{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.3,
  "eval_steps": 500,
  "global_step": 30000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.001,
      "grad_norm": 20.30853843688965,
      "learning_rate": 2.97e-05,
      "loss": 6.3768,
      "num_input_tokens_seen": 6553600,
      "step": 100,
      "train_runtime": 67.7249,
      "train_tokens_per_second": 96767.992
    },
    {
      "epoch": 0.002,
      "grad_norm": 12.960590362548828,
      "learning_rate": 5.97e-05,
      "loss": 1.5751,
      "num_input_tokens_seen": 13107200,
      "step": 200,
      "train_runtime": 121.8001,
      "train_tokens_per_second": 107612.389
    },
    {
      "epoch": 0.003,
      "grad_norm": 4.234757900238037,
      "learning_rate": 8.969999999999998e-05,
      "loss": 1.3419,
      "num_input_tokens_seen": 19660800,
      "step": 300,
      "train_runtime": 175.7236,
      "train_tokens_per_second": 111884.791
    },
    {
      "epoch": 0.004,
      "grad_norm": 2.8189597129821777,
      "learning_rate": 0.0001197,
      "loss": 1.1467,
      "num_input_tokens_seen": 26214400,
      "step": 400,
      "train_runtime": 229.814,
      "train_tokens_per_second": 114067.895
    },
    {
      "epoch": 0.005,
      "grad_norm": 4.593051910400391,
      "learning_rate": 0.00014969999999999998,
      "loss": 1.02,
      "num_input_tokens_seen": 32768000,
      "step": 500,
      "train_runtime": 288.5437,
      "train_tokens_per_second": 113563.389
    },
    {
      "epoch": 0.006,
      "grad_norm": 1.745302677154541,
      "learning_rate": 0.00017969999999999998,
      "loss": 0.9381,
      "num_input_tokens_seen": 39321600,
      "step": 600,
      "train_runtime": 343.6994,
      "train_tokens_per_second": 114406.961
    },
    {
      "epoch": 0.007,
      "grad_norm": 2.180215358734131,
      "learning_rate": 0.00020969999999999997,
      "loss": 0.8912,
      "num_input_tokens_seen": 45875200,
      "step": 700,
      "train_runtime": 398.5625,
      "train_tokens_per_second": 115101.641
    },
    {
      "epoch": 0.008,
      "grad_norm": 1.5782365798950195,
      "learning_rate": 0.0002397,
      "loss": 0.8525,
      "num_input_tokens_seen": 52428800,
      "step": 800,
      "train_runtime": 453.9393,
      "train_tokens_per_second": 115497.369
    },
    {
      "epoch": 0.009,
      "grad_norm": 1.306846022605896,
      "learning_rate": 0.0002697,
      "loss": 0.8411,
      "num_input_tokens_seen": 58982400,
      "step": 900,
      "train_runtime": 508.8336,
      "train_tokens_per_second": 115916.867
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.9485424160957336,
      "learning_rate": 0.00029969999999999997,
      "loss": 0.8207,
      "num_input_tokens_seen": 65536000,
      "step": 1000,
      "train_runtime": 569.5888,
      "train_tokens_per_second": 115058.451
    },
    {
      "epoch": 0.011,
      "grad_norm": 0.8611180186271667,
      "learning_rate": 0.00029999925978027874,
      "loss": 0.8158,
      "num_input_tokens_seen": 72089600,
      "step": 1100,
      "train_runtime": 623.5912,
      "train_tokens_per_second": 115603.946
    },
    {
      "epoch": 0.012,
      "grad_norm": 0.8997637033462524,
      "learning_rate": 0.0002999970091452017,
      "loss": 0.7948,
      "num_input_tokens_seen": 78643200,
      "step": 1200,
      "train_runtime": 679.5203,
      "train_tokens_per_second": 115733.415
    },
    {
      "epoch": 0.013,
      "grad_norm": 0.6279132962226868,
      "learning_rate": 0.00029999324804190795,
      "loss": 0.7852,
      "num_input_tokens_seen": 85196800,
      "step": 1300,
      "train_runtime": 739.7161,
      "train_tokens_per_second": 115174.993
    },
    {
      "epoch": 0.014,
      "grad_norm": 0.5959413051605225,
      "learning_rate": 0.0002999879765082716,
      "loss": 0.7736,
      "num_input_tokens_seen": 91750400,
      "step": 1400,
      "train_runtime": 794.2065,
      "train_tokens_per_second": 115524.612
    },
    {
      "epoch": 0.015,
      "grad_norm": 0.6557334065437317,
      "learning_rate": 0.000299981194597377,
      "loss": 0.7631,
      "num_input_tokens_seen": 98304000,
      "step": 1500,
      "train_runtime": 847.9465,
      "train_tokens_per_second": 115931.844
    },
    {
      "epoch": 0.016,
      "grad_norm": 0.7466850876808167,
      "learning_rate": 0.0002999729023775179,
      "loss": 0.7564,
      "num_input_tokens_seen": 104857600,
      "step": 1600,
      "train_runtime": 907.419,
      "train_tokens_per_second": 115555.874
    },
    {
      "epoch": 0.017,
      "grad_norm": 0.6613496541976929,
      "learning_rate": 0.0002999630999321969,
      "loss": 0.7664,
      "num_input_tokens_seen": 111411200,
      "step": 1700,
      "train_runtime": 957.6918,
      "train_tokens_per_second": 116333.047
    },
    {
      "epoch": 0.018,
      "grad_norm": 0.49886658787727356,
      "learning_rate": 0.00029995178736012443,
      "loss": 0.7577,
      "num_input_tokens_seen": 117964800,
      "step": 1800,
      "train_runtime": 1017.5123,
      "train_tokens_per_second": 115934.516
    },
    {
      "epoch": 0.019,
      "grad_norm": 0.5621941089630127,
      "learning_rate": 0.0002999389647752181,
      "loss": 0.743,
      "num_input_tokens_seen": 124518400,
      "step": 1900,
      "train_runtime": 1070.9552,
      "train_tokens_per_second": 116268.538
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.46181556582450867,
      "learning_rate": 0.00029992463230660104,
      "loss": 0.7429,
      "num_input_tokens_seen": 131072000,
      "step": 2000,
      "train_runtime": 1130.6338,
      "train_tokens_per_second": 115927.894
    },
    {
      "epoch": 0.021,
      "grad_norm": 0.4571639895439148,
      "learning_rate": 0.00029990879009860117,
      "loss": 0.7349,
      "num_input_tokens_seen": 137625600,
      "step": 2100,
      "train_runtime": 1185.0559,
      "train_tokens_per_second": 116134.269
    },
    {
      "epoch": 0.022,
      "grad_norm": 0.5650346875190735,
      "learning_rate": 0.0002998914383107493,
      "loss": 0.7373,
      "num_input_tokens_seen": 144179200,
      "step": 2200,
      "train_runtime": 1239.4558,
      "train_tokens_per_second": 116324.603
    },
    {
      "epoch": 0.023,
      "grad_norm": 0.5640336871147156,
      "learning_rate": 0.0002998725771177778,
      "loss": 0.7383,
      "num_input_tokens_seen": 150732800,
      "step": 2300,
      "train_runtime": 1294.8284,
      "train_tokens_per_second": 116411.407
    },
    {
      "epoch": 0.024,
      "grad_norm": 0.4288141429424286,
      "learning_rate": 0.00029985220670961847,
      "loss": 0.7256,
      "num_input_tokens_seen": 157286400,
      "step": 2400,
      "train_runtime": 1353.9338,
      "train_tokens_per_second": 116169.935
    },
    {
      "epoch": 0.025,
      "grad_norm": 0.7532113194465637,
      "learning_rate": 0.0002998303272914014,
      "loss": 0.7301,
      "num_input_tokens_seen": 163840000,
      "step": 2500,
      "train_runtime": 1408.3918,
      "train_tokens_per_second": 116331.266
    },
    {
      "epoch": 0.026,
      "grad_norm": 0.42602330446243286,
      "learning_rate": 0.00029980693908345185,
      "loss": 0.7238,
      "num_input_tokens_seen": 170393600,
      "step": 2600,
      "train_runtime": 1461.9698,
      "train_tokens_per_second": 116550.699
    },
    {
      "epoch": 0.027,
      "grad_norm": 0.655644953250885,
      "learning_rate": 0.00029978204232128895,
      "loss": 0.7213,
      "num_input_tokens_seen": 176947200,
      "step": 2700,
      "train_runtime": 1523.0525,
      "train_tokens_per_second": 116179.321
    },
    {
      "epoch": 0.028,
      "grad_norm": 0.3962925970554352,
      "learning_rate": 0.0002997556372556227,
      "loss": 0.7238,
      "num_input_tokens_seen": 183500800,
      "step": 2800,
      "train_runtime": 1577.6914,
      "train_tokens_per_second": 116309.692
    },
    {
      "epoch": 0.029,
      "grad_norm": 0.44488221406936646,
      "learning_rate": 0.0002997277241523519,
      "loss": 0.7218,
      "num_input_tokens_seen": 190054400,
      "step": 2900,
      "train_runtime": 1631.4556,
      "train_tokens_per_second": 116493.763
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.3886267840862274,
      "learning_rate": 0.00029969830329256125,
      "loss": 0.7091,
      "num_input_tokens_seen": 196608000,
      "step": 3000,
      "train_runtime": 1686.3399,
      "train_tokens_per_second": 116588.595
    },
    {
      "epoch": 0.031,
      "grad_norm": 0.46353062987327576,
      "learning_rate": 0.00029966737497251836,
      "loss": 0.7132,
      "num_input_tokens_seen": 203161600,
      "step": 3100,
      "train_runtime": 1745.3481,
      "train_tokens_per_second": 116401.765
    },
    {
      "epoch": 0.032,
      "grad_norm": 0.45476260781288147,
      "learning_rate": 0.0002996349395036711,
      "loss": 0.7099,
      "num_input_tokens_seen": 209715200,
      "step": 3200,
      "train_runtime": 1799.8354,
      "train_tokens_per_second": 116519.096
    },
    {
      "epoch": 0.033,
      "grad_norm": 0.5311458706855774,
      "learning_rate": 0.00029960099721264435,
      "loss": 0.7162,
      "num_input_tokens_seen": 216268800,
      "step": 3300,
      "train_runtime": 1854.9316,
      "train_tokens_per_second": 116591.252
    },
    {
      "epoch": 0.034,
      "grad_norm": 0.4546497166156769,
      "learning_rate": 0.0002995655484412365,
      "loss": 0.7046,
      "num_input_tokens_seen": 222822400,
      "step": 3400,
      "train_runtime": 1909.1868,
      "train_tokens_per_second": 116710.63
    },
    {
      "epoch": 0.035,
      "grad_norm": 0.6349391937255859,
      "learning_rate": 0.00029952859354641636,
      "loss": 0.7073,
      "num_input_tokens_seen": 229376000,
      "step": 3500,
      "train_runtime": 1969.0427,
      "train_tokens_per_second": 116491.125
    },
    {
      "epoch": 0.036,
      "grad_norm": 0.493557870388031,
      "learning_rate": 0.00029949013290031924,
      "loss": 0.7021,
      "num_input_tokens_seen": 235929600,
      "step": 3600,
      "train_runtime": 2023.4924,
      "train_tokens_per_second": 116595.251
    },
    {
      "epoch": 0.037,
      "grad_norm": 0.3123897314071655,
      "learning_rate": 0.00029945016689024353,
      "loss": 0.696,
      "num_input_tokens_seen": 242483200,
      "step": 3700,
      "train_runtime": 2077.3277,
      "train_tokens_per_second": 116728.432
    },
    {
      "epoch": 0.038,
      "grad_norm": 0.43099701404571533,
      "learning_rate": 0.0002994086959186464,
      "loss": 0.6997,
      "num_input_tokens_seen": 249036800,
      "step": 3800,
