{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.263574064312071,
  "eval_steps": 500,
  "global_step": 2500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 0.8691443920135498,
      "epoch": 0.021085925144965736,
      "grad_norm": 47.5,
      "learning_rate": 7.2e-07,
      "loss": 2.1268569946289064,
      "mean_token_accuracy": 0.5856592752039432,
      "num_tokens": 53385.0,
      "step": 10
    },
    {
      "entropy": 0.9018128983676433,
      "epoch": 0.04217185028993147,
      "grad_norm": 49.25,
      "learning_rate": 1.52e-06,
      "loss": 2.1235822677612304,
      "mean_token_accuracy": 0.5826416034251451,
      "num_tokens": 108949.0,
      "step": 20
    },
    {
      "entropy": 0.9313402399420738,
      "epoch": 0.0632577754348972,
      "grad_norm": 38.25,
      "learning_rate": 2.3200000000000002e-06,
      "loss": 1.9997358322143555,
      "mean_token_accuracy": 0.5817345686256885,
      "num_tokens": 162034.0,
      "step": 30
    },
    {
      "entropy": 0.9819381944835186,
      "epoch": 0.08434370057986294,
      "grad_norm": 28.25,
      "learning_rate": 3.12e-06,
      "loss": 1.8760208129882812,
      "mean_token_accuracy": 0.5874959267675877,
      "num_tokens": 214165.0,
      "step": 40
    },
    {
      "entropy": 1.0624087303876877,
      "epoch": 0.10542962572482868,
      "grad_norm": 14.1875,
      "learning_rate": 3.920000000000001e-06,
      "loss": 1.6169570922851562,
      "mean_token_accuracy": 0.601640059798956,
      "num_tokens": 265527.0,
      "step": 50
    },
    {
      "entropy": 1.1721510097384453,
      "epoch": 0.1265155508697944,
      "grad_norm": 11.3125,
      "learning_rate": 4.7200000000000005e-06,
      "loss": 1.4622878074645995,
      "mean_token_accuracy": 0.6209174185991287,
      "num_tokens": 319341.0,
      "step": 60
    },
    {
      "entropy": 1.2009219758212566,
      "epoch": 0.14760147601476015,
      "grad_norm": 9.0625,
      "learning_rate": 5.5200000000000005e-06,
      "loss": 1.2933337211608886,
      "mean_token_accuracy": 0.6467074818909169,
      "num_tokens": 373709.0,
      "step": 70
    },
    {
      "entropy": 1.1439887225627898,
      "epoch": 0.16868740115972589,
      "grad_norm": 9.3125,
      "learning_rate": 6.3200000000000005e-06,
      "loss": 1.2084498405456543,
      "mean_token_accuracy": 0.6660686552524566,
      "num_tokens": 425923.0,
      "step": 80
    },
    {
      "entropy": 1.0713621236383915,
      "epoch": 0.18977332630469163,
      "grad_norm": 7.84375,
      "learning_rate": 7.1200000000000004e-06,
      "loss": 1.1175435066223145,
      "mean_token_accuracy": 0.6812180563807487,
      "num_tokens": 480012.0,
      "step": 90
    },
    {
      "entropy": 1.0369222231209279,
      "epoch": 0.21085925144965736,
      "grad_norm": 7.09375,
      "learning_rate": 7.92e-06,
      "loss": 1.062850570678711,
      "mean_token_accuracy": 0.6899570092558861,
      "num_tokens": 533890.0,
      "step": 100
    },
    {
      "entropy": 0.9894454136490822,
      "epoch": 0.23194517659462308,
      "grad_norm": 7.75,
      "learning_rate": 8.720000000000001e-06,
      "loss": 1.0150799751281738,
      "mean_token_accuracy": 0.7009239844977856,
      "num_tokens": 590654.0,
      "step": 110
    },
    {
      "entropy": 0.9410129584372043,
      "epoch": 0.2530311017395888,
      "grad_norm": 7.28125,
      "learning_rate": 9.52e-06,
      "loss": 0.9810515403747558,
      "mean_token_accuracy": 0.709618617594242,
      "num_tokens": 643080.0,
      "step": 120
    },
    {
      "entropy": 0.9265115469694137,
      "epoch": 0.2741170268845546,
      "grad_norm": 7.46875,
      "learning_rate": 1.0320000000000001e-05,
      "loss": 0.9485825538635254,
      "mean_token_accuracy": 0.7162063598632813,
      "num_tokens": 697254.0,
      "step": 130
    },
    {
      "entropy": 0.9009825974702835,
      "epoch": 0.2952029520295203,
      "grad_norm": 6.96875,
      "learning_rate": 1.1120000000000002e-05,
      "loss": 0.9119878768920898,
      "mean_token_accuracy": 0.7238413639366627,
      "num_tokens": 751774.0,
      "step": 140
    },
    {
      "entropy": 0.9040703341364861,
      "epoch": 0.316288877174486,
      "grad_norm": 7.21875,
      "learning_rate": 1.1920000000000001e-05,
      "loss": 0.9283308029174805,
      "mean_token_accuracy": 0.7166316144168376,
      "num_tokens": 808927.0,
      "step": 150
    },
    {
      "entropy": 0.8878034949302673,
      "epoch": 0.33737480231945177,
      "grad_norm": 7.375,
      "learning_rate": 1.2720000000000002e-05,
      "loss": 0.9068665504455566,
      "mean_token_accuracy": 0.7254431843757629,
      "num_tokens": 858559.0,
      "step": 160
    },
    {
      "entropy": 0.887480016052723,
      "epoch": 0.3584607274644175,
      "grad_norm": 7.65625,
      "learning_rate": 1.3520000000000003e-05,
      "loss": 0.8971159934997559,
      "mean_token_accuracy": 0.7258390121161937,
      "num_tokens": 912285.0,
      "step": 170
    },
    {
      "entropy": 0.8527634374797344,
      "epoch": 0.37954665260938325,
      "grad_norm": 7.6875,
      "learning_rate": 1.432e-05,
      "loss": 0.8586221694946289,
      "mean_token_accuracy": 0.7359919100999832,
      "num_tokens": 969385.0,
      "step": 180
    },
    {
      "entropy": 0.8445627830922604,
      "epoch": 0.40063257775434896,
      "grad_norm": 7.03125,
      "learning_rate": 1.5120000000000001e-05,
      "loss": 0.855714225769043,
      "mean_token_accuracy": 0.7341729424893856,
      "num_tokens": 1026106.0,
      "step": 190
    },
    {
      "entropy": 0.841130618005991,
      "epoch": 0.42171850289931473,
      "grad_norm": 7.21875,
      "learning_rate": 1.5920000000000003e-05,
      "loss": 0.8620488166809082,
      "mean_token_accuracy": 0.7361006513237953,
      "num_tokens": 1079743.0,
      "step": 200
    },
    {
      "entropy": 0.831258487701416,
      "epoch": 0.44280442804428044,
      "grad_norm": 7.0,
      "learning_rate": 1.672e-05,
      "loss": 0.8441590309143067,
      "mean_token_accuracy": 0.7379129216074943,
      "num_tokens": 1131453.0,
      "step": 210
    },
    {
      "entropy": 0.79731200709939,
      "epoch": 0.46389035318924615,
      "grad_norm": 7.875,
      "learning_rate": 1.752e-05,
      "loss": 0.8135190963745117,
      "mean_token_accuracy": 0.7497627973556519,
      "num_tokens": 1182486.0,
      "step": 220
    },
    {
      "entropy": 0.8269097700715065,
      "epoch": 0.4849762783342119,
      "grad_norm": 7.5625,
      "learning_rate": 1.832e-05,
      "loss": 0.8399410247802734,
      "mean_token_accuracy": 0.7389348305761814,
      "num_tokens": 1238136.0,
      "step": 230
    },
    {
      "entropy": 0.8093138128519058,
      "epoch": 0.5060622034791776,
      "grad_norm": 7.40625,
      "learning_rate": 1.912e-05,
      "loss": 0.810025691986084,
      "mean_token_accuracy": 0.7460841409862041,
      "num_tokens": 1294939.0,
      "step": 240
    },
    {
      "entropy": 0.8069386176764966,
      "epoch": 0.5271481286241434,
      "grad_norm": 6.84375,
      "learning_rate": 1.9920000000000002e-05,
      "loss": 0.8215888023376465,
      "mean_token_accuracy": 0.7438369721174241,
      "num_tokens": 1346462.0,
      "step": 250
    },
    {
      "entropy": 0.8088659413158894,
      "epoch": 0.5482340537691092,
      "grad_norm": 6.75,
      "learning_rate": 1.9920000000000002e-05,
      "loss": 0.8201449394226075,
      "mean_token_accuracy": 0.7393839374184609,
      "num_tokens": 1397444.0,
      "step": 260
    },
    {
      "entropy": 0.7969542622566224,
      "epoch": 0.5693199789140748,
      "grad_norm": 6.59375,
      "learning_rate": 1.9831111111111113e-05,
      "loss": 0.8009759902954101,
      "mean_token_accuracy": 0.7512488305568695,
      "num_tokens": 1453900.0,
      "step": 270
    },
    {
      "entropy": 0.780416651815176,
      "epoch": 0.5904059040590406,
      "grad_norm": 7.21875,
      "learning_rate": 1.9742222222222223e-05,
      "loss": 0.7874263763427735,
      "mean_token_accuracy": 0.7526524804532528,
      "num_tokens": 1505529.0,
      "step": 280
    },
    {
      "entropy": 0.7752278268337249,
      "epoch": 0.6114918292040064,
      "grad_norm": 6.875,
      "learning_rate": 1.9653333333333334e-05,
      "loss": 0.8075678825378418,
      "mean_token_accuracy": 0.746569037437439,
      "num_tokens": 1562473.0,
      "step": 290
    },
    {
      "entropy": 0.7895354829728604,
      "epoch": 0.632577754348972,
      "grad_norm": 6.6875,
      "learning_rate": 1.9564444444444444e-05,
      "loss": 0.7880948543548584,
      "mean_token_accuracy": 0.7494045957922936,
      "num_tokens": 1613320.0,
      "step": 300
    },
    {
      "entropy": 0.7653828673064709,
      "epoch": 0.6536636794939378,
      "grad_norm": 6.78125,
      "learning_rate": 1.9475555555555558e-05,
      "loss": 0.7780828475952148,
      "mean_token_accuracy": 0.7556171745061875,
      "num_tokens": 1662847.0,
      "step": 310
    },
    {
      "entropy": 0.7652887552976608,
      "epoch": 0.6747496046389035,
