16977795958518982,
      "num_tokens": 27009329.0,
      "step": 10650
    },
    {
      "entropy": 6.07104811668396,
      "epoch": 1.2296595499134448,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048604051520361846,
      "loss": 5.9469,
      "mean_token_accuracy": 0.17188530862331391,
      "num_tokens": 27022261.0,
      "step": 10655
    },
    {
      "entropy": 6.119796419143677,
      "epoch": 1.2302365839584535,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048602620326926484,
      "loss": 6.0257,
      "mean_token_accuracy": 0.1737441673874855,
      "num_tokens": 27035006.0,
      "step": 10660
    },
    {
      "entropy": 6.093335390090942,
      "epoch": 1.2308136180034621,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048601188423714965,
      "loss": 5.9034,
      "mean_token_accuracy": 0.17869824469089507,
      "num_tokens": 27048481.0,
      "step": 10665
    },
    {
      "entropy": 6.171976137161255,
      "epoch": 1.231390652048471,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004859975581077545,
      "loss": 6.0815,
      "mean_token_accuracy": 0.16199043095111848,
      "num_tokens": 27061083.0,
      "step": 10670
    },
    {
      "entropy": 6.175375986099243,
      "epoch": 1.2319676860934794,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004859832248815614,
      "loss": 5.9926,
      "mean_token_accuracy": 0.16824596226215363,
      "num_tokens": 27073547.0,
      "step": 10675
    },
    {
      "entropy": 6.076800107955933,
      "epoch": 1.2325447201384883,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004859688845590522,
      "loss": 5.9731,
      "mean_token_accuracy": 0.17331507354974746,
      "num_tokens": 27085739.0,
      "step": 10680
    },
    {
      "entropy": 6.041712474822998,
      "epoch": 1.233121754183497,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00048595453714070944,
      "loss": 5.9714,
      "mean_token_accuracy": 0.17364395558834075,
      "num_tokens": 27097331.0,
      "step": 10685
    },
    {
      "entropy": 6.197361183166504,
      "epoch": 1.2336987882285055,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004859401826270156,
      "loss": 5.9975,
      "mean_token_accuracy": 0.1749812513589859,
      "num_tokens": 27109181.0,
      "step": 10690
    },
    {
      "entropy": 6.0757262229919435,
      "epoch": 1.2342758222735142,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004859258210184536,
      "loss": 5.8947,
      "mean_token_accuracy": 0.1778571218252182,
      "num_tokens": 27120978.0,
      "step": 10695
    },
    {
      "entropy": 6.1293861865997314,
      "epoch": 1.2348528563185228,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048591145231550623,
      "loss": 5.9443,
      "mean_token_accuracy": 0.17495268434286118,
      "num_tokens": 27135243.0,
      "step": 10700
    },
    {
      "entropy": 5.980443143844605,
      "epoch": 1.2354298903635315,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048589707651865697,
      "loss": 5.8813,
      "mean_token_accuracy": 0.18009669929742814,
      "num_tokens": 27149366.0,
      "step": 10705
    },
    {
      "entropy": 6.084422159194946,
      "epoch": 1.23600692440854,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004858826936283893,
      "loss": 5.9886,
      "mean_token_accuracy": 0.16944239139556885,
      "num_tokens": 27162552.0,
      "step": 10710
    },
    {
      "entropy": 6.136016225814819,
      "epoch": 1.2365839584535487,
      "grad_norm": 0.9375,
      "learning_rate": 0.000485868303645187,
      "loss": 5.9622,
      "mean_token_accuracy": 0.17089103609323503,
      "num_tokens": 27175312.0,
      "step": 10715
    },
    {
      "entropy": 6.081205034255982,
      "epoch": 1.2371609924985574,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048585390656953397,
      "loss": 5.9417,
      "mean_token_accuracy": 0.16926106065511703,
      "num_tokens": 27187730.0,
      "step": 10720
    },
    {
      "entropy": 6.088982057571411,
      "epoch": 1.237738026543566,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004858395024019147,
      "loss": 5.9405,
      "mean_token_accuracy": 0.17881813645362854,
      "num_tokens": 27200191.0,
      "step": 10725
    },
    {
      "entropy": 6.11422963142395,
      "epoch": 1.2383150605885747,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004858250911428133,
      "loss": 5.9824,
      "mean_token_accuracy": 0.17729408144950867,
      "num_tokens": 27211905.0,
      "step": 10730
    },
    {
      "entropy": 6.140657663345337,
      "epoch": 1.2388920946335833,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004858106727927148,
      "loss": 5.9549,
      "mean_token_accuracy": 0.1747775673866272,
      "num_tokens": 27224153.0,
      "step": 10735
    },
    {
      "entropy": 6.108855867385865,
      "epoch": 1.239469128678592,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000485796247352104,
      "loss": 5.9685,
      "mean_token_accuracy": 0.16635477542877197,
      "num_tokens": 27237167.0,
      "step": 10740
    },
    {
      "entropy": 6.093992424011231,
      "epoch": 1.2400461627236008,
      "grad_norm": 1.0,
      "learning_rate": 0.0004857818148214662,
      "loss": 5.9883,
      "mean_token_accuracy": 0.17342361062765121,
      "num_tokens": 27247956.0,
      "step": 10745
    },
    {
      "entropy": 6.072857332229614,
      "epoch": 1.2406231967686094,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004857673752012866,
      "loss": 5.989,
      "mean_token_accuracy": 0.16709066033363343,
      "num_tokens": 27260357.0,
      "step": 10750
    },
    {
      "entropy": 6.104099941253662,
      "epoch": 1.241200230813618,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048575292849205114,
      "loss": 5.9179,
      "mean_token_accuracy": 0.16976598501205445,
      "num_tokens": 27273953.0,
      "step": 10755
    },
    {
      "entropy": 6.091964960098267,
      "epoch": 1.2417772648586267,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004857384746942456,
      "loss": 5.9673,
      "mean_token_accuracy": 0.17539182007312776,
      "num_tokens": 27285550.0,
      "step": 10760
    },
    {
      "entropy": 6.131041049957275,
      "epoch": 1.2423542989036354,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004857240138083562,
      "loss": 6.0342,
      "mean_token_accuracy": 0.1663442313671112,
      "num_tokens": 27297817.0,
      "step": 10765
    },
    {
      "entropy": 6.101998138427734,
      "epoch": 1.242931332948644,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004857095458348692,
      "loss": 6.0681,
      "mean_token_accuracy": 0.16690385192632676,
      "num_tokens": 27310398.0,
      "step": 10770
    },
    {
      "entropy": 6.161550569534302,
      "epoch": 1.2435083669936526,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048569507077427134,
      "loss": 6.0153,
      "mean_token_accuracy": 0.16498113572597503,
      "num_tokens": 27322462.0,
      "step": 10775
    },
    {
      "entropy": 6.240576934814453,
      "epoch": 1.2440854010386613,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004856805886270494,
      "loss": 6.029,
      "mean_token_accuracy": 0.16795228272676468,
      "num_tokens": 27335301.0,
      "step": 10780
    },
    {
      "entropy": 6.058903932571411,
      "epoch": 1.24466243508367,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004856660993936905,
      "loss": 5.9021,
      "mean_token_accuracy": 0.1822773978114128,
      "num_tokens": 27347425.0,
      "step": 10785
    },
    {
      "entropy": 6.121625185012817,
      "epoch": 1.2452394691286786,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000485651603074682,
      "loss": 6.0041,
      "mean_token_accuracy": 0.17138130217790604,
      "num_tokens": 27361149.0,
      "step": 10790
    },
    {
      "entropy": 6.089787435531616,
      "epoch": 1.2458165031736872,
      "grad_norm": 1.0,
      "learning_rate": 0.0004856370996705115,
      "loss": 5.9743,
      "mean_token_accuracy": 0.17579420804977416,
      "num_tokens": 27373055.0,
      "step": 10795
    },
    {
      "entropy": 6.083675527572632,
      "epoch": 1.2463935372186958,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004856225891816667,
      "loss": 6.0204,
      "mean_token_accuracy": 0.17585015445947647,
      "num_tokens": 27385279.0,
      "step": 10800
    },
    {
      "entropy": 6.152761697769165,
      "epoch": 1.2469705712637045,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004856080716086358,
      "loss": 6.0186,
      "mean_token_accuracy": 0.17063578367233276,
      "num_tokens": 27399035.0,
      "step": 10805
    },
    {
      "entropy": 6.150375509262085,
      "epoch": 1.2475476053087133,
      "grad_norm": 1.046875,
      "learning_rate": 0.000485593546951907,
      "loss": 6.002,
      "mean_token_accuracy": 0.17371326833963394,
      "num_tokens": 27410473.0,
      "step": 10810
    },
    {
      "entropy": 6.136099052429199,
      "epoch": 1.2481246393537218,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004855790152119688,
      "loss": 5.9527,
      "mean_token_accuracy": 0.17234997749328612,
      "num_tokens": 27422194.0,
      "step": 10815
    },
    {
      "entropy": 6.054213380813598,
      "epoch": 1.2487016733987306,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004855644763893102,
      "loss": 5.9846,
      "mean_token_accuracy": 0.17384154945611954,
      "num_tokens": 27434800.0,
      "step": 10820
    },
    {
      "entropy": 6.022722816467285,
      "epoch": 1.2492787074437393,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048554993048441987,
      "loss": 5.8581,
      "mean_token_accuracy": 0.17802471965551375,
      "num_tokens": 27447002.0,
      "step": 10825
    },
    {
      "entropy": 6.186601305007935,
      "epoch": 1.249855741488748,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004855353774977873,
      "loss": 6.0513,
      "mean_token_accuracy": 0.16859824508428572,
      "num_tokens": 27461461.0,
      "step": 10830
    },
    {
      "entropy": 6.12995491027832,
      "epoch": 1.2504327755337565,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048552081742990184,
      "loss": 6.0576,
      "mean_token_accuracy": 0.15986740216612816,
      "num_tokens": 27475062.0,
      "step": 10835
    },
    {
      "entropy": 6.108346271514892,
      "epoch": 1.2510098095787652,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048550625028125327,
      "loss": 5.9106,
      "mean_token_accuracy": 0.1790057376027107,
      "num_tokens": 27487812.0,
      "step": 10840
    },
    {
      "entropy": 6.062467813491821,
      "epoch": 1.2515868436237738,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004854916760523315,
      "loss": 5.9577,
      "mean_token_accuracy": 0.17829149961471558,
      "num_tokens": 27501351.0,
      "step": 10845
    },
    {
      "entropy": 6.118744039535523,
      "epoch": 1.2521638776687825,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048547709474362684,
      "loss": 5.9284,
      "mean_token_accuracy": 0.17315473556518554,
      "num_tokens": 27514651.0,
      "step": 10850
    },
    {
      "entropy": 6.165411710739136,
      "epoch": 1.252740911713791,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004854625063556296,
      "loss": 6.0572,
      "mean_token_accuracy": 0.16677621155977249,
      "num_tokens": 27527037.0,
      "step": 10855
    },
    {
      "entropy": 6.214095735549927,
      "epoch": 1.2533179457587997,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004854479108888305,
      "loss": 6.0343,
      "mean_token_accuracy": 0.16586532071232796,
      "num_tokens": 27540800.0,
      "step": 10860
    },
    {
      "entropy": 6.1121422290802006,
      "epoch": 1.2538949798038084,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048543330834372047,
      "loss": 5.9428,
      "mean_token_accuracy": 0.17423375993967055,
      "num_tokens": 27553873.0,
      "step": 10865
    },
    {
      "entropy": 6.0320068359375,
      "epoch": 1.254472013848817,
      "grad_norm": 1.0,
      "learning_rate": 0.00048541869872079064,
      "loss": 5.9914,
      "mean_token_accuracy": 0.17038509547710418,
      "num_tokens": 27566336.0,
      "step": 10870
    },
    {
      "entropy": 6.173742961883545,
      "epoch": 1.2550490478938257,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004854040820205324,
      "loss": 6.0707,
      "mean_token_accuracy": 0.16543821692466737,
      "num_tokens": 27578677.0,
      "step": 10875
    },
    {
      "entropy": 6.198507785797119,
      "epoch": 1.2556260819388343,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004853894582434373,
      "loss": 6.0411,
      "mean_token_accuracy": 0.17079235911369323,
      "num_tokens": 27590734.0,
      "step": 10880
    },
    {
      "entropy": 6.0929481983184814,
      "epoch": 1.2562031159838432,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004853748273899974,
      "loss": 5.9616,
      "mean_token_accuracy": 0.1683901444077492,
      "num_tokens": 27603193.0,
      "step": 10885
    },
    {
      "entropy": 6.046877145767212,
      "epoch": 1.2567801500288516,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004853601894607046,
      "loss": 5.9329,
      "mean_token_accuracy": 0.17547256648540496,
      "num_tokens": 27615248.0,
      "step": 10890
    },
    {
      "entropy": 6.166851282119751,
      "epoch": 1.2573571840738604,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004853455444560514,
      "loss": 6.0319,
      "mean_token_accuracy": 0.17462890446186066,
      "num_tokens": 27627384.0,
      "step": 10895
    },
    {
      "entropy": 6.131621313095093,
      "epoch": 1.257934218118869,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004853308923765302,
      "loss": 6.0354,
      "mean_token_accuracy": 0.16517208963632585,
      "num_tokens": 27641014.0,
      "step": 10900
    },
    {
      "entropy": 6.13031153678894,
      "epoch": 1.2585112521638777,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000485316233222634,
      "loss": 5.8822,
      "mean_token_accuracy": 0.17947525829076766,
      "num_tokens": 27652325.0,
      "step": 10905
    },
    {
      "entropy": 6.081737518310547,
      "epoch": 1.2590882862088864,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004853015669948557,
      "loss": 5.9786,
      "mean_token_accuracy": 0.17024474143981932,
      "num_tokens": 27664365.0,
      "step": 10910
    },
    {
      "entropy": 6.068706035614014,
      "epoch": 1.259665320253895,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048528689369368863,
      "loss": 5.9032,
      "mean_token_accuracy": 0.17318783402442933,
      "num_tokens": 27677657.0,
      "step": 10915
    },
    {
      "entropy": 6.083243942260742,
      "epoch": 1.2602423542989036,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004852722133196264,
      "loss": 5.8337,
      "mean_token_accuracy": 0.18861999064683915,
      "num_tokens": 27689598.0,
      "step": 10920
    },
    {
      "entropy": 6.042051792144775,
      "epoch": 1.2608193883439123,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004852575258731627,
      "loss": 5.8701,
      "mean_token_accuracy": 0.18486067056655883,
      "num_tokens": 27702573.0,
      "step": 10925
    },
    {
      "entropy": 6.080828332901001,
      "epoch": 1.261396422388921,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004852428313547916,
      "loss": 5.9896,
      "mean_token_accuracy": 0.17420790791511537,
      "num_tokens": 27714506.0,
      "step": 10930
    },
    {
      "entropy": 5.950251865386963,
      "epoch": 1.2619734564339296,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048522812976500726,
      "loss": 5.6788,
      "mean_token_accuracy": 0.2015139251947403,
      "num_tokens": 27728308.0,
      "step": 10935
    },
    {
      "entropy": 6.0726526260375975,
      "epoch": 1.2625504904789382,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048521342110430417,
      "loss": 5.9849,
      "mean_token_accuracy": 0.17030115723609923,
      "num_tokens": 27741521.0,
      "step": 10940
    },
    {
      "entropy": 6.171085453033447,
      "epoch": 1.2631275245239468,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048519870537317713,
      "loss": 6.0093,
      "mean_token_accuracy": 0.16484422534704207,
      "num_tokens": 27754033.0,
      "step": 10945
    },
    {
      "entropy": 6.161662244796753,
      "epoch": 1.2637045585689557,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048518398257212103,
      "loss": 6.0246,
      "mean_token_accuracy": 0.16985787451267242,
      "num_tokens": 27766557.0,
      "step": 10950
    },
    {
      "entropy": 6.089897155761719,
      "epoch": 1.2642815926139641,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004851692527016311,
      "loss": 5.9533,
      "mean_token_accuracy": 0.17847632318735124,
      "num_tokens": 27778018.0,
      "step": 10955
    },
    {
      "entropy": 6.050190782546997,
      "epoch": 1.264858626658973,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004851545157622027,
      "loss": 5.907,
      "mean_token_accuracy": 0.17910790741443633,
      "num_tokens": 27789514.0,
      "step": 10960
    },
    {
      "entropy": 6.123341512680054,
      "epoch": 1.2654356607039816,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004851397717543316,
      "loss": 5.9674,
      "mean_token_accuracy": 0.17060527801513672,
      "num_tokens": 27802759.0,
      "step": 10965
    },
    {
      "entropy": 6.063577795028687,
      "epoch": 1.2660126947489903,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004851250206785137,
      "loss": 5.9303,
      "mean_token_accuracy": 0.17530115693807602,
      "num_tokens": 27817064.0,
      "step": 10970
    },
    {
      "entropy": 6.141218137741089,
      "epoch": 1.266589728793999,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048511026253524503,
      "loss": 5.9708,
      "mean_token_accuracy": 0.16788488626480103,
      "num_tokens": 27830627.0,
      "step": 10975
    },
    {
      "entropy": 6.03924069404602,
      "epoch": 1.2671667628390075,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004850954973250221,
      "loss": 5.8493,
      "mean_token_accuracy": 0.18568919152021407,
      "num_tokens": 27842745.0,
      "step": 10980
    },
    {
      "entropy": 6.044141817092895,
      "epoch": 1.2677437968840162,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048508072504834144,
      "loss": 5.98,
      "mean_token_accuracy": 0.17580776810646057,
      "num_tokens": 27855720.0,
      "step": 10985
    },
    {
      "entropy": 6.091658973693848,
      "epoch": 1.2683208309290248,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048506594570569997,
      "loss": 5.9353,
      "mean_token_accuracy": 0.17962995618581773,
      "num_tokens": 27869019.0,
      "step": 10990
    },
    {
      "entropy": 6.147141075134277,
      "epoch": 1.2688978649740335,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004850511592975947,
      "loss": 6.0388,
      "mean_token_accuracy": 0.1666327863931656,
      "num_tokens": 27883563.0,
      "step": 10995
    },
    {
      "entropy": 6.02319188117981,
      "epoch": 1.269474899019042,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004850363658245231,
      "loss": 5.9481,
      "mean_token_accuracy": 0.17660248577594756,
      "num_tokens": 27895980.0,
      "step": 11000
    },
    {
      "entropy": 6.049231147766113,
      "epoch": 1.2700519330640507,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004850215652869826,
      "loss": 6.012,
      "mean_token_accuracy": 0.16709204614162446,
      "num_tokens": 27907535.0,
      "step": 11005
    },
    {
      "entropy": 6.219564485549927,
      "epoch": 1.2706289671090594,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004850067576854711,
      "loss": 5.9535,
      "mean_token_accuracy": 0.1725838750600815,
      "num_tokens": 27920665.0,
      "step": 11010
    },
    {
      "entropy": 6.025489759445191,
      "epoch": 1.271206001154068,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004849919430204867,
      "loss": 5.8718,
      "mean_token_accuracy": 0.17963292002677916,
      "num_tokens": 27932178.0,
      "step": 11015
    },
    {
      "entropy": 6.04595832824707,
      "epoch": 1.2717830351990767,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004849771212925275,
      "loss": 5.9167,
      "mean_token_accuracy": 0.17644584476947783,
      "num_tokens": 27945319.0,
      "step": 11020
    },
    {
      "entropy": 6.061441230773926,
      "epoch": 1.2723600692440855,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048496229250209223,
      "loss": 5.9243,
      "mean_token_accuracy": 0.1741519570350647,
      "num_tokens": 27957765.0,
      "step": 11025
    },
    {
      "entropy": 6.075029039382935,
      "epoch": 1.272937103289094,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004849474566496795,
      "loss": 5.9381,
      "mean_token_accuracy": 0.1798078417778015,
      "num_tokens": 27971306.0,
      "step": 11030
    },
    {
      "entropy": 6.056388664245605,
      "epoch": 1.2735141373341028,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004849326137357883,
      "loss": 5.8735,
      "mean_token_accuracy": 0.1768690049648285,
      "num_tokens": 27982998.0,
      "step": 11035
    },
    {
      "entropy": 6.063432359695435,
      "epoch": 1.2740911713791114,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000484917763760918,
      "loss": 5.9894,
      "mean_token_accuracy": 0.17349187135696412,
      "num_tokens": 27995792.0,
      "step": 11040
    },
    {
      "entropy": 5.992597770690918,
      "epoch": 1.27466820542412,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048490290672556784,
      "loss": 5.8631,
      "mean_token_accuracy": 0.1780084028840065,
      "num_tokens": 28008882.0,
      "step": 11045
    },
    {
      "entropy": 6.150570011138916,
      "epoch": 1.2752452394691287,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004848880426302378,
      "loss": 5.9962,
      "mean_token_accuracy": 0.17136159986257554,
      "num_tokens": 28022768.0,
      "step": 11050
    },
    {
      "entropy": 6.089174842834472,
      "epoch": 1.2758222735141374,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004848731714754277,
      "loss": 5.9229,
      "mean_token_accuracy": 0.17786931693553926,
      "num_tokens": 28035196.0,
      "step": 11055
    },
    {
      "entropy": 6.207886028289795,
      "epoch": 1.276399307559146,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004848582932616377,
      "loss": 6.1183,
      "mean_token_accuracy": 0.1631258398294449,
      "num_tokens": 28048089.0,
      "step": 11060
    },
    {
      "entropy": 6.16239972114563,
      "epoch": 1.2769763416041546,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004848434079893682,
      "loss": 5.9949,
      "mean_token_accuracy": 0.1683859869837761,
      "num_tokens": 28060140.0,
      "step": 11065
    },
    {
      "entropy": 6.128098630905152,
      "epoch": 1.2775533756491633,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004848285156591201,
      "loss": 5.9666,
      "mean_token_accuracy": 0.17058906108140945,
      "num_tokens": 28072448.0,
      "step": 11070
    },
    {
      "entropy": 6.111254453659058,
      "epoch": 1.278130409694172,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048481361627139384,
      "loss": 6.0565,
      "mean_token_accuracy": 0.1653997540473938,
      "num_tokens": 28084696.0,
      "step": 11075
    },
    {
      "entropy": 6.13771390914917,
      "epoch": 1.2787074437391805,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048479870982669096,
      "loss": 5.9901,
      "mean_token_accuracy": 0.17498163729906083,
      "num_tokens": 28097834.0,
      "step": 11080
    },
    {
      "entropy": 6.061259078979492,
      "epoch": 1.2792844777841892,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048478379632551266,
      "loss": 5.8859,
      "mean_token_accuracy": 0.18365999013185502,
      "num_tokens": 28110066.0,
      "step": 11085
    },
    {
      "entropy": 6.118547153472901,
      "epoch": 1.279861511829198,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004847688757683606,
      "loss": 5.9995,
      "mean_token_accuracy": 0.1729790076613426,
      "num_tokens": 28122617.0,
      "step": 11090
    },
    {
      "entropy": 6.073926830291748,
      "epoch": 1.2804385458742065,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004847539481557366,
      "loss": 5.8913,
      "mean_token_accuracy": 0.17460058778524398,
      "num_tokens": 28134987.0,
      "step": 11095
    },
    {
      "entropy": 6.113080263137817,
      "epoch": 1.2810155799192153,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048473901348814277,
      "loss": 5.9764,
      "mean_token_accuracy": 0.16729729622602463,
      "num_tokens": 28147478.0,
      "step": 11100
    },
    {
      "entropy": 6.115693616867065,
      "epoch": 1.281592613964224,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004847240717660814,
      "loss": 6.0103,
      "mean_token_accuracy": 0.17114701569080354,
      "num_tokens": 28159010.0,
      "step": 11105
    },
    {
      "entropy": 6.085699892044067,
      "epoch": 1.2821696480092326,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048470912299005493,
      "loss": 6.0182,
      "mean_token_accuracy": 0.1666300743818283,
      "num_tokens": 28170579.0,
      "step": 11110
    },
    {
      "entropy": 6.153729057312011,
      "epoch": 1.2827466820542412,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048469416716056635,
      "loss": 5.9545,
      "mean_token_accuracy": 0.17459358721971513,
      "num_tokens": 28182569.0,
      "step": 11115
    },
    {
      "entropy": 6.171100854873657,
      "epoch": 1.28332371609925,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004846792042781187,
      "loss": 6.038,
      "mean_token_accuracy": 0.16944910660386087,
      "num_tokens": 28195736.0,
      "step": 11120
    },
    {
      "entropy": 6.119990634918213,
      "epoch": 1.2839007501442585,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048466423434321513,
      "loss": 6.0174,
      "mean_token_accuracy": 0.17535480260848998,
      "num_tokens": 28210944.0,
      "step": 11125
    },
    {
      "entropy": 6.114975643157959,
      "epoch": 1.2844777841892672,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048464925735635907,
      "loss": 5.9543,
      "mean_token_accuracy": 0.1677554503083229,
      "num_tokens": 28223674.0,
      "step": 11130
    },
    {
      "entropy": 6.151212739944458,
      "epoch": 1.2850548182342758,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048463427331805445,
      "loss": 6.0385,
      "mean_token_accuracy": 0.16458675563335418,
      "num_tokens": 28236643.0,
      "step": 11135
    },
    {
      "entropy": 6.110540962219238,
      "epoch": 1.2856318522792844,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004846192822288052,
      "loss": 6.0161,
      "mean_token_accuracy": 0.165218223631382,
      "num_tokens": 28249678.0,
      "step": 11140
    },
    {
      "entropy": 6.07472333908081,
      "epoch": 1.286208886324293,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004846042840891155,
      "loss": 5.9557,
      "mean_token_accuracy": 0.17722685337066652,
      "num_tokens": 28262571.0,
      "step": 11145
    },
    {
      "entropy": 6.1123682975769045,
      "epoch": 1.2867859203693017,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004845892788994899,
      "loss": 5.9534,
      "mean_token_accuracy": 0.17424528002738954,
      "num_tokens": 28275450.0,
      "step": 11150
    },
    {
      "entropy": 6.122414827346802,
      "epoch": 1.2873629544143104,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004845742666604329,
      "loss": 5.9411,
      "mean_token_accuracy": 0.17285699695348739,
      "num_tokens": 28288224.0,
      "step": 11155
    },
    {
      "entropy": 5.938300704956054,
      "epoch": 1.287939988459319,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048455924737244953,
      "loss": 5.7517,
      "mean_token_accuracy": 0.18958992511034012,
      "num_tokens": 28300515.0,
      "step": 11160
    },
    {
      "entropy": 6.079554843902588,
      "epoch": 1.2885170225043279,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048454422103604505,
      "loss": 5.9853,
      "mean_token_accuracy": 0.1725201666355133,
      "num_tokens": 28313729.0,
      "step": 11165
    },
    {
      "entropy": 6.11998438835144,
      "epoch": 1.2890940565493363,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004845291876517247,
      "loss": 5.9138,
      "mean_token_accuracy": 0.17606235295534134,
      "num_tokens": 28325872.0,
      "step": 11170
    },
    {
      "entropy": 6.150653886795044,
      "epoch": 1.2896710905943451,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048451414721999424,
      "loss": 5.9568,
      "mean_token_accuracy": 0.17766788154840468,
      "num_tokens": 28337820.0,
      "step": 11175
    },
    {
      "entropy": 6.111855316162109,
      "epoch": 1.2902481246393538,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048449909974135954,
      "loss": 5.9912,
      "mean_token_accuracy": 0.1744311884045601,
      "num_tokens": 28350608.0,
      "step": 11180
    },
    {
      "entropy": 6.043523263931275,
      "epoch": 1.2908251586843624,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004844840452163267,
      "loss": 5.9204,
      "mean_token_accuracy": 0.17294538915157318,
      "num_tokens": 28362229.0,
      "step": 11185
    },
    {
      "entropy": 6.06131763458252,
      "epoch": 1.291402192729371,
      "grad_norm": 1.046875,
      "learning_rate": 0.000484468983645402,
      "loss": 5.9125,
      "mean_token_accuracy": 0.17595116347074508,
      "num_tokens": 28374767.0,
      "step": 11190
    },
    {
      "entropy": 5.985938501358032,
      "epoch": 1.2919792267743797,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048445391502909213,
      "loss": 5.8536,
      "mean_token_accuracy": 0.17375739961862563,
      "num_tokens": 28387401.0,
      "step": 11195
    },
    {
      "entropy": 6.020863246917725,
      "epoch": 1.2925562608193883,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048443883936790386,
      "loss": 5.9171,
      "mean_token_accuracy": 0.1787354215979576,
      "num_tokens": 28400123.0,
      "step": 11200
    },
    {
      "entropy": 6.109019374847412,
      "epoch": 1.293133294864397,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048442375666234427,
      "loss": 5.9894,
      "mean_token_accuracy": 0.17528459727764129,
      "num_tokens": 28411607.0,
      "step": 11205
    },
    {
      "entropy": 6.185568761825562,
      "epoch": 1.2937103289094056,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004844086669129206,
      "loss": 5.9793,
      "mean_token_accuracy": 0.16734876185655595,
      "num_tokens": 28424299.0,
      "step": 11210
    },
    {
      "entropy": 6.174504709243775,
      "epoch": 1.2942873629544143,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048439357012014045,
      "loss": 6.0329,
      "mean_token_accuracy": 0.17625446170568465,
      "num_tokens": 28438646.0,
      "step": 11215
    },
    {
      "entropy": 6.066485118865967,
      "epoch": 1.294864396999423,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004843784662845116,
      "loss": 5.9809,
      "mean_token_accuracy": 0.16953236758708953,
      "num_tokens": 28452870.0,
      "step": 11220
    },
    {
      "entropy": 6.151538896560669,
      "epoch": 1.2954414310444315,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000484363355406542,
      "loss": 6.044,
      "mean_token_accuracy": 0.17070204764604568,
      "num_tokens": 28465391.0,
      "step": 11225
    },
    {
      "entropy": 6.0952962875366214,
      "epoch": 1.2960184650894404,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048434823748674,
      "loss": 5.9026,
      "mean_token_accuracy": 0.17612581998109816,
      "num_tokens": 28477458.0,
      "step": 11230
    },
    {
      "entropy": 6.197121667861938,
      "epoch": 1.2965954991344488,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048433311252561403,
      "loss": 5.9551,
      "mean_token_accuracy": 0.17088288813829422,
      "num_tokens": 28489092.0,
      "step": 11235
    },
    {
      "entropy": 6.137632274627686,
      "epoch": 1.2971725331794577,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004843179805236728,
      "loss": 5.9698,
      "mean_token_accuracy": 0.17735913544893264,
      "num_tokens": 28501990.0,
      "step": 11240
    },
    {
      "entropy": 6.115343523025513,
      "epoch": 1.297749567224466,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048430284148142516,
      "loss": 5.9368,
      "mean_token_accuracy": 0.1745120257139206,
      "num_tokens": 28515984.0,
      "step": 11245
    },
    {
      "entropy": 6.03889479637146,
      "epoch": 1.298326601269475,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004842876953993805,
      "loss": 5.8908,
      "mean_token_accuracy": 0.17861931324005126,
      "num_tokens": 28528093.0,
      "step": 11250
    },
    {
      "entropy": 6.12768874168396,
      "epoch": 1.2989036353144836,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004842725422780482,
      "loss": 5.9573,
      "mean_token_accuracy": 0.17614713609218596,
      "num_tokens": 28541177.0,
      "step": 11255
    },
    {
      "entropy": 6.093069267272949,
      "epoch": 1.2994806693594922,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004842573821179378,
      "loss": 5.8855,
      "mean_token_accuracy": 0.17719563394784926,
      "num_tokens": 28553033.0,
      "step": 11260
    },
    {
      "entropy": 5.958415412902832,
      "epoch": 1.3000577034045009,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004842422149195593,
      "loss": 5.8198,
      "mean_token_accuracy": 0.18456507474184036,
      "num_tokens": 28565787.0,
      "step": 11265
    },
    {
      "entropy": 6.117810392379761,
      "epoch": 1.3006347374495095,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048422704068342294,
      "loss": 5.9565,
      "mean_token_accuracy": 0.1704675704240799,
      "num_tokens": 28578823.0,
      "step": 11270
    },
    {
      "entropy": 6.14793438911438,
      "epoch": 1.3012117714945182,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004842118594100389,
      "loss": 5.9851,
      "mean_token_accuracy": 0.17299738079309462,
      "num_tokens": 28590873.0,
      "step": 11275
    },
    {
      "entropy": 6.146806812286377,
      "epoch": 1.3017888055395268,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048419667109991793,
      "loss": 5.9377,
      "mean_token_accuracy": 0.17301224917173386,
      "num_tokens": 28603202.0,
      "step": 11280
    },
    {
      "entropy": 6.11052131652832,
      "epoch": 1.3023658395845354,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048418147575357085,
      "loss": 6.0146,
      "mean_token_accuracy": 0.16819313168525696,
      "num_tokens": 28616803.0,
      "step": 11285
    },
    {
      "entropy": 6.138533973693848,
      "epoch": 1.302942873629544,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004841662733715087,
      "loss": 6.0039,
      "mean_token_accuracy": 0.17240726947784424,
      "num_tokens": 28629439.0,
      "step": 11290
    },
    {
      "entropy": 6.108653593063354,
      "epoch": 1.3035199076745527,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048415106395424294,
      "loss": 5.9245,
      "mean_token_accuracy": 0.18144268840551375,
      "num_tokens": 28642002.0,
      "step": 11295
    },
    {
      "entropy": 6.055723333358765,
      "epoch": 1.3040969417195614,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048413584750228494,
      "loss": 6.0138,
      "mean_token_accuracy": 0.16678009331226348,
      "num_tokens": 28653906.0,
      "step": 11300
    },
    {
      "entropy": 6.138033723831176,
      "epoch": 1.3046739757645702,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048412062401614653,
      "loss": 5.9371,
      "mean_token_accuracy": 0.17298426777124404,
      "num_tokens": 28666085.0,
      "step": 11305
    },
    {
      "entropy": 6.141763734817505,
      "epoch": 1.3052510098095786,
      "grad_norm": 1.0,
      "learning_rate": 0.00048410539349634,
      "loss": 5.9811,
      "mean_token_accuracy": 0.170349383354187,
      "num_tokens": 28679298.0,
      "step": 11310
    },
    {
      "entropy": 6.130426454544067,
      "epoch": 1.3058280438545875,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048409015594337725,
      "loss": 5.9109,
      "mean_token_accuracy": 0.17495015263557434,
      "num_tokens": 28691146.0,
      "step": 11315
    },
    {
      "entropy": 6.03031210899353,
      "epoch": 1.3064050778995961,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000484074911357771,
      "loss": 5.8865,
      "mean_token_accuracy": 0.18226094543933868,
      "num_tokens": 28705259.0,
      "step": 11320
    },
    {
      "entropy": 6.11535382270813,
      "epoch": 1.3069821119446048,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048405965974003404,
      "loss": 5.9363,
      "mean_token_accuracy": 0.17273859977722167,
      "num_tokens": 28718004.0,
      "step": 11325
    },
    {
      "entropy": 6.244010591506958,
      "epoch": 1.3075591459896134,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048404440109067927,
      "loss": 6.0409,
      "mean_token_accuracy": 0.15987419933080674,
      "num_tokens": 28730442.0,
      "step": 11330
    },
    {
      "entropy": 6.031905746459961,
      "epoch": 1.308136180034622,
      "grad_norm": 1.0,
      "learning_rate": 0.0004840291354102198,
      "loss": 5.9444,
      "mean_token_accuracy": 0.17410803586244583,
      "num_tokens": 28743078.0,
      "step": 11335
    },
    {
      "entropy": 6.0483558654785154,
      "epoch": 1.3087132140796307,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004840138626991693,
      "loss": 5.9021,
      "mean_token_accuracy": 0.18282301723957062,
      "num_tokens": 28756975.0,
      "step": 11340
    },
    {
      "entropy": 6.080908489227295,
      "epoch": 1.3092902481246393,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004839985829580413,
      "loss": 5.9448,
      "mean_token_accuracy": 0.1796576648950577,
      "num_tokens": 28770830.0,
      "step": 11345
    },
    {
      "entropy": 6.0359063148498535,
      "epoch": 1.309867282169648,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048398329618734977,
      "loss": 5.9358,
      "mean_token_accuracy": 0.17064609229564667,
      "num_tokens": 28782894.0,
      "step": 11350
    },
    {
      "entropy": 6.145987796783447,
      "epoch": 1.3104443162146566,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004839680023876089,
      "loss": 5.9889,
      "mean_token_accuracy": 0.17035564482212068,
      "num_tokens": 28796412.0,
      "step": 11355
    },
    {
      "entropy": 6.185467386245728,
      "epoch": 1.3110213502596653,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004839527015593331,
      "loss": 5.9832,
      "mean_token_accuracy": 0.16961285918951036,
      "num_tokens": 28809004.0,
      "step": 11360
    },
    {
      "entropy": 6.018322706222534,
      "epoch": 1.311598384304674,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048393739370303684,
      "loss": 5.9085,
      "mean_token_accuracy": 0.18176113367080687,
      "num_tokens": 28821104.0,
      "step": 11365
    },
    {
      "entropy": 6.0850245475769045,
      "epoch": 1.3121754183496828,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004839220788192353,
      "loss": 5.9354,
      "mean_token_accuracy": 0.1715235397219658,
      "num_tokens": 28833413.0,
      "step": 11370
    },
    {
      "entropy": 6.1291341304779055,
      "epoch": 1.3127524523946912,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048390675690844335,
      "loss": 5.9735,
      "mean_token_accuracy": 0.17504852265119553,
      "num_tokens": 28846233.0,
      "step": 11375
    },
    {
      "entropy": 6.120204734802246,
      "epoch": 1.3133294864397,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004838914279711764,
      "loss": 5.9769,
      "mean_token_accuracy": 0.17033064216375352,
      "num_tokens": 28859296.0,
      "step": 11380
    },
    {
      "entropy": 6.095674753189087,
      "epoch": 1.3139065204847085,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048387609200795003,
      "loss": 5.947,
      "mean_token_accuracy": 0.17082445323467255,
      "num_tokens": 28872871.0,
      "step": 11385
    },
    {
      "entropy": 6.117646074295044,
      "epoch": 1.3144835545297173,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048386074901928,
      "loss": 5.9383,
      "mean_token_accuracy": 0.1711859956383705,
      "num_tokens": 28885842.0,
      "step": 11390
    },
    {
      "entropy": 6.050236988067627,
      "epoch": 1.315060588574726,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004838453990056825,
      "loss": 5.938,
      "mean_token_accuracy": 0.17580579817295075,
      "num_tokens": 28898030.0,
      "step": 11395
    },
    {
      "entropy": 6.096662950515747,
      "epoch": 1.3156376226197346,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048383004196767373,
      "loss": 6.0101,
      "mean_token_accuracy": 0.16708967387676238,
      "num_tokens": 28910562.0,
      "step": 11400
    },
    {
      "entropy": 6.263904237747193,
      "epoch": 1.3162146566647432,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004838146779057702,
      "loss": 6.0455,
      "mean_token_accuracy": 0.16449646949768065,
      "num_tokens": 28923169.0,
      "step": 11405
    },
    {
      "entropy": 6.145656538009644,
      "epoch": 1.3167916907097519,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004837993068204887,
      "loss": 6.0154,
      "mean_token_accuracy": 0.17170106023550033,
      "num_tokens": 28935840.0,
      "step": 11410
    },
    {
      "entropy": 6.010359764099121,
      "epoch": 1.3173687247547605,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048378392871234634,
      "loss": 5.8508,
      "mean_token_accuracy": 0.17636101841926574,
      "num_tokens": 28948533.0,
      "step": 11415
    },
    {
      "entropy": 6.074927043914795,
      "epoch": 1.3179457587997692,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004837685435818601,
      "loss": 5.9046,
      "mean_token_accuracy": 0.1742495134472847,
      "num_tokens": 28961060.0,
      "step": 11420
    },
    {
      "entropy": 6.116287612915039,
      "epoch": 1.3185227928447778,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004837531514295477,
      "loss": 5.9217,
      "mean_token_accuracy": 0.18382218480110168,
      "num_tokens": 28973570.0,
      "step": 11425
    },
    {
      "entropy": 6.012993860244751,
      "epoch": 1.3190998268897864,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004837377522559267,
      "loss": 5.9371,
      "mean_token_accuracy": 0.17799128890037536,
      "num_tokens": 28987519.0,
      "step": 11430
    },
    {
      "entropy": 6.135233449935913,
      "epoch": 1.319676860934795,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048372234606151507,
      "loss": 5.9212,
      "mean_token_accuracy": 0.17561387717723848,
      "num_tokens": 29000790.0,
      "step": 11435
    },
    {
      "entropy": 6.113107776641845,
      "epoch": 1.3202538949798037,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048370693284683106,
      "loss": 5.9352,
      "mean_token_accuracy": 0.16696703881025315,
      "num_tokens": 29013263.0,
      "step": 11440
    },
    {
      "entropy": 6.054227828979492,
      "epoch": 1.3208309290248126,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048369151261239303,
      "loss": 5.9277,
      "mean_token_accuracy": 0.18207932710647584,
      "num_tokens": 29024958.0,
      "step": 11445
    },
    {
      "entropy": 6.039257335662842,
      "epoch": 1.321407963069821,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004836760853587196,
      "loss": 5.9473,
      "mean_token_accuracy": 0.17116033732891084,
      "num_tokens": 29037752.0,
      "step": 11450
    },
    {
      "entropy": 6.065895938873291,
      "epoch": 1.3219849971148299,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048366065108632967,
      "loss": 5.8377,
      "mean_token_accuracy": 0.1877012297511101,
      "num_tokens": 29050872.0,
      "step": 11455
    },
    {
      "entropy": 6.133243656158447,
      "epoch": 1.3225620311598385,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048364520979574246,
      "loss": 5.9109,
      "mean_token_accuracy": 0.17161595225334167,
      "num_tokens": 29062472.0,
      "step": 11460
    },
    {
      "entropy": 6.0585527420043945,
      "epoch": 1.3231390652048471,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048362976148747715,
      "loss": 5.9763,
      "mean_token_accuracy": 0.17766943722963333,
      "num_tokens": 29075392.0,
      "step": 11465
    },
    {
      "entropy": 6.078961515426636,
      "epoch": 1.3237160992498558,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004836143061620536,
      "loss": 5.9323,
      "mean_token_accuracy": 0.16850050985813142,
      "num_tokens": 29088221.0,
      "step": 11470
    },
    {
      "entropy": 6.141566324234009,
      "epoch": 1.3242931332948644,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004835988438199914,
      "loss": 5.9718,
      "mean_token_accuracy": 0.17352256178855896,
      "num_tokens": 29101050.0,
      "step": 11475
    },
    {
      "entropy": 6.108234834671021,
      "epoch": 1.324870167339873,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004835833744618107,
      "loss": 5.9898,
      "mean_token_accuracy": 0.16549343168735503,
      "num_tokens": 29113928.0,
      "step": 11480
    },
    {
      "entropy": 6.158697175979614,
      "epoch": 1.3254472013848817,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004835678980880318,
      "loss": 6.0603,
      "mean_token_accuracy": 0.16534637957811354,
      "num_tokens": 29127196.0,
      "step": 11485
    },
    {
      "entropy": 6.051623153686523,
      "epoch": 1.3260242354298903,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004835524146991753,
      "loss": 5.928,
      "mean_token_accuracy": 0.1737898290157318,
      "num_tokens": 29140498.0,
      "step": 11490
    },
    {
      "entropy": 6.034381675720215,
      "epoch": 1.326601269474899,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048353692429576185,
      "loss": 5.8036,
      "mean_token_accuracy": 0.18179915398359298,
      "num_tokens": 29153910.0,
      "step": 11495
    },
    {
      "entropy": 6.107257890701294,
      "epoch": 1.3271783035199076,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004835214268783126,
      "loss": 5.9683,
      "mean_token_accuracy": 0.17448195815086365,
      "num_tokens": 29166289.0,
      "step": 11500
    },
    {
      "entropy": 6.126091337203979,
      "epoch": 1.3277553375649163,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048350592244734866,
      "loss": 5.9732,
      "mean_token_accuracy": 0.17407142817974092,
      "num_tokens": 29178321.0,
      "step": 11505
    },
    {
      "entropy": 6.120612096786499,
      "epoch": 1.3283323716099251,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004834904110033917,
      "loss": 5.937,
      "mean_token_accuracy": 0.17456878274679183,
      "num_tokens": 29190396.0,
      "step": 11510
    },
    {
      "entropy": 6.0490258693695065,
      "epoch": 1.3289094056549335,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048347489254696327,
      "loss": 5.9779,
      "mean_token_accuracy": 0.1751318782567978,
      "num_tokens": 29202698.0,
      "step": 11515
    },
    {
      "entropy": 6.082028722763061,
      "epoch": 1.3294864396999424,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004834593670785854,
      "loss": 5.8875,
      "mean_token_accuracy": 0.1831155925989151,
      "num_tokens": 29214508.0,
      "step": 11520
    },
    {
      "entropy": 6.16449728012085,
      "epoch": 1.3300634737449508,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048344383459878024,
      "loss": 5.9692,
      "mean_token_accuracy": 0.17426125407218934,
      "num_tokens": 29225683.0,
      "step": 11525
    },
    {
      "entropy": 6.023352241516113,
      "epoch": 1.3306405077899597,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048342829510807024,
      "loss": 5.8738,
      "mean_token_accuracy": 0.18996732532978058,
      "num_tokens": 29239497.0,
      "step": 11530
    },
    {
      "entropy": 6.059785079956055,
      "epoch": 1.3312175418349683,
      "grad_norm": 0.953125,
      "learning_rate": 0.000483412748606978,
      "loss": 5.9596,
      "mean_token_accuracy": 0.1758654996752739,
      "num_tokens": 29252230.0,
      "step": 11535
    },
    {
      "entropy": 6.080663585662842,
      "epoch": 1.331794575879977,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004833971950960266,
      "loss": 5.9373,
      "mean_token_accuracy": 0.17315787076950073,
      "num_tokens": 29265349.0,
      "step": 11540
    },
    {
      "entropy": 6.086237001419067,
      "epoch": 1.3323716099249856,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004833816345757391,
      "loss": 5.9472,
      "mean_token_accuracy": 0.17665637284517288,
      "num_tokens": 29278291.0,
      "step": 11545
    },
    {
      "entropy": 6.080244684219361,
      "epoch": 1.3329486439699942,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004833660670466387,
      "loss": 5.9777,
      "mean_token_accuracy": 0.17158039063215255,
      "num_tokens": 29290580.0,
      "step": 11550
    },
    {
      "entropy": 6.18371729850769,
      "epoch": 1.3335256780150029,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004833504925092492,
      "loss": 6.0971,
      "mean_token_accuracy": 0.1635443538427353,
      "num_tokens": 29305738.0,
      "step": 11555
    },
    {
      "entropy": 6.046380996704102,
      "epoch": 1.3341027120600115,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004833349109640944,
      "loss": 5.8798,
      "mean_token_accuracy": 0.17971539497375488,
      "num_tokens": 29316533.0,
      "step": 11560
    },
    {
      "entropy": 6.054925632476807,
      "epoch": 1.3346797461050202,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004833193224116983,
      "loss": 5.9235,
      "mean_token_accuracy": 0.17894544452428818,
      "num_tokens": 29329723.0,
      "step": 11565
    },
    {
      "entropy": 6.1368358612060545,
      "epoch": 1.3352567801500288,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048330372685258526,
      "loss": 5.9807,
      "mean_token_accuracy": 0.17134260535240173,
      "num_tokens": 29342830.0,
      "step": 11570
    },
    {
      "entropy": 5.956214857101441,
      "epoch": 1.3358338141950374,
      "grad_norm": 1.0,
      "learning_rate": 0.0004832881242872799,
      "loss": 5.8467,
      "mean_token_accuracy": 0.18215566724538804,
      "num_tokens": 29354096.0,
      "step": 11575
    },
    {
      "entropy": 6.07789249420166,
      "epoch": 1.336410848240046,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004832725147163069,
      "loss": 5.9013,
      "mean_token_accuracy": 0.17742161750793456,
      "num_tokens": 29367173.0,
      "step": 11580
    },
    {
      "entropy": 6.098205614089966,
      "epoch": 1.336987882285055,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048325689814019134,
      "loss": 5.9297,
      "mean_token_accuracy": 0.17371537536382675,
      "num_tokens": 29378958.0,
      "step": 11585
    },
    {
      "entropy": 6.092908668518066,
      "epoch": 1.3375649163300634,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004832412745594585,
      "loss": 5.9391,
      "mean_token_accuracy": 0.17739199697971345,
      "num_tokens": 29391360.0,
      "step": 11590
    },
    {
      "entropy": 6.097428274154663,
      "epoch": 1.3381419503750722,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048322564397463376,
      "loss": 5.9768,
      "mean_token_accuracy": 0.17219074964523315,
      "num_tokens": 29404040.0,
      "step": 11595
    },
    {
      "entropy": 6.130678272247314,
      "epoch": 1.3387189844200809,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048321000638624296,
      "loss": 5.9772,
      "mean_token_accuracy": 0.16908041685819625,
      "num_tokens": 29417960.0,
      "step": 11600
    },
    {
      "entropy": 6.109904384613037,
      "epoch": 1.3392960184650895,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000483194361794812,
      "loss": 5.9476,
      "mean_token_accuracy": 0.17486595511436462,
      "num_tokens": 29431453.0,
      "step": 11605
    },
    {
      "entropy": 6.077830219268799,
      "epoch": 1.3398730525100981,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000483178710200867,
      "loss": 6.0007,
      "mean_token_accuracy": 0.1671554908156395,
      "num_tokens": 29443716.0,
      "step": 11610
    },
    {
      "entropy": 6.044572496414185,
      "epoch": 1.3404500865551068,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004831630516049346,
      "loss": 5.8816,
      "mean_token_accuracy": 0.1758398562669754,
      "num_tokens": 29455588.0,
      "step": 11615
    },
    {
      "entropy": 5.980679512023926,
      "epoch": 1.3410271206001154,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004831473860075414,
      "loss": 5.8177,
      "mean_token_accuracy": 0.18274110108613967,
      "num_tokens": 29469472.0,
      "step": 11620
    },
    {
      "entropy": 6.076540851593018,
      "epoch": 1.341604154645124,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048313171340921417,
      "loss": 5.9185,
      "mean_token_accuracy": 0.18089540153741837,
      "num_tokens": 29482432.0,
      "step": 11625
    },
    {
      "entropy": 6.105173206329345,
      "epoch": 1.3421811886901327,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048311603381048025,
      "loss": 5.8896,
      "mean_token_accuracy": 0.18112491220235824,
      "num_tokens": 29495425.0,
      "step": 11630
    },
    {
      "entropy": 6.096925163269043,
      "epoch": 1.3427582227351413,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004831003472118668,
      "loss": 5.9812,
      "mean_token_accuracy": 0.17457438856363297,
      "num_tokens": 29507336.0,
      "step": 11635
    },
    {
      "entropy": 5.967815637588501,
      "epoch": 1.34333525678015,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004830846536139016,
      "loss": 5.8969,
      "mean_token_accuracy": 0.17934599816799163,
      "num_tokens": 29519857.0,
      "step": 11640
    },
    {
      "entropy": 6.059503412246704,
      "epoch": 1.3439122908251586,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004830689530171124,
      "loss": 5.9127,
      "mean_token_accuracy": 0.17375947833061217,
      "num_tokens": 29531842.0,
      "step": 11645
    },
    {
      "entropy": 6.131421232223511,
      "epoch": 1.3444893248701673,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004830532454220273,
      "loss": 5.9855,
      "mean_token_accuracy": 0.17311149537563325,
      "num_tokens": 29543721.0,
      "step": 11650
    },
    {
      "entropy": 6.1412577629089355,
      "epoch": 1.345066358915176,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048303753082917474,
      "loss": 5.9575,
      "mean_token_accuracy": 0.17456770986318587,
      "num_tokens": 29555713.0,
      "step": 11655
    },
    {
      "entropy": 6.0892298221588135,
      "epoch": 1.3456433929601848,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048302180923908306,
      "loss": 5.9783,
      "mean_token_accuracy": 0.17275725156068802,
      "num_tokens": 29568983.0,
      "step": 11660
    },
    {
      "entropy": 6.118532800674439,
      "epoch": 1.3462204270051932,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048300608065228126,
      "loss": 5.9885,
      "mean_token_accuracy": 0.17179093807935714,
      "num_tokens": 29580778.0,
      "step": 11665
    },
    {
      "entropy": 6.104513168334961,
      "epoch": 1.346797461050202,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048299034506929815,
      "loss": 5.9824,
      "mean_token_accuracy": 0.17020383477210999,
      "num_tokens": 29594395.0,
      "step": 11670
    },
    {
      "entropy": 5.954204082489014,
      "epoch": 1.3473744950952107,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048297460249066315,
      "loss": 5.849,
      "mean_token_accuracy": 0.1876319870352745,
      "num_tokens": 29607588.0,
      "step": 11675
    },
    {
      "entropy": 6.124476289749145,
      "epoch": 1.3479515291402193,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004829588529169057,
      "loss": 5.9501,
      "mean_token_accuracy": 0.17769657224416732,
      "num_tokens": 29618822.0,
      "step": 11680
    },
    {
      "entropy": 5.9971160888671875,
      "epoch": 1.348528563185228,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004829430963485555,
      "loss": 5.9161,
      "mean_token_accuracy": 0.17655061334371566,
      "num_tokens": 29631759.0,
      "step": 11685
    },
    {
      "entropy": 6.100216913223266,
      "epoch": 1.3491055972302366,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004829273327861426,
      "loss": 5.9363,
      "mean_token_accuracy": 0.16798125952482224,
      "num_tokens": 29644381.0,
      "step": 11690
    },
    {
      "entropy": 6.094891548156738,
      "epoch": 1.3496826312752452,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004829115622301971,
      "loss": 5.9498,
      "mean_token_accuracy": 0.17301006764173507,
      "num_tokens": 29657370.0,
      "step": 11695
    },
    {
      "entropy": 6.128429841995239,
      "epoch": 1.3502596653202539,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048289578468124956,
      "loss": 5.9765,
      "mean_token_accuracy": 0.1713126465678215,
      "num_tokens": 29669438.0,
      "step": 11700
    },
    {
      "entropy": 6.122415447235108,
      "epoch": 1.3508366993652625,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048288000013983046,
      "loss": 5.9878,
      "mean_token_accuracy": 0.1747574269771576,
      "num_tokens": 29681277.0,
      "step": 11705
    },
    {
      "entropy": 6.032140779495239,
      "epoch": 1.3514137334102712,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048286420860647086,
      "loss": 5.8453,
      "mean_token_accuracy": 0.18025387674570084,
      "num_tokens": 29693523.0,
      "step": 11710
    },
    {
      "entropy": 6.202200126647949,
      "epoch": 1.3519907674552798,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048284841008170185,
      "loss": 6.0388,
      "mean_token_accuracy": 0.16537994295358657,
      "num_tokens": 29707589.0,
      "step": 11715
    },
    {
      "entropy": 6.108614063262939,
      "epoch": 1.3525678015002884,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048283260456605487,
      "loss": 6.0315,
      "mean_token_accuracy": 0.1682431697845459,
      "num_tokens": 29719267.0,
      "step": 11720
    },
    {
      "entropy": 6.057359457015991,
      "epoch": 1.3531448355452973,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004828167920600614,
      "loss": 5.8577,
      "mean_token_accuracy": 0.17803561091423034,
      "num_tokens": 29731575.0,
      "step": 11725
    },
    {
      "entropy": 6.124388885498047,
      "epoch": 1.3537218695903057,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004828009725642534,
      "loss": 5.9841,
      "mean_token_accuracy": 0.16690063774585723,
      "num_tokens": 29744762.0,
      "step": 11730
    },
    {
      "entropy": 6.115428352355957,
      "epoch": 1.3542989036353146,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004827851460791628,
      "loss": 6.0293,
      "mean_token_accuracy": 0.17205984592437745,
      "num_tokens": 29758300.0,
      "step": 11735
    },
    {
      "entropy": 6.081460189819336,
      "epoch": 1.3548759376803232,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048276931260532213,
      "loss": 5.9089,
      "mean_token_accuracy": 0.17634457647800444,
      "num_tokens": 29771489.0,
      "step": 11740
    },
    {
      "entropy": 6.101334428787231,
      "epoch": 1.3554529717253319,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004827534721432639,
      "loss": 6.009,
      "mean_token_accuracy": 0.1757515624165535,
      "num_tokens": 29783063.0,
      "step": 11745
    },
    {
      "entropy": 6.147397422790528,
      "epoch": 1.3560300057703405,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004827376246935207,
      "loss": 6.0265,
      "mean_token_accuracy": 0.17187947779893875,
      "num_tokens": 29797136.0,
      "step": 11750
    },
    {
      "entropy": 6.129934549331665,
      "epoch": 1.3566070398153491,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004827217702566257,
      "loss": 5.9334,
      "mean_token_accuracy": 0.18125496059656143,
      "num_tokens": 29809580.0,
      "step": 11755
    },
    {
      "entropy": 5.962800645828247,
      "epoch": 1.3571840738603578,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048270590883311217,
      "loss": 5.8398,
      "mean_token_accuracy": 0.1791318029165268,
      "num_tokens": 29822400.0,
      "step": 11760
    },
    {
      "entropy": 6.05812635421753,
      "epoch": 1.3577611079053664,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048269004042351363,
      "loss": 5.9422,
      "mean_token_accuracy": 0.17107647508382798,
      "num_tokens": 29834392.0,
      "step": 11765
    },
    {
      "entropy": 6.162915325164795,
      "epoch": 1.358338141950375,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004826741650283637,
      "loss": 5.9496,
      "mean_token_accuracy": 0.17571107894182206,
      "num_tokens": 29846626.0,
      "step": 11770
    },
    {
      "entropy": 6.08893232345581,
      "epoch": 1.3589151759953837,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004826582826481963,
      "loss": 5.9008,
      "mean_token_accuracy": 0.17556214779615403,
      "num_tokens": 29858988.0,
      "step": 11775
    },
    {
      "entropy": 5.954388284683228,
      "epoch": 1.3594922100403923,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004826423932835458,
      "loss": 5.9047,
      "mean_token_accuracy": 0.17963224202394484,
      "num_tokens": 29871875.0,
      "step": 11780
    },
    {
      "entropy": 6.087636947631836,
      "epoch": 1.360069244085401,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048262649693494653,
      "loss": 5.9047,
      "mean_token_accuracy": 0.17776284366846085,
      "num_tokens": 29884757.0,
      "step": 11785
    },
    {
      "entropy": 6.103380918502808,
      "epoch": 1.3606462781304096,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004826105936029332,
      "loss": 5.9405,
      "mean_token_accuracy": 0.17620218694210052,
      "num_tokens": 29897309.0,
      "step": 11790
    },
    {
      "entropy": 6.1345278263092045,
      "epoch": 1.3612233121754183,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004825946832880406,
      "loss": 6.0101,
      "mean_token_accuracy": 0.16618997752666473,
      "num_tokens": 29910005.0,
      "step": 11795
    },
    {
      "entropy": 6.063724136352539,
      "epoch": 1.3618003462204271,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048257876599080404,
      "loss": 5.9496,
      "mean_token_accuracy": 0.1740277275443077,
      "num_tokens": 29921912.0,
      "step": 11800
    },
    {
      "entropy": 6.127210712432861,
      "epoch": 1.3623773802654355,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048256284171175874,
      "loss": 5.9446,
      "mean_token_accuracy": 0.1731536105275154,
      "num_tokens": 29934399.0,
      "step": 11805
    },
    {
      "entropy": 6.090194511413574,
      "epoch": 1.3629544143104444,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048254691045144035,
      "loss": 5.8791,
      "mean_token_accuracy": 0.17328746169805526,
      "num_tokens": 29946660.0,
      "step": 11810
    },
    {
      "entropy": 6.0298761367797855,
      "epoch": 1.363531448355453,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004825309722103848,
      "loss": 5.9027,
      "mean_token_accuracy": 0.1751670151948929,
      "num_tokens": 29958382.0,
      "step": 11815
    },
    {
      "entropy": 6.122469091415406,
      "epoch": 1.3641084824004617,
      "grad_norm": 1.015625,
      "learning_rate": 0.000482515026989128,
      "loss": 5.911,
      "mean_token_accuracy": 0.1743503525853157,
      "num_tokens": 29971241.0,
      "step": 11820
    },
    {
      "entropy": 6.140733098983764,
      "epoch": 1.3646855164454703,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048249907478820634,
      "loss": 5.9848,
      "mean_token_accuracy": 0.17422153949737548,
      "num_tokens": 29984145.0,
      "step": 11825
    },
    {
      "entropy": 6.044448375701904,
      "epoch": 1.365262550490479,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048248311560815637,
      "loss": 5.8688,
      "mean_token_accuracy": 0.1806940332055092,
      "num_tokens": 29997092.0,
      "step": 11830
    },
    {
      "entropy": 5.958268308639527,
      "epoch": 1.3658395845354876,
      "grad_norm": 0.875,
      "learning_rate": 0.0004824671494495149,
      "loss": 5.9323,
      "mean_token_accuracy": 0.1785980373620987,
      "num_tokens": 30009369.0,
      "step": 11835
    },
    {
      "entropy": 6.067847490310669,
      "epoch": 1.3664166185804962,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004824511763128189,
      "loss": 5.8787,
      "mean_token_accuracy": 0.17705994546413423,
      "num_tokens": 30021423.0,
      "step": 11840
    },
    {
      "entropy": 6.128710985183716,
      "epoch": 1.3669936526255049,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048243519619860567,
      "loss": 5.9497,
      "mean_token_accuracy": 0.17178182452917098,
      "num_tokens": 30034426.0,
      "step": 11845
    },
    {
      "entropy": 6.061038017272949,
      "epoch": 1.3675706866705135,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004824192091074126,
      "loss": 5.9023,
      "mean_token_accuracy": 0.17695294916629792,
      "num_tokens": 30046687.0,
      "step": 11850
    },
    {
      "entropy": 6.0966493606567385,
      "epoch": 1.3681477207155222,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004824032150397775,
      "loss": 5.9996,
      "mean_token_accuracy": 0.16679150909185408,
      "num_tokens": 30058815.0,
      "step": 11855
    },
    {
      "entropy": 6.179765796661377,
      "epoch": 1.3687247547605308,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048238721399623824,
      "loss": 5.9966,
      "mean_token_accuracy": 0.16816366463899612,
      "num_tokens": 30070715.0,
      "step": 11860
    },
    {
      "entropy": 6.090771675109863,
      "epoch": 1.3693017888055397,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048237120597733316,
      "loss": 5.8575,
      "mean_token_accuracy": 0.1778775066137314,
      "num_tokens": 30083636.0,
      "step": 11865
    },
    {
      "entropy": 6.08870759010315,
      "epoch": 1.369878822850548,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004823551909836005,
      "loss": 5.9725,
      "mean_token_accuracy": 0.17038166970014573,
      "num_tokens": 30094956.0,
      "step": 11870
    },
    {
      "entropy": 6.125296115875244,
      "epoch": 1.370455856895557,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048233916901557896,
      "loss": 5.8939,
      "mean_token_accuracy": 0.17389345169067383,
      "num_tokens": 30107947.0,
      "step": 11875
    },
    {
      "entropy": 6.136224031448364,
      "epoch": 1.3710328909405656,
      "grad_norm": 1.0,
      "learning_rate": 0.00048232314007380753,
      "loss": 6.0046,
      "mean_token_accuracy": 0.1672964096069336,
      "num_tokens": 30120557.0,
      "step": 11880
    },
    {
      "entropy": 6.119308996200561,
      "epoch": 1.3716099249855742,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048230710415882524,
      "loss": 5.9991,
      "mean_token_accuracy": 0.17866672426462174,
      "num_tokens": 30131734.0,
      "step": 11885
    },
    {
      "entropy": 6.100404500961304,
      "epoch": 1.3721869590305829,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048229106127117144,
      "loss": 5.8817,
      "mean_token_accuracy": 0.18228523582220077,
      "num_tokens": 30144025.0,
      "step": 11890
    },
    {
      "entropy": 6.126446723937988,
      "epoch": 1.3727639930755915,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004822750114113858,
      "loss": 5.9689,
      "mean_token_accuracy": 0.17440032809972764,
      "num_tokens": 30156246.0,
      "step": 11895
    },
    {
      "entropy": 6.0651209354400635,
      "epoch": 1.3733410271206001,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004822589545800081,
      "loss": 5.9259,
      "mean_token_accuracy": 0.17684071511030197,
      "num_tokens": 30169349.0,
      "step": 11900
    },
    {
      "entropy": 6.133235168457031,
      "epoch": 1.3739180611656088,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004822428907775784,
      "loss": 5.9515,
      "mean_token_accuracy": 0.17447723299264908,
      "num_tokens": 30181821.0,
      "step": 11905
    },
    {
      "entropy": 6.129652404785157,
      "epoch": 1.3744950952106174,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048222682000463704,
      "loss": 5.9157,
      "mean_token_accuracy": 0.17929187119007112,
      "num_tokens": 30194278.0,
      "step": 11910
    },
    {
      "entropy": 6.08655195236206,
      "epoch": 1.375072129255626,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004822107422617245,
      "loss": 5.942,
      "mean_token_accuracy": 0.17582537084817887,
      "num_tokens": 30206706.0,
      "step": 11915
    },
    {
      "entropy": 6.033338975906372,
      "epoch": 1.3756491633006347,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048219465754938156,
      "loss": 5.9233,
      "mean_token_accuracy": 0.17808154821395875,
      "num_tokens": 30220356.0,
      "step": 11920
    },
    {
      "entropy": 6.135046768188476,
      "epoch": 1.3762261973456433,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048217856586814926,
      "loss": 5.9635,
      "mean_token_accuracy": 0.17709923088550567,
      "num_tokens": 30232821.0,
      "step": 11925
    },
    {
      "entropy": 6.006950950622558,
      "epoch": 1.376803231390652,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048216246721856875,
      "loss": 5.8958,
      "mean_token_accuracy": 0.17723796516656876,
      "num_tokens": 30246094.0,
      "step": 11930
    },
    {
      "entropy": 6.101513004302978,
      "epoch": 1.3773802654356606,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048214636160118164,
      "loss": 5.9255,
      "mean_token_accuracy": 0.18114928156137466,
      "num_tokens": 30259313.0,
      "step": 11935
    },
    {
      "entropy": 6.177270364761353,
      "epoch": 1.3779572994806695,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004821302490165295,
      "loss": 5.9935,
      "mean_token_accuracy": 0.166851706802845,
      "num_tokens": 30274045.0,
      "step": 11940
    },
    {
      "entropy": 6.090599536895752,
      "epoch": 1.378534333525678,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004821141294651544,
      "loss": 5.9265,
      "mean_token_accuracy": 0.17465987503528596,
      "num_tokens": 30286250.0,
      "step": 11945
    },
    {
      "entropy": 6.016452121734619,
