{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.61,
  "eval_steps": 500,
  "global_step": 61000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.001,
      "grad_norm": 17.39519691467285,
      "learning_rate": 2.97e-05,
      "loss": 9.7941,
      "num_input_tokens_seen": 6553600,
      "step": 100,
      "train_runtime": 74.0623,
      "train_tokens_per_second": 88487.632
    },
    {
      "epoch": 0.002,
      "grad_norm": 10.212440490722656,
      "learning_rate": 5.97e-05,
      "loss": 1.0389,
      "num_input_tokens_seen": 13107200,
      "step": 200,
      "train_runtime": 135.0365,
      "train_tokens_per_second": 97064.126
    },
    {
      "epoch": 0.003,
      "grad_norm": 6.982235908508301,
      "learning_rate": 8.969999999999998e-05,
      "loss": 0.7951,
      "num_input_tokens_seen": 19660800,
      "step": 300,
      "train_runtime": 196.4342,
      "train_tokens_per_second": 100088.472
    },
    {
      "epoch": 0.004,
      "grad_norm": 2.089735507965088,
      "learning_rate": 0.0001197,
      "loss": 0.6341,
      "num_input_tokens_seen": 26214400,
      "step": 400,
      "train_runtime": 257.5653,
      "train_tokens_per_second": 101777.682
    },
    {
      "epoch": 0.005,
      "grad_norm": 2.6269969940185547,
      "learning_rate": 0.00014969999999999998,
      "loss": 0.5353,
      "num_input_tokens_seen": 32768000,
      "step": 500,
      "train_runtime": 323.599,
      "train_tokens_per_second": 101261.143
    },
    {
      "epoch": 0.006,
      "grad_norm": 0.9126470685005188,
      "learning_rate": 0.00017969999999999998,
      "loss": 0.4822,
      "num_input_tokens_seen": 39321600,
      "step": 600,
      "train_runtime": 385.3073,
      "train_tokens_per_second": 102052.566
    },
    {
      "epoch": 0.007,
      "grad_norm": 0.7452394366264343,
      "learning_rate": 0.00020969999999999997,
      "loss": 0.4534,
      "num_input_tokens_seen": 45875200,
      "step": 700,
      "train_runtime": 447.534,
      "train_tokens_per_second": 102506.63
    },
    {
      "epoch": 0.008,
      "grad_norm": 0.6909123659133911,
      "learning_rate": 0.0002397,
      "loss": 0.4323,
      "num_input_tokens_seen": 52428800,
      "step": 800,
      "train_runtime": 510.1043,
      "train_tokens_per_second": 102780.558
    },
    {
      "epoch": 0.009,
      "grad_norm": 0.5689504146575928,
      "learning_rate": 0.0002697,
      "loss": 0.4262,
      "num_input_tokens_seen": 58982400,
      "step": 900,
      "train_runtime": 571.3595,
      "train_tokens_per_second": 103231.669
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.42208704352378845,
      "learning_rate": 0.00029969999999999997,
      "loss": 0.4158,
      "num_input_tokens_seen": 65536000,
      "step": 1000,
      "train_runtime": 638.5123,
      "train_tokens_per_second": 102638.586
    },
    {
      "epoch": 0.011,
      "grad_norm": 0.4542798399925232,
      "learning_rate": 0.00029999925978027874,
      "loss": 0.4127,
      "num_input_tokens_seen": 72089600,
      "step": 1100,
      "train_runtime": 698.6527,
      "train_tokens_per_second": 103183.742
    },
    {
      "epoch": 0.012,
      "grad_norm": 0.4086480736732483,
      "learning_rate": 0.0002999970091452017,
      "loss": 0.4018,
      "num_input_tokens_seen": 78643200,
      "step": 1200,
      "train_runtime": 761.7182,
      "train_tokens_per_second": 103244.485
    },
    {
      "epoch": 0.013,
      "grad_norm": 0.37623685598373413,
      "learning_rate": 0.00029999324804190795,
      "loss": 0.3969,
      "num_input_tokens_seen": 85196800,
      "step": 1300,
      "train_runtime": 827.9033,
      "train_tokens_per_second": 102906.7
    },
    {
      "epoch": 0.014,
      "grad_norm": 0.3346163332462311,
      "learning_rate": 0.0002999879765082716,
      "loss": 0.3906,
      "num_input_tokens_seen": 91750400,
      "step": 1400,
      "train_runtime": 889.5401,
      "train_tokens_per_second": 103143.635
    },
    {
      "epoch": 0.015,
      "grad_norm": 0.4093320369720459,
      "learning_rate": 0.000299981194597377,
      "loss": 0.3852,
      "num_input_tokens_seen": 98304000,
      "step": 1500,
      "train_runtime": 950.9359,
      "train_tokens_per_second": 103376.055
    },
    {
      "epoch": 0.016,
      "grad_norm": 0.3808560371398926,
      "learning_rate": 0.0002999729023775179,
      "loss": 0.3819,
      "num_input_tokens_seen": 104857600,
      "step": 1600,
      "train_runtime": 1017.4047,
      "train_tokens_per_second": 103063.807
    },
    {
      "epoch": 0.017,
      "grad_norm": 0.3014701306819916,
      "learning_rate": 0.0002999630999321969,
      "loss": 0.387,
      "num_input_tokens_seen": 111411200,
      "step": 1700,
      "train_runtime": 1075.027,
      "train_tokens_per_second": 103635.721
    },
    {
      "epoch": 0.018,
      "grad_norm": 0.25073230266571045,
      "learning_rate": 0.00029995178736012443,
      "loss": 0.382,
      "num_input_tokens_seen": 117964800,
      "step": 1800,
      "train_runtime": 1141.6684,
      "train_tokens_per_second": 103326.671
    },
    {
      "epoch": 0.019,
      "grad_norm": 0.2569698989391327,
      "learning_rate": 0.0002999389647752181,
      "loss": 0.3745,
      "num_input_tokens_seen": 124518400,
      "step": 1900,
      "train_runtime": 1202.9974,
      "train_tokens_per_second": 103506.793
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.2895148694515228,
      "learning_rate": 0.00029992463230660104,
      "loss": 0.3747,
      "num_input_tokens_seen": 131072000,
      "step": 2000,
      "train_runtime": 1271.272,
      "train_tokens_per_second": 103103.035
    },
    {
      "epoch": 0.021,
      "grad_norm": 0.28352853655815125,
      "learning_rate": 0.00029990879009860117,
      "loss": 0.3701,
      "num_input_tokens_seen": 137625600,
      "step": 2100,
      "train_runtime": 1335.8501,
      "train_tokens_per_second": 103024.736
    },
    {
      "epoch": 0.022,
      "grad_norm": 0.2598542273044586,
      "learning_rate": 0.0002998914383107493,
      "loss": 0.3715,
      "num_input_tokens_seen": 144179200,
      "step": 2200,
      "train_runtime": 1400.0516,
      "train_tokens_per_second": 102981.347
    },
    {
      "epoch": 0.023,
      "grad_norm": 0.300857275724411,
      "learning_rate": 0.0002998725771177778,
      "loss": 0.3723,
      "num_input_tokens_seen": 150732800,
      "step": 2300,
      "train_runtime": 1465.03,
      "train_tokens_per_second": 102887.178
    },
    {
      "epoch": 0.024,
      "grad_norm": 0.19827991724014282,
      "learning_rate": 0.00029985220670961847,
      "loss": 0.3654,
      "num_input_tokens_seen": 157286400,
      "step": 2400,
      "train_runtime": 1534.4652,
      "train_tokens_per_second": 102502.423
    },
    {
      "epoch": 0.025,
      "grad_norm": 0.36876365542411804,
      "learning_rate": 0.0002998303272914014,
      "loss": 0.368,
      "num_input_tokens_seen": 163840000,
      "step": 2500,
      "train_runtime": 1598.5928,
      "train_tokens_per_second": 102490.141
    },
    {
      "epoch": 0.026,
      "grad_norm": 0.23755036294460297,
      "learning_rate": 0.00029980693908345185,
      "loss": 0.3648,
      "num_input_tokens_seen": 170393600,
      "step": 2600,
      "train_runtime": 1661.9675,
      "train_tokens_per_second": 102525.227
    },
    {
      "epoch": 0.027,
      "grad_norm": 0.3921568691730499,
      "learning_rate": 0.00029978204232128895,
      "loss": 0.3633,
      "num_input_tokens_seen": 176947200,
      "step": 2700,
      "train_runtime": 1731.9606,
      "train_tokens_per_second": 102165.837
    },
    {
      "epoch": 0.028,
      "grad_norm": 0.1964094191789627,
      "learning_rate": 0.0002997556372556227,
      "loss": 0.365,
      "num_input_tokens_seen": 183500800,
      "step": 2800,
      "train_runtime": 1796.4926,
      "train_tokens_per_second": 102143.922
    },
    {
      "epoch": 0.029,
      "grad_norm": 0.2469199150800705,
      "learning_rate": 0.0002997277241523519,
      "loss": 0.364,
      "num_input_tokens_seen": 190054400,
      "step": 2900,
      "train_runtime": 1860.3342,
      "train_tokens_per_second": 102161.428
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.19437766075134277,
      "learning_rate": 0.00029969830329256125,
      "loss": 0.3574,
      "num_input_tokens_seen": 196608000,
      "step": 3000,
      "train_runtime": 1924.7283,
      "train_tokens_per_second": 102148.444
    },
    {
      "epoch": 0.031,
      "grad_norm": 0.23198598623275757,
      "learning_rate": 0.00029966737497251836,
      "loss": 0.3599,
      "num_input_tokens_seen": 203161600,
      "step": 3100,
      "train_runtime": 1993.345,
      "train_tokens_per_second": 101919.94
    },
    {
      "epoch": 0.032,
      "grad_norm": 0.22857527434825897,
      "learning_rate": 0.0002996349395036711,
      "loss": 0.3579,
      "num_input_tokens_seen": 209715200,
      "step": 3200,
      "train_runtime": 2057.8023,
      "train_tokens_per_second": 101912.218
    },
    {
      "epoch": 0.033,
      "grad_norm": 0.24812710285186768,
      "learning_rate": 0.00029960099721264435,
      "loss": 0.3612,
      "num_input_tokens_seen": 216268800,
      "step": 3300,
      "train_runtime": 2121.9536,
      "train_tokens_per_second": 101919.666
    },
    {
      "epoch": 0.034,
      "grad_norm": 0.21982239186763763,
      "learning_rate": 0.0002995655484412365,
      "loss": 0.3554,
      "num_input_tokens_seen": 222822400,
      "step": 3400,
      "train_runtime": 2186.6347,
      "train_tokens_per_second": 101901.979
    },
    {
      "epoch": 0.035,
      "grad_norm": 0.3460980951786041,
      "learning_rate": 0.00029952859354641636,
      "loss": 0.3568,
      "num_input_tokens_seen": 229376000,
      "step": 3500,
      "train_runtime": 2256.5384,
      "train_tokens_per_second": 101649.502
    },
    {
      "epoch": 0.036,
      "grad_norm": 0.25577911734580994,
      "learning_rate": 0.00029949013290031924,
      "loss": 0.354,
      "num_input_tokens_seen": 235929600,
      "step": 3600,
      "train_runtime": 2320.5776,
      "train_tokens_per_second": 101668.483
    },
    {
      "epoch": 0.037,
      "grad_norm": 0.16108086705207825,
      "learning_rate": 0.00029945016689024353,
      "loss": 0.3509,
      "num_input_tokens_seen": 242483200,
      "step": 3700,
      "train_runtime": 2383.8992,
      "train_tokens_per_second": 101717.051
    },
    {
      "epoch": 0.038,
      "grad_norm": 0.2431662529706955,
      "learning_rate": 0.0002994086959186464,
      "loss": 0.3527,
      "num_input_tokens_seen": 249036800,
      "step": 3800,
      "train_runtime": 2448.8427,
      "train_tokens_per_second": 101695.71
    },
    {
      "epoch": 0.039,
      "grad_norm": 0.18574966490268707,
      "learning_rate": 0.00029936572040314014,
      "loss": 0.3546,
      "num_input_tokens_seen": 255590400,
      "step": 3900,
      "train_runtime": 2518.1288,
      "train_tokens_per_second": 101500.13
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.15902996063232422,
      "learning_rate": 0.0002993212407764877,
      "loss": 0.3519,
      "num_input_tokens_seen": 262144000,
      "step": 4000,
      "train_runtime": 2581.8809,
      "train_tokens_per_second": 101532.18
    },
    {
      "epoch": 0.041,
      "grad_norm": 0.21019065380096436,
      "learning_rate": 0.00029927525748659834,
      "loss": 0.3567,
      "num_input_tokens_seen": 268697600,
      "step": 4100,
      "train_runtime": 2646.5068,
      "train_tokens_per_second": 101529.154
    },
    {
      "epoch": 0.042,
      "grad_norm": 0.18648174405097961,
      "learning_rate": 0.0002992277709965234,
      "loss": 0.3512,
      "num_input_tokens_seen": 275251200,
      "step": 4200,
      "train_runtime": 2710.4754,
      "train_tokens_per_second": 101550.895
    },
    {
      "epoch": 0.043,
      "grad_norm": 0.21123889088630676,
      "learning_rate": 0.0002991787817844513,
      "loss": 0.3521,
      "num_input_tokens_seen": 281804800,
      "step": 4300,
      "train_runtime": 2780.6173,
      "train_tokens_per_second": 101346.13
    },
    {
      "epoch": 0.044,
      "grad_norm": 0.22183509171009064,
      "learning_rate": 0.0002991282903437028,
      "loss": 0.3486,
      "num_input_tokens_seen": 288358400,
      "step": 4400,
      "train_runtime": 2843.584,
      "train_tokens_per_second": 101406.674
    },
    {
      "epoch": 0.045,
      "grad_norm": 0.19213925302028656,
      "learning_rate": 0.0002990762971827262,
      "loss": 0.3481,
      "num_input_tokens_seen": 294912000,
      "step": 4500,
      "train_runtime": 2906.5309,
      "train_tokens_per_second": 101465.29
    },
    {
      "epoch": 0.046,
      "grad_norm": 0.16215530037879944,
      "learning_rate": 0.00029902280282509197,
      "loss": 0.3506,
      "num_input_tokens_seen": 301465600,
      "step": 4600,
      "train_runtime": 2977.8135,
      "train_tokens_per_second": 101237.232
    },
    {
      "epoch": 0.047,
      "grad_norm": 0.17120705544948578,
      "learning_rate": 0.0002989678078094878,
      "loss": 0.3433,
      "num_input_tokens_seen": 308019200,
      "step": 4700,
      "train_runtime": 3040.7538,
      "train_tokens_per_second": 101296.988
    },
    {
      "epoch": 0.048,
      "grad_norm": 0.26389873027801514,
      "learning_rate": 0.00029891131268971284,
      "loss": 0.345,
      "num_input_tokens_seen": 314572800,
      "step": 4800,
      "train_runtime": 3104.3446,
      "train_tokens_per_second": 101333.081
    },
    {
      "epoch": 0.049,
      "grad_norm": 0.1639779806137085,
      "learning_rate": 0.0002988533180346723,
      "loss": 0.3431,
      "num_input_tokens_seen": 321126400,
      "step": 4900,
      "train_runtime": 3172.6385,
      "train_tokens_per_second": 101217.457
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.21486082673072815,
      "learning_rate": 0.0002987938244283717,
      "loss": 0.3413,
      "num_input_tokens_seen": 327680000,
      "step": 5000,
      "train_runtime": 3237.5961,
      "train_tokens_per_second": 101210.896
    },
    {
      "epoch": 0.051,
      "grad_norm": 0.20326170325279236,
      "learning_rate": 0.00029873283246991105,
      "loss": 0.3457,
      "num_input_tokens_seen": 334233600,
      "step": 5100,
      "train_runtime": 3302.3096,
      "train_tokens_per_second": 101212.074
    },
    {
      "epoch": 0.052,
      "grad_norm": 0.171161487698555,
      "learning_rate": 0.0002986703427734787,
      "loss": 0.345,
      "num_input_tokens_seen": 340787200,
      "step": 5200,
      "train_runtime": 3367.4928,
      "train_tokens_per_second": 101199.089
    },
    {
      "epoch": 0.053,
      "grad_norm": 0.19781792163848877,
      "learning_rate": 0.00029860635596834517,
      "loss": 0.3455,
      "num_input_tokens_seen": 347340800,
      "step": 5300,
      "train_runtime": 3430.9148,
      "train_tokens_per_second": 101238.538
    },
    {
      "epoch": 0.054,
      "grad_norm": 0.1795511543750763,
      "learning_rate": 0.0002985408726988569,
      "loss": 0.3439,
      "num_input_tokens_seen": 353894400,
      "step": 5400,
      "train_runtime": 3498.4556,
      "train_tokens_per_second": 101157.322
    },
    {
      "epoch": 0.055,
      "grad_norm": 0.1671728938817978,
      "learning_rate": 0.0002984738936244296,
      "loss": 0.3422,
      "num_input_tokens_seen": 360448000,
      "step": 5500,
      "train_runtime": 3561.4394,
      "train_tokens_per_second": 101208.516
    },
    {
      "epoch": 0.056,
      "grad_norm": 0.17824003100395203,
      "learning_rate": 0.0002984054194195419,
      "loss": 0.3489,
      "num_input_tokens_seen": 367001600,
      "step": 5600,
      "train_runtime": 3625.8956,
      "train_tokens_per_second": 101216.814
    },
    {
      "epoch": 0.057,
      "grad_norm": 0.1654757708311081,
      "learning_rate": 0.0002983354507737283,
      "loss": 0.3463,
      "num_input_tokens_seen": 373555200,
      "step": 5700,
      "train_runtime": 3690.173,
      "train_tokens_per_second": 101229.725
    },
    {
      "epoch": 0.058,
      "grad_norm": 0.2033533751964569,
      "learning_rate": 0.00029826398839157215,
      "loss": 0.3462,
      "num_input_tokens_seen": 380108800,
      "step": 5800,
      "train_runtime": 3759.2019,
      "train_tokens_per_second": 101114.229
    },
    {
      "epoch": 0.059,
      "grad_norm": 0.19753150641918182,
      "learning_rate": 0.000298191032992699,
      "loss": 0.3436,
      "num_input_tokens_seen": 386662400,
      "step": 5900,
      "train_runtime": 3822.1964,
      "train_tokens_per_second": 101162.357
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.13978537917137146,
      "learning_rate": 0.0002981165853117688,
      "loss": 0.3393,
      "num_input_tokens_seen": 393216000,
      "step": 6000,
      "train_runtime": 3890.9859,
      "train_tokens_per_second": 101058.192
    },
    {
      "epoch": 0.061,
      "grad_norm": 0.28539636731147766,
      "learning_rate": 0.000298040646098469,
      "loss": 0.3419,
      "num_input_tokens_seen": 399769600,
      "step": 6100,
      "train_runtime": 3955.42,
      "train_tokens_per_second": 101068.813
    },
    {
      "epoch": 0.062,
      "grad_norm": 0.14195021986961365,
      "learning_rate": 0.0002979632161175064,
      "loss": 0.3408,
      "num_input_tokens_seen": 406323200,
      "step": 6200,
      "train_runtime": 4019.3462,
      "train_tokens_per_second": 101091.865
    },
    {
      "epoch": 0.063,
      "grad_norm": 0.26058393716812134,
      "learning_rate": 0.0002978842961486003,
      "loss": 0.3411,
      "num_input_tokens_seen": 412876800,
      "step": 6300,
      "train_runtime": 4082.619,
      "train_tokens_per_second": 101130.379
    },
    {
      "epoch": 0.064,
      "grad_norm": 0.1645655333995819,
      "learning_rate": 0.0002978038869864738,
      "loss": 0.3392,
      "num_input_tokens_seen": 419430400,
      "step": 6400,
      "train_runtime": 4152.2955,
      "train_tokens_per_second": 101011.694
    },
    {
      "epoch": 0.065,
      "grad_norm": 0.1678280532360077,
      "learning_rate": 0.0002977219894408463,
      "loss": 0.338,
      "num_input_tokens_seen": 425984000,
      "step": 6500,
      "train_runtime": 4215.8141,
      "train_tokens_per_second": 101044.304
    },
    {
      "epoch": 0.066,
      "grad_norm": 0.19337573647499084,
      "learning_rate": 0.0002976386043364251,
      "loss": 0.3424,
      "num_input_tokens_seen": 432537600,
      "step": 6600,
      "train_runtime": 4278.8465,
      "train_tokens_per_second": 101087.432
    },
    {
      "epoch": 0.067,
      "grad_norm": 0.14295175671577454,
      "learning_rate": 0.00029755373251289733,
      "loss": 0.3443,
      "num_input_tokens_seen": 439091200,
      "step": 6700,
      "train_runtime": 4348.6665,
      "train_tokens_per_second": 100971.459
    },
    {
      "epoch": 0.068,
      "grad_norm": 0.22164900600910187,
      "learning_rate": 0.0002974673748249213,
      "loss": 0.339,
      "num_input_tokens_seen": 445644800,
      "step": 6800,
      "train_runtime": 4413.12,
      "train_tokens_per_second": 100981.799
    },
    {
      "epoch": 0.069,
      "grad_norm": 0.1831408590078354,
      "learning_rate": 0.00029737953214211804,
      "loss": 0.3398,
      "num_input_tokens_seen": 452198400,
      "step": 6900,
      "train_runtime": 4477.6672,
      "train_tokens_per_second": 100989.73
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.21329298615455627,
      "learning_rate": 0.0002972902053490623,
      "loss": 0.3372,
      "num_input_tokens_seen": 458752000,
      "step": 7000,
      "train_runtime": 4541.4752,
      "train_tokens_per_second": 101013.873
    },
    {
      "epoch": 0.071,
      "grad_norm": 0.16601704061031342,
      "learning_rate": 0.00029719939534527393,
      "loss": 0.3436,
      "num_input_tokens_seen": 465305600,
      "step": 7100,
      "train_runtime": 4607.1943,
      "train_tokens_per_second": 100995.436
    },
    {
      "epoch": 0.072,
      "grad_norm": 0.2303948849439621,
      "learning_rate": 0.00029710710304520866,
      "loss": 0.339,
      "num_input_tokens_seen": 471859200,
      "step": 7200,
      "train_runtime": 4672.0421,
      "train_tokens_per_second": 100996.349
    },
    {
      "epoch": 0.073,
      "grad_norm": 0.21449029445648193,
      "learning_rate": 0.00029701332937824885,
      "loss": 0.336,
      "num_input_tokens_seen": 478412800,
      "step": 7300,
      "train_runtime": 4742.0375,
      "train_tokens_per_second": 100887.605
    },
    {
      "epoch": 0.074,
      "grad_norm": 0.1367533802986145,
      "learning_rate": 0.0002969180752886944,
      "loss": 0.3397,
      "num_input_tokens_seen": 484966400,
      "step": 7400,
      "train_runtime": 4805.1341,
      "train_tokens_per_second": 100926.716
    },
    {
      "epoch": 0.075,
      "grad_norm": 0.1852603256702423,
      "learning_rate": 0.0002968213417357529,
      "loss": 0.34,
      "num_input_tokens_seen": 491520000,
      "step": 7500,
      "train_runtime": 4867.6611,
      "train_tokens_per_second": 100976.628
    },
    {
      "epoch": 0.076,
      "grad_norm": 0.18590585887432098,
      "learning_rate": 0.00029672312969353015,
      "loss": 0.3375,
      "num_input_tokens_seen": 498073600,
      "step": 7600,
      "train_runtime": 4938.9456,
      "train_tokens_per_second": 100846.14
    },
    {
      "epoch": 0.077,
      "grad_norm": 0.17078232765197754,
      "learning_rate": 0.00029662344015102027,
      "loss": 0.3374,
      "num_input_tokens_seen": 504627200,
      "step": 7700,
      "train_runtime": 5003.5948,
      "train_tokens_per_second": 100852.931
