{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.2410655095522208,
  "eval_steps": 300,
  "global_step": 3000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 8.035516985074027e-05,
      "grad_norm": 0.4128876030445099,
      "learning_rate": 0.0,
      "loss": 1.8204,
      "num_input_tokens_seen": 3020,
      "step": 1,
      "train_runtime": 4.1565,
      "train_tokens_per_second": 726.569
    },
    {
      "epoch": 0.00016071033970148054,
      "grad_norm": 0.4843665063381195,
      "learning_rate": 4e-05,
      "loss": 1.8788,
      "num_input_tokens_seen": 6288,
      "step": 2,
      "train_runtime": 7.5142,
      "train_tokens_per_second": 836.817
    },
    {
      "epoch": 0.0002410655095522208,
      "grad_norm": 0.40459391474723816,
      "learning_rate": 8e-05,
      "loss": 1.7231,
      "num_input_tokens_seen": 8462,
      "step": 3,
      "train_runtime": 9.4044,
      "train_tokens_per_second": 899.788
    },
    {
      "epoch": 0.0003214206794029611,
      "grad_norm": 0.5273571610450745,
      "learning_rate": 0.00012,
      "loss": 2.0439,
      "num_input_tokens_seen": 10932,
      "step": 4,
      "train_runtime": 11.4295,
      "train_tokens_per_second": 956.472
    },
    {
      "epoch": 0.0004017758492537014,
      "grad_norm": 0.653209388256073,
      "learning_rate": 0.00016,
      "loss": 2.0319,
      "num_input_tokens_seen": 12380,
      "step": 5,
      "train_runtime": 12.9859,
      "train_tokens_per_second": 953.345
    },
    {
      "epoch": 0.0004821310191044416,
      "grad_norm": 0.9386360049247742,
      "learning_rate": 0.0002,
      "loss": 1.5702,
      "num_input_tokens_seen": 14486,
      "step": 6,
      "train_runtime": 14.8546,
      "train_tokens_per_second": 975.184
    },
    {
      "epoch": 0.0005624861889551819,
      "grad_norm": 0.7347410917282104,
      "learning_rate": 0.00019993322203672788,
      "loss": 1.4188,
      "num_input_tokens_seen": 16916,
      "step": 7,
      "train_runtime": 16.7998,
      "train_tokens_per_second": 1006.916
    },
    {
      "epoch": 0.0006428413588059222,
      "grad_norm": 0.7510082721710205,
      "learning_rate": 0.00019986644407345576,
      "loss": 1.6411,
      "num_input_tokens_seen": 19304,
      "step": 8,
      "train_runtime": 18.9414,
      "train_tokens_per_second": 1019.145
    },
    {
      "epoch": 0.0007231965286566625,
      "grad_norm": 0.531984269618988,
      "learning_rate": 0.00019979966611018366,
      "loss": 1.3169,
      "num_input_tokens_seen": 22836,
      "step": 9,
      "train_runtime": 21.6639,
      "train_tokens_per_second": 1054.103
    },
    {
      "epoch": 0.0008035516985074027,
      "grad_norm": 0.5088340044021606,
      "learning_rate": 0.00019973288814691153,
      "loss": 1.1896,
      "num_input_tokens_seen": 24708,
      "step": 10,
      "train_runtime": 23.3512,
      "train_tokens_per_second": 1058.105
    },
    {
      "epoch": 0.000883906868358143,
      "grad_norm": 0.5455313920974731,
      "learning_rate": 0.0001996661101836394,
      "loss": 1.3183,
      "num_input_tokens_seen": 26740,
      "step": 11,
      "train_runtime": 25.1284,
      "train_tokens_per_second": 1064.134
    },
    {
      "epoch": 0.0009642620382088832,
      "grad_norm": 0.811518132686615,
      "learning_rate": 0.00019959933222036728,
      "loss": 1.4893,
      "num_input_tokens_seen": 30814,
      "step": 12,
      "train_runtime": 28.0711,
      "train_tokens_per_second": 1097.711
    },
    {
      "epoch": 0.0010446172080596236,
      "grad_norm": 0.7564262747764587,
      "learning_rate": 0.00019953255425709515,
      "loss": 1.1969,
      "num_input_tokens_seen": 34248,
      "step": 13,
      "train_runtime": 30.648,
      "train_tokens_per_second": 1117.464
    },
    {
      "epoch": 0.0011249723779103637,
      "grad_norm": 1.155938744544983,
      "learning_rate": 0.00019946577629382305,
      "loss": 1.2152,
      "num_input_tokens_seen": 36942,
      "step": 14,
      "train_runtime": 32.8559,
      "train_tokens_per_second": 1124.364
    },
    {
      "epoch": 0.001205327547761104,
      "grad_norm": 0.6982361078262329,
      "learning_rate": 0.00019939899833055092,
      "loss": 1.3557,
      "num_input_tokens_seen": 43012,
      "step": 15,
      "train_runtime": 38.7669,
      "train_tokens_per_second": 1109.503
    },
    {
      "epoch": 0.0012856827176118443,
      "grad_norm": 0.7386748194694519,
      "learning_rate": 0.00019933222036727882,
      "loss": 1.3978,
      "num_input_tokens_seen": 45962,
      "step": 16,
      "train_runtime": 41.1696,
      "train_tokens_per_second": 1116.405
    },
    {
      "epoch": 0.0013660378874625846,
      "grad_norm": 0.3979336619377136,
      "learning_rate": 0.0001992654424040067,
      "loss": 1.4051,
      "num_input_tokens_seen": 49486,
      "step": 17,
      "train_runtime": 43.7809,
      "train_tokens_per_second": 1130.31
    },
    {
      "epoch": 0.001446393057313325,
      "grad_norm": 0.587181568145752,
      "learning_rate": 0.00019919866444073457,
      "loss": 1.384,
      "num_input_tokens_seen": 53358,
      "step": 18,
      "train_runtime": 46.7014,
      "train_tokens_per_second": 1142.534
    },
    {
      "epoch": 0.0015267482271640652,
      "grad_norm": 0.32670366764068604,
      "learning_rate": 0.00019913188647746244,
      "loss": 1.1467,
      "num_input_tokens_seen": 57148,
      "step": 19,
      "train_runtime": 49.547,
      "train_tokens_per_second": 1153.411
    },
    {
      "epoch": 0.0016071033970148055,
      "grad_norm": 0.40024834871292114,
      "learning_rate": 0.00019906510851419034,
      "loss": 1.3787,
      "num_input_tokens_seen": 59690,
      "step": 20,
      "train_runtime": 51.6745,
      "train_tokens_per_second": 1155.114
    },
    {
      "epoch": 0.0016874585668655458,
      "grad_norm": 0.5168688297271729,
      "learning_rate": 0.00019899833055091822,
      "loss": 1.1428,
      "num_input_tokens_seen": 63540,
      "step": 21,
      "train_runtime": 54.5401,
      "train_tokens_per_second": 1165.014
    },
    {
      "epoch": 0.001767813736716286,
      "grad_norm": 0.39956149458885193,
      "learning_rate": 0.0001989315525876461,
      "loss": 1.0249,
      "num_input_tokens_seen": 65924,
      "step": 22,
      "train_runtime": 56.6016,
      "train_tokens_per_second": 1164.703
    },
    {
      "epoch": 0.0018481689065670262,
      "grad_norm": 0.28881531953811646,
      "learning_rate": 0.00019886477462437396,
      "loss": 0.806,
      "num_input_tokens_seen": 70882,
      "step": 23,
      "train_runtime": 60.2487,
      "train_tokens_per_second": 1176.489
    },
    {
      "epoch": 0.0019285240764177665,
      "grad_norm": 0.3325786590576172,
      "learning_rate": 0.00019879799666110183,
      "loss": 1.0302,
      "num_input_tokens_seen": 74916,
      "step": 24,
      "train_runtime": 63.2047,
      "train_tokens_per_second": 1185.292
    },
    {
      "epoch": 0.0020088792462685068,
      "grad_norm": 0.29011744260787964,
      "learning_rate": 0.00019873121869782974,
      "loss": 1.0027,
      "num_input_tokens_seen": 77098,
      "step": 25,
      "train_runtime": 65.1412,
      "train_tokens_per_second": 1183.553
    },
    {
      "epoch": 0.0020892344161192473,
      "grad_norm": 0.2955094277858734,
      "learning_rate": 0.0001986644407345576,
      "loss": 1.3089,
      "num_input_tokens_seen": 79924,
      "step": 26,
      "train_runtime": 67.3225,
      "train_tokens_per_second": 1187.182
    },
    {
      "epoch": 0.0021695895859699874,
      "grad_norm": 0.4490657448768616,
      "learning_rate": 0.00019859766277128548,
      "loss": 1.4656,
      "num_input_tokens_seen": 82532,
      "step": 27,
      "train_runtime": 69.5014,
      "train_tokens_per_second": 1187.487
    },
    {
      "epoch": 0.0022499447558207274,
      "grad_norm": 0.4933200776576996,
      "learning_rate": 0.00019853088480801335,
      "loss": 1.2859,
      "num_input_tokens_seen": 84328,
      "step": 28,
      "train_runtime": 71.2185,
      "train_tokens_per_second": 1184.074
    },
    {
      "epoch": 0.002330299925671468,
      "grad_norm": 0.33834826946258545,
      "learning_rate": 0.00019846410684474123,
      "loss": 1.4659,
      "num_input_tokens_seen": 87268,
      "step": 29,
      "train_runtime": 73.4964,
      "train_tokens_per_second": 1187.378
    },
    {
      "epoch": 0.002410655095522208,
      "grad_norm": 0.4045095443725586,
      "learning_rate": 0.00019839732888146913,
      "loss": 1.2821,
      "num_input_tokens_seen": 90546,
      "step": 30,
      "train_runtime": 75.9674,
      "train_tokens_per_second": 1191.905
    },
    {
      "epoch": 0.0024910102653729485,
      "grad_norm": 0.4099307656288147,
      "learning_rate": 0.000198330550918197,
      "loss": 1.2785,
      "num_input_tokens_seen": 92580,
      "step": 31,
      "train_runtime": 80.094,
      "train_tokens_per_second": 1155.891
    },
    {
      "epoch": 0.0025713654352236886,
      "grad_norm": 0.34756144881248474,
      "learning_rate": 0.00019826377295492487,
      "loss": 1.307,
      "num_input_tokens_seen": 95030,
      "step": 32,
      "train_runtime": 82.0563,
      "train_tokens_per_second": 1158.107
    },
    {
      "epoch": 0.002651720605074429,
      "grad_norm": 0.34918075799942017,
      "learning_rate": 0.00019819699499165277,
      "loss": 1.1396,
      "num_input_tokens_seen": 96986,
      "step": 33,
      "train_runtime": 83.8307,
      "train_tokens_per_second": 1156.927
    },
    {
      "epoch": 0.002732075774925169,
      "grad_norm": 0.3395019471645355,
      "learning_rate": 0.00019813021702838065,
      "loss": 0.7633,
      "num_input_tokens_seen": 98324,
      "step": 34,
      "train_runtime": 85.4198,
      "train_tokens_per_second": 1151.068
    },
    {
      "epoch": 0.0028124309447759097,
      "grad_norm": 0.34623047709465027,
      "learning_rate": 0.00019806343906510852,
      "loss": 1.2512,
      "num_input_tokens_seen": 101938,
      "step": 35,
      "train_runtime": 88.129,
      "train_tokens_per_second": 1156.691
    },
    {
      "epoch": 0.00289278611462665,
      "grad_norm": 0.27665939927101135,
      "learning_rate": 0.00019799666110183642,
      "loss": 1.0116,
      "num_input_tokens_seen": 103720,
      "step": 36,
      "train_runtime": 89.9509,
      "train_tokens_per_second": 1153.074
    },
    {
      "epoch": 0.00297314128447739,
      "grad_norm": 0.30108535289764404,
      "learning_rate": 0.0001979298831385643,
      "loss": 1.0475,
      "num_input_tokens_seen": 105896,
      "step": 37,
      "train_runtime": 91.8679,
      "train_tokens_per_second": 1152.699
    },
    {
      "epoch": 0.0030534964543281304,
      "grad_norm": 0.30923861265182495,
      "learning_rate": 0.00019786310517529217,
      "loss": 1.1604,
      "num_input_tokens_seen": 108132,
      "step": 38,
      "train_runtime": 93.8496,
      "train_tokens_per_second": 1152.184
    },
    {
      "epoch": 0.0031338516241788705,
      "grad_norm": 0.33123621344566345,
      "learning_rate": 0.00019779632721202004,
      "loss": 1.1684,
      "num_input_tokens_seen": 114642,
      "step": 39,
      "train_runtime": 98.1366,
      "train_tokens_per_second": 1168.188
    },
    {
      "epoch": 0.003214206794029611,
      "grad_norm": 0.3305603861808777,
      "learning_rate": 0.00019772954924874791,
      "loss": 0.9128,
      "num_input_tokens_seen": 117000,
      "step": 40,
      "train_runtime": 100.1499,
      "train_tokens_per_second": 1168.248
    },
    {
      "epoch": 0.003294561963880351,
      "grad_norm": 0.3243100345134735,
      "learning_rate": 0.00019766277128547581,
      "loss": 1.018,
      "num_input_tokens_seen": 119622,
      "step": 41,
      "train_runtime": 102.2941,
      "train_tokens_per_second": 1169.393
    },
    {
      "epoch": 0.0033749171337310916,
      "grad_norm": 0.30681395530700684,
      "learning_rate": 0.0001975959933222037,
      "loss": 0.8678,
      "num_input_tokens_seen": 121540,
      "step": 42,
      "train_runtime": 104.0763,
      "train_tokens_per_second": 1167.797
    },
    {
      "epoch": 0.0034552723035818317,
      "grad_norm": 0.3122865557670593,
      "learning_rate": 0.00019752921535893156,
      "loss": 1.2514,
      "num_input_tokens_seen": 125898,
      "step": 43,
      "train_runtime": 107.1933,
      "train_tokens_per_second": 1174.495
    },
    {
      "epoch": 0.003535627473432572,
      "grad_norm": 0.26632049679756165,
      "learning_rate": 0.00019746243739565943,
      "loss": 1.0827,
      "num_input_tokens_seen": 128672,
      "step": 44,
      "train_runtime": 109.4139,
      "train_tokens_per_second": 1176.012
    },
    {
      "epoch": 0.0036159826432833123,
      "grad_norm": 0.28619447350502014,
      "learning_rate": 0.0001973956594323873,
      "loss": 1.1041,
      "num_input_tokens_seen": 130568,
      "step": 45,
      "train_runtime": 111.1803,
      "train_tokens_per_second": 1174.381
    },
    {
      "epoch": 0.0036963378131340523,
      "grad_norm": 0.3083089292049408,
      "learning_rate": 0.0001973288814691152,
      "loss": 0.891,
      "num_input_tokens_seen": 132298,
      "step": 46,
      "train_runtime": 112.8204,
      "train_tokens_per_second": 1172.642
    },
    {
      "epoch": 0.003776692982984793,
      "grad_norm": 0.27593693137168884,
      "learning_rate": 0.00019726210350584308,
      "loss": 1.1675,
      "num_input_tokens_seen": 134276,
      "step": 47,
      "train_runtime": 114.6997,
      "train_tokens_per_second": 1170.675
    },
    {
      "epoch": 0.003857048152835533,
      "grad_norm": 0.33598679304122925,
      "learning_rate": 0.00019719532554257095,
      "loss": 0.9671,
      "num_input_tokens_seen": 137872,
      "step": 48,
      "train_runtime": 117.4017,
      "train_tokens_per_second": 1174.361
    },
    {
      "epoch": 0.0039374033226862734,
      "grad_norm": 0.34018149971961975,
      "learning_rate": 0.00019712854757929883,
      "loss": 1.2053,
      "num_input_tokens_seen": 140504,
      "step": 49,
      "train_runtime": 119.5089,
      "train_tokens_per_second": 1175.678
    },
    {
      "epoch": 0.0040177584925370135,
      "grad_norm": 0.2602927088737488,
      "learning_rate": 0.00019706176961602673,
      "loss": 1.0059,
      "num_input_tokens_seen": 144068,
      "step": 50,
      "train_runtime": 122.149,
      "train_tokens_per_second": 1179.445
    },
    {
      "epoch": 0.004098113662387754,
      "grad_norm": 0.32150211930274963,
      "learning_rate": 0.0001969949916527546,
      "loss": 1.3919,
      "num_input_tokens_seen": 147256,
      "step": 51,
      "train_runtime": 124.6096,
      "train_tokens_per_second": 1181.739
    },
    {
      "epoch": 0.0041784688322384946,
      "grad_norm": 0.29502740502357483,
      "learning_rate": 0.0001969282136894825,
      "loss": 0.9057,
      "num_input_tokens_seen": 150536,
      "step": 52,
      "train_runtime": 127.0806,
      "train_tokens_per_second": 1184.571
    },
    {
      "epoch": 0.004258824002089235,
      "grad_norm": 0.3121987581253052,
      "learning_rate": 0.00019686143572621037,
      "loss": 1.0761,
      "num_input_tokens_seen": 152722,
      "step": 53,
      "train_runtime": 129.0448,
      "train_tokens_per_second": 1183.481
    },
    {
      "epoch": 0.004339179171939975,
      "grad_norm": 0.3160906732082367,
      "learning_rate": 0.00019679465776293825,
      "loss": 1.3607,
      "num_input_tokens_seen": 155360,
      "step": 54,
      "train_runtime": 131.1821,
      "train_tokens_per_second": 1184.308
    },
    {
      "epoch": 0.004419534341790715,
      "grad_norm": 0.3124387264251709,
      "learning_rate": 0.00019672787979966612,
      "loss": 1.0816,
      "num_input_tokens_seen": 157556,
      "step": 55,
      "train_runtime": 133.0249,
      "train_tokens_per_second": 1184.41
    },
    {
      "epoch": 0.004499889511641455,
      "grad_norm": 0.3181786835193634,
      "learning_rate": 0.000196661101836394,
      "loss": 1.1499,
      "num_input_tokens_seen": 159108,
      "step": 56,
      "train_runtime": 134.668,
      "train_tokens_per_second": 1181.484
    },
    {
      "epoch": 0.004580244681492196,
      "grad_norm": 0.2405022531747818,
      "learning_rate": 0.0001965943238731219,
      "loss": 0.996,
      "num_input_tokens_seen": 161932,
      "step": 57,
      "train_runtime": 136.9022,
      "train_tokens_per_second": 1182.83
    },
    {
      "epoch": 0.004660599851342936,
      "grad_norm": 0.4457870125770569,
      "learning_rate": 0.00019652754590984977,
      "loss": 1.2403,
      "num_input_tokens_seen": 165060,
      "step": 58,
      "train_runtime": 139.3245,
      "train_tokens_per_second": 1184.717
    },
    {
      "epoch": 0.004740955021193676,
      "grad_norm": 0.38461732864379883,
      "learning_rate": 0.00019646076794657764,
      "loss": 1.1906,
      "num_input_tokens_seen": 166838,
      "step": 59,
      "train_runtime": 141.049,
      "train_tokens_per_second": 1182.838
    },
    {
      "epoch": 0.004821310191044416,
      "grad_norm": 0.3107485771179199,
      "learning_rate": 0.0001963939899833055,
      "loss": 0.8156,
      "num_input_tokens_seen": 169110,
      "step": 60,
      "train_runtime": 143.0768,
      "train_tokens_per_second": 1181.953
    },
    {
      "epoch": 0.004901665360895157,
      "grad_norm": 0.33994483947753906,
      "learning_rate": 0.00019632721202003339,
      "loss": 1.1714,
      "num_input_tokens_seen": 174954,
      "step": 61,
      "train_runtime": 149.0535,
      "train_tokens_per_second": 1173.767
    },
    {
      "epoch": 0.004982020530745897,
      "grad_norm": 0.44787871837615967,
      "learning_rate": 0.00019626043405676129,
      "loss": 1.1095,
      "num_input_tokens_seen": 177472,
      "step": 62,
      "train_runtime": 151.142,
      "train_tokens_per_second": 1174.207
    },
    {
      "epoch": 0.005062375700596637,
      "grad_norm": 0.27905064821243286,
      "learning_rate": 0.00019619365609348916,
      "loss": 1.1697,
      "num_input_tokens_seen": 180142,
      "step": 63,
      "train_runtime": 153.2885,
      "train_tokens_per_second": 1175.183
    },
    {
      "epoch": 0.005142730870447377,
      "grad_norm": 0.37700754404067993,
      "learning_rate": 0.00019612687813021703,
      "loss": 0.8613,
      "num_input_tokens_seen": 182542,
      "step": 64,
      "train_runtime": 155.2662,
      "train_tokens_per_second": 1175.671
    },
    {
      "epoch": 0.005223086040298117,
      "grad_norm": 0.38357803225517273,
      "learning_rate": 0.0001960601001669449,
      "loss": 1.415,
      "num_input_tokens_seen": 185674,
      "step": 65,
      "train_runtime": 157.8417,
      "train_tokens_per_second": 1176.33
    },
    {
      "epoch": 0.005303441210148858,
      "grad_norm": 0.331096351146698,
      "learning_rate": 0.00019599332220367278,
      "loss": 1.138,
      "num_input_tokens_seen": 188782,
      "step": 66,
      "train_runtime": 160.1648,
      "train_tokens_per_second": 1178.673
    },
    {
      "epoch": 0.005383796379999598,
      "grad_norm": 0.30474039912223816,
      "learning_rate": 0.00019592654424040068,
      "loss": 0.9835,
      "num_input_tokens_seen": 191548,
      "step": 67,
      "train_runtime": 162.3347,
      "train_tokens_per_second": 1179.957
    },
    {
      "epoch": 0.005464151549850338,
      "grad_norm": 0.3312320113182068,
      "learning_rate": 0.00019585976627712855,
      "loss": 1.3151,
      "num_input_tokens_seen": 193904,
      "step": 68,
      "train_runtime": 164.3436,
      "train_tokens_per_second": 1179.87
    },
    {
      "epoch": 0.0055445067197010785,
      "grad_norm": 0.33529728651046753,
      "learning_rate": 0.00019579298831385645,
      "loss": 0.7548,
      "num_input_tokens_seen": 195936,
      "step": 69,
      "train_runtime": 166.103,
      "train_tokens_per_second": 1179.605
    },
    {
      "epoch": 0.0056248618895518195,
      "grad_norm": 0.4673730432987213,
      "learning_rate": 0.00019572621035058433,
      "loss": 1.0031,
      "num_input_tokens_seen": 198776,
      "step": 70,
      "train_runtime": 168.4298,
      "train_tokens_per_second": 1180.171
    },
    {
      "epoch": 0.0057052170594025595,
      "grad_norm": 0.3669169247150421,
      "learning_rate": 0.0001956594323873122,
      "loss": 1.3102,
      "num_input_tokens_seen": 200468,
      "step": 71,
      "train_runtime": 170.1819,
      "train_tokens_per_second": 1177.963
    },
    {
      "epoch": 0.0057855722292533,
      "grad_norm": 0.32449740171432495,
      "learning_rate": 0.00019559265442404007,
      "loss": 1.3044,
      "num_input_tokens_seen": 202402,
      "step": 72,
      "train_runtime": 171.9566,
      "train_tokens_per_second": 1177.053
    },
    {
      "epoch": 0.00586592739910404,
      "grad_norm": 0.317749947309494,
      "learning_rate": 0.00019552587646076797,
      "loss": 0.9175,
      "num_input_tokens_seen": 204974,
      "step": 73,
      "train_runtime": 174.1032,
      "train_tokens_per_second": 1177.313
    },
    {
      "epoch": 0.00594628256895478,
      "grad_norm": 0.2952556312084198,
      "learning_rate": 0.00019545909849749584,
      "loss": 1.0741,
      "num_input_tokens_seen": 207182,
      "step": 74,
      "train_runtime": 176.0295,
      "train_tokens_per_second": 1176.973
    },
    {
      "epoch": 0.006026637738805521,
      "grad_norm": 0.3129689693450928,
      "learning_rate": 0.00019539232053422372,
      "loss": 1.3219,
      "num_input_tokens_seen": 209668,
      "step": 75,
      "train_runtime": 178.1761,
      "train_tokens_per_second": 1176.746
    },
    {
      "epoch": 0.006106992908656261,
      "grad_norm": 0.30917924642562866,
      "learning_rate": 0.0001953255425709516,
      "loss": 0.9812,
      "num_input_tokens_seen": 211166,
      "step": 76,
      "train_runtime": 179.7681,
      "train_tokens_per_second": 1174.658
    },
    {
      "epoch": 0.006187348078507001,
      "grad_norm": 0.3077734708786011,
      "learning_rate": 0.00019525876460767946,
      "loss": 0.8593,
      "num_input_tokens_seen": 213850,
      "step": 77,
      "train_runtime": 181.9352,
      "train_tokens_per_second": 1175.419
    },
    {
      "epoch": 0.006267703248357741,
      "grad_norm": 0.32136771082878113,
      "learning_rate": 0.00019519198664440736,
      "loss": 0.8924,
      "num_input_tokens_seen": 215464,
      "step": 78,
      "train_runtime": 183.5224,
      "train_tokens_per_second": 1174.048
    },
    {
      "epoch": 0.006348058418208482,
      "grad_norm": 0.3433220684528351,
      "learning_rate": 0.00019512520868113524,
      "loss": 1.2776,
      "num_input_tokens_seen": 217552,
      "step": 79,
      "train_runtime": 185.4837,
      "train_tokens_per_second": 1172.89
    },
    {
      "epoch": 0.006428413588059222,
      "grad_norm": 0.2551766037940979,
      "learning_rate": 0.0001950584307178631,
      "loss": 0.9551,
      "num_input_tokens_seen": 219780,
      "step": 80,
      "train_runtime": 187.3526,
      "train_tokens_per_second": 1173.082
    },
    {
      "epoch": 0.006508768757909962,
      "grad_norm": 0.32899704575538635,
      "learning_rate": 0.00019499165275459098,
      "loss": 1.1009,
      "num_input_tokens_seen": 221666,
      "step": 81,
      "train_runtime": 189.228,
      "train_tokens_per_second": 1171.423
    },
    {
      "epoch": 0.006589123927760702,
      "grad_norm": 0.2957477271556854,
      "learning_rate": 0.00019492487479131886,
      "loss": 1.0102,
      "num_input_tokens_seen": 224776,
      "step": 82,
      "train_runtime": 191.6613,
      "train_tokens_per_second": 1172.777
    },
    {
      "epoch": 0.006669479097611442,
      "grad_norm": 0.3482891321182251,
      "learning_rate": 0.00019485809682804673,
      "loss": 1.2924,
      "num_input_tokens_seen": 226698,
      "step": 83,
      "train_runtime": 193.4219,
      "train_tokens_per_second": 1172.039
    },
    {
      "epoch": 0.006749834267462183,
      "grad_norm": 0.3823198080062866,
      "learning_rate": 0.00019479131886477463,
      "loss": 1.0649,
      "num_input_tokens_seen": 228926,
      "step": 84,
      "train_runtime": 195.3491,
      "train_tokens_per_second": 1171.881
    },
    {
      "epoch": 0.006830189437312923,
      "grad_norm": 0.35176682472229004,
      "learning_rate": 0.0001947245409015025,
      "loss": 0.7604,
      "num_input_tokens_seen": 230892,
      "step": 85,
      "train_runtime": 197.2893,
      "train_tokens_per_second": 1170.322
    },
    {
      "epoch": 0.006910544607163663,
      "grad_norm": 0.31647080183029175,
      "learning_rate": 0.0001946577629382304,
      "loss": 1.1499,
      "num_input_tokens_seen": 232544,
      "step": 86,
      "train_runtime": 198.9968,
      "train_tokens_per_second": 1168.582
    },
    {
      "epoch": 0.006990899777014403,
      "grad_norm": 0.32222849130630493,
      "learning_rate": 0.00019459098497495828,
      "loss": 1.3333,
      "num_input_tokens_seen": 235528,
      "step": 87,
      "train_runtime": 201.3497,
      "train_tokens_per_second": 1169.746
    },
    {
      "epoch": 0.007071254946865144,
      "grad_norm": 0.24119481444358826,
      "learning_rate": 0.00019452420701168615,
      "loss": 0.8085,
      "num_input_tokens_seen": 241730,
      "step": 88,
      "train_runtime": 205.582,
      "train_tokens_per_second": 1175.832
    },
    {
      "epoch": 0.0071516101167158844,
      "grad_norm": 0.32314327359199524,
      "learning_rate": 0.00019445742904841405,
      "loss": 1.0329,
      "num_input_tokens_seen": 246178,
      "step": 89,
      "train_runtime": 208.8028,
      "train_tokens_per_second": 1178.997
    },
    {
      "epoch": 0.0072319652865666245,
      "grad_norm": 0.33118972182273865,
      "learning_rate": 0.00019439065108514192,
      "loss": 0.7723,
      "num_input_tokens_seen": 248024,
      "step": 90,
      "train_runtime": 210.5811,
      "train_tokens_per_second": 1177.807
    },
    {
      "epoch": 0.007312320456417365,
      "grad_norm": 0.31792575120925903,
      "learning_rate": 0.0001943238731218698,
      "loss": 1.2614,
      "num_input_tokens_seen": 250440,
      "step": 91,
      "train_runtime": 214.4372,
      "train_tokens_per_second": 1167.894
    },
    {
      "epoch": 0.007392675626268105,
      "grad_norm": 0.2944311499595642,
      "learning_rate": 0.00019425709515859767,
      "loss": 1.1398,
      "num_input_tokens_seen": 252556,
      "step": 92,
      "train_runtime": 216.2752,
      "train_tokens_per_second": 1167.753
    },
    {
      "epoch": 0.007473030796118846,
      "grad_norm": 0.36758142709732056,
      "learning_rate": 0.00019419031719532554,
      "loss": 1.2504,
      "num_input_tokens_seen": 254592,
      "step": 93,
      "train_runtime": 218.1443,
      "train_tokens_per_second": 1167.081
    },
    {
      "epoch": 0.007553385965969586,
      "grad_norm": 0.36635059118270874,
      "learning_rate": 0.00019412353923205344,
      "loss": 1.2812,
      "num_input_tokens_seen": 259620,
      "step": 94,
      "train_runtime": 221.5746,
      "train_tokens_per_second": 1171.705
    },
    {
      "epoch": 0.007633741135820326,
      "grad_norm": 0.3519682288169861,
      "learning_rate": 0.00019405676126878132,
      "loss": 0.7904,
      "num_input_tokens_seen": 265416,
      "step": 95,
      "train_runtime": 225.4702,
      "train_tokens_per_second": 1177.167
    },
    {
      "epoch": 0.007714096305671066,
      "grad_norm": 0.3626551926136017,
      "learning_rate": 0.0001939899833055092,
      "loss": 1.0009,
      "num_input_tokens_seen": 267690,
      "step": 96,
      "train_runtime": 227.3784,
      "train_tokens_per_second": 1177.288
    },
    {
      "epoch": 0.007794451475521807,
      "grad_norm": 0.3471205234527588,
      "learning_rate": 0.00019392320534223706,
      "loss": 0.7001,
      "num_input_tokens_seen": 269884,
      "step": 97,
      "train_runtime": 229.3942,
      "train_tokens_per_second": 1176.507
    },
    {
      "epoch": 0.007874806645372547,
      "grad_norm": 0.34355437755584717,
      "learning_rate": 0.00019385642737896494,
      "loss": 1.0947,
      "num_input_tokens_seen": 272076,
      "step": 98,
      "train_runtime": 231.2879,
      "train_tokens_per_second": 1176.352
    },
    {
      "epoch": 0.007955161815223287,
      "grad_norm": 0.41802075505256653,
      "learning_rate": 0.0001937896494156928,
      "loss": 1.0463,
      "num_input_tokens_seen": 273562,
      "step": 99,
      "train_runtime": 232.8923,
      "train_tokens_per_second": 1174.629
    },
    {
      "epoch": 0.008035516985074027,
      "grad_norm": 0.30516642332077026,
      "learning_rate": 0.0001937228714524207,
      "loss": 0.9816,
      "num_input_tokens_seen": 277596,
      "step": 100,
      "train_runtime": 235.8828,
      "train_tokens_per_second": 1176.839
    },
    {
      "epoch": 0.008115872154924767,
      "grad_norm": 0.31250709295272827,
      "learning_rate": 0.00019365609348914858,
      "loss": 1.2959,
      "num_input_tokens_seen": 279602,
      "step": 101,
      "train_runtime": 237.637,
      "train_tokens_per_second": 1176.593
    },
    {
      "epoch": 0.008196227324775507,
      "grad_norm": 0.28079161047935486,
      "learning_rate": 0.00019358931552587646,
      "loss": 0.9055,
      "num_input_tokens_seen": 282844,
      "step": 102,
      "train_runtime": 240.1527,
      "train_tokens_per_second": 1177.767
    },
    {
      "epoch": 0.008276582494626247,
      "grad_norm": 0.32835954427719116,
      "learning_rate": 0.00019352253756260436,
      "loss": 1.2387,
      "num_input_tokens_seen": 285008,
      "step": 103,
      "train_runtime": 242.1092,
      "train_tokens_per_second": 1177.188
    },
    {
      "epoch": 0.008356937664476989,
      "grad_norm": 0.36610671877861023,
      "learning_rate": 0.00019345575959933223,
      "loss": 1.1193,
      "num_input_tokens_seen": 286942,
      "step": 104,
      "train_runtime": 243.8693,
      "train_tokens_per_second": 1176.622
    },
    {
      "epoch": 0.00843729283432773,
      "grad_norm": 0.3345883786678314,
      "learning_rate": 0.00019338898163606013,
      "loss": 1.1626,
      "num_input_tokens_seen": 289080,
      "step": 105,
      "train_runtime": 245.8204,
      "train_tokens_per_second": 1175.981
    },
    {
      "epoch": 0.00851764800417847,
      "grad_norm": 0.2844889461994171,
      "learning_rate": 0.000193322203672788,
      "loss": 1.0387,
      "num_input_tokens_seen": 291750,
      "step": 106,
      "train_runtime": 247.9457,
      "train_tokens_per_second": 1176.669
    },
    {
      "epoch": 0.00859800317402921,
      "grad_norm": 0.3191639482975006,
      "learning_rate": 0.00019325542570951588,
      "loss": 1.1567,
      "num_input_tokens_seen": 293930,
      "step": 107,
      "train_runtime": 249.8932,
      "train_tokens_per_second": 1176.223
    },
    {
      "epoch": 0.00867835834387995,
      "grad_norm": 0.34993913769721985,
      "learning_rate": 0.00019318864774624375,
      "loss": 1.6206,
      "num_input_tokens_seen": 297240,
      "step": 108,
      "train_runtime": 252.2674,
      "train_tokens_per_second": 1178.273
    },
    {
      "epoch": 0.00875871351373069,
      "grad_norm": 0.3329659104347229,
      "learning_rate": 0.00019312186978297162,
      "loss": 0.8939,
      "num_input_tokens_seen": 300264,
      "step": 109,
      "train_runtime": 254.5719,
      "train_tokens_per_second": 1179.486
    },
    {
      "epoch": 0.00883906868358143,
      "grad_norm": 0.44520723819732666,
      "learning_rate": 0.00019305509181969952,
      "loss": 0.9183,
      "num_input_tokens_seen": 304144,
      "step": 110,
      "train_runtime": 257.3764,
      "train_tokens_per_second": 1181.709
    },
    {
      "epoch": 0.00891942385343217,
      "grad_norm": 0.364200621843338,
      "learning_rate": 0.0001929883138564274,
      "loss": 1.2748,
      "num_input_tokens_seen": 306332,
      "step": 111,
      "train_runtime": 259.1637,
      "train_tokens_per_second": 1182.002
    },
    {
      "epoch": 0.00899977902328291,
      "grad_norm": 0.3243260085582733,
      "learning_rate": 0.00019292153589315527,
      "loss": 1.3284,
      "num_input_tokens_seen": 309172,
      "step": 112,
      "train_runtime": 261.3851,
      "train_tokens_per_second": 1182.822
    },
    {
      "epoch": 0.009080134193133652,
      "grad_norm": 0.3614398241043091,
      "learning_rate": 0.00019285475792988314,
      "loss": 1.1274,
      "num_input_tokens_seen": 311626,
      "step": 113,
      "train_runtime": 263.4309,
      "train_tokens_per_second": 1182.952
    },
    {
      "epoch": 0.009160489362984392,
      "grad_norm": 0.36545294523239136,
      "learning_rate": 0.00019278797996661101,
      "loss": 1.3002,
      "num_input_tokens_seen": 313606,
      "step": 114,
      "train_runtime": 265.2996,
      "train_tokens_per_second": 1182.082
    },
    {
      "epoch": 0.009240844532835132,
      "grad_norm": 0.2833547294139862,
      "learning_rate": 0.0001927212020033389,
      "loss": 0.9341,
      "num_input_tokens_seen": 317028,
      "step": 115,
      "train_runtime": 267.9413,
      "train_tokens_per_second": 1183.199
    },
    {
      "epoch": 0.009321199702685872,
      "grad_norm": 0.31258538365364075,
      "learning_rate": 0.0001926544240400668,
      "loss": 1.0246,
      "num_input_tokens_seen": 319196,
      "step": 116,
      "train_runtime": 269.816,
      "train_tokens_per_second": 1183.014
    },
    {
      "epoch": 0.009401554872536612,
      "grad_norm": 0.29805701971054077,
      "learning_rate": 0.00019258764607679466,
      "loss": 0.8631,
      "num_input_tokens_seen": 321744,
      "step": 117,
      "train_runtime": 271.9051,
      "train_tokens_per_second": 1183.295
    },
    {
      "epoch": 0.009481910042387352,
      "grad_norm": 0.31408268213272095,
      "learning_rate": 0.00019252086811352253,
      "loss": 1.2062,
      "num_input_tokens_seen": 323606,
      "step": 118,
      "train_runtime": 273.5706,
      "train_tokens_per_second": 1182.897
    },
    {
      "epoch": 0.009562265212238092,
      "grad_norm": 0.31819990277290344,
      "learning_rate": 0.0001924540901502504,
      "loss": 0.8167,
      "num_input_tokens_seen": 325606,
      "step": 119,
      "train_runtime": 275.3904,
      "train_tokens_per_second": 1182.343
    },
    {
      "epoch": 0.009642620382088832,
      "grad_norm": 0.45045605301856995,
      "learning_rate": 0.0001923873121869783,
      "loss": 1.0418,
      "num_input_tokens_seen": 328614,
      "step": 120,
      "train_runtime": 277.709,
      "train_tokens_per_second": 1183.303
    },
    {
      "epoch": 0.009722975551939572,
      "grad_norm": 0.4011978507041931,
      "learning_rate": 0.00019232053422370618,
      "loss": 1.2648,
      "num_input_tokens_seen": 329944,
      "step": 121,
      "train_runtime": 280.8675,
      "train_tokens_per_second": 1174.732
    },
    {
      "epoch": 0.009803330721790314,
      "grad_norm": 0.31260213255882263,
      "learning_rate": 0.00019225375626043408,
      "loss": 0.8947,
      "num_input_tokens_seen": 332712,
      "step": 122,
      "train_runtime": 283.0597,
      "train_tokens_per_second": 1175.413
    },
    {
      "epoch": 0.009883685891641054,
      "grad_norm": 0.26335906982421875,
      "learning_rate": 0.00019218697829716195,
      "loss": 1.1746,
      "num_input_tokens_seen": 335566,
      "step": 123,
      "train_runtime": 285.3939,
      "train_tokens_per_second": 1175.8
    },
    {
      "epoch": 0.009964041061491794,
      "grad_norm": 0.341661661863327,
      "learning_rate": 0.00019212020033388983,
      "loss": 0.9389,
      "num_input_tokens_seen": 337592,
      "step": 124,
      "train_runtime": 287.21,
      "train_tokens_per_second": 1175.419
    },
    {
      "epoch": 0.010044396231342534,
      "grad_norm": 0.2743532061576843,
      "learning_rate": 0.0001920534223706177,
      "loss": 1.0696,
      "num_input_tokens_seen": 342232,
      "step": 125,
      "train_runtime": 290.5217,
      "train_tokens_per_second": 1177.991
    },
    {
      "epoch": 0.010124751401193274,
      "grad_norm": 0.32186877727508545,
      "learning_rate": 0.0001919866444073456,
      "loss": 1.1878,
      "num_input_tokens_seen": 348666,
      "step": 126,
      "train_runtime": 294.8735,
      "train_tokens_per_second": 1182.426
    },
    {
      "epoch": 0.010205106571044014,
      "grad_norm": 0.32328155636787415,
      "learning_rate": 0.00019191986644407347,
      "loss": 1.1599,
      "num_input_tokens_seen": 351168,
      "step": 127,
      "train_runtime": 296.876,
      "train_tokens_per_second": 1182.878
    },
    {
      "epoch": 0.010285461740894754,
      "grad_norm": 0.24393251538276672,
      "learning_rate": 0.00019185308848080135,
      "loss": 0.9055,
      "num_input_tokens_seen": 353856,
      "step": 128,
      "train_runtime": 299.0515,
      "train_tokens_per_second": 1183.261
    },
    {
      "epoch": 0.010365816910745495,
      "grad_norm": 0.3262992203235626,
      "learning_rate": 0.00019178631051752922,
      "loss": 1.0087,
      "num_input_tokens_seen": 356904,
      "step": 129,
      "train_runtime": 301.5106,
      "train_tokens_per_second": 1183.719
    },
    {
      "epoch": 0.010446172080596235,
      "grad_norm": 0.3335343599319458,
      "learning_rate": 0.0001917195325542571,
      "loss": 0.7588,
      "num_input_tokens_seen": 358514,
      "step": 130,
      "train_runtime": 303.1449,
      "train_tokens_per_second": 1182.649
    },
    {
      "epoch": 0.010526527250446976,
      "grad_norm": 0.3395357131958008,
      "learning_rate": 0.00019165275459098497,
      "loss": 0.9075,
      "num_input_tokens_seen": 361170,
      "step": 131,
      "train_runtime": 305.3375,
      "train_tokens_per_second": 1182.855
    },
    {
      "epoch": 0.010606882420297717,
      "grad_norm": 0.2968093752861023,
      "learning_rate": 0.00019158597662771287,
      "loss": 1.2704,
      "num_input_tokens_seen": 364662,
      "step": 132,
      "train_runtime": 307.9228,
      "train_tokens_per_second": 1184.264
    },
    {
      "epoch": 0.010687237590148457,
      "grad_norm": 0.28696516156196594,
      "learning_rate": 0.00019151919866444074,
      "loss": 0.974,
      "num_input_tokens_seen": 366440,
      "step": 133,
      "train_runtime": 309.6872,
      "train_tokens_per_second": 1183.259
    },
    {
      "epoch": 0.010767592759999197,
      "grad_norm": 0.39053821563720703,
      "learning_rate": 0.0001914524207011686,
      "loss": 1.1736,
      "num_input_tokens_seen": 369316,
      "step": 134,
      "train_runtime": 311.9466,
      "train_tokens_per_second": 1183.908
    },
    {
      "epoch": 0.010847947929849937,
      "grad_norm": 0.2857292890548706,
      "learning_rate": 0.0001913856427378965,
      "loss": 1.2166,
      "num_input_tokens_seen": 374450,
      "step": 135,
      "train_runtime": 315.4388,
      "train_tokens_per_second": 1187.076
    },
    {
      "epoch": 0.010928303099700677,
      "grad_norm": 0.31891822814941406,
      "learning_rate": 0.0001913188647746244,
      "loss": 1.1599,
      "num_input_tokens_seen": 376760,
      "step": 136,
      "train_runtime": 317.4662,
      "train_tokens_per_second": 1186.772
    },
    {
      "epoch": 0.011008658269551417,
      "grad_norm": 0.38524264097213745,
      "learning_rate": 0.00019125208681135226,
      "loss": 0.9643,
      "num_input_tokens_seen": 383920,
      "step": 137,
      "train_runtime": 322.2268,
      "train_tokens_per_second": 1191.459
    },
    {
      "epoch": 0.011089013439402157,
      "grad_norm": 0.27233901619911194,
      "learning_rate": 0.00019118530884808016,
      "loss": 0.9706,
      "num_input_tokens_seen": 387242,
      "step": 138,
      "train_runtime": 324.7669,
      "train_tokens_per_second": 1192.369
    },
    {
      "epoch": 0.011169368609252897,
      "grad_norm": 0.31465741991996765,
      "learning_rate": 0.00019111853088480803,
      "loss": 0.817,
      "num_input_tokens_seen": 389096,
      "step": 139,
      "train_runtime": 326.4619,
      "train_tokens_per_second": 1191.857
    },
    {
      "epoch": 0.011249723779103639,
      "grad_norm": 0.33668479323387146,
      "learning_rate": 0.0001910517529215359,
      "loss": 0.9297,
      "num_input_tokens_seen": 390466,
      "step": 140,
      "train_runtime": 328.0146,
      "train_tokens_per_second": 1190.392
    },
    {
      "epoch": 0.011330078948954379,
      "grad_norm": 0.3122538924217224,
      "learning_rate": 0.00019098497495826378,