      "grad_norm": 6.375,
      "learning_rate": 1.938666666666667e-05,
      "loss": 0.7539787292480469,
      "mean_token_accuracy": 0.7582654722034932,
      "num_tokens": 1718553.0,
      "step": 320
    },
    {
      "entropy": 0.7076033115386963,
      "epoch": 0.6958355297838693,
      "grad_norm": 7.15625,
      "learning_rate": 1.929777777777778e-05,
      "loss": 0.7396605491638184,
      "mean_token_accuracy": 0.7659220136702061,
      "num_tokens": 1771024.0,
      "step": 330
    },
    {
      "entropy": 0.7509547933936119,
      "epoch": 0.716921454928835,
      "grad_norm": 6.15625,
      "learning_rate": 1.920888888888889e-05,
      "loss": 0.7459388256072998,
      "mean_token_accuracy": 0.7644543245434761,
      "num_tokens": 1821303.0,
      "step": 340
    },
    {
      "entropy": 0.7591827407479286,
      "epoch": 0.7380073800738007,
      "grad_norm": 7.03125,
      "learning_rate": 1.912e-05,
      "loss": 0.7627717971801757,
      "mean_token_accuracy": 0.7567056350409984,
      "num_tokens": 1868674.0,
      "step": 350
    },
    {
      "entropy": 0.7316406652331352,
      "epoch": 0.7590933052187665,
      "grad_norm": 6.59375,
      "learning_rate": 1.9031111111111114e-05,
      "loss": 0.7425707340240478,
      "mean_token_accuracy": 0.7653280735015869,
      "num_tokens": 1924619.0,
      "step": 360
    },
    {
      "entropy": 0.7539067789912224,
      "epoch": 0.7801792303637322,
      "grad_norm": 7.0625,
      "learning_rate": 1.8942222222222224e-05,
      "loss": 0.7716987609863282,
      "mean_token_accuracy": 0.7524559155106545,
      "num_tokens": 1978795.0,
      "step": 370
    },
    {
      "entropy": 0.7687754213809967,
      "epoch": 0.8012651555086979,
      "grad_norm": 7.1875,
      "learning_rate": 1.8853333333333335e-05,
      "loss": 0.7734373569488525,
      "mean_token_accuracy": 0.7497503250837326,
      "num_tokens": 2029994.0,
      "step": 380
    },
    {
      "entropy": 0.7508607231080532,
      "epoch": 0.8223510806536637,
      "grad_norm": 7.09375,
      "learning_rate": 1.8764444444444445e-05,
      "loss": 0.7575721740722656,
      "mean_token_accuracy": 0.7572125047445297,
      "num_tokens": 2082653.0,
      "step": 390
    },
    {
      "entropy": 0.7252861328423024,
      "epoch": 0.8434370057986295,
      "grad_norm": 6.5625,
      "learning_rate": 1.867555555555556e-05,
      "loss": 0.7381186485290527,
      "mean_token_accuracy": 0.7612978041172027,
      "num_tokens": 2135276.0,
      "step": 400
    },
    {
      "entropy": 0.7339485622942448,
      "epoch": 0.8645229309435951,
      "grad_norm": 6.625,
      "learning_rate": 1.858666666666667e-05,
      "loss": 0.7348906993865967,
      "mean_token_accuracy": 0.7663561321794987,
      "num_tokens": 2187466.0,
      "step": 410
    },
    {
      "entropy": 0.7272624362260103,
      "epoch": 0.8856088560885609,
      "grad_norm": 6.90625,
      "learning_rate": 1.849777777777778e-05,
      "loss": 0.749025821685791,
      "mean_token_accuracy": 0.7617379955947399,
      "num_tokens": 2237434.0,
      "step": 420
    },
    {
      "entropy": 0.7477380312979222,
      "epoch": 0.9066947812335266,
      "grad_norm": 6.34375,
      "learning_rate": 1.840888888888889e-05,
      "loss": 0.7498275756835937,
      "mean_token_accuracy": 0.7604753144085408,
      "num_tokens": 2288985.0,
      "step": 430
    },
    {
      "entropy": 0.7346535548567772,
      "epoch": 0.9277807063784923,
      "grad_norm": 6.9375,
      "learning_rate": 1.832e-05,
      "loss": 0.739972448348999,
      "mean_token_accuracy": 0.7658773727715016,
      "num_tokens": 2341187.0,
      "step": 440
    },
    {
      "entropy": 0.7167556572705507,
      "epoch": 0.9488666315234581,
      "grad_norm": 6.4375,
      "learning_rate": 1.823111111111111e-05,
      "loss": 0.7350784778594971,
      "mean_token_accuracy": 0.7644179269671441,
      "num_tokens": 2393385.0,
      "step": 450
    },
    {
      "entropy": 0.7537519961595536,
      "epoch": 0.9699525566684238,
      "grad_norm": 7.28125,
      "learning_rate": 1.814222222222222e-05,
      "loss": 0.757652759552002,
      "mean_token_accuracy": 0.7587068513035774,
      "num_tokens": 2442248.0,
      "step": 460
    },
    {
      "entropy": 0.72038439437747,
      "epoch": 0.9910384818133896,
      "grad_norm": 7.40625,
      "learning_rate": 1.8053333333333332e-05,
      "loss": 0.7361824035644531,
      "mean_token_accuracy": 0.7609050884842873,
      "num_tokens": 2498394.0,
      "step": 470
    },
    {
      "entropy": 0.735570513718837,
      "epoch": 1.0105429625724829,
      "grad_norm": 6.875,
      "learning_rate": 1.7964444444444446e-05,
      "loss": 0.730868148803711,
      "mean_token_accuracy": 0.7710521068121936,
      "num_tokens": 2545422.0,
      "step": 480
    },
    {
      "entropy": 0.6775223582983017,
      "epoch": 1.0316288877174487,
      "grad_norm": 7.125,
      "learning_rate": 1.7875555555555556e-05,
      "loss": 0.6740085124969483,
      "mean_token_accuracy": 0.7795436836779117,
      "num_tokens": 2597144.0,
      "step": 490
    },
    {
      "entropy": 0.6628748029470444,
      "epoch": 1.0527148128624144,
      "grad_norm": 6.84375,
      "learning_rate": 1.7786666666666667e-05,
      "loss": 0.6774313926696778,
      "mean_token_accuracy": 0.7789938881993294,
      "num_tokens": 2651938.0,
      "step": 500
    },
    {
      "entropy": 0.6695188779383898,
      "epoch": 1.07380073800738,
      "grad_norm": 7.0,
      "learning_rate": 1.7697777777777777e-05,
      "loss": 0.664512825012207,
      "mean_token_accuracy": 0.7832319870591163,
      "num_tokens": 2704948.0,
      "step": 510
    },
    {
      "entropy": 0.6812492057681083,
      "epoch": 1.0948866631523457,
      "grad_norm": 7.0,
      "learning_rate": 1.760888888888889e-05,
      "loss": 0.7036191940307617,
      "mean_token_accuracy": 0.7707728065550328,
      "num_tokens": 2757576.0,
      "step": 520
    },
    {
      "entropy": 0.6861097872257232,
      "epoch": 1.1159725882973115,
      "grad_norm": 7.3125,
      "learning_rate": 1.752e-05,
      "loss": 0.6651991844177246,
      "mean_token_accuracy": 0.7841649115085602,
      "num_tokens": 2808989.0,
      "step": 530
    },
    {
      "entropy": 0.6694958411157131,
      "epoch": 1.1370585134422773,
      "grad_norm": 6.78125,
      "learning_rate": 1.7431111111111112e-05,
      "loss": 0.6833772659301758,
      "mean_token_accuracy": 0.7775424145162105,
      "num_tokens": 2864507.0,
      "step": 540
    },
    {
      "entropy": 0.6568067740648985,
      "epoch": 1.158144438587243,
      "grad_norm": 6.21875,
      "learning_rate": 1.7342222222222226e-05,
      "loss": 0.6552294254302978,
      "mean_token_accuracy": 0.7832994520664215,
      "num_tokens": 2919201.0,
      "step": 550
    },
    {
      "entropy": 0.6432577680796385,
      "epoch": 1.1792303637322088,
      "grad_norm": 6.65625,
      "learning_rate": 1.7253333333333336e-05,
      "loss": 0.6560451030731201,
      "mean_token_accuracy": 0.7863748580217361,
      "num_tokens": 2972434.0,
      "step": 560
    },
    {
      "entropy": 0.6655397050082683,
      "epoch": 1.2003162888771746,
      "grad_norm": 6.09375,
      "learning_rate": 1.7164444444444447e-05,
      "loss": 0.6878513336181641,
      "mean_token_accuracy": 0.7753379993140698,
      "num_tokens": 3025099.0,
      "step": 570
    },
    {
      "entropy": 0.686093359068036,
      "epoch": 1.2214022140221403,
      "grad_norm": 7.09375,
      "learning_rate": 1.7075555555555557e-05,
      "loss": 0.6911579132080078,
      "mean_token_accuracy": 0.7787310972809791,
      "num_tokens": 3076307.0,
      "step": 580
    },
    {
      "entropy": 0.6812980778515338,
      "epoch": 1.2424881391671059,
      "grad_norm": 7.125,
      "learning_rate": 1.6986666666666668e-05,
      "loss": 0.6704156398773193,
      "mean_token_accuracy": 0.7793935135006904,
      "num_tokens": 3130053.0,
      "step": 590
    },
    {
      "entropy": 0.6409769468009472,
      "epoch": 1.2635740643120716,
      "grad_norm": 7.4375,
      "learning_rate": 1.6897777777777778e-05,
      "loss": 0.661864185333252,
      "mean_token_accuracy": 0.7865670621395111,
      "num_tokens": 3183331.0,
      "step": 600
    },
    {
      "entropy": 0.6799454756081105,
      "epoch": 1.2846599894570374,
      "grad_norm": 6.65625,
      "learning_rate": 1.680888888888889e-05,
      "loss": 0.6818383693695068,
      "mean_token_accuracy": 0.778287123888731,
      "num_tokens": 3239561.0,
      "step": 610
    },
    {
      "entropy": 0.6814555354416371,
      "epoch": 1.3057459146020032,
      "grad_norm": 6.3125,
      "learning_rate": 1.672e-05,
      "loss": 0.679498815536499,
      "mean_token_accuracy": 0.7788635492324829,
      "num_tokens": 3292996.0,
      "step": 620
    },
    {
      "entropy": 0.667108029127121,
      "epoch": 1.326831839746969,
      "grad_norm": 6.90625,
      "learning_rate": 1.6631111111111113e-05,
      "loss": 0.6758651733398438,