      "train_runtime": 2132.1624,
      "train_tokens_per_second": 116800.109
    },
    {
      "epoch": 0.039,
      "grad_norm": 0.3274790346622467,
      "learning_rate": 0.00029936572040314014,
      "loss": 0.7034,
      "num_input_tokens_seen": 255590400,
      "step": 3900,
      "train_runtime": 2193.0795,
      "train_tokens_per_second": 116544.063
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.33723658323287964,
      "learning_rate": 0.0002993212407764877,
      "loss": 0.6977,
      "num_input_tokens_seen": 262144000,
      "step": 4000,
      "train_runtime": 2246.9274,
      "train_tokens_per_second": 116667.767
    },
    {
      "epoch": 0.041,
      "grad_norm": 0.42106112837791443,
      "learning_rate": 0.00029927525748659834,
      "loss": 0.7078,
      "num_input_tokens_seen": 268697600,
      "step": 4100,
      "train_runtime": 2302.3386,
      "train_tokens_per_second": 116706.38
    },
    {
      "epoch": 0.042,
      "grad_norm": 0.3567107915878296,
      "learning_rate": 0.0002992277709965234,
      "loss": 0.6967,
      "num_input_tokens_seen": 275251200,
      "step": 4200,
      "train_runtime": 2356.6313,
      "train_tokens_per_second": 116798.585
    },
    {
      "epoch": 0.043,
      "grad_norm": 0.4269777834415436,
      "learning_rate": 0.0002991787817844513,
      "loss": 0.6976,
      "num_input_tokens_seen": 281804800,
      "step": 4300,
      "train_runtime": 2417.993,
      "train_tokens_per_second": 116544.92
    },
    {
      "epoch": 0.044,
      "grad_norm": 0.4035900831222534,
      "learning_rate": 0.0002991282903437028,
      "loss": 0.6915,
      "num_input_tokens_seen": 288358400,
      "step": 4400,
      "train_runtime": 2471.1514,
      "train_tokens_per_second": 116689.896
    },
    {
      "epoch": 0.045,
      "grad_norm": 0.3197948634624481,
      "learning_rate": 0.0002990762971827262,
      "loss": 0.6905,
      "num_input_tokens_seen": 294912000,
      "step": 4500,
      "train_runtime": 2525.6119,
      "train_tokens_per_second": 116768.533
    },
    {
      "epoch": 0.046,
      "grad_norm": 0.3118247389793396,
      "learning_rate": 0.00029902280282509197,
      "loss": 0.6959,
      "num_input_tokens_seen": 301465600,
      "step": 4600,
      "train_runtime": 2587.1477,
      "train_tokens_per_second": 116524.311
    },
    {
      "epoch": 0.047,
      "grad_norm": 0.38194310665130615,
      "learning_rate": 0.0002989678078094878,
      "loss": 0.6812,
      "num_input_tokens_seen": 308019200,
      "step": 4700,
      "train_runtime": 2645.2255,
      "train_tokens_per_second": 116443.456
    },
    {
      "epoch": 0.048,
      "grad_norm": 0.5399206876754761,
      "learning_rate": 0.00029891131268971284,
      "loss": 0.6843,
      "num_input_tokens_seen": 314572800,
      "step": 4800,
      "train_runtime": 2702.2839,
      "train_tokens_per_second": 116409.974
    },
    {
      "epoch": 0.049,
      "grad_norm": 0.3325801491737366,
      "learning_rate": 0.0002988533180346723,
      "loss": 0.6803,
      "num_input_tokens_seen": 321126400,
      "step": 4900,
      "train_runtime": 2761.723,
      "train_tokens_per_second": 116277.558
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.41960838437080383,
      "learning_rate": 0.0002987938244283717,
      "loss": 0.6772,
      "num_input_tokens_seen": 327680000,
      "step": 5000,
      "train_runtime": 2818.6212,
      "train_tokens_per_second": 116255.424
    },
    {
      "epoch": 0.051,
      "grad_norm": 0.40982258319854736,
      "learning_rate": 0.00029873283246991105,
      "loss": 0.6861,
      "num_input_tokens_seen": 334233600,
      "step": 5100,
      "train_runtime": 2873.5528,
      "train_tokens_per_second": 116313.715
    },
    {
      "epoch": 0.052,
      "grad_norm": 0.31062355637550354,
      "learning_rate": 0.0002986703427734787,
      "loss": 0.6852,
      "num_input_tokens_seen": 340787200,
      "step": 5200,
      "train_runtime": 2929.1895,
      "train_tokens_per_second": 116341.805
    },
    {
      "epoch": 0.053,
      "grad_norm": 0.3270035684108734,
      "learning_rate": 0.00029860635596834517,
      "loss": 0.6854,
      "num_input_tokens_seen": 347340800,
      "step": 5300,
      "train_runtime": 2983.4815,
      "train_tokens_per_second": 116421.301
    },
    {
      "epoch": 0.054,
      "grad_norm": 0.3153812289237976,
      "learning_rate": 0.0002985408726988569,
      "loss": 0.6826,
      "num_input_tokens_seen": 353894400,
      "step": 5400,
      "train_runtime": 3042.2104,
      "train_tokens_per_second": 116328.049
    },
    {
      "epoch": 0.055,
      "grad_norm": 0.328072190284729,
      "learning_rate": 0.0002984738936244296,
      "loss": 0.6796,
      "num_input_tokens_seen": 360448000,
      "step": 5500,
      "train_runtime": 3096.1295,
      "train_tokens_per_second": 116418.904
    },
    {
      "epoch": 0.056,
      "grad_norm": 0.32828816771507263,
      "learning_rate": 0.0002984054194195419,
      "loss": 0.6927,
      "num_input_tokens_seen": 367001600,
      "step": 5600,
      "train_runtime": 3151.7383,
      "train_tokens_per_second": 116444.184
    },
    {
      "epoch": 0.057,
      "grad_norm": 0.3481093645095825,
      "learning_rate": 0.0002983354507737283,
      "loss": 0.6876,
      "num_input_tokens_seen": 373555200,
      "step": 5700,
      "train_runtime": 3207.4,
      "train_tokens_per_second": 116466.67
    },
    {
      "epoch": 0.058,
      "grad_norm": 0.4054113030433655,
      "learning_rate": 0.00029826398839157215,
      "loss": 0.6868,
      "num_input_tokens_seen": 380108800,
      "step": 5800,
      "train_runtime": 3267.0971,
      "train_tokens_per_second": 116344.508
    },
    {
      "epoch": 0.059,
      "grad_norm": 0.3671651780605316,
      "learning_rate": 0.000298191032992699,
      "loss": 0.6818,
      "num_input_tokens_seen": 386662400,
      "step": 5900,
      "train_runtime": 3320.6147,
      "train_tokens_per_second": 116443.019
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.2815225422382355,
      "learning_rate": 0.0002981165853117688,
      "loss": 0.6735,
      "num_input_tokens_seen": 393216000,
      "step": 6000,
      "train_runtime": 3380.4372,
      "train_tokens_per_second": 116321.049
    },
    {
      "epoch": 0.061,
      "grad_norm": 0.5300637483596802,
      "learning_rate": 0.000298040646098469,
      "loss": 0.6792,
      "num_input_tokens_seen": 399769600,
      "step": 6100,
      "train_runtime": 3434.9665,
      "train_tokens_per_second": 116382.388
    },
    {
      "epoch": 0.062,
      "grad_norm": 0.28088632225990295,
      "learning_rate": 0.0002979632161175064,
      "loss": 0.6763,
      "num_input_tokens_seen": 406323200,
      "step": 6200,
      "train_runtime": 3490.2084,
      "train_tokens_per_second": 116418.034
    },
    {
      "epoch": 0.063,
      "grad_norm": 0.5535532236099243,
      "learning_rate": 0.0002978842961486003,
      "loss": 0.6769,
      "num_input_tokens_seen": 412876800,
      "step": 6300,
      "train_runtime": 3587.6493,
      "train_tokens_per_second": 115082.821
    },
    {
      "epoch": 0.064,
      "grad_norm": 0.3341565728187561,
      "learning_rate": 0.0002978038869864738,
      "loss": 0.6734,
      "num_input_tokens_seen": 419430400,
      "step": 6400,
      "train_runtime": 3650.2458,
      "train_tokens_per_second": 114904.702
    },
    {
      "epoch": 0.065,
      "grad_norm": 0.34299543499946594,
      "learning_rate": 0.0002977219894408463,
      "loss": 0.671,
      "num_input_tokens_seen": 425984000,
      "step": 6500,
      "train_runtime": 3704.075,
      "train_tokens_per_second": 115004.151
    },
    {
      "epoch": 0.066,
      "grad_norm": 0.3468911349773407,
      "learning_rate": 0.0002976386043364251,
      "loss": 0.6799,
      "num_input_tokens_seen": 432537600,
      "step": 6600,
      "train_runtime": 3758.2332,
      "train_tokens_per_second": 115090.677
    },
    {
      "epoch": 0.067,
      "grad_norm": 0.29494985938072205,
      "learning_rate": 0.00029755373251289733,
      "loss": 0.6836,
      "num_input_tokens_seen": 439091200,
      "step": 6700,
      "train_runtime": 3818.8992,
      "train_tokens_per_second": 114978.475
    },
    {
      "epoch": 0.068,
      "grad_norm": 0.43209826946258545,
      "learning_rate": 0.0002974673748249213,
      "loss": 0.6721,
      "num_input_tokens_seen": 445644800,
      "step": 6800,
      "train_runtime": 3873.8852,
      "train_tokens_per_second": 115038.204
    },
    {
      "epoch": 0.069,
      "grad_norm": 0.36356812715530396,
      "learning_rate": 0.00029737953214211804,
      "loss": 0.6742,
      "num_input_tokens_seen": 452198400,
      "step": 6900,
      "train_runtime": 3928.8659,
      "train_tokens_per_second": 115096.42
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.4425353705883026,
      "learning_rate": 0.0002972902053490623,
      "loss": 0.6691,
      "num_input_tokens_seen": 458752000,
      "step": 7000,
      "train_runtime": 3982.4622,
      "train_tokens_per_second": 115193.058
    },
    {
      "epoch": 0.071,
      "grad_norm": 0.3378838300704956,
      "learning_rate": 0.00029719939534527393,
      "loss": 0.6817,
      "num_input_tokens_seen": 465305600,
      "step": 7100,
      "train_runtime": 4038.9423,
      "train_tokens_per_second": 115204.814
    },
    {
      "epoch": 0.072,
      "grad_norm": 0.46528419852256775,
      "learning_rate": 0.00029710710304520866,
      "loss": 0.6731,
      "num_input_tokens_seen": 471859200,
      "step": 7200,
      "train_runtime": 4094.7269,
      "train_tokens_per_second": 115235.817
    },
    {
      "epoch": 0.073,
      "grad_norm": 0.45732709765434265,
      "learning_rate": 0.00029701332937824885,
      "loss": 0.6667,
      "num_input_tokens_seen": 478412800,
      "step": 7300,
      "train_runtime": 4155.1372,
      "train_tokens_per_second": 115137.666
    },
    {
      "epoch": 0.074,
      "grad_norm": 0.28316208720207214,
      "learning_rate": 0.0002969180752886944,
      "loss": 0.6735,
      "num_input_tokens_seen": 484966400,
      "step": 7400,
      "train_runtime": 4208.4074,
      "train_tokens_per_second": 115237.512
    },
    {
      "epoch": 0.075,
      "grad_norm": 0.35296332836151123,
      "learning_rate": 0.0002968213417357529,
      "loss": 0.674,
      "num_input_tokens_seen": 491520000,
      "step": 7500,
      "train_runtime": 4262.7857,
      "train_tokens_per_second": 115304.882
    },
    {
      "epoch": 0.076,
      "grad_norm": 0.37089261412620544,
      "learning_rate": 0.00029672312969353015,
      "loss": 0.6696,
      "num_input_tokens_seen": 498073600,