    },
    {
      "epoch": 0.078,
      "grad_norm": 0.14574670791625977,
      "learning_rate": 0.00029652227411209594,
      "loss": 0.3369,
      "num_input_tokens_seen": 511180800,
      "step": 7800,
      "train_runtime": 5067.2522,
      "train_tokens_per_second": 100879.289
    },
    {
      "epoch": 0.079,
      "grad_norm": 0.1603483408689499,
      "learning_rate": 0.0002964196325954979,
      "loss": 0.3352,
      "num_input_tokens_seen": 517734400,
      "step": 7900,
      "train_runtime": 5131.2908,
      "train_tokens_per_second": 100897.497
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.16576310992240906,
      "learning_rate": 0.0002963155166348253,
      "loss": 0.3376,
      "num_input_tokens_seen": 524288000,
      "step": 8000,
      "train_runtime": 5200.6662,
      "train_tokens_per_second": 100811.699
    },
    {
      "epoch": 0.081,
      "grad_norm": 0.31833919882774353,
      "learning_rate": 0.0002962099272785246,
      "loss": 0.3382,
      "num_input_tokens_seen": 530841600,
      "step": 8100,
      "train_runtime": 5266.7639,
      "train_tokens_per_second": 100790.849
    },
    {
      "epoch": 0.082,
      "grad_norm": 0.14755409955978394,
      "learning_rate": 0.0002961028655898794,
      "loss": 0.3348,
      "num_input_tokens_seen": 537395200,
      "step": 8200,
      "train_runtime": 5331.3948,
      "train_tokens_per_second": 100798.238
    },
    {
      "epoch": 0.083,
      "grad_norm": 0.2060171663761139,
      "learning_rate": 0.0002959943326469998,
      "loss": 0.3338,
      "num_input_tokens_seen": 543948800,
      "step": 8300,
      "train_runtime": 5395.0396,
      "train_tokens_per_second": 100823.876
    },
    {
      "epoch": 0.084,
      "grad_norm": 0.16461625695228577,
      "learning_rate": 0.0002958843295428112,
      "loss": 0.3326,
      "num_input_tokens_seen": 550502400,
      "step": 8400,
      "train_runtime": 5458.2259,
      "train_tokens_per_second": 100857.387
    },
    {
      "epoch": 0.085,
      "grad_norm": 0.15455660223960876,
      "learning_rate": 0.0002957728573850438,
      "loss": 0.3339,
      "num_input_tokens_seen": 557056000,
      "step": 8500,
      "train_runtime": 5527.7417,
      "train_tokens_per_second": 100774.607
    },
    {
      "epoch": 0.086,
      "grad_norm": 0.17872081696987152,
      "learning_rate": 0.0002956599172962209,
      "loss": 0.3404,
      "num_input_tokens_seen": 563609600,
      "step": 8600,
      "train_runtime": 5593.3318,
      "train_tokens_per_second": 100764.557
    },
    {
      "epoch": 0.087,
      "grad_norm": 0.19022491574287415,
      "learning_rate": 0.0002955455104136479,
      "loss": 0.3329,
      "num_input_tokens_seen": 570163200,
      "step": 8700,
      "train_runtime": 5659.0887,
      "train_tokens_per_second": 100751.77
    },
    {
      "epoch": 0.088,
      "grad_norm": 0.14710059762001038,
      "learning_rate": 0.00029542963788940096,
      "loss": 0.3323,
      "num_input_tokens_seen": 576716800,
      "step": 8800,
      "train_runtime": 5722.168,
      "train_tokens_per_second": 100786.415
    },
    {
      "epoch": 0.089,
      "grad_norm": 0.1998033970594406,
      "learning_rate": 0.00029531230089031505,
      "loss": 0.3378,
      "num_input_tokens_seen": 583270400,
      "step": 8900,
      "train_runtime": 5787.7324,
      "train_tokens_per_second": 100777.016
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.125193253159523,
      "learning_rate": 0.0002951935005979724,
      "loss": 0.3325,
      "num_input_tokens_seen": 589824000,
      "step": 9000,
      "train_runtime": 5855.8455,
      "train_tokens_per_second": 100723.968
    },
    {
      "epoch": 0.091,
      "grad_norm": 0.19552631676197052,
      "learning_rate": 0.0002950732382086907,
      "loss": 0.3316,
      "num_input_tokens_seen": 596377600,
      "step": 9100,
      "train_runtime": 5921.9714,
      "train_tokens_per_second": 100705.923
    },
    {
      "epoch": 0.092,
      "grad_norm": 0.16468137502670288,
      "learning_rate": 0.0002949515149335108,
      "loss": 0.3349,
      "num_input_tokens_seen": 602931200,
      "step": 9200,
      "train_runtime": 5986.1243,
      "train_tokens_per_second": 100721.464
    },
    {
      "epoch": 0.093,
      "grad_norm": 0.1658785343170166,
      "learning_rate": 0.0002948283319981848,
      "loss": 0.3281,
      "num_input_tokens_seen": 609484800,
      "step": 9300,
      "train_runtime": 6050.7028,
      "train_tokens_per_second": 100729.588
    },
    {
      "epoch": 0.094,
      "grad_norm": 0.16668474674224854,
      "learning_rate": 0.00029470369064316354,
      "loss": 0.3301,
      "num_input_tokens_seen": 616038400,
      "step": 9400,
      "train_runtime": 6115.0892,
      "train_tokens_per_second": 100740.706
    },
    {
      "epoch": 0.095,
      "grad_norm": 0.16522246599197388,
      "learning_rate": 0.00029457759212358397,
      "loss": 0.3305,
      "num_input_tokens_seen": 622592000,
      "step": 9500,
      "train_runtime": 6183.2082,
      "train_tokens_per_second": 100690.77
    },
    {
      "epoch": 0.096,
      "grad_norm": 0.2229623645544052,
      "learning_rate": 0.00029445003770925686,
      "loss": 0.3289,
      "num_input_tokens_seen": 629145600,
      "step": 9600,
      "train_runtime": 6247.5147,
      "train_tokens_per_second": 100703.341
    },
    {
      "epoch": 0.097,
      "grad_norm": 0.16620689630508423,
      "learning_rate": 0.00029432102868465367,
      "loss": 0.3299,
      "num_input_tokens_seen": 635699200,
      "step": 9700,
      "train_runtime": 6312.7504,
      "train_tokens_per_second": 100700.829
    },
    {
      "epoch": 0.098,
      "grad_norm": 0.15970012545585632,
      "learning_rate": 0.0002941905663488939,
      "loss": 0.3292,
      "num_input_tokens_seen": 642252800,
      "step": 9800,
      "train_runtime": 6382.1987,
      "train_tokens_per_second": 100631.903
    },
    {
      "epoch": 0.099,
      "grad_norm": 0.14614014327526093,
      "learning_rate": 0.0002940586520157318,
      "loss": 0.3329,
      "num_input_tokens_seen": 648806400,
      "step": 9900,
      "train_runtime": 6445.6924,
      "train_tokens_per_second": 100657.362
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.16558828949928284,
      "learning_rate": 0.00029392528701354325,
      "loss": 0.3286,
      "num_input_tokens_seen": 655360000,
      "step": 10000,
      "train_runtime": 6509.151,
      "train_tokens_per_second": 100682.87
    },
    {
      "epoch": 0.101,
      "grad_norm": 0.1442118138074875,
      "learning_rate": 0.00029379047268531243,
      "loss": 0.3314,
      "num_input_tokens_seen": 661913600,
      "step": 10100,
      "train_runtime": 6575.3071,
      "train_tokens_per_second": 100666.568
    },
    {
      "epoch": 0.102,
      "grad_norm": 0.16007182002067566,
      "learning_rate": 0.00029365421038861795,
      "loss": 0.3326,
      "num_input_tokens_seen": 668467200,
      "step": 10200,
      "train_runtime": 6639.6314,
      "train_tokens_per_second": 100678.359
    },
    {
      "epoch": 0.103,
      "grad_norm": 0.1417239010334015,
      "learning_rate": 0.0002935165014956198,
      "loss": 0.3292,
      "num_input_tokens_seen": 675020800,
      "step": 10300,
      "train_runtime": 6704.2875,
      "train_tokens_per_second": 100684.942
    },
    {
      "epoch": 0.104,
      "grad_norm": 0.20092202723026276,
      "learning_rate": 0.0002933773473930448,
      "loss": 0.3251,
      "num_input_tokens_seen": 681574400,
      "step": 10400,
      "train_runtime": 6769.9733,
      "train_tokens_per_second": 100676.083
    },
    {
      "epoch": 0.105,
      "grad_norm": 0.12387008965015411,
      "learning_rate": 0.0002932367494821734,
      "loss": 0.3302,
      "num_input_tokens_seen": 688128000,
      "step": 10500,
      "train_runtime": 6840.7627,
      "train_tokens_per_second": 100592.292
    },
    {
      "epoch": 0.106,
      "grad_norm": 0.17865417897701263,
      "learning_rate": 0.00029309470917882497,
      "loss": 0.328,
      "num_input_tokens_seen": 694681600,
      "step": 10600,
      "train_runtime": 6905.9119,
      "train_tokens_per_second": 100592.305
    },
    {
      "epoch": 0.107,
      "grad_norm": 0.14125974476337433,
      "learning_rate": 0.0002929512279133437,
      "loss": 0.3296,
      "num_input_tokens_seen": 701235200,
      "step": 10700,
      "train_runtime": 6969.9941,
      "train_tokens_per_second": 100607.718
    },
    {
      "epoch": 0.108,
      "grad_norm": 0.15725336968898773,
      "learning_rate": 0.0002928063071305844,
      "loss": 0.3279,
      "num_input_tokens_seen": 707788800,
      "step": 10800,
      "train_runtime": 7032.9479,
      "train_tokens_per_second": 100638.994
    },
    {
      "epoch": 0.109,
      "grad_norm": 0.15254800021648407,
      "learning_rate": 0.0002926599482898978,
      "loss": 0.3276,
      "num_input_tokens_seen": 714342400,
      "step": 10900,
      "train_runtime": 7097.644,
      "train_tokens_per_second": 100645.002
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.23630526661872864,
      "learning_rate": 0.00029251215286511573,
      "loss": 0.3278,
      "num_input_tokens_seen": 720896000,
      "step": 11000,
      "train_runtime": 7167.7206,
      "train_tokens_per_second": 100575.348
    },
    {
      "epoch": 0.111,
      "grad_norm": 0.14799726009368896,
      "learning_rate": 0.00029236292234453647,
      "loss": 0.3264,
      "num_input_tokens_seen": 727449600,
      "step": 11100,
      "train_runtime": 7232.1207,
      "train_tokens_per_second": 100585.932
    },
    {
      "epoch": 0.112,
      "grad_norm": 0.17712198197841644,
      "learning_rate": 0.0002922122582309097,
      "loss": 0.3304,
      "num_input_tokens_seen": 734003200,
      "step": 11200,
      "train_runtime": 7296.7016,
      "train_tokens_per_second": 100593.835
    },
    {
      "epoch": 0.113,
      "grad_norm": 0.1620536595582962,
      "learning_rate": 0.0002920601620414215,
      "loss": 0.3266,
      "num_input_tokens_seen": 740556800,
      "step": 11300,
      "train_runtime": 7359.3874,
      "train_tokens_per_second": 100627.506
    },
    {
      "epoch": 0.114,
      "grad_norm": 0.1695978045463562,
      "learning_rate": 0.0002919066353076786,
      "loss": 0.3269,
      "num_input_tokens_seen": 747110400,
      "step": 11400,
      "train_runtime": 7425.5624,
      "train_tokens_per_second": 100613.308
    },
    {
      "epoch": 0.115,
      "grad_norm": 0.23728708922863007,
      "learning_rate": 0.00029175167957569366,
      "loss": 0.3269,
      "num_input_tokens_seen": 753664000,
      "step": 11500,
      "train_runtime": 7489.1752,
      "train_tokens_per_second": 100633.779
    },
    {
      "epoch": 0.116,
      "grad_norm": 0.14579418301582336,
      "learning_rate": 0.0002915952964058691,
      "loss": 0.3254,
      "num_input_tokens_seen": 760217600,
      "step": 11600,
      "train_runtime": 7559.1466,
      "train_tokens_per_second": 100569.237
    },
    {
      "epoch": 0.117,
      "grad_norm": 0.15569131076335907,
      "learning_rate": 0.00029143748737298173,
      "loss": 0.3309,
      "num_input_tokens_seen": 766771200,
      "step": 11700,
      "train_runtime": 7625.7219,
      "train_tokens_per_second": 100550.638
    },
    {
      "epoch": 0.118,
      "grad_norm": 0.15939873456954956,
      "learning_rate": 0.00029127825406616677,
      "loss": 0.3251,
      "num_input_tokens_seen": 773324800,
      "step": 11800,
      "train_runtime": 7690.5664,
      "train_tokens_per_second": 100554.987
    },
    {
      "epoch": 0.119,
      "grad_norm": 0.1355784833431244,
      "learning_rate": 0.0002911175980889019,
      "loss": 0.3287,
      "num_input_tokens_seen": 779878400,
      "step": 11900,
      "train_runtime": 7753.5378,
      "train_tokens_per_second": 100583.556
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.19504176080226898,
      "learning_rate": 0.00029095552105899095,
      "loss": 0.325,
      "num_input_tokens_seen": 786432000,
      "step": 12000,
      "train_runtime": 7817.9364,
      "train_tokens_per_second": 100593.297
    },
    {
      "epoch": 0.121,
      "grad_norm": 0.1594318449497223,
      "learning_rate": 0.0002907920246085478,
      "loss": 0.3242,
      "num_input_tokens_seen": 792985600,
      "step": 12100,
      "train_runtime": 7887.1116,
      "train_tokens_per_second": 100541.953
    },
    {
      "epoch": 0.122,
      "grad_norm": 0.15172167122364044,
      "learning_rate": 0.00029062711038397996,
      "loss": 0.3325,
      "num_input_tokens_seen": 799539200,
      "step": 12200,
      "train_runtime": 7952.1371,
      "train_tokens_per_second": 100543.94
    },
    {
      "epoch": 0.123,
      "grad_norm": 0.13253241777420044,
      "learning_rate": 0.00029046078004597175,
      "loss": 0.3239,
      "num_input_tokens_seen": 806092800,
      "step": 12300,
      "train_runtime": 8016.3597,
      "train_tokens_per_second": 100555.966
    },
    {
      "epoch": 0.124,
      "grad_norm": 0.2943899929523468,
      "learning_rate": 0.00029029303526946796,
      "loss": 0.3238,
      "num_input_tokens_seen": 812646400,
      "step": 12400,
      "train_runtime": 8079.6597,
      "train_tokens_per_second": 100579.286
    },
    {
      "epoch": 0.125,
      "grad_norm": 0.1583172082901001,
      "learning_rate": 0.0002901238777436565,
      "loss": 0.3217,
      "num_input_tokens_seen": 819200000,
      "step": 12500,
      "train_runtime": 8148.9297,
      "train_tokens_per_second": 100528.539
    },
    {
      "epoch": 0.126,
      "grad_norm": 0.1598382592201233,
      "learning_rate": 0.00028995330917195184,
      "loss": 0.3245,
      "num_input_tokens_seen": 825753600,
      "step": 12600,
      "train_runtime": 8213.0201,
      "train_tokens_per_second": 100542.016
    },
    {
      "epoch": 0.127,
      "grad_norm": 0.13507018983364105,
      "learning_rate": 0.00028978133127197765,
      "loss": 0.3247,
      "num_input_tokens_seen": 832307200,
      "step": 12700,
      "train_runtime": 8277.3925,
      "train_tokens_per_second": 100551.859
    },
    {
      "epoch": 0.128,
      "grad_norm": 0.1688830703496933,
      "learning_rate": 0.0002896079457755493,
      "loss": 0.3258,
      "num_input_tokens_seen": 838860800,
      "step": 12800,
      "train_runtime": 8342.3491,
      "train_tokens_per_second": 100554.507
    },
    {
      "epoch": 0.129,
      "grad_norm": 0.2753322422504425,
      "learning_rate": 0.000289433154428657,
      "loss": 0.3249,
      "num_input_tokens_seen": 845414400,
      "step": 12900,
      "train_runtime": 8406.9898,
      "train_tokens_per_second": 100560.892
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.20588786900043488,
      "learning_rate": 0.0002892569589914476,
      "loss": 0.3232,
      "num_input_tokens_seen": 851968000,
      "step": 13000,
      "train_runtime": 8475.9626,
      "train_tokens_per_second": 100515.781
    },
    {
      "epoch": 0.131,
      "grad_norm": 0.1462445855140686,
      "learning_rate": 0.0002890793612382072,
      "loss": 0.3239,
      "num_input_tokens_seen": 858521600,
      "step": 13100,
      "train_runtime": 8539.9861,
      "train_tokens_per_second": 100529.625
    },
    {
      "epoch": 0.132,
      "grad_norm": 0.11379440873861313,
      "learning_rate": 0.0002889003629573432,
      "loss": 0.3249,
      "num_input_tokens_seen": 865075200,
      "step": 13200,
      "train_runtime": 8604.867,
      "train_tokens_per_second": 100533.244
    },
    {
      "epoch": 0.133,
      "grad_norm": 0.12769202888011932,
      "learning_rate": 0.00028871996595136626,
      "loss": 0.327,
      "num_input_tokens_seen": 871628800,
      "step": 13300,
      "train_runtime": 8669.3605,
      "train_tokens_per_second": 100541.303
    },
    {
      "epoch": 0.134,
      "grad_norm": 0.14837151765823364,
      "learning_rate": 0.0002885381720368723,
      "loss": 0.321,
      "num_input_tokens_seen": 878182400,
      "step": 13400,
      "train_runtime": 8738.2624,
      "train_tokens_per_second": 100498.515
    },
    {
      "epoch": 0.135,
      "grad_norm": 0.1538904309272766,
      "learning_rate": 0.000288354983044524,
      "loss": 0.3207,
      "num_input_tokens_seen": 884736000,
      "step": 13500,
      "train_runtime": 8802.2586,
      "train_tokens_per_second": 100512.385
    },
    {
      "epoch": 0.136,
      "grad_norm": 0.12802962958812714,
      "learning_rate": 0.00028817040081903245,
      "loss": 0.3241,
      "num_input_tokens_seen": 891289600,
      "step": 13600,
      "train_runtime": 8866.1163,
      "train_tokens_per_second": 100527.624
    },
    {
      "epoch": 0.137,
      "grad_norm": 0.35466450452804565,
      "learning_rate": 0.00028798442721913867,
      "loss": 0.3214,
      "num_input_tokens_seen": 897843200,
      "step": 13700,
      "train_runtime": 8930.5828,
      "train_tokens_per_second": 100535.79
    },
    {
      "epoch": 0.138,
      "grad_norm": 0.13867586851119995,
      "learning_rate": 0.00028779706411759465,
      "loss": 0.3199,
      "num_input_tokens_seen": 904396800,
      "step": 13800,
      "train_runtime": 9001.3287,
      "train_tokens_per_second": 100473.7
    },
    {
      "epoch": 0.139,
      "grad_norm": 0.2114623785018921,
      "learning_rate": 0.00028760831340114484,
      "loss": 0.3234,
      "num_input_tokens_seen": 910950400,
      "step": 13900,
      "train_runtime": 9066.3163,
      "train_tokens_per_second": 100476.353
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.14202618598937988,
      "learning_rate": 0.00028741817697050683,
      "loss": 0.3232,
      "num_input_tokens_seen": 917504000,
      "step": 14000,
      "train_runtime": 9130.2003,
      "train_tokens_per_second": 100491.114
    },
    {
      "epoch": 0.141,
      "grad_norm": 0.1686236560344696,
      "learning_rate": 0.00028722665674035233,
      "loss": 0.3203,
      "num_input_tokens_seen": 924057600,
      "step": 14100,
      "train_runtime": 9195.1426,
      "train_tokens_per_second": 100494.102
    },
    {
      "epoch": 0.142,
      "grad_norm": 0.14483292400836945,
      "learning_rate": 0.0002870337546392879,
      "loss": 0.3321,
      "num_input_tokens_seen": 930611200,
      "step": 14200,
      "train_runtime": 9259.404,
      "train_tokens_per_second": 100504.438
    },
    {
      "epoch": 0.143,
      "grad_norm": 0.12517394125461578,
      "learning_rate": 0.00028683947260983576,
      "loss": 0.3233,
      "num_input_tokens_seen": 937164800,
      "step": 14300,
      "train_runtime": 9324.1454,
      "train_tokens_per_second": 100509.458
    },
    {
      "epoch": 0.144,
      "grad_norm": 0.24776680767536163,
      "learning_rate": 0.00028664381260841356,
      "loss": 0.3192,
      "num_input_tokens_seen": 943718400,
      "step": 14400,
      "train_runtime": 9393.645,
      "train_tokens_per_second": 100463.494
    },
    {
      "epoch": 0.145,
      "grad_norm": 0.4200928807258606,
      "learning_rate": 0.0002864467766053154,
      "loss": 0.321,
      "num_input_tokens_seen": 950272000,
      "step": 14500,
      "train_runtime": 9456.5857,
      "train_tokens_per_second": 100487.853
    },
    {
      "epoch": 0.146,
      "grad_norm": 0.14573471248149872,
      "learning_rate": 0.00028624836658469165,
      "loss": 0.3198,
      "num_input_tokens_seen": 956825600,
      "step": 14600,
      "train_runtime": 9525.9633,
      "train_tokens_per_second": 100443.973
    },
    {
      "epoch": 0.147,
      "grad_norm": 0.1546989232301712,
      "learning_rate": 0.00028604858454452906,
      "loss": 0.3267,
      "num_input_tokens_seen": 963379200,
      "step": 14700,
      "train_runtime": 9585.7512,
      "train_tokens_per_second": 100501.169
    },
    {
      "epoch": 0.148,
      "grad_norm": 0.172988623380661,
      "learning_rate": 0.00028584743249663057,
      "loss": 0.3222,
      "num_input_tokens_seen": 969932800,
      "step": 14800,
      "train_runtime": 9650.7111,
      "train_tokens_per_second": 100503.765
    },
    {
      "epoch": 0.149,
      "grad_norm": 0.19345735013484955,
      "learning_rate": 0.000285644912466595,
      "loss": 0.3194,
      "num_input_tokens_seen": 976486400,
      "step": 14900,
      "train_runtime": 9721.1196,
      "train_tokens_per_second": 100449.994
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.13317954540252686,
      "learning_rate": 0.00028544102649379684,
      "loss": 0.3236,
      "num_input_tokens_seen": 983040000,
      "step": 15000,
      "train_runtime": 9784.7921,
      "train_tokens_per_second": 100466.11
    },
    {
      "epoch": 0.151,
      "grad_norm": 0.17458604276180267,
      "learning_rate": 0.00028523577663136556,
      "loss": 0.3208,
      "num_input_tokens_seen": 989593600,
      "step": 15100,
      "train_runtime": 9853.1273,
      "train_tokens_per_second": 100434.468
    },
    {
      "epoch": 0.152,
      "grad_norm": 0.1358109712600708,
      "learning_rate": 0.000285029164946165,
      "loss": 0.3237,
      "num_input_tokens_seen": 996147200,
      "step": 15200,
      "train_runtime": 9917.7044,
      "train_tokens_per_second": 100441.307
    },
    {
      "epoch": 0.153,
      "grad_norm": 0.16100633144378662,
      "learning_rate": 0.0002848211935187725,
      "loss": 0.3267,
      "num_input_tokens_seen": 1002700800,
      "step": 15300,
      "train_runtime": 9982.8922,
      "train_tokens_per_second": 100441.914
    },
    {
      "epoch": 0.154,
      "grad_norm": 0.20419622957706451,
      "learning_rate": 0.0002846118644434581,
      "loss": 0.3193,
      "num_input_tokens_seen": 1009254400,
      "step": 15400,