      "loss": 1.0915,
      "num_input_tokens_seen": 392606,
      "step": 141,
      "train_runtime": 329.8811,
      "train_tokens_per_second": 1190.144
    },
    {
      "epoch": 0.011410434118805119,
      "grad_norm": 0.30943194031715393,
      "learning_rate": 0.00019091819699499168,
      "loss": 1.1691,
      "num_input_tokens_seen": 396696,
      "step": 142,
      "train_runtime": 332.8156,
      "train_tokens_per_second": 1191.939
    },
    {
      "epoch": 0.01149078928865586,
      "grad_norm": 0.3253512680530548,
      "learning_rate": 0.00019085141903171955,
      "loss": 1.3012,
      "num_input_tokens_seen": 401946,
      "step": 143,
      "train_runtime": 336.627,
      "train_tokens_per_second": 1194.04
    },
    {
      "epoch": 0.0115711444585066,
      "grad_norm": 0.25847896933555603,
      "learning_rate": 0.00019078464106844743,
      "loss": 0.9076,
      "num_input_tokens_seen": 404700,
      "step": 144,
      "train_runtime": 338.7673,
      "train_tokens_per_second": 1194.625
    },
    {
      "epoch": 0.01165149962835734,
      "grad_norm": 0.3235805034637451,
      "learning_rate": 0.0001907178631051753,
      "loss": 0.8834,
      "num_input_tokens_seen": 408034,
      "step": 145,
      "train_runtime": 341.3179,
      "train_tokens_per_second": 1195.466
    },
    {
      "epoch": 0.01173185479820808,
      "grad_norm": 0.3012896776199341,
      "learning_rate": 0.00019065108514190317,
      "loss": 0.8598,
      "num_input_tokens_seen": 410702,
      "step": 146,
      "train_runtime": 343.4148,
      "train_tokens_per_second": 1195.936
    },
    {
      "epoch": 0.01181220996805882,
      "grad_norm": 0.2586335837841034,
      "learning_rate": 0.00019058430717863107,
      "loss": 0.8836,
      "num_input_tokens_seen": 413772,
      "step": 147,
      "train_runtime": 345.8469,
      "train_tokens_per_second": 1196.402
    },
    {
      "epoch": 0.01189256513790956,
      "grad_norm": 0.32044851779937744,
      "learning_rate": 0.00019051752921535895,
      "loss": 1.0863,
      "num_input_tokens_seen": 416638,
      "step": 148,
      "train_runtime": 348.0787,
      "train_tokens_per_second": 1196.965
    },
    {
      "epoch": 0.011972920307760301,
      "grad_norm": 0.2788418233394623,
      "learning_rate": 0.00019045075125208682,
      "loss": 1.0831,
      "num_input_tokens_seen": 419928,
      "step": 149,
      "train_runtime": 350.5712,
      "train_tokens_per_second": 1197.839
    },
    {
      "epoch": 0.012053275477611041,
      "grad_norm": 0.30677565932273865,
      "learning_rate": 0.0001903839732888147,
      "loss": 0.9189,
      "num_input_tokens_seen": 421844,
      "step": 150,
      "train_runtime": 352.38,
      "train_tokens_per_second": 1197.128
    },
    {
      "epoch": 0.012133630647461782,
      "grad_norm": 0.32193708419799805,
      "learning_rate": 0.00019031719532554257,
      "loss": 1.1098,
      "num_input_tokens_seen": 427896,
      "step": 151,
      "train_runtime": 358.4671,
      "train_tokens_per_second": 1193.683
    },
    {
      "epoch": 0.012213985817312522,
      "grad_norm": 0.3394554555416107,
      "learning_rate": 0.00019025041736227044,
      "loss": 1.2618,
      "num_input_tokens_seen": 431144,
      "step": 152,
      "train_runtime": 360.9563,
      "train_tokens_per_second": 1194.449
    },
    {
      "epoch": 0.012294340987163262,
      "grad_norm": 0.2948708236217499,
      "learning_rate": 0.00019018363939899834,
      "loss": 1.1895,
      "num_input_tokens_seen": 434692,
      "step": 153,
      "train_runtime": 363.6825,
      "train_tokens_per_second": 1195.252
    },
    {
      "epoch": 0.012374696157014002,
      "grad_norm": 0.29126468300819397,
      "learning_rate": 0.0001901168614357262,
      "loss": 1.0718,
      "num_input_tokens_seen": 437218,
      "step": 154,
      "train_runtime": 365.779,
      "train_tokens_per_second": 1195.306
    },
    {
      "epoch": 0.012455051326864742,
      "grad_norm": 0.3046357333660126,
      "learning_rate": 0.0001900500834724541,
      "loss": 1.0429,
      "num_input_tokens_seen": 438958,
      "step": 155,
      "train_runtime": 367.4695,
      "train_tokens_per_second": 1194.543
    },
    {
      "epoch": 0.012535406496715482,
      "grad_norm": 0.32712283730506897,
      "learning_rate": 0.00018998330550918199,
      "loss": 1.1488,
      "num_input_tokens_seen": 441698,
      "step": 156,
      "train_runtime": 369.7698,
      "train_tokens_per_second": 1194.522
    },
    {
      "epoch": 0.012615761666566222,
      "grad_norm": 0.2856847047805786,
      "learning_rate": 0.00018991652754590986,
      "loss": 0.813,
      "num_input_tokens_seen": 444172,
      "step": 157,
      "train_runtime": 371.8292,
      "train_tokens_per_second": 1194.559
    },
    {
      "epoch": 0.012696116836416964,
      "grad_norm": 0.2785474359989166,
      "learning_rate": 0.00018984974958263776,
      "loss": 0.9779,
      "num_input_tokens_seen": 447652,
      "step": 158,
      "train_runtime": 374.403,
      "train_tokens_per_second": 1195.642
    },
    {
      "epoch": 0.012776472006267704,
      "grad_norm": 0.3558673858642578,
      "learning_rate": 0.00018978297161936563,
      "loss": 0.8613,
      "num_input_tokens_seen": 449884,
      "step": 159,
      "train_runtime": 376.4077,
      "train_tokens_per_second": 1195.204
    },
    {
      "epoch": 0.012856827176118444,
      "grad_norm": 0.30430328845977783,
      "learning_rate": 0.0001897161936560935,
      "loss": 1.1641,
      "num_input_tokens_seen": 452756,
      "step": 160,
      "train_runtime": 378.6837,
      "train_tokens_per_second": 1195.605
    },
    {
      "epoch": 0.012937182345969184,
      "grad_norm": 0.33249416947364807,
      "learning_rate": 0.00018964941569282138,
      "loss": 1.5722,
      "num_input_tokens_seen": 457040,
      "step": 161,
      "train_runtime": 381.8514,
      "train_tokens_per_second": 1196.905
    },
    {
      "epoch": 0.013017537515819924,
      "grad_norm": 0.33412277698516846,
      "learning_rate": 0.00018958263772954925,
      "loss": 0.9908,
      "num_input_tokens_seen": 459366,
      "step": 162,
      "train_runtime": 383.8328,
      "train_tokens_per_second": 1196.787
    },
    {
      "epoch": 0.013097892685670664,
      "grad_norm": 0.3346627652645111,
      "learning_rate": 0.00018951585976627715,
      "loss": 1.0819,
      "num_input_tokens_seen": 461756,
      "step": 163,
      "train_runtime": 385.8181,
      "train_tokens_per_second": 1196.823
    },
    {
      "epoch": 0.013178247855521404,
      "grad_norm": 0.40962570905685425,
      "learning_rate": 0.00018944908180300502,
      "loss": 0.851,
      "num_input_tokens_seen": 464706,
      "step": 164,
      "train_runtime": 388.1203,
      "train_tokens_per_second": 1197.325
    },
    {
      "epoch": 0.013258603025372144,
      "grad_norm": 0.2635403275489807,
      "learning_rate": 0.0001893823038397329,
      "loss": 1.0716,
      "num_input_tokens_seen": 468260,
      "step": 165,
      "train_runtime": 390.6874,
      "train_tokens_per_second": 1198.554
    },
    {
      "epoch": 0.013338958195222884,
      "grad_norm": 0.367654412984848,
      "learning_rate": 0.00018931552587646077,
      "loss": 0.7616,
      "num_input_tokens_seen": 471848,
      "step": 166,
      "train_runtime": 393.284,
      "train_tokens_per_second": 1199.764
    },
    {
      "epoch": 0.013419313365073626,
      "grad_norm": 0.30167949199676514,
      "learning_rate": 0.00018924874791318864,
      "loss": 1.2558,
      "num_input_tokens_seen": 475074,
      "step": 167,
      "train_runtime": 395.8428,
      "train_tokens_per_second": 1200.158
    },
    {
      "epoch": 0.013499668534924366,
      "grad_norm": 0.365676611661911,
      "learning_rate": 0.00018918196994991652,
      "loss": 0.9429,
      "num_input_tokens_seen": 477110,
      "step": 168,
      "train_runtime": 397.6989,
      "train_tokens_per_second": 1199.676
    },
    {
      "epoch": 0.013580023704775106,
      "grad_norm": 0.26204466819763184,
      "learning_rate": 0.00018911519198664442,
      "loss": 0.9637,
      "num_input_tokens_seen": 479824,
      "step": 169,
      "train_runtime": 399.87,
      "train_tokens_per_second": 1199.95
    },
    {
      "epoch": 0.013660378874625847,
      "grad_norm": 0.3446420133113861,
      "learning_rate": 0.0001890484140233723,
      "loss": 1.0273,
      "num_input_tokens_seen": 481880,
      "step": 170,
      "train_runtime": 401.7023,
      "train_tokens_per_second": 1199.595
    },
    {
      "epoch": 0.013740734044476587,
      "grad_norm": 0.3590293824672699,
      "learning_rate": 0.00018898163606010016,
      "loss": 0.8755,
      "num_input_tokens_seen": 484256,
      "step": 171,
      "train_runtime": 403.7885,
      "train_tokens_per_second": 1199.281
    },
    {
      "epoch": 0.013821089214327327,
      "grad_norm": 0.2834423780441284,
      "learning_rate": 0.00018891485809682806,
      "loss": 0.8606,
      "num_input_tokens_seen": 486474,
      "step": 172,
      "train_runtime": 405.7457,
      "train_tokens_per_second": 1198.963
    },
    {
      "epoch": 0.013901444384178067,
      "grad_norm": 0.3650365471839905,
      "learning_rate": 0.00018884808013355594,
      "loss": 0.9288,
      "num_input_tokens_seen": 488262,
      "step": 173,
      "train_runtime": 407.4692,
      "train_tokens_per_second": 1198.279
    },
    {
      "epoch": 0.013981799554028807,
      "grad_norm": 0.29482585191726685,
      "learning_rate": 0.00018878130217028384,
      "loss": 0.9901,
      "num_input_tokens_seen": 491614,
      "step": 174,
      "train_runtime": 410.0341,
      "train_tokens_per_second": 1198.959
    },
    {
      "epoch": 0.014062154723879547,
      "grad_norm": 0.32307374477386475,
      "learning_rate": 0.0001887145242070117,
      "loss": 1.1529,
      "num_input_tokens_seen": 494446,
      "step": 175,
      "train_runtime": 412.2213,
      "train_tokens_per_second": 1199.468
    },
    {
      "epoch": 0.014142509893730289,
      "grad_norm": 0.3395145833492279,
      "learning_rate": 0.00018864774624373958,
      "loss": 0.9121,
      "num_input_tokens_seen": 496538,
      "step": 176,
      "train_runtime": 414.066,
      "train_tokens_per_second": 1199.176
    },
    {
      "epoch": 0.014222865063581029,
      "grad_norm": 0.31804996728897095,
      "learning_rate": 0.00018858096828046746,
      "loss": 1.0448,
      "num_input_tokens_seen": 498296,
      "step": 177,
      "train_runtime": 415.8093,
      "train_tokens_per_second": 1198.376
    },
    {
      "epoch": 0.014303220233431769,
      "grad_norm": 0.2677456736564636,
      "learning_rate": 0.00018851419031719533,
      "loss": 1.2028,
      "num_input_tokens_seen": 501642,
      "step": 178,
      "train_runtime": 418.2921,
      "train_tokens_per_second": 1199.262
    },
    {
      "epoch": 0.014383575403282509,
      "grad_norm": 0.3364807367324829,
      "learning_rate": 0.00018844741235392323,
      "loss": 1.1838,
      "num_input_tokens_seen": 504436,
      "step": 179,
      "train_runtime": 420.5919,
      "train_tokens_per_second": 1199.348
    },
    {
      "epoch": 0.014463930573133249,
      "grad_norm": 0.2928422689437866,
      "learning_rate": 0.0001883806343906511,
      "loss": 1.1621,
      "num_input_tokens_seen": 507202,
      "step": 180,
      "train_runtime": 422.8354,
      "train_tokens_per_second": 1199.526
    },
    {
      "epoch": 0.014544285742983989,
      "grad_norm": 0.2664826214313507,
      "learning_rate": 0.00018831385642737898,
      "loss": 1.001,
      "num_input_tokens_seen": 509714,
      "step": 181,
      "train_runtime": 426.5843,
      "train_tokens_per_second": 1194.873
    },
    {
      "epoch": 0.01462464091283473,
      "grad_norm": 0.3463704288005829,
      "learning_rate": 0.00018824707846410685,
      "loss": 1.3769,
      "num_input_tokens_seen": 513034,
      "step": 182,
      "train_runtime": 429.0651,
      "train_tokens_per_second": 1195.702
    },
    {
      "epoch": 0.01470499608268547,
      "grad_norm": 0.29105862975120544,
      "learning_rate": 0.00018818030050083472,
      "loss": 1.1422,
      "num_input_tokens_seen": 515310,
      "step": 183,
      "train_runtime": 430.9312,
      "train_tokens_per_second": 1195.806
    },
    {
      "epoch": 0.01478535125253621,
      "grad_norm": 0.3170328736305237,
      "learning_rate": 0.0001881135225375626,
      "loss": 1.0611,
      "num_input_tokens_seen": 516950,
      "step": 184,
      "train_runtime": 432.6182,
      "train_tokens_per_second": 1194.934
    },
    {
      "epoch": 0.014865706422386951,
      "grad_norm": 0.28122666478157043,
      "learning_rate": 0.0001880467445742905,
      "loss": 0.8899,
      "num_input_tokens_seen": 518780,
      "step": 185,
      "train_runtime": 434.3792,
      "train_tokens_per_second": 1194.302
    },
    {
      "epoch": 0.014946061592237691,
      "grad_norm": 0.4011630117893219,
      "learning_rate": 0.00018797996661101837,
      "loss": 1.0401,
      "num_input_tokens_seen": 520284,
      "step": 186,
      "train_runtime": 435.9646,
      "train_tokens_per_second": 1193.409
    },
    {
      "epoch": 0.015026416762088431,
      "grad_norm": 0.3490202724933624,
      "learning_rate": 0.00018791318864774624,
      "loss": 0.9352,
      "num_input_tokens_seen": 522202,
      "step": 187,
      "train_runtime": 437.705,
      "train_tokens_per_second": 1193.045
    },
    {
      "epoch": 0.015106771931939171,
      "grad_norm": 0.3363865911960602,
      "learning_rate": 0.00018784641068447412,
      "loss": 1.0091,
      "num_input_tokens_seen": 525100,
      "step": 188,
      "train_runtime": 439.9542,
      "train_tokens_per_second": 1193.533
    },
    {
      "epoch": 0.015187127101789911,
      "grad_norm": 0.32519596815109253,
      "learning_rate": 0.00018777963272120202,
      "loss": 0.9339,
      "num_input_tokens_seen": 530716,
      "step": 189,
      "train_runtime": 443.8903,
      "train_tokens_per_second": 1195.602
    },
    {
      "epoch": 0.015267482271640652,
      "grad_norm": 0.34321466088294983,
      "learning_rate": 0.0001877128547579299,
      "loss": 0.927,
      "num_input_tokens_seen": 532866,
      "step": 190,
      "train_runtime": 445.7307,
      "train_tokens_per_second": 1195.489
    },
    {
      "epoch": 0.015347837441491392,
      "grad_norm": 0.2840273976325989,
      "learning_rate": 0.0001876460767946578,
      "loss": 0.9089,
      "num_input_tokens_seen": 535414,
      "step": 191,
      "train_runtime": 447.8213,
      "train_tokens_per_second": 1195.597
    },
    {
      "epoch": 0.015428192611342132,
      "grad_norm": 0.3532767593860626,
      "learning_rate": 0.00018757929883138566,
      "loss": 1.1514,
      "num_input_tokens_seen": 536984,
      "step": 192,
      "train_runtime": 449.4759,
      "train_tokens_per_second": 1194.689
    },
    {
      "epoch": 0.015508547781192872,
      "grad_norm": 0.27947568893432617,
      "learning_rate": 0.00018751252086811354,
      "loss": 1.1501,
      "num_input_tokens_seen": 538716,
      "step": 193,
      "train_runtime": 451.1462,
      "train_tokens_per_second": 1194.105
    },
    {
      "epoch": 0.015588902951043614,
      "grad_norm": 0.32376599311828613,
      "learning_rate": 0.0001874457429048414,
      "loss": 0.5927,
      "num_input_tokens_seen": 540080,
      "step": 194,
      "train_runtime": 452.7433,
      "train_tokens_per_second": 1192.905
    },
    {
      "epoch": 0.015669258120894352,
      "grad_norm": 0.28643298149108887,
      "learning_rate": 0.0001873789649415693,
      "loss": 1.0052,
      "num_input_tokens_seen": 541804,
      "step": 195,
      "train_runtime": 454.4569,
      "train_tokens_per_second": 1192.201
    },
    {
      "epoch": 0.015749613290745094,
      "grad_norm": 0.2806437313556671,
      "learning_rate": 0.00018731218697829718,
      "loss": 0.9636,
      "num_input_tokens_seen": 543062,
      "step": 196,
      "train_runtime": 455.9817,
      "train_tokens_per_second": 1190.973
    },
    {
      "epoch": 0.015829968460595832,
      "grad_norm": 0.2808949649333954,
      "learning_rate": 0.00018724540901502506,
      "loss": 1.0987,
      "num_input_tokens_seen": 545946,
      "step": 197,
      "train_runtime": 458.2294,
      "train_tokens_per_second": 1191.425
    },
    {
      "epoch": 0.015910323630446574,
      "grad_norm": 0.27494966983795166,
      "learning_rate": 0.00018717863105175293,
      "loss": 0.9189,
      "num_input_tokens_seen": 548450,
      "step": 198,
      "train_runtime": 460.2307,
      "train_tokens_per_second": 1191.685
    },
    {
      "epoch": 0.015990678800297316,
      "grad_norm": 0.29499420523643494,
      "learning_rate": 0.0001871118530884808,
      "loss": 1.1379,
      "num_input_tokens_seen": 550444,
      "step": 199,
      "train_runtime": 462.0599,
      "train_tokens_per_second": 1191.283
    },
    {
      "epoch": 0.016071033970148054,
      "grad_norm": 0.34383973479270935,
      "learning_rate": 0.00018704507512520868,
      "loss": 0.9679,
      "num_input_tokens_seen": 552642,
      "step": 200,
      "train_runtime": 463.9201,
      "train_tokens_per_second": 1191.244
    },
    {
      "epoch": 0.016151389139998796,
      "grad_norm": 0.351194828748703,
      "learning_rate": 0.00018697829716193658,
      "loss": 1.1505,
      "num_input_tokens_seen": 554544,
      "step": 201,
      "train_runtime": 465.8094,
      "train_tokens_per_second": 1190.495
    },
    {
      "epoch": 0.016231744309849534,
      "grad_norm": 0.32537057995796204,
      "learning_rate": 0.00018691151919866445,
      "loss": 1.2466,
      "num_input_tokens_seen": 557158,
      "step": 202,
      "train_runtime": 467.9495,
      "train_tokens_per_second": 1190.637
    },
    {
      "epoch": 0.016312099479700276,
      "grad_norm": 0.341013103723526,
      "learning_rate": 0.00018684474123539232,
      "loss": 0.8909,
      "num_input_tokens_seen": 558518,
      "step": 203,
      "train_runtime": 469.5307,
      "train_tokens_per_second": 1189.524
    },
    {
      "epoch": 0.016392454649551014,
      "grad_norm": 0.2822452485561371,
      "learning_rate": 0.0001867779632721202,
      "loss": 0.7545,
      "num_input_tokens_seen": 560902,
      "step": 204,
      "train_runtime": 471.5518,
      "train_tokens_per_second": 1189.481
    },
    {
      "epoch": 0.016472809819401756,
      "grad_norm": 0.31220871210098267,
      "learning_rate": 0.00018671118530884807,
      "loss": 0.9268,
      "num_input_tokens_seen": 562980,
      "step": 205,
      "train_runtime": 473.4079,
      "train_tokens_per_second": 1189.207
    },
    {
      "epoch": 0.016553164989252495,
      "grad_norm": 0.33676648139953613,
      "learning_rate": 0.00018664440734557597,
      "loss": 0.9465,
      "num_input_tokens_seen": 565404,
      "step": 206,
      "train_runtime": 475.4152,
      "train_tokens_per_second": 1189.285
    },
    {
      "epoch": 0.016633520159103236,
      "grad_norm": 0.267469197511673,
      "learning_rate": 0.00018657762938230384,
      "loss": 0.8469,
      "num_input_tokens_seen": 568626,
      "step": 207,
      "train_runtime": 477.9041,
      "train_tokens_per_second": 1189.833
    },
    {
      "epoch": 0.016713875328953978,
      "grad_norm": 0.3146427273750305,
      "learning_rate": 0.00018651085141903174,
      "loss": 1.2071,
      "num_input_tokens_seen": 571642,
      "step": 208,
      "train_runtime": 480.2594,
      "train_tokens_per_second": 1190.278
    },
    {
      "epoch": 0.016794230498804717,
      "grad_norm": 0.32051125168800354,
      "learning_rate": 0.00018644407345575962,
      "loss": 1.1623,
      "num_input_tokens_seen": 574268,
      "step": 209,
      "train_runtime": 482.4211,
      "train_tokens_per_second": 1190.387
    },
    {
      "epoch": 0.01687458566865546,
      "grad_norm": 0.3071257174015045,
      "learning_rate": 0.0001863772954924875,
      "loss": 1.1845,
      "num_input_tokens_seen": 577408,
      "step": 210,
      "train_runtime": 484.818,
      "train_tokens_per_second": 1190.979
    },
    {
      "epoch": 0.016954940838506197,
      "grad_norm": 0.4016221761703491,
      "learning_rate": 0.0001863105175292154,
      "loss": 1.075,
      "num_input_tokens_seen": 580100,
      "step": 211,
      "train_runtime": 489.0016,
      "train_tokens_per_second": 1186.295
    },
    {
      "epoch": 0.01703529600835694,
      "grad_norm": 0.2990512549877167,
      "learning_rate": 0.00018624373956594326,
      "loss": 0.9753,
      "num_input_tokens_seen": 583272,
      "step": 212,
      "train_runtime": 491.3423,
      "train_tokens_per_second": 1187.099
    },
    {
      "epoch": 0.017115651178207677,
      "grad_norm": 0.3335602581501007,
      "learning_rate": 0.00018617696160267113,
      "loss": 0.7289,
      "num_input_tokens_seen": 585576,
      "step": 213,
      "train_runtime": 493.2548,
      "train_tokens_per_second": 1187.167
    },
    {
      "epoch": 0.01719600634805842,
      "grad_norm": 0.31806111335754395,
      "learning_rate": 0.000186110183639399,
      "loss": 0.8652,
      "num_input_tokens_seen": 586932,
      "step": 214,
      "train_runtime": 494.7909,
      "train_tokens_per_second": 1186.222
    },
    {
      "epoch": 0.017276361517909157,
      "grad_norm": 0.269388884305954,
      "learning_rate": 0.00018604340567612688,
      "loss": 1.3159,
      "num_input_tokens_seen": 590226,
      "step": 215,
      "train_runtime": 497.2321,
      "train_tokens_per_second": 1187.023
    },
    {
      "epoch": 0.0173567166877599,
      "grad_norm": 0.3230859041213989,
      "learning_rate": 0.00018597662771285475,
      "loss": 1.2652,
      "num_input_tokens_seen": 592584,
      "step": 216,
      "train_runtime": 499.1562,
      "train_tokens_per_second": 1187.171
    },
    {
      "epoch": 0.01743707185761064,
      "grad_norm": 0.3253515064716339,
      "learning_rate": 0.00018590984974958265,
      "loss": 1.0018,
      "num_input_tokens_seen": 594752,
      "step": 217,
      "train_runtime": 500.9867,
      "train_tokens_per_second": 1187.161
    },
    {
      "epoch": 0.01751742702746138,
      "grad_norm": 0.29371151328086853,
      "learning_rate": 0.00018584307178631053,
      "loss": 0.9875,
      "num_input_tokens_seen": 596604,
      "step": 218,
      "train_runtime": 502.7494,
      "train_tokens_per_second": 1186.683
    },
    {
      "epoch": 0.01759778219731212,
      "grad_norm": 0.33814385533332825,
      "learning_rate": 0.0001857762938230384,
      "loss": 1.0848,
      "num_input_tokens_seen": 600334,
      "step": 219,
      "train_runtime": 505.3906,
      "train_tokens_per_second": 1187.861
    },
    {
      "epoch": 0.01767813736716286,
      "grad_norm": 0.26476287841796875,
      "learning_rate": 0.00018570951585976627,
      "loss": 0.9078,
      "num_input_tokens_seen": 603930,
      "step": 220,
      "train_runtime": 508.0612,
      "train_tokens_per_second": 1188.695
    },
    {
      "epoch": 0.0177584925370136,
      "grad_norm": 0.2639309763908386,
      "learning_rate": 0.00018564273789649415,
      "loss": 1.0194,
      "num_input_tokens_seen": 606094,
      "step": 221,
      "train_runtime": 509.9427,
      "train_tokens_per_second": 1188.553
    },
    {
      "epoch": 0.01783884770686434,
      "grad_norm": 0.3658522367477417,
      "learning_rate": 0.00018557595993322205,
      "loss": 1.3372,
      "num_input_tokens_seen": 609150,
      "step": 222,
      "train_runtime": 512.2491,
      "train_tokens_per_second": 1189.168
    },
    {
      "epoch": 0.01791920287671508,
      "grad_norm": 0.2818155288696289,
      "learning_rate": 0.00018550918196994992,
      "loss": 1.0582,
      "num_input_tokens_seen": 611726,
      "step": 223,
      "train_runtime": 514.3235,
      "train_tokens_per_second": 1189.38
    },
    {
      "epoch": 0.01799955804656582,
      "grad_norm": 0.281306654214859,
      "learning_rate": 0.0001854424040066778,
      "loss": 0.9165,
      "num_input_tokens_seen": 613594,
      "step": 224,
      "train_runtime": 516.0901,
      "train_tokens_per_second": 1188.928
    },
    {
      "epoch": 0.01807991321641656,
      "grad_norm": 0.3171606957912445,
      "learning_rate": 0.0001853756260434057,
      "loss": 0.9054,
      "num_input_tokens_seen": 615324,
      "step": 225,
      "train_runtime": 517.8348,
      "train_tokens_per_second": 1188.263
    },
    {
      "epoch": 0.018160268386267303,
      "grad_norm": 0.2836587131023407,
      "learning_rate": 0.00018530884808013357,
      "loss": 1.0919,
      "num_input_tokens_seen": 618166,
      "step": 226,
      "train_runtime": 519.9689,
      "train_tokens_per_second": 1188.852
    },
    {
      "epoch": 0.01824062355611804,
      "grad_norm": 0.29988783597946167,
      "learning_rate": 0.00018524207011686147,
      "loss": 0.7693,
      "num_input_tokens_seen": 620906,
      "step": 227,
      "train_runtime": 522.1826,
      "train_tokens_per_second": 1189.059
    },
    {
      "epoch": 0.018320978725968783,
      "grad_norm": 0.38843727111816406,
      "learning_rate": 0.00018517529215358934,
      "loss": 0.92,
      "num_input_tokens_seen": 623114,
      "step": 228,
      "train_runtime": 524.1536,
      "train_tokens_per_second": 1188.8
    },
    {
      "epoch": 0.01840133389581952,
      "grad_norm": 0.3710363507270813,
      "learning_rate": 0.00018510851419031721,
      "loss": 0.9383,
      "num_input_tokens_seen": 624454,
      "step": 229,
      "train_runtime": 525.6716,
      "train_tokens_per_second": 1187.917
    },
    {
      "epoch": 0.018481689065670263,
      "grad_norm": 0.34894946217536926,
      "learning_rate": 0.0001850417362270451,
      "loss": 1.4647,
      "num_input_tokens_seen": 626452,
      "step": 230,
      "train_runtime": 527.5335,
      "train_tokens_per_second": 1187.511
    },
    {
      "epoch": 0.018562044235521002,
      "grad_norm": 0.3306509554386139,
      "learning_rate": 0.00018497495826377296,
      "loss": 1.0398,
      "num_input_tokens_seen": 630858,
      "step": 231,
      "train_runtime": 530.764,
      "train_tokens_per_second": 1188.585
    },
    {
      "epoch": 0.018642399405371744,
      "grad_norm": 0.3057507276535034,
      "learning_rate": 0.00018490818030050083,
      "loss": 1.0419,
      "num_input_tokens_seen": 633720,
      "step": 232,
      "train_runtime": 533.0,
      "train_tokens_per_second": 1188.968
    },
    {
      "epoch": 0.018722754575222482,
      "grad_norm": 0.3533569872379303,
      "learning_rate": 0.00018484140233722873,
      "loss": 1.3783,
      "num_input_tokens_seen": 635986,
      "step": 233,
      "train_runtime": 535.048,
      "train_tokens_per_second": 1188.652
    },
    {
      "epoch": 0.018803109745073224,
      "grad_norm": 0.33996251225471497,
      "learning_rate": 0.0001847746243739566,
      "loss": 1.0849,
      "num_input_tokens_seen": 637676,
      "step": 234,
      "train_runtime": 536.751,
      "train_tokens_per_second": 1188.03
    },
    {
      "epoch": 0.018883464914923966,
      "grad_norm": 0.2771894335746765,
      "learning_rate": 0.00018470784641068448,
      "loss": 0.9898,
      "num_input_tokens_seen": 640458,
      "step": 235,
      "train_runtime": 538.9257,
      "train_tokens_per_second": 1188.398
    },
    {
      "epoch": 0.018963820084774704,
      "grad_norm": 0.351846843957901,
      "learning_rate": 0.00018464106844741235,
      "loss": 1.2405,
      "num_input_tokens_seen": 642002,
      "step": 236,
      "train_runtime": 540.5796,
      "train_tokens_per_second": 1187.618
    },
    {
      "epoch": 0.019044175254625446,
      "grad_norm": 0.3286455273628235,
      "learning_rate": 0.00018457429048414023,
      "loss": 0.8548,
      "num_input_tokens_seen": 646542,
      "step": 237,
      "train_runtime": 543.7546,
      "train_tokens_per_second": 1189.033
    },
    {
      "epoch": 0.019124530424476184,
      "grad_norm": 0.2703714966773987,
      "learning_rate": 0.00018450751252086813,
      "loss": 0.7789,
      "num_input_tokens_seen": 648010,
      "step": 238,
      "train_runtime": 545.3085,
      "train_tokens_per_second": 1188.336
    },
    {
      "epoch": 0.019204885594326926,
      "grad_norm": 0.3055046498775482,
      "learning_rate": 0.000184440734557596,
      "loss": 1.2842,
      "num_input_tokens_seen": 651002,
      "step": 239,
      "train_runtime": 547.6009,
      "train_tokens_per_second": 1188.826
    },
    {
      "epoch": 0.019285240764177664,
      "grad_norm": 0.2989618480205536,
      "learning_rate": 0.00018437395659432387,
      "loss": 0.7324,
      "num_input_tokens_seen": 654196,
      "step": 240,
      "train_runtime": 550.0684,
      "train_tokens_per_second": 1189.299
    },
    {
      "epoch": 0.019365595934028406,
      "grad_norm": 0.30718129873275757,
      "learning_rate": 0.00018430717863105175,
      "loss": 0.6282,
      "num_input_tokens_seen": 656554,
      "step": 241,
      "train_runtime": 554.0852,
      "train_tokens_per_second": 1184.933
    },
    {
      "epoch": 0.019445951103879144,
      "grad_norm": 0.3032449781894684,
      "learning_rate": 0.00018424040066777965,
      "loss": 0.8998,
      "num_input_tokens_seen": 658716,
      "step": 242,
      "train_runtime": 555.9485,
      "train_tokens_per_second": 1184.851
    },
    {
      "epoch": 0.019526306273729886,
      "grad_norm": 0.36071792244911194,
      "learning_rate": 0.00018417362270450752,
      "loss": 1.0962,
      "num_input_tokens_seen": 660638,
      "step": 243,
      "train_runtime": 557.7411,
      "train_tokens_per_second": 1184.489
    },
    {
      "epoch": 0.019606661443580628,
      "grad_norm": 0.3375242352485657,
      "learning_rate": 0.00018410684474123542,
      "loss": 1.0222,
      "num_input_tokens_seen": 663794,
      "step": 244,
      "train_runtime": 560.31,
      "train_tokens_per_second": 1184.691
    },
    {
      "epoch": 0.019687016613431366,
      "grad_norm": 0.3033391833305359,
      "learning_rate": 0.0001840400667779633,
      "loss": 0.9836,
      "num_input_tokens_seen": 665898,
      "step": 245,
      "train_runtime": 562.1405,
      "train_tokens_per_second": 1184.576
    },
    {
      "epoch": 0.019767371783282108,
      "grad_norm": 0.334856241941452,
      "learning_rate": 0.00018397328881469117,
      "loss": 1.0768,
      "num_input_tokens_seen": 669356,
      "step": 246,
      "train_runtime": 564.7073,
      "train_tokens_per_second": 1185.315
    },
    {
      "epoch": 0.019847726953132847,
      "grad_norm": 0.32017403841018677,
      "learning_rate": 0.00018390651085141904,
      "loss": 0.9337,
      "num_input_tokens_seen": 670858,
      "step": 247,
      "train_runtime": 566.3387,
      "train_tokens_per_second": 1184.553
    },
    {
      "epoch": 0.01992808212298359,
      "grad_norm": 0.28297939896583557,
      "learning_rate": 0.0001838397328881469,
      "loss": 1.06,
      "num_input_tokens_seen": 673118,
      "step": 248,
      "train_runtime": 568.2626,
      "train_tokens_per_second": 1184.519
    },
    {
      "epoch": 0.020008437292834327,
      "grad_norm": 0.31982824206352234,
      "learning_rate": 0.0001837729549248748,
      "loss": 1.0177,
      "num_input_tokens_seen": 675548,
      "step": 249,
      "train_runtime": 570.3117,
      "train_tokens_per_second": 1184.524
    },
    {
      "epoch": 0.02008879246268507,
      "grad_norm": 0.28836411237716675,
      "learning_rate": 0.00018370617696160269,
      "loss": 0.9755,
      "num_input_tokens_seen": 679092,
      "step": 250,
      "train_runtime": 572.8814,
      "train_tokens_per_second": 1185.397
    },
    {
      "epoch": 0.020169147632535807,
      "grad_norm": 0.32343533635139465,
      "learning_rate": 0.00018363939899833056,
      "loss": 1.1175,
      "num_input_tokens_seen": 683370,
      "step": 251,
      "train_runtime": 575.9968,
      "train_tokens_per_second": 1186.413
    },
    {
      "epoch": 0.02024950280238655,
      "grad_norm": 0.25986310839653015,
      "learning_rate": 0.00018357262103505843,
      "loss": 1.0566,
      "num_input_tokens_seen": 686506,
      "step": 252,
      "train_runtime": 578.3604,
      "train_tokens_per_second": 1186.987
    },
    {
      "epoch": 0.02032985797223729,
      "grad_norm": 0.4172917902469635,
      "learning_rate": 0.0001835058430717863,
      "loss": 1.1435,
      "num_input_tokens_seen": 688146,
      "step": 253,
      "train_runtime": 579.9753,
      "train_tokens_per_second": 1186.509
    },
    {
      "epoch": 0.02041021314208803,
      "grad_norm": 0.3311367928981781,
      "learning_rate": 0.0001834390651085142,
      "loss": 1.1981,
      "num_input_tokens_seen": 690826,
      "step": 254,
      "train_runtime": 582.1721,
      "train_tokens_per_second": 1186.635
    },
    {
      "epoch": 0.02049056831193877,
      "grad_norm": 0.32272371649742126,
      "learning_rate": 0.00018337228714524208,
      "loss": 1.0856,
      "num_input_tokens_seen": 696758,
      "step": 255,
      "train_runtime": 586.1914,
      "train_tokens_per_second": 1188.619
    },
    {
      "epoch": 0.02057092348178951,
      "grad_norm": 0.3319406807422638,
      "learning_rate": 0.00018330550918196995,
      "loss": 0.9676,
      "num_input_tokens_seen": 699308,
      "step": 256,
      "train_runtime": 588.2945,
      "train_tokens_per_second": 1188.704
    },
    {
      "epoch": 0.02065127865164025,
      "grad_norm": 0.28314608335494995,
      "learning_rate": 0.00018323873121869782,
      "loss": 1.0579,
      "num_input_tokens_seen": 701988,
      "step": 257,
      "train_runtime": 590.4355,
      "train_tokens_per_second": 1188.933
    },
    {
      "epoch": 0.02073163382149099,
      "grad_norm": 0.3084408938884735,
      "learning_rate": 0.0001831719532554257,
      "loss": 1.0021,
      "num_input_tokens_seen": 705378,
      "step": 258,
      "train_runtime": 593.0866,
      "train_tokens_per_second": 1189.334
    },
    {
      "epoch": 0.02081198899134173,
      "grad_norm": 0.33266472816467285,
      "learning_rate": 0.0001831051752921536,
      "loss": 0.9928,
      "num_input_tokens_seen": 708452,
      "step": 259,
      "train_runtime": 595.4499,
      "train_tokens_per_second": 1189.776
    },
    {
      "epoch": 0.02089234416119247,
      "grad_norm": 0.33789265155792236,
      "learning_rate": 0.00018303839732888147,
      "loss": 1.1864,
      "num_input_tokens_seen": 711596,
      "step": 260,
      "train_runtime": 597.8571,
      "train_tokens_per_second": 1190.244
    },
    {
      "epoch": 0.02097269933104321,
      "grad_norm": 0.39183464646339417,
      "learning_rate": 0.00018297161936560937,
      "loss": 1.1375,
      "num_input_tokens_seen": 713536,
      "step": 261,
      "train_runtime": 599.6624,
      "train_tokens_per_second": 1189.896
    },
    {
      "epoch": 0.021053054500893953,
      "grad_norm": 0.2988393306732178,
      "learning_rate": 0.00018290484140233724,
      "loss": 1.0411,
      "num_input_tokens_seen": 715772,
      "step": 262,
      "train_runtime": 601.5042,
      "train_tokens_per_second": 1189.97
    },
    {
      "epoch": 0.02113340967074469,
      "grad_norm": 0.35692253708839417,
      "learning_rate": 0.00018283806343906512,
      "loss": 1.3377,
      "num_input_tokens_seen": 718176,
      "step": 263,
      "train_runtime": 603.5783,
      "train_tokens_per_second": 1189.864
    },
    {
      "epoch": 0.021213764840595433,
      "grad_norm": 0.3549998998641968,
      "learning_rate": 0.000182771285475793,
      "loss": 1.0436,
      "num_input_tokens_seen": 720824,
      "step": 264,
      "train_runtime": 605.6188,
      "train_tokens_per_second": 1190.227
    },
    {
      "epoch": 0.02129412001044617,
      "grad_norm": 0.39884763956069946,
      "learning_rate": 0.0001827045075125209,
      "loss": 1.0585,
      "num_input_tokens_seen": 722544,
      "step": 265,
      "train_runtime": 607.3114,
      "train_tokens_per_second": 1189.742
    },
    {
      "epoch": 0.021374475180296913,
      "grad_norm": 0.3078259229660034,
      "learning_rate": 0.00018263772954924876,
      "loss": 1.2007,
      "num_input_tokens_seen": 725812,
      "step": 266,
      "train_runtime": 609.7726,
      "train_tokens_per_second": 1190.299
    },
    {
      "epoch": 0.02145483035014765,
      "grad_norm": 0.3153076767921448,
      "learning_rate": 0.00018257095158597664,
      "loss": 0.9561,
      "num_input_tokens_seen": 729636,
      "step": 267,
      "train_runtime": 612.5845,
      "train_tokens_per_second": 1191.078
    },
    {
      "epoch": 0.021535185519998393,
      "grad_norm": 0.3072142004966736,
      "learning_rate": 0.0001825041736227045,
      "loss": 1.1039,
      "num_input_tokens_seen": 731774,
      "step": 268,
      "train_runtime": 614.4373,
      "train_tokens_per_second": 1190.966
    },
    {
      "epoch": 0.021615540689849132,
      "grad_norm": 0.2938554584980011,
      "learning_rate": 0.00018243739565943238,
      "loss": 0.7892,
      "num_input_tokens_seen": 734640,
      "step": 269,
      "train_runtime": 616.8038,
      "train_tokens_per_second": 1191.043
    },
    {
      "epoch": 0.021695895859699874,
      "grad_norm": 0.3175390958786011,
      "learning_rate": 0.00018237061769616028,
      "loss": 1.1768,
      "num_input_tokens_seen": 738162,
      "step": 270,
      "train_runtime": 619.4736,
      "train_tokens_per_second": 1191.596
    },
    {
      "epoch": 0.021776251029550615,
      "grad_norm": 0.359017550945282,
      "learning_rate": 0.00018230383973288816,
      "loss": 1.4663,
      "num_input_tokens_seen": 741614,
      "step": 271,
      "train_runtime": 624.2311,
      "train_tokens_per_second": 1188.044
    },
    {
      "epoch": 0.021856606199401354,
      "grad_norm": 0.31477251648902893,
      "learning_rate": 0.00018223706176961603,
      "loss": 1.1051,
      "num_input_tokens_seen": 744968,
      "step": 272,
      "train_runtime": 626.7102,
      "train_tokens_per_second": 1188.696
    },
    {
      "epoch": 0.021936961369252096,
      "grad_norm": 0.40492677688598633,
      "learning_rate": 0.0001821702838063439,
      "loss": 0.792,
      "num_input_tokens_seen": 746498,
      "step": 273,
      "train_runtime": 628.3129,
      "train_tokens_per_second": 1188.099
    },
    {
      "epoch": 0.022017316539102834,
      "grad_norm": 0.29795950651168823,
      "learning_rate": 0.00018210350584307178,
      "loss": 0.8987,
      "num_input_tokens_seen": 749370,
      "step": 274,
      "train_runtime": 630.5416,
      "train_tokens_per_second": 1188.455
    },
    {
      "epoch": 0.022097671708953576,
      "grad_norm": 0.21813727915287018,
      "learning_rate": 0.00018203672787979968,
      "loss": 0.8571,
      "num_input_tokens_seen": 752410,
      "step": 275,
      "train_runtime": 632.9991,
      "train_tokens_per_second": 1188.643
    },
    {
      "epoch": 0.022178026878804314,
      "grad_norm": 0.46228572726249695,
      "learning_rate": 0.00018196994991652755,
      "loss": 0.8525,
      "num_input_tokens_seen": 756972,
      "step": 276,
      "train_runtime": 636.2182,
      "train_tokens_per_second": 1189.799
    },
    {
      "epoch": 0.022258382048655056,
      "grad_norm": 0.3270361125469208,
      "learning_rate": 0.00018190317195325542,
      "loss": 1.276,
      "num_input_tokens_seen": 759264,
      "step": 277,
      "train_runtime": 638.1625,
      "train_tokens_per_second": 1189.766
    },
    {
      "epoch": 0.022338737218505794,
      "grad_norm": 0.3066183626651764,
      "learning_rate": 0.00018183639398998332,
      "loss": 0.84,
      "num_input_tokens_seen": 762356,
      "step": 278,
      "train_runtime": 640.515,
      "train_tokens_per_second": 1190.223
    },
    {
      "epoch": 0.022419092388356536,
      "grad_norm": 0.3379145860671997,
      "learning_rate": 0.0001817696160267112,
      "loss": 0.845,
      "num_input_tokens_seen": 764308,
      "step": 279,
      "train_runtime": 642.3225,
      "train_tokens_per_second": 1189.913
    },
    {
      "epoch": 0.022499447558207278,
      "grad_norm": 0.2823447287082672,
      "learning_rate": 0.0001817028380634391,
      "loss": 1.3112,
      "num_input_tokens_seen": 766402,
      "step": 280,
      "train_runtime": 644.0831,
      "train_tokens_per_second": 1189.912
    },
    {
      "epoch": 0.022579802728058016,
      "grad_norm": 0.3513202965259552,
      "learning_rate": 0.00018163606010016697,
      "loss": 0.9946,
      "num_input_tokens_seen": 769308,