      "step": 7600,
      "train_runtime": 4324.696,
      "train_tokens_per_second": 115169.622
    },
    {
      "epoch": 0.077,
      "grad_norm": 0.3451579809188843,
      "learning_rate": 0.00029662344015102027,
      "loss": 0.669,
      "num_input_tokens_seen": 504627200,
      "step": 7700,
      "train_runtime": 4380.0469,
      "train_tokens_per_second": 115210.456
    },
    {
      "epoch": 0.078,
      "grad_norm": 0.2890200912952423,
      "learning_rate": 0.00029652227411209594,
      "loss": 0.6672,
      "num_input_tokens_seen": 511180800,
      "step": 7800,
      "train_runtime": 4433.6829,
      "train_tokens_per_second": 115294.849
    },
    {
      "epoch": 0.079,
      "grad_norm": 0.3042987287044525,
      "learning_rate": 0.0002964196325954979,
      "loss": 0.6642,
      "num_input_tokens_seen": 517734400,
      "step": 7900,
      "train_runtime": 4488.6662,
      "train_tokens_per_second": 115342.593
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.34017419815063477,
      "learning_rate": 0.0002963155166348253,
      "loss": 0.6689,
      "num_input_tokens_seen": 524288000,
      "step": 8000,
      "train_runtime": 4548.362,
      "train_tokens_per_second": 115269.63
    },
    {
      "epoch": 0.081,
      "grad_norm": 0.5584019422531128,
      "learning_rate": 0.0002962099272785246,
      "loss": 0.6703,
      "num_input_tokens_seen": 530841600,
      "step": 8100,
      "train_runtime": 4605.2398,
      "train_tokens_per_second": 115269.047
    },
    {
      "epoch": 0.082,
      "grad_norm": 0.2589603364467621,
      "learning_rate": 0.0002961028655898794,
      "loss": 0.6634,
      "num_input_tokens_seen": 537395200,
      "step": 8200,
      "train_runtime": 4660.2455,
      "train_tokens_per_second": 115314.783
    },
    {
      "epoch": 0.083,
      "grad_norm": 0.41600939631462097,
      "learning_rate": 0.0002959943326469998,
      "loss": 0.6611,
      "num_input_tokens_seen": 543948800,
      "step": 8300,
      "train_runtime": 4714.0847,
      "train_tokens_per_second": 115388.0
    },
    {
      "epoch": 0.084,
      "grad_norm": 0.28658175468444824,
      "learning_rate": 0.0002958843295428112,
      "loss": 0.659,
      "num_input_tokens_seen": 550502400,
      "step": 8400,
      "train_runtime": 4768.8365,
      "train_tokens_per_second": 115437.466
    },
    {
      "epoch": 0.085,
      "grad_norm": 0.290019154548645,
      "learning_rate": 0.0002957728573850438,
      "loss": 0.661,
      "num_input_tokens_seen": 557056000,
      "step": 8500,
      "train_runtime": 4828.5019,
      "train_tokens_per_second": 115368.289
    },
    {
      "epoch": 0.086,
      "grad_norm": 0.3445529341697693,
      "learning_rate": 0.0002956599172962209,
      "loss": 0.6745,
      "num_input_tokens_seen": 563609600,
      "step": 8600,
      "train_runtime": 4883.6807,
      "train_tokens_per_second": 115406.727
    },
    {
      "epoch": 0.087,
      "grad_norm": 0.3387536406517029,
      "learning_rate": 0.0002955455104136479,
      "loss": 0.6596,
      "num_input_tokens_seen": 570163200,
      "step": 8700,
      "train_runtime": 4940.0949,
      "train_tokens_per_second": 115415.435
    },
    {
      "epoch": 0.088,
      "grad_norm": 0.3031691312789917,
      "learning_rate": 0.00029542963788940096,
      "loss": 0.658,
      "num_input_tokens_seen": 576716800,
      "step": 8800,
      "train_runtime": 4993.542,
      "train_tokens_per_second": 115492.529
    },
    {
      "epoch": 0.089,
      "grad_norm": 0.3859446048736572,
      "learning_rate": 0.00029531230089031505,
      "loss": 0.6684,
      "num_input_tokens_seen": 583270400,
      "step": 8900,
      "train_runtime": 5048.437,
      "train_tokens_per_second": 115534.848
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.28773999214172363,
      "learning_rate": 0.0002951935005979724,
      "loss": 0.6585,
      "num_input_tokens_seen": 589824000,
      "step": 9000,
      "train_runtime": 5107.8168,
      "train_tokens_per_second": 115474.775
    },
    {
      "epoch": 0.091,
      "grad_norm": 0.3639370799064636,
      "learning_rate": 0.0002950732382086907,
      "loss": 0.6573,
      "num_input_tokens_seen": 596377600,
      "step": 9100,
      "train_runtime": 5164.5839,
      "train_tokens_per_second": 115474.473
    },
    {
      "epoch": 0.092,
      "grad_norm": 0.3325030505657196,
      "learning_rate": 0.0002949515149335108,
      "loss": 0.6634,
      "num_input_tokens_seen": 602931200,
      "step": 9200,
      "train_runtime": 5219.0432,
      "train_tokens_per_second": 115525.237
    },
    {
      "epoch": 0.093,
      "grad_norm": 0.3296424448490143,
      "learning_rate": 0.0002948283319981848,
      "loss": 0.65,
      "num_input_tokens_seen": 609484800,
      "step": 9300,
      "train_runtime": 5273.9916,
      "train_tokens_per_second": 115564.234
    },
    {
      "epoch": 0.094,
      "grad_norm": 0.323024719953537,
      "learning_rate": 0.00029470369064316354,
      "loss": 0.6533,
      "num_input_tokens_seen": 616038400,
      "step": 9400,
      "train_runtime": 5327.8342,
      "train_tokens_per_second": 115626.421
    },
    {
      "epoch": 0.095,
      "grad_norm": 0.31108859181404114,
      "learning_rate": 0.00029457759212358397,
      "loss": 0.6549,
      "num_input_tokens_seen": 622592000,
      "step": 9500,
      "train_runtime": 5386.4463,
      "train_tokens_per_second": 115584.927
    },
    {
      "epoch": 0.096,
      "grad_norm": 0.43114957213401794,
      "learning_rate": 0.00029445003770925686,
      "loss": 0.6518,
      "num_input_tokens_seen": 629145600,
      "step": 9600,
      "train_runtime": 5441.6464,
      "train_tokens_per_second": 115616.773
    },
    {
      "epoch": 0.097,
      "grad_norm": 0.3269326090812683,
      "learning_rate": 0.00029432102868465367,
      "loss": 0.6539,
      "num_input_tokens_seen": 635699200,
      "step": 9700,
      "train_runtime": 5497.0335,
      "train_tokens_per_second": 115644.046
    },
    {
      "epoch": 0.098,
      "grad_norm": 0.3119284510612488,
      "learning_rate": 0.0002941905663488939,
      "loss": 0.6525,
      "num_input_tokens_seen": 642252800,
      "step": 9800,
      "train_runtime": 5557.4204,
      "train_tokens_per_second": 115566.712
    },
    {
      "epoch": 0.099,
      "grad_norm": 0.30968910455703735,
      "learning_rate": 0.0002940586520157318,
      "loss": 0.6593,
      "num_input_tokens_seen": 648806400,
      "step": 9900,
      "train_runtime": 5612.1781,
      "train_tokens_per_second": 115606.881
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.3167921006679535,
      "learning_rate": 0.00029392528701354325,
      "loss": 0.6516,
      "num_input_tokens_seen": 655360000,
      "step": 10000,
      "train_runtime": 5665.7095,
      "train_tokens_per_second": 115671.303
    },
    {
      "epoch": 0.101,
      "grad_norm": 0.3000660538673401,
      "learning_rate": 0.00029379047268531243,
      "loss": 0.6579,
      "num_input_tokens_seen": 661913600,
      "step": 10100,
      "train_runtime": 5721.9414,
      "train_tokens_per_second": 115679.899
    },
    {
      "epoch": 0.102,
      "grad_norm": 0.30189189314842224,
      "learning_rate": 0.00029365421038861795,
      "loss": 0.6594,
      "num_input_tokens_seen": 668467200,
      "step": 10200,
      "train_runtime": 5777.1263,
      "train_tokens_per_second": 115709.294
    },
    {
      "epoch": 0.103,
      "grad_norm": 0.2689358592033386,
      "learning_rate": 0.0002935165014956198,
      "loss": 0.6527,
      "num_input_tokens_seen": 675020800,
      "step": 10300,
      "train_runtime": 5832.7222,
      "train_tokens_per_second": 115729.977
    },
    {
      "epoch": 0.104,
      "grad_norm": 0.40012845396995544,
      "learning_rate": 0.0002933773473930448,
      "loss": 0.6442,
      "num_input_tokens_seen": 681574400,
      "step": 10400,
      "train_runtime": 5888.393,
      "train_tokens_per_second": 115748.796
    },
    {
      "epoch": 0.105,
      "grad_norm": 0.26425933837890625,
      "learning_rate": 0.0002932367494821734,
      "loss": 0.655,
      "num_input_tokens_seen": 688128000,
      "step": 10500,
      "train_runtime": 5949.719,
      "train_tokens_per_second": 115657.227
    },
    {
      "epoch": 0.106,
      "grad_norm": 0.3352152705192566,
      "learning_rate": 0.00029309470917882497,
      "loss": 0.6513,
      "num_input_tokens_seen": 694681600,
      "step": 10600,
      "train_runtime": 6005.1104,
      "train_tokens_per_second": 115681.738
    },
    {
      "epoch": 0.107,
      "grad_norm": 0.2809161841869354,
      "learning_rate": 0.0002929512279133437,
      "loss": 0.6544,
      "num_input_tokens_seen": 701235200,
      "step": 10700,
      "train_runtime": 6060.3088,
      "train_tokens_per_second": 115709.482
    },
    {
      "epoch": 0.108,
      "grad_norm": 0.29120925068855286,
      "learning_rate": 0.0002928063071305844,
      "loss": 0.6502,
      "num_input_tokens_seen": 707788800,
      "step": 10800,
      "train_runtime": 6114.2187,
      "train_tokens_per_second": 115761.119
    },
    {
      "epoch": 0.109,
      "grad_norm": 0.312067449092865,
      "learning_rate": 0.0002926599482898978,
      "loss": 0.6495,
      "num_input_tokens_seen": 714342400,
      "step": 10900,
      "train_runtime": 6169.1634,
      "train_tokens_per_second": 115792.426
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.4553526043891907,
      "learning_rate": 0.00029251215286511573,
      "loss": 0.6508,
      "num_input_tokens_seen": 720896000,
      "step": 11000,
      "train_runtime": 6230.1722,
      "train_tokens_per_second": 115710.445
    },
    {
      "epoch": 0.111,
      "grad_norm": 0.2880023419857025,
      "learning_rate": 0.00029236292234453647,
      "loss": 0.6476,
      "num_input_tokens_seen": 727449600,
      "step": 11100,
      "train_runtime": 6284.648,
      "train_tokens_per_second": 115750.253
    },
    {
      "epoch": 0.112,
      "grad_norm": 0.3704841136932373,
      "learning_rate": 0.0002922122582309097,
      "loss": 0.6559,
      "num_input_tokens_seen": 734003200,
      "step": 11200,
      "train_runtime": 6340.8669,
      "train_tokens_per_second": 115757.547
    },
    {
      "epoch": 0.113,
      "grad_norm": 0.34455180168151855,
      "learning_rate": 0.0002920601620414215,
      "loss": 0.6482,
      "num_input_tokens_seen": 740556800,
      "step": 11300,
      "train_runtime": 6393.5327,
      "train_tokens_per_second": 115829.046
    },
    {
      "epoch": 0.114,
      "grad_norm": 0.3323168456554413,
      "learning_rate": 0.0002919066353076786,