      "mean_token_accuracy": 0.7792007848620415,
      "num_tokens": 3345736.0,
      "step": 630
    },
    {
      "entropy": 0.6692709110677242,
      "epoch": 1.3479177648919347,
      "grad_norm": 6.96875,
      "learning_rate": 1.6542222222222223e-05,
      "loss": 0.6796128273010253,
      "mean_token_accuracy": 0.7766051791608334,
      "num_tokens": 3398076.0,
      "step": 640
    },
    {
      "entropy": 0.6564048878848553,
      "epoch": 1.3690036900369003,
      "grad_norm": 6.9375,
      "learning_rate": 1.6453333333333334e-05,
      "loss": 0.6633264064788819,
      "mean_token_accuracy": 0.7854946702718735,
      "num_tokens": 3450796.0,
      "step": 650
    },
    {
      "entropy": 0.6611979272216558,
      "epoch": 1.390089615181866,
      "grad_norm": 6.65625,
      "learning_rate": 1.6364444444444444e-05,
      "loss": 0.6661429405212402,
      "mean_token_accuracy": 0.7835663430392742,
      "num_tokens": 3503863.0,
      "step": 660
    },
    {
      "entropy": 0.6534961331635714,
      "epoch": 1.4111755403268318,
      "grad_norm": 6.9375,
      "learning_rate": 1.6275555555555558e-05,
      "loss": 0.6586530208587646,
      "mean_token_accuracy": 0.7844067431986332,
      "num_tokens": 3556265.0,
      "step": 670
    },
    {
      "entropy": 0.6516873501241207,
      "epoch": 1.4322614654717976,
      "grad_norm": 6.40625,
      "learning_rate": 1.618666666666667e-05,
      "loss": 0.6522529125213623,
      "mean_token_accuracy": 0.7855845786631107,
      "num_tokens": 3611423.0,
      "step": 680
    },
    {
      "entropy": 0.6764013767242432,
      "epoch": 1.4533473906167633,
      "grad_norm": 6.71875,
      "learning_rate": 1.609777777777778e-05,
      "loss": 0.6884442806243897,
      "mean_token_accuracy": 0.7756352722644806,
      "num_tokens": 3666630.0,
      "step": 690
    },
    {
      "entropy": 0.6554486058652401,
      "epoch": 1.474433315761729,
      "grad_norm": 7.40625,
      "learning_rate": 1.600888888888889e-05,
      "loss": 0.6563316822052002,
      "mean_token_accuracy": 0.7854672558605671,
      "num_tokens": 3718719.0,
      "step": 700
    },
    {
      "entropy": 0.6385185327380896,
      "epoch": 1.4955192409066949,
      "grad_norm": 6.59375,
      "learning_rate": 1.5920000000000003e-05,
      "loss": 0.6513954639434815,
      "mean_token_accuracy": 0.7851097173988819,
      "num_tokens": 3771822.0,
      "step": 710
    },
    {
      "entropy": 0.654766371101141,
      "epoch": 1.5166051660516606,
      "grad_norm": 6.5,
      "learning_rate": 1.5831111111111114e-05,
      "loss": 0.6622748851776123,
      "mean_token_accuracy": 0.7847663886845112,
      "num_tokens": 3823795.0,
      "step": 720
    },
    {
      "entropy": 0.6597310703247785,
      "epoch": 1.5376910911966264,
      "grad_norm": 6.375,
      "learning_rate": 1.5742222222222224e-05,
      "loss": 0.6634318351745605,
      "mean_token_accuracy": 0.7861979939043522,
      "num_tokens": 3875784.0,
      "step": 730
    },
    {
      "entropy": 0.6637851595878601,
      "epoch": 1.558777016341592,
      "grad_norm": 6.9375,
      "learning_rate": 1.5653333333333335e-05,
      "loss": 0.6621220111846924,
      "mean_token_accuracy": 0.7886320613324642,
      "num_tokens": 3929412.0,
      "step": 740
    },
    {
      "entropy": 0.637692479044199,
      "epoch": 1.5798629414865577,
      "grad_norm": 6.59375,
      "learning_rate": 1.5564444444444445e-05,
      "loss": 0.6402386665344239,
      "mean_token_accuracy": 0.7895873270928859,
      "num_tokens": 3982472.0,
      "step": 750
    },
    {
      "entropy": 0.6578179821372032,
      "epoch": 1.6009488666315235,
      "grad_norm": 7.25,
      "learning_rate": 1.5475555555555556e-05,
      "loss": 0.6751936435699463,
      "mean_token_accuracy": 0.7776105716824532,
      "num_tokens": 4037829.0,
      "step": 760
    },
    {
      "entropy": 0.6653612300753593,
      "epoch": 1.6220347917764892,
      "grad_norm": 6.9375,
      "learning_rate": 1.5386666666666666e-05,
      "loss": 0.6717413425445556,
      "mean_token_accuracy": 0.777862798422575,
      "num_tokens": 4093735.0,
      "step": 770
    },
    {
      "entropy": 0.6561707239598036,
      "epoch": 1.6431207169214548,
      "grad_norm": 7.03125,
      "learning_rate": 1.5297777777777777e-05,
      "loss": 0.6501868724822998,
      "mean_token_accuracy": 0.7863908343017101,
      "num_tokens": 4148410.0,
      "step": 780
    },
    {
      "entropy": 0.6655665650963783,
      "epoch": 1.6642066420664205,
      "grad_norm": 6.59375,
      "learning_rate": 1.520888888888889e-05,
      "loss": 0.679684829711914,
      "mean_token_accuracy": 0.778567586839199,
      "num_tokens": 4202907.0,
      "step": 790
    },
    {
      "entropy": 0.6781473312526941,
      "epoch": 1.6852925672113863,
      "grad_norm": 6.875,
      "learning_rate": 1.5120000000000001e-05,
      "loss": 0.6722404479980468,
      "mean_token_accuracy": 0.7763243250548839,
      "num_tokens": 4256213.0,
      "step": 800
    },
    {
      "entropy": 0.6321199681609869,
      "epoch": 1.706378492356352,
      "grad_norm": 6.78125,
      "learning_rate": 1.5031111111111111e-05,
      "loss": 0.6404708385467529,
      "mean_token_accuracy": 0.7903632469475269,
      "num_tokens": 4309268.0,
      "step": 810
    },
    {
      "entropy": 0.6424700867384672,
      "epoch": 1.7274644175013179,
      "grad_norm": 7.15625,
      "learning_rate": 1.4942222222222222e-05,
      "loss": 0.6444578170776367,
      "mean_token_accuracy": 0.7881530813872815,
      "num_tokens": 4360761.0,
      "step": 820
    },
    {
      "entropy": 0.6615507304668427,
      "epoch": 1.7485503426462836,
      "grad_norm": 6.0625,
      "learning_rate": 1.4853333333333336e-05,
      "loss": 0.6604718208312989,
      "mean_token_accuracy": 0.782805722206831,
      "num_tokens": 4411678.0,
      "step": 830
    },
    {
      "entropy": 0.6196760050952435,
      "epoch": 1.7696362677912494,
      "grad_norm": 6.6875,
      "learning_rate": 1.4764444444444446e-05,
      "loss": 0.6261836051940918,
      "mean_token_accuracy": 0.7926589228212834,
      "num_tokens": 4467254.0,
      "step": 840
    },
    {
      "entropy": 0.6139154363423586,
      "epoch": 1.7907221929362152,
      "grad_norm": 7.21875,
      "learning_rate": 1.4675555555555557e-05,
      "loss": 0.6301932811737061,
      "mean_token_accuracy": 0.7918383464217186,
      "num_tokens": 4517383.0,
      "step": 850
    },
    {
      "entropy": 0.6410171519964933,
      "epoch": 1.811808118081181,
      "grad_norm": 7.1875,
      "learning_rate": 1.4586666666666667e-05,
      "loss": 0.655810022354126,
      "mean_token_accuracy": 0.7856725618243218,
      "num_tokens": 4569980.0,
      "step": 860
    },
    {
      "entropy": 0.6666022066026926,
      "epoch": 1.8328940432261467,
      "grad_norm": 6.5625,
      "learning_rate": 1.449777777777778e-05,
      "loss": 0.6677482604980469,
      "mean_token_accuracy": 0.7798060193657875,
      "num_tokens": 4622612.0,
      "step": 870
    },
    {
      "entropy": 0.6398684818297624,
      "epoch": 1.8539799683711122,
      "grad_norm": 7.84375,
      "learning_rate": 1.440888888888889e-05,
      "loss": 0.6501501083374024,
      "mean_token_accuracy": 0.7857734531164169,
      "num_tokens": 4676525.0,
      "step": 880
    },
    {
      "entropy": 0.6285852700471878,
      "epoch": 1.875065893516078,
      "grad_norm": 7.0625,
      "learning_rate": 1.432e-05,
      "loss": 0.6422877311706543,
      "mean_token_accuracy": 0.7884316690266132,
      "num_tokens": 4734090.0,
      "step": 890
    },
    {
      "entropy": 0.6409385818988085,
      "epoch": 1.8961518186610438,
      "grad_norm": 7.9375,
      "learning_rate": 1.4231111111111114e-05,
      "loss": 0.6352181434631348,
      "mean_token_accuracy": 0.7952412232756615,
      "num_tokens": 4788537.0,
      "step": 900
    },
    {
      "entropy": 0.6456746399402619,
      "epoch": 1.9172377438060095,
      "grad_norm": 6.90625,
      "learning_rate": 1.4142222222222224e-05,
      "loss": 0.6564863204956055,
      "mean_token_accuracy": 0.789187741279602,
      "num_tokens": 4840342.0,
      "step": 910
    },
    {
      "entropy": 0.6360713161528111,
      "epoch": 1.938323668950975,
      "grad_norm": 6.3125,
      "learning_rate": 1.4053333333333335e-05,
      "loss": 0.6300373554229737,
      "mean_token_accuracy": 0.7876299776136875,
      "num_tokens": 4891524.0,
      "step": 920
    },
    {
      "entropy": 0.6285211775451899,
      "epoch": 1.9594095940959408,
      "grad_norm": 7.78125,
      "learning_rate": 1.3964444444444445e-05,
      "loss": 0.6337003231048584,
      "mean_token_accuracy": 0.7889726974070073,
      "num_tokens": 4939573.0,
      "step": 930
    },
    {
      "entropy": 0.6301535632461309,
      "epoch": 1.9804955192409066,
      "grad_norm": 6.84375,
      "learning_rate": 1.3875555555555557e-05,
      "loss": 0.6374162673950196,
      "mean_token_accuracy": 0.7898242607712745,
      "num_tokens": 4991267.0,
      "step": 940
    },