      "step": 281,
      "train_runtime": 646.3636,
      "train_tokens_per_second": 1190.209
    },
    {
      "epoch": 0.022660157897908758,
      "grad_norm": 0.3789612650871277,
      "learning_rate": 0.00018156928213689484,
      "loss": 1.0793,
      "num_input_tokens_seen": 771698,
      "step": 282,
      "train_runtime": 648.3166,
      "train_tokens_per_second": 1190.31
    },
    {
      "epoch": 0.022740513067759496,
      "grad_norm": 0.3214729130268097,
      "learning_rate": 0.00018150250417362272,
      "loss": 1.2683,
      "num_input_tokens_seen": 778998,
      "step": 283,
      "train_runtime": 653.0352,
      "train_tokens_per_second": 1192.888
    },
    {
      "epoch": 0.022820868237610238,
      "grad_norm": 0.3571290373802185,
      "learning_rate": 0.0001814357262103506,
      "loss": 0.8057,
      "num_input_tokens_seen": 780600,
      "step": 284,
      "train_runtime": 654.6897,
      "train_tokens_per_second": 1192.321
    },
    {
      "epoch": 0.022901223407460976,
      "grad_norm": 0.37717893719673157,
      "learning_rate": 0.00018136894824707846,
      "loss": 1.0761,
      "num_input_tokens_seen": 783222,
      "step": 285,
      "train_runtime": 656.8363,
      "train_tokens_per_second": 1192.416
    },
    {
      "epoch": 0.02298157857731172,
      "grad_norm": 0.3153490424156189,
      "learning_rate": 0.00018130217028380636,
      "loss": 0.9528,
      "num_input_tokens_seen": 786730,
      "step": 286,
      "train_runtime": 659.4083,
      "train_tokens_per_second": 1193.085
    },
    {
      "epoch": 0.023061933747162457,
      "grad_norm": 0.29828205704689026,
      "learning_rate": 0.00018123539232053424,
      "loss": 0.9196,
      "num_input_tokens_seen": 789450,
      "step": 287,
      "train_runtime": 661.5361,
      "train_tokens_per_second": 1193.359
    },
    {
      "epoch": 0.0231422889170132,
      "grad_norm": 0.3584853708744049,
      "learning_rate": 0.0001811686143572621,
      "loss": 0.8471,
      "num_input_tokens_seen": 791610,
      "step": 288,
      "train_runtime": 663.4202,
      "train_tokens_per_second": 1193.226
    },
    {
      "epoch": 0.02322264408686394,
      "grad_norm": 0.33623620867729187,
      "learning_rate": 0.00018110183639398998,
      "loss": 1.0002,
      "num_input_tokens_seen": 794294,
      "step": 289,
      "train_runtime": 665.5696,
      "train_tokens_per_second": 1193.405
    },
    {
      "epoch": 0.02330299925671468,
      "grad_norm": 0.2922380566596985,
      "learning_rate": 0.00018103505843071786,
      "loss": 1.0132,
      "num_input_tokens_seen": 798442,
      "step": 290,
      "train_runtime": 668.5575,
      "train_tokens_per_second": 1194.276
    },
    {
      "epoch": 0.02338335442656542,
      "grad_norm": 0.2733694314956665,
      "learning_rate": 0.00018096828046744576,
      "loss": 1.0303,
      "num_input_tokens_seen": 801948,
      "step": 291,
      "train_runtime": 671.1696,
      "train_tokens_per_second": 1194.852
    },
    {
      "epoch": 0.02346370959641616,
      "grad_norm": 0.3377852141857147,
      "learning_rate": 0.00018090150250417363,
      "loss": 1.1365,
      "num_input_tokens_seen": 803772,
      "step": 292,
      "train_runtime": 672.8719,
      "train_tokens_per_second": 1194.539
    },
    {
      "epoch": 0.0235440647662669,
      "grad_norm": 0.3539382219314575,
      "learning_rate": 0.0001808347245409015,
      "loss": 1.3856,
      "num_input_tokens_seen": 805678,
      "step": 293,
      "train_runtime": 674.6333,
      "train_tokens_per_second": 1194.246
    },
    {
      "epoch": 0.02362441993611764,
      "grad_norm": 0.3829708695411682,
      "learning_rate": 0.00018076794657762938,
      "loss": 1.179,
      "num_input_tokens_seen": 807536,
      "step": 294,
      "train_runtime": 676.3396,
      "train_tokens_per_second": 1193.98
    },
    {
      "epoch": 0.02370477510596838,
      "grad_norm": 0.2746177911758423,
      "learning_rate": 0.00018070116861435728,
      "loss": 0.9804,
      "num_input_tokens_seen": 810270,
      "step": 295,
      "train_runtime": 678.4794,
      "train_tokens_per_second": 1194.244
    },
    {
      "epoch": 0.02378513027581912,
      "grad_norm": 0.2421218901872635,
      "learning_rate": 0.00018063439065108515,
      "loss": 0.8228,
      "num_input_tokens_seen": 812892,
      "step": 296,
      "train_runtime": 680.595,
      "train_tokens_per_second": 1194.384
    },
    {
      "epoch": 0.02386548544566986,
      "grad_norm": 0.3055904507637024,
      "learning_rate": 0.00018056761268781305,
      "loss": 0.9661,
      "num_input_tokens_seen": 815328,
      "step": 297,
      "train_runtime": 682.5564,
      "train_tokens_per_second": 1194.521
    },
    {
      "epoch": 0.023945840615520603,
      "grad_norm": 0.3203045129776001,
      "learning_rate": 0.00018050083472454092,
      "loss": 1.0487,
      "num_input_tokens_seen": 818542,
      "step": 298,
      "train_runtime": 685.0442,
      "train_tokens_per_second": 1194.875
    },
    {
      "epoch": 0.02402619578537134,
      "grad_norm": 0.33606308698654175,
      "learning_rate": 0.0001804340567612688,
      "loss": 0.7003,
      "num_input_tokens_seen": 819804,
      "step": 299,
      "train_runtime": 686.5544,
      "train_tokens_per_second": 1194.085
    },
    {
      "epoch": 0.024106550955222083,
      "grad_norm": 0.35947442054748535,
      "learning_rate": 0.00018036727879799667,
      "loss": 1.0502,
      "num_input_tokens_seen": 822530,
      "step": 300,
      "train_runtime": 688.7295,
      "train_tokens_per_second": 1194.271
    },
    {
      "epoch": 0.024106550955222083,
      "eval_loss": 1.0293227434158325,
      "eval_runtime": 92.507,
      "eval_samples_per_second": 10.745,
      "eval_steps_per_second": 5.373,
      "num_input_tokens_seen": 822530,
      "step": 300
    },
    {
      "epoch": 0.02418690612507282,
      "grad_norm": 0.3234299421310425,
      "learning_rate": 0.00018030050083472454,
      "loss": 0.8288,
      "num_input_tokens_seen": 825642,
      "step": 301,
      "train_runtime": 785.6794,
      "train_tokens_per_second": 1050.864
    },
    {
      "epoch": 0.024267261294923563,
      "grad_norm": 0.3050432801246643,
      "learning_rate": 0.00018023372287145244,
      "loss": 1.1029,
      "num_input_tokens_seen": 829728,
      "step": 302,
      "train_runtime": 788.6291,
      "train_tokens_per_second": 1052.114
    },
    {
      "epoch": 0.0243476164647743,
      "grad_norm": 0.30831417441368103,
      "learning_rate": 0.00018016694490818031,
      "loss": 1.0772,
      "num_input_tokens_seen": 831968,
      "step": 303,
      "train_runtime": 790.5661,
      "train_tokens_per_second": 1052.37
    },
    {
      "epoch": 0.024427971634625043,
      "grad_norm": 0.31089532375335693,
      "learning_rate": 0.0001801001669449082,
      "loss": 0.9571,
      "num_input_tokens_seen": 835942,
      "step": 304,
      "train_runtime": 793.4084,
      "train_tokens_per_second": 1053.609
    },
    {
      "epoch": 0.02450832680447578,
      "grad_norm": 0.2574862837791443,
      "learning_rate": 0.00018003338898163606,
      "loss": 1.0147,
      "num_input_tokens_seen": 838214,
      "step": 305,
      "train_runtime": 795.3659,
      "train_tokens_per_second": 1053.872
    },
    {
      "epoch": 0.024588681974326523,
      "grad_norm": 0.29710450768470764,
      "learning_rate": 0.00017996661101836393,
      "loss": 0.9261,
      "num_input_tokens_seen": 843000,
      "step": 306,
      "train_runtime": 798.8386,
      "train_tokens_per_second": 1055.282
    },
    {
      "epoch": 0.024669037144177265,
      "grad_norm": 0.27685678005218506,
      "learning_rate": 0.00017989983305509183,
      "loss": 1.0539,
      "num_input_tokens_seen": 845588,
      "step": 307,
      "train_runtime": 800.9657,
      "train_tokens_per_second": 1055.711
    },
    {
      "epoch": 0.024749392314028004,
      "grad_norm": 0.2785474956035614,
      "learning_rate": 0.0001798330550918197,
      "loss": 0.7323,
      "num_input_tokens_seen": 848704,
      "step": 308,
      "train_runtime": 803.3106,
      "train_tokens_per_second": 1056.508
    },
    {
      "epoch": 0.024829747483878745,
      "grad_norm": 0.3003632128238678,
      "learning_rate": 0.00017976627712854758,
      "loss": 0.8072,
      "num_input_tokens_seen": 850446,
      "step": 309,
      "train_runtime": 804.9444,
      "train_tokens_per_second": 1056.528
    },
    {
      "epoch": 0.024910102653729484,
      "grad_norm": 0.3557335138320923,
      "learning_rate": 0.00017969949916527545,
      "loss": 1.1393,
      "num_input_tokens_seen": 852032,
      "step": 310,
      "train_runtime": 806.5353,
      "train_tokens_per_second": 1056.41
    },
    {
      "epoch": 0.024990457823580225,
      "grad_norm": 0.28027603030204773,
      "learning_rate": 0.00017963272120200333,
      "loss": 1.0839,
      "num_input_tokens_seen": 855012,
      "step": 311,
      "train_runtime": 808.852,
      "train_tokens_per_second": 1057.069
    },
    {
      "epoch": 0.025070812993430964,
      "grad_norm": 0.3183228373527527,
      "learning_rate": 0.00017956594323873123,
      "loss": 1.2603,
      "num_input_tokens_seen": 858100,
      "step": 312,
      "train_runtime": 811.2805,
      "train_tokens_per_second": 1057.711
    },
    {
      "epoch": 0.025151168163281706,
      "grad_norm": 0.29947277903556824,
      "learning_rate": 0.0001794991652754591,
      "loss": 0.9422,
      "num_input_tokens_seen": 860210,
      "step": 313,
      "train_runtime": 813.1143,
      "train_tokens_per_second": 1057.92
    },
    {
      "epoch": 0.025231523333132444,
      "grad_norm": 0.2728103697299957,
      "learning_rate": 0.000179432387312187,
      "loss": 0.9389,
      "num_input_tokens_seen": 863722,
      "step": 314,
      "train_runtime": 815.7772,
      "train_tokens_per_second": 1058.772
    },
    {
      "epoch": 0.025311878502983186,
      "grad_norm": 0.33725112676620483,
      "learning_rate": 0.00017936560934891487,
      "loss": 1.2485,
      "num_input_tokens_seen": 868400,
      "step": 315,
      "train_runtime": 819.2078,
      "train_tokens_per_second": 1060.049
    },
    {
      "epoch": 0.025392233672833928,
      "grad_norm": 0.25919032096862793,
      "learning_rate": 0.00017929883138564275,
      "loss": 0.9498,
      "num_input_tokens_seen": 870832,
      "step": 316,
      "train_runtime": 821.221,
      "train_tokens_per_second": 1060.411
    },
    {
      "epoch": 0.025472588842684666,
      "grad_norm": 0.26820915937423706,
      "learning_rate": 0.00017923205342237062,
      "loss": 0.7763,
      "num_input_tokens_seen": 874886,
      "step": 317,
      "train_runtime": 824.1521,
      "train_tokens_per_second": 1061.559
    },
    {
      "epoch": 0.025552944012535408,
      "grad_norm": 0.3074873387813568,
      "learning_rate": 0.00017916527545909852,
      "loss": 0.992,
      "num_input_tokens_seen": 877822,
      "step": 318,
      "train_runtime": 826.512,
      "train_tokens_per_second": 1062.08
    },
    {
      "epoch": 0.025633299182386146,
      "grad_norm": 0.2895393967628479,
      "learning_rate": 0.0001790984974958264,
      "loss": 1.0058,
      "num_input_tokens_seen": 879982,
      "step": 319,
      "train_runtime": 828.3689,
      "train_tokens_per_second": 1062.307
    },
    {
      "epoch": 0.025713654352236888,
      "grad_norm": 0.33440056443214417,
      "learning_rate": 0.00017903171953255427,
      "loss": 0.998,
      "num_input_tokens_seen": 881992,
      "step": 320,
      "train_runtime": 830.2461,
      "train_tokens_per_second": 1062.326
    },
    {
      "epoch": 0.025794009522087626,
      "grad_norm": 0.2001473754644394,
      "learning_rate": 0.00017896494156928214,
      "loss": 0.769,
      "num_input_tokens_seen": 885918,
      "step": 321,
      "train_runtime": 833.1032,
      "train_tokens_per_second": 1063.395
    },
    {
      "epoch": 0.025874364691938368,
      "grad_norm": 0.30161628127098083,
      "learning_rate": 0.00017889816360601,
      "loss": 0.9834,
      "num_input_tokens_seen": 888442,
      "step": 322,
      "train_runtime": 835.1223,
      "train_tokens_per_second": 1063.847
    },
    {
      "epoch": 0.025954719861789106,
      "grad_norm": 0.26389217376708984,
      "learning_rate": 0.0001788313856427379,
      "loss": 0.9027,
      "num_input_tokens_seen": 891640,
      "step": 323,
      "train_runtime": 837.49,
      "train_tokens_per_second": 1064.657
    },
    {
      "epoch": 0.026035075031639848,
      "grad_norm": 0.2665056586265564,
      "learning_rate": 0.0001787646076794658,
      "loss": 0.6519,
      "num_input_tokens_seen": 894146,
      "step": 324,
      "train_runtime": 839.5513,
      "train_tokens_per_second": 1065.028
    },
    {
      "epoch": 0.02611543020149059,
      "grad_norm": 0.3544711768627167,
      "learning_rate": 0.00017869782971619366,
      "loss": 1.0825,
      "num_input_tokens_seen": 897002,
      "step": 325,
      "train_runtime": 841.9102,
      "train_tokens_per_second": 1065.437
    },
    {
      "epoch": 0.02619578537134133,
      "grad_norm": 0.28575438261032104,
      "learning_rate": 0.00017863105175292153,
      "loss": 0.7889,
      "num_input_tokens_seen": 898690,
      "step": 326,
      "train_runtime": 843.4952,
      "train_tokens_per_second": 1065.436
    },
    {
      "epoch": 0.02627614054119207,
      "grad_norm": 0.31976303458213806,
      "learning_rate": 0.0001785642737896494,
      "loss": 0.8223,
      "num_input_tokens_seen": 902416,
      "step": 327,
      "train_runtime": 846.41,
      "train_tokens_per_second": 1066.169
    },
    {
      "epoch": 0.02635649571104281,
      "grad_norm": 0.37153351306915283,
      "learning_rate": 0.0001784974958263773,
      "loss": 1.2682,
      "num_input_tokens_seen": 904900,
      "step": 328,
      "train_runtime": 848.4684,
      "train_tokens_per_second": 1066.51
    },
    {
      "epoch": 0.02643685088089355,
      "grad_norm": 0.3395041823387146,
      "learning_rate": 0.00017843071786310518,
      "loss": 0.7636,
      "num_input_tokens_seen": 908254,
      "step": 329,
      "train_runtime": 850.9843,
      "train_tokens_per_second": 1067.298
    },
    {
      "epoch": 0.02651720605074429,
      "grad_norm": 0.2999211847782135,
      "learning_rate": 0.00017836393989983305,
      "loss": 1.2016,
      "num_input_tokens_seen": 913526,
      "step": 330,
      "train_runtime": 854.8068,
      "train_tokens_per_second": 1068.693
    },
    {
      "epoch": 0.02659756122059503,
      "grad_norm": 0.4062979519367218,
      "learning_rate": 0.00017829716193656095,
      "loss": 1.3509,
      "num_input_tokens_seen": 915430,
      "step": 331,
      "train_runtime": 858.6115,
      "train_tokens_per_second": 1066.175
    },
    {
      "epoch": 0.02667791639044577,
      "grad_norm": 0.36649632453918457,
      "learning_rate": 0.00017823038397328883,
      "loss": 1.2088,
      "num_input_tokens_seen": 917560,
      "step": 332,
      "train_runtime": 860.456,
      "train_tokens_per_second": 1066.365
    },
    {
      "epoch": 0.02675827156029651,
      "grad_norm": 0.3506964147090912,
      "learning_rate": 0.0001781636060100167,
      "loss": 0.8204,
      "num_input_tokens_seen": 919932,
      "step": 333,
      "train_runtime": 862.4173,
      "train_tokens_per_second": 1066.69
    },
    {
      "epoch": 0.026838626730147253,
      "grad_norm": 0.3504076600074768,
      "learning_rate": 0.0001780968280467446,
      "loss": 1.1844,
      "num_input_tokens_seen": 923320,
      "step": 334,
      "train_runtime": 864.9706,
      "train_tokens_per_second": 1067.458
    },
    {
      "epoch": 0.02691898189999799,
      "grad_norm": 0.34691742062568665,
      "learning_rate": 0.00017803005008347247,
      "loss": 1.1981,
      "num_input_tokens_seen": 925438,
      "step": 335,
      "train_runtime": 866.9339,
      "train_tokens_per_second": 1067.484
    },
    {
      "epoch": 0.026999337069848733,
      "grad_norm": 0.28551656007766724,
      "learning_rate": 0.00017796327212020035,
      "loss": 0.9535,
      "num_input_tokens_seen": 927510,
      "step": 336,
      "train_runtime": 868.7185,
      "train_tokens_per_second": 1067.676
    },
    {
      "epoch": 0.02707969223969947,
      "grad_norm": 0.2888050079345703,
      "learning_rate": 0.00017789649415692822,
      "loss": 1.2472,
      "num_input_tokens_seen": 929556,
      "step": 337,
      "train_runtime": 870.5581,
      "train_tokens_per_second": 1067.77
    },
    {
      "epoch": 0.027160047409550213,
      "grad_norm": 0.27506229281425476,
      "learning_rate": 0.0001778297161936561,
      "loss": 1.2203,
      "num_input_tokens_seen": 931564,
      "step": 338,
      "train_runtime": 872.3983,
      "train_tokens_per_second": 1067.82
    },
    {
      "epoch": 0.02724040257940095,
      "grad_norm": 0.376938134431839,
      "learning_rate": 0.000177762938230384,
      "loss": 1.1158,
      "num_input_tokens_seen": 933568,
      "step": 339,
      "train_runtime": 874.2015,
      "train_tokens_per_second": 1067.909
    },
    {
      "epoch": 0.027320757749251693,
      "grad_norm": 0.3374652862548828,
      "learning_rate": 0.00017769616026711187,
      "loss": 1.1596,
      "num_input_tokens_seen": 936760,
      "step": 340,
      "train_runtime": 876.6754,
      "train_tokens_per_second": 1068.537
    },
    {
      "epoch": 0.02740111291910243,
      "grad_norm": 0.3184716999530792,
      "learning_rate": 0.00017762938230383974,
      "loss": 0.9784,
      "num_input_tokens_seen": 940762,
      "step": 341,
      "train_runtime": 879.5921,
      "train_tokens_per_second": 1069.543
    },
    {
      "epoch": 0.027481468088953173,
      "grad_norm": 0.27784883975982666,
      "learning_rate": 0.0001775626043405676,
      "loss": 0.9596,
      "num_input_tokens_seen": 944180,
      "step": 342,
      "train_runtime": 882.14,
      "train_tokens_per_second": 1070.329
    },
    {
      "epoch": 0.027561823258803915,
      "grad_norm": 0.31485989689826965,
      "learning_rate": 0.00017749582637729548,
      "loss": 0.9454,
      "num_input_tokens_seen": 947532,
      "step": 343,
      "train_runtime": 884.7641,
      "train_tokens_per_second": 1070.943
    },
    {
      "epoch": 0.027642178428654653,
      "grad_norm": 0.19721388816833496,
      "learning_rate": 0.00017742904841402339,
      "loss": 0.8979,
      "num_input_tokens_seen": 950782,
      "step": 344,
      "train_runtime": 887.1238,
      "train_tokens_per_second": 1071.758
    },
    {
      "epoch": 0.027722533598505395,
      "grad_norm": 0.3029862642288208,
      "learning_rate": 0.00017736227045075126,
      "loss": 1.3457,
      "num_input_tokens_seen": 953466,
      "step": 345,
      "train_runtime": 889.2656,
      "train_tokens_per_second": 1072.195
    },
    {
      "epoch": 0.027802888768356133,
      "grad_norm": 0.2497890740633011,
      "learning_rate": 0.00017729549248747913,
      "loss": 0.932,
      "num_input_tokens_seen": 955468,
      "step": 346,
      "train_runtime": 891.0254,
      "train_tokens_per_second": 1072.324
    },
    {
      "epoch": 0.027883243938206875,
      "grad_norm": 0.2995472848415375,
      "learning_rate": 0.000177228714524207,
      "loss": 1.104,
      "num_input_tokens_seen": 962602,
      "step": 347,
      "train_runtime": 895.8467,
      "train_tokens_per_second": 1074.516
    },
    {
      "epoch": 0.027963599108057614,
      "grad_norm": 0.3298209309577942,
      "learning_rate": 0.0001771619365609349,
      "loss": 1.446,
      "num_input_tokens_seen": 965170,
      "step": 348,
      "train_runtime": 897.9165,
      "train_tokens_per_second": 1074.899
    },
    {
      "epoch": 0.028043954277908355,
      "grad_norm": 0.35635843873023987,
      "learning_rate": 0.00017709515859766278,
      "loss": 1.196,
      "num_input_tokens_seen": 966402,
      "step": 349,
      "train_runtime": 899.4278,
      "train_tokens_per_second": 1074.463
    },
    {
      "epoch": 0.028124309447759094,
      "grad_norm": 0.2697749137878418,
      "learning_rate": 0.00017702838063439068,
      "loss": 1.1712,
      "num_input_tokens_seen": 969120,
      "step": 350,
      "train_runtime": 901.6435,
      "train_tokens_per_second": 1074.837
    },
    {
      "epoch": 0.028204664617609836,
      "grad_norm": 0.28007960319519043,
      "learning_rate": 0.00017696160267111855,
      "loss": 0.7704,
      "num_input_tokens_seen": 971174,
      "step": 351,
      "train_runtime": 903.4927,
      "train_tokens_per_second": 1074.911
    },
    {
      "epoch": 0.028285019787460577,
      "grad_norm": 0.2861485779285431,
      "learning_rate": 0.00017689482470784642,
      "loss": 1.1288,
      "num_input_tokens_seen": 974448,
      "step": 352,
      "train_runtime": 906.0007,
      "train_tokens_per_second": 1075.549
    },
    {
      "epoch": 0.028365374957311316,
      "grad_norm": 0.2961346507072449,
      "learning_rate": 0.0001768280467445743,
      "loss": 1.252,
      "num_input_tokens_seen": 977222,
      "step": 353,
      "train_runtime": 908.1532,
      "train_tokens_per_second": 1076.054
    },
    {
      "epoch": 0.028445730127162058,
      "grad_norm": 0.2412026822566986,
      "learning_rate": 0.00017676126878130217,
      "loss": 0.9611,
      "num_input_tokens_seen": 979666,
      "step": 354,
      "train_runtime": 910.2101,
      "train_tokens_per_second": 1076.308
    },
    {
      "epoch": 0.028526085297012796,
      "grad_norm": 0.40000274777412415,
      "learning_rate": 0.00017669449081803007,
      "loss": 1.3821,
      "num_input_tokens_seen": 980940,
      "step": 355,
      "train_runtime": 911.7758,
      "train_tokens_per_second": 1075.857
    },
    {
      "epoch": 0.028606440466863538,
      "grad_norm": 0.3051069974899292,
      "learning_rate": 0.00017662771285475794,
      "loss": 0.8413,
      "num_input_tokens_seen": 983684,
      "step": 356,
      "train_runtime": 913.9508,
      "train_tokens_per_second": 1076.299
    },
    {
      "epoch": 0.028686795636714276,
      "grad_norm": 0.31597766280174255,
      "learning_rate": 0.00017656093489148582,
      "loss": 1.2544,
      "num_input_tokens_seen": 993122,
      "step": 357,
      "train_runtime": 919.9445,
      "train_tokens_per_second": 1079.546
    },
    {
      "epoch": 0.028767150806565018,
      "grad_norm": 0.34993791580200195,
      "learning_rate": 0.0001764941569282137,
      "loss": 1.3095,
      "num_input_tokens_seen": 995658,
      "step": 358,
      "train_runtime": 922.0586,
      "train_tokens_per_second": 1079.821
    },
    {
      "epoch": 0.028847505976415756,
      "grad_norm": 0.31747668981552124,
      "learning_rate": 0.00017642737896494156,
      "loss": 1.0146,
      "num_input_tokens_seen": 998980,
      "step": 359,
      "train_runtime": 924.6271,
      "train_tokens_per_second": 1080.414
    },
    {
      "epoch": 0.028927861146266498,
      "grad_norm": 0.2664702832698822,
      "learning_rate": 0.00017636060100166946,
      "loss": 0.8058,
      "num_input_tokens_seen": 1001000,
      "step": 360,
      "train_runtime": 926.5061,
      "train_tokens_per_second": 1080.403
    },
    {
      "epoch": 0.02900821631611724,
      "grad_norm": 0.3068900406360626,
      "learning_rate": 0.00017629382303839734,
      "loss": 1.2296,
      "num_input_tokens_seen": 1004382,
      "step": 361,
      "train_runtime": 930.9966,
      "train_tokens_per_second": 1078.825
    },
    {
      "epoch": 0.029088571485967978,
      "grad_norm": 0.3020861744880676,
      "learning_rate": 0.0001762270450751252,
      "loss": 1.3893,
      "num_input_tokens_seen": 1007248,
      "step": 362,
      "train_runtime": 933.244,
      "train_tokens_per_second": 1079.298
    },
    {
      "epoch": 0.02916892665581872,
      "grad_norm": 0.3030184805393219,
      "learning_rate": 0.00017616026711185308,
      "loss": 1.4272,
      "num_input_tokens_seen": 1011076,
      "step": 363,
      "train_runtime": 936.0694,
      "train_tokens_per_second": 1080.129
    },
    {
      "epoch": 0.02924928182566946,
      "grad_norm": 0.32471176981925964,
      "learning_rate": 0.00017609348914858096,
      "loss": 1.0103,
      "num_input_tokens_seen": 1013570,
      "step": 364,
      "train_runtime": 938.0863,
      "train_tokens_per_second": 1080.466
    },
    {
      "epoch": 0.0293296369955202,
      "grad_norm": 0.3173401653766632,
      "learning_rate": 0.00017602671118530886,
      "loss": 0.9379,
      "num_input_tokens_seen": 1014974,
      "step": 365,
      "train_runtime": 939.6768,
      "train_tokens_per_second": 1080.131
    },
    {
      "epoch": 0.02940999216537094,
      "grad_norm": 0.28528130054473877,
      "learning_rate": 0.00017595993322203673,
      "loss": 0.7587,
      "num_input_tokens_seen": 1017044,
      "step": 366,
      "train_runtime": 941.6503,
      "train_tokens_per_second": 1080.065
    },
    {
      "epoch": 0.02949034733522168,
      "grad_norm": 0.2877449691295624,
      "learning_rate": 0.00017589315525876463,
      "loss": 0.9675,
      "num_input_tokens_seen": 1021034,
      "step": 367,
      "train_runtime": 944.3924,
      "train_tokens_per_second": 1081.154
    },
    {
      "epoch": 0.02957070250507242,
      "grad_norm": 0.26632797718048096,
      "learning_rate": 0.0001758263772954925,
      "loss": 0.7783,
      "num_input_tokens_seen": 1022808,
      "step": 368,
      "train_runtime": 946.0748,
      "train_tokens_per_second": 1081.107
    },
    {
      "epoch": 0.02965105767492316,
      "grad_norm": 0.31062689423561096,
      "learning_rate": 0.00017575959933222038,
      "loss": 1.214,
      "num_input_tokens_seen": 1026104,
      "step": 369,
      "train_runtime": 948.4748,
      "train_tokens_per_second": 1081.846
    },
    {
      "epoch": 0.029731412844773902,
      "grad_norm": 0.2866829037666321,
      "learning_rate": 0.00017569282136894825,
      "loss": 0.7996,
      "num_input_tokens_seen": 1029384,
      "step": 370,
      "train_runtime": 951.1018,
      "train_tokens_per_second": 1082.307
    },
    {
      "epoch": 0.02981176801462464,
      "grad_norm": 0.31888410449028015,
      "learning_rate": 0.00017562604340567615,
      "loss": 1.1984,
      "num_input_tokens_seen": 1033068,
      "step": 371,
      "train_runtime": 953.7773,
      "train_tokens_per_second": 1083.133
    },
    {
      "epoch": 0.029892123184475382,
      "grad_norm": 0.2944906949996948,
      "learning_rate": 0.00017555926544240402,
      "loss": 1.2214,
      "num_input_tokens_seen": 1035278,
      "step": 372,
      "train_runtime": 955.6776,
      "train_tokens_per_second": 1083.292
    },
    {
      "epoch": 0.02997247835432612,
      "grad_norm": 0.3593326508998871,
      "learning_rate": 0.0001754924874791319,
      "loss": 1.0988,
      "num_input_tokens_seen": 1038318,
      "step": 373,
      "train_runtime": 958.0543,
      "train_tokens_per_second": 1083.778
    },
    {
      "epoch": 0.030052833524176863,
      "grad_norm": 0.309621125459671,
      "learning_rate": 0.00017542570951585977,
      "loss": 0.9025,
      "num_input_tokens_seen": 1041572,
      "step": 374,
      "train_runtime": 960.4868,
      "train_tokens_per_second": 1084.421
    },
    {
      "epoch": 0.0301331886940276,
      "grad_norm": 0.2950892448425293,
      "learning_rate": 0.00017535893155258764,
      "loss": 1.1809,
      "num_input_tokens_seen": 1043442,
      "step": 375,
      "train_runtime": 962.2298,
      "train_tokens_per_second": 1084.4
    },
    {
      "epoch": 0.030213543863878343,
      "grad_norm": 0.3762933909893036,
      "learning_rate": 0.00017529215358931554,
      "loss": 0.98,
      "num_input_tokens_seen": 1045694,
      "step": 376,
      "train_runtime": 964.0901,
      "train_tokens_per_second": 1084.643
    },
    {
      "epoch": 0.03029389903372908,
      "grad_norm": 0.2948756515979767,
      "learning_rate": 0.00017522537562604342,
      "loss": 0.9154,
      "num_input_tokens_seen": 1049460,
      "step": 377,
      "train_runtime": 966.8047,
      "train_tokens_per_second": 1085.493
    },
    {
      "epoch": 0.030374254203579823,
      "grad_norm": 0.266676664352417,
      "learning_rate": 0.0001751585976627713,
      "loss": 1.0154,
      "num_input_tokens_seen": 1053400,
      "step": 378,
      "train_runtime": 969.5925,
      "train_tokens_per_second": 1086.436
    },
    {
      "epoch": 0.030454609373430565,
      "grad_norm": 0.271295428276062,
      "learning_rate": 0.00017509181969949916,
      "loss": 0.9774,
      "num_input_tokens_seen": 1056186,
      "step": 379,
      "train_runtime": 971.7497,
      "train_tokens_per_second": 1086.891
    },
    {
      "epoch": 0.030534964543281303,
      "grad_norm": 0.2567802369594574,
      "learning_rate": 0.00017502504173622704,
      "loss": 0.7366,
      "num_input_tokens_seen": 1058786,
      "step": 380,
      "train_runtime": 973.8713,
      "train_tokens_per_second": 1087.193
    },
    {
      "epoch": 0.030615319713132045,
      "grad_norm": 0.2612038850784302,
      "learning_rate": 0.0001749582637729549,
      "loss": 0.9533,
      "num_input_tokens_seen": 1061502,
      "step": 381,
      "train_runtime": 976.0026,
      "train_tokens_per_second": 1087.602
    },
    {
      "epoch": 0.030695674882982783,
      "grad_norm": 0.3431708514690399,
      "learning_rate": 0.0001748914858096828,
      "loss": 0.9277,
      "num_input_tokens_seen": 1063254,
      "step": 382,
      "train_runtime": 977.7576,
      "train_tokens_per_second": 1087.441
    },
    {
      "epoch": 0.030776030052833525,
      "grad_norm": 0.3861244320869446,
      "learning_rate": 0.0001748247078464107,
      "loss": 1.2894,
      "num_input_tokens_seen": 1066208,
      "step": 383,
      "train_runtime": 980.0461,
      "train_tokens_per_second": 1087.916
    },
    {
      "epoch": 0.030856385222684263,
      "grad_norm": 0.2675583064556122,
      "learning_rate": 0.00017475792988313858,
      "loss": 1.1537,
      "num_input_tokens_seen": 1069864,
      "step": 384,
      "train_runtime": 982.6449,
      "train_tokens_per_second": 1088.759
    },
    {
      "epoch": 0.030936740392535005,
      "grad_norm": 0.3754398226737976,
      "learning_rate": 0.00017469115191986646,
      "loss": 1.0408,
      "num_input_tokens_seen": 1072406,
      "step": 385,
      "train_runtime": 984.8168,
      "train_tokens_per_second": 1088.94
    },
    {
      "epoch": 0.031017095562385744,
      "grad_norm": 0.24042372405529022,
      "learning_rate": 0.00017462437395659433,
      "loss": 1.3946,
      "num_input_tokens_seen": 1075636,
      "step": 386,
      "train_runtime": 987.3118,
      "train_tokens_per_second": 1089.459
    },
    {
      "epoch": 0.031097450732236485,
      "grad_norm": 0.3622994124889374,
      "learning_rate": 0.00017455759599332223,
      "loss": 1.0261,
      "num_input_tokens_seen": 1078594,
      "step": 387,
      "train_runtime": 989.7139,
      "train_tokens_per_second": 1089.804
    },
    {
      "epoch": 0.031177805902087227,
      "grad_norm": 0.2808270752429962,
      "learning_rate": 0.0001744908180300501,
      "loss": 1.1029,
      "num_input_tokens_seen": 1080282,
      "step": 388,
      "train_runtime": 991.353,
      "train_tokens_per_second": 1089.705
    },
    {
      "epoch": 0.03125816107193797,
      "grad_norm": 0.2996695637702942,
      "learning_rate": 0.00017442404006677798,
      "loss": 1.0891,
      "num_input_tokens_seen": 1084780,
      "step": 389,
      "train_runtime": 994.6542,
      "train_tokens_per_second": 1090.61
    },
    {
      "epoch": 0.031338516241788704,
      "grad_norm": 0.28290489315986633,
      "learning_rate": 0.00017435726210350585,
      "loss": 0.9752,
      "num_input_tokens_seen": 1087136,
      "step": 390,
      "train_runtime": 996.5787,
      "train_tokens_per_second": 1090.868
    },
    {
      "epoch": 0.031418871411639446,
      "grad_norm": 0.36534032225608826,
      "learning_rate": 0.00017429048414023372,
      "loss": 1.2168,
      "num_input_tokens_seen": 1090300,
      "step": 391,
      "train_runtime": 1000.7808,
      "train_tokens_per_second": 1089.449
    },
    {
      "epoch": 0.03149922658149019,
      "grad_norm": 0.34648290276527405,
      "learning_rate": 0.00017422370617696162,
      "loss": 0.9941,
      "num_input_tokens_seen": 1092990,
      "step": 392,
      "train_runtime": 1002.9913,
      "train_tokens_per_second": 1089.73
    },
    {
      "epoch": 0.03157958175134093,
      "grad_norm": 0.3221375644207001,
      "learning_rate": 0.0001741569282136895,
      "loss": 0.9356,
      "num_input_tokens_seen": 1095638,
      "step": 393,
      "train_runtime": 1005.08,
      "train_tokens_per_second": 1090.1
    },
    {
      "epoch": 0.031659936921191664,
      "grad_norm": 0.2946549952030182,
      "learning_rate": 0.00017409015025041737,
      "loss": 1.0739,
      "num_input_tokens_seen": 1097618,
      "step": 394,
      "train_runtime": 1006.8478,
      "train_tokens_per_second": 1090.153
    },
    {
      "epoch": 0.031740292091042406,
      "grad_norm": 0.3398057222366333,
      "learning_rate": 0.00017402337228714524,
      "loss": 1.0839,
      "num_input_tokens_seen": 1098974,
      "step": 395,
      "train_runtime": 1008.3661,
      "train_tokens_per_second": 1089.856
    },
    {
      "epoch": 0.03182064726089315,
      "grad_norm": 0.28257885575294495,
      "learning_rate": 0.00017395659432387311,
      "loss": 1.1825,
      "num_input_tokens_seen": 1101014,
      "step": 396,
      "train_runtime": 1010.2335,
      "train_tokens_per_second": 1089.861
    },
    {
      "epoch": 0.03190100243074389,
      "grad_norm": 0.2921408712863922,
      "learning_rate": 0.00017388981636060101,
      "loss": 1.7143,
      "num_input_tokens_seen": 1106898,
      "step": 397,
      "train_runtime": 1014.1899,
      "train_tokens_per_second": 1091.411
    },
    {
      "epoch": 0.03198135760059463,
      "grad_norm": 0.3177490532398224,
      "learning_rate": 0.0001738230383973289,
      "loss": 1.1459,
      "num_input_tokens_seen": 1110030,
      "step": 398,
      "train_runtime": 1016.5793,
      "train_tokens_per_second": 1091.927
    },
    {
      "epoch": 0.032061712770445366,
      "grad_norm": 0.27498766779899597,
      "learning_rate": 0.00017375626043405676,
      "loss": 0.9801,
      "num_input_tokens_seen": 1112316,
      "step": 399,
      "train_runtime": 1018.5545,
      "train_tokens_per_second": 1092.054
    },
    {
      "epoch": 0.03214206794029611,
      "grad_norm": 0.2847306430339813,
      "learning_rate": 0.00017368948247078466,
      "loss": 0.9599,
      "num_input_tokens_seen": 1114706,
      "step": 400,
      "train_runtime": 1020.5936,
      "train_tokens_per_second": 1092.213
    },
    {
      "epoch": 0.03222242311014685,
      "grad_norm": 0.34009799361228943,
      "learning_rate": 0.00017362270450751253,
      "loss": 1.1011,
      "num_input_tokens_seen": 1120778,
      "step": 401,
      "train_runtime": 1024.7879,
      "train_tokens_per_second": 1093.668
    },
    {
      "epoch": 0.03230277827999759,
      "grad_norm": 0.35099753737449646,
      "learning_rate": 0.0001735559265442404,
      "loss": 1.3141,
      "num_input_tokens_seen": 1123376,
      "step": 402,
      "train_runtime": 1026.9283,
      "train_tokens_per_second": 1093.919
    },
    {
      "epoch": 0.03238313344984833,
      "grad_norm": 0.3260703384876251,
      "learning_rate": 0.0001734891485809683,
      "loss": 1.0699,
      "num_input_tokens_seen": 1125148,
      "step": 403,
      "train_runtime": 1028.5868,
      "train_tokens_per_second": 1093.878
    },
    {
      "epoch": 0.03246348861969907,
      "grad_norm": 0.29309403896331787,
      "learning_rate": 0.00017342237061769618,
      "loss": 1.0256,
      "num_input_tokens_seen": 1127176,
      "step": 404,
      "train_runtime": 1030.4677,
      "train_tokens_per_second": 1093.849
    },
    {
      "epoch": 0.03254384378954981,
      "grad_norm": 0.2610347270965576,
      "learning_rate": 0.00017335559265442405,
      "loss": 0.8452,
      "num_input_tokens_seen": 1131924,
      "step": 405,
      "train_runtime": 1033.8426,
      "train_tokens_per_second": 1094.871
    },
    {
      "epoch": 0.03262419895940055,
      "grad_norm": 0.3089810609817505,
      "learning_rate": 0.00017328881469115193,
      "loss": 0.9333,
      "num_input_tokens_seen": 1133338,
      "step": 406,
      "train_runtime": 1035.4445,
      "train_tokens_per_second": 1094.542
    },
    {
      "epoch": 0.032704554129251294,
      "grad_norm": 0.32210832834243774,
      "learning_rate": 0.0001732220367278798,
      "loss": 0.9726,
      "num_input_tokens_seen": 1136126,
      "step": 407,
      "train_runtime": 1037.6272,
      "train_tokens_per_second": 1094.927
    },
    {
      "epoch": 0.03278490929910203,
      "grad_norm": 0.3733350336551666,
      "learning_rate": 0.0001731552587646077,
      "loss": 1.0942,
      "num_input_tokens_seen": 1137842,
      "step": 408,
      "train_runtime": 1039.3386,
      "train_tokens_per_second": 1094.775
    },
    {
      "epoch": 0.03286526446895277,
      "grad_norm": 0.28001129627227783,
      "learning_rate": 0.00017308848080133557,
      "loss": 1.0348,
      "num_input_tokens_seen": 1141572,
      "step": 409,
      "train_runtime": 1042.0283,
      "train_tokens_per_second": 1095.529
    },
    {
      "epoch": 0.03294561963880351,
      "grad_norm": 0.27658599615097046,
      "learning_rate": 0.00017302170283806345,
      "loss": 0.8904,
      "num_input_tokens_seen": 1144452,
      "step": 410,
      "train_runtime": 1044.2155,
      "train_tokens_per_second": 1095.992
    },
    {
      "epoch": 0.033025974808654254,
      "grad_norm": 0.3635813891887665,
      "learning_rate": 0.00017295492487479132,
      "loss": 1.3809,
      "num_input_tokens_seen": 1148208,
      "step": 411,
      "train_runtime": 1047.0373,
      "train_tokens_per_second": 1096.626
    },
    {
      "epoch": 0.03310632997850499,
      "grad_norm": 0.3085966110229492,
      "learning_rate": 0.0001728881469115192,
      "loss": 0.9932,
      "num_input_tokens_seen": 1150816,
      "step": 412,
      "train_runtime": 1049.1355,
      "train_tokens_per_second": 1096.918
    },
    {
      "epoch": 0.03318668514835573,
      "grad_norm": 0.4016687870025635,
      "learning_rate": 0.0001728213689482471,
      "loss": 1.1162,
      "num_input_tokens_seen": 1153398,
      "step": 413,
      "train_runtime": 1051.2539,
      "train_tokens_per_second": 1097.164
    },
    {
      "epoch": 0.03326704031820647,
      "grad_norm": 0.3800343871116638,
      "learning_rate": 0.00017275459098497497,
      "loss": 1.2136,
      "num_input_tokens_seen": 1156048,
      "step": 414,
      "train_runtime": 1053.4026,
      "train_tokens_per_second": 1097.442
    },
    {
      "epoch": 0.033347395488057215,
      "grad_norm": 0.3973337411880493,
      "learning_rate": 0.00017268781302170284,
      "loss": 1.0981,
      "num_input_tokens_seen": 1158302,
      "step": 415,
      "train_runtime": 1055.3362,
      "train_tokens_per_second": 1097.567
    },
    {
      "epoch": 0.033427750657907956,
      "grad_norm": 0.3727961480617523,
      "learning_rate": 0.0001726210350584307,
      "loss": 1.3403,
      "num_input_tokens_seen": 1161096,
      "step": 416,
      "train_runtime": 1057.5223,
      "train_tokens_per_second": 1097.94
    },
    {
      "epoch": 0.03350810582775869,
      "grad_norm": 0.3895726203918457,
      "learning_rate": 0.0001725542570951586,
      "loss": 1.013,
      "num_input_tokens_seen": 1162742,
      "step": 417,
      "train_runtime": 1059.2234,
      "train_tokens_per_second": 1097.731
    },
    {
      "epoch": 0.03358846099760943,
      "grad_norm": 0.3150062561035156,
      "learning_rate": 0.0001724874791318865,
      "loss": 1.2591,
      "num_input_tokens_seen": 1165064,
      "step": 418,
      "train_runtime": 1061.1738,
      "train_tokens_per_second": 1097.901
    },
    {
      "epoch": 0.033668816167460175,
      "grad_norm": 0.3188002407550812,
      "learning_rate": 0.0001724207011686144,
      "loss": 1.1988,
      "num_input_tokens_seen": 1168084,
      "step": 419,
      "train_runtime": 1063.5283,
      "train_tokens_per_second": 1098.31
    },
    {
      "epoch": 0.03374917133731092,
      "grad_norm": 0.3131507933139801,
      "learning_rate": 0.00017235392320534226,
      "loss": 0.7852,
      "num_input_tokens_seen": 1171426,
      "step": 420,