      "epoch": 1.3791113675706868,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048209800294759836,
      "loss": 5.9669,
      "mean_token_accuracy": 0.17568169683218002,
      "num_tokens": 30299417.0,
      "step": 11950
    },
    {
      "entropy": 6.073246097564697,
      "epoch": 1.3796884016156954,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004820818694644039,
      "loss": 5.9116,
      "mean_token_accuracy": 0.17077578604221344,
      "num_tokens": 30311465.0,
      "step": 11955
    },
    {
      "entropy": 5.998990249633789,
      "epoch": 1.380265435660704,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048206572901611364,
      "loss": 5.8401,
      "mean_token_accuracy": 0.18066396266222,
      "num_tokens": 30325941.0,
      "step": 11960
    },
    {
      "entropy": 6.092406463623047,
      "epoch": 1.3808424697057127,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048204958160327034,
      "loss": 6.073,
      "mean_token_accuracy": 0.16636116951704025,
      "num_tokens": 30339238.0,
      "step": 11965
    },
    {
      "entropy": 6.127943229675293,
      "epoch": 1.3814195037507213,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048203342722641726,
      "loss": 5.8838,
      "mean_token_accuracy": 0.17834559828042984,
      "num_tokens": 30351082.0,
      "step": 11970
    },
    {
      "entropy": 6.0980922222137455,
      "epoch": 1.38199653779573,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048201726588609776,
      "loss": 5.9265,
      "mean_token_accuracy": 0.17330157980322838,
      "num_tokens": 30364472.0,
      "step": 11975
    },
    {
      "entropy": 6.045344734191895,
      "epoch": 1.3825735718407386,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048200109758285534,
      "loss": 5.8833,
      "mean_token_accuracy": 0.17546160221099855,
      "num_tokens": 30377113.0,
      "step": 11980
    },
    {
      "entropy": 6.15612759590149,
      "epoch": 1.3831506058857472,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004819849223172337,
      "loss": 5.9581,
      "mean_token_accuracy": 0.17162193953990937,
      "num_tokens": 30389313.0,
      "step": 11985
    },
    {
      "entropy": 6.089161777496338,
      "epoch": 1.3837276399307559,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048196874008977716,
      "loss": 6.0768,
      "mean_token_accuracy": 0.16370799839496614,
      "num_tokens": 30402227.0,
      "step": 11990
    },
    {
      "entropy": 6.077267980575561,
      "epoch": 1.3843046739757645,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004819525509010298,
      "loss": 5.9584,
      "mean_token_accuracy": 0.17190659493207933,
      "num_tokens": 30414719.0,
      "step": 11995
    },
    {
      "entropy": 6.102652597427368,
      "epoch": 1.3848817080207732,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004819363547515361,
      "loss": 5.9883,
      "mean_token_accuracy": 0.1682654172182083,
      "num_tokens": 30427989.0,
      "step": 12000
    },
    {
      "epoch": 1.3848817080207732,
      "eval_entropy": 6.023790257201467,
      "eval_loss": 6.071915149688721,
      "eval_mean_token_accuracy": 0.17303504268719555,
      "eval_num_tokens": 30427989.0,
      "eval_runtime": 22.8251,
      "eval_samples_per_second": 1232.679,
      "eval_steps_per_second": 154.085,
      "step": 12000
    },
    {
      "entropy": 6.093491554260254,
      "epoch": 1.385458742065782,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000481920151641841,
      "loss": 5.8983,
      "mean_token_accuracy": 0.17043909132480622,
      "num_tokens": 30441097.0,
      "step": 12005
    },
    {
      "entropy": 5.964703845977783,
      "epoch": 1.3860357761107904,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048190394157248935,
      "loss": 5.7828,
      "mean_token_accuracy": 0.1835661932826042,
      "num_tokens": 30453535.0,
      "step": 12010
    },
    {
      "entropy": 6.0704926490783695,
      "epoch": 1.3866128101557993,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004818877245440264,
      "loss": 6.0001,
      "mean_token_accuracy": 0.17165828496217728,
      "num_tokens": 30466350.0,
      "step": 12015
    },
    {
      "entropy": 6.206932306289673,
      "epoch": 1.3871898442008077,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048187150055699763,
      "loss": 6.0075,
      "mean_token_accuracy": 0.16973246186971663,
      "num_tokens": 30479183.0,
      "step": 12020
    },
    {
      "entropy": 6.084630680084229,
      "epoch": 1.3877668782458166,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004818552696119487,
      "loss": 5.9107,
      "mean_token_accuracy": 0.1781137093901634,
      "num_tokens": 30491903.0,
      "step": 12025
    },
    {
      "entropy": 6.049559736251831,
      "epoch": 1.3883439122908252,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004818390317094255,
      "loss": 5.9017,
      "mean_token_accuracy": 0.17801452428102493,
      "num_tokens": 30504424.0,
      "step": 12030
    },
    {
      "entropy": 6.069197797775269,
      "epoch": 1.3889209463358339,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004818227868499742,
      "loss": 5.9278,
      "mean_token_accuracy": 0.17522621750831605,
      "num_tokens": 30517339.0,
      "step": 12035
    },
    {
      "entropy": 6.057424592971802,
      "epoch": 1.3894979803808425,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004818065350341412,
      "loss": 5.9463,
      "mean_token_accuracy": 0.16934870481491088,
      "num_tokens": 30529218.0,
      "step": 12040
    },
    {
      "entropy": 6.088070011138916,
      "epoch": 1.3900750144258511,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048179027626247325,
      "loss": 5.8788,
      "mean_token_accuracy": 0.17842864245176315,
      "num_tokens": 30540601.0,
      "step": 12045
    },
    {
      "entropy": 5.958896017074585,
      "epoch": 1.3906520484708598,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004817740105355169,
      "loss": 5.8622,
      "mean_token_accuracy": 0.17979412525892258,
      "num_tokens": 30552834.0,
      "step": 12050
    },
    {
      "entropy": 6.034915256500244,
      "epoch": 1.3912290825158684,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048175773785381947,
      "loss": 5.9271,
      "mean_token_accuracy": 0.18114378899335862,
      "num_tokens": 30563799.0,
      "step": 12055
    },
    {
      "entropy": 6.083360481262207,
      "epoch": 1.391806116560877,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048174145821792833,
      "loss": 5.9061,
      "mean_token_accuracy": 0.1791294515132904,
      "num_tokens": 30576637.0,
      "step": 12060
    },
    {
      "entropy": 6.126413488388062,
      "epoch": 1.3923831506058857,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004817251716283908,
      "loss": 6.0126,
      "mean_token_accuracy": 0.17122441232204438,
      "num_tokens": 30589720.0,
      "step": 12065
    },
    {
      "entropy": 6.1004551410675045,
      "epoch": 1.3929601846508943,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004817088780857548,
      "loss": 6.0141,
      "mean_token_accuracy": 0.17372102290391922,
      "num_tokens": 30602895.0,
      "step": 12070
    },
    {
      "entropy": 6.0461390018463135,
      "epoch": 1.393537218695903,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048169257759056845,
      "loss": 5.923,
      "mean_token_accuracy": 0.18145256340503693,
      "num_tokens": 30615411.0,
      "step": 12075
    },
    {
      "entropy": 6.08439884185791,
      "epoch": 1.3941142527409118,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048167627014337994,
      "loss": 5.9294,
      "mean_token_accuracy": 0.17341874837875365,
      "num_tokens": 30627682.0,
      "step": 12080
    },
    {
      "entropy": 6.101941537857056,
      "epoch": 1.3946912867859202,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00048165995574473764,
      "loss": 6.0609,
      "mean_token_accuracy": 0.16635080575942993,
      "num_tokens": 30641723.0,
      "step": 12085
    },
    {
      "entropy": 6.1498425006866455,
      "epoch": 1.395268320830929,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048164363439519043,
      "loss": 6.0061,
      "mean_token_accuracy": 0.170177860558033,
      "num_tokens": 30654926.0,
      "step": 12090
    },
    {
      "entropy": 6.110022258758545,
      "epoch": 1.3958453548759377,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004816273060952873,
      "loss": 5.9537,
      "mean_token_accuracy": 0.1740071251988411,
      "num_tokens": 30667641.0,
      "step": 12095
    },
    {
      "entropy": 6.130577468872071,
      "epoch": 1.3964223889209464,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048161097084557726,
      "loss": 5.9536,
      "mean_token_accuracy": 0.17439624667167664,
      "num_tokens": 30680841.0,
      "step": 12100
    },
    {
      "entropy": 6.087655401229858,
      "epoch": 1.396999422965955,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048159462864660993,
      "loss": 5.9179,
      "mean_token_accuracy": 0.1790216714143753,
      "num_tokens": 30695829.0,
      "step": 12105
    },
    {
      "entropy": 6.0643167972564695,
      "epoch": 1.3975764570109637,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004815782794989348,
      "loss": 5.8669,
      "mean_token_accuracy": 0.17555442154407502,
      "num_tokens": 30707853.0,
      "step": 12110
    },
    {
      "entropy": 6.04755744934082,
      "epoch": 1.3981534910559723,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004815619234031019,
      "loss": 5.8595,
      "mean_token_accuracy": 0.17693040072917937,
      "num_tokens": 30720568.0,
      "step": 12115
    },
    {
      "entropy": 6.046544790267944,
      "epoch": 1.398730525100981,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004815455603596613,
      "loss": 5.9262,
      "mean_token_accuracy": 0.17478361278772353,
      "num_tokens": 30734282.0,
      "step": 12120
    },
    {
      "entropy": 6.130530214309692,
      "epoch": 1.3993075591459896,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004815291903691634,
      "loss": 5.9178,
      "mean_token_accuracy": 0.17716023325920105,
      "num_tokens": 30746186.0,
      "step": 12125
    },
    {
      "entropy": 6.050559616088867,
      "epoch": 1.3998845931909982,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048151281343215873,
      "loss": 5.9461,
      "mean_token_accuracy": 0.1730263724923134,
      "num_tokens": 30757697.0,
      "step": 12130
    },
    {
      "entropy": 6.092675638198853,
      "epoch": 1.4004616272360069,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004814964295491982,
      "loss": 5.9258,
      "mean_token_accuracy": 0.18009327352046967,
      "num_tokens": 30770068.0,
      "step": 12135
    },
    {
      "entropy": 6.049663639068603,
      "epoch": 1.4010386612810155,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048148003872083286,
      "loss": 5.852,
      "mean_token_accuracy": 0.18247767984867097,
      "num_tokens": 30782135.0,
      "step": 12140
    },
    {
      "entropy": 6.094912767410278,
      "epoch": 1.4016156953260244,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048146364094761384,
      "loss": 5.9835,
      "mean_token_accuracy": 0.17268931716680527,
      "num_tokens": 30793448.0,
      "step": 12145
    },
    {
      "entropy": 6.103264904022216,
      "epoch": 1.4021927293710328,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048144723623009297,
      "loss": 5.8902,
      "mean_token_accuracy": 0.17312564700841904,
      "num_tokens": 30806280.0,
      "step": 12150
    },
    {
      "entropy": 6.012092590332031,
      "epoch": 1.4027697634160416,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004814308245688218,
      "loss": 5.9074,
      "mean_token_accuracy": 0.17844756096601486,
      "num_tokens": 30820394.0,
      "step": 12155
    },
    {
      "entropy": 6.12517237663269,
      "epoch": 1.40334679746105,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048141440596435235,
      "loss": 5.9085,
      "mean_token_accuracy": 0.17849789559841156,
      "num_tokens": 30832343.0,
      "step": 12160
    },
    {
      "entropy": 6.139100551605225,
      "epoch": 1.403923831506059,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004813979804172369,
      "loss": 5.9456,
      "mean_token_accuracy": 0.17201208025217057,
      "num_tokens": 30844833.0,
      "step": 12165
    },
    {
      "entropy": 6.029835414886475,
      "epoch": 1.4045008655510676,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048138154792802795,
      "loss": 5.8691,
      "mean_token_accuracy": 0.17654473930597306,
      "num_tokens": 30856899.0,
      "step": 12170
    },
    {
      "entropy": 6.076024580001831,
      "epoch": 1.4050778995960762,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004813651084972781,
      "loss": 5.9183,
      "mean_token_accuracy": 0.17868135124444962,
      "num_tokens": 30869172.0,
      "step": 12175
    },
    {
      "entropy": 6.122739934921265,
      "epoch": 1.4056549336410848,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048134866212554036,
      "loss": 5.9137,
      "mean_token_accuracy": 0.17689942121505736,
      "num_tokens": 30880751.0,
      "step": 12180
    },
    {
      "entropy": 6.103661823272705,
      "epoch": 1.4062319676860935,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004813322088133679,
      "loss": 5.9223,
      "mean_token_accuracy": 0.16530330777168273,
      "num_tokens": 30893583.0,
      "step": 12185
    },
    {
      "entropy": 6.113063097000122,
      "epoch": 1.4068090017311021,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048131574856131407,
      "loss": 5.9882,
      "mean_token_accuracy": 0.17224605828523637,
      "num_tokens": 30906150.0,
      "step": 12190
    },
    {
      "entropy": 6.090886449813842,
      "epoch": 1.4073860357761108,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004812992813699324,
      "loss": 5.9463,
      "mean_token_accuracy": 0.17186954915523528,
      "num_tokens": 30918505.0,
      "step": 12195
    },
    {
      "entropy": 6.107962608337402,
      "epoch": 1.4079630698211194,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000481282807239777,
      "loss": 5.9581,
      "mean_token_accuracy": 0.17071547061204911,
      "num_tokens": 30932519.0,
      "step": 12200
    },
    {
      "entropy": 6.100613975524903,
      "epoch": 1.408540103866128,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004812663261714019,
      "loss": 5.971,
      "mean_token_accuracy": 0.171573406457901,
      "num_tokens": 30944482.0,
      "step": 12205
    },
    {
      "entropy": 6.073848533630371,
      "epoch": 1.4091171379111367,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004812498381653613,
      "loss": 5.8798,
      "mean_token_accuracy": 0.17763544768095016,
      "num_tokens": 30956726.0,
      "step": 12210
    },
    {
      "entropy": 6.025596809387207,
      "epoch": 1.4096941719561453,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004812333432222098,
      "loss": 5.8918,
      "mean_token_accuracy": 0.17340152710676193,
      "num_tokens": 30968557.0,
      "step": 12215
    },
    {
      "entropy": 6.125025510787964,
      "epoch": 1.4102712060011542,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004812168413425023,
      "loss": 5.9656,
      "mean_token_accuracy": 0.17555570751428604,
      "num_tokens": 30981442.0,
      "step": 12220
    },
    {
      "entropy": 6.070878648757935,
      "epoch": 1.4108482400461626,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048120033252679374,
      "loss": 5.8745,
      "mean_token_accuracy": 0.169204081594944,
      "num_tokens": 30994126.0,
      "step": 12225
    },
    {
      "entropy": 6.114383506774902,
      "epoch": 1.4114252740911715,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048118381677563946,
      "loss": 5.9251,
      "mean_token_accuracy": 0.1811545178294182,
      "num_tokens": 31006688.0,
      "step": 12230
    },
    {
      "entropy": 6.139276075363159,
      "epoch": 1.41200230813618,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004811672940895949,
      "loss": 6.0082,
      "mean_token_accuracy": 0.16701795160770416,
      "num_tokens": 31019922.0,
      "step": 12235
    },
    {
      "entropy": 6.164082145690918,
      "epoch": 1.4125793421811887,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004811507644692158,
      "loss": 6.0182,
      "mean_token_accuracy": 0.16474975049495696,
      "num_tokens": 31032068.0,
      "step": 12240
    },
    {
      "entropy": 6.067570495605469,
      "epoch": 1.4131563762261974,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004811342279150581,
      "loss": 5.8818,
      "mean_token_accuracy": 0.17567080855369568,
      "num_tokens": 31045547.0,
      "step": 12245
    },
    {
      "entropy": 6.092800903320312,
      "epoch": 1.413733410271206,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004811176844276781,
      "loss": 5.9826,
      "mean_token_accuracy": 0.17210223376750947,
      "num_tokens": 31059219.0,
      "step": 12250
    },
    {
      "entropy": 6.152126264572144,
      "epoch": 1.4143104443162147,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004811011340076322,
      "loss": 5.9732,
      "mean_token_accuracy": 0.1719807893037796,
      "num_tokens": 31072472.0,
      "step": 12255
    },
    {
      "entropy": 6.129370927810669,
      "epoch": 1.4148874783612233,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048108457665547695,
      "loss": 6.011,
      "mean_token_accuracy": 0.1647595539689064,
      "num_tokens": 31085394.0,
      "step": 12260
    },
    {
      "entropy": 6.194111442565918,
      "epoch": 1.415464512406232,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004810680123717694,
      "loss": 5.994,
      "mean_token_accuracy": 0.1681814044713974,
      "num_tokens": 31098651.0,
      "step": 12265
    },
    {
      "entropy": 6.056366443634033,
      "epoch": 1.4160415464512406,
      "grad_norm": 1.0,
      "learning_rate": 0.0004810514411570666,
      "loss": 5.9539,
      "mean_token_accuracy": 0.17994170337915422,
      "num_tokens": 31111297.0,
      "step": 12270
    },
    {
      "entropy": 6.046131134033203,
      "epoch": 1.4166185804962492,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004810348630119259,
      "loss": 5.9573,
      "mean_token_accuracy": 0.1732081577181816,
      "num_tokens": 31124247.0,
      "step": 12275
    },
    {
      "entropy": 6.1294591426849365,
      "epoch": 1.4171956145412579,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048101827793690493,
      "loss": 5.9117,
      "mean_token_accuracy": 0.17428311407566072,
      "num_tokens": 31136106.0,
      "step": 12280
    },
    {
      "entropy": 6.040399980545044,
      "epoch": 1.4177726485862667,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004810016859325615,
      "loss": 5.8285,
      "mean_token_accuracy": 0.18322467505931855,
      "num_tokens": 31148561.0,
      "step": 12285
    },
    {
      "entropy": 6.117893838882447,
      "epoch": 1.4183496826312751,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004809850869994537,
      "loss": 5.9559,
      "mean_token_accuracy": 0.16528882533311845,
      "num_tokens": 31162289.0,
      "step": 12290
    },
    {
      "entropy": 6.140533494949341,
      "epoch": 1.418926716676284,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004809684811381398,
      "loss": 5.9152,
      "mean_token_accuracy": 0.1732899785041809,
      "num_tokens": 31174013.0,
      "step": 12295
    },
    {
      "entropy": 5.999874639511108,
      "epoch": 1.4195037507212924,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004809518683491785,
      "loss": 5.8619,
      "mean_token_accuracy": 0.18032350093126298,
      "num_tokens": 31186299.0,
      "step": 12300
    },
    {
      "entropy": 6.094850254058838,
      "epoch": 1.4200807847663013,
      "grad_norm": 1.0,
      "learning_rate": 0.00048093524863312823,
      "loss": 5.9533,
      "mean_token_accuracy": 0.17016388922929765,
      "num_tokens": 31198458.0,
      "step": 12305
    },
    {
      "entropy": 6.1386736869812015,
      "epoch": 1.42065781881131,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004809186219905482,
      "loss": 6.0032,
      "mean_token_accuracy": 0.16769521534442902,
      "num_tokens": 31210712.0,
      "step": 12310
    },
    {
      "entropy": 6.069535541534424,
      "epoch": 1.4212348528563186,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004809019884219976,
      "loss": 5.8747,
      "mean_token_accuracy": 0.18012098968029022,
      "num_tokens": 31223460.0,
      "step": 12315
    },
    {
      "entropy": 5.985542058944702,
      "epoch": 1.4218118869013272,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048088534792803595,
      "loss": 5.9335,
      "mean_token_accuracy": 0.17480566799640657,
      "num_tokens": 31235298.0,
      "step": 12320
    },
    {
      "entropy": 6.1801306247711185,
      "epoch": 1.4223889209463358,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048086870050922286,
      "loss": 5.9634,
      "mean_token_accuracy": 0.16771591752767562,
      "num_tokens": 31247164.0,
      "step": 12325
    },
    {
      "entropy": 6.113706302642822,
      "epoch": 1.4229659549913445,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048085204616611833,
      "loss": 6.0013,
      "mean_token_accuracy": 0.17126764804124833,
      "num_tokens": 31259963.0,
      "step": 12330
    },
    {
      "entropy": 6.072208452224731,
      "epoch": 1.4235429890363531,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048083538489928246,
      "loss": 5.9327,
      "mean_token_accuracy": 0.17276113033294677,
      "num_tokens": 31272646.0,
      "step": 12335
    },
    {
      "entropy": 6.028815031051636,
      "epoch": 1.4241200230813618,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004808187167092757,
      "loss": 5.8601,
      "mean_token_accuracy": 0.18015699684619904,
      "num_tokens": 31286474.0,
      "step": 12340
    },
    {
      "entropy": 6.025919389724732,
      "epoch": 1.4246970571263704,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004808020415966586,
      "loss": 5.8205,
      "mean_token_accuracy": 0.17833875864744186,
      "num_tokens": 31300086.0,
      "step": 12345
    },
    {
      "entropy": 6.072927665710449,
      "epoch": 1.425274091171379,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000480785359561992,
      "loss": 5.8202,
      "mean_token_accuracy": 0.18715763986110687,
      "num_tokens": 31312562.0,
      "step": 12350
    },
    {
      "entropy": 6.108096647262573,
      "epoch": 1.4258511252163877,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048076867060583704,
      "loss": 5.9683,
      "mean_token_accuracy": 0.1724245309829712,
      "num_tokens": 31325200.0,
      "step": 12355
    },
    {
      "entropy": 6.031751346588135,
      "epoch": 1.4264281592613965,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004807519747287551,
      "loss": 5.9124,
      "mean_token_accuracy": 0.17800545543432236,
      "num_tokens": 31337361.0,
      "step": 12360
    },
    {
      "entropy": 6.033226394653321,
      "epoch": 1.427005193306405,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004807352719313078,
      "loss": 5.8964,
      "mean_token_accuracy": 0.17548862844705582,
      "num_tokens": 31351411.0,
      "step": 12365
    },
    {
      "entropy": 6.045888471603393,
      "epoch": 1.4275822273514138,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004807185622140567,
      "loss": 5.9114,
      "mean_token_accuracy": 0.1785559043288231,
      "num_tokens": 31364616.0,
      "step": 12370
    },
    {
      "entropy": 6.059982585906982,
      "epoch": 1.4281592613964225,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048070184557756396,
      "loss": 5.8679,
      "mean_token_accuracy": 0.1801156744360924,
      "num_tokens": 31376515.0,
      "step": 12375
    },
    {
      "entropy": 6.086584091186523,
      "epoch": 1.428736295441431,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048068512202239177,
      "loss": 5.9141,
      "mean_token_accuracy": 0.17790561318397521,
      "num_tokens": 31388408.0,
      "step": 12380
    },
    {
      "entropy": 6.043620443344116,
      "epoch": 1.4293133294864397,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004806683915491028,
      "loss": 5.9097,
      "mean_token_accuracy": 0.17231618165969848,
      "num_tokens": 31401001.0,
      "step": 12385
    },
    {
      "entropy": 6.081690406799316,
      "epoch": 1.4298903635314484,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004806516541582596,
      "loss": 5.8808,
      "mean_token_accuracy": 0.17417093515396118,
      "num_tokens": 31413259.0,
      "step": 12390
    },
    {
      "entropy": 6.201285266876221,
      "epoch": 1.430467397576457,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048063490985042506,
      "loss": 6.0619,
      "mean_token_accuracy": 0.1634480506181717,
      "num_tokens": 31425997.0,
      "step": 12395
    },
    {
      "entropy": 6.09340615272522,
      "epoch": 1.4310444316214657,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004806181586261626,
      "loss": 5.9724,
      "mean_token_accuracy": 0.16944300383329391,
      "num_tokens": 31439967.0,
      "step": 12400
    },
    {
      "entropy": 6.107439613342285,
      "epoch": 1.4316214656664743,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048060140048603544,
      "loss": 5.9375,
      "mean_token_accuracy": 0.17209382504224777,
      "num_tokens": 31453122.0,
      "step": 12405
    },
    {
      "entropy": 6.152525329589844,
      "epoch": 1.432198499711483,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004805846354306073,
      "loss": 5.9535,
      "mean_token_accuracy": 0.17317837178707124,
      "num_tokens": 31465043.0,
      "step": 12410
    },
    {
      "entropy": 6.021739912033081,
      "epoch": 1.4327755337564916,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048056786346044214,
      "loss": 5.8488,
      "mean_token_accuracy": 0.18081784397363662,
      "num_tokens": 31477003.0,
      "step": 12415
    },
    {
      "entropy": 6.055218601226807,
      "epoch": 1.4333525678015002,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048055108457610395,
      "loss": 5.967,
      "mean_token_accuracy": 0.17251208573579788,
      "num_tokens": 31491612.0,
      "step": 12420
    },
    {
      "entropy": 6.08023886680603,
      "epoch": 1.4339296018465089,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004805342987781571,
      "loss": 5.914,
      "mean_token_accuracy": 0.17360394299030305,
      "num_tokens": 31504488.0,
      "step": 12425
    },
    {
      "entropy": 6.111648464202881,
      "epoch": 1.4345066358915175,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004805175060671663,
      "loss": 6.0096,
      "mean_token_accuracy": 0.16792135685682297,
      "num_tokens": 31517770.0,
      "step": 12430
    },
    {
      "entropy": 6.005486726760864,
      "epoch": 1.4350836699365264,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004805007064436962,
      "loss": 5.7384,
      "mean_token_accuracy": 0.17957783937454225,
      "num_tokens": 31530262.0,
      "step": 12435
    },
    {
      "entropy": 6.022184038162232,
      "epoch": 1.4356607039815348,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004804838999083119,
      "loss": 5.8976,
      "mean_token_accuracy": 0.16769999414682388,
      "num_tokens": 31543988.0,
      "step": 12440
    },
    {
      "entropy": 5.980929231643676,
      "epoch": 1.4362377380265436,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004804670864615787,
      "loss": 5.8708,
      "mean_token_accuracy": 0.17612825334072113,
      "num_tokens": 31555690.0,
      "step": 12445
    },
    {
      "entropy": 6.01245608329773,
      "epoch": 1.4368147720715523,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048045026610406223,
      "loss": 5.9402,
      "mean_token_accuracy": 0.1745675638318062,
      "num_tokens": 31568542.0,
      "step": 12450
    },
    {
      "entropy": 6.082572984695434,
      "epoch": 1.437391806116561,
      "grad_norm": 0.90625,
      "learning_rate": 0.000480433438836328,
      "loss": 5.9718,
      "mean_token_accuracy": 0.17191985994577408,
      "num_tokens": 31580538.0,
      "step": 12455
    },
    {
      "entropy": 6.126201915740967,
      "epoch": 1.4379688401615696,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048041660465894206,
      "loss": 5.8816,
      "mean_token_accuracy": 0.1763559341430664,
      "num_tokens": 31593414.0,
      "step": 12460
    },
    {
      "entropy": 6.107641649246216,
      "epoch": 1.4385458742065782,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004803997635724707,
      "loss": 5.9404,
      "mean_token_accuracy": 0.171015864610672,
      "num_tokens": 31605931.0,
      "step": 12465
    },
    {
      "entropy": 6.090613746643067,
      "epoch": 1.4391229082515868,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004803829155774804,
      "loss": 5.8653,
      "mean_token_accuracy": 0.17159179002046585,
      "num_tokens": 31617949.0,
      "step": 12470
    },
    {
      "entropy": 6.048834037780762,
      "epoch": 1.4396999422965955,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004803660606745377,
      "loss": 5.9434,
      "mean_token_accuracy": 0.17393881678581238,
      "num_tokens": 31630429.0,
      "step": 12475
    },
    {
      "entropy": 6.029767942428589,
      "epoch": 1.4402769763416041,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048034919886420953,
      "loss": 5.9174,
      "mean_token_accuracy": 0.17770526558160782,
      "num_tokens": 31641430.0,
      "step": 12480
    },
    {
      "entropy": 6.147845506668091,
      "epoch": 1.4408540103866128,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048033233014706304,
      "loss": 6.0341,
      "mean_token_accuracy": 0.16704342365264893,
      "num_tokens": 31653465.0,
      "step": 12485
    },
    {
      "entropy": 6.126414585113525,
      "epoch": 1.4414310444316214,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048031545452366565,
      "loss": 5.9194,
      "mean_token_accuracy": 0.17453492432832718,
      "num_tokens": 31665550.0,
      "step": 12490
    },
    {
      "entropy": 5.982351636886596,
      "epoch": 1.44200807847663,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048029857199458493,
      "loss": 5.9174,
      "mean_token_accuracy": 0.18255743384361267,
      "num_tokens": 31677991.0,
      "step": 12495
    },
    {
      "entropy": 6.130622339248657,
      "epoch": 1.442585112521639,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048028168256038873,
      "loss": 5.9453,
      "mean_token_accuracy": 0.17678724825382233,
      "num_tokens": 31690148.0,
      "step": 12500
    },
    {
      "entropy": 6.130819797515869,
      "epoch": 1.4431621465666473,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048026478622164513,
      "loss": 5.9802,
      "mean_token_accuracy": 0.17068534791469575,
      "num_tokens": 31703632.0,
      "step": 12505
    },
    {
      "entropy": 6.071145629882812,
      "epoch": 1.4437391806116562,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048024788297892234,
      "loss": 5.9257,
      "mean_token_accuracy": 0.17986937016248702,
      "num_tokens": 31714722.0,
      "step": 12510
    },
    {
      "entropy": 6.119812059402466,
      "epoch": 1.4443162146566648,
      "grad_norm": 1.015625,
      "learning_rate": 0.000480230972832789,
      "loss": 5.8896,
      "mean_token_accuracy": 0.17709065079689026,
      "num_tokens": 31726554.0,
      "step": 12515
    },
    {
      "entropy": 6.180822277069092,
      "epoch": 1.4448932487016735,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048021405578381384,
      "loss": 6.021,
      "mean_token_accuracy": 0.1725223705172539,
      "num_tokens": 31738604.0,
      "step": 12520
    },
    {
      "entropy": 5.9732238292694095,
      "epoch": 1.445470282746682,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004801971318325659,
      "loss": 5.8802,
      "mean_token_accuracy": 0.18279252499341964,
      "num_tokens": 31750084.0,
      "step": 12525
    },
    {
      "entropy": 6.084851169586182,
      "epoch": 1.4460473167916907,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004801802009796142,
      "loss": 5.9157,
      "mean_token_accuracy": 0.17345007210969926,
      "num_tokens": 31764068.0,
      "step": 12530
    },
    {
      "entropy": 6.040093803405762,
      "epoch": 1.4466243508366994,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004801632632255285,
      "loss": 5.8797,
      "mean_token_accuracy": 0.17544397711753845,
      "num_tokens": 31777425.0,
      "step": 12535
    },
    {
      "entropy": 6.0521464347839355,
      "epoch": 1.447201384881708,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004801463185708783,
      "loss": 5.9125,
      "mean_token_accuracy": 0.17761372178792953,
      "num_tokens": 31789767.0,
      "step": 12540
    },
    {
      "entropy": 5.992119264602661,
      "epoch": 1.4477784189267167,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048012936701623363,
      "loss": 5.8343,
      "mean_token_accuracy": 0.18532239496707917,
      "num_tokens": 31801399.0,
      "step": 12545
    },
    {
      "entropy": 6.060655021667481,
      "epoch": 1.4483554529717253,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048011240856216456,
      "loss": 5.9441,
      "mean_token_accuracy": 0.1721680074930191,
      "num_tokens": 31814260.0,
      "step": 12550
    },
    {
      "entropy": 6.135332250595093,
      "epoch": 1.448932487016734,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048009544320924154,
      "loss": 5.9015,
      "mean_token_accuracy": 0.1749567672610283,
      "num_tokens": 31826674.0,
      "step": 12555
    },
    {
      "entropy": 6.12035870552063,
      "epoch": 1.4495095210617426,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004800784709580351,
      "loss": 6.0311,
      "mean_token_accuracy": 0.16694096177816392,
      "num_tokens": 31839951.0,
      "step": 12560
    },
    {
      "entropy": 6.0713475227355955,
      "epoch": 1.4500865551067512,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004800614918091161,
      "loss": 5.8913,
      "mean_token_accuracy": 0.17824322879314422,
      "num_tokens": 31852344.0,
      "step": 12565
    },
    {
      "entropy": 6.022281169891357,
      "epoch": 1.4506635891517599,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004800445057630558,
      "loss": 5.8266,
      "mean_token_accuracy": 0.18461204171180726,
      "num_tokens": 31865134.0,
      "step": 12570
    },
    {
      "entropy": 5.999623250961304,
      "epoch": 1.4512406231967687,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004800275128204254,
      "loss": 5.8637,
      "mean_token_accuracy": 0.18167271465063095,
      "num_tokens": 31877395.0,
      "step": 12575
    },
    {
      "entropy": 6.090144491195678,
      "epoch": 1.4518176572417771,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048001051298179637,
      "loss": 5.8883,
      "mean_token_accuracy": 0.17698944061994554,
      "num_tokens": 31889421.0,
      "step": 12580
    },
    {
      "entropy": 6.109578275680542,
      "epoch": 1.452394691286786,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004799935062477407,
      "loss": 5.9428,
      "mean_token_accuracy": 0.1730349436402321,
      "num_tokens": 31903004.0,
      "step": 12585
    },
    {
      "entropy": 6.124676370620728,
      "epoch": 1.4529717253317946,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047997649261883013,
      "loss": 5.933,
      "mean_token_accuracy": 0.17646005302667617,
      "num_tokens": 31914682.0,
      "step": 12590
    },
    {
      "entropy": 6.168044996261597,
      "epoch": 1.4535487593768033,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004799594720956371,
      "loss": 5.9614,
      "mean_token_accuracy": 0.17037569880485534,
      "num_tokens": 31928861.0,
      "step": 12595
    },
    {
      "entropy": 6.008680200576782,
      "epoch": 1.454125793421812,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047994244467873407,
      "loss": 5.9108,
      "mean_token_accuracy": 0.17247474044561387,
      "num_tokens": 31942043.0,
      "step": 12600
    },
    {
      "entropy": 6.135958194732666,
      "epoch": 1.4547028274668206,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047992541036869364,
      "loss": 5.9828,
      "mean_token_accuracy": 0.17243419140577315,
      "num_tokens": 31954453.0,
      "step": 12605
    },
    {
      "entropy": 6.066493129730224,
      "epoch": 1.4552798615118292,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004799083691660889,
      "loss": 5.9082,
      "mean_token_accuracy": 0.16739271432161332,
      "num_tokens": 31967998.0,
      "step": 12610
    },
    {
      "entropy": 6.057908248901367,
      "epoch": 1.4558568955568378,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004798913210714929,
      "loss": 5.9712,
      "mean_token_accuracy": 0.16803003549575807,
      "num_tokens": 31980445.0,
      "step": 12615
    },
    {
      "entropy": 6.038349866867065,
      "epoch": 1.4564339296018465,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047987426608547915,
      "loss": 5.9001,
      "mean_token_accuracy": 0.18142564296722413,
      "num_tokens": 31993706.0,
      "step": 12620
    },
    {
      "entropy": 6.109736728668213,
      "epoch": 1.4570109636468551,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004798572042086212,
      "loss": 5.8813,
      "mean_token_accuracy": 0.17724963277578354,
      "num_tokens": 32005189.0,
      "step": 12625
    },
    {
      "entropy": 6.048584461212158,
      "epoch": 1.4575879976918638,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047984013544149286,
      "loss": 5.9824,
      "mean_token_accuracy": 0.17215955555438994,
      "num_tokens": 32017883.0,
      "step": 12630
    },
    {
      "entropy": 6.03533787727356,
      "epoch": 1.4581650317368724,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047982305978466836,
      "loss": 5.9217,
      "mean_token_accuracy": 0.1753133624792099,
      "num_tokens": 32030701.0,
      "step": 12635
    },
    {
      "entropy": 6.042423391342163,
      "epoch": 1.4587420657818813,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047980597723872205,
      "loss": 5.9216,
      "mean_token_accuracy": 0.1705354019999504,
      "num_tokens": 32042726.0,
      "step": 12640
    },
    {
      "entropy": 6.086380243301392,
      "epoch": 1.4593190998268897,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004797888878042283,
      "loss": 5.8469,
      "mean_token_accuracy": 0.17806642204523088,
      "num_tokens": 32055158.0,
      "step": 12645
    },
    {
      "entropy": 6.061453056335449,
      "epoch": 1.4598961338718985,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047977179148176217,
      "loss": 5.9205,
      "mean_token_accuracy": 0.17423218935728074,
      "num_tokens": 32068416.0,
      "step": 12650
    },
    {
      "entropy": 6.112088394165039,
      "epoch": 1.4604731679169072,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004797546882718985,
      "loss": 5.931,
      "mean_token_accuracy": 0.17192683964967728,
      "num_tokens": 32081800.0,
      "step": 12655
    },
    {
      "entropy": 6.124772691726685,
      "epoch": 1.4610502019619158,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00047973757817521256,
      "loss": 5.9893,
      "mean_token_accuracy": 0.16981735080480576,
      "num_tokens": 32095134.0,
      "step": 12660
    },
    {
      "entropy": 5.9755542278289795,
      "epoch": 1.4616272360069245,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004797204611922799,
      "loss": 5.8488,
      "mean_token_accuracy": 0.17988146245479583,
      "num_tokens": 32106685.0,
      "step": 12665
    },
    {
      "entropy": 6.090330934524536,
      "epoch": 1.462204270051933,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047970333732367626,
      "loss": 5.9628,
      "mean_token_accuracy": 0.17426787465810775,
      "num_tokens": 32120072.0,
      "step": 12670
    },
    {
      "entropy": 6.133768606185913,
      "epoch": 1.4627813040969417,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047968620656997754,
      "loss": 5.8949,
      "mean_token_accuracy": 0.1740260064601898,
      "num_tokens": 32133205.0,
      "step": 12675
    },
    {
      "entropy": 6.02523889541626,
      "epoch": 1.4633583381419504,
      "grad_norm": 1.0,
      "learning_rate": 0.00047966906893175994,
      "loss": 5.8634,
      "mean_token_accuracy": 0.18402136713266373,
      "num_tokens": 32147674.0,
      "step": 12680
    },
    {
      "entropy": 6.076717472076416,
      "epoch": 1.463935372186959,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004796519244095998,
      "loss": 5.9017,
      "mean_token_accuracy": 0.17213110327720643,
      "num_tokens": 32159689.0,
      "step": 12685
    },
    {
      "entropy": 6.069744682312011,
      "epoch": 1.4645124062319677,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047963477300407394,
      "loss": 5.9537,
      "mean_token_accuracy": 0.17260026782751084,
      "num_tokens": 32172641.0,
      "step": 12690
    },
    {
      "entropy": 6.075920343399048,
      "epoch": 1.4650894402769763,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047961761471575903,
      "loss": 5.925,
      "mean_token_accuracy": 0.17798299342393875,
      "num_tokens": 32184512.0,
      "step": 12695
    },
    {
      "entropy": 6.088216066360474,
      "epoch": 1.465666474321985,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047960044954523237,
      "loss": 5.9096,
      "mean_token_accuracy": 0.18035248965024947,
      "num_tokens": 32195908.0,
      "step": 12700
    },
    {
      "entropy": 6.06147346496582,
      "epoch": 1.4662435083669936,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047958327749307117,
      "loss": 5.9005,
      "mean_token_accuracy": 0.1695326089859009,
      "num_tokens": 32208855.0,
      "step": 12705
    },
    {
      "entropy": 6.061979961395264,
      "epoch": 1.4668205424120022,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047956609855985305,
      "loss": 5.9113,
      "mean_token_accuracy": 0.17111039906740189,
      "num_tokens": 32222769.0,
      "step": 12710
    },
    {
      "entropy": 6.143628454208374,
      "epoch": 1.467397576457011,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004795489127461559,
      "loss": 5.9294,
      "mean_token_accuracy": 0.17359299808740616,
      "num_tokens": 32235192.0,
      "step": 12715
    },
    {
      "entropy": 6.0017719745635985,
      "epoch": 1.4679746105020195,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047953172005255756,
      "loss": 5.9158,
      "mean_token_accuracy": 0.17549984008073807,
      "num_tokens": 32249196.0,
      "step": 12720
    },
    {
      "entropy": 6.177342748641967,
      "epoch": 1.4685516445470284,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004795145204796365,
      "loss": 5.975,
      "mean_token_accuracy": 0.17217395901679994,
      "num_tokens": 32261591.0,
      "step": 12725
    },
    {
      "entropy": 6.118365526199341,
      "epoch": 1.469128678592037,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004794973140279711,
      "loss": 5.9862,
      "mean_token_accuracy": 0.16865849643945693,
      "num_tokens": 32275409.0,
      "step": 12730
    },
    {
      "entropy": 6.106164216995239,
      "epoch": 1.4697057126370456,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047948010069814005,
      "loss": 5.8101,
      "mean_token_accuracy": 0.18006059527397156,
      "num_tokens": 32287897.0,
      "step": 12735
    },
    {
      "entropy": 6.064902257919312,
      "epoch": 1.4702827466820543,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004794628804907225,
      "loss": 6.0175,
      "mean_token_accuracy": 0.1707676187157631,
      "num_tokens": 32300595.0,
      "step": 12740
    },
    {
      "entropy": 6.054769086837768,
      "epoch": 1.470859780727063,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047944565340629755,
      "loss": 5.8923,
      "mean_token_accuracy": 0.1801082119345665,
      "num_tokens": 32312374.0,
      "step": 12745
    },
    {
      "entropy": 6.104706621170044,
      "epoch": 1.4714368147720716,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047942841944544453,
      "loss": 5.8902,
      "mean_token_accuracy": 0.17718605399131776,
      "num_tokens": 32326526.0,
      "step": 12750
    },
    {
      "entropy": 6.110076475143432,
      "epoch": 1.4720138488170802,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004794111786087431,
      "loss": 5.9776,
      "mean_token_accuracy": 0.16989217698574066,
      "num_tokens": 32339591.0,
      "step": 12755
    },
    {
      "entropy": 6.079367208480835,
      "epoch": 1.4725908828620888,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004793939308967733,
      "loss": 5.9315,
      "mean_token_accuracy": 0.17519567310810089,
      "num_tokens": 32352802.0,
      "step": 12760
    },
    {
      "entropy": 6.036862087249756,
      "epoch": 1.4731679169070975,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004793766763101152,
      "loss": 5.9029,
      "mean_token_accuracy": 0.16931188404560088,
      "num_tokens": 32365951.0,
      "step": 12765
    },
    {
      "entropy": 6.124979448318482,
      "epoch": 1.4737449509521061,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047935941484934905,
      "loss": 5.9483,
      "mean_token_accuracy": 0.1734633281826973,
      "num_tokens": 32378573.0,
      "step": 12770
    },
    {
      "entropy": 6.071750497817993,
      "epoch": 1.4743219849971148,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047934214651505547,
      "loss": 5.8949,
      "mean_token_accuracy": 0.17485958933830262,
      "num_tokens": 32392135.0,
      "step": 12775
    },
    {
      "entropy": 6.070564937591553,
      "epoch": 1.4748990190421236,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047932487130781533,
      "loss": 5.9993,
      "mean_token_accuracy": 0.17158770114183425,
      "num_tokens": 32404821.0,
      "step": 12780
    },
    {
      "entropy": 6.031760597229004,
      "epoch": 1.475476053087132,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004793075892282097,
      "loss": 5.8877,
      "mean_token_accuracy": 0.1766712933778763,
      "num_tokens": 32417118.0,
      "step": 12785
    },
    {
      "entropy": 6.091771745681763,
      "epoch": 1.476053087132141,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004792903002768197,
      "loss": 5.8588,
      "mean_token_accuracy": 0.18056404143571853,
      "num_tokens": 32428272.0,
      "step": 12790
    },
    {
      "entropy": 6.05395131111145,
      "epoch": 1.4766301211771495,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047927300445422687,
      "loss": 5.9479,
      "mean_token_accuracy": 0.17518570572137832,
      "num_tokens": 32441038.0,
      "step": 12795
    },
    {
      "entropy": 6.016059398651123,
      "epoch": 1.4772071552221582,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004792557017610131,
      "loss": 5.8377,
      "mean_token_accuracy": 0.1764357417821884,
      "num_tokens": 32452959.0,
      "step": 12800
    },
    {
      "entropy": 6.067819738388062,
      "epoch": 1.4777841892671668,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047923839219776027,
      "loss": 5.9653,
      "mean_token_accuracy": 0.17642948925495147,
      "num_tokens": 32465398.0,
      "step": 12805
    },
    {
      "entropy": 6.034655380249023,
      "epoch": 1.4783612233121755,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004792210757650506,
      "loss": 5.8298,
      "mean_token_accuracy": 0.17612256109714508,
      "num_tokens": 32477128.0,
      "step": 12810
    },
    {
      "entropy": 6.040339946746826,
      "epoch": 1.478938257357184,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047920375246346636,
      "loss": 5.9158,
      "mean_token_accuracy": 0.178180530667305,
      "num_tokens": 32489831.0,
      "step": 12815
    },
    {
      "entropy": 6.105113172531128,
      "epoch": 1.4795152914021927,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047918642229359044,
      "loss": 5.9781,
      "mean_token_accuracy": 0.1660287633538246,
      "num_tokens": 32502239.0,
      "step": 12820
    },
    {
      "entropy": 6.09044919013977,
      "epoch": 1.4800923254472014,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004791690852560056,
      "loss": 5.9173,
      "mean_token_accuracy": 0.17184444069862365,
      "num_tokens": 32515374.0,
      "step": 12825
    },
    {
      "entropy": 6.079788398742676,
      "epoch": 1.48066935949221,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000479151741351295,
      "loss": 5.9398,
      "mean_token_accuracy": 0.1705206051468849,
      "num_tokens": 32527208.0,
      "step": 12830
    },
    {
      "entropy": 6.1947925090789795,
      "epoch": 1.4812463935372187,
      "grad_norm": 1.015625,
      "learning_rate": 0.000479134390580042,
      "loss": 6.0286,
      "mean_token_accuracy": 0.16896121352910995,
      "num_tokens": 32540834.0,
      "step": 12835
    },
    {
      "entropy": 6.028968477249146,
      "epoch": 1.4818234275822273,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047911703294283015,
      "loss": 5.8551,
      "mean_token_accuracy": 0.1789850190281868,
      "num_tokens": 32552833.0,
      "step": 12840
    },
    {
      "entropy": 5.935985326766968,
      "epoch": 1.482400461627236,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047909966844024334,
      "loss": 5.8614,
      "mean_token_accuracy": 0.17835307270288467,
      "num_tokens": 32565484.0,
      "step": 12845
    },
    {
      "entropy": 6.133987188339233,
      "epoch": 1.4829774956722446,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047908229707286547,
      "loss": 6.0085,
      "mean_token_accuracy": 0.1672040656208992,
      "num_tokens": 32579099.0,
      "step": 12850
    },
    {
      "entropy": 6.137663698196411,
      "epoch": 1.4835545297172534,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000479064918841281,
      "loss": 5.872,
      "mean_token_accuracy": 0.17899580299854279,
      "num_tokens": 32591530.0,
      "step": 12855
    },
    {
      "entropy": 6.078830862045288,
      "epoch": 1.4841315637622619,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047904753374607427,
      "loss": 6.0172,
      "mean_token_accuracy": 0.17025664150714875,
      "num_tokens": 32603980.0,
      "step": 12860
    },
    {
      "entropy": 6.118191719055176,
      "epoch": 1.4847085978072707,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047903014178783015,
      "loss": 5.9256,
      "mean_token_accuracy": 0.1736728996038437,
      "num_tokens": 32617593.0,
      "step": 12865
    },
    {
      "entropy": 6.135264778137207,
      "epoch": 1.4852856318522794,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004790127429671335,
      "loss": 5.9086,
      "mean_token_accuracy": 0.1773558095097542,
      "num_tokens": 32629813.0,
      "step": 12870
    },
    {
      "entropy": 6.055281257629394,
      "epoch": 1.485862665897288,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004789953372845697,
      "loss": 5.9681,
      "mean_token_accuracy": 0.17167186439037324,
      "num_tokens": 32640982.0,
      "step": 12875
    },
    {
      "entropy": 6.02922248840332,
      "epoch": 1.4864396999422966,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000478977924740724,
      "loss": 5.836,
      "mean_token_accuracy": 0.17965785712003707,
      "num_tokens": 32652717.0,
      "step": 12880
    },
    {
      "entropy": 6.097882890701294,
      "epoch": 1.4870167339873053,
      "grad_norm": 1.0,
      "learning_rate": 0.0004789605053361821,
      "loss": 5.9338,
      "mean_token_accuracy": 0.17405622601509094,
      "num_tokens": 32665531.0,
      "step": 12885
    },
    {
      "entropy": 6.071781063079834,
      "epoch": 1.487593768032314,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004789430790715299,
      "loss": 5.8954,
      "mean_token_accuracy": 0.1777450382709503,
      "num_tokens": 32679296.0,
      "step": 12890
    },
    {
      "entropy": 6.091601991653443,
      "epoch": 1.4881708020773226,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047892564594735355,
      "loss": 5.8913,
      "mean_token_accuracy": 0.18283336907625197,
      "num_tokens": 32691806.0,
      "step": 12895
    },
    {
      "entropy": 5.998292446136475,
      "epoch": 1.4887478361223312,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047890820596423943,
      "loss": 5.8912,
      "mean_token_accuracy": 0.1796869471669197,
      "num_tokens": 32704280.0,
      "step": 12900
    },
    {
      "entropy": 6.0124204635620115,
      "epoch": 1.4893248701673398,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000478890759122774,
      "loss": 5.8817,
      "mean_token_accuracy": 0.17614697515964509,
      "num_tokens": 32716101.0,
      "step": 12905
    },
    {
      "entropy": 6.135642004013062,
      "epoch": 1.4899019042123485,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004788733054235443,
      "loss": 5.9454,
      "mean_token_accuracy": 0.17057950496673585,
      "num_tokens": 32728063.0,
      "step": 12910
    },
    {
      "entropy": 6.066538667678833,
      "epoch": 1.4904789382573571,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047885584486713717,
      "loss": 5.8709,
      "mean_token_accuracy": 0.17541065961122512,
      "num_tokens": 32742084.0,
      "step": 12915
    },
    {
      "entropy": 6.05385274887085,
      "epoch": 1.491055972302366,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004788383774541399,
      "loss": 5.9053,
      "mean_token_accuracy": 0.1840200886130333,
      "num_tokens": 32754708.0,
      "step": 12920
    },
    {
      "entropy": 6.064989852905273,
      "epoch": 1.4916330063473744,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004788209031851402,
      "loss": 5.8777,
      "mean_token_accuracy": 0.17674598097801208,
      "num_tokens": 32767326.0,
      "step": 12925
    },
    {
      "entropy": 6.076388311386109,
      "epoch": 1.4922100403923833,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004788034220607256,
      "loss": 5.9667,
      "mean_token_accuracy": 0.17857249826192856,
      "num_tokens": 32781334.0,
      "step": 12930
    },
    {
      "entropy": 6.112741136550904,
      "epoch": 1.4927870744373917,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047878593408148405,
      "loss": 5.9786,
      "mean_token_accuracy": 0.18228708803653718,
      "num_tokens": 32793795.0,
      "step": 12935
    },
    {
      "entropy": 6.053544902801514,
      "epoch": 1.4933641084824005,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047876843924800393,
      "loss": 5.8636,
      "mean_token_accuracy": 0.1786147654056549,
      "num_tokens": 32807484.0,
      "step": 12940
    },
    {
      "entropy": 5.996946811676025,
      "epoch": 1.4939411425274092,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004787509375608735,
      "loss": 5.8966,
      "mean_token_accuracy": 0.18170347958803176,
      "num_tokens": 32819129.0,
      "step": 12945
    },
    {
      "entropy": 6.0466532707214355,
      "epoch": 1.4945181765724178,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047873342902068157,
      "loss": 5.9213,
      "mean_token_accuracy": 0.1744529649615288,
      "num_tokens": 32832077.0,
      "step": 12950
    },
    {
      "entropy": 6.038098955154419,
      "epoch": 1.4950952106174265,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047871591362801694,
      "loss": 5.9135,
      "mean_token_accuracy": 0.17597944736480714,
      "num_tokens": 32845629.0,
      "step": 12955
    },
    {
      "entropy": 6.0727417945861815,
      "epoch": 1.495672244662435,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004786983913834687,
      "loss": 5.9032,
      "mean_token_accuracy": 0.1805686980485916,
      "num_tokens": 32857233.0,
      "step": 12960
    },
    {
      "entropy": 6.060959100723267,
      "epoch": 1.4962492787074437,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047868086228762633,
      "loss": 5.9399,
      "mean_token_accuracy": 0.1719135895371437,
      "num_tokens": 32869210.0,
      "step": 12965
    },
    {
      "entropy": 6.089103317260742,
      "epoch": 1.4968263127524524,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047866332634107926,
      "loss": 5.9841,
      "mean_token_accuracy": 0.17466243952512742,
      "num_tokens": 32881633.0,
      "step": 12970
    },
    {
      "entropy": 6.159843778610229,
      "epoch": 1.497403346797461,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047864578354441743,
      "loss": 5.8655,
      "mean_token_accuracy": 0.1833904892206192,
      "num_tokens": 32893904.0,
      "step": 12975
    },
    {
      "entropy": 6.034606647491455,
      "epoch": 1.4979803808424696,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004786282338982308,
      "loss": 5.9194,
      "mean_token_accuracy": 0.17909137159585953,
      "num_tokens": 32906612.0,
      "step": 12980
    },
    {
      "entropy": 6.079008007049561,
      "epoch": 1.4985574148874783,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004786106774031096,
      "loss": 5.9742,
      "mean_token_accuracy": 0.1760239914059639,
      "num_tokens": 32919677.0,
      "step": 12985
    },
    {
      "entropy": 6.048500728607178,
      "epoch": 1.499134448932487,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004785931140596445,
      "loss": 5.8568,
      "mean_token_accuracy": 0.17758766263723375,
      "num_tokens": 32932385.0,
      "step": 12990
    },
    {
      "entropy": 6.032617425918579,
      "epoch": 1.4997114829774958,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047857554386842606,
      "loss": 5.8916,
      "mean_token_accuracy": 0.18109066933393478,
      "num_tokens": 32945524.0,
      "step": 12995
    },
    {
      "entropy": 6.068087816238403,
      "epoch": 1.5002885170225042,
      "grad_norm": 1.0,
      "learning_rate": 0.00047855796683004534,
      "loss": 5.894,
      "mean_token_accuracy": 0.1754665270447731,
      "num_tokens": 32957871.0,
      "step": 13000
    },
    {
      "entropy": 6.0700243473052975,
      "epoch": 1.500865551067513,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047854038294509356,
      "loss": 5.9615,
      "mean_token_accuracy": 0.17432401776313783,
      "num_tokens": 32971727.0,
      "step": 13005
    },
    {
      "entropy": 6.0010889053344725,
      "epoch": 1.5014425851125215,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004785227922141621,
      "loss": 5.839,
      "mean_token_accuracy": 0.18153130859136582,
      "num_tokens": 32984884.0,
      "step": 13010
    },
    {
      "entropy": 6.076522350311279,
      "epoch": 1.5020196191575304,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047850519463784263,
      "loss": 5.9608,
      "mean_token_accuracy": 0.17547461539506912,
      "num_tokens": 32997475.0,
      "step": 13015
    },
    {
      "entropy": 6.097279262542725,
      "epoch": 1.502596653202539,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047848759021672693,
      "loss": 5.9537,
      "mean_token_accuracy": 0.17340280413627623,
      "num_tokens": 33010466.0,
      "step": 13020
    },
    {
      "entropy": 6.1239526748657225,
      "epoch": 1.5031736872475476,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004784699789514073,
      "loss": 5.9377,
      "mean_token_accuracy": 0.17342619746923446,
      "num_tokens": 33023806.0,
      "step": 13025
    },
    {
      "entropy": 6.101027727127075,
      "epoch": 1.5037507212925563,
      "grad_norm": 0.96875,
      "learning_rate": 0.000478452360842476,
      "loss": 5.8771,
      "mean_token_accuracy": 0.1744084894657135,
      "num_tokens": 33036132.0,
      "step": 13030
    },
    {
      "entropy": 6.047265863418579,
      "epoch": 1.504327755337565,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047843473589052557,
      "loss": 5.9701,
      "mean_token_accuracy": 0.17032059878110886,
      "num_tokens": 33049706.0,
      "step": 13035
    },
    {
      "entropy": 6.120639133453369,
      "epoch": 1.5049047893825735,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047841710409614893,
      "loss": 5.881,
      "mean_token_accuracy": 0.17759590297937394,
      "num_tokens": 33061725.0,
      "step": 13040
    },
    {
      "entropy": 6.1116883754730225,
      "epoch": 1.5054818234275822,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004783994654599389,
      "loss": 5.8988,
      "mean_token_accuracy": 0.18071457743644714,
      "num_tokens": 33074908.0,
      "step": 13045
    },
    {
      "entropy": 6.005496501922607,
      "epoch": 1.5060588574725908,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047838181998248897,
      "loss": 5.9261,
      "mean_token_accuracy": 0.17357082664966583,
      "num_tokens": 33086631.0,
      "step": 13050
    },
    {
      "entropy": 6.084824085235596,
      "epoch": 1.5066358915175995,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004783641676643925,
      "loss": 5.9254,
      "mean_token_accuracy": 0.17593357264995574,
      "num_tokens": 33100109.0,
      "step": 13055
    },
    {
      "entropy": 6.086726665496826,
      "epoch": 1.5072129255626083,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047834650850624334,
      "loss": 5.8628,
      "mean_token_accuracy": 0.17900268584489823,
      "num_tokens": 33111953.0,
      "step": 13060
    },
    {
      "entropy": 6.024261379241944,
      "epoch": 1.5077899596076167,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004783288425086353,
      "loss": 5.8619,
      "mean_token_accuracy": 0.18280904293060302,
      "num_tokens": 33124563.0,
      "step": 13065
    },
    {
      "entropy": 6.122851181030273,
      "epoch": 1.5083669936526256,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004783111696721627,
      "loss": 5.9629,
      "mean_token_accuracy": 0.17729966193437577,
      "num_tokens": 33137636.0,
      "step": 13070
    },
    {
      "entropy": 6.127305793762207,
      "epoch": 1.508944027697634,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004782934899974199,
      "loss": 6.0132,
      "mean_token_accuracy": 0.17273994535207748,
      "num_tokens": 33150209.0,
      "step": 13075
    },
    {
      "entropy": 5.928384447097779,
      "epoch": 1.5095210617426429,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047827580348500147,
      "loss": 5.7964,
      "mean_token_accuracy": 0.18683600425720215,
      "num_tokens": 33162424.0,
      "step": 13080
    },
    {
      "entropy": 6.027935552597046,
      "epoch": 1.5100980957876513,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047825811013550246,
      "loss": 5.8902,
      "mean_token_accuracy": 0.18491662740707399,
      "num_tokens": 33174923.0,
      "step": 13085
    },
    {
      "entropy": 6.072844457626343,
      "epoch": 1.5106751298326602,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047824040994951786,
      "loss": 5.8023,
      "mean_token_accuracy": 0.17709645628929138,
      "num_tokens": 33188022.0,
      "step": 13090
    },
    {
      "entropy": 6.069659566879272,
      "epoch": 1.5112521638776688,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004782227029276429,
      "loss": 5.9235,
      "mean_token_accuracy": 0.17771101891994476,
      "num_tokens": 33201657.0,
      "step": 13095
    },
    {
      "entropy": 6.096413516998291,
      "epoch": 1.5118291979226774,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047820498907047345,
      "loss": 5.9055,
      "mean_token_accuracy": 0.17591238021850586,
      "num_tokens": 33214692.0,
      "step": 13100
    },
    {
      "entropy": 6.079206085205078,
      "epoch": 1.512406231967686,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000478187268378605,
      "loss": 5.9304,
      "mean_token_accuracy": 0.17662953436374665,
      "num_tokens": 33227419.0,
      "step": 13105
    },
    {
      "entropy": 6.138459825515747,
      "epoch": 1.5129832660126947,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047816954085263375,
      "loss": 5.9331,
      "mean_token_accuracy": 0.18027483820915222,
      "num_tokens": 33239859.0,
      "step": 13110
    },
    {
      "entropy": 6.0682234287261965,
      "epoch": 1.5135603000577034,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004781518064931559,
      "loss": 5.9052,
      "mean_token_accuracy": 0.17642598301172258,
      "num_tokens": 33252872.0,
      "step": 13115
    },
    {
      "entropy": 6.042830610275269,
      "epoch": 1.514137334102712,
      "grad_norm": 0.96875,
      "learning_rate": 0.000478134065300768,
      "loss": 5.832,
      "mean_token_accuracy": 0.18356831669807433,
      "num_tokens": 33264220.0,
      "step": 13120
    },
    {
      "entropy": 6.087376546859741,
      "epoch": 1.5147143681477209,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004781163172760667,
      "loss": 6.001,
      "mean_token_accuracy": 0.1670086994767189,
      "num_tokens": 33276379.0,
      "step": 13125
    },
    {
      "entropy": 6.016605377197266,
      "epoch": 1.5152914021927293,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047809856241964893,
      "loss": 5.8613,
      "mean_token_accuracy": 0.17730607092380524,
      "num_tokens": 33290487.0,
      "step": 13130
    },
    {
      "entropy": 6.133427858352661,
      "epoch": 1.5158684362377381,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004780808007321119,
      "loss": 6.0191,
      "mean_token_accuracy": 0.16540180742740632,
      "num_tokens": 33302560.0,
      "step": 13135
    },
    {
      "entropy": 6.107445812225341,
      "epoch": 1.5164454702827466,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004780630322140531,
      "loss": 5.9571,
      "mean_token_accuracy": 0.17098450511693955,
      "num_tokens": 33315077.0,
      "step": 13140
    },
    {
      "entropy": 6.05032205581665,
      "epoch": 1.5170225043277554,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047804525686607,
      "loss": 5.9501,
      "mean_token_accuracy": 0.17436959594488144,
      "num_tokens": 33326769.0,
      "step": 13145
    },
    {
      "entropy": 5.9553306102752686,
      "epoch": 1.5175995383727638,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004780274746887606,
      "loss": 5.7373,
      "mean_token_accuracy": 0.1891272783279419,
      "num_tokens": 33340723.0,
      "step": 13150
    },
    {
      "entropy": 6.211591958999634,
      "epoch": 1.5181765724177727,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047800968568272284,
      "loss": 5.9682,
      "mean_token_accuracy": 0.1752364382147789,
      "num_tokens": 33353894.0,
      "step": 13155
    },
    {
      "entropy": 6.182546997070313,
      "epoch": 1.5187536064627813,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004779918898485551,
      "loss": 6.0225,
      "mean_token_accuracy": 0.16770537495613097,
      "num_tokens": 33366134.0,
      "step": 13160
    },
    {
      "entropy": 6.119816827774048,
      "epoch": 1.51933064050779,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047797408718685614,
      "loss": 5.9021,
      "mean_token_accuracy": 0.1767203077673912,
      "num_tokens": 33377604.0,
      "step": 13165
    },
    {
      "entropy": 6.075493621826172,
      "epoch": 1.5199076745527986,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047795627769822447,
      "loss": 5.9458,
      "mean_token_accuracy": 0.17003180384635924,
      "num_tokens": 33389616.0,
      "step": 13170
    },
    {
      "entropy": 6.054233264923096,
      "epoch": 1.5204847085978073,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047793846138325925,
      "loss": 5.898,
      "mean_token_accuracy": 0.17698230147361754,
      "num_tokens": 33401493.0,
      "step": 13175
    },
    {
      "entropy": 6.077927589416504,
      "epoch": 1.521061742642816,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004779206382425598,
      "loss": 5.8804,
      "mean_token_accuracy": 0.18109677881002426,
      "num_tokens": 33414088.0,
      "step": 13180
    },
    {
      "entropy": 6.109663486480713,
      "epoch": 1.5216387766878245,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004779028082767253,
      "loss": 5.9408,
      "mean_token_accuracy": 0.1743367925286293,
      "num_tokens": 33426481.0,
      "step": 13185
    },
    {
      "entropy": 6.061102819442749,
      "epoch": 1.5222158107328332,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004778849714863557,
      "loss": 5.8639,
      "mean_token_accuracy": 0.1749799817800522,
      "num_tokens": 33437876.0,
      "step": 13190
    },
    {
      "entropy": 6.096581506729126,
      "epoch": 1.5227928447778418,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004778671278720508,
      "loss": 5.9467,
      "mean_token_accuracy": 0.17177701294422149,
      "num_tokens": 33450164.0,
      "step": 13195
    },
    {
      "entropy": 6.080563259124756,
      "epoch": 1.5233698788228507,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004778492774344109,
      "loss": 5.8411,
      "mean_token_accuracy": 0.18225359916687012,
      "num_tokens": 33462583.0,
      "step": 13200
    },
    {
      "entropy": 5.990028524398804,
      "epoch": 1.523946912867859,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004778314201740363,
      "loss": 5.9182,
      "mean_token_accuracy": 0.18244762271642684,
      "num_tokens": 33474099.0,
      "step": 13205
    },
    {
      "entropy": 6.08252100944519,
      "epoch": 1.524523946912868,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047781355609152764,
      "loss": 5.9067,
      "mean_token_accuracy": 0.17280192226171492,
      "num_tokens": 33486565.0,
      "step": 13210
    },
    {
      "entropy": 6.096342468261719,
      "epoch": 1.5251009809578764,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004777956851874858,
      "loss": 5.8566,
      "mean_token_accuracy": 0.17550449222326278,
      "num_tokens": 33498703.0,
      "step": 13215
    },
    {
      "entropy": 6.0455546379089355,
      "epoch": 1.5256780150028852,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047777780746251176,
      "loss": 5.9262,
      "mean_token_accuracy": 0.17824054062366484,
      "num_tokens": 33511431.0,
      "step": 13220
    },
    {
      "entropy": 6.001427793502808,
      "epoch": 1.5262550490478937,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047775992291720687,
      "loss": 5.8456,
      "mean_token_accuracy": 0.17821435928344725,
      "num_tokens": 33524771.0,
      "step": 13225
    },
    {
      "entropy": 6.078997659683227,
      "epoch": 1.5268320830929025,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004777420315521728,
      "loss": 5.8633,
      "mean_token_accuracy": 0.18156085759401322,
      "num_tokens": 33538362.0,
      "step": 13230
    },
    {
      "entropy": 6.054742908477783,
      "epoch": 1.5274091171379112,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004777241333680111,
      "loss": 5.9425,
      "mean_token_accuracy": 0.17418718636035918,
      "num_tokens": 33550905.0,
      "step": 13235
    },
    {
      "entropy": 6.120349788665772,
      "epoch": 1.5279861511829198,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004777062283653239,