    {
      "entropy": 0.6537058566873138,
      "epoch": 2.0,
      "grad_norm": 13.3125,
      "learning_rate": 1.3786666666666668e-05,
      "loss": 0.641156530380249,
      "mean_token_accuracy": 0.7822844466647586,
      "num_tokens": 5036304.0,
      "step": 950
    },
    {
      "entropy": 0.6411334682255984,
      "epoch": 2.0210859251449658,
      "grad_norm": 7.1875,
      "learning_rate": 1.3697777777777778e-05,
      "loss": 0.6083747386932373,
      "mean_token_accuracy": 0.8002627611160278,
      "num_tokens": 5089737.0,
      "step": 960
    },
    {
      "entropy": 0.5841103594750165,
      "epoch": 2.0421718502899315,
      "grad_norm": 6.53125,
      "learning_rate": 1.3608888888888889e-05,
      "loss": 0.6012133121490478,
      "mean_token_accuracy": 0.8027123242616654,
      "num_tokens": 5143330.0,
      "step": 970
    },
    {
      "entropy": 0.590392829477787,
      "epoch": 2.0632577754348973,
      "grad_norm": 6.46875,
      "learning_rate": 1.3520000000000003e-05,
      "loss": 0.5829256534576416,
      "mean_token_accuracy": 0.8068650290369987,
      "num_tokens": 5198614.0,
      "step": 980
    },
    {
      "entropy": 0.6148912344127894,
      "epoch": 2.084343700579863,
      "grad_norm": 6.59375,
      "learning_rate": 1.3431111111111113e-05,
      "loss": 0.6117685317993165,
      "mean_token_accuracy": 0.7951427385210991,
      "num_tokens": 5252755.0,
      "step": 990
    },
    {
      "entropy": 0.5893940377980471,
      "epoch": 2.105429625724829,
      "grad_norm": 6.625,
      "learning_rate": 1.3342222222222224e-05,
      "loss": 0.5790496349334717,
      "mean_token_accuracy": 0.809234394133091,
      "num_tokens": 5304859.0,
      "step": 1000
    },
    {
      "entropy": 0.587013079598546,
      "epoch": 2.1265155508697946,
      "grad_norm": 6.9375,
      "learning_rate": 1.3253333333333334e-05,
      "loss": 0.5892822265625,
      "mean_token_accuracy": 0.8040889374911785,
      "num_tokens": 5357877.0,
      "step": 1010
    },
    {
      "entropy": 0.5877181258052587,
      "epoch": 2.14760147601476,
      "grad_norm": 6.09375,
      "learning_rate": 1.3164444444444446e-05,
      "loss": 0.595665693283081,
      "mean_token_accuracy": 0.8036225430667401,
      "num_tokens": 5413504.0,
      "step": 1020
    },
    {
      "entropy": 0.5848948258906603,
      "epoch": 2.1686874011597257,
      "grad_norm": 6.4375,
      "learning_rate": 1.3075555555555557e-05,
      "loss": 0.5771171092987061,
      "mean_token_accuracy": 0.8047491163015366,
      "num_tokens": 5466549.0,
      "step": 1030
    },
    {
      "entropy": 0.5815035384148359,
      "epoch": 2.1897733263046915,
      "grad_norm": 6.96875,
      "learning_rate": 1.2986666666666667e-05,
      "loss": 0.5768617153167724,
      "mean_token_accuracy": 0.8057669088244438,
      "num_tokens": 5519515.0,
      "step": 1040
    },
    {
      "entropy": 0.58335498906672,
      "epoch": 2.2108592514496572,
      "grad_norm": 7.6875,
      "learning_rate": 1.2897777777777778e-05,
      "loss": 0.6007503509521485,
      "mean_token_accuracy": 0.8005498602986336,
      "num_tokens": 5570188.0,
      "step": 1050
    },
    {
      "entropy": 0.6068603809922933,
      "epoch": 2.231945176594623,
      "grad_norm": 6.5625,
      "learning_rate": 1.2808888888888891e-05,
      "loss": 0.6024115562438965,
      "mean_token_accuracy": 0.7973541341722011,
      "num_tokens": 5624190.0,
      "step": 1060
    },
    {
      "entropy": 0.5774707745760679,
      "epoch": 2.2530311017395888,
      "grad_norm": 6.53125,
      "learning_rate": 1.2720000000000002e-05,
      "loss": 0.5705479145050049,
      "mean_token_accuracy": 0.8100002221763134,
      "num_tokens": 5675656.0,
      "step": 1070
    },
    {
      "entropy": 0.5851722996681928,
      "epoch": 2.2741170268845545,
      "grad_norm": 7.3125,
      "learning_rate": 1.2631111111111112e-05,
      "loss": 0.5905722141265869,
      "mean_token_accuracy": 0.8069308035075664,
      "num_tokens": 5728878.0,
      "step": 1080
    },
    {
      "entropy": 0.594813584536314,
      "epoch": 2.2952029520295203,
      "grad_norm": 6.875,
      "learning_rate": 1.2542222222222223e-05,
      "loss": 0.5923901557922363,
      "mean_token_accuracy": 0.8034852437675,
      "num_tokens": 5783257.0,
      "step": 1090
    },
    {
      "entropy": 0.5841789487749338,
      "epoch": 2.316288877174486,
      "grad_norm": 7.0,
      "learning_rate": 1.2453333333333335e-05,
      "loss": 0.5864346504211426,
      "mean_token_accuracy": 0.8033530995249748,
      "num_tokens": 5834581.0,
      "step": 1100
    },
    {
      "entropy": 0.6010521166026592,
      "epoch": 2.337374802319452,
      "grad_norm": 6.9375,
      "learning_rate": 1.2364444444444445e-05,
      "loss": 0.5941137313842774,
      "mean_token_accuracy": 0.8033333070576191,
      "num_tokens": 5887458.0,
      "step": 1110
    },
    {
      "entropy": 0.5919341973960399,
      "epoch": 2.3584607274644176,
      "grad_norm": 8.75,
      "learning_rate": 1.2275555555555556e-05,
      "loss": 0.6103182315826416,
      "mean_token_accuracy": 0.7999237082898617,
      "num_tokens": 5941284.0,
      "step": 1120
    },
    {
      "entropy": 0.5942621320486069,
      "epoch": 2.3795466526093834,
      "grad_norm": 7.375,
      "learning_rate": 1.2186666666666666e-05,
      "loss": 0.5983363151550293,
      "mean_token_accuracy": 0.8016222022473812,
      "num_tokens": 5995817.0,
      "step": 1130
    },
    {
      "entropy": 0.6046227600425482,
      "epoch": 2.400632577754349,
      "grad_norm": 6.40625,
      "learning_rate": 1.209777777777778e-05,
      "loss": 0.6026669025421143,
      "mean_token_accuracy": 0.8024684950709343,
      "num_tokens": 6050783.0,
      "step": 1140
    },
    {
      "entropy": 0.5998306684195995,
      "epoch": 2.421718502899315,
      "grad_norm": 6.59375,
      "learning_rate": 1.200888888888889e-05,
      "loss": 0.5893473625183105,
      "mean_token_accuracy": 0.803314708918333,
      "num_tokens": 6103371.0,
      "step": 1150
    },
    {
      "entropy": 0.593693134188652,
      "epoch": 2.4428044280442807,
      "grad_norm": 6.5,
      "learning_rate": 1.1920000000000001e-05,
      "loss": 0.6076583862304688,
      "mean_token_accuracy": 0.7995239071547985,
      "num_tokens": 6155502.0,
      "step": 1160
    },
    {
      "entropy": 0.5827030725777149,
      "epoch": 2.463890353189246,
      "grad_norm": 6.84375,
      "learning_rate": 1.1831111111111112e-05,
      "loss": 0.5882821559906006,
      "mean_token_accuracy": 0.7929977595806121,
      "num_tokens": 6208060.0,
      "step": 1170
    },
    {
      "entropy": 0.5910260166972876,
      "epoch": 2.4849762783342118,
      "grad_norm": 7.53125,
      "learning_rate": 1.1742222222222224e-05,
      "loss": 0.5860947132110595,
      "mean_token_accuracy": 0.8049415498971939,
      "num_tokens": 6261588.0,
      "step": 1180
    },
    {
      "entropy": 0.5704694345593453,
      "epoch": 2.5060622034791775,
      "grad_norm": 6.75,
      "learning_rate": 1.1653333333333334e-05,
      "loss": 0.5749064445495605,
      "mean_token_accuracy": 0.8084491945803165,
      "num_tokens": 6314963.0,
      "step": 1190
    },
    {
      "entropy": 0.597440430521965,
      "epoch": 2.5271481286241433,
      "grad_norm": 6.5,
      "learning_rate": 1.1564444444444445e-05,
      "loss": 0.5958696842193604,
      "mean_token_accuracy": 0.8057812102138996,
      "num_tokens": 6368741.0,
      "step": 1200
    },
    {
      "entropy": 0.6070981964468956,
      "epoch": 2.548234053769109,
      "grad_norm": 7.96875,
      "learning_rate": 1.1475555555555557e-05,
      "loss": 0.6143896102905273,
      "mean_token_accuracy": 0.7980599500238895,
      "num_tokens": 6425424.0,
      "step": 1210
    },
    {
      "entropy": 0.5948727998882533,
      "epoch": 2.569319978914075,
      "grad_norm": 6.5,
      "learning_rate": 1.1386666666666669e-05,
      "loss": 0.5917119026184082,
      "mean_token_accuracy": 0.8025051333010197,
      "num_tokens": 6476263.0,
      "step": 1220
    },
    {
      "entropy": 0.5811738461256027,
      "epoch": 2.5904059040590406,
      "grad_norm": 7.21875,
      "learning_rate": 1.129777777777778e-05,
      "loss": 0.5865864753723145,
      "mean_token_accuracy": 0.8044181451201439,
      "num_tokens": 6528012.0,
      "step": 1230
    },
    {
      "entropy": 0.5937653455883265,
      "epoch": 2.6114918292040064,
      "grad_norm": 6.59375,
      "learning_rate": 1.120888888888889e-05,
      "loss": 0.5843213558197021,
      "mean_token_accuracy": 0.8025668740272522,
      "num_tokens": 6583421.0,
      "step": 1240
    },
    {
      "entropy": 0.573565949127078,
      "epoch": 2.632577754348972,
      "grad_norm": 6.5625,
      "learning_rate": 1.1120000000000002e-05,
      "loss": 0.5826557159423829,
      "mean_token_accuracy": 0.8070438347756863,
      "num_tokens": 6636858.0,
      "step": 1250
    },
    {
      "entropy": 0.6216809719800949,
      "epoch": 2.653663679493938,