      "loss": 0.6489,
      "num_input_tokens_seen": 747110400,
      "step": 11400,
      "train_runtime": 6449.7175,
      "train_tokens_per_second": 115836.144
    },
    {
      "epoch": 0.115,
      "grad_norm": 0.4480803608894348,
      "learning_rate": 0.00029175167957569366,
      "loss": 0.6492,
      "num_input_tokens_seen": 753664000,
      "step": 11500,
      "train_runtime": 6503.6523,
      "train_tokens_per_second": 115883.194
    },
    {
      "epoch": 0.116,
      "grad_norm": 0.2677996754646301,
      "learning_rate": 0.0002915952964058691,
      "loss": 0.6462,
      "num_input_tokens_seen": 760217600,
      "step": 11600,
      "train_runtime": 6565.0474,
      "train_tokens_per_second": 115797.731
    },
    {
      "epoch": 0.117,
      "grad_norm": 0.2838340699672699,
      "learning_rate": 0.00029143748737298173,
      "loss": 0.6574,
      "num_input_tokens_seen": 766771200,
      "step": 11700,
      "train_runtime": 6620.4342,
      "train_tokens_per_second": 115818.868
    },
    {
      "epoch": 0.118,
      "grad_norm": 0.3221312165260315,
      "learning_rate": 0.00029127825406616677,
      "loss": 0.6456,
      "num_input_tokens_seen": 773324800,
      "step": 11800,
      "train_runtime": 6676.1674,
      "train_tokens_per_second": 115833.645
    },
    {
      "epoch": 0.119,
      "grad_norm": 0.2825019657611847,
      "learning_rate": 0.0002911175980889019,
      "loss": 0.6526,
      "num_input_tokens_seen": 779878400,
      "step": 11900,
      "train_runtime": 6730.5574,
      "train_tokens_per_second": 115871.295
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.38775157928466797,
      "learning_rate": 0.00029095552105899095,
      "loss": 0.6454,
      "num_input_tokens_seen": 786432000,
      "step": 12000,
      "train_runtime": 6784.0533,
      "train_tokens_per_second": 115923.617
    },
    {
      "epoch": 0.121,
      "grad_norm": 0.32168275117874146,
      "learning_rate": 0.0002907920246085478,
      "loss": 0.644,
      "num_input_tokens_seen": 792985600,
      "step": 12100,
      "train_runtime": 6844.3129,
      "train_tokens_per_second": 115860.512
    },
    {
      "epoch": 0.122,
      "grad_norm": 0.3085991442203522,
      "learning_rate": 0.00029062711038397996,
      "loss": 0.6609,
      "num_input_tokens_seen": 799539200,
      "step": 12200,
      "train_runtime": 6899.148,
      "train_tokens_per_second": 115889.557
    },
    {
      "epoch": 0.123,
      "grad_norm": 0.26165974140167236,
      "learning_rate": 0.00029046078004597175,
      "loss": 0.6436,
      "num_input_tokens_seen": 806092800,
      "step": 12300,
      "train_runtime": 6953.2129,
      "train_tokens_per_second": 115930.983
    },
    {
      "epoch": 0.124,
      "grad_norm": 0.5689886212348938,
      "learning_rate": 0.00029029303526946796,
      "loss": 0.6432,
      "num_input_tokens_seen": 812646400,
      "step": 12400,
      "train_runtime": 7008.0879,
      "train_tokens_per_second": 115958.363
    },
    {
      "epoch": 0.125,
      "grad_norm": 0.3152099847793579,
      "learning_rate": 0.0002901238777436565,
      "loss": 0.6391,
      "num_input_tokens_seen": 819200000,
      "step": 12500,
      "train_runtime": 7067.2869,
      "train_tokens_per_second": 115914.355
    },
    {
      "epoch": 0.126,
      "grad_norm": 0.28095683455467224,
      "learning_rate": 0.00028995330917195184,
      "loss": 0.645,
      "num_input_tokens_seen": 825753600,
      "step": 12600,
      "train_runtime": 7122.3035,
      "train_tokens_per_second": 115939.121
    },
    {
      "epoch": 0.127,
      "grad_norm": 0.271088570356369,
      "learning_rate": 0.00028978133127197765,
      "loss": 0.6452,
      "num_input_tokens_seen": 832307200,
      "step": 12700,
      "train_runtime": 7177.0079,
      "train_tokens_per_second": 115968.55
    },
    {
      "epoch": 0.128,
      "grad_norm": 0.3357028067111969,
      "learning_rate": 0.0002896079457755493,
      "loss": 0.6477,
      "num_input_tokens_seen": 838860800,
      "step": 12800,
      "train_runtime": 7232.8071,
      "train_tokens_per_second": 115979.977
    },
    {
      "epoch": 0.129,
      "grad_norm": 0.4945499897003174,
      "learning_rate": 0.000289433154428657,
      "loss": 0.6455,
      "num_input_tokens_seen": 845414400,
      "step": 12900,
      "train_runtime": 7287.7238,
      "train_tokens_per_second": 116005.274
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.43597710132598877,
      "learning_rate": 0.0002892569589914476,
      "loss": 0.6422,
      "num_input_tokens_seen": 851968000,
      "step": 13000,
      "train_runtime": 7347.1365,
      "train_tokens_per_second": 115959.19
    },
    {
      "epoch": 0.131,
      "grad_norm": 0.2706185281276703,
      "learning_rate": 0.0002890793612382072,
      "loss": 0.6434,
      "num_input_tokens_seen": 858521600,
      "step": 13100,
      "train_runtime": 7401.7618,
      "train_tokens_per_second": 115988.818
    },
    {
      "epoch": 0.132,
      "grad_norm": 0.22793404757976532,
      "learning_rate": 0.0002889003629573432,
      "loss": 0.646,
      "num_input_tokens_seen": 865075200,
      "step": 13200,
      "train_runtime": 7457.327,
      "train_tokens_per_second": 116003.389
    },
    {
      "epoch": 0.133,
      "grad_norm": 0.2563342750072479,
      "learning_rate": 0.00028871996595136626,
      "loss": 0.6503,
      "num_input_tokens_seen": 871628800,
      "step": 13300,
      "train_runtime": 7511.6003,
      "train_tokens_per_second": 116037.697
    },
    {
      "epoch": 0.134,
      "grad_norm": 0.30800577998161316,
      "learning_rate": 0.0002885381720368723,
      "loss": 0.6378,
      "num_input_tokens_seen": 878182400,
      "step": 13400,
      "train_runtime": 7571.585,
      "train_tokens_per_second": 115983.959
    },
    {
      "epoch": 0.135,
      "grad_norm": 0.2970588505268097,
      "learning_rate": 0.000288354983044524,
      "loss": 0.6375,
      "num_input_tokens_seen": 884736000,
      "step": 13500,
      "train_runtime": 7625.682,
      "train_tokens_per_second": 116020.573
    },
    {
      "epoch": 0.136,
      "grad_norm": 0.24373579025268555,
      "learning_rate": 0.00028817040081903245,
      "loss": 0.6436,
      "num_input_tokens_seen": 891289600,
      "step": 13600,
      "train_runtime": 7679.3213,
      "train_tokens_per_second": 116063.591
    },
    {
      "epoch": 0.137,
      "grad_norm": 0.6528549194335938,
      "learning_rate": 0.00028798442721913867,
      "loss": 0.639,
      "num_input_tokens_seen": 897843200,
      "step": 13700,
      "train_runtime": 7735.7681,
      "train_tokens_per_second": 116063.872
    },
    {
      "epoch": 0.138,
      "grad_norm": 0.2750374376773834,
      "learning_rate": 0.00028779706411759465,
      "loss": 0.6362,
      "num_input_tokens_seen": 904396800,
      "step": 13800,
      "train_runtime": 7796.3703,
      "train_tokens_per_second": 116002.288
    },
    {
      "epoch": 0.139,
      "grad_norm": 0.3658313751220703,
      "learning_rate": 0.00028760831340114484,
      "loss": 0.6424,
      "num_input_tokens_seen": 910950400,
      "step": 13900,
      "train_runtime": 7851.8089,
      "train_tokens_per_second": 116017.903
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.2838168740272522,
      "learning_rate": 0.00028741817697050683,
      "loss": 0.6424,
      "num_input_tokens_seen": 917504000,
      "step": 14000,
      "train_runtime": 7906.2377,
      "train_tokens_per_second": 116048.117
    },
    {
      "epoch": 0.141,
      "grad_norm": 0.30455613136291504,
      "learning_rate": 0.00028722665674035233,
      "loss": 0.637,
      "num_input_tokens_seen": 924057600,
      "step": 14100,
      "train_runtime": 7961.9367,
      "train_tokens_per_second": 116059.401
    },
    {
      "epoch": 0.142,
      "grad_norm": 0.3287622630596161,
      "learning_rate": 0.0002870337546392879,
      "loss": 0.6612,
      "num_input_tokens_seen": 930611200,
      "step": 14200,
      "train_runtime": 8016.7903,
      "train_tokens_per_second": 116082.767
    },
    {
      "epoch": 0.143,
      "grad_norm": 0.2602579593658447,
      "learning_rate": 0.00028683947260983576,
      "loss": 0.6433,
      "num_input_tokens_seen": 937164800,
      "step": 14300,
      "train_runtime": 8072.3352,
      "train_tokens_per_second": 116095.873
    },
    {
      "epoch": 0.144,
      "grad_norm": 0.4514355957508087,
      "learning_rate": 0.00028664381260841356,
      "loss": 0.6346,
      "num_input_tokens_seen": 943718400,
      "step": 14400,
      "train_runtime": 8133.1299,
      "train_tokens_per_second": 116033.853
    },
    {
      "epoch": 0.145,
      "grad_norm": 0.7613756060600281,
      "learning_rate": 0.0002864467766053154,
      "loss": 0.6383,
      "num_input_tokens_seen": 950272000,
      "step": 14500,
      "train_runtime": 8186.629,
      "train_tokens_per_second": 116076.103
    },
    {
      "epoch": 0.146,
      "grad_norm": 0.28044015169143677,
      "learning_rate": 0.00028624836658469165,
      "loss": 0.6363,
      "num_input_tokens_seen": 956825600,
      "step": 14600,
      "train_runtime": 8246.5064,
      "train_tokens_per_second": 116027.995
    },
    {
      "epoch": 0.147,
      "grad_norm": 0.3139192759990692,
      "learning_rate": 0.00028604858454452906,
      "loss": 0.65,
      "num_input_tokens_seen": 963379200,
      "step": 14700,
      "train_runtime": 8296.4517,
      "train_tokens_per_second": 116119.424
    },
    {
      "epoch": 0.148,
      "grad_norm": 0.33300819993019104,
      "learning_rate": 0.00028584743249663057,
      "loss": 0.6409,
      "num_input_tokens_seen": 969932800,
      "step": 14800,
      "train_runtime": 8351.7667,
      "train_tokens_per_second": 116135.045
    },
    {
      "epoch": 0.149,
      "grad_norm": 0.35637542605400085,
      "learning_rate": 0.000285644912466595,
      "loss": 0.6356,
      "num_input_tokens_seen": 976486400,
      "step": 14900,
      "train_runtime": 8412.3756,
      "train_tokens_per_second": 116077.366
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.2518846392631531,
      "learning_rate": 0.00028544102649379684,
      "loss": 0.6435,
      "num_input_tokens_seen": 983040000,
      "step": 15000,
      "train_runtime": 8466.8259,
      "train_tokens_per_second": 116104.904
    },
    {
      "epoch": 0.151,
      "grad_norm": 0.2983591854572296,
      "learning_rate": 0.00028523577663136556,
      "loss": 0.638,
      "num_input_tokens_seen": 989593600,
      "step": 15100,
      "train_runtime": 8526.2848,
      "train_tokens_per_second": 116063.869
    },
    {