      "train_runtime": 10046.3454,
      "train_tokens_per_second": 100459.855
    },
    {
      "epoch": 0.155,
      "grad_norm": 0.17805695533752441,
      "learning_rate": 0.00028440117982816326,
      "loss": 0.3159,
      "num_input_tokens_seen": 1015808000,
      "step": 15500,
      "train_runtime": 10110.0124,
      "train_tokens_per_second": 100475.446
    },
    {
      "epoch": 0.156,
      "grad_norm": 0.17533563077449799,
      "learning_rate": 0.0002841891417944796,
      "loss": 0.3216,
      "num_input_tokens_seen": 1022361600,
      "step": 15600,
      "train_runtime": 10178.7469,
      "train_tokens_per_second": 100440.812
    },
    {
      "epoch": 0.157,
      "grad_norm": 0.13143610954284668,
      "learning_rate": 0.0002839757524776279,
      "loss": 0.3234,
      "num_input_tokens_seen": 1028915200,
      "step": 15700,
      "train_runtime": 10243.1395,
      "train_tokens_per_second": 100449.203
    },
    {
      "epoch": 0.158,
      "grad_norm": 0.13563373684883118,
      "learning_rate": 0.0002837610140264361,
      "loss": 0.3194,
      "num_input_tokens_seen": 1035468800,
      "step": 15800,
      "train_runtime": 10307.5423,
      "train_tokens_per_second": 100457.39
    },
    {
      "epoch": 0.159,
      "grad_norm": 0.14616088569164276,
      "learning_rate": 0.0002835449286033182,
      "loss": 0.3178,
      "num_input_tokens_seen": 1042022400,
      "step": 15900,
      "train_runtime": 10378.0909,
      "train_tokens_per_second": 100405.982
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.1539888232946396,
      "learning_rate": 0.0002833274983842518,
      "loss": 0.3156,
      "num_input_tokens_seen": 1048576000,
      "step": 16000,
      "train_runtime": 10441.484,
      "train_tokens_per_second": 100424.039
    },
    {
      "epoch": 0.161,
      "grad_norm": 0.15786372125148773,
      "learning_rate": 0.0002831087255587569,
      "loss": 0.318,
      "num_input_tokens_seen": 1055129600,
      "step": 16100,
      "train_runtime": 10505.72,
      "train_tokens_per_second": 100433.821
    },
    {
      "epoch": 0.162,
      "grad_norm": 0.14359760284423828,
      "learning_rate": 0.0002828886123298734,
      "loss": 0.3179,
      "num_input_tokens_seen": 1061683200,
      "step": 16200,
      "train_runtime": 10570.7713,
      "train_tokens_per_second": 100435.736
    },
    {
      "epoch": 0.163,
      "grad_norm": 0.1415397673845291,
      "learning_rate": 0.00028266716091413906,
      "loss": 0.32,
      "num_input_tokens_seen": 1068236800,
      "step": 16300,
      "train_runtime": 10635.2645,
      "train_tokens_per_second": 100442.899
    },
    {
      "epoch": 0.164,
      "grad_norm": 0.1199110895395279,
      "learning_rate": 0.0002824443735415673,
      "loss": 0.3188,
      "num_input_tokens_seen": 1074790400,
      "step": 16400,
      "train_runtime": 10704.7074,
      "train_tokens_per_second": 100403.529
    },
    {
      "epoch": 0.165,
      "grad_norm": 0.18369431793689728,
      "learning_rate": 0.0002822202524556243,
      "loss": 0.3208,
      "num_input_tokens_seen": 1081344000,
      "step": 16500,
      "train_runtime": 10770.1863,
      "train_tokens_per_second": 100401.606
    },
    {
      "epoch": 0.166,
      "grad_norm": 0.2615172266960144,
      "learning_rate": 0.00028199479991320695,
      "loss": 0.3224,
      "num_input_tokens_seen": 1087897600,
      "step": 16600,
      "train_runtime": 10834.6749,
      "train_tokens_per_second": 100408.883
    },
    {
      "epoch": 0.167,
      "grad_norm": 0.1250002384185791,
      "learning_rate": 0.00028176801818461994,
      "loss": 0.3171,
      "num_input_tokens_seen": 1094451200,
      "step": 16700,
      "train_runtime": 10899.3075,
      "train_tokens_per_second": 100414.747
    },
    {
      "epoch": 0.168,
      "grad_norm": 0.14198775589466095,
      "learning_rate": 0.00028153990955355273,
      "loss": 0.3194,
      "num_input_tokens_seen": 1101004800,
      "step": 16800,
      "train_runtime": 10964.3423,
      "train_tokens_per_second": 100416.858
    },
    {
      "epoch": 0.169,
      "grad_norm": 0.14076939225196838,
      "learning_rate": 0.00028131047631705665,
      "loss": 0.3189,
      "num_input_tokens_seen": 1107558400,
      "step": 16900,
      "train_runtime": 11033.6033,
      "train_tokens_per_second": 100380.48
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.13334921002388,
      "learning_rate": 0.00028107972078552187,
      "loss": 0.3198,
      "num_input_tokens_seen": 1114112000,
      "step": 17000,
      "train_runtime": 11098.612,
      "train_tokens_per_second": 100383.003
    },
    {
      "epoch": 0.171,
      "grad_norm": 0.13615840673446655,
      "learning_rate": 0.0002808476452826541,
      "loss": 0.3168,
      "num_input_tokens_seen": 1120665600,
      "step": 17100,
      "train_runtime": 11161.3832,
      "train_tokens_per_second": 100405.62
    },
    {
      "epoch": 0.172,
      "grad_norm": 0.14747090637683868,
      "learning_rate": 0.00028061425214545094,
      "loss": 0.3163,
      "num_input_tokens_seen": 1127219200,
      "step": 17200,
      "train_runtime": 11231.5954,
      "train_tokens_per_second": 100361.45
    },
    {
      "epoch": 0.173,
      "grad_norm": 0.15957149863243103,
      "learning_rate": 0.00028037954372417883,
      "loss": 0.317,
      "num_input_tokens_seen": 1133772800,
      "step": 17300,
      "train_runtime": 11295.5019,
      "train_tokens_per_second": 100373.831
    },
    {
      "epoch": 0.174,
      "grad_norm": 0.20420241355895996,
      "learning_rate": 0.0002801435223823488,
      "loss": 0.3207,
      "num_input_tokens_seen": 1140326400,
      "step": 17400,
      "train_runtime": 11360.8649,
      "train_tokens_per_second": 100373.203
    },
    {
      "epoch": 0.175,
      "grad_norm": 0.20070046186447144,
      "learning_rate": 0.00027990619049669336,
      "loss": 0.3206,
      "num_input_tokens_seen": 1146880000,
      "step": 17500,
      "train_runtime": 11424.854,
      "train_tokens_per_second": 100384.652
    },
    {
      "epoch": 0.176,
      "grad_norm": 0.13903649151325226,
      "learning_rate": 0.00027966755045714177,
      "loss": 0.3227,
      "num_input_tokens_seen": 1153433600,
      "step": 17600,
      "train_runtime": 11488.6874,
      "train_tokens_per_second": 100397.336
    },
    {
      "epoch": 0.177,
      "grad_norm": 0.15853877365589142,
      "learning_rate": 0.00027942760466679673,
      "loss": 0.3168,
      "num_input_tokens_seen": 1159987200,
      "step": 17700,
      "train_runtime": 11559.2862,
      "train_tokens_per_second": 100351.11
    },
    {
      "epoch": 0.178,
      "grad_norm": 0.14262589812278748,
      "learning_rate": 0.00027918635554190956,
      "loss": 0.3235,
      "num_input_tokens_seen": 1166540800,
      "step": 17800,
      "train_runtime": 11622.4751,
      "train_tokens_per_second": 100369.395
    },
    {
      "epoch": 0.179,
      "grad_norm": 0.14338357746601105,
      "learning_rate": 0.00027894380551185636,
      "loss": 0.3204,
      "num_input_tokens_seen": 1173094400,
      "step": 17900,
      "train_runtime": 11687.9668,
      "train_tokens_per_second": 100367.705
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.12374505400657654,
      "learning_rate": 0.00027869995701911314,
      "loss": 0.3156,
      "num_input_tokens_seen": 1179648000,
      "step": 18000,
      "train_runtime": 11751.6619,
      "train_tokens_per_second": 100381.377
    },
    {
      "epoch": 0.181,
      "grad_norm": 0.11708634346723557,
      "learning_rate": 0.0002784548125192316,
      "loss": 0.3145,
      "num_input_tokens_seen": 1186201600,
      "step": 18100,
      "train_runtime": 11816.0633,
      "train_tokens_per_second": 100388.9
    },
    {
      "epoch": 0.182,
      "grad_norm": 0.1318449079990387,
      "learning_rate": 0.0002782083744808141,
      "loss": 0.3159,
      "num_input_tokens_seen": 1192755200,
      "step": 18200,
      "train_runtime": 11887.7736,
      "train_tokens_per_second": 100334.616
    },
    {
      "epoch": 0.183,
      "grad_norm": 0.3383175730705261,
      "learning_rate": 0.000277960645385489,
      "loss": 0.3191,
      "num_input_tokens_seen": 1199308800,
      "step": 18300,
      "train_runtime": 11953.3207,
      "train_tokens_per_second": 100332.688
    },
    {
      "epoch": 0.184,
      "grad_norm": 0.13779285550117493,
      "learning_rate": 0.00027771162772788544,
      "loss": 0.3168,
      "num_input_tokens_seen": 1205862400,
      "step": 18400,
      "train_runtime": 12016.7432,
      "train_tokens_per_second": 100348.521
    },
    {
      "epoch": 0.185,
      "grad_norm": 0.15161630511283875,
      "learning_rate": 0.00027746132401560857,
      "loss": 0.3146,
      "num_input_tokens_seen": 1212416000,
      "step": 18500,
      "train_runtime": 12081.3443,
      "train_tokens_per_second": 100354.395
    },
    {
      "epoch": 0.186,
      "grad_norm": 0.1523953378200531,
      "learning_rate": 0.0002772097367692139,
      "loss": 0.3172,
      "num_input_tokens_seen": 1218969600,
      "step": 18600,
      "train_runtime": 12145.9663,
      "train_tokens_per_second": 100360.035
    },
    {
      "epoch": 0.187,
      "grad_norm": 0.12802754342556,
      "learning_rate": 0.00027695686852218226,
      "loss": 0.3198,
      "num_input_tokens_seen": 1225523200,
      "step": 18700,
      "train_runtime": 12215.5887,
      "train_tokens_per_second": 100324.53
    },
    {
      "epoch": 0.188,
      "grad_norm": 0.13653679192066193,
      "learning_rate": 0.00027670272182089416,
      "loss": 0.319,
      "num_input_tokens_seen": 1232076800,
      "step": 18800,
      "train_runtime": 12280.146,
      "train_tokens_per_second": 100330.794
    },
    {
      "epoch": 0.189,
      "grad_norm": 0.15152159333229065,
      "learning_rate": 0.0002764472992246039,
      "loss": 0.3165,
      "num_input_tokens_seen": 1238630400,
      "step": 18900,
      "train_runtime": 12344.6292,
      "train_tokens_per_second": 100337.594
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.13211041688919067,
      "learning_rate": 0.0002761906033054143,
      "loss": 0.3161,
      "num_input_tokens_seen": 1245184000,
      "step": 19000,
      "train_runtime": 12407.4556,
      "train_tokens_per_second": 100357.724
    },
    {
      "epoch": 0.191,
      "grad_norm": 0.19933822751045227,
      "learning_rate": 0.00027593263664825045,
      "loss": 0.3173,
      "num_input_tokens_seen": 1251737600,
      "step": 19100,
      "train_runtime": 12472.5241,
      "train_tokens_per_second": 100359.606
    },
    {
      "epoch": 0.192,
      "grad_norm": 0.1472938358783722,
      "learning_rate": 0.00027567340185083363,
      "loss": 0.3157,
      "num_input_tokens_seen": 1258291200,
      "step": 19200,
      "train_runtime": 12542.0532,
      "train_tokens_per_second": 100325.774
    },
    {
      "epoch": 0.193,
      "grad_norm": 0.1466071903705597,
      "learning_rate": 0.00027541290152365537,
      "loss": 0.3188,
      "num_input_tokens_seen": 1264844800,
      "step": 19300,
      "train_runtime": 12606.5735,
      "train_tokens_per_second": 100332.164
    },
    {
      "epoch": 0.194,
      "grad_norm": 0.1384386122226715,
      "learning_rate": 0.00027515113828995117,
      "loss": 0.318,
      "num_input_tokens_seen": 1271398400,
      "step": 19400,
      "train_runtime": 12672.5058,
      "train_tokens_per_second": 100327.309
    },
    {
      "epoch": 0.195,
      "grad_norm": 0.16287657618522644,
      "learning_rate": 0.00027488811478567374,
      "loss": 0.3153,
      "num_input_tokens_seen": 1277952000,
      "step": 19500,
      "train_runtime": 12735.4985,
      "train_tokens_per_second": 100345.66
    },
    {
      "epoch": 0.196,
      "grad_norm": 0.14955779910087585,
      "learning_rate": 0.0002746238336594671,
      "loss": 0.3144,
      "num_input_tokens_seen": 1284505600,
      "step": 19600,
      "train_runtime": 12804.8911,
      "train_tokens_per_second": 100313.669
    },
    {
      "epoch": 0.197,
      "grad_norm": 0.15176887810230255,
      "learning_rate": 0.00027435829757263894,
      "loss": 0.3172,
      "num_input_tokens_seen": 1291059200,
      "step": 19700,
      "train_runtime": 12869.0984,
      "train_tokens_per_second": 100322.428
    },
    {
      "epoch": 0.198,
      "grad_norm": 0.12215608358383179,
      "learning_rate": 0.0002740915091991349,
      "loss": 0.3182,
      "num_input_tokens_seen": 1297612800,
      "step": 19800,
      "train_runtime": 12932.8746,
      "train_tokens_per_second": 100334.446
    },
    {
      "epoch": 0.199,
      "grad_norm": 0.248954638838768,
      "learning_rate": 0.0002738234712255109,
      "loss": 0.3171,
      "num_input_tokens_seen": 1304166400,
      "step": 19900,
      "train_runtime": 13003.7739,
      "train_tokens_per_second": 100291.378
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.18855011463165283,
      "learning_rate": 0.00027355418635090635,
      "loss": 0.3181,
      "num_input_tokens_seen": 1310720000,
      "step": 20000,
      "train_runtime": 13068.3505,
      "train_tokens_per_second": 100297.279
    },
    {
      "epoch": 0.201,
      "grad_norm": 0.17624643445014954,
      "learning_rate": 0.000273283657287017,
      "loss": 0.3147,
      "num_input_tokens_seen": 1317273600,
      "step": 20100,
      "train_runtime": 13133.7291,
      "train_tokens_per_second": 100296.998
    },
    {
      "epoch": 0.202,
      "grad_norm": 0.12586164474487305,
      "learning_rate": 0.00027301188675806745,
      "loss": 0.3203,
      "num_input_tokens_seen": 1323827200,
      "step": 20200,
      "train_runtime": 13197.5369,
      "train_tokens_per_second": 100308.657
    },
    {
      "epoch": 0.203,
      "grad_norm": 0.13073797523975372,
      "learning_rate": 0.0002727388775007839,
      "loss": 0.3149,
      "num_input_tokens_seen": 1330380800,
      "step": 20300,
      "train_runtime": 13261.8266,
      "train_tokens_per_second": 100316.558
    },
    {
      "epoch": 0.204,
      "grad_norm": 0.12983232736587524,
      "learning_rate": 0.0002724646322643666,
      "loss": 0.3157,
      "num_input_tokens_seen": 1336934400,
      "step": 20400,
      "train_runtime": 13325.295,
      "train_tokens_per_second": 100330.567
    },
    {
      "epoch": 0.205,
      "grad_norm": 0.2400187999010086,
      "learning_rate": 0.000272189153810462,
      "loss": 0.3178,
      "num_input_tokens_seen": 1343488000,
      "step": 20500,
      "train_runtime": 13395.2424,
      "train_tokens_per_second": 100295.908
    },
    {
      "epoch": 0.206,
      "grad_norm": 0.11757266521453857,
      "learning_rate": 0.0002719124449131351,
      "loss": 0.3164,
      "num_input_tokens_seen": 1350041600,
      "step": 20600,
      "train_runtime": 13459.4754,
      "train_tokens_per_second": 100304.177
    },
    {
      "epoch": 0.207,
      "grad_norm": 0.1606636494398117,
      "learning_rate": 0.00027163450835884144,
      "loss": 0.3146,
      "num_input_tokens_seen": 1356595200,
      "step": 20700,
      "train_runtime": 13524.1715,
      "train_tokens_per_second": 100308.932
    },
    {
      "epoch": 0.208,
      "grad_norm": 0.1295078545808792,
      "learning_rate": 0.00027135534694639894,
      "loss": 0.3175,
      "num_input_tokens_seen": 1363148800,
      "step": 20800,
      "train_runtime": 13588.4538,
      "train_tokens_per_second": 100316.697
    },
    {
      "epoch": 0.209,
      "grad_norm": 0.18409083783626556,
      "learning_rate": 0.00027107496348696003,
      "loss": 0.3189,
      "num_input_tokens_seen": 1369702400,
      "step": 20900,
      "train_runtime": 13653.2417,
      "train_tokens_per_second": 100320.673
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.12083840370178223,
      "learning_rate": 0.00027079336080398296,
      "loss": 0.3139,
      "num_input_tokens_seen": 1376256000,
      "step": 21000,
      "train_runtime": 13723.0075,
      "train_tokens_per_second": 100288.22
    },
    {
      "epoch": 0.211,
      "grad_norm": 0.16270384192466736,
      "learning_rate": 0.00027051054173320366,
      "loss": 0.3147,
      "num_input_tokens_seen": 1382809600,
      "step": 21100,
      "train_runtime": 13787.7693,
      "train_tokens_per_second": 100292.482
    },
    {
      "epoch": 0.212,
      "grad_norm": 0.12299864739179611,
      "learning_rate": 0.000270226509122607,
      "loss": 0.3137,
      "num_input_tokens_seen": 1389363200,
      "step": 21200,
      "train_runtime": 13851.6298,
      "train_tokens_per_second": 100303.229
    },
    {
      "epoch": 0.213,
      "grad_norm": 0.12248677760362625,
      "learning_rate": 0.0002699412658323983,
      "loss": 0.3177,
      "num_input_tokens_seen": 1395916800,
      "step": 21300,
      "train_runtime": 13915.8434,
      "train_tokens_per_second": 100311.333
    },
    {
      "epoch": 0.214,
      "grad_norm": 0.13090935349464417,
      "learning_rate": 0.00026965481473497423,
      "loss": 0.3146,
      "num_input_tokens_seen": 1402470400,
      "step": 21400,
      "train_runtime": 13985.645,
      "train_tokens_per_second": 100279.28
    },
    {
      "epoch": 0.215,
      "grad_norm": 0.1279245913028717,
      "learning_rate": 0.0002693671587148942,
      "loss": 0.3128,
      "num_input_tokens_seen": 1409024000,
      "step": 21500,
      "train_runtime": 14050.4506,
      "train_tokens_per_second": 100283.19
    },
    {
      "epoch": 0.216,
      "grad_norm": 0.15504342317581177,
      "learning_rate": 0.0002690783006688511,
      "loss": 0.3145,
      "num_input_tokens_seen": 1415577600,
      "step": 21600,
      "train_runtime": 14115.855,
      "train_tokens_per_second": 100282.81
    },
    {
      "epoch": 0.217,
      "grad_norm": 0.1325046420097351,
      "learning_rate": 0.0002687882435056423,
      "loss": 0.3138,
      "num_input_tokens_seen": 1422131200,
      "step": 21700,
      "train_runtime": 14179.61,
      "train_tokens_per_second": 100294.098
    },
    {
      "epoch": 0.218,
      "grad_norm": 0.17374184727668762,
      "learning_rate": 0.0002684969901461402,
      "loss": 0.3179,
      "num_input_tokens_seen": 1428684800,
      "step": 21800,
      "train_runtime": 14245.0199,
      "train_tokens_per_second": 100293.633
    },
    {
      "epoch": 0.219,
      "grad_norm": 0.16908228397369385,
      "learning_rate": 0.000268204543523263,
      "loss": 0.3182,
      "num_input_tokens_seen": 1435238400,
      "step": 21900,
      "train_runtime": 14310.1147,
      "train_tokens_per_second": 100295.381
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.15052039921283722,
      "learning_rate": 0.0002679109065819447,
      "loss": 0.3148,
      "num_input_tokens_seen": 1441792000,
      "step": 22000,
      "train_runtime": 14374.221,
      "train_tokens_per_second": 100304.01
    },
    {
      "epoch": 0.221,
      "grad_norm": 0.1661474108695984,
      "learning_rate": 0.0002676160822791062,
      "loss": 0.3142,
      "num_input_tokens_seen": 1448345600,
      "step": 22100,
      "train_runtime": 14445.9108,
      "train_tokens_per_second": 100259.902
    },
    {
      "epoch": 0.222,
      "grad_norm": 0.16423378884792328,
      "learning_rate": 0.00026732007358362496,
      "loss": 0.323,
      "num_input_tokens_seen": 1454899200,
      "step": 22200,
      "train_runtime": 14510.5733,
      "train_tokens_per_second": 100264.763
    },
    {
      "epoch": 0.223,
      "grad_norm": 0.14868460595607758,
      "learning_rate": 0.0002670228834763052,
      "loss": 0.3155,
      "num_input_tokens_seen": 1461452800,
      "step": 22300,
      "train_runtime": 14575.7382,
      "train_tokens_per_second": 100266.126
    },
    {
      "epoch": 0.224,
      "grad_norm": 0.1287386268377304,
      "learning_rate": 0.00026672451494984804,
      "loss": 0.3152,
      "num_input_tokens_seen": 1468006400,
      "step": 22400,
      "train_runtime": 14639.7379,
      "train_tokens_per_second": 100275.456
    },
    {
      "epoch": 0.225,
      "grad_norm": 0.14276720583438873,
      "learning_rate": 0.0002664249710088213,
      "loss": 0.3131,
      "num_input_tokens_seen": 1474560000,
      "step": 22500,
      "train_runtime": 14703.588,
      "train_tokens_per_second": 100285.726
    },
    {
      "epoch": 0.226,
      "grad_norm": 0.1419740915298462,
      "learning_rate": 0.00026612425466962893,
      "loss": 0.3112,
      "num_input_tokens_seen": 1481113600,
      "step": 22600,
      "train_runtime": 14773.1939,
      "train_tokens_per_second": 100256.83
    },
    {
      "epoch": 0.227,
      "grad_norm": 0.12067803740501404,
      "learning_rate": 0.00026582236896048134,
      "loss": 0.3122,
      "num_input_tokens_seen": 1487667200,
      "step": 22700,
      "train_runtime": 14837.1829,
      "train_tokens_per_second": 100266.15
    },
    {
      "epoch": 0.228,
      "grad_norm": 0.1338939219713211,
      "learning_rate": 0.00026551931692136413,
      "loss": 0.3128,
      "num_input_tokens_seen": 1494220800,
      "step": 22800,
      "train_runtime": 14900.9562,
      "train_tokens_per_second": 100276.84
    },
    {
      "epoch": 0.229,
      "grad_norm": 0.16754469275474548,
      "learning_rate": 0.00026521510160400804,
      "loss": 0.3133,
      "num_input_tokens_seen": 1500774400,
      "step": 22900,
      "train_runtime": 14965.1238,
      "train_tokens_per_second": 100284.797
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.12648451328277588,
      "learning_rate": 0.00026490972607185793,
      "loss": 0.311,
      "num_input_tokens_seen": 1507328000,
      "step": 23000,
      "train_runtime": 15034.861,
      "train_tokens_per_second": 100255.533