      "grad_norm": 7.09375,
      "learning_rate": 1.1031111111111112e-05,
      "loss": 0.6125890254974365,
      "mean_token_accuracy": 0.7970895677804947,
      "num_tokens": 6692168.0,
      "step": 1260
    },
    {
      "entropy": 0.5998566258698702,
      "epoch": 2.6747496046389037,
      "grad_norm": 7.0625,
      "learning_rate": 1.0942222222222223e-05,
      "loss": 0.5921674728393554,
      "mean_token_accuracy": 0.8061520479619503,
      "num_tokens": 6745834.0,
      "step": 1270
    },
    {
      "entropy": 0.5880189586430788,
      "epoch": 2.6958355297838694,
      "grad_norm": 7.59375,
      "learning_rate": 1.0853333333333333e-05,
      "loss": 0.5986497402191162,
      "mean_token_accuracy": 0.8006219923496246,
      "num_tokens": 6800244.0,
      "step": 1280
    },
    {
      "entropy": 0.5981763809919357,
      "epoch": 2.716921454928835,
      "grad_norm": 6.6875,
      "learning_rate": 1.0764444444444445e-05,
      "loss": 0.6020579814910889,
      "mean_token_accuracy": 0.8032843962311744,
      "num_tokens": 6853454.0,
      "step": 1290
    },
    {
      "entropy": 0.5963539965450764,
      "epoch": 2.7380073800738005,
      "grad_norm": 6.59375,
      "learning_rate": 1.0675555555555558e-05,
      "loss": 0.5908561706542969,
      "mean_token_accuracy": 0.8011696465313435,
      "num_tokens": 6904473.0,
      "step": 1300
    },
    {
      "entropy": 0.5899917531758547,
      "epoch": 2.7590933052187667,
      "grad_norm": 6.6875,
      "learning_rate": 1.0586666666666668e-05,
      "loss": 0.5744058132171631,
      "mean_token_accuracy": 0.8053683824837208,
      "num_tokens": 6956756.0,
      "step": 1310
    },
    {
      "entropy": 0.582448274642229,
      "epoch": 2.780179230363732,
      "grad_norm": 6.78125,
      "learning_rate": 1.0497777777777779e-05,
      "loss": 0.5848373889923095,
      "mean_token_accuracy": 0.8054254874587059,
      "num_tokens": 7008611.0,
      "step": 1320
    },
    {
      "entropy": 0.5906594522297383,
      "epoch": 2.801265155508698,
      "grad_norm": 7.75,
      "learning_rate": 1.040888888888889e-05,
      "loss": 0.6006948471069335,
      "mean_token_accuracy": 0.7971956506371498,
      "num_tokens": 7062235.0,
      "step": 1330
    },
    {
      "entropy": 0.5976275954395532,
      "epoch": 2.8223510806536636,
      "grad_norm": 6.34375,
      "learning_rate": 1.0320000000000001e-05,
      "loss": 0.5997600078582763,
      "mean_token_accuracy": 0.7998531877994537,
      "num_tokens": 7115737.0,
      "step": 1340
    },
    {
      "entropy": 0.6079881630837918,
      "epoch": 2.8434370057986293,
      "grad_norm": 7.0,
      "learning_rate": 1.0231111111111112e-05,
      "loss": 0.6024625778198243,
      "mean_token_accuracy": 0.7988371469080449,
      "num_tokens": 7166870.0,
      "step": 1350
    },
    {
      "entropy": 0.6036868289113044,
      "epoch": 2.864522930943595,
      "grad_norm": 6.6875,
      "learning_rate": 1.0142222222222222e-05,
      "loss": 0.5962271213531494,
      "mean_token_accuracy": 0.8011133573949337,
      "num_tokens": 7219785.0,
      "step": 1360
    },
    {
      "entropy": 0.5746289525181055,
      "epoch": 2.885608856088561,
      "grad_norm": 7.5625,
      "learning_rate": 1.0053333333333334e-05,
      "loss": 0.5769401073455811,
      "mean_token_accuracy": 0.809252617508173,
      "num_tokens": 7275411.0,
      "step": 1370
    },
    {
      "entropy": 0.5846736520528794,
      "epoch": 2.9066947812335266,
      "grad_norm": 7.125,
      "learning_rate": 9.964444444444445e-06,
      "loss": 0.5899319171905517,
      "mean_token_accuracy": 0.8028473742306232,
      "num_tokens": 7326752.0,
      "step": 1380
    },
    {
      "entropy": 0.5813163112848997,
      "epoch": 2.9277807063784924,
      "grad_norm": 7.46875,
      "learning_rate": 9.875555555555557e-06,
      "loss": 0.5799898147583008,
      "mean_token_accuracy": 0.8026903085410595,
      "num_tokens": 7375608.0,
      "step": 1390
    },
    {
      "entropy": 0.5724870227277279,
      "epoch": 2.948866631523458,
      "grad_norm": 6.90625,
      "learning_rate": 9.786666666666667e-06,
      "loss": 0.5783823490142822,
      "mean_token_accuracy": 0.8059644535183906,
      "num_tokens": 7430145.0,
      "step": 1400
    },
    {
      "entropy": 0.5778653729707003,
      "epoch": 2.969952556668424,
      "grad_norm": 6.84375,
      "learning_rate": 9.697777777777778e-06,
      "loss": 0.5640594005584717,
      "mean_token_accuracy": 0.8100298568606377,
      "num_tokens": 7482976.0,
      "step": 1410
    },
    {
      "entropy": 0.5823766525834799,
      "epoch": 2.9910384818133897,
      "grad_norm": 6.5625,
      "learning_rate": 9.60888888888889e-06,
      "loss": 0.5931865692138671,
      "mean_token_accuracy": 0.8023995622992516,
      "num_tokens": 7533043.0,
      "step": 1420
    },
    {
      "entropy": 0.5924574036050487,
      "epoch": 3.0105429625724827,
      "grad_norm": 6.53125,
      "learning_rate": 9.52e-06,
      "loss": 0.5867043972015381,
      "mean_token_accuracy": 0.803342911842707,
      "num_tokens": 7581993.0,
      "step": 1430
    },
    {
      "entropy": 0.579814862459898,
      "epoch": 3.0316288877174484,
      "grad_norm": 6.34375,
      "learning_rate": 9.431111111111112e-06,
      "loss": 0.5730488300323486,
      "mean_token_accuracy": 0.8101175747811794,
      "num_tokens": 7639343.0,
      "step": 1440
    },
    {
      "entropy": 0.560778708383441,
      "epoch": 3.052714812862414,
      "grad_norm": 7.40625,
      "learning_rate": 9.342222222222223e-06,
      "loss": 0.5500640392303466,
      "mean_token_accuracy": 0.8135757654905319,
      "num_tokens": 7690695.0,
      "step": 1450
    },
    {
      "entropy": 0.5529176875948906,
      "epoch": 3.07380073800738,
      "grad_norm": 7.21875,
      "learning_rate": 9.253333333333333e-06,
      "loss": 0.549665117263794,
      "mean_token_accuracy": 0.81509620398283,
      "num_tokens": 7746705.0,
      "step": 1460
    },
    {
      "entropy": 0.5382925808429718,
      "epoch": 3.0948866631523457,
      "grad_norm": 7.5,
      "learning_rate": 9.164444444444446e-06,
      "loss": 0.54415283203125,
      "mean_token_accuracy": 0.819620954990387,
      "num_tokens": 7799255.0,
      "step": 1470
    },
    {
      "entropy": 0.551479097828269,
      "epoch": 3.1159725882973115,
      "grad_norm": 7.125,
      "learning_rate": 9.075555555555556e-06,
      "loss": 0.5520545959472656,
      "mean_token_accuracy": 0.8147840730845928,
      "num_tokens": 7852753.0,
      "step": 1480
    },
    {
      "entropy": 0.571834359690547,
      "epoch": 3.1370585134422773,
      "grad_norm": 7.59375,
      "learning_rate": 8.986666666666666e-06,
      "loss": 0.5553510189056396,
      "mean_token_accuracy": 0.8117231354117393,
      "num_tokens": 7904298.0,
      "step": 1490
    },
    {
      "entropy": 0.5676855646073818,
      "epoch": 3.158144438587243,
      "grad_norm": 7.75,
      "learning_rate": 8.897777777777779e-06,
      "loss": 0.5595914840698242,
      "mean_token_accuracy": 0.8130396403372288,
      "num_tokens": 7956705.0,
      "step": 1500
    },
    {
      "entropy": 0.5613869089633227,
      "epoch": 3.179230363732209,
      "grad_norm": 7.78125,
      "learning_rate": 8.80888888888889e-06,
      "loss": 0.544979190826416,
      "mean_token_accuracy": 0.8170888110995292,
      "num_tokens": 8008023.0,
      "step": 1510
    },
    {
      "entropy": 0.570792680233717,
      "epoch": 3.2003162888771746,
      "grad_norm": 6.75,
      "learning_rate": 8.720000000000001e-06,
      "loss": 0.560499906539917,
      "mean_token_accuracy": 0.8146887816488743,
      "num_tokens": 8058344.0,
      "step": 1520
    },
    {
      "entropy": 0.5498528245836496,
      "epoch": 3.2214022140221403,
      "grad_norm": 7.15625,
      "learning_rate": 8.631111111111112e-06,
      "loss": 0.5479342937469482,
      "mean_token_accuracy": 0.8137361407279968,
      "num_tokens": 8110587.0,
      "step": 1530
    },
    {
      "entropy": 0.5499754682183265,
      "epoch": 3.242488139167106,
      "grad_norm": 6.9375,
      "learning_rate": 8.542222222222222e-06,
      "loss": 0.5445993423461915,
      "mean_token_accuracy": 0.8179522223770619,
      "num_tokens": 8160497.0,
      "step": 1540
    },
    {
      "entropy": 0.572133193910122,
      "epoch": 3.263574064312072,
      "grad_norm": 6.84375,
      "learning_rate": 8.453333333333334e-06,
      "loss": 0.5740565776824951,
      "mean_token_accuracy": 0.8101011097431183,
      "num_tokens": 8214878.0,
      "step": 1550
    },
    {
      "entropy": 0.5666386324912309,
      "epoch": 3.2846599894570376,
      "grad_norm": 7.59375,
      "learning_rate": 8.364444444444445e-06,
      "loss": 0.5723002433776856,
      "mean_token_accuracy": 0.8091295748949051,
      "num_tokens": 8267314.0,
      "step": 1560
    },
    {
      "entropy": 0.5659929499030113,
      "epoch": 3.305745914602003,
      "grad_norm": 6.96875,
      "learning_rate": 8.275555555555557e-06,
      "loss": 0.5623570442199707,