      "loss": 5.9679,
      "mean_token_accuracy": 0.17858754545450212,
      "num_tokens": 33563665.0,
      "step": 13240
    },
    {
      "entropy": 6.056065464019776,
      "epoch": 1.5285631852279284,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047768831654471333,
      "loss": 5.8715,
      "mean_token_accuracy": 0.18302759677171707,
      "num_tokens": 33576957.0,
      "step": 13245
    },
    {
      "entropy": 6.056236696243286,
      "epoch": 1.529140219272937,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047767039790678204,
      "loss": 5.898,
      "mean_token_accuracy": 0.16999716460704803,
      "num_tokens": 33589641.0,
      "step": 13250
    },
    {
      "entropy": 6.052677202224731,
      "epoch": 1.5297172533179457,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047765247245213255,
      "loss": 5.8195,
      "mean_token_accuracy": 0.18229652494192122,
      "num_tokens": 33602377.0,
      "step": 13255
    },
    {
      "entropy": 6.064414739608765,
      "epoch": 1.5302942873629544,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004776345401813678,
      "loss": 5.8585,
      "mean_token_accuracy": 0.17987353205680848,
      "num_tokens": 33614796.0,
      "step": 13260
    },
    {
      "entropy": 6.1753137588500975,
      "epoch": 1.5308713214079632,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047761660109509095,
      "loss": 6.0024,
      "mean_token_accuracy": 0.17107590436935424,
      "num_tokens": 33626988.0,
      "step": 13265
    },
    {
      "entropy": 6.0178674221038815,
      "epoch": 1.5314483554529716,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004775986551939054,
      "loss": 5.8733,
      "mean_token_accuracy": 0.17519475817680358,
      "num_tokens": 33641376.0,
      "step": 13270
    },
    {
      "entropy": 6.086547565460205,
      "epoch": 1.5320253894979805,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047758070247841465,
      "loss": 5.9116,
      "mean_token_accuracy": 0.17213350534439087,
      "num_tokens": 33655155.0,
      "step": 13275
    },
    {
      "entropy": 6.098645973205566,
      "epoch": 1.532602423542989,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047756274294922266,
      "loss": 5.8802,
      "mean_token_accuracy": 0.17963136434555055,
      "num_tokens": 33668480.0,
      "step": 13280
    },
    {
      "entropy": 5.9868268966674805,
      "epoch": 1.5331794575879978,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004775447766069334,
      "loss": 5.81,
      "mean_token_accuracy": 0.1825267717242241,
      "num_tokens": 33680919.0,
      "step": 13285
    },
    {
      "entropy": 5.970165729522705,
      "epoch": 1.5337564916330062,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047752680345215115,
      "loss": 5.898,
      "mean_token_accuracy": 0.17485707402229309,
      "num_tokens": 33693161.0,
      "step": 13290
    },
    {
      "entropy": 6.1339739799499515,
      "epoch": 1.534333525678015,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004775088234854805,
      "loss": 5.9846,
      "mean_token_accuracy": 0.17738447040319444,
      "num_tokens": 33707537.0,
      "step": 13295
    },
    {
      "entropy": 6.125365495681763,
      "epoch": 1.5349105597230237,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004774908367075262,
      "loss": 5.86,
      "mean_token_accuracy": 0.17240599244832994,
      "num_tokens": 33719501.0,
      "step": 13300
    },
    {
      "entropy": 6.132128715515137,
      "epoch": 1.5354875937680323,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004774728431188931,
      "loss": 5.9898,
      "mean_token_accuracy": 0.16844378858804704,
      "num_tokens": 33731938.0,
      "step": 13305
    },
    {
      "entropy": 6.11914095878601,
      "epoch": 1.536064627813041,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047745484272018664,
      "loss": 5.9771,
      "mean_token_accuracy": 0.17454359233379363,
      "num_tokens": 33745509.0,
      "step": 13310
    },
    {
      "entropy": 6.087774324417114,
      "epoch": 1.5366416618580496,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004774368355120119,
      "loss": 5.8978,
      "mean_token_accuracy": 0.17703245878219603,
      "num_tokens": 33757818.0,
      "step": 13315
    },
    {
      "entropy": 6.057894611358643,
      "epoch": 1.5372186959030583,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000477418821494975,
      "loss": 5.8555,
      "mean_token_accuracy": 0.17535060048103332,
      "num_tokens": 33769840.0,
      "step": 13320
    },
    {
      "entropy": 6.058853054046631,
      "epoch": 1.537795729948067,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004774008006696814,
      "loss": 5.9061,
      "mean_token_accuracy": 0.1735802933573723,
      "num_tokens": 33783713.0,
      "step": 13325
    },
    {
      "entropy": 5.973485469818115,
      "epoch": 1.5383727639930755,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004773827730367375,
      "loss": 5.8976,
      "mean_token_accuracy": 0.18004723638296127,
      "num_tokens": 33796005.0,
      "step": 13330
    },
    {
      "entropy": 6.0902787208557125,
      "epoch": 1.5389497980380842,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047736473859674955,
      "loss": 5.8458,
      "mean_token_accuracy": 0.1806577354669571,
      "num_tokens": 33807295.0,
      "step": 13335
    },
    {
      "entropy": 6.059075927734375,
      "epoch": 1.539526832083093,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047734669735032404,
      "loss": 5.81,
      "mean_token_accuracy": 0.1782573029398918,
      "num_tokens": 33819772.0,
      "step": 13340
    },
    {
      "entropy": 6.029000997543335,
      "epoch": 1.5401038661281015,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047732864929806796,
      "loss": 5.9333,
      "mean_token_accuracy": 0.17572814375162124,
      "num_tokens": 33831820.0,
      "step": 13345
    },
    {
      "entropy": 6.122318696975708,
      "epoch": 1.5406809001731103,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004773105944405883,
      "loss": 5.9787,
      "mean_token_accuracy": 0.17256004959344864,
      "num_tokens": 33845753.0,
      "step": 13350
    },
    {
      "entropy": 6.171608972549438,
      "epoch": 1.5412579342181187,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047729253277849226,
      "loss": 5.9083,
      "mean_token_accuracy": 0.17647509574890136,
      "num_tokens": 33859438.0,
      "step": 13355
    },
    {
      "entropy": 6.023289775848388,
      "epoch": 1.5418349682631276,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047727446431238734,
      "loss": 5.8857,
      "mean_token_accuracy": 0.18259461969137192,
      "num_tokens": 33872062.0,
      "step": 13360
    },
    {
      "entropy": 6.014738273620606,
      "epoch": 1.542412002308136,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004772563890428813,
      "loss": 5.8454,
      "mean_token_accuracy": 0.18652074337005614,
      "num_tokens": 33883840.0,
      "step": 13365
    },
    {
      "entropy": 6.131973028182983,
      "epoch": 1.5429890363531449,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004772383069705821,
      "loss": 5.9585,
      "mean_token_accuracy": 0.16837052255868912,
      "num_tokens": 33896907.0,
      "step": 13370
    },
    {
      "entropy": 6.032290315628051,
      "epoch": 1.5435660703981535,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004772202180960978,
      "loss": 5.9237,
      "mean_token_accuracy": 0.17497534304857254,
      "num_tokens": 33912135.0,
      "step": 13375
    },
    {
      "entropy": 6.0658266067504885,
      "epoch": 1.5441431044431622,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047720212242003703,
      "loss": 5.8645,
      "mean_token_accuracy": 0.18023320883512498,
      "num_tokens": 33925014.0,
      "step": 13380
    },
    {
      "entropy": 6.1423811435699465,
      "epoch": 1.5447201384881708,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047718401994300825,
      "loss": 5.9735,
      "mean_token_accuracy": 0.172008416056633,
      "num_tokens": 33937429.0,
      "step": 13385
    },
    {
      "entropy": 6.034386205673218,
      "epoch": 1.5452971725331794,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047716591066562043,
      "loss": 5.9134,
      "mean_token_accuracy": 0.1797392800450325,
      "num_tokens": 33951030.0,
      "step": 13390
    },
    {
      "entropy": 6.105065488815308,
      "epoch": 1.545874206578188,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047714779458848255,
      "loss": 5.9928,
      "mean_token_accuracy": 0.16981230080127716,
      "num_tokens": 33962930.0,
      "step": 13395
    },
    {
      "entropy": 6.075588130950928,
      "epoch": 1.5464512406231967,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004771296717122041,
      "loss": 5.8817,
      "mean_token_accuracy": 0.17833095788955688,
      "num_tokens": 33977113.0,
      "step": 13400
    },
    {
      "entropy": 5.923658847808838,
      "epoch": 1.5470282746682056,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004771115420373945,
      "loss": 5.6427,
      "mean_token_accuracy": 0.19028390496969222,
      "num_tokens": 33989091.0,
      "step": 13405
    },
    {
      "entropy": 6.065774393081665,
      "epoch": 1.547605308713214,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047709340556466366,
      "loss": 5.9845,
      "mean_token_accuracy": 0.17547981590032577,
      "num_tokens": 34002051.0,
      "step": 13410
    },
    {
      "entropy": 6.1873057842254635,
      "epoch": 1.5481823427582229,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047707526229462144,
      "loss": 5.9746,
      "mean_token_accuracy": 0.16953601390123368,
      "num_tokens": 34014482.0,
      "step": 13415
    },
    {
      "entropy": 6.090724945068359,
      "epoch": 1.5487593768032313,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047705711222787816,
      "loss": 5.9122,
      "mean_token_accuracy": 0.18190435767173768,
      "num_tokens": 34025665.0,
      "step": 13420
    },
    {
      "entropy": 6.0527599334716795,
      "epoch": 1.5493364108482401,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047703895536504423,
      "loss": 5.8758,
      "mean_token_accuracy": 0.17877805083990098,
      "num_tokens": 34038163.0,
      "step": 13425
    },
    {
      "entropy": 6.081933450698853,
      "epoch": 1.5499134448932486,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004770207917067305,
      "loss": 5.9442,
      "mean_token_accuracy": 0.17934029102325438,
      "num_tokens": 34050774.0,
      "step": 13430
    },
    {
      "entropy": 6.1415973663330075,
      "epoch": 1.5504904789382574,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047700262125354765,
      "loss": 5.9511,
      "mean_token_accuracy": 0.17057330459356307,
      "num_tokens": 34062480.0,
      "step": 13435
    },
    {
      "entropy": 6.07702112197876,
      "epoch": 1.551067512983266,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047698444400610707,
      "loss": 5.8799,
      "mean_token_accuracy": 0.18241317421197892,
      "num_tokens": 34075314.0,
      "step": 13440
    },
    {
      "entropy": 6.007231760025024,
      "epoch": 1.5516445470282747,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047696625996502,
      "loss": 5.8239,
      "mean_token_accuracy": 0.17951334714889527,
      "num_tokens": 34087528.0,
      "step": 13445
    },
    {
      "entropy": 5.996708011627197,
      "epoch": 1.5522215810732833,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047694806913089815,
      "loss": 5.7972,
      "mean_token_accuracy": 0.18027949780225755,
      "num_tokens": 34101740.0,
      "step": 13450
    },
    {
      "entropy": 6.146956586837769,
      "epoch": 1.552798615118292,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004769298715043533,
      "loss": 5.9113,
      "mean_token_accuracy": 0.1722385048866272,
      "num_tokens": 34113042.0,
      "step": 13455
    },
    {
      "entropy": 6.039388227462768,
      "epoch": 1.5533756491633006,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004769116670859975,
      "loss": 5.9914,
      "mean_token_accuracy": 0.17165443152189255,
      "num_tokens": 34126801.0,
      "step": 13460
    },
    {
      "entropy": 6.091321086883545,
      "epoch": 1.5539526832083093,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047689345587644314,
      "loss": 5.9714,
      "mean_token_accuracy": 0.174395789206028,
      "num_tokens": 34139469.0,
      "step": 13465
    },
    {
      "entropy": 6.113066005706787,
      "epoch": 1.554529717253318,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047687523787630256,
      "loss": 5.9237,
      "mean_token_accuracy": 0.17263210713863372,
      "num_tokens": 34151940.0,
      "step": 13470
    },
    {
      "entropy": 6.042326641082764,
      "epoch": 1.5551067512983265,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004768570130861887,
      "loss": 5.8537,
      "mean_token_accuracy": 0.1777663916349411,
      "num_tokens": 34164375.0,
      "step": 13475
    },
    {
      "entropy": 6.086356687545776,
      "epoch": 1.5556837853433354,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004768387815067144,
      "loss": 5.9376,
      "mean_token_accuracy": 0.172316275537014,
      "num_tokens": 34176585.0,
      "step": 13480
    },
    {
      "entropy": 6.146156644821167,
      "epoch": 1.5562608193883438,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047682054313849304,
      "loss": 5.9406,
      "mean_token_accuracy": 0.1742989495396614,
      "num_tokens": 34188518.0,
      "step": 13485
    },
    {
      "entropy": 6.044810247421265,
      "epoch": 1.5568378534333527,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004768022979821379,
      "loss": 5.8133,
      "mean_token_accuracy": 0.18448566794395446,
      "num_tokens": 34200865.0,
      "step": 13490
    },
    {
      "entropy": 6.06164436340332,
      "epoch": 1.557414887478361,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004767840460382628,
      "loss": 5.8839,
      "mean_token_accuracy": 0.18351510316133499,
      "num_tokens": 34212766.0,
      "step": 13495
    },
    {
      "entropy": 6.048433256149292,
      "epoch": 1.55799192152337,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004767657873074815,
      "loss": 5.9133,
      "mean_token_accuracy": 0.17739587277173996,
      "num_tokens": 34225401.0,
      "step": 13500
    },
    {
      "entropy": 6.053851985931397,
      "epoch": 1.5585689555683784,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004767475217904081,
      "loss": 5.9036,
      "mean_token_accuracy": 0.1756637677550316,
      "num_tokens": 34237474.0,
      "step": 13505
    },
    {
      "entropy": 6.133122348785401,
      "epoch": 1.5591459896133872,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047672924948765705,
      "loss": 5.9823,
      "mean_token_accuracy": 0.16657297164201737,
      "num_tokens": 34249598.0,
      "step": 13510
    },
    {
      "entropy": 6.1387372493743895,
      "epoch": 1.5597230236583959,
      "grad_norm": 1.0,
      "learning_rate": 0.00047671097039984293,
      "loss": 5.8653,
      "mean_token_accuracy": 0.17887792140245437,
      "num_tokens": 34261773.0,
      "step": 13515
    },
    {
      "entropy": 6.10162467956543,
      "epoch": 1.5603000577034045,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047669268452758053,
      "loss": 5.9445,
      "mean_token_accuracy": 0.17530276179313659,
      "num_tokens": 34275398.0,
      "step": 13520
    },
    {
      "entropy": 6.056762266159057,
      "epoch": 1.5608770917484132,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047667439187148476,
      "loss": 5.8777,
      "mean_token_accuracy": 0.17667476385831832,
      "num_tokens": 34288934.0,
      "step": 13525
    },
    {
      "entropy": 6.045689964294434,
      "epoch": 1.5614541257934218,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004766560924321711,
      "loss": 5.8736,
      "mean_token_accuracy": 0.17737327963113786,
      "num_tokens": 34302187.0,
      "step": 13530
    },
    {
      "entropy": 6.030709171295166,
      "epoch": 1.5620311598384304,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047663778621025496,
      "loss": 5.8898,
      "mean_token_accuracy": 0.1733369782567024,
      "num_tokens": 34314745.0,
      "step": 13535
    },
    {
      "entropy": 6.126426935195923,
      "epoch": 1.562608193883439,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004766194732063519,
      "loss": 5.869,
      "mean_token_accuracy": 0.1819433018565178,
      "num_tokens": 34326771.0,
      "step": 13540
    },
    {
      "entropy": 6.082818794250488,
      "epoch": 1.563185227928448,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047660115342107814,
      "loss": 5.8742,
      "mean_token_accuracy": 0.1774468705058098,
      "num_tokens": 34339452.0,
      "step": 13545
    },
    {
      "entropy": 6.080442285537719,
      "epoch": 1.5637622619734564,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047658282685504973,
      "loss": 5.9215,
      "mean_token_accuracy": 0.16776583939790726,
      "num_tokens": 34351872.0,
      "step": 13550
    },
    {
      "entropy": 6.060977458953857,
      "epoch": 1.5643392960184652,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000476564493508883,
      "loss": 5.8888,
      "mean_token_accuracy": 0.17701526433229448,
      "num_tokens": 34363963.0,
      "step": 13555
    },
    {
      "entropy": 6.108565998077393,
      "epoch": 1.5649163300634736,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047654615338319466,
      "loss": 5.9712,
      "mean_token_accuracy": 0.1666532039642334,
      "num_tokens": 34377245.0,
      "step": 13560
    },
    {
      "entropy": 6.034358835220337,
      "epoch": 1.5654933641084825,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004765278064786016,
      "loss": 5.8585,
      "mean_token_accuracy": 0.1741995856165886,
      "num_tokens": 34391047.0,
      "step": 13565
    },
    {
      "entropy": 6.11386137008667,
      "epoch": 1.566070398153491,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047650945279572085,
      "loss": 5.9144,
      "mean_token_accuracy": 0.16956199705600739,
      "num_tokens": 34404655.0,
      "step": 13570
    },
    {
      "entropy": 6.106350755691528,
      "epoch": 1.5666474321984998,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004764910923351698,
      "loss": 5.8423,
      "mean_token_accuracy": 0.17780797630548478,
      "num_tokens": 34417902.0,
      "step": 13575
    },
    {
      "entropy": 6.087349510192871,
      "epoch": 1.5672244662435084,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047647272509756584,
      "loss": 5.8849,
      "mean_token_accuracy": 0.1813091143965721,
      "num_tokens": 34430630.0,
      "step": 13580
    },
    {
      "entropy": 6.019781160354614,
      "epoch": 1.567801500288517,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004764543510835269,
      "loss": 5.9075,
      "mean_token_accuracy": 0.17631940245628358,
      "num_tokens": 34444320.0,
      "step": 13585
    },
    {
      "entropy": 6.048609590530395,
      "epoch": 1.5683785343335257,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000476435970293671,
      "loss": 5.8133,
      "mean_token_accuracy": 0.18302533626556397,
      "num_tokens": 34457081.0,
      "step": 13590
    },
    {
      "entropy": 6.016569805145264,
      "epoch": 1.5689555683785343,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047641758272861626,
      "loss": 5.8389,
      "mean_token_accuracy": 0.18321243971586226,
      "num_tokens": 34469560.0,
      "step": 13595
    },
    {
      "entropy": 6.05223913192749,
      "epoch": 1.569532602423543,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004763991883889811,
      "loss": 5.9002,
      "mean_token_accuracy": 0.17379349321126938,
      "num_tokens": 34482036.0,
      "step": 13600
    },
    {
      "entropy": 6.048428153991699,
      "epoch": 1.5701096364685516,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004763807872753843,
      "loss": 5.8927,
      "mean_token_accuracy": 0.17731169015169143,
      "num_tokens": 34494316.0,
      "step": 13605
    },
    {
      "entropy": 6.0263885974884035,
      "epoch": 1.5706866705135603,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047636237938844484,
      "loss": 5.8368,
      "mean_token_accuracy": 0.1783609852194786,
      "num_tokens": 34507026.0,
      "step": 13610
    },
    {
      "entropy": 6.1095623016357425,
      "epoch": 1.571263704558569,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004763439647287817,
      "loss": 5.8523,
      "mean_token_accuracy": 0.1773489311337471,
      "num_tokens": 34519910.0,
      "step": 13615
    },
    {
      "entropy": 6.008932828903198,
      "epoch": 1.5718407386035778,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004763255432970144,
      "loss": 5.892,
      "mean_token_accuracy": 0.17871266156435012,
      "num_tokens": 34534832.0,
      "step": 13620
    },
    {
      "entropy": 6.049214792251587,
      "epoch": 1.5724177726485862,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047630711509376235,
      "loss": 5.9089,
      "mean_token_accuracy": 0.17490043640136718,
      "num_tokens": 34548559.0,
      "step": 13625
    },
    {
      "entropy": 6.08085994720459,
      "epoch": 1.572994806693595,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004762886801196455,
      "loss": 5.9386,
      "mean_token_accuracy": 0.1765395686030388,
      "num_tokens": 34561131.0,
      "step": 13630
    },
    {
      "entropy": 6.140205144882202,
      "epoch": 1.5735718407386035,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047627023837528386,
      "loss": 5.8916,
      "mean_token_accuracy": 0.173751600086689,
      "num_tokens": 34572863.0,
      "step": 13635
    },
    {
      "entropy": 5.981328010559082,
      "epoch": 1.5741488747836123,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047625178986129775,
      "loss": 5.7412,
      "mean_token_accuracy": 0.1841447740793228,
      "num_tokens": 34586277.0,
      "step": 13640
    },
    {
      "entropy": 5.9721653938293455,
      "epoch": 1.5747259088286207,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004762333345783078,
      "loss": 5.8523,
      "mean_token_accuracy": 0.17884965986013412,
      "num_tokens": 34598112.0,
      "step": 13645
    },
    {
      "entropy": 6.031538820266723,
      "epoch": 1.5753029428736296,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047621487252693436,
      "loss": 5.8497,
      "mean_token_accuracy": 0.18430056124925615,
      "num_tokens": 34610569.0,
      "step": 13650
    },
    {
      "entropy": 6.090058422088623,
      "epoch": 1.5758799769186382,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047619640370779876,
      "loss": 5.912,
      "mean_token_accuracy": 0.17911432236433028,
      "num_tokens": 34623148.0,
      "step": 13655
    },
    {
      "entropy": 5.98702974319458,
      "epoch": 1.5764570109636469,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047617792812152207,
      "loss": 5.858,
      "mean_token_accuracy": 0.1803685024380684,
      "num_tokens": 34635260.0,
      "step": 13660
    },
    {
      "entropy": 5.934999561309814,
      "epoch": 1.5770340450086555,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004761594457687256,
      "loss": 5.8016,
      "mean_token_accuracy": 0.18361441045999527,
      "num_tokens": 34648717.0,
      "step": 13665
    },
    {
      "entropy": 6.078234148025513,
      "epoch": 1.5776110790536642,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004761409566500313,
      "loss": 5.9183,
      "mean_token_accuracy": 0.1722771182656288,
      "num_tokens": 34662442.0,
      "step": 13670
    },
    {
      "entropy": 6.129392051696778,
      "epoch": 1.5781881130986728,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047612246076606066,
      "loss": 5.8598,
      "mean_token_accuracy": 0.1768808364868164,
      "num_tokens": 34675213.0,
      "step": 13675
    },
    {
      "entropy": 6.035598230361939,
      "epoch": 1.5787651471436814,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047610395811743594,
      "loss": 5.841,
      "mean_token_accuracy": 0.1749825358390808,
      "num_tokens": 34687959.0,
      "step": 13680
    },
    {
      "entropy": 6.004653024673462,
      "epoch": 1.5793421811886903,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047608544870477956,
      "loss": 5.835,
      "mean_token_accuracy": 0.18075956106185914,
      "num_tokens": 34700242.0,
      "step": 13685
    },
    {
      "entropy": 6.011392641067505,
      "epoch": 1.5799192152336987,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047606693252871395,
      "loss": 5.8584,
      "mean_token_accuracy": 0.17911642640829087,
      "num_tokens": 34713245.0,
      "step": 13690
    },
    {
      "entropy": 6.04226245880127,
      "epoch": 1.5804962492787076,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000476048409589862,
      "loss": 5.9353,
      "mean_token_accuracy": 0.17338775098323822,
      "num_tokens": 34725248.0,
      "step": 13695
    },
    {
      "entropy": 6.165214204788208,
      "epoch": 1.581073283323716,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004760298798888466,
      "loss": 5.9707,
      "mean_token_accuracy": 0.1719074457883835,
      "num_tokens": 34737801.0,
      "step": 13700
    },
    {
      "entropy": 6.080092239379883,
      "epoch": 1.5816503173687249,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004760113434262911,
      "loss": 5.9159,
      "mean_token_accuracy": 0.1728976398706436,
      "num_tokens": 34749504.0,
      "step": 13705
    },
    {
      "entropy": 6.048469972610474,
      "epoch": 1.5822273514137333,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047599280020281894,
      "loss": 5.9161,
      "mean_token_accuracy": 0.17429527789354324,
      "num_tokens": 34761238.0,
      "step": 13710
    },
    {
      "entropy": 6.109679508209228,
      "epoch": 1.5828043854587421,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047597425021905364,
      "loss": 5.9259,
      "mean_token_accuracy": 0.17530940622091293,
      "num_tokens": 34773662.0,
      "step": 13715
    },
    {
      "entropy": 6.055896806716919,
      "epoch": 1.5833814195037508,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004759556934756194,
      "loss": 5.9238,
      "mean_token_accuracy": 0.17185305655002595,
      "num_tokens": 34786774.0,
      "step": 13720
    },
    {
      "entropy": 6.041871500015259,
      "epoch": 1.5839584535487594,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047593712997314017,
      "loss": 5.8656,
      "mean_token_accuracy": 0.18314026892185212,
      "num_tokens": 34798169.0,
      "step": 13725
    },
    {
      "entropy": 6.0185106754302975,
      "epoch": 1.584535487593768,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047591855971224035,
      "loss": 5.8355,
      "mean_token_accuracy": 0.1787860631942749,
      "num_tokens": 34809870.0,
      "step": 13730
    },
    {
      "entropy": 6.082783460617065,
      "epoch": 1.5851125216387767,
      "grad_norm": 1.0,
      "learning_rate": 0.0004758999826935446,
      "loss": 5.9861,
      "mean_token_accuracy": 0.17320285886526107,
      "num_tokens": 34821684.0,
      "step": 13735
    },
    {
      "entropy": 6.079405975341797,
      "epoch": 1.5856895556837853,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004758813989176778,
      "loss": 5.907,
      "mean_token_accuracy": 0.1729307472705841,
      "num_tokens": 34834905.0,
      "step": 13740
    },
    {
      "entropy": 5.99232029914856,
      "epoch": 1.586266589728794,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047586280838526483,
      "loss": 5.8607,
      "mean_token_accuracy": 0.17246422320604324,
      "num_tokens": 34846797.0,
      "step": 13745
    },
    {
      "entropy": 6.092650413513184,
      "epoch": 1.5868436237738026,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004758442110969312,
      "loss": 5.9074,
      "mean_token_accuracy": 0.1704322561621666,
      "num_tokens": 34859261.0,
      "step": 13750
    },
    {
      "entropy": 6.1012345314025875,
      "epoch": 1.5874206578188113,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004758256070533022,
      "loss": 5.9402,
      "mean_token_accuracy": 0.17145328968763351,
      "num_tokens": 34871775.0,
      "step": 13755
    },
    {
      "entropy": 6.026003646850586,
      "epoch": 1.5879976918638201,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004758069962550037,
      "loss": 5.7903,
      "mean_token_accuracy": 0.1818581312894821,
      "num_tokens": 34884461.0,
      "step": 13760
    },
    {
      "entropy": 6.078271389007568,
      "epoch": 1.5885747259088285,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047578837870266156,
      "loss": 5.9112,
      "mean_token_accuracy": 0.17359140515327454,
      "num_tokens": 34896772.0,
      "step": 13765
    },
    {
      "entropy": 6.060633516311645,
      "epoch": 1.5891517599538374,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047576975439690206,
      "loss": 5.9362,
      "mean_token_accuracy": 0.1681361213326454,
      "num_tokens": 34909914.0,
      "step": 13770
    },
    {
      "entropy": 6.111034345626831,
      "epoch": 1.5897287939988458,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047575112333835164,
      "loss": 5.9517,
      "mean_token_accuracy": 0.1690449595451355,
      "num_tokens": 34922298.0,
      "step": 13775
    },
    {
      "entropy": 6.159610271453857,
      "epoch": 1.5903058280438547,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047573248552763684,
      "loss": 6.017,
      "mean_token_accuracy": 0.17719260156154631,
      "num_tokens": 34935159.0,
      "step": 13780
    },
    {
      "entropy": 6.017228031158448,
      "epoch": 1.590882862088863,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047571384096538474,
      "loss": 5.7912,
      "mean_token_accuracy": 0.17786884307861328,
      "num_tokens": 34948347.0,
      "step": 13785
    },
    {
      "entropy": 6.0053074836730955,
      "epoch": 1.591459896133872,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004756951896522222,
      "loss": 5.8981,
      "mean_token_accuracy": 0.17603261917829513,
      "num_tokens": 34961489.0,
      "step": 13790
    },
    {
      "entropy": 6.170222759246826,
      "epoch": 1.5920369301788806,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004756765315887766,
      "loss": 6.0135,
      "mean_token_accuracy": 0.16643890589475632,
      "num_tokens": 34974458.0,
      "step": 13795
    },
    {
      "entropy": 6.148947620391846,
      "epoch": 1.5926139642238892,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004756578667756756,
      "loss": 5.9019,
      "mean_token_accuracy": 0.17909768968820572,
      "num_tokens": 34988498.0,
      "step": 13800
    },
    {
      "entropy": 5.868844127655029,
      "epoch": 1.5931909982688979,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004756391952135469,
      "loss": 5.7388,
      "mean_token_accuracy": 0.19357212781906127,
      "num_tokens": 35002101.0,
      "step": 13805
    },
    {
      "entropy": 6.068660593032837,
      "epoch": 1.5937680323139065,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047562051690301855,
      "loss": 5.961,
      "mean_token_accuracy": 0.17820781618356704,
      "num_tokens": 35014079.0,
      "step": 13810
    },
    {
      "entropy": 6.093256235122681,
      "epoch": 1.5943450663589152,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047560183184471873,
      "loss": 5.8493,
      "mean_token_accuracy": 0.1808098018169403,
      "num_tokens": 35027455.0,
      "step": 13815
    },
    {
      "entropy": 5.964543962478638,
      "epoch": 1.5949221004039238,
      "grad_norm": 0.890625,
      "learning_rate": 0.000475583140039276,
      "loss": 5.819,
      "mean_token_accuracy": 0.18177127838134766,
      "num_tokens": 35040699.0,
      "step": 13820
    },
    {
      "entropy": 6.089759588241577,
      "epoch": 1.5954991344489324,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047556444148731897,
      "loss": 5.9448,
      "mean_token_accuracy": 0.17098997980356218,
      "num_tokens": 35052758.0,
      "step": 13825
    },
    {
      "entropy": 6.014875698089599,
      "epoch": 1.596076168493941,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004755457361894766,
      "loss": 5.924,
      "mean_token_accuracy": 0.17579948008060456,
      "num_tokens": 35065393.0,
      "step": 13830
    },
    {
      "entropy": 6.13860330581665,
      "epoch": 1.59665320253895,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047552702414637793,
      "loss": 5.9387,
      "mean_token_accuracy": 0.17041729688644408,
      "num_tokens": 35077708.0,
      "step": 13835
    },
    {
      "entropy": 6.091925430297851,
      "epoch": 1.5972302365839584,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047550830535865245,
      "loss": 5.9227,
      "mean_token_accuracy": 0.17138339728116989,
      "num_tokens": 35088870.0,
      "step": 13840
    },
    {
      "entropy": 6.029639816284179,
      "epoch": 1.5978072706289672,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047548957982692974,
      "loss": 5.8355,
      "mean_token_accuracy": 0.18020908683538436,
      "num_tokens": 35100128.0,
      "step": 13845
    },
    {
      "entropy": 6.07875657081604,
      "epoch": 1.5983843046739756,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004754708475518396,
      "loss": 5.9955,
      "mean_token_accuracy": 0.1758660688996315,
      "num_tokens": 35113539.0,
      "step": 13850
    },
    {
      "entropy": 6.1427453517913815,
      "epoch": 1.5989613387189845,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047545210853401214,
      "loss": 5.9556,
      "mean_token_accuracy": 0.17159054577350616,
      "num_tokens": 35126897.0,
      "step": 13855
    },
    {
      "entropy": 6.067446517944336,
      "epoch": 1.599538372763993,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047543336277407753,
      "loss": 5.8949,
      "mean_token_accuracy": 0.18084877282381057,
      "num_tokens": 35140197.0,
      "step": 13860
    },
    {
      "entropy": 6.026173639297485,
      "epoch": 1.6001154068090018,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047541461027266624,
      "loss": 5.8835,
      "mean_token_accuracy": 0.1788007602095604,
      "num_tokens": 35151928.0,
      "step": 13865
    },
    {
      "entropy": 6.112354516983032,
      "epoch": 1.6006924408540104,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004753958510304091,
      "loss": 5.9406,
      "mean_token_accuracy": 0.17329683899879456,
      "num_tokens": 35165045.0,
      "step": 13870
    },
    {
      "entropy": 6.100225591659546,
      "epoch": 1.601269474899019,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047537708504793714,
      "loss": 5.8334,
      "mean_token_accuracy": 0.18541805744171141,
      "num_tokens": 35177631.0,
      "step": 13875
    },
    {
      "entropy": 6.002701711654663,
      "epoch": 1.6018465089440277,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047535831232588146,
      "loss": 5.9094,
      "mean_token_accuracy": 0.1702704280614853,
      "num_tokens": 35190166.0,
      "step": 13880
    },
    {
      "entropy": 6.137913417816162,
      "epoch": 1.6024235429890363,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047533953286487345,
      "loss": 5.9539,
      "mean_token_accuracy": 0.1759255900979042,
      "num_tokens": 35202604.0,
      "step": 13885
    },
    {
      "entropy": 5.991219568252563,
      "epoch": 1.603000577034045,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004753207466655447,
      "loss": 5.7588,
      "mean_token_accuracy": 0.18425407707691194,
      "num_tokens": 35214686.0,
      "step": 13890
    },
    {
      "entropy": 6.042591857910156,
      "epoch": 1.6035776110790536,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004753019537285273,
      "loss": 5.9649,
      "mean_token_accuracy": 0.17304341048002242,
      "num_tokens": 35226230.0,
      "step": 13895
    },
    {
      "entropy": 6.09741177558899,
      "epoch": 1.6041546451240625,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047528315405445296,
      "loss": 5.925,
      "mean_token_accuracy": 0.1723446935415268,
      "num_tokens": 35240273.0,
      "step": 13900
    },
    {
      "entropy": 6.093927335739136,
      "epoch": 1.604731679169071,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047526434764395435,
      "loss": 5.9418,
      "mean_token_accuracy": 0.17055649310350418,
      "num_tokens": 35253095.0,
      "step": 13905
    },
    {
      "entropy": 6.080751943588257,
      "epoch": 1.6053087132140798,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047524553449766386,
      "loss": 5.8588,
      "mean_token_accuracy": 0.18328708857297898,
      "num_tokens": 35265770.0,
      "step": 13910
    },
    {
      "entropy": 6.007266235351563,
      "epoch": 1.6058857472590882,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047522671461621433,
      "loss": 5.8953,
      "mean_token_accuracy": 0.18225565403699875,
      "num_tokens": 35278800.0,
      "step": 13915
    },
    {
      "entropy": 6.080861759185791,
      "epoch": 1.606462781304097,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004752078880002386,
      "loss": 5.8319,
      "mean_token_accuracy": 0.17970171719789504,
      "num_tokens": 35290423.0,
      "step": 13920
    },
    {
      "entropy": 6.036775827407837,
      "epoch": 1.6070398153491054,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00047518905465037005,
      "loss": 5.8968,
      "mean_token_accuracy": 0.18257960826158523,
      "num_tokens": 35303679.0,
      "step": 13925
    },
    {
      "entropy": 6.0855405807495115,
      "epoch": 1.6076168493941143,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047517021456724214,
      "loss": 5.9874,
      "mean_token_accuracy": 0.17240236103534698,
      "num_tokens": 35316900.0,
      "step": 13930
    },
    {
      "entropy": 6.091698455810547,
      "epoch": 1.608193883439123,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047515136775148843,
      "loss": 5.932,
      "mean_token_accuracy": 0.1755893349647522,
      "num_tokens": 35329952.0,
      "step": 13935
    },
    {
      "entropy": 6.0558922290802,
      "epoch": 1.6087709174841316,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004751325142037429,
      "loss": 5.9361,
      "mean_token_accuracy": 0.1725696861743927,
      "num_tokens": 35342107.0,
      "step": 13940
    },
    {
      "entropy": 6.102687931060791,
      "epoch": 1.6093479515291402,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047511365392463974,
      "loss": 5.8387,
      "mean_token_accuracy": 0.18006399422883987,
      "num_tokens": 35354996.0,
      "step": 13945
    },
    {
      "entropy": 5.999194240570068,
      "epoch": 1.6099249855741489,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047509478691481317,
      "loss": 5.9075,
      "mean_token_accuracy": 0.17710723578929902,
      "num_tokens": 35367332.0,
      "step": 13950
    },
    {
      "entropy": 6.029749917984009,
      "epoch": 1.6105020196191575,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047507591317489783,
      "loss": 5.8457,
      "mean_token_accuracy": 0.17658497095108033,
      "num_tokens": 35379469.0,
      "step": 13955
    },
    {
      "entropy": 6.1228736400604244,
      "epoch": 1.6110790536641661,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004750570327055286,
      "loss": 5.8085,
      "mean_token_accuracy": 0.1829373374581337,
      "num_tokens": 35392579.0,
      "step": 13960
    },
    {
      "entropy": 6.047231769561767,
      "epoch": 1.6116560877091748,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047503814550734034,
      "loss": 5.9008,
      "mean_token_accuracy": 0.17175330519676207,
      "num_tokens": 35405421.0,
      "step": 13965
    },
    {
      "entropy": 6.018987703323364,
      "epoch": 1.6122331217541834,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004750192515809685,
      "loss": 5.9715,
      "mean_token_accuracy": 0.17282398790121078,
      "num_tokens": 35418046.0,
      "step": 13970
    },
    {
      "entropy": 6.124209070205689,
      "epoch": 1.6128101557991923,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047500035092704843,
      "loss": 5.8812,
      "mean_token_accuracy": 0.17559023350477218,
      "num_tokens": 35430346.0,
      "step": 13975
    },
    {
      "entropy": 6.078659772872925,
      "epoch": 1.6133871898442007,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004749814435462159,
      "loss": 5.9151,
      "mean_token_accuracy": 0.1656491607427597,
      "num_tokens": 35443607.0,
      "step": 13980
    },
    {
      "entropy": 6.043546438217163,
      "epoch": 1.6139642238892096,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004749625294391069,
      "loss": 5.8523,
      "mean_token_accuracy": 0.17862537801265715,
      "num_tokens": 35455890.0,
      "step": 13985
    },
    {
      "entropy": 5.975907516479492,
      "epoch": 1.614541257934218,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047494360860635755,
      "loss": 5.8833,
      "mean_token_accuracy": 0.18577795326709748,
      "num_tokens": 35470340.0,
      "step": 13990
    },
    {
      "entropy": 6.086055707931519,
      "epoch": 1.6151182919792268,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004749246810486042,
      "loss": 5.8784,
      "mean_token_accuracy": 0.17569545209407805,
      "num_tokens": 35482827.0,
      "step": 13995
    },
    {
      "entropy": 6.148967170715332,
      "epoch": 1.6156953260242353,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004749057467664836,
      "loss": 5.9578,
      "mean_token_accuracy": 0.17188836634159088,
      "num_tokens": 35494289.0,
      "step": 14000
    },
    {
      "entropy": 5.997967720031738,
      "epoch": 1.6162723600692441,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047488680576063247,
      "loss": 5.8441,
      "mean_token_accuracy": 0.177097150683403,
      "num_tokens": 35506541.0,
      "step": 14005
    },
    {
      "entropy": 6.0019043445587155,
      "epoch": 1.6168493941142528,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004748678580316878,
      "loss": 5.7981,
      "mean_token_accuracy": 0.183547006547451,
      "num_tokens": 35519628.0,
      "step": 14010
    },
    {
      "entropy": 6.065429544448852,
      "epoch": 1.6174264281592614,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047484890358028714,
      "loss": 5.8004,
      "mean_token_accuracy": 0.17986395955085754,
      "num_tokens": 35532525.0,
      "step": 14015
    },
    {
      "entropy": 6.053888511657715,
      "epoch": 1.61800346220427,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004748299424070678,
      "loss": 5.913,
      "mean_token_accuracy": 0.1770218253135681,
      "num_tokens": 35544046.0,
      "step": 14020
    },
    {
      "entropy": 6.078341245651245,
      "epoch": 1.6185804962492787,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004748109745126677,
      "loss": 5.9197,
      "mean_token_accuracy": 0.17828752994537353,
      "num_tokens": 35557784.0,
      "step": 14025
    },
    {
      "entropy": 6.098750734329224,
      "epoch": 1.6191575302942873,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047479199989772464,
      "loss": 5.9578,
      "mean_token_accuracy": 0.17179629057645798,
      "num_tokens": 35571472.0,
      "step": 14030
    },
    {
      "entropy": 6.00908522605896,
      "epoch": 1.619734564339296,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000474773018562877,
      "loss": 5.7811,
      "mean_token_accuracy": 0.19541188478469848,
      "num_tokens": 35584992.0,
      "step": 14035
    },
    {
      "entropy": 6.008803462982177,
      "epoch": 1.6203115983843048,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047475403050876297,
      "loss": 5.8032,
      "mean_token_accuracy": 0.1761187806725502,
      "num_tokens": 35596643.0,
      "step": 14040
    },
    {
      "entropy": 6.10283932685852,
      "epoch": 1.6208886324293132,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004747350357360214,
      "loss": 5.9443,
      "mean_token_accuracy": 0.1755638524889946,
      "num_tokens": 35608796.0,
      "step": 14045
    },
    {
      "entropy": 6.009304571151733,
      "epoch": 1.621465666474322,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004747160342452912,
      "loss": 5.7988,
      "mean_token_accuracy": 0.1836042210459709,
      "num_tokens": 35620008.0,
      "step": 14050
    },
    {
      "entropy": 6.0905333995819095,
      "epoch": 1.6220427005193305,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047469702603721134,
      "loss": 5.8635,
      "mean_token_accuracy": 0.17672292292118072,
      "num_tokens": 35632751.0,
      "step": 14055
    },
    {
      "entropy": 5.955360364913941,
      "epoch": 1.6226197345643394,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004746780111124212,
      "loss": 5.841,
      "mean_token_accuracy": 0.17916012555360794,
      "num_tokens": 35645031.0,
      "step": 14060
    },
    {
      "entropy": 6.044130516052246,
      "epoch": 1.6231967686093478,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047465898947156033,
      "loss": 5.8527,
      "mean_token_accuracy": 0.17793368697166442,
      "num_tokens": 35656731.0,
      "step": 14065
    },
    {
      "entropy": 6.103395557403564,
      "epoch": 1.6237738026543567,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047463996111526854,
      "loss": 5.8728,
      "mean_token_accuracy": 0.17957282960414886,
      "num_tokens": 35668371.0,
      "step": 14070
    },
    {
      "entropy": 6.076901865005493,
      "epoch": 1.6243508366993653,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047462092604418576,
      "loss": 5.9509,
      "mean_token_accuracy": 0.1767246440052986,
      "num_tokens": 35681339.0,
      "step": 14075
    },
    {
      "entropy": 6.102066230773926,
      "epoch": 1.624927870744374,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047460188425895236,
      "loss": 5.9101,
      "mean_token_accuracy": 0.1754898652434349,
      "num_tokens": 35693775.0,
      "step": 14080
    },
    {
      "entropy": 5.961327934265137,
      "epoch": 1.6255049047893826,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047458283576020874,
      "loss": 5.7838,
      "mean_token_accuracy": 0.19257943332195282,
      "num_tokens": 35705707.0,
      "step": 14085
    },
    {
      "entropy": 5.989688920974731,
      "epoch": 1.6260819388343912,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047456378054859554,
      "loss": 5.8215,
      "mean_token_accuracy": 0.18398423939943315,
      "num_tokens": 35717793.0,
      "step": 14090
    },
    {
      "entropy": 6.0407380104064945,
      "epoch": 1.6266589728793999,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047454471862475375,
      "loss": 5.8388,
      "mean_token_accuracy": 0.1814399316906929,
      "num_tokens": 35730762.0,
      "step": 14095
    },
    {
      "entropy": 6.09248719215393,
      "epoch": 1.6272360069244085,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047452564998932446,
      "loss": 5.8533,
      "mean_token_accuracy": 0.17506527453660964,
      "num_tokens": 35742852.0,
      "step": 14100
    },
    {
      "entropy": 6.109133005142212,
      "epoch": 1.6278130409694171,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000474506574642949,
      "loss": 5.9284,
      "mean_token_accuracy": 0.1666823610663414,
      "num_tokens": 35756416.0,
      "step": 14105
    },
    {
      "entropy": 6.10202841758728,
      "epoch": 1.6283900750144258,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000474487492586269,
      "loss": 5.9365,
      "mean_token_accuracy": 0.1692310929298401,
      "num_tokens": 35768347.0,
      "step": 14110
    },
    {
      "entropy": 6.044725561141968,
      "epoch": 1.6289671090594346,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004744684038199263,
      "loss": 5.851,
      "mean_token_accuracy": 0.17192990332841873,
      "num_tokens": 35781228.0,
      "step": 14115
    },
    {
      "entropy": 6.067657995223999,
      "epoch": 1.629544143104443,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047444930834456293,
      "loss": 5.8622,
      "mean_token_accuracy": 0.17533150017261506,
      "num_tokens": 35793757.0,
      "step": 14120
    },
    {
      "entropy": 6.120788145065307,
      "epoch": 1.630121177149452,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004744302061608212,
      "loss": 5.9107,
      "mean_token_accuracy": 0.17259672433137893,
      "num_tokens": 35805321.0,
      "step": 14125
    },
    {
      "entropy": 6.051059532165527,
      "epoch": 1.6306982111944603,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047441109726934345,
      "loss": 5.8403,
      "mean_token_accuracy": 0.1835612267255783,
      "num_tokens": 35817755.0,
      "step": 14130
    },
    {
      "entropy": 6.019346046447754,
      "epoch": 1.6312752452394692,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00047439198167077256,
      "loss": 5.8883,
      "mean_token_accuracy": 0.1750844195485115,
      "num_tokens": 35831856.0,
      "step": 14135
    },
    {
      "entropy": 6.0478509902954105,
      "epoch": 1.6318522792844776,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004743728593657514,
      "loss": 5.9312,
      "mean_token_accuracy": 0.1804079830646515,
      "num_tokens": 35844613.0,
      "step": 14140
    },
    {
      "entropy": 6.030668687820435,
      "epoch": 1.6324293133294865,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004743537303549231,
      "loss": 5.8321,
      "mean_token_accuracy": 0.18044550865888595,
      "num_tokens": 35856928.0,
      "step": 14145
    },
    {
      "entropy": 6.014145231246948,
      "epoch": 1.6330063473744951,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004743345946389311,
      "loss": 5.8596,
      "mean_token_accuracy": 0.1766326054930687,
      "num_tokens": 35870161.0,
      "step": 14150
    },
    {
      "entropy": 6.114483976364136,
      "epoch": 1.6335833814195038,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047431545221841907,
      "loss": 5.906,
      "mean_token_accuracy": 0.17500362396240235,
      "num_tokens": 35883176.0,
      "step": 14155
    },
    {
      "entropy": 6.034779500961304,
      "epoch": 1.6341604154645124,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047429630309403086,
      "loss": 5.8062,
      "mean_token_accuracy": 0.17787060141563416,
      "num_tokens": 35895741.0,
      "step": 14160
    },
    {
      "entropy": 6.076469135284424,
      "epoch": 1.634737449509521,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047427714726641044,
      "loss": 5.8989,
      "mean_token_accuracy": 0.17581305652856827,
      "num_tokens": 35908921.0,
      "step": 14165
    },
    {
      "entropy": 6.096293640136719,
      "epoch": 1.6353144835545297,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047425798473620215,
      "loss": 5.8827,
      "mean_token_accuracy": 0.18694284111261367,
      "num_tokens": 35921388.0,
      "step": 14170
    },
    {
      "entropy": 6.064926671981811,
      "epoch": 1.6358915175995383,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004742388155040505,
      "loss": 5.9082,
      "mean_token_accuracy": 0.17920336723327637,
      "num_tokens": 35933880.0,
      "step": 14175
    },
    {
      "entropy": 6.015205430984497,
      "epoch": 1.6364685516445472,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047421963957060026,
      "loss": 5.8531,
      "mean_token_accuracy": 0.17704266011714936,
      "num_tokens": 35946690.0,
      "step": 14180
    },
    {
      "entropy": 6.090030670166016,
      "epoch": 1.6370455856895556,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004742004569364964,
      "loss": 5.8925,
      "mean_token_accuracy": 0.17510934323072433,
      "num_tokens": 35958437.0,
      "step": 14185
    },
    {
      "entropy": 6.106938886642456,
      "epoch": 1.6376226197345645,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047418126760238416,
      "loss": 5.8629,
      "mean_token_accuracy": 0.1792551949620247,
      "num_tokens": 35969734.0,
      "step": 14190
    },
    {
      "entropy": 5.977389526367188,
      "epoch": 1.6381996537795729,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047416207156890887,
      "loss": 5.8387,
      "mean_token_accuracy": 0.18467945158481597,
      "num_tokens": 35983059.0,
      "step": 14195
    },
    {
      "entropy": 6.145402431488037,
      "epoch": 1.6387766878245817,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004741428688367164,
      "loss": 5.8821,
      "mean_token_accuracy": 0.17565348148345947,
      "num_tokens": 35996207.0,
      "step": 14200
    },
    {
      "entropy": 6.1338025569915775,
      "epoch": 1.6393537218695902,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047412365940645225,
      "loss": 5.9051,
      "mean_token_accuracy": 0.17503598630428313,
      "num_tokens": 36008328.0,
      "step": 14205
    },
    {
      "entropy": 6.079263210296631,
      "epoch": 1.639930755914599,
      "grad_norm": 1.0,
      "learning_rate": 0.00047410444327876286,
      "loss": 5.9962,
      "mean_token_accuracy": 0.1768765464425087,
      "num_tokens": 36021384.0,
      "step": 14210
    },
    {
      "entropy": 6.024386930465698,
      "epoch": 1.6405077899596077,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047408522045429426,
      "loss": 5.883,
      "mean_token_accuracy": 0.17830740362405778,
      "num_tokens": 36034101.0,
      "step": 14215
    },
    {
      "entropy": 6.072831726074218,
      "epoch": 1.6410848240046163,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004740659909336933,
      "loss": 5.8047,
      "mean_token_accuracy": 0.1767995148897171,
      "num_tokens": 36046504.0,
      "step": 14220
    },
    {
      "entropy": 6.086407804489136,
      "epoch": 1.641661858049625,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047404675471760655,
      "loss": 5.8989,
      "mean_token_accuracy": 0.1768716737627983,
      "num_tokens": 36058759.0,
      "step": 14225
    },
    {
      "entropy": 6.057722330093384,
      "epoch": 1.6422388920946336,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004740275118066811,
      "loss": 5.8979,
      "mean_token_accuracy": 0.1821932837367058,
      "num_tokens": 36070238.0,
      "step": 14230
    },
    {
      "entropy": 6.0589007377624515,
      "epoch": 1.6428159261396422,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047400826220156416,
      "loss": 5.9055,
      "mean_token_accuracy": 0.17538813799619674,
      "num_tokens": 36083543.0,
      "step": 14235
    },
    {
      "entropy": 6.035336923599243,
      "epoch": 1.6433929601846509,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047398900590290313,
      "loss": 5.7479,
      "mean_token_accuracy": 0.18545111864805222,
      "num_tokens": 36095250.0,
      "step": 14240
    },
    {
      "entropy": 6.002789068222046,
      "epoch": 1.6439699942296595,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047396974291134575,
      "loss": 5.9022,
      "mean_token_accuracy": 0.18134790062904357,
      "num_tokens": 36107919.0,
      "step": 14245
    },
    {
      "entropy": 6.145536947250366,
      "epoch": 1.6445470282746681,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047395047322754,
      "loss": 5.9477,
      "mean_token_accuracy": 0.17147489488124848,
      "num_tokens": 36120284.0,
      "step": 14250
    },
    {
      "entropy": 6.172828817367554,
      "epoch": 1.645124062319677,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047393119685213374,
      "loss": 5.9213,
      "mean_token_accuracy": 0.17762400954961777,
      "num_tokens": 36133710.0,
      "step": 14255
    },
    {
      "entropy": 6.059191560745239,
      "epoch": 1.6457010963646854,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004739119137857756,
      "loss": 5.8837,
      "mean_token_accuracy": 0.17916419357061386,
      "num_tokens": 36147042.0,
      "step": 14260
    },
    {
      "entropy": 6.069198989868164,
      "epoch": 1.6462781304096943,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047389262402911404,
      "loss": 5.9146,
      "mean_token_accuracy": 0.17229731231927872,
      "num_tokens": 36159053.0,
      "step": 14265
    },
    {
      "entropy": 6.047973585128784,
      "epoch": 1.6468551644547027,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047387332758279784,
      "loss": 5.7984,
      "mean_token_accuracy": 0.17994379997253418,
      "num_tokens": 36171321.0,
      "step": 14270
    },
    {
      "entropy": 5.993471002578735,
      "epoch": 1.6474321984997116,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047385402444747606,
      "loss": 5.7559,
      "mean_token_accuracy": 0.1900617390871048,
      "num_tokens": 36183341.0,
      "step": 14275
    },
    {
      "entropy": 6.005244159698487,
      "epoch": 1.64800923254472,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047383471462379803,
      "loss": 5.8331,
      "mean_token_accuracy": 0.17995353192090988,
      "num_tokens": 36195407.0,
      "step": 14280
    },
    {
      "entropy": 6.087130498886109,
      "epoch": 1.6485862665897288,
      "grad_norm": 1.03125,
      "learning_rate": 0.000473815398112413,
      "loss": 5.9335,
      "mean_token_accuracy": 0.16984787732362747,
      "num_tokens": 36208736.0,
      "step": 14285
    },
    {
      "entropy": 6.083852243423462,
      "epoch": 1.6491633006347375,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004737960749139708,
      "loss": 5.8696,
      "mean_token_accuracy": 0.1803152084350586,
      "num_tokens": 36220022.0,
      "step": 14290
    },
    {
      "entropy": 5.995620965957642,
      "epoch": 1.6497403346797461,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004737767450291215,
      "loss": 5.7931,
      "mean_token_accuracy": 0.17577860206365586,
      "num_tokens": 36232105.0,
      "step": 14295
    },
    {
      "entropy": 6.023604154586792,
      "epoch": 1.6503173687247548,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047375740845851506,
      "loss": 5.8663,
      "mean_token_accuracy": 0.17927693128585814,
      "num_tokens": 36243892.0,
      "step": 14300
    },
    {
      "entropy": 6.059008932113647,
      "epoch": 1.6508944027697634,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004737380652028019,
      "loss": 5.9325,
      "mean_token_accuracy": 0.1714098185300827,
      "num_tokens": 36255365.0,
      "step": 14305
    },
    {
      "entropy": 6.104949998855591,
      "epoch": 1.651471436814772,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047371871526263263,
      "loss": 5.9262,
      "mean_token_accuracy": 0.17757108062505722,
      "num_tokens": 36268238.0,
      "step": 14310
    },
    {
      "entropy": 6.120410442352295,
      "epoch": 1.6520484708597807,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004736993586386581,
      "loss": 5.9221,
      "mean_token_accuracy": 0.1765486031770706,
      "num_tokens": 36280294.0,
      "step": 14315
    },
    {
      "entropy": 6.027853488922119,
      "epoch": 1.6526255049047895,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047367999533152934,
      "loss": 5.8092,
      "mean_token_accuracy": 0.18136782944202423,
      "num_tokens": 36293159.0,
      "step": 14320
    },
    {
      "entropy": 6.0950737476348875,
      "epoch": 1.653202538949798,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047366062534189756,
      "loss": 5.9417,
      "mean_token_accuracy": 0.1704432874917984,
      "num_tokens": 36306492.0,
      "step": 14325
    },
    {
      "entropy": 6.072968339920044,
      "epoch": 1.6537795729948068,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047364124867041446,
      "loss": 5.8755,
      "mean_token_accuracy": 0.171821166574955,
      "num_tokens": 36319020.0,
      "step": 14330
    },
    {
      "entropy": 6.041807365417481,
      "epoch": 1.6543566070398152,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047362186531773156,
      "loss": 5.9489,
      "mean_token_accuracy": 0.16833515465259552,
      "num_tokens": 36332163.0,
      "step": 14335
    },
    {
      "entropy": 6.080481386184692,
      "epoch": 1.654933641084824,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004736024752845008,
      "loss": 5.8727,
      "mean_token_accuracy": 0.18257838487625122,
      "num_tokens": 36345880.0,
      "step": 14340
    },
    {
      "entropy": 6.046308517456055,
      "epoch": 1.6555106751298325,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004735830785713746,
      "loss": 5.8633,
      "mean_token_accuracy": 0.18226323574781417,
      "num_tokens": 36358927.0,
      "step": 14345
    },
    {
      "entropy": 6.069298171997071,
      "epoch": 1.6560877091748414,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004735636751790051,
      "loss": 5.9088,
      "mean_token_accuracy": 0.1722438171505928,
      "num_tokens": 36371009.0,
      "step": 14350
    },
    {
      "entropy": 6.134785509109497,
      "epoch": 1.65666474321985,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000473544265108045,
      "loss": 5.9459,
      "mean_token_accuracy": 0.17300676107406615,
      "num_tokens": 36385538.0,
      "step": 14355
    },
    {
      "entropy": 6.040959262847901,
      "epoch": 1.6572417772648587,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047352484835914716,
      "loss": 5.8698,
      "mean_token_accuracy": 0.17529087662696838,
      "num_tokens": 36398265.0,
      "step": 14360
    },
    {
      "entropy": 6.039867496490478,
      "epoch": 1.6578188113098673,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004735054249329647,
      "loss": 5.8683,
      "mean_token_accuracy": 0.17899925261735916,
      "num_tokens": 36410691.0,
      "step": 14365
    },
    {
      "entropy": 6.072979259490967,
      "epoch": 1.658395845354876,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047348599483015087,
      "loss": 5.8335,
      "mean_token_accuracy": 0.1773272916674614,
      "num_tokens": 36423531.0,
      "step": 14370
    },
    {
      "entropy": 6.058085250854492,
      "epoch": 1.6589728793998846,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047346655805135916,
      "loss": 5.8908,
      "mean_token_accuracy": 0.18027470409870147,
      "num_tokens": 36437429.0,
      "step": 14375
    },
    {
      "entropy": 5.984629774093628,
      "epoch": 1.6595499134448932,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004734471145972434,
      "loss": 5.7802,
      "mean_token_accuracy": 0.1838893324136734,
      "num_tokens": 36449264.0,
      "step": 14380
    },
    {
      "entropy": 6.027268886566162,
      "epoch": 1.6601269474899019,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047342766446845753,
      "loss": 5.8098,
      "mean_token_accuracy": 0.18468562960624696,
      "num_tokens": 36461132.0,
      "step": 14385
    },
    {
      "entropy": 6.0705695152282715,
      "epoch": 1.6607039815349105,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004734082076656557,
      "loss": 5.9044,
      "mean_token_accuracy": 0.18105503767728806,
      "num_tokens": 36474195.0,
      "step": 14390
    },
    {
      "entropy": 6.082694721221924,
      "epoch": 1.6612810155799194,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047338874418949235,
      "loss": 5.9932,
      "mean_token_accuracy": 0.17038628160953523,
      "num_tokens": 36486197.0,
      "step": 14395
    },
    {
      "entropy": 6.041064357757568,
      "epoch": 1.6618580496249278,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047336927404062213,
      "loss": 5.8173,
      "mean_token_accuracy": 0.18693056255578994,
      "num_tokens": 36499215.0,
      "step": 14400
    },
    {
      "entropy": 6.065183639526367,
      "epoch": 1.6624350836699366,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047334979721969995,
      "loss": 5.9295,
      "mean_token_accuracy": 0.17181039452552796,
      "num_tokens": 36511872.0,
      "step": 14405
    },
    {
      "entropy": 6.11459641456604,
      "epoch": 1.663012117714945,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004733303137273808,
      "loss": 5.923,
      "mean_token_accuracy": 0.1792657658457756,
      "num_tokens": 36524313.0,
      "step": 14410
    },
    {
      "entropy": 5.990611362457275,
      "epoch": 1.663589151759954,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047331082356432015,
      "loss": 5.8534,
      "mean_token_accuracy": 0.18302565366029738,
      "num_tokens": 36538532.0,
      "step": 14415
    },
    {
      "entropy": 6.1371361255645756,
      "epoch": 1.6641661858049623,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004732913267311734,
      "loss": 5.9077,
      "mean_token_accuracy": 0.17350229173898696,
      "num_tokens": 36551117.0,
      "step": 14420
    },
    {
      "entropy": 6.087515830993652,
      "epoch": 1.6647432198499712,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004732718232285964,
      "loss": 5.7981,
      "mean_token_accuracy": 0.17448867708444596,
      "num_tokens": 36563704.0,
      "step": 14425
    },
    {
      "entropy": 5.9836955070495605,
      "epoch": 1.6653202538949798,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047325231305724507,
      "loss": 5.9442,
      "mean_token_accuracy": 0.175588358938694,
      "num_tokens": 36575952.0,
      "step": 14430
    },
    {
      "entropy": 6.055558824539185,
      "epoch": 1.6658972879399885,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004732327962177757,
      "loss": 5.7742,
      "mean_token_accuracy": 0.18464993983507155,
      "num_tokens": 36590181.0,
      "step": 14435
    },
    {
      "entropy": 6.026779365539551,
      "epoch": 1.6664743219849971,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004732132727108447,
      "loss": 5.8574,
      "mean_token_accuracy": 0.17366963624954224,
      "num_tokens": 36601986.0,
      "step": 14440
    },
    {
      "entropy": 6.104998397827148,
      "epoch": 1.6670513560300058,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047319374253710874,
      "loss": 5.9164,
      "mean_token_accuracy": 0.1731383353471756,
      "num_tokens": 36614135.0,
      "step": 14445
    },
    {
      "entropy": 6.112477445602417,
      "epoch": 1.6676283900750144,
      "grad_norm": 1.0,
      "learning_rate": 0.0004731742056972247,
      "loss": 5.9529,
      "mean_token_accuracy": 0.17732008397579194,
      "num_tokens": 36625952.0,
      "step": 14450
    },
    {
      "entropy": 6.090760517120361,
      "epoch": 1.668205424120023,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004731546621918497,
      "loss": 5.9092,
      "mean_token_accuracy": 0.1757805034518242,
      "num_tokens": 36638945.0,
      "step": 14455
    },
    {
      "entropy": 6.128685569763183,
      "epoch": 1.668782458165032,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000473135112021641,
      "loss": 5.9352,
      "mean_token_accuracy": 0.17241512686014177,
      "num_tokens": 36652550.0,
      "step": 14460
    },
    {
      "entropy": 6.093085479736328,
      "epoch": 1.6693594922100403,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047311555518725617,
      "loss": 5.8892,
      "mean_token_accuracy": 0.17684196978807448,
      "num_tokens": 36665260.0,
      "step": 14465
    },
    {
      "entropy": 6.1125630855560305,
      "epoch": 1.6699365262550492,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047309599168935323,
      "loss": 5.8716,
      "mean_token_accuracy": 0.17868396192789077,
      "num_tokens": 36679275.0,
      "step": 14470
    },
    {
      "entropy": 6.002921104431152,
      "epoch": 1.6705135603000576,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047307642152858993,
      "loss": 5.8352,
      "mean_token_accuracy": 0.18537280708551407,
      "num_tokens": 36691849.0,
      "step": 14475
    },
    {
      "entropy": 6.064547443389893,
      "epoch": 1.6710905943450665,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047305684470562453,
      "loss": 5.8603,
      "mean_token_accuracy": 0.18693628311157226,
      "num_tokens": 36703618.0,
      "step": 14480
    },
    {
      "entropy": 6.034905576705933,
      "epoch": 1.6716676283900749,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004730372612211156,
      "loss": 5.8465,
      "mean_token_accuracy": 0.17593854963779448,
      "num_tokens": 36716332.0,
      "step": 14485
    },
    {
      "entropy": 5.945816850662231,
      "epoch": 1.6722446624350837,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047301767107572174,
      "loss": 5.7929,
      "mean_token_accuracy": 0.18382909744977952,
      "num_tokens": 36729136.0,
      "step": 14490
    },
    {
      "entropy": 6.028561925888061,
      "epoch": 1.6728216964800924,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004729980742701018,
      "loss": 5.9196,
      "mean_token_accuracy": 0.18047868013381957,
      "num_tokens": 36743061.0,
      "step": 14495
    },
    {
      "entropy": 6.136689281463623,
      "epoch": 1.673398730525101,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004729784708049151,
      "loss": 5.863,
      "mean_token_accuracy": 0.18827371597290038,
      "num_tokens": 36756151.0,
      "step": 14500
    },
    {
      "entropy": 6.090974473953247,
      "epoch": 1.6739757645701097,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004729588606808209,
      "loss": 5.8794,
      "mean_token_accuracy": 0.17531039863824843,
      "num_tokens": 36768008.0,
      "step": 14505
    },
    {
      "entropy": 5.977316761016846,
      "epoch": 1.6745527986151183,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047293924389847864,
      "loss": 5.7738,
      "mean_token_accuracy": 0.1848136678338051,
      "num_tokens": 36780694.0,
      "step": 14510
    },
    {
      "entropy": 5.948745536804199,
      "epoch": 1.675129832660127,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004729196204585483,
      "loss": 5.8401,
      "mean_token_accuracy": 0.18342776596546173,
      "num_tokens": 36793040.0,
      "step": 14515
    },
    {
      "entropy": 6.070938205718994,
      "epoch": 1.6757068667051356,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047289999036168983,
      "loss": 5.9284,
      "mean_token_accuracy": 0.1756843239068985,
      "num_tokens": 36806293.0,
      "step": 14520
    },
    {
      "entropy": 6.07716064453125,
      "epoch": 1.6762839007501442,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004728803536085634,
      "loss": 5.8347,
      "mean_token_accuracy": 0.1787402883172035,
      "num_tokens": 36818717.0,
      "step": 14525
    },
    {
      "entropy": 6.068575143814087,
      "epoch": 1.6768609347951529,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047286071019982974,
      "loss": 5.8723,
      "mean_token_accuracy": 0.17294562757015228,
      "num_tokens": 36832984.0,
      "step": 14530
    },
    {
      "entropy": 6.046890497207642,
      "epoch": 1.6774379688401617,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047284106013614926,
      "loss": 5.8309,
      "mean_token_accuracy": 0.1792532667517662,
      "num_tokens": 36844088.0,