    },
    {
      "epoch": 0.231,
      "grad_norm": 0.12040221691131592,
      "learning_rate": 0.0002646031934000421,
      "loss": 0.3166,
      "num_input_tokens_seen": 1513881600,
      "step": 23100,
      "train_runtime": 15099.2676,
      "train_tokens_per_second": 100261.922
    },
    {
      "epoch": 0.232,
      "grad_norm": 0.12486282736063004,
      "learning_rate": 0.00026429550667534095,
      "loss": 0.3151,
      "num_input_tokens_seen": 1520435200,
      "step": 23200,
      "train_runtime": 15164.1184,
      "train_tokens_per_second": 100265.321
    },
    {
      "epoch": 0.233,
      "grad_norm": 0.18211719393730164,
      "learning_rate": 0.0002639866689961565,
      "loss": 0.3117,
      "num_input_tokens_seen": 1526988800,
      "step": 23300,
      "train_runtime": 15229.7058,
      "train_tokens_per_second": 100263.841
    },
    {
      "epoch": 0.234,
      "grad_norm": 0.13128802180290222,
      "learning_rate": 0.00026367668347248083,
      "loss": 0.3125,
      "num_input_tokens_seen": 1533542400,
      "step": 23400,
      "train_runtime": 15293.6404,
      "train_tokens_per_second": 100273.209
    },
    {
      "epoch": 0.235,
      "grad_norm": 0.11493753641843796,
      "learning_rate": 0.0002633655532258646,
      "loss": 0.317,
      "num_input_tokens_seen": 1540096000,
      "step": 23500,
      "train_runtime": 15365.113,
      "train_tokens_per_second": 100233.301
    },
    {
      "epoch": 0.236,
      "grad_norm": 0.15309779345989227,
      "learning_rate": 0.000263053281389386,
      "loss": 0.3136,
      "num_input_tokens_seen": 1546649600,
      "step": 23600,
      "train_runtime": 15428.6523,
      "train_tokens_per_second": 100245.282
    },
    {
      "epoch": 0.237,
      "grad_norm": 0.15829730033874512,
      "learning_rate": 0.0002627398711076189,
      "loss": 0.3098,
      "num_input_tokens_seen": 1553203200,
      "step": 23700,
      "train_runtime": 15493.1944,
      "train_tokens_per_second": 100250.675
    },
    {
      "epoch": 0.238,
      "grad_norm": 0.13252806663513184,
      "learning_rate": 0.0002624253255366014,
      "loss": 0.3096,
      "num_input_tokens_seen": 1559756800,
      "step": 23800,
      "train_runtime": 15556.5037,
      "train_tokens_per_second": 100263.969
    },
    {
      "epoch": 0.239,
      "grad_norm": 0.18889528512954712,
      "learning_rate": 0.0002621096478438039,
      "loss": 0.3146,
      "num_input_tokens_seen": 1566310400,
      "step": 23900,
      "train_runtime": 15621.7412,
      "train_tokens_per_second": 100264.777
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.16285447776317596,
      "learning_rate": 0.00026179284120809727,
      "loss": 0.3168,
      "num_input_tokens_seen": 1572864000,
      "step": 24000,
      "train_runtime": 15687.4424,
      "train_tokens_per_second": 100262.615
    },
    {
      "epoch": 0.241,
      "grad_norm": 0.14852070808410645,
      "learning_rate": 0.0002614749088197208,
      "loss": 0.3115,
      "num_input_tokens_seen": 1579417600,
      "step": 24100,
      "train_runtime": 15752.1472,
      "train_tokens_per_second": 100266.813
    },
    {
      "epoch": 0.242,
      "grad_norm": 0.22735795378684998,
      "learning_rate": 0.00026115585388025015,
      "loss": 0.3099,
      "num_input_tokens_seen": 1585971200,
      "step": 24200,
      "train_runtime": 15823.0117,
      "train_tokens_per_second": 100231.943
    },
    {
      "epoch": 0.243,
      "grad_norm": 0.16086964309215546,
      "learning_rate": 0.00026083567960256493,
      "loss": 0.3107,
      "num_input_tokens_seen": 1592524800,
      "step": 24300,
      "train_runtime": 15889.3517,
      "train_tokens_per_second": 100225.914
    },
    {
      "epoch": 0.244,
      "grad_norm": 0.15085358917713165,
      "learning_rate": 0.00026051438921081667,
      "loss": 0.3112,
      "num_input_tokens_seen": 1599078400,
      "step": 24400,
      "train_runtime": 15954.2137,
      "train_tokens_per_second": 100229.22
    },
    {
      "epoch": 0.245,
      "grad_norm": 0.14889656007289886,
      "learning_rate": 0.00026019198594039595,
      "loss": 0.3147,
      "num_input_tokens_seen": 1605632000,
      "step": 24500,
      "train_runtime": 16020.1883,
      "train_tokens_per_second": 100225.539
    },
    {
      "epoch": 0.246,
      "grad_norm": 0.15055876970291138,
      "learning_rate": 0.00025986847303790026,
      "loss": 0.3125,
      "num_input_tokens_seen": 1612185600,
      "step": 24600,
      "train_runtime": 16084.1346,
      "train_tokens_per_second": 100234.525
    },
    {
      "epoch": 0.247,
      "grad_norm": 0.14507324993610382,
      "learning_rate": 0.00025954385376110076,
      "loss": 0.3115,
      "num_input_tokens_seen": 1618739200,
      "step": 24700,
      "train_runtime": 16148.9618,
      "train_tokens_per_second": 100237.973
    },
    {
      "epoch": 0.248,
      "grad_norm": 0.1229107677936554,
      "learning_rate": 0.00025921813137891005,
      "loss": 0.3147,
      "num_input_tokens_seen": 1625292800,
      "step": 24800,
      "train_runtime": 16214.7466,
      "train_tokens_per_second": 100235.473
    },
    {
      "epoch": 0.249,
      "grad_norm": 0.1423114389181137,
      "learning_rate": 0.000258891309171349,
      "loss": 0.3127,
      "num_input_tokens_seen": 1631846400,
      "step": 24900,
      "train_runtime": 16278.9968,
      "train_tokens_per_second": 100242.442
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.15807275474071503,
      "learning_rate": 0.00025856339042951344,
      "loss": 0.3088,
      "num_input_tokens_seen": 1638400000,
      "step": 25000,
      "train_runtime": 16343.5944,
      "train_tokens_per_second": 100247.226
    },
    {
      "epoch": 0.251,
      "grad_norm": 0.15635885298252106,
      "learning_rate": 0.0002582343784555415,
      "loss": 0.3105,
      "num_input_tokens_seen": 1644953600,
      "step": 25100,
      "train_runtime": 16414.1861,
      "train_tokens_per_second": 100215.362
    },
    {
      "epoch": 0.252,
      "grad_norm": 0.13579483330249786,
      "learning_rate": 0.00025790427656258017,
      "loss": 0.3159,
      "num_input_tokens_seen": 1651507200,
      "step": 25200,
      "train_runtime": 16478.0373,
      "train_tokens_per_second": 100224.752
    },
    {
      "epoch": 0.253,
      "grad_norm": 0.14977572858333588,
      "learning_rate": 0.00025757308807475185,
      "loss": 0.3115,
      "num_input_tokens_seen": 1658060800,
      "step": 25300,
      "train_runtime": 16542.7006,
      "train_tokens_per_second": 100229.149
    },
    {
      "epoch": 0.254,
      "grad_norm": 0.1324361115694046,
      "learning_rate": 0.00025724081632712086,
      "loss": 0.3108,
      "num_input_tokens_seen": 1664614400,
      "step": 25400,
      "train_runtime": 16607.2591,
      "train_tokens_per_second": 100234.144
    },
    {
      "epoch": 0.255,
      "grad_norm": 0.12053392827510834,
      "learning_rate": 0.0002569074646656601,
      "loss": 0.3081,
      "num_input_tokens_seen": 1671168000,
      "step": 25500,
      "train_runtime": 16676.4765,
      "train_tokens_per_second": 100211.096
    },
    {
      "epoch": 0.256,
      "grad_norm": 0.16214688122272491,
      "learning_rate": 0.00025657303644721695,
      "loss": 0.3154,
      "num_input_tokens_seen": 1677721600,
      "step": 25600,
      "train_runtime": 16741.4269,
      "train_tokens_per_second": 100213.776
    },
    {
      "epoch": 0.257,
      "grad_norm": 0.13730435073375702,
      "learning_rate": 0.00025623753503948004,
      "loss": 0.3159,
      "num_input_tokens_seen": 1684275200,
      "step": 25700,
      "train_runtime": 16805.4849,
      "train_tokens_per_second": 100221.755
    },
    {
      "epoch": 0.258,
      "grad_norm": 0.16218283772468567,
      "learning_rate": 0.00025590096382094475,
      "loss": 0.3111,
      "num_input_tokens_seen": 1690828800,
      "step": 25800,
      "train_runtime": 16869.8548,
      "train_tokens_per_second": 100227.821
    },
    {
      "epoch": 0.259,
      "grad_norm": 0.15016646683216095,
      "learning_rate": 0.00025556332618087945,
      "loss": 0.3106,
      "num_input_tokens_seen": 1697382400,
      "step": 25900,
      "train_runtime": 16938.0105,
      "train_tokens_per_second": 100211.439
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.1398506760597229,
      "learning_rate": 0.00025522462551929155,
      "loss": 0.313,
      "num_input_tokens_seen": 1703936000,
      "step": 26000,
      "train_runtime": 17003.6995,
      "train_tokens_per_second": 100209.722
    },
    {
      "epoch": 0.261,
      "grad_norm": 0.12380320578813553,
      "learning_rate": 0.00025488486524689283,
      "loss": 0.3133,
      "num_input_tokens_seen": 1710489600,
      "step": 26100,
      "train_runtime": 17069.3522,
      "train_tokens_per_second": 100208.232
    },
    {
      "epoch": 0.262,
      "grad_norm": 0.14536257088184357,
      "learning_rate": 0.00025454404878506555,
      "loss": 0.3115,
      "num_input_tokens_seen": 1717043200,
      "step": 26200,
      "train_runtime": 17132.7395,
      "train_tokens_per_second": 100220.003
    },
    {
      "epoch": 0.263,
      "grad_norm": 0.14442390203475952,
      "learning_rate": 0.0002542021795658276,
      "loss": 0.311,
      "num_input_tokens_seen": 1723596800,
      "step": 26300,
      "train_runtime": 17196.4745,
      "train_tokens_per_second": 100229.66
    },
    {
      "epoch": 0.264,
      "grad_norm": 0.12595972418785095,
      "learning_rate": 0.0002538592610317984,
      "loss": 0.3118,
      "num_input_tokens_seen": 1730150400,
      "step": 26400,
      "train_runtime": 17266.9358,
      "train_tokens_per_second": 100200.199
    },
    {
      "epoch": 0.265,
      "grad_norm": 0.1587669402360916,
      "learning_rate": 0.00025351529663616355,
      "loss": 0.3132,
      "num_input_tokens_seen": 1736704000,
      "step": 26500,
      "train_runtime": 17331.5833,
      "train_tokens_per_second": 100204.578
    },
    {
      "epoch": 0.266,
      "grad_norm": 0.1406719982624054,
      "learning_rate": 0.00025317028984264087,
      "loss": 0.3099,
      "num_input_tokens_seen": 1743257600,
      "step": 26600,
      "train_runtime": 17395.5945,
      "train_tokens_per_second": 100212.591
    },
    {
      "epoch": 0.267,
      "grad_norm": 0.1677832007408142,
      "learning_rate": 0.0002528242441254448,
      "loss": 0.309,
      "num_input_tokens_seen": 1749811200,
      "step": 26700,
      "train_runtime": 17459.1185,
      "train_tokens_per_second": 100223.342
    },
    {
      "epoch": 0.268,
      "grad_norm": 0.13640043139457703,
      "learning_rate": 0.000252477162969252,
      "loss": 0.3112,
      "num_input_tokens_seen": 1756364800,
      "step": 26800,
      "train_runtime": 17523.2088,
      "train_tokens_per_second": 100230.775
    },
    {
      "epoch": 0.269,
      "grad_norm": 0.12981313467025757,
      "learning_rate": 0.00025212904986916584,
      "loss": 0.3124,
      "num_input_tokens_seen": 1762918400,
      "step": 26900,
      "train_runtime": 17587.6922,
      "train_tokens_per_second": 100235.914
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.14338868856430054,
      "learning_rate": 0.00025177990833068133,
      "loss": 0.3124,
      "num_input_tokens_seen": 1769472000,
      "step": 27000,
      "train_runtime": 17658.758,
      "train_tokens_per_second": 100203.649
    },
    {
      "epoch": 0.271,
      "grad_norm": 0.17518877983093262,
      "learning_rate": 0.0002514297418696499,
      "loss": 0.3076,
      "num_input_tokens_seen": 1776025600,
      "step": 27100,
      "train_runtime": 17723.3886,
      "train_tokens_per_second": 100208.016
    },
    {
      "epoch": 0.272,
      "grad_norm": 0.1369880735874176,
      "learning_rate": 0.0002510785540122439,
      "loss": 0.3114,
      "num_input_tokens_seen": 1782579200,
      "step": 27200,
      "train_runtime": 17786.611,
      "train_tokens_per_second": 100220.283
    },
    {
      "epoch": 0.273,
      "grad_norm": 0.15111377835273743,
      "learning_rate": 0.0002507263482949212,
      "loss": 0.3144,
      "num_input_tokens_seen": 1789132800,
      "step": 27300,
      "train_runtime": 17852.1418,
      "train_tokens_per_second": 100219.504
    },
    {
      "epoch": 0.274,
      "grad_norm": 0.140447199344635,
      "learning_rate": 0.0002503731282643894,
      "loss": 0.3103,
      "num_input_tokens_seen": 1795686400,
      "step": 27400,
      "train_runtime": 17917.1236,
      "train_tokens_per_second": 100221.801
    },
    {
      "epoch": 0.275,
      "grad_norm": 0.1373315006494522,
      "learning_rate": 0.0002500188974775704,
      "loss": 0.3095,
      "num_input_tokens_seen": 1802240000,
      "step": 27500,
      "train_runtime": 17981.4799,
      "train_tokens_per_second": 100227.568
    },
    {
      "epoch": 0.276,
      "grad_norm": 0.1453147530555725,
      "learning_rate": 0.00024966365950156416,
      "loss": 0.3085,
      "num_input_tokens_seen": 1808793600,
      "step": 27600,
      "train_runtime": 18052.109,
      "train_tokens_per_second": 100198.464
    },
    {
      "epoch": 0.277,
      "grad_norm": 0.19097484648227692,
      "learning_rate": 0.00024930741791361326,
      "loss": 0.3128,
      "num_input_tokens_seen": 1815347200,
      "step": 27700,
      "train_runtime": 18117.9773,
      "train_tokens_per_second": 100195.909
    },
    {
      "epoch": 0.278,
      "grad_norm": 0.2222718745470047,
      "learning_rate": 0.0002489501763010664,
      "loss": 0.3107,
      "num_input_tokens_seen": 1821900800,
      "step": 27800,
      "train_runtime": 18178.1946,
      "train_tokens_per_second": 100224.519
    },
    {
      "epoch": 0.279,
      "grad_norm": 0.16960225999355316,
      "learning_rate": 0.00024859193826134285,
      "loss": 0.3093,
      "num_input_tokens_seen": 1828454400,
      "step": 27900,
      "train_runtime": 18248.1866,
      "train_tokens_per_second": 100199.238
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.15540289878845215,
      "learning_rate": 0.00024823270740189556,
      "loss": 0.3084,
      "num_input_tokens_seen": 1835008000,
      "step": 28000,
      "train_runtime": 18313.0722,
      "train_tokens_per_second": 100202.084
    },
    {
      "epoch": 0.281,
      "grad_norm": 0.1421203911304474,
      "learning_rate": 0.00024787248734017527,
      "loss": 0.3119,
      "num_input_tokens_seen": 1841561600,
      "step": 28100,
      "train_runtime": 18377.039,
      "train_tokens_per_second": 100209.919
    },
    {
      "epoch": 0.282,
      "grad_norm": 0.131204292178154,
      "learning_rate": 0.0002475112817035941,
      "loss": 0.3127,
      "num_input_tokens_seen": 1848115200,
      "step": 28200,
      "train_runtime": 18441.4656,
      "train_tokens_per_second": 100215.202
    },
    {
      "epoch": 0.283,
      "grad_norm": 0.1507508009672165,
      "learning_rate": 0.0002471490941294887,
      "loss": 0.3118,
      "num_input_tokens_seen": 1854668800,
      "step": 28300,
      "train_runtime": 18511.3095,
      "train_tokens_per_second": 100191.118
    },
    {
      "epoch": 0.284,
      "grad_norm": 0.12522923946380615,
      "learning_rate": 0.000246785928265084,
      "loss": 0.3104,
      "num_input_tokens_seen": 1861222400,
      "step": 28400,
      "train_runtime": 18574.4697,
      "train_tokens_per_second": 100203.259
    },
    {
      "epoch": 0.285,
      "grad_norm": 0.2087126076221466,
      "learning_rate": 0.0002464217877674562,
      "loss": 0.3132,
      "num_input_tokens_seen": 1867776000,
      "step": 28500,
      "train_runtime": 18638.8332,
      "train_tokens_per_second": 100208.848
    },
    {
      "epoch": 0.286,
      "grad_norm": 0.1495303064584732,
      "learning_rate": 0.0002460566763034961,
      "loss": 0.3159,
      "num_input_tokens_seen": 1874329600,
      "step": 28600,
      "train_runtime": 18703.8924,
      "train_tokens_per_second": 100210.671
    },
    {
      "epoch": 0.287,
      "grad_norm": 0.14563380181789398,
      "learning_rate": 0.00024569059754987196,
      "loss": 0.3116,
      "num_input_tokens_seen": 1880883200,
      "step": 28700,
      "train_runtime": 18774.7813,
      "train_tokens_per_second": 100181.364
    },
    {
      "epoch": 0.288,
      "grad_norm": 0.12803615629673004,
      "learning_rate": 0.00024532355519299296,
      "loss": 0.3099,
      "num_input_tokens_seen": 1887436800,
      "step": 28800,
      "train_runtime": 18838.435,
      "train_tokens_per_second": 100190.743
    },
    {
      "epoch": 0.289,
      "grad_norm": 0.5618897676467896,
      "learning_rate": 0.0002449555529289714,
      "loss": 0.3129,
      "num_input_tokens_seen": 1893990400,
      "step": 28900,
      "train_runtime": 18901.8999,
      "train_tokens_per_second": 100201.06
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.15488959848880768,
      "learning_rate": 0.0002445865944635861,
      "loss": 0.3155,
      "num_input_tokens_seen": 1900544000,
      "step": 29000,
      "train_runtime": 18967.9894,
      "train_tokens_per_second": 100197.441
    },
    {
      "epoch": 0.291,
      "grad_norm": 0.13676992058753967,
      "learning_rate": 0.0002442166835122446,
      "loss": 0.3101,
      "num_input_tokens_seen": 1907097600,
      "step": 29100,
      "train_runtime": 19031.1664,
      "train_tokens_per_second": 100209.181
    },
    {
      "epoch": 0.292,
      "grad_norm": 0.11402736604213715,
      "learning_rate": 0.00024384582379994614,
      "loss": 0.3094,
      "num_input_tokens_seen": 1913651200,
      "step": 29200,
      "train_runtime": 19096.1775,
      "train_tokens_per_second": 100211.218
    },
    {
      "epoch": 0.293,
      "grad_norm": 0.1358448714017868,
      "learning_rate": 0.00024347401906124388,
      "loss": 0.309,
      "num_input_tokens_seen": 1920204800,
      "step": 29300,
      "train_runtime": 19165.3098,
      "train_tokens_per_second": 100191.691
    },
    {
      "epoch": 0.294,
      "grad_norm": 0.14608891308307648,
      "learning_rate": 0.0002431012730402075,
      "loss": 0.3119,
      "num_input_tokens_seen": 1926758400,
      "step": 29400,
      "train_runtime": 19230.3069,
      "train_tokens_per_second": 100193.845
    },
    {
      "epoch": 0.295,
      "grad_norm": 0.1501711755990982,
      "learning_rate": 0.00024272758949038517,
      "loss": 0.3091,
      "num_input_tokens_seen": 1933312000,
      "step": 29500,
      "train_runtime": 19294.7627,
      "train_tokens_per_second": 100198.796
    },
    {
      "epoch": 0.296,
      "grad_norm": 0.1614496409893036,
      "learning_rate": 0.00024235297217476616,
      "loss": 0.3104,
      "num_input_tokens_seen": 1939865600,
      "step": 29600,
      "train_runtime": 19364.7415,
      "train_tokens_per_second": 100175.135
    },
    {
      "epoch": 0.297,
      "grad_norm": 0.11902807652950287,
      "learning_rate": 0.00024197742486574268,
      "loss": 0.3126,
      "num_input_tokens_seen": 1946419200,
      "step": 29700,
      "train_runtime": 19429.1038,
      "train_tokens_per_second": 100180.596
    },
    {
      "epoch": 0.298,
      "grad_norm": 0.12998123466968536,
      "learning_rate": 0.0002416009513450719,
      "loss": 0.3102,
      "num_input_tokens_seen": 1952972800,
      "step": 29800,
      "train_runtime": 19494.2244,
      "train_tokens_per_second": 100182.124
    },
    {
      "epoch": 0.299,
      "grad_norm": 0.2079559862613678,
      "learning_rate": 0.00024122355540383806,
      "loss": 0.311,
      "num_input_tokens_seen": 1959526400,
      "step": 29900,
      "train_runtime": 19559.2072,
      "train_tokens_per_second": 100184.347
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.15128397941589355,
      "learning_rate": 0.00024084524084241405,
      "loss": 0.3076,
      "num_input_tokens_seen": 1966080000,
      "step": 30000,
      "train_runtime": 19623.3669,
      "train_tokens_per_second": 100190.758
    },
    {
      "epoch": 0.301,
      "grad_norm": 0.13512304425239563,
      "learning_rate": 0.00024046601147042332,
      "loss": 0.3119,
      "num_input_tokens_seen": 1972633600,
      "step": 30100,
      "train_runtime": 19688.91,
      "train_tokens_per_second": 100190.086
    },
    {
      "epoch": 0.302,
      "grad_norm": 0.12716713547706604,
      "learning_rate": 0.0002400858711067015,
      "loss": 0.3093,
      "num_input_tokens_seen": 1979187200,
      "step": 30200,
      "train_runtime": 19753.5863,
      "train_tokens_per_second": 100193.816
    },
    {
      "epoch": 0.303,
      "grad_norm": 0.1301889717578888,
      "learning_rate": 0.00023970482357925772,
      "loss": 0.31,
      "num_input_tokens_seen": 1985740800,
      "step": 30300,
      "train_runtime": 19823.6081,
      "train_tokens_per_second": 100170.503
    },
    {
      "epoch": 0.304,
      "grad_norm": 0.13871292769908905,
      "learning_rate": 0.00023932287272523646,
      "loss": 0.3084,
      "num_input_tokens_seen": 1992294400,
      "step": 30400,
      "train_runtime": 19887.7656,
      "train_tokens_per_second": 100176.885
    },
    {
      "epoch": 0.305,
      "grad_norm": 0.12449346482753754,
      "learning_rate": 0.00023894002239087847,
      "loss": 0.3276,
      "num_input_tokens_seen": 1998848000,
      "step": 30500,
      "train_runtime": 19952.5714,
      "train_tokens_per_second": 100179.97
    },
    {
      "epoch": 0.306,
      "grad_norm": 0.1523977369070053,
      "learning_rate": 0.0002385562764314825,
      "loss": 0.3097,
      "num_input_tokens_seen": 2005401600,
      "step": 30600,
      "train_runtime": 20017.8352,
      "train_tokens_per_second": 100180.743
    },
    {
      "epoch": 0.307,