      "train_runtime": 1066.1926,
      "train_tokens_per_second": 1098.7
    },
    {
      "epoch": 0.03382952650716165,
      "grad_norm": 0.2633776366710663,
      "learning_rate": 0.00017228714524207013,
      "loss": 0.9786,
      "num_input_tokens_seen": 1174218,
      "step": 421,
      "train_runtime": 1070.1204,
      "train_tokens_per_second": 1097.277
    },
    {
      "epoch": 0.03390988167701239,
      "grad_norm": 0.3124426305294037,
      "learning_rate": 0.000172220367278798,
      "loss": 0.9216,
      "num_input_tokens_seen": 1176310,
      "step": 422,
      "train_runtime": 1071.9478,
      "train_tokens_per_second": 1097.358
    },
    {
      "epoch": 0.033990236846863135,
      "grad_norm": 0.2875818908214569,
      "learning_rate": 0.00017215358931552588,
      "loss": 1.2885,
      "num_input_tokens_seen": 1178900,
      "step": 423,
      "train_runtime": 1074.1468,
      "train_tokens_per_second": 1097.522
    },
    {
      "epoch": 0.03407059201671388,
      "grad_norm": 0.3103838562965393,
      "learning_rate": 0.00017208681135225378,
      "loss": 1.1525,
      "num_input_tokens_seen": 1184014,
      "step": 424,
      "train_runtime": 1077.6489,
      "train_tokens_per_second": 1098.701
    },
    {
      "epoch": 0.03415094718656462,
      "grad_norm": 0.29413601756095886,
      "learning_rate": 0.00017202003338898165,
      "loss": 0.9123,
      "num_input_tokens_seen": 1186588,
      "step": 425,
      "train_runtime": 1079.6745,
      "train_tokens_per_second": 1099.024
    },
    {
      "epoch": 0.034231302356415354,
      "grad_norm": 0.3262699246406555,
      "learning_rate": 0.00017195325542570953,
      "loss": 0.7304,
      "num_input_tokens_seen": 1189358,
      "step": 426,
      "train_runtime": 1081.9085,
      "train_tokens_per_second": 1099.315
    },
    {
      "epoch": 0.034311657526266096,
      "grad_norm": 0.3086264431476593,
      "learning_rate": 0.0001718864774624374,
      "loss": 1.3913,
      "num_input_tokens_seen": 1192308,
      "step": 427,
      "train_runtime": 1084.2987,
      "train_tokens_per_second": 1099.612
    },
    {
      "epoch": 0.03439201269611684,
      "grad_norm": 0.31988200545310974,
      "learning_rate": 0.00017181969949916527,
      "loss": 0.9717,
      "num_input_tokens_seen": 1195212,
      "step": 428,
      "train_runtime": 1086.614,
      "train_tokens_per_second": 1099.942
    },
    {
      "epoch": 0.03447236786596758,
      "grad_norm": 0.3014238476753235,
      "learning_rate": 0.00017175292153589317,
      "loss": 1.1486,
      "num_input_tokens_seen": 1197536,
      "step": 429,
      "train_runtime": 1088.5881,
      "train_tokens_per_second": 1100.082
    },
    {
      "epoch": 0.034552723035818314,
      "grad_norm": 0.34671786427497864,
      "learning_rate": 0.00017168614357262105,
      "loss": 1.1268,
      "num_input_tokens_seen": 1199868,
      "step": 430,
      "train_runtime": 1090.5466,
      "train_tokens_per_second": 1100.245
    },
    {
      "epoch": 0.034633078205669056,
      "grad_norm": 0.2531111240386963,
      "learning_rate": 0.00017161936560934892,
      "loss": 0.8155,
      "num_input_tokens_seen": 1202948,
      "step": 431,
      "train_runtime": 1092.9097,
      "train_tokens_per_second": 1100.684
    },
    {
      "epoch": 0.0347134333755198,
      "grad_norm": 0.29365605115890503,
      "learning_rate": 0.0001715525876460768,
      "loss": 1.0678,
      "num_input_tokens_seen": 1206882,
      "step": 432,
      "train_runtime": 1095.7879,
      "train_tokens_per_second": 1101.383
    },
    {
      "epoch": 0.03479378854537054,
      "grad_norm": 0.3197197914123535,
      "learning_rate": 0.00017148580968280467,
      "loss": 1.0508,
      "num_input_tokens_seen": 1210226,
      "step": 433,
      "train_runtime": 1098.2378,
      "train_tokens_per_second": 1101.971
    },
    {
      "epoch": 0.03487414371522128,
      "grad_norm": 0.3036319613456726,
      "learning_rate": 0.00017141903171953257,
      "loss": 1.1495,
      "num_input_tokens_seen": 1213026,
      "step": 434,
      "train_runtime": 1100.4866,
      "train_tokens_per_second": 1102.263
    },
    {
      "epoch": 0.034954498885072016,
      "grad_norm": 0.276408851146698,
      "learning_rate": 0.00017135225375626044,
      "loss": 0.9933,
      "num_input_tokens_seen": 1216442,
      "step": 435,
      "train_runtime": 1103.0565,
      "train_tokens_per_second": 1102.792
    },
    {
      "epoch": 0.03503485405492276,
      "grad_norm": 0.3260977864265442,
      "learning_rate": 0.00017128547579298834,
      "loss": 1.1636,
      "num_input_tokens_seen": 1219046,
      "step": 436,
      "train_runtime": 1105.1811,
      "train_tokens_per_second": 1103.028
    },
    {
      "epoch": 0.0351152092247735,
      "grad_norm": 0.28073257207870483,
      "learning_rate": 0.0001712186978297162,
      "loss": 1.2499,
      "num_input_tokens_seen": 1221778,
      "step": 437,
      "train_runtime": 1107.3901,
      "train_tokens_per_second": 1103.295
    },
    {
      "epoch": 0.03519556439462424,
      "grad_norm": 0.30442941188812256,
      "learning_rate": 0.00017115191986644409,
      "loss": 1.2515,
      "num_input_tokens_seen": 1224002,
      "step": 438,
      "train_runtime": 1109.1972,
      "train_tokens_per_second": 1103.503
    },
    {
      "epoch": 0.035275919564474977,
      "grad_norm": 0.2952291667461395,
      "learning_rate": 0.00017108514190317196,
      "loss": 0.8443,
      "num_input_tokens_seen": 1225982,
      "step": 439,
      "train_runtime": 1110.9581,
      "train_tokens_per_second": 1103.536
    },
    {
      "epoch": 0.03535627473432572,
      "grad_norm": 0.25175273418426514,
      "learning_rate": 0.00017101836393989986,
      "loss": 0.9596,
      "num_input_tokens_seen": 1230048,
      "step": 440,
      "train_runtime": 1113.8367,
      "train_tokens_per_second": 1104.334
    },
    {
      "epoch": 0.03543662990417646,
      "grad_norm": 0.25086110830307007,
      "learning_rate": 0.00017095158597662773,
      "loss": 0.8363,
      "num_input_tokens_seen": 1232236,
      "step": 441,
      "train_runtime": 1115.7358,
      "train_tokens_per_second": 1104.416
    },
    {
      "epoch": 0.0355169850740272,
      "grad_norm": 0.26771441102027893,
      "learning_rate": 0.0001708848080133556,
      "loss": 0.7641,
      "num_input_tokens_seen": 1234444,
      "step": 442,
      "train_runtime": 1117.6826,
      "train_tokens_per_second": 1104.467
    },
    {
      "epoch": 0.035597340243877944,
      "grad_norm": 0.27989158034324646,
      "learning_rate": 0.00017081803005008348,
      "loss": 1.0552,
      "num_input_tokens_seen": 1237656,
      "step": 443,
      "train_runtime": 1120.1697,
      "train_tokens_per_second": 1104.883
    },
    {
      "epoch": 0.03567769541372868,
      "grad_norm": 0.2957555949687958,
      "learning_rate": 0.00017075125208681135,
      "loss": 1.1023,
      "num_input_tokens_seen": 1241284,
      "step": 444,
      "train_runtime": 1122.8789,
      "train_tokens_per_second": 1105.448
    },
    {
      "epoch": 0.03575805058357942,
      "grad_norm": 0.2737371027469635,
      "learning_rate": 0.00017068447412353925,
      "loss": 0.9989,
      "num_input_tokens_seen": 1242916,
      "step": 445,
      "train_runtime": 1124.5852,
      "train_tokens_per_second": 1105.222
    },
    {
      "epoch": 0.03583840575343016,
      "grad_norm": 0.3018834590911865,
      "learning_rate": 0.00017061769616026712,
      "loss": 0.8744,
      "num_input_tokens_seen": 1245290,
      "step": 446,
      "train_runtime": 1126.5345,
      "train_tokens_per_second": 1105.417
    },
    {
      "epoch": 0.035918760923280904,
      "grad_norm": 0.3098128139972687,
      "learning_rate": 0.000170550918196995,
      "loss": 0.9994,
      "num_input_tokens_seen": 1247416,
      "step": 447,
      "train_runtime": 1128.505,
      "train_tokens_per_second": 1105.37
    },
    {
      "epoch": 0.03599911609313164,
      "grad_norm": 0.2849488854408264,
      "learning_rate": 0.00017048414023372287,
      "loss": 1.3004,
      "num_input_tokens_seen": 1251064,
      "step": 448,
      "train_runtime": 1131.1497,
      "train_tokens_per_second": 1106.011
    },
    {
      "epoch": 0.03607947126298238,
      "grad_norm": 0.2536231577396393,
      "learning_rate": 0.00017041736227045074,
      "loss": 0.9552,
      "num_input_tokens_seen": 1254712,
      "step": 449,
      "train_runtime": 1133.8587,
      "train_tokens_per_second": 1106.586
    },
    {
      "epoch": 0.03615982643283312,
      "grad_norm": 0.3291691839694977,
      "learning_rate": 0.00017035058430717862,
      "loss": 1.2501,
      "num_input_tokens_seen": 1257184,
      "step": 450,
      "train_runtime": 1135.9122,
      "train_tokens_per_second": 1106.762
    },
    {
      "epoch": 0.036240181602683864,
      "grad_norm": 0.3126014173030853,
      "learning_rate": 0.00017028380634390652,
      "loss": 1.1347,
      "num_input_tokens_seen": 1259704,
      "step": 451,
      "train_runtime": 1139.7684,
      "train_tokens_per_second": 1105.228
    },
    {
      "epoch": 0.036320536772534606,
      "grad_norm": 0.2651938498020172,
      "learning_rate": 0.0001702170283806344,
      "loss": 0.8355,
      "num_input_tokens_seen": 1263596,
      "step": 452,
      "train_runtime": 1142.6277,
      "train_tokens_per_second": 1105.869
    },
    {
      "epoch": 0.03640089194238534,
      "grad_norm": 0.2972799241542816,
      "learning_rate": 0.0001701502504173623,
      "loss": 1.2333,
      "num_input_tokens_seen": 1265162,
      "step": 453,
      "train_runtime": 1144.2087,
      "train_tokens_per_second": 1105.709
    },
    {
      "epoch": 0.03648124711223608,
      "grad_norm": 0.3837432563304901,
      "learning_rate": 0.00017008347245409016,
      "loss": 1.1638,
      "num_input_tokens_seen": 1266794,
      "step": 454,
      "train_runtime": 1145.8047,
      "train_tokens_per_second": 1105.593
    },
    {
      "epoch": 0.036561602282086825,
      "grad_norm": 0.31418415904045105,
      "learning_rate": 0.00017001669449081804,
      "loss": 1.2247,
      "num_input_tokens_seen": 1270012,
      "step": 455,
      "train_runtime": 1148.2069,
      "train_tokens_per_second": 1106.083
    },
    {
      "epoch": 0.03664195745193757,
      "grad_norm": 0.2486877590417862,
      "learning_rate": 0.00016994991652754594,
      "loss": 0.9665,
      "num_input_tokens_seen": 1272992,
      "step": 456,
      "train_runtime": 1150.5529,
      "train_tokens_per_second": 1106.418
    },
    {
      "epoch": 0.0367223126217883,
      "grad_norm": 0.28888317942619324,
      "learning_rate": 0.0001698831385642738,
      "loss": 0.895,
      "num_input_tokens_seen": 1274966,
      "step": 457,
      "train_runtime": 1152.3425,
      "train_tokens_per_second": 1106.412
    },
    {
      "epoch": 0.03680266779163904,
      "grad_norm": 0.34526416659355164,
      "learning_rate": 0.00016981636060100168,
      "loss": 1.1872,
      "num_input_tokens_seen": 1276664,
      "step": 458,
      "train_runtime": 1154.0102,
      "train_tokens_per_second": 1106.285
    },
    {
      "epoch": 0.036883022961489785,
      "grad_norm": 0.3177397549152374,
      "learning_rate": 0.00016974958263772956,
      "loss": 0.8193,
      "num_input_tokens_seen": 1280068,
      "step": 459,
      "train_runtime": 1156.5444,
      "train_tokens_per_second": 1106.804
    },
    {
      "epoch": 0.03696337813134053,
      "grad_norm": 0.3397356867790222,
      "learning_rate": 0.00016968280467445743,
      "loss": 1.0335,
      "num_input_tokens_seen": 1282256,
      "step": 460,
      "train_runtime": 1158.471,
      "train_tokens_per_second": 1106.852
    },
    {
      "epoch": 0.03704373330119127,
      "grad_norm": 0.31023797392845154,
      "learning_rate": 0.00016961602671118533,
      "loss": 1.0502,
      "num_input_tokens_seen": 1285318,
      "step": 461,
      "train_runtime": 1160.9668,
      "train_tokens_per_second": 1107.11
    },
    {
      "epoch": 0.037124088471042004,
      "grad_norm": 0.28655844926834106,
      "learning_rate": 0.0001695492487479132,
      "loss": 0.9002,
      "num_input_tokens_seen": 1288374,
      "step": 462,
      "train_runtime": 1163.3335,
      "train_tokens_per_second": 1107.485
    },
    {
      "epoch": 0.037204443640892745,
      "grad_norm": 0.3392660617828369,
      "learning_rate": 0.00016948247078464108,
      "loss": 0.6588,
      "num_input_tokens_seen": 1290312,
      "step": 463,
      "train_runtime": 1165.1734,
      "train_tokens_per_second": 1107.399
    },
    {
      "epoch": 0.03728479881074349,
      "grad_norm": 0.2751123905181885,
      "learning_rate": 0.00016941569282136895,
      "loss": 0.7814,
      "num_input_tokens_seen": 1292578,
      "step": 464,
      "train_runtime": 1167.1163,
      "train_tokens_per_second": 1107.497
    },
    {
      "epoch": 0.03736515398059423,
      "grad_norm": 0.3018333315849304,
      "learning_rate": 0.00016934891485809682,
      "loss": 1.0923,
      "num_input_tokens_seen": 1294750,
      "step": 465,
      "train_runtime": 1169.0036,
      "train_tokens_per_second": 1107.567
    },
    {
      "epoch": 0.037445509150444964,
      "grad_norm": 0.34172528982162476,
      "learning_rate": 0.0001692821368948247,
      "loss": 1.1338,
      "num_input_tokens_seen": 1298036,
      "step": 466,
      "train_runtime": 1171.5677,
      "train_tokens_per_second": 1107.948
    },
    {
      "epoch": 0.037525864320295706,
      "grad_norm": 0.25886115431785583,
      "learning_rate": 0.0001692153589315526,
      "loss": 0.7873,
      "num_input_tokens_seen": 1300234,
      "step": 467,
      "train_runtime": 1173.4752,
      "train_tokens_per_second": 1108.02
    },
    {
      "epoch": 0.03760621949014645,
      "grad_norm": 0.2959878444671631,
      "learning_rate": 0.00016914858096828047,
      "loss": 0.891,
      "num_input_tokens_seen": 1302488,
      "step": 468,
      "train_runtime": 1175.4823,
      "train_tokens_per_second": 1108.046
    },
    {
      "epoch": 0.03768657465999719,
      "grad_norm": 0.3088524639606476,
      "learning_rate": 0.00016908180300500834,
      "loss": 1.0366,
      "num_input_tokens_seen": 1304818,
      "step": 469,
      "train_runtime": 1177.4553,
      "train_tokens_per_second": 1108.168
    },
    {
      "epoch": 0.03776692982984793,
      "grad_norm": 0.3135266602039337,
      "learning_rate": 0.00016901502504173624,
      "loss": 1.0494,
      "num_input_tokens_seen": 1307896,
      "step": 470,
      "train_runtime": 1179.8193,
      "train_tokens_per_second": 1108.556
    },
    {
      "epoch": 0.037847284999698666,
      "grad_norm": 0.2991373836994171,
      "learning_rate": 0.00016894824707846412,
      "loss": 1.2399,
      "num_input_tokens_seen": 1311892,
      "step": 471,
      "train_runtime": 1182.6972,
      "train_tokens_per_second": 1109.237
    },
    {
      "epoch": 0.03792764016954941,
      "grad_norm": 0.35278165340423584,
      "learning_rate": 0.00016888146911519202,
      "loss": 0.7783,
      "num_input_tokens_seen": 1314402,
      "step": 472,
      "train_runtime": 1184.7445,
      "train_tokens_per_second": 1109.439
    },
    {
      "epoch": 0.03800799533940015,
      "grad_norm": 0.2868156135082245,
      "learning_rate": 0.0001688146911519199,
      "loss": 1.238,
      "num_input_tokens_seen": 1317072,
      "step": 473,
      "train_runtime": 1186.826,
      "train_tokens_per_second": 1109.743
    },
    {
      "epoch": 0.03808835050925089,
      "grad_norm": 0.2835414707660675,
      "learning_rate": 0.00016874791318864776,
      "loss": 1.1234,
      "num_input_tokens_seen": 1320792,
      "step": 474,
      "train_runtime": 1189.6151,
      "train_tokens_per_second": 1110.268
    },
    {
      "epoch": 0.038168705679101626,
      "grad_norm": 0.32327404618263245,
      "learning_rate": 0.00016868113522537564,
      "loss": 0.9026,
      "num_input_tokens_seen": 1322676,
      "step": 475,
      "train_runtime": 1191.4105,
      "train_tokens_per_second": 1110.177
    },
    {
      "epoch": 0.03824906084895237,
      "grad_norm": 0.3147921562194824,
      "learning_rate": 0.0001686143572621035,
      "loss": 1.2454,
      "num_input_tokens_seen": 1325490,
      "step": 476,
      "train_runtime": 1193.6387,
      "train_tokens_per_second": 1110.462
    },
    {
      "epoch": 0.03832941601880311,
      "grad_norm": 0.3412136137485504,
      "learning_rate": 0.0001685475792988314,
      "loss": 0.9937,
      "num_input_tokens_seen": 1329268,
      "step": 477,
      "train_runtime": 1196.4157,
      "train_tokens_per_second": 1111.042
    },
    {
      "epoch": 0.03840977118865385,
      "grad_norm": 0.39507076144218445,
      "learning_rate": 0.00016848080133555928,
      "loss": 1.3339,
      "num_input_tokens_seen": 1331396,
      "step": 478,
      "train_runtime": 1198.2903,
      "train_tokens_per_second": 1111.08
    },
    {
      "epoch": 0.038490126358504594,
      "grad_norm": 0.3128511309623718,
      "learning_rate": 0.00016841402337228716,
      "loss": 1.3723,
      "num_input_tokens_seen": 1334386,
      "step": 479,
      "train_runtime": 1200.6237,
      "train_tokens_per_second": 1111.411
    },
    {
      "epoch": 0.03857048152835533,
      "grad_norm": 0.3059634268283844,
      "learning_rate": 0.00016834724540901503,
      "loss": 0.8782,
      "num_input_tokens_seen": 1337446,
      "step": 480,
      "train_runtime": 1202.9349,
      "train_tokens_per_second": 1111.819
    },
    {
      "epoch": 0.03865083669820607,
      "grad_norm": 0.3931996822357178,
      "learning_rate": 0.0001682804674457429,
      "loss": 0.8992,
      "num_input_tokens_seen": 1339258,
      "step": 481,
      "train_runtime": 1206.7839,
      "train_tokens_per_second": 1109.774
    },
    {
      "epoch": 0.03873119186805681,
      "grad_norm": 0.26833397150039673,
      "learning_rate": 0.00016821368948247077,
      "loss": 1.3182,
      "num_input_tokens_seen": 1342016,
      "step": 482,
      "train_runtime": 1208.9671,
      "train_tokens_per_second": 1110.052
    },
    {
      "epoch": 0.038811547037907554,
      "grad_norm": 0.283633828163147,
      "learning_rate": 0.00016814691151919868,
      "loss": 1.2343,
      "num_input_tokens_seen": 1345168,
      "step": 483,
      "train_runtime": 1211.3739,
      "train_tokens_per_second": 1110.448
    },
    {
      "epoch": 0.03889190220775829,
      "grad_norm": 0.24241532385349274,
      "learning_rate": 0.00016808013355592655,
      "loss": 0.6693,
      "num_input_tokens_seen": 1348988,
      "step": 484,
      "train_runtime": 1214.089,
      "train_tokens_per_second": 1111.111
    },
    {
      "epoch": 0.03897225737760903,
      "grad_norm": 0.30962204933166504,
      "learning_rate": 0.00016801335559265442,
      "loss": 0.794,
      "num_input_tokens_seen": 1351492,
      "step": 485,
      "train_runtime": 1216.1875,
      "train_tokens_per_second": 1111.253
    },
    {
      "epoch": 0.03905261254745977,
      "grad_norm": 0.3316614627838135,
      "learning_rate": 0.0001679465776293823,
      "loss": 1.1496,
      "num_input_tokens_seen": 1352990,
      "step": 486,
      "train_runtime": 1217.7667,
      "train_tokens_per_second": 1111.042
    },
    {
      "epoch": 0.039132967717310514,
      "grad_norm": 0.3337160050868988,
      "learning_rate": 0.0001678797996661102,
      "loss": 1.3319,
      "num_input_tokens_seen": 1355490,
      "step": 487,
      "train_runtime": 1219.8451,
      "train_tokens_per_second": 1111.199
    },
    {
      "epoch": 0.039213322887161256,
      "grad_norm": 0.33112165331840515,
      "learning_rate": 0.00016781302170283807,
      "loss": 1.2246,
      "num_input_tokens_seen": 1358564,
      "step": 488,
      "train_runtime": 1222.1597,
      "train_tokens_per_second": 1111.609
    },
    {
      "epoch": 0.03929367805701199,
      "grad_norm": 0.3269559442996979,
      "learning_rate": 0.00016774624373956597,
      "loss": 1.1281,
      "num_input_tokens_seen": 1361516,
      "step": 489,
      "train_runtime": 1224.4886,
      "train_tokens_per_second": 1111.906
    },
    {
      "epoch": 0.03937403322686273,
      "grad_norm": 0.31837156414985657,
      "learning_rate": 0.00016767946577629384,
      "loss": 1.0632,
      "num_input_tokens_seen": 1363794,
      "step": 490,
      "train_runtime": 1226.4391,
      "train_tokens_per_second": 1111.995
    },
    {
      "epoch": 0.039454388396713475,
      "grad_norm": 0.38089337944984436,
      "learning_rate": 0.00016761268781302171,
      "loss": 0.9146,
      "num_input_tokens_seen": 1366056,
      "step": 491,
      "train_runtime": 1228.3994,
      "train_tokens_per_second": 1112.062
    },
    {
      "epoch": 0.039534743566564216,
      "grad_norm": 0.42928919196128845,
      "learning_rate": 0.0001675459098497496,
      "loss": 0.9499,
      "num_input_tokens_seen": 1368594,
      "step": 492,
      "train_runtime": 1230.6799,
      "train_tokens_per_second": 1112.063
    },
    {
      "epoch": 0.03961509873641495,
      "grad_norm": 0.2621776759624481,
      "learning_rate": 0.0001674791318864775,
      "loss": 0.6951,
      "num_input_tokens_seen": 1372604,
      "step": 493,
      "train_runtime": 1233.5747,
      "train_tokens_per_second": 1112.704
    },
    {
      "epoch": 0.03969545390626569,
      "grad_norm": 0.40490996837615967,
      "learning_rate": 0.00016741235392320536,
      "loss": 1.3043,
      "num_input_tokens_seen": 1375610,
      "step": 494,
      "train_runtime": 1235.8826,
      "train_tokens_per_second": 1113.059
    },
    {
      "epoch": 0.039775809076116435,
      "grad_norm": 0.2883574962615967,
      "learning_rate": 0.00016734557595993323,
      "loss": 1.1695,
      "num_input_tokens_seen": 1378200,
      "step": 495,
      "train_runtime": 1237.9826,
      "train_tokens_per_second": 1113.263
    },
    {
      "epoch": 0.03985616424596718,
      "grad_norm": 0.36347633600234985,
      "learning_rate": 0.0001672787979966611,
      "loss": 1.1713,
      "num_input_tokens_seen": 1379932,
      "step": 496,
      "train_runtime": 1239.5904,
      "train_tokens_per_second": 1113.216
    },
    {
      "epoch": 0.03993651941581792,
      "grad_norm": 0.3250490725040436,
      "learning_rate": 0.00016721202003338898,
      "loss": 1.0827,
      "num_input_tokens_seen": 1381588,
      "step": 497,
      "train_runtime": 1241.2933,
      "train_tokens_per_second": 1113.023
    },
    {
      "epoch": 0.04001687458566865,
      "grad_norm": 0.31661948561668396,
      "learning_rate": 0.00016714524207011685,
      "loss": 0.9298,
      "num_input_tokens_seen": 1386302,
      "step": 498,
      "train_runtime": 1244.6013,
      "train_tokens_per_second": 1113.852
    },
    {
      "epoch": 0.040097229755519395,
      "grad_norm": 0.2635970711708069,
      "learning_rate": 0.00016707846410684475,
      "loss": 0.6909,
      "num_input_tokens_seen": 1388112,
      "step": 499,
      "train_runtime": 1246.2907,
      "train_tokens_per_second": 1113.795
    },
    {
      "epoch": 0.04017758492537014,
      "grad_norm": 0.285051554441452,
      "learning_rate": 0.00016701168614357263,
      "loss": 0.7754,
      "num_input_tokens_seen": 1391612,
      "step": 500,
      "train_runtime": 1248.9314,
      "train_tokens_per_second": 1114.242
    },
    {
      "epoch": 0.04025794009522088,
      "grad_norm": 0.3328403830528259,
      "learning_rate": 0.0001669449081803005,
      "loss": 1.4458,
      "num_input_tokens_seen": 1394868,
      "step": 501,
      "train_runtime": 1251.3154,
      "train_tokens_per_second": 1114.721
    },
    {
      "epoch": 0.040338295265071614,
      "grad_norm": 0.31206661462783813,
      "learning_rate": 0.00016687813021702837,
      "loss": 0.9529,
      "num_input_tokens_seen": 1397966,
      "step": 502,
      "train_runtime": 1253.7554,
      "train_tokens_per_second": 1115.023
    },
    {
      "epoch": 0.040418650434922355,
      "grad_norm": 0.3316718339920044,
      "learning_rate": 0.00016681135225375625,
      "loss": 1.2912,
      "num_input_tokens_seen": 1401884,
      "step": 503,
      "train_runtime": 1256.7215,
      "train_tokens_per_second": 1115.509
    },
    {
      "epoch": 0.0404990056047731,
      "grad_norm": 0.28495049476623535,
      "learning_rate": 0.00016674457429048415,
      "loss": 0.9016,
      "num_input_tokens_seen": 1404772,
      "step": 504,
      "train_runtime": 1258.8812,
      "train_tokens_per_second": 1115.889
    },
    {
      "epoch": 0.04057936077462384,
      "grad_norm": 0.30233266949653625,
      "learning_rate": 0.00016667779632721202,
      "loss": 1.0187,
      "num_input_tokens_seen": 1409318,
      "step": 505,
      "train_runtime": 1262.1301,
      "train_tokens_per_second": 1116.619
    },
    {
      "epoch": 0.04065971594447458,
      "grad_norm": 0.3341182768344879,
      "learning_rate": 0.00016661101836393992,
      "loss": 0.9593,
      "num_input_tokens_seen": 1412278,
      "step": 506,
      "train_runtime": 1264.4717,
      "train_tokens_per_second": 1116.892
    },
    {
      "epoch": 0.040740071114325316,
      "grad_norm": 0.28704357147216797,
      "learning_rate": 0.0001665442404006678,
      "loss": 1.001,
      "num_input_tokens_seen": 1415504,
      "step": 507,
      "train_runtime": 1266.8747,
      "train_tokens_per_second": 1117.32
    },
    {
      "epoch": 0.04082042628417606,
      "grad_norm": 0.26951321959495544,
      "learning_rate": 0.00016647746243739567,
      "loss": 0.9587,
      "num_input_tokens_seen": 1418684,
      "step": 508,
      "train_runtime": 1269.3245,
      "train_tokens_per_second": 1117.668
    },
    {
      "epoch": 0.0409007814540268,
      "grad_norm": 0.3407914936542511,
      "learning_rate": 0.00016641068447412357,
      "loss": 0.9534,
      "num_input_tokens_seen": 1421036,
      "step": 509,
      "train_runtime": 1271.3605,
      "train_tokens_per_second": 1117.729
    },
    {
      "epoch": 0.04098113662387754,
      "grad_norm": 0.33112484216690063,
      "learning_rate": 0.00016634390651085144,
      "loss": 0.8905,
      "num_input_tokens_seen": 1423300,
      "step": 510,
      "train_runtime": 1273.3026,
      "train_tokens_per_second": 1117.802
    },
    {
      "epoch": 0.041061491793728276,
      "grad_norm": 0.27271634340286255,
      "learning_rate": 0.0001662771285475793,
      "loss": 0.7882,
      "num_input_tokens_seen": 1426964,
      "step": 511,
      "train_runtime": 1278.2192,
      "train_tokens_per_second": 1116.369
    },
    {
      "epoch": 0.04114184696357902,
      "grad_norm": 0.2489190697669983,
      "learning_rate": 0.00016621035058430719,
      "loss": 0.7609,
      "num_input_tokens_seen": 1429752,
      "step": 512,
      "train_runtime": 1280.3282,
      "train_tokens_per_second": 1116.707
    },
    {
      "epoch": 0.04122220213342976,
      "grad_norm": 0.4314664602279663,
      "learning_rate": 0.00016614357262103506,
      "loss": 0.9647,
      "num_input_tokens_seen": 1432418,
      "step": 513,
      "train_runtime": 1282.5147,
      "train_tokens_per_second": 1116.882
    },
    {
      "epoch": 0.0413025573032805,
      "grad_norm": 0.37440216541290283,
      "learning_rate": 0.00016607679465776293,
      "loss": 1.3032,
      "num_input_tokens_seen": 1436178,
      "step": 514,
      "train_runtime": 1285.2827,
      "train_tokens_per_second": 1117.402
    },
    {
      "epoch": 0.04138291247313124,
      "grad_norm": 0.3475145697593689,
      "learning_rate": 0.00016601001669449083,
      "loss": 0.8235,
      "num_input_tokens_seen": 1438236,
      "step": 515,
      "train_runtime": 1287.0789,
      "train_tokens_per_second": 1117.442
    },
    {
      "epoch": 0.04146326764298198,
      "grad_norm": 0.29458504915237427,
      "learning_rate": 0.0001659432387312187,
      "loss": 1.0016,
      "num_input_tokens_seen": 1441264,
      "step": 516,
      "train_runtime": 1289.4622,
      "train_tokens_per_second": 1117.725
    },
    {
      "epoch": 0.04154362281283272,
      "grad_norm": 0.2998146414756775,
      "learning_rate": 0.00016587646076794658,
      "loss": 1.1158,
      "num_input_tokens_seen": 1444966,
      "step": 517,
      "train_runtime": 1292.2153,
      "train_tokens_per_second": 1118.208
    },
    {
      "epoch": 0.04162397798268346,
      "grad_norm": 0.3156857490539551,
      "learning_rate": 0.00016580968280467445,
      "loss": 0.9576,
      "num_input_tokens_seen": 1448814,
      "step": 518,
      "train_runtime": 1295.0533,
      "train_tokens_per_second": 1118.729
    },
    {
      "epoch": 0.041704333152534204,
      "grad_norm": 0.2988636791706085,
      "learning_rate": 0.00016574290484140233,
      "loss": 0.9327,
      "num_input_tokens_seen": 1452354,
      "step": 519,
      "train_runtime": 1297.6684,
      "train_tokens_per_second": 1119.203
    },
    {
      "epoch": 0.04178468832238494,
      "grad_norm": 0.29688844084739685,
      "learning_rate": 0.00016567612687813023,
      "loss": 1.2435,
      "num_input_tokens_seen": 1454658,
      "step": 520,
      "train_runtime": 1299.6956,
      "train_tokens_per_second": 1119.23
    },
    {
      "epoch": 0.04186504349223568,
      "grad_norm": 0.32755959033966064,
      "learning_rate": 0.0001656093489148581,
      "loss": 0.9829,
      "num_input_tokens_seen": 1456450,
      "step": 521,
      "train_runtime": 1301.4736,
      "train_tokens_per_second": 1119.078
    },
    {
      "epoch": 0.04194539866208642,
      "grad_norm": 0.33042067289352417,
      "learning_rate": 0.00016554257095158597,
      "loss": 1.1854,
      "num_input_tokens_seen": 1459110,
      "step": 522,
      "train_runtime": 1303.6084,
      "train_tokens_per_second": 1119.285
    },
    {
      "epoch": 0.042025753831937164,
      "grad_norm": 0.24316900968551636,
      "learning_rate": 0.00016547579298831387,
      "loss": 1.0388,
      "num_input_tokens_seen": 1461406,
      "step": 523,
      "train_runtime": 1305.4827,
      "train_tokens_per_second": 1119.437
    },
    {
      "epoch": 0.042106109001787906,
      "grad_norm": 0.2608068883419037,
      "learning_rate": 0.00016540901502504175,
      "loss": 0.797,
      "num_input_tokens_seen": 1463864,
      "step": 524,
      "train_runtime": 1307.4925,
      "train_tokens_per_second": 1119.596
    },
    {
      "epoch": 0.04218646417163864,
      "grad_norm": 0.2739805281162262,
      "learning_rate": 0.00016534223706176965,
      "loss": 0.9095,
      "num_input_tokens_seen": 1466956,
      "step": 525,
      "train_runtime": 1309.8869,
      "train_tokens_per_second": 1119.91
    },
    {
      "epoch": 0.04226681934148938,
      "grad_norm": 0.29957881569862366,
      "learning_rate": 0.00016527545909849752,
      "loss": 0.9904,
      "num_input_tokens_seen": 1470008,
      "step": 526,
      "train_runtime": 1312.2745,
      "train_tokens_per_second": 1120.199
    },
    {
      "epoch": 0.042347174511340124,
      "grad_norm": 0.31610366702079773,
      "learning_rate": 0.0001652086811352254,
      "loss": 1.329,
      "num_input_tokens_seen": 1471884,
      "step": 527,
      "train_runtime": 1313.9712,
      "train_tokens_per_second": 1120.18
    },
    {
      "epoch": 0.042427529681190866,
      "grad_norm": 0.2806980609893799,
      "learning_rate": 0.00016514190317195327,
      "loss": 0.8342,
      "num_input_tokens_seen": 1473638,
      "step": 528,
      "train_runtime": 1315.6966,
      "train_tokens_per_second": 1120.044
    },
    {
      "epoch": 0.0425078848510416,
      "grad_norm": 0.34865710139274597,
      "learning_rate": 0.00016507512520868114,
      "loss": 1.0074,
      "num_input_tokens_seen": 1477422,
      "step": 529,
      "train_runtime": 1318.4171,
      "train_tokens_per_second": 1120.603
    },
    {
      "epoch": 0.04258824002089234,
      "grad_norm": 0.29557326436042786,
      "learning_rate": 0.00016500834724540904,
      "loss": 1.0376,
      "num_input_tokens_seen": 1480596,
      "step": 530,
      "train_runtime": 1320.7853,
      "train_tokens_per_second": 1120.997
    },
    {
      "epoch": 0.042668595190743085,
      "grad_norm": 0.31514546275138855,
      "learning_rate": 0.0001649415692821369,
      "loss": 0.7571,
      "num_input_tokens_seen": 1484062,
      "step": 531,
      "train_runtime": 1323.5221,
      "train_tokens_per_second": 1121.297
    },
    {
      "epoch": 0.042748950360593826,
      "grad_norm": 0.3336114287376404,
      "learning_rate": 0.00016487479131886478,
      "loss": 1.2614,
      "num_input_tokens_seen": 1489870,
      "step": 532,
      "train_runtime": 1327.4899,
      "train_tokens_per_second": 1122.321
    },
    {
      "epoch": 0.04282930553044457,
      "grad_norm": 0.34417012333869934,
      "learning_rate": 0.00016480801335559266,
      "loss": 1.4205,
      "num_input_tokens_seen": 1492616,
      "step": 533,
      "train_runtime": 1329.7479,
      "train_tokens_per_second": 1122.48
    },
    {
      "epoch": 0.0429096607002953,
      "grad_norm": 0.34672799706459045,
      "learning_rate": 0.00016474123539232053,
      "loss": 1.0132,
      "num_input_tokens_seen": 1496298,
      "step": 534,
      "train_runtime": 1332.4516,
      "train_tokens_per_second": 1122.966
    },
    {
      "epoch": 0.042990015870146045,
      "grad_norm": 0.3274341821670532,
      "learning_rate": 0.0001646744574290484,
      "loss": 0.9711,
      "num_input_tokens_seen": 1498178,
      "step": 535,
      "train_runtime": 1334.293,
      "train_tokens_per_second": 1122.825
    },
    {
      "epoch": 0.04307037103999679,
      "grad_norm": 0.31558096408843994,
      "learning_rate": 0.0001646076794657763,
      "loss": 1.1928,
      "num_input_tokens_seen": 1500716,
      "step": 536,
      "train_runtime": 1336.2966,
      "train_tokens_per_second": 1123.041
    },
    {
      "epoch": 0.04315072620984753,
      "grad_norm": 0.2852008044719696,
      "learning_rate": 0.00016454090150250418,
      "loss": 0.9901,
      "num_input_tokens_seen": 1503318,
      "step": 537,
      "train_runtime": 1338.4725,
      "train_tokens_per_second": 1123.159
    },
    {
      "epoch": 0.043231081379698263,
      "grad_norm": 0.26600342988967896,
      "learning_rate": 0.00016447412353923205,
      "loss": 0.9885,
      "num_input_tokens_seen": 1506788,
      "step": 538,
      "train_runtime": 1341.1218,
      "train_tokens_per_second": 1123.528
    },
    {
      "epoch": 0.043311436549549005,
      "grad_norm": 0.257356196641922,
      "learning_rate": 0.00016440734557595992,
      "loss": 0.6195,
      "num_input_tokens_seen": 1508954,
      "step": 539,
      "train_runtime": 1342.9701,
      "train_tokens_per_second": 1123.595
    },
    {
      "epoch": 0.04339179171939975,
      "grad_norm": 0.36241650581359863,
      "learning_rate": 0.00016434056761268782,
      "loss": 0.8107,
      "num_input_tokens_seen": 1514520,
      "step": 540,
      "train_runtime": 1346.8254,
      "train_tokens_per_second": 1124.511
    },
    {
      "epoch": 0.04347214688925049,
      "grad_norm": 0.30938225984573364,
      "learning_rate": 0.0001642737896494157,
      "loss": 0.9419,
      "num_input_tokens_seen": 1517146,
      "step": 541,
      "train_runtime": 1351.0155,
      "train_tokens_per_second": 1122.967
    },
    {
      "epoch": 0.04355250205910123,
      "grad_norm": 0.30551084876060486,
      "learning_rate": 0.0001642070116861436,
      "loss": 0.7459,
      "num_input_tokens_seen": 1518960,
      "step": 542,
      "train_runtime": 1352.7273,
      "train_tokens_per_second": 1122.887
    },
    {
      "epoch": 0.043632857228951966,
      "grad_norm": 0.3435930013656616,
      "learning_rate": 0.00016414023372287147,
      "loss": 0.8315,
      "num_input_tokens_seen": 1521962,
      "step": 543,
      "train_runtime": 1355.0508,
      "train_tokens_per_second": 1123.177
    },
    {
      "epoch": 0.04371321239880271,
      "grad_norm": 0.3326947093009949,
      "learning_rate": 0.00016407345575959934,
      "loss": 1.3582,
      "num_input_tokens_seen": 1524492,
      "step": 544,
      "train_runtime": 1357.1273,
      "train_tokens_per_second": 1123.323
    },
    {
      "epoch": 0.04379356756865345,
      "grad_norm": 0.2878602147102356,
      "learning_rate": 0.00016400667779632722,
      "loss": 1.0477,
      "num_input_tokens_seen": 1526732,
      "step": 545,
      "train_runtime": 1359.0122,
      "train_tokens_per_second": 1123.413
    },
    {
      "epoch": 0.04387392273850419,
      "grad_norm": 0.35480237007141113,
      "learning_rate": 0.00016393989983305512,
      "loss": 0.922,
      "num_input_tokens_seen": 1528872,
      "step": 546,
      "train_runtime": 1360.8368,
      "train_tokens_per_second": 1123.479
    },
    {
      "epoch": 0.043954277908354926,
      "grad_norm": 0.40055155754089355,
      "learning_rate": 0.000163873121869783,
      "loss": 1.1884,
      "num_input_tokens_seen": 1531854,
      "step": 547,
      "train_runtime": 1363.1985,
      "train_tokens_per_second": 1123.72
    },
    {
      "epoch": 0.04403463307820567,
      "grad_norm": 0.2950405776500702,
      "learning_rate": 0.00016380634390651086,
      "loss": 1.0828,
      "num_input_tokens_seen": 1534416,
      "step": 548,
      "train_runtime": 1365.2619,
      "train_tokens_per_second": 1123.899
    },
    {
      "epoch": 0.04411498824805641,
      "grad_norm": 0.3034539520740509,
      "learning_rate": 0.00016373956594323874,
      "loss": 1.0698,
      "num_input_tokens_seen": 1538264,
      "step": 549,
      "train_runtime": 1368.0231,
      "train_tokens_per_second": 1124.443
    },
    {
      "epoch": 0.04419534341790715,
      "grad_norm": 0.3034869432449341,
      "learning_rate": 0.0001636727879799666,
      "loss": 0.8759,
      "num_input_tokens_seen": 1542434,
      "step": 550,
      "train_runtime": 1371.052,
      "train_tokens_per_second": 1125.0
    },
    {
      "epoch": 0.04427569858775789,
      "grad_norm": 0.4061206579208374,
      "learning_rate": 0.00016360601001669448,
      "loss": 1.0151,
      "num_input_tokens_seen": 1543824,
      "step": 551,
      "train_runtime": 1372.6307,
      "train_tokens_per_second": 1124.719
    },
    {
      "epoch": 0.04435605375760863,
      "grad_norm": 0.26810020208358765,
      "learning_rate": 0.00016353923205342238,
      "loss": 0.7532,
      "num_input_tokens_seen": 1546394,
      "step": 552,
      "train_runtime": 1374.7568,
      "train_tokens_per_second": 1124.849
    },
    {
      "epoch": 0.04443640892745937,
      "grad_norm": 0.3383218050003052,
      "learning_rate": 0.00016347245409015026,
      "loss": 0.9789,
      "num_input_tokens_seen": 1548030,
      "step": 553,
      "train_runtime": 1376.5011,
      "train_tokens_per_second": 1124.612
    },
    {
      "epoch": 0.04451676409731011,
      "grad_norm": 0.5009167194366455,
      "learning_rate": 0.00016340567612687813,
      "loss": 1.1864,
      "num_input_tokens_seen": 1549794,
      "step": 554,
      "train_runtime": 1378.2539,
      "train_tokens_per_second": 1124.462
    },
    {
      "epoch": 0.044597119267160854,
      "grad_norm": 0.3085728585720062,
      "learning_rate": 0.000163338898163606,
      "loss": 0.8579,
      "num_input_tokens_seen": 1552074,
      "step": 555,
      "train_runtime": 1380.2922,
      "train_tokens_per_second": 1124.453
    },
    {
      "epoch": 0.04467747443701159,
      "grad_norm": 0.2872931659221649,
      "learning_rate": 0.00016327212020033388,
      "loss": 0.9283,
      "num_input_tokens_seen": 1556052,
      "step": 556,
      "train_runtime": 1383.1055,
      "train_tokens_per_second": 1125.042
    },
    {
      "epoch": 0.04475782960686233,
      "grad_norm": 0.33563920855522156,
      "learning_rate": 0.00016320534223706178,
      "loss": 0.9859,
      "num_input_tokens_seen": 1558014,
      "step": 557,
      "train_runtime": 1384.9456,
      "train_tokens_per_second": 1124.964
    },
    {
      "epoch": 0.04483818477671307,
      "grad_norm": 0.30859917402267456,
      "learning_rate": 0.00016313856427378965,
      "loss": 1.0199,
      "num_input_tokens_seen": 1559734,
      "step": 558,
      "train_runtime": 1386.5523,
      "train_tokens_per_second": 1124.901
    },
    {
      "epoch": 0.044918539946563814,
      "grad_norm": 0.40110355615615845,
      "learning_rate": 0.00016307178631051755,
      "loss": 1.0607,
      "num_input_tokens_seen": 1562466,
      "step": 559,
      "train_runtime": 1388.7789,
      "train_tokens_per_second": 1125.065
    },
    {
      "epoch": 0.044998895116414556,