      "mean_token_accuracy": 0.8093601904809475,
      "num_tokens": 8320145.0,
      "step": 1570
    },
    {
      "entropy": 0.5665242202579975,
      "epoch": 3.3268318397469687,
      "grad_norm": 7.0625,
      "learning_rate": 8.186666666666667e-06,
      "loss": 0.5534502983093261,
      "mean_token_accuracy": 0.8129527278244495,
      "num_tokens": 8370852.0,
      "step": 1580
    },
    {
      "entropy": 0.5829939197748899,
      "epoch": 3.3479177648919345,
      "grad_norm": 7.40625,
      "learning_rate": 8.09777777777778e-06,
      "loss": 0.5778002738952637,
      "mean_token_accuracy": 0.8072332389652729,
      "num_tokens": 8424209.0,
      "step": 1590
    },
    {
      "entropy": 0.5479087311774492,
      "epoch": 3.3690036900369003,
      "grad_norm": 6.15625,
      "learning_rate": 8.00888888888889e-06,
      "loss": 0.5359113216400146,
      "mean_token_accuracy": 0.819999934732914,
      "num_tokens": 8476617.0,
      "step": 1600
    },
    {
      "entropy": 0.5520685013383627,
      "epoch": 3.390089615181866,
      "grad_norm": 6.53125,
      "learning_rate": 7.92e-06,
      "loss": 0.540053415298462,
      "mean_token_accuracy": 0.817080944776535,
      "num_tokens": 8529074.0,
      "step": 1610
    },
    {
      "entropy": 0.5533283669501543,
      "epoch": 3.411175540326832,
      "grad_norm": 7.25,
      "learning_rate": 7.831111111111111e-06,
      "loss": 0.5461609363555908,
      "mean_token_accuracy": 0.8130916491150856,
      "num_tokens": 8582416.0,
      "step": 1620
    },
    {
      "entropy": 0.5606811255216598,
      "epoch": 3.4322614654717976,
      "grad_norm": 8.25,
      "learning_rate": 7.742222222222223e-06,
      "loss": 0.565118408203125,
      "mean_token_accuracy": 0.813844607770443,
      "num_tokens": 8634249.0,
      "step": 1630
    },
    {
      "entropy": 0.5458527609705925,
      "epoch": 3.4533473906167633,
      "grad_norm": 7.6875,
      "learning_rate": 7.653333333333333e-06,
      "loss": 0.5410670757293701,
      "mean_token_accuracy": 0.8188620381057262,
      "num_tokens": 8685168.0,
      "step": 1640
    },
    {
      "entropy": 0.5717878874391318,
      "epoch": 3.474433315761729,
      "grad_norm": 7.6875,
      "learning_rate": 7.564444444444446e-06,
      "loss": 0.5788788795471191,
      "mean_token_accuracy": 0.8078153640031814,
      "num_tokens": 8740562.0,
      "step": 1650
    },
    {
      "entropy": 0.5646973233669996,
      "epoch": 3.495519240906695,
      "grad_norm": 6.75,
      "learning_rate": 7.475555555555556e-06,
      "loss": 0.5595187187194824,
      "mean_token_accuracy": 0.8135132856667042,
      "num_tokens": 8794411.0,
      "step": 1660
    },
    {
      "entropy": 0.5641634043306112,
      "epoch": 3.5166051660516606,
      "grad_norm": 6.6875,
      "learning_rate": 7.386666666666667e-06,
      "loss": 0.5511392116546631,
      "mean_token_accuracy": 0.8159678690135479,
      "num_tokens": 8844547.0,
      "step": 1670
    },
    {
      "entropy": 0.5724293068051338,
      "epoch": 3.5376910911966264,
      "grad_norm": 6.875,
      "learning_rate": 7.297777777777778e-06,
      "loss": 0.5668179988861084,
      "mean_token_accuracy": 0.8106696806848049,
      "num_tokens": 8902311.0,
      "step": 1680
    },
    {
      "entropy": 0.5610478781163692,
      "epoch": 3.558777016341592,
      "grad_norm": 6.96875,
      "learning_rate": 7.20888888888889e-06,
      "loss": 0.5528025150299072,
      "mean_token_accuracy": 0.8120494581758976,
      "num_tokens": 8957569.0,
      "step": 1690
    },
    {
      "entropy": 0.5370181340724229,
      "epoch": 3.5798629414865575,
      "grad_norm": 7.28125,
      "learning_rate": 7.1200000000000004e-06,
      "loss": 0.5477783203125,
      "mean_token_accuracy": 0.8162322834134101,
      "num_tokens": 9009083.0,
      "step": 1700
    },
    {
      "entropy": 0.5545388571918011,
      "epoch": 3.6009488666315237,
      "grad_norm": 6.5,
      "learning_rate": 7.031111111111112e-06,
      "loss": 0.5430552005767822,
      "mean_token_accuracy": 0.8191993653774261,
      "num_tokens": 9061980.0,
      "step": 1710
    },
    {
      "entropy": 0.5662489753216505,
      "epoch": 3.622034791776489,
      "grad_norm": 7.3125,
      "learning_rate": 6.942222222222222e-06,
      "loss": 0.5623671531677246,
      "mean_token_accuracy": 0.8118648871779441,
      "num_tokens": 9117653.0,
      "step": 1720
    },
    {
      "entropy": 0.5656806524842978,
      "epoch": 3.643120716921455,
      "grad_norm": 7.40625,
      "learning_rate": 6.853333333333334e-06,
      "loss": 0.5693060398101807,
      "mean_token_accuracy": 0.8121738277375699,
      "num_tokens": 9168884.0,
      "step": 1730
    },
    {
      "entropy": 0.5669469233602286,
      "epoch": 3.6642066420664205,
      "grad_norm": 6.875,
      "learning_rate": 6.764444444444445e-06,
      "loss": 0.5708917617797852,
      "mean_token_accuracy": 0.8130646869540215,
      "num_tokens": 9221794.0,
      "step": 1740
    },
    {
      "entropy": 0.5680123705416917,
      "epoch": 3.6852925672113863,
      "grad_norm": 7.3125,
      "learning_rate": 6.675555555555556e-06,
      "loss": 0.565749979019165,
      "mean_token_accuracy": 0.8114100478589534,
      "num_tokens": 9277320.0,
      "step": 1750
    },
    {
      "entropy": 0.5670568168163299,
      "epoch": 3.706378492356352,
      "grad_norm": 6.8125,
      "learning_rate": 6.5866666666666666e-06,
      "loss": 0.5531346797943115,
      "mean_token_accuracy": 0.815721620619297,
      "num_tokens": 9328654.0,
      "step": 1760
    },
    {
      "entropy": 0.5794487085193396,
      "epoch": 3.727464417501318,
      "grad_norm": 7.25,
      "learning_rate": 6.497777777777779e-06,
      "loss": 0.5695443630218506,
      "mean_token_accuracy": 0.8125023365020752,
      "num_tokens": 9382781.0,
      "step": 1770
    },
    {
      "entropy": 0.5529596284031868,
      "epoch": 3.7485503426462836,
      "grad_norm": 7.21875,
      "learning_rate": 6.408888888888889e-06,
      "loss": 0.5538197040557862,
      "mean_token_accuracy": 0.8132442943751812,
      "num_tokens": 9434446.0,
      "step": 1780
    },
    {
      "entropy": 0.554557716101408,
      "epoch": 3.7696362677912494,
      "grad_norm": 7.34375,
      "learning_rate": 6.3200000000000005e-06,
      "loss": 0.5493638515472412,
      "mean_token_accuracy": 0.8154775679111481,
      "num_tokens": 9488504.0,
      "step": 1790
    },
    {
      "entropy": 0.5696977652609349,
      "epoch": 3.790722192936215,
      "grad_norm": 6.28125,
      "learning_rate": 6.231111111111111e-06,
      "loss": 0.5688616275787354,
      "mean_token_accuracy": 0.8128006041049958,
      "num_tokens": 9545943.0,
      "step": 1800
    },
    {
      "entropy": 0.5715777482837439,
      "epoch": 3.811808118081181,
      "grad_norm": 6.78125,
      "learning_rate": 6.142222222222223e-06,
      "loss": 0.5713305473327637,
      "mean_token_accuracy": 0.811994157731533,
      "num_tokens": 9597820.0,
      "step": 1810
    },
    {
      "entropy": 0.5638129018247128,
      "epoch": 3.8328940432261467,
      "grad_norm": 7.25,
      "learning_rate": 6.0533333333333335e-06,
      "loss": 0.5548198699951172,
      "mean_token_accuracy": 0.8136275447905064,
      "num_tokens": 9651480.0,
      "step": 1820
    },
    {
      "entropy": 0.5706254750490188,
      "epoch": 3.853979968371112,
      "grad_norm": 6.8125,
      "learning_rate": 5.964444444444445e-06,
      "loss": 0.5752524852752685,
      "mean_token_accuracy": 0.8053847931325435,
      "num_tokens": 9706114.0,
      "step": 1830
    },
    {
      "entropy": 0.5612337995320559,
      "epoch": 3.875065893516078,
      "grad_norm": 7.0,
      "learning_rate": 5.875555555555556e-06,
      "loss": 0.5460340499877929,
      "mean_token_accuracy": 0.8159381657838821,
      "num_tokens": 9760141.0,
      "step": 1840
    },
    {
      "entropy": 0.5514645285904407,
      "epoch": 3.8961518186610435,
      "grad_norm": 8.0,
      "learning_rate": 5.7866666666666674e-06,
      "loss": 0.5530869960784912,
      "mean_token_accuracy": 0.8169759206473828,
      "num_tokens": 9812466.0,
      "step": 1850
    },
    {
      "entropy": 0.5574501186609269,
      "epoch": 3.9172377438060098,
      "grad_norm": 7.5,
      "learning_rate": 5.697777777777779e-06,
      "loss": 0.5557053089141846,
      "mean_token_accuracy": 0.8142473429441452,
      "num_tokens": 9863547.0,
      "step": 1860
    },
    {
      "entropy": 0.560401076450944,
      "epoch": 3.938323668950975,
      "grad_norm": 6.53125,
      "learning_rate": 5.608888888888889e-06,
      "loss": 0.5598348140716553,
      "mean_token_accuracy": 0.810575645416975,
      "num_tokens": 9917210.0,
      "step": 1870
    },
    {
      "entropy": 0.5611382886767388,
      "epoch": 3.959409594095941,
      "grad_norm": 6.875,
      "learning_rate": 5.5200000000000005e-06,
      "loss": 0.5510398387908936,
      "mean_token_accuracy": 0.8139252245426178,
      "num_tokens": 9971753.0,
      "step": 1880
    },