      "step": 14535
    },
    {
      "entropy": 6.081466150283814,
      "epoch": 1.6780150028851701,
      "grad_norm": 0.90625,
      "learning_rate": 0.000472821403418183,
      "loss": 5.965,
      "mean_token_accuracy": 0.1728912800550461,
      "num_tokens": 36856846.0,
      "step": 14540
    },
    {
      "entropy": 6.088641405105591,
      "epoch": 1.678592036930179,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004728017400465921,
      "loss": 5.8383,
      "mean_token_accuracy": 0.17599605470895768,
      "num_tokens": 36870023.0,
      "step": 14545
    },
    {
      "entropy": 6.089526891708374,
      "epoch": 1.6791690709751874,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047278207002203796,
      "loss": 5.9216,
      "mean_token_accuracy": 0.17570875734090804,
      "num_tokens": 36883298.0,
      "step": 14550
    },
    {
      "entropy": 6.060133600234986,
      "epoch": 1.6797461050201963,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047276239334518216,
      "loss": 5.8835,
      "mean_token_accuracy": 0.1784228801727295,
      "num_tokens": 36894714.0,
      "step": 14555
    },
    {
      "entropy": 5.959715938568115,
      "epoch": 1.6803231390652047,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047274271001668646,
      "loss": 5.7355,
      "mean_token_accuracy": 0.19264876544475557,
      "num_tokens": 36908332.0,
      "step": 14560
    },
    {
      "entropy": 6.018507194519043,
      "epoch": 1.6809001731102136,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047272302003721286,
      "loss": 5.8497,
      "mean_token_accuracy": 0.17790351808071136,
      "num_tokens": 36921235.0,
      "step": 14565
    },
    {
      "entropy": 6.07934799194336,
      "epoch": 1.6814772071552222,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047270332340742377,
      "loss": 5.939,
      "mean_token_accuracy": 0.16910149306058883,
      "num_tokens": 36934580.0,
      "step": 14570
    },
    {
      "entropy": 6.098540496826172,
      "epoch": 1.6820542412002308,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047268362012798157,
      "loss": 5.8792,
      "mean_token_accuracy": 0.17883206754922867,
      "num_tokens": 36946936.0,
      "step": 14575
    },
    {
      "entropy": 5.926387786865234,
      "epoch": 1.6826312752452395,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004726639101995491,
      "loss": 5.7996,
      "mean_token_accuracy": 0.18647063821554183,
      "num_tokens": 36960179.0,
      "step": 14580
    },
    {
      "entropy": 6.082691478729248,
      "epoch": 1.6832083092902481,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047264419362278906,
      "loss": 5.9071,
      "mean_token_accuracy": 0.18075298964977266,
      "num_tokens": 36972739.0,
      "step": 14585
    },
    {
      "entropy": 5.994677591323852,
      "epoch": 1.6837853433352568,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047262447039836484,
      "loss": 5.8025,
      "mean_token_accuracy": 0.17602844387292862,
      "num_tokens": 36987314.0,
      "step": 14590
    },
    {
      "entropy": 6.057880592346192,
      "epoch": 1.6843623773802654,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047260474052693963,
      "loss": 5.872,
      "mean_token_accuracy": 0.17974022775888443,
      "num_tokens": 36999050.0,
      "step": 14595
    },
    {
      "entropy": 6.053677320480347,
      "epoch": 1.684939411425274,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047258500400917724,
      "loss": 5.8391,
      "mean_token_accuracy": 0.17832357585430145,
      "num_tokens": 37011905.0,
      "step": 14600
    },
    {
      "entropy": 6.112119817733765,
      "epoch": 1.6855164454702827,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047256526084574137,
      "loss": 5.8921,
      "mean_token_accuracy": 0.17971248924732208,
      "num_tokens": 37023926.0,
      "step": 14605
    },
    {
      "entropy": 6.095492172241211,
      "epoch": 1.6860934795152915,
      "grad_norm": 1.0,
      "learning_rate": 0.00047254551103729597,
      "loss": 5.8998,
      "mean_token_accuracy": 0.17007645219564438,
      "num_tokens": 37037734.0,
      "step": 14610
    },
    {
      "entropy": 6.109078693389892,
      "epoch": 1.6866705135603,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004725257545845055,
      "loss": 5.9543,
      "mean_token_accuracy": 0.17537282556295394,
      "num_tokens": 37049689.0,
      "step": 14615
    },
    {
      "entropy": 6.134923458099365,
      "epoch": 1.6872475476053088,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047250599148803443,
      "loss": 5.9625,
      "mean_token_accuracy": 0.1695738360285759,
      "num_tokens": 37063999.0,
      "step": 14620
    },
    {
      "entropy": 5.962371873855591,
      "epoch": 1.6878245816503172,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047248622174854746,
      "loss": 5.8302,
      "mean_token_accuracy": 0.17880164384841918,
      "num_tokens": 37076855.0,
      "step": 14625
    },
    {
      "entropy": 6.1008376121521,
      "epoch": 1.688401615695326,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004724664453667094,
      "loss": 5.8798,
      "mean_token_accuracy": 0.18154750764369965,
      "num_tokens": 37089235.0,
      "step": 14630
    },
    {
      "entropy": 6.078764152526856,
      "epoch": 1.6889786497403345,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004724466623431857,
      "loss": 5.9164,
      "mean_token_accuracy": 0.17736300975084304,
      "num_tokens": 37102011.0,
      "step": 14635
    },
    {
      "entropy": 5.988099002838135,
      "epoch": 1.6895556837853434,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004724268726786415,
      "loss": 5.781,
      "mean_token_accuracy": 0.18803667277097702,
      "num_tokens": 37115246.0,
      "step": 14640
    },
    {
      "entropy": 6.137476205825806,
      "epoch": 1.690132717830352,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004724070763737424,
      "loss": 5.9979,
      "mean_token_accuracy": 0.1695678263902664,
      "num_tokens": 37127705.0,
      "step": 14645
    },
    {
      "entropy": 6.0985329151153564,
      "epoch": 1.6907097518753607,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004723872734291545,
      "loss": 5.8435,
      "mean_token_accuracy": 0.18003033399581908,
      "num_tokens": 37141136.0,
      "step": 14650
    },
    {
      "entropy": 6.008995246887207,
      "epoch": 1.6912867859203693,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047236746384554364,
      "loss": 5.8218,
      "mean_token_accuracy": 0.17580111175775529,
      "num_tokens": 37154242.0,
      "step": 14655
    },
    {
      "entropy": 6.004652976989746,
      "epoch": 1.691863819965378,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004723476476235762,
      "loss": 5.8264,
      "mean_token_accuracy": 0.18464961051940917,
      "num_tokens": 37166254.0,
      "step": 14660
    },
    {
      "entropy": 6.119229936599732,
      "epoch": 1.6924408540103866,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004723278247639186,
      "loss": 5.9127,
      "mean_token_accuracy": 0.176186203956604,
      "num_tokens": 37179579.0,
      "step": 14665
    },
    {
      "entropy": 6.020057678222656,
      "epoch": 1.6930178880553952,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004723079952672377,
      "loss": 5.8213,
      "mean_token_accuracy": 0.18570153713226317,
      "num_tokens": 37191478.0,
      "step": 14670
    },
    {
      "entropy": 5.968469667434692,
      "epoch": 1.693594922100404,
      "grad_norm": 1.0,
      "learning_rate": 0.00047228815913420035,
      "loss": 5.7154,
      "mean_token_accuracy": 0.18630415946245193,
      "num_tokens": 37203785.0,
      "step": 14675
    },
    {
      "entropy": 6.031087923049927,
      "epoch": 1.6941719561454125,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004722683163654738,
      "loss": 5.7934,
      "mean_token_accuracy": 0.18254295289516448,
      "num_tokens": 37216298.0,
      "step": 14680
    },
    {
      "entropy": 6.083994150161743,
      "epoch": 1.6947489901904214,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004722484669617254,
      "loss": 5.9019,
      "mean_token_accuracy": 0.18348269611597062,
      "num_tokens": 37228659.0,
      "step": 14685
    },
    {
      "entropy": 6.013123941421509,
      "epoch": 1.6953260242354298,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047222861092362277,
      "loss": 5.8683,
      "mean_token_accuracy": 0.18049643635749818,
      "num_tokens": 37240763.0,
      "step": 14690
    },
    {
      "entropy": 6.0135657787323,
      "epoch": 1.6959030582804386,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047220874825183387,
      "loss": 5.7905,
      "mean_token_accuracy": 0.18084255009889602,
      "num_tokens": 37252238.0,
      "step": 14695
    },
    {
      "entropy": 6.0948036193847654,
      "epoch": 1.696480092325447,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047218887894702656,
      "loss": 5.9224,
      "mean_token_accuracy": 0.174990476667881,
      "num_tokens": 37264475.0,
      "step": 14700
    },
    {
      "entropy": 6.074303007125854,
      "epoch": 1.697057126370456,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004721690030098693,
      "loss": 5.8574,
      "mean_token_accuracy": 0.1801070123910904,
      "num_tokens": 37277338.0,
      "step": 14705
    },
    {
      "entropy": 6.077761459350586,
      "epoch": 1.6976341604154646,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004721491204410305,
      "loss": 5.9079,
      "mean_token_accuracy": 0.17569748163223267,
      "num_tokens": 37289985.0,
      "step": 14710
    },
    {
      "entropy": 6.086352825164795,
      "epoch": 1.6982111944604732,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047212923124117915,
      "loss": 5.912,
      "mean_token_accuracy": 0.17878119349479676,
      "num_tokens": 37303177.0,
      "step": 14715
    },
    {
      "entropy": 6.072368812561035,
      "epoch": 1.6987882285054818,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004721093354109839,
      "loss": 5.8335,
      "mean_token_accuracy": 0.18163759261369705,
      "num_tokens": 37316662.0,
      "step": 14720
    },
    {
      "entropy": 6.073123598098755,
      "epoch": 1.6993652625504905,
      "grad_norm": 1.0,
      "learning_rate": 0.00047208943295111406,
      "loss": 5.8632,
      "mean_token_accuracy": 0.18174527436494828,
      "num_tokens": 37327782.0,
      "step": 14725
    },
    {
      "entropy": 6.039751815795898,
      "epoch": 1.6999422965954991,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047206952386223905,
      "loss": 5.7697,
      "mean_token_accuracy": 0.18593833446502686,
      "num_tokens": 37340650.0,
      "step": 14730
    },
    {
      "entropy": 6.022388172149658,
      "epoch": 1.7005193306405078,
      "grad_norm": 0.875,
      "learning_rate": 0.0004720496081450285,
      "loss": 5.9096,
      "mean_token_accuracy": 0.17580284625291825,
      "num_tokens": 37354330.0,
      "step": 14735
    },
    {
      "entropy": 6.085542392730713,
      "epoch": 1.7010963646855164,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047202968580015224,
      "loss": 5.8827,
      "mean_token_accuracy": 0.17684022337198257,
      "num_tokens": 37367447.0,
      "step": 14740
    },
    {
      "entropy": 6.071901750564575,
      "epoch": 1.701673398730525,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047200975682828045,
      "loss": 5.8098,
      "mean_token_accuracy": 0.18558563739061357,
      "num_tokens": 37380171.0,
      "step": 14745
    },
    {
      "entropy": 6.072594785690308,
      "epoch": 1.702250432775534,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004719898212300832,
      "loss": 5.8835,
      "mean_token_accuracy": 0.17697048783302308,
      "num_tokens": 37393042.0,
      "step": 14750
    },
    {
      "entropy": 6.075764942169189,
      "epoch": 1.7028274668205423,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047196987900623134,
      "loss": 5.9396,
      "mean_token_accuracy": 0.1766821950674057,
      "num_tokens": 37406854.0,
      "step": 14755
    },
    {
      "entropy": 6.098237037658691,
      "epoch": 1.7034045008655512,
      "grad_norm": 1.0,
      "learning_rate": 0.00047194993015739527,
      "loss": 5.9266,
      "mean_token_accuracy": 0.17287011593580245,
      "num_tokens": 37419765.0,
      "step": 14760
    },
    {
      "entropy": 6.10421347618103,
      "epoch": 1.7039815349105596,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047192997468424624,
      "loss": 5.8561,
      "mean_token_accuracy": 0.18376631438732147,
      "num_tokens": 37432280.0,
      "step": 14765
    },
    {
      "entropy": 6.080137395858765,
      "epoch": 1.7045585689555685,
      "grad_norm": 1.0,
      "learning_rate": 0.0004719100125874553,
      "loss": 5.9377,
      "mean_token_accuracy": 0.17264190763235093,
      "num_tokens": 37444315.0,
      "step": 14770
    },
    {
      "entropy": 6.131002378463745,
      "epoch": 1.7051356030005769,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004718900438676939,
      "loss": 5.9443,
      "mean_token_accuracy": 0.1707395553588867,
      "num_tokens": 37457588.0,
      "step": 14775
    },
    {
      "entropy": 6.086152410507202,
      "epoch": 1.7057126370455857,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004718700685256337,
      "loss": 5.864,
      "mean_token_accuracy": 0.17592869848012924,
      "num_tokens": 37471740.0,
      "step": 14780
    },
    {
      "entropy": 6.053951072692871,
      "epoch": 1.7062896710905944,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004718500865619465,
      "loss": 5.8476,
      "mean_token_accuracy": 0.1741065576672554,
      "num_tokens": 37484835.0,
      "step": 14785
    },
    {
      "entropy": 6.080347156524658,
      "epoch": 1.706866705135603,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004718300979773044,
      "loss": 5.8631,
      "mean_token_accuracy": 0.17933740317821503,
      "num_tokens": 37497987.0,
      "step": 14790
    },
    {
      "entropy": 6.07176833152771,
      "epoch": 1.7074437391806117,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047181010277237977,
      "loss": 5.8596,
      "mean_token_accuracy": 0.18199979662895202,
      "num_tokens": 37510454.0,
      "step": 14795
    },
    {
      "entropy": 6.0047344207763675,
      "epoch": 1.7080207732256203,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004717901009478451,
      "loss": 5.8146,
      "mean_token_accuracy": 0.18236078917980195,
      "num_tokens": 37523959.0,
      "step": 14800
    },
    {
      "entropy": 6.115893602371216,
      "epoch": 1.708597807270629,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047177009250437313,
      "loss": 5.953,
      "mean_token_accuracy": 0.1759780988097191,
      "num_tokens": 37538116.0,
      "step": 14805
    },
    {
      "entropy": 6.126485967636109,
      "epoch": 1.7091748413156376,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047175007744263683,
      "loss": 5.8709,
      "mean_token_accuracy": 0.17192399650812148,
      "num_tokens": 37549940.0,
      "step": 14810
    },
    {
      "entropy": 6.07864203453064,
      "epoch": 1.7097518753606464,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047173005576330935,
      "loss": 5.8909,
      "mean_token_accuracy": 0.1718295007944107,
      "num_tokens": 37562854.0,
      "step": 14815
    },
    {
      "entropy": 5.98805422782898,
      "epoch": 1.7103289094056549,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047171002746706424,
      "loss": 5.7723,
      "mean_token_accuracy": 0.18562911450862885,
      "num_tokens": 37575639.0,
      "step": 14820
    },
    {
      "entropy": 6.0082015037536625,
      "epoch": 1.7109059434506637,
      "grad_norm": 0.84375,
      "learning_rate": 0.00047168999255457506,
      "loss": 5.8011,
      "mean_token_accuracy": 0.18458645641803742,
      "num_tokens": 37588555.0,
      "step": 14825
    },
    {
      "entropy": 6.111817455291748,
      "epoch": 1.7114829774956721,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047166995102651565,
      "loss": 5.8946,
      "mean_token_accuracy": 0.17755903005599977,
      "num_tokens": 37602636.0,
      "step": 14830
    },
    {
      "entropy": 6.002731418609619,
      "epoch": 1.712060011540681,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004716499028835601,
      "loss": 5.839,
      "mean_token_accuracy": 0.17560880929231643,
      "num_tokens": 37616769.0,
      "step": 14835
    },
    {
      "entropy": 6.15225510597229,
      "epoch": 1.7126370455856894,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004716298481263828,
      "loss": 5.9062,
      "mean_token_accuracy": 0.17974839508533477,
      "num_tokens": 37628173.0,
      "step": 14840
    },
    {
      "entropy": 6.052190113067627,
      "epoch": 1.7132140796306983,
      "grad_norm": 1.0,
      "learning_rate": 0.0004716097867556583,
      "loss": 5.8269,
      "mean_token_accuracy": 0.17960754930973052,
      "num_tokens": 37640953.0,
      "step": 14845
    },
    {
      "entropy": 6.0859509944915775,
      "epoch": 1.713791113675707,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047158971877206114,
      "loss": 5.9073,
      "mean_token_accuracy": 0.17382419407367705,
      "num_tokens": 37655091.0,
      "step": 14850
    },
    {
      "entropy": 6.149226808547974,
      "epoch": 1.7143681477207156,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004715696441762665,
      "loss": 5.9137,
      "mean_token_accuracy": 0.17798158675432205,
      "num_tokens": 37668120.0,
      "step": 14855
    },
    {
      "entropy": 6.060831880569458,
      "epoch": 1.7149451817657242,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047154956296894954,
      "loss": 5.7545,
      "mean_token_accuracy": 0.18553408086299897,
      "num_tokens": 37679844.0,
      "step": 14860
    },
    {
      "entropy": 5.999103307723999,
      "epoch": 1.7155222158107328,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004715294751507856,
      "loss": 5.8836,
      "mean_token_accuracy": 0.17819452583789824,
      "num_tokens": 37692756.0,
      "step": 14865
    },
    {
      "entropy": 6.108115148544312,
      "epoch": 1.7160992498557415,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004715093807224505,
      "loss": 5.9213,
      "mean_token_accuracy": 0.17663307040929793,
      "num_tokens": 37704525.0,
      "step": 14870
    },
    {
      "entropy": 6.103940486907959,
      "epoch": 1.71667628390075,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047148927968462,
      "loss": 5.898,
      "mean_token_accuracy": 0.17944020181894302,
      "num_tokens": 37715980.0,
      "step": 14875
    },
    {
      "entropy": 6.01836576461792,
      "epoch": 1.7172533179457588,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047146917203797,
      "loss": 5.7528,
      "mean_token_accuracy": 0.18468107432126998,
      "num_tokens": 37728448.0,
      "step": 14880
    },
    {
      "entropy": 6.06011734008789,
      "epoch": 1.7178303519907674,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004714490577831771,
      "loss": 5.8458,
      "mean_token_accuracy": 0.17664414644241333,
      "num_tokens": 37741798.0,
      "step": 14885
    },
    {
      "entropy": 5.990115165710449,
      "epoch": 1.7184073860357763,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004714289369209177,
      "loss": 5.79,
      "mean_token_accuracy": 0.18968200087547302,
      "num_tokens": 37755056.0,
      "step": 14890
    },
    {
      "entropy": 6.100113105773926,
      "epoch": 1.7189844200807847,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047140880945186863,
      "loss": 5.9336,
      "mean_token_accuracy": 0.16627510786056518,
      "num_tokens": 37767554.0,
      "step": 14895
    },
    {
      "entropy": 6.136945533752441,
      "epoch": 1.7195614541257935,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047138867537670676,
      "loss": 5.8896,
      "mean_token_accuracy": 0.17914667278528212,
      "num_tokens": 37781430.0,
      "step": 14900
    },
    {
      "entropy": 5.905944919586181,
      "epoch": 1.720138488170802,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047136853469610933,
      "loss": 5.7394,
      "mean_token_accuracy": 0.18020859062671662,
      "num_tokens": 37793138.0,
      "step": 14905
    },
    {
      "entropy": 6.079594564437866,
      "epoch": 1.7207155222158108,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047134838741075383,
      "loss": 5.9191,
      "mean_token_accuracy": 0.17323111444711686,
      "num_tokens": 37806685.0,
      "step": 14910
    },
    {
      "entropy": 5.965782642364502,
      "epoch": 1.7212925562608192,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047132823352131787,
      "loss": 5.7454,
      "mean_token_accuracy": 0.1905008301138878,
      "num_tokens": 37818571.0,
      "step": 14915
    },
    {
      "entropy": 6.015972471237182,
      "epoch": 1.721869590305828,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004713080730284793,
      "loss": 5.8796,
      "mean_token_accuracy": 0.1792006552219391,
      "num_tokens": 37831245.0,
      "step": 14920
    },
    {
      "entropy": 6.14883861541748,
      "epoch": 1.7224466243508367,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047128790593291617,
      "loss": 5.9241,
      "mean_token_accuracy": 0.17393770515918733,
      "num_tokens": 37844140.0,
      "step": 14925
    },
    {
      "entropy": 6.134219980239868,
      "epoch": 1.7230236583958454,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047126773223530677,
      "loss": 5.9253,
      "mean_token_accuracy": 0.1774785429239273,
      "num_tokens": 37856783.0,
      "step": 14930
    },
    {
      "entropy": 6.0744922161102295,
      "epoch": 1.723600692440854,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047124755193632975,
      "loss": 5.921,
      "mean_token_accuracy": 0.1763788342475891,
      "num_tokens": 37871219.0,
      "step": 14935
    },
    {
      "entropy": 6.035473346710205,
      "epoch": 1.7241777264858626,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047122736503666377,
      "loss": 5.8722,
      "mean_token_accuracy": 0.18346308022737504,
      "num_tokens": 37882510.0,
      "step": 14940
    },
    {
      "entropy": 6.123553085327148,
      "epoch": 1.7247547605308713,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004712071715369878,
      "loss": 5.9033,
      "mean_token_accuracy": 0.17857804298400878,
      "num_tokens": 37894785.0,
      "step": 14945
    },
    {
      "entropy": 5.991181516647339,
      "epoch": 1.72533179457588,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004711869714379812,
      "loss": 5.8046,
      "mean_token_accuracy": 0.17936460077762603,
      "num_tokens": 37907598.0,
      "step": 14950
    },
    {
      "entropy": 6.020657300949097,
      "epoch": 1.7259088286208888,
      "grad_norm": 1.0,
      "learning_rate": 0.0004711667647403232,
      "loss": 5.8365,
      "mean_token_accuracy": 0.1728009521961212,
      "num_tokens": 37920399.0,
      "step": 14955
    },
    {
      "entropy": 6.106961727142334,
      "epoch": 1.7264858626658972,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047114655144469354,
      "loss": 5.9179,
      "mean_token_accuracy": 0.17654111087322236,
      "num_tokens": 37934157.0,
      "step": 14960
    },
    {
      "entropy": 6.074722576141357,
      "epoch": 1.727062896710906,
      "grad_norm": 1.0,
      "learning_rate": 0.00047112633155177203,
      "loss": 5.8089,
      "mean_token_accuracy": 0.17849755436182022,
      "num_tokens": 37946058.0,
      "step": 14965
    },
    {
      "entropy": 6.057895278930664,
      "epoch": 1.7276399307559145,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004711061050622388,
      "loss": 5.8791,
      "mean_token_accuracy": 0.174577496945858,
      "num_tokens": 37958549.0,
      "step": 14970
    },
    {
      "entropy": 6.057162570953369,
      "epoch": 1.7282169648009233,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047108587197677404,
      "loss": 5.8576,
      "mean_token_accuracy": 0.1771823614835739,
      "num_tokens": 37971633.0,
      "step": 14975
    },
    {
      "entropy": 6.040497255325318,
      "epoch": 1.7287939988459318,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047106563229605845,
      "loss": 5.7853,
      "mean_token_accuracy": 0.18077824264764786,
      "num_tokens": 37984048.0,
      "step": 14980
    },
    {
      "entropy": 6.078268718719483,
      "epoch": 1.7293710328909406,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047104538602077276,
      "loss": 5.8765,
      "mean_token_accuracy": 0.17665274292230607,
      "num_tokens": 37996092.0,
      "step": 14985
    },
    {
      "entropy": 5.972409725189209,
      "epoch": 1.7299480669359493,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004710251331515978,
      "loss": 5.8049,
      "mean_token_accuracy": 0.18591080158948897,
      "num_tokens": 38008769.0,
      "step": 14990
    },
    {
      "entropy": 6.038018274307251,
      "epoch": 1.730525100980958,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047100487368921485,
      "loss": 5.7818,
      "mean_token_accuracy": 0.1865561082959175,
      "num_tokens": 38021344.0,
      "step": 14995
    },
    {
      "entropy": 6.137103271484375,
      "epoch": 1.7311021350259665,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004709846076343055,
      "loss": 5.8879,
      "mean_token_accuracy": 0.1766980767250061,
      "num_tokens": 38033839.0,
      "step": 15000
    },
    {
      "epoch": 1.7311021350259665,
      "eval_entropy": 5.966758595226094,
      "eval_loss": 5.98366117477417,
      "eval_mean_token_accuracy": 0.17819551700916283,
      "eval_num_tokens": 38033839.0,
      "eval_runtime": 22.8972,
      "eval_samples_per_second": 1228.796,
      "eval_steps_per_second": 153.599,
      "step": 15000
    },
    {
      "entropy": 6.08551926612854,
      "epoch": 1.7316791690709752,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047096433498755103,
      "loss": 5.9192,
      "mean_token_accuracy": 0.1751316651701927,
      "num_tokens": 38047026.0,
      "step": 15005
    },
    {
      "entropy": 6.01678729057312,
      "epoch": 1.7322562031159838,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047094405574963364,
      "loss": 5.831,
      "mean_token_accuracy": 0.18445987105369568,
      "num_tokens": 38059089.0,
      "step": 15010
    },
    {
      "entropy": 6.022669458389283,
      "epoch": 1.7328332371609925,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047092376992123516,
      "loss": 5.8318,
      "mean_token_accuracy": 0.17793042063713074,
      "num_tokens": 38072925.0,
      "step": 15015
    },
    {
      "entropy": 6.052866744995117,
      "epoch": 1.733410271206001,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047090347750303803,
      "loss": 5.8388,
      "mean_token_accuracy": 0.18262719064950944,
      "num_tokens": 38084532.0,
      "step": 15020
    },
    {
      "entropy": 6.069236326217651,
      "epoch": 1.7339873052510097,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004708831784957247,
      "loss": 5.8651,
      "mean_token_accuracy": 0.17873544842004777,
      "num_tokens": 38098024.0,
      "step": 15025
    },
    {
      "entropy": 6.04683837890625,
      "epoch": 1.7345643392960186,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004708628728999781,
      "loss": 5.8638,
      "mean_token_accuracy": 0.17402407824993132,
      "num_tokens": 38110718.0,
      "step": 15030
    },
    {
      "entropy": 6.090602827072144,
      "epoch": 1.735141373341027,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004708425607164809,
      "loss": 5.9249,
      "mean_token_accuracy": 0.17607097029685975,
      "num_tokens": 38124807.0,
      "step": 15035
    },
    {
      "entropy": 6.07017502784729,
      "epoch": 1.7357184073860359,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004708222419459165,
      "loss": 5.8618,
      "mean_token_accuracy": 0.17373752743005752,
      "num_tokens": 38136660.0,
      "step": 15040
    },
    {
      "entropy": 6.035074377059937,
      "epoch": 1.7362954414310443,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004708019165889683,
      "loss": 5.8944,
      "mean_token_accuracy": 0.17979559451341628,
      "num_tokens": 38148154.0,
      "step": 15045
    },
    {
      "entropy": 6.0491241931915285,
      "epoch": 1.7368724754760532,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004707815846463199,
      "loss": 5.776,
      "mean_token_accuracy": 0.18361513465642929,
      "num_tokens": 38161092.0,
      "step": 15050
    },
    {
      "entropy": 6.050814056396485,
      "epoch": 1.7374495095210616,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004707612461186552,
      "loss": 5.9027,
      "mean_token_accuracy": 0.1813490331172943,
      "num_tokens": 38172349.0,
      "step": 15055
    },
    {
      "entropy": 6.065812826156616,
      "epoch": 1.7380265435660704,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047074090100665805,
      "loss": 5.902,
      "mean_token_accuracy": 0.17493660598993302,
      "num_tokens": 38184266.0,
      "step": 15060
    },
    {
      "entropy": 5.978083992004395,
      "epoch": 1.738603577611079,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047072054931101306,
      "loss": 5.7511,
      "mean_token_accuracy": 0.18896911889314652,
      "num_tokens": 38196865.0,
      "step": 15065
    },
    {
      "entropy": 6.043002557754517,
      "epoch": 1.7391806116560877,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004707001910324046,
      "loss": 5.8064,
      "mean_token_accuracy": 0.18126082122325898,
      "num_tokens": 38208416.0,
      "step": 15070
    },
    {
      "entropy": 6.043419027328492,
      "epoch": 1.7397576457010964,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047067982617151737,
      "loss": 5.8338,
      "mean_token_accuracy": 0.1798071011900902,
      "num_tokens": 38220649.0,
      "step": 15075
    },
    {
      "entropy": 6.055264329910278,
      "epoch": 1.740334679746105,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004706594547290365,
      "loss": 5.8592,
      "mean_token_accuracy": 0.18013397455215455,
      "num_tokens": 38233538.0,
      "step": 15080
    },
    {
      "entropy": 5.978543424606324,
      "epoch": 1.7409117137911136,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047063907670564695,
      "loss": 5.7211,
      "mean_token_accuracy": 0.18684158772230147,
      "num_tokens": 38245603.0,
      "step": 15085
    },
    {
      "entropy": 6.0789251804351805,
      "epoch": 1.7414887478361223,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004706186921020342,
      "loss": 5.914,
      "mean_token_accuracy": 0.1770313337445259,
      "num_tokens": 38258201.0,
      "step": 15090
    },
    {
      "entropy": 6.070990514755249,
      "epoch": 1.7420657818811311,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047059830091888407,
      "loss": 5.8371,
      "mean_token_accuracy": 0.1840934932231903,
      "num_tokens": 38270659.0,
      "step": 15095
    },
    {
      "entropy": 6.05364236831665,
      "epoch": 1.7426428159261396,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004705779031568821,
      "loss": 5.8504,
      "mean_token_accuracy": 0.17698458284139634,
      "num_tokens": 38283101.0,
      "step": 15100
    },
    {
      "entropy": 6.04548487663269,
      "epoch": 1.7432198499711484,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047055749881671463,
      "loss": 5.784,
      "mean_token_accuracy": 0.19270895123481752,
      "num_tokens": 38296855.0,
      "step": 15105
    },
    {
      "entropy": 5.98614821434021,
      "epoch": 1.7437968840161568,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004705370878990677,
      "loss": 5.7242,
      "mean_token_accuracy": 0.1890035167336464,
      "num_tokens": 38308354.0,
      "step": 15110
    },
    {
      "entropy": 6.01570873260498,
      "epoch": 1.7443739180611657,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047051667040462806,
      "loss": 5.887,
      "mean_token_accuracy": 0.17889729291200637,
      "num_tokens": 38321239.0,
      "step": 15115
    },
    {
      "entropy": 6.140709447860718,
      "epoch": 1.7449509521061741,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047049624633408224,
      "loss": 5.922,
      "mean_token_accuracy": 0.17176299542188644,
      "num_tokens": 38333672.0,
      "step": 15120
    },
    {
      "entropy": 6.036281156539917,
      "epoch": 1.745527986151183,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047047581568811724,
      "loss": 5.8611,
      "mean_token_accuracy": 0.1780383750796318,
      "num_tokens": 38346809.0,
      "step": 15125
    },
    {
      "entropy": 6.064457654953003,
      "epoch": 1.7461050201961916,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047045537846742043,
      "loss": 5.8557,
      "mean_token_accuracy": 0.17343237102031708,
      "num_tokens": 38359129.0,
      "step": 15130
    },
    {
      "entropy": 6.134432649612426,
      "epoch": 1.7466820542412003,
      "grad_norm": 0.984375,
      "learning_rate": 0.000470434934672679,
      "loss": 5.9598,
      "mean_token_accuracy": 0.17527367174625397,
      "num_tokens": 38371621.0,
      "step": 15135
    },
    {
      "entropy": 6.099580907821656,
      "epoch": 1.747259088286209,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047041448430458054,
      "loss": 5.9017,
      "mean_token_accuracy": 0.17456192374229432,
      "num_tokens": 38384579.0,
      "step": 15140
    },
    {
      "entropy": 6.134019136428833,
      "epoch": 1.7478361223312175,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047039402736381305,
      "loss": 5.9496,
      "mean_token_accuracy": 0.17220059782266617,
      "num_tokens": 38396765.0,
      "step": 15145
    },
    {
      "entropy": 6.077568864822387,
      "epoch": 1.7484131563762262,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004703735638510645,
      "loss": 5.833,
      "mean_token_accuracy": 0.18297814279794694,
      "num_tokens": 38409778.0,
      "step": 15150
    },
    {
      "entropy": 6.079481697082519,
      "epoch": 1.7489901904212348,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004703530937670232,
      "loss": 5.8424,
      "mean_token_accuracy": 0.1740758791565895,
      "num_tokens": 38421729.0,
      "step": 15155
    },
    {
      "entropy": 6.082099437713623,
      "epoch": 1.7495672244662435,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004703326171123776,
      "loss": 5.8518,
      "mean_token_accuracy": 0.17609713971614838,
      "num_tokens": 38435136.0,
      "step": 15160
    },
    {
      "entropy": 6.112727785110474,
      "epoch": 1.750144258511252,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004703121338878164,
      "loss": 5.835,
      "mean_token_accuracy": 0.1800225406885147,
      "num_tokens": 38447362.0,
      "step": 15165
    },
    {
      "entropy": 5.9935276985168455,
      "epoch": 1.750721292556261,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004702916440940286,
      "loss": 5.7682,
      "mean_token_accuracy": 0.18413769751787185,
      "num_tokens": 38459796.0,
      "step": 15170
    },
    {
      "entropy": 6.078943109512329,
      "epoch": 1.7512983266012694,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004702711477317034,
      "loss": 5.987,
      "mean_token_accuracy": 0.16556783020496368,
      "num_tokens": 38472536.0,
      "step": 15175
    },
    {
      "entropy": 6.118814134597779,
      "epoch": 1.7518753606462782,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004702506448015301,
      "loss": 5.8731,
      "mean_token_accuracy": 0.17335542887449265,
      "num_tokens": 38485107.0,
      "step": 15180
    },
    {
      "entropy": 6.084583902359009,
      "epoch": 1.7524523946912867,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047023013530419843,
      "loss": 5.9395,
      "mean_token_accuracy": 0.17048297822475433,
      "num_tokens": 38498560.0,
      "step": 15185
    },
    {
      "entropy": 5.939894390106201,
      "epoch": 1.7530294287362955,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004702096192403981,
      "loss": 5.6348,
      "mean_token_accuracy": 0.1919153854250908,
      "num_tokens": 38510553.0,
      "step": 15190
    },
    {
      "entropy": 6.009325838088989,
      "epoch": 1.753606462781304,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004701890966108192,
      "loss": 5.8515,
      "mean_token_accuracy": 0.18088147044181824,
      "num_tokens": 38524191.0,
      "step": 15195
    },
    {
      "entropy": 6.0358314037323,
      "epoch": 1.7541834968263128,
      "grad_norm": 0.953125,
      "learning_rate": 0.000470168567416152,
      "loss": 5.8704,
      "mean_token_accuracy": 0.18294660300016402,
      "num_tokens": 38535654.0,
      "step": 15200
    },
    {
      "entropy": 6.086613750457763,
      "epoch": 1.7547605308713214,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004701480316570869,
      "loss": 5.9253,
      "mean_token_accuracy": 0.17463968694210052,
      "num_tokens": 38548596.0,
      "step": 15205
    },
    {
      "entropy": 5.995666456222534,
      "epoch": 1.75533756491633,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004701274893343147,
      "loss": 5.8312,
      "mean_token_accuracy": 0.17816859483718872,
      "num_tokens": 38559852.0,
      "step": 15210
    },
    {
      "entropy": 5.927803087234497,
      "epoch": 1.7559145989613387,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047010694044852643,
      "loss": 5.8014,
      "mean_token_accuracy": 0.18001902848482132,
      "num_tokens": 38571773.0,
      "step": 15215
    },
    {
      "entropy": 6.108920526504517,
      "epoch": 1.7564916330063474,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000470086385000413,
      "loss": 5.836,
      "mean_token_accuracy": 0.18273732513189317,
      "num_tokens": 38583898.0,
      "step": 15220
    },
    {
      "entropy": 6.079107856750488,
      "epoch": 1.757068667051356,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004700658229906661,
      "loss": 5.8929,
      "mean_token_accuracy": 0.1771309494972229,
      "num_tokens": 38596573.0,
      "step": 15225
    },
    {
      "entropy": 6.112817573547363,
      "epoch": 1.7576457010963646,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047004525441997694,
      "loss": 5.8864,
      "mean_token_accuracy": 0.17313503324985505,
      "num_tokens": 38610083.0,
      "step": 15230
    },
    {
      "entropy": 6.120206737518311,
      "epoch": 1.7582227351413735,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004700246792890376,
      "loss": 5.8809,
      "mean_token_accuracy": 0.17861181795597075,
      "num_tokens": 38622710.0,
      "step": 15235
    },
    {
      "entropy": 6.048461246490478,
      "epoch": 1.758799769186382,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004700040975985401,
      "loss": 5.8452,
      "mean_token_accuracy": 0.18395235687494277,
      "num_tokens": 38634633.0,
      "step": 15240
    },
    {
      "entropy": 6.023446130752563,
      "epoch": 1.7593768032313908,
      "grad_norm": 1.0,
      "learning_rate": 0.00046998350934917654,
      "loss": 5.8693,
      "mean_token_accuracy": 0.18143530040979386,
      "num_tokens": 38646836.0,
      "step": 15245
    },
    {
      "entropy": 6.0506744384765625,
      "epoch": 1.7599538372763992,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046996291454163956,
      "loss": 5.847,
      "mean_token_accuracy": 0.18208872079849242,
      "num_tokens": 38659251.0,
      "step": 15250
    },
    {
      "entropy": 6.099902725219726,
      "epoch": 1.760530871321408,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004699423131766218,
      "loss": 5.8962,
      "mean_token_accuracy": 0.17486644983291627,
      "num_tokens": 38672464.0,
      "step": 15255
    },
    {
      "entropy": 6.064452600479126,
      "epoch": 1.7611079053664165,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004699217052548161,
      "loss": 5.8869,
      "mean_token_accuracy": 0.17572104185819626,
      "num_tokens": 38684989.0,
      "step": 15260
    },
    {
      "entropy": 6.066902732849121,
      "epoch": 1.7616849394114253,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046990109077691577,
      "loss": 5.9375,
      "mean_token_accuracy": 0.17806067317724228,
      "num_tokens": 38698528.0,
      "step": 15265
    },
    {
      "entropy": 6.0771803855896,
      "epoch": 1.762261973456434,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046988046974361406,
      "loss": 5.8186,
      "mean_token_accuracy": 0.17879218608140945,
      "num_tokens": 38710525.0,
      "step": 15270
    },
    {
      "entropy": 6.05092568397522,
      "epoch": 1.7628390075014426,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004698598421556045,
      "loss": 5.8278,
      "mean_token_accuracy": 0.17911438941955565,
      "num_tokens": 38723668.0,
      "step": 15275
    },
    {
      "entropy": 6.099549293518066,
      "epoch": 1.7634160415464513,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004698392080135809,
      "loss": 5.8873,
      "mean_token_accuracy": 0.17584020793437957,
      "num_tokens": 38735136.0,
      "step": 15280
    },
    {
      "entropy": 6.04482250213623,
      "epoch": 1.76399307559146,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004698185673182374,
      "loss": 5.8682,
      "mean_token_accuracy": 0.17925010323524476,
      "num_tokens": 38748118.0,
      "step": 15285
    },
    {
      "entropy": 6.018276119232178,
      "epoch": 1.7645701096364685,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046979792007026817,
      "loss": 5.8148,
      "mean_token_accuracy": 0.1807548612356186,
      "num_tokens": 38759685.0,
      "step": 15290
    },
    {
      "entropy": 6.047301244735718,
      "epoch": 1.7651471436814772,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004697772662703676,
      "loss": 5.7362,
      "mean_token_accuracy": 0.19055562019348143,
      "num_tokens": 38772541.0,
      "step": 15295
    },
    {
      "entropy": 6.116163873672486,
      "epoch": 1.7657241777264858,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046975660591923047,
      "loss": 5.8899,
      "mean_token_accuracy": 0.1762281760573387,
      "num_tokens": 38785319.0,
      "step": 15300
    },
    {
      "entropy": 6.032145023345947,
      "epoch": 1.7663012117714945,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004697359390175516,
      "loss": 5.8851,
      "mean_token_accuracy": 0.1739386186003685,
      "num_tokens": 38797979.0,
      "step": 15305
    },
    {
      "entropy": 5.970040321350098,
      "epoch": 1.7668782458165033,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046971526556602625,
      "loss": 5.7229,
      "mean_token_accuracy": 0.19173863530158997,
      "num_tokens": 38810797.0,
      "step": 15310
    },
    {
      "entropy": 6.033888530731201,
      "epoch": 1.7674552798615117,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004696945855653495,
      "loss": 5.8109,
      "mean_token_accuracy": 0.18025269657373427,
      "num_tokens": 38822409.0,
      "step": 15315
    },
    {
      "entropy": 6.0703253746032715,
      "epoch": 1.7680323139065206,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004696738990162172,
      "loss": 5.9177,
      "mean_token_accuracy": 0.1735719159245491,
      "num_tokens": 38835952.0,
      "step": 15320
    },
    {
      "entropy": 6.086167573928833,
      "epoch": 1.768609347951529,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000469653205919325,
      "loss": 5.8631,
      "mean_token_accuracy": 0.17888627350330352,
      "num_tokens": 38848778.0,
      "step": 15325
    },
    {
      "entropy": 6.03140664100647,
      "epoch": 1.7691863819965379,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046963250627536884,
      "loss": 5.8542,
      "mean_token_accuracy": 0.1804684579372406,
      "num_tokens": 38860885.0,
      "step": 15330
    },
    {
      "entropy": 6.081646347045899,
      "epoch": 1.7697634160415463,
      "grad_norm": 1.0,
      "learning_rate": 0.0004696118000850451,
      "loss": 5.8,
      "mean_token_accuracy": 0.18154779374599456,
      "num_tokens": 38873939.0,
      "step": 15335
    },
    {
      "entropy": 6.1259620666503904,
      "epoch": 1.7703404500865552,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046959108734905003,
      "loss": 5.8927,
      "mean_token_accuracy": 0.17331063151359558,
      "num_tokens": 38887648.0,
      "step": 15340
    },
    {
      "entropy": 6.053666830062866,
      "epoch": 1.7709174841315638,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046957036806808045,
      "loss": 5.9092,
      "mean_token_accuracy": 0.17477380186319352,
      "num_tokens": 38899908.0,
      "step": 15345
    },
    {
      "entropy": 6.157666206359863,
      "epoch": 1.7714945181765724,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004695496422428332,
      "loss": 5.9076,
      "mean_token_accuracy": 0.1748862311244011,
      "num_tokens": 38910918.0,
      "step": 15350
    },
    {
      "entropy": 6.0673305034637455,
      "epoch": 1.772071552221581,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004695289098740054,
      "loss": 5.8816,
      "mean_token_accuracy": 0.17632643282413482,
      "num_tokens": 38923427.0,
      "step": 15355
    },
    {
      "entropy": 5.973150014877319,
      "epoch": 1.7726485862665897,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004695081709622943,
      "loss": 5.8059,
      "mean_token_accuracy": 0.18723804652690887,
      "num_tokens": 38935308.0,
      "step": 15360
    },
    {
      "entropy": 6.1497673988342285,
      "epoch": 1.7732256203115984,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046948742550839744,
      "loss": 5.9407,
      "mean_token_accuracy": 0.17909758239984513,
      "num_tokens": 38948444.0,
      "step": 15365
    },
    {
      "entropy": 6.052219009399414,
      "epoch": 1.773802654356607,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004694666735130127,
      "loss": 5.8433,
      "mean_token_accuracy": 0.17491151243448258,
      "num_tokens": 38959888.0,
      "step": 15370
    },
    {
      "entropy": 5.971717405319214,
      "epoch": 1.7743796884016156,
      "grad_norm": 1.28125,
      "learning_rate": 0.000469445914976838,
      "loss": 5.7335,
      "mean_token_accuracy": 0.1883981317281723,
      "num_tokens": 38971769.0,
      "step": 15375
    },
    {
      "entropy": 6.059285831451416,
      "epoch": 1.7749567224466243,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004694251499005715,
      "loss": 5.8098,
      "mean_token_accuracy": 0.18303107917308808,
      "num_tokens": 38984718.0,
      "step": 15380
    },
    {
      "entropy": 6.02830548286438,
      "epoch": 1.7755337564916331,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004694043782849116,
      "loss": 5.8366,
      "mean_token_accuracy": 0.17478239685297012,
      "num_tokens": 38996670.0,
      "step": 15385
    },
    {
      "entropy": 5.979550743103028,
      "epoch": 1.7761107905366416,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046938360013055715,
      "loss": 5.789,
      "mean_token_accuracy": 0.1873162254691124,
      "num_tokens": 39009270.0,
      "step": 15390
    },
    {
      "entropy": 6.077267742156982,
      "epoch": 1.7766878245816504,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046936281543820673,
      "loss": 5.8489,
      "mean_token_accuracy": 0.17917584031820297,
      "num_tokens": 39022332.0,
      "step": 15395
    },
    {
      "entropy": 6.059001779556274,
      "epoch": 1.7772648586266588,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046934202420855967,
      "loss": 5.7951,
      "mean_token_accuracy": 0.1782673105597496,
      "num_tokens": 39034964.0,
      "step": 15400
    },
    {
      "entropy": 6.021873569488525,
      "epoch": 1.7778418926716677,
      "grad_norm": 2.828125,
      "learning_rate": 0.00046932122644231507,
      "loss": 5.9077,
      "mean_token_accuracy": 0.1787031263113022,
      "num_tokens": 39048705.0,
      "step": 15405
    },
    {
      "entropy": 6.0342669010162355,
      "epoch": 1.7784189267166761,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046930042214017256,
      "loss": 5.8779,
      "mean_token_accuracy": 0.17730460613965987,
      "num_tokens": 39061994.0,
      "step": 15410
    },
    {
      "entropy": 6.068991708755493,
      "epoch": 1.778995960761685,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004692796113028319,
      "loss": 5.8399,
      "mean_token_accuracy": 0.18004249334335326,
      "num_tokens": 39075279.0,
      "step": 15415
    },
    {
      "entropy": 6.106332540512085,
      "epoch": 1.7795729948066936,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000469258793930993,
      "loss": 5.8797,
      "mean_token_accuracy": 0.1724288910627365,
      "num_tokens": 39087914.0,
      "step": 15420
    },
    {
      "entropy": 6.02064094543457,
      "epoch": 1.7801500288517023,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046923797002535605,
      "loss": 5.8358,
      "mean_token_accuracy": 0.18486146330833436,
      "num_tokens": 39100302.0,
      "step": 15425
    },
    {
      "entropy": 6.008547592163086,
      "epoch": 1.780727062896711,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004692171395866214,
      "loss": 5.8932,
      "mean_token_accuracy": 0.1791265442967415,
      "num_tokens": 39114073.0,
      "step": 15430
    },
    {
      "entropy": 6.040784645080566,
      "epoch": 1.7813040969417195,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046919630261548986,
      "loss": 5.8698,
      "mean_token_accuracy": 0.18200439661741258,
      "num_tokens": 39126213.0,
      "step": 15435
    },
    {
      "entropy": 6.032877254486084,
      "epoch": 1.7818811309867282,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046917545911266207,
      "loss": 5.8307,
      "mean_token_accuracy": 0.178719100356102,
      "num_tokens": 39138600.0,
      "step": 15440
    },
    {
      "entropy": 6.052349042892456,
      "epoch": 1.7824581650317368,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046915460907883923,
      "loss": 5.8095,
      "mean_token_accuracy": 0.18662605732679366,
      "num_tokens": 39149356.0,
      "step": 15445
    },
    {
      "entropy": 6.0392174243927,
      "epoch": 1.7830351990767457,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046913375251472246,
      "loss": 5.7992,
      "mean_token_accuracy": 0.17511518597602843,
      "num_tokens": 39162141.0,
      "step": 15450
    },
    {
      "entropy": 5.985444688796997,
      "epoch": 1.783612233121754,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046911288942101345,
      "loss": 5.849,
      "mean_token_accuracy": 0.17972800731658936,
      "num_tokens": 39173714.0,
      "step": 15455
    },
    {
      "entropy": 6.051642036437988,
      "epoch": 1.784189267166763,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004690920197984138,
      "loss": 5.7585,
      "mean_token_accuracy": 0.18328466266393661,
      "num_tokens": 39186498.0,
      "step": 15460
    },
    {
      "entropy": 6.008974599838257,
      "epoch": 1.7847663012117714,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004690711436476254,
      "loss": 5.8413,
      "mean_token_accuracy": 0.18659435510635375,
      "num_tokens": 39199762.0,
      "step": 15465
    },
    {
      "entropy": 5.932628154754639,
      "epoch": 1.7853433352567802,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046905026096935056,
      "loss": 5.7905,
      "mean_token_accuracy": 0.18133017867803575,
      "num_tokens": 39213236.0,
      "step": 15470
    },
    {
      "entropy": 6.0883159160614015,
      "epoch": 1.7859203693017887,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004690293717642916,
      "loss": 5.8782,
      "mean_token_accuracy": 0.17507740259170532,
      "num_tokens": 39225538.0,
      "step": 15475
    },
    {
      "entropy": 6.1186144828796385,
      "epoch": 1.7864974033467975,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000469008476033151,
      "loss": 5.8749,
      "mean_token_accuracy": 0.17825197130441667,
      "num_tokens": 39237889.0,
      "step": 15480
    },
    {
      "entropy": 6.029528713226318,
      "epoch": 1.7870744373918062,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046898757377663176,
      "loss": 5.8822,
      "mean_token_accuracy": 0.17928979247808458,
      "num_tokens": 39250686.0,
      "step": 15485
    },
    {
      "entropy": 6.005362606048584,
      "epoch": 1.7876514714368148,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046896666499543683,
      "loss": 5.793,
      "mean_token_accuracy": 0.1799032986164093,
      "num_tokens": 39262708.0,
      "step": 15490
    },
    {
      "entropy": 5.9179400444030765,
      "epoch": 1.7882285054818234,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00046894574969026946,
      "loss": 5.6617,
      "mean_token_accuracy": 0.19557317346334457,
      "num_tokens": 39274778.0,
      "step": 15495
    },
    {
      "entropy": 6.039992904663086,
      "epoch": 1.788805539526832,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046892482786183313,
      "loss": 5.8962,
      "mean_token_accuracy": 0.17480368316173553,
      "num_tokens": 39288088.0,
      "step": 15500
    },
    {
      "entropy": 5.999263858795166,
      "epoch": 1.7893825735718407,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046890389951083155,
      "loss": 5.7799,
      "mean_token_accuracy": 0.19062529951334,
      "num_tokens": 39300828.0,
      "step": 15505
    },
    {
      "entropy": 6.03234076499939,
      "epoch": 1.7899596076168494,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046888296463796857,
      "loss": 5.7858,
      "mean_token_accuracy": 0.18807779848575593,
      "num_tokens": 39313839.0,
      "step": 15510
    },
    {
      "entropy": 5.9806556224823,
      "epoch": 1.790536641661858,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004688620232439485,
      "loss": 5.8717,
      "mean_token_accuracy": 0.1791326195001602,
      "num_tokens": 39326115.0,
      "step": 15515
    },
    {
      "entropy": 6.152474594116211,
      "epoch": 1.7911136757068666,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046884107532947547,
      "loss": 5.8599,
      "mean_token_accuracy": 0.1788884460926056,
      "num_tokens": 39338756.0,
      "step": 15520
    },
    {
      "entropy": 6.1109377384185795,
      "epoch": 1.7916907097518755,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046882012089525415,
      "loss": 5.8948,
      "mean_token_accuracy": 0.16849719882011413,
      "num_tokens": 39352406.0,
      "step": 15525
    },
    {
      "entropy": 6.026819038391113,
      "epoch": 1.792267743796884,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004687991599419895,
      "loss": 5.8676,
      "mean_token_accuracy": 0.17431113123893738,
      "num_tokens": 39366532.0,
      "step": 15530
    },
    {
      "entropy": 6.075605154037476,
      "epoch": 1.7928447778418928,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046877819247038624,
      "loss": 5.875,
      "mean_token_accuracy": 0.1821760728955269,
      "num_tokens": 39378730.0,
      "step": 15535
    },
    {
      "entropy": 6.106965637207031,
      "epoch": 1.7934218118869012,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004687572184811497,
      "loss": 5.8833,
      "mean_token_accuracy": 0.17908782958984376,
      "num_tokens": 39391448.0,
      "step": 15540
    },
    {
      "entropy": 6.046957015991211,
      "epoch": 1.79399884593191,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046873623797498545,
      "loss": 5.8272,
      "mean_token_accuracy": 0.1828412339091301,
      "num_tokens": 39405120.0,
      "step": 15545
    },
    {
      "entropy": 6.089573907852173,
      "epoch": 1.7945758799769185,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000468715250952599,
      "loss": 5.8793,
      "mean_token_accuracy": 0.17551148682832718,
      "num_tokens": 39417508.0,
      "step": 15550
    },
    {
      "entropy": 6.079168319702148,
      "epoch": 1.7951529140219273,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046869425741469635,
      "loss": 5.8617,
      "mean_token_accuracy": 0.17495205402374267,
      "num_tokens": 39430185.0,
      "step": 15555
    },
    {
      "entropy": 6.086002206802368,
      "epoch": 1.795729948066936,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004686732573619835,
      "loss": 5.9942,
      "mean_token_accuracy": 0.17707456946372985,
      "num_tokens": 39442855.0,
      "step": 15560
    },
    {
      "entropy": 6.041767644882202,
      "epoch": 1.7963069821119446,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004686522507951669,
      "loss": 5.7869,
      "mean_token_accuracy": 0.18758971691131593,
      "num_tokens": 39454855.0,
      "step": 15565
    },
    {
      "entropy": 6.043534612655639,
      "epoch": 1.7968840161569533,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004686312377149531,
      "loss": 5.8343,
      "mean_token_accuracy": 0.17684232741594313,
      "num_tokens": 39466905.0,
      "step": 15570
    },
    {
      "entropy": 6.068918895721436,
      "epoch": 1.797461050201962,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046861021812204874,
      "loss": 5.8143,
      "mean_token_accuracy": 0.18126294612884522,
      "num_tokens": 39479907.0,
      "step": 15575
    },
    {
      "entropy": 6.011303186416626,
      "epoch": 1.7980380842469705,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046858919201716085,
      "loss": 5.7618,
      "mean_token_accuracy": 0.17967075407505034,
      "num_tokens": 39491934.0,
      "step": 15580
    },
    {
      "entropy": 6.0332972526550295,
      "epoch": 1.7986151182919792,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004685681594009966,
      "loss": 5.8617,
      "mean_token_accuracy": 0.18586660772562028,
      "num_tokens": 39503669.0,
      "step": 15585
    },
    {
      "entropy": 6.074440383911133,
      "epoch": 1.799192152336988,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046854712027426357,
      "loss": 5.8884,
      "mean_token_accuracy": 0.17828327864408494,
      "num_tokens": 39515555.0,
      "step": 15590
    },
    {
      "entropy": 6.004092502593994,
      "epoch": 1.7997691863819965,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046852607463766923,
      "loss": 5.8172,
      "mean_token_accuracy": 0.18024124056100846,
      "num_tokens": 39528867.0,
      "step": 15595
    },
    {
      "entropy": 6.0933550834655765,
      "epoch": 1.8003462204270053,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004685050224919215,
      "loss": 5.8569,
      "mean_token_accuracy": 0.1799604371190071,
      "num_tokens": 39541768.0,
      "step": 15600
    },
    {
      "entropy": 6.105311012268066,
      "epoch": 1.8009232544720137,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046848396383772844,
      "loss": 5.837,
      "mean_token_accuracy": 0.18166967779397963,
      "num_tokens": 39553578.0,
      "step": 15605
    },
    {
      "entropy": 6.071601390838623,
      "epoch": 1.8015002885170226,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004684628986757984,
      "loss": 5.8352,
      "mean_token_accuracy": 0.18051561564207078,
      "num_tokens": 39565846.0,
      "step": 15610
    },
    {
      "entropy": 6.131592416763306,
      "epoch": 1.802077322562031,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004684418270068398,
      "loss": 5.8815,
      "mean_token_accuracy": 0.17616225332021712,
      "num_tokens": 39578536.0,
      "step": 15615
    },
    {
      "entropy": 5.962609004974365,
      "epoch": 1.8026543566070399,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004684207488315615,
      "loss": 5.8234,
      "mean_token_accuracy": 0.18232211321592331,
      "num_tokens": 39591233.0,
      "step": 15620
    },
    {
      "entropy": 6.0452563762664795,
      "epoch": 1.8032313906520485,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004683996641506724,
      "loss": 5.8093,
      "mean_token_accuracy": 0.17528978288173674,
      "num_tokens": 39602335.0,
      "step": 15625
    },
    {
      "entropy": 5.97154803276062,
      "epoch": 1.8038084246970572,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046837857296488163,
      "loss": 5.7345,
      "mean_token_accuracy": 0.18521395921707154,
      "num_tokens": 39614278.0,
      "step": 15630
    },
    {
      "entropy": 5.968838834762574,
      "epoch": 1.8043854587420658,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046835747527489857,
      "loss": 5.8267,
      "mean_token_accuracy": 0.1789167195558548,
      "num_tokens": 39627218.0,
      "step": 15635
    },
    {
      "entropy": 5.990184211730957,
      "epoch": 1.8049624927870744,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000468336371081433,
      "loss": 5.8552,
      "mean_token_accuracy": 0.18183627873659133,
      "num_tokens": 39639607.0,
      "step": 15640
    },
    {
      "entropy": 6.047556304931641,
      "epoch": 1.805539526832083,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046831526038519444,
      "loss": 5.8487,
      "mean_token_accuracy": 0.17250668108463288,
      "num_tokens": 39652748.0,
      "step": 15645
    },
    {
      "entropy": 6.09375376701355,
      "epoch": 1.8061165608770917,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004682941431868933,
      "loss": 5.8172,
      "mean_token_accuracy": 0.1831415444612503,
      "num_tokens": 39665984.0,
      "step": 15650
    },
    {
      "entropy": 6.025468540191651,
      "epoch": 1.8066935949221004,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046827301948723963,
      "loss": 5.8143,
      "mean_token_accuracy": 0.18363720178604126,
      "num_tokens": 39680017.0,
      "step": 15655
    },
    {
      "entropy": 6.0484872341156,
      "epoch": 1.807270628967109,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046825188928694393,
      "loss": 5.8353,
      "mean_token_accuracy": 0.1790148064494133,
      "num_tokens": 39694070.0,
      "step": 15660
    },
    {
      "entropy": 6.0738471984863285,
      "epoch": 1.8078476630121179,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004682307525867169,
      "loss": 5.8374,
      "mean_token_accuracy": 0.17122248858213424,
      "num_tokens": 39706341.0,
      "step": 15665
    },
    {
      "entropy": 6.055118560791016,
      "epoch": 1.8084246970571263,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004682096093872695,
      "loss": 5.9295,
      "mean_token_accuracy": 0.17942884862422942,
      "num_tokens": 39719552.0,
      "step": 15670
    },
    {
      "entropy": 6.150817012786865,
      "epoch": 1.8090017311021351,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046818845968931284,
      "loss": 5.9384,
      "mean_token_accuracy": 0.1695065289735794,
      "num_tokens": 39732424.0,
      "step": 15675
    },
    {
      "entropy": 6.050227069854737,
      "epoch": 1.8095787651471436,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046816730349355843,
      "loss": 5.8374,
      "mean_token_accuracy": 0.18148971647024154,
      "num_tokens": 39745088.0,
      "step": 15680
    },
    {
      "entropy": 6.0705187797546385,
      "epoch": 1.8101557991921524,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046814614080071756,
      "loss": 5.8374,
      "mean_token_accuracy": 0.18692173957824706,
      "num_tokens": 39756912.0,
      "step": 15685
    },
    {
      "entropy": 6.012263679504395,
      "epoch": 1.8107328332371608,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046812497161150224,
      "loss": 5.7899,
      "mean_token_accuracy": 0.1817648872733116,
      "num_tokens": 39768224.0,
      "step": 15690
    },
    {
      "entropy": 5.990527868270874,
      "epoch": 1.8113098672821697,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046810379592662445,
      "loss": 5.8051,
      "mean_token_accuracy": 0.1769183963537216,
      "num_tokens": 39780486.0,
      "step": 15695
    },
    {
      "entropy": 6.122638273239136,
      "epoch": 1.8118869013271783,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046808261374679637,
      "loss": 5.8348,
      "mean_token_accuracy": 0.1783178463578224,
      "num_tokens": 39793572.0,
      "step": 15700
    },
    {
      "entropy": 6.0269934177398685,
      "epoch": 1.812463935372187,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004680614250727305,
      "loss": 5.8119,
      "mean_token_accuracy": 0.1820087641477585,
      "num_tokens": 39806064.0,
      "step": 15705
    },
    {
      "entropy": 5.985729360580445,
      "epoch": 1.8130409694171956,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004680402299051395,
      "loss": 5.8225,
      "mean_token_accuracy": 0.1798326000571251,
      "num_tokens": 39818638.0,
      "step": 15710
    },
    {
      "entropy": 6.077349424362183,
      "epoch": 1.8136180034622043,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004680190282447363,
      "loss": 5.8953,
      "mean_token_accuracy": 0.17430179119110106,
      "num_tokens": 39831788.0,
      "step": 15715
    },
    {
      "entropy": 6.027905464172363,
      "epoch": 1.814195037507213,
      "grad_norm": 0.875,
      "learning_rate": 0.0004679978200922339,
      "loss": 5.8097,
      "mean_token_accuracy": 0.17993560433387756,
      "num_tokens": 39844043.0,
      "step": 15720
    },
    {
      "entropy": 6.039300584793091,
      "epoch": 1.8147720715522215,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004679766054483457,
      "loss": 5.8897,
      "mean_token_accuracy": 0.1796398714184761,
      "num_tokens": 39857351.0,
      "step": 15725
    },
    {
      "entropy": 6.021570110321045,
      "epoch": 1.8153491055972304,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004679553843137852,
      "loss": 5.8013,
      "mean_token_accuracy": 0.18483318090438844,
      "num_tokens": 39868975.0,
      "step": 15730
    },
    {
      "entropy": 5.973925018310547,
      "epoch": 1.8159261396422388,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046793415668926625,
      "loss": 5.7701,
      "mean_token_accuracy": 0.18482405990362166,
      "num_tokens": 39880995.0,
      "step": 15735
    },
    {
      "entropy": 6.039802837371826,
      "epoch": 1.8165031736872477,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004679129225755028,
      "loss": 5.7701,
      "mean_token_accuracy": 0.18159797936677932,
      "num_tokens": 39893821.0,
      "step": 15740
    },
    {
      "entropy": 6.044361305236817,
      "epoch": 1.817080207732256,
      "grad_norm": 0.9375,
      "learning_rate": 0.000467891681973209,
      "loss": 5.8439,
      "mean_token_accuracy": 0.18538560420274736,
      "num_tokens": 39906938.0,
      "step": 15745
    },
    {
      "entropy": 6.0200530052185055,
      "epoch": 1.817657241777265,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046787043488309926,
      "loss": 5.7998,
      "mean_token_accuracy": 0.1817298486828804,
      "num_tokens": 39920647.0,
      "step": 15750
    },
    {
      "entropy": 6.083376789093018,
      "epoch": 1.8182342758222734,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004678491813058882,
      "loss": 5.8077,
      "mean_token_accuracy": 0.1737175241112709,
      "num_tokens": 39932925.0,
      "step": 15755
    },
    {
      "entropy": 6.06862325668335,
      "epoch": 1.8188113098672822,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004678279212422908,
      "loss": 5.8677,
      "mean_token_accuracy": 0.17698773592710496,
      "num_tokens": 39945950.0,
      "step": 15760
    },
    {
      "entropy": 6.0465010643005375,
      "epoch": 1.8193883439122909,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004678066546930221,
      "loss": 5.8566,
      "mean_token_accuracy": 0.18024222254753114,
      "num_tokens": 39957231.0,
      "step": 15765
    },
    {
      "entropy": 6.045978212356568,
      "epoch": 1.8199653779572995,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046778538165879725,
      "loss": 5.8041,
      "mean_token_accuracy": 0.1852501705288887,
      "num_tokens": 39969390.0,
      "step": 15770
    },
    {
      "entropy": 6.027174806594848,
      "epoch": 1.8205424120023082,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046776410214033185,
      "loss": 5.8924,
      "mean_token_accuracy": 0.17558384388685228,
      "num_tokens": 39982893.0,
      "step": 15775
    },
    {
      "entropy": 6.098627281188965,
      "epoch": 1.8211194460473168,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004677428161383417,
      "loss": 5.9181,
      "mean_token_accuracy": 0.17253419756889343,
      "num_tokens": 39996413.0,
      "step": 15780
    },
    {
      "entropy": 6.092543745040894,
      "epoch": 1.8216964800923254,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004677215236535426,
      "loss": 5.8845,
      "mean_token_accuracy": 0.17856497764587403,
      "num_tokens": 40008662.0,
      "step": 15785
    },
    {
      "entropy": 6.053987503051758,
      "epoch": 1.822273514137334,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004677002246866508,
      "loss": 5.8417,
      "mean_token_accuracy": 0.18015456795692444,
      "num_tokens": 40021511.0,
      "step": 15790
    },
    {
      "entropy": 6.044880676269531,
      "epoch": 1.8228505481823427,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046767891923838264,
      "loss": 5.8211,
      "mean_token_accuracy": 0.17729813158512114,
      "num_tokens": 40034205.0,
      "step": 15795
    },
    {
      "entropy": 6.07872314453125,
      "epoch": 1.8234275822273514,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004676576073094548,
      "loss": 5.873,
      "mean_token_accuracy": 0.17684656530618667,
      "num_tokens": 40046507.0,
      "step": 15800
    },
    {
      "entropy": 6.054924774169922,
      "epoch": 1.8240046162723602,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046763628890058396,
      "loss": 5.7832,
      "mean_token_accuracy": 0.17904330044984818,
      "num_tokens": 40058561.0,
      "step": 15805
    },
    {
      "entropy": 6.06187686920166,
      "epoch": 1.8245816503173686,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046761496401248734,
      "loss": 5.8846,
      "mean_token_accuracy": 0.18583040535449982,
      "num_tokens": 40070351.0,
      "step": 15810
    },
    {
      "entropy": 5.938790321350098,
      "epoch": 1.8251586843623775,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000467593632645882,
      "loss": 5.719,
      "mean_token_accuracy": 0.1951930806040764,
      "num_tokens": 40084057.0,
      "step": 15815
    },
    {
      "entropy": 6.137287950515747,
      "epoch": 1.825735718407386,
      "grad_norm": 1.0,
      "learning_rate": 0.0004675722948014855,
      "loss": 5.954,
      "mean_token_accuracy": 0.170821350812912,
      "num_tokens": 40097584.0,
      "step": 15820
    },
    {
      "entropy": 5.95341649055481,
      "epoch": 1.8263127524523948,
      "grad_norm": 1.0,
      "learning_rate": 0.00046755095048001556,
      "loss": 5.8239,
      "mean_token_accuracy": 0.18519600480794907,
      "num_tokens": 40110303.0,
      "step": 15825
    },
    {
      "entropy": 6.084471893310547,