      "grad_norm": 0.1439458280801773,
      "learning_rate": 0.00023817163871136596,
      "loss": 0.3048,
      "num_input_tokens_seen": 2011955200,
      "step": 30700,
      "train_runtime": 20081.8889,
      "train_tokens_per_second": 100187.548
    },
    {
      "epoch": 0.308,
      "grad_norm": 0.12756380438804626,
      "learning_rate": 0.00023778611310382652,
      "loss": 0.3075,
      "num_input_tokens_seen": 2018508800,
      "step": 30800,
      "train_runtime": 20145.6107,
      "train_tokens_per_second": 100195.96
    },
    {
      "epoch": 0.309,
      "grad_norm": 0.14607320725917816,
      "learning_rate": 0.0002373997034911027,
      "loss": 0.3139,
      "num_input_tokens_seen": 2025062400,
      "step": 30900,
      "train_runtime": 20210.9796,
      "train_tokens_per_second": 100196.153
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.12456675618886948,
      "learning_rate": 0.00023701241376433506,
      "loss": 0.3089,
      "num_input_tokens_seen": 2031616000,
      "step": 31000,
      "train_runtime": 20281.0675,
      "train_tokens_per_second": 100173.031
    },
    {
      "epoch": 0.311,
      "grad_norm": 0.13834626972675323,
      "learning_rate": 0.0002366242478235268,
      "loss": 0.3066,
      "num_input_tokens_seen": 2038169600,
      "step": 31100,
      "train_runtime": 20346.0263,
      "train_tokens_per_second": 100175.315
    },
    {
      "epoch": 0.312,
      "grad_norm": 0.1534184068441391,
      "learning_rate": 0.00023623520957750471,
      "loss": 0.3082,
      "num_input_tokens_seen": 2044723200,
      "step": 31200,
      "train_runtime": 20409.76,
      "train_tokens_per_second": 100183.598
    },
    {
      "epoch": 0.313,
      "grad_norm": 0.12966671586036682,
      "learning_rate": 0.00023584530294387953,
      "loss": 0.3126,
      "num_input_tokens_seen": 2051276800,
      "step": 31300,
      "train_runtime": 20475.6348,
      "train_tokens_per_second": 100181.353
    },
    {
      "epoch": 0.314,
      "grad_norm": 0.14474999904632568,
      "learning_rate": 0.00023545453184900682,
      "loss": 0.3091,
      "num_input_tokens_seen": 2057830400,
      "step": 31400,
      "train_runtime": 20539.196,
      "train_tokens_per_second": 100190.407
    },
    {
      "epoch": 0.315,
      "grad_norm": 0.13208946585655212,
      "learning_rate": 0.00023506290022794706,
      "loss": 0.3095,
      "num_input_tokens_seen": 2064384000,
      "step": 31500,
      "train_runtime": 20604.221,
      "train_tokens_per_second": 100192.286
    },
    {
      "epoch": 0.316,
      "grad_norm": 0.15090374648571014,
      "learning_rate": 0.00023467041202442643,
      "loss": 0.3073,
      "num_input_tokens_seen": 2070937600,
      "step": 31600,
      "train_runtime": 20674.5759,
      "train_tokens_per_second": 100168.323
    },
    {
      "epoch": 0.317,
      "grad_norm": 0.18638543784618378,
      "learning_rate": 0.00023427707119079669,
      "loss": 0.312,
      "num_input_tokens_seen": 2077491200,
      "step": 31700,
      "train_runtime": 20738.8671,
      "train_tokens_per_second": 100173.804
    },
    {
      "epoch": 0.318,
      "grad_norm": 0.1385478377342224,
      "learning_rate": 0.0002338828816879957,
      "loss": 0.3095,
      "num_input_tokens_seen": 2084044800,
      "step": 31800,
      "train_runtime": 20802.7906,
      "train_tokens_per_second": 100181.021
    },
    {
      "epoch": 0.319,
      "grad_norm": 0.15265443921089172,
      "learning_rate": 0.00023348784748550744,
      "loss": 0.3103,
      "num_input_tokens_seen": 2090598400,
      "step": 31900,
      "train_runtime": 20868.0311,
      "train_tokens_per_second": 100181.871
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.15918248891830444,
      "learning_rate": 0.00023309197256132184,
      "loss": 0.3102,
      "num_input_tokens_seen": 2097152000,
      "step": 32000,
      "train_runtime": 20937.8931,
      "train_tokens_per_second": 100160.603
    },
    {
      "epoch": 0.321,
      "grad_norm": 0.14801020920276642,
      "learning_rate": 0.00023269526090189505,
      "loss": 0.3147,
      "num_input_tokens_seen": 2103705600,
      "step": 32100,
      "train_runtime": 21002.9142,
      "train_tokens_per_second": 100162.557
    },
    {
      "epoch": 0.322,
      "grad_norm": 0.18616679310798645,
      "learning_rate": 0.00023229771650210907,
      "loss": 0.3099,
      "num_input_tokens_seen": 2110259200,
      "step": 32200,
      "train_runtime": 21067.872,
      "train_tokens_per_second": 100164.801
    },
    {
      "epoch": 0.323,
      "grad_norm": 0.13931268453598022,
      "learning_rate": 0.00023189934336523163,
      "loss": 0.3115,
      "num_input_tokens_seen": 2116812800,
      "step": 32300,
      "train_runtime": 21131.2256,
      "train_tokens_per_second": 100174.634
    },
    {
      "epoch": 0.324,
      "grad_norm": 0.1734631061553955,
      "learning_rate": 0.00023150014550287574,
      "loss": 0.3112,
      "num_input_tokens_seen": 2123366400,
      "step": 32400,
      "train_runtime": 21201.6285,
      "train_tokens_per_second": 100151.099
    },
    {
      "epoch": 0.325,
      "grad_norm": 0.13876596093177795,
      "learning_rate": 0.00023110012693495943,
      "loss": 0.31,
      "num_input_tokens_seen": 2129920000,
      "step": 32500,
      "train_runtime": 21265.8205,
      "train_tokens_per_second": 100156.963
    },
    {
      "epoch": 0.326,
      "grad_norm": 0.20441171526908875,
      "learning_rate": 0.00023069929168966527,
      "loss": 0.3095,
      "num_input_tokens_seen": 2136473600,
      "step": 32600,
      "train_runtime": 21329.6315,
      "train_tokens_per_second": 100164.581
    },
    {
      "epoch": 0.327,
      "grad_norm": 0.12022672593593597,
      "learning_rate": 0.0002302976438033997,
      "loss": 0.3089,
      "num_input_tokens_seen": 2143027200,
      "step": 32700,
      "train_runtime": 21394.0086,
      "train_tokens_per_second": 100169.502
    },
    {
      "epoch": 0.328,
      "grad_norm": 0.23158074915409088,
      "learning_rate": 0.0002298951873207525,
      "loss": 0.3121,
      "num_input_tokens_seen": 2149580800,
      "step": 32800,
      "train_runtime": 21459.8938,
      "train_tokens_per_second": 100167.355
    },
    {
      "epoch": 0.329,
      "grad_norm": 0.11978685855865479,
      "learning_rate": 0.00022949192629445606,
      "loss": 0.308,
      "num_input_tokens_seen": 2156134400,
      "step": 32900,
      "train_runtime": 21524.2825,
      "train_tokens_per_second": 100172.185
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.16882842779159546,
      "learning_rate": 0.0002290878647853443,
      "loss": 0.3076,
      "num_input_tokens_seen": 2162688000,
      "step": 33000,
      "train_runtime": 21595.0222,
      "train_tokens_per_second": 100147.524
    },
    {
      "epoch": 0.331,
      "grad_norm": 0.1368299126625061,
      "learning_rate": 0.00022868300686231224,
      "loss": 0.3078,
      "num_input_tokens_seen": 2169241600,
      "step": 33100,
      "train_runtime": 21659.0361,
      "train_tokens_per_second": 100154.115
    },
    {
      "epoch": 0.332,
      "grad_norm": 0.13301041722297668,
      "learning_rate": 0.00022827735660227457,
      "loss": 0.3103,
      "num_input_tokens_seen": 2175795200,
      "step": 33200,
      "train_runtime": 21723.8934,
      "train_tokens_per_second": 100156.779
    },
    {
      "epoch": 0.333,
      "grad_norm": 0.13545189797878265,
      "learning_rate": 0.000227870918090125,
      "loss": 0.3068,
      "num_input_tokens_seen": 2182348800,
      "step": 33300,
      "train_runtime": 21788.4359,
      "train_tokens_per_second": 100160.875
    },
    {
      "epoch": 0.334,
      "grad_norm": 0.2138141542673111,
      "learning_rate": 0.00022746369541869476,
      "loss": 0.3059,
      "num_input_tokens_seen": 2188902400,
      "step": 33400,
      "train_runtime": 21853.4857,
      "train_tokens_per_second": 100162.621
    },
    {
      "epoch": 0.335,
      "grad_norm": 0.1255991905927658,
      "learning_rate": 0.00022705569268871163,
      "loss": 0.3099,
      "num_input_tokens_seen": 2195456000,
      "step": 33500,
      "train_runtime": 21918.1728,
      "train_tokens_per_second": 100166.014
    },
    {
      "epoch": 0.336,
      "grad_norm": 0.1330287754535675,
      "learning_rate": 0.00022664691400875865,
      "loss": 0.3093,
      "num_input_tokens_seen": 2202009600,
      "step": 33600,
      "train_runtime": 21987.6743,
      "train_tokens_per_second": 100147.454
    },
    {
      "epoch": 0.337,
      "grad_norm": 0.1321260631084442,
      "learning_rate": 0.00022623736349523254,
      "loss": 0.3109,
      "num_input_tokens_seen": 2208563200,
      "step": 33700,
      "train_runtime": 22052.5483,
      "train_tokens_per_second": 100150.022
    },
    {
      "epoch": 0.338,
      "grad_norm": 0.13865865767002106,
      "learning_rate": 0.00022582704527230238,
      "loss": 0.3068,
      "num_input_tokens_seen": 2215116800,
      "step": 33800,
      "train_runtime": 22117.0958,
      "train_tokens_per_second": 100154.054
    },
    {
      "epoch": 0.339,
      "grad_norm": 0.13597998023033142,
      "learning_rate": 0.0002254159634718682,
      "loss": 0.3061,
      "num_input_tokens_seen": 2221670400,
      "step": 33900,
      "train_runtime": 22180.0605,
      "train_tokens_per_second": 100165.209
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.14176584780216217,
      "learning_rate": 0.00022500412223351915,
      "loss": 0.3114,
      "num_input_tokens_seen": 2228224000,
      "step": 34000,
      "train_runtime": 22251.2759,
      "train_tokens_per_second": 100139.157
    },
    {
      "epoch": 0.341,
      "grad_norm": 0.13006241619586945,
      "learning_rate": 0.0002245915257044919,
      "loss": 0.3071,
      "num_input_tokens_seen": 2234777600,
      "step": 34100,
      "train_runtime": 22315.7056,
      "train_tokens_per_second": 100143.712
    },
    {
      "epoch": 0.342,
      "grad_norm": 0.186634823679924,
      "learning_rate": 0.00022417817803962892,
      "loss": 0.3032,
      "num_input_tokens_seen": 2241331200,
      "step": 34200,
      "train_runtime": 22380.1064,
      "train_tokens_per_second": 100148.371
    },
    {
      "epoch": 0.343,
      "grad_norm": 0.1767393946647644,
      "learning_rate": 0.0002237640834013366,
      "loss": 0.3085,
      "num_input_tokens_seen": 2247884800,
      "step": 34300,
      "train_runtime": 22444.6012,
      "train_tokens_per_second": 100152.584
    },
    {
      "epoch": 0.344,
      "grad_norm": 0.15075454115867615,
      "learning_rate": 0.0002233492459595434,
      "loss": 0.3099,
      "num_input_tokens_seen": 2254438400,
      "step": 34400,
      "train_runtime": 22509.6493,
      "train_tokens_per_second": 100154.31
    },
    {
      "epoch": 0.345,
      "grad_norm": 0.15754783153533936,
      "learning_rate": 0.00022293366989165772,
      "loss": 0.307,
      "num_input_tokens_seen": 2260992000,
      "step": 34500,
      "train_runtime": 22579.4848,
      "train_tokens_per_second": 100134.791
    },
    {
      "epoch": 0.346,
      "grad_norm": 0.13372038304805756,
      "learning_rate": 0.00022251735938252587,
      "loss": 0.3066,
      "num_input_tokens_seen": 2267545600,
      "step": 34600,
      "train_runtime": 22643.953,
      "train_tokens_per_second": 100139.123
    },
    {
      "epoch": 0.347,
      "grad_norm": 0.17753738164901733,
      "learning_rate": 0.0002221003186243902,
      "loss": 0.3087,
      "num_input_tokens_seen": 2274099200,
      "step": 34700,
      "train_runtime": 22708.6869,
      "train_tokens_per_second": 100142.259
    },
    {
      "epoch": 0.348,
      "grad_norm": 0.1375788450241089,
      "learning_rate": 0.00022168255181684643,
      "loss": 0.3064,
      "num_input_tokens_seen": 2280652800,
      "step": 34800,
      "train_runtime": 22774.2018,
      "train_tokens_per_second": 100141.942
    },
    {
      "epoch": 0.349,
      "grad_norm": 0.14929898083209991,
      "learning_rate": 0.00022126406316680172,
      "loss": 0.3108,
      "num_input_tokens_seen": 2287206400,
      "step": 34900,
      "train_runtime": 22839.776,
      "train_tokens_per_second": 100141.367
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.15789327025413513,
      "learning_rate": 0.00022084485688843208,
      "loss": 0.3082,
      "num_input_tokens_seen": 2293760000,
      "step": 35000,
      "train_runtime": 22904.3853,
      "train_tokens_per_second": 100145.015
    },
    {
      "epoch": 0.351,
      "grad_norm": 0.1339723765850067,
      "learning_rate": 0.00022042493720314003,
      "loss": 0.3127,
      "num_input_tokens_seen": 2300313600,
      "step": 35100,
      "train_runtime": 22968.8594,
      "train_tokens_per_second": 100149.231
    },
    {
      "epoch": 0.352,
      "grad_norm": 0.14159700274467468,
      "learning_rate": 0.00022000430833951228,
      "loss": 0.3096,
      "num_input_tokens_seen": 2306867200,
      "step": 35200,
      "train_runtime": 23033.0283,
      "train_tokens_per_second": 100154.751
    },
    {
      "epoch": 0.353,
      "grad_norm": 0.17289403080940247,
      "learning_rate": 0.00021958297453327673,
      "loss": 0.3058,
      "num_input_tokens_seen": 2313420800,
      "step": 35300,
      "train_runtime": 23103.5037,
      "train_tokens_per_second": 100132.899
    },
    {
      "epoch": 0.354,
      "grad_norm": 0.1353076845407486,
      "learning_rate": 0.00021916094002726012,
      "loss": 0.3048,
      "num_input_tokens_seen": 2319974400,
      "step": 35400,
      "train_runtime": 23166.8292,
      "train_tokens_per_second": 100142.077
    },
    {
      "epoch": 0.355,
      "grad_norm": 0.12303294241428375,
      "learning_rate": 0.00021873820907134534,
      "loss": 0.3102,
      "num_input_tokens_seen": 2326528000,
      "step": 35500,
      "train_runtime": 23232.6655,
      "train_tokens_per_second": 100140.382
    },
    {
      "epoch": 0.356,
      "grad_norm": 0.14765286445617676,
      "learning_rate": 0.0002183147859224283,
      "loss": 0.3106,
      "num_input_tokens_seen": 2333081600,
      "step": 35600,
      "train_runtime": 23296.4196,
      "train_tokens_per_second": 100147.647
    },
    {
      "epoch": 0.357,
      "grad_norm": 0.13833215832710266,
      "learning_rate": 0.00021789067484437544,
      "loss": 0.3055,
      "num_input_tokens_seen": 2339635200,
      "step": 35700,
      "train_runtime": 23361.5704,
      "train_tokens_per_second": 100148.884
    },
    {
      "epoch": 0.358,
      "grad_norm": 0.13157132267951965,
      "learning_rate": 0.00021746588010798068,
      "loss": 0.3081,
      "num_input_tokens_seen": 2346188800,
      "step": 35800,
      "train_runtime": 23430.7927,
      "train_tokens_per_second": 100132.711
    },
    {
      "epoch": 0.359,
      "grad_norm": 0.12913836538791656,
      "learning_rate": 0.00021704040599092216,
      "loss": 0.3094,
      "num_input_tokens_seen": 2352742400,
      "step": 35900,
      "train_runtime": 23495.4052,
      "train_tokens_per_second": 100136.277
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.13528013229370117,
      "learning_rate": 0.00021661425677771965,
      "loss": 0.3061,
      "num_input_tokens_seen": 2359296000,
      "step": 36000,
      "train_runtime": 23559.8424,
      "train_tokens_per_second": 100140.568
    },
    {
      "epoch": 0.361,
      "grad_norm": 0.15519119799137115,
      "learning_rate": 0.00021618743675969095,
      "loss": 0.3065,
      "num_input_tokens_seen": 2365849600,
      "step": 36100,
      "train_runtime": 23624.7603,
      "train_tokens_per_second": 100142.798
    },
    {
      "epoch": 0.362,
      "grad_norm": 0.14744772017002106,
      "learning_rate": 0.0002157599502349089,
      "loss": 0.3068,
      "num_input_tokens_seen": 2372403200,
      "step": 36200,
      "train_runtime": 23688.8845,
      "train_tokens_per_second": 100148.371
    },
    {
      "epoch": 0.363,
      "grad_norm": 0.13838911056518555,
      "learning_rate": 0.00021533180150815802,
      "loss": 0.3097,
      "num_input_tokens_seen": 2378956800,
      "step": 36300,
      "train_runtime": 23759.9908,
      "train_tokens_per_second": 100124.483
    },
    {
      "epoch": 0.364,
      "grad_norm": 0.12536117434501648,
      "learning_rate": 0.00021490299489089132,
      "loss": 0.3067,
      "num_input_tokens_seen": 2385510400,
      "step": 36400,
      "train_runtime": 23823.7123,
      "train_tokens_per_second": 100131.767
    },
    {
      "epoch": 0.365,
      "grad_norm": 0.14205192029476166,
      "learning_rate": 0.00021447353470118656,
      "loss": 0.3049,
      "num_input_tokens_seen": 2392064000,
      "step": 36500,
      "train_runtime": 23887.5453,
      "train_tokens_per_second": 100138.544
    },
    {
      "epoch": 0.366,
      "grad_norm": 0.11950815469026566,
      "learning_rate": 0.00021404342526370326,
      "loss": 0.3072,
      "num_input_tokens_seen": 2398617600,
      "step": 36600,
      "train_runtime": 23951.3108,
      "train_tokens_per_second": 100145.567
    },
    {
      "epoch": 0.367,
      "grad_norm": 0.1286599189043045,
      "learning_rate": 0.00021361267090963846,
      "loss": 0.3096,
      "num_input_tokens_seen": 2405171200,
      "step": 36700,
      "train_runtime": 24016.5354,
      "train_tokens_per_second": 100146.468
    },
    {
      "epoch": 0.368,
      "grad_norm": 0.12663663923740387,
      "learning_rate": 0.0002131812759766839,
      "loss": 0.3054,
      "num_input_tokens_seen": 2411724800,
      "step": 36800,
      "train_runtime": 24085.8974,
      "train_tokens_per_second": 100130.162
    },
    {
      "epoch": 0.369,
      "grad_norm": 0.16495896875858307,
      "learning_rate": 0.00021274924480898169,
      "loss": 0.3037,
      "num_input_tokens_seen": 2418278400,
      "step": 36900,
      "train_runtime": 24149.4634,
      "train_tokens_per_second": 100137.977
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.13351881504058838,
      "learning_rate": 0.00021231658175708087,
      "loss": 0.309,
      "num_input_tokens_seen": 2424832000,
      "step": 37000,
      "train_runtime": 24214.3635,
      "train_tokens_per_second": 100140.233
    },
    {
      "epoch": 0.371,
      "grad_norm": 0.13137440383434296,
      "learning_rate": 0.00021188329117789357,
      "loss": 0.3061,
      "num_input_tokens_seen": 2431385600,
      "step": 37100,
      "train_runtime": 24284.8537,
      "train_tokens_per_second": 100119.426
    },
    {
      "epoch": 0.372,
      "grad_norm": 0.17069390416145325,
      "learning_rate": 0.0002114493774346512,
      "loss": 0.3075,
      "num_input_tokens_seen": 2437939200,
      "step": 37200,
      "train_runtime": 24349.7441,
      "train_tokens_per_second": 100121.759
    },
    {
      "epoch": 0.373,
      "grad_norm": 0.13554754853248596,
      "learning_rate": 0.00021101484489686025,
      "loss": 0.3056,
      "num_input_tokens_seen": 2444492800,
      "step": 37300,
      "train_runtime": 24413.4106,
      "train_tokens_per_second": 100129.099
    },
    {
      "epoch": 0.374,
      "grad_norm": 0.24161159992218018,
      "learning_rate": 0.00021057969794025866,
      "loss": 0.3084,
      "num_input_tokens_seen": 2451046400,
      "step": 37400,
      "train_runtime": 24479.2787,
      "train_tokens_per_second": 100127.395
    },
    {
      "epoch": 0.375,
      "grad_norm": 0.11480960994958878,
      "learning_rate": 0.00021014394094677128,
      "loss": 0.3065,
      "num_input_tokens_seen": 2457600000,
      "step": 37500,
      "train_runtime": 24543.1085,
      "train_tokens_per_second": 100134.015
    },
    {
      "epoch": 0.376,
      "grad_norm": 0.1333978921175003,
      "learning_rate": 0.00020970757830446633,
      "loss": 0.3047,
      "num_input_tokens_seen": 2464153600,
      "step": 37600,
      "train_runtime": 24612.4036,
      "train_tokens_per_second": 100118.365
    },
    {
      "epoch": 0.377,
      "grad_norm": 0.1306515485048294,
      "learning_rate": 0.00020927061440751072,
      "loss": 0.3039,
      "num_input_tokens_seen": 2470707200,
      "step": 37700,
      "train_runtime": 24676.7406,
      "train_tokens_per_second": 100122.915
    },
    {
      "epoch": 0.378,
      "grad_norm": 0.19177651405334473,
      "learning_rate": 0.00020883305365612602,
      "loss": 0.3091,
      "num_input_tokens_seen": 2477260800,
      "step": 37800,
      "train_runtime": 24742.4612,
      "train_tokens_per_second": 100121.842
    },
    {
      "epoch": 0.379,
      "grad_norm": 0.14794479310512543,
      "learning_rate": 0.00020839490045654425,
      "loss": 0.3103,
      "num_input_tokens_seen": 2483814400,
      "step": 37900,
      "train_runtime": 24807.833,
      "train_tokens_per_second": 100122.183
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.1391579508781433,
      "learning_rate": 0.00020795615922096313,
      "loss": 0.305,
      "num_input_tokens_seen": 2490368000,
      "step": 38000,
      "train_runtime": 24871.0815,
      "train_tokens_per_second": 100131.07
    },
    {
      "epoch": 0.381,
      "grad_norm": 0.14466038346290588,
      "learning_rate": 0.00020751683436750207,
      "loss": 0.3066,
      "num_input_tokens_seen": 2496921600,
      "step": 38100,
      "train_runtime": 24941.5584,
      "train_tokens_per_second": 100110.89
    },
    {
      "epoch": 0.382,
      "grad_norm": 0.14706650376319885,
      "learning_rate": 0.00020707693032015752,
      "loss": 0.3131,
      "num_input_tokens_seen": 2503475200,
      "step": 38200,
      "train_runtime": 25006.658,
      "train_tokens_per_second": 100112.346
    },
    {
      "epoch": 0.383,