      "grad_norm": 0.5635671615600586,
      "learning_rate": 0.00016300500834724542,
      "loss": 1.2245,
      "num_input_tokens_seen": 1564642,
      "step": 560,
      "train_runtime": 1390.6863,
      "train_tokens_per_second": 1125.086
    },
    {
      "epoch": 0.04507925028626529,
      "grad_norm": 0.4702927768230438,
      "learning_rate": 0.0001629382303839733,
      "loss": 0.7973,
      "num_input_tokens_seen": 1567322,
      "step": 561,
      "train_runtime": 1392.774,
      "train_tokens_per_second": 1125.324
    },
    {
      "epoch": 0.04515960545611603,
      "grad_norm": 0.3103383481502533,
      "learning_rate": 0.0001628714524207012,
      "loss": 1.0033,
      "num_input_tokens_seen": 1570110,
      "step": 562,
      "train_runtime": 1394.8792,
      "train_tokens_per_second": 1125.624
    },
    {
      "epoch": 0.045239960625966774,
      "grad_norm": 0.37203335762023926,
      "learning_rate": 0.00016280467445742907,
      "loss": 0.924,
      "num_input_tokens_seen": 1571484,
      "step": 563,
      "train_runtime": 1396.4245,
      "train_tokens_per_second": 1125.363
    },
    {
      "epoch": 0.045320315795817516,
      "grad_norm": 0.276150643825531,
      "learning_rate": 0.00016273789649415694,
      "loss": 0.9249,
      "num_input_tokens_seen": 1573506,
      "step": 564,
      "train_runtime": 1398.1345,
      "train_tokens_per_second": 1125.433
    },
    {
      "epoch": 0.04540067096566825,
      "grad_norm": 0.3232124149799347,
      "learning_rate": 0.00016267111853088482,
      "loss": 1.3818,
      "num_input_tokens_seen": 1576608,
      "step": 565,
      "train_runtime": 1400.5972,
      "train_tokens_per_second": 1125.668
    },
    {
      "epoch": 0.04548102613551899,
      "grad_norm": 0.3408963680267334,
      "learning_rate": 0.0001626043405676127,
      "loss": 1.213,
      "num_input_tokens_seen": 1578858,
      "step": 566,
      "train_runtime": 1402.6764,
      "train_tokens_per_second": 1125.604
    },
    {
      "epoch": 0.045561381305369734,
      "grad_norm": 0.26941266655921936,
      "learning_rate": 0.00016253756260434056,
      "loss": 0.9142,
      "num_input_tokens_seen": 1581898,
      "step": 567,
      "train_runtime": 1404.953,
      "train_tokens_per_second": 1125.944
    },
    {
      "epoch": 0.045641736475220476,
      "grad_norm": 0.2751913070678711,
      "learning_rate": 0.00016247078464106846,
      "loss": 0.5027,
      "num_input_tokens_seen": 1583358,
      "step": 568,
      "train_runtime": 1406.5934,
      "train_tokens_per_second": 1125.669
    },
    {
      "epoch": 0.04572209164507122,
      "grad_norm": 0.25977325439453125,
      "learning_rate": 0.00016240400667779634,
      "loss": 0.57,
      "num_input_tokens_seen": 1586514,
      "step": 569,
      "train_runtime": 1409.088,
      "train_tokens_per_second": 1125.916
    },
    {
      "epoch": 0.04580244681492195,
      "grad_norm": 0.28802600502967834,
      "learning_rate": 0.0001623372287145242,
      "loss": 0.9104,
      "num_input_tokens_seen": 1588972,
      "step": 570,
      "train_runtime": 1411.0445,
      "train_tokens_per_second": 1126.096
    },
    {
      "epoch": 0.045882801984772695,
      "grad_norm": 0.334716796875,
      "learning_rate": 0.00016227045075125208,
      "loss": 1.2295,
      "num_input_tokens_seen": 1592682,
      "step": 571,
      "train_runtime": 1415.8663,
      "train_tokens_per_second": 1124.882
    },
    {
      "epoch": 0.04596315715462344,
      "grad_norm": 0.35369017720222473,
      "learning_rate": 0.00016220367278797996,
      "loss": 0.9268,
      "num_input_tokens_seen": 1594704,
      "step": 572,
      "train_runtime": 1417.7121,
      "train_tokens_per_second": 1124.843
    },
    {
      "epoch": 0.04604351232447418,
      "grad_norm": 0.30256491899490356,
      "learning_rate": 0.00016213689482470786,
      "loss": 0.8199,
      "num_input_tokens_seen": 1597136,
      "step": 573,
      "train_runtime": 1419.6891,
      "train_tokens_per_second": 1124.99
    },
    {
      "epoch": 0.04612386749432491,
      "grad_norm": 0.45038849115371704,
      "learning_rate": 0.00016207011686143573,
      "loss": 0.7788,
      "num_input_tokens_seen": 1599230,
      "step": 574,
      "train_runtime": 1421.5377,
      "train_tokens_per_second": 1125.0
    },
    {
      "epoch": 0.046204222664175655,
      "grad_norm": 0.30017659068107605,
      "learning_rate": 0.0001620033388981636,
      "loss": 1.1283,
      "num_input_tokens_seen": 1601638,
      "step": 575,
      "train_runtime": 1423.5342,
      "train_tokens_per_second": 1125.114
    },
    {
      "epoch": 0.0462845778340264,
      "grad_norm": 0.3505881726741791,
      "learning_rate": 0.0001619365609348915,
      "loss": 1.1406,
      "num_input_tokens_seen": 1605180,
      "step": 576,
      "train_runtime": 1426.2371,
      "train_tokens_per_second": 1125.465
    },
    {
      "epoch": 0.04636493300387714,
      "grad_norm": 0.2925964295864105,
      "learning_rate": 0.00016186978297161938,
      "loss": 0.921,
      "num_input_tokens_seen": 1607854,
      "step": 577,
      "train_runtime": 1428.5118,
      "train_tokens_per_second": 1125.545
    },
    {
      "epoch": 0.04644528817372788,
      "grad_norm": 0.27441343665122986,
      "learning_rate": 0.00016180300500834728,
      "loss": 0.9909,
      "num_input_tokens_seen": 1609836,
      "step": 578,
      "train_runtime": 1430.3275,
      "train_tokens_per_second": 1125.502
    },
    {
      "epoch": 0.046525643343578615,
      "grad_norm": 0.3009042739868164,
      "learning_rate": 0.00016173622704507515,
      "loss": 0.9685,
      "num_input_tokens_seen": 1612046,
      "step": 579,
      "train_runtime": 1432.1628,
      "train_tokens_per_second": 1125.602
    },
    {
      "epoch": 0.04660599851342936,
      "grad_norm": 0.2828468382358551,
      "learning_rate": 0.00016166944908180302,
      "loss": 1.087,
      "num_input_tokens_seen": 1617292,
      "step": 580,
      "train_runtime": 1435.8583,
      "train_tokens_per_second": 1126.359
    },
    {
      "epoch": 0.0466863536832801,
      "grad_norm": 0.21674178540706635,
      "learning_rate": 0.0001616026711185309,
      "loss": 0.5879,
      "num_input_tokens_seen": 1621710,
      "step": 581,
      "train_runtime": 1438.8474,
      "train_tokens_per_second": 1127.09
    },
    {
      "epoch": 0.04676670885313084,
      "grad_norm": 0.37525784969329834,
      "learning_rate": 0.00016153589315525877,
      "loss": 0.8292,
      "num_input_tokens_seen": 1625344,
      "step": 582,
      "train_runtime": 1441.5424,
      "train_tokens_per_second": 1127.503
    },
    {
      "epoch": 0.046847064022981576,
      "grad_norm": 0.3150649666786194,
      "learning_rate": 0.00016146911519198664,
      "loss": 1.2162,
      "num_input_tokens_seen": 1628084,
      "step": 583,
      "train_runtime": 1443.7235,
      "train_tokens_per_second": 1127.698
    },
    {
      "epoch": 0.04692741919283232,
      "grad_norm": 0.29897260665893555,
      "learning_rate": 0.00016140233722871454,
      "loss": 1.0659,
      "num_input_tokens_seen": 1630780,
      "step": 584,
      "train_runtime": 1445.8491,
      "train_tokens_per_second": 1127.905
    },
    {
      "epoch": 0.04700777436268306,
      "grad_norm": 0.27850741147994995,
      "learning_rate": 0.00016133555926544241,
      "loss": 1.1754,
      "num_input_tokens_seen": 1633554,
      "step": 585,
      "train_runtime": 1448.0572,
      "train_tokens_per_second": 1128.1
    },
    {
      "epoch": 0.0470881295325338,
      "grad_norm": 0.3504570424556732,
      "learning_rate": 0.0001612687813021703,
      "loss": 0.9959,
      "num_input_tokens_seen": 1635914,
      "step": 586,
      "train_runtime": 1450.1219,
      "train_tokens_per_second": 1128.122
    },
    {
      "epoch": 0.04716848470238454,
      "grad_norm": 0.29737618565559387,
      "learning_rate": 0.00016120200333889816,
      "loss": 1.1948,
      "num_input_tokens_seen": 1638884,
      "step": 587,
      "train_runtime": 1452.4476,
      "train_tokens_per_second": 1128.36
    },
    {
      "epoch": 0.04724883987223528,
      "grad_norm": 0.3277566134929657,
      "learning_rate": 0.00016113522537562603,
      "loss": 1.3117,
      "num_input_tokens_seen": 1642050,
      "step": 588,
      "train_runtime": 1454.8081,
      "train_tokens_per_second": 1128.706
    },
    {
      "epoch": 0.04732919504208602,
      "grad_norm": 0.34753653407096863,
      "learning_rate": 0.00016106844741235393,
      "loss": 1.056,
      "num_input_tokens_seen": 1643628,
      "step": 589,
      "train_runtime": 1456.4866,
      "train_tokens_per_second": 1128.488
    },
    {
      "epoch": 0.04740955021193676,
      "grad_norm": 0.2398485243320465,
      "learning_rate": 0.0001610016694490818,
      "loss": 0.6869,
      "num_input_tokens_seen": 1645652,
      "step": 590,
      "train_runtime": 1458.3916,
      "train_tokens_per_second": 1128.402
    },
    {
      "epoch": 0.0474899053817875,
      "grad_norm": 0.26019415259361267,
      "learning_rate": 0.00016093489148580968,
      "loss": 0.7328,
      "num_input_tokens_seen": 1648080,
      "step": 591,
      "train_runtime": 1460.4683,
      "train_tokens_per_second": 1128.46
    },
    {
      "epoch": 0.04757026055163824,
      "grad_norm": 0.269775927066803,
      "learning_rate": 0.00016086811352253755,
      "loss": 0.9202,
      "num_input_tokens_seen": 1650988,
      "step": 592,
      "train_runtime": 1462.6425,
      "train_tokens_per_second": 1128.771
    },
    {
      "epoch": 0.04765061572148898,
      "grad_norm": 0.2770116627216339,
      "learning_rate": 0.00016080133555926545,
      "loss": 0.8383,
      "num_input_tokens_seen": 1653452,
      "step": 593,
      "train_runtime": 1464.6359,
      "train_tokens_per_second": 1128.917
    },
    {
      "epoch": 0.04773097089133972,
      "grad_norm": 0.36373430490493774,
      "learning_rate": 0.00016073455759599333,
      "loss": 0.8562,
      "num_input_tokens_seen": 1655594,
      "step": 594,
      "train_runtime": 1466.5373,
      "train_tokens_per_second": 1128.914
    },
    {
      "epoch": 0.047811326061190464,
      "grad_norm": 0.28949663043022156,
      "learning_rate": 0.00016066777963272123,
      "loss": 1.1579,
      "num_input_tokens_seen": 1657722,
      "step": 595,
      "train_runtime": 1468.3706,
      "train_tokens_per_second": 1128.953
    },
    {
      "epoch": 0.047891681231041205,
      "grad_norm": 0.30830082297325134,
      "learning_rate": 0.0001606010016694491,
      "loss": 1.1471,
      "num_input_tokens_seen": 1659910,
      "step": 596,
      "train_runtime": 1470.3253,
      "train_tokens_per_second": 1128.941
    },
    {
      "epoch": 0.04797203640089194,
      "grad_norm": 0.347534716129303,
      "learning_rate": 0.00016053422370617697,
      "loss": 1.2735,
      "num_input_tokens_seen": 1663292,
      "step": 597,
      "train_runtime": 1472.8028,
      "train_tokens_per_second": 1129.338
    },
    {
      "epoch": 0.04805239157074268,
      "grad_norm": 0.4051826298236847,
      "learning_rate": 0.00016046744574290485,
      "loss": 0.7937,
      "num_input_tokens_seen": 1665434,
      "step": 598,
      "train_runtime": 1474.6824,
      "train_tokens_per_second": 1129.351
    },
    {
      "epoch": 0.048132746740593424,
      "grad_norm": 0.28954237699508667,
      "learning_rate": 0.00016040066777963272,
      "loss": 0.8046,
      "num_input_tokens_seen": 1668146,
      "step": 599,
      "train_runtime": 1476.8764,
      "train_tokens_per_second": 1129.51
    },
    {
      "epoch": 0.048213101910444166,
      "grad_norm": 0.31506192684173584,
      "learning_rate": 0.00016033388981636062,
      "loss": 0.9025,
      "num_input_tokens_seen": 1669976,
      "step": 600,
      "train_runtime": 1478.5583,
      "train_tokens_per_second": 1129.462
    },
    {
      "epoch": 0.048213101910444166,
      "eval_loss": 1.0045510530471802,
      "eval_runtime": 92.4855,
      "eval_samples_per_second": 10.748,
      "eval_steps_per_second": 5.374,
      "num_input_tokens_seen": 1669976,
      "step": 600
    },
    {
      "epoch": 0.0482934570802949,
      "grad_norm": 0.2952899634838104,
      "learning_rate": 0.0001602671118530885,
      "loss": 0.9183,
      "num_input_tokens_seen": 1672228,
      "step": 601,
      "train_runtime": 1575.0426,
      "train_tokens_per_second": 1061.703
    },
    {
      "epoch": 0.04837381225014564,
      "grad_norm": 0.37234246730804443,
      "learning_rate": 0.00016020033388981637,
      "loss": 1.015,
      "num_input_tokens_seen": 1675632,
      "step": 602,
      "train_runtime": 1577.6241,
      "train_tokens_per_second": 1062.124
    },
    {
      "epoch": 0.048454167419996384,
      "grad_norm": 0.3364265561103821,
      "learning_rate": 0.00016013355592654424,
      "loss": 0.8864,
      "num_input_tokens_seen": 1679108,
      "step": 603,
      "train_runtime": 1580.2599,
      "train_tokens_per_second": 1062.552
    },
    {
      "epoch": 0.048534522589847126,
      "grad_norm": 0.25584179162979126,
      "learning_rate": 0.0001600667779632721,
      "loss": 0.8234,
      "num_input_tokens_seen": 1682176,
      "step": 604,
      "train_runtime": 1582.6328,
      "train_tokens_per_second": 1062.897
    },
    {
      "epoch": 0.04861487775969787,
      "grad_norm": 0.2942083775997162,
      "learning_rate": 0.00016,
      "loss": 0.8421,
      "num_input_tokens_seen": 1685508,
      "step": 605,
      "train_runtime": 1585.0741,
      "train_tokens_per_second": 1063.362
    },
    {
      "epoch": 0.0486952329295486,
      "grad_norm": 0.32731980085372925,
      "learning_rate": 0.00015993322203672789,
      "loss": 0.8268,
      "num_input_tokens_seen": 1687128,
      "step": 606,
      "train_runtime": 1586.6616,
      "train_tokens_per_second": 1063.319
    },
    {
      "epoch": 0.048775588099399345,
      "grad_norm": 0.3258304297924042,
      "learning_rate": 0.00015986644407345576,
      "loss": 0.993,
      "num_input_tokens_seen": 1689832,
      "step": 607,
      "train_runtime": 1588.8773,
      "train_tokens_per_second": 1063.538
    },
    {
      "epoch": 0.048855943269250086,
      "grad_norm": 0.39065971970558167,
      "learning_rate": 0.00015979966611018363,
      "loss": 1.0378,
      "num_input_tokens_seen": 1692632,
      "step": 608,
      "train_runtime": 1591.0951,
      "train_tokens_per_second": 1063.816
    },
    {
      "epoch": 0.04893629843910083,
      "grad_norm": 0.30499234795570374,
      "learning_rate": 0.0001597328881469115,
      "loss": 0.9139,
      "num_input_tokens_seen": 1694372,
      "step": 609,
      "train_runtime": 1592.7801,
      "train_tokens_per_second": 1063.783
    },
    {
      "epoch": 0.04901665360895156,
      "grad_norm": 0.29981762170791626,
      "learning_rate": 0.0001596661101836394,
      "loss": 0.8803,
      "num_input_tokens_seen": 1696954,
      "step": 610,
      "train_runtime": 1594.8667,
      "train_tokens_per_second": 1064.01
    },
    {
      "epoch": 0.049097008778802305,
      "grad_norm": 0.27179670333862305,
      "learning_rate": 0.00015959933222036728,
      "loss": 0.7833,
      "num_input_tokens_seen": 1700184,
      "step": 611,
      "train_runtime": 1597.3495,
      "train_tokens_per_second": 1064.378
    },
    {
      "epoch": 0.04917736394865305,
      "grad_norm": 0.2974316477775574,
      "learning_rate": 0.00015953255425709518,
      "loss": 0.7587,
      "num_input_tokens_seen": 1703114,
      "step": 612,
      "train_runtime": 1599.7017,
      "train_tokens_per_second": 1064.645
    },
    {
      "epoch": 0.04925771911850379,
      "grad_norm": 0.29061052203178406,
      "learning_rate": 0.00015946577629382305,
      "loss": 1.0839,
      "num_input_tokens_seen": 1706324,
      "step": 613,
      "train_runtime": 1602.0616,
      "train_tokens_per_second": 1065.08
    },
    {
      "epoch": 0.04933807428835453,
      "grad_norm": 0.31000691652297974,
      "learning_rate": 0.00015939899833055093,
      "loss": 1.1214,
      "num_input_tokens_seen": 1707674,
      "step": 614,
      "train_runtime": 1603.6048,
      "train_tokens_per_second": 1064.897
    },
    {
      "epoch": 0.049418429458205265,
      "grad_norm": 0.30445772409439087,
      "learning_rate": 0.0001593322203672788,
      "loss": 1.2307,
      "num_input_tokens_seen": 1709922,
      "step": 615,
      "train_runtime": 1605.5598,
      "train_tokens_per_second": 1065.0
    },
    {
      "epoch": 0.04949878462805601,
      "grad_norm": 0.3193359971046448,
      "learning_rate": 0.0001592654424040067,
      "loss": 0.8299,
      "num_input_tokens_seen": 1712346,
      "step": 616,
      "train_runtime": 1607.5376,
      "train_tokens_per_second": 1065.198
    },
    {
      "epoch": 0.04957913979790675,
      "grad_norm": 0.33989283442497253,
      "learning_rate": 0.00015919866444073457,
      "loss": 1.0256,
      "num_input_tokens_seen": 1714932,
      "step": 617,
      "train_runtime": 1609.6151,
      "train_tokens_per_second": 1065.43
    },
    {
      "epoch": 0.04965949496775749,
      "grad_norm": 0.2855062186717987,
      "learning_rate": 0.00015913188647746245,
      "loss": 1.0201,
      "num_input_tokens_seen": 1716728,
      "step": 618,
      "train_runtime": 1611.4335,
      "train_tokens_per_second": 1065.342
    },
    {
      "epoch": 0.049739850137608226,
      "grad_norm": 0.30604907870292664,
      "learning_rate": 0.00015906510851419032,
      "loss": 1.2372,
      "num_input_tokens_seen": 1720144,
      "step": 619,
      "train_runtime": 1613.9733,
      "train_tokens_per_second": 1065.782
    },
    {
      "epoch": 0.04982020530745897,
      "grad_norm": 0.28403913974761963,
      "learning_rate": 0.0001589983305509182,
      "loss": 0.9487,
      "num_input_tokens_seen": 1723338,
      "step": 620,
      "train_runtime": 1616.3836,
      "train_tokens_per_second": 1066.169
    },
    {
      "epoch": 0.04990056047730971,
      "grad_norm": 0.298054963350296,
      "learning_rate": 0.0001589315525876461,
      "loss": 1.1147,
      "num_input_tokens_seen": 1725456,
      "step": 621,
      "train_runtime": 1618.2846,
      "train_tokens_per_second": 1066.225
    },
    {
      "epoch": 0.04998091564716045,
      "grad_norm": 0.3101802468299866,
      "learning_rate": 0.00015886477462437397,
      "loss": 0.8578,
      "num_input_tokens_seen": 1727466,
      "step": 622,
      "train_runtime": 1620.136,
      "train_tokens_per_second": 1066.248
    },
    {
      "epoch": 0.05006127081701119,
      "grad_norm": 0.318127304315567,
      "learning_rate": 0.00015879799666110184,
      "loss": 1.1982,
      "num_input_tokens_seen": 1730740,
      "step": 623,
      "train_runtime": 1622.6157,
      "train_tokens_per_second": 1066.636
    },
    {
      "epoch": 0.05014162598686193,
      "grad_norm": 0.2606166899204254,
      "learning_rate": 0.0001587312186978297,
      "loss": 0.8735,
      "num_input_tokens_seen": 1732610,
      "step": 624,
      "train_runtime": 1624.4236,
      "train_tokens_per_second": 1066.6
    },
    {
      "epoch": 0.05022198115671267,
      "grad_norm": 0.2978360950946808,
      "learning_rate": 0.00015866444073455758,
      "loss": 0.947,
      "num_input_tokens_seen": 1734532,
      "step": 625,
      "train_runtime": 1626.2926,
      "train_tokens_per_second": 1066.556
    },
    {
      "epoch": 0.05030233632656341,
      "grad_norm": 0.24004898965358734,
      "learning_rate": 0.00015859766277128548,
      "loss": 0.7807,
      "num_input_tokens_seen": 1736410,
      "step": 626,
      "train_runtime": 1627.9219,
      "train_tokens_per_second": 1066.642
    },
    {
      "epoch": 0.05038269149641415,
      "grad_norm": 0.32084575295448303,
      "learning_rate": 0.00015853088480801336,
      "loss": 1.1471,
      "num_input_tokens_seen": 1739552,
      "step": 627,
      "train_runtime": 1630.3991,
      "train_tokens_per_second": 1066.949
    },
    {
      "epoch": 0.05046304666626489,
      "grad_norm": 0.2868972420692444,
      "learning_rate": 0.00015846410684474123,
      "loss": 0.9481,
      "num_input_tokens_seen": 1741818,
      "step": 628,
      "train_runtime": 1632.3224,
      "train_tokens_per_second": 1067.08
    },
    {
      "epoch": 0.05054340183611563,
      "grad_norm": 0.28634634613990784,
      "learning_rate": 0.00015839732888146913,
      "loss": 0.8362,
      "num_input_tokens_seen": 1744374,
      "step": 629,
      "train_runtime": 1634.5174,
      "train_tokens_per_second": 1067.21
    },
    {
      "epoch": 0.05062375700596637,
      "grad_norm": 0.28812044858932495,
      "learning_rate": 0.000158330550918197,
      "loss": 1.1386,
      "num_input_tokens_seen": 1746560,
      "step": 630,
      "train_runtime": 1636.4281,
      "train_tokens_per_second": 1067.3
    },
    {
      "epoch": 0.05070411217581711,
      "grad_norm": 0.25034233927726746,
      "learning_rate": 0.00015826377295492488,
      "loss": 1.0815,
      "num_input_tokens_seen": 1750862,
      "step": 631,
      "train_runtime": 1641.2506,
      "train_tokens_per_second": 1066.785
    },
    {
      "epoch": 0.050784467345667855,
      "grad_norm": 0.3361033797264099,
      "learning_rate": 0.00015819699499165278,
      "loss": 0.9215,
      "num_input_tokens_seen": 1756678,
      "step": 632,
      "train_runtime": 1645.3083,
      "train_tokens_per_second": 1067.689
    },
    {
      "epoch": 0.05086482251551859,
      "grad_norm": 0.30996039509773254,
      "learning_rate": 0.00015813021702838065,
      "loss": 1.1946,
      "num_input_tokens_seen": 1758616,
      "step": 633,
      "train_runtime": 1647.0662,
      "train_tokens_per_second": 1067.726
    },
    {
      "epoch": 0.05094517768536933,
      "grad_norm": 0.40377989411354065,
      "learning_rate": 0.00015806343906510852,
      "loss": 0.8723,
      "num_input_tokens_seen": 1761132,
      "step": 634,
      "train_runtime": 1649.1157,
      "train_tokens_per_second": 1067.925
    },
    {
      "epoch": 0.051025532855220074,
      "grad_norm": 0.37052226066589355,
      "learning_rate": 0.0001579966611018364,
      "loss": 1.3183,
      "num_input_tokens_seen": 1764814,
      "step": 635,
      "train_runtime": 1651.7167,
      "train_tokens_per_second": 1068.473
    },
    {
      "epoch": 0.051105888025070816,
      "grad_norm": 0.2999049723148346,
      "learning_rate": 0.00015792988313856427,
      "loss": 1.1808,
      "num_input_tokens_seen": 1767792,
      "step": 636,
      "train_runtime": 1653.9548,
      "train_tokens_per_second": 1068.827
    },
    {
      "epoch": 0.05118624319492155,
      "grad_norm": 0.29953721165657043,
      "learning_rate": 0.00015786310517529217,
      "loss": 1.1399,
      "num_input_tokens_seen": 1771904,
      "step": 637,
      "train_runtime": 1656.8735,
      "train_tokens_per_second": 1069.426
    },
    {
      "epoch": 0.05126659836477229,
      "grad_norm": 0.3269258439540863,
      "learning_rate": 0.00015779632721202004,
      "loss": 1.1387,
      "num_input_tokens_seen": 1777092,
      "step": 638,
      "train_runtime": 1660.5696,
      "train_tokens_per_second": 1070.17
    },
    {
      "epoch": 0.051346953534623034,
      "grad_norm": 0.399961918592453,
      "learning_rate": 0.00015772954924874792,
      "loss": 1.0384,
      "num_input_tokens_seen": 1779902,
      "step": 639,
      "train_runtime": 1662.7712,
      "train_tokens_per_second": 1070.443
    },
    {
      "epoch": 0.051427308704473776,
      "grad_norm": 0.3907172977924347,
      "learning_rate": 0.0001576627712854758,
      "loss": 0.7542,
      "num_input_tokens_seen": 1783698,
      "step": 640,
      "train_runtime": 1665.4589,
      "train_tokens_per_second": 1070.995
    },
    {
      "epoch": 0.05150766387432452,
      "grad_norm": 0.3268730640411377,
      "learning_rate": 0.00015759599332220366,
      "loss": 1.0357,
      "num_input_tokens_seen": 1785036,
      "step": 641,
      "train_runtime": 1667.0452,
      "train_tokens_per_second": 1070.778
    },
    {
      "epoch": 0.05158801904417525,
      "grad_norm": 0.33701813220977783,
      "learning_rate": 0.00015752921535893156,
      "loss": 1.0021,
      "num_input_tokens_seen": 1787714,
      "step": 642,
      "train_runtime": 1669.1885,
      "train_tokens_per_second": 1071.008
    },
    {
      "epoch": 0.051668374214025994,
      "grad_norm": 0.38054177165031433,
      "learning_rate": 0.00015746243739565944,
      "loss": 1.1016,
      "num_input_tokens_seen": 1790836,
      "step": 643,
      "train_runtime": 1671.5851,
      "train_tokens_per_second": 1071.34
    },
    {
      "epoch": 0.051748729383876736,
      "grad_norm": 0.3670094609260559,
      "learning_rate": 0.0001573956594323873,
      "loss": 1.0972,
      "num_input_tokens_seen": 1793486,
      "step": 644,
      "train_runtime": 1673.8086,
      "train_tokens_per_second": 1071.5
    },
    {
      "epoch": 0.05182908455372748,
      "grad_norm": 0.31700095534324646,
      "learning_rate": 0.0001573288814691152,
      "loss": 1.1621,
      "num_input_tokens_seen": 1797676,
      "step": 645,
      "train_runtime": 1676.7705,
      "train_tokens_per_second": 1072.106
    },
    {
      "epoch": 0.05190943972357821,
      "grad_norm": 0.3346264064311981,
      "learning_rate": 0.00015726210350584308,
      "loss": 1.2192,
      "num_input_tokens_seen": 1799684,
      "step": 646,
      "train_runtime": 1678.5158,
      "train_tokens_per_second": 1072.188
    },
    {
      "epoch": 0.051989794893428955,
      "grad_norm": 0.3344186842441559,
      "learning_rate": 0.00015719532554257096,
      "loss": 0.9437,
      "num_input_tokens_seen": 1801374,
      "step": 647,
      "train_runtime": 1680.1845,
      "train_tokens_per_second": 1072.129
    },
    {
      "epoch": 0.052070150063279697,
      "grad_norm": 0.30313628911972046,
      "learning_rate": 0.00015712854757929886,
      "loss": 0.9575,
      "num_input_tokens_seen": 1803918,
      "step": 648,
      "train_runtime": 1682.2983,
      "train_tokens_per_second": 1072.294
    },
    {
      "epoch": 0.05215050523313044,
      "grad_norm": 0.38865846395492554,
      "learning_rate": 0.00015706176961602673,
      "loss": 0.8839,
      "num_input_tokens_seen": 1806028,
      "step": 649,
      "train_runtime": 1684.2041,
      "train_tokens_per_second": 1072.333
    },
    {
      "epoch": 0.05223086040298118,
      "grad_norm": 0.34449660778045654,
      "learning_rate": 0.0001569949916527546,
      "loss": 1.1493,
      "num_input_tokens_seen": 1807402,
      "step": 650,
      "train_runtime": 1685.7496,
      "train_tokens_per_second": 1072.165
    },
    {
      "epoch": 0.052311215572831915,
      "grad_norm": 0.2894439101219177,
      "learning_rate": 0.00015692821368948248,
      "loss": 0.8203,
      "num_input_tokens_seen": 1811010,
      "step": 651,
      "train_runtime": 1688.3898,
      "train_tokens_per_second": 1072.626
    },
    {
      "epoch": 0.05239157074268266,
      "grad_norm": 0.34291529655456543,
      "learning_rate": 0.00015686143572621035,
      "loss": 0.9161,
      "num_input_tokens_seen": 1813494,
      "step": 652,
      "train_runtime": 1690.5124,
      "train_tokens_per_second": 1072.748
    },
    {
      "epoch": 0.0524719259125334,
      "grad_norm": 0.3010275959968567,
      "learning_rate": 0.00015679465776293825,
      "loss": 1.0819,
      "num_input_tokens_seen": 1815446,
      "step": 653,
      "train_runtime": 1692.2704,
      "train_tokens_per_second": 1072.787
    },
    {
      "epoch": 0.05255228108238414,
      "grad_norm": 0.25478747487068176,
      "learning_rate": 0.00015672787979966612,
      "loss": 0.8383,
      "num_input_tokens_seen": 1819932,
      "step": 654,
      "train_runtime": 1695.4802,
      "train_tokens_per_second": 1073.402
    },
    {
      "epoch": 0.052632636252234875,
      "grad_norm": 0.3816944360733032,
      "learning_rate": 0.000156661101836394,
      "loss": 1.0901,
      "num_input_tokens_seen": 1822734,
      "step": 655,
      "train_runtime": 1697.7839,
      "train_tokens_per_second": 1073.596
    },
    {
      "epoch": 0.05271299142208562,
      "grad_norm": 0.2838401794433594,
      "learning_rate": 0.00015659432387312187,
      "loss": 1.0233,
      "num_input_tokens_seen": 1826098,
      "step": 656,
      "train_runtime": 1700.2787,
      "train_tokens_per_second": 1073.999
    },
    {
      "epoch": 0.05279334659193636,
      "grad_norm": 0.31507235765457153,
      "learning_rate": 0.00015652754590984974,
      "loss": 1.0798,
      "num_input_tokens_seen": 1827972,
      "step": 657,
      "train_runtime": 1702.0232,
      "train_tokens_per_second": 1073.999
    },
    {
      "epoch": 0.0528737017617871,
      "grad_norm": 0.2731901705265045,
      "learning_rate": 0.00015646076794657764,
      "loss": 0.6823,
      "num_input_tokens_seen": 1829970,
      "step": 658,
      "train_runtime": 1703.8753,
      "train_tokens_per_second": 1074.005
    },
    {
      "epoch": 0.05295405693163784,
      "grad_norm": 0.31667497754096985,
      "learning_rate": 0.00015639398998330552,
      "loss": 1.0495,
      "num_input_tokens_seen": 1833766,
      "step": 659,
      "train_runtime": 1706.7872,
      "train_tokens_per_second": 1074.396
    },
    {
      "epoch": 0.05303441210148858,
      "grad_norm": 0.3218362331390381,
      "learning_rate": 0.0001563272120200334,
      "loss": 0.9003,
      "num_input_tokens_seen": 1836690,
      "step": 660,
      "train_runtime": 1709.1734,
      "train_tokens_per_second": 1074.607
    },
    {
      "epoch": 0.05311476727133932,
      "grad_norm": 0.27151763439178467,
      "learning_rate": 0.00015626043405676126,
      "loss": 0.6526,
      "num_input_tokens_seen": 1838682,
      "step": 661,
      "train_runtime": 1713.0454,
      "train_tokens_per_second": 1073.341
    },
    {
      "epoch": 0.05319512244119006,
      "grad_norm": 0.35131528973579407,
      "learning_rate": 0.00015619365609348916,
      "loss": 0.94,
      "num_input_tokens_seen": 1842712,
      "step": 662,
      "train_runtime": 1716.0381,
      "train_tokens_per_second": 1073.818
    },
    {
      "epoch": 0.0532754776110408,
      "grad_norm": 0.2772160768508911,
      "learning_rate": 0.00015612687813021704,
      "loss": 1.0933,
      "num_input_tokens_seen": 1845834,
      "step": 663,
      "train_runtime": 1718.4847,
      "train_tokens_per_second": 1074.106
    },
    {
      "epoch": 0.05335583278089154,
      "grad_norm": 0.3245062530040741,
      "learning_rate": 0.00015606010016694494,
      "loss": 1.1518,
      "num_input_tokens_seen": 1850994,
      "step": 664,
      "train_runtime": 1722.107,
      "train_tokens_per_second": 1074.843
    },
    {
      "epoch": 0.05343618795074228,
      "grad_norm": 0.3193010687828064,
      "learning_rate": 0.0001559933222036728,
      "loss": 0.9863,
      "num_input_tokens_seen": 1854218,
      "step": 665,
      "train_runtime": 1724.6527,
      "train_tokens_per_second": 1075.125
    },
    {
      "epoch": 0.05351654312059302,
      "grad_norm": 0.2925742268562317,
      "learning_rate": 0.00015592654424040068,
      "loss": 0.8213,
      "num_input_tokens_seen": 1858166,
      "step": 666,
      "train_runtime": 1727.4576,
      "train_tokens_per_second": 1075.665
    },
    {
      "epoch": 0.05359689829044376,
      "grad_norm": 0.3380376398563385,
      "learning_rate": 0.00015585976627712856,
      "loss": 0.9427,
      "num_input_tokens_seen": 1860388,
      "step": 667,
      "train_runtime": 1729.3953,
      "train_tokens_per_second": 1075.745
    },
    {
      "epoch": 0.053677253460294505,
      "grad_norm": 0.32776203751564026,
      "learning_rate": 0.00015579298831385643,
      "loss": 1.367,
      "num_input_tokens_seen": 1863720,
      "step": 668,
      "train_runtime": 1731.8818,
      "train_tokens_per_second": 1076.124
    },
    {
      "epoch": 0.05375760863014524,
      "grad_norm": 0.31421107053756714,
      "learning_rate": 0.00015572621035058433,
      "loss": 1.0622,
      "num_input_tokens_seen": 1865502,
      "step": 669,
      "train_runtime": 1733.5408,
      "train_tokens_per_second": 1076.122
    },
    {
      "epoch": 0.05383796379999598,
      "grad_norm": 0.28330931067466736,
      "learning_rate": 0.0001556594323873122,
      "loss": 1.2096,
      "num_input_tokens_seen": 1868070,
      "step": 670,
      "train_runtime": 1735.6739,
      "train_tokens_per_second": 1076.279
    },
    {
      "epoch": 0.053918318969846724,
      "grad_norm": 0.2609613537788391,
      "learning_rate": 0.00015559265442404007,
      "loss": 0.6405,
      "num_input_tokens_seen": 1869382,
      "step": 671,
      "train_runtime": 1737.2445,
      "train_tokens_per_second": 1076.062
    },
    {
      "epoch": 0.053998674139697465,
      "grad_norm": 0.318636953830719,
      "learning_rate": 0.00015552587646076795,
      "loss": 0.9964,
      "num_input_tokens_seen": 1872516,
      "step": 672,
      "train_runtime": 1739.6236,
      "train_tokens_per_second": 1076.391
    },
    {
      "epoch": 0.0540790293095482,
      "grad_norm": 0.26012375950813293,
      "learning_rate": 0.00015545909849749582,
      "loss": 0.933,
      "num_input_tokens_seen": 1874650,
      "step": 673,
      "train_runtime": 1741.5153,
      "train_tokens_per_second": 1076.448
    },
    {
      "epoch": 0.05415938447939894,
      "grad_norm": 0.32276418805122375,
      "learning_rate": 0.00015539232053422372,
      "loss": 1.0842,
      "num_input_tokens_seen": 1876620,
      "step": 674,
      "train_runtime": 1743.3273,
      "train_tokens_per_second": 1076.459
    },
    {
      "epoch": 0.054239739649249684,
      "grad_norm": 0.3392653167247772,
      "learning_rate": 0.0001553255425709516,
      "loss": 1.0656,
      "num_input_tokens_seen": 1881422,
      "step": 675,
      "train_runtime": 1746.7302,
      "train_tokens_per_second": 1077.111
    },
    {
      "epoch": 0.054320094819100426,
      "grad_norm": 0.29520127177238464,
      "learning_rate": 0.00015525876460767947,
      "loss": 0.6711,
      "num_input_tokens_seen": 1884100,
      "step": 676,
      "train_runtime": 1748.9369,
      "train_tokens_per_second": 1077.283
    },
    {
      "epoch": 0.05440044998895117,
      "grad_norm": 0.3347858786582947,
      "learning_rate": 0.00015519198664440734,
      "loss": 1.1263,
      "num_input_tokens_seen": 1885718,
      "step": 677,
      "train_runtime": 1750.5285,
      "train_tokens_per_second": 1077.228
    },
    {
      "epoch": 0.0544808051588019,
      "grad_norm": 0.3290030360221863,
      "learning_rate": 0.00015512520868113521,
      "loss": 0.9578,
      "num_input_tokens_seen": 1888086,
      "step": 678,
      "train_runtime": 1752.5106,
      "train_tokens_per_second": 1077.361
    },
    {
      "epoch": 0.054561160328652644,
      "grad_norm": 0.3620326817035675,
      "learning_rate": 0.00015505843071786311,
      "loss": 0.9817,
      "num_input_tokens_seen": 1890442,
      "step": 679,
      "train_runtime": 1754.5567,
      "train_tokens_per_second": 1077.447
    },
    {
      "epoch": 0.054641515498503386,
      "grad_norm": 0.36778369545936584,
      "learning_rate": 0.000154991652754591,
      "loss": 0.838,
      "num_input_tokens_seen": 1892284,
      "step": 680,
      "train_runtime": 1756.3317,
      "train_tokens_per_second": 1077.407
    },
    {
      "epoch": 0.05472187066835413,
      "grad_norm": 0.31623950600624084,
      "learning_rate": 0.0001549248747913189,
      "loss": 0.9384,
      "num_input_tokens_seen": 1894934,
      "step": 681,
      "train_runtime": 1758.5071,
      "train_tokens_per_second": 1077.581
    },
    {
      "epoch": 0.05480222583820486,
      "grad_norm": 0.2577507495880127,
      "learning_rate": 0.00015485809682804676,
      "loss": 0.8484,
      "num_input_tokens_seen": 1898802,
      "step": 682,
      "train_runtime": 1761.357,
      "train_tokens_per_second": 1078.034
    },
    {
      "epoch": 0.054882581008055605,
      "grad_norm": 0.2751832902431488,
      "learning_rate": 0.00015479131886477463,
      "loss": 0.7455,
      "num_input_tokens_seen": 1900668,
      "step": 683,
      "train_runtime": 1762.9767,
      "train_tokens_per_second": 1078.102
    },
    {
      "epoch": 0.054962936177906346,
      "grad_norm": 0.3682342767715454,
      "learning_rate": 0.0001547245409015025,
      "loss": 0.9043,
      "num_input_tokens_seen": 1904252,
      "step": 684,
      "train_runtime": 1765.5917,
      "train_tokens_per_second": 1078.535
    },
    {
      "epoch": 0.05504329134775709,
      "grad_norm": 0.29235321283340454,
      "learning_rate": 0.0001546577629382304,
      "loss": 1.0878,
      "num_input_tokens_seen": 1907134,
      "step": 685,
      "train_runtime": 1767.8719,
      "train_tokens_per_second": 1078.774
    },
    {
      "epoch": 0.05512364651760783,
      "grad_norm": 0.27555689215660095,
      "learning_rate": 0.00015459098497495828,
      "loss": 0.803,
      "num_input_tokens_seen": 1910678,
      "step": 686,
      "train_runtime": 1770.5152,
      "train_tokens_per_second": 1079.165
    },
    {
      "epoch": 0.055204001687458565,
      "grad_norm": 0.32898277044296265,
      "learning_rate": 0.00015452420701168615,
      "loss": 0.8662,
      "num_input_tokens_seen": 1912310,
      "step": 687,
      "train_runtime": 1772.1262,
      "train_tokens_per_second": 1079.105
    },
    {
      "epoch": 0.05528435685730931,
      "grad_norm": 0.2872154414653778,
      "learning_rate": 0.00015445742904841403,
      "loss": 1.1046,
      "num_input_tokens_seen": 1915508,
      "step": 688,
      "train_runtime": 1774.4972,
      "train_tokens_per_second": 1079.465
    },
    {
      "epoch": 0.05536471202716005,
      "grad_norm": 0.29981961846351624,
      "learning_rate": 0.0001543906510851419,
      "loss": 1.1274,
      "num_input_tokens_seen": 1918116,
      "step": 689,
      "train_runtime": 1776.6914,
      "train_tokens_per_second": 1079.6
    },
    {
      "epoch": 0.05544506719701079,
      "grad_norm": 0.33591628074645996,
      "learning_rate": 0.0001543238731218698,
      "loss": 0.9346,
      "num_input_tokens_seen": 1919452,
      "step": 690,
      "train_runtime": 1778.2472,
      "train_tokens_per_second": 1079.407
    },
    {
      "epoch": 0.055525422366861525,
      "grad_norm": 0.284965455532074,
      "learning_rate": 0.00015425709515859767,
      "loss": 1.0246,
      "num_input_tokens_seen": 1922190,
      "step": 691,
      "train_runtime": 1782.4841,
      "train_tokens_per_second": 1078.377
    },
    {
      "epoch": 0.05560577753671227,
      "grad_norm": 0.27644267678260803,
      "learning_rate": 0.00015419031719532555,
      "loss": 0.9134,
      "num_input_tokens_seen": 1925062,
      "step": 692,
      "train_runtime": 1784.6903,
      "train_tokens_per_second": 1078.653
    },
    {
      "epoch": 0.05568613270656301,
      "grad_norm": 0.33998218178749084,
      "learning_rate": 0.00015412353923205342,
      "loss": 0.8353,
      "num_input_tokens_seen": 1927318,
      "step": 693,
      "train_runtime": 1786.609,
      "train_tokens_per_second": 1078.758
    },
    {
      "epoch": 0.05576648787641375,
      "grad_norm": 0.2729622423648834,
      "learning_rate": 0.0001540567612687813,
      "loss": 0.7494,
      "num_input_tokens_seen": 1931602,
      "step": 694,
      "train_runtime": 1789.702,
      "train_tokens_per_second": 1079.287
    },
    {
      "epoch": 0.05584684304626449,
      "grad_norm": 0.28827887773513794,
      "learning_rate": 0.0001539899833055092,
      "loss": 0.9188,
      "num_input_tokens_seen": 1933984,
      "step": 695,
      "train_runtime": 1791.7153,
      "train_tokens_per_second": 1079.404
    },
    {
      "epoch": 0.05592719821611523,
      "grad_norm": 0.4026768207550049,
      "learning_rate": 0.00015392320534223707,
      "loss": 0.9791,
      "num_input_tokens_seen": 1936502,
      "step": 696,
      "train_runtime": 1793.9386,
      "train_tokens_per_second": 1079.469
    },
    {
      "epoch": 0.05600755338596597,
      "grad_norm": 0.2798472046852112,
      "learning_rate": 0.00015385642737896494,
      "loss": 0.62,
      "num_input_tokens_seen": 1938680,
      "step": 697,
      "train_runtime": 1795.8759,
      "train_tokens_per_second": 1079.518
    },
    {
      "epoch": 0.05608790855581671,
      "grad_norm": 0.2956175208091736,
      "learning_rate": 0.00015378964941569284,
      "loss": 0.6289,
      "num_input_tokens_seen": 1940370,
      "step": 698,
      "train_runtime": 1797.5764,