      "epoch": 1.8268897864974032,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004675295996821899,
      "loss": 5.9233,
      "mean_token_accuracy": 0.17486847341060638,
      "num_tokens": 40122594.0,
      "step": 15830
    },
    {
      "entropy": 6.096218776702881,
      "epoch": 1.827466820542412,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000467508242408727,
      "loss": 5.7879,
      "mean_token_accuracy": 0.18510335832834243,
      "num_tokens": 40133737.0,
      "step": 15835
    },
    {
      "entropy": 6.1139054775238035,
      "epoch": 1.8280438545874207,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004674868786603448,
      "loss": 5.8618,
      "mean_token_accuracy": 0.17935024350881576,
      "num_tokens": 40144891.0,
      "step": 15840
    },
    {
      "entropy": 6.067246055603027,
      "epoch": 1.8286208886324293,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004674655084377622,
      "loss": 5.9651,
      "mean_token_accuracy": 0.17418448477983475,
      "num_tokens": 40158217.0,
      "step": 15845
    },
    {
      "entropy": 6.096169710159302,
      "epoch": 1.829197922677438,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004674441317416978,
      "loss": 5.8183,
      "mean_token_accuracy": 0.18234384506940843,
      "num_tokens": 40170733.0,
      "step": 15850
    },
    {
      "entropy": 6.030439805984497,
      "epoch": 1.8297749567224466,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00046742274857287057,
      "loss": 5.8666,
      "mean_token_accuracy": 0.18111820966005326,
      "num_tokens": 40183204.0,
      "step": 15855
    },
    {
      "entropy": 6.063157653808593,
      "epoch": 1.8303519907674553,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004674013589319998,
      "loss": 5.9167,
      "mean_token_accuracy": 0.17580896466970444,
      "num_tokens": 40194816.0,
      "step": 15860
    },
    {
      "entropy": 6.0145985126495365,
      "epoch": 1.830929024812464,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004673799628198049,
      "loss": 5.8018,
      "mean_token_accuracy": 0.1839042529463768,
      "num_tokens": 40205796.0,
      "step": 15865
    },
    {
      "entropy": 6.02322940826416,
      "epoch": 1.8315060588574728,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046735856023700546,
      "loss": 5.7981,
      "mean_token_accuracy": 0.17602976113557817,
      "num_tokens": 40218232.0,
      "step": 15870
    },
    {
      "entropy": 6.111505651473999,
      "epoch": 1.8320830929024812,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004673371511843215,
      "loss": 5.8791,
      "mean_token_accuracy": 0.18025649785995485,
      "num_tokens": 40230441.0,
      "step": 15875
    },
    {
      "entropy": 6.082293605804443,
      "epoch": 1.83266012694749,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004673157356624729,
      "loss": 5.8846,
      "mean_token_accuracy": 0.17131757885217666,
      "num_tokens": 40241858.0,
      "step": 15880
    },
    {
      "entropy": 6.0725068092346195,
      "epoch": 1.8332371609924984,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004672943136721801,
      "loss": 5.7978,
      "mean_token_accuracy": 0.1779753178358078,
      "num_tokens": 40254310.0,
      "step": 15885
    },
    {
      "entropy": 5.99890923500061,
      "epoch": 1.8338141950375073,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004672728852141636,
      "loss": 5.8578,
      "mean_token_accuracy": 0.17791000306606292,
      "num_tokens": 40267851.0,
      "step": 15890
    },
    {
      "entropy": 6.053861427307129,
      "epoch": 1.8343912290825157,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046725145028914404,
      "loss": 5.8601,
      "mean_token_accuracy": 0.1875164270401001,
      "num_tokens": 40279899.0,
      "step": 15895
    },
    {
      "entropy": 6.060376262664795,
      "epoch": 1.8349682631275246,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004672300088978425,
      "loss": 5.838,
      "mean_token_accuracy": 0.17466944456100464,
      "num_tokens": 40291996.0,
      "step": 15900
    },
    {
      "entropy": 6.053700399398804,
      "epoch": 1.8355452971725332,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004672085610409801,
      "loss": 5.7504,
      "mean_token_accuracy": 0.1875486433506012,
      "num_tokens": 40304426.0,
      "step": 15905
    },
    {
      "entropy": 5.9922538757324215,
      "epoch": 1.8361223312175419,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046718710671927815,
      "loss": 5.8481,
      "mean_token_accuracy": 0.17591640204191208,
      "num_tokens": 40318287.0,
      "step": 15910
    },
    {
      "entropy": 6.03928279876709,
      "epoch": 1.8366993652625505,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046716564593345843,
      "loss": 5.7841,
      "mean_token_accuracy": 0.18345965892076493,
      "num_tokens": 40330906.0,
      "step": 15915
    },
    {
      "entropy": 5.974145174026489,
      "epoch": 1.8372763993075591,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046714417868424265,
      "loss": 5.8069,
      "mean_token_accuracy": 0.18108908236026763,
      "num_tokens": 40342310.0,
      "step": 15920
    },
    {
      "entropy": 6.0805967330932615,
      "epoch": 1.8378534333525678,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046712270497235284,
      "loss": 5.8853,
      "mean_token_accuracy": 0.17639820724725724,
      "num_tokens": 40355596.0,
      "step": 15925
    },
    {
      "entropy": 6.109693956375122,
      "epoch": 1.8384304673975764,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004671012247985112,
      "loss": 5.8689,
      "mean_token_accuracy": 0.1779905751347542,
      "num_tokens": 40368841.0,
      "step": 15930
    },
    {
      "entropy": 6.043868494033814,
      "epoch": 1.839007501442585,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046707973816344044,
      "loss": 5.816,
      "mean_token_accuracy": 0.17921997159719466,
      "num_tokens": 40381496.0,
      "step": 15935
    },
    {
      "entropy": 6.127592468261719,
      "epoch": 1.8395845354875937,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046705824506786304,
      "loss": 5.879,
      "mean_token_accuracy": 0.17880403101444245,
      "num_tokens": 40393750.0,
      "step": 15940
    },
    {
      "entropy": 5.917369556427002,
      "epoch": 1.8401615695326026,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046703674551250193,
      "loss": 5.6807,
      "mean_token_accuracy": 0.19033827930688857,
      "num_tokens": 40406262.0,
      "step": 15945
    },
    {
      "entropy": 5.903103017807007,
      "epoch": 1.840738603577611,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004670152394980803,
      "loss": 5.6821,
      "mean_token_accuracy": 0.1855830132961273,
      "num_tokens": 40418776.0,
      "step": 15950
    },
    {
      "entropy": 6.069427585601806,
      "epoch": 1.8413156376226198,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004669937270253214,
      "loss": 5.9045,
      "mean_token_accuracy": 0.17174598276615144,
      "num_tokens": 40432300.0,
      "step": 15955
    },
    {
      "entropy": 6.075692510604858,
      "epoch": 1.8418926716676283,
      "grad_norm": 0.921875,
      "learning_rate": 0.000466972208094949,
      "loss": 5.9188,
      "mean_token_accuracy": 0.17251038402318955,
      "num_tokens": 40446215.0,
      "step": 15960
    },
    {
      "entropy": 6.104417562484741,
      "epoch": 1.8424697057126371,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046695068270768665,
      "loss": 5.9068,
      "mean_token_accuracy": 0.17366523891687394,
      "num_tokens": 40457932.0,
      "step": 15965
    },
    {
      "entropy": 6.041384363174439,
      "epoch": 1.8430467397576455,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046692915086425846,
      "loss": 5.7981,
      "mean_token_accuracy": 0.1833280235528946,
      "num_tokens": 40469897.0,
      "step": 15970
    },
    {
      "entropy": 6.063959217071533,
      "epoch": 1.8436237738026544,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046690761256538857,
      "loss": 5.8518,
      "mean_token_accuracy": 0.17779323756694793,
      "num_tokens": 40483104.0,
      "step": 15975
    },
    {
      "entropy": 5.9873377799987795,
      "epoch": 1.844200807847663,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004668860678118015,
      "loss": 5.76,
      "mean_token_accuracy": 0.1885027378797531,
      "num_tokens": 40494844.0,
      "step": 15980
    },
    {
      "entropy": 6.071859550476074,
      "epoch": 1.8447778418926717,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046686451660422185,
      "loss": 5.8816,
      "mean_token_accuracy": 0.17762214839458465,
      "num_tokens": 40506972.0,
      "step": 15985
    },
    {
      "entropy": 6.028794813156128,
      "epoch": 1.8453548759376803,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00046684295894337455,
      "loss": 5.8392,
      "mean_token_accuracy": 0.183400397002697,
      "num_tokens": 40519575.0,
      "step": 15990
    },
    {
      "entropy": 6.043858337402344,
      "epoch": 1.845931909982689,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004668213948299845,
      "loss": 5.7662,
      "mean_token_accuracy": 0.18693499118089676,
      "num_tokens": 40532895.0,
      "step": 15995
    },
    {
      "entropy": 6.059022998809814,
      "epoch": 1.8465089440276976,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004667998242647772,
      "loss": 5.7994,
      "mean_token_accuracy": 0.1798090249300003,
      "num_tokens": 40544315.0,
      "step": 16000
    },
    {
      "entropy": 6.047746467590332,
      "epoch": 1.8470859780727062,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000466778247248478,
      "loss": 5.869,
      "mean_token_accuracy": 0.1767059877514839,
      "num_tokens": 40557942.0,
      "step": 16005
    },
    {
      "entropy": 6.108817434310913,
      "epoch": 1.847663012117715,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046675666378181275,
      "loss": 5.9317,
      "mean_token_accuracy": 0.16908392310142517,
      "num_tokens": 40570364.0,
      "step": 16010
    },
    {
      "entropy": 6.093663120269776,
      "epoch": 1.8482400461627235,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004667350738655074,
      "loss": 5.9279,
      "mean_token_accuracy": 0.17202756851911544,
      "num_tokens": 40583259.0,
      "step": 16015
    },
    {
      "entropy": 6.0244832038879395,
      "epoch": 1.8488170802077324,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046671347750028806,
      "loss": 5.7993,
      "mean_token_accuracy": 0.18064562380313873,
      "num_tokens": 40595810.0,
      "step": 16020
    },
    {
      "entropy": 6.057338428497315,
      "epoch": 1.8493941142527408,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004666918746868811,
      "loss": 5.8589,
      "mean_token_accuracy": 0.1784820646047592,
      "num_tokens": 40606856.0,
      "step": 16025
    },
    {
      "entropy": 6.09908971786499,
      "epoch": 1.8499711482977497,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004666702654260133,
      "loss": 5.8755,
      "mean_token_accuracy": 0.17277028709650039,
      "num_tokens": 40618764.0,
      "step": 16030
    },
    {
      "entropy": 6.141882133483887,
      "epoch": 1.850548182342758,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046664864971841113,
      "loss": 5.8581,
      "mean_token_accuracy": 0.18288937509059905,
      "num_tokens": 40631833.0,
      "step": 16035
    },
    {
      "entropy": 6.138440275192261,
      "epoch": 1.851125216387767,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046662702756480195,
      "loss": 5.883,
      "mean_token_accuracy": 0.17805315405130387,
      "num_tokens": 40644215.0,
      "step": 16040
    },
    {
      "entropy": 6.0266601085662845,
      "epoch": 1.8517022504327756,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046660539896591286,
      "loss": 5.7813,
      "mean_token_accuracy": 0.18519651442766188,
      "num_tokens": 40656473.0,
      "step": 16045
    },
    {
      "entropy": 6.102505779266357,
      "epoch": 1.8522792844777842,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004665837639224713,
      "loss": 5.8078,
      "mean_token_accuracy": 0.18442520946264268,
      "num_tokens": 40668949.0,
      "step": 16050
    },
    {
      "entropy": 6.121508598327637,
      "epoch": 1.8528563185227929,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046656212243520505,
      "loss": 5.8569,
      "mean_token_accuracy": 0.17951979041099547,
      "num_tokens": 40681550.0,
      "step": 16055
    },
    {
      "entropy": 5.923879766464234,
      "epoch": 1.8534333525678015,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046654047450484193,
      "loss": 5.7451,
      "mean_token_accuracy": 0.18278682976961136,
      "num_tokens": 40695197.0,
      "step": 16060
    },
    {
      "entropy": 6.075685024261475,
      "epoch": 1.8540103866128101,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004665188201321102,
      "loss": 5.8039,
      "mean_token_accuracy": 0.1821405827999115,
      "num_tokens": 40707091.0,
      "step": 16065
    },
    {
      "entropy": 6.0385089874267575,
      "epoch": 1.8545874206578188,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046649715931773795,
      "loss": 5.7097,
      "mean_token_accuracy": 0.1867336019873619,
      "num_tokens": 40719929.0,
      "step": 16070
    },
    {
      "entropy": 5.96746129989624,
      "epoch": 1.8551644547028274,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000466475492062454,
      "loss": 5.8149,
      "mean_token_accuracy": 0.18618215769529342,
      "num_tokens": 40732897.0,
      "step": 16075
    },
    {
      "entropy": 5.974911022186279,
      "epoch": 1.855741488747836,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046645381836698684,
      "loss": 5.759,
      "mean_token_accuracy": 0.18704587668180467,
      "num_tokens": 40746427.0,
      "step": 16080
    },
    {
      "entropy": 6.043270444869995,
      "epoch": 1.856318522792845,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004664321382320657,
      "loss": 5.826,
      "mean_token_accuracy": 0.18285732269287108,
      "num_tokens": 40758538.0,
      "step": 16085
    },
    {
      "entropy": 6.04480357170105,
      "epoch": 1.8568955568378533,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046641045165841965,
      "loss": 5.9203,
      "mean_token_accuracy": 0.17706322222948073,
      "num_tokens": 40772008.0,
      "step": 16090
    },
    {
      "entropy": 6.09946346282959,
      "epoch": 1.8574725908828622,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046638875864677814,
      "loss": 5.8744,
      "mean_token_accuracy": 0.16923159509897232,
      "num_tokens": 40785654.0,
      "step": 16095
    },
    {
      "entropy": 6.180622577667236,
      "epoch": 1.8580496249278706,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004663670591978708,
      "loss": 5.9826,
      "mean_token_accuracy": 0.16907162219285965,
      "num_tokens": 40799394.0,
      "step": 16100
    },
    {
      "entropy": 6.1322098731994625,
      "epoch": 1.8586266589728795,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004663453533124275,
      "loss": 5.9321,
      "mean_token_accuracy": 0.17194543033838272,
      "num_tokens": 40812054.0,
      "step": 16105
    },
    {
      "entropy": 6.073370122909546,
      "epoch": 1.859203693017888,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004663236409911783,
      "loss": 5.9565,
      "mean_token_accuracy": 0.1739451602101326,
      "num_tokens": 40825138.0,
      "step": 16110
    },
    {
      "entropy": 6.146795272827148,
      "epoch": 1.8597807270628968,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046630192223485345,
      "loss": 5.8833,
      "mean_token_accuracy": 0.17568397521972656,
      "num_tokens": 40837766.0,
      "step": 16115
    },
    {
      "entropy": 6.037355041503906,
      "epoch": 1.8603577611079054,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046628019704418345,
      "loss": 5.8394,
      "mean_token_accuracy": 0.1798497825860977,
      "num_tokens": 40851042.0,
      "step": 16120
    },
    {
      "entropy": 6.0473363399505615,
      "epoch": 1.860934795152914,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000466258465419899,
      "loss": 5.7634,
      "mean_token_accuracy": 0.1885899558663368,
      "num_tokens": 40864074.0,
      "step": 16125
    },
    {
      "entropy": 6.107569980621338,
      "epoch": 1.8615118291979227,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004662367273627312,
      "loss": 5.8123,
      "mean_token_accuracy": 0.18879611045122147,
      "num_tokens": 40876404.0,
      "step": 16130
    },
    {
      "entropy": 6.032333135604858,
      "epoch": 1.8620888632429313,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000466214982873411,
      "loss": 5.8111,
      "mean_token_accuracy": 0.1827455446124077,
      "num_tokens": 40889684.0,
      "step": 16135
    },
    {
      "entropy": 5.976114225387573,
      "epoch": 1.86266589728794,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046619323195266975,
      "loss": 5.7629,
      "mean_token_accuracy": 0.18625782132148744,
      "num_tokens": 40902223.0,
      "step": 16140
    },
    {
      "entropy": 6.04302659034729,
      "epoch": 1.8632429313329486,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004661714746012392,
      "loss": 5.9025,
      "mean_token_accuracy": 0.17759815752506256,
      "num_tokens": 40916251.0,
      "step": 16145
    },
    {
      "entropy": 6.055775451660156,
      "epoch": 1.8638199653779572,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000466149710819851,
      "loss": 5.8935,
      "mean_token_accuracy": 0.17854576855897902,
      "num_tokens": 40929248.0,
      "step": 16150
    },
    {
      "entropy": 6.099144220352173,
      "epoch": 1.8643969994229659,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004661279406092373,
      "loss": 5.7998,
      "mean_token_accuracy": 0.182799631357193,
      "num_tokens": 40943575.0,
      "step": 16155
    },
    {
      "entropy": 6.128057241439819,
      "epoch": 1.8649740334679747,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046610616397013027,
      "loss": 5.8503,
      "mean_token_accuracy": 0.17482273429632186,
      "num_tokens": 40956256.0,
      "step": 16160
    },
    {
      "entropy": 6.068996381759644,
      "epoch": 1.8655510675129832,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004660843809032623,
      "loss": 5.9018,
      "mean_token_accuracy": 0.17754170447587966,
      "num_tokens": 40968786.0,
      "step": 16165
    },
    {
      "entropy": 6.061064100265503,
      "epoch": 1.866128101557992,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004660625914093661,
      "loss": 5.9127,
      "mean_token_accuracy": 0.17889419049024582,
      "num_tokens": 40982221.0,
      "step": 16170
    },
    {
      "entropy": 6.052145910263062,
      "epoch": 1.8667051356030004,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004660407954891745,
      "loss": 5.8793,
      "mean_token_accuracy": 0.17607972472906114,
      "num_tokens": 40995226.0,
      "step": 16175
    },
    {
      "entropy": 6.068248748779297,
      "epoch": 1.8672821696480093,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004660189931434207,
      "loss": 5.8004,
      "mean_token_accuracy": 0.18212874978780746,
      "num_tokens": 41007251.0,
      "step": 16180
    },
    {
      "entropy": 6.0908308029174805,
      "epoch": 1.8678592036930177,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004659971843728379,
      "loss": 5.877,
      "mean_token_accuracy": 0.17980101257562636,
      "num_tokens": 41020218.0,
      "step": 16185
    },
    {
      "entropy": 6.0212737083435055,
      "epoch": 1.8684362377380266,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004659753691781597,
      "loss": 5.8402,
      "mean_token_accuracy": 0.17672305554151535,
      "num_tokens": 41032862.0,
      "step": 16190
    },
    {
      "entropy": 5.9973221778869625,
      "epoch": 1.8690132717830352,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004659535475601198,
      "loss": 5.7216,
      "mean_token_accuracy": 0.1869569316506386,
      "num_tokens": 41045992.0,
      "step": 16195
    },
    {
      "entropy": 5.953974056243896,
      "epoch": 1.8695903058280439,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046593171951945226,
      "loss": 5.7646,
      "mean_token_accuracy": 0.18351998776197434,
      "num_tokens": 41057902.0,
      "step": 16200
    },
    {
      "entropy": 5.971217441558838,
      "epoch": 1.8701673398730525,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00046590988505689114,
      "loss": 5.7855,
      "mean_token_accuracy": 0.18416197448968888,
      "num_tokens": 41070318.0,
      "step": 16205
    },
    {
      "entropy": 6.050753259658814,
      "epoch": 1.8707443739180611,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046588804417317084,
      "loss": 5.846,
      "mean_token_accuracy": 0.17704913318157195,
      "num_tokens": 41083918.0,
      "step": 16210
    },
    {
      "entropy": 6.001997375488282,
      "epoch": 1.8713214079630698,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046586619686902597,
      "loss": 5.7983,
      "mean_token_accuracy": 0.18218977153301238,
      "num_tokens": 41097730.0,
      "step": 16215
    },
    {
      "entropy": 6.108248090744018,
      "epoch": 1.8718984420080784,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046584434314519144,
      "loss": 5.9073,
      "mean_token_accuracy": 0.1716300591826439,
      "num_tokens": 41110795.0,
      "step": 16220
    },
    {
      "entropy": 6.068462991714478,
      "epoch": 1.8724754760530873,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004658224830024022,
      "loss": 5.8032,
      "mean_token_accuracy": 0.1841044083237648,
      "num_tokens": 41123216.0,
      "step": 16225
    },
    {
      "entropy": 6.059892702102661,
      "epoch": 1.8730525100980957,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004658006164413935,
      "loss": 5.8929,
      "mean_token_accuracy": 0.17776608169078828,
      "num_tokens": 41135293.0,
      "step": 16230
    },
    {
      "entropy": 5.978354978561401,
      "epoch": 1.8736295441431046,
      "grad_norm": 1.0,
      "learning_rate": 0.0004657787434629009,
      "loss": 5.8281,
      "mean_token_accuracy": 0.19038864374160766,
      "num_tokens": 41149475.0,
      "step": 16235
    },
    {
      "entropy": 6.167351531982422,
      "epoch": 1.874206578188113,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046575686406765993,
      "loss": 5.891,
      "mean_token_accuracy": 0.17185684144496918,
      "num_tokens": 41162380.0,
      "step": 16240
    },
    {
      "entropy": 6.076961708068848,
      "epoch": 1.8747836122331218,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046573497825640664,
      "loss": 5.7293,
      "mean_token_accuracy": 0.18280680924654008,
      "num_tokens": 41175483.0,
      "step": 16245
    },
    {
      "entropy": 5.86759352684021,
      "epoch": 1.8753606462781303,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004657130860298771,
      "loss": 5.7267,
      "mean_token_accuracy": 0.19026558697223664,
      "num_tokens": 41189342.0,
      "step": 16250
    },
    {
      "entropy": 6.011432552337647,
      "epoch": 1.8759376803231391,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004656911873888077,
      "loss": 5.8099,
      "mean_token_accuracy": 0.18776544481515883,
      "num_tokens": 41202342.0,
      "step": 16255
    },
    {
      "entropy": 5.991354942321777,
      "epoch": 1.8765147143681478,
      "grad_norm": 1.0,
      "learning_rate": 0.0004656692823339349,
      "loss": 5.7609,
      "mean_token_accuracy": 0.17845608294010162,
      "num_tokens": 41215006.0,
      "step": 16260
    },
    {
      "entropy": 6.0740437507629395,
      "epoch": 1.8770917484131564,
      "grad_norm": 0.875,
      "learning_rate": 0.0004656473708659955,
      "loss": 5.8367,
      "mean_token_accuracy": 0.18067946434020996,
      "num_tokens": 41227735.0,
      "step": 16265
    },
    {
      "entropy": 6.031647968292236,
      "epoch": 1.877668782458165,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00046562545298572646,
      "loss": 5.7994,
      "mean_token_accuracy": 0.18391980677843095,
      "num_tokens": 41240731.0,
      "step": 16270
    },
    {
      "entropy": 6.061124563217163,
      "epoch": 1.8782458165031737,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000465603528693865,
      "loss": 5.8394,
      "mean_token_accuracy": 0.17864254415035247,
      "num_tokens": 41253190.0,
      "step": 16275
    },
    {
      "entropy": 6.058759975433349,
      "epoch": 1.8788228505481823,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046558159799114853,
      "loss": 5.8582,
      "mean_token_accuracy": 0.17517795115709306,
      "num_tokens": 41265224.0,
      "step": 16280
    },
    {
      "entropy": 5.935038805007935,
      "epoch": 1.879399884593191,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004655596608783147,
      "loss": 5.7458,
      "mean_token_accuracy": 0.19064777493476867,
      "num_tokens": 41277865.0,
      "step": 16285
    },
    {
      "entropy": 6.002727937698364,
      "epoch": 1.8799769186381996,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046553771735610136,
      "loss": 5.7812,
      "mean_token_accuracy": 0.1791316121816635,
      "num_tokens": 41291660.0,
      "step": 16290
    },
    {
      "entropy": 6.07350263595581,
      "epoch": 1.8805539526832082,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004655157674252465,
      "loss": 5.9045,
      "mean_token_accuracy": 0.17297184467315674,
      "num_tokens": 41304284.0,
      "step": 16295
    },
    {
      "entropy": 6.041435670852661,
      "epoch": 1.881130986728217,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046549381108648843,
      "loss": 5.802,
      "mean_token_accuracy": 0.18783484995365143,
      "num_tokens": 41316541.0,
      "step": 16300
    },
    {
      "entropy": 6.044201326370239,
      "epoch": 1.8817080207732255,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004654718483405656,
      "loss": 5.8215,
      "mean_token_accuracy": 0.18328167051076888,
      "num_tokens": 41329809.0,
      "step": 16305
    },
    {
      "entropy": 6.021008396148682,
      "epoch": 1.8822850548182344,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046544987918821685,
      "loss": 5.7746,
      "mean_token_accuracy": 0.1796517327427864,
      "num_tokens": 41341883.0,
      "step": 16310
    },
    {
      "entropy": 5.9968537330627445,
      "epoch": 1.8828620888632428,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000465427903630181,
      "loss": 5.7471,
      "mean_token_accuracy": 0.18361966013908387,
      "num_tokens": 41354161.0,
      "step": 16315
    },
    {
      "entropy": 6.016858339309692,
      "epoch": 1.8834391229082517,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004654059216671972,
      "loss": 5.6936,
      "mean_token_accuracy": 0.19069059193134308,
      "num_tokens": 41366737.0,
      "step": 16320
    },
    {
      "entropy": 6.076669359207154,
      "epoch": 1.88401615695326,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004653839333000048,
      "loss": 5.8362,
      "mean_token_accuracy": 0.18188464790582656,
      "num_tokens": 41380525.0,
      "step": 16325
    },
    {
      "entropy": 6.005020236968994,
      "epoch": 1.884593190998269,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046536193852934334,
      "loss": 5.7492,
      "mean_token_accuracy": 0.18200652450323104,
      "num_tokens": 41393361.0,
      "step": 16330
    },
    {
      "entropy": 6.008965730667114,
      "epoch": 1.8851702250432776,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046533993735595273,
      "loss": 5.8402,
      "mean_token_accuracy": 0.18226926475763322,
      "num_tokens": 41405919.0,
      "step": 16335
    },
    {
      "entropy": 6.047534036636352,
      "epoch": 1.8857472590882862,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046531792978057277,
      "loss": 5.8123,
      "mean_token_accuracy": 0.182885105907917,
      "num_tokens": 41417870.0,
      "step": 16340
    },
    {
      "entropy": 5.974366855621338,
      "epoch": 1.8863242931332949,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004652959158039438,
      "loss": 5.8034,
      "mean_token_accuracy": 0.19051939398050308,
      "num_tokens": 41429242.0,
      "step": 16345
    },
    {
      "entropy": 6.107768917083741,
      "epoch": 1.8869013271783035,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004652738954268062,
      "loss": 5.8961,
      "mean_token_accuracy": 0.17473069429397584,
      "num_tokens": 41442073.0,
      "step": 16350
    },
    {
      "entropy": 6.0322259902954105,
      "epoch": 1.8874783612233121,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046525186864990073,
      "loss": 5.8994,
      "mean_token_accuracy": 0.1744527742266655,
      "num_tokens": 41453429.0,
      "step": 16355
    },
    {
      "entropy": 6.027315902709961,
      "epoch": 1.8880553952683208,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046522983547396806,
      "loss": 5.8262,
      "mean_token_accuracy": 0.18459306061267852,
      "num_tokens": 41464940.0,
      "step": 16360
    },
    {
      "entropy": 6.103169107437134,
      "epoch": 1.8886324293133296,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004652077958997494,
      "loss": 5.8418,
      "mean_token_accuracy": 0.1721346780657768,
      "num_tokens": 41477562.0,
      "step": 16365
    },
    {
      "entropy": 6.035250234603882,
      "epoch": 1.889209463358338,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046518574992798604,
      "loss": 5.8143,
      "mean_token_accuracy": 0.1887367010116577,
      "num_tokens": 41490299.0,
      "step": 16370
    },
    {
      "entropy": 6.012840843200683,
      "epoch": 1.889786497403347,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046516369755941945,
      "loss": 5.7314,
      "mean_token_accuracy": 0.18549880087375642,
      "num_tokens": 41501305.0,
      "step": 16375
    },
    {
      "entropy": 6.1314490795135494,
      "epoch": 1.8903635314483553,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004651416387947914,
      "loss": 5.9142,
      "mean_token_accuracy": 0.17746112048625945,
      "num_tokens": 41515390.0,
      "step": 16380
    },
    {
      "entropy": 6.062517261505127,
      "epoch": 1.8909405654933642,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004651195736348437,
      "loss": 5.9033,
      "mean_token_accuracy": 0.18322975933551788,
      "num_tokens": 41528247.0,
      "step": 16385
    },
    {
      "entropy": 5.99382758140564,
      "epoch": 1.8915175995383726,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004650975020803186,
      "loss": 5.7791,
      "mean_token_accuracy": 0.1803511068224907,
      "num_tokens": 41542249.0,
      "step": 16390
    },
    {
      "entropy": 6.015912675857544,
      "epoch": 1.8920946335833815,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004650754241319584,
      "loss": 5.8197,
      "mean_token_accuracy": 0.1793270081281662,
      "num_tokens": 41554835.0,
      "step": 16395
    },
    {
      "entropy": 6.116751337051392,
      "epoch": 1.8926716676283901,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046505333979050573,
      "loss": 5.8658,
      "mean_token_accuracy": 0.17514849603176116,
      "num_tokens": 41566427.0,
      "step": 16400
    },
    {
      "entropy": 6.047809028625489,
      "epoch": 1.8932487016733988,
      "grad_norm": 0.875,
      "learning_rate": 0.0004650312490567035,
      "loss": 5.8908,
      "mean_token_accuracy": 0.17499162405729293,
      "num_tokens": 41580193.0,
      "step": 16405
    },
    {
      "entropy": 6.016090726852417,
      "epoch": 1.8938257357184074,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004650091519312945,
      "loss": 5.7264,
      "mean_token_accuracy": 0.18968772888183594,
      "num_tokens": 41594048.0,
      "step": 16410
    },
    {
      "entropy": 6.106095695495606,
      "epoch": 1.894402769763416,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046498704841502203,
      "loss": 5.8816,
      "mean_token_accuracy": 0.1731876850128174,
      "num_tokens": 41607283.0,
      "step": 16415
    },
    {
      "entropy": 6.023270177841186,
      "epoch": 1.8949798038084247,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004649649385086296,
      "loss": 5.7331,
      "mean_token_accuracy": 0.19095669984817504,
      "num_tokens": 41619844.0,
      "step": 16420
    },
    {
      "entropy": 5.8834144115448,
      "epoch": 1.8955568378534333,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004649428222128608,
      "loss": 5.6864,
      "mean_token_accuracy": 0.1901530221104622,
      "num_tokens": 41631764.0,
      "step": 16425
    },
    {
      "entropy": 6.080666589736938,
      "epoch": 1.896133871898442,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046492069952845953,
      "loss": 5.8589,
      "mean_token_accuracy": 0.17537793815135955,
      "num_tokens": 41643703.0,
      "step": 16430
    },
    {
      "entropy": 6.07476487159729,
      "epoch": 1.8967109059434506,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046489857045617,
      "loss": 5.8442,
      "mean_token_accuracy": 0.18390381336212158,
      "num_tokens": 41657057.0,
      "step": 16435
    },
    {
      "entropy": 6.026254320144654,
      "epoch": 1.8972879399884595,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046487643499673625,
      "loss": 5.86,
      "mean_token_accuracy": 0.17803988754749298,
      "num_tokens": 41669870.0,
      "step": 16440
    },
    {
      "entropy": 6.076148986816406,
      "epoch": 1.8978649740334679,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004648542931509029,
      "loss": 5.7975,
      "mean_token_accuracy": 0.17993361055850982,
      "num_tokens": 41681015.0,
      "step": 16445
    },
    {
      "entropy": 6.036432409286499,
      "epoch": 1.8984420080784767,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004648321449194148,
      "loss": 5.829,
      "mean_token_accuracy": 0.18756165653467177,
      "num_tokens": 41694116.0,
      "step": 16450
    },
    {
      "entropy": 6.0329348087310795,
      "epoch": 1.8990190421234852,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046480999030301673,
      "loss": 5.8442,
      "mean_token_accuracy": 0.17569254338741302,
      "num_tokens": 41707294.0,
      "step": 16455
    },
    {
      "entropy": 6.140571641921997,
      "epoch": 1.899596076168494,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046478782930245395,
      "loss": 5.9128,
      "mean_token_accuracy": 0.17540325671434404,
      "num_tokens": 41719837.0,
      "step": 16460
    },
    {
      "entropy": 6.0995622158050535,
      "epoch": 1.9001731102135024,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046476566191847176,
      "loss": 5.7909,
      "mean_token_accuracy": 0.17919419407844545,
      "num_tokens": 41732308.0,
      "step": 16465
    },
    {
      "entropy": 5.997230386734008,
      "epoch": 1.9007501442585113,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004647434881518159,
      "loss": 5.7824,
      "mean_token_accuracy": 0.1817353442311287,
      "num_tokens": 41745051.0,
      "step": 16470
    },
    {
      "entropy": 6.046524095535278,
      "epoch": 1.90132717830352,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046472130800323194,
      "loss": 5.8205,
      "mean_token_accuracy": 0.1840731531381607,
      "num_tokens": 41757545.0,
      "step": 16475
    },
    {
      "entropy": 6.034554386138916,
      "epoch": 1.9019042123485286,
      "grad_norm": 0.875,
      "learning_rate": 0.0004646991214734661,
      "loss": 5.8196,
      "mean_token_accuracy": 0.1793336719274521,
      "num_tokens": 41772157.0,
      "step": 16480
    },
    {
      "entropy": 6.047153759002685,
      "epoch": 1.9024812463935372,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004646769285632645,
      "loss": 5.8226,
      "mean_token_accuracy": 0.18367716819047927,
      "num_tokens": 41785404.0,
      "step": 16485
    },
    {
      "entropy": 5.983449649810791,
      "epoch": 1.9030582804385459,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004646547292733737,
      "loss": 5.7598,
      "mean_token_accuracy": 0.19147514551877975,
      "num_tokens": 41797758.0,
      "step": 16490
    },
    {
      "entropy": 6.015515947341919,
      "epoch": 1.9036353144835545,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004646325236045402,
      "loss": 5.8757,
      "mean_token_accuracy": 0.18391702473163604,
      "num_tokens": 41811162.0,
      "step": 16495
    },
    {
      "entropy": 6.00651912689209,
      "epoch": 1.9042123485285631,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000464610311557511,
      "loss": 5.7371,
      "mean_token_accuracy": 0.18388524055480956,
      "num_tokens": 41823304.0,
      "step": 16500
    },
    {
      "entropy": 5.933037185668946,
      "epoch": 1.904789382573572,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004645880931330331,
      "loss": 5.7139,
      "mean_token_accuracy": 0.18777222484350203,
      "num_tokens": 41837027.0,
      "step": 16505
    },
    {
      "entropy": 6.0248188972473145,
      "epoch": 1.9053664166185804,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004645658683318539,
      "loss": 5.8442,
      "mean_token_accuracy": 0.1826476648449898,
      "num_tokens": 41851887.0,
      "step": 16510
    },
    {
      "entropy": 6.090163373947144,
      "epoch": 1.9059434506635893,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004645436371547209,
      "loss": 5.7959,
      "mean_token_accuracy": 0.18397875130176544,
      "num_tokens": 41864908.0,
      "step": 16515
    },
    {
      "entropy": 6.111636590957642,
      "epoch": 1.9065204847085977,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046452139960238186,
      "loss": 5.8276,
      "mean_token_accuracy": 0.18277143836021423,
      "num_tokens": 41877604.0,
      "step": 16520
    },
    {
      "entropy": 6.034374713897705,
      "epoch": 1.9070975187536066,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00046449915567558467,
      "loss": 5.8239,
      "mean_token_accuracy": 0.1749449521303177,
      "num_tokens": 41890580.0,
      "step": 16525
    },
    {
      "entropy": 5.989800548553466,
      "epoch": 1.907674552798615,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004644769053750775,
      "loss": 5.8208,
      "mean_token_accuracy": 0.17860884368419647,
      "num_tokens": 41903833.0,
      "step": 16530
    },
    {
      "entropy": 6.044780826568603,
      "epoch": 1.9082515868436238,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004644546487016087,
      "loss": 5.8193,
      "mean_token_accuracy": 0.17884487956762313,
      "num_tokens": 41916094.0,
      "step": 16535
    },
    {
      "entropy": 6.0077831745147705,
      "epoch": 1.9088286208886325,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00046443238565592687,
      "loss": 5.7737,
      "mean_token_accuracy": 0.18054241985082625,
      "num_tokens": 41929199.0,
      "step": 16540
    },
    {
      "entropy": 6.051807165145874,
      "epoch": 1.9094056549336411,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046441011623878087,
      "loss": 5.8056,
      "mean_token_accuracy": 0.178181754052639,
      "num_tokens": 41942254.0,
      "step": 16545
    },
    {
      "entropy": 6.102326393127441,
      "epoch": 1.9099826889786498,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004643878404509197,
      "loss": 5.8335,
      "mean_token_accuracy": 0.17778859436511993,
      "num_tokens": 41954129.0,
      "step": 16550
    },
    {
      "entropy": 6.075048828125,
      "epoch": 1.9105597230236584,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046436555829309264,
      "loss": 5.8272,
      "mean_token_accuracy": 0.17996654510498047,
      "num_tokens": 41966489.0,
      "step": 16555
    },
    {
      "entropy": 5.968257856369019,
      "epoch": 1.911136757068667,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000464343269766049,
      "loss": 5.7627,
      "mean_token_accuracy": 0.18915678262710572,
      "num_tokens": 41980235.0,
      "step": 16560
    },
    {
      "entropy": 6.0489274024963375,
      "epoch": 1.9117137911136757,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046432097487053857,
      "loss": 5.7581,
      "mean_token_accuracy": 0.18709539622068405,
      "num_tokens": 41992531.0,
      "step": 16565
    },
    {
      "entropy": 6.046317100524902,
      "epoch": 1.9122908251586843,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046429867360731124,
      "loss": 5.8404,
      "mean_token_accuracy": 0.18301450610160827,
      "num_tokens": 42004825.0,
      "step": 16570
    },
    {
      "entropy": 6.061797332763672,
      "epoch": 1.912867859203693,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046427636597711695,
      "loss": 5.7836,
      "mean_token_accuracy": 0.18293384909629823,
      "num_tokens": 42017275.0,
      "step": 16575
    },
    {
      "entropy": 6.012455463409424,
      "epoch": 1.9134448932487018,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046425405198070624,
      "loss": 5.801,
      "mean_token_accuracy": 0.19055292010307312,
      "num_tokens": 42031102.0,
      "step": 16580
    },
    {
      "entropy": 6.051971864700318,
      "epoch": 1.9140219272937102,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046423173161882944,
      "loss": 5.8457,
      "mean_token_accuracy": 0.17437489926815034,
      "num_tokens": 42043276.0,
      "step": 16585
    },
    {
      "entropy": 6.097778844833374,
      "epoch": 1.914598961338719,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046420940489223735,
      "loss": 5.7796,
      "mean_token_accuracy": 0.18555857837200165,
      "num_tokens": 42055634.0,
      "step": 16590
    },
    {
      "entropy": 6.0181280136108395,
      "epoch": 1.9151759953837275,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004641870718016809,
      "loss": 5.7375,
      "mean_token_accuracy": 0.18321775197982787,
      "num_tokens": 42068910.0,
      "step": 16595
    },
    {
      "entropy": 5.930918979644775,
      "epoch": 1.9157530294287364,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004641647323479113,
      "loss": 5.7165,
      "mean_token_accuracy": 0.1831022560596466,
      "num_tokens": 42081008.0,
      "step": 16600
    },
    {
      "entropy": 5.9950813293457035,
      "epoch": 1.9163300634737448,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004641423865316798,
      "loss": 5.8112,
      "mean_token_accuracy": 0.1840982422232628,
      "num_tokens": 42093982.0,
      "step": 16605
    },
    {
      "entropy": 6.049228525161743,
      "epoch": 1.9169070975187537,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004641200343537381,
      "loss": 5.8075,
      "mean_token_accuracy": 0.18128660321235657,
      "num_tokens": 42105441.0,
      "step": 16610
    },
    {
      "entropy": 6.109910202026367,
      "epoch": 1.9174841315637623,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000464097675814838,
      "loss": 5.8297,
      "mean_token_accuracy": 0.1800685629248619,
      "num_tokens": 42117636.0,
      "step": 16615
    },
    {
      "entropy": 6.052115631103516,
      "epoch": 1.918061165608771,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004640753109157316,
      "loss": 5.8225,
      "mean_token_accuracy": 0.17841600328683854,
      "num_tokens": 42130009.0,
      "step": 16620
    },
    {
      "entropy": 6.027559280395508,
      "epoch": 1.9186381996537796,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046405293965717093,
      "loss": 5.7964,
      "mean_token_accuracy": 0.18438176214694976,
      "num_tokens": 42142181.0,
      "step": 16625
    },
    {
      "entropy": 6.029458856582641,
      "epoch": 1.9192152336987882,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004640305620399086,
      "loss": 5.8416,
      "mean_token_accuracy": 0.1830775499343872,
      "num_tokens": 42155626.0,
      "step": 16630
    },
    {
      "entropy": 6.014059686660767,
      "epoch": 1.9197922677437969,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004640081780646971,
      "loss": 5.8333,
      "mean_token_accuracy": 0.18027895241975783,
      "num_tokens": 42169309.0,
      "step": 16635
    },
    {
      "entropy": 6.10810432434082,
      "epoch": 1.9203693017888055,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046398578773228954,
      "loss": 5.8506,
      "mean_token_accuracy": 0.18482111692428588,
      "num_tokens": 42182372.0,
      "step": 16640
    },
    {
      "entropy": 5.948515892028809,
      "epoch": 1.9209463358338144,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046396339104343886,
      "loss": 5.6901,
      "mean_token_accuracy": 0.18857464343309402,
      "num_tokens": 42194792.0,
      "step": 16645
    },
    {
      "entropy": 6.0487837314605715,
      "epoch": 1.9215233698788228,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004639409879988984,
      "loss": 5.8506,
      "mean_token_accuracy": 0.1803712859749794,
      "num_tokens": 42207194.0,
      "step": 16650
    },
    {
      "entropy": 6.020299291610717,
      "epoch": 1.9221004039238316,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004639185785994216,
      "loss": 5.7996,
      "mean_token_accuracy": 0.1860925540328026,
      "num_tokens": 42219058.0,
      "step": 16655
    },
    {
      "entropy": 6.012314796447754,
      "epoch": 1.92267743796884,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046389616284576226,
      "loss": 5.7764,
      "mean_token_accuracy": 0.19074427336454391,
      "num_tokens": 42230726.0,
      "step": 16660
    },
    {
      "entropy": 5.898144960403442,
      "epoch": 1.923254472013849,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046387374073867435,
      "loss": 5.6227,
      "mean_token_accuracy": 0.20040173083543777,
      "num_tokens": 42243859.0,
      "step": 16665
    },
    {
      "entropy": 6.0709311962127686,
      "epoch": 1.9238315060588573,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046385131227891197,
      "loss": 5.8716,
      "mean_token_accuracy": 0.1749637871980667,
      "num_tokens": 42257253.0,
      "step": 16670
    },
    {
      "entropy": 6.083539533615112,
      "epoch": 1.9244085401038662,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004638288774672295,
      "loss": 5.8593,
      "mean_token_accuracy": 0.17598051875829696,
      "num_tokens": 42269202.0,
      "step": 16675
    },
    {
      "entropy": 6.056668615341186,
      "epoch": 1.9249855741488748,
      "grad_norm": 0.875,
      "learning_rate": 0.0004638064363043816,
      "loss": 5.8114,
      "mean_token_accuracy": 0.18566043376922609,
      "num_tokens": 42282714.0,
      "step": 16680
    },
    {
      "entropy": 6.092519235610962,
      "epoch": 1.9255626081938835,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046378398879112293,
      "loss": 5.8408,
      "mean_token_accuracy": 0.17610444724559784,
      "num_tokens": 42297272.0,
      "step": 16685
    },
    {
      "entropy": 6.052019453048706,
      "epoch": 1.9261396422388921,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004637615349282086,
      "loss": 5.8641,
      "mean_token_accuracy": 0.18478135168552398,
      "num_tokens": 42308938.0,
      "step": 16690
    },
    {
      "entropy": 5.999912548065185,
      "epoch": 1.9267166762839008,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004637390747163938,
      "loss": 5.8291,
      "mean_token_accuracy": 0.17928818762302398,
      "num_tokens": 42321310.0,
      "step": 16695
    },
    {
      "entropy": 5.930019330978394,
      "epoch": 1.9272937103289094,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004637166081564339,
      "loss": 5.7797,
      "mean_token_accuracy": 0.19012727439403534,
      "num_tokens": 42332760.0,
      "step": 16700
    },
    {
      "entropy": 6.0544960498809814,
      "epoch": 1.927870744373918,
      "grad_norm": 0.859375,
      "learning_rate": 0.00046369413524908473,
      "loss": 5.7725,
      "mean_token_accuracy": 0.18740156888961793,
      "num_tokens": 42346993.0,
      "step": 16705
    },
    {
      "entropy": 6.048433017730713,
      "epoch": 1.9284477784189267,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046367165599510207,
      "loss": 5.8476,
      "mean_token_accuracy": 0.1843491941690445,
      "num_tokens": 42358701.0,
      "step": 16710
    },
    {
      "entropy": 6.061404514312744,
      "epoch": 1.9290248124639353,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004636491703952419,
      "loss": 5.8897,
      "mean_token_accuracy": 0.1851487085223198,
      "num_tokens": 42371429.0,
      "step": 16715
    },
    {
      "entropy": 6.058820486068726,
      "epoch": 1.9296018465089442,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046362667845026057,
      "loss": 5.8181,
      "mean_token_accuracy": 0.18111080080270767,
      "num_tokens": 42383346.0,
      "step": 16720
    },
    {
      "entropy": 6.007965183258056,
      "epoch": 1.9301788805539526,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046360418016091467,
      "loss": 5.7957,
      "mean_token_accuracy": 0.18789906054735184,
      "num_tokens": 42395247.0,
      "step": 16725
    },
    {
      "entropy": 6.112230443954468,
      "epoch": 1.9307559145989615,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046358167552796085,
      "loss": 5.7899,
      "mean_token_accuracy": 0.18112557828426362,
      "num_tokens": 42408009.0,
      "step": 16730
    },
    {
      "entropy": 6.0204085350036625,
      "epoch": 1.9313329486439699,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046355916455215597,
      "loss": 5.8318,
      "mean_token_accuracy": 0.1760340467095375,
      "num_tokens": 42420367.0,
      "step": 16735
    },
    {
      "entropy": 6.060974264144898,
      "epoch": 1.9319099826889787,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004635366472342573,
      "loss": 5.8797,
      "mean_token_accuracy": 0.1790865868330002,
      "num_tokens": 42433298.0,
      "step": 16740
    },
    {
      "entropy": 6.068971014022827,
      "epoch": 1.9324870167339872,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004635141235750222,
      "loss": 5.856,
      "mean_token_accuracy": 0.18127947598695754,
      "num_tokens": 42446422.0,
      "step": 16745
    },
    {
      "entropy": 6.050488519668579,
      "epoch": 1.933064050778996,
      "grad_norm": 1.0,
      "learning_rate": 0.00046349159357520815,
      "loss": 5.8477,
      "mean_token_accuracy": 0.1733735144138336,
      "num_tokens": 42458874.0,
      "step": 16750
    },
    {
      "entropy": 6.05755205154419,
      "epoch": 1.9336410848240047,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000463469057235573,
      "loss": 5.7886,
      "mean_token_accuracy": 0.1818990081548691,
      "num_tokens": 42472093.0,
      "step": 16755
    },
    {
      "entropy": 5.915979433059692,
      "epoch": 1.9342181188690133,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00046344651455687476,
      "loss": 5.7356,
      "mean_token_accuracy": 0.18848615288734435,
      "num_tokens": 42485396.0,
      "step": 16760
    },
    {
      "entropy": 6.053208351135254,
      "epoch": 1.934795152914022,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046342396553987155,
      "loss": 5.8409,
      "mean_token_accuracy": 0.1814213678240776,
      "num_tokens": 42498379.0,
      "step": 16765
    },
    {
      "entropy": 6.094767618179321,
      "epoch": 1.9353721869590306,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046340141018532186,
      "loss": 5.8039,
      "mean_token_accuracy": 0.1839069440960884,
      "num_tokens": 42512125.0,
      "step": 16770
    },
    {
      "entropy": 5.976181507110596,
      "epoch": 1.9359492210040392,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004633788484939843,
      "loss": 5.7707,
      "mean_token_accuracy": 0.18441660553216935,
      "num_tokens": 42525539.0,
      "step": 16775
    },
    {
      "entropy": 5.978352785110474,
      "epoch": 1.9365262550490479,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046335628046661776,
      "loss": 5.7271,
      "mean_token_accuracy": 0.18636305630207062,
      "num_tokens": 42538399.0,
      "step": 16780
    },
    {
      "entropy": 6.108949899673462,
      "epoch": 1.9371032890940567,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046333370610398135,
      "loss": 5.8853,
      "mean_token_accuracy": 0.1792922481894493,
      "num_tokens": 42550607.0,
      "step": 16785
    },
    {
      "entropy": 6.184437990188599,
      "epoch": 1.9376803231390651,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004633111254068342,
      "loss": 5.9311,
      "mean_token_accuracy": 0.16933584809303284,
      "num_tokens": 42563113.0,
      "step": 16790
    },
    {
      "entropy": 5.995370197296142,
      "epoch": 1.938257357184074,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004632885383759359,
      "loss": 5.778,
      "mean_token_accuracy": 0.1862604171037674,
      "num_tokens": 42574792.0,
      "step": 16795
    },
    {
      "entropy": 5.965495014190674,
      "epoch": 1.9388343912290824,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00046326594501204624,
      "loss": 5.7109,
      "mean_token_accuracy": 0.18787965625524522,
      "num_tokens": 42587360.0,
      "step": 16800
    },
    {
      "entropy": 6.024766540527343,
      "epoch": 1.9394114252740913,
      "grad_norm": 0.921875,
      "learning_rate": 0.000463243345315925,
      "loss": 5.7877,
      "mean_token_accuracy": 0.1800096333026886,
      "num_tokens": 42599901.0,
      "step": 16805
    },
    {
      "entropy": 6.041310262680054,
      "epoch": 1.9399884593190997,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046322073928833224,
      "loss": 5.8071,
      "mean_token_accuracy": 0.18491238951683045,
      "num_tokens": 42612127.0,
      "step": 16810
    },
    {
      "entropy": 6.123684549331665,
      "epoch": 1.9405654933641086,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004631981269300285,
      "loss": 5.9093,
      "mean_token_accuracy": 0.1760391980409622,
      "num_tokens": 42624086.0,
      "step": 16815
    },
    {
      "entropy": 6.120830011367798,
      "epoch": 1.9411425274091172,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004631755082417742,
      "loss": 5.8669,
      "mean_token_accuracy": 0.18229341357946396,
      "num_tokens": 42637402.0,
      "step": 16820
    },
    {
      "entropy": 6.044957685470581,
      "epoch": 1.9417195614541258,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004631528832243302,
      "loss": 5.8382,
      "mean_token_accuracy": 0.17690208405256272,
      "num_tokens": 42648162.0,
      "step": 16825
    },
    {
      "entropy": 5.982219934463501,
      "epoch": 1.9422965954991345,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046313025187845735,
      "loss": 5.8045,
      "mean_token_accuracy": 0.18325186371803284,
      "num_tokens": 42658787.0,
      "step": 16830
    },
    {
      "entropy": 6.046440505981446,
      "epoch": 1.942873629544143,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046310761420491705,
      "loss": 5.839,
      "mean_token_accuracy": 0.17343674302101136,
      "num_tokens": 42672036.0,
      "step": 16835
    },
    {
      "entropy": 6.076276445388794,
      "epoch": 1.9434506635891518,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004630849702044705,
      "loss": 5.8029,
      "mean_token_accuracy": 0.17841587960720062,
      "num_tokens": 42683691.0,
      "step": 16840
    },
    {
      "entropy": 5.98874044418335,
      "epoch": 1.9440276976341604,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046306231987787937,
      "loss": 5.7988,
      "mean_token_accuracy": 0.18201770931482314,
      "num_tokens": 42697469.0,
      "step": 16845
    },
    {
      "entropy": 6.041253995895386,
      "epoch": 1.944604731679169,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046303966322590556,
      "loss": 5.8432,
      "mean_token_accuracy": 0.17816432118415831,
      "num_tokens": 42710792.0,
      "step": 16850
    },
    {
      "entropy": 6.034585189819336,
      "epoch": 1.9451817657241777,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004630170002493111,
      "loss": 5.7903,
      "mean_token_accuracy": 0.18545068502426149,
      "num_tokens": 42722549.0,
      "step": 16855
    },
    {
      "entropy": 6.008102130889893,
      "epoch": 1.9457587997691865,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046299433094885826,
      "loss": 5.796,
      "mean_token_accuracy": 0.18120642006397247,
      "num_tokens": 42736113.0,
      "step": 16860
    },
    {
      "entropy": 6.049204969406128,
      "epoch": 1.946335833814195,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046297165532530944,
      "loss": 5.7748,
      "mean_token_accuracy": 0.18600474447011947,
      "num_tokens": 42748801.0,
      "step": 16865
    },
    {
      "entropy": 6.051483011245727,
      "epoch": 1.9469128678592038,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004629489733794274,
      "loss": 5.8368,
      "mean_token_accuracy": 0.1803561493754387,
      "num_tokens": 42760738.0,
      "step": 16870
    },
    {
      "entropy": 6.077905654907227,
      "epoch": 1.9474899019042122,
      "grad_norm": 1.0,
      "learning_rate": 0.000462926285111975,
      "loss": 5.8386,
      "mean_token_accuracy": 0.17866179645061492,
      "num_tokens": 42772470.0,
      "step": 16875
    },
    {
      "entropy": 6.051263380050659,
      "epoch": 1.948066935949221,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046290359052371535,
      "loss": 5.8836,
      "mean_token_accuracy": 0.18016574233770372,
      "num_tokens": 42784919.0,
      "step": 16880
    },
    {
      "entropy": 6.033104133605957,
      "epoch": 1.9486439699942295,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004628808896154117,
      "loss": 5.7445,
      "mean_token_accuracy": 0.18956043124198912,
      "num_tokens": 42797722.0,
      "step": 16885
    },
    {
      "entropy": 6.078951168060303,
      "epoch": 1.9492210040392384,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004628581823878277,
      "loss": 5.8569,
      "mean_token_accuracy": 0.17718900591135026,
      "num_tokens": 42809539.0,
      "step": 16890
    },
    {
      "entropy": 5.967656183242798,
      "epoch": 1.949798038084247,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000462835468841727,
      "loss": 5.6044,
      "mean_token_accuracy": 0.18973815441131592,
      "num_tokens": 42821519.0,
      "step": 16895
    },
    {
      "entropy": 5.998979139328003,
      "epoch": 1.9503750721292556,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004628127489778737,
      "loss": 5.7371,
      "mean_token_accuracy": 0.18968395441770552,
      "num_tokens": 42833966.0,
      "step": 16900
    },
    {
      "entropy": 5.9654953956604,
      "epoch": 1.9509521061742643,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004627900227970318,
      "loss": 5.7899,
      "mean_token_accuracy": 0.18958741873502732,
      "num_tokens": 42846162.0,
      "step": 16905
    },
    {
      "entropy": 6.022485399246216,
      "epoch": 1.951529140219273,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046276729029996576,
      "loss": 5.8484,
      "mean_token_accuracy": 0.18595175445079803,
      "num_tokens": 42860019.0,
      "step": 16910
    },
    {
      "entropy": 6.069406270980835,
      "epoch": 1.9521061742642816,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046274455148744025,
      "loss": 5.8957,
      "mean_token_accuracy": 0.17203402370214463,
      "num_tokens": 42872699.0,
      "step": 16915
    },
    {
      "entropy": 6.034542274475098,
      "epoch": 1.9526832083092902,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046272180636022,
      "loss": 5.85,
      "mean_token_accuracy": 0.18414656072854996,
      "num_tokens": 42884907.0,
      "step": 16920
    },
    {
      "entropy": 6.078352737426758,
      "epoch": 1.953260242354299,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046269905491907,
      "loss": 5.8457,
      "mean_token_accuracy": 0.18092133104801178,
      "num_tokens": 42898387.0,
      "step": 16925
    },
    {
      "entropy": 6.116144227981567,
      "epoch": 1.9538372763993075,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004626762971647555,
      "loss": 5.8607,
      "mean_token_accuracy": 0.17819654494524,
      "num_tokens": 42910845.0,
      "step": 16930
    },
    {
      "entropy": 6.017821741104126,
      "epoch": 1.9544143104443163,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046265353309804205,
      "loss": 5.8261,
      "mean_token_accuracy": 0.1807723045349121,
      "num_tokens": 42922494.0,
      "step": 16935
    },
    {
      "entropy": 6.030594825744629,
      "epoch": 1.9549913444893248,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004626307627196952,
      "loss": 5.7892,
      "mean_token_accuracy": 0.17765034288167952,
      "num_tokens": 42934618.0,
      "step": 16940
    },
    {
      "entropy": 6.065912199020386,
      "epoch": 1.9555683785343336,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004626079860304808,
      "loss": 5.7853,
      "mean_token_accuracy": 0.17963001132011414,
      "num_tokens": 42947518.0,
      "step": 16945
    },
    {
      "entropy": 5.923216342926025,
      "epoch": 1.956145412579342,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00046258520303116496,
      "loss": 5.7664,
      "mean_token_accuracy": 0.18624502569437026,
      "num_tokens": 42960068.0,
      "step": 16950
    },
    {
      "entropy": 6.055463266372681,
      "epoch": 1.956722446624351,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004625624137225141,
      "loss": 5.7989,
      "mean_token_accuracy": 0.17723031342029572,
      "num_tokens": 42972982.0,
      "step": 16955
    },
    {
      "entropy": 5.970288467407227,
      "epoch": 1.9572994806693593,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004625396181052945,
      "loss": 5.7252,
      "mean_token_accuracy": 0.1974362090229988,
      "num_tokens": 42985040.0,
      "step": 16960
    },
    {
      "entropy": 6.091972494125367,
      "epoch": 1.9578765147143682,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046251681618027316,
      "loss": 5.8686,
      "mean_token_accuracy": 0.17186680436134338,
      "num_tokens": 42998503.0,
      "step": 16965
    },
    {
      "entropy": 5.991856718063355,
      "epoch": 1.9584535487593768,
      "grad_norm": 0.875,
      "learning_rate": 0.0004624940079482167,
      "loss": 5.7083,
      "mean_token_accuracy": 0.19384026229381562,
      "num_tokens": 43011334.0,
      "step": 16970
    },
    {
      "entropy": 5.931790733337403,
      "epoch": 1.9590305828043855,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046247119340989254,
      "loss": 5.7326,
      "mean_token_accuracy": 0.1932584449648857,
      "num_tokens": 43024772.0,
      "step": 16975
    },
    {
      "entropy": 6.1196709156036375,
      "epoch": 1.959607616849394,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004624483725660678,
      "loss": 5.7956,
      "mean_token_accuracy": 0.18500371873378754,
      "num_tokens": 43036733.0,
      "step": 16980
    },
    {
      "entropy": 6.04849271774292,
      "epoch": 1.9601846508944027,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004624255454175102,
      "loss": 5.8172,
      "mean_token_accuracy": 0.18441128432750703,
      "num_tokens": 43050482.0,
      "step": 16985
    },
    {
      "entropy": 6.140427827835083,
      "epoch": 1.9607616849394114,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004624027119649875,
      "loss": 5.896,
      "mean_token_accuracy": 0.1736210837960243,
      "num_tokens": 43062826.0,
      "step": 16990
    },
    {
      "entropy": 6.025839233398438,
      "epoch": 1.96133871898442,
      "grad_norm": 0.890625,
      "learning_rate": 0.00046237987220926766,
      "loss": 5.7652,
      "mean_token_accuracy": 0.18915013074874878,
      "num_tokens": 43077134.0,
      "step": 16995
    },
    {
      "entropy": 5.92431058883667,
      "epoch": 1.9619157530294289,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046235702615111886,
      "loss": 5.7232,
      "mean_token_accuracy": 0.19181123673915862,
      "num_tokens": 43089871.0,
      "step": 17000
    },
    {
      "entropy": 6.055762577056885,
      "epoch": 1.9624927870744373,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004623341737913096,
      "loss": 5.8377,
      "mean_token_accuracy": 0.18268973678350447,
      "num_tokens": 43102977.0,
      "step": 17005
    },
    {
      "entropy": 6.033108282089233,
      "epoch": 1.9630698211194462,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004623113151306085,
      "loss": 5.7729,
      "mean_token_accuracy": 0.18458595871925354,
      "num_tokens": 43114685.0,
      "step": 17010
    },
    {
      "entropy": 6.015901565551758,
      "epoch": 1.9636468551644546,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046228845016978425,
      "loss": 5.7531,
      "mean_token_accuracy": 0.18066166788339616,
      "num_tokens": 43127208.0,
      "step": 17015
    },
    {
      "entropy": 6.035107183456421,
      "epoch": 1.9642238892094634,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004622655789096061,
      "loss": 5.7212,
      "mean_token_accuracy": 0.18810130208730697,
      "num_tokens": 43140732.0,
      "step": 17020
    },
    {
      "entropy": 5.9659887790679935,
      "epoch": 1.9648009232544719,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046224270135084315,
      "loss": 5.7972,
      "mean_token_accuracy": 0.1785309538245201,
      "num_tokens": 43152746.0,
      "step": 17025
    },
    {
      "entropy": 6.004025793075561,
      "epoch": 1.9653779572994807,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046221981749426503,
      "loss": 5.7562,
      "mean_token_accuracy": 0.17872935384511948,
      "num_tokens": 43165608.0,
      "step": 17030
    },
    {
      "entropy": 6.054522466659546,
      "epoch": 1.9659549913444894,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00046219692734064124,
      "loss": 5.7963,
      "mean_token_accuracy": 0.18241801857948303,
      "num_tokens": 43178805.0,
      "step": 17035
    },
    {
      "entropy": 6.08264856338501,
      "epoch": 1.966532025389498,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004621740308907419,
      "loss": 5.8754,
      "mean_token_accuracy": 0.17564467787742616,
      "num_tokens": 43191366.0,
      "step": 17040
    },
    {
      "entropy": 6.054989194869995,
      "epoch": 1.9671090594345066,
      "grad_norm": 0.875,
      "learning_rate": 0.0004621511281453369,
      "loss": 5.8313,
      "mean_token_accuracy": 0.1780903622508049,
      "num_tokens": 43203748.0,
      "step": 17045
    },
    {
      "entropy": 6.013095903396606,
      "epoch": 1.9676860934795153,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004621282191051967,
      "loss": 5.7751,
      "mean_token_accuracy": 0.18758614808321,
      "num_tokens": 43215661.0,
      "step": 17050
    },
    {
      "entropy": 6.049832773208618,
      "epoch": 1.968263127524524,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004621053037710918,
      "loss": 5.8488,
      "mean_token_accuracy": 0.1818351075053215,
      "num_tokens": 43227869.0,
      "step": 17055
    },
    {
      "entropy": 6.037087726593017,
      "epoch": 1.9688401615695326,
      "grad_norm": 0.953125,
      "learning_rate": 0.000462082382143793,
      "loss": 5.8163,
      "mean_token_accuracy": 0.17590783536434174,
      "num_tokens": 43240620.0,
      "step": 17060
    },
    {
      "entropy": 6.0272666931152346,
      "epoch": 1.9694171956145412,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046205945422407113,
      "loss": 5.767,
      "mean_token_accuracy": 0.1855676144361496,
      "num_tokens": 43252523.0,
      "step": 17065
    },
    {
      "entropy": 6.040353631973266,
      "epoch": 1.9699942296595498,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046203652001269754,
      "loss": 5.8222,
      "mean_token_accuracy": 0.1859998255968094,
      "num_tokens": 43265731.0,
      "step": 17070
    },
    {
      "entropy": 5.921057271957397,
      "epoch": 1.9705712637045587,
      "grad_norm": 1.0,
      "learning_rate": 0.00046201357951044337,
      "loss": 5.6916,
      "mean_token_accuracy": 0.1887800082564354,
      "num_tokens": 43277929.0,
      "step": 17075
    },
    {
      "entropy": 6.057554292678833,
      "epoch": 1.9711482977495671,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046199063271808047,
      "loss": 5.835,
      "mean_token_accuracy": 0.18094127923250197,
      "num_tokens": 43290011.0,
      "step": 17080
    },
    {
      "entropy": 6.045626306533814,
      "epoch": 1.971725331794576,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004619676796363804,
      "loss": 5.8121,
      "mean_token_accuracy": 0.179393170773983,
      "num_tokens": 43303205.0,
      "step": 17085
    },
    {
      "entropy": 5.9444986343383786,
      "epoch": 1.9723023658395844,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046194472026611546,
      "loss": 5.6811,
      "mean_token_accuracy": 0.20005128979682923,
      "num_tokens": 43316214.0,
      "step": 17090
    },
    {
      "entropy": 5.974051284790039,
      "epoch": 1.9728793998845933,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004619217546080576,
      "loss": 5.7567,
      "mean_token_accuracy": 0.185201096534729,
      "num_tokens": 43328497.0,
      "step": 17095
    },
    {
      "entropy": 6.057856607437134,
      "epoch": 1.9734564339296017,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004618987826629794,
      "loss": 5.8399,
      "mean_token_accuracy": 0.17756510227918626,
      "num_tokens": 43341081.0,
      "step": 17100
    },
    {
      "entropy": 6.092422866821289,
      "epoch": 1.9740334679746105,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046187580443165344,
      "loss": 5.8931,
      "mean_token_accuracy": 0.17274920791387557,
      "num_tokens": 43354001.0,
      "step": 17105
    },
    {
      "entropy": 5.944503211975098,
      "epoch": 1.9746105020196192,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004618528199148527,
      "loss": 5.6655,
      "mean_token_accuracy": 0.19117288291454315,
      "num_tokens": 43368925.0,
      "step": 17110
    },
    {
      "entropy": 5.944300985336303,
      "epoch": 1.9751875360646278,
      "grad_norm": 0.90625,
      "learning_rate": 0.00046182982911335016,
      "loss": 5.7302,
      "mean_token_accuracy": 0.18855586498975754,
      "num_tokens": 43382063.0,
      "step": 17115
    },
    {
      "entropy": 6.076223564147949,
      "epoch": 1.9757645701096365,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046180683202791905,
      "loss": 5.7876,
      "mean_token_accuracy": 0.1801442325115204,
      "num_tokens": 43393810.0,
      "step": 17120
    },
    {
      "entropy": 6.0048305034637455,
      "epoch": 1.976341604154645,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000461783828659333,
      "loss": 5.8132,