      "grad_norm": 0.1455349326133728,
      "learning_rate": 0.00020663645150875834,
      "loss": 0.3058,
      "num_input_tokens_seen": 2510028800,
      "step": 38300,
      "train_runtime": 25070.3473,
      "train_tokens_per_second": 100119.427
    },
    {
      "epoch": 0.384,
      "grad_norm": 0.13858123123645782,
      "learning_rate": 0.00020619540236892125,
      "loss": 0.3066,
      "num_input_tokens_seen": 2516582400,
      "step": 38400,
      "train_runtime": 25135.6982,
      "train_tokens_per_second": 100119.853
    },
    {
      "epoch": 0.385,
      "grad_norm": 0.17408473789691925,
      "learning_rate": 0.00020575378734200616,
      "loss": 0.3068,
      "num_input_tokens_seen": 2523136000,
      "step": 38500,
      "train_runtime": 25206.1351,
      "train_tokens_per_second": 100100.075
    },
    {
      "epoch": 0.386,
      "grad_norm": 0.12729153037071228,
      "learning_rate": 0.0002053116108750715,
      "loss": 0.3062,
      "num_input_tokens_seen": 2529689600,
      "step": 38600,
      "train_runtime": 25270.823,
      "train_tokens_per_second": 100103.174
    },
    {
      "epoch": 0.387,
      "grad_norm": 0.15452224016189575,
      "learning_rate": 0.0002048688774208294,
      "loss": 0.3029,
      "num_input_tokens_seen": 2536243200,
      "step": 38700,
      "train_runtime": 25334.6018,
      "train_tokens_per_second": 100109.851
    },
    {
      "epoch": 0.388,
      "grad_norm": 0.11749983578920364,
      "learning_rate": 0.0002044255914376009,
      "loss": 0.3055,
      "num_input_tokens_seen": 2542796800,
      "step": 38800,
      "train_runtime": 25398.9456,
      "train_tokens_per_second": 100114.266
    },
    {
      "epoch": 0.389,
      "grad_norm": 0.12558670341968536,
      "learning_rate": 0.00020398175738927082,
      "loss": 0.307,
      "num_input_tokens_seen": 2549350400,
      "step": 38900,
      "train_runtime": 25469.3443,
      "train_tokens_per_second": 100094.858
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.11652723699808121,
      "learning_rate": 0.00020353737974524312,
      "loss": 0.3059,
      "num_input_tokens_seen": 2555904000,
      "step": 39000,
      "train_runtime": 25534.1962,
      "train_tokens_per_second": 100097.296
    },
    {
      "epoch": 0.391,
      "grad_norm": 0.14530417323112488,
      "learning_rate": 0.00020309246298039584,
      "loss": 0.3043,
      "num_input_tokens_seen": 2562457600,
      "step": 39100,
      "train_runtime": 25597.7668,
      "train_tokens_per_second": 100104.733
    },
    {
      "epoch": 0.392,
      "grad_norm": 0.2145591825246811,
      "learning_rate": 0.0002026470115750357,
      "loss": 0.3097,
      "num_input_tokens_seen": 2569011200,
      "step": 39200,
      "train_runtime": 25662.2383,
      "train_tokens_per_second": 100108.618
    },
    {
      "epoch": 0.393,
      "grad_norm": 0.13407446444034576,
      "learning_rate": 0.0002022010300148535,
      "loss": 0.3072,
      "num_input_tokens_seen": 2575564800,
      "step": 39300,
      "train_runtime": 25726.7635,
      "train_tokens_per_second": 100112.274
    },
    {
      "epoch": 0.394,
      "grad_norm": 0.20070548355579376,
      "learning_rate": 0.0002017545227908786,
      "loss": 0.3042,
      "num_input_tokens_seen": 2582118400,
      "step": 39400,
      "train_runtime": 25798.3829,
      "train_tokens_per_second": 100088.382
    },
    {
      "epoch": 0.395,
      "grad_norm": 0.12969562411308289,
      "learning_rate": 0.00020130749439943376,
      "loss": 0.3025,
      "num_input_tokens_seen": 2588672000,
      "step": 39500,
      "train_runtime": 25861.9837,
      "train_tokens_per_second": 100095.647
    },
    {
      "epoch": 0.396,
      "grad_norm": 0.22430787980556488,
      "learning_rate": 0.00020085994934208998,
      "loss": 0.3075,
      "num_input_tokens_seen": 2595225600,
      "step": 39600,
      "train_runtime": 25927.1388,
      "train_tokens_per_second": 100096.876
    },
    {
      "epoch": 0.397,
      "grad_norm": 0.1543964445590973,
      "learning_rate": 0.00020041189212562094,
      "loss": 0.3061,
      "num_input_tokens_seen": 2601779200,
      "step": 39700,
      "train_runtime": 25990.8084,
      "train_tokens_per_second": 100103.82
    },
    {
      "epoch": 0.398,
      "grad_norm": 0.17474599182605743,
      "learning_rate": 0.0001999633272619579,
      "loss": 0.3026,
      "num_input_tokens_seen": 2608332800,
      "step": 39800,
      "train_runtime": 26055.1661,
      "train_tokens_per_second": 100108.086
    },
    {
      "epoch": 0.399,
      "grad_norm": 0.12200487405061722,
      "learning_rate": 0.00019951425926814404,
      "loss": 0.3051,
      "num_input_tokens_seen": 2614886400,
      "step": 39900,
      "train_runtime": 26125.5167,
      "train_tokens_per_second": 100089.366
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.12909364700317383,
      "learning_rate": 0.00019906469266628904,
      "loss": 0.3083,
      "num_input_tokens_seen": 2621440000,
      "step": 40000,
      "train_runtime": 26189.9855,
      "train_tokens_per_second": 100093.221
    },
    {
      "epoch": 0.401,
      "grad_norm": 0.14507311582565308,
      "learning_rate": 0.0001986146319835236,
      "loss": 0.3063,
      "num_input_tokens_seen": 2627993600,
      "step": 40100,
      "train_runtime": 26254.1189,
      "train_tokens_per_second": 100098.335
    },
    {
      "epoch": 0.402,
      "grad_norm": 0.15015749633312225,
      "learning_rate": 0.00019816408175195383,
      "loss": 0.3024,
      "num_input_tokens_seen": 2634547200,
      "step": 40200,
      "train_runtime": 26317.4656,
      "train_tokens_per_second": 100106.417
    },
    {
      "epoch": 0.403,
      "grad_norm": 0.1793050467967987,
      "learning_rate": 0.0001977130465086155,
      "loss": 0.3058,
      "num_input_tokens_seen": 2641100800,
      "step": 40300,
      "train_runtime": 26387.6285,
      "train_tokens_per_second": 100088.6
    },
    {
      "epoch": 0.404,
      "grad_norm": 0.13494957983493805,
      "learning_rate": 0.0001972615307954286,
      "loss": 0.3058,
      "num_input_tokens_seen": 2647654400,
      "step": 40400,
      "train_runtime": 26452.3646,
      "train_tokens_per_second": 100091.407
    },
    {
      "epoch": 0.405,
      "grad_norm": 0.15225248038768768,
      "learning_rate": 0.00019680953915915124,
      "loss": 0.3032,
      "num_input_tokens_seen": 2654208000,
      "step": 40500,
      "train_runtime": 26516.6796,
      "train_tokens_per_second": 100095.79
    },
    {
      "epoch": 0.406,
      "grad_norm": 0.15482735633850098,
      "learning_rate": 0.00019635707615133427,
      "loss": 0.3061,
      "num_input_tokens_seen": 2660761600,
      "step": 40600,
      "train_runtime": 26585.3848,
      "train_tokens_per_second": 100083.622
    },
    {
      "epoch": 0.407,
      "grad_norm": 0.15725013613700867,
      "learning_rate": 0.00019590414632827513,
      "loss": 0.3101,
      "num_input_tokens_seen": 2667315200,
      "step": 40700,
      "train_runtime": 26649.9092,
      "train_tokens_per_second": 100087.215
    },
    {
      "epoch": 0.408,
      "grad_norm": 0.16835036873817444,
      "learning_rate": 0.00019545075425097204,
      "loss": 0.3049,
      "num_input_tokens_seen": 2673868800,
      "step": 40800,
      "train_runtime": 26714.9814,
      "train_tokens_per_second": 100088.739
    },
    {
      "epoch": 0.409,
      "grad_norm": 0.167361319065094,
      "learning_rate": 0.00019499690448507827,
      "loss": 0.3027,
      "num_input_tokens_seen": 2680422400,
      "step": 40900,
      "train_runtime": 26779.2716,
      "train_tokens_per_second": 100093.178
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.1781291663646698,
      "learning_rate": 0.00019454260160085588,
      "loss": 0.3005,
      "num_input_tokens_seen": 2686976000,
      "step": 41000,
      "train_runtime": 26843.9197,
      "train_tokens_per_second": 100096.261
    },
    {
      "epoch": 0.411,
      "grad_norm": 0.1289975345134735,
      "learning_rate": 0.0001940878501731299,
      "loss": 0.3085,
      "num_input_tokens_seen": 2693529600,
      "step": 41100,
      "train_runtime": 26914.2047,
      "train_tokens_per_second": 100078.365
    },
    {
      "epoch": 0.412,
      "grad_norm": 0.12804220616817474,
      "learning_rate": 0.00019363265478124214,
      "loss": 0.3062,
      "num_input_tokens_seen": 2700083200,
      "step": 41200,
      "train_runtime": 26979.3069,
      "train_tokens_per_second": 100079.784
    },
    {
      "epoch": 0.413,
      "grad_norm": 0.14838483929634094,
      "learning_rate": 0.00019317702000900516,
      "loss": 0.3065,
      "num_input_tokens_seen": 2706636800,
      "step": 41300,
      "train_runtime": 27043.7101,
      "train_tokens_per_second": 100083.783
    },
    {
      "epoch": 0.414,
      "grad_norm": 0.3049434423446655,
      "learning_rate": 0.000192720950444656,
      "loss": 0.3075,
      "num_input_tokens_seen": 2713190400,
      "step": 41400,
      "train_runtime": 27108.2869,
      "train_tokens_per_second": 100087.121
    },
    {
      "epoch": 0.415,
      "grad_norm": 0.16474822163581848,
      "learning_rate": 0.00019226445068081018,
      "loss": 0.3087,
      "num_input_tokens_seen": 2719744000,
      "step": 41500,
      "train_runtime": 27173.4382,
      "train_tokens_per_second": 100088.328
    },
    {
      "epoch": 0.416,
      "grad_norm": 0.18445253372192383,
      "learning_rate": 0.00019180752531441523,
      "loss": 0.3065,
      "num_input_tokens_seen": 2726297600,
      "step": 41600,
      "train_runtime": 27237.7945,
      "train_tokens_per_second": 100092.45
    },
    {
      "epoch": 0.417,
      "grad_norm": 0.1226682960987091,
      "learning_rate": 0.00019135017894670456,
      "loss": 0.3062,
      "num_input_tokens_seen": 2732851200,
      "step": 41700,
      "train_runtime": 27307.5255,
      "train_tokens_per_second": 100076.852
    },
    {
      "epoch": 0.418,
      "grad_norm": 0.12846247851848602,
      "learning_rate": 0.0001908924161831509,
      "loss": 0.3064,
      "num_input_tokens_seen": 2739404800,
      "step": 41800,
      "train_runtime": 27371.4125,
      "train_tokens_per_second": 100082.698
    },
    {
      "epoch": 0.419,
      "grad_norm": 0.14241133630275726,
      "learning_rate": 0.0001904342416334203,
      "loss": 0.3048,
      "num_input_tokens_seen": 2745958400,
      "step": 41900,
      "train_runtime": 27436.5912,
      "train_tokens_per_second": 100083.803
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.19496770203113556,
      "learning_rate": 0.00018997565991132532,
      "loss": 0.3046,
      "num_input_tokens_seen": 2752512000,
      "step": 42000,
      "train_runtime": 27500.5131,
      "train_tokens_per_second": 100089.478
    },
    {
      "epoch": 0.421,
      "grad_norm": 0.16859756410121918,
      "learning_rate": 0.0001895166756347789,
      "loss": 0.3082,
      "num_input_tokens_seen": 2759065600,
      "step": 42100,
      "train_runtime": 27570.8932,
      "train_tokens_per_second": 100071.68
    },
    {
      "epoch": 0.422,
      "grad_norm": 0.13300351798534393,
      "learning_rate": 0.0001890572934257475,
      "loss": 0.3065,
      "num_input_tokens_seen": 2765619200,
      "step": 42200,
      "train_runtime": 27634.6434,
      "train_tokens_per_second": 100077.977
    },
    {
      "epoch": 0.423,
      "grad_norm": 0.14460822939872742,
      "learning_rate": 0.00018859751791020497,
      "loss": 0.3055,
      "num_input_tokens_seen": 2772172800,
      "step": 42300,
      "train_runtime": 27700.3395,
      "train_tokens_per_second": 100077.214
    },
    {
      "epoch": 0.424,
      "grad_norm": 0.1369091421365738,
      "learning_rate": 0.0001881373537180856,
      "loss": 0.3026,
      "num_input_tokens_seen": 2778726400,
      "step": 42400,
      "train_runtime": 27764.0211,
      "train_tokens_per_second": 100083.716
    },
    {
      "epoch": 0.425,
      "grad_norm": 0.15593157708644867,
      "learning_rate": 0.00018767680548323766,
      "loss": 0.3014,
      "num_input_tokens_seen": 2785280000,
      "step": 42500,
      "train_runtime": 27828.3317,
      "train_tokens_per_second": 100087.926
    },
    {
      "epoch": 0.426,
      "grad_norm": 0.18689674139022827,
      "learning_rate": 0.0001872158778433768,
      "loss": 0.3041,
      "num_input_tokens_seen": 2791833600,
      "step": 42600,
      "train_runtime": 27897.9539,
      "train_tokens_per_second": 100073.059
    },
    {
      "epoch": 0.427,
      "grad_norm": 0.1532142609357834,
      "learning_rate": 0.0001867545754400392,
      "loss": 0.3041,
      "num_input_tokens_seen": 2798387200,
      "step": 42700,
      "train_runtime": 27964.2157,
      "train_tokens_per_second": 100070.291
    },
    {
      "epoch": 0.428,
      "grad_norm": 0.12894967198371887,
      "learning_rate": 0.000186292902918535,
      "loss": 0.3047,
      "num_input_tokens_seen": 2804940800,
      "step": 42800,
      "train_runtime": 28028.1798,
      "train_tokens_per_second": 100075.739
    },
    {
      "epoch": 0.429,
      "grad_norm": 0.14526289701461792,
      "learning_rate": 0.00018583086492790136,
      "loss": 0.3097,
      "num_input_tokens_seen": 2811494400,
      "step": 42900,
      "train_runtime": 28093.2724,
      "train_tokens_per_second": 100077.142
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.15546266734600067,
      "learning_rate": 0.00018536846612085566,
      "loss": 0.3066,
      "num_input_tokens_seen": 2818048000,
      "step": 43000,
      "train_runtime": 28157.8145,
      "train_tokens_per_second": 100080.495
    },
    {
      "epoch": 0.431,
      "grad_norm": 0.16307438910007477,
      "learning_rate": 0.00018490571115374878,
      "loss": 0.3073,
      "num_input_tokens_seen": 2824601600,
      "step": 43100,
      "train_runtime": 28227.9591,
      "train_tokens_per_second": 100063.968
    },
    {
      "epoch": 0.432,
      "grad_norm": 0.1360054761171341,
      "learning_rate": 0.00018444260468651816,
      "loss": 0.3013,
      "num_input_tokens_seen": 2831155200,
      "step": 43200,
      "train_runtime": 28291.3921,
      "train_tokens_per_second": 100071.258
    },
    {
      "epoch": 0.433,
      "grad_norm": 0.1404498666524887,
      "learning_rate": 0.00018397915138264068,
      "loss": 0.3066,
      "num_input_tokens_seen": 2837708800,
      "step": 43300,
      "train_runtime": 28355.3195,
      "train_tokens_per_second": 100076.771
    },
    {
      "epoch": 0.434,
      "grad_norm": 0.1926499307155609,
      "learning_rate": 0.00018351535590908606,
      "loss": 0.3012,
      "num_input_tokens_seen": 2844262400,
      "step": 43400,
      "train_runtime": 28420.6726,
      "train_tokens_per_second": 100077.237
    },
    {
      "epoch": 0.435,
      "grad_norm": 0.13713879883289337,
      "learning_rate": 0.00018305122293626948,
      "loss": 0.3029,
      "num_input_tokens_seen": 2850816000,
      "step": 43500,
      "train_runtime": 28490.1826,
      "train_tokens_per_second": 100063.1
    },
    {
      "epoch": 0.436,
      "grad_norm": 0.1541578322649002,
      "learning_rate": 0.00018258675713800492,
      "loss": 0.3061,
      "num_input_tokens_seen": 2857369600,
      "step": 43600,
      "train_runtime": 28555.7903,
      "train_tokens_per_second": 100062.704
    },
    {
      "epoch": 0.437,
      "grad_norm": 0.14117270708084106,
      "learning_rate": 0.00018212196319145773,
      "loss": 0.3053,
      "num_input_tokens_seen": 2863923200,
      "step": 43700,
      "train_runtime": 28622.1811,
      "train_tokens_per_second": 100059.572
    },
    {
      "epoch": 0.438,
      "grad_norm": 0.14943140745162964,
      "learning_rate": 0.00018165684577709778,
      "loss": 0.3043,
      "num_input_tokens_seen": 2870476800,
      "step": 43800,
      "train_runtime": 28686.5648,
      "train_tokens_per_second": 100063.455
    },
    {
      "epoch": 0.439,
      "grad_norm": 0.14043770730495453,
      "learning_rate": 0.0001811914095786524,
      "loss": 0.3048,
      "num_input_tokens_seen": 2877030400,
      "step": 43900,
      "train_runtime": 28751.3532,
      "train_tokens_per_second": 100065.913
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.17811591923236847,
      "learning_rate": 0.0001807256592830588,
      "loss": 0.3088,
      "num_input_tokens_seen": 2883584000,
      "step": 44000,
      "train_runtime": 28815.5193,
      "train_tokens_per_second": 100070.52
    },
    {
      "epoch": 0.441,
      "grad_norm": 0.14588113129138947,
      "learning_rate": 0.00018025959958041732,
      "loss": 0.3017,
      "num_input_tokens_seen": 2890137600,
      "step": 44100,
      "train_runtime": 28880.019,
      "train_tokens_per_second": 100073.951
    },
    {
      "epoch": 0.442,
      "grad_norm": 0.22986213862895966,
      "learning_rate": 0.00017979323516394407,
      "loss": 0.3049,
      "num_input_tokens_seen": 2896691200,
      "step": 44200,
      "train_runtime": 28945.7871,
      "train_tokens_per_second": 100072.981
    },
    {
      "epoch": 0.443,
      "grad_norm": 0.853501558303833,
      "learning_rate": 0.00017932657072992344,
      "loss": 0.3081,
      "num_input_tokens_seen": 2903244800,
      "step": 44300,
      "train_runtime": 29016.3509,
      "train_tokens_per_second": 100055.476
    },
    {
      "epoch": 0.444,
      "grad_norm": 0.15835335850715637,
      "learning_rate": 0.00017885961097766117,
      "loss": 0.3035,
      "num_input_tokens_seen": 2909798400,
      "step": 44400,
      "train_runtime": 29079.9877,
      "train_tokens_per_second": 100061.886
    },
    {
      "epoch": 0.445,
      "grad_norm": 0.25418880581855774,
      "learning_rate": 0.00017839236060943674,
      "loss": 0.3014,
      "num_input_tokens_seen": 2916352000,
      "step": 44500,
      "train_runtime": 29144.3776,
      "train_tokens_per_second": 100065.681
    },
    {
      "epoch": 0.446,
      "grad_norm": 0.14922253787517548,
      "learning_rate": 0.0001779248243304562,
      "loss": 0.3038,
      "num_input_tokens_seen": 2922905600,
      "step": 44600,
      "train_runtime": 29208.2393,
      "train_tokens_per_second": 100071.27
    },
    {
      "epoch": 0.447,
      "grad_norm": 0.14103923738002777,
      "learning_rate": 0.00017745700684880465,
      "loss": 0.3064,
      "num_input_tokens_seen": 2929459200,
      "step": 44700,
      "train_runtime": 29273.1105,
      "train_tokens_per_second": 100073.383
    },
    {
      "epoch": 0.448,
      "grad_norm": 0.15813007950782776,
      "learning_rate": 0.000176988912875399,
      "loss": 0.3049,
      "num_input_tokens_seen": 2936012800,
      "step": 44800,
      "train_runtime": 29342.9224,
      "train_tokens_per_second": 100058.636
    },
    {
      "epoch": 0.449,
      "grad_norm": 0.1471075564622879,
      "learning_rate": 0.00017652054712394028,
      "loss": 0.3029,
      "num_input_tokens_seen": 2942566400,
      "step": 44900,
      "train_runtime": 29408.1792,
      "train_tokens_per_second": 100059.455
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.16910097002983093,
      "learning_rate": 0.0001760519143108665,
      "loss": 0.3026,
      "num_input_tokens_seen": 2949120000,
      "step": 45000,
      "train_runtime": 29472.6802,
      "train_tokens_per_second": 100062.837
    },
    {
      "epoch": 0.451,
      "grad_norm": 0.15087512135505676,
      "learning_rate": 0.00017558301915530483,
      "loss": 0.305,
      "num_input_tokens_seen": 2955673600,
      "step": 45100,
      "train_runtime": 29537.0324,
      "train_tokens_per_second": 100066.708
    },
    {
      "epoch": 0.452,
      "grad_norm": 0.16292531788349152,
      "learning_rate": 0.00017511386637902428,
      "loss": 0.305,
      "num_input_tokens_seen": 2962227200,
      "step": 45200,
      "train_runtime": 29600.4356,
      "train_tokens_per_second": 100073.77
    },
    {
      "epoch": 0.453,
      "grad_norm": 0.14504611492156982,
      "learning_rate": 0.00017464446070638814,
      "loss": 0.3061,
      "num_input_tokens_seen": 2968780800,
      "step": 45300,
      "train_runtime": 29670.2849,
      "train_tokens_per_second": 100059.06
    },
    {
      "epoch": 0.454,
      "grad_norm": 0.14068329334259033,
      "learning_rate": 0.00017417480686430622,
      "loss": 0.3096,
      "num_input_tokens_seen": 2975334400,
      "step": 45400,
      "train_runtime": 29735.31,
      "train_tokens_per_second": 100060.648
    },
    {
      "epoch": 0.455,
      "grad_norm": 0.139748677611351,
      "learning_rate": 0.00017370490958218765,
      "loss": 0.3027,
      "num_input_tokens_seen": 2981888000,
      "step": 45500,
      "train_runtime": 29800.4491,
      "train_tokens_per_second": 100061.848
    },
    {
      "epoch": 0.456,
      "grad_norm": 0.1487821340560913,
      "learning_rate": 0.00017323477359189272,
      "loss": 0.3023,
      "num_input_tokens_seen": 2988441600,
      "step": 45600,
      "train_runtime": 29869.053,
      "train_tokens_per_second": 100051.434
    },
    {
      "epoch": 0.457,
      "grad_norm": 0.15015476942062378,
      "learning_rate": 0.00017276440362768564,
      "loss": 0.3028,
      "num_input_tokens_seen": 2994995200,
      "step": 45700,
      "train_runtime": 29933.644,
      "train_tokens_per_second": 100054.481
    },
    {
      "epoch": 0.458,
      "grad_norm": 0.1298416256904602,
      "learning_rate": 0.0001722938044261868,
      "loss": 0.3058,
      "num_input_tokens_seen": 3001548800,
      "step": 45800,
      "train_runtime": 29997.6813,
      "train_tokens_per_second": 100059.36
    },
    {
      "epoch": 0.459,
      "grad_norm": 0.1956530213356018,