    {
      "entropy": 0.571844469755888,
      "epoch": 3.9804955192409066,
      "grad_norm": 7.15625,
      "learning_rate": 5.431111111111112e-06,
      "loss": 0.5825337409973145,
      "mean_token_accuracy": 0.8051016978919506,
      "num_tokens": 10024121.0,
      "step": 1890
    },
    {
      "entropy": 0.5610769511880102,
      "epoch": 4.0,
      "grad_norm": 15.75,
      "learning_rate": 5.342222222222223e-06,
      "loss": 0.5544355392456055,
      "mean_token_accuracy": 0.8145432029221509,
      "num_tokens": 10072608.0,
      "step": 1900
    },
    {
      "entropy": 0.5582417894154788,
      "epoch": 4.021085925144965,
      "grad_norm": 6.5625,
      "learning_rate": 5.2533333333333336e-06,
      "loss": 0.5517781734466553,
      "mean_token_accuracy": 0.8166630238294601,
      "num_tokens": 10127149.0,
      "step": 1910
    },
    {
      "entropy": 0.5561550367623568,
      "epoch": 4.0421718502899315,
      "grad_norm": 6.5,
      "learning_rate": 5.164444444444445e-06,
      "loss": 0.5391493797302246,
      "mean_token_accuracy": 0.8172842212021351,
      "num_tokens": 10178921.0,
      "step": 1920
    },
    {
      "entropy": 0.5536927301436663,
      "epoch": 4.063257775434897,
      "grad_norm": 6.8125,
      "learning_rate": 5.075555555555556e-06,
      "loss": 0.5435996532440186,
      "mean_token_accuracy": 0.8193235926330089,
      "num_tokens": 10230953.0,
      "step": 1930
    },
    {
      "entropy": 0.5522860966622829,
      "epoch": 4.084343700579863,
      "grad_norm": 6.65625,
      "learning_rate": 4.986666666666667e-06,
      "loss": 0.5417564868927002,
      "mean_token_accuracy": 0.8215342923998833,
      "num_tokens": 10285145.0,
      "step": 1940
    },
    {
      "entropy": 0.558988232910633,
      "epoch": 4.105429625724828,
      "grad_norm": 8.25,
      "learning_rate": 4.897777777777778e-06,
      "loss": 0.5513458251953125,
      "mean_token_accuracy": 0.8168663889169693,
      "num_tokens": 10340678.0,
      "step": 1950
    },
    {
      "entropy": 0.5479944169521331,
      "epoch": 4.126515550869795,
      "grad_norm": 7.6875,
      "learning_rate": 4.808888888888889e-06,
      "loss": 0.5411903858184814,
      "mean_token_accuracy": 0.8173033416271209,
      "num_tokens": 10396265.0,
      "step": 1960
    },
    {
      "entropy": 0.5479635439813137,
      "epoch": 4.14760147601476,
      "grad_norm": 6.6875,
      "learning_rate": 4.7200000000000005e-06,
      "loss": 0.5325856685638428,
      "mean_token_accuracy": 0.8196660950779915,
      "num_tokens": 10448162.0,
      "step": 1970
    },
    {
      "entropy": 0.5631415072828532,
      "epoch": 4.168687401159726,
      "grad_norm": 6.96875,
      "learning_rate": 4.631111111111111e-06,
      "loss": 0.5563766956329346,
      "mean_token_accuracy": 0.8139985457062722,
      "num_tokens": 10500711.0,
      "step": 1980
    },
    {
      "entropy": 0.5530602999031544,
      "epoch": 4.1897733263046915,
      "grad_norm": 6.78125,
      "learning_rate": 4.542222222222223e-06,
      "loss": 0.5429882049560547,
      "mean_token_accuracy": 0.815962927043438,
      "num_tokens": 10554919.0,
      "step": 1990
    },
    {
      "entropy": 0.5508173421025276,
      "epoch": 4.210859251449658,
      "grad_norm": 6.90625,
      "learning_rate": 4.453333333333334e-06,
      "loss": 0.5390869140625,
      "mean_token_accuracy": 0.8184089511632919,
      "num_tokens": 10606269.0,
      "step": 2000
    },
    {
      "entropy": 0.5496918559074402,
      "epoch": 4.231945176594623,
      "grad_norm": 7.0,
      "learning_rate": 4.364444444444445e-06,
      "loss": 0.5486071109771729,
      "mean_token_accuracy": 0.8196936115622521,
      "num_tokens": 10660269.0,
      "step": 2010
    },
    {
      "entropy": 0.5489282257854938,
      "epoch": 4.253031101739589,
      "grad_norm": 7.4375,
      "learning_rate": 4.275555555555556e-06,
      "loss": 0.5514675617218018,
      "mean_token_accuracy": 0.8145210847258568,
      "num_tokens": 10711759.0,
      "step": 2020
    },
    {
      "entropy": 0.5529059153050184,
      "epoch": 4.2741170268845545,
      "grad_norm": 6.78125,
      "learning_rate": 4.1866666666666675e-06,
      "loss": 0.5444664001464844,
      "mean_token_accuracy": 0.8184945359826088,
      "num_tokens": 10763911.0,
      "step": 2030
    },
    {
      "entropy": 0.551506045460701,
      "epoch": 4.29520295202952,
      "grad_norm": 7.34375,
      "learning_rate": 4.097777777777778e-06,
      "loss": 0.548325777053833,
      "mean_token_accuracy": 0.8162619814276695,
      "num_tokens": 10817880.0,
      "step": 2040
    },
    {
      "entropy": 0.5396009232848883,
      "epoch": 4.316288877174486,
      "grad_norm": 6.8125,
      "learning_rate": 4.008888888888889e-06,
      "loss": 0.5268721103668212,
      "mean_token_accuracy": 0.8218203999102116,
      "num_tokens": 10870365.0,
      "step": 2050
    },
    {
      "entropy": 0.5537472911179065,
      "epoch": 4.337374802319451,
      "grad_norm": 6.65625,
      "learning_rate": 3.920000000000001e-06,
      "loss": 0.5522951126098633,
      "mean_token_accuracy": 0.8159066334366798,
      "num_tokens": 10922363.0,
      "step": 2060
    },
    {
      "entropy": 0.5330416887998581,
      "epoch": 4.358460727464418,
      "grad_norm": 7.21875,
      "learning_rate": 3.831111111111112e-06,
      "loss": 0.5207313537597656,
      "mean_token_accuracy": 0.8252225182950497,
      "num_tokens": 10972251.0,
      "step": 2070
    },
    {
      "entropy": 0.5454271405935287,
      "epoch": 4.379546652609383,
      "grad_norm": 7.125,
      "learning_rate": 3.7422222222222228e-06,
      "loss": 0.5341221809387207,
      "mean_token_accuracy": 0.8211991436779499,
      "num_tokens": 11025983.0,
      "step": 2080
    },
    {
      "entropy": 0.5519603937864304,
      "epoch": 4.400632577754349,
      "grad_norm": 7.5625,
      "learning_rate": 3.6533333333333336e-06,
      "loss": 0.5493678569793701,
      "mean_token_accuracy": 0.8145092785358429,
      "num_tokens": 11081248.0,
      "step": 2090
    },
    {
      "entropy": 0.5668473180383444,
      "epoch": 4.4217185028993145,
      "grad_norm": 7.21875,
      "learning_rate": 3.564444444444445e-06,
      "loss": 0.5566862106323243,
      "mean_token_accuracy": 0.813294805586338,
      "num_tokens": 11135369.0,
      "step": 2100
    },
    {
      "entropy": 0.5516544353216887,
      "epoch": 4.442804428044281,
      "grad_norm": 7.21875,
      "learning_rate": 3.475555555555556e-06,
      "loss": 0.5478980541229248,
      "mean_token_accuracy": 0.8169082552194595,
      "num_tokens": 11190765.0,
      "step": 2110
    },
    {
      "entropy": 0.5469749603420496,
      "epoch": 4.463890353189246,
      "grad_norm": 8.1875,
      "learning_rate": 3.386666666666667e-06,
      "loss": 0.5484558582305908,
      "mean_token_accuracy": 0.8153667166829109,
      "num_tokens": 11242129.0,
      "step": 2120
    },
    {
      "entropy": 0.5624320391565562,
      "epoch": 4.484976278334212,
      "grad_norm": 7.15625,
      "learning_rate": 3.297777777777778e-06,
      "loss": 0.5431473255157471,
      "mean_token_accuracy": 0.8167137958109378,
      "num_tokens": 11296409.0,
      "step": 2130
    },
    {
      "entropy": 0.5575876057147979,
      "epoch": 4.5060622034791775,
      "grad_norm": 6.625,
      "learning_rate": 3.2088888888888893e-06,
      "loss": 0.546560525894165,
      "mean_token_accuracy": 0.8190927796065808,
      "num_tokens": 11351788.0,
      "step": 2140
    },
    {
      "entropy": 0.5531029619276524,
      "epoch": 4.527148128624144,
      "grad_norm": 6.6875,
      "learning_rate": 3.12e-06,
      "loss": 0.5480860233306885,
      "mean_token_accuracy": 0.8185123346745968,
      "num_tokens": 11405274.0,
      "step": 2150
    },
    {
      "entropy": 0.5629217986017465,
      "epoch": 4.548234053769109,
      "grad_norm": 7.40625,
      "learning_rate": 3.0311111111111115e-06,
      "loss": 0.5423421859741211,
      "mean_token_accuracy": 0.8194163963198662,
      "num_tokens": 11459305.0,
      "step": 2160
    },
    {
      "entropy": 0.5574934646487236,
      "epoch": 4.569319978914075,
      "grad_norm": 7.40625,
      "learning_rate": 2.9422222222222224e-06,
      "loss": 0.5479726791381836,
      "mean_token_accuracy": 0.8106288813054562,
      "num_tokens": 11512317.0,
      "step": 2170
    },
    {
      "entropy": 0.5610745966434478,
      "epoch": 4.590405904059041,
      "grad_norm": 7.96875,
      "learning_rate": 2.8533333333333337e-06,
      "loss": 0.5556921005249024,
      "mean_token_accuracy": 0.8137760147452354,
      "num_tokens": 11566146.0,
      "step": 2180
    },
    {
      "entropy": 0.5264041472226382,
      "epoch": 4.611491829204006,
      "grad_norm": 6.71875,
      "learning_rate": 2.7644444444444446e-06,
      "loss": 0.535123348236084,
      "mean_token_accuracy": 0.8231086760759354,
      "num_tokens": 11619757.0,
      "step": 2190
    },
    {
      "entropy": 0.5513654734939337,
      "epoch": 4.632577754348972,