      "mean_token_accuracy": 0.18633317053318024,
      "num_tokens": 43406500.0,
      "step": 17125
    },
    {
      "entropy": 5.996311855316162,
      "epoch": 1.9769186381996537,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046176081900836565,
      "loss": 5.8168,
      "mean_token_accuracy": 0.18064135909080506,
      "num_tokens": 43418284.0,
      "step": 17130
    },
    {
      "entropy": 6.042291641235352,
      "epoch": 1.9774956722446624,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046173780307579086,
      "loss": 5.7782,
      "mean_token_accuracy": 0.18679904490709304,
      "num_tokens": 43430913.0,
      "step": 17135
    },
    {
      "entropy": 5.990366888046265,
      "epoch": 1.9780727062896712,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004617147808623829,
      "loss": 5.7559,
      "mean_token_accuracy": 0.1905573472380638,
      "num_tokens": 43444637.0,
      "step": 17140
    },
    {
      "entropy": 6.0072136402130125,
      "epoch": 1.9786497403346797,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046169175236891605,
      "loss": 5.8125,
      "mean_token_accuracy": 0.1850295528769493,
      "num_tokens": 43457725.0,
      "step": 17145
    },
    {
      "entropy": 6.0557976245880125,
      "epoch": 1.9792267743796885,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004616687175961648,
      "loss": 5.8499,
      "mean_token_accuracy": 0.17723384201526643,
      "num_tokens": 43469902.0,
      "step": 17150
    },
    {
      "entropy": 6.0423469066619875,
      "epoch": 1.979803808424697,
      "grad_norm": 1.0,
      "learning_rate": 0.0004616456765449039,
      "loss": 5.8186,
      "mean_token_accuracy": 0.1741511657834053,
      "num_tokens": 43482959.0,
      "step": 17155
    },
    {
      "entropy": 6.0511149883270265,
      "epoch": 1.9803808424697058,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046162262921590845,
      "loss": 5.8421,
      "mean_token_accuracy": 0.18103216141462325,
      "num_tokens": 43496051.0,
      "step": 17160
    },
    {
      "entropy": 6.117519903182983,
      "epoch": 1.9809578765147142,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004615995756099535,
      "loss": 5.8427,
      "mean_token_accuracy": 0.1856980249285698,
      "num_tokens": 43509031.0,
      "step": 17165
    },
    {
      "entropy": 5.990596961975098,
      "epoch": 1.981534910559723,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004615765157278146,
      "loss": 5.7319,
      "mean_token_accuracy": 0.18005537241697311,
      "num_tokens": 43521577.0,
      "step": 17170
    },
    {
      "entropy": 6.058952760696411,
      "epoch": 1.9821119446047317,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046155344957026726,
      "loss": 5.8247,
      "mean_token_accuracy": 0.1800488770008087,
      "num_tokens": 43536029.0,
      "step": 17175
    },
    {
      "entropy": 5.995803785324097,
      "epoch": 1.9826889786497404,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004615303771380873,
      "loss": 5.761,
      "mean_token_accuracy": 0.17971590757369996,
      "num_tokens": 43548663.0,
      "step": 17180
    },
    {
      "entropy": 6.047090339660644,
      "epoch": 1.983266012694749,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046150729843205077,
      "loss": 5.8145,
      "mean_token_accuracy": 0.183537033200264,
      "num_tokens": 43560302.0,
      "step": 17185
    },
    {
      "entropy": 6.036820554733277,
      "epoch": 1.9838430467397576,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046148421345293384,
      "loss": 5.8094,
      "mean_token_accuracy": 0.18047595620155335,
      "num_tokens": 43572657.0,
      "step": 17190
    },
    {
      "entropy": 6.054653310775757,
      "epoch": 1.9844200807847663,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004614611222015131,
      "loss": 5.8704,
      "mean_token_accuracy": 0.18218297511339188,
      "num_tokens": 43584973.0,
      "step": 17195
    },
    {
      "entropy": 6.074311971664429,
      "epoch": 1.984997114829775,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046143802467856507,
      "loss": 5.8565,
      "mean_token_accuracy": 0.18148746192455292,
      "num_tokens": 43597457.0,
      "step": 17200
    },
    {
      "entropy": 6.067594575881958,
      "epoch": 1.9855741488747836,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046141492088486673,
      "loss": 5.7883,
      "mean_token_accuracy": 0.18180698454380034,
      "num_tokens": 43610256.0,
      "step": 17205
    },
    {
      "entropy": 6.083819627761841,
      "epoch": 1.9861511829197922,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004613918108211951,
      "loss": 5.8203,
      "mean_token_accuracy": 0.17717413157224654,
      "num_tokens": 43622765.0,
      "step": 17210
    },
    {
      "entropy": 6.064636087417602,
      "epoch": 1.986728216964801,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004613686944883275,
      "loss": 5.8012,
      "mean_token_accuracy": 0.17770864814519882,
      "num_tokens": 43634353.0,
      "step": 17215
    },
    {
      "entropy": 6.083303308486938,
      "epoch": 1.9873052510098095,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046134557188704146,
      "loss": 5.8523,
      "mean_token_accuracy": 0.17714426368474961,
      "num_tokens": 43647778.0,
      "step": 17220
    },
    {
      "entropy": 6.048518323898316,
      "epoch": 1.9878822850548183,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046132244301811466,
      "loss": 5.7874,
      "mean_token_accuracy": 0.1894538789987564,
      "num_tokens": 43659759.0,
      "step": 17225
    },
    {
      "entropy": 5.997383117675781,
      "epoch": 1.9884593190998268,
      "grad_norm": 0.84375,
      "learning_rate": 0.00046129930788232497,
      "loss": 5.717,
      "mean_token_accuracy": 0.18646375387907027,
      "num_tokens": 43673223.0,
      "step": 17230
    },
    {
      "entropy": 6.0200427055358885,
      "epoch": 1.9890363531448356,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046127616648045073,
      "loss": 5.8093,
      "mean_token_accuracy": 0.18413543105125427,
      "num_tokens": 43684657.0,
      "step": 17235
    },
    {
      "entropy": 6.091105175018311,
      "epoch": 1.989613387189844,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046125301881327007,
      "loss": 5.8956,
      "mean_token_accuracy": 0.1752494052052498,
      "num_tokens": 43697583.0,
      "step": 17240
    },
    {
      "entropy": 6.007848644256592,
      "epoch": 1.990190421234853,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046122986488156167,
      "loss": 5.7411,
      "mean_token_accuracy": 0.19191174656152726,
      "num_tokens": 43711175.0,
      "step": 17245
    },
    {
      "entropy": 6.001803207397461,
      "epoch": 1.9907674552798615,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046120670468610426,
      "loss": 5.7743,
      "mean_token_accuracy": 0.19252148568630217,
      "num_tokens": 43724595.0,
      "step": 17250
    },
    {
      "entropy": 6.059439468383789,
      "epoch": 1.9913444893248702,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046118353822767683,
      "loss": 5.7409,
      "mean_token_accuracy": 0.1942279502749443,
      "num_tokens": 43737192.0,
      "step": 17255
    },
    {
      "entropy": 5.987417221069336,
      "epoch": 1.9919215233698788,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004611603655070586,
      "loss": 5.7618,
      "mean_token_accuracy": 0.1917675405740738,
      "num_tokens": 43749543.0,
      "step": 17260
    },
    {
      "entropy": 5.96321587562561,
      "epoch": 1.9924985574148875,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046113718652502896,
      "loss": 5.8151,
      "mean_token_accuracy": 0.18014004826545715,
      "num_tokens": 43762754.0,
      "step": 17265
    },
    {
      "entropy": 6.102833604812622,
      "epoch": 1.993075591459896,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004611140012823675,
      "loss": 5.8083,
      "mean_token_accuracy": 0.18339821845293044,
      "num_tokens": 43774489.0,
      "step": 17270
    },
    {
      "entropy": 6.027917385101318,
      "epoch": 1.9936526255049047,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046109080977985395,
      "loss": 5.7901,
      "mean_token_accuracy": 0.18506036400794984,
      "num_tokens": 43786519.0,
      "step": 17275
    },
    {
      "entropy": 5.901705169677735,
      "epoch": 1.9942296595499136,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046106761201826854,
      "loss": 5.7573,
      "mean_token_accuracy": 0.18901724219322205,
      "num_tokens": 43800359.0,
      "step": 17280
    },
    {
      "entropy": 5.941084146499634,
      "epoch": 1.994806693594922,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046104440799839145,
      "loss": 5.7161,
      "mean_token_accuracy": 0.1894305720925331,
      "num_tokens": 43812130.0,
      "step": 17285
    },
    {
      "entropy": 5.980419301986695,
      "epoch": 1.9953837276399309,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000461021197721003,
      "loss": 5.7344,
      "mean_token_accuracy": 0.1925340101122856,
      "num_tokens": 43823544.0,
      "step": 17290
    },
    {
      "entropy": 6.10966010093689,
      "epoch": 1.9959607616849393,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046099798118688407,
      "loss": 5.8535,
      "mean_token_accuracy": 0.18141087591648103,
      "num_tokens": 43836283.0,
      "step": 17295
    },
    {
      "entropy": 6.020470905303955,
      "epoch": 1.9965377957299482,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004609747583968154,
      "loss": 5.8606,
      "mean_token_accuracy": 0.17749748528003692,
      "num_tokens": 43849077.0,
      "step": 17300
    },
    {
      "entropy": 6.082731485366821,
      "epoch": 1.9971148297749566,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004609515293515781,
      "loss": 5.8058,
      "mean_token_accuracy": 0.18178121298551558,
      "num_tokens": 43862805.0,
      "step": 17305
    },
    {
      "entropy": 6.121027183532715,
      "epoch": 1.9976918638199654,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004609282940519535,
      "loss": 5.8307,
      "mean_token_accuracy": 0.18025550693273545,
      "num_tokens": 43875420.0,
      "step": 17310
    },
    {
      "entropy": 6.0624958038330075,
      "epoch": 1.998268897864974,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004609050524987231,
      "loss": 5.8414,
      "mean_token_accuracy": 0.17818181961774826,
      "num_tokens": 43888445.0,
      "step": 17315
    },
    {
      "entropy": 6.056692981719971,
      "epoch": 1.9988459319099827,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004608818046926686,
      "loss": 5.8773,
      "mean_token_accuracy": 0.17983170002698898,
      "num_tokens": 43900943.0,
      "step": 17320
    },
    {
      "entropy": 6.016945123672485,
      "epoch": 1.9994229659549914,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004608585506345719,
      "loss": 5.7955,
      "mean_token_accuracy": 0.19324122369289398,
      "num_tokens": 43913951.0,
      "step": 17325
    },
    {
      "entropy": 6.043436336517334,
      "epoch": 2.0,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004608352903252152,
      "loss": 5.7908,
      "mean_token_accuracy": 0.1832709103822708,
      "num_tokens": 43926234.0,
      "step": 17330
    },
    {
      "entropy": 5.944561004638672,
      "epoch": 2.000577034045009,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046081202376538084,
      "loss": 5.6601,
      "mean_token_accuracy": 0.19420283734798433,
      "num_tokens": 43939519.0,
      "step": 17335
    },
    {
      "entropy": 6.012294578552246,
      "epoch": 2.0011540680900173,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004607887509558513,
      "loss": 5.675,
      "mean_token_accuracy": 0.19301000833511353,
      "num_tokens": 43953209.0,
      "step": 17340
    },
    {
      "entropy": 6.036614179611206,
      "epoch": 2.001731102135026,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004607654718974094,
      "loss": 5.6689,
      "mean_token_accuracy": 0.18590746968984603,
      "num_tokens": 43965578.0,
      "step": 17345
    },
    {
      "entropy": 5.992333459854126,
      "epoch": 2.0023081361800346,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004607421865908382,
      "loss": 5.6468,
      "mean_token_accuracy": 0.1917512908577919,
      "num_tokens": 43978508.0,
      "step": 17350
    },
    {
      "entropy": 6.0887082576751705,
      "epoch": 2.0028851702250434,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004607188950369207,
      "loss": 5.7872,
      "mean_token_accuracy": 0.18215877413749695,
      "num_tokens": 43992266.0,
      "step": 17355
    },
    {
      "entropy": 6.058491945266724,
      "epoch": 2.003462204270052,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004606955972364405,
      "loss": 5.7318,
      "mean_token_accuracy": 0.1816852256655693,
      "num_tokens": 44004603.0,
      "step": 17360
    },
    {
      "entropy": 6.044751453399658,
      "epoch": 2.0040392383150607,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004606722931901812,
      "loss": 5.6882,
      "mean_token_accuracy": 0.18352509737014772,
      "num_tokens": 44017578.0,
      "step": 17365
    },
    {
      "entropy": 6.038043212890625,
      "epoch": 2.004616272360069,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004606489828989264,
      "loss": 5.7312,
      "mean_token_accuracy": 0.19145991057157516,
      "num_tokens": 44031745.0,
      "step": 17370
    },
    {
      "entropy": 6.027770948410034,
      "epoch": 2.005193306405078,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004606256663634604,
      "loss": 5.7467,
      "mean_token_accuracy": 0.183679062128067,
      "num_tokens": 44043027.0,
      "step": 17375
    },
    {
      "entropy": 6.011777305603028,
      "epoch": 2.0057703404500864,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004606023435845672,
      "loss": 5.7082,
      "mean_token_accuracy": 0.18609317541122436,
      "num_tokens": 44056211.0,
      "step": 17380
    },
    {
      "entropy": 5.9904265880584715,
      "epoch": 2.0063473744950953,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004605790145630314,
      "loss": 5.5747,
      "mean_token_accuracy": 0.20142478942871095,
      "num_tokens": 44069861.0,
      "step": 17385
    },
    {
      "entropy": 6.036700201034546,
      "epoch": 2.0069244085401037,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004605556792996377,
      "loss": 5.7264,
      "mean_token_accuracy": 0.17982960045337676,
      "num_tokens": 44083373.0,
      "step": 17390
    },
    {
      "entropy": 5.936181974411011,
      "epoch": 2.0075014425851125,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004605323377951709,
      "loss": 5.6267,
      "mean_token_accuracy": 0.1944442257285118,
      "num_tokens": 44096138.0,
      "step": 17395
    },
    {
      "entropy": 5.976179599761963,
      "epoch": 2.008078476630121,
      "grad_norm": 0.921875,
      "learning_rate": 0.000460508990050416,
      "loss": 5.6717,
      "mean_token_accuracy": 0.18520279228687286,
      "num_tokens": 44107591.0,
      "step": 17400
    },
    {
      "entropy": 5.992496538162231,
      "epoch": 2.00865551067513,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004604856360661584,
      "loss": 5.6885,
      "mean_token_accuracy": 0.18762126863002776,
      "num_tokens": 44119733.0,
      "step": 17405
    },
    {
      "entropy": 6.1146995544433596,
      "epoch": 2.0092325447201387,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004604622758431835,
      "loss": 5.8237,
      "mean_token_accuracy": 0.1806732937693596,
      "num_tokens": 44131967.0,
      "step": 17410
    },
    {
      "entropy": 6.032386207580567,
      "epoch": 2.009809578765147,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046043890938227724,
      "loss": 5.7118,
      "mean_token_accuracy": 0.1906834825873375,
      "num_tokens": 44144426.0,
      "step": 17415
    },
    {
      "entropy": 5.990420150756836,
      "epoch": 2.010386612810156,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046041553668422526,
      "loss": 5.771,
      "mean_token_accuracy": 0.18210591971874238,
      "num_tokens": 44156298.0,
      "step": 17420
    },
    {
      "entropy": 5.965189409255982,
      "epoch": 2.0109636468551644,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046039215774981376,
      "loss": 5.6574,
      "mean_token_accuracy": 0.18526073098182677,
      "num_tokens": 44167955.0,
      "step": 17425
    },
    {
      "entropy": 6.01398777961731,
      "epoch": 2.0115406809001732,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046036877257982917,
      "loss": 5.6888,
      "mean_token_accuracy": 0.1865089401602745,
      "num_tokens": 44180555.0,
      "step": 17430
    },
    {
      "entropy": 5.925980043411255,
      "epoch": 2.0121177149451817,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000460345381175058,
      "loss": 5.6899,
      "mean_token_accuracy": 0.18701752424240112,
      "num_tokens": 44194335.0,
      "step": 17435
    },
    {
      "entropy": 6.087516736984253,
      "epoch": 2.0126947489901905,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000460321983536287,
      "loss": 5.7319,
      "mean_token_accuracy": 0.1843874216079712,
      "num_tokens": 44207258.0,
      "step": 17440
    },
    {
      "entropy": 6.081299448013306,
      "epoch": 2.013271783035199,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046029857966430315,
      "loss": 5.7506,
      "mean_token_accuracy": 0.18663651794195174,
      "num_tokens": 44220424.0,
      "step": 17445
    },
    {
      "entropy": 5.9247886657714846,
      "epoch": 2.013848817080208,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00046027516955989364,
      "loss": 5.6564,
      "mean_token_accuracy": 0.18876769691705703,
      "num_tokens": 44233394.0,
      "step": 17450
    },
    {
      "entropy": 6.065911197662354,
      "epoch": 2.014425851125216,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046025175322384577,
      "loss": 5.6912,
      "mean_token_accuracy": 0.1761830374598503,
      "num_tokens": 44246255.0,
      "step": 17455
    },
    {
      "entropy": 6.022132396697998,
      "epoch": 2.015002885170225,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046022833065694727,
      "loss": 5.6994,
      "mean_token_accuracy": 0.18506648987531663,
      "num_tokens": 44258315.0,
      "step": 17460
    },
    {
      "entropy": 6.0035851955413815,
      "epoch": 2.0155799192152335,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046020490185998575,
      "loss": 5.7825,
      "mean_token_accuracy": 0.18078363537788392,
      "num_tokens": 44271667.0,
      "step": 17465
    },
    {
      "entropy": 6.07629246711731,
      "epoch": 2.0161569532602424,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004601814668337495,
      "loss": 5.7255,
      "mean_token_accuracy": 0.1853678733110428,
      "num_tokens": 44284831.0,
      "step": 17470
    },
    {
      "entropy": 6.029196691513062,
      "epoch": 2.016733987305251,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046015802557902654,
      "loss": 5.7412,
      "mean_token_accuracy": 0.18108316212892533,
      "num_tokens": 44297161.0,
      "step": 17475
    },
    {
      "entropy": 6.049780941009521,
      "epoch": 2.0173110213502596,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046013457809660537,
      "loss": 5.6824,
      "mean_token_accuracy": 0.18761288672685622,
      "num_tokens": 44309537.0,
      "step": 17480
    },
    {
      "entropy": 5.9421937465667725,
      "epoch": 2.0178880553952685,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046011112438727454,
      "loss": 5.6847,
      "mean_token_accuracy": 0.1931302949786186,
      "num_tokens": 44320761.0,
      "step": 17485
    },
    {
      "entropy": 5.935798788070679,
      "epoch": 2.018465089440277,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004600876644518231,
      "loss": 5.5752,
      "mean_token_accuracy": 0.19838424026966095,
      "num_tokens": 44333323.0,
      "step": 17490
    },
    {
      "entropy": 5.908039951324463,
      "epoch": 2.0190421234852858,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046006419829104,
      "loss": 5.5205,
      "mean_token_accuracy": 0.203137344121933,
      "num_tokens": 44345919.0,
      "step": 17495
    },
    {
      "entropy": 6.042596769332886,
      "epoch": 2.019619157530294,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004600407259057145,
      "loss": 5.7686,
      "mean_token_accuracy": 0.18061307817697525,
      "num_tokens": 44359702.0,
      "step": 17500
    },
    {
      "entropy": 5.954426050186157,
      "epoch": 2.020196191575303,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004600172472966361,
      "loss": 5.6414,
      "mean_token_accuracy": 0.19359543025493622,
      "num_tokens": 44374186.0,
      "step": 17505
    },
    {
      "entropy": 6.008376693725586,
      "epoch": 2.0207732256203115,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045999376246459446,
      "loss": 5.7914,
      "mean_token_accuracy": 0.1805202320218086,
      "num_tokens": 44387870.0,
      "step": 17510
    },
    {
      "entropy": 6.013973808288574,
      "epoch": 2.0213502596653203,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004599702714103795,
      "loss": 5.7321,
      "mean_token_accuracy": 0.18433096557855605,
      "num_tokens": 44400706.0,
      "step": 17515
    },
    {
      "entropy": 6.073158931732178,
      "epoch": 2.0219272937103288,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004599467741347814,
      "loss": 5.709,
      "mean_token_accuracy": 0.1823631927371025,
      "num_tokens": 44413166.0,
      "step": 17520
    },
    {
      "entropy": 6.01109676361084,
      "epoch": 2.0225043277553376,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004599232706385904,
      "loss": 5.7045,
      "mean_token_accuracy": 0.19090095311403274,
      "num_tokens": 44425265.0,
      "step": 17525
    },
    {
      "entropy": 5.937154006958008,
      "epoch": 2.023081361800346,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000459899760922597,
      "loss": 5.6341,
      "mean_token_accuracy": 0.19117631912231445,
      "num_tokens": 44436990.0,
      "step": 17530
    },
    {
      "entropy": 6.032291603088379,
      "epoch": 2.023658395845355,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004598762449875921,
      "loss": 5.7001,
      "mean_token_accuracy": 0.18903365284204482,
      "num_tokens": 44450503.0,
      "step": 17535
    },
    {
      "entropy": 6.034704303741455,
      "epoch": 2.0242354298903633,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004598527228343665,
      "loss": 5.7416,
      "mean_token_accuracy": 0.17908364534378052,
      "num_tokens": 44463559.0,
      "step": 17540
    },
    {
      "entropy": 6.070844507217407,
      "epoch": 2.024812463935372,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004598291944637112,
      "loss": 5.7741,
      "mean_token_accuracy": 0.1829705372452736,
      "num_tokens": 44475064.0,
      "step": 17545
    },
    {
      "entropy": 6.036838340759277,
      "epoch": 2.025389497980381,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045980565987641797,
      "loss": 5.6826,
      "mean_token_accuracy": 0.18891073167324066,
      "num_tokens": 44487094.0,
      "step": 17550
    },
    {
      "entropy": 6.023197507858276,
      "epoch": 2.0259665320253895,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045978211907327804,
      "loss": 5.7047,
      "mean_token_accuracy": 0.18487124145030975,
      "num_tokens": 44498950.0,
      "step": 17555
    },
    {
      "entropy": 5.997314453125,
      "epoch": 2.0265435660703983,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004597585720550834,
      "loss": 5.726,
      "mean_token_accuracy": 0.18498716801404952,
      "num_tokens": 44511277.0,
      "step": 17560
    },
    {
      "entropy": 5.946838235855102,
      "epoch": 2.0271206001154067,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000459735018822626,
      "loss": 5.6723,
      "mean_token_accuracy": 0.18394143953919412,
      "num_tokens": 44525595.0,
      "step": 17565
    },
    {
      "entropy": 6.061409664154053,
      "epoch": 2.0276976341604156,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045971145937669794,
      "loss": 5.7369,
      "mean_token_accuracy": 0.18297409862279893,
      "num_tokens": 44537470.0,
      "step": 17570
    },
    {
      "entropy": 6.082262420654297,
      "epoch": 2.028274668205424,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004596878937180917,
      "loss": 5.7135,
      "mean_token_accuracy": 0.17975378632545472,
      "num_tokens": 44550594.0,
      "step": 17575
    },
    {
      "entropy": 6.03918080329895,
      "epoch": 2.028851702250433,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004596643218475999,
      "loss": 5.7204,
      "mean_token_accuracy": 0.18140195906162263,
      "num_tokens": 44563204.0,
      "step": 17580
    },
    {
      "entropy": 6.00481390953064,
      "epoch": 2.0294287362954413,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045964074376601534,
      "loss": 5.7097,
      "mean_token_accuracy": 0.1847568243741989,
      "num_tokens": 44577377.0,
      "step": 17585
    },
    {
      "entropy": 6.067275810241699,
      "epoch": 2.03000577034045,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045961715947413106,
      "loss": 5.7248,
      "mean_token_accuracy": 0.18429731577634811,
      "num_tokens": 44589658.0,
      "step": 17590
    },
    {
      "entropy": 5.971717500686646,
      "epoch": 2.0305828043854586,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004595935689727404,
      "loss": 5.7363,
      "mean_token_accuracy": 0.1904805526137352,
      "num_tokens": 44602477.0,
      "step": 17595
    },
    {
      "entropy": 6.041388893127442,
      "epoch": 2.0311598384304674,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004595699722626367,
      "loss": 5.6925,
      "mean_token_accuracy": 0.183915413916111,
      "num_tokens": 44613734.0,
      "step": 17600
    },
    {
      "entropy": 6.071640968322754,
      "epoch": 2.031736872475476,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00045954636934461367,
      "loss": 5.8225,
      "mean_token_accuracy": 0.1763857051730156,
      "num_tokens": 44626608.0,
      "step": 17605
    },
    {
      "entropy": 5.924884796142578,
      "epoch": 2.0323139065204847,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004595227602194652,
      "loss": 5.6219,
      "mean_token_accuracy": 0.20038360804319383,
      "num_tokens": 44640982.0,
      "step": 17610
    },
    {
      "entropy": 6.010447216033936,
      "epoch": 2.0328909405654936,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004594991448879853,
      "loss": 5.7418,
      "mean_token_accuracy": 0.182910718023777,
      "num_tokens": 44653546.0,
      "step": 17615
    },
    {
      "entropy": 5.965117120742798,
      "epoch": 2.033467974610502,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004594755233509683,
      "loss": 5.6459,
      "mean_token_accuracy": 0.1902913361787796,
      "num_tokens": 44667555.0,
      "step": 17620
    },
    {
      "entropy": 5.9174095630645756,
      "epoch": 2.034045008655511,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045945189560920875,
      "loss": 5.5877,
      "mean_token_accuracy": 0.1865845337510109,
      "num_tokens": 44680205.0,
      "step": 17625
    },
    {
      "entropy": 6.069635915756225,
      "epoch": 2.0346220427005193,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004594282616635013,
      "loss": 5.7189,
      "mean_token_accuracy": 0.18286188691854477,
      "num_tokens": 44692838.0,
      "step": 17630
    },
    {
      "entropy": 6.016884994506836,
      "epoch": 2.035199076745528,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004594046215146409,
      "loss": 5.7272,
      "mean_token_accuracy": 0.18433081358671188,
      "num_tokens": 44706010.0,
      "step": 17635
    },
    {
      "entropy": 5.979556655883789,
      "epoch": 2.0357761107905366,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045938097516342257,
      "loss": 5.6919,
      "mean_token_accuracy": 0.19550836831331253,
      "num_tokens": 44718857.0,
      "step": 17640
    },
    {
      "entropy": 6.070746612548828,
      "epoch": 2.0363531448355454,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045935732261064184,
      "loss": 5.7006,
      "mean_token_accuracy": 0.18868672400712966,
      "num_tokens": 44730876.0,
      "step": 17645
    },
    {
      "entropy": 5.930169486999512,
      "epoch": 2.036930178880554,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045933366385709405,
      "loss": 5.6456,
      "mean_token_accuracy": 0.18749533742666244,
      "num_tokens": 44743236.0,
      "step": 17650
    },
    {
      "entropy": 5.908342075347901,
      "epoch": 2.0375072129255627,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004593099989035751,
      "loss": 5.6284,
      "mean_token_accuracy": 0.19837864339351655,
      "num_tokens": 44755762.0,
      "step": 17655
    },
    {
      "entropy": 6.058293056488037,
      "epoch": 2.038084246970571,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004592863277508809,
      "loss": 5.7553,
      "mean_token_accuracy": 0.18033153414726258,
      "num_tokens": 44768439.0,
      "step": 17660
    },
    {
      "entropy": 6.039272308349609,
      "epoch": 2.03866128101558,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004592626503998076,
      "loss": 5.7117,
      "mean_token_accuracy": 0.18270668685436248,
      "num_tokens": 44781751.0,
      "step": 17665
    },
    {
      "entropy": 5.9539844512939455,
      "epoch": 2.0392383150605884,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004592389668511515,
      "loss": 5.5641,
      "mean_token_accuracy": 0.19032059460878373,
      "num_tokens": 44794221.0,
      "step": 17670
    },
    {
      "entropy": 5.984038639068603,
      "epoch": 2.0398153491055973,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004592152771057093,
      "loss": 5.7445,
      "mean_token_accuracy": 0.1776629090309143,
      "num_tokens": 44807027.0,
      "step": 17675
    },
    {
      "entropy": 6.010822868347168,
      "epoch": 2.0403923831506057,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045919158116427785,
      "loss": 5.7343,
      "mean_token_accuracy": 0.18576721996068954,
      "num_tokens": 44820399.0,
      "step": 17680
    },
    {
      "entropy": 6.0934093475341795,
      "epoch": 2.0409694171956145,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00045916787902765396,
      "loss": 5.8031,
      "mean_token_accuracy": 0.17448977530002593,
      "num_tokens": 44833179.0,
      "step": 17685
    },
    {
      "entropy": 5.952419471740723,
      "epoch": 2.0415464512406234,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004591441706966349,
      "loss": 5.6325,
      "mean_token_accuracy": 0.2013050362467766,
      "num_tokens": 44846169.0,
      "step": 17690
    },
    {
      "entropy": 5.882620573043823,
      "epoch": 2.042123485285632,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004591204561720183,
      "loss": 5.5874,
      "mean_token_accuracy": 0.1909885361790657,
      "num_tokens": 44858400.0,
      "step": 17695
    },
    {
      "entropy": 6.060600090026855,
      "epoch": 2.0427005193306407,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004590967354546015,
      "loss": 5.7841,
      "mean_token_accuracy": 0.18193771094083785,
      "num_tokens": 44870291.0,
      "step": 17700
    },
    {
      "entropy": 6.089450693130493,
      "epoch": 2.043277553375649,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004590730085451824,
      "loss": 5.7458,
      "mean_token_accuracy": 0.18323152810335158,
      "num_tokens": 44883893.0,
      "step": 17705
    },
    {
      "entropy": 6.003972864151001,
      "epoch": 2.043854587420658,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045904927544455914,
      "loss": 5.6487,
      "mean_token_accuracy": 0.18641297966241838,
      "num_tokens": 44896075.0,
      "step": 17710
    },
    {
      "entropy": 5.882717704772949,
      "epoch": 2.0444316214656664,
      "grad_norm": 1.0,
      "learning_rate": 0.00045902553615353005,
      "loss": 5.6307,
      "mean_token_accuracy": 0.18795052021741868,
      "num_tokens": 44908752.0,
      "step": 17715
    },
    {
      "entropy": 6.0706864356994625,
      "epoch": 2.0450086555106752,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004590017906728933,
      "loss": 5.7041,
      "mean_token_accuracy": 0.19492311775684357,
      "num_tokens": 44920782.0,
      "step": 17720
    },
    {
      "entropy": 5.986864280700684,
      "epoch": 2.0455856895556837,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045897803900344774,
      "loss": 5.6309,
      "mean_token_accuracy": 0.19004185050725936,
      "num_tokens": 44934565.0,
      "step": 17725
    },
    {
      "entropy": 5.917514801025391,
      "epoch": 2.0461627236006925,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004589542811459923,
      "loss": 5.6646,
      "mean_token_accuracy": 0.191831174492836,
      "num_tokens": 44948232.0,
      "step": 17730
    },
    {
      "entropy": 5.999948978424072,
      "epoch": 2.046739757645701,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004589305171013259,
      "loss": 5.6573,
      "mean_token_accuracy": 0.18444900512695311,
      "num_tokens": 44960092.0,
      "step": 17735
    },
    {
      "entropy": 6.058301639556885,
      "epoch": 2.04731679169071,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000458906746870248,
      "loss": 5.7767,
      "mean_token_accuracy": 0.1754949301481247,
      "num_tokens": 44972065.0,
      "step": 17740
    },
    {
      "entropy": 5.987555599212646,
      "epoch": 2.047893825735718,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000458882970453558,
      "loss": 5.7479,
      "mean_token_accuracy": 0.19060574173927308,
      "num_tokens": 44985142.0,
      "step": 17745
    },
    {
      "entropy": 6.050002479553223,
      "epoch": 2.048470859780727,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004588591878520556,
      "loss": 5.7763,
      "mean_token_accuracy": 0.18163920640945436,
      "num_tokens": 44998131.0,
      "step": 17750
    },
    {
      "entropy": 6.016087865829467,
      "epoch": 2.049047893825736,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004588353990665407,
      "loss": 5.6532,
      "mean_token_accuracy": 0.18914062529802322,
      "num_tokens": 45012715.0,
      "step": 17755
    },
    {
      "entropy": 6.039240407943725,
      "epoch": 2.0496249278707444,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004588116040978136,
      "loss": 5.7812,
      "mean_token_accuracy": 0.17659887224435805,
      "num_tokens": 45025214.0,
      "step": 17760
    },
    {
      "entropy": 6.0181725025177,
      "epoch": 2.050201961915753,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00045878780294667437,
      "loss": 5.6378,
      "mean_token_accuracy": 0.19634415209293365,
      "num_tokens": 45037956.0,
      "step": 17765
    },
    {
      "entropy": 5.9880119323730465,
      "epoch": 2.0507789959607616,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004587639956139237,
      "loss": 5.7719,
      "mean_token_accuracy": 0.18094383031129838,
      "num_tokens": 45052074.0,
      "step": 17770
    },
    {
      "entropy": 6.004917907714844,
      "epoch": 2.0513560300057705,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004587401821003624,
      "loss": 5.6961,
      "mean_token_accuracy": 0.1851331725716591,
      "num_tokens": 45064408.0,
      "step": 17775
    },
    {
      "entropy": 5.986521863937378,
      "epoch": 2.051933064050779,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045871636240679133,
      "loss": 5.651,
      "mean_token_accuracy": 0.1947542831301689,
      "num_tokens": 45077944.0,
      "step": 17780
    },
    {
      "entropy": 5.93346905708313,
      "epoch": 2.0525100980957878,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004586925365340117,
      "loss": 5.6456,
      "mean_token_accuracy": 0.1897893115878105,
      "num_tokens": 45090682.0,
      "step": 17785
    },
    {
      "entropy": 5.960285711288452,
      "epoch": 2.053087132140796,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004586687044828247,
      "loss": 5.6615,
      "mean_token_accuracy": 0.19371493458747863,
      "num_tokens": 45102567.0,
      "step": 17790
    },
    {
      "entropy": 5.96713342666626,
      "epoch": 2.053664166185805,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004586448662540322,
      "loss": 5.6469,
      "mean_token_accuracy": 0.19119673669338227,
      "num_tokens": 45114585.0,
      "step": 17795
    },
    {
      "entropy": 6.06060938835144,
      "epoch": 2.0542412002308135,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004586210218484358,
      "loss": 5.7782,
      "mean_token_accuracy": 0.17690440714359285,
      "num_tokens": 45128491.0,
      "step": 17800
    },
    {
      "entropy": 5.99797739982605,
      "epoch": 2.0548182342758223,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045859717126683745,
      "loss": 5.7573,
      "mean_token_accuracy": 0.17794754803180696,
      "num_tokens": 45141646.0,
      "step": 17805
    },
    {
      "entropy": 5.955329179763794,
      "epoch": 2.0553952683208307,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045857331451003953,
      "loss": 5.6144,
      "mean_token_accuracy": 0.19593816697597505,
      "num_tokens": 45153797.0,
      "step": 17810
    },
    {
      "entropy": 5.986756896972656,
      "epoch": 2.0559723023658396,
      "grad_norm": 0.90625,
      "learning_rate": 0.00045854945157884435,
      "loss": 5.6975,
      "mean_token_accuracy": 0.18754651099443437,
      "num_tokens": 45166626.0,
      "step": 17815
    },
    {
      "entropy": 6.003169393539428,
      "epoch": 2.056549336410848,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045852558247405455,
      "loss": 5.6933,
      "mean_token_accuracy": 0.1856832891702652,
      "num_tokens": 45181974.0,
      "step": 17820
    },
    {
      "entropy": 6.129568004608155,
      "epoch": 2.057126370455857,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045850170719647287,
      "loss": 5.7837,
      "mean_token_accuracy": 0.18432683795690535,
      "num_tokens": 45194406.0,
      "step": 17825
    },
    {
      "entropy": 5.976993465423584,
      "epoch": 2.0577034045008658,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045847782574690254,
      "loss": 5.6802,
      "mean_token_accuracy": 0.18917316496372222,
      "num_tokens": 45206777.0,
      "step": 17830
    },
    {
      "entropy": 5.952489852905273,
      "epoch": 2.058280438545874,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045845393812614664,
      "loss": 5.6625,
      "mean_token_accuracy": 0.19009070694446564,
      "num_tokens": 45219718.0,
      "step": 17835
    },
    {
      "entropy": 6.068571472167969,
      "epoch": 2.058857472590883,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004584300443350086,
      "loss": 5.7155,
      "mean_token_accuracy": 0.1864481419324875,
      "num_tokens": 45231038.0,
      "step": 17840
    },
    {
      "entropy": 5.983989000320435,
      "epoch": 2.0594345066358914,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004584061443742922,
      "loss": 5.7408,
      "mean_token_accuracy": 0.18815728276968002,
      "num_tokens": 45242835.0,
      "step": 17845
    },
    {
      "entropy": 5.979442071914673,
      "epoch": 2.0600115406809003,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045838223824480124,
      "loss": 5.7318,
      "mean_token_accuracy": 0.18548285514116286,
      "num_tokens": 45254857.0,
      "step": 17850
    },
    {
      "entropy": 5.973767375946045,
      "epoch": 2.0605885747259087,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004583583259473397,
      "loss": 5.6665,
      "mean_token_accuracy": 0.1925843760371208,
      "num_tokens": 45267897.0,
      "step": 17855
    },
    {
      "entropy": 6.006917285919189,
      "epoch": 2.0611656087709176,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045833440748271203,
      "loss": 5.7209,
      "mean_token_accuracy": 0.193971711397171,
      "num_tokens": 45280209.0,
      "step": 17860
    },
    {
      "entropy": 5.97929368019104,
      "epoch": 2.061742642815926,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045831048285172266,
      "loss": 5.6807,
      "mean_token_accuracy": 0.19400661587715148,
      "num_tokens": 45291834.0,
      "step": 17865
    },
    {
      "entropy": 6.022850370407104,
      "epoch": 2.062319676860935,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004582865520551762,
      "loss": 5.7798,
      "mean_token_accuracy": 0.18204474002122878,
      "num_tokens": 45305057.0,
      "step": 17870
    },
    {
      "entropy": 5.952797985076904,
      "epoch": 2.0628967109059433,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045826261509387755,
      "loss": 5.6619,
      "mean_token_accuracy": 0.18860250413417817,
      "num_tokens": 45318212.0,
      "step": 17875
    },
    {
      "entropy": 6.122391653060913,
      "epoch": 2.063473744950952,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045823867196863197,
      "loss": 5.7763,
      "mean_token_accuracy": 0.1813092678785324,
      "num_tokens": 45330981.0,
      "step": 17880
    },
    {
      "entropy": 5.9185404777526855,
      "epoch": 2.0640507789959606,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004582147226802446,
      "loss": 5.6212,
      "mean_token_accuracy": 0.1904293864965439,
      "num_tokens": 45344404.0,
      "step": 17885
    },
    {
      "entropy": 5.880917310714722,
      "epoch": 2.0646278130409694,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004581907672295211,
      "loss": 5.6443,
      "mean_token_accuracy": 0.19268029928207397,
      "num_tokens": 45357708.0,
      "step": 17890
    },
    {
      "entropy": 5.876305770874024,
      "epoch": 2.065204847085978,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045816680561726716,
      "loss": 5.5595,
      "mean_token_accuracy": 0.20058406591415406,
      "num_tokens": 45370625.0,
      "step": 17895
    },
    {
      "entropy": 6.032795143127442,
      "epoch": 2.0657818811309867,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004581428378442886,
      "loss": 5.6543,
      "mean_token_accuracy": 0.18789996802806855,
      "num_tokens": 45381864.0,
      "step": 17900
    },
    {
      "entropy": 6.0216062545776365,
      "epoch": 2.0663589151759956,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00045811886391139173,
      "loss": 5.718,
      "mean_token_accuracy": 0.18320301175117493,
      "num_tokens": 45394624.0,
      "step": 17905
    },
    {
      "entropy": 5.906604194641114,
      "epoch": 2.066935949221004,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045809488381938284,
      "loss": 5.5583,
      "mean_token_accuracy": 0.20106221437454225,
      "num_tokens": 45408551.0,
      "step": 17910
    },
    {
      "entropy": 6.019377279281616,
      "epoch": 2.067512983266013,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004580708975690684,
      "loss": 5.7446,
      "mean_token_accuracy": 0.1816095605492592,
      "num_tokens": 45420432.0,
      "step": 17915
    },
    {
      "entropy": 6.069661283493042,
      "epoch": 2.0680900173110213,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004580469051612554,
      "loss": 5.8592,
      "mean_token_accuracy": 0.17064070403575898,
      "num_tokens": 45432914.0,
      "step": 17920
    },
    {
      "entropy": 6.000480651855469,
      "epoch": 2.06866705135603,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045802290659675057,
      "loss": 5.6882,
      "mean_token_accuracy": 0.1879230722784996,
      "num_tokens": 45445501.0,
      "step": 17925
    },
    {
      "entropy": 5.89355354309082,
      "epoch": 2.0692440854010385,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045799890187636123,
      "loss": 5.598,
      "mean_token_accuracy": 0.1891181692481041,
      "num_tokens": 45458920.0,
      "step": 17930
    },
    {
      "entropy": 6.1179811477661135,
      "epoch": 2.0698211194460474,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045797489100089464,
      "loss": 5.8012,
      "mean_token_accuracy": 0.17863436192274093,
      "num_tokens": 45471438.0,
      "step": 17935
    },
    {
      "entropy": 5.991161108016968,
      "epoch": 2.070398153491056,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004579508739711585,
      "loss": 5.6757,
      "mean_token_accuracy": 0.1856082260608673,
      "num_tokens": 45483956.0,
      "step": 17940
    },
    {
      "entropy": 5.996302127838135,
      "epoch": 2.0709751875360647,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045792685078796075,
      "loss": 5.7626,
      "mean_token_accuracy": 0.1835334599018097,
      "num_tokens": 45496499.0,
      "step": 17945
    },
    {
      "entropy": 6.039648914337159,
      "epoch": 2.071552221581073,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000457902821452109,
      "loss": 5.676,
      "mean_token_accuracy": 0.19217363893985748,
      "num_tokens": 45509069.0,
      "step": 17950
    },
    {
      "entropy": 6.076765727996826,
      "epoch": 2.072129255626082,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045787878596441193,
      "loss": 5.8101,
      "mean_token_accuracy": 0.1798417642712593,
      "num_tokens": 45521707.0,
      "step": 17955
    },
    {
      "entropy": 5.94257607460022,
      "epoch": 2.0727062896710904,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004578547443256776,
      "loss": 5.6257,
      "mean_token_accuracy": 0.199801105260849,
      "num_tokens": 45533618.0,
      "step": 17960
    },
    {
      "entropy": 5.950657558441162,
      "epoch": 2.0732833237160992,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004578306965367148,
      "loss": 5.7307,
      "mean_token_accuracy": 0.18980734944343566,
      "num_tokens": 45546487.0,
      "step": 17965
    },
    {
      "entropy": 6.023340082168579,
      "epoch": 2.073860357761108,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045780664259833235,
      "loss": 5.7154,
      "mean_token_accuracy": 0.18379884511232375,
      "num_tokens": 45560258.0,
      "step": 17970
    },
    {
      "entropy": 6.073582553863526,
      "epoch": 2.0744373918061165,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045778258251133924,
      "loss": 5.7847,
      "mean_token_accuracy": 0.17456829249858857,
      "num_tokens": 45573152.0,
      "step": 17975
    },
    {
      "entropy": 6.138666439056396,
      "epoch": 2.0750144258511254,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045775851627654474,
      "loss": 5.7896,
      "mean_token_accuracy": 0.17493618577718734,
      "num_tokens": 45586042.0,
      "step": 17980
    },
    {
      "entropy": 6.065514945983887,
      "epoch": 2.075591459896134,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004577344438947584,
      "loss": 5.814,
      "mean_token_accuracy": 0.17758093476295472,
      "num_tokens": 45598761.0,
      "step": 17985
    },
    {
      "entropy": 5.920669412612915,
      "epoch": 2.0761684939411427,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045771036536678984,
      "loss": 5.6549,
      "mean_token_accuracy": 0.19194121807813644,
      "num_tokens": 45610988.0,
      "step": 17990
    },
    {
      "entropy": 5.957716417312622,
      "epoch": 2.076745527986151,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00045768628069344885,
      "loss": 5.6678,
      "mean_token_accuracy": 0.1952780455350876,
      "num_tokens": 45623689.0,
      "step": 17995
    },
    {
      "entropy": 6.0375429630279545,
      "epoch": 2.07732256203116,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004576621898755455,
      "loss": 5.7503,
      "mean_token_accuracy": 0.1843549683690071,
      "num_tokens": 45637283.0,
      "step": 18000
    },
    {
      "epoch": 2.07732256203116,
      "eval_entropy": 5.880664597392794,
      "eval_loss": 5.906169414520264,
      "eval_mean_token_accuracy": 0.1827783616943423,
      "eval_num_tokens": 45637283.0,
      "eval_runtime": 22.7366,
      "eval_samples_per_second": 1237.476,
      "eval_steps_per_second": 154.685,
      "step": 18000
    },
    {
      "entropy": 6.048853397369385,
      "epoch": 2.0778995960761684,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045763809291389024,
      "loss": 5.7626,
      "mean_token_accuracy": 0.1728976011276245,
      "num_tokens": 45651139.0,
      "step": 18005
    },
    {
      "entropy": 6.04798150062561,
      "epoch": 2.0784766301211772,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004576139898092933,
      "loss": 5.7295,
      "mean_token_accuracy": 0.18367347717285157,
      "num_tokens": 45663696.0,
      "step": 18010
    },
    {
      "entropy": 6.007404088973999,
      "epoch": 2.0790536641661856,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004575898805625657,
      "loss": 5.7379,
      "mean_token_accuracy": 0.1890614867210388,
      "num_tokens": 45676272.0,
      "step": 18015
    },
    {
      "entropy": 5.992618989944458,
      "epoch": 2.0796306982111945,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045756576517451804,
      "loss": 5.7247,
      "mean_token_accuracy": 0.18777452707290648,
      "num_tokens": 45688031.0,
      "step": 18020
    },
    {
      "entropy": 6.0242023944854735,
      "epoch": 2.080207732256203,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045754164364596156,
      "loss": 5.7889,
      "mean_token_accuracy": 0.17941420525312424,
      "num_tokens": 45700286.0,
      "step": 18025
    },
    {
      "entropy": 6.030111455917359,
      "epoch": 2.080784766301212,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004575175159777076,
      "loss": 5.7181,
      "mean_token_accuracy": 0.18589846938848495,
      "num_tokens": 45713935.0,
      "step": 18030
    },
    {
      "entropy": 5.9831019878387455,
      "epoch": 2.0813618003462206,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004574933821705676,
      "loss": 5.671,
      "mean_token_accuracy": 0.18654076606035233,
      "num_tokens": 45726815.0,
      "step": 18035
    },
    {
      "entropy": 5.960329055786133,
      "epoch": 2.081938834391229,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004574692422253534,
      "loss": 5.6469,
      "mean_token_accuracy": 0.191339348256588,
      "num_tokens": 45738727.0,
      "step": 18040
    },
    {
      "entropy": 5.982663822174072,
      "epoch": 2.082515868436238,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045744509614287687,
      "loss": 5.6904,
      "mean_token_accuracy": 0.18464579582214355,
      "num_tokens": 45751221.0,
      "step": 18045
    },
    {
      "entropy": 6.051972055435181,
      "epoch": 2.0830929024812463,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004574209439239501,
      "loss": 5.6674,
      "mean_token_accuracy": 0.18977078944444656,
      "num_tokens": 45764680.0,
      "step": 18050
    },
    {
      "entropy": 6.0074684619903564,
      "epoch": 2.083669936526255,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00045739678556938563,
      "loss": 5.6993,
      "mean_token_accuracy": 0.18339624404907226,
      "num_tokens": 45778324.0,
      "step": 18055
    },
    {
      "entropy": 5.919332647323609,
      "epoch": 2.0842469705712636,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045737262107999575,
      "loss": 5.5794,
      "mean_token_accuracy": 0.19093694686889648,
      "num_tokens": 45789412.0,
      "step": 18060
    },
    {
      "entropy": 6.023279809951783,
      "epoch": 2.0848240046162725,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004573484504565934,
      "loss": 5.82,
      "mean_token_accuracy": 0.17860755324363708,
      "num_tokens": 45802264.0,
      "step": 18065
    },
    {
      "entropy": 6.044946002960205,
      "epoch": 2.085401038661281,
      "grad_norm": 1.0,
      "learning_rate": 0.0004573242736999915,
      "loss": 5.6905,
      "mean_token_accuracy": 0.18967233449220658,
      "num_tokens": 45815411.0,
      "step": 18070
    },
    {
      "entropy": 6.046158504486084,
      "epoch": 2.0859780727062898,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045730009081100314,
      "loss": 5.801,
      "mean_token_accuracy": 0.1804548680782318,
      "num_tokens": 45828001.0,
      "step": 18075
    },
    {
      "entropy": 6.019017934799194,
      "epoch": 2.086555106751298,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045727590179044195,
      "loss": 5.7343,
      "mean_token_accuracy": 0.18603432476520537,
      "num_tokens": 45841001.0,
      "step": 18080
    },
    {
      "entropy": 6.050055885314942,
      "epoch": 2.087132140796307,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004572517066391211,
      "loss": 5.775,
      "mean_token_accuracy": 0.17991660684347152,
      "num_tokens": 45855160.0,
      "step": 18085
    },
    {
      "entropy": 5.897989940643311,
      "epoch": 2.0877091748413155,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045722750535785484,
      "loss": 5.56,
      "mean_token_accuracy": 0.1968957081437111,
      "num_tokens": 45867822.0,
      "step": 18090
    },
    {
      "entropy": 6.054330539703369,
      "epoch": 2.0882862088863243,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045720329794745685,
      "loss": 5.7573,
      "mean_token_accuracy": 0.17509395033121108,
      "num_tokens": 45878990.0,
      "step": 18095
    },
    {
      "entropy": 6.041588068008423,
      "epoch": 2.0888632429313327,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004571790844087415,
      "loss": 5.7092,
      "mean_token_accuracy": 0.1836043506860733,
      "num_tokens": 45891411.0,
      "step": 18100
    },
    {
      "entropy": 5.984649801254273,
      "epoch": 2.0894402769763416,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004571548647425231,
      "loss": 5.6511,
      "mean_token_accuracy": 0.19124047905206681,
      "num_tokens": 45903589.0,
      "step": 18105
    },
    {
      "entropy": 6.012734746932983,
      "epoch": 2.0900173110213505,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004571306389496164,
      "loss": 5.6685,
      "mean_token_accuracy": 0.18652251064777375,
      "num_tokens": 45915454.0,
      "step": 18110
    },
    {
      "entropy": 6.0199378490447994,
      "epoch": 2.090594345066359,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045710640703083594,
      "loss": 5.6831,
      "mean_token_accuracy": 0.18773744702339173,
      "num_tokens": 45928974.0,
      "step": 18115
    },
    {
      "entropy": 6.050688457489014,
      "epoch": 2.0911713791113677,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00045708216898699707,
      "loss": 5.7161,
      "mean_token_accuracy": 0.18538530617952348,
      "num_tokens": 45941040.0,
      "step": 18120
    },
    {
      "entropy": 6.020025777816772,
      "epoch": 2.091748413156376,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045705792481891483,
      "loss": 5.7435,
      "mean_token_accuracy": 0.18697226345539092,
      "num_tokens": 45954760.0,
      "step": 18125
    },
    {
      "entropy": 5.952204465866089,
      "epoch": 2.092325447201385,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00045703367452740477,
      "loss": 5.7838,
      "mean_token_accuracy": 0.186946402490139,
      "num_tokens": 45967735.0,
      "step": 18130
    },
    {
      "entropy": 6.086184453964234,
      "epoch": 2.0929024812463934,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045700941811328247,
      "loss": 5.8153,
      "mean_token_accuracy": 0.1794430807232857,
      "num_tokens": 45979318.0,
      "step": 18135
    },
    {
      "entropy": 5.933572816848755,
      "epoch": 2.0934795152914023,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045698515557736374,
      "loss": 5.6699,
      "mean_token_accuracy": 0.18824813663959503,
      "num_tokens": 45992487.0,
      "step": 18140
    },
    {
      "entropy": 6.065004158020019,
      "epoch": 2.0940565493364107,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045696088692046477,
      "loss": 5.7687,
      "mean_token_accuracy": 0.18396323770284653,
      "num_tokens": 46004103.0,
      "step": 18145
    },
    {
      "entropy": 6.078880262374878,
      "epoch": 2.0946335833814196,
      "grad_norm": 0.90625,
      "learning_rate": 0.00045693661214340174,
      "loss": 5.7471,
      "mean_token_accuracy": 0.18101374953985214,
      "num_tokens": 46016498.0,
      "step": 18150
    },
    {
      "entropy": 5.985243082046509,
      "epoch": 2.095210617426428,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045691233124699117,
      "loss": 5.6873,
      "mean_token_accuracy": 0.19274849444627762,
      "num_tokens": 46028179.0,
      "step": 18155
    },
    {
      "entropy": 5.9811958312988285,
      "epoch": 2.095787651471437,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004568880442320497,
      "loss": 5.6849,
      "mean_token_accuracy": 0.18147507160902024,
      "num_tokens": 46040152.0,
      "step": 18160
    },
    {
      "entropy": 6.0245216369628904,
      "epoch": 2.0963646855164453,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045686375109939417,
      "loss": 5.7051,
      "mean_token_accuracy": 0.18797954618930818,
      "num_tokens": 46052663.0,
      "step": 18165
    },
    {
      "entropy": 5.905424976348877,
      "epoch": 2.096941719561454,
      "grad_norm": 1.0,
      "learning_rate": 0.0004568394518498417,
      "loss": 5.6844,
      "mean_token_accuracy": 0.19021470099687576,
      "num_tokens": 46066793.0,
      "step": 18170
    },
    {
      "entropy": 6.023455667495727,
      "epoch": 2.0975187536064626,
      "grad_norm": 0.890625,
      "learning_rate": 0.00045681514648420965,
      "loss": 5.7662,
      "mean_token_accuracy": 0.18414924293756485,
      "num_tokens": 46080913.0,
      "step": 18175
    },
    {
      "entropy": 6.044397306442261,
      "epoch": 2.0980957876514714,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004567908350033154,
      "loss": 5.7383,
      "mean_token_accuracy": 0.1795857697725296,
      "num_tokens": 46093344.0,
      "step": 18180
    },
    {
      "entropy": 6.016541767120361,
      "epoch": 2.0986728216964803,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004567665174079767,
      "loss": 5.6626,
      "mean_token_accuracy": 0.18950953483581542,
      "num_tokens": 46106967.0,
      "step": 18185
    },
    {
      "entropy": 6.051772785186768,
      "epoch": 2.0992498557414887,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045674219369901153,
      "loss": 5.6912,
      "mean_token_accuracy": 0.19235837906599046,
      "num_tokens": 46118564.0,
      "step": 18190
    },
    {
      "entropy": 6.047540283203125,
      "epoch": 2.0998268897864976,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004567178638772379,
      "loss": 5.7849,
      "mean_token_accuracy": 0.17638496309518814,
      "num_tokens": 46130057.0,
      "step": 18195
    },
    {
      "entropy": 5.929427719116211,
      "epoch": 2.100403923831506,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004566935279434742,
      "loss": 5.6508,
      "mean_token_accuracy": 0.19615066796541214,
      "num_tokens": 46142730.0,
      "step": 18200
    },
    {
      "entropy": 6.0279358386993405,
      "epoch": 2.100980957876515,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045666918589853895,
      "loss": 5.6847,
      "mean_token_accuracy": 0.19306991696357728,
      "num_tokens": 46153669.0,
      "step": 18205
    },
    {
      "entropy": 6.019779062271118,
      "epoch": 2.1015579919215233,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045664483774325094,
      "loss": 5.7261,
      "mean_token_accuracy": 0.18208801597356797,
      "num_tokens": 46166472.0,
      "step": 18210
    },
    {
      "entropy": 6.041446495056152,
      "epoch": 2.102135025966532,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004566204834784289,
      "loss": 5.7199,
      "mean_token_accuracy": 0.18697401881217957,
      "num_tokens": 46178417.0,
      "step": 18215
    },
    {
      "entropy": 5.98527045249939,
      "epoch": 2.1027120600115405,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00045659612310489225,
      "loss": 5.6606,
      "mean_token_accuracy": 0.18842491507530212,
      "num_tokens": 46191537.0,
      "step": 18220
    },
    {
      "entropy": 5.943756437301635,
      "epoch": 2.1032890940565494,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00045657175662346014,
      "loss": 5.6394,
      "mean_token_accuracy": 0.19221796691417695,
      "num_tokens": 46205716.0,
      "step": 18225
    },
    {
      "entropy": 6.072571706771851,
      "epoch": 2.103866128101558,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004565473840349522,
      "loss": 5.7277,
      "mean_token_accuracy": 0.18538048267364501,
      "num_tokens": 46219050.0,
      "step": 18230
    },
    {
      "entropy": 6.055739259719848,
      "epoch": 2.1044431621465667,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045652300534018816,
      "loss": 5.7196,
      "mean_token_accuracy": 0.1873373046517372,
      "num_tokens": 46231045.0,
      "step": 18235
    },
    {
      "entropy": 6.1495918273925785,
      "epoch": 2.105020196191575,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000456498620539988,
      "loss": 5.8472,
      "mean_token_accuracy": 0.17814819663763046,
      "num_tokens": 46244629.0,
      "step": 18240
    },
    {
      "entropy": 6.071480655670166,
      "epoch": 2.105597230236584,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004564742296351719,
      "loss": 5.7191,
      "mean_token_accuracy": 0.1898820072412491,
      "num_tokens": 46256960.0,
      "step": 18245
    },
    {
      "entropy": 5.989773225784302,
      "epoch": 2.106174264281593,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045644983262656014,
      "loss": 5.682,
      "mean_token_accuracy": 0.19029029309749604,
      "num_tokens": 46268558.0,
      "step": 18250
    },
    {
      "entropy": 6.071680498123169,
      "epoch": 2.1067512983266012,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004564254295149735,
      "loss": 5.7338,
      "mean_token_accuracy": 0.1861613929271698,
      "num_tokens": 46281529.0,
      "step": 18255
    },
    {
      "entropy": 6.013559198379516,
      "epoch": 2.10732833237161,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045640102030123264,
      "loss": 5.6842,
      "mean_token_accuracy": 0.19009953439235688,
      "num_tokens": 46293598.0,
      "step": 18260
    },
    {
      "entropy": 5.992179727554321,
      "epoch": 2.1079053664166185,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045637660498615865,
      "loss": 5.6429,
      "mean_token_accuracy": 0.19381584525108336,
      "num_tokens": 46304252.0,
      "step": 18265
    },
    {
      "entropy": 6.03651270866394,
      "epoch": 2.1084824004616274,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004563521835705725,
      "loss": 5.728,
      "mean_token_accuracy": 0.19038084149360657,
      "num_tokens": 46316295.0,
      "step": 18270
    },
    {
      "entropy": 5.98615255355835,
      "epoch": 2.109059434506636,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045632775605529587,
      "loss": 5.6653,
      "mean_token_accuracy": 0.185970276594162,
      "num_tokens": 46328929.0,
      "step": 18275
    },
    {
      "entropy": 6.083420705795288,
      "epoch": 2.1096364685516447,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004563033224411501,
      "loss": 5.754,
      "mean_token_accuracy": 0.18342412561178206,
      "num_tokens": 46340784.0,
      "step": 18280
    },
    {
      "entropy": 6.021325492858887,
      "epoch": 2.110213502596653,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004562788827289572,
      "loss": 5.6767,
      "mean_token_accuracy": 0.18768522441387175,
      "num_tokens": 46354188.0,
      "step": 18285
    },
    {
      "entropy": 5.910429573059082,
      "epoch": 2.110790536641662,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045625443691953914,
      "loss": 5.5908,
      "mean_token_accuracy": 0.201395583152771,
      "num_tokens": 46366076.0,
      "step": 18290
    },
    {
      "entropy": 5.978281450271607,
      "epoch": 2.1113675706866704,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004562299850137181,
      "loss": 5.6086,
      "mean_token_accuracy": 0.19262748658657075,
      "num_tokens": 46376954.0,
      "step": 18295
    },
    {
      "entropy": 6.028190946578979,
      "epoch": 2.111944604731679,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004562055270123166,
      "loss": 5.6747,
      "mean_token_accuracy": 0.18833606392145158,
      "num_tokens": 46387951.0,
      "step": 18300
    },
    {
      "entropy": 6.029117107391357,
      "epoch": 2.1125216387766876,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045618106291615715,
      "loss": 5.752,
      "mean_token_accuracy": 0.1882779747247696,
      "num_tokens": 46399889.0,
      "step": 18305
    },
    {
      "entropy": 5.98864221572876,
      "epoch": 2.1130986728216965,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004561565927260627,
      "loss": 5.7426,
      "mean_token_accuracy": 0.18718918263912201,
      "num_tokens": 46412420.0,
      "step": 18310
    },
    {
      "entropy": 6.015537738800049,
      "epoch": 2.1136757068667054,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004561321164428561,
      "loss": 5.7239,
      "mean_token_accuracy": 0.18623706102371215,
      "num_tokens": 46424758.0,
      "step": 18315
    },
    {
      "entropy": 5.990384483337403,
      "epoch": 2.114252740911714,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004561076340673609,
      "loss": 5.7419,
      "mean_token_accuracy": 0.18392558842897416,
      "num_tokens": 46437163.0,
      "step": 18320
    },
    {
      "entropy": 6.024917411804199,
      "epoch": 2.1148297749567226,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004560831456004003,
      "loss": 5.6559,
      "mean_token_accuracy": 0.1888328969478607,
      "num_tokens": 46450780.0,
      "step": 18325
    },
    {
      "entropy": 6.0346321105957035,
      "epoch": 2.115406809001731,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004560586510427981,
      "loss": 5.7031,
      "mean_token_accuracy": 0.18672369718551635,
      "num_tokens": 46463879.0,
      "step": 18330
    },
    {
      "entropy": 5.9857580184936525,
      "epoch": 2.11598384304674,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004560341503953781,
      "loss": 5.7205,
      "mean_token_accuracy": 0.18481241017580033,
      "num_tokens": 46477624.0,
      "step": 18335
    },
    {
      "entropy": 6.035448980331421,
      "epoch": 2.1165608770917483,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004560096436589643,
      "loss": 5.6879,
      "mean_token_accuracy": 0.18533761352300643,
      "num_tokens": 46489745.0,
      "step": 18340
    },
    {
      "entropy": 6.011020278930664,
      "epoch": 2.117137911136757,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045598513083438115,
      "loss": 5.6673,
      "mean_token_accuracy": 0.19509464651346206,
      "num_tokens": 46501429.0,
      "step": 18345
    },
    {
      "entropy": 5.992949962615967,
      "epoch": 2.1177149451817656,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00045596061192245297,
      "loss": 5.7663,
      "mean_token_accuracy": 0.18114303499460221,
      "num_tokens": 46515094.0,
      "step": 18350
    },
    {
      "entropy": 5.979162168502808,
      "epoch": 2.1182919792267745,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004559360869240045,
      "loss": 5.6704,
      "mean_token_accuracy": 0.190386763215065,
      "num_tokens": 46527659.0,
      "step": 18355
    },
    {
      "entropy": 5.968454837799072,
      "epoch": 2.118869013271783,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004559115558398606,
      "loss": 5.6556,
      "mean_token_accuracy": 0.18969476968050003,
      "num_tokens": 46540190.0,
      "step": 18360
    },
    {
      "entropy": 6.08863410949707,
      "epoch": 2.1194460473167918,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004558870186708465,
      "loss": 5.7678,
      "mean_token_accuracy": 0.18254510164260865,
      "num_tokens": 46552349.0,
      "step": 18365
    },
    {
      "entropy": 6.06099944114685,
      "epoch": 2.1200230813618,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045586247541778724,
      "loss": 5.7548,
      "mean_token_accuracy": 0.18356144428253174,
      "num_tokens": 46564693.0,
      "step": 18370
    },
    {
      "entropy": 5.910883569717408,
      "epoch": 2.120600115406809,
      "grad_norm": 1.0,
      "learning_rate": 0.0004558379260815085,
      "loss": 5.6226,
      "mean_token_accuracy": 0.19525169134140014,
      "num_tokens": 46577035.0,
      "step": 18375
    },
    {
      "entropy": 5.981589365005493,
      "epoch": 2.1211771494518175,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004558133706628359,
      "loss": 5.6907,
      "mean_token_accuracy": 0.17863125652074813,
      "num_tokens": 46589670.0,
      "step": 18380
    },
    {
      "entropy": 6.092907190322876,
      "epoch": 2.1217541834968263,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045578880916259554,
      "loss": 5.7235,
      "mean_token_accuracy": 0.18174250274896622,
      "num_tokens": 46601320.0,
      "step": 18385
    },
    {
      "entropy": 5.990592622756958,
      "epoch": 2.122331217541835,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004557642415816132,
      "loss": 5.7166,
      "mean_token_accuracy": 0.1862447202205658,
      "num_tokens": 46615065.0,
      "step": 18390
    },
    {
      "entropy": 5.93634614944458,
      "epoch": 2.1229082515868436,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004557396679207155,
      "loss": 5.6462,
      "mean_token_accuracy": 0.1924818903207779,
      "num_tokens": 46628847.0,
      "step": 18395
    },
    {
      "entropy": 6.016889333724976,
      "epoch": 2.1234852856318525,
      "grad_norm": 1.0,
      "learning_rate": 0.00045571508818072887,
      "loss": 5.6745,
      "mean_token_accuracy": 0.18570567667484283,
      "num_tokens": 46642214.0,
      "step": 18400
    },
    {
      "entropy": 6.046145963668823,
      "epoch": 2.124062319676861,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004556905023624799,
      "loss": 5.6781,
      "mean_token_accuracy": 0.18970150947570802,
      "num_tokens": 46655920.0,
      "step": 18405
    },
    {
      "entropy": 6.018705272674561,
      "epoch": 2.1246393537218697,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045566591046679577,
      "loss": 5.7428,
      "mean_token_accuracy": 0.18459266126155854,
      "num_tokens": 46668006.0,
      "step": 18410
    },
    {
      "entropy": 5.948866176605224,
      "epoch": 2.125216387766878,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00045564131249450343,
      "loss": 5.6065,
      "mean_token_accuracy": 0.19558244198560715,
      "num_tokens": 46681523.0,
      "step": 18415
    },
    {