      "learning_rate": 0.0001718229807263249,
      "loss": 0.3033,
      "num_input_tokens_seen": 3008102400,
      "step": 45900,
      "train_runtime": 30067.1877,
      "train_tokens_per_second": 100046.018
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.15267929434776306,
      "learning_rate": 0.0001713519372692894,
      "loss": 0.3028,
      "num_input_tokens_seen": 3014656000,
      "step": 46000,
      "train_runtime": 30131.0143,
      "train_tokens_per_second": 100051.594
    },
    {
      "epoch": 0.461,
      "grad_norm": 0.13846905529499054,
      "learning_rate": 0.0001708806787984826,
      "loss": 0.3036,
      "num_input_tokens_seen": 3021209600,
      "step": 46100,
      "train_runtime": 30195.5066,
      "train_tokens_per_second": 100054.94
    },
    {
      "epoch": 0.462,
      "grad_norm": 0.13704828917980194,
      "learning_rate": 0.00017040921005947212,
      "loss": 0.3094,
      "num_input_tokens_seen": 3027763200,
      "step": 46200,
      "train_runtime": 30260.3523,
      "train_tokens_per_second": 100057.104
    },
    {
      "epoch": 0.463,
      "grad_norm": 0.15288543701171875,
      "learning_rate": 0.0001699375357999429,
      "loss": 0.3014,
      "num_input_tokens_seen": 3034316800,
      "step": 46300,
      "train_runtime": 30325.5675,
      "train_tokens_per_second": 100058.039
    },
    {
      "epoch": 0.464,
      "grad_norm": 0.19963988661766052,
      "learning_rate": 0.0001694656607696496,
      "loss": 0.3061,
      "num_input_tokens_seen": 3040870400,
      "step": 46400,
      "train_runtime": 30399.8434,
      "train_tokens_per_second": 100029.147
    },
    {
      "epoch": 0.465,
      "grad_norm": 0.14533430337905884,
      "learning_rate": 0.0001689935897203684,
      "loss": 0.3056,
      "num_input_tokens_seen": 3047424000,
      "step": 46500,
      "train_runtime": 30464.3563,
      "train_tokens_per_second": 100032.443
    },
    {
      "epoch": 0.466,
      "grad_norm": 0.14005503058433533,
      "learning_rate": 0.0001685213274058496,
      "loss": 0.3016,
      "num_input_tokens_seen": 3053977600,
      "step": 46600,
      "train_runtime": 30528.7292,
      "train_tokens_per_second": 100036.185
    },
    {
      "epoch": 0.467,
      "grad_norm": 0.17612388730049133,
      "learning_rate": 0.00016804887858176944,
      "loss": 0.3006,
      "num_input_tokens_seen": 3060531200,
      "step": 46700,
      "train_runtime": 30592.7142,
      "train_tokens_per_second": 100041.179
    },
    {
      "epoch": 0.468,
      "grad_norm": 0.13526348769664764,
      "learning_rate": 0.00016757624800568238,
      "loss": 0.3001,
      "num_input_tokens_seen": 3067084800,
      "step": 46800,
      "train_runtime": 30656.5144,
      "train_tokens_per_second": 100046.755
    },
    {
      "epoch": 0.469,
      "grad_norm": 0.6205772161483765,
      "learning_rate": 0.00016710344043697301,
      "loss": 0.3016,
      "num_input_tokens_seen": 3073638400,
      "step": 46900,
      "train_runtime": 30727.0215,
      "train_tokens_per_second": 100030.47
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.15328101813793182,
      "learning_rate": 0.0001666304606368083,
      "loss": 0.3049,
      "num_input_tokens_seen": 3080192000,
      "step": 47000,
      "train_runtime": 30792.0203,
      "train_tokens_per_second": 100032.15
    },
    {
      "epoch": 0.471,
      "grad_norm": 0.1804981380701065,
      "learning_rate": 0.00016615731336808962,
      "loss": 0.3008,
      "num_input_tokens_seen": 3086745600,
      "step": 47100,
      "train_runtime": 30856.1119,
      "train_tokens_per_second": 100036.764
    },
    {
      "epoch": 0.472,
      "grad_norm": 0.1460595428943634,
      "learning_rate": 0.0001656840033954047,
      "loss": 0.2996,
      "num_input_tokens_seen": 3093299200,
      "step": 47200,
      "train_runtime": 30922.3293,
      "train_tokens_per_second": 100034.482
    },
    {
      "epoch": 0.473,
      "grad_norm": 0.17493313550949097,
      "learning_rate": 0.00016521053548497973,
      "loss": 0.3005,
      "num_input_tokens_seen": 3099852800,
      "step": 47300,
      "train_runtime": 30985.6891,
      "train_tokens_per_second": 100041.435
    },
    {
      "epoch": 0.474,
      "grad_norm": 0.11990969628095627,
      "learning_rate": 0.0001647369144046313,
      "loss": 0.2995,
      "num_input_tokens_seen": 3106406400,
      "step": 47400,
      "train_runtime": 31056.5152,
      "train_tokens_per_second": 100024.307
    },
    {
      "epoch": 0.475,
      "grad_norm": 0.15634778141975403,
      "learning_rate": 0.00016426314492371842,
      "loss": 0.3054,
      "num_input_tokens_seen": 3112960000,
      "step": 47500,
      "train_runtime": 31121.0302,
      "train_tokens_per_second": 100027.537
    },
    {
      "epoch": 0.476,
      "grad_norm": 0.14218732714653015,
      "learning_rate": 0.0001637892318130945,
      "loss": 0.3036,
      "num_input_tokens_seen": 3119513600,
      "step": 47600,
      "train_runtime": 31185.6411,
      "train_tokens_per_second": 100030.446
    },
    {
      "epoch": 0.477,
      "grad_norm": 0.147688090801239,
      "learning_rate": 0.00016331517984505934,
      "loss": 0.3003,
      "num_input_tokens_seen": 3126067200,
      "step": 47700,
      "train_runtime": 31250.7507,
      "train_tokens_per_second": 100031.748
    },
    {
      "epoch": 0.478,
      "grad_norm": 0.1728331595659256,
      "learning_rate": 0.00016284099379331092,
      "loss": 0.2997,
      "num_input_tokens_seen": 3132620800,
      "step": 47800,
      "train_runtime": 31321.2751,
      "train_tokens_per_second": 100015.749
    },
    {
      "epoch": 0.479,
      "grad_norm": 0.12835726141929626,
      "learning_rate": 0.00016236667843289759,
      "loss": 0.2989,
      "num_input_tokens_seen": 3139174400,
      "step": 47900,
      "train_runtime": 31386.2974,
      "train_tokens_per_second": 100017.353
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.13368946313858032,
      "learning_rate": 0.00016189223854016973,
      "loss": 0.3078,
      "num_input_tokens_seen": 3145728000,
      "step": 48000,
      "train_runtime": 31451.659,
      "train_tokens_per_second": 100017.872
    },
    {
      "epoch": 0.481,
      "grad_norm": 0.12727653980255127,
      "learning_rate": 0.00016141767889273182,
      "loss": 0.3017,
      "num_input_tokens_seen": 3152281600,
      "step": 48100,
      "train_runtime": 31516.5086,
      "train_tokens_per_second": 100020.013
    },
    {
      "epoch": 0.482,
      "grad_norm": 0.16222263872623444,
      "learning_rate": 0.00016094300426939417,
      "loss": 0.3009,
      "num_input_tokens_seen": 3158835200,
      "step": 48200,
      "train_runtime": 31581.3453,
      "train_tokens_per_second": 100022.186
    },
    {
      "epoch": 0.483,
      "grad_norm": 0.15287387371063232,
      "learning_rate": 0.00016046821945012505,
      "loss": 0.2975,
      "num_input_tokens_seen": 3165388800,
      "step": 48300,
      "train_runtime": 31645.8484,
      "train_tokens_per_second": 100025.405
    },
    {
      "epoch": 0.484,
      "grad_norm": 0.13035738468170166,
      "learning_rate": 0.00015999332921600226,
      "loss": 0.3046,
      "num_input_tokens_seen": 3171942400,
      "step": 48400,
      "train_runtime": 31716.5254,
      "train_tokens_per_second": 100009.139
    },
    {
      "epoch": 0.485,
      "grad_norm": 0.16508948802947998,
      "learning_rate": 0.00015951833834916532,
      "loss": 0.3061,
      "num_input_tokens_seen": 3178496000,
      "step": 48500,
      "train_runtime": 31781.7614,
      "train_tokens_per_second": 100010.064
    },
    {
      "epoch": 0.486,
      "grad_norm": 0.1543286293745041,
      "learning_rate": 0.00015904325163276672,
      "loss": 0.2995,
      "num_input_tokens_seen": 3185049600,
      "step": 48600,
      "train_runtime": 31847.2029,
      "train_tokens_per_second": 100010.34
    },
    {
      "epoch": 0.487,
      "grad_norm": 0.13470540940761566,
      "learning_rate": 0.00015856807385092466,
      "loss": 0.3067,
      "num_input_tokens_seen": 3191603200,
      "step": 48700,
      "train_runtime": 31911.0411,
      "train_tokens_per_second": 100015.64
    },
    {
      "epoch": 0.488,
      "grad_norm": 0.15521059930324554,
      "learning_rate": 0.00015809280978867405,
      "loss": 0.3009,
      "num_input_tokens_seen": 3198156800,
      "step": 48800,
      "train_runtime": 31975.3091,
      "train_tokens_per_second": 100019.574
    },
    {
      "epoch": 0.489,
      "grad_norm": 0.16505663096904755,
      "learning_rate": 0.0001576174642319187,
      "loss": 0.3019,
      "num_input_tokens_seen": 3204710400,
      "step": 48900,
      "train_runtime": 32039.3359,
      "train_tokens_per_second": 100024.246
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.15701062977313995,
      "learning_rate": 0.0001571420419673831,
      "loss": 0.3025,
      "num_input_tokens_seen": 3211264000,
      "step": 49000,
      "train_runtime": 32104.9123,
      "train_tokens_per_second": 100024.07
    },
    {
      "epoch": 0.491,
      "grad_norm": 0.22376379370689392,
      "learning_rate": 0.0001566665477825642,
      "loss": 0.3035,
      "num_input_tokens_seen": 3217817600,
      "step": 49100,
      "train_runtime": 32177.5739,
      "train_tokens_per_second": 100001.871
    },
    {
      "epoch": 0.492,
      "grad_norm": 0.1716614067554474,
      "learning_rate": 0.0001561909864656831,
      "loss": 0.3046,
      "num_input_tokens_seen": 3224371200,
      "step": 49200,
      "train_runtime": 32241.8903,
      "train_tokens_per_second": 100005.65
    },
    {
      "epoch": 0.493,
      "grad_norm": 0.17557290196418762,
      "learning_rate": 0.00015571536280563705,
      "loss": 0.2987,
      "num_input_tokens_seen": 3230924800,
      "step": 49300,
      "train_runtime": 32307.4373,
      "train_tokens_per_second": 100005.605
    },
    {
      "epoch": 0.494,
      "grad_norm": 0.16884572803974152,
      "learning_rate": 0.000155239681591951,
      "loss": 0.2986,
      "num_input_tokens_seen": 3237478400,
      "step": 49400,
      "train_runtime": 32371.4412,
      "train_tokens_per_second": 100010.326
    },
    {
      "epoch": 0.495,
      "grad_norm": 0.15279650688171387,
      "learning_rate": 0.00015476394761472953,
      "loss": 0.2982,
      "num_input_tokens_seen": 3244032000,
      "step": 49500,
      "train_runtime": 32436.5241,
      "train_tokens_per_second": 100011.702
    },
    {
      "epoch": 0.496,
      "grad_norm": 0.1866491436958313,
      "learning_rate": 0.00015428816566460843,
      "loss": 0.3038,
      "num_input_tokens_seen": 3250585600,
      "step": 49600,
      "train_runtime": 32508.3167,
      "train_tokens_per_second": 99992.43
    },
    {
      "epoch": 0.497,
      "grad_norm": 0.14084835350513458,
      "learning_rate": 0.00015381234053270669,
      "loss": 0.3027,
      "num_input_tokens_seen": 3257139200,
      "step": 49700,
      "train_runtime": 32572.1194,
      "train_tokens_per_second": 99997.767
    },
    {
      "epoch": 0.498,
      "grad_norm": 0.16111333668231964,
      "learning_rate": 0.0001533364770105781,
      "loss": 0.3015,
      "num_input_tokens_seen": 3263692800,
      "step": 49800,
      "train_runtime": 32637.2501,
      "train_tokens_per_second": 99999.013
    },
    {
      "epoch": 0.499,
      "grad_norm": 0.14655210077762604,
      "learning_rate": 0.0001528605798901631,
      "loss": 0.3012,
      "num_input_tokens_seen": 3270246400,
      "step": 49900,
      "train_runtime": 32707.4201,
      "train_tokens_per_second": 99984.847
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.1385914832353592,
      "learning_rate": 0.00015238465396374027,
      "loss": 0.3027,
      "num_input_tokens_seen": 3276800000,
      "step": 50000,
      "train_runtime": 32772.7798,
      "train_tokens_per_second": 99985.415
    },
    {
      "epoch": 0.501,
      "grad_norm": 0.1433262825012207,
      "learning_rate": 0.00015190870402387858,
      "loss": 0.3006,
      "num_input_tokens_seen": 3283353600,
      "step": 50100,
      "train_runtime": 32837.3412,
      "train_tokens_per_second": 99988.412
    },
    {
      "epoch": 0.502,
      "grad_norm": 0.15529057383537292,
      "learning_rate": 0.00015143273486338857,
      "loss": 0.2995,
      "num_input_tokens_seen": 3289907200,
      "step": 50200,
      "train_runtime": 32902.1033,
      "train_tokens_per_second": 99990.787
    },
    {
      "epoch": 0.503,
      "grad_norm": 0.1301671862602234,
      "learning_rate": 0.00015095675127527438,
      "loss": 0.3055,
      "num_input_tokens_seen": 3296460800,
      "step": 50300,
      "train_runtime": 32967.0743,
      "train_tokens_per_second": 99992.519
    },
    {
      "epoch": 0.504,
      "grad_norm": 0.1454419493675232,
      "learning_rate": 0.00015048075805268547,
      "loss": 0.3036,
      "num_input_tokens_seen": 3303014400,
      "step": 50400,
      "train_runtime": 33033.1243,
      "train_tokens_per_second": 99990.978
    },
    {
      "epoch": 0.505,
      "grad_norm": 0.1473357379436493,
      "learning_rate": 0.00015000475998886825,
      "loss": 0.3018,
      "num_input_tokens_seen": 3309568000,
      "step": 50500,
      "train_runtime": 33105.2406,
      "train_tokens_per_second": 99971.121
    },
    {
      "epoch": 0.506,
      "grad_norm": 0.13996386528015137,
      "learning_rate": 0.00014952876187711804,
      "loss": 0.2974,
      "num_input_tokens_seen": 3316121600,
      "step": 50600,
      "train_runtime": 33169.1198,
      "train_tokens_per_second": 99976.171
    },
    {
      "epoch": 0.507,
      "grad_norm": 0.14000660181045532,
      "learning_rate": 0.00014905276851073053,
      "loss": 0.2992,
      "num_input_tokens_seen": 3322675200,
      "step": 50700,
      "train_runtime": 33234.0005,
      "train_tokens_per_second": 99978.19
    },
    {
      "epoch": 0.508,
      "grad_norm": 0.14661286771297455,
      "learning_rate": 0.00014857678468295352,
      "loss": 0.3045,
      "num_input_tokens_seen": 3329228800,
      "step": 50800,
      "train_runtime": 33299.7758,
      "train_tokens_per_second": 99977.514
    },
    {
      "epoch": 0.509,
      "grad_norm": 0.15111635625362396,
      "learning_rate": 0.00014810081518693902,
      "loss": 0.3006,
      "num_input_tokens_seen": 3335782400,
      "step": 50900,
      "train_runtime": 33370.9097,
      "train_tokens_per_second": 99960.787
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.12965109944343567,
      "learning_rate": 0.0001476248648156945,
      "loss": 0.2986,
      "num_input_tokens_seen": 3342336000,
      "step": 51000,
      "train_runtime": 33435.7602,
      "train_tokens_per_second": 99962.913
    },
    {
      "epoch": 0.511,
      "grad_norm": 0.13791891932487488,
      "learning_rate": 0.00014714893836203485,
      "loss": 0.2994,
      "num_input_tokens_seen": 3348889600,
      "step": 51100,
      "train_runtime": 33500.2878,
      "train_tokens_per_second": 99965.995
    },
    {
      "epoch": 0.512,
      "grad_norm": 0.1420348435640335,
      "learning_rate": 0.0001466730406185343,
      "loss": 0.2996,
      "num_input_tokens_seen": 3355443200,
      "step": 51200,
      "train_runtime": 33564.5521,
      "train_tokens_per_second": 99969.849
    },
    {
      "epoch": 0.513,
      "grad_norm": 0.1938745528459549,
      "learning_rate": 0.0001461971763774778,
      "loss": 0.3007,
      "num_input_tokens_seen": 3361996800,
      "step": 51300,
      "train_runtime": 33630.8004,
      "train_tokens_per_second": 99967.79
    },
    {
      "epoch": 0.514,
      "grad_norm": 0.1449531763792038,
      "learning_rate": 0.0001457213504308129,
      "loss": 0.3011,
      "num_input_tokens_seen": 3368550400,
      "step": 51400,
      "train_runtime": 33696.4447,
      "train_tokens_per_second": 99967.532
    },
    {
      "epoch": 0.515,
      "grad_norm": 0.16473324596881866,
      "learning_rate": 0.00014524556757010177,
      "loss": 0.3005,
      "num_input_tokens_seen": 3375104000,
      "step": 51500,
      "train_runtime": 33766.6492,
      "train_tokens_per_second": 99953.773
    },
    {
      "epoch": 0.516,
      "grad_norm": 0.1542610377073288,
      "learning_rate": 0.00014476983258647234,
      "loss": 0.3012,
      "num_input_tokens_seen": 3381657600,
      "step": 51600,
      "train_runtime": 33832.0917,
      "train_tokens_per_second": 99954.139
    },
    {
      "epoch": 0.517,
      "grad_norm": 0.1388223022222519,
      "learning_rate": 0.0001442941502705707,
      "loss": 0.3031,
      "num_input_tokens_seen": 3388211200,
      "step": 51700,
      "train_runtime": 33896.7212,
      "train_tokens_per_second": 99956.901
    },
    {
      "epoch": 0.518,
      "grad_norm": 0.19452647864818573,
      "learning_rate": 0.0001438185254125125,
      "loss": 0.3011,
      "num_input_tokens_seen": 3394764800,
      "step": 51800,
      "train_runtime": 33962.0557,
      "train_tokens_per_second": 99957.577
    },
    {
      "epoch": 0.519,
      "grad_norm": 0.16043786704540253,
      "learning_rate": 0.00014334296280183473,
      "loss": 0.2997,
      "num_input_tokens_seen": 3401318400,
      "step": 51900,
      "train_runtime": 34027.5551,
      "train_tokens_per_second": 99957.766
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.19769923388957977,
      "learning_rate": 0.00014286746722744768,
      "loss": 0.3007,
      "num_input_tokens_seen": 3407872000,
      "step": 52000,
      "train_runtime": 34098.2307,
      "train_tokens_per_second": 99942.781
    },
    {
      "epoch": 0.521,
      "grad_norm": 0.1524592489004135,
      "learning_rate": 0.00014239204347758647,
      "loss": 0.299,
      "num_input_tokens_seen": 3414425600,
      "step": 52100,
      "train_runtime": 34164.2522,
      "train_tokens_per_second": 99941.47
    },
    {
      "epoch": 0.522,
      "grad_norm": 0.14221727848052979,
      "learning_rate": 0.00014191669633976294,
      "loss": 0.3029,
      "num_input_tokens_seen": 3420979200,
      "step": 52200,
      "train_runtime": 34227.7165,
      "train_tokens_per_second": 99947.632
    },
    {
      "epoch": 0.523,
      "grad_norm": 0.15958262979984283,
      "learning_rate": 0.00014144143060071756,
      "loss": 0.3005,
      "num_input_tokens_seen": 3427532800,
      "step": 52300,
      "train_runtime": 34292.9446,
      "train_tokens_per_second": 99948.629
    },
    {
      "epoch": 0.524,
      "grad_norm": 0.1545192301273346,
      "learning_rate": 0.000140966251046371,
      "loss": 0.3024,
      "num_input_tokens_seen": 3434086400,
      "step": 52400,
      "train_runtime": 34357.5392,
      "train_tokens_per_second": 99951.466
    },
    {
      "epoch": 0.525,
      "grad_norm": 0.14636173844337463,
      "learning_rate": 0.0001404911624617761,
      "loss": 0.2967,
      "num_input_tokens_seen": 3440640000,
      "step": 52500,
      "train_runtime": 34423.9361,
      "train_tokens_per_second": 99949.058
    },
    {
      "epoch": 0.526,
      "grad_norm": 0.26764926314353943,
      "learning_rate": 0.00014001616963106966,
      "loss": 0.2982,
      "num_input_tokens_seen": 3447193600,
      "step": 52600,
      "train_runtime": 34489.4544,
      "train_tokens_per_second": 99949.206
    },
    {
      "epoch": 0.527,
      "grad_norm": 0.20636320114135742,
      "learning_rate": 0.00013954127733742416,
      "loss": 0.3011,
      "num_input_tokens_seen": 3453747200,
      "step": 52700,
      "train_runtime": 34559.9071,
      "train_tokens_per_second": 99935.083
    },
    {
      "epoch": 0.528,
      "grad_norm": 0.1523534059524536,
      "learning_rate": 0.0001390664903629998,
      "loss": 0.3042,
      "num_input_tokens_seen": 3460300800,
      "step": 52800,
      "train_runtime": 34624.4507,
      "train_tokens_per_second": 99938.071
    },
    {
      "epoch": 0.529,
      "grad_norm": 0.15213948488235474,
      "learning_rate": 0.0001385918134888961,
      "loss": 0.3024,
      "num_input_tokens_seen": 3466854400,
      "step": 52900,
      "train_runtime": 34690.2273,
      "train_tokens_per_second": 99937.495
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.14115960896015167,
      "learning_rate": 0.00013811725149510387,
      "loss": 0.2999,
      "num_input_tokens_seen": 3473408000,
      "step": 53000,
      "train_runtime": 34756.5786,
      "train_tokens_per_second": 99935.268
    },
    {
      "epoch": 0.531,
      "grad_norm": 0.16747893393039703,
      "learning_rate": 0.0001376428091604572,
      "loss": 0.3011,
      "num_input_tokens_seen": 3479961600,
      "step": 53100,
      "train_runtime": 34823.0381,
      "train_tokens_per_second": 99932.74
    },
    {
      "epoch": 0.532,
      "grad_norm": 0.1266140639781952,
      "learning_rate": 0.00013716849126258512,
      "loss": 0.2985,
      "num_input_tokens_seen": 3486515200,
      "step": 53200,
      "train_runtime": 34892.7557,
      "train_tokens_per_second": 99920.89
    },
    {
      "epoch": 0.533,
      "grad_norm": 0.14753171801567078,
      "learning_rate": 0.00013669430257786354,
      "loss": 0.2996,
      "num_input_tokens_seen": 3493068800,
      "step": 53300,
      "train_runtime": 34957.0461,
      "train_tokens_per_second": 99924.599
    },
    {
      "epoch": 0.534,
      "grad_norm": 0.2617182731628418,
      "learning_rate": 0.00013622024788136728,
      "loss": 0.3027,
      "num_input_tokens_seen": 3499622400,
      "step": 53400,
      "train_runtime": 35022.8837,
      "train_tokens_per_second": 99923.879
    },
    {
      "epoch": 0.535,
      "grad_norm": 0.17150761187076569,
      "learning_rate": 0.00013574633194682185,