      "train_tokens_per_second": 1079.437
    },
    {
      "epoch": 0.05616826372566745,
      "grad_norm": 0.30674660205841064,
      "learning_rate": 0.0001537228714524207,
      "loss": 1.1346,
      "num_input_tokens_seen": 1944056,
      "step": 699,
      "train_runtime": 1800.1653,
      "train_tokens_per_second": 1079.932
    },
    {
      "epoch": 0.05624861889551819,
      "grad_norm": 0.3235774040222168,
      "learning_rate": 0.00015365609348914859,
      "loss": 1.1422,
      "num_input_tokens_seen": 1945484,
      "step": 700,
      "train_runtime": 1801.7645,
      "train_tokens_per_second": 1079.766
    },
    {
      "epoch": 0.05632897406536893,
      "grad_norm": 0.3246361017227173,
      "learning_rate": 0.00015358931552587649,
      "loss": 1.037,
      "num_input_tokens_seen": 1950982,
      "step": 701,
      "train_runtime": 1805.6746,
      "train_tokens_per_second": 1080.473
    },
    {
      "epoch": 0.05640932923521967,
      "grad_norm": 0.28568196296691895,
      "learning_rate": 0.00015352253756260436,
      "loss": 0.928,
      "num_input_tokens_seen": 1954032,
      "step": 702,
      "train_runtime": 1807.9373,
      "train_tokens_per_second": 1080.807
    },
    {
      "epoch": 0.05648968440507041,
      "grad_norm": 0.34797027707099915,
      "learning_rate": 0.00015345575959933223,
      "loss": 0.9213,
      "num_input_tokens_seen": 1955984,
      "step": 703,
      "train_runtime": 1809.842,
      "train_tokens_per_second": 1080.748
    },
    {
      "epoch": 0.056570039574921155,
      "grad_norm": 0.30620652437210083,
      "learning_rate": 0.0001533889816360601,
      "loss": 0.8303,
      "num_input_tokens_seen": 1959298,
      "step": 704,
      "train_runtime": 1812.4911,
      "train_tokens_per_second": 1080.997
    },
    {
      "epoch": 0.05665039474477189,
      "grad_norm": 0.26757800579071045,
      "learning_rate": 0.00015332220367278798,
      "loss": 1.0902,
      "num_input_tokens_seen": 1962378,
      "step": 705,
      "train_runtime": 1814.8604,
      "train_tokens_per_second": 1081.283
    },
    {
      "epoch": 0.05673074991462263,
      "grad_norm": 0.3470252454280853,
      "learning_rate": 0.00015325542570951588,
      "loss": 0.7748,
      "num_input_tokens_seen": 1965148,
      "step": 706,
      "train_runtime": 1817.1073,
      "train_tokens_per_second": 1081.471
    },
    {
      "epoch": 0.05681110508447337,
      "grad_norm": 0.257053405046463,
      "learning_rate": 0.00015318864774624375,
      "loss": 0.7623,
      "num_input_tokens_seen": 1967994,
      "step": 707,
      "train_runtime": 1819.4047,
      "train_tokens_per_second": 1081.669
    },
    {
      "epoch": 0.056891460254324115,
      "grad_norm": 0.3352760672569275,
      "learning_rate": 0.00015312186978297163,
      "loss": 1.1276,
      "num_input_tokens_seen": 1971064,
      "step": 708,
      "train_runtime": 1821.7562,
      "train_tokens_per_second": 1081.958
    },
    {
      "epoch": 0.05697181542417485,
      "grad_norm": 0.3215519189834595,
      "learning_rate": 0.0001530550918196995,
      "loss": 1.0408,
      "num_input_tokens_seen": 1973634,
      "step": 709,
      "train_runtime": 1823.853,
      "train_tokens_per_second": 1082.123
    },
    {
      "epoch": 0.05705217059402559,
      "grad_norm": 0.29732009768486023,
      "learning_rate": 0.00015298831385642737,
      "loss": 1.2322,
      "num_input_tokens_seen": 1975384,
      "step": 710,
      "train_runtime": 1825.5454,
      "train_tokens_per_second": 1082.079
    },
    {
      "epoch": 0.057132525763876334,
      "grad_norm": 0.35030630230903625,
      "learning_rate": 0.00015292153589315527,
      "loss": 0.8895,
      "num_input_tokens_seen": 1978254,
      "step": 711,
      "train_runtime": 1827.8538,
      "train_tokens_per_second": 1082.282
    },
    {
      "epoch": 0.057212880933727076,
      "grad_norm": 0.32095322012901306,
      "learning_rate": 0.00015285475792988315,
      "loss": 1.0253,
      "num_input_tokens_seen": 1980222,
      "step": 712,
      "train_runtime": 1829.64,
      "train_tokens_per_second": 1082.301
    },
    {
      "epoch": 0.05729323610357782,
      "grad_norm": 0.2621558606624603,
      "learning_rate": 0.00015278797996661102,
      "loss": 1.0712,
      "num_input_tokens_seen": 1983714,
      "step": 713,
      "train_runtime": 1832.1453,
      "train_tokens_per_second": 1082.727
    },
    {
      "epoch": 0.05737359127342855,
      "grad_norm": 0.30798768997192383,
      "learning_rate": 0.0001527212020033389,
      "loss": 1.0327,
      "num_input_tokens_seen": 1985694,
      "step": 714,
      "train_runtime": 1833.9312,
      "train_tokens_per_second": 1082.753
    },
    {
      "epoch": 0.057453946443279294,
      "grad_norm": 0.32821840047836304,
      "learning_rate": 0.0001526544240400668,
      "loss": 0.9513,
      "num_input_tokens_seen": 1988750,
      "step": 715,
      "train_runtime": 1836.2632,
      "train_tokens_per_second": 1083.042
    },
    {
      "epoch": 0.057534301613130036,
      "grad_norm": 0.35471609234809875,
      "learning_rate": 0.00015258764607679466,
      "loss": 0.9607,
      "num_input_tokens_seen": 1991266,
      "step": 716,
      "train_runtime": 1838.2892,
      "train_tokens_per_second": 1083.217
    },
    {
      "epoch": 0.05761465678298078,
      "grad_norm": 0.29893821477890015,
      "learning_rate": 0.00015252086811352257,
      "loss": 0.7843,
      "num_input_tokens_seen": 1993566,
      "step": 717,
      "train_runtime": 1840.2644,
      "train_tokens_per_second": 1083.304
    },
    {
      "epoch": 0.05769501195283151,
      "grad_norm": 0.33154726028442383,
      "learning_rate": 0.00015245409015025044,
      "loss": 1.2394,
      "num_input_tokens_seen": 1995860,
      "step": 718,
      "train_runtime": 1842.17,
      "train_tokens_per_second": 1083.429
    },
    {
      "epoch": 0.057775367122682254,
      "grad_norm": 0.3338896632194519,
      "learning_rate": 0.0001523873121869783,
      "loss": 0.7933,
      "num_input_tokens_seen": 1998962,
      "step": 719,
      "train_runtime": 1844.537,
      "train_tokens_per_second": 1083.72
    },
    {
      "epoch": 0.057855722292532996,
      "grad_norm": 0.3402370512485504,
      "learning_rate": 0.00015232053422370618,
      "loss": 1.0473,
      "num_input_tokens_seen": 2001728,
      "step": 720,
      "train_runtime": 1846.8124,
      "train_tokens_per_second": 1083.883
    },
    {
      "epoch": 0.05793607746238374,
      "grad_norm": 0.3279224932193756,
      "learning_rate": 0.00015225375626043406,
      "loss": 0.8423,
      "num_input_tokens_seen": 2004578,
      "step": 721,
      "train_runtime": 1851.0791,
      "train_tokens_per_second": 1082.924
    },
    {
      "epoch": 0.05801643263223448,
      "grad_norm": 0.27328982949256897,
      "learning_rate": 0.00015218697829716196,
      "loss": 0.9153,
      "num_input_tokens_seen": 2007970,
      "step": 722,
      "train_runtime": 1853.6423,
      "train_tokens_per_second": 1083.256
    },
    {
      "epoch": 0.058096787802085215,
      "grad_norm": 0.3246921896934509,
      "learning_rate": 0.00015212020033388983,
      "loss": 1.1111,
      "num_input_tokens_seen": 2009966,
      "step": 723,
      "train_runtime": 1855.5032,
      "train_tokens_per_second": 1083.246
    },
    {
      "epoch": 0.058177142971935956,
      "grad_norm": 0.29969558119773865,
      "learning_rate": 0.0001520534223706177,
      "loss": 1.1185,
      "num_input_tokens_seen": 2011548,
      "step": 724,
      "train_runtime": 1857.1507,
      "train_tokens_per_second": 1083.137
    },
    {
      "epoch": 0.0582574981417867,
      "grad_norm": 0.25052857398986816,
      "learning_rate": 0.00015198664440734558,
      "loss": 0.8242,
      "num_input_tokens_seen": 2014494,
      "step": 725,
      "train_runtime": 1859.4042,
      "train_tokens_per_second": 1083.408
    },
    {
      "epoch": 0.05833785331163744,
      "grad_norm": 0.2533136010169983,
      "learning_rate": 0.00015191986644407345,
      "loss": 1.0544,
      "num_input_tokens_seen": 2018804,
      "step": 726,
      "train_runtime": 1862.4596,
      "train_tokens_per_second": 1083.945
    },
    {
      "epoch": 0.058418208481488175,
      "grad_norm": 0.34300217032432556,
      "learning_rate": 0.00015185308848080135,
      "loss": 1.1558,
      "num_input_tokens_seen": 2021872,
      "step": 727,
      "train_runtime": 1864.9797,
      "train_tokens_per_second": 1084.125
    },
    {
      "epoch": 0.05849856365133892,
      "grad_norm": 0.2861180007457733,
      "learning_rate": 0.00015178631051752922,
      "loss": 0.8775,
      "num_input_tokens_seen": 2024396,
      "step": 728,
      "train_runtime": 1867.0564,
      "train_tokens_per_second": 1084.271
    },
    {
      "epoch": 0.05857891882118966,
      "grad_norm": 0.2502950131893158,
      "learning_rate": 0.0001517195325542571,
      "loss": 0.7155,
      "num_input_tokens_seen": 2028560,
      "step": 729,
      "train_runtime": 1870.0691,
      "train_tokens_per_second": 1084.751
    },
    {
      "epoch": 0.0586592739910404,
      "grad_norm": 0.3140285909175873,
      "learning_rate": 0.00015165275459098497,
      "loss": 0.9702,
      "num_input_tokens_seen": 2031816,
      "step": 730,
      "train_runtime": 1872.5611,
      "train_tokens_per_second": 1085.047
    },
    {
      "epoch": 0.05873962916089114,
      "grad_norm": 0.3011912405490875,
      "learning_rate": 0.00015158597662771284,
      "loss": 1.1461,
      "num_input_tokens_seen": 2033714,
      "step": 731,
      "train_runtime": 1874.2976,
      "train_tokens_per_second": 1085.054
    },
    {
      "epoch": 0.05881998433074188,
      "grad_norm": 0.3443948030471802,
      "learning_rate": 0.00015151919866444074,
      "loss": 1.106,
      "num_input_tokens_seen": 2039640,
      "step": 732,
      "train_runtime": 1878.3622,
      "train_tokens_per_second": 1085.861
    },
    {
      "epoch": 0.05890033950059262,
      "grad_norm": 0.30849480628967285,
      "learning_rate": 0.00015145242070116862,
      "loss": 1.1929,
      "num_input_tokens_seen": 2043664,
      "step": 733,
      "train_runtime": 1881.3566,
      "train_tokens_per_second": 1086.271
    },
    {
      "epoch": 0.05898069467044336,
      "grad_norm": 0.3384644687175751,
      "learning_rate": 0.00015138564273789652,
      "loss": 1.3346,
      "num_input_tokens_seen": 2046676,
      "step": 734,
      "train_runtime": 1883.7806,
      "train_tokens_per_second": 1086.473
    },
    {
      "epoch": 0.0590610498402941,
      "grad_norm": 0.41189488768577576,
      "learning_rate": 0.0001513188647746244,
      "loss": 1.1487,
      "num_input_tokens_seen": 2048630,
      "step": 735,
      "train_runtime": 1885.5806,
      "train_tokens_per_second": 1086.472
    },
    {
      "epoch": 0.05914140501014484,
      "grad_norm": 0.35140398144721985,
      "learning_rate": 0.00015125208681135226,
      "loss": 0.7905,
      "num_input_tokens_seen": 2050468,
      "step": 736,
      "train_runtime": 1887.3125,
      "train_tokens_per_second": 1086.449
    },
    {
      "epoch": 0.05922176017999558,
      "grad_norm": 0.29333919286727905,
      "learning_rate": 0.00015118530884808014,
      "loss": 0.7653,
      "num_input_tokens_seen": 2053514,
      "step": 737,
      "train_runtime": 1889.6361,
      "train_tokens_per_second": 1086.725
    },
    {
      "epoch": 0.05930211534984632,
      "grad_norm": 0.27940815687179565,
      "learning_rate": 0.00015111853088480804,
      "loss": 1.1097,
      "num_input_tokens_seen": 2056204,
      "step": 738,
      "train_runtime": 1891.8235,
      "train_tokens_per_second": 1086.89
    },
    {
      "epoch": 0.05938247051969706,
      "grad_norm": 0.27439242601394653,
      "learning_rate": 0.0001510517529215359,
      "loss": 0.7571,
      "num_input_tokens_seen": 2058580,
      "step": 739,
      "train_runtime": 1893.8655,
      "train_tokens_per_second": 1086.973
    },
    {
      "epoch": 0.059462825689547805,
      "grad_norm": 0.5973687171936035,
      "learning_rate": 0.00015098497495826378,
      "loss": 0.8022,
      "num_input_tokens_seen": 2061114,
      "step": 740,
      "train_runtime": 1896.0041,
      "train_tokens_per_second": 1087.083
    },
    {
      "epoch": 0.05954318085939854,
      "grad_norm": 0.3319377303123474,
      "learning_rate": 0.00015091819699499166,
      "loss": 0.9576,
      "num_input_tokens_seen": 2063538,
      "step": 741,
      "train_runtime": 1898.041,
      "train_tokens_per_second": 1087.194
    },
    {
      "epoch": 0.05962353602924928,
      "grad_norm": 0.27499696612358093,
      "learning_rate": 0.00015085141903171953,
      "loss": 0.8339,
      "num_input_tokens_seen": 2067446,
      "step": 742,
      "train_runtime": 1900.8523,
      "train_tokens_per_second": 1087.642
    },
    {
      "epoch": 0.05970389119910002,
      "grad_norm": 0.3294494152069092,
      "learning_rate": 0.00015078464106844743,
      "loss": 0.8107,
      "num_input_tokens_seen": 2069534,
      "step": 743,
      "train_runtime": 1902.6727,
      "train_tokens_per_second": 1087.698
    },
    {
      "epoch": 0.059784246368950765,
      "grad_norm": 0.3478166162967682,
      "learning_rate": 0.0001507178631051753,
      "loss": 0.803,
      "num_input_tokens_seen": 2071896,
      "step": 744,
      "train_runtime": 1904.7234,
      "train_tokens_per_second": 1087.767
    },
    {
      "epoch": 0.0598646015388015,
      "grad_norm": 0.3554915189743042,
      "learning_rate": 0.00015065108514190318,
      "loss": 1.0449,
      "num_input_tokens_seen": 2073966,
      "step": 745,
      "train_runtime": 1906.6912,
      "train_tokens_per_second": 1087.73
    },
    {
      "epoch": 0.05994495670865224,
      "grad_norm": 0.30656698346138,
      "learning_rate": 0.00015058430717863105,
      "loss": 0.7917,
      "num_input_tokens_seen": 2076060,
      "step": 746,
      "train_runtime": 1908.6257,
      "train_tokens_per_second": 1087.725
    },
    {
      "epoch": 0.060025311878502983,
      "grad_norm": 0.2801983952522278,
      "learning_rate": 0.00015051752921535892,
      "loss": 0.7432,
      "num_input_tokens_seen": 2078862,
      "step": 747,
      "train_runtime": 1910.8057,
      "train_tokens_per_second": 1087.951
    },
    {
      "epoch": 0.060105667048353725,
      "grad_norm": 0.47750338912010193,
      "learning_rate": 0.0001504507512520868,
      "loss": 1.0171,
      "num_input_tokens_seen": 2080784,
      "step": 748,
      "train_runtime": 1912.4467,
      "train_tokens_per_second": 1088.022
    },
    {
      "epoch": 0.06018602221820447,
      "grad_norm": 0.38815444707870483,
      "learning_rate": 0.0001503839732888147,
      "loss": 1.089,
      "num_input_tokens_seen": 2083682,
      "step": 749,
      "train_runtime": 1914.7004,
      "train_tokens_per_second": 1088.255
    },
    {
      "epoch": 0.0602663773880552,
      "grad_norm": 0.30164188146591187,
      "learning_rate": 0.00015031719532554257,
      "loss": 1.0412,
      "num_input_tokens_seen": 2086678,
      "step": 750,
      "train_runtime": 1917.0235,
      "train_tokens_per_second": 1088.499
    },
    {
      "epoch": 0.060346732557905944,
      "grad_norm": 0.32808375358581543,
      "learning_rate": 0.00015025041736227047,
      "loss": 1.1034,
      "num_input_tokens_seen": 2088932,
      "step": 751,
      "train_runtime": 1920.9797,
      "train_tokens_per_second": 1087.431
    },
    {
      "epoch": 0.060427087727756686,
      "grad_norm": 0.28214800357818604,
      "learning_rate": 0.00015018363939899834,
      "loss": 0.8384,
      "num_input_tokens_seen": 2091816,
      "step": 752,
      "train_runtime": 1923.1554,
      "train_tokens_per_second": 1087.7
    },
    {
      "epoch": 0.06050744289760743,
      "grad_norm": 0.33014342188835144,
      "learning_rate": 0.00015011686143572622,
      "loss": 1.1402,
      "num_input_tokens_seen": 2095270,
      "step": 753,
      "train_runtime": 1925.7468,
      "train_tokens_per_second": 1088.03
    },
    {
      "epoch": 0.06058779806745816,
      "grad_norm": 0.3177398443222046,
      "learning_rate": 0.00015005008347245412,
      "loss": 1.0321,
      "num_input_tokens_seen": 2098510,
      "step": 754,
      "train_runtime": 1928.2026,
      "train_tokens_per_second": 1088.324
    },
    {
      "epoch": 0.060668153237308904,
      "grad_norm": 0.3043687343597412,
      "learning_rate": 0.000149983305509182,
      "loss": 1.0418,
      "num_input_tokens_seen": 2100818,
      "step": 755,
      "train_runtime": 1930.0889,
      "train_tokens_per_second": 1088.457
    },
    {
      "epoch": 0.060748508407159646,
      "grad_norm": 0.3117942810058594,
      "learning_rate": 0.00014991652754590986,
      "loss": 1.0242,
      "num_input_tokens_seen": 2103630,
      "step": 756,
      "train_runtime": 1932.282,
      "train_tokens_per_second": 1088.677
    },
    {
      "epoch": 0.06082886357701039,
      "grad_norm": 0.2615673542022705,
      "learning_rate": 0.00014984974958263774,
      "loss": 0.8929,
      "num_input_tokens_seen": 2105904,
      "step": 757,
      "train_runtime": 1934.1837,
      "train_tokens_per_second": 1088.782
    },
    {
      "epoch": 0.06090921874686113,
      "grad_norm": 0.3243381977081299,
      "learning_rate": 0.0001497829716193656,
      "loss": 1.1386,
      "num_input_tokens_seen": 2108018,
      "step": 758,
      "train_runtime": 1936.0094,
      "train_tokens_per_second": 1088.847
    },
    {
      "epoch": 0.060989573916711864,
      "grad_norm": 0.330251544713974,
      "learning_rate": 0.0001497161936560935,
      "loss": 0.9637,
      "num_input_tokens_seen": 2111362,
      "step": 759,
      "train_runtime": 1938.6033,
      "train_tokens_per_second": 1089.115
    },
    {
      "epoch": 0.061069929086562606,
      "grad_norm": 0.34466591477394104,
      "learning_rate": 0.00014964941569282138,
      "loss": 1.0165,
      "num_input_tokens_seen": 2114330,
      "step": 760,
      "train_runtime": 1941.0526,
      "train_tokens_per_second": 1089.27
    },
    {
      "epoch": 0.06115028425641335,
      "grad_norm": 0.29236334562301636,
      "learning_rate": 0.00014958263772954926,
      "loss": 0.9817,
      "num_input_tokens_seen": 2117744,
      "step": 761,
      "train_runtime": 1943.6514,
      "train_tokens_per_second": 1089.57
    },
    {
      "epoch": 0.06123063942626409,
      "grad_norm": 0.3150699734687805,
      "learning_rate": 0.00014951585976627713,
      "loss": 1.1979,
      "num_input_tokens_seen": 2120630,
      "step": 762,
      "train_runtime": 1945.8991,
      "train_tokens_per_second": 1089.794
    },
    {
      "epoch": 0.061310994596114825,
      "grad_norm": 0.3132155239582062,
      "learning_rate": 0.000149449081803005,
      "loss": 0.8408,
      "num_input_tokens_seen": 2125434,
      "step": 763,
      "train_runtime": 1949.221,
      "train_tokens_per_second": 1090.402
    },
    {
      "epoch": 0.06139134976596557,
      "grad_norm": 0.3346937894821167,
      "learning_rate": 0.00014938230383973287,
      "loss": 1.2044,
      "num_input_tokens_seen": 2128796,
      "step": 764,
      "train_runtime": 1951.791,
      "train_tokens_per_second": 1090.688
    },
    {
      "epoch": 0.06147170493581631,
      "grad_norm": 0.29328790307044983,
      "learning_rate": 0.00014931552587646077,
      "loss": 1.1102,
      "num_input_tokens_seen": 2132024,
      "step": 765,
      "train_runtime": 1954.229,
      "train_tokens_per_second": 1090.98
    },
    {
      "epoch": 0.06155206010566705,
      "grad_norm": 0.2834538221359253,
      "learning_rate": 0.00014924874791318865,
      "loss": 0.8466,
      "num_input_tokens_seen": 2133818,
      "step": 766,
      "train_runtime": 1955.9092,
      "train_tokens_per_second": 1090.96
    },
    {
      "epoch": 0.06163241527551779,
      "grad_norm": 0.3271242082118988,
      "learning_rate": 0.00014918196994991652,
      "loss": 1.1796,
      "num_input_tokens_seen": 2136520,
      "step": 767,
      "train_runtime": 1958.137,
      "train_tokens_per_second": 1091.098
    },
    {
      "epoch": 0.06171277044536853,
      "grad_norm": 0.3325490355491638,
      "learning_rate": 0.00014911519198664442,
      "loss": 1.1547,
      "num_input_tokens_seen": 2138466,
      "step": 768,
      "train_runtime": 1960.0398,
      "train_tokens_per_second": 1091.032
    },
    {
      "epoch": 0.06179312561521927,
      "grad_norm": 0.34948816895484924,
      "learning_rate": 0.0001490484140233723,
      "loss": 1.0936,
      "num_input_tokens_seen": 2139834,
      "step": 769,
      "train_runtime": 1961.5708,
      "train_tokens_per_second": 1090.878
    },
    {
      "epoch": 0.06187348078507001,
      "grad_norm": 0.26612240076065063,
      "learning_rate": 0.0001489816360601002,
      "loss": 1.1135,
      "num_input_tokens_seen": 2142074,
      "step": 770,
      "train_runtime": 1963.5128,
      "train_tokens_per_second": 1090.94
    },
    {
      "epoch": 0.06195383595492075,
      "grad_norm": 0.26699098944664,
      "learning_rate": 0.00014891485809682807,
      "loss": 1.0005,
      "num_input_tokens_seen": 2144910,
      "step": 771,
      "train_runtime": 1965.7524,
      "train_tokens_per_second": 1091.139
    },
    {
      "epoch": 0.06203419112477149,
      "grad_norm": 0.26087722182273865,
      "learning_rate": 0.00014884808013355594,
      "loss": 1.0274,
      "num_input_tokens_seen": 2148416,
      "step": 772,
      "train_runtime": 1968.3805,
      "train_tokens_per_second": 1091.464
    },
    {
      "epoch": 0.06211454629462223,
      "grad_norm": 0.305115282535553,
      "learning_rate": 0.00014878130217028381,
      "loss": 0.804,
      "num_input_tokens_seen": 2150900,
      "step": 773,
      "train_runtime": 1970.4444,
      "train_tokens_per_second": 1091.581
    },
    {
      "epoch": 0.06219490146447297,
      "grad_norm": 0.29959383606910706,
      "learning_rate": 0.0001487145242070117,
      "loss": 0.6794,
      "num_input_tokens_seen": 2157162,
      "step": 774,
      "train_runtime": 1974.6936,
      "train_tokens_per_second": 1092.403
    },
    {
      "epoch": 0.06227525663432371,
      "grad_norm": 0.27511703968048096,
      "learning_rate": 0.0001486477462437396,
      "loss": 0.64,
      "num_input_tokens_seen": 2159236,
      "step": 775,
      "train_runtime": 1976.6404,
      "train_tokens_per_second": 1092.377
    },
    {
      "epoch": 0.062355611804174454,
      "grad_norm": 0.3052910268306732,
      "learning_rate": 0.00014858096828046746,
      "loss": 1.0433,
      "num_input_tokens_seen": 2160942,
      "step": 776,
      "train_runtime": 1978.3747,
      "train_tokens_per_second": 1092.281
    },
    {
      "epoch": 0.06243596697402519,
      "grad_norm": 0.27722468972206116,
      "learning_rate": 0.00014851419031719533,
      "loss": 0.7702,
      "num_input_tokens_seen": 2163482,
      "step": 777,
      "train_runtime": 1980.4575,
      "train_tokens_per_second": 1092.415
    },
    {
      "epoch": 0.06251632214387594,
      "grad_norm": 0.35401254892349243,
      "learning_rate": 0.0001484474123539232,
      "loss": 0.9084,
      "num_input_tokens_seen": 2166868,
      "step": 778,
      "train_runtime": 1983.0423,
      "train_tokens_per_second": 1092.699
    },
    {
      "epoch": 0.06259667731372667,
      "grad_norm": 0.3272496461868286,
      "learning_rate": 0.00014838063439065108,
      "loss": 1.0999,
      "num_input_tokens_seen": 2169036,
      "step": 779,
      "train_runtime": 1984.9872,
      "train_tokens_per_second": 1092.72
    },
    {
      "epoch": 0.06267703248357741,
      "grad_norm": 0.3062959909439087,
      "learning_rate": 0.00014831385642737895,
      "loss": 1.2502,
      "num_input_tokens_seen": 2173368,
      "step": 780,
      "train_runtime": 1988.1783,
      "train_tokens_per_second": 1093.145
    },
    {
      "epoch": 0.06275738765342816,
      "grad_norm": 0.30199354887008667,
      "learning_rate": 0.00014824707846410685,
      "loss": 0.9818,
      "num_input_tokens_seen": 2177228,
      "step": 781,
      "train_runtime": 1992.9748,
      "train_tokens_per_second": 1092.451
    },
    {
      "epoch": 0.06283774282327889,
      "grad_norm": 0.3133285343647003,
      "learning_rate": 0.00014818030050083473,
      "loss": 0.9363,
      "num_input_tokens_seen": 2179588,
      "step": 782,
      "train_runtime": 1995.0025,
      "train_tokens_per_second": 1092.524
    },
    {
      "epoch": 0.06291809799312963,
      "grad_norm": 0.2926086187362671,
      "learning_rate": 0.0001481135225375626,
      "loss": 0.7605,
      "num_input_tokens_seen": 2183504,
      "step": 783,
      "train_runtime": 1997.794,
      "train_tokens_per_second": 1092.958
    },
    {
      "epoch": 0.06299845316298038,
      "grad_norm": 0.3522408902645111,
      "learning_rate": 0.00014804674457429047,
      "loss": 0.9815,
      "num_input_tokens_seen": 2185204,
      "step": 784,
      "train_runtime": 1999.4975,
      "train_tokens_per_second": 1092.877
    },
    {
      "epoch": 0.06307880833283111,
      "grad_norm": 0.2630411386489868,
      "learning_rate": 0.00014797996661101837,
      "loss": 0.6635,
      "num_input_tokens_seen": 2187770,
      "step": 785,
      "train_runtime": 2001.5965,
      "train_tokens_per_second": 1093.012
    },
    {
      "epoch": 0.06315916350268186,
      "grad_norm": 0.3483324944972992,
      "learning_rate": 0.00014791318864774625,
      "loss": 0.6638,
      "num_input_tokens_seen": 2189540,
      "step": 786,
      "train_runtime": 2003.2525,
      "train_tokens_per_second": 1092.993
    },
    {
      "epoch": 0.0632395186725326,
      "grad_norm": 0.3385311961174011,
      "learning_rate": 0.00014784641068447415,
      "loss": 1.081,
      "num_input_tokens_seen": 2192502,
      "step": 787,
      "train_runtime": 2005.533,
      "train_tokens_per_second": 1093.227
    },
    {
      "epoch": 0.06331987384238333,
      "grad_norm": 0.29116401076316833,
      "learning_rate": 0.00014777963272120202,
      "loss": 0.8055,
      "num_input_tokens_seen": 2195678,
      "step": 788,
      "train_runtime": 2007.9418,
      "train_tokens_per_second": 1093.497
    },
    {
      "epoch": 0.06340022901223408,
      "grad_norm": 0.3660394549369812,
      "learning_rate": 0.0001477128547579299,
      "loss": 1.0744,
      "num_input_tokens_seen": 2197978,
      "step": 789,
      "train_runtime": 2009.9164,
      "train_tokens_per_second": 1093.567
    },
    {
      "epoch": 0.06348058418208481,
      "grad_norm": 0.3061630427837372,
      "learning_rate": 0.00014764607679465777,
      "loss": 1.111,
      "num_input_tokens_seen": 2201604,
      "step": 790,
      "train_runtime": 2012.5559,
      "train_tokens_per_second": 1093.934
    },
    {
      "epoch": 0.06356093935193556,
      "grad_norm": 0.43108677864074707,
      "learning_rate": 0.00014757929883138567,
      "loss": 1.1627,
      "num_input_tokens_seen": 2203232,
      "step": 791,
      "train_runtime": 2014.2338,
      "train_tokens_per_second": 1093.831
    },
    {
      "epoch": 0.0636412945217863,
      "grad_norm": 0.28936949372291565,
      "learning_rate": 0.00014751252086811354,
      "loss": 1.0865,
      "num_input_tokens_seen": 2206308,
      "step": 792,
      "train_runtime": 2016.6608,
      "train_tokens_per_second": 1094.04
    },
    {
      "epoch": 0.06372164969163703,
      "grad_norm": 0.3124532997608185,
      "learning_rate": 0.0001474457429048414,
      "loss": 0.9604,
      "num_input_tokens_seen": 2208956,
      "step": 793,
      "train_runtime": 2018.8626,
      "train_tokens_per_second": 1094.159
    },
    {
      "epoch": 0.06380200486148778,
      "grad_norm": 0.3006727695465088,
      "learning_rate": 0.00014737896494156929,
      "loss": 1.2855,
      "num_input_tokens_seen": 2212122,
      "step": 794,
      "train_runtime": 2021.3389,
      "train_tokens_per_second": 1094.385
    },
    {
      "epoch": 0.06388236003133851,
      "grad_norm": 0.31648802757263184,
      "learning_rate": 0.00014731218697829716,
      "loss": 0.8528,
      "num_input_tokens_seen": 2214924,
      "step": 795,
      "train_runtime": 2023.5258,
      "train_tokens_per_second": 1094.586
    },
    {
      "epoch": 0.06396271520118926,
      "grad_norm": 0.34675687551498413,
      "learning_rate": 0.00014724540901502506,
      "loss": 1.1535,
      "num_input_tokens_seen": 2217172,
      "step": 796,
      "train_runtime": 2025.438,
      "train_tokens_per_second": 1094.663
    },
    {
      "epoch": 0.06404307037104,
      "grad_norm": 0.32003170251846313,
      "learning_rate": 0.00014717863105175293,
      "loss": 1.238,
      "num_input_tokens_seen": 2219220,
      "step": 797,
      "train_runtime": 2027.283,
      "train_tokens_per_second": 1094.677
    },
    {
      "epoch": 0.06412342554089073,
      "grad_norm": 0.3504277467727661,
      "learning_rate": 0.0001471118530884808,
      "loss": 1.2829,
      "num_input_tokens_seen": 2223568,
      "step": 798,
      "train_runtime": 2030.2734,
      "train_tokens_per_second": 1095.206
    },
    {
      "epoch": 0.06420378071074148,
      "grad_norm": 0.4509447515010834,
      "learning_rate": 0.00014704507512520868,
      "loss": 1.118,
      "num_input_tokens_seen": 2228636,
      "step": 799,
      "train_runtime": 2033.9184,
      "train_tokens_per_second": 1095.735
    },
    {
      "epoch": 0.06428413588059222,
      "grad_norm": 0.38358405232429504,
      "learning_rate": 0.00014697829716193655,
      "loss": 1.2002,
      "num_input_tokens_seen": 2231550,
      "step": 800,
      "train_runtime": 2036.2424,
      "train_tokens_per_second": 1095.916
    },
    {
      "epoch": 0.06436449105044295,
      "grad_norm": 1.3257673978805542,
      "learning_rate": 0.00014691151919866443,
      "loss": 1.0104,
      "num_input_tokens_seen": 2233714,
      "step": 801,
      "train_runtime": 2038.1132,
      "train_tokens_per_second": 1095.971
    },
    {
      "epoch": 0.0644448462202937,
      "grad_norm": 0.338990181684494,
      "learning_rate": 0.00014684474123539233,
      "loss": 1.186,
      "num_input_tokens_seen": 2238176,
      "step": 802,
      "train_runtime": 2041.2052,
      "train_tokens_per_second": 1096.497
    },
    {
      "epoch": 0.06452520139014443,
      "grad_norm": 0.3138526976108551,
      "learning_rate": 0.0001467779632721202,
      "loss": 1.0222,
      "num_input_tokens_seen": 2240232,
      "step": 803,
      "train_runtime": 2043.1096,
      "train_tokens_per_second": 1096.482
    },
    {
      "epoch": 0.06460555655999518,
      "grad_norm": 0.33103451132774353,
      "learning_rate": 0.0001467111853088481,
      "loss": 0.9648,
      "num_input_tokens_seen": 2244260,
      "step": 804,
      "train_runtime": 2046.0091,
      "train_tokens_per_second": 1096.896
    },
    {
      "epoch": 0.06468591172984592,
      "grad_norm": 0.3298628628253937,
      "learning_rate": 0.00014664440734557597,
      "loss": 0.997,
      "num_input_tokens_seen": 2247018,
      "step": 805,
      "train_runtime": 2048.1157,
      "train_tokens_per_second": 1097.115
    },
    {
      "epoch": 0.06476626689969665,
      "grad_norm": 0.4388108253479004,
      "learning_rate": 0.00014657762938230385,
      "loss": 1.0147,
      "num_input_tokens_seen": 2249964,
      "step": 806,
      "train_runtime": 2050.5383,
      "train_tokens_per_second": 1097.255
    },
    {
      "epoch": 0.0648466220695474,
      "grad_norm": 0.3108203709125519,
      "learning_rate": 0.00014651085141903175,
      "loss": 1.1797,
      "num_input_tokens_seen": 2252174,
      "step": 807,
      "train_runtime": 2052.5013,
      "train_tokens_per_second": 1097.283
    },
    {
      "epoch": 0.06492697723939814,
      "grad_norm": 0.28939804434776306,
      "learning_rate": 0.00014644407345575962,
      "loss": 0.8583,
      "num_input_tokens_seen": 2254734,
      "step": 808,
      "train_runtime": 2054.6798,
      "train_tokens_per_second": 1097.365
    },
    {
      "epoch": 0.06500733240924889,
      "grad_norm": 0.32528600096702576,
      "learning_rate": 0.0001463772954924875,
      "loss": 0.9182,
      "num_input_tokens_seen": 2256938,
      "step": 809,
      "train_runtime": 2056.6395,
      "train_tokens_per_second": 1097.391
    },
    {
      "epoch": 0.06508768757909962,
      "grad_norm": 0.3474142849445343,
      "learning_rate": 0.00014631051752921536,
      "loss": 1.0004,
      "num_input_tokens_seen": 2259454,
      "step": 810,
      "train_runtime": 2058.7849,
      "train_tokens_per_second": 1097.47
    },
    {
      "epoch": 0.06516804274895036,
      "grad_norm": 0.2869245111942291,
      "learning_rate": 0.00014624373956594324,
      "loss": 0.8194,
      "num_input_tokens_seen": 2262010,
      "step": 811,
      "train_runtime": 2062.842,
      "train_tokens_per_second": 1096.55
    },
    {
      "epoch": 0.0652483979188011,
      "grad_norm": 0.287260502576828,
      "learning_rate": 0.00014617696160267114,
      "loss": 0.5982,
      "num_input_tokens_seen": 2264084,
      "step": 812,
      "train_runtime": 2064.7321,
      "train_tokens_per_second": 1096.551
    },
    {
      "epoch": 0.06532875308865184,
      "grad_norm": 0.3509330451488495,
      "learning_rate": 0.000146110183639399,
      "loss": 0.9707,
      "num_input_tokens_seen": 2267016,
      "step": 813,
      "train_runtime": 2067.0734,
      "train_tokens_per_second": 1096.727
    },
    {
      "epoch": 0.06540910825850259,
      "grad_norm": 0.2859979271888733,
      "learning_rate": 0.00014604340567612688,
      "loss": 1.0721,
      "num_input_tokens_seen": 2269022,
      "step": 814,
      "train_runtime": 2068.8138,
      "train_tokens_per_second": 1096.774
    },
    {
      "epoch": 0.06548946342835332,
      "grad_norm": 0.3086937367916107,
      "learning_rate": 0.00014597662771285476,
      "loss": 1.0999,
      "num_input_tokens_seen": 2271010,
      "step": 815,
      "train_runtime": 2070.6374,
      "train_tokens_per_second": 1096.769
    },
    {
      "epoch": 0.06556981859820406,
      "grad_norm": 0.3894175589084625,
      "learning_rate": 0.00014590984974958263,
      "loss": 0.9667,
      "num_input_tokens_seen": 2272986,
      "step": 816,
      "train_runtime": 2072.5372,
      "train_tokens_per_second": 1096.717
    },
    {
      "epoch": 0.0656501737680548,
      "grad_norm": 0.29864999651908875,
      "learning_rate": 0.0001458430717863105,
      "loss": 1.132,
      "num_input_tokens_seen": 2276418,
      "step": 817,
      "train_runtime": 2075.0689,
      "train_tokens_per_second": 1097.032
    },
    {
      "epoch": 0.06573052893790554,
      "grad_norm": 0.2906742990016937,
      "learning_rate": 0.0001457762938230384,
      "loss": 1.0294,
      "num_input_tokens_seen": 2279166,
      "step": 818,
      "train_runtime": 2077.2069,
      "train_tokens_per_second": 1097.226
    },
    {
      "epoch": 0.06581088410775628,
      "grad_norm": 0.35557055473327637,
      "learning_rate": 0.00014570951585976628,
      "loss": 1.4576,
      "num_input_tokens_seen": 2282628,
      "step": 819,
      "train_runtime": 2079.7498,
      "train_tokens_per_second": 1097.549
    },
    {
      "epoch": 0.06589123927760702,
      "grad_norm": 0.33739808201789856,
      "learning_rate": 0.00014564273789649415,
      "loss": 1.0374,
      "num_input_tokens_seen": 2284800,
      "step": 820,
      "train_runtime": 2081.7058,
      "train_tokens_per_second": 1097.561
    },
    {
      "epoch": 0.06597159444745776,
      "grad_norm": 0.3174346089363098,
      "learning_rate": 0.00014557595993322205,
      "loss": 1.2244,
      "num_input_tokens_seen": 2286730,
      "step": 821,
      "train_runtime": 2083.4607,
      "train_tokens_per_second": 1097.563
    },
    {
      "epoch": 0.06605194961730851,
      "grad_norm": 0.3701056241989136,
      "learning_rate": 0.00014550918196994992,
      "loss": 1.0022,
      "num_input_tokens_seen": 2288096,
      "step": 822,
      "train_runtime": 2085.0081,
      "train_tokens_per_second": 1097.404
    },
    {
      "epoch": 0.06613230478715924,
      "grad_norm": 0.362595796585083,
      "learning_rate": 0.00014544240400667782,
      "loss": 1.1043,
      "num_input_tokens_seen": 2293658,
      "step": 823,
      "train_runtime": 2088.7749,
      "train_tokens_per_second": 1098.088
    },
    {
      "epoch": 0.06621265995700998,
      "grad_norm": 0.3538409471511841,
      "learning_rate": 0.0001453756260434057,
      "loss": 0.7711,
      "num_input_tokens_seen": 2297144,
      "step": 824,
      "train_runtime": 2091.3574,
      "train_tokens_per_second": 1098.399
    },
    {
      "epoch": 0.06629301512686073,
      "grad_norm": 0.3676832318305969,
      "learning_rate": 0.00014530884808013357,
      "loss": 0.9033,
      "num_input_tokens_seen": 2298690,
      "step": 825,
      "train_runtime": 2092.9404,
      "train_tokens_per_second": 1098.306
    },
    {
      "epoch": 0.06637337029671146,
      "grad_norm": 0.3476060628890991,
      "learning_rate": 0.00014524207011686144,
      "loss": 0.937,
      "num_input_tokens_seen": 2302140,
      "step": 826,
      "train_runtime": 2095.5397,
      "train_tokens_per_second": 1098.59
    },
    {
      "epoch": 0.06645372546656221,
      "grad_norm": 0.3231962323188782,
      "learning_rate": 0.00014517529215358932,
      "loss": 1.0176,
      "num_input_tokens_seen": 2306764,
      "step": 827,
      "train_runtime": 2098.8486,
      "train_tokens_per_second": 1099.062
    },
    {
      "epoch": 0.06653408063641295,
      "grad_norm": 0.297827810049057,
      "learning_rate": 0.00014510851419031722,
      "loss": 1.3588,
      "num_input_tokens_seen": 2310012,
      "step": 828,
      "train_runtime": 2101.2968,
      "train_tokens_per_second": 1099.327
    },
    {
      "epoch": 0.06661443580626368,
      "grad_norm": 0.3019084334373474,
      "learning_rate": 0.0001450417362270451,
      "loss": 1.1462,
      "num_input_tokens_seen": 2313408,
      "step": 829,
      "train_runtime": 2103.8391,
      "train_tokens_per_second": 1099.613
    },
    {
      "epoch": 0.06669479097611443,
      "grad_norm": 0.30470922589302063,
      "learning_rate": 0.00014497495826377296,
      "loss": 0.6886,
      "num_input_tokens_seen": 2315554,
      "step": 830,
      "train_runtime": 2105.8151,
      "train_tokens_per_second": 1099.6
    },
    {
      "epoch": 0.06677514614596516,
      "grad_norm": 0.28943684697151184,
      "learning_rate": 0.00014490818030050084,
      "loss": 0.826,
      "num_input_tokens_seen": 2318524,
      "step": 831,
      "train_runtime": 2108.13,
      "train_tokens_per_second": 1099.801
    },
    {
      "epoch": 0.06685550131581591,
      "grad_norm": 0.24569806456565857,
      "learning_rate": 0.0001448414023372287,
      "loss": 1.0543,
      "num_input_tokens_seen": 2322384,
      "step": 832,
      "train_runtime": 2110.9425,
      "train_tokens_per_second": 1100.165
    },
    {
      "epoch": 0.06693585648566665,
      "grad_norm": 0.36814194917678833,
      "learning_rate": 0.00014477462437395658,
      "loss": 1.1182,
      "num_input_tokens_seen": 2324442,
      "step": 833,
      "train_runtime": 2112.8781,
      "train_tokens_per_second": 1100.131
    },
    {
      "epoch": 0.06701621165551738,
      "grad_norm": 0.31107985973358154,
      "learning_rate": 0.00014470784641068448,
      "loss": 1.1417,
      "num_input_tokens_seen": 2327314,
      "step": 834,
      "train_runtime": 2115.1806,
      "train_tokens_per_second": 1100.291
    },
    {
      "epoch": 0.06709656682536813,
      "grad_norm": 0.31709203124046326,
      "learning_rate": 0.00014464106844741236,
      "loss": 1.2616,
      "num_input_tokens_seen": 2330790,
      "step": 835,
      "train_runtime": 2117.7567,
      "train_tokens_per_second": 1100.594
    },
    {
      "epoch": 0.06717692199521887,
      "grad_norm": 0.2921311557292938,
      "learning_rate": 0.00014457429048414023,
      "loss": 1.684,
      "num_input_tokens_seen": 2334704,
      "step": 836,
      "train_runtime": 2120.5867,
      "train_tokens_per_second": 1100.971
    },
    {
      "epoch": 0.0672572771650696,
      "grad_norm": 0.33907264471054077,
      "learning_rate": 0.0001445075125208681,
      "loss": 1.0741,
      "num_input_tokens_seen": 2336548,
      "step": 837,
      "train_runtime": 2122.2753,
      "train_tokens_per_second": 1100.964
    },
    {
      "epoch": 0.06733763233492035,
      "grad_norm": 0.37785765528678894,
      "learning_rate": 0.000144440734557596,