      "entropy": 6.077988481521606,
      "epoch": 2.125793421811887,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004556167084464302,
      "loss": 5.6924,
      "mean_token_accuracy": 0.1846438840031624,
      "num_tokens": 46693904.0,
      "step": 18420
    },
    {
      "entropy": 5.935725736618042,
      "epoch": 2.1263704558568954,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004555920983234038,
      "loss": 5.7082,
      "mean_token_accuracy": 0.18182436674833297,
      "num_tokens": 46705210.0,
      "step": 18425
    },
    {
      "entropy": 5.971320676803589,
      "epoch": 2.1269474899019043,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045556748212625183,
      "loss": 5.6906,
      "mean_token_accuracy": 0.18957730233669282,
      "num_tokens": 46717594.0,
      "step": 18430
    },
    {
      "entropy": 6.104868984222412,
      "epoch": 2.1275245239469127,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004555428598558023,
      "loss": 5.8119,
      "mean_token_accuracy": 0.17719220668077468,
      "num_tokens": 46730363.0,
      "step": 18435
    },
    {
      "entropy": 5.966627883911133,
      "epoch": 2.1281015579919216,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004555182315128833,
      "loss": 5.699,
      "mean_token_accuracy": 0.18805057257413865,
      "num_tokens": 46742554.0,
      "step": 18440
    },
    {
      "entropy": 6.058917570114136,
      "epoch": 2.12867859203693,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004554935970983234,
      "loss": 5.7467,
      "mean_token_accuracy": 0.19433538764715194,
      "num_tokens": 46754907.0,
      "step": 18445
    },
    {
      "entropy": 6.039881181716919,
      "epoch": 2.129255626081939,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004554689566129509,
      "loss": 5.7387,
      "mean_token_accuracy": 0.1805693820118904,
      "num_tokens": 46767463.0,
      "step": 18450
    },
    {
      "entropy": 6.056182479858398,
      "epoch": 2.1298326601269473,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045544431005759467,
      "loss": 5.7733,
      "mean_token_accuracy": 0.18174730390310287,
      "num_tokens": 46781523.0,
      "step": 18455
    },
    {
      "entropy": 6.061337852478028,
      "epoch": 2.130409694171956,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00045541965743308376,
      "loss": 5.7831,
      "mean_token_accuracy": 0.18946727067232133,
      "num_tokens": 46796557.0,
      "step": 18460
    },
    {
      "entropy": 6.052516317367553,
      "epoch": 2.130986728216965,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004553949987402473,
      "loss": 5.7166,
      "mean_token_accuracy": 0.18634486198425293,
      "num_tokens": 46809142.0,
      "step": 18465
    },
    {
      "entropy": 6.029892492294311,
      "epoch": 2.1315637622619734,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004553703339799146,
      "loss": 5.762,
      "mean_token_accuracy": 0.18468051999807358,
      "num_tokens": 46821654.0,
      "step": 18470
    },
    {
      "entropy": 5.980151510238647,
      "epoch": 2.1321407963069823,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004553456631529154,
      "loss": 5.6525,
      "mean_token_accuracy": 0.18921638429164886,
      "num_tokens": 46833262.0,
      "step": 18475
    },
    {
      "entropy": 6.056187486648559,
      "epoch": 2.1327178303519907,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004553209862600794,
      "loss": 5.7672,
      "mean_token_accuracy": 0.18176563382148742,
      "num_tokens": 46846194.0,
      "step": 18480
    },
    {
      "entropy": 5.948853492736816,
      "epoch": 2.1332948643969996,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004552963033022365,
      "loss": 5.6721,
      "mean_token_accuracy": 0.19187997430562972,
      "num_tokens": 46858395.0,
      "step": 18485
    },
    {
      "entropy": 6.062136125564575,
      "epoch": 2.133871898442008,
      "grad_norm": 0.890625,
      "learning_rate": 0.00045527161428021706,
      "loss": 5.7365,
      "mean_token_accuracy": 0.1916806474328041,
      "num_tokens": 46871609.0,
      "step": 18490
    },
    {
      "entropy": 6.057358694076538,
      "epoch": 2.134448932487017,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004552469191948514,
      "loss": 5.7611,
      "mean_token_accuracy": 0.17908270359039308,
      "num_tokens": 46883177.0,
      "step": 18495
    },
    {
      "entropy": 6.020691061019898,
      "epoch": 2.1350259665320253,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004552222180469702,
      "loss": 5.6163,
      "mean_token_accuracy": 0.1873266205191612,
      "num_tokens": 46895678.0,
      "step": 18500
    },
    {
      "entropy": 5.921075344085693,
      "epoch": 2.135603000577034,
      "grad_norm": 0.90625,
      "learning_rate": 0.00045519751083740413,
      "loss": 5.652,
      "mean_token_accuracy": 0.18498406410217286,
      "num_tokens": 46909041.0,
      "step": 18505
    },
    {
      "entropy": 6.021319961547851,
      "epoch": 2.1361800346220425,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045517279756698435,
      "loss": 5.6339,
      "mean_token_accuracy": 0.20026166141033172,
      "num_tokens": 46921386.0,
      "step": 18510
    },
    {
      "entropy": 5.978417682647705,
      "epoch": 2.1367570686670514,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00045514807823654203,
      "loss": 5.6612,
      "mean_token_accuracy": 0.19413836747407914,
      "num_tokens": 46933627.0,
      "step": 18515
    },
    {
      "entropy": 6.008830642700195,
      "epoch": 2.13733410271206,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004551233528469086,
      "loss": 5.7514,
      "mean_token_accuracy": 0.1833950772881508,
      "num_tokens": 46945108.0,
      "step": 18520
    },
    {
      "entropy": 6.016637849807739,
      "epoch": 2.1379111367570687,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045509862139891553,
      "loss": 5.6554,
      "mean_token_accuracy": 0.1937614068388939,
      "num_tokens": 46958366.0,
      "step": 18525
    },
    {
      "entropy": 5.990306997299195,
      "epoch": 2.138488170802077,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045507388389339496,
      "loss": 5.6764,
      "mean_token_accuracy": 0.18769063353538512,
      "num_tokens": 46972001.0,
      "step": 18530
    },
    {
      "entropy": 5.982241201400757,
      "epoch": 2.139065204847086,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045504914033117866,
      "loss": 5.6821,
      "mean_token_accuracy": 0.19092099368572235,
      "num_tokens": 46984365.0,
      "step": 18535
    },
    {
      "entropy": 5.969756412506103,
      "epoch": 2.139642238892095,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00045502439071309897,
      "loss": 5.616,
      "mean_token_accuracy": 0.1971512869000435,
      "num_tokens": 46996838.0,
      "step": 18540
    },
    {
      "entropy": 5.990099477767944,
      "epoch": 2.1402192729371032,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045499963503998835,
      "loss": 5.7622,
      "mean_token_accuracy": 0.18322264701128005,
      "num_tokens": 47008979.0,
      "step": 18545
    },
    {
      "entropy": 6.022986030578613,
      "epoch": 2.140796306982112,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004549748733126794,
      "loss": 5.7233,
      "mean_token_accuracy": 0.18511332273483277,
      "num_tokens": 47021092.0,
      "step": 18550
    },
    {
      "entropy": 5.912018156051635,
      "epoch": 2.1413733410271205,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000454950105532005,
      "loss": 5.6358,
      "mean_token_accuracy": 0.1909983053803444,
      "num_tokens": 47034277.0,
      "step": 18555
    },
    {
      "entropy": 5.9719085693359375,
      "epoch": 2.1419503750721294,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045492533169879816,
      "loss": 5.6757,
      "mean_token_accuracy": 0.18602865040302277,
      "num_tokens": 47046783.0,
      "step": 18560
    },
    {
      "entropy": 5.982133626937866,
      "epoch": 2.142527409117138,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045490055181389216,
      "loss": 5.7271,
      "mean_token_accuracy": 0.18602306842803956,
      "num_tokens": 47059358.0,
      "step": 18565
    },
    {
      "entropy": 6.072206306457519,
      "epoch": 2.1431044431621467,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045487576587812046,
      "loss": 5.6987,
      "mean_token_accuracy": 0.18129154443740844,
      "num_tokens": 47071492.0,
      "step": 18570
    },
    {
      "entropy": 6.018304967880249,
      "epoch": 2.143681477207155,
      "grad_norm": 1.0,
      "learning_rate": 0.00045485097389231675,
      "loss": 5.7175,
      "mean_token_accuracy": 0.1831903040409088,
      "num_tokens": 47083196.0,
      "step": 18575
    },
    {
      "entropy": 6.009262323379517,
      "epoch": 2.144258511252164,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004548261758573149,
      "loss": 5.7149,
      "mean_token_accuracy": 0.19636280685663224,
      "num_tokens": 47095193.0,
      "step": 18580
    },
    {
      "entropy": 6.066697072982788,
      "epoch": 2.1448355452971724,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004548013717739489,
      "loss": 5.7361,
      "mean_token_accuracy": 0.1875574991106987,
      "num_tokens": 47107751.0,
      "step": 18585
    },
    {
      "entropy": 5.972162294387817,
      "epoch": 2.145412579342181,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004547765616430531,
      "loss": 5.6953,
      "mean_token_accuracy": 0.19365291744470597,
      "num_tokens": 47121336.0,
      "step": 18590
    },
    {
      "entropy": 5.982786321640015,
      "epoch": 2.14598961338719,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004547517454654619,
      "loss": 5.6965,
      "mean_token_accuracy": 0.18653467446565627,
      "num_tokens": 47135565.0,
      "step": 18595
    },
    {
      "entropy": 6.0438196659088135,
      "epoch": 2.1465666474321985,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045472692324201005,
      "loss": 5.7388,
      "mean_token_accuracy": 0.18625867813825608,
      "num_tokens": 47147703.0,
      "step": 18600
    },
    {
      "entropy": 6.049138927459717,
      "epoch": 2.1471436814772074,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045470209497353243,
      "loss": 5.6677,
      "mean_token_accuracy": 0.1892435595393181,
      "num_tokens": 47159309.0,
      "step": 18605
    },
    {
      "entropy": 6.041308450698852,
      "epoch": 2.1477207155222158,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004546772606608641,
      "loss": 5.7549,
      "mean_token_accuracy": 0.1886858180165291,
      "num_tokens": 47171275.0,
      "step": 18610
    },
    {
      "entropy": 5.980893611907959,
      "epoch": 2.1482977495672246,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004546524203048404,
      "loss": 5.6816,
      "mean_token_accuracy": 0.19062912911176683,
      "num_tokens": 47183556.0,
      "step": 18615
    },
    {
      "entropy": 6.038990497589111,
      "epoch": 2.148874783612233,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004546275739062967,
      "loss": 5.678,
      "mean_token_accuracy": 0.1897209882736206,
      "num_tokens": 47196863.0,
      "step": 18620
    },
    {
      "entropy": 6.053529500961304,
      "epoch": 2.149451817657242,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004546027214660688,
      "loss": 5.7472,
      "mean_token_accuracy": 0.1814189523458481,
      "num_tokens": 47208717.0,
      "step": 18625
    },
    {
      "entropy": 5.9938740730285645,
      "epoch": 2.1500288517022503,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004545778629849926,
      "loss": 5.7277,
      "mean_token_accuracy": 0.18487774729728698,
      "num_tokens": 47222404.0,
      "step": 18630
    },
    {
      "entropy": 6.0173876762390135,
      "epoch": 2.150605885747259,
      "grad_norm": 1.0,
      "learning_rate": 0.0004545529984639042,
      "loss": 5.6906,
      "mean_token_accuracy": 0.18889817893505095,
      "num_tokens": 47235005.0,
      "step": 18635
    },
    {
      "entropy": 6.067893838882446,
      "epoch": 2.1511829197922676,
      "grad_norm": 0.875,
      "learning_rate": 0.0004545281279036398,
      "loss": 5.7468,
      "mean_token_accuracy": 0.18828802555799484,
      "num_tokens": 47248591.0,
      "step": 18640
    },
    {
      "entropy": 5.954255151748657,
      "epoch": 2.1517599538372765,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004545032513050361,
      "loss": 5.6362,
      "mean_token_accuracy": 0.18907804638147355,
      "num_tokens": 47261231.0,
      "step": 18645
    },
    {
      "entropy": 6.017440748214722,
      "epoch": 2.152336987882285,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004544783686689296,
      "loss": 5.6708,
      "mean_token_accuracy": 0.18650169521570206,
      "num_tokens": 47272891.0,
      "step": 18650
    },
    {
      "entropy": 6.073746967315674,
      "epoch": 2.1529140219272938,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045445347999615736,
      "loss": 5.7276,
      "mean_token_accuracy": 0.181544291973114,
      "num_tokens": 47284730.0,
      "step": 18655
    },
    {
      "entropy": 6.03571891784668,
      "epoch": 2.153491055972302,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045442858528755653,
      "loss": 5.6948,
      "mean_token_accuracy": 0.18449441492557525,
      "num_tokens": 47297485.0,
      "step": 18660
    },
    {
      "entropy": 5.920697164535523,
      "epoch": 2.154068090017311,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045440368454396435,
      "loss": 5.6279,
      "mean_token_accuracy": 0.18921328634023665,
      "num_tokens": 47310925.0,
      "step": 18665
    },
    {
      "entropy": 5.980650472640991,
      "epoch": 2.15464512406232,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004543787777662183,
      "loss": 5.6547,
      "mean_token_accuracy": 0.18732201009988786,
      "num_tokens": 47322645.0,
      "step": 18670
    },
    {
      "entropy": 5.988466262817383,
      "epoch": 2.1552221581073283,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045435386495515617,
      "loss": 5.7319,
      "mean_token_accuracy": 0.18790375292301179,
      "num_tokens": 47334582.0,
      "step": 18675
    },
    {
      "entropy": 5.931600666046142,
      "epoch": 2.155799192152337,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004543289461116159,
      "loss": 5.6069,
      "mean_token_accuracy": 0.19037822037935256,
      "num_tokens": 47347792.0,
      "step": 18680
    },
    {
      "entropy": 5.967264127731323,
      "epoch": 2.1563762261973456,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004543040212364356,
      "loss": 5.718,
      "mean_token_accuracy": 0.1824642539024353,
      "num_tokens": 47361064.0,
      "step": 18685
    },
    {
      "entropy": 6.032902336120605,
      "epoch": 2.1569532602423545,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004542790903304536,
      "loss": 5.7371,
      "mean_token_accuracy": 0.18738179951906203,
      "num_tokens": 47374329.0,
      "step": 18690
    },
    {
      "entropy": 6.02009391784668,
      "epoch": 2.157530294287363,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004542541533945085,
      "loss": 5.7427,
      "mean_token_accuracy": 0.18601662367582322,
      "num_tokens": 47387755.0,
      "step": 18695
    },
    {
      "entropy": 6.039984130859375,
      "epoch": 2.1581073283323717,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00045422921042943885,
      "loss": 5.7232,
      "mean_token_accuracy": 0.18195971846580505,
      "num_tokens": 47400970.0,
      "step": 18700
    },
    {
      "entropy": 5.886466693878174,
      "epoch": 2.15868436237738,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000454204261436084,
      "loss": 5.5827,
      "mean_token_accuracy": 0.1956717550754547,
      "num_tokens": 47414091.0,
      "step": 18705
    },
    {
      "entropy": 5.953662919998169,
      "epoch": 2.159261396422389,
      "grad_norm": 0.90625,
      "learning_rate": 0.00045417930641528255,
      "loss": 5.513,
      "mean_token_accuracy": 0.20684197396039963,
      "num_tokens": 47428485.0,
      "step": 18710
    },
    {
      "entropy": 5.981724786758423,
      "epoch": 2.1598384304673974,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045415434536787424,
      "loss": 5.7425,
      "mean_token_accuracy": 0.1869402199983597,
      "num_tokens": 47442474.0,
      "step": 18715
    },
    {
      "entropy": 5.990594577789307,
      "epoch": 2.1604154645124063,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004541293782946985,
      "loss": 5.75,
      "mean_token_accuracy": 0.1854146048426628,
      "num_tokens": 47456286.0,
      "step": 18720
    },
    {
      "entropy": 6.006946992874146,
      "epoch": 2.1609924985574147,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004541044051965952,
      "loss": 5.6479,
      "mean_token_accuracy": 0.19618288278579712,
      "num_tokens": 47468946.0,
      "step": 18725
    },
    {
      "entropy": 6.087666416168213,
      "epoch": 2.1615695326024236,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004540794260744041,
      "loss": 5.631,
      "mean_token_accuracy": 0.1918164849281311,
      "num_tokens": 47481877.0,
      "step": 18730
    },
    {
      "entropy": 5.9110908031463625,
      "epoch": 2.162146566647432,
      "grad_norm": 1.0,
      "learning_rate": 0.0004540544409289655,
      "loss": 5.6745,
      "mean_token_accuracy": 0.19598102569580078,
      "num_tokens": 47493564.0,
      "step": 18735
    },
    {
      "entropy": 5.9637696743011475,
      "epoch": 2.162723600692441,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004540294497611197,
      "loss": 5.7308,
      "mean_token_accuracy": 0.18277894258499144,
      "num_tokens": 47506565.0,
      "step": 18740
    },
    {
      "entropy": 5.995489311218262,
      "epoch": 2.1633006347374497,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045400445257170725,
      "loss": 5.6691,
      "mean_token_accuracy": 0.1910385712981224,
      "num_tokens": 47518384.0,
      "step": 18745
    },
    {
      "entropy": 5.948680210113525,
      "epoch": 2.163877668782458,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004539794493615689,
      "loss": 5.6605,
      "mean_token_accuracy": 0.1844969227910042,
      "num_tokens": 47530260.0,
      "step": 18750
    },
    {
      "entropy": 6.05945315361023,
      "epoch": 2.164454702827467,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004539544401315458,
      "loss": 5.7263,
      "mean_token_accuracy": 0.18397010266780853,
      "num_tokens": 47544651.0,
      "step": 18755
    },
    {
      "entropy": 6.0348375797271725,
      "epoch": 2.1650317368724754,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000453929424882479,
      "loss": 5.7302,
      "mean_token_accuracy": 0.18119997829198836,
      "num_tokens": 47556854.0,
      "step": 18760
    },
    {
      "entropy": 6.00861234664917,
      "epoch": 2.1656087709174843,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045390440361520987,
      "loss": 5.7477,
      "mean_token_accuracy": 0.18471066653728485,
      "num_tokens": 47569372.0,
      "step": 18765
    },
    {
      "entropy": 5.991002464294434,
      "epoch": 2.1661858049624927,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004538793763305799,
      "loss": 5.6717,
      "mean_token_accuracy": 0.18278762102127075,
      "num_tokens": 47581111.0,
      "step": 18770
    },
    {
      "entropy": 6.021338033676147,
      "epoch": 2.1667628390075016,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045385434302943104,
      "loss": 5.7187,
      "mean_token_accuracy": 0.18790800124406815,
      "num_tokens": 47593224.0,
      "step": 18775
    },
    {
      "entropy": 6.0478739738464355,
      "epoch": 2.16733987305251,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004538293037126052,
      "loss": 5.6447,
      "mean_token_accuracy": 0.19186590611934662,
      "num_tokens": 47606959.0,
      "step": 18780
    },
    {
      "entropy": 5.93533239364624,
      "epoch": 2.167916907097519,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045380425838094444,
      "loss": 5.6772,
      "mean_token_accuracy": 0.18490293622016907,
      "num_tokens": 47620159.0,
      "step": 18785
    },
    {
      "entropy": 6.028489589691162,
      "epoch": 2.1684939411425272,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045377920703529133,
      "loss": 5.7751,
      "mean_token_accuracy": 0.1816497728228569,
      "num_tokens": 47632325.0,
      "step": 18790
    },
    {
      "entropy": 6.01787371635437,
      "epoch": 2.169070975187536,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004537541496764885,
      "loss": 5.6626,
      "mean_token_accuracy": 0.1925953671336174,
      "num_tokens": 47644247.0,
      "step": 18795
    },
    {
      "entropy": 6.000169277191162,
      "epoch": 2.1696480092325445,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004537290863053786,
      "loss": 5.7174,
      "mean_token_accuracy": 0.1880468860268593,
      "num_tokens": 47655724.0,
      "step": 18800
    },
    {
      "entropy": 5.881081438064575,
      "epoch": 2.1702250432775534,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045370401692280455,
      "loss": 5.6351,
      "mean_token_accuracy": 0.18675459623336793,
      "num_tokens": 47667298.0,
      "step": 18805
    },
    {
      "entropy": 6.006354475021363,
      "epoch": 2.170802077322562,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045367894152960976,
      "loss": 5.6935,
      "mean_token_accuracy": 0.18990926444530487,
      "num_tokens": 47680393.0,
      "step": 18810
    },
    {
      "entropy": 5.931335401535034,
      "epoch": 2.1713791113675707,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045365386012663744,
      "loss": 5.6488,
      "mean_token_accuracy": 0.19712566286325456,
      "num_tokens": 47691766.0,
      "step": 18815
    },
    {
      "entropy": 6.016427230834961,
      "epoch": 2.1719561454125795,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00045362877271473137,
      "loss": 5.7433,
      "mean_token_accuracy": 0.18534431159496306,
      "num_tokens": 47704710.0,
      "step": 18820
    },
    {
      "entropy": 5.941805601119995,
      "epoch": 2.172533179457588,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045360367929473517,
      "loss": 5.6346,
      "mean_token_accuracy": 0.18922779858112335,
      "num_tokens": 47716869.0,
      "step": 18825
    },
    {
      "entropy": 5.957417392730713,
      "epoch": 2.173110213502597,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000453578579867493,
      "loss": 5.6393,
      "mean_token_accuracy": 0.18775125741958618,
      "num_tokens": 47729294.0,
      "step": 18830
    },
    {
      "entropy": 5.994666194915771,
      "epoch": 2.1736872475476052,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045355347443384896,
      "loss": 5.759,
      "mean_token_accuracy": 0.18355831056833266,
      "num_tokens": 47741130.0,
      "step": 18835
    },
    {
      "entropy": 5.985246372222901,
      "epoch": 2.174264281592614,
      "grad_norm": 0.875,
      "learning_rate": 0.00045352836299464755,
      "loss": 5.709,
      "mean_token_accuracy": 0.1882963478565216,
      "num_tokens": 47753337.0,
      "step": 18840
    },
    {
      "entropy": 5.988741159439087,
      "epoch": 2.1748413156376225,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004535032455507333,
      "loss": 5.6881,
      "mean_token_accuracy": 0.182587231695652,
      "num_tokens": 47766105.0,
      "step": 18845
    },
    {
      "entropy": 5.983407306671142,
      "epoch": 2.1754183496826314,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045347812210295107,
      "loss": 5.6548,
      "mean_token_accuracy": 0.1949616402387619,
      "num_tokens": 47777887.0,
      "step": 18850
    },
    {
      "entropy": 6.05370717048645,
      "epoch": 2.17599538372764,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00045345299265214583,
      "loss": 5.7394,
      "mean_token_accuracy": 0.18910082578659057,
      "num_tokens": 47789860.0,
      "step": 18855
    },
    {
      "entropy": 5.996746921539307,
      "epoch": 2.1765724177726486,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045342785719916284,
      "loss": 5.6845,
      "mean_token_accuracy": 0.18357667475938796,
      "num_tokens": 47801987.0,
      "step": 18860
    },
    {
      "entropy": 6.0607177257537845,
      "epoch": 2.177149451817657,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045340271574484755,
      "loss": 5.7633,
      "mean_token_accuracy": 0.18708713203668595,
      "num_tokens": 47814586.0,
      "step": 18865
    },
    {
      "entropy": 5.9876776218414305,
      "epoch": 2.177726485862666,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004533775682900454,
      "loss": 5.6444,
      "mean_token_accuracy": 0.18875792026519775,
      "num_tokens": 47827138.0,
      "step": 18870
    },
    {
      "entropy": 6.033636999130249,
      "epoch": 2.1783035199076743,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004533524148356025,
      "loss": 5.7691,
      "mean_token_accuracy": 0.1851331189274788,
      "num_tokens": 47839748.0,
      "step": 18875
    },
    {
      "entropy": 5.8488870620727536,
      "epoch": 2.178880553952683,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004533272553823646,
      "loss": 5.5206,
      "mean_token_accuracy": 0.1994289517402649,
      "num_tokens": 47852542.0,
      "step": 18880
    },
    {
      "entropy": 6.017760801315307,
      "epoch": 2.179457587997692,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045330208993117816,
      "loss": 5.7094,
      "mean_token_accuracy": 0.18834460973739625,
      "num_tokens": 47866192.0,
      "step": 18885
    },
    {
      "entropy": 5.927276134490967,
      "epoch": 2.1800346220427005,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004532769184828894,
      "loss": 5.6169,
      "mean_token_accuracy": 0.20162469893693924,
      "num_tokens": 47878164.0,
      "step": 18890
    },
    {
      "entropy": 5.946986198425293,
      "epoch": 2.1806116560877093,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004532517410383451,
      "loss": 5.6731,
      "mean_token_accuracy": 0.18163194954395295,
      "num_tokens": 47890040.0,
      "step": 18895
    },
    {
      "entropy": 6.00893497467041,
      "epoch": 2.1811886901327178,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000453226557598392,
      "loss": 5.6857,
      "mean_token_accuracy": 0.18261783868074416,
      "num_tokens": 47902949.0,
      "step": 18900
    },
    {
      "entropy": 6.046803665161133,
      "epoch": 2.1817657241777266,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004532013681638771,
      "loss": 5.7141,
      "mean_token_accuracy": 0.18881515562534332,
      "num_tokens": 47915674.0,
      "step": 18905
    },
    {
      "entropy": 5.987967014312744,
      "epoch": 2.182342758222735,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004531761727356478,
      "loss": 5.6142,
      "mean_token_accuracy": 0.19461592435836791,
      "num_tokens": 47929358.0,
      "step": 18910
    },
    {
      "entropy": 6.043885946273804,
      "epoch": 2.182919792267744,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004531509713145514,
      "loss": 5.7407,
      "mean_token_accuracy": 0.18649691492319107,
      "num_tokens": 47940790.0,
      "step": 18915
    },
    {
      "entropy": 6.031921625137329,
      "epoch": 2.1834968263127523,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045312576390143557,
      "loss": 5.6977,
      "mean_token_accuracy": 0.18730615973472595,
      "num_tokens": 47952651.0,
      "step": 18920
    },
    {
      "entropy": 6.064148426055908,
      "epoch": 2.184073860357761,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045310055049714815,
      "loss": 5.7214,
      "mean_token_accuracy": 0.18128344267606736,
      "num_tokens": 47964378.0,
      "step": 18925
    },
    {
      "entropy": 6.103727245330811,
      "epoch": 2.1846508944027696,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045307533110253726,
      "loss": 5.8113,
      "mean_token_accuracy": 0.18211405724287033,
      "num_tokens": 47976724.0,
      "step": 18930
    },
    {
      "entropy": 5.980295181274414,
      "epoch": 2.1852279284477785,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00045305010571845096,
      "loss": 5.7173,
      "mean_token_accuracy": 0.18438691049814224,
      "num_tokens": 47989364.0,
      "step": 18935
    },
    {
      "entropy": 6.006443166732788,
      "epoch": 2.185804962492787,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004530248743457378,
      "loss": 5.7076,
      "mean_token_accuracy": 0.19357595443725586,
      "num_tokens": 48001845.0,
      "step": 18940
    },
    {
      "entropy": 6.034667491912842,
      "epoch": 2.1863819965377957,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004529996369852465,
      "loss": 5.7699,
      "mean_token_accuracy": 0.18304516971111298,
      "num_tokens": 48016789.0,
      "step": 18945
    },
    {
      "entropy": 6.085127162933349,
      "epoch": 2.1869590305828046,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045297439363782576,
      "loss": 5.7774,
      "mean_token_accuracy": 0.17666388750076295,
      "num_tokens": 48029016.0,
      "step": 18950
    },
    {
      "entropy": 5.879226398468018,
      "epoch": 2.187536064627813,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004529491443043247,
      "loss": 5.6099,
      "mean_token_accuracy": 0.19980298429727555,
      "num_tokens": 48041998.0,
      "step": 18955
    },
    {
      "entropy": 6.042237138748169,
      "epoch": 2.188113098672822,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004529238889855926,
      "loss": 5.7602,
      "mean_token_accuracy": 0.18491050899028777,
      "num_tokens": 48054716.0,
      "step": 18960
    },
    {
      "entropy": 6.007125473022461,
      "epoch": 2.1886901327178303,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004528986276824789,
      "loss": 5.6549,
      "mean_token_accuracy": 0.1938191443681717,
      "num_tokens": 48068444.0,
      "step": 18965
    },
    {
      "entropy": 6.032671022415161,
      "epoch": 2.189267166762839,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004528733603958331,
      "loss": 5.6715,
      "mean_token_accuracy": 0.19485439509153366,
      "num_tokens": 48081494.0,
      "step": 18970
    },
    {
      "entropy": 6.000243091583252,
      "epoch": 2.1898442008078476,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004528480871265053,
      "loss": 5.6855,
      "mean_token_accuracy": 0.18609120398759843,
      "num_tokens": 48094092.0,
      "step": 18975
    },
    {
      "entropy": 5.895876073837281,
      "epoch": 2.1904212348528564,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045282280787534537,
      "loss": 5.6004,
      "mean_token_accuracy": 0.19726919680833815,
      "num_tokens": 48106061.0,
      "step": 18980
    },
    {
      "entropy": 5.93585057258606,
      "epoch": 2.190998268897865,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004527975226432036,
      "loss": 5.6449,
      "mean_token_accuracy": 0.1912421390414238,
      "num_tokens": 48118686.0,
      "step": 18985
    },
    {
      "entropy": 6.064427137374878,
      "epoch": 2.1915753029428737,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045277223143093057,
      "loss": 5.692,
      "mean_token_accuracy": 0.1886579677462578,
      "num_tokens": 48130946.0,
      "step": 18990
    },
    {
      "entropy": 5.954942274093628,
      "epoch": 2.192152336987882,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045274693423937674,
      "loss": 5.5821,
      "mean_token_accuracy": 0.20205695629119874,
      "num_tokens": 48143923.0,
      "step": 18995
    },
    {
      "entropy": 6.0250465869903564,
      "epoch": 2.192729371032891,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045272163106939314,
      "loss": 5.7507,
      "mean_token_accuracy": 0.17731713205575944,
      "num_tokens": 48156089.0,
      "step": 19000
    },
    {
      "entropy": 5.988602018356323,
      "epoch": 2.1933064050778994,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045269632192183076,
      "loss": 5.6302,
      "mean_token_accuracy": 0.19669615775346755,
      "num_tokens": 48169930.0,
      "step": 19005
    },
    {
      "entropy": 6.07318081855774,
      "epoch": 2.1938834391229083,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045267100679754075,
      "loss": 5.7009,
      "mean_token_accuracy": 0.18296739012002944,
      "num_tokens": 48181378.0,
      "step": 19010
    },
    {
      "entropy": 5.940174007415772,
      "epoch": 2.1944604731679167,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004526456856973748,
      "loss": 5.7242,
      "mean_token_accuracy": 0.1859443485736847,
      "num_tokens": 48193141.0,
      "step": 19015
    },
    {
      "entropy": 5.942185926437378,
      "epoch": 2.1950375072129256,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004526203586221844,
      "loss": 5.5933,
      "mean_token_accuracy": 0.2029839873313904,
      "num_tokens": 48206817.0,
      "step": 19020
    },
    {
      "entropy": 6.041692018508911,
      "epoch": 2.1956145412579344,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045259502557282145,
      "loss": 5.7028,
      "mean_token_accuracy": 0.18940547555685044,
      "num_tokens": 48220104.0,
      "step": 19025
    },
    {
      "entropy": 6.0428550243377686,
      "epoch": 2.196191575302943,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004525696865501381,
      "loss": 5.7299,
      "mean_token_accuracy": 0.18368273377418518,
      "num_tokens": 48232446.0,
      "step": 19030
    },
    {
      "entropy": 5.957457208633423,
      "epoch": 2.1967686093479517,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004525443415549865,
      "loss": 5.731,
      "mean_token_accuracy": 0.1855126917362213,
      "num_tokens": 48244071.0,
      "step": 19035
    },
    {
      "entropy": 6.04968228340149,
      "epoch": 2.19734564339296,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00045251899058821915,
      "loss": 5.6989,
      "mean_token_accuracy": 0.18192910104990007,
      "num_tokens": 48256808.0,
      "step": 19040
    },
    {
      "entropy": 6.048809766769409,
      "epoch": 2.197922677437969,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004524936336506887,
      "loss": 5.6177,
      "mean_token_accuracy": 0.19199782013893127,
      "num_tokens": 48269635.0,
      "step": 19045
    },
    {
      "entropy": 5.938378047943115,
      "epoch": 2.1984997114829774,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004524682707432482,
      "loss": 5.7029,
      "mean_token_accuracy": 0.18706714510917663,
      "num_tokens": 48282211.0,
      "step": 19050
    },
    {
      "entropy": 6.009896087646484,
      "epoch": 2.1990767455279863,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045244290186675034,
      "loss": 5.7859,
      "mean_token_accuracy": 0.17765019685029984,
      "num_tokens": 48295617.0,
      "step": 19055
    },
    {
      "entropy": 6.1145110607147215,
      "epoch": 2.1996537795729947,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004524175270220489,
      "loss": 5.7344,
      "mean_token_accuracy": 0.1834876909852028,
      "num_tokens": 48307885.0,
      "step": 19060
    },
    {
      "entropy": 6.018986368179322,
      "epoch": 2.2002308136180035,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045239214620999683,
      "loss": 5.6633,
      "mean_token_accuracy": 0.19506632387638093,
      "num_tokens": 48320451.0,
      "step": 19065
    },
    {
      "entropy": 5.958942794799805,
      "epoch": 2.200807847663012,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004523667594314481,
      "loss": 5.6336,
      "mean_token_accuracy": 0.1905920147895813,
      "num_tokens": 48333283.0,
      "step": 19070
    },
    {
      "entropy": 6.1205836772918705,
      "epoch": 2.201384881708021,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00045234136668725655,
      "loss": 5.776,
      "mean_token_accuracy": 0.18089927732944489,
      "num_tokens": 48347197.0,
      "step": 19075
    },
    {
      "entropy": 6.064835500717163,
      "epoch": 2.2019619157530292,
      "grad_norm": 1.0,
      "learning_rate": 0.00045231596797827616,
      "loss": 5.796,
      "mean_token_accuracy": 0.1805758461356163,
      "num_tokens": 48359356.0,
      "step": 19080
    },
    {
      "entropy": 5.995683574676514,
      "epoch": 2.202538949798038,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045229056330536134,
      "loss": 5.7118,
      "mean_token_accuracy": 0.1867974057793617,
      "num_tokens": 48373343.0,
      "step": 19085
    },
    {
      "entropy": 5.984430170059204,
      "epoch": 2.2031159838430465,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045226515266936644,
      "loss": 5.6259,
      "mean_token_accuracy": 0.1912284106016159,
      "num_tokens": 48385614.0,
      "step": 19090
    },
    {
      "entropy": 5.9917560577392575,
      "epoch": 2.2036930178880554,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004522397360711462,
      "loss": 5.6742,
      "mean_token_accuracy": 0.18881919533014296,
      "num_tokens": 48397814.0,
      "step": 19095
    },
    {
      "entropy": 5.982270002365112,
      "epoch": 2.2042700519330642,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004522143135115555,
      "loss": 5.6831,
      "mean_token_accuracy": 0.19666559100151063,
      "num_tokens": 48411092.0,
      "step": 19100
    },
    {
      "entropy": 6.036627578735351,
      "epoch": 2.2048470859780727,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045218888499144933,
      "loss": 5.6948,
      "mean_token_accuracy": 0.18556898981332778,
      "num_tokens": 48423203.0,
      "step": 19105
    },
    {
      "entropy": 5.969345045089722,
      "epoch": 2.2054241200230815,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045216345051168314,
      "loss": 5.6899,
      "mean_token_accuracy": 0.19395052045583724,
      "num_tokens": 48435562.0,
      "step": 19110
    },
    {
      "entropy": 5.94818754196167,
      "epoch": 2.20600115406809,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004521380100731122,
      "loss": 5.6814,
      "mean_token_accuracy": 0.18750114887952804,
      "num_tokens": 48449021.0,
      "step": 19115
    },
    {
      "entropy": 6.020796632766723,
      "epoch": 2.206578188113099,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00045211256367659234,
      "loss": 5.6428,
      "mean_token_accuracy": 0.19194164872169495,
      "num_tokens": 48462411.0,
      "step": 19120
    },
    {
      "entropy": 6.020190668106079,
      "epoch": 2.207155222158107,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004520871113229793,
      "loss": 5.7068,
      "mean_token_accuracy": 0.18544422537088395,
      "num_tokens": 48474397.0,
      "step": 19125
    },
    {
      "entropy": 6.091891050338745,
      "epoch": 2.207732256203116,
      "grad_norm": 0.81640625,
      "learning_rate": 0.00045206165301312927,
      "loss": 5.7779,
      "mean_token_accuracy": 0.1839219808578491,
      "num_tokens": 48490549.0,
      "step": 19130
    },
    {
      "entropy": 6.1061601638793945,
      "epoch": 2.2083092902481245,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004520361887478985,
      "loss": 5.7257,
      "mean_token_accuracy": 0.18886110186576843,
      "num_tokens": 48503692.0,
      "step": 19135
    },
    {
      "entropy": 5.94710521697998,
      "epoch": 2.2088863242931334,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004520107185281435,
      "loss": 5.6502,
      "mean_token_accuracy": 0.19022322148084642,
      "num_tokens": 48515574.0,
      "step": 19140
    },
    {
      "entropy": 6.031595993041992,
      "epoch": 2.209463358338142,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004519852423547208,
      "loss": 5.7655,
      "mean_token_accuracy": 0.18012748956680297,
      "num_tokens": 48527832.0,
      "step": 19145
    },
    {
      "entropy": 6.054387950897217,
      "epoch": 2.2100403923831506,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004519597602284875,
      "loss": 5.6824,
      "mean_token_accuracy": 0.18968423157930375,
      "num_tokens": 48539558.0,
      "step": 19150
    },
    {
      "entropy": 5.963400173187256,
      "epoch": 2.210617426428159,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004519342721503005,
      "loss": 5.6587,
      "mean_token_accuracy": 0.19415226429700852,
      "num_tokens": 48553360.0,
      "step": 19155
    },
    {
      "entropy": 6.017910957336426,
      "epoch": 2.211194460473168,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004519087781210171,
      "loss": 5.7187,
      "mean_token_accuracy": 0.17883104234933853,
      "num_tokens": 48567416.0,
      "step": 19160
    },
    {
      "entropy": 6.020536804199219,
      "epoch": 2.2117714945181763,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000451883278141495,
      "loss": 5.7026,
      "mean_token_accuracy": 0.19276584088802337,
      "num_tokens": 48580086.0,
      "step": 19165
    },
    {
      "entropy": 5.951890087127685,
      "epoch": 2.212348528563185,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045185777221259157,
      "loss": 5.5901,
      "mean_token_accuracy": 0.2003543496131897,
      "num_tokens": 48591758.0,
      "step": 19170
    },
    {
      "entropy": 5.839439344406128,
      "epoch": 2.212925562608194,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004518322603351648,
      "loss": 5.5364,
      "mean_token_accuracy": 0.20399542152881622,
      "num_tokens": 48604994.0,
      "step": 19175
    },
    {
      "entropy": 6.024185752868652,
      "epoch": 2.2135025966532025,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045180674251007287,
      "loss": 5.7714,
      "mean_token_accuracy": 0.1824466586112976,
      "num_tokens": 48617652.0,
      "step": 19180
    },
    {
      "entropy": 5.9852907180786135,
      "epoch": 2.2140796306982113,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045178121873817395,
      "loss": 5.6141,
      "mean_token_accuracy": 0.19718071669340134,
      "num_tokens": 48629696.0,
      "step": 19185
    },
    {
      "entropy": 5.956853008270263,
      "epoch": 2.2146566647432198,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004517556890203265,
      "loss": 5.702,
      "mean_token_accuracy": 0.18472678512334822,
      "num_tokens": 48641297.0,
      "step": 19190
    },
    {
      "entropy": 6.02914719581604,
      "epoch": 2.2152336987882286,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045173015335738925,
      "loss": 5.7103,
      "mean_token_accuracy": 0.19034913331270217,
      "num_tokens": 48654384.0,
      "step": 19195
    },
    {
      "entropy": 5.967139959335327,
      "epoch": 2.215810732833237,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045170461175022116,
      "loss": 5.5925,
      "mean_token_accuracy": 0.19663489311933519,
      "num_tokens": 48666868.0,
      "step": 19200
    },
    {
      "entropy": 6.043667364120483,
      "epoch": 2.216387766878246,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004516790641996812,
      "loss": 5.7304,
      "mean_token_accuracy": 0.18787663877010347,
      "num_tokens": 48679780.0,
      "step": 19205
    },
    {
      "entropy": 5.9972248554229735,
      "epoch": 2.2169648009232543,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004516535107066286,
      "loss": 5.7045,
      "mean_token_accuracy": 0.1850156933069229,
      "num_tokens": 48692438.0,
      "step": 19210
    },
    {
      "entropy": 6.000919389724731,
      "epoch": 2.217541834968263,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045162795127192296,
      "loss": 5.7029,
      "mean_token_accuracy": 0.1894032523036003,
      "num_tokens": 48705627.0,
      "step": 19215
    },
    {
      "entropy": 6.026567506790161,
      "epoch": 2.2181188690132716,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000451602385896424,
      "loss": 5.6663,
      "mean_token_accuracy": 0.1768063023686409,
      "num_tokens": 48717481.0,
      "step": 19220
    },
    {
      "entropy": 5.938123178482056,
      "epoch": 2.2186959030582805,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00045157681458099145,
      "loss": 5.6848,
      "mean_token_accuracy": 0.183728164434433,
      "num_tokens": 48729821.0,
      "step": 19225
    },
    {
      "entropy": 6.0174112796783445,
      "epoch": 2.2192729371032893,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004515512373264854,
      "loss": 5.6744,
      "mean_token_accuracy": 0.18976697474718093,
      "num_tokens": 48742632.0,
      "step": 19230
    },
    {
      "entropy": 5.998519420623779,
      "epoch": 2.2198499711482977,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00045152565413376623,
      "loss": 5.6306,
      "mean_token_accuracy": 0.19520811587572098,
      "num_tokens": 48754712.0,
      "step": 19235
    },
    {
      "entropy": 5.9644848823547365,
      "epoch": 2.2204270051933066,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004515000650036944,
      "loss": 5.6645,
      "mean_token_accuracy": 0.18843511044979094,
      "num_tokens": 48767170.0,
      "step": 19240
    },
    {
      "entropy": 6.08558497428894,
      "epoch": 2.221004039238315,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004514744699371305,
      "loss": 5.8183,
      "mean_token_accuracy": 0.17158350199460984,
      "num_tokens": 48780392.0,
      "step": 19245
    },
    {
      "entropy": 6.121714401245117,
      "epoch": 2.221581073283324,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045144886893493547,
      "loss": 5.7161,
      "mean_token_accuracy": 0.18549460619688035,
      "num_tokens": 48792100.0,
      "step": 19250
    },
    {
      "entropy": 5.996174287796021,
      "epoch": 2.2221581073283323,
      "grad_norm": 0.859375,
      "learning_rate": 0.00045142326199797047,
      "loss": 5.7126,
      "mean_token_accuracy": 0.19352317750453948,
      "num_tokens": 48805192.0,
      "step": 19255
    },
    {
      "entropy": 5.957972860336303,
      "epoch": 2.222735141373341,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00045139764912709656,
      "loss": 5.5797,
      "mean_token_accuracy": 0.19542951732873917,
      "num_tokens": 48819735.0,
      "step": 19260
    },
    {
      "entropy": 5.987135791778565,
      "epoch": 2.2233121754183496,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004513720303231754,
      "loss": 5.6968,
      "mean_token_accuracy": 0.18352420330047609,
      "num_tokens": 48833785.0,
      "step": 19265
    },
    {
      "entropy": 6.027687358856201,
      "epoch": 2.2238892094633584,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045134640558706864,
      "loss": 5.6903,
      "mean_token_accuracy": 0.18623915165662766,
      "num_tokens": 48846491.0,
      "step": 19270
    },
    {
      "entropy": 6.005075883865357,
      "epoch": 2.224466243508367,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000451320774919638,
      "loss": 5.6084,
      "mean_token_accuracy": 0.189390167593956,
      "num_tokens": 48858269.0,
      "step": 19275
    },
    {
      "entropy": 5.982754468917847,
      "epoch": 2.2250432775533757,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00045129513832174587,
      "loss": 5.7678,
      "mean_token_accuracy": 0.17832827270030976,
      "num_tokens": 48870051.0,
      "step": 19280
    },
    {
      "entropy": 6.02019624710083,
      "epoch": 2.225620311598384,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045126949579425414,
      "loss": 5.7127,
      "mean_token_accuracy": 0.18822382390499115,
      "num_tokens": 48882933.0,
      "step": 19285
    },
    {
      "entropy": 5.998995733261109,
      "epoch": 2.226197345643393,
      "grad_norm": 0.90625,
      "learning_rate": 0.00045124384733802563,
      "loss": 5.6952,
      "mean_token_accuracy": 0.18356683105230331,
      "num_tokens": 48895641.0,
      "step": 19290
    },
    {
      "entropy": 6.057963466644287,
      "epoch": 2.2267743796884014,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004512181929539228,
      "loss": 5.7761,
      "mean_token_accuracy": 0.18233993500471116,
      "num_tokens": 48907509.0,
      "step": 19295
    },
    {
      "entropy": 6.022281312942505,
      "epoch": 2.2273514137334103,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045119253264280855,
      "loss": 5.6498,
      "mean_token_accuracy": 0.1901828721165657,
      "num_tokens": 48920270.0,
      "step": 19300
    },
    {
      "entropy": 5.946374225616455,
      "epoch": 2.227928447778419,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045116686640554607,
      "loss": 5.6203,
      "mean_token_accuracy": 0.19598630964756011,
      "num_tokens": 48932630.0,
      "step": 19305
    },
    {
      "entropy": 6.025989437103272,
      "epoch": 2.2285054818234276,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045114119424299843,
      "loss": 5.6844,
      "mean_token_accuracy": 0.19058180153369902,
      "num_tokens": 48944759.0,
      "step": 19310
    },
    {
      "entropy": 5.833686351776123,
      "epoch": 2.2290825158684364,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004511155161560293,
      "loss": 5.5838,
      "mean_token_accuracy": 0.20254546403884888,
      "num_tokens": 48956835.0,
      "step": 19315
    },
    {
      "entropy": 5.9103399276733395,
      "epoch": 2.229659549913445,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045108983214550225,
      "loss": 5.6176,
      "mean_token_accuracy": 0.19416672140359878,
      "num_tokens": 48970552.0,
      "step": 19320
    },
    {
      "entropy": 6.042256546020508,
      "epoch": 2.2302365839584537,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004510641422122811,
      "loss": 5.745,
      "mean_token_accuracy": 0.18928102403879166,
      "num_tokens": 48981206.0,
      "step": 19325
    },
    {
      "entropy": 6.034929275512695,
      "epoch": 2.230813618003462,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045103844635723005,
      "loss": 5.6642,
      "mean_token_accuracy": 0.18891657441854476,
      "num_tokens": 48992849.0,
      "step": 19330
    },
    {
      "entropy": 6.050493764877319,
      "epoch": 2.231390652048471,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045101274458121335,
      "loss": 5.8342,
      "mean_token_accuracy": 0.18349453955888748,
      "num_tokens": 49006094.0,
      "step": 19335
    },
    {
      "entropy": 5.940919542312622,
      "epoch": 2.2319676860934794,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045098703688509537,
      "loss": 5.672,
      "mean_token_accuracy": 0.18957189470529556,
      "num_tokens": 49019118.0,
      "step": 19340
    },
    {
      "entropy": 5.952958965301514,
      "epoch": 2.2325447201384883,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004509613232697408,
      "loss": 5.5878,
      "mean_token_accuracy": 0.1967113196849823,
      "num_tokens": 49031217.0,
      "step": 19345
    },
    {
      "entropy": 6.040749454498291,
      "epoch": 2.2331217541834967,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004509356037360145,
      "loss": 5.7068,
      "mean_token_accuracy": 0.18772808015346526,
      "num_tokens": 49043251.0,
      "step": 19350
    },
    {
      "entropy": 6.038175487518311,
      "epoch": 2.2336987882285055,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004509098782847816,
      "loss": 5.6942,
      "mean_token_accuracy": 0.18514969646930696,
      "num_tokens": 49055227.0,
      "step": 19355
    },
    {
      "entropy": 6.0091157913208,
      "epoch": 2.234275822273514,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004508841469169073,
      "loss": 5.7399,
      "mean_token_accuracy": 0.1856280267238617,
      "num_tokens": 49067810.0,
      "step": 19360
    },
    {
      "entropy": 5.979104900360108,
      "epoch": 2.234852856318523,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045085840963325716,
      "loss": 5.6308,
      "mean_token_accuracy": 0.19960149377584457,
      "num_tokens": 49079898.0,
      "step": 19365
    },
    {
      "entropy": 6.05658450126648,
      "epoch": 2.2354298903635312,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004508326664346967,
      "loss": 5.7488,
      "mean_token_accuracy": 0.18467224687337874,
      "num_tokens": 49092961.0,
      "step": 19370
    },
    {
      "entropy": 5.956911945343018,
      "epoch": 2.23600692440854,
      "grad_norm": 0.84375,
      "learning_rate": 0.00045080691732209176,
      "loss": 5.6576,
      "mean_token_accuracy": 0.19287675023078918,
      "num_tokens": 49106670.0,
      "step": 19375
    },
    {
      "entropy": 5.860464763641358,
      "epoch": 2.236583958453549,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045078116229630864,
      "loss": 5.6149,
      "mean_token_accuracy": 0.192200168967247,
      "num_tokens": 49119947.0,
      "step": 19380
    },
    {
      "entropy": 6.0130188941955565,
      "epoch": 2.2371609924985574,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045075540135821343,
      "loss": 5.7151,
      "mean_token_accuracy": 0.18441500663757324,
      "num_tokens": 49132486.0,
      "step": 19385
    },
    {
      "entropy": 6.006196451187134,
      "epoch": 2.2377380265435662,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004507296345086725,
      "loss": 5.6905,
      "mean_token_accuracy": 0.18315815776586533,
      "num_tokens": 49145121.0,
      "step": 19390
    },
    {
      "entropy": 6.020493984222412,
      "epoch": 2.2383150605885747,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004507038617485526,
      "loss": 5.7229,
      "mean_token_accuracy": 0.18449845016002656,
      "num_tokens": 49157247.0,
      "step": 19395
    },
    {
      "entropy": 6.055055809020996,
      "epoch": 2.2388920946335835,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045067808307872064,
      "loss": 5.7891,
      "mean_token_accuracy": 0.18034157603979112,
      "num_tokens": 49170886.0,
      "step": 19400
    },
    {
      "entropy": 6.040819311141968,
      "epoch": 2.239469128678592,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045065229850004365,
      "loss": 5.7117,
      "mean_token_accuracy": 0.17809169441461564,
      "num_tokens": 49183087.0,
      "step": 19405
    },
    {
      "entropy": 6.056982517242432,
      "epoch": 2.240046162723601,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004506265080133888,
      "loss": 5.6658,
      "mean_token_accuracy": 0.18811472207307817,
      "num_tokens": 49194703.0,
      "step": 19410
    },
    {
      "entropy": 6.006521320343017,
      "epoch": 2.240623196768609,
      "grad_norm": 1.0,
      "learning_rate": 0.00045060071161962364,
      "loss": 5.7822,
      "mean_token_accuracy": 0.18367432653903962,
      "num_tokens": 49208459.0,
      "step": 19415
    },
    {
      "entropy": 5.932523393630982,
      "epoch": 2.241200230813618,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004505749093196158,
      "loss": 5.5775,
      "mean_token_accuracy": 0.19648521095514299,
      "num_tokens": 49221043.0,
      "step": 19420
    },
    {
      "entropy": 6.109450340270996,
      "epoch": 2.2417772648586265,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000450549101114233,
      "loss": 5.734,
      "mean_token_accuracy": 0.18743004351854325,
      "num_tokens": 49233738.0,
      "step": 19425
    },
    {
      "entropy": 5.990781354904175,
      "epoch": 2.2423542989036354,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004505232870043434,
      "loss": 5.7198,
      "mean_token_accuracy": 0.18616331666707991,
      "num_tokens": 49246244.0,
      "step": 19430
    },
    {
      "entropy": 6.014692401885986,
      "epoch": 2.2429313329486438,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004504974669908152,
      "loss": 5.7402,
      "mean_token_accuracy": 0.18278931677341462,
      "num_tokens": 49258126.0,
      "step": 19435
    },
    {
      "entropy": 6.047117567062378,
      "epoch": 2.2435083669936526,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045047164107451696,
      "loss": 5.7377,
      "mean_token_accuracy": 0.19140591770410537,
      "num_tokens": 49270303.0,
      "step": 19440
    },
    {
      "entropy": 5.878828001022339,
      "epoch": 2.244085401038661,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004504458092563172,
      "loss": 5.5083,
      "mean_token_accuracy": 0.20176081955432892,
      "num_tokens": 49283295.0,
      "step": 19445
    },
    {
      "entropy": 6.083686971664429,
      "epoch": 2.24466243508367,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045041997153708475,
      "loss": 5.7062,
      "mean_token_accuracy": 0.18729586899280548,
      "num_tokens": 49296003.0,
      "step": 19450
    },
    {
      "entropy": 6.0752167224884035,
      "epoch": 2.2452394691286788,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045039412791768877,
      "loss": 5.7561,
      "mean_token_accuracy": 0.19029568284749984,
      "num_tokens": 49307982.0,
      "step": 19455
    },
    {
      "entropy": 6.006900787353516,
      "epoch": 2.245816503173687,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045036827839899833,
      "loss": 5.6919,
      "mean_token_accuracy": 0.18658557534217834,
      "num_tokens": 49319220.0,
      "step": 19460
    },
    {
      "entropy": 5.954504728317261,
      "epoch": 2.246393537218696,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045034242298188303,
      "loss": 5.67,
      "mean_token_accuracy": 0.19045452773571014,
      "num_tokens": 49331683.0,
      "step": 19465
    },
    {
      "entropy": 5.992480659484864,
      "epoch": 2.2469705712637045,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004503165616672124,
      "loss": 5.6025,
      "mean_token_accuracy": 0.19973125755786897,
      "num_tokens": 49344815.0,
      "step": 19470
    },
    {
      "entropy": 6.04161901473999,
      "epoch": 2.2475476053087133,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004502906944558563,
      "loss": 5.748,
      "mean_token_accuracy": 0.1864808738231659,
      "num_tokens": 49357399.0,
      "step": 19475
    },
    {
      "entropy": 6.019393825531006,
      "epoch": 2.2481246393537218,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004502648213486848,
      "loss": 5.7139,
      "mean_token_accuracy": 0.1883710280060768,
      "num_tokens": 49369641.0,
      "step": 19480
    },
    {
      "entropy": 5.99062876701355,
      "epoch": 2.2487016733987306,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045023894234656807,
      "loss": 5.6946,
      "mean_token_accuracy": 0.1910533621907234,
      "num_tokens": 49381043.0,
      "step": 19485
    },
    {
      "entropy": 6.041792249679565,
      "epoch": 2.249278707443739,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004502130574503766,
      "loss": 5.7247,
      "mean_token_accuracy": 0.18494125008583068,
      "num_tokens": 49392862.0,
      "step": 19490
    },
    {
      "entropy": 6.062245655059814,
      "epoch": 2.249855741488748,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004501871666609809,
      "loss": 5.7651,
      "mean_token_accuracy": 0.18619535118341446,
      "num_tokens": 49405566.0,
      "step": 19495
    },
    {
      "entropy": 6.038840055465698,
      "epoch": 2.2504327755337563,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004501612699792519,
      "loss": 5.7096,
      "mean_token_accuracy": 0.18405784368515016,
      "num_tokens": 49418048.0,
      "step": 19500
    },
    {
      "entropy": 5.961416244506836,
      "epoch": 2.251009809578765,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004501353674060606,
      "loss": 5.6825,
      "mean_token_accuracy": 0.1893472284078598,
      "num_tokens": 49430129.0,
      "step": 19505
    },
    {
      "entropy": 6.005632591247559,
      "epoch": 2.251586843623774,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004501094589422782,
      "loss": 5.7696,
      "mean_token_accuracy": 0.1801247075200081,
      "num_tokens": 49443047.0,
      "step": 19510
    },
    {
      "entropy": 6.054523468017578,
      "epoch": 2.2521638776687825,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00045008354458877614,
      "loss": 5.7306,
      "mean_token_accuracy": 0.18874183744192125,
      "num_tokens": 49456061.0,
      "step": 19515
    },
    {
      "entropy": 5.896962642669678,
      "epoch": 2.252740911713791,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000450057624346426,
      "loss": 5.4909,
      "mean_token_accuracy": 0.20230573862791063,
      "num_tokens": 49470044.0,
      "step": 19520
    },
    {
      "entropy": 5.968057155609131,
      "epoch": 2.2533179457587997,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045003169821609967,
      "loss": 5.6992,
      "mean_token_accuracy": 0.1845711350440979,
      "num_tokens": 49482232.0,
      "step": 19525
    },
    {
      "entropy": 5.9649145126342775,
      "epoch": 2.2538949798038086,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045000576619866914,
      "loss": 5.649,
      "mean_token_accuracy": 0.18737708032131195,
      "num_tokens": 49493639.0,
      "step": 19530
    },
    {
      "entropy": 6.039490604400635,
      "epoch": 2.254472013848817,
      "grad_norm": 0.875,
      "learning_rate": 0.00044997982829500657,
      "loss": 5.7831,
      "mean_token_accuracy": 0.1838113322854042,
      "num_tokens": 49506233.0,
      "step": 19535
    },
    {
      "entropy": 6.011043739318848,
      "epoch": 2.255049047893826,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044995388450598436,
      "loss": 5.6975,
      "mean_token_accuracy": 0.1865554615855217,
      "num_tokens": 49518479.0,
      "step": 19540
    },
    {
      "entropy": 6.005075454711914,
      "epoch": 2.2556260819388343,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004499279348324751,
      "loss": 5.7085,
      "mean_token_accuracy": 0.18798978477716446,
      "num_tokens": 49531293.0,
      "step": 19545
    },
    {
      "entropy": 5.889176750183106,
      "epoch": 2.256203115983843,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044990197927535173,
      "loss": 5.5829,
      "mean_token_accuracy": 0.1966006800532341,
      "num_tokens": 49543492.0,
      "step": 19550
    },
    {
      "entropy": 6.003212404251099,
      "epoch": 2.2567801500288516,
      "grad_norm": 0.921875,
      "learning_rate": 0.00044987601783548703,
      "loss": 5.7331,
      "mean_token_accuracy": 0.18933305740356446,
      "num_tokens": 49555460.0,
      "step": 19555
    },
    {
      "entropy": 5.887110042572021,
      "epoch": 2.2573571840738604,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004498500505137545,
      "loss": 5.5556,
      "mean_token_accuracy": 0.19345985502004623,
      "num_tokens": 49568077.0,
      "step": 19560
    },
    {
      "entropy": 5.961208915710449,
      "epoch": 2.257934218118869,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004498240773110273,
      "loss": 5.6827,
      "mean_token_accuracy": 0.19075668901205062,
      "num_tokens": 49580201.0,
      "step": 19565
    },
    {
      "entropy": 5.969878911972046,
      "epoch": 2.2585112521638777,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004497980982281791,
      "loss": 5.644,
      "mean_token_accuracy": 0.18944551050662994,
      "num_tokens": 49592469.0,
      "step": 19570
    },
    {
      "entropy": 5.990805053710938,
      "epoch": 2.259088286208886,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004497721132660837,
      "loss": 5.7218,
      "mean_token_accuracy": 0.18673073053359984,
      "num_tokens": 49604961.0,
      "step": 19575
    },
    {
      "entropy": 6.037886667251587,
      "epoch": 2.259665320253895,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044974612242561516,
      "loss": 5.7103,
      "mean_token_accuracy": 0.18756490498781203,
      "num_tokens": 49617435.0,
      "step": 19580
    },
    {
      "entropy": 6.014698123931884,
      "epoch": 2.260242354298904,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004497201257076475,
      "loss": 5.7241,
      "mean_token_accuracy": 0.1855972960591316,
      "num_tokens": 49629799.0,
      "step": 19585
    },
    {
      "entropy": 5.8581945419311525,
      "epoch": 2.2608193883439123,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004496941231130552,
      "loss": 5.5522,
      "mean_token_accuracy": 0.19231831431388854,
      "num_tokens": 49642937.0,
      "step": 19590
    },
    {
      "entropy": 5.960637903213501,
      "epoch": 2.261396422388921,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004496681146427129,
      "loss": 5.6504,
      "mean_token_accuracy": 0.19083957523107528,
      "num_tokens": 49655731.0,
      "step": 19595
    },
    {
      "entropy": 6.073339366912842,
      "epoch": 2.2619734564339296,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00044964210029749523,
      "loss": 5.6871,
      "mean_token_accuracy": 0.18497314155101777,
      "num_tokens": 49668515.0,
      "step": 19600
    },
    {
      "entropy": 6.033228254318237,
      "epoch": 2.2625504904789384,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00044961608007827736,
      "loss": 5.7178,
      "mean_token_accuracy": 0.1855931907892227,
      "num_tokens": 49681246.0,
      "step": 19605
    },
    {
      "entropy": 5.948070335388183,
      "epoch": 2.263127524523947,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004495900539859344,
      "loss": 5.702,
      "mean_token_accuracy": 0.1878760948777199,
      "num_tokens": 49694234.0,
      "step": 19610
    },
    {
      "entropy": 5.94187183380127,
      "epoch": 2.2637045585689557,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044956402202134157,
      "loss": 5.5759,
      "mean_token_accuracy": 0.19712469279766082,
      "num_tokens": 49707511.0,
      "step": 19615
    },
    {
      "entropy": 5.989843845367432,
      "epoch": 2.264281592613964,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00044953798418537465,
      "loss": 5.6797,
      "mean_token_accuracy": 0.19812555611133575,
      "num_tokens": 49720348.0,
      "step": 19620
    },
    {
      "entropy": 5.874147176742554,
      "epoch": 2.264858626658973,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004495119404789093,
      "loss": 5.5422,
      "mean_token_accuracy": 0.18751270323991776,
      "num_tokens": 49733777.0,
      "step": 19625
    },
    {
      "entropy": 6.040910863876343,
      "epoch": 2.2654356607039814,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004494858909028216,
      "loss": 5.6466,
      "mean_token_accuracy": 0.1891347125172615,
      "num_tokens": 49746962.0,
      "step": 19630
    },
    {
      "entropy": 6.069393587112427,
      "epoch": 2.2660126947489903,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004494598354579875,
      "loss": 5.7462,
      "mean_token_accuracy": 0.18031825721263886,
      "num_tokens": 49760211.0,
      "step": 19635
    },
    {
      "entropy": 5.939764070510864,
      "epoch": 2.2665897287939987,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004494337741452836,
      "loss": 5.6078,
      "mean_token_accuracy": 0.19920677691698074,
      "num_tokens": 49773215.0,
      "step": 19640
    },
    {
      "entropy": 6.030709552764892,
      "epoch": 2.2671667628390075,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004494077069655863,
      "loss": 5.6811,
      "mean_token_accuracy": 0.18590768724679946,
      "num_tokens": 49785521.0,
      "step": 19645
    },
    {
      "entropy": 5.938182735443116,
      "epoch": 2.267743796884016,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004493816339197724,
      "loss": 5.6478,
      "mean_token_accuracy": 0.19264764785766603,
      "num_tokens": 49796449.0,
      "step": 19650
    },
    {
      "entropy": 5.9930580139160154,
      "epoch": 2.268320830929025,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044935555500871897,
      "loss": 5.7237,
      "mean_token_accuracy": 0.1861321210861206,
      "num_tokens": 49810222.0,
      "step": 19655
    },
    {
      "entropy": 6.031548643112183,
      "epoch": 2.2688978649740337,
      "grad_norm": 0.875,
      "learning_rate": 0.000449329470233303,
      "loss": 5.6491,
      "mean_token_accuracy": 0.1902469277381897,
      "num_tokens": 49822799.0,
      "step": 19660
    },
    {
      "entropy": 5.983332252502441,
      "epoch": 2.269474899019042,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044930337959440183,
      "loss": 5.6645,
      "mean_token_accuracy": 0.1808748185634613,
      "num_tokens": 49835782.0,
      "step": 19665
    },
    {
      "entropy": 6.0095597267150875,
      "epoch": 2.270051933064051,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004492772830928931,
      "loss": 5.7018,
      "mean_token_accuracy": 0.18447400331497193,
      "num_tokens": 49849204.0,
      "step": 19670
    },
    {
      "entropy": 6.030408191680908,
      "epoch": 2.2706289671090594,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044925118072965456,
      "loss": 5.7782,
      "mean_token_accuracy": 0.18234920352697373,
      "num_tokens": 49860769.0,
      "step": 19675
    },
    {
      "entropy": 6.010703372955322,
      "epoch": 2.2712060011540682,
      "grad_norm": 0.9375,
      "learning_rate": 0.000449225072505564,
      "loss": 5.6713,
      "mean_token_accuracy": 0.1898810014128685,
      "num_tokens": 49873923.0,
      "step": 19680
    },
    {
      "entropy": 5.997634792327881,
      "epoch": 2.2717830351990767,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00044919895842149976,
      "loss": 5.6432,
      "mean_token_accuracy": 0.19750694334506988,
      "num_tokens": 49886968.0,
      "step": 19685
    },
    {
      "entropy": 5.89441065788269,
      "epoch": 2.2723600692440855,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00044917283847834005,
      "loss": 5.5927,
      "mean_token_accuracy": 0.196426597237587,
      "num_tokens": 49899096.0,
      "step": 19690
    },
    {
      "entropy": 5.984251546859741,
      "epoch": 2.272937103289094,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004491467126769635,
      "loss": 5.6481,
      "mean_token_accuracy": 0.18489282131195067,
      "num_tokens": 49911887.0,
      "step": 19695
    },
    {
      "entropy": 5.989725923538208,
      "epoch": 2.273514137334103,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00044912058101824866,
      "loss": 5.6289,
      "mean_token_accuracy": 0.1994626134634018,
      "num_tokens": 49925188.0,
      "step": 19700
    },
    {
      "entropy": 5.9687357425689695,
      "epoch": 2.274091171379111,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00044909444350307463,
      "loss": 5.6623,
      "mean_token_accuracy": 0.18800243586301804,
      "num_tokens": 49938568.0,
      "step": 19705
    },
    {
      "entropy": 5.982844018936158,
      "epoch": 2.27466820542412,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004490683001323205,
      "loss": 5.5998,
      "mean_token_accuracy": 0.19410659223794938,
      "num_tokens": 49950630.0,
      "step": 19710
    },
    {
      "entropy": 5.989612197875976,
      "epoch": 2.2752452394691285,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044904215090686554,