      "epoch": 0.152,
      "grad_norm": 0.26403748989105225,
      "learning_rate": 0.000285029164946165,
      "loss": 0.6439,
      "num_input_tokens_seen": 996147200,
      "step": 15200,
      "train_runtime": 8581.358,
      "train_tokens_per_second": 116082.7
    },
    {
      "epoch": 0.153,
      "grad_norm": 0.3652186393737793,
      "learning_rate": 0.0002848211935187725,
      "loss": 0.6499,
      "num_input_tokens_seen": 1002700800,
      "step": 15300,
      "train_runtime": 8637.0354,
      "train_tokens_per_second": 116093.167
    },
    {
      "epoch": 0.154,
      "grad_norm": 0.3816509246826172,
      "learning_rate": 0.0002846118644434581,
      "loss": 0.6355,
      "num_input_tokens_seen": 1009254400,
      "step": 15400,
      "train_runtime": 8691.2267,
      "train_tokens_per_second": 116123.355
    },
    {
      "epoch": 0.155,
      "grad_norm": 0.35392388701438904,
      "learning_rate": 0.00028440117982816326,
      "loss": 0.6286,
      "num_input_tokens_seen": 1015808000,
      "step": 15500,
      "train_runtime": 8745.8114,
      "train_tokens_per_second": 116147.943
    },
    {
      "epoch": 0.156,
      "grad_norm": 0.34099990129470825,
      "learning_rate": 0.0002841891417944796,
      "loss": 0.6396,
      "num_input_tokens_seen": 1022361600,
      "step": 15600,
      "train_runtime": 8805.4075,
      "train_tokens_per_second": 116106.108
    },
    {
      "epoch": 0.157,
      "grad_norm": 0.2872321605682373,
      "learning_rate": 0.0002839757524776279,
      "loss": 0.6436,
      "num_input_tokens_seen": 1028915200,
      "step": 15700,
      "train_runtime": 8860.6468,
      "train_tokens_per_second": 116121.907
    },
    {
      "epoch": 0.158,
      "grad_norm": 0.2647675573825836,
      "learning_rate": 0.0002837610140264361,
      "loss": 0.636,
      "num_input_tokens_seen": 1035468800,
      "step": 15800,
      "train_runtime": 8915.1622,
      "train_tokens_per_second": 116146.94
    },
    {
      "epoch": 0.159,
      "grad_norm": 0.28176337480545044,
      "learning_rate": 0.0002835449286033182,
      "loss": 0.6322,
      "num_input_tokens_seen": 1042022400,
      "step": 15900,
      "train_runtime": 8975.5504,
      "train_tokens_per_second": 116095.655
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.31414464116096497,
      "learning_rate": 0.0002833274983842518,
      "loss": 0.6282,
      "num_input_tokens_seen": 1048576000,
      "step": 16000,
      "train_runtime": 9029.8157,
      "train_tokens_per_second": 116123.744
    },
    {
      "epoch": 0.161,
      "grad_norm": 0.3065328598022461,
      "learning_rate": 0.0002831087255587569,
      "loss": 0.6333,
      "num_input_tokens_seen": 1055129600,
      "step": 16100,
      "train_runtime": 9084.3619,
      "train_tokens_per_second": 116147.905
    },
    {
      "epoch": 0.162,
      "grad_norm": 0.28930503129959106,
      "learning_rate": 0.0002828886123298734,
      "loss": 0.6333,
      "num_input_tokens_seen": 1061683200,
      "step": 16200,
      "train_runtime": 9139.4895,
      "train_tokens_per_second": 116164.388
    },
    {
      "epoch": 0.163,
      "grad_norm": 0.28349098563194275,
      "learning_rate": 0.00028266716091413906,
      "loss": 0.6368,
      "num_input_tokens_seen": 1068236800,
      "step": 16300,
      "train_runtime": 9194.7322,
      "train_tokens_per_second": 116179.219
    },
    {
      "epoch": 0.164,
      "grad_norm": 0.2510078251361847,
      "learning_rate": 0.0002824443735415673,
      "loss": 0.6348,
      "num_input_tokens_seen": 1074790400,
      "step": 16400,
      "train_runtime": 9254.9047,
      "train_tokens_per_second": 116131.979
    },
    {
      "epoch": 0.165,
      "grad_norm": 0.35605230927467346,
      "learning_rate": 0.0002822202524556243,
      "loss": 0.639,
      "num_input_tokens_seen": 1081344000,
      "step": 16500,
      "train_runtime": 9310.1385,
      "train_tokens_per_second": 116146.929
    },
    {
      "epoch": 0.166,
      "grad_norm": 0.518841564655304,
      "learning_rate": 0.00028199479991320695,
      "loss": 0.6424,
      "num_input_tokens_seen": 1087897600,
      "step": 16600,
      "train_runtime": 9365.0545,
      "train_tokens_per_second": 116165.645
    },
    {
      "epoch": 0.167,
      "grad_norm": 0.2293197512626648,
      "learning_rate": 0.00028176801818461994,
      "loss": 0.6316,
      "num_input_tokens_seen": 1094451200,
      "step": 16700,
      "train_runtime": 9419.4884,
      "train_tokens_per_second": 116190.09
    },
    {
      "epoch": 0.168,
      "grad_norm": 0.2790850102901459,
      "learning_rate": 0.00028153990955355273,
      "loss": 0.636,
      "num_input_tokens_seen": 1101004800,
      "step": 16800,
      "train_runtime": 9476.046,
      "train_tokens_per_second": 116188.207
    },
    {
      "epoch": 0.169,
      "grad_norm": 0.2565983235836029,
      "learning_rate": 0.00028131047631705665,
      "loss": 0.6351,
      "num_input_tokens_seen": 1107558400,
      "step": 16900,
      "train_runtime": 9536.06,
      "train_tokens_per_second": 116144.236
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.2592833638191223,
      "learning_rate": 0.00028107972078552187,
      "loss": 0.637,
      "num_input_tokens_seen": 1114112000,
      "step": 17000,
      "train_runtime": 9591.4651,
      "train_tokens_per_second": 116156.603
    },
    {
      "epoch": 0.171,
      "grad_norm": 0.2744060754776001,
      "learning_rate": 0.0002808476452826541,
      "loss": 0.6313,
      "num_input_tokens_seen": 1120665600,
      "step": 17100,
      "train_runtime": 9644.9448,
      "train_tokens_per_second": 116192.018
    },
    {
      "epoch": 0.172,
      "grad_norm": 0.2991725504398346,
      "learning_rate": 0.00028061425214545094,
      "loss": 0.6298,
      "num_input_tokens_seen": 1127219200,
      "step": 17200,
      "train_runtime": 9706.1229,
      "train_tokens_per_second": 116134.857
    },
    {
      "epoch": 0.173,
      "grad_norm": 0.3391658365726471,
      "learning_rate": 0.00028037954372417883,
      "loss": 0.6316,
      "num_input_tokens_seen": 1133772800,
      "step": 17300,
      "train_runtime": 9760.541,
      "train_tokens_per_second": 116158.807
    },
    {
      "epoch": 0.174,
      "grad_norm": 0.3975388705730438,
      "learning_rate": 0.0002801435223823488,
      "loss": 0.639,
      "num_input_tokens_seen": 1140326400,
      "step": 17400,
      "train_runtime": 9816.1595,
      "train_tokens_per_second": 116168.284
    },
    {
      "epoch": 0.175,
      "grad_norm": 0.3754902482032776,
      "learning_rate": 0.00027990619049669336,
      "loss": 0.6391,
      "num_input_tokens_seen": 1146880000,
      "step": 17500,
      "train_runtime": 9870.38,
      "train_tokens_per_second": 116194.108
    },
    {
      "epoch": 0.176,
      "grad_norm": 0.2742336094379425,
      "learning_rate": 0.00027966755045714177,
      "loss": 0.6425,
      "num_input_tokens_seen": 1153433600,
      "step": 17600,
      "train_runtime": 9924.9127,
      "train_tokens_per_second": 116215.994
    },
    {
      "epoch": 0.177,
      "grad_norm": 0.31107792258262634,
      "learning_rate": 0.00027942760466679673,
      "loss": 0.6312,
      "num_input_tokens_seen": 1159987200,
      "step": 17700,
      "train_runtime": 9985.7287,
      "train_tokens_per_second": 116164.502
    },
    {
      "epoch": 0.178,
      "grad_norm": 0.2781549096107483,
      "learning_rate": 0.00027918635554190956,
      "loss": 0.6447,
      "num_input_tokens_seen": 1166540800,
      "step": 17800,
      "train_runtime": 10039.8303,
      "train_tokens_per_second": 116191.287
    },
    {
      "epoch": 0.179,
      "grad_norm": 0.28636643290519714,
      "learning_rate": 0.00027894380551185636,
      "loss": 0.6383,
      "num_input_tokens_seen": 1173094400,
      "step": 17900,
      "train_runtime": 10096.4761,
      "train_tokens_per_second": 116188.498
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.25008583068847656,
      "learning_rate": 0.00027869995701911314,
      "loss": 0.6286,
      "num_input_tokens_seen": 1179648000,
      "step": 18000,
      "train_runtime": 10150.5104,
      "train_tokens_per_second": 116215.634
    },
    {
      "epoch": 0.181,
      "grad_norm": 0.2600831985473633,
      "learning_rate": 0.0002784548125192316,
      "loss": 0.6265,
      "num_input_tokens_seen": 1186201600,
      "step": 18100,
      "train_runtime": 10205.6407,
      "train_tokens_per_second": 116229.998
    },
    {
      "epoch": 0.182,
      "grad_norm": 0.2586776614189148,
      "learning_rate": 0.0002782083744808141,
      "loss": 0.6292,
      "num_input_tokens_seen": 1192755200,
      "step": 18200,
      "train_runtime": 10267.3133,
      "train_tokens_per_second": 116170.138
    },
    {
      "epoch": 0.183,
      "grad_norm": 0.6283107399940491,
      "learning_rate": 0.000277960645385489,
      "loss": 0.6364,
      "num_input_tokens_seen": 1199308800,
      "step": 18300,
      "train_runtime": 10323.5921,
      "train_tokens_per_second": 116171.657
    },
    {
      "epoch": 0.184,
      "grad_norm": 0.2882465124130249,
      "learning_rate": 0.00027771162772788544,
      "loss": 0.6318,
      "num_input_tokens_seen": 1205862400,
      "step": 18400,
      "train_runtime": 10377.5746,
      "train_tokens_per_second": 116198.866
    },
    {
      "epoch": 0.185,
      "grad_norm": 0.3091796934604645,
      "learning_rate": 0.00027746132401560857,
      "loss": 0.6269,
      "num_input_tokens_seen": 1212416000,
      "step": 18500,
      "train_runtime": 10432.3665,
      "train_tokens_per_second": 116216.776
    },
    {
      "epoch": 0.186,
      "grad_norm": 0.280862957239151,
      "learning_rate": 0.0002772097367692139,
      "loss": 0.6323,
      "num_input_tokens_seen": 1218969600,
      "step": 18600,
      "train_runtime": 10487.2106,
      "train_tokens_per_second": 116233.921
    },
    {
      "epoch": 0.187,
      "grad_norm": 0.24253763258457184,
      "learning_rate": 0.00027695686852218226,
      "loss": 0.6374,
      "num_input_tokens_seen": 1225523200,
      "step": 18700,
      "train_runtime": 10547.3913,
      "train_tokens_per_second": 116192.067
    },
    {
      "epoch": 0.188,
      "grad_norm": 0.26167231798171997,
      "learning_rate": 0.00027670272182089416,
      "loss": 0.6357,
      "num_input_tokens_seen": 1232076800,
      "step": 18800,
      "train_runtime": 10602.8292,
      "train_tokens_per_second": 116202.645
    },
    {
      "epoch": 0.189,
      "grad_norm": 0.31137940287590027,
      "learning_rate": 0.0002764472992246039,
      "loss": 0.6305,