      "grad_norm": 7.28125,
      "learning_rate": 2.675555555555556e-06,
      "loss": 0.5477256774902344,
      "mean_token_accuracy": 0.8165041409432888,
      "num_tokens": 11673528.0,
      "step": 2200
    },
    {
      "entropy": 0.5440514117479325,
      "epoch": 4.653663679493937,
      "grad_norm": 7.15625,
      "learning_rate": 2.5866666666666667e-06,
      "loss": 0.5353331565856934,
      "mean_token_accuracy": 0.8190776027739048,
      "num_tokens": 11728320.0,
      "step": 2210
    },
    {
      "entropy": 0.5514211151748896,
      "epoch": 4.674749604638904,
      "grad_norm": 6.6875,
      "learning_rate": 2.497777777777778e-06,
      "loss": 0.5415604114532471,
      "mean_token_accuracy": 0.8205962501466274,
      "num_tokens": 11780028.0,
      "step": 2220
    },
    {
      "entropy": 0.5382947374135256,
      "epoch": 4.695835529783869,
      "grad_norm": 8.375,
      "learning_rate": 2.4088888888888894e-06,
      "loss": 0.5256577491760254,
      "mean_token_accuracy": 0.8235763169825077,
      "num_tokens": 11830895.0,
      "step": 2230
    },
    {
      "entropy": 0.54562176913023,
      "epoch": 4.716921454928835,
      "grad_norm": 7.1875,
      "learning_rate": 2.3200000000000002e-06,
      "loss": 0.5267295837402344,
      "mean_token_accuracy": 0.8207360960543155,
      "num_tokens": 11882155.0,
      "step": 2240
    },
    {
      "entropy": 0.5469693087041378,
      "epoch": 4.7380073800738005,
      "grad_norm": 7.28125,
      "learning_rate": 2.2311111111111115e-06,
      "loss": 0.5502583026885987,
      "mean_token_accuracy": 0.812764473259449,
      "num_tokens": 11936554.0,
      "step": 2250
    },
    {
      "entropy": 0.5447085045278073,
      "epoch": 4.759093305218767,
      "grad_norm": 6.5,
      "learning_rate": 2.1422222222222224e-06,
      "loss": 0.5389336585998535,
      "mean_token_accuracy": 0.820573166757822,
      "num_tokens": 11988156.0,
      "step": 2260
    },
    {
      "entropy": 0.5513703927397728,
      "epoch": 4.780179230363732,
      "grad_norm": 7.59375,
      "learning_rate": 2.0533333333333337e-06,
      "loss": 0.5426520347595215,
      "mean_token_accuracy": 0.8161885485053062,
      "num_tokens": 12041543.0,
      "step": 2270
    },
    {
      "entropy": 0.5453945960849523,
      "epoch": 4.801265155508698,
      "grad_norm": 8.3125,
      "learning_rate": 1.9644444444444446e-06,
      "loss": 0.5514724731445313,
      "mean_token_accuracy": 0.8171669006347656,
      "num_tokens": 12092858.0,
      "step": 2280
    },
    {
      "entropy": 0.5482615809887648,
      "epoch": 4.822351080653664,
      "grad_norm": 8.1875,
      "learning_rate": 1.8755555555555557e-06,
      "loss": 0.5525861263275147,
      "mean_token_accuracy": 0.8130425401031971,
      "num_tokens": 12144862.0,
      "step": 2290
    },
    {
      "entropy": 0.5636871963739395,
      "epoch": 4.84343700579863,
      "grad_norm": 7.34375,
      "learning_rate": 1.7866666666666668e-06,
      "loss": 0.5567370414733886,
      "mean_token_accuracy": 0.8114767424762249,
      "num_tokens": 12197932.0,
      "step": 2300
    },
    {
      "entropy": 0.5445570610463619,
      "epoch": 4.864522930943595,
      "grad_norm": 6.875,
      "learning_rate": 1.6977777777777779e-06,
      "loss": 0.5326248645782471,
      "mean_token_accuracy": 0.8205311246216297,
      "num_tokens": 12249536.0,
      "step": 2310
    },
    {
      "entropy": 0.567267632111907,
      "epoch": 4.885608856088561,
      "grad_norm": 7.125,
      "learning_rate": 1.608888888888889e-06,
      "loss": 0.5614665031433106,
      "mean_token_accuracy": 0.8123429246246815,
      "num_tokens": 12302328.0,
      "step": 2320
    },
    {
      "entropy": 0.5418668888509274,
      "epoch": 4.906694781233527,
      "grad_norm": 7.09375,
      "learning_rate": 1.52e-06,
      "loss": 0.5339171409606933,
      "mean_token_accuracy": 0.821372539550066,
      "num_tokens": 12354907.0,
      "step": 2330
    },
    {
      "entropy": 0.5500333663076162,
      "epoch": 4.927780706378492,
      "grad_norm": 7.4375,
      "learning_rate": 1.4311111111111111e-06,
      "loss": 0.5362411499023437,
      "mean_token_accuracy": 0.8219335168600083,
      "num_tokens": 12409864.0,
      "step": 2340
    },
    {
      "entropy": 0.551620377600193,
      "epoch": 4.948866631523458,
      "grad_norm": 7.5,
      "learning_rate": 1.3422222222222222e-06,
      "loss": 0.5413876056671143,
      "mean_token_accuracy": 0.8180868111550808,
      "num_tokens": 12463400.0,
      "step": 2350
    },
    {
      "entropy": 0.5467831913381815,
      "epoch": 4.9699525566684235,
      "grad_norm": 7.0,
      "learning_rate": 1.2533333333333333e-06,
      "loss": 0.5441878318786622,
      "mean_token_accuracy": 0.8187509059906006,
      "num_tokens": 12515674.0,
      "step": 2360
    },
    {
      "entropy": 0.5409926638007164,
      "epoch": 4.99103848181339,
      "grad_norm": 6.65625,
      "learning_rate": 1.1644444444444446e-06,
      "loss": 0.5396465301513672,
      "mean_token_accuracy": 0.8193774409592152,
      "num_tokens": 12568384.0,
      "step": 2370
    },
    {
      "entropy": 0.5556026499819111,
      "epoch": 5.010542962572483,
      "grad_norm": 7.59375,
      "learning_rate": 1.0755555555555557e-06,
      "loss": 0.549100399017334,
      "mean_token_accuracy": 0.8160284652903274,
      "num_tokens": 12616599.0,
      "step": 2380
    },
    {
      "entropy": 0.5403375245630742,
      "epoch": 5.031628887717448,
      "grad_norm": 6.6875,
      "learning_rate": 9.866666666666668e-07,
      "loss": 0.5360908508300781,
      "mean_token_accuracy": 0.8230746820569038,
      "num_tokens": 12670346.0,
      "step": 2390
    },
    {
      "entropy": 0.5554473079741001,
      "epoch": 5.052714812862415,
      "grad_norm": 6.84375,
      "learning_rate": 8.977777777777778e-07,
      "loss": 0.5354344844818115,
      "mean_token_accuracy": 0.82228574603796,
      "num_tokens": 12722166.0,
      "step": 2400
    },
    {
      "entropy": 0.5599666737020016,
      "epoch": 5.07380073800738,
      "grad_norm": 6.5,
      "learning_rate": 8.088888888888889e-07,
      "loss": 0.5506783962249756,
      "mean_token_accuracy": 0.8167558930814266,
      "num_tokens": 12772478.0,
      "step": 2410
    },
    {
      "entropy": 0.5462040238082408,
      "epoch": 5.094886663152346,
      "grad_norm": 7.03125,
      "learning_rate": 7.2e-07,
      "loss": 0.5262123584747315,
      "mean_token_accuracy": 0.8223468787968159,
      "num_tokens": 12828399.0,
      "step": 2420
    },
    {
      "entropy": 0.5505207426846027,
      "epoch": 5.1159725882973115,
      "grad_norm": 7.6875,
      "learning_rate": 6.311111111111112e-07,
      "loss": 0.5430339813232422,
      "mean_token_accuracy": 0.8183211967349052,
      "num_tokens": 12882883.0,
      "step": 2430
    },
    {
      "entropy": 0.5480441506952047,
      "epoch": 5.137058513442278,
      "grad_norm": 7.4375,
      "learning_rate": 5.422222222222223e-07,
      "loss": 0.5462999820709229,
      "mean_token_accuracy": 0.817756624519825,
      "num_tokens": 12931900.0,
      "step": 2440
    },
    {
      "entropy": 0.5532797310501337,
      "epoch": 5.158144438587243,
      "grad_norm": 7.625,
      "learning_rate": 4.533333333333334e-07,
      "loss": 0.5435749530792237,
      "mean_token_accuracy": 0.816430265456438,
      "num_tokens": 12983728.0,
      "step": 2450
    },
    {
      "entropy": 0.5532240316271781,
      "epoch": 5.179230363732208,
      "grad_norm": 7.4375,
      "learning_rate": 3.644444444444445e-07,
      "loss": 0.556332015991211,
      "mean_token_accuracy": 0.8138340756297111,
      "num_tokens": 13033055.0,
      "step": 2460
    },
    {
      "entropy": 0.5663877837359905,
      "epoch": 5.200316288877175,
      "grad_norm": 7.5,
      "learning_rate": 2.7555555555555555e-07,
      "loss": 0.5551712036132812,
      "mean_token_accuracy": 0.8142496928572655,
      "num_tokens": 13088474.0,
      "step": 2470
    },
    {
      "entropy": 0.540856496989727,
      "epoch": 5.22140221402214,
      "grad_norm": 6.65625,
      "learning_rate": 1.866666666666667e-07,
      "loss": 0.5255017280578613,
      "mean_token_accuracy": 0.8260384507477283,
      "num_tokens": 13143575.0,
      "step": 2480
    },
    {
      "entropy": 0.5562540978193283,
      "epoch": 5.242488139167106,
      "grad_norm": 7.21875,
      "learning_rate": 9.777777777777779e-08,
      "loss": 0.5427064895629883,
      "mean_token_accuracy": 0.8161332339048386,
      "num_tokens": 13194995.0,
      "step": 2490
    },
    {
      "entropy": 0.5543773144483566,
      "epoch": 5.263574064312071,
      "grad_norm": 7.40625,
      "learning_rate": 8.88888888888889e-09,
      "loss": 0.5387593269348144,
      "mean_token_accuracy": 0.8183573611080647,
      "num_tokens": 13248069.0,
      "step": 2500
    }
  ],
  "logging_steps": 10,
  "max_steps": 2500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.3554811993811558e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}