      "loss": 5.6494,
      "mean_token_accuracy": 0.192630897462368,
      "num_tokens": 49962242.0,
      "step": 19715
    },
    {
      "entropy": 5.981988716125488,
      "epoch": 2.2758222735141374,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044901599582758926,
      "loss": 5.7229,
      "mean_token_accuracy": 0.18031760156154633,
      "num_tokens": 49974604.0,
      "step": 19720
    },
    {
      "entropy": 5.949802446365356,
      "epoch": 2.2763993075591458,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00044898983489537143,
      "loss": 5.6689,
      "mean_token_accuracy": 0.1892044261097908,
      "num_tokens": 49986326.0,
      "step": 19725
    },
    {
      "entropy": 6.023297500610352,
      "epoch": 2.2769763416041546,
      "grad_norm": 1.09375,
      "learning_rate": 0.000448963668111092,
      "loss": 5.65,
      "mean_token_accuracy": 0.1980791911482811,
      "num_tokens": 49998752.0,
      "step": 19730
    },
    {
      "entropy": 5.972213172912598,
      "epoch": 2.2775533756491635,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00044893749547563085,
      "loss": 5.7468,
      "mean_token_accuracy": 0.18335980772972107,
      "num_tokens": 50011839.0,
      "step": 19735
    },
    {
      "entropy": 5.97639102935791,
      "epoch": 2.278130409694172,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044891131698986846,
      "loss": 5.6213,
      "mean_token_accuracy": 0.19201520085334778,
      "num_tokens": 50024554.0,
      "step": 19740
    },
    {
      "entropy": 6.041115427017212,
      "epoch": 2.2787074437391808,
      "grad_norm": 1.0,
      "learning_rate": 0.0004488851326546854,
      "loss": 5.7523,
      "mean_token_accuracy": 0.1877804309129715,
      "num_tokens": 50038868.0,
      "step": 19745
    },
    {
      "entropy": 6.036338043212891,
      "epoch": 2.279284477784189,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004488589424709622,
      "loss": 5.6751,
      "mean_token_accuracy": 0.19028718024492264,
      "num_tokens": 50050556.0,
      "step": 19750
    },
    {
      "entropy": 6.0639410495758055,
      "epoch": 2.279861511829198,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004488327464395799,
      "loss": 5.7113,
      "mean_token_accuracy": 0.1872240573167801,
      "num_tokens": 50063169.0,
      "step": 19755
    },
    {
      "entropy": 6.0162170886993405,
      "epoch": 2.2804385458742065,
      "grad_norm": 1.0,
      "learning_rate": 0.0004488065445614195,
      "loss": 5.7457,
      "mean_token_accuracy": 0.18508987873792648,
      "num_tokens": 50076563.0,
      "step": 19760
    },
    {
      "entropy": 5.9870819568634035,
      "epoch": 2.2810155799192153,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004487803368373623,
      "loss": 5.6972,
      "mean_token_accuracy": 0.18701732754707337,
      "num_tokens": 50089584.0,
      "step": 19765
    },
    {
      "entropy": 5.97633728981018,
      "epoch": 2.2815926139642237,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004487541232682898,
      "loss": 5.633,
      "mean_token_accuracy": 0.18684376776218414,
      "num_tokens": 50101566.0,
      "step": 19770
    },
    {
      "entropy": 6.027882862091064,
      "epoch": 2.2821696480092326,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004487279038550836,
      "loss": 5.701,
      "mean_token_accuracy": 0.19523897767066956,
      "num_tokens": 50114378.0,
      "step": 19775
    },
    {
      "entropy": 6.000160598754883,
      "epoch": 2.282746682054241,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004487016785986258,
      "loss": 5.6551,
      "mean_token_accuracy": 0.1923314392566681,
      "num_tokens": 50126658.0,
      "step": 19780
    },
    {
      "entropy": 5.98645133972168,
      "epoch": 2.28332371609925,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044867544749979817,
      "loss": 5.6966,
      "mean_token_accuracy": 0.18773801177740096,
      "num_tokens": 50139802.0,
      "step": 19785
    },
    {
      "entropy": 6.1028032302856445,
      "epoch": 2.2839007501442588,
      "grad_norm": 1.0,
      "learning_rate": 0.00044864921055948317,
      "loss": 5.7977,
      "mean_token_accuracy": 0.18053346276283264,
      "num_tokens": 50152158.0,
      "step": 19790
    },
    {
      "entropy": 6.0473404884338375,
      "epoch": 2.284477784189267,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044862296777856326,
      "loss": 5.6634,
      "mean_token_accuracy": 0.18888448029756547,
      "num_tokens": 50165099.0,
      "step": 19795
    },
    {
      "entropy": 5.937266063690186,
      "epoch": 2.2850548182342756,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004485967191579211,
      "loss": 5.6554,
      "mean_token_accuracy": 0.18916501700878144,
      "num_tokens": 50177073.0,
      "step": 19800
    },
    {
      "entropy": 6.0406135559082035,
      "epoch": 2.2856318522792844,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004485704646984394,
      "loss": 5.7279,
      "mean_token_accuracy": 0.18644598722457886,
      "num_tokens": 50190458.0,
      "step": 19805
    },
    {
      "entropy": 5.969699811935425,
      "epoch": 2.2862088863242933,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004485442044010015,
      "loss": 5.6228,
      "mean_token_accuracy": 0.1918807476758957,
      "num_tokens": 50201726.0,
      "step": 19810
    },
    {
      "entropy": 6.016246175765991,
      "epoch": 2.2867859203693017,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004485179382664904,
      "loss": 5.6616,
      "mean_token_accuracy": 0.19440938979387284,
      "num_tokens": 50215550.0,
      "step": 19815
    },
    {
      "entropy": 5.993291759490967,
      "epoch": 2.2873629544143106,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004484916662957896,
      "loss": 5.6723,
      "mean_token_accuracy": 0.18349820971488953,
      "num_tokens": 50227982.0,
      "step": 19820
    },
    {
      "entropy": 6.034884643554688,
      "epoch": 2.287939988459319,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004484653884897829,
      "loss": 5.6696,
      "mean_token_accuracy": 0.18665958642959596,
      "num_tokens": 50239804.0,
      "step": 19825
    },
    {
      "entropy": 6.075146579742432,
      "epoch": 2.288517022504328,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044843910484935394,
      "loss": 5.716,
      "mean_token_accuracy": 0.19153106659650804,
      "num_tokens": 50252192.0,
      "step": 19830
    },
    {
      "entropy": 6.075097370147705,
      "epoch": 2.2890940565493363,
      "grad_norm": 1.0,
      "learning_rate": 0.0004484128153753868,
      "loss": 5.7583,
      "mean_token_accuracy": 0.181975756585598,
      "num_tokens": 50265194.0,
      "step": 19835
    },
    {
      "entropy": 6.000475597381592,
      "epoch": 2.289671090594345,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044838652006876583,
      "loss": 5.6453,
      "mean_token_accuracy": 0.19204287230968475,
      "num_tokens": 50277704.0,
      "step": 19840
    },
    {
      "entropy": 5.9271547317504885,
      "epoch": 2.2902481246393536,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044836021893037537,
      "loss": 5.6279,
      "mean_token_accuracy": 0.19254519790410995,
      "num_tokens": 50290670.0,
      "step": 19845
    },
    {
      "entropy": 5.905674743652344,
      "epoch": 2.2908251586843624,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004483339119611001,
      "loss": 5.6224,
      "mean_token_accuracy": 0.19302036762237548,
      "num_tokens": 50302764.0,
      "step": 19850
    },
    {
      "entropy": 6.004446411132813,
      "epoch": 2.291402192729371,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044830759916182476,
      "loss": 5.6833,
      "mean_token_accuracy": 0.19413198977708818,
      "num_tokens": 50316824.0,
      "step": 19855
    },
    {
      "entropy": 6.124077177047729,
      "epoch": 2.2919792267743797,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004482812805334344,
      "loss": 5.7341,
      "mean_token_accuracy": 0.1839630201458931,
      "num_tokens": 50329259.0,
      "step": 19860
    },
    {
      "entropy": 5.965328645706177,
      "epoch": 2.2925562608193886,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004482549560768142,
      "loss": 5.6268,
      "mean_token_accuracy": 0.20081087350845336,
      "num_tokens": 50341165.0,
      "step": 19865
    },
    {
      "entropy": 5.846808052062988,
      "epoch": 2.293133294864397,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004482286257928497,
      "loss": 5.6312,
      "mean_token_accuracy": 0.1967452809214592,
      "num_tokens": 50354655.0,
      "step": 19870
    },
    {
      "entropy": 6.057693910598755,
      "epoch": 2.293710328909406,
      "grad_norm": 1.0,
      "learning_rate": 0.0004482022896824263,
      "loss": 5.7325,
      "mean_token_accuracy": 0.1845332995057106,
      "num_tokens": 50366917.0,
      "step": 19875
    },
    {
      "entropy": 6.058557605743408,
      "epoch": 2.2942873629544143,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00044817594774643004,
      "loss": 5.7726,
      "mean_token_accuracy": 0.18246468752622605,
      "num_tokens": 50380415.0,
      "step": 19880
    },
    {
      "entropy": 6.004601907730103,
      "epoch": 2.294864396999423,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00044814959998574675,
      "loss": 5.6364,
      "mean_token_accuracy": 0.19406939893960953,
      "num_tokens": 50393051.0,
      "step": 19885
    },
    {
      "entropy": 6.033665704727173,
      "epoch": 2.2954414310444315,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00044812324640126265,
      "loss": 5.7393,
      "mean_token_accuracy": 0.18072236478328704,
      "num_tokens": 50407517.0,
      "step": 19890
    },
    {
      "entropy": 6.055517911911011,
      "epoch": 2.2960184650894404,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004480968869938642,
      "loss": 5.7511,
      "mean_token_accuracy": 0.18511457443237306,
      "num_tokens": 50419656.0,
      "step": 19895
    },
    {
      "entropy": 6.012159395217895,
      "epoch": 2.296595499134449,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00044807052176443793,
      "loss": 5.5961,
      "mean_token_accuracy": 0.1847260281443596,
      "num_tokens": 50433798.0,
      "step": 19900
    },
    {
      "entropy": 6.07460584640503,
      "epoch": 2.2971725331794577,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00044804415071387067,
      "loss": 5.7412,
      "mean_token_accuracy": 0.18035354912281037,
      "num_tokens": 50445459.0,
      "step": 19905
    },
    {
      "entropy": 5.98373851776123,
      "epoch": 2.297749567224466,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004480177738430492,
      "loss": 5.6309,
      "mean_token_accuracy": 0.19788738936185837,
      "num_tokens": 50458649.0,
      "step": 19910
    },
    {
      "entropy": 5.86762056350708,
      "epoch": 2.298326601269475,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044799139115286104,
      "loss": 5.5207,
      "mean_token_accuracy": 0.20949087291955948,
      "num_tokens": 50471878.0,
      "step": 19915
    },
    {
      "entropy": 6.0376941680908205,
      "epoch": 2.2989036353144834,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004479650026441933,
      "loss": 5.7293,
      "mean_token_accuracy": 0.1906494379043579,
      "num_tokens": 50484033.0,
      "step": 19920
    },
    {
      "entropy": 6.056472110748291,
      "epoch": 2.2994806693594922,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044793860831793356,
      "loss": 5.7058,
      "mean_token_accuracy": 0.18846760541200638,
      "num_tokens": 50496072.0,
      "step": 19925
    },
    {
      "entropy": 6.032245922088623,
      "epoch": 2.3000577034045007,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044791220817496963,
      "loss": 5.6601,
      "mean_token_accuracy": 0.1949844554066658,
      "num_tokens": 50508572.0,
      "step": 19930
    },
    {
      "entropy": 6.033547401428223,
      "epoch": 2.3006347374495095,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004478858022161895,
      "loss": 5.7759,
      "mean_token_accuracy": 0.18765847831964494,
      "num_tokens": 50521946.0,
      "step": 19935
    },
    {
      "entropy": 6.023372840881348,
      "epoch": 2.3012117714945184,
      "grad_norm": 1.0,
      "learning_rate": 0.0004478593904424813,
      "loss": 5.6761,
      "mean_token_accuracy": 0.18883833438158035,
      "num_tokens": 50534517.0,
      "step": 19940
    },
    {
      "entropy": 5.985220909118652,
      "epoch": 2.301788805539527,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004478329728547334,
      "loss": 5.6922,
      "mean_token_accuracy": 0.1892163038253784,
      "num_tokens": 50546479.0,
      "step": 19945
    },
    {
      "entropy": 6.091574239730835,
      "epoch": 2.3023658395845357,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044780654945383424,
      "loss": 5.7902,
      "mean_token_accuracy": 0.1810247167944908,
      "num_tokens": 50559579.0,
      "step": 19950
    },
    {
      "entropy": 6.0166724681854244,
      "epoch": 2.302942873629544,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044778012024067267,
      "loss": 5.6182,
      "mean_token_accuracy": 0.18541074693202972,
      "num_tokens": 50572810.0,
      "step": 19955
    },
    {
      "entropy": 5.988966941833496,
      "epoch": 2.303519907674553,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004477536852161376,
      "loss": 5.6677,
      "mean_token_accuracy": 0.18990887105464935,
      "num_tokens": 50585514.0,
      "step": 19960
    },
    {
      "entropy": 6.019472885131836,
      "epoch": 2.3040969417195614,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004477272443811181,
      "loss": 5.7106,
      "mean_token_accuracy": 0.19112055003643036,
      "num_tokens": 50597897.0,
      "step": 19965
    },
    {
      "entropy": 5.948458099365235,
      "epoch": 2.3046739757645702,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004477007977365035,
      "loss": 5.6845,
      "mean_token_accuracy": 0.18707336038351058,
      "num_tokens": 50610251.0,
      "step": 19970
    },
    {
      "entropy": 5.995520830154419,
      "epoch": 2.3052510098095786,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004476743452831834,
      "loss": 5.6742,
      "mean_token_accuracy": 0.1870413526892662,
      "num_tokens": 50622212.0,
      "step": 19975
    },
    {
      "entropy": 5.991366577148438,
      "epoch": 2.3058280438545875,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00044764788702204747,
      "loss": 5.6679,
      "mean_token_accuracy": 0.18325578719377517,
      "num_tokens": 50636648.0,
      "step": 19980
    },
    {
      "entropy": 6.074189138412476,
      "epoch": 2.306405077899596,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004476214229539856,
      "loss": 5.7331,
      "mean_token_accuracy": 0.19208849221467972,
      "num_tokens": 50648033.0,
      "step": 19985
    },
    {
      "entropy": 5.971599340438843,
      "epoch": 2.306982111944605,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004475949530798879,
      "loss": 5.7198,
      "mean_token_accuracy": 0.18865778148174286,
      "num_tokens": 50660547.0,
      "step": 19990
    },
    {
      "entropy": 5.976639366149902,
      "epoch": 2.307559145989613,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044756847740064475,
      "loss": 5.6442,
      "mean_token_accuracy": 0.19408462047576905,
      "num_tokens": 50672194.0,
      "step": 19995
    },
    {
      "entropy": 5.954980087280274,
      "epoch": 2.308136180034622,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004475419959171465,
      "loss": 5.6683,
      "mean_token_accuracy": 0.19476917535066604,
      "num_tokens": 50685896.0,
      "step": 20000
    },
    {
      "entropy": 5.914623737335205,
      "epoch": 2.3087132140796305,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000447515508630284,
      "loss": 5.612,
      "mean_token_accuracy": 0.1903280273079872,
      "num_tokens": 50699255.0,
      "step": 20005
    },
    {
      "entropy": 5.994593238830566,
      "epoch": 2.3092902481246393,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044748901554094806,
      "loss": 5.6117,
      "mean_token_accuracy": 0.19207348078489303,
      "num_tokens": 50711098.0,
      "step": 20010
    },
    {
      "entropy": 6.079914903640747,
      "epoch": 2.309867282169648,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004474625166500297,
      "loss": 5.8361,
      "mean_token_accuracy": 0.17811717242002487,
      "num_tokens": 50724821.0,
      "step": 20015
    },
    {
      "entropy": 6.057621860504151,
      "epoch": 2.3104443162146566,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044743601195842026,
      "loss": 5.7061,
      "mean_token_accuracy": 0.1867959052324295,
      "num_tokens": 50736572.0,
      "step": 20020
    },
    {
      "entropy": 5.807083702087402,
      "epoch": 2.3110213502596655,
      "grad_norm": 0.859375,
      "learning_rate": 0.00044740950146701127,
      "loss": 5.4422,
      "mean_token_accuracy": 0.20899612158536912,
      "num_tokens": 50749743.0,
      "step": 20025
    },
    {
      "entropy": 6.016130065917968,
      "epoch": 2.311598384304674,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004473829851766943,
      "loss": 5.6504,
      "mean_token_accuracy": 0.19711169898509978,
      "num_tokens": 50761558.0,
      "step": 20030
    },
    {
      "entropy": 6.005746221542358,
      "epoch": 2.3121754183496828,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004473564630883612,
      "loss": 5.7366,
      "mean_token_accuracy": 0.18144539296627044,
      "num_tokens": 50773963.0,
      "step": 20035
    },
    {
      "entropy": 5.981596994400024,
      "epoch": 2.312752452394691,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004473299352029042,
      "loss": 5.6866,
      "mean_token_accuracy": 0.18796083629131316,
      "num_tokens": 50787160.0,
      "step": 20040
    },
    {
      "entropy": 5.9934327602386475,
      "epoch": 2.3133294864397,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004473034015212154,
      "loss": 5.5719,
      "mean_token_accuracy": 0.19589619785547258,
      "num_tokens": 50800014.0,
      "step": 20045
    },
    {
      "entropy": 5.915900754928589,
      "epoch": 2.3139065204847085,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004472768620441872,
      "loss": 5.5901,
      "mean_token_accuracy": 0.19380107074975966,
      "num_tokens": 50812149.0,
      "step": 20050
    },
    {
      "entropy": 6.007445049285889,
      "epoch": 2.3144835545297173,
      "grad_norm": 0.890625,
      "learning_rate": 0.00044725031677271234,
      "loss": 5.7257,
      "mean_token_accuracy": 0.1878676474094391,
      "num_tokens": 50825044.0,
      "step": 20055
    },
    {
      "entropy": 6.0918385028839115,
      "epoch": 2.3150605885747257,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004472237657076837,
      "loss": 5.789,
      "mean_token_accuracy": 0.18238871395587922,
      "num_tokens": 50838813.0,
      "step": 20060
    },
    {
      "entropy": 5.980409240722656,
      "epoch": 2.3156376226197346,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004471972088499942,
      "loss": 5.5776,
      "mean_token_accuracy": 0.18552233874797822,
      "num_tokens": 50853029.0,
      "step": 20065
    },
    {
      "entropy": 5.989367580413818,
      "epoch": 2.3162146566647435,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004471706462005371,
      "loss": 5.6622,
      "mean_token_accuracy": 0.1867926225066185,
      "num_tokens": 50866383.0,
      "step": 20070
    },
    {
      "entropy": 5.97119779586792,
      "epoch": 2.316791690709752,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004471440777602059,
      "loss": 5.6874,
      "mean_token_accuracy": 0.19191924482584,
      "num_tokens": 50880537.0,
      "step": 20075
    },
    {
      "entropy": 5.965939331054687,
      "epoch": 2.3173687247547603,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044711750352989407,
      "loss": 5.6527,
      "mean_token_accuracy": 0.1954786866903305,
      "num_tokens": 50892213.0,
      "step": 20080
    },
    {
      "entropy": 5.985074758529663,
      "epoch": 2.317945758799769,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004470909235104956,
      "loss": 5.6806,
      "mean_token_accuracy": 0.18766402453184128,
      "num_tokens": 50904022.0,
      "step": 20085
    },
    {
      "entropy": 5.97315993309021,
      "epoch": 2.318522792844778,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004470643377029043,
      "loss": 5.6594,
      "mean_token_accuracy": 0.1852023720741272,
      "num_tokens": 50916766.0,
      "step": 20090
    },
    {
      "entropy": 5.971874570846557,
      "epoch": 2.3190998268897864,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004470377461080145,
      "loss": 5.6184,
      "mean_token_accuracy": 0.19405989944934846,
      "num_tokens": 50928842.0,
      "step": 20095
    },
    {
      "entropy": 5.930401039123535,
      "epoch": 2.3196768609347953,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004470111487267206,
      "loss": 5.6735,
      "mean_token_accuracy": 0.18988527953624726,
      "num_tokens": 50941079.0,
      "step": 20100
    },
    {
      "entropy": 6.002483701705932,
      "epoch": 2.3202538949798037,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004469845455599171,
      "loss": 5.662,
      "mean_token_accuracy": 0.18476664423942565,
      "num_tokens": 50953299.0,
      "step": 20105
    },
    {
      "entropy": 6.026644468307495,
      "epoch": 2.3208309290248126,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004469579366084989,
      "loss": 5.6484,
      "mean_token_accuracy": 0.18848707973957063,
      "num_tokens": 50965550.0,
      "step": 20110
    },
    {
      "entropy": 5.949587392807007,
      "epoch": 2.321407963069821,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044693132187336094,
      "loss": 5.622,
      "mean_token_accuracy": 0.19643386006355285,
      "num_tokens": 50978454.0,
      "step": 20115
    },
    {
      "entropy": 6.001140356063843,
      "epoch": 2.32198499711483,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004469047013553983,
      "loss": 5.6768,
      "mean_token_accuracy": 0.18956706672906876,
      "num_tokens": 50990473.0,
      "step": 20120
    },
    {
      "entropy": 5.981013393402099,
      "epoch": 2.3225620311598383,
      "grad_norm": 1.0,
      "learning_rate": 0.00044687807505550646,
      "loss": 5.7467,
      "mean_token_accuracy": 0.1879773110151291,
      "num_tokens": 51003914.0,
      "step": 20125
    },
    {
      "entropy": 6.046197366714478,
      "epoch": 2.323139065204847,
      "grad_norm": 0.921875,
      "learning_rate": 0.00044685144297458096,
      "loss": 5.7642,
      "mean_token_accuracy": 0.18143218904733657,
      "num_tokens": 51016874.0,
      "step": 20130
    },
    {
      "entropy": 6.037942552566529,
      "epoch": 2.3237160992498556,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044682480511351754,
      "loss": 5.6925,
      "mean_token_accuracy": 0.19093703776597976,
      "num_tokens": 51030568.0,
      "step": 20135
    },
    {
      "entropy": 6.065595436096191,
      "epoch": 2.3242931332948644,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004467981614732121,
      "loss": 5.6758,
      "mean_token_accuracy": 0.18532338440418245,
      "num_tokens": 51043876.0,
      "step": 20140
    },
    {
      "entropy": 6.017436552047729,
      "epoch": 2.3248701673398733,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00044677151205456086,
      "loss": 5.7123,
      "mean_token_accuracy": 0.1866360515356064,
      "num_tokens": 51056693.0,
      "step": 20145
    },
    {
      "entropy": 6.0116270065307615,
      "epoch": 2.3254472013848817,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004467448568584601,
      "loss": 5.7116,
      "mean_token_accuracy": 0.18866565972566604,
      "num_tokens": 51068878.0,
      "step": 20150
    },
    {
      "entropy": 5.996644735336304,
      "epoch": 2.32602423542989,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004467181958858064,
      "loss": 5.739,
      "mean_token_accuracy": 0.18351015001535415,
      "num_tokens": 51082234.0,
      "step": 20155
    },
    {
      "entropy": 6.008231973648071,
      "epoch": 2.326601269474899,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004466915291374965,
      "loss": 5.7061,
      "mean_token_accuracy": 0.18806238770484923,
      "num_tokens": 51095469.0,
      "step": 20160
    },
    {
      "entropy": 6.06663932800293,
      "epoch": 2.327178303519908,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004466648566144273,
      "loss": 5.6518,
      "mean_token_accuracy": 0.18935106098651885,
      "num_tokens": 51107302.0,
      "step": 20165
    },
    {
      "entropy": 6.012594318389892,
      "epoch": 2.3277553375649163,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000446638178317496,
      "loss": 5.7677,
      "mean_token_accuracy": 0.17738774865865709,
      "num_tokens": 51121222.0,
      "step": 20170
    },
    {
      "entropy": 6.066937112808228,
      "epoch": 2.328332371609925,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00044661149424759974,
      "loss": 5.7203,
      "mean_token_accuracy": 0.189773128926754,
      "num_tokens": 51134113.0,
      "step": 20175
    },
    {
      "entropy": 5.9701526165008545,
      "epoch": 2.3289094056549335,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004465848044056361,
      "loss": 5.6413,
      "mean_token_accuracy": 0.19023274332284928,
      "num_tokens": 51145978.0,
      "step": 20180
    },
    {
      "entropy": 5.863338422775269,
      "epoch": 2.3294864396999424,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004465581087925028,
      "loss": 5.5756,
      "mean_token_accuracy": 0.19540849179029465,
      "num_tokens": 51158016.0,
      "step": 20185
    },
    {
      "entropy": 6.014245939254761,
      "epoch": 2.330063473744951,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004465314074090978,
      "loss": 5.6662,
      "mean_token_accuracy": 0.19576810151338578,
      "num_tokens": 51170222.0,
      "step": 20190
    },
    {
      "entropy": 5.99976544380188,
      "epoch": 2.3306405077899597,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044650470025631904,
      "loss": 5.6103,
      "mean_token_accuracy": 0.19109719395637512,
      "num_tokens": 51183560.0,
      "step": 20195
    },
    {
      "entropy": 5.962846565246582,
      "epoch": 2.331217541834968,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004464779873350649,
      "loss": 5.6422,
      "mean_token_accuracy": 0.19249949604272842,
      "num_tokens": 51195532.0,
      "step": 20200
    },
    {
      "entropy": 6.029008865356445,
      "epoch": 2.331794575879977,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004464512686462339,
      "loss": 5.7008,
      "mean_token_accuracy": 0.1868092104792595,
      "num_tokens": 51207078.0,
      "step": 20205
    },
    {
      "entropy": 5.974702787399292,
      "epoch": 2.3323716099249854,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00044642454419072455,
      "loss": 5.7221,
      "mean_token_accuracy": 0.17796826213598252,
      "num_tokens": 51220642.0,
      "step": 20210
    },
    {
      "entropy": 6.038258028030396,
      "epoch": 2.3329486439699942,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004463978139694358,
      "loss": 5.7074,
      "mean_token_accuracy": 0.18571033030748368,
      "num_tokens": 51234127.0,
      "step": 20215
    },
    {
      "entropy": 6.033366966247558,
      "epoch": 2.333525678015003,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044637107798326675,
      "loss": 5.6594,
      "mean_token_accuracy": 0.1872406542301178,
      "num_tokens": 51246645.0,
      "step": 20220
    },
    {
      "entropy": 5.942353963851929,
      "epoch": 2.3341027120600115,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004463443362331166,
      "loss": 5.5956,
      "mean_token_accuracy": 0.19952472746372224,
      "num_tokens": 51259392.0,
      "step": 20225
    },
    {
      "entropy": 5.922240209579468,
      "epoch": 2.3346797461050204,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044631758871988486,
      "loss": 5.6493,
      "mean_token_accuracy": 0.1865500718355179,
      "num_tokens": 51273073.0,
      "step": 20230
    },
    {
      "entropy": 5.9896955490112305,
      "epoch": 2.335256780150029,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004462908354444711,
      "loss": 5.6739,
      "mean_token_accuracy": 0.19293740540742874,
      "num_tokens": 51286350.0,
      "step": 20235
    },
    {
      "entropy": 5.941976070404053,
      "epoch": 2.3358338141950377,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004462640764077751,
      "loss": 5.5602,
      "mean_token_accuracy": 0.2030545249581337,
      "num_tokens": 51298262.0,
      "step": 20240
    },
    {
      "entropy": 6.026712131500244,
      "epoch": 2.336410848240046,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004462373116106971,
      "loss": 5.7089,
      "mean_token_accuracy": 0.18382593840360642,
      "num_tokens": 51310162.0,
      "step": 20245
    },
    {
      "entropy": 5.9759049892425535,
      "epoch": 2.336987882285055,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044621054105413704,
      "loss": 5.5954,
      "mean_token_accuracy": 0.19266701638698577,
      "num_tokens": 51323468.0,
      "step": 20250
    },
    {
      "entropy": 6.042396783828735,
      "epoch": 2.3375649163300634,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044618376473899555,
      "loss": 5.7039,
      "mean_token_accuracy": 0.18453248739242553,
      "num_tokens": 51336674.0,
      "step": 20255
    },
    {
      "entropy": 6.0035162448883055,
      "epoch": 2.338141950375072,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004461569826661732,
      "loss": 5.7392,
      "mean_token_accuracy": 0.1827777311205864,
      "num_tokens": 51349950.0,
      "step": 20260
    },
    {
      "entropy": 6.005234050750732,
      "epoch": 2.3387189844200806,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004461301948365707,
      "loss": 5.7216,
      "mean_token_accuracy": 0.18666368573904038,
      "num_tokens": 51362191.0,
      "step": 20265
    },
    {
      "entropy": 5.967562246322632,
      "epoch": 2.3392960184650895,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004461034012510891,
      "loss": 5.6053,
      "mean_token_accuracy": 0.1955225184559822,
      "num_tokens": 51374325.0,
      "step": 20270
    },
    {
      "entropy": 5.908226585388183,
      "epoch": 2.339873052510098,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044607660191062963,
      "loss": 5.5016,
      "mean_token_accuracy": 0.20404043942689895,
      "num_tokens": 51386632.0,
      "step": 20275
    },
    {
      "entropy": 6.028873777389526,
      "epoch": 2.340450086555107,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044604979681609364,
      "loss": 5.7284,
      "mean_token_accuracy": 0.18442705571651458,
      "num_tokens": 51397876.0,
      "step": 20280
    },
    {
      "entropy": 5.997422409057617,
      "epoch": 2.341027120600115,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044602298596838264,
      "loss": 5.6376,
      "mean_token_accuracy": 0.19605884253978728,
      "num_tokens": 51411466.0,
      "step": 20285
    },
    {
      "entropy": 5.912942743301391,
      "epoch": 2.341604154645124,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044599616936839853,
      "loss": 5.5544,
      "mean_token_accuracy": 0.1968266859650612,
      "num_tokens": 51423241.0,
      "step": 20290
    },
    {
      "entropy": 5.925179290771484,
      "epoch": 2.342181188690133,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004459693470170432,
      "loss": 5.6933,
      "mean_token_accuracy": 0.18970109820365905,
      "num_tokens": 51436056.0,
      "step": 20295
    },
    {
      "entropy": 6.059657859802246,
      "epoch": 2.3427582227351413,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004459425189152187,
      "loss": 5.7054,
      "mean_token_accuracy": 0.1905315950512886,
      "num_tokens": 51449986.0,
      "step": 20300
    },
    {
      "entropy": 5.831287670135498,
      "epoch": 2.34333525678015,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00044591568506382757,
      "loss": 5.5033,
      "mean_token_accuracy": 0.20874383598566054,
      "num_tokens": 51463784.0,
      "step": 20305
    },
    {
      "entropy": 5.87903790473938,
      "epoch": 2.3439122908251586,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044588884546377226,
      "loss": 5.5528,
      "mean_token_accuracy": 0.2086120441555977,
      "num_tokens": 51476772.0,
      "step": 20310
    },
    {
      "entropy": 6.024144983291626,
      "epoch": 2.3444893248701675,
      "grad_norm": 1.0,
      "learning_rate": 0.0004458620001159555,
      "loss": 5.685,
      "mean_token_accuracy": 0.19044898748397826,
      "num_tokens": 51489889.0,
      "step": 20315
    },
    {
      "entropy": 6.018150186538696,
      "epoch": 2.345066358915176,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004458351490212803,
      "loss": 5.7106,
      "mean_token_accuracy": 0.18736932426691055,
      "num_tokens": 51503462.0,
      "step": 20320
    },
    {
      "entropy": 5.918054580688477,
      "epoch": 2.3456433929601848,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044580829218064964,
      "loss": 5.61,
      "mean_token_accuracy": 0.19015377312898635,
      "num_tokens": 51515338.0,
      "step": 20325
    },
    {
      "entropy": 5.928893375396728,
      "epoch": 2.346220427005193,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000445781429594967,
      "loss": 5.6189,
      "mean_token_accuracy": 0.1986027717590332,
      "num_tokens": 51527990.0,
      "step": 20330
    },
    {
      "entropy": 6.008607387542725,
      "epoch": 2.346797461050202,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004457545612651357,
      "loss": 5.6985,
      "mean_token_accuracy": 0.1866683468222618,
      "num_tokens": 51542316.0,
      "step": 20335
    },
    {
      "entropy": 5.9535891056060795,
      "epoch": 2.3473744950952105,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00044572768719205964,
      "loss": 5.6313,
      "mean_token_accuracy": 0.1888444185256958,
      "num_tokens": 51556268.0,
      "step": 20340
    },
    {
      "entropy": 6.074997138977051,
      "epoch": 2.3479515291402193,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044570080737664264,
      "loss": 5.7355,
      "mean_token_accuracy": 0.18668777793645858,
      "num_tokens": 51568511.0,
      "step": 20345
    },
    {
      "entropy": 6.028747415542602,
      "epoch": 2.3485285631852277,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00044567392181978874,
      "loss": 5.7158,
      "mean_token_accuracy": 0.18676182478666306,
      "num_tokens": 51581915.0,
      "step": 20350
    },
    {
      "entropy": 5.9520079612731935,
      "epoch": 2.3491055972302366,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044564703052240223,
      "loss": 5.5612,
      "mean_token_accuracy": 0.19253202229738237,
      "num_tokens": 51593544.0,
      "step": 20355
    },
    {
      "entropy": 5.918116712570191,
      "epoch": 2.349682631275245,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004456201334853876,
      "loss": 5.6559,
      "mean_token_accuracy": 0.19101679623126983,
      "num_tokens": 51605722.0,
      "step": 20360
    },
    {
      "entropy": 5.868543434143066,
      "epoch": 2.350259665320254,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044559323070964956,
      "loss": 5.5461,
      "mean_token_accuracy": 0.20295370370149612,
      "num_tokens": 51618277.0,
      "step": 20365
    },
    {
      "entropy": 6.011160898208618,
      "epoch": 2.3508366993652627,
      "grad_norm": 0.953125,
      "learning_rate": 0.000445566322196093,
      "loss": 5.6418,
      "mean_token_accuracy": 0.19960096031427382,
      "num_tokens": 51629960.0,
      "step": 20370
    },
    {
      "entropy": 5.8711200714111325,
      "epoch": 2.351413733410271,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004455394079456228,
      "loss": 5.5692,
      "mean_token_accuracy": 0.1938213050365448,
      "num_tokens": 51643089.0,
      "step": 20375
    },
    {
      "entropy": 5.994217395782471,
      "epoch": 2.35199076745528,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044551248795914446,
      "loss": 5.7413,
      "mean_token_accuracy": 0.18273805677890778,
      "num_tokens": 51655581.0,
      "step": 20380
    },
    {
      "entropy": 5.978893566131592,
      "epoch": 2.3525678015002884,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004454855622375632,
      "loss": 5.6208,
      "mean_token_accuracy": 0.18950185030698777,
      "num_tokens": 51667693.0,
      "step": 20385
    },
    {
      "entropy": 5.97436032295227,
      "epoch": 2.3531448355452973,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004454586307817848,
      "loss": 5.6914,
      "mean_token_accuracy": 0.18687188625335693,
      "num_tokens": 51680479.0,
      "step": 20390
    },
    {
      "entropy": 6.030154991149902,
      "epoch": 2.3537218695903057,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000445431693592715,
      "loss": 5.7447,
      "mean_token_accuracy": 0.1878482073545456,
      "num_tokens": 51693053.0,
      "step": 20395
    },
    {
      "entropy": 5.990267562866211,
      "epoch": 2.3542989036353146,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004454047506712598,
      "loss": 5.7178,
      "mean_token_accuracy": 0.18702870458364487,
      "num_tokens": 51704974.0,
      "step": 20400
    },
    {
      "entropy": 6.028567123413086,
      "epoch": 2.354875937680323,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044537780201832545,
      "loss": 5.6454,
      "mean_token_accuracy": 0.1914108067750931,
      "num_tokens": 51717190.0,
      "step": 20405
    },
    {
      "entropy": 5.994177436828613,
      "epoch": 2.355452971725332,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004453508476348184,
      "loss": 5.7097,
      "mean_token_accuracy": 0.1873905897140503,
      "num_tokens": 51730462.0,
      "step": 20410
    },
    {
      "entropy": 6.02954740524292,
      "epoch": 2.3560300057703403,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004453238875216452,
      "loss": 5.7303,
      "mean_token_accuracy": 0.18603793531656265,
      "num_tokens": 51743723.0,
      "step": 20415
    },
    {
      "entropy": 5.986447048187256,
      "epoch": 2.356607039815349,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004452969216797127,
      "loss": 5.6685,
      "mean_token_accuracy": 0.19376004338264466,
      "num_tokens": 51756959.0,
      "step": 20420
    },
    {
      "entropy": 5.95242075920105,
      "epoch": 2.357184073860358,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004452699501099277,
      "loss": 5.7002,
      "mean_token_accuracy": 0.18131199479103088,
      "num_tokens": 51768282.0,
      "step": 20425
    },
    {
      "entropy": 6.047074985504151,
      "epoch": 2.3577611079053664,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044524297281319766,
      "loss": 5.7013,
      "mean_token_accuracy": 0.19015496224164963,
      "num_tokens": 51780729.0,
      "step": 20430
    },
    {
      "entropy": 6.0384564876556395,
      "epoch": 2.358338141950375,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044521598979042963,
      "loss": 5.7791,
      "mean_token_accuracy": 0.18238887041807175,
      "num_tokens": 51793661.0,
      "step": 20435
    },
    {
      "entropy": 6.064259481430054,
      "epoch": 2.3589151759953837,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044518900104253154,
      "loss": 5.7154,
      "mean_token_accuracy": 0.18589272052049638,
      "num_tokens": 51806027.0,
      "step": 20440
    },
    {
      "entropy": 6.036914730072022,
      "epoch": 2.3594922100403926,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004451620065704108,
      "loss": 5.7688,
      "mean_token_accuracy": 0.1836748331785202,
      "num_tokens": 51817495.0,
      "step": 20445
    },
    {
      "entropy": 5.97684907913208,
      "epoch": 2.360069244085401,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044513500637497546,
      "loss": 5.7049,
      "mean_token_accuracy": 0.18662225902080537,
      "num_tokens": 51829948.0,
      "step": 20450
    },
    {
      "entropy": 5.99874529838562,
      "epoch": 2.36064627813041,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044510800045713373,
      "loss": 5.7746,
      "mean_token_accuracy": 0.18649423122406006,
      "num_tokens": 51842615.0,
      "step": 20455
    },
    {
      "entropy": 6.0690391063690186,
      "epoch": 2.3612233121754183,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044508098881779396,
      "loss": 5.661,
      "mean_token_accuracy": 0.19255159497261048,
      "num_tokens": 51853785.0,
      "step": 20460
    },
    {
      "entropy": 5.936944150924683,
      "epoch": 2.361800346220427,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004450539714578645,
      "loss": 5.5804,
      "mean_token_accuracy": 0.19726670533418655,
      "num_tokens": 51867427.0,
      "step": 20465
    },
    {
      "entropy": 6.077270126342773,
      "epoch": 2.3623773802654355,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004450269483782543,
      "loss": 5.8277,
      "mean_token_accuracy": 0.17756775468587876,
      "num_tokens": 51880241.0,
      "step": 20470
    },
    {
      "entropy": 5.987586498260498,
      "epoch": 2.3629544143104444,
      "grad_norm": 0.875,
      "learning_rate": 0.0004449999195798721,
      "loss": 5.668,
      "mean_token_accuracy": 0.19133693724870682,
      "num_tokens": 51891927.0,
      "step": 20475
    },
    {
      "entropy": 5.985398674011231,
      "epoch": 2.363531448355453,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044497288506362706,
      "loss": 5.7006,
      "mean_token_accuracy": 0.18982720524072647,
      "num_tokens": 51905633.0,
      "step": 20480
    },
    {
      "entropy": 6.013924646377563,
      "epoch": 2.3641084824004617,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004449458448304284,
      "loss": 5.7435,
      "mean_token_accuracy": 0.1776958927512169,
      "num_tokens": 51919398.0,
      "step": 20485
    },
    {
      "entropy": 5.865958595275879,
      "epoch": 2.36468551644547,
      "grad_norm": 0.921875,
      "learning_rate": 0.00044491879888118574,
      "loss": 5.4633,
      "mean_token_accuracy": 0.20870193541049958,
      "num_tokens": 51932732.0,
      "step": 20490
    },
    {
      "entropy": 5.928377294540406,
      "epoch": 2.365262550490479,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004448917472168087,
      "loss": 5.6044,
      "mean_token_accuracy": 0.19486753791570663,
      "num_tokens": 51945830.0,
      "step": 20495
    },
    {
      "entropy": 6.054198408126831,
      "epoch": 2.365839584535488,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044486468983820707,
      "loss": 5.6271,
      "mean_token_accuracy": 0.19218454658985137,
      "num_tokens": 51957643.0,
      "step": 20500
    },
    {
      "entropy": 5.96296010017395,
      "epoch": 2.3664166185804962,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000444837626746291,
      "loss": 5.6367,
      "mean_token_accuracy": 0.18630898743867874,
      "num_tokens": 51969840.0,
      "step": 20505
    },
    {
      "entropy": 5.968493413925171,
      "epoch": 2.366993652625505,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004448105579419707,
      "loss": 5.6933,
      "mean_token_accuracy": 0.18852551132440568,
      "num_tokens": 51981250.0,
      "step": 20510
    },
    {
      "entropy": 6.037579727172852,
      "epoch": 2.3675706866705135,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004447834834261567,
      "loss": 5.6925,
      "mean_token_accuracy": 0.1882827639579773,
      "num_tokens": 51993312.0,
      "step": 20515
    },
    {
      "entropy": 5.946306276321411,
      "epoch": 2.3681477207155224,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004447564031997595,
      "loss": 5.6519,
      "mean_token_accuracy": 0.19652391970157623,
      "num_tokens": 52006697.0,
      "step": 20520
    },
    {
      "entropy": 5.966462469100952,
      "epoch": 2.368724754760531,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044472931726369,
      "loss": 5.5906,
      "mean_token_accuracy": 0.19143429547548294,
      "num_tokens": 52018938.0,
      "step": 20525
    },
    {
      "entropy": 5.901692819595337,
      "epoch": 2.3693017888055397,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044470222561885926,
      "loss": 5.6065,
      "mean_token_accuracy": 0.20193671733140944,
      "num_tokens": 52031309.0,
      "step": 20530
    },
    {
      "entropy": 6.027193403244018,
      "epoch": 2.369878822850548,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00044467512826617845,
      "loss": 5.6615,
      "mean_token_accuracy": 0.1927194118499756,
      "num_tokens": 52043661.0,
      "step": 20535
    },
    {
      "entropy": 5.960782623291015,
      "epoch": 2.370455856895557,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044464802520655897,
      "loss": 5.6534,
      "mean_token_accuracy": 0.19271975457668306,
      "num_tokens": 52055049.0,
      "step": 20540
    },
    {
      "entropy": 5.892357730865479,
      "epoch": 2.3710328909405654,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004446209164409124,
      "loss": 5.665,
      "mean_token_accuracy": 0.1921023339033127,
      "num_tokens": 52068188.0,
      "step": 20545
    },
    {
      "entropy": 6.066678190231324,
      "epoch": 2.371609924985574,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004445938019701506,
      "loss": 5.7103,
      "mean_token_accuracy": 0.18981400430202483,
      "num_tokens": 52080662.0,
      "step": 20550
    },
    {
      "entropy": 6.009578466415405,
      "epoch": 2.3721869590305826,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004445666817951855,
      "loss": 5.7085,
      "mean_token_accuracy": 0.19093797206878663,
      "num_tokens": 52093088.0,
      "step": 20555
    },
    {
      "entropy": 5.9241838455200195,
      "epoch": 2.3727639930755915,
      "grad_norm": 1.0,
      "learning_rate": 0.0004445395559169293,
      "loss": 5.623,
      "mean_token_accuracy": 0.19211751371622085,
      "num_tokens": 52105796.0,
      "step": 20560
    },
    {
      "entropy": 5.99240288734436,
      "epoch": 2.3733410271206,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004445124243362943,
      "loss": 5.6653,
      "mean_token_accuracy": 0.1920988827943802,
      "num_tokens": 52118146.0,
      "step": 20565
    },
    {
      "entropy": 5.968817663192749,
      "epoch": 2.3739180611656088,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004444852870541932,
      "loss": 5.6306,
      "mean_token_accuracy": 0.19264880269765855,
      "num_tokens": 52131033.0,
      "step": 20570
    },
    {
      "entropy": 6.027396154403687,
      "epoch": 2.3744950952106176,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004444581440715386,
      "loss": 5.7118,
      "mean_token_accuracy": 0.18831726312637329,
      "num_tokens": 52143742.0,
      "step": 20575
    },
    {
      "entropy": 5.969668245315551,
      "epoch": 2.375072129255626,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044443099538924347,
      "loss": 5.5736,
      "mean_token_accuracy": 0.19735325127840042,
      "num_tokens": 52156091.0,
      "step": 20580
    },
    {
      "entropy": 6.001495122909546,
      "epoch": 2.375649163300635,
      "grad_norm": 1.0,
      "learning_rate": 0.0004444038410082211,
      "loss": 5.7296,
      "mean_token_accuracy": 0.18331396579742432,
      "num_tokens": 52168404.0,
      "step": 20585
    },
    {
      "entropy": 5.934076642990112,
      "epoch": 2.3762261973456433,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004443766809293846,
      "loss": 5.5968,
      "mean_token_accuracy": 0.20053293853998183,
      "num_tokens": 52180660.0,
      "step": 20590
    },
    {
      "entropy": 6.003503656387329,
      "epoch": 2.376803231390652,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004443495151536475,
      "loss": 5.6719,
      "mean_token_accuracy": 0.1982205852866173,
      "num_tokens": 52193752.0,
      "step": 20595
    },
    {
      "entropy": 6.0551975727081295,
      "epoch": 2.3773802654356606,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004443223436819237,
      "loss": 5.6954,
      "mean_token_accuracy": 0.18968099355697632,
      "num_tokens": 52205686.0,
      "step": 20600
    },
    {
      "entropy": 5.991528177261353,
      "epoch": 2.3779572994806695,
      "grad_norm": 0.921875,
      "learning_rate": 0.00044429516651512687,
      "loss": 5.6716,
      "mean_token_accuracy": 0.1905240759253502,
      "num_tokens": 52218743.0,
      "step": 20605
    },
    {
      "entropy": 6.02947506904602,
      "epoch": 2.378534333525678,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00044426798365417133,
      "loss": 5.5772,
      "mean_token_accuracy": 0.19006024301052094,
      "num_tokens": 52231760.0,
      "step": 20610
    },
    {
      "entropy": 5.990348482131958,
      "epoch": 2.3791113675706868,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00044424079509997104,
      "loss": 5.681,
      "mean_token_accuracy": 0.185404933989048,
      "num_tokens": 52243146.0,
      "step": 20615
    },
    {
      "entropy": 5.995209360122681,
      "epoch": 2.379688401615695,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004442136008534409,
      "loss": 5.7263,
      "mean_token_accuracy": 0.18622615039348603,
      "num_tokens": 52255330.0,
      "step": 20620
    },
    {
      "entropy": 6.031816196441651,
      "epoch": 2.380265435660704,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044418640091549525,
      "loss": 5.6798,
      "mean_token_accuracy": 0.19001232832670212,
      "num_tokens": 52266563.0,
      "step": 20625
    },
    {
      "entropy": 6.007872486114502,
      "epoch": 2.3808424697057124,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000444159195287049,
      "loss": 5.7445,
      "mean_token_accuracy": 0.18851658403873445,
      "num_tokens": 52279137.0,
      "step": 20630
    },
    {
      "entropy": 5.952458620071411,
      "epoch": 2.3814195037507213,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004441319839690173,
      "loss": 5.667,
      "mean_token_accuracy": 0.1959312900900841,
      "num_tokens": 52293317.0,
      "step": 20635
    },
    {
      "entropy": 6.045389175415039,
      "epoch": 2.3819965377957297,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004441047669623153,
      "loss": 5.7078,
      "mean_token_accuracy": 0.18266079276800157,
      "num_tokens": 52306383.0,
      "step": 20640
    },
    {
      "entropy": 5.958673048019409,
      "epoch": 2.3825735718407386,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044407754426785845,
      "loss": 5.5942,
      "mean_token_accuracy": 0.19515654146671296,
      "num_tokens": 52318446.0,
      "step": 20645
    },
    {
      "entropy": 6.003369951248169,
      "epoch": 2.3831506058857475,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004440503158865625,
      "loss": 5.6315,
      "mean_token_accuracy": 0.1883022129535675,
      "num_tokens": 52330761.0,
      "step": 20650
    },
    {
      "entropy": 6.044725179672241,
      "epoch": 2.383727639930756,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00044402308181934295,
      "loss": 5.7,
      "mean_token_accuracy": 0.1817420855164528,
      "num_tokens": 52342906.0,
      "step": 20655
    },
    {
      "entropy": 5.943409013748169,
      "epoch": 2.3843046739757647,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004439958420671162,
      "loss": 5.6472,
      "mean_token_accuracy": 0.19289316982030869,
      "num_tokens": 52355928.0,
      "step": 20660
    },
    {
      "entropy": 5.9810339450836185,
      "epoch": 2.384881708020773,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00044396859663079814,
      "loss": 5.6865,
      "mean_token_accuracy": 0.18732011020183564,
      "num_tokens": 52368370.0,
      "step": 20665
    },
    {
      "entropy": 6.085014867782593,
      "epoch": 2.385458742065782,
      "grad_norm": 0.890625,
      "learning_rate": 0.00044394134551130533,
      "loss": 5.7468,
      "mean_token_accuracy": 0.17752742171287536,
      "num_tokens": 52382636.0,
      "step": 20670
    },
    {
      "entropy": 5.990962123870849,
      "epoch": 2.3860357761107904,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004439140887095542,
      "loss": 5.6968,
      "mean_token_accuracy": 0.19116786122322083,
      "num_tokens": 52395745.0,
      "step": 20675
    },
    {
      "entropy": 5.966981983184814,
      "epoch": 2.3866128101557993,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044388682622646165,
      "loss": 5.6104,
      "mean_token_accuracy": 0.19393228739500046,
      "num_tokens": 52408929.0,
      "step": 20680
    },
    {
      "entropy": 5.899444723129273,
      "epoch": 2.3871898442008077,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004438595580629446,
      "loss": 5.5965,
      "mean_token_accuracy": 0.1946858286857605,
      "num_tokens": 52421626.0,
      "step": 20685
    },
    {
      "entropy": 6.099765396118164,
      "epoch": 2.3877668782458166,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004438322842199202,
      "loss": 5.7658,
      "mean_token_accuracy": 0.1797529324889183,
      "num_tokens": 52434139.0,
      "step": 20690
    },
    {
      "entropy": 5.967930173873901,
      "epoch": 2.388343912290825,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004438050046983057,
      "loss": 5.6266,
      "mean_token_accuracy": 0.19512900114059448,
      "num_tokens": 52446141.0,
      "step": 20695
    },
    {
      "entropy": 5.942644643783569,
      "epoch": 2.388920946335834,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00044377771949901876,
      "loss": 5.6565,
      "mean_token_accuracy": 0.19210680574178696,
      "num_tokens": 52459110.0,
      "step": 20700
    },
    {
      "entropy": 5.999210023880005,
      "epoch": 2.3894979803808427,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00044375042862297703,
      "loss": 5.7698,
      "mean_token_accuracy": 0.18338969945907593,
      "num_tokens": 52471985.0,
      "step": 20705
    },
    {
      "entropy": 6.060536193847656,
      "epoch": 2.390075014425851,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00044372313207109856,
      "loss": 5.7011,
      "mean_token_accuracy": 0.19356985986232758,
      "num_tokens": 52484226.0,
      "step": 20710
    },
    {
      "entropy": 5.972516775131226,
      "epoch": 2.3906520484708595,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044369582984430127,
      "loss": 5.6298,
      "mean_token_accuracy": 0.19238045960664749,
      "num_tokens": 52496141.0,
      "step": 20715
    },
    {
      "entropy": 6.003650140762329,
      "epoch": 2.3912290825158684,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044366852194350354,
      "loss": 5.6857,
      "mean_token_accuracy": 0.18424241244792938,
      "num_tokens": 52509231.0,
      "step": 20720
    },
    {
      "entropy": 5.959863662719727,
      "epoch": 2.3918061165608773,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004436412083696239,
      "loss": 5.6339,
      "mean_token_accuracy": 0.19150403439998626,
      "num_tokens": 52522619.0,
      "step": 20725
    },
    {
      "entropy": 6.037461853027343,
      "epoch": 2.3923831506058857,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044361388912358095,
      "loss": 5.6896,
      "mean_token_accuracy": 0.1898658439517021,
      "num_tokens": 52534241.0,
      "step": 20730
    },
    {
      "entropy": 6.105290365219116,
      "epoch": 2.3929601846508946,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004435865642062936,
      "loss": 5.6813,
      "mean_token_accuracy": 0.18596949130296708,
      "num_tokens": 52546306.0,
      "step": 20735
    },
    {
      "entropy": 6.042927551269531,
      "epoch": 2.393537218695903,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004435592336186809,
      "loss": 5.7512,
      "mean_token_accuracy": 0.17757801711559296,
      "num_tokens": 52559855.0,
      "step": 20740
    },
    {
      "entropy": 5.950037145614624,
      "epoch": 2.394114252740912,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004435318973616622,
      "loss": 5.565,
      "mean_token_accuracy": 0.2020915448665619,
      "num_tokens": 52574071.0,
      "step": 20745
    },
    {
      "entropy": 5.934414720535278,
      "epoch": 2.3946912867859202,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044350455543615665,
      "loss": 5.6362,
      "mean_token_accuracy": 0.19383638948202134,
      "num_tokens": 52585983.0,
      "step": 20750
    },
    {
      "entropy": 5.996983575820923,
      "epoch": 2.395268320830929,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004434772078430843,
      "loss": 5.5724,
      "mean_token_accuracy": 0.19496240317821503,
      "num_tokens": 52598399.0,
      "step": 20755
    },
    {
      "entropy": 6.021268272399903,
      "epoch": 2.3958453548759375,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004434498545833646,
      "loss": 5.6866,
      "mean_token_accuracy": 0.189863383769989,
      "num_tokens": 52610172.0,
      "step": 20760
    },
    {
      "entropy": 5.991119432449341,
      "epoch": 2.3964223889209464,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004434224956579177,
      "loss": 5.6716,
      "mean_token_accuracy": 0.19683381766080857,
      "num_tokens": 52622626.0,
      "step": 20765
    },
    {
      "entropy": 6.002688789367676,
      "epoch": 2.396999422965955,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004433951310676639,
      "loss": 5.6617,
      "mean_token_accuracy": 0.18929496556520461,
      "num_tokens": 52633893.0,
      "step": 20770
    },
    {
      "entropy": 6.033216524124145,
      "epoch": 2.3975764570109637,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00044336776081352347,
      "loss": 5.6339,
      "mean_token_accuracy": 0.19052568674087525,
      "num_tokens": 52645674.0,
      "step": 20775
    },
    {
      "entropy": 6.044572830200195,
      "epoch": 2.3981534910559725,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044334038489641706,
      "loss": 5.7362,
      "mean_token_accuracy": 0.18531822562217712,
      "num_tokens": 52659030.0,
      "step": 20780
    },
    {
      "entropy": 5.965080213546753,
      "epoch": 2.398730525100981,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044331300331726544,
      "loss": 5.6345,
      "mean_token_accuracy": 0.19408441483974456,
      "num_tokens": 52670680.0,
      "step": 20785
    },
    {
      "entropy": 6.017557430267334,
      "epoch": 2.3993075591459894,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044328561607698947,
      "loss": 5.7733,
      "mean_token_accuracy": 0.18214710801839828,
      "num_tokens": 52682860.0,
      "step": 20790
    },
    {
      "entropy": 5.949009275436401,
      "epoch": 2.3998845931909982,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004432582231765105,
      "loss": 5.6138,
      "mean_token_accuracy": 0.19202758818864823,
      "num_tokens": 52695566.0,
      "step": 20795
    },
    {
      "entropy": 5.979013967514038,
      "epoch": 2.400461627236007,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044323082461674974,
      "loss": 5.5822,
      "mean_token_accuracy": 0.1971833422780037,
      "num_tokens": 52708409.0,
      "step": 20800
    },
    {
      "entropy": 5.9186173439025875,
      "epoch": 2.4010386612810155,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004432034203986287,
      "loss": 5.6376,
      "mean_token_accuracy": 0.19338284730911254,
      "num_tokens": 52721637.0,
      "step": 20805
    },
    {
      "entropy": 5.932637500762939,
      "epoch": 2.4016156953260244,
      "grad_norm": 1.0,
      "learning_rate": 0.0004431760105230693,
      "loss": 5.6866,
      "mean_token_accuracy": 0.1851627305150032,
      "num_tokens": 52732157.0,
      "step": 20810
    },
    {
      "entropy": 5.992417621612549,
      "epoch": 2.402192729371033,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00044314859499099325,
      "loss": 5.696,
      "mean_token_accuracy": 0.18770527988672256,
      "num_tokens": 52745038.0,
      "step": 20815
    },
    {
      "entropy": 6.063066387176514,
      "epoch": 2.4027697634160416,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044312117380332274,
      "loss": 5.6726,
      "mean_token_accuracy": 0.19015658348798753,
      "num_tokens": 52757381.0,
      "step": 20820
    },
    {
      "entropy": 5.92491340637207,
      "epoch": 2.40334679746105,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044309374696098,
      "loss": 5.6013,
      "mean_token_accuracy": 0.2000294417142868,
      "num_tokens": 52770604.0,
      "step": 20825
    },
    {
      "entropy": 5.913321208953858,
      "epoch": 2.403923831506059,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004430663144648876,
      "loss": 5.6041,
      "mean_token_accuracy": 0.20210540443658828,
      "num_tokens": 52784634.0,
      "step": 20830
    },
    {
      "entropy": 5.95510311126709,
      "epoch": 2.4045008655510673,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00044303887631596823,
      "loss": 5.6481,
      "mean_token_accuracy": 0.19303112626075744,
      "num_tokens": 52798417.0,
      "step": 20835
    },
    {
      "entropy": 6.080182027816773,
      "epoch": 2.405077899596076,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004430114325151447,
      "loss": 5.7997,
      "mean_token_accuracy": 0.1815366640686989,
      "num_tokens": 52810885.0,
      "step": 20840
    },
    {
      "entropy": 5.9650794506073,
      "epoch": 2.4056549336410846,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004429839830633401,
      "loss": 5.6428,
      "mean_token_accuracy": 0.19498305767774582,
      "num_tokens": 52824450.0,
      "step": 20845
    },
    {
      "entropy": 5.997240209579468,
      "epoch": 2.4062319676860935,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004429565279614777,
      "loss": 5.676,
      "mean_token_accuracy": 0.18849873989820481,
      "num_tokens": 52837053.0,
      "step": 20850
    },
    {
      "entropy": 6.011818695068359,
      "epoch": 2.4068090017311023,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00044292906721048094,
      "loss": 5.6896,
      "mean_token_accuracy": 0.18508405834436417,
      "num_tokens": 52851695.0,
      "step": 20855
    },
    {
      "entropy": 6.06520791053772,
      "epoch": 2.4073860357761108,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004429016008112733,
      "loss": 5.6609,
      "mean_token_accuracy": 0.18886839151382445,
      "num_tokens": 52862999.0,
      "step": 20860
    },
    {
      "entropy": 5.9042871475219725,
      "epoch": 2.4079630698211196,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004428741287647788,
      "loss": 5.6087,
      "mean_token_accuracy": 0.198186057806015,
      "num_tokens": 52875224.0,
      "step": 20865
    },
    {
      "entropy": 5.982259273529053,
      "epoch": 2.408540103866128,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00044284665107192136,
      "loss": 5.6902,
      "mean_token_accuracy": 0.18410100191831588,
      "num_tokens": 52887296.0,
      "step": 20870
    },
    {
      "entropy": 5.994974040985108,
      "epoch": 2.409117137911137,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004428191677336251,
      "loss": 5.5888,
      "mean_token_accuracy": 0.1945437788963318,
      "num_tokens": 52900670.0,
      "step": 20875
    },
    {
      "entropy": 6.067313241958618,
      "epoch": 2.4096941719561453,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004427916787508146,
      "loss": 5.7801,
      "mean_token_accuracy": 0.18830416053533555,
      "num_tokens": 52914515.0,
      "step": 20880
    },
    {
      "entropy": 5.909205198287964,
      "epoch": 2.410271206001154,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004427641841244144,
      "loss": 5.5797,
      "mean_token_accuracy": 0.19459755271673201,
      "num_tokens": 52928448.0,
      "step": 20885
    },
    {
      "entropy": 6.002679634094238,
      "epoch": 2.4108482400461626,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000442736683855349,
      "loss": 5.6594,
      "mean_token_accuracy": 0.1948203593492508,
      "num_tokens": 52941381.0,
      "step": 20890
    },
    {
      "entropy": 5.910699033737183,
      "epoch": 2.4114252740911715,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004427091779445437,
      "loss": 5.5971,
      "mean_token_accuracy": 0.18868958204984665,
      "num_tokens": 52954959.0,
      "step": 20895
    },
    {
      "entropy": 6.021023464202881,
      "epoch": 2.41200230813618,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004426816663929235,
      "loss": 5.65,
      "mean_token_accuracy": 0.1949334278702736,
      "num_tokens": 52967956.0,
      "step": 20900
    },
    {
      "entropy": 6.022794961929321,
      "epoch": 2.4125793421811887,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044265414920141366,
      "loss": 5.6982,
      "mean_token_accuracy": 0.19125588983297348,
      "num_tokens": 52980997.0,
      "step": 20905
    },
    {
      "entropy": 6.046177625656128,
      "epoch": 2.413156376226197,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00044262662637093987,
      "loss": 5.694,
      "mean_token_accuracy": 0.1861991360783577,
      "num_tokens": 52994636.0,
      "step": 20910
    },
    {
      "entropy": 5.975221252441406,
      "epoch": 2.413733410271206,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044259909790242776,
      "loss": 5.6318,
      "mean_token_accuracy": 0.1857071414589882,
      "num_tokens": 53007065.0,
      "step": 20915
    },
    {
      "entropy": 5.983144903182984,
      "epoch": 2.4143104443162144,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004425715637968032,
      "loss": 5.6376,
      "mean_token_accuracy": 0.18444152921438217,
      "num_tokens": 53020479.0,
      "step": 20920
    },
    {
      "entropy": 6.009814691543579,
      "epoch": 2.4148874783612233,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004425440240549923,
      "loss": 5.7273,
      "mean_token_accuracy": 0.18924343585968018,
      "num_tokens": 53036390.0,
      "step": 20925
    },
    {
      "entropy": 5.986808347702026,
      "epoch": 2.415464512406232,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044251647867792147,
      "loss": 5.6219,
      "mean_token_accuracy": 0.19008978754281997,
      "num_tokens": 53048604.0,
      "step": 20930
    },
    {
      "entropy": 6.017729616165161,
      "epoch": 2.4160415464512406,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044248892766651706,
      "loss": 5.6807,
      "mean_token_accuracy": 0.19031698852777482,
      "num_tokens": 53062774.0,
      "step": 20935
    },
    {
      "entropy": 6.001942300796509,
      "epoch": 2.4166185804962494,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004424613710217057,
      "loss": 5.67,
      "mean_token_accuracy": 0.18617767095565796,
      "num_tokens": 53076701.0,
      "step": 20940
    },
    {
      "entropy": 6.0974773406982425,
      "epoch": 2.417195614541258,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044243380874441437,
      "loss": 5.7232,
      "mean_token_accuracy": 0.19026297181844712,
      "num_tokens": 53089950.0,
      "step": 20945
    },
    {
      "entropy": 5.920363998413086,
      "epoch": 2.4177726485862667,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00044240624083557,
      "loss": 5.5557,
      "mean_token_accuracy": 0.19442303329706193,
      "num_tokens": 53103622.0,
      "step": 20950
    },
    {
      "entropy": 6.0251014709472654,
      "epoch": 2.418349682631275,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044237866729609994,
      "loss": 5.7605,
      "mean_token_accuracy": 0.19088042974472047,
      "num_tokens": 53116382.0,
      "step": 20955
    },
    {
      "entropy": 6.078667783737183,
      "epoch": 2.418926716676284,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004423510881269315,
      "loss": 5.6781,
      "mean_token_accuracy": 0.19093145728111266,
      "num_tokens": 53128759.0,
      "step": 20960
    },
    {
      "entropy": 6.007024526596069,
      "epoch": 2.4195037507212924,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004423235033289924,
      "loss": 5.6647,
      "mean_token_accuracy": 0.1980356603860855,
      "num_tokens": 53141131.0,
      "step": 20965
    },
    {
      "entropy": 5.973045921325683,
      "epoch": 2.4200807847663013,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004422959129032103,
      "loss": 5.6915,
      "mean_token_accuracy": 0.18837860673666001,
      "num_tokens": 53152486.0,
      "step": 20970
    },
    {
      "entropy": 5.956928539276123,
      "epoch": 2.4206578188113097,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044226831685051333,
      "loss": 5.5933,
      "mean_token_accuracy": 0.19217808097600936,
      "num_tokens": 53164897.0,
      "step": 20975
    },
    {
      "entropy": 6.035703229904175,
      "epoch": 2.4212348528563186,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004422407151718296,
      "loss": 5.7471,
      "mean_token_accuracy": 0.18574887216091157,
      "num_tokens": 53177759.0,
      "step": 20980
    },
    {
      "entropy": 6.077817106246949,
      "epoch": 2.4218118869013274,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044221310786808746,
      "loss": 5.7199,
      "mean_token_accuracy": 0.19029134064912795,
      "num_tokens": 53191985.0,
      "step": 20985
    },
    {
      "entropy": 5.937718915939331,
      "epoch": 2.422388920946336,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004421854949402155,
      "loss": 5.5452,
      "mean_token_accuracy": 0.19829166829586028,
      "num_tokens": 53203531.0,
      "step": 20990
    },
    {
      "entropy": 6.001355075836182,
      "epoch": 2.4229659549913443,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044215787638914245,
      "loss": 5.655,
      "mean_token_accuracy": 0.18430670648813247,
      "num_tokens": 53216440.0,
      "step": 20995
    },
    {
      "entropy": 6.047285556793213,
      "epoch": 2.423542989036353,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004421302522157973,
      "loss": 5.7208,
      "mean_token_accuracy": 0.17981597185134887,
      "num_tokens": 53230674.0,
      "step": 21000
    },
    {
      "epoch": 2.423542989036353,
      "eval_entropy": 5.851141214167357,
      "eval_loss": 5.831363201141357,
      "eval_mean_token_accuracy": 0.18799451582466367,
      "eval_num_tokens": 53230674.0,
      "eval_runtime": 22.7472,
      "eval_samples_per_second": 1236.899,
      "eval_steps_per_second": 154.612,
      "step": 21000
    }
  ],
  "logging_steps": 5,
  "max_steps": 86650,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 3000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 7.8499687550976e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}