      "num_input_tokens_seen": 1238630400,
      "step": 18900,
      "train_runtime": 10657.9196,
      "train_tokens_per_second": 116216.903
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.2601984441280365,
      "learning_rate": 0.0002761906033054143,
      "loss": 0.6301,
      "num_input_tokens_seen": 1245184000,
      "step": 19000,
      "train_runtime": 10712.1812,
      "train_tokens_per_second": 116240.005
    },
    {
      "epoch": 0.191,
      "grad_norm": 0.42287951707839966,
      "learning_rate": 0.00027593263664825045,
      "loss": 0.6324,
      "num_input_tokens_seen": 1251737600,
      "step": 19100,
      "train_runtime": 10766.7423,
      "train_tokens_per_second": 116259.642
    },
    {
      "epoch": 0.192,
      "grad_norm": 0.28982725739479065,
      "learning_rate": 0.00027567340185083363,
      "loss": 0.6291,
      "num_input_tokens_seen": 1258291200,
      "step": 19200,
      "train_runtime": 10826.785,
      "train_tokens_per_second": 116220.208
    },
    {
      "epoch": 0.193,
      "grad_norm": 0.2896488308906555,
      "learning_rate": 0.00027541290152365537,
      "loss": 0.6354,
      "num_input_tokens_seen": 1264844800,
      "step": 19300,
      "train_runtime": 10881.8628,
      "train_tokens_per_second": 116234.216
    },
    {
      "epoch": 0.194,
      "grad_norm": 0.27890026569366455,
      "learning_rate": 0.00027515113828995117,
      "loss": 0.6339,
      "num_input_tokens_seen": 1271398400,
      "step": 19400,
      "train_runtime": 10937.449,
      "train_tokens_per_second": 116242.682
    },
    {
      "epoch": 0.195,
      "grad_norm": 0.2991676926612854,
      "learning_rate": 0.00027488811478567374,
      "loss": 0.6282,
      "num_input_tokens_seen": 1277952000,
      "step": 19500,
      "train_runtime": 10991.5778,
      "train_tokens_per_second": 116266.475
    },
    {
      "epoch": 0.196,
      "grad_norm": 0.2899467349052429,
      "learning_rate": 0.0002746238336594671,
      "loss": 0.6264,
      "num_input_tokens_seen": 1284505600,
      "step": 19600,
      "train_runtime": 11050.8685,
      "train_tokens_per_second": 116235.714
    },
    {
      "epoch": 0.197,
      "grad_norm": 0.32908034324645996,
      "learning_rate": 0.00027435829757263894,
      "loss": 0.6321,
      "num_input_tokens_seen": 1291059200,
      "step": 19700,
      "train_runtime": 11105.6721,
      "train_tokens_per_second": 116252.234
    },
    {
      "epoch": 0.198,
      "grad_norm": 0.23919178545475006,
      "learning_rate": 0.0002740915091991349,
      "loss": 0.6342,
      "num_input_tokens_seen": 1297612800,
      "step": 19800,
      "train_runtime": 11160.1183,
      "train_tokens_per_second": 116272.316
    },
    {
      "epoch": 0.199,
      "grad_norm": 0.5094599723815918,
      "learning_rate": 0.0002738234712255109,
      "loss": 0.6317,
      "num_input_tokens_seen": 1304166400,
      "step": 19900,
      "train_runtime": 11221.0166,
      "train_tokens_per_second": 116225.333
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.37415555119514465,
      "learning_rate": 0.00027355418635090635,
      "loss": 0.634,
      "num_input_tokens_seen": 1310720000,
      "step": 20000,
      "train_runtime": 11276.9767,
      "train_tokens_per_second": 116229.734
    },
    {
      "epoch": 0.201,
      "grad_norm": 0.39444148540496826,
      "learning_rate": 0.000273283657287017,
      "loss": 0.6272,
      "num_input_tokens_seen": 1317273600,
      "step": 20100,
      "train_runtime": 11332.3114,
      "train_tokens_per_second": 116240.505
    },
    {
      "epoch": 0.202,
      "grad_norm": 0.2751108705997467,
      "learning_rate": 0.00027301188675806745,
      "loss": 0.6385,
      "num_input_tokens_seen": 1323827200,
      "step": 20200,
      "train_runtime": 11387.621,
      "train_tokens_per_second": 116251.427
    },
    {
      "epoch": 0.203,
      "grad_norm": 0.2579997479915619,
      "learning_rate": 0.0002727388775007839,
      "loss": 0.6278,
      "num_input_tokens_seen": 1330380800,
      "step": 20300,
      "train_runtime": 11441.506,
      "train_tokens_per_second": 116276.721
    },
    {
      "epoch": 0.204,
      "grad_norm": 0.27806708216667175,
      "learning_rate": 0.0002724646322643666,
      "loss": 0.6292,
      "num_input_tokens_seen": 1336934400,
      "step": 20400,
      "train_runtime": 11496.0966,
      "train_tokens_per_second": 116294.638
    },
    {
      "epoch": 0.205,
      "grad_norm": 0.48101556301116943,
      "learning_rate": 0.000272189153810462,
      "loss": 0.6335,
      "num_input_tokens_seen": 1343488000,
      "step": 20500,
      "train_runtime": 11557.1454,
      "train_tokens_per_second": 116247.391
    },
    {
      "epoch": 0.206,
      "grad_norm": 0.24878458678722382,
      "learning_rate": 0.0002719124449131351,
      "loss": 0.6309,
      "num_input_tokens_seen": 1350041600,
      "step": 20600,
      "train_runtime": 11613.0753,
      "train_tokens_per_second": 116251.86
    },
    {
      "epoch": 0.207,
      "grad_norm": 0.2999299466609955,
      "learning_rate": 0.00027163450835884144,
      "loss": 0.627,
      "num_input_tokens_seen": 1356595200,
      "step": 20700,
      "train_runtime": 11667.4657,
      "train_tokens_per_second": 116271.625
    },
    {
      "epoch": 0.208,
      "grad_norm": 0.2567562162876129,
      "learning_rate": 0.00027135534694639894,
      "loss": 0.6326,
      "num_input_tokens_seen": 1363148800,
      "step": 20800,
      "train_runtime": 11723.8934,
      "train_tokens_per_second": 116271.0
    },
    {
      "epoch": 0.209,
      "grad_norm": 0.3484431207180023,
      "learning_rate": 0.00027107496348696003,
      "loss": 0.6356,
      "num_input_tokens_seen": 1369702400,
      "step": 20900,
      "train_runtime": 11777.6307,
      "train_tokens_per_second": 116296.939
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.23648300766944885,
      "learning_rate": 0.00027079336080398296,
      "loss": 0.6256,
      "num_input_tokens_seen": 1376256000,
      "step": 21000,
      "train_runtime": 11838.2167,
      "train_tokens_per_second": 116255.348
    },
    {
      "epoch": 0.211,
      "grad_norm": 0.33549532294273376,
      "learning_rate": 0.00027051054173320366,
      "loss": 0.6271,
      "num_input_tokens_seen": 1382809600,
      "step": 21100,
      "train_runtime": 11893.4157,
      "train_tokens_per_second": 116266.819
    },
    {
      "epoch": 0.212,
      "grad_norm": 0.2517142593860626,
      "learning_rate": 0.000270226509122607,
      "loss": 0.6254,
      "num_input_tokens_seen": 1389363200,
      "step": 21200,
      "train_runtime": 11947.4932,
      "train_tokens_per_second": 116289.098
    },
    {
      "epoch": 0.213,
      "grad_norm": 0.2526894807815552,
      "learning_rate": 0.0002699412658323983,
      "loss": 0.633,
      "num_input_tokens_seen": 1395916800,
      "step": 21300,
      "train_runtime": 12002.0117,
      "train_tokens_per_second": 116306.902
    },
    {
      "epoch": 0.214,
      "grad_norm": 0.283974289894104,
      "learning_rate": 0.00026965481473497423,
      "loss": 0.6273,
      "num_input_tokens_seen": 1402470400,
      "step": 21400,
      "train_runtime": 12063.5743,
      "train_tokens_per_second": 116256.622
    },
    {
      "epoch": 0.215,
      "grad_norm": 0.24663467705249786,
      "learning_rate": 0.0002693671587148942,
      "loss": 0.6236,
      "num_input_tokens_seen": 1409024000,
      "step": 21500,
      "train_runtime": 12118.0936,
      "train_tokens_per_second": 116274.395
    },
    {
      "epoch": 0.216,
      "grad_norm": 0.30240553617477417,
      "learning_rate": 0.0002690783006688511,
      "loss": 0.6272,
      "num_input_tokens_seen": 1415577600,
      "step": 21600,
      "train_runtime": 12174.4024,
      "train_tokens_per_second": 116274.915
    },
    {
      "epoch": 0.217,
      "grad_norm": 0.2617557644844055,
      "learning_rate": 0.0002687882435056423,
      "loss": 0.6256,
      "num_input_tokens_seen": 1422131200,
      "step": 21700,
      "train_runtime": 12229.4977,
      "train_tokens_per_second": 116286.967
    },
    {
      "epoch": 0.218,
      "grad_norm": 0.3469904661178589,
      "learning_rate": 0.0002684969901461402,
      "loss": 0.634,
      "num_input_tokens_seen": 1428684800,
      "step": 21800,
      "train_runtime": 12284.4989,
      "train_tokens_per_second": 116299.803
    },
    {
      "epoch": 0.219,
      "grad_norm": 0.32695531845092773,
      "learning_rate": 0.000268204543523263,
      "loss": 0.6343,
      "num_input_tokens_seen": 1435238400,
      "step": 21900,
      "train_runtime": 12340.2707,
      "train_tokens_per_second": 116305.261
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.2909862995147705,
      "learning_rate": 0.0002679109065819447,
      "loss": 0.6275,
      "num_input_tokens_seen": 1441792000,
      "step": 22000,
      "train_runtime": 12395.2144,
      "train_tokens_per_second": 116318.44
    },
    {
      "epoch": 0.221,
      "grad_norm": 0.32462117075920105,
      "learning_rate": 0.0002676160822791062,
      "loss": 0.6265,
      "num_input_tokens_seen": 1448345600,
      "step": 22100,
      "train_runtime": 12456.5575,
      "train_tokens_per_second": 116271.738
    },
    {
      "epoch": 0.222,
      "grad_norm": 0.32560595870018005,
      "learning_rate": 0.00026732007358362496,
      "loss": 0.6434,
      "num_input_tokens_seen": 1454899200,
      "step": 22200,
      "train_runtime": 12512.3995,
      "train_tokens_per_second": 116276.594
    },
    {
      "epoch": 0.223,
      "grad_norm": 0.2810288369655609,
      "learning_rate": 0.0002670228834763052,
      "loss": 0.6289,
      "num_input_tokens_seen": 1461452800,
      "step": 22300,
      "train_runtime": 12567.5646,
      "train_tokens_per_second": 116287.67
    },
    {
      "epoch": 0.224,
      "grad_norm": 0.2596298158168793,
      "learning_rate": 0.00026672451494984804,
      "loss": 0.6287,
      "num_input_tokens_seen": 1468006400,
      "step": 22400,
      "train_runtime": 12622.4346,
      "train_tokens_per_second": 116301.367
    },
    {
      "epoch": 0.225,
      "grad_norm": 0.2850113809108734,
      "learning_rate": 0.0002664249710088213,
      "loss": 0.6242,
      "num_input_tokens_seen": 1474560000,
      "step": 22500,
      "train_runtime": 12676.5798,
      "train_tokens_per_second": 116321.596
    },
    {
      "epoch": 0.226,
      "grad_norm": 0.27683302760124207,
      "learning_rate": 0.00026612425466962893,
      "loss": 0.6206,
      "num_input_tokens_seen": 1481113600,
      "step": 22600,
      "train_runtime": 12737.6237,
      "train_tokens_per_second": 116278.643