      "loss": 0.9078,
      "num_input_tokens_seen": 2339354,
      "step": 838,
      "train_runtime": 2124.4297,
      "train_tokens_per_second": 1101.168
    },
    {
      "epoch": 0.06741798750477108,
      "grad_norm": 0.28260868787765503,
      "learning_rate": 0.00014437395659432388,
      "loss": 0.5368,
      "num_input_tokens_seen": 2341304,
      "step": 839,
      "train_runtime": 2126.1999,
      "train_tokens_per_second": 1101.168
    },
    {
      "epoch": 0.06749834267462183,
      "grad_norm": 0.32741856575012207,
      "learning_rate": 0.00014430717863105178,
      "loss": 0.8695,
      "num_input_tokens_seen": 2344036,
      "step": 840,
      "train_runtime": 2128.4882,
      "train_tokens_per_second": 1101.268
    },
    {
      "epoch": 0.06757869784447257,
      "grad_norm": 0.2809388041496277,
      "learning_rate": 0.00014424040066777965,
      "loss": 0.9008,
      "num_input_tokens_seen": 2347434,
      "step": 841,
      "train_runtime": 2133.0161,
      "train_tokens_per_second": 1100.523
    },
    {
      "epoch": 0.0676590530143233,
      "grad_norm": 0.319504052400589,
      "learning_rate": 0.00014417362270450752,
      "loss": 0.8029,
      "num_input_tokens_seen": 2349046,
      "step": 842,
      "train_runtime": 2134.6633,
      "train_tokens_per_second": 1100.429
    },
    {
      "epoch": 0.06773940818417405,
      "grad_norm": 0.39870065450668335,
      "learning_rate": 0.0001441068447412354,
      "loss": 0.8163,
      "num_input_tokens_seen": 2350854,
      "step": 843,
      "train_runtime": 2136.3911,
      "train_tokens_per_second": 1100.386
    },
    {
      "epoch": 0.06781976335402479,
      "grad_norm": 0.3060488700866699,
      "learning_rate": 0.0001440400667779633,
      "loss": 1.4375,
      "num_input_tokens_seen": 2352832,
      "step": 844,
      "train_runtime": 2138.0359,
      "train_tokens_per_second": 1100.464
    },
    {
      "epoch": 0.06790011852387554,
      "grad_norm": 0.3320007622241974,
      "learning_rate": 0.00014397328881469117,
      "loss": 1.0134,
      "num_input_tokens_seen": 2355290,
      "step": 845,
      "train_runtime": 2140.1806,
      "train_tokens_per_second": 1100.51
    },
    {
      "epoch": 0.06798047369372627,
      "grad_norm": 0.27483120560646057,
      "learning_rate": 0.00014390651085141904,
      "loss": 1.0031,
      "num_input_tokens_seen": 2358390,
      "step": 846,
      "train_runtime": 2142.6555,
      "train_tokens_per_second": 1100.686
    },
    {
      "epoch": 0.068060828863577,
      "grad_norm": 0.2573234736919403,
      "learning_rate": 0.00014383973288814692,
      "loss": 0.7291,
      "num_input_tokens_seen": 2363332,
      "step": 847,
      "train_runtime": 2145.9755,
      "train_tokens_per_second": 1101.286
    },
    {
      "epoch": 0.06814118403342775,
      "grad_norm": 0.36865949630737305,
      "learning_rate": 0.0001437729549248748,
      "loss": 0.9043,
      "num_input_tokens_seen": 2364422,
      "step": 848,
      "train_runtime": 2147.4875,
      "train_tokens_per_second": 1101.018
    },
    {
      "epoch": 0.06822153920327849,
      "grad_norm": 0.34308263659477234,
      "learning_rate": 0.00014370617696160266,
      "loss": 0.9577,
      "num_input_tokens_seen": 2368038,
      "step": 849,
      "train_runtime": 2150.1985,
      "train_tokens_per_second": 1101.311
    },
    {
      "epoch": 0.06830189437312924,
      "grad_norm": 0.31565070152282715,
      "learning_rate": 0.00014363939899833056,
      "loss": 0.8389,
      "num_input_tokens_seen": 2369928,
      "step": 850,
      "train_runtime": 2151.9561,
      "train_tokens_per_second": 1101.29
    },
    {
      "epoch": 0.06838224954297997,
      "grad_norm": 0.3276442289352417,
      "learning_rate": 0.00014357262103505844,
      "loss": 1.0645,
      "num_input_tokens_seen": 2372630,
      "step": 851,
      "train_runtime": 2154.1483,
      "train_tokens_per_second": 1101.424
    },
    {
      "epoch": 0.06846260471283071,
      "grad_norm": 0.33249008655548096,
      "learning_rate": 0.0001435058430717863,
      "loss": 0.9825,
      "num_input_tokens_seen": 2375532,
      "step": 852,
      "train_runtime": 2156.4452,
      "train_tokens_per_second": 1101.596
    },
    {
      "epoch": 0.06854295988268146,
      "grad_norm": 0.30959081649780273,
      "learning_rate": 0.00014343906510851418,
      "loss": 0.7941,
      "num_input_tokens_seen": 2377566,
      "step": 853,
      "train_runtime": 2158.2389,
      "train_tokens_per_second": 1101.623
    },
    {
      "epoch": 0.06862331505253219,
      "grad_norm": 0.33025220036506653,
      "learning_rate": 0.00014337228714524205,
      "loss": 1.0709,
      "num_input_tokens_seen": 2379800,
      "step": 854,
      "train_runtime": 2160.1969,
      "train_tokens_per_second": 1101.659
    },
    {
      "epoch": 0.06870367022238293,
      "grad_norm": 0.32086995244026184,
      "learning_rate": 0.00014330550918196995,
      "loss": 1.1182,
      "num_input_tokens_seen": 2382506,
      "step": 855,
      "train_runtime": 2162.4812,
      "train_tokens_per_second": 1101.746
    },
    {
      "epoch": 0.06878402539223367,
      "grad_norm": 0.2943474054336548,
      "learning_rate": 0.00014323873121869783,
      "loss": 0.9666,
      "num_input_tokens_seen": 2383976,
      "step": 856,
      "train_runtime": 2164.0846,
      "train_tokens_per_second": 1101.609
    },
    {
      "epoch": 0.06886438056208441,
      "grad_norm": 0.25427353382110596,
      "learning_rate": 0.00014317195325542573,
      "loss": 0.8151,
      "num_input_tokens_seen": 2386344,
      "step": 857,
      "train_runtime": 2166.0732,
      "train_tokens_per_second": 1101.691
    },
    {
      "epoch": 0.06894473573193516,
      "grad_norm": 0.3360339403152466,
      "learning_rate": 0.0001431051752921536,
      "loss": 1.0084,
      "num_input_tokens_seen": 2388192,
      "step": 858,
      "train_runtime": 2167.7965,
      "train_tokens_per_second": 1101.668
    },
    {
      "epoch": 0.0690250909017859,
      "grad_norm": 0.2708817422389984,
      "learning_rate": 0.00014303839732888147,
      "loss": 0.7753,
      "num_input_tokens_seen": 2390812,
      "step": 859,
      "train_runtime": 2169.9736,
      "train_tokens_per_second": 1101.77
    },
    {
      "epoch": 0.06910544607163663,
      "grad_norm": 0.28567513823509216,
      "learning_rate": 0.00014297161936560937,
      "loss": 0.9032,
      "num_input_tokens_seen": 2393582,
      "step": 860,
      "train_runtime": 2172.2403,
      "train_tokens_per_second": 1101.896
    },
    {
      "epoch": 0.06918580124148738,
      "grad_norm": 0.3475951850414276,
      "learning_rate": 0.00014290484140233725,
      "loss": 1.1451,
      "num_input_tokens_seen": 2395954,
      "step": 861,
      "train_runtime": 2174.2098,
      "train_tokens_per_second": 1101.988
    },
    {
      "epoch": 0.06926615641133811,
      "grad_norm": 0.27370399236679077,
      "learning_rate": 0.00014283806343906512,
      "loss": 0.6476,
      "num_input_tokens_seen": 2399068,
      "step": 862,
      "train_runtime": 2176.6151,
      "train_tokens_per_second": 1102.201
    },
    {
      "epoch": 0.06934651158118886,
      "grad_norm": 0.3076470196247101,
      "learning_rate": 0.000142771285475793,
      "loss": 1.3084,
      "num_input_tokens_seen": 2402940,
      "step": 863,
      "train_runtime": 2179.3847,
      "train_tokens_per_second": 1102.577
    },
    {
      "epoch": 0.0694268667510396,
      "grad_norm": 0.26172661781311035,
      "learning_rate": 0.00014270450751252087,
      "loss": 0.9147,
      "num_input_tokens_seen": 2405548,
      "step": 864,
      "train_runtime": 2181.4025,
      "train_tokens_per_second": 1102.753
    },
    {
      "epoch": 0.06950722192089033,
      "grad_norm": 0.23512691259384155,
      "learning_rate": 0.00014263772954924874,
      "loss": 0.5001,
      "num_input_tokens_seen": 2408136,
      "step": 865,
      "train_runtime": 2183.4872,
      "train_tokens_per_second": 1102.885
    },
    {
      "epoch": 0.06958757709074108,
      "grad_norm": 0.32899776101112366,
      "learning_rate": 0.00014257095158597664,
      "loss": 0.7307,
      "num_input_tokens_seen": 2411650,
      "step": 866,
      "train_runtime": 2186.067,
      "train_tokens_per_second": 1103.191
    },
    {
      "epoch": 0.06966793226059181,
      "grad_norm": 0.3800621032714844,
      "learning_rate": 0.00014250417362270451,
      "loss": 1.2247,
      "num_input_tokens_seen": 2413314,
      "step": 867,
      "train_runtime": 2187.77,
      "train_tokens_per_second": 1103.093
    },
    {
      "epoch": 0.06974828743044256,
      "grad_norm": 0.3513256311416626,
      "learning_rate": 0.0001424373956594324,
      "loss": 1.1149,
      "num_input_tokens_seen": 2416568,
      "step": 868,
      "train_runtime": 2190.2241,
      "train_tokens_per_second": 1103.343
    },
    {
      "epoch": 0.0698286426002933,
      "grad_norm": 0.2976699769496918,
      "learning_rate": 0.00014237061769616026,
      "loss": 1.0455,
      "num_input_tokens_seen": 2418772,
      "step": 869,
      "train_runtime": 2192.1403,
      "train_tokens_per_second": 1103.384
    },
    {
      "epoch": 0.06990899777014403,
      "grad_norm": 0.3433803617954254,
      "learning_rate": 0.00014230383973288813,
      "loss": 0.7543,
      "num_input_tokens_seen": 2422658,
      "step": 870,
      "train_runtime": 2194.9572,
      "train_tokens_per_second": 1103.738
    },
    {
      "epoch": 0.06998935293999478,
      "grad_norm": 0.2978561520576477,
      "learning_rate": 0.00014223706176961603,
      "loss": 0.9657,
      "num_input_tokens_seen": 2426162,
      "step": 871,
      "train_runtime": 2199.5277,
      "train_tokens_per_second": 1103.038
    },
    {
      "epoch": 0.07006970810984552,
      "grad_norm": 0.32957741618156433,
      "learning_rate": 0.0001421702838063439,
      "loss": 1.1739,
      "num_input_tokens_seen": 2429586,
      "step": 872,
      "train_runtime": 2202.1639,
      "train_tokens_per_second": 1103.272
    },
    {
      "epoch": 0.07015006327969625,
      "grad_norm": 0.27114084362983704,
      "learning_rate": 0.00014210350584307178,
      "loss": 0.817,
      "num_input_tokens_seen": 2432888,
      "step": 873,
      "train_runtime": 2204.6522,
      "train_tokens_per_second": 1103.525
    },
    {
      "epoch": 0.070230418449547,
      "grad_norm": 0.29244133830070496,
      "learning_rate": 0.00014203672787979968,
      "loss": 0.8771,
      "num_input_tokens_seen": 2437370,
      "step": 874,
      "train_runtime": 2207.811,
      "train_tokens_per_second": 1103.976
    },
    {
      "epoch": 0.07031077361939773,
      "grad_norm": 0.32263773679733276,
      "learning_rate": 0.00014196994991652755,
      "loss": 1.2008,
      "num_input_tokens_seen": 2440254,
      "step": 875,
      "train_runtime": 2210.0591,
      "train_tokens_per_second": 1104.158
    },
    {
      "epoch": 0.07039112878924848,
      "grad_norm": 0.31919553875923157,
      "learning_rate": 0.00014190317195325545,
      "loss": 1.3151,
      "num_input_tokens_seen": 2443340,
      "step": 876,
      "train_runtime": 2212.5313,
      "train_tokens_per_second": 1104.319
    },
    {
      "epoch": 0.07047148395909922,
      "grad_norm": 0.35774099826812744,
      "learning_rate": 0.00014183639398998333,
      "loss": 1.3196,
      "num_input_tokens_seen": 2448938,
      "step": 877,
      "train_runtime": 2216.4178,
      "train_tokens_per_second": 1104.908
    },
    {
      "epoch": 0.07055183912894995,
      "grad_norm": 0.3260195255279541,
      "learning_rate": 0.0001417696160267112,
      "loss": 1.1791,
      "num_input_tokens_seen": 2450678,
      "step": 878,
      "train_runtime": 2218.0835,
      "train_tokens_per_second": 1104.863
    },
    {
      "epoch": 0.0706321942988007,
      "grad_norm": 0.329819917678833,
      "learning_rate": 0.00014170283806343907,
      "loss": 0.8675,
      "num_input_tokens_seen": 2452838,
      "step": 879,
      "train_runtime": 2219.9794,
      "train_tokens_per_second": 1104.892
    },
    {
      "epoch": 0.07071254946865144,
      "grad_norm": 0.28507351875305176,
      "learning_rate": 0.00014163606010016695,
      "loss": 0.8498,
      "num_input_tokens_seen": 2455320,
      "step": 880,
      "train_runtime": 2222.0154,
      "train_tokens_per_second": 1104.997
    },
    {
      "epoch": 0.07079290463850219,
      "grad_norm": 0.33230507373809814,
      "learning_rate": 0.00014156928213689482,
      "loss": 0.9981,
      "num_input_tokens_seen": 2458492,
      "step": 881,
      "train_runtime": 2224.3685,
      "train_tokens_per_second": 1105.254
    },
    {
      "epoch": 0.07087325980835292,
      "grad_norm": 0.3293096423149109,
      "learning_rate": 0.00014150250417362272,
      "loss": 0.9813,
      "num_input_tokens_seen": 2461146,
      "step": 882,
      "train_runtime": 2226.5517,
      "train_tokens_per_second": 1105.362
    },
    {
      "epoch": 0.07095361497820366,
      "grad_norm": 0.2984834909439087,
      "learning_rate": 0.0001414357262103506,
      "loss": 0.8875,
      "num_input_tokens_seen": 2463364,
      "step": 883,
      "train_runtime": 2228.509,
      "train_tokens_per_second": 1105.387
    },
    {
      "epoch": 0.0710339701480544,
      "grad_norm": 0.3159421980381012,
      "learning_rate": 0.00014136894824707847,
      "loss": 0.9008,
      "num_input_tokens_seen": 2468600,
      "step": 884,
      "train_runtime": 2232.2104,
      "train_tokens_per_second": 1105.899
    },
    {
      "epoch": 0.07111432531790514,
      "grad_norm": 0.2863374650478363,
      "learning_rate": 0.00014130217028380634,
      "loss": 0.71,
      "num_input_tokens_seen": 2471004,
      "step": 885,
      "train_runtime": 2234.2492,
      "train_tokens_per_second": 1105.966
    },
    {
      "epoch": 0.07119468048775589,
      "grad_norm": 0.35992884635925293,
      "learning_rate": 0.0001412353923205342,
      "loss": 0.9864,
      "num_input_tokens_seen": 2472852,
      "step": 886,
      "train_runtime": 2236.0199,
      "train_tokens_per_second": 1105.917
    },
    {
      "epoch": 0.07127503565760662,
      "grad_norm": 0.2793021500110626,
      "learning_rate": 0.0001411686143572621,
      "loss": 1.023,
      "num_input_tokens_seen": 2475582,
      "step": 887,
      "train_runtime": 2238.2781,
      "train_tokens_per_second": 1106.021
    },
    {
      "epoch": 0.07135539082745736,
      "grad_norm": 0.2968974709510803,
      "learning_rate": 0.00014110183639398999,
      "loss": 1.1568,
      "num_input_tokens_seen": 2477236,
      "step": 888,
      "train_runtime": 2239.9187,
      "train_tokens_per_second": 1105.949
    },
    {
      "epoch": 0.0714357459973081,
      "grad_norm": 0.2545320987701416,
      "learning_rate": 0.00014103505843071786,
      "loss": 0.9085,
      "num_input_tokens_seen": 2480568,
      "step": 889,
      "train_runtime": 2242.4181,
      "train_tokens_per_second": 1106.202
    },
    {
      "epoch": 0.07151610116715884,
      "grad_norm": 0.30533871054649353,
      "learning_rate": 0.00014096828046744576,
      "loss": 0.9795,
      "num_input_tokens_seen": 2483616,
      "step": 890,
      "train_runtime": 2244.7653,
      "train_tokens_per_second": 1106.403
    },
    {
      "epoch": 0.07159645633700958,
      "grad_norm": 0.32545360922813416,
      "learning_rate": 0.00014090150250417363,
      "loss": 1.3072,
      "num_input_tokens_seen": 2486592,
      "step": 891,
      "train_runtime": 2247.1278,
      "train_tokens_per_second": 1106.565
    },
    {
      "epoch": 0.07167681150686032,
      "grad_norm": 0.39703962206840515,
      "learning_rate": 0.00014083472454090153,
      "loss": 0.9259,
      "num_input_tokens_seen": 2489772,
      "step": 892,
      "train_runtime": 2249.528,
      "train_tokens_per_second": 1106.797
    },
    {
      "epoch": 0.07175716667671106,
      "grad_norm": 0.30668574571609497,
      "learning_rate": 0.0001407679465776294,
      "loss": 0.7198,
      "num_input_tokens_seen": 2491744,
      "step": 893,
      "train_runtime": 2251.3359,
      "train_tokens_per_second": 1106.785
    },
    {
      "epoch": 0.07183752184656181,
      "grad_norm": 0.3326931297779083,
      "learning_rate": 0.00014070116861435728,
      "loss": 0.8413,
      "num_input_tokens_seen": 2493878,
      "step": 894,
      "train_runtime": 2253.2432,
      "train_tokens_per_second": 1106.795
    },
    {
      "epoch": 0.07191787701641254,
      "grad_norm": 0.3320900797843933,
      "learning_rate": 0.00014063439065108515,
      "loss": 1.0939,
      "num_input_tokens_seen": 2497272,
      "step": 895,
      "train_runtime": 2255.8107,
      "train_tokens_per_second": 1107.04
    },
    {
      "epoch": 0.07199823218626328,
      "grad_norm": 0.2334388792514801,
      "learning_rate": 0.00014056761268781303,
      "loss": 0.6992,
      "num_input_tokens_seen": 2500572,
      "step": 896,
      "train_runtime": 2258.2501,
      "train_tokens_per_second": 1107.305
    },
    {
      "epoch": 0.07207858735611403,
      "grad_norm": 0.2968538701534271,
      "learning_rate": 0.0001405008347245409,
      "loss": 0.9063,
      "num_input_tokens_seen": 2503412,
      "step": 897,
      "train_runtime": 2260.531,
      "train_tokens_per_second": 1107.444
    },
    {
      "epoch": 0.07215894252596476,
      "grad_norm": 0.4175359010696411,
      "learning_rate": 0.0001404340567612688,
      "loss": 1.0619,
      "num_input_tokens_seen": 2504978,
      "step": 898,
      "train_runtime": 2262.1536,
      "train_tokens_per_second": 1107.342
    },
    {
      "epoch": 0.07223929769581551,
      "grad_norm": 0.29439496994018555,
      "learning_rate": 0.00014036727879799667,
      "loss": 1.1068,
      "num_input_tokens_seen": 2507652,
      "step": 899,
      "train_runtime": 2264.3545,
      "train_tokens_per_second": 1107.447
    },
    {
      "epoch": 0.07231965286566625,
      "grad_norm": 0.25840285420417786,
      "learning_rate": 0.00014030050083472454,
      "loss": 0.6788,
      "num_input_tokens_seen": 2510880,
      "step": 900,
      "train_runtime": 2266.7892,
      "train_tokens_per_second": 1107.681
    },
    {
      "epoch": 0.07231965286566625,
      "eval_loss": 0.9866039752960205,
      "eval_runtime": 92.2639,
      "eval_samples_per_second": 10.773,
      "eval_steps_per_second": 5.387,
      "num_input_tokens_seen": 2510880,
      "step": 900
    },
    {
      "epoch": 0.07240000803551698,
      "grad_norm": 0.3053082525730133,
      "learning_rate": 0.00014023372287145242,
      "loss": 0.8406,
      "num_input_tokens_seen": 2515564,
      "step": 901,
      "train_runtime": 2364.3843,
      "train_tokens_per_second": 1063.94
    },
    {
      "epoch": 0.07248036320536773,
      "grad_norm": 0.3301785886287689,
      "learning_rate": 0.0001401669449081803,
      "loss": 0.9861,
      "num_input_tokens_seen": 2517432,
      "step": 902,
      "train_runtime": 2366.1807,
      "train_tokens_per_second": 1063.922
    },
    {
      "epoch": 0.07256071837521846,
      "grad_norm": 0.25372639298439026,
      "learning_rate": 0.0001401001669449082,
      "loss": 0.75,
      "num_input_tokens_seen": 2520406,
      "step": 903,
      "train_runtime": 2368.4827,
      "train_tokens_per_second": 1064.144
    },
    {
      "epoch": 0.07264107354506921,
      "grad_norm": 0.2701776921749115,
      "learning_rate": 0.00014003338898163606,
      "loss": 1.0623,
      "num_input_tokens_seen": 2522934,
      "step": 904,
      "train_runtime": 2370.5351,
      "train_tokens_per_second": 1064.289
    },
    {
      "epoch": 0.07272142871491995,
      "grad_norm": 0.4370821714401245,
      "learning_rate": 0.00013996661101836394,
      "loss": 1.2123,
      "num_input_tokens_seen": 2524668,
      "step": 905,
      "train_runtime": 2372.2502,
      "train_tokens_per_second": 1064.25
    },
    {
      "epoch": 0.07280178388477068,
      "grad_norm": 0.28252699971199036,
      "learning_rate": 0.0001398998330550918,
      "loss": 0.9468,
      "num_input_tokens_seen": 2527960,
      "step": 906,
      "train_runtime": 2374.7204,
      "train_tokens_per_second": 1064.53
    },
    {
      "epoch": 0.07288213905462143,
      "grad_norm": 0.42438647150993347,
      "learning_rate": 0.0001398330550918197,
      "loss": 0.9752,
      "num_input_tokens_seen": 2529190,
      "step": 907,
      "train_runtime": 2376.2171,
      "train_tokens_per_second": 1064.377
    },
    {
      "epoch": 0.07296249422447217,
      "grad_norm": 0.31667259335517883,
      "learning_rate": 0.00013976627712854758,
      "loss": 1.1575,
      "num_input_tokens_seen": 2531280,
      "step": 908,
      "train_runtime": 2378.042,
      "train_tokens_per_second": 1064.439
    },
    {
      "epoch": 0.0730428493943229,
      "grad_norm": 0.2693456709384918,
      "learning_rate": 0.00013969949916527548,
      "loss": 0.8908,
      "num_input_tokens_seen": 2534944,
      "step": 909,
      "train_runtime": 2380.7252,
      "train_tokens_per_second": 1064.778
    },
    {
      "epoch": 0.07312320456417365,
      "grad_norm": 0.30570200085639954,
      "learning_rate": 0.00013963272120200336,
      "loss": 0.8179,
      "num_input_tokens_seen": 2537856,
      "step": 910,
      "train_runtime": 2382.9982,
      "train_tokens_per_second": 1064.984
    },
    {
      "epoch": 0.07320355973402438,
      "grad_norm": 0.33213672041893005,
      "learning_rate": 0.00013956594323873123,
      "loss": 0.9409,
      "num_input_tokens_seen": 2540764,
      "step": 911,
      "train_runtime": 2385.2901,
      "train_tokens_per_second": 1065.18
    },
    {
      "epoch": 0.07328391490387513,
      "grad_norm": 0.2745407819747925,
      "learning_rate": 0.0001394991652754591,
      "loss": 0.8521,
      "num_input_tokens_seen": 2544394,
      "step": 912,
      "train_runtime": 2387.9627,
      "train_tokens_per_second": 1065.508
    },
    {
      "epoch": 0.07336427007372587,
      "grad_norm": 0.27919888496398926,
      "learning_rate": 0.00013943238731218698,
      "loss": 0.8954,
      "num_input_tokens_seen": 2546268,
      "step": 913,
      "train_runtime": 2389.7563,
      "train_tokens_per_second": 1065.493
    },
    {
      "epoch": 0.0734446252435766,
      "grad_norm": 0.3274606466293335,
      "learning_rate": 0.00013936560934891488,
      "loss": 1.2778,
      "num_input_tokens_seen": 2548368,
      "step": 914,
      "train_runtime": 2391.5747,
      "train_tokens_per_second": 1065.561
    },
    {
      "epoch": 0.07352498041342735,
      "grad_norm": 0.2568143606185913,
      "learning_rate": 0.00013929883138564275,
      "loss": 0.9203,
      "num_input_tokens_seen": 2552286,
      "step": 915,
      "train_runtime": 2394.4307,
      "train_tokens_per_second": 1065.926
    },
    {
      "epoch": 0.07360533558327809,
      "grad_norm": 0.32585129141807556,
      "learning_rate": 0.00013923205342237062,
      "loss": 1.0161,
      "num_input_tokens_seen": 2555594,
      "step": 916,
      "train_runtime": 2396.8311,
      "train_tokens_per_second": 1066.239
    },
    {
      "epoch": 0.07368569075312884,
      "grad_norm": 0.3598208427429199,
      "learning_rate": 0.0001391652754590985,
      "loss": 1.2984,
      "num_input_tokens_seen": 2557780,
      "step": 917,
      "train_runtime": 2398.7323,
      "train_tokens_per_second": 1066.305
    },
    {
      "epoch": 0.07376604592297957,
      "grad_norm": 3.1584792137145996,
      "learning_rate": 0.00013909849749582637,
      "loss": 0.9168,
      "num_input_tokens_seen": 2559650,
      "step": 918,
      "train_runtime": 2400.505,
      "train_tokens_per_second": 1066.296
    },
    {
      "epoch": 0.0738464010928303,
      "grad_norm": 0.20963475108146667,
      "learning_rate": 0.00013903171953255427,
      "loss": 0.7538,
      "num_input_tokens_seen": 2562186,
      "step": 919,
      "train_runtime": 2402.557,
      "train_tokens_per_second": 1066.441
    },
    {
      "epoch": 0.07392675626268105,
      "grad_norm": 0.33692628145217896,
      "learning_rate": 0.00013896494156928214,
      "loss": 0.9828,
      "num_input_tokens_seen": 2564266,
      "step": 920,
      "train_runtime": 2404.3244,
      "train_tokens_per_second": 1066.522
    },
    {
      "epoch": 0.07400711143253179,
      "grad_norm": 0.2947012186050415,
      "learning_rate": 0.00013889816360601002,
      "loss": 0.9959,
      "num_input_tokens_seen": 2567370,
      "step": 921,
      "train_runtime": 2406.7523,
      "train_tokens_per_second": 1066.736
    },
    {
      "epoch": 0.07408746660238254,
      "grad_norm": 0.24923081696033478,
      "learning_rate": 0.0001388313856427379,
      "loss": 0.7778,
      "num_input_tokens_seen": 2569422,
      "step": 922,
      "train_runtime": 2408.5658,
      "train_tokens_per_second": 1066.785
    },
    {
      "epoch": 0.07416782177223327,
      "grad_norm": 0.3389129042625427,
      "learning_rate": 0.00013876460767946576,
      "loss": 0.8421,
      "num_input_tokens_seen": 2573156,
      "step": 923,
      "train_runtime": 2411.4311,
      "train_tokens_per_second": 1067.066
    },
    {
      "epoch": 0.07424817694208401,
      "grad_norm": 0.33941352367401123,
      "learning_rate": 0.00013869782971619366,
      "loss": 0.7385,
      "num_input_tokens_seen": 2574820,
      "step": 924,
      "train_runtime": 2413.0289,
      "train_tokens_per_second": 1067.049
    },
    {
      "epoch": 0.07432853211193476,
      "grad_norm": 0.25243544578552246,
      "learning_rate": 0.00013863105175292154,
      "loss": 0.84,
      "num_input_tokens_seen": 2577314,
      "step": 925,
      "train_runtime": 2415.0644,
      "train_tokens_per_second": 1067.182
    },
    {
      "epoch": 0.07440888728178549,
      "grad_norm": 0.31679514050483704,
      "learning_rate": 0.00013856427378964944,
      "loss": 1.0994,
      "num_input_tokens_seen": 2579960,
      "step": 926,
      "train_runtime": 2417.1973,
      "train_tokens_per_second": 1067.335
    },
    {
      "epoch": 0.07448924245163623,
      "grad_norm": 0.31016841530799866,
      "learning_rate": 0.0001384974958263773,
      "loss": 1.0724,
      "num_input_tokens_seen": 2583246,
      "step": 927,
      "train_runtime": 2419.6613,
      "train_tokens_per_second": 1067.606
    },
    {
      "epoch": 0.07456959762148697,
      "grad_norm": 0.31450945138931274,
      "learning_rate": 0.00013843071786310518,
      "loss": 1.0615,
      "num_input_tokens_seen": 2585238,
      "step": 928,
      "train_runtime": 2421.4737,
      "train_tokens_per_second": 1067.63
    },
    {
      "epoch": 0.07464995279133771,
      "grad_norm": 0.33551156520843506,
      "learning_rate": 0.00013836393989983308,
      "loss": 1.165,
      "num_input_tokens_seen": 2588802,
      "step": 929,
      "train_runtime": 2424.0847,
      "train_tokens_per_second": 1067.95
    },
    {
      "epoch": 0.07473030796118846,
      "grad_norm": 0.31785404682159424,
      "learning_rate": 0.00013829716193656096,
      "loss": 0.8766,
      "num_input_tokens_seen": 2591080,
      "step": 930,
      "train_runtime": 2425.9993,
      "train_tokens_per_second": 1068.046
    },
    {
      "epoch": 0.07481066313103919,
      "grad_norm": 0.3630451560020447,
      "learning_rate": 0.00013823038397328883,
      "loss": 1.1808,
      "num_input_tokens_seen": 2593852,
      "step": 931,
      "train_runtime": 2430.233,
      "train_tokens_per_second": 1067.326
    },
    {
      "epoch": 0.07489101830088993,
      "grad_norm": 0.3400619328022003,
      "learning_rate": 0.0001381636060100167,
      "loss": 0.9443,
      "num_input_tokens_seen": 2595960,
      "step": 932,
      "train_runtime": 2432.1372,
      "train_tokens_per_second": 1067.358
    },
    {
      "epoch": 0.07497137347074068,
      "grad_norm": 0.3520766496658325,
      "learning_rate": 0.00013809682804674458,
      "loss": 1.0372,
      "num_input_tokens_seen": 2597788,
      "step": 933,
      "train_runtime": 2433.888,
      "train_tokens_per_second": 1067.341
    },
    {
      "epoch": 0.07505172864059141,
      "grad_norm": 0.3400426208972931,
      "learning_rate": 0.00013803005008347245,
      "loss": 1.0806,
      "num_input_tokens_seen": 2599384,
      "step": 934,
      "train_runtime": 2435.563,
      "train_tokens_per_second": 1067.262
    },
    {
      "epoch": 0.07513208381044216,
      "grad_norm": 0.2729679048061371,
      "learning_rate": 0.00013796327212020035,
      "loss": 0.7861,
      "num_input_tokens_seen": 2601622,
      "step": 935,
      "train_runtime": 2437.4584,
      "train_tokens_per_second": 1067.35
    },
    {
      "epoch": 0.0752124389802929,
      "grad_norm": 0.30108392238616943,
      "learning_rate": 0.00013789649415692822,
      "loss": 0.9608,
      "num_input_tokens_seen": 2603164,
      "step": 936,
      "train_runtime": 2439.0252,
      "train_tokens_per_second": 1067.297
    },
    {
      "epoch": 0.07529279415014363,
      "grad_norm": 0.35101914405822754,
      "learning_rate": 0.0001378297161936561,
      "loss": 0.8204,
      "num_input_tokens_seen": 2605520,
      "step": 937,
      "train_runtime": 2440.953,
      "train_tokens_per_second": 1067.419
    },
    {
      "epoch": 0.07537314931999438,
      "grad_norm": 0.2950875163078308,
      "learning_rate": 0.00013776293823038397,
      "loss": 0.9251,
      "num_input_tokens_seen": 2608150,
      "step": 938,
      "train_runtime": 2443.0413,
      "train_tokens_per_second": 1067.583
    },
    {
      "epoch": 0.07545350448984511,
      "grad_norm": 0.36441686749458313,
      "learning_rate": 0.00013769616026711184,
      "loss": 0.9219,
      "num_input_tokens_seen": 2611550,
      "step": 939,
      "train_runtime": 2445.6027,
      "train_tokens_per_second": 1067.855
    },
    {
      "epoch": 0.07553385965969586,
      "grad_norm": 0.33548998832702637,
      "learning_rate": 0.00013762938230383974,
      "loss": 0.7137,
      "num_input_tokens_seen": 2615298,
      "step": 940,
      "train_runtime": 2448.4347,
      "train_tokens_per_second": 1068.151
    },
    {
      "epoch": 0.0756142148295466,
      "grad_norm": 0.35290947556495667,
      "learning_rate": 0.00013756260434056762,
      "loss": 1.067,
      "num_input_tokens_seen": 2617814,
      "step": 941,
      "train_runtime": 2450.4719,
      "train_tokens_per_second": 1068.29
    },
    {
      "epoch": 0.07569456999939733,
      "grad_norm": 0.3726901710033417,
      "learning_rate": 0.0001374958263772955,
      "loss": 1.0097,
      "num_input_tokens_seen": 2620472,
      "step": 942,
      "train_runtime": 2452.5967,
      "train_tokens_per_second": 1068.448
    },
    {
      "epoch": 0.07577492516924808,
      "grad_norm": 0.34148970246315,
      "learning_rate": 0.0001374290484140234,
      "loss": 0.7774,
      "num_input_tokens_seen": 2622866,
      "step": 943,
      "train_runtime": 2454.4761,
      "train_tokens_per_second": 1068.605
    },
    {
      "epoch": 0.07585528033909882,
      "grad_norm": 0.30516350269317627,
      "learning_rate": 0.00013736227045075126,
      "loss": 1.0457,
      "num_input_tokens_seen": 2626722,
      "step": 944,
      "train_runtime": 2457.2673,
      "train_tokens_per_second": 1068.961
    },
    {
      "epoch": 0.07593563550894955,
      "grad_norm": 0.3085141181945801,
      "learning_rate": 0.00013729549248747916,
      "loss": 0.8317,
      "num_input_tokens_seen": 2629280,
      "step": 945,
      "train_runtime": 2459.3703,
      "train_tokens_per_second": 1069.087
    },
    {
      "epoch": 0.0760159906788003,
      "grad_norm": 0.2886466383934021,
      "learning_rate": 0.00013722871452420704,
      "loss": 1.0737,
      "num_input_tokens_seen": 2632166,
      "step": 946,
      "train_runtime": 2461.6671,
      "train_tokens_per_second": 1069.262
    },
    {
      "epoch": 0.07609634584865103,
      "grad_norm": 0.34234121441841125,
      "learning_rate": 0.0001371619365609349,
      "loss": 1.1575,
      "num_input_tokens_seen": 2634576,
      "step": 947,
      "train_runtime": 2463.6806,
      "train_tokens_per_second": 1069.366
    },
    {
      "epoch": 0.07617670101850178,
      "grad_norm": 0.33628860116004944,
      "learning_rate": 0.00013709515859766278,
      "loss": 0.769,
      "num_input_tokens_seen": 2638766,
      "step": 948,
      "train_runtime": 2466.7,
      "train_tokens_per_second": 1069.756
    },
    {
      "epoch": 0.07625705618835252,
      "grad_norm": 0.30892881751060486,
      "learning_rate": 0.00013702838063439065,
      "loss": 0.8611,
      "num_input_tokens_seen": 2640364,
      "step": 949,
      "train_runtime": 2468.3586,
      "train_tokens_per_second": 1069.684
    },
    {
      "epoch": 0.07633741135820325,
      "grad_norm": 0.3496696949005127,
      "learning_rate": 0.00013696160267111853,
      "loss": 1.0535,
      "num_input_tokens_seen": 2642190,
      "step": 950,
      "train_runtime": 2470.1155,
      "train_tokens_per_second": 1069.663
    },
    {
      "epoch": 0.076417766528054,
      "grad_norm": 0.37416523694992065,
      "learning_rate": 0.00013689482470784643,
      "loss": 0.9155,
      "num_input_tokens_seen": 2644842,
      "step": 951,
      "train_runtime": 2472.3448,
      "train_tokens_per_second": 1069.771
    },
    {
      "epoch": 0.07649812169790474,
      "grad_norm": 0.32983753085136414,
      "learning_rate": 0.0001368280467445743,
      "loss": 1.0483,
      "num_input_tokens_seen": 2648056,
      "step": 952,
      "train_runtime": 2474.7585,
      "train_tokens_per_second": 1070.026
    },
    {
      "epoch": 0.07657847686775549,
      "grad_norm": 0.30027881264686584,
      "learning_rate": 0.00013676126878130217,
      "loss": 1.0072,
      "num_input_tokens_seen": 2651264,
      "step": 953,
      "train_runtime": 2477.2327,
      "train_tokens_per_second": 1070.252
    },
    {
      "epoch": 0.07665883203760622,
      "grad_norm": 0.2947307825088501,
      "learning_rate": 0.00013669449081803005,
      "loss": 0.8835,
      "num_input_tokens_seen": 2653084,
      "step": 954,
      "train_runtime": 2479.0202,
      "train_tokens_per_second": 1070.215
    },
    {
      "epoch": 0.07673918720745695,
      "grad_norm": 0.3923710584640503,
      "learning_rate": 0.00013662771285475792,
      "loss": 1.1733,
      "num_input_tokens_seen": 2655462,
      "step": 955,
      "train_runtime": 2480.9973,
      "train_tokens_per_second": 1070.32
    },
    {
      "epoch": 0.0768195423773077,
      "grad_norm": 0.3278249204158783,
      "learning_rate": 0.00013656093489148582,
      "loss": 0.8441,
      "num_input_tokens_seen": 2660804,
      "step": 956,
      "train_runtime": 2484.6748,
      "train_tokens_per_second": 1070.886
    },
    {
      "epoch": 0.07689989754715844,
      "grad_norm": 0.3549589514732361,
      "learning_rate": 0.0001364941569282137,
      "loss": 1.2839,
      "num_input_tokens_seen": 2663256,
      "step": 957,
      "train_runtime": 2486.7036,
      "train_tokens_per_second": 1070.999
    },
    {
      "epoch": 0.07698025271700919,
      "grad_norm": 0.2980327010154724,
      "learning_rate": 0.00013642737896494157,
      "loss": 0.885,
      "num_input_tokens_seen": 2665716,
      "step": 958,
      "train_runtime": 2488.7584,
      "train_tokens_per_second": 1071.103
    },
    {
      "epoch": 0.07706060788685992,
      "grad_norm": 0.26402148604393005,
      "learning_rate": 0.00013636060100166944,
      "loss": 0.8327,
      "num_input_tokens_seen": 2671606,
      "step": 959,
      "train_runtime": 2492.6877,
      "train_tokens_per_second": 1071.777
    },
    {
      "epoch": 0.07714096305671066,
      "grad_norm": 0.2681846618652344,
      "learning_rate": 0.00013629382303839734,
      "loss": 1.0438,
      "num_input_tokens_seen": 2674746,
      "step": 960,
      "train_runtime": 2495.1259,
      "train_tokens_per_second": 1071.988
    },
    {
      "epoch": 0.0772213182265614,
      "grad_norm": 0.2988883852958679,
      "learning_rate": 0.00013622704507512521,
      "loss": 0.9291,
      "num_input_tokens_seen": 2677738,
      "step": 961,
      "train_runtime": 2499.5586,
      "train_tokens_per_second": 1071.284
    },
    {
      "epoch": 0.07730167339641214,
      "grad_norm": 0.32819023728370667,
      "learning_rate": 0.00013616026711185311,
      "loss": 1.003,
      "num_input_tokens_seen": 2680480,
      "step": 962,
      "train_runtime": 2501.7862,
      "train_tokens_per_second": 1071.427
    },
    {
      "epoch": 0.07738202856626288,
      "grad_norm": 0.33189529180526733,
      "learning_rate": 0.000136093489148581,
      "loss": 1.0143,
      "num_input_tokens_seen": 2683808,
      "step": 963,
      "train_runtime": 2504.2769,
      "train_tokens_per_second": 1071.69
    },
    {
      "epoch": 0.07746238373611362,
      "grad_norm": 0.410520076751709,
      "learning_rate": 0.00013602671118530886,
      "loss": 0.9478,
      "num_input_tokens_seen": 2684988,
      "step": 964,
      "train_runtime": 2505.7765,
      "train_tokens_per_second": 1071.519
    },
    {
      "epoch": 0.07754273890596436,
      "grad_norm": 0.3094470798969269,
      "learning_rate": 0.00013595993322203673,
      "loss": 0.9913,
      "num_input_tokens_seen": 2687546,
      "step": 965,
      "train_runtime": 2507.8293,
      "train_tokens_per_second": 1071.662
    },
    {
      "epoch": 0.07762309407581511,
      "grad_norm": 0.29441124200820923,
      "learning_rate": 0.0001358931552587646,
      "loss": 1.0393,
      "num_input_tokens_seen": 2689618,
      "step": 966,
      "train_runtime": 2509.7037,
      "train_tokens_per_second": 1071.687
    },
    {
      "epoch": 0.07770344924566584,
      "grad_norm": 0.3093130886554718,
      "learning_rate": 0.0001358263772954925,
      "loss": 0.9381,
      "num_input_tokens_seen": 2692652,
      "step": 967,
      "train_runtime": 2512.1247,
      "train_tokens_per_second": 1071.862
    },
    {
      "epoch": 0.07778380441551658,
      "grad_norm": 0.28437837958335876,
      "learning_rate": 0.00013575959933222038,
      "loss": 1.3758,
      "num_input_tokens_seen": 2696322,
      "step": 968,
      "train_runtime": 2514.8131,
      "train_tokens_per_second": 1072.176
    },
    {
      "epoch": 0.07786415958536733,
      "grad_norm": 0.2774305045604706,
      "learning_rate": 0.00013569282136894825,
      "loss": 1.0696,
      "num_input_tokens_seen": 2698352,
      "step": 969,
      "train_runtime": 2516.584,
      "train_tokens_per_second": 1072.228
    },
    {
      "epoch": 0.07794451475521806,
      "grad_norm": 0.26283907890319824,
      "learning_rate": 0.00013562604340567613,
      "loss": 0.8386,
      "num_input_tokens_seen": 2700488,
      "step": 970,
      "train_runtime": 2518.421,
      "train_tokens_per_second": 1072.294
    },
    {
      "epoch": 0.07802486992506881,
      "grad_norm": 0.2612572908401489,
      "learning_rate": 0.000135559265442404,
      "loss": 0.9145,
      "num_input_tokens_seen": 2703334,
      "step": 971,
      "train_runtime": 2520.606,
      "train_tokens_per_second": 1072.494
    },
    {
      "epoch": 0.07810522509491954,
      "grad_norm": 0.28983965516090393,
      "learning_rate": 0.0001354924874791319,
      "loss": 0.9132,
      "num_input_tokens_seen": 2705586,
      "step": 972,
      "train_runtime": 2522.4971,
      "train_tokens_per_second": 1072.582
    },
    {
      "epoch": 0.07818558026477028,
      "grad_norm": 0.3406868278980255,
      "learning_rate": 0.00013542570951585977,
      "loss": 0.9522,
      "num_input_tokens_seen": 2707800,
      "step": 973,
      "train_runtime": 2524.4546,
      "train_tokens_per_second": 1072.628
    },
    {
      "epoch": 0.07826593543462103,
      "grad_norm": 0.2974483370780945,
      "learning_rate": 0.00013535893155258765,
      "loss": 0.909,
      "num_input_tokens_seen": 2710136,
      "step": 974,
      "train_runtime": 2526.4256,
      "train_tokens_per_second": 1072.716
    },
    {
      "epoch": 0.07834629060447176,
      "grad_norm": 0.3680464029312134,
      "learning_rate": 0.00013529215358931552,
      "loss": 1.008,
      "num_input_tokens_seen": 2711794,
      "step": 975,
      "train_runtime": 2528.0481,
      "train_tokens_per_second": 1072.683
    },
    {
      "epoch": 0.07842664577432251,
      "grad_norm": 0.33887895941734314,
      "learning_rate": 0.0001352253756260434,
      "loss": 0.8406,
      "num_input_tokens_seen": 2715442,
      "step": 976,
      "train_runtime": 2530.7166,
      "train_tokens_per_second": 1072.993
    },
    {
      "epoch": 0.07850700094417325,