      "loss": 0.3027,
      "num_input_tokens_seen": 3506176000,
      "step": 53500,
      "train_runtime": 35088.2396,
      "train_tokens_per_second": 99924.534
    },
    {
      "epoch": 0.536,
      "grad_norm": 0.16566570103168488,
      "learning_rate": 0.0001352725595465555,
      "loss": 0.2999,
      "num_input_tokens_seen": 3512729600,
      "step": 53600,
      "train_runtime": 35153.6189,
      "train_tokens_per_second": 99925.12
    },
    {
      "epoch": 0.537,
      "grad_norm": 0.13577675819396973,
      "learning_rate": 0.000134798935451451,
      "loss": 0.2969,
      "num_input_tokens_seen": 3519283200,
      "step": 53700,
      "train_runtime": 35225.0068,
      "train_tokens_per_second": 99908.659
    },
    {
      "epoch": 0.538,
      "grad_norm": 0.20843537151813507,
      "learning_rate": 0.00013432546443089768,
      "loss": 0.2967,
      "num_input_tokens_seen": 3525836800,
      "step": 53800,
      "train_runtime": 35288.0858,
      "train_tokens_per_second": 99915.785
    },
    {
      "epoch": 0.539,
      "grad_norm": 0.15664201974868774,
      "learning_rate": 0.0001338521512527436,
      "loss": 0.3007,
      "num_input_tokens_seen": 3532390400,
      "step": 53900,
      "train_runtime": 35353.7477,
      "train_tokens_per_second": 99915.586
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.14205297827720642,
      "learning_rate": 0.00013337900068324712,
      "loss": 0.3001,
      "num_input_tokens_seen": 3538944000,
      "step": 54000,
      "train_runtime": 35423.5891,
      "train_tokens_per_second": 99903.598
    },
    {
      "epoch": 0.541,
      "grad_norm": 0.13229498267173767,
      "learning_rate": 0.00013290601748702918,
      "loss": 0.2931,
      "num_input_tokens_seen": 3545497600,
      "step": 54100,
      "train_runtime": 35489.6646,
      "train_tokens_per_second": 99902.257
    },
    {
      "epoch": 0.542,
      "grad_norm": 0.1380510926246643,
      "learning_rate": 0.00013243320642702543,
      "loss": 0.3116,
      "num_input_tokens_seen": 3552051200,
      "step": 54200,
      "train_runtime": 35554.9224,
      "train_tokens_per_second": 99903.219
    },
    {
      "epoch": 0.543,
      "grad_norm": 0.16735288500785828,
      "learning_rate": 0.0001319605722644379,
      "loss": 0.2998,
      "num_input_tokens_seen": 3558604800,
      "step": 54300,
      "train_runtime": 35619.4728,
      "train_tokens_per_second": 99906.161
    },
    {
      "epoch": 0.544,
      "grad_norm": 0.17502574622631073,
      "learning_rate": 0.0001314881197586874,
      "loss": 0.3004,
      "num_input_tokens_seen": 3565158400,
      "step": 54400,
      "train_runtime": 35685.8161,
      "train_tokens_per_second": 99904.074
    },
    {
      "epoch": 0.545,
      "grad_norm": 0.14805424213409424,
      "learning_rate": 0.0001310158536673654,
      "loss": 0.2983,
      "num_input_tokens_seen": 3571712000,
      "step": 54500,
      "train_runtime": 35750.1467,
      "train_tokens_per_second": 99907.618
    },
    {
      "epoch": 0.546,
      "grad_norm": 0.1533045917749405,
      "learning_rate": 0.0001305437787461862,
      "loss": 0.2976,
      "num_input_tokens_seen": 3578265600,
      "step": 54600,
      "train_runtime": 35816.4973,
      "train_tokens_per_second": 99905.515
    },
    {
      "epoch": 0.547,
      "grad_norm": 0.18475773930549622,
      "learning_rate": 0.00013007189974893903,
      "loss": 0.2951,
      "num_input_tokens_seen": 3584819200,
      "step": 54700,
      "train_runtime": 35886.6478,
      "train_tokens_per_second": 99892.841
    },
    {
      "epoch": 0.548,
      "grad_norm": 0.13913068175315857,
      "learning_rate": 0.00012960022142744016,
      "loss": 0.297,
      "num_input_tokens_seen": 3591372800,
      "step": 54800,
      "train_runtime": 35950.7798,
      "train_tokens_per_second": 99896.937
    },
    {
      "epoch": 0.549,
      "grad_norm": 0.15448203682899475,
      "learning_rate": 0.00012912874853148506,
      "loss": 0.303,
      "num_input_tokens_seen": 3597926400,
      "step": 54900,
      "train_runtime": 36015.8762,
      "train_tokens_per_second": 99898.344
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.15416036546230316,
      "learning_rate": 0.00012865748580880053,
      "loss": 0.2979,
      "num_input_tokens_seen": 3604480000,
      "step": 55000,
      "train_runtime": 36080.201,
      "train_tokens_per_second": 99901.883
    },
    {
      "epoch": 0.551,
      "grad_norm": 0.14506150782108307,
      "learning_rate": 0.0001281864380049969,
      "loss": 0.2983,
      "num_input_tokens_seen": 3611033600,
      "step": 55100,
      "train_runtime": 36150.2521,
      "train_tokens_per_second": 99889.583
    },
    {
      "epoch": 0.552,
      "grad_norm": 0.17357710003852844,
      "learning_rate": 0.00012771560986352042,
      "loss": 0.2986,
      "num_input_tokens_seen": 3617587200,
      "step": 55200,
      "train_runtime": 36215.2659,
      "train_tokens_per_second": 99891.223
    },
    {
      "epoch": 0.553,
      "grad_norm": 0.16711916029453278,
      "learning_rate": 0.0001272450061256052,
      "loss": 0.2979,
      "num_input_tokens_seen": 3624140800,
      "step": 55300,
      "train_runtime": 36279.3222,
      "train_tokens_per_second": 99895.494
    },
    {
      "epoch": 0.554,
      "grad_norm": 0.1502256691455841,
      "learning_rate": 0.00012677463153022565,
      "loss": 0.3007,
      "num_input_tokens_seen": 3630694400,
      "step": 55400,
      "train_runtime": 36345.9552,
      "train_tokens_per_second": 99892.667
    },
    {
      "epoch": 0.555,
      "grad_norm": 0.15480037033557892,
      "learning_rate": 0.0001263044908140488,
      "loss": 0.2975,
      "num_input_tokens_seen": 3637248000,
      "step": 55500,
      "train_runtime": 36415.9598,
      "train_tokens_per_second": 99880.602
    },
    {
      "epoch": 0.556,
      "grad_norm": 0.15693609416484833,
      "learning_rate": 0.00012583458871138632,
      "loss": 0.2978,
      "num_input_tokens_seen": 3643801600,
      "step": 55600,
      "train_runtime": 36480.5541,
      "train_tokens_per_second": 99883.395
    },
    {
      "epoch": 0.557,
      "grad_norm": 0.147445410490036,
      "learning_rate": 0.00012536492995414723,
      "loss": 0.2991,
      "num_input_tokens_seen": 3650355200,
      "step": 55700,
      "train_runtime": 36545.2319,
      "train_tokens_per_second": 99885.95
    },
    {
      "epoch": 0.558,
      "grad_norm": 0.13640980422496796,
      "learning_rate": 0.00012489551927179007,
      "loss": 0.2987,
      "num_input_tokens_seen": 3656908800,
      "step": 55800,
      "train_runtime": 36611.0993,
      "train_tokens_per_second": 99885.25
    },
    {
      "epoch": 0.559,
      "grad_norm": 0.14373840391635895,
      "learning_rate": 0.00012442636139127508,
      "loss": 0.3,
      "num_input_tokens_seen": 3663462400,
      "step": 55900,
      "train_runtime": 36676.4606,
      "train_tokens_per_second": 99885.931
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.14679211378097534,
      "learning_rate": 0.00012395746103701695,
      "loss": 0.2996,
      "num_input_tokens_seen": 3670016000,
      "step": 56000,
      "train_runtime": 36748.2938,
      "train_tokens_per_second": 99869.018
    },
    {
      "epoch": 0.561,
      "grad_norm": 0.15536077320575714,
      "learning_rate": 0.00012348882293083708,
      "loss": 0.2953,
      "num_input_tokens_seen": 3676569600,
      "step": 56100,
      "train_runtime": 36813.4246,
      "train_tokens_per_second": 99870.35
    },
    {
      "epoch": 0.562,
      "grad_norm": 0.16678054630756378,
      "learning_rate": 0.00012302045179191594,
      "loss": 0.2969,
      "num_input_tokens_seen": 3683123200,
      "step": 56200,
      "train_runtime": 36877.8431,
      "train_tokens_per_second": 99873.607
    },
    {
      "epoch": 0.563,
      "grad_norm": 0.15781697630882263,
      "learning_rate": 0.00012255235233674572,
      "loss": 0.2972,
      "num_input_tokens_seen": 3689676800,
      "step": 56300,
      "train_runtime": 36943.2178,
      "train_tokens_per_second": 99874.267
    },
    {
      "epoch": 0.564,
      "grad_norm": 0.13541863858699799,
      "learning_rate": 0.00012208452927908278,
      "loss": 0.302,
      "num_input_tokens_seen": 3696230400,
      "step": 56400,
      "train_runtime": 37008.8029,
      "train_tokens_per_second": 99874.357
    },
    {
      "epoch": 0.565,
      "grad_norm": 0.1400034874677658,
      "learning_rate": 0.00012161698732990003,
      "loss": 0.3,
      "num_input_tokens_seen": 3702784000,
      "step": 56500,
      "train_runtime": 37078.9889,
      "train_tokens_per_second": 99862.054
    },
    {
      "epoch": 0.566,
      "grad_norm": 0.1511828452348709,
      "learning_rate": 0.00012114973119733987,
      "loss": 0.3017,
      "num_input_tokens_seen": 3709337600,
      "step": 56600,
      "train_runtime": 37144.0507,
      "train_tokens_per_second": 99863.573
    },
    {
      "epoch": 0.567,
      "grad_norm": 0.15576902031898499,
      "learning_rate": 0.00012068276558666616,
      "loss": 0.2981,
      "num_input_tokens_seen": 3715891200,
      "step": 56700,
      "train_runtime": 37206.97,
      "train_tokens_per_second": 99870.836
    },
    {
      "epoch": 0.568,
      "grad_norm": 0.24084219336509705,
      "learning_rate": 0.00012021609520021752,
      "loss": 0.3025,
      "num_input_tokens_seen": 3722444800,
      "step": 56800,
      "train_runtime": 37278.1305,
      "train_tokens_per_second": 99855.995
    },
    {
      "epoch": 0.569,
      "grad_norm": 0.16832643747329712,
      "learning_rate": 0.00011974972473735957,
      "loss": 0.301,
      "num_input_tokens_seen": 3728998400,
      "step": 56900,
      "train_runtime": 37343.2452,
      "train_tokens_per_second": 99857.374
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.18326181173324585,
      "learning_rate": 0.00011928365889443764,
      "loss": 0.2987,
      "num_input_tokens_seen": 3735552000,
      "step": 57000,
      "train_runtime": 37407.594,
      "train_tokens_per_second": 99860.793
    },
    {
      "epoch": 0.571,
      "grad_norm": 0.15526984632015228,
      "learning_rate": 0.00011881790236472966,
      "loss": 0.2991,
      "num_input_tokens_seen": 3742105600,
      "step": 57100,
      "train_runtime": 37474.3952,
      "train_tokens_per_second": 99857.665
    },
    {
      "epoch": 0.572,
      "grad_norm": 0.18177416920661926,
      "learning_rate": 0.00011835245983839869,
      "loss": 0.3002,
      "num_input_tokens_seen": 3748659200,
      "step": 57200,
      "train_runtime": 37538.8922,
      "train_tokens_per_second": 99860.677
    },
    {
      "epoch": 0.573,
      "grad_norm": 0.1915498822927475,
      "learning_rate": 0.00011788733600244575,
      "loss": 0.2986,
      "num_input_tokens_seen": 3755212800,
      "step": 57300,
      "train_runtime": 37605.3867,
      "train_tokens_per_second": 99858.375
    },
    {
      "epoch": 0.574,
      "grad_norm": 0.15175184607505798,
      "learning_rate": 0.00011742253554066278,
      "loss": 0.3015,
      "num_input_tokens_seen": 3761766400,
      "step": 57400,
      "train_runtime": 37678.0051,
      "train_tokens_per_second": 99839.851
    },
    {
      "epoch": 0.575,
      "grad_norm": 0.16369026899337769,
      "learning_rate": 0.00011695806313358523,
      "loss": 0.3003,
      "num_input_tokens_seen": 3768320000,
      "step": 57500,
      "train_runtime": 37742.0245,
      "train_tokens_per_second": 99844.141
    },
    {
      "epoch": 0.576,
      "grad_norm": 0.16646848618984222,
      "learning_rate": 0.00011649392345844506,
      "loss": 0.2972,
      "num_input_tokens_seen": 3774873600,
      "step": 57600,
      "train_runtime": 37807.5481,
      "train_tokens_per_second": 99844.444
    },
    {
      "epoch": 0.577,
      "grad_norm": 0.14035099744796753,
      "learning_rate": 0.00011603012118912372,
      "loss": 0.2985,
      "num_input_tokens_seen": 3781427200,
      "step": 57700,
      "train_runtime": 37871.8826,
      "train_tokens_per_second": 99847.88
    },
    {
      "epoch": 0.578,
      "grad_norm": 0.14899714291095734,
      "learning_rate": 0.00011556666099610485,
      "loss": 0.3008,
      "num_input_tokens_seen": 3787980800,
      "step": 57800,
      "train_runtime": 37943.2827,
      "train_tokens_per_second": 99832.712
    },
    {
      "epoch": 0.579,
      "grad_norm": 0.15600667893886566,
      "learning_rate": 0.00011510354754642745,
      "loss": 0.303,
      "num_input_tokens_seen": 3794534400,
      "step": 57900,
      "train_runtime": 38008.9332,
      "train_tokens_per_second": 99832.699
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.1631072610616684,
      "learning_rate": 0.00011464078550363887,
      "loss": 0.2978,
      "num_input_tokens_seen": 3801088000,
      "step": 58000,
      "train_runtime": 38073.7575,
      "train_tokens_per_second": 99834.853
    },
    {
      "epoch": 0.581,
      "grad_norm": 0.1560899019241333,
      "learning_rate": 0.0001141783795277477,
      "loss": 0.299,
      "num_input_tokens_seen": 3807641600,
      "step": 58100,
      "train_runtime": 38139.694,
      "train_tokens_per_second": 99834.089
    },
    {
      "epoch": 0.582,
      "grad_norm": 0.1506076604127884,
      "learning_rate": 0.00011371633427517696,
      "loss": 0.2985,
      "num_input_tokens_seen": 3814195200,
      "step": 58200,
      "train_runtime": 38209.9556,
      "train_tokens_per_second": 99822.026
    },
    {
      "epoch": 0.583,
      "grad_norm": 0.16049940884113312,
      "learning_rate": 0.00011325465439871731,
      "loss": 0.2998,
      "num_input_tokens_seen": 3820748800,
      "step": 58300,
      "train_runtime": 38274.5015,
      "train_tokens_per_second": 99824.913
    },
    {
      "epoch": 0.584,
      "grad_norm": 0.15604519844055176,
      "learning_rate": 0.00011279334454747989,
      "loss": 0.2969,
      "num_input_tokens_seen": 3827302400,
      "step": 58400,
      "train_runtime": 38341.4547,
      "train_tokens_per_second": 99821.523
    },
    {
      "epoch": 0.585,
      "grad_norm": 0.15963351726531982,
      "learning_rate": 0.00011233240936684981,
      "loss": 0.2988,
      "num_input_tokens_seen": 3833856000,
      "step": 58500,
      "train_runtime": 38406.0222,
      "train_tokens_per_second": 99824.345
    },
    {
      "epoch": 0.586,
      "grad_norm": 0.15443411469459534,
      "learning_rate": 0.00011187185349843916,
      "loss": 0.298,
      "num_input_tokens_seen": 3840409600,
      "step": 58600,
      "train_runtime": 38472.0656,
      "train_tokens_per_second": 99823.327
    },
    {
      "epoch": 0.587,
      "grad_norm": 0.15459220111370087,
      "learning_rate": 0.00011141168158004053,
      "loss": 0.3004,
      "num_input_tokens_seen": 3846963200,
      "step": 58700,
      "train_runtime": 38542.0532,
      "train_tokens_per_second": 99812.098
    },
    {
      "epoch": 0.588,
      "grad_norm": 0.16199928522109985,
      "learning_rate": 0.00011095189824557998,
      "loss": 0.2985,
      "num_input_tokens_seen": 3853516800,
      "step": 58800,
      "train_runtime": 38609.4411,
      "train_tokens_per_second": 99807.63
    },
    {
      "epoch": 0.589,
      "grad_norm": 0.2209610939025879,
      "learning_rate": 0.00011049250812507054,
      "loss": 0.3005,
      "num_input_tokens_seen": 3860070400,
      "step": 58900,
      "train_runtime": 38675.4402,
      "train_tokens_per_second": 99806.761
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.22285670042037964,
      "learning_rate": 0.00011003351584456571,
      "loss": 0.298,
      "num_input_tokens_seen": 3866624000,
      "step": 59000,
      "train_runtime": 38740.3065,
      "train_tokens_per_second": 99808.813
    },
    {
      "epoch": 0.591,
      "grad_norm": 0.2148812711238861,
      "learning_rate": 0.0001095749260261126,
      "loss": 0.2966,
      "num_input_tokens_seen": 3873177600,
      "step": 59100,
      "train_runtime": 38806.3344,
      "train_tokens_per_second": 99807.871
    },
    {
      "epoch": 0.592,
      "grad_norm": 0.21284043788909912,
      "learning_rate": 0.00010911674328770559,
      "loss": 0.3009,
      "num_input_tokens_seen": 3879731200,
      "step": 59200,
      "train_runtime": 38871.8466,
      "train_tokens_per_second": 99808.256
    },
    {
      "epoch": 0.593,
      "grad_norm": 0.1655593365430832,
      "learning_rate": 0.00010865897224323979,
      "loss": 0.2981,
      "num_input_tokens_seen": 3886284800,
      "step": 59300,
      "train_runtime": 38937.7196,
      "train_tokens_per_second": 99807.714
    },
    {
      "epoch": 0.594,
      "grad_norm": 0.17153207957744598,
      "learning_rate": 0.00010820161750246453,
      "loss": 0.3042,
      "num_input_tokens_seen": 3892838400,
      "step": 59400,
      "train_runtime": 39004.8582,
      "train_tokens_per_second": 99803.937
    },
    {
      "epoch": 0.595,
      "grad_norm": 0.15362666547298431,
      "learning_rate": 0.00010774468367093696,
      "loss": 0.3001,
      "num_input_tokens_seen": 3899392000,
      "step": 59500,
      "train_runtime": 39068.7475,
      "train_tokens_per_second": 99808.472
    },
    {
      "epoch": 0.596,
      "grad_norm": 0.15481388568878174,
      "learning_rate": 0.00010728817534997573,
      "loss": 0.2973,
      "num_input_tokens_seen": 3905945600,
      "step": 59600,
      "train_runtime": 39137.2916,
      "train_tokens_per_second": 99801.122
    },
    {
      "epoch": 0.597,
      "grad_norm": 0.1292748749256134,
      "learning_rate": 0.00010683209713661453,
      "loss": 0.2993,
      "num_input_tokens_seen": 3912499200,
      "step": 59700,
      "train_runtime": 39198.2818,
      "train_tokens_per_second": 99813.028
    },
    {
      "epoch": 0.598,
      "grad_norm": 0.14853951334953308,
      "learning_rate": 0.00010637645362355589,
      "loss": 0.2967,
      "num_input_tokens_seen": 3919052800,
      "step": 59800,
      "train_runtime": 39262.6162,
      "train_tokens_per_second": 99816.395
    },
    {
      "epoch": 0.599,
      "grad_norm": 0.13745439052581787,
      "learning_rate": 0.00010592124939912497,
      "loss": 0.3023,
      "num_input_tokens_seen": 3925606400,
      "step": 59900,
      "train_runtime": 39328.4755,
      "train_tokens_per_second": 99815.88
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.14352121949195862,
      "learning_rate": 0.00010546648904722326,
      "loss": 0.2973,
      "num_input_tokens_seen": 3932160000,
      "step": 60000,
      "train_runtime": 39393.6967,
      "train_tokens_per_second": 99816.984
    },
    {
      "epoch": 0.601,
      "grad_norm": 0.16375063359737396,
      "learning_rate": 0.0001050121771472824,
      "loss": 0.2934,
      "num_input_tokens_seen": 3938713600,
      "step": 60100,
      "train_runtime": 39465.7876,
      "train_tokens_per_second": 99800.709
    },
    {
      "epoch": 0.602,
      "grad_norm": 0.144679456949234,
      "learning_rate": 0.0001045583182742182,
      "loss": 0.2983,
      "num_input_tokens_seen": 3945267200,
      "step": 60200,
      "train_runtime": 39531.166,
      "train_tokens_per_second": 99801.438
    },
    {
      "epoch": 0.603,
      "grad_norm": 0.33903974294662476,
      "learning_rate": 0.00010410491699838448,
      "loss": 0.2981,
      "num_input_tokens_seen": 3951820800,
      "step": 60300,
      "train_runtime": 39596.8662,
      "train_tokens_per_second": 99801.352
    },
    {
      "epoch": 0.604,
      "grad_norm": 0.1823410987854004,
      "learning_rate": 0.00010365197788552707,
      "loss": 0.2986,
      "num_input_tokens_seen": 3958374400,
      "step": 60400,
      "train_runtime": 39664.1206,
      "train_tokens_per_second": 99797.357
    },
    {
      "epoch": 0.605,
      "grad_norm": 0.18758277595043182,
      "learning_rate": 0.00010319950549673778,
      "loss": 0.2967,
      "num_input_tokens_seen": 3964928000,
      "step": 60500,
      "train_runtime": 39728.4695,
      "train_tokens_per_second": 99800.673
    },
    {
      "epoch": 0.606,
      "grad_norm": 0.173909991979599,
      "learning_rate": 0.00010274750438840855,
      "loss": 0.2981,
      "num_input_tokens_seen": 3971481600,
      "step": 60600,
      "train_runtime": 39794.5098,
      "train_tokens_per_second": 99799.737
    },
    {
      "epoch": 0.607,
      "grad_norm": 0.14504651725292206,
      "learning_rate": 0.00010229597911218554,
      "loss": 0.2967,
      "num_input_tokens_seen": 3978035200,
      "step": 60700,
      "train_runtime": 39864.8024,
      "train_tokens_per_second": 99788.158
    },
    {
      "epoch": 0.608,
      "grad_norm": 0.1418026238679886,
      "learning_rate": 0.00010184493421492324,
      "loss": 0.2976,
      "num_input_tokens_seen": 3984588800,
      "step": 60800,
      "train_runtime": 39931.2064,
      "train_tokens_per_second": 99786.337
    },
    {
      "epoch": 0.609,
      "grad_norm": 0.18415790796279907,
      "learning_rate": 0.0001013943742386388,
      "loss": 0.2997,
      "num_input_tokens_seen": 3991142400,
      "step": 60900,
      "train_runtime": 39996.7127,
      "train_tokens_per_second": 99786.761
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.14107364416122437,
      "learning_rate": 0.00010094430372046616,
      "loss": 0.2979,
      "num_input_tokens_seen": 3997696000,
      "step": 61000,
      "train_runtime": 40068.6157,
      "train_tokens_per_second": 99771.253
    }
  ],
  "logging_steps": 100,
  "max_steps": 100000,
  "num_input_tokens_seen": 3997696000,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 7.5963643723776e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}