    },
    {
      "epoch": 0.227,
      "grad_norm": 0.239657461643219,
      "learning_rate": 0.00026582236896048134,
      "loss": 0.6225,
      "num_input_tokens_seen": 1487667200,
      "step": 22700,
      "train_runtime": 12792.2969,
      "train_tokens_per_second": 116293.987
    },
    {
      "epoch": 0.228,
      "grad_norm": 0.23831160366535187,
      "learning_rate": 0.00026551931692136413,
      "loss": 0.6239,
      "num_input_tokens_seen": 1494220800,
      "step": 22800,
      "train_runtime": 12846.1745,
      "train_tokens_per_second": 116316.402
    },
    {
      "epoch": 0.229,
      "grad_norm": 0.34589797258377075,
      "learning_rate": 0.00026521510160400804,
      "loss": 0.6247,
      "num_input_tokens_seen": 1500774400,
      "step": 22900,
      "train_runtime": 12900.799,
      "train_tokens_per_second": 116331.895
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.2497359812259674,
      "learning_rate": 0.00026490972607185793,
      "loss": 0.62,
      "num_input_tokens_seen": 1507328000,
      "step": 23000,
      "train_runtime": 12961.1601,
      "train_tokens_per_second": 116295.763
    },
    {
      "epoch": 0.231,
      "grad_norm": 0.24375373125076294,
      "learning_rate": 0.0002646031934000421,
      "loss": 0.6315,
      "num_input_tokens_seen": 1513881600,
      "step": 23100,
      "train_runtime": 13015.5663,
      "train_tokens_per_second": 116313.156
    },
    {
      "epoch": 0.232,
      "grad_norm": 0.26220017671585083,
      "learning_rate": 0.00026429550667534095,
      "loss": 0.6285,
      "num_input_tokens_seen": 1520435200,
      "step": 23200,
      "train_runtime": 13070.9925,
      "train_tokens_per_second": 116321.327
    },
    {
      "epoch": 0.233,
      "grad_norm": 0.3479808568954468,
      "learning_rate": 0.0002639866689961565,
      "loss": 0.6212,
      "num_input_tokens_seen": 1526988800,
      "step": 23300,
      "train_runtime": 13127.7973,
      "train_tokens_per_second": 116317.213
    },
    {
      "epoch": 0.234,
      "grad_norm": 0.2807121276855469,
      "learning_rate": 0.00026367668347248083,
      "loss": 0.6233,
      "num_input_tokens_seen": 1533542400,
      "step": 23400,
      "train_runtime": 13181.874,
      "train_tokens_per_second": 116337.206
    },
    {
      "epoch": 0.235,
      "grad_norm": 0.22995448112487793,
      "learning_rate": 0.0002633655532258646,
      "loss": 0.6322,
      "num_input_tokens_seen": 1540096000,
      "step": 23500,
      "train_runtime": 13243.5947,
      "train_tokens_per_second": 116289.877
    },
    {
      "epoch": 0.236,
      "grad_norm": 0.2977929413318634,
      "learning_rate": 0.000263053281389386,
      "loss": 0.6255,
      "num_input_tokens_seen": 1546649600,
      "step": 23600,
      "train_runtime": 13297.1642,
      "train_tokens_per_second": 116314.244
    },
    {
      "epoch": 0.237,
      "grad_norm": 0.2991272211074829,
      "learning_rate": 0.0002627398711076189,
      "loss": 0.6178,
      "num_input_tokens_seen": 1553203200,
      "step": 23700,
      "train_runtime": 13352.7857,
      "train_tokens_per_second": 116320.537
    },
    {
      "epoch": 0.238,
      "grad_norm": 0.26155802607536316,
      "learning_rate": 0.0002624253255366014,
      "loss": 0.6173,
      "num_input_tokens_seen": 1559756800,
      "step": 23800,
      "train_runtime": 13406.4815,
      "train_tokens_per_second": 116343.486
    },
    {
      "epoch": 0.239,
      "grad_norm": 0.3786696493625641,
      "learning_rate": 0.0002621096478438039,
      "loss": 0.6275,
      "num_input_tokens_seen": 1566310400,
      "step": 23900,
      "train_runtime": 13461.9758,
      "train_tokens_per_second": 116350.707
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.347703754901886,
      "learning_rate": 0.00026179284120809727,
      "loss": 0.6317,
      "num_input_tokens_seen": 1572864000,
      "step": 24000,
      "train_runtime": 13518.3676,
      "train_tokens_per_second": 116350.143
    },
    {
      "epoch": 0.241,
      "grad_norm": 0.2794933021068573,
      "learning_rate": 0.0002614749088197208,
      "loss": 0.6213,
      "num_input_tokens_seen": 1579417600,
      "step": 24100,
      "train_runtime": 13573.1619,
      "train_tokens_per_second": 116363.277
    },
    {
      "epoch": 0.242,
      "grad_norm": 0.4467460811138153,
      "learning_rate": 0.00026115585388025015,
      "loss": 0.618,
      "num_input_tokens_seen": 1585971200,
      "step": 24200,
      "train_runtime": 13633.9103,
      "train_tokens_per_second": 116325.483
    },
    {
      "epoch": 0.243,
      "grad_norm": 0.3274633288383484,
      "learning_rate": 0.00026083567960256493,
      "loss": 0.6198,
      "num_input_tokens_seen": 1592524800,
      "step": 24300,
      "train_runtime": 13690.8223,
      "train_tokens_per_second": 116320.61
    },
    {
      "epoch": 0.244,
      "grad_norm": 0.26471319794654846,
      "learning_rate": 0.00026051438921081667,
      "loss": 0.621,
      "num_input_tokens_seen": 1599078400,
      "step": 24400,
      "train_runtime": 13746.056,
      "train_tokens_per_second": 116329.979
    },
    {
      "epoch": 0.245,
      "grad_norm": 0.29486411809921265,
      "learning_rate": 0.00026019198594039595,
      "loss": 0.6275,
      "num_input_tokens_seen": 1605632000,
      "step": 24500,
      "train_runtime": 13802.5927,
      "train_tokens_per_second": 116328.29
    },
    {
      "epoch": 0.246,
      "grad_norm": 0.31198298931121826,
      "learning_rate": 0.00025986847303790026,
      "loss": 0.6232,
      "num_input_tokens_seen": 1612185600,
      "step": 24600,
      "train_runtime": 13856.7259,
      "train_tokens_per_second": 116346.792
    },
    {
      "epoch": 0.247,
      "grad_norm": 0.28400614857673645,
      "learning_rate": 0.00025954385376110076,
      "loss": 0.6213,
      "num_input_tokens_seen": 1618739200,
      "step": 24700,
      "train_runtime": 13911.4019,
      "train_tokens_per_second": 116360.609
    },
    {
      "epoch": 0.248,
      "grad_norm": 0.24376364052295685,
      "learning_rate": 0.00025921813137891005,
      "loss": 0.6279,
      "num_input_tokens_seen": 1625292800,
      "step": 24800,
      "train_runtime": 13968.2353,
      "train_tokens_per_second": 116356.345
    },
    {
      "epoch": 0.249,
      "grad_norm": 0.28743863105773926,
      "learning_rate": 0.000258891309171349,
      "loss": 0.6238,
      "num_input_tokens_seen": 1631846400,
      "step": 24900,
      "train_runtime": 14022.4846,
      "train_tokens_per_second": 116373.557
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.33175599575042725,
      "learning_rate": 0.00025856339042951344,
      "loss": 0.6159,
      "num_input_tokens_seen": 1638400000,
      "step": 25000,
      "train_runtime": 14076.9493,
      "train_tokens_per_second": 116388.855
    },
    {
      "epoch": 0.251,
      "grad_norm": 0.31394195556640625,
      "learning_rate": 0.0002582343784555415,
      "loss": 0.6194,
      "num_input_tokens_seen": 1644953600,
      "step": 25100,
      "train_runtime": 14137.6518,
      "train_tokens_per_second": 116352.675
    },
    {
      "epoch": 0.252,
      "grad_norm": 0.2717108428478241,
      "learning_rate": 0.00025790427656258017,
      "loss": 0.63,
      "num_input_tokens_seen": 1651507200,
      "step": 25200,
      "train_runtime": 14191.6974,
      "train_tokens_per_second": 116371.365
    },
    {
      "epoch": 0.253,
      "grad_norm": 0.2907046377658844,
      "learning_rate": 0.00025757308807475185,
      "loss": 0.6214,
      "num_input_tokens_seen": 1658060800,
      "step": 25300,
      "train_runtime": 14247.6012,
      "train_tokens_per_second": 116374.734
    },
    {
      "epoch": 0.254,
      "grad_norm": 0.25791919231414795,
      "learning_rate": 0.00025724081632712086,
      "loss": 0.6197,
      "num_input_tokens_seen": 1664614400,
      "step": 25400,
      "train_runtime": 14302.1699,
      "train_tokens_per_second": 116388.94
    },
    {
      "epoch": 0.255,
      "grad_norm": 0.23399506509304047,
      "learning_rate": 0.0002569074646656601,
      "loss": 0.6145,
      "num_input_tokens_seen": 1671168000,
      "step": 25500,
      "train_runtime": 14362.8828,
      "train_tokens_per_second": 116353.243
    },
    {
      "epoch": 0.256,
      "grad_norm": 0.3080284297466278,
      "learning_rate": 0.00025657303644721695,
      "loss": 0.6288,
      "num_input_tokens_seen": 1677721600,
      "step": 25600,
      "train_runtime": 14418.2746,
      "train_tokens_per_second": 116360.775
    },
    {
      "epoch": 0.257,
      "grad_norm": 0.27764490246772766,
      "learning_rate": 0.00025623753503948004,
      "loss": 0.6302,
      "num_input_tokens_seen": 1684275200,
      "step": 25700,
      "train_runtime": 14472.1462,
      "train_tokens_per_second": 116380.471
    },
    {
      "epoch": 0.258,
      "grad_norm": 0.3530017137527466,
      "learning_rate": 0.00025590096382094475,
      "loss": 0.6205,
      "num_input_tokens_seen": 1690828800,
      "step": 25800,
      "train_runtime": 14527.093,
      "train_tokens_per_second": 116391.407
    },
    {
      "epoch": 0.259,
      "grad_norm": 0.28179532289505005,
      "learning_rate": 0.00025556332618087945,
      "loss": 0.6196,
      "num_input_tokens_seen": 1697382400,
      "step": 25900,
      "train_runtime": 14586.2883,
      "train_tokens_per_second": 116368.357
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.27701336145401,
      "learning_rate": 0.00025522462551929155,
      "loss": 0.6244,
      "num_input_tokens_seen": 1703936000,
      "step": 26000,
      "train_runtime": 14641.6003,
      "train_tokens_per_second": 116376.35
    },
    {
      "epoch": 0.261,
      "grad_norm": 0.2609594464302063,
      "learning_rate": 0.00025488486524689283,
      "loss": 0.625,
      "num_input_tokens_seen": 1710489600,
      "step": 26100,
      "train_runtime": 14698.0992,
      "train_tokens_per_second": 116374.884
    },
    {
      "epoch": 0.262,
      "grad_norm": 0.2739205062389374,
      "learning_rate": 0.00025454404878506555,
      "loss": 0.6214,
      "num_input_tokens_seen": 1717043200,
      "step": 26200,
      "train_runtime": 14752.0923,
      "train_tokens_per_second": 116393.198
    },
    {
      "epoch": 0.263,
      "grad_norm": 0.27936309576034546,
      "learning_rate": 0.0002542021795658276,
      "loss": 0.6203,
      "num_input_tokens_seen": 1723596800,
      "step": 26300,
      "train_runtime": 14806.7928,
      "train_tokens_per_second": 116405.816
    },
    {
      "epoch": 0.264,
      "grad_norm": 0.24206426739692688,
      "learning_rate": 0.0002538592610317984,