      "grad_norm": 0.2796920835971832,
      "learning_rate": 0.0001351585976627713,
      "loss": 0.8131,
      "num_input_tokens_seen": 2717936,
      "step": 977,
      "train_runtime": 2532.7658,
      "train_tokens_per_second": 1073.11
    },
    {
      "epoch": 0.07858735611402398,
      "grad_norm": 0.3388163447380066,
      "learning_rate": 0.00013509181969949917,
      "loss": 0.8993,
      "num_input_tokens_seen": 2720616,
      "step": 978,
      "train_runtime": 2535.0221,
      "train_tokens_per_second": 1073.212
    },
    {
      "epoch": 0.07866771128387473,
      "grad_norm": 0.2970505654811859,
      "learning_rate": 0.00013502504173622707,
      "loss": 0.8783,
      "num_input_tokens_seen": 2725144,
      "step": 979,
      "train_runtime": 2538.1962,
      "train_tokens_per_second": 1073.654
    },
    {
      "epoch": 0.07874806645372547,
      "grad_norm": 0.43267375230789185,
      "learning_rate": 0.00013495826377295494,
      "loss": 0.9771,
      "num_input_tokens_seen": 2727974,
      "step": 980,
      "train_runtime": 2540.4159,
      "train_tokens_per_second": 1073.83
    },
    {
      "epoch": 0.0788284216235762,
      "grad_norm": 0.2954391837120056,
      "learning_rate": 0.0001348914858096828,
      "loss": 1.058,
      "num_input_tokens_seen": 2732958,
      "step": 981,
      "train_runtime": 2543.9824,
      "train_tokens_per_second": 1074.283
    },
    {
      "epoch": 0.07890877679342695,
      "grad_norm": 0.39759600162506104,
      "learning_rate": 0.00013482470784641069,
      "loss": 1.1552,
      "num_input_tokens_seen": 2735778,
      "step": 982,
      "train_runtime": 2546.1563,
      "train_tokens_per_second": 1074.474
    },
    {
      "epoch": 0.07898913196327768,
      "grad_norm": 0.37654632329940796,
      "learning_rate": 0.00013475792988313859,
      "loss": 0.9169,
      "num_input_tokens_seen": 2738094,
      "step": 983,
      "train_runtime": 2548.0408,
      "train_tokens_per_second": 1074.588
    },
    {
      "epoch": 0.07906948713312843,
      "grad_norm": 0.40086641907691956,
      "learning_rate": 0.00013469115191986646,
      "loss": 1.1409,
      "num_input_tokens_seen": 2741050,
      "step": 984,
      "train_runtime": 2550.365,
      "train_tokens_per_second": 1074.768
    },
    {
      "epoch": 0.07914984230297917,
      "grad_norm": 0.27053409814834595,
      "learning_rate": 0.00013462437395659433,
      "loss": 0.8318,
      "num_input_tokens_seen": 2745116,
      "step": 985,
      "train_runtime": 2553.2183,
      "train_tokens_per_second": 1075.159
    },
    {
      "epoch": 0.0792301974728299,
      "grad_norm": 0.28262272477149963,
      "learning_rate": 0.0001345575959933222,
      "loss": 1.2787,
      "num_input_tokens_seen": 2748234,
      "step": 986,
      "train_runtime": 2555.6442,
      "train_tokens_per_second": 1075.359
    },
    {
      "epoch": 0.07931055264268065,
      "grad_norm": 0.39394891262054443,
      "learning_rate": 0.00013449081803005008,
      "loss": 0.9399,
      "num_input_tokens_seen": 2749862,
      "step": 987,
      "train_runtime": 2557.3673,
      "train_tokens_per_second": 1075.271
    },
    {
      "epoch": 0.07939090781253139,
      "grad_norm": 0.2508150637149811,
      "learning_rate": 0.00013442404006677798,
      "loss": 0.9904,
      "num_input_tokens_seen": 2752260,
      "step": 988,
      "train_runtime": 2559.3405,
      "train_tokens_per_second": 1075.379
    },
    {
      "epoch": 0.07947126298238213,
      "grad_norm": 0.43806296586990356,
      "learning_rate": 0.00013435726210350585,
      "loss": 1.0792,
      "num_input_tokens_seen": 2754188,
      "step": 989,
      "train_runtime": 2561.1655,
      "train_tokens_per_second": 1075.365
    },
    {
      "epoch": 0.07955161815223287,
      "grad_norm": 0.3231634497642517,
      "learning_rate": 0.00013429048414023373,
      "loss": 0.8406,
      "num_input_tokens_seen": 2756030,
      "step": 990,
      "train_runtime": 2562.9015,
      "train_tokens_per_second": 1075.355
    },
    {
      "epoch": 0.0796319733220836,
      "grad_norm": 0.35421857237815857,
      "learning_rate": 0.0001342237061769616,
      "loss": 0.8888,
      "num_input_tokens_seen": 2757438,
      "step": 991,
      "train_runtime": 2566.5337,
      "train_tokens_per_second": 1074.382
    },
    {
      "epoch": 0.07971232849193435,
      "grad_norm": 0.3755453824996948,
      "learning_rate": 0.00013415692821368947,
      "loss": 0.7254,
      "num_input_tokens_seen": 2759520,
      "step": 992,
      "train_runtime": 2568.2994,
      "train_tokens_per_second": 1074.454
    },
    {
      "epoch": 0.07979268366178509,
      "grad_norm": 0.33065688610076904,
      "learning_rate": 0.00013409015025041737,
      "loss": 0.9412,
      "num_input_tokens_seen": 2762334,
      "step": 993,
      "train_runtime": 2570.5521,
      "train_tokens_per_second": 1074.607
    },
    {
      "epoch": 0.07987303883163584,
      "grad_norm": 0.2998780906200409,
      "learning_rate": 0.00013402337228714524,
      "loss": 0.7833,
      "num_input_tokens_seen": 2764744,
      "step": 994,
      "train_runtime": 2572.6205,
      "train_tokens_per_second": 1074.68
    },
    {
      "epoch": 0.07995339400148657,
      "grad_norm": 0.32594361901283264,
      "learning_rate": 0.00013395659432387312,
      "loss": 0.8737,
      "num_input_tokens_seen": 2766820,
      "step": 995,
      "train_runtime": 2574.3804,
      "train_tokens_per_second": 1074.752
    },
    {
      "epoch": 0.0800337491713373,
      "grad_norm": 0.3067326545715332,
      "learning_rate": 0.00013388981636060102,
      "loss": 1.0962,
      "num_input_tokens_seen": 2769598,
      "step": 996,
      "train_runtime": 2576.5668,
      "train_tokens_per_second": 1074.918
    },
    {
      "epoch": 0.08011410434118806,
      "grad_norm": 0.3515946865081787,
      "learning_rate": 0.0001338230383973289,
      "loss": 0.8304,
      "num_input_tokens_seen": 2771802,
      "step": 997,
      "train_runtime": 2578.6034,
      "train_tokens_per_second": 1074.924
    },
    {
      "epoch": 0.08019445951103879,
      "grad_norm": 0.2858158349990845,
      "learning_rate": 0.00013375626043405676,
      "loss": 1.0681,
      "num_input_tokens_seen": 2774884,
      "step": 998,
      "train_runtime": 2580.9382,
      "train_tokens_per_second": 1075.145
    },
    {
      "epoch": 0.08027481468088953,
      "grad_norm": 0.6481911540031433,
      "learning_rate": 0.00013368948247078466,
      "loss": 0.974,
      "num_input_tokens_seen": 2776190,
      "step": 999,
      "train_runtime": 2582.5019,
      "train_tokens_per_second": 1075.0
    },
    {
      "epoch": 0.08035516985074027,
      "grad_norm": 0.34825989603996277,
      "learning_rate": 0.00013362270450751254,
      "loss": 1.092,
      "num_input_tokens_seen": 2781156,
      "step": 1000,
      "train_runtime": 2586.0123,
      "train_tokens_per_second": 1075.461
    },
    {
      "epoch": 0.08043552502059101,
      "grad_norm": 0.34092801809310913,
      "learning_rate": 0.0001335559265442404,
      "loss": 1.0392,
      "num_input_tokens_seen": 2783550,
      "step": 1001,
      "train_runtime": 2588.0704,
      "train_tokens_per_second": 1075.531
    },
    {
      "epoch": 0.08051588019044176,
      "grad_norm": 0.3506810963153839,
      "learning_rate": 0.00013348914858096828,
      "loss": 0.8791,
      "num_input_tokens_seen": 2786234,
      "step": 1002,
      "train_runtime": 2590.2442,
      "train_tokens_per_second": 1075.665
    },
    {
      "epoch": 0.08059623536029249,
      "grad_norm": 0.3020326793193817,
      "learning_rate": 0.00013342237061769616,
      "loss": 1.0414,
      "num_input_tokens_seen": 2788480,
      "step": 1003,
      "train_runtime": 2592.1898,
      "train_tokens_per_second": 1075.724
    },
    {
      "epoch": 0.08067659053014323,
      "grad_norm": 0.3670845627784729,
      "learning_rate": 0.00013335559265442406,
      "loss": 1.2268,
      "num_input_tokens_seen": 2791858,
      "step": 1004,
      "train_runtime": 2594.8598,
      "train_tokens_per_second": 1075.919
    },
    {
      "epoch": 0.08075694569999398,
      "grad_norm": 0.29267269372940063,
      "learning_rate": 0.00013328881469115193,
      "loss": 0.8361,
      "num_input_tokens_seen": 2794236,
      "step": 1005,
      "train_runtime": 2596.8873,
      "train_tokens_per_second": 1075.994
    },
    {
      "epoch": 0.08083730086984471,
      "grad_norm": 0.3259250819683075,
      "learning_rate": 0.0001332220367278798,
      "loss": 1.2422,
      "num_input_tokens_seen": 2796898,
      "step": 1006,
      "train_runtime": 2599.064,
      "train_tokens_per_second": 1076.117
    },
    {
      "epoch": 0.08091765603969546,
      "grad_norm": 0.35236838459968567,
      "learning_rate": 0.00013315525876460768,
      "loss": 0.932,
      "num_input_tokens_seen": 2799064,
      "step": 1007,
      "train_runtime": 2600.9463,
      "train_tokens_per_second": 1076.171
    },
    {
      "epoch": 0.0809980112095462,
      "grad_norm": 0.3951377868652344,
      "learning_rate": 0.00013308848080133555,
      "loss": 0.8327,
      "num_input_tokens_seen": 2803022,
      "step": 1008,
      "train_runtime": 2603.9105,
      "train_tokens_per_second": 1076.466
    },
    {
      "epoch": 0.08107836637939693,
      "grad_norm": 0.4352874755859375,
      "learning_rate": 0.00013302170283806345,
      "loss": 1.0081,
      "num_input_tokens_seen": 2804650,
      "step": 1009,
      "train_runtime": 2605.5357,
      "train_tokens_per_second": 1076.42
    },
    {
      "epoch": 0.08115872154924768,
      "grad_norm": 0.31826362013816833,
      "learning_rate": 0.00013295492487479132,
      "loss": 1.2703,
      "num_input_tokens_seen": 2808310,
      "step": 1010,
      "train_runtime": 2608.2061,
      "train_tokens_per_second": 1076.721
    },
    {
      "epoch": 0.08123907671909841,
      "grad_norm": 0.35901400446891785,
      "learning_rate": 0.0001328881469115192,
      "loss": 0.898,
      "num_input_tokens_seen": 2810310,
      "step": 1011,
      "train_runtime": 2610.0179,
      "train_tokens_per_second": 1076.74
    },
    {
      "epoch": 0.08131943188894916,
      "grad_norm": 0.2934917211532593,
      "learning_rate": 0.00013282136894824707,
      "loss": 0.8468,
      "num_input_tokens_seen": 2812368,
      "step": 1012,
      "train_runtime": 2611.8821,
      "train_tokens_per_second": 1076.759
    },
    {
      "epoch": 0.0813997870587999,
      "grad_norm": 0.4374319911003113,
      "learning_rate": 0.00013275459098497497,
      "loss": 1.1719,
      "num_input_tokens_seen": 2813738,
      "step": 1013,
      "train_runtime": 2613.4076,
      "train_tokens_per_second": 1076.655
    },
    {
      "epoch": 0.08148014222865063,
      "grad_norm": 0.28275415301322937,
      "learning_rate": 0.00013268781302170284,
      "loss": 0.8692,
      "num_input_tokens_seen": 2815598,
      "step": 1014,
      "train_runtime": 2615.0373,
      "train_tokens_per_second": 1076.695
    },
    {
      "epoch": 0.08156049739850138,
      "grad_norm": 0.21789546310901642,
      "learning_rate": 0.00013262103505843074,
      "loss": 0.7414,
      "num_input_tokens_seen": 2817622,
      "step": 1015,
      "train_runtime": 2616.8942,
      "train_tokens_per_second": 1076.705
    },
    {
      "epoch": 0.08164085256835212,
      "grad_norm": 0.4071328639984131,
      "learning_rate": 0.00013255425709515862,
      "loss": 0.9186,
      "num_input_tokens_seen": 2821024,
      "step": 1016,
      "train_runtime": 2619.4291,
      "train_tokens_per_second": 1076.961
    },
    {
      "epoch": 0.08172120773820285,
      "grad_norm": 0.26125869154930115,
      "learning_rate": 0.0001324874791318865,
      "loss": 0.7718,
      "num_input_tokens_seen": 2823164,
      "step": 1017,
      "train_runtime": 2621.2648,
      "train_tokens_per_second": 1077.024
    },
    {
      "epoch": 0.0818015629080536,
      "grad_norm": 0.3403930962085724,
      "learning_rate": 0.00013242070116861436,
      "loss": 1.052,
      "num_input_tokens_seen": 2826500,
      "step": 1018,
      "train_runtime": 2623.7693,
      "train_tokens_per_second": 1077.267
    },
    {
      "epoch": 0.08188191807790433,
      "grad_norm": 0.2818782329559326,
      "learning_rate": 0.00013235392320534224,
      "loss": 1.0326,
      "num_input_tokens_seen": 2829252,
      "step": 1019,
      "train_runtime": 2626.0117,
      "train_tokens_per_second": 1077.395
    },
    {
      "epoch": 0.08196227324775508,
      "grad_norm": 0.30011674761772156,
      "learning_rate": 0.00013228714524207014,
      "loss": 1.2725,
      "num_input_tokens_seen": 2831818,
      "step": 1020,
      "train_runtime": 2628.1214,
      "train_tokens_per_second": 1077.507
    },
    {
      "epoch": 0.08204262841760582,
      "grad_norm": 0.27111268043518066,
      "learning_rate": 0.000132220367278798,
      "loss": 0.9214,
      "num_input_tokens_seen": 2837286,
      "step": 1021,
      "train_runtime": 2633.9711,
      "train_tokens_per_second": 1077.189
    },
    {
      "epoch": 0.08212298358745655,
      "grad_norm": 0.2721482515335083,
      "learning_rate": 0.00013215358931552588,
      "loss": 1.0397,
      "num_input_tokens_seen": 2839606,
      "step": 1022,
      "train_runtime": 2635.9691,
      "train_tokens_per_second": 1077.253
    },
    {
      "epoch": 0.0822033387573073,
      "grad_norm": 0.27940985560417175,
      "learning_rate": 0.00013208681135225376,
      "loss": 0.8657,
      "num_input_tokens_seen": 2841590,
      "step": 1023,
      "train_runtime": 2637.8243,
      "train_tokens_per_second": 1077.248
    },
    {
      "epoch": 0.08228369392715804,
      "grad_norm": 0.332803875207901,
      "learning_rate": 0.00013202003338898163,
      "loss": 0.6836,
      "num_input_tokens_seen": 2846138,
      "step": 1024,
      "train_runtime": 2641.0347,
      "train_tokens_per_second": 1077.66
    },
    {
      "epoch": 0.08236404909700878,
      "grad_norm": 0.29152676463127136,
      "learning_rate": 0.00013195325542570953,
      "loss": 0.8369,
      "num_input_tokens_seen": 2848412,
      "step": 1025,
      "train_runtime": 2642.9532,
      "train_tokens_per_second": 1077.738
    },
    {
      "epoch": 0.08244440426685952,
      "grad_norm": 0.37724512815475464,
      "learning_rate": 0.0001318864774624374,
      "loss": 0.8647,
      "num_input_tokens_seen": 2849874,
      "step": 1026,
      "train_runtime": 2644.5686,
      "train_tokens_per_second": 1077.633
    },
    {
      "epoch": 0.08252475943671025,
      "grad_norm": 0.33222779631614685,
      "learning_rate": 0.00013181969949916528,
      "loss": 1.2321,
      "num_input_tokens_seen": 2852100,
      "step": 1027,
      "train_runtime": 2646.5123,
      "train_tokens_per_second": 1077.682
    },
    {
      "epoch": 0.082605114606561,
      "grad_norm": 0.31807512044906616,
      "learning_rate": 0.00013175292153589315,
      "loss": 0.8616,
      "num_input_tokens_seen": 2854402,
      "step": 1028,
      "train_runtime": 2648.4239,
      "train_tokens_per_second": 1077.774
    },
    {
      "epoch": 0.08268546977641174,
      "grad_norm": 0.30015531182289124,
      "learning_rate": 0.00013168614357262102,
      "loss": 1.0101,
      "num_input_tokens_seen": 2856872,
      "step": 1029,
      "train_runtime": 2650.4748,
      "train_tokens_per_second": 1077.872
    },
    {
      "epoch": 0.08276582494626249,
      "grad_norm": 0.34851160645484924,
      "learning_rate": 0.00013161936560934892,
      "loss": 1.0764,
      "num_input_tokens_seen": 2858760,
      "step": 1030,
      "train_runtime": 2652.2334,
      "train_tokens_per_second": 1077.869
    },
    {
      "epoch": 0.08284618011611322,
      "grad_norm": 0.3319004774093628,
      "learning_rate": 0.0001315525876460768,
      "loss": 1.1199,
      "num_input_tokens_seen": 2860558,
      "step": 1031,
      "train_runtime": 2653.9857,
      "train_tokens_per_second": 1077.835
    },
    {
      "epoch": 0.08292653528596396,
      "grad_norm": 0.3031421899795532,
      "learning_rate": 0.0001314858096828047,
      "loss": 0.8352,
      "num_input_tokens_seen": 2864206,
      "step": 1032,
      "train_runtime": 2656.7358,
      "train_tokens_per_second": 1078.092
    },
    {
      "epoch": 0.0830068904558147,
      "grad_norm": 0.2702726125717163,
      "learning_rate": 0.00013141903171953257,
      "loss": 1.04,
      "num_input_tokens_seen": 2867676,
      "step": 1033,
      "train_runtime": 2659.2742,
      "train_tokens_per_second": 1078.368
    },
    {
      "epoch": 0.08308724562566544,
      "grad_norm": 0.30745938420295715,
      "learning_rate": 0.00013135225375626044,
      "loss": 1.1677,
      "num_input_tokens_seen": 2870066,
      "step": 1034,
      "train_runtime": 2661.2829,
      "train_tokens_per_second": 1078.452
    },
    {
      "epoch": 0.08316760079551618,
      "grad_norm": 0.3097253143787384,
      "learning_rate": 0.00013128547579298832,
      "loss": 0.7924,
      "num_input_tokens_seen": 2873278,
      "step": 1035,
      "train_runtime": 2663.7629,
      "train_tokens_per_second": 1078.654
    },
    {
      "epoch": 0.08324795596536692,
      "grad_norm": 0.31835857033729553,
      "learning_rate": 0.00013121869782971622,
      "loss": 1.0011,
      "num_input_tokens_seen": 2874902,
      "step": 1036,
      "train_runtime": 2665.4132,
      "train_tokens_per_second": 1078.595
    },
    {
      "epoch": 0.08332831113521766,
      "grad_norm": 0.2660655379295349,
      "learning_rate": 0.0001311519198664441,
      "loss": 0.8712,
      "num_input_tokens_seen": 2877546,
      "step": 1037,
      "train_runtime": 2667.4818,
      "train_tokens_per_second": 1078.75
    },
    {
      "epoch": 0.08340866630506841,
      "grad_norm": 0.31649038195610046,
      "learning_rate": 0.00013108514190317196,
      "loss": 0.5954,
      "num_input_tokens_seen": 2879968,
      "step": 1038,
      "train_runtime": 2669.5734,
      "train_tokens_per_second": 1078.812
    },
    {
      "epoch": 0.08348902147491914,
      "grad_norm": 0.27658918499946594,
      "learning_rate": 0.00013101836393989983,
      "loss": 0.7088,
      "num_input_tokens_seen": 2881680,
      "step": 1039,
      "train_runtime": 2671.244,
      "train_tokens_per_second": 1078.778
    },
    {
      "epoch": 0.08356937664476988,
      "grad_norm": 0.2922908365726471,
      "learning_rate": 0.0001309515859766277,
      "loss": 1.2907,
      "num_input_tokens_seen": 2884004,
      "step": 1040,
      "train_runtime": 2673.2697,
      "train_tokens_per_second": 1078.83
    },
    {
      "epoch": 0.08364973181462063,
      "grad_norm": 0.23947563767433167,
      "learning_rate": 0.0001308848080133556,
      "loss": 0.7296,
      "num_input_tokens_seen": 2886704,
      "step": 1041,
      "train_runtime": 2675.4457,
      "train_tokens_per_second": 1078.962
    },
    {
      "epoch": 0.08373008698447136,
      "grad_norm": 0.27451804280281067,
      "learning_rate": 0.00013081803005008348,
      "loss": 1.0104,
      "num_input_tokens_seen": 2888982,
      "step": 1042,
      "train_runtime": 2677.5056,
      "train_tokens_per_second": 1078.983
    },
    {
      "epoch": 0.08381044215432211,
      "grad_norm": 0.2693738341331482,
      "learning_rate": 0.00013075125208681135,
      "loss": 1.0469,
      "num_input_tokens_seen": 2893346,
      "step": 1043,
      "train_runtime": 2680.46,
      "train_tokens_per_second": 1079.421
    },
    {
      "epoch": 0.08389079732417284,
      "grad_norm": 0.29715558886528015,
      "learning_rate": 0.00013068447412353923,
      "loss": 1.0376,
      "num_input_tokens_seen": 2895484,
      "step": 1044,
      "train_runtime": 2682.423,
      "train_tokens_per_second": 1079.429
    },
    {
      "epoch": 0.08397115249402358,
      "grad_norm": 0.3529125452041626,
      "learning_rate": 0.0001306176961602671,
      "loss": 1.3031,
      "num_input_tokens_seen": 2898552,
      "step": 1045,
      "train_runtime": 2684.8308,
      "train_tokens_per_second": 1079.603
    },
    {
      "epoch": 0.08405150766387433,
      "grad_norm": 0.34995678067207336,
      "learning_rate": 0.00013055091819699497,
      "loss": 0.8914,
      "num_input_tokens_seen": 2901326,
      "step": 1046,
      "train_runtime": 2687.041,
      "train_tokens_per_second": 1079.748
    },
    {
      "epoch": 0.08413186283372506,
      "grad_norm": 0.272154837846756,
      "learning_rate": 0.00013048414023372287,
      "loss": 0.8414,
      "num_input_tokens_seen": 2903690,
      "step": 1047,
      "train_runtime": 2689.0487,
      "train_tokens_per_second": 1079.821
    },
    {
      "epoch": 0.08421221800357581,
      "grad_norm": 0.3116409182548523,
      "learning_rate": 0.00013041736227045075,
      "loss": 1.2821,
      "num_input_tokens_seen": 2906882,
      "step": 1048,
      "train_runtime": 2691.4399,
      "train_tokens_per_second": 1080.047
    },
    {
      "epoch": 0.08429257317342655,
      "grad_norm": 0.2630103528499603,
      "learning_rate": 0.00013035058430717865,
      "loss": 0.9734,
      "num_input_tokens_seen": 2909530,
      "step": 1049,
      "train_runtime": 2693.5598,
      "train_tokens_per_second": 1080.18
    },
    {
      "epoch": 0.08437292834327728,
      "grad_norm": 0.26427966356277466,
      "learning_rate": 0.00013028380634390652,
      "loss": 1.2673,
      "num_input_tokens_seen": 2911538,
      "step": 1050,
      "train_runtime": 2695.3087,
      "train_tokens_per_second": 1080.224
    },
    {
      "epoch": 0.08445328351312803,
      "grad_norm": 0.2952263653278351,
      "learning_rate": 0.0001302170283806344,
      "loss": 1.144,
      "num_input_tokens_seen": 2913298,
      "step": 1051,
      "train_runtime": 2699.0488,
      "train_tokens_per_second": 1079.38
    },
    {
      "epoch": 0.08453363868297877,
      "grad_norm": 0.36667385697364807,
      "learning_rate": 0.0001301502504173623,
      "loss": 0.9793,
      "num_input_tokens_seen": 2919372,
      "step": 1052,
      "train_runtime": 2703.297,
      "train_tokens_per_second": 1079.93
    },
    {
      "epoch": 0.0846139938528295,
      "grad_norm": 0.33305513858795166,
      "learning_rate": 0.00013008347245409017,
      "loss": 1.3704,
      "num_input_tokens_seen": 2922472,
      "step": 1053,
      "train_runtime": 2705.785,
      "train_tokens_per_second": 1080.083
    },
    {
      "epoch": 0.08469434902268025,
      "grad_norm": 0.38013726472854614,
      "learning_rate": 0.00013001669449081804,
      "loss": 0.9353,
      "num_input_tokens_seen": 2924566,
      "step": 1054,
      "train_runtime": 2707.7092,
      "train_tokens_per_second": 1080.089
    },
    {
      "epoch": 0.08477470419253098,
      "grad_norm": 0.28573718667030334,
      "learning_rate": 0.00012994991652754591,
      "loss": 0.9211,
      "num_input_tokens_seen": 2927362,
      "step": 1055,
      "train_runtime": 2709.8846,
      "train_tokens_per_second": 1080.253
    },
    {
      "epoch": 0.08485505936238173,
      "grad_norm": 0.31847864389419556,
      "learning_rate": 0.0001298831385642738,
      "loss": 0.9339,
      "num_input_tokens_seen": 2930960,
      "step": 1056,
      "train_runtime": 2712.6034,
      "train_tokens_per_second": 1080.497
    },
    {
      "epoch": 0.08493541453223247,
      "grad_norm": 0.44332391023635864,
      "learning_rate": 0.0001298163606010017,
      "loss": 1.288,
      "num_input_tokens_seen": 2933888,
      "step": 1057,
      "train_runtime": 2714.9189,
      "train_tokens_per_second": 1080.654
    },
    {
      "epoch": 0.0850157697020832,
      "grad_norm": 0.350253701210022,
      "learning_rate": 0.00012974958263772956,
      "loss": 0.9685,
      "num_input_tokens_seen": 2937338,
      "step": 1058,
      "train_runtime": 2717.4842,
      "train_tokens_per_second": 1080.903
    },
    {
      "epoch": 0.08509612487193395,
      "grad_norm": 0.33024874329566956,
      "learning_rate": 0.00012968280467445743,
      "loss": 0.7487,
      "num_input_tokens_seen": 2938738,
      "step": 1059,
      "train_runtime": 2719.0708,
      "train_tokens_per_second": 1080.788
    },
    {
      "epoch": 0.08517648004178469,
      "grad_norm": 0.30797895789146423,
      "learning_rate": 0.0001296160267111853,
      "loss": 1.0426,
      "num_input_tokens_seen": 2942108,
      "step": 1060,
      "train_runtime": 2721.5573,
      "train_tokens_per_second": 1081.038
    },
    {
      "epoch": 0.08525683521163543,
      "grad_norm": 0.33872172236442566,
      "learning_rate": 0.00012954924874791318,
      "loss": 0.9575,
      "num_input_tokens_seen": 2945742,
      "step": 1061,
      "train_runtime": 2724.2489,
      "train_tokens_per_second": 1081.304
    },
    {
      "epoch": 0.08533719038148617,
      "grad_norm": 0.38300707936286926,
      "learning_rate": 0.00012948247078464108,
      "loss": 1.0997,
      "num_input_tokens_seen": 2948428,
      "step": 1062,
      "train_runtime": 2726.5297,
      "train_tokens_per_second": 1081.385
    },
    {
      "epoch": 0.0854175455513369,
      "grad_norm": 0.34665048122406006,
      "learning_rate": 0.00012941569282136895,
      "loss": 0.9976,
      "num_input_tokens_seen": 2949936,
      "step": 1063,
      "train_runtime": 2728.1047,
      "train_tokens_per_second": 1081.313
    },
    {
      "epoch": 0.08549790072118765,
      "grad_norm": 0.4175337255001068,
      "learning_rate": 0.00012934891485809683,
      "loss": 0.9155,
      "num_input_tokens_seen": 2952068,
      "step": 1064,
      "train_runtime": 2729.9211,
      "train_tokens_per_second": 1081.375
    },
    {
      "epoch": 0.08557825589103839,
      "grad_norm": 0.29677116870880127,
      "learning_rate": 0.0001292821368948247,
      "loss": 1.0279,
      "num_input_tokens_seen": 2957814,
      "step": 1065,
      "train_runtime": 2733.866,
      "train_tokens_per_second": 1081.916
    },
    {
      "epoch": 0.08565861106088914,
      "grad_norm": 0.27838361263275146,
      "learning_rate": 0.0001292153589315526,
      "loss": 0.8967,
      "num_input_tokens_seen": 2961052,
      "step": 1066,
      "train_runtime": 2736.39,
      "train_tokens_per_second": 1082.102
    },
    {
      "epoch": 0.08573896623073987,
      "grad_norm": 0.3175829350948334,
      "learning_rate": 0.00012914858096828047,
      "loss": 1.0931,
      "num_input_tokens_seen": 2963148,
      "step": 1067,
      "train_runtime": 2738.1507,
      "train_tokens_per_second": 1082.171
    },
    {
      "epoch": 0.0858193214005906,
      "grad_norm": 0.33555886149406433,
      "learning_rate": 0.00012908180300500837,
      "loss": 1.1348,
      "num_input_tokens_seen": 2965132,
      "step": 1068,
      "train_runtime": 2739.9372,
      "train_tokens_per_second": 1082.19
    },
    {
      "epoch": 0.08589967657044136,
      "grad_norm": 0.3252342939376831,
      "learning_rate": 0.00012901502504173625,
      "loss": 1.1991,
      "num_input_tokens_seen": 2968356,
      "step": 1069,
      "train_runtime": 2742.3451,
      "train_tokens_per_second": 1082.415
    },
    {
      "epoch": 0.08598003174029209,
      "grad_norm": 0.34085896611213684,
      "learning_rate": 0.00012894824707846412,
      "loss": 1.0614,
      "num_input_tokens_seen": 2971714,
      "step": 1070,
      "train_runtime": 2744.7657,
      "train_tokens_per_second": 1082.684
    },
    {
      "epoch": 0.08606038691014282,
      "grad_norm": 0.3247336447238922,
      "learning_rate": 0.000128881469115192,
      "loss": 1.2979,
      "num_input_tokens_seen": 2977932,
      "step": 1071,
      "train_runtime": 2748.9803,
      "train_tokens_per_second": 1083.286
    },
    {
      "epoch": 0.08614074207999357,
      "grad_norm": 0.285835862159729,
      "learning_rate": 0.00012881469115191987,
      "loss": 1.0365,
      "num_input_tokens_seen": 2980568,
      "step": 1072,
      "train_runtime": 2751.0834,
      "train_tokens_per_second": 1083.416
    },
    {
      "epoch": 0.08622109724984431,
      "grad_norm": 0.30813121795654297,
      "learning_rate": 0.00012874791318864777,
      "loss": 1.0571,
      "num_input_tokens_seen": 2983094,
      "step": 1073,
      "train_runtime": 2753.1286,
      "train_tokens_per_second": 1083.529
    },
    {
      "epoch": 0.08630145241969506,
      "grad_norm": 0.2712572515010834,
      "learning_rate": 0.00012868113522537564,
      "loss": 0.9591,
      "num_input_tokens_seen": 2985844,
      "step": 1074,
      "train_runtime": 2755.2385,
      "train_tokens_per_second": 1083.697
    },
    {
      "epoch": 0.08638180758954579,
      "grad_norm": 0.255834698677063,
      "learning_rate": 0.0001286143572621035,
      "loss": 0.7757,
      "num_input_tokens_seen": 2989678,
      "step": 1075,
      "train_runtime": 2757.9776,
      "train_tokens_per_second": 1084.011
    },
    {
      "epoch": 0.08646216275939653,
      "grad_norm": 0.3725673258304596,
      "learning_rate": 0.00012854757929883139,
      "loss": 0.892,
      "num_input_tokens_seen": 2993680,
      "step": 1076,
      "train_runtime": 2760.8726,
      "train_tokens_per_second": 1084.324
    },
    {
      "epoch": 0.08654251792924728,
      "grad_norm": 0.41319602727890015,
      "learning_rate": 0.00012848080133555926,
      "loss": 0.882,
      "num_input_tokens_seen": 2995424,
      "step": 1077,
      "train_runtime": 2762.5437,
      "train_tokens_per_second": 1084.299
    },
    {
      "epoch": 0.08662287309909801,
      "grad_norm": 0.3238431215286255,
      "learning_rate": 0.00012841402337228716,
      "loss": 1.0179,
      "num_input_tokens_seen": 2997696,
      "step": 1078,
      "train_runtime": 2764.5115,
      "train_tokens_per_second": 1084.349
    },
    {
      "epoch": 0.08670322826894876,
      "grad_norm": 0.2749386429786682,
      "learning_rate": 0.00012834724540901503,
      "loss": 1.0802,
      "num_input_tokens_seen": 3001614,
      "step": 1079,
      "train_runtime": 2767.2697,
      "train_tokens_per_second": 1084.684
    },
    {
      "epoch": 0.0867835834387995,
      "grad_norm": 0.38312584161758423,
      "learning_rate": 0.0001282804674457429,
      "loss": 0.9611,
      "num_input_tokens_seen": 3003716,
      "step": 1080,
      "train_runtime": 2769.0475,
      "train_tokens_per_second": 1084.747
    },
    {
      "epoch": 0.08686393860865023,
      "grad_norm": 0.29507017135620117,
      "learning_rate": 0.00012821368948247078,
      "loss": 0.8433,
      "num_input_tokens_seen": 3007974,
      "step": 1081,
      "train_runtime": 2773.7463,
      "train_tokens_per_second": 1084.445
    },
    {
      "epoch": 0.08694429377850098,
      "grad_norm": 0.3189148008823395,
      "learning_rate": 0.00012814691151919865,
      "loss": 0.9217,
      "num_input_tokens_seen": 3010416,
      "step": 1082,
      "train_runtime": 2775.8468,
      "train_tokens_per_second": 1084.504
    },
    {
      "epoch": 0.08702464894835171,
      "grad_norm": 0.35395893454551697,
      "learning_rate": 0.00012808013355592655,
      "loss": 1.0264,
      "num_input_tokens_seen": 3013942,
      "step": 1083,
      "train_runtime": 2778.4236,
      "train_tokens_per_second": 1084.767
    },
    {
      "epoch": 0.08710500411820246,
      "grad_norm": 0.3388271927833557,
      "learning_rate": 0.00012801335559265442,
      "loss": 1.0092,
      "num_input_tokens_seen": 3019748,
      "step": 1084,
      "train_runtime": 2782.4385,
      "train_tokens_per_second": 1085.288
    },
    {
      "epoch": 0.0871853592880532,
      "grad_norm": 0.30448269844055176,
      "learning_rate": 0.00012794657762938233,
      "loss": 1.0393,
      "num_input_tokens_seen": 3023376,
      "step": 1085,
      "train_runtime": 2785.1904,
      "train_tokens_per_second": 1085.519
    },
    {
      "epoch": 0.08726571445790393,
      "grad_norm": 0.3304644823074341,
      "learning_rate": 0.0001278797996661102,
      "loss": 1.1707,
      "num_input_tokens_seen": 3025506,
      "step": 1086,
      "train_runtime": 2787.0267,
      "train_tokens_per_second": 1085.568
    },
    {
      "epoch": 0.08734606962775468,
      "grad_norm": 0.3211642801761627,
      "learning_rate": 0.00012781302170283807,
      "loss": 0.755,
      "num_input_tokens_seen": 3029170,
      "step": 1087,
      "train_runtime": 2789.8789,
      "train_tokens_per_second": 1085.771
    },
    {
      "epoch": 0.08742642479760541,
      "grad_norm": 0.35865727066993713,
      "learning_rate": 0.00012774624373956594,
      "loss": 1.0895,
      "num_input_tokens_seen": 3030380,
      "step": 1088,
      "train_runtime": 2791.378,
      "train_tokens_per_second": 1085.622
    },
    {
      "epoch": 0.08750677996745615,
      "grad_norm": 0.40979212522506714,
      "learning_rate": 0.00012767946577629384,
      "loss": 0.7962,
      "num_input_tokens_seen": 3035056,
      "step": 1089,
      "train_runtime": 2794.7372,
      "train_tokens_per_second": 1085.99
    },
    {
      "epoch": 0.0875871351373069,
      "grad_norm": 0.364322304725647,
      "learning_rate": 0.00012761268781302172,
      "loss": 0.6807,
      "num_input_tokens_seen": 3039682,
      "step": 1090,
      "train_runtime": 2797.9686,
      "train_tokens_per_second": 1086.389
    },
    {
      "epoch": 0.08766749030715763,
      "grad_norm": 0.25292372703552246,
      "learning_rate": 0.0001275459098497496,
      "loss": 0.6023,
      "num_input_tokens_seen": 3042474,
      "step": 1091,
      "train_runtime": 2800.2429,
      "train_tokens_per_second": 1086.504
    },
    {
      "epoch": 0.08774784547700838,
      "grad_norm": 0.3244749903678894,
      "learning_rate": 0.00012747913188647746,
      "loss": 0.8558,
      "num_input_tokens_seen": 3045692,
      "step": 1092,
      "train_runtime": 2802.7945,
      "train_tokens_per_second": 1086.663
    },
    {
      "epoch": 0.08782820064685912,
      "grad_norm": 0.3067663013935089,
      "learning_rate": 0.00012741235392320534,
      "loss": 1.2547,
      "num_input_tokens_seen": 3049432,
      "step": 1093,
      "train_runtime": 2805.4878,
      "train_tokens_per_second": 1086.953
    },
    {
      "epoch": 0.08790855581670985,
      "grad_norm": 0.33565008640289307,
      "learning_rate": 0.00012734557595993324,
      "loss": 1.1355,
      "num_input_tokens_seen": 3051074,
      "step": 1094,
      "train_runtime": 2807.1266,
      "train_tokens_per_second": 1086.903
    },
    {
      "epoch": 0.0879889109865606,
      "grad_norm": 0.2515357434749603,
      "learning_rate": 0.0001272787979966611,
      "loss": 0.9789,
      "num_input_tokens_seen": 3054176,
      "step": 1095,
      "train_runtime": 2809.5162,
      "train_tokens_per_second": 1087.083
    },
    {
      "epoch": 0.08806926615641134,
      "grad_norm": 0.318328320980072,
      "learning_rate": 0.00012721202003338898,
      "loss": 0.9675,
      "num_input_tokens_seen": 3056492,
      "step": 1096,
      "train_runtime": 2811.5311,
      "train_tokens_per_second": 1087.127
    },
    {
      "epoch": 0.08814962132626208,
      "grad_norm": 0.28950101137161255,
      "learning_rate": 0.00012714524207011686,
      "loss": 0.9639,
      "num_input_tokens_seen": 3058622,
      "step": 1097,
      "train_runtime": 2813.4483,
      "train_tokens_per_second": 1087.143
    },
    {
      "epoch": 0.08822997649611282,
      "grad_norm": 0.26731500029563904,
      "learning_rate": 0.00012707846410684473,
      "loss": 0.8594,
      "num_input_tokens_seen": 3060882,
      "step": 1098,
      "train_runtime": 2815.2939,
      "train_tokens_per_second": 1087.234
    },
    {
      "epoch": 0.08831033166596355,
      "grad_norm": 0.3259464204311371,
      "learning_rate": 0.0001270116861435726,
      "loss": 0.8474,
      "num_input_tokens_seen": 3064758,
      "step": 1099,
      "train_runtime": 2817.9789,
      "train_tokens_per_second": 1087.573
    },
    {
      "epoch": 0.0883906868358143,
      "grad_norm": 0.2628142237663269,
      "learning_rate": 0.0001269449081803005,
      "loss": 0.9793,
      "num_input_tokens_seen": 3067526,
      "step": 1100,
      "train_runtime": 2820.2414,
      "train_tokens_per_second": 1087.682
    },
    {
      "epoch": 0.08847104200566504,
      "grad_norm": 0.2963213622570038,
      "learning_rate": 0.00012687813021702838,
      "loss": 1.0457,
      "num_input_tokens_seen": 3072786,
      "step": 1101,
      "train_runtime": 2823.9783,
      "train_tokens_per_second": 1088.105
    },
    {
      "epoch": 0.08855139717551579,
      "grad_norm": 0.25485289096832275,
      "learning_rate": 0.00012681135225375628,
      "loss": 0.8066,
      "num_input_tokens_seen": 3075326,
      "step": 1102,
      "train_runtime": 2826.029,
      "train_tokens_per_second": 1088.215
    },
    {
      "epoch": 0.08863175234536652,
      "grad_norm": 0.3377240002155304,
      "learning_rate": 0.00012674457429048415,
      "loss": 1.0257,
      "num_input_tokens_seen": 3077822,
      "step": 1103,
      "train_runtime": 2828.1255,
      "train_tokens_per_second": 1088.29
    },
    {
      "epoch": 0.08871210751521726,
      "grad_norm": 0.3242761194705963,
      "learning_rate": 0.00012667779632721202,
      "loss": 0.9341,
      "num_input_tokens_seen": 3080172,
      "step": 1104,
      "train_runtime": 2830.1085,
      "train_tokens_per_second": 1088.358
    },
    {
      "epoch": 0.088792462685068,
      "grad_norm": 0.3196955621242523,
      "learning_rate": 0.00012661101836393992,
      "loss": 0.8233,
      "num_input_tokens_seen": 3081838,
      "step": 1105,
      "train_runtime": 2831.8054,
      "train_tokens_per_second": 1088.294
    },
    {
      "epoch": 0.08887281785491874,
      "grad_norm": 0.31218263506889343,
      "learning_rate": 0.0001265442404006678,
      "loss": 0.697,
      "num_input_tokens_seen": 3083180,
      "step": 1106,
      "train_runtime": 2833.3457,
      "train_tokens_per_second": 1088.176
    },
    {
      "epoch": 0.08895317302476947,
      "grad_norm": 0.25134435296058655,
      "learning_rate": 0.00012647746243739567,
      "loss": 0.9046,
      "num_input_tokens_seen": 3085902,
      "step": 1107,
      "train_runtime": 2835.5446,
      "train_tokens_per_second": 1088.293
    },
    {
      "epoch": 0.08903352819462022,
      "grad_norm": 0.3291313648223877,
      "learning_rate": 0.00012641068447412354,
      "loss": 1.0468,
      "num_input_tokens_seen": 3088724,
      "step": 1108,
      "train_runtime": 2837.8289,
      "train_tokens_per_second": 1088.411
    },
    {
      "epoch": 0.08911388336447096,
      "grad_norm": 0.27928370237350464,
      "learning_rate": 0.00012634390651085142,
      "loss": 0.5629,
      "num_input_tokens_seen": 3091956,
      "step": 1109,
      "train_runtime": 2840.291,
      "train_tokens_per_second": 1088.605
    },
    {
      "epoch": 0.08919423853432171,
      "grad_norm": 0.28240346908569336,
      "learning_rate": 0.00012627712854757932,
      "loss": 0.8145,
      "num_input_tokens_seen": 3093976,
      "step": 1110,
      "train_runtime": 2842.1335,
      "train_tokens_per_second": 1088.61
    },
    {
      "epoch": 0.08927459370417244,
      "grad_norm": 0.3529825508594513,
      "learning_rate": 0.0001262103505843072,
      "loss": 1.157,
      "num_input_tokens_seen": 3095586,
      "step": 1111,
      "train_runtime": 2845.8065,
      "train_tokens_per_second": 1087.771
    },
    {
      "epoch": 0.08935494887402318,
      "grad_norm": 0.36176350712776184,
      "learning_rate": 0.00012614357262103506,
      "loss": 0.6121,
      "num_input_tokens_seen": 3097942,
      "step": 1112,
      "train_runtime": 2847.8101,
      "train_tokens_per_second": 1087.833
    },
    {
      "epoch": 0.08943530404387393,
      "grad_norm": 0.3233826756477356,
      "learning_rate": 0.00012607679465776294,
      "loss": 0.8317,
      "num_input_tokens_seen": 3100626,
      "step": 1113,
      "train_runtime": 2850.0094,
      "train_tokens_per_second": 1087.935
    },
    {
      "epoch": 0.08951565921372466,
      "grad_norm": 0.3124631345272064,
      "learning_rate": 0.0001260100166944908,
      "loss": 0.8077,
      "num_input_tokens_seen": 3103032,
      "step": 1114,
      "train_runtime": 2851.9984,
      "train_tokens_per_second": 1088.02
    },
    {
      "epoch": 0.08959601438357541,
      "grad_norm": 0.4034174084663391,
      "learning_rate": 0.00012594323873121868,
      "loss": 1.0622,
      "num_input_tokens_seen": 3105442,
      "step": 1115,
      "train_runtime": 2854.1264,
      "train_tokens_per_second": 1088.053
    },
    {
      "epoch": 0.08967636955342614,
      "grad_norm": 0.34770938754081726,
      "learning_rate