      "loss": 0.622,
      "num_input_tokens_seen": 1730150400,
      "step": 26400,
      "train_runtime": 14868.0925,
      "train_tokens_per_second": 116366.669
    },
    {
      "epoch": 0.265,
      "grad_norm": 0.29860714077949524,
      "learning_rate": 0.00025351529663616355,
      "loss": 0.6244,
      "num_input_tokens_seen": 1736704000,
      "step": 26500,
      "train_runtime": 14923.7941,
      "train_tokens_per_second": 116371.479
    },
    {
      "epoch": 0.266,
      "grad_norm": 0.2928908169269562,
      "learning_rate": 0.00025317028984264087,
      "loss": 0.6183,
      "num_input_tokens_seen": 1743257600,
      "step": 26600,
      "train_runtime": 14978.8307,
      "train_tokens_per_second": 116381.421
    },
    {
      "epoch": 0.267,
      "grad_norm": 0.32871657609939575,
      "learning_rate": 0.0002528242441254448,
      "loss": 0.6163,
      "num_input_tokens_seen": 1749811200,
      "step": 26700,
      "train_runtime": 15032.6295,
      "train_tokens_per_second": 116400.873
    },
    {
      "epoch": 0.268,
      "grad_norm": 0.26268327236175537,
      "learning_rate": 0.000252477162969252,
      "loss": 0.6211,
      "num_input_tokens_seen": 1756364800,
      "step": 26800,
      "train_runtime": 15087.2644,
      "train_tokens_per_second": 116413.735
    },
    {
      "epoch": 0.269,
      "grad_norm": 0.248652845621109,
      "learning_rate": 0.00025212904986916584,
      "loss": 0.6231,
      "num_input_tokens_seen": 1762918400,
      "step": 26900,
      "train_runtime": 15142.5638,
      "train_tokens_per_second": 116421.395
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.28964775800704956,
      "learning_rate": 0.00025177990833068133,
      "loss": 0.6233,
      "num_input_tokens_seen": 1769472000,
      "step": 27000,
      "train_runtime": 15204.3907,
      "train_tokens_per_second": 116379.014
    },
    {
      "epoch": 0.271,
      "grad_norm": 0.3479403555393219,
      "learning_rate": 0.0002514297418696499,
      "loss": 0.6136,
      "num_input_tokens_seen": 1776025600,
      "step": 27100,
      "train_runtime": 15259.9986,
      "train_tokens_per_second": 116384.388
    },
    {
      "epoch": 0.272,
      "grad_norm": 0.2727760076522827,
      "learning_rate": 0.0002510785540122439,
      "loss": 0.6215,
      "num_input_tokens_seen": 1782579200,
      "step": 27200,
      "train_runtime": 15313.7249,
      "train_tokens_per_second": 116404.024
    },
    {
      "epoch": 0.273,
      "grad_norm": 0.2885916531085968,
      "learning_rate": 0.0002507263482949212,
      "loss": 0.6269,
      "num_input_tokens_seen": 1789132800,
      "step": 27300,
      "train_runtime": 15370.4778,
      "train_tokens_per_second": 116400.598
    },
    {
      "epoch": 0.274,
      "grad_norm": 0.2588540017604828,
      "learning_rate": 0.0002503731282643894,
      "loss": 0.619,
      "num_input_tokens_seen": 1795686400,
      "step": 27400,
      "train_runtime": 15425.7714,
      "train_tokens_per_second": 116408.208
    },
    {
      "epoch": 0.275,
      "grad_norm": 0.29681408405303955,
      "learning_rate": 0.0002500188974775704,
      "loss": 0.6175,
      "num_input_tokens_seen": 1802240000,
      "step": 27500,
      "train_runtime": 15480.68,
      "train_tokens_per_second": 116418.658
    },
    {
      "epoch": 0.276,
      "grad_norm": 0.319444864988327,
      "learning_rate": 0.00024966365950156416,
      "loss": 0.6155,
      "num_input_tokens_seen": 1808793600,
      "step": 27600,
      "train_runtime": 15541.0108,
      "train_tokens_per_second": 116388.414
    },
    {
      "epoch": 0.277,
      "grad_norm": 0.3779368996620178,
      "learning_rate": 0.00024930741791361326,
      "loss": 0.6238,
      "num_input_tokens_seen": 1815347200,
      "step": 27700,
      "train_runtime": 15597.4681,
      "train_tokens_per_second": 116387.3
    },
    {
      "epoch": 0.278,
      "grad_norm": 0.4451746940612793,
      "learning_rate": 0.0002489501763010664,
      "loss": 0.6201,
      "num_input_tokens_seen": 1821900800,
      "step": 27800,
      "train_runtime": 15648.3014,
      "train_tokens_per_second": 116428.024
    },
    {
      "epoch": 0.279,
      "grad_norm": 0.35030296444892883,
      "learning_rate": 0.00024859193826134285,
      "loss": 0.6173,
      "num_input_tokens_seen": 1828454400,
      "step": 27900,
      "train_runtime": 15708.8593,
      "train_tokens_per_second": 116396.383
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.2962912321090698,
      "learning_rate": 0.00024823270740189556,
      "loss": 0.6154,
      "num_input_tokens_seen": 1835008000,
      "step": 28000,
      "train_runtime": 15764.0565,
      "train_tokens_per_second": 116404.556
    },
    {
      "epoch": 0.281,
      "grad_norm": 0.2939753234386444,
      "learning_rate": 0.00024787248734017527,
      "loss": 0.6222,
      "num_input_tokens_seen": 1841561600,
      "step": 28100,
      "train_runtime": 15818.596,
      "train_tokens_per_second": 116417.513
    },
    {
      "epoch": 0.282,
      "grad_norm": 0.24470455944538116,
      "learning_rate": 0.0002475112817035941,
      "loss": 0.6238,
      "num_input_tokens_seen": 1848115200,
      "step": 28200,
      "train_runtime": 15872.9134,
      "train_tokens_per_second": 116432.01
    },
    {
      "epoch": 0.283,
      "grad_norm": 0.31226930022239685,
      "learning_rate": 0.0002471490941294887,
      "loss": 0.6212,
      "num_input_tokens_seen": 1854668800,
      "step": 28300,
      "train_runtime": 15933.0328,
      "train_tokens_per_second": 116404.003
    },
    {
      "epoch": 0.284,
      "grad_norm": 0.2585756182670593,
      "learning_rate": 0.000246785928265084,
      "loss": 0.6191,
      "num_input_tokens_seen": 1861222400,
      "step": 28400,
      "train_runtime": 15987.3279,
      "train_tokens_per_second": 116418.604
    },
    {
      "epoch": 0.285,
      "grad_norm": 0.4430062770843506,
      "learning_rate": 0.0002464217877674562,
      "loss": 0.6249,
      "num_input_tokens_seen": 1867776000,
      "step": 28500,
      "train_runtime": 16041.837,
      "train_tokens_per_second": 116431.553
    },
    {
      "epoch": 0.286,
      "grad_norm": 0.31980106234550476,
      "learning_rate": 0.0002460566763034961,
      "loss": 0.6303,
      "num_input_tokens_seen": 1874329600,
      "step": 28600,
      "train_runtime": 16098.6135,
      "train_tokens_per_second": 116428.014
    },
    {
      "epoch": 0.287,
      "grad_norm": 0.26904401183128357,
      "learning_rate": 0.00024569059754987196,
      "loss": 0.6214,
      "num_input_tokens_seen": 1880883200,
      "step": 28700,
      "train_runtime": 16159.9296,
      "train_tokens_per_second": 116391.794
    },
    {
      "epoch": 0.288,
      "grad_norm": 0.2635185420513153,
      "learning_rate": 0.00024532355519299296,
      "loss": 0.6181,
      "num_input_tokens_seen": 1887436800,
      "step": 28800,
      "train_runtime": 16214.468,
      "train_tokens_per_second": 116404.485
    },
    {
      "epoch": 0.289,
      "grad_norm": 1.0330250263214111,
      "learning_rate": 0.0002449555529289714,
      "loss": 0.6241,
      "num_input_tokens_seen": 1893990400,
      "step": 28900,
      "train_runtime": 16268.5971,
      "train_tokens_per_second": 116420.02
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.3014289438724518,
      "learning_rate": 0.0002445865944635861,
      "loss": 0.6295,
      "num_input_tokens_seen": 1900544000,
      "step": 29000,
      "train_runtime": 16324.9935,
      "train_tokens_per_second": 116419.281
    },
    {
      "epoch": 0.291,
      "grad_norm": 0.2632603347301483,
      "learning_rate": 0.0002442166835122446,
      "loss": 0.6188,
      "num_input_tokens_seen": 1907097600,
      "step": 29100,
      "train_runtime": 16378.6859,
      "train_tokens_per_second": 116437.766
    },
    {
      "epoch": 0.292,
      "grad_norm": 0.22746869921684265,
      "learning_rate": 0.00024384582379994614,
      "loss": 0.6172,
      "num_input_tokens_seen": 1913651200,
      "step": 29200,
      "train_runtime": 16434.3593,
      "train_tokens_per_second": 116442.093
    },
    {
      "epoch": 0.293,
      "grad_norm": 0.26981279253959656,
      "learning_rate": 0.00024347401906124388,
      "loss": 0.6161,
      "num_input_tokens_seen": 1920204800,
      "step": 29300,
      "train_runtime": 16494.2752,
      "train_tokens_per_second": 116416.44
    },
    {
      "epoch": 0.294,
      "grad_norm": 0.30709710717201233,
      "learning_rate": 0.0002431012730402075,
      "loss": 0.6225,
      "num_input_tokens_seen": 1926758400,
      "step": 29400,
      "train_runtime": 16549.5618,
      "train_tokens_per_second": 116423.53
    },
    {
      "epoch": 0.295,
      "grad_norm": 0.27512043714523315,
      "learning_rate": 0.00024272758949038517,
      "loss": 0.6167,
      "num_input_tokens_seen": 1933312000,
      "step": 29500,
      "train_runtime": 16604.0285,
      "train_tokens_per_second": 116436.321
    },
    {
      "epoch": 0.296,
      "grad_norm": 0.33047741651535034,
      "learning_rate": 0.00024235297217476616,
      "loss": 0.6194,
      "num_input_tokens_seen": 1939865600,
      "step": 29600,
      "train_runtime": 16665.4033,
      "train_tokens_per_second": 116400.76
    },
    {
      "epoch": 0.297,
      "grad_norm": 0.23181872069835663,
      "learning_rate": 0.00024197742486574268,
      "loss": 0.6235,
      "num_input_tokens_seen": 1946419200,
      "step": 29700,
      "train_runtime": 16719.5543,
      "train_tokens_per_second": 116415.735
    },
    {
      "epoch": 0.298,
      "grad_norm": 0.25960394740104675,
      "learning_rate": 0.0002416009513450719,
      "loss": 0.6189,
      "num_input_tokens_seen": 1952972800,
      "step": 29800,
      "train_runtime": 16775.5144,
      "train_tokens_per_second": 116418.058
    },
    {
      "epoch": 0.299,
      "grad_norm": 0.3980496823787689,
      "learning_rate": 0.00024122355540383806,
      "loss": 0.6205,
      "num_input_tokens_seen": 1959526400,
      "step": 29900,
      "train_runtime": 16831.9075,
      "train_tokens_per_second": 116417.369
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.274140328168869,
      "learning_rate": 0.00024084524084241405,
      "loss": 0.6137,
      "num_input_tokens_seen": 1966080000,
      "step": 30000,
      "train_runtime": 16885.9089,
      "train_tokens_per_second": 116433.176
    }
  ],
  "logging_steps": 100,
  "max_steps": 100000,
  "num_input_tokens_seen": 1966080000,
  "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": 3.755848237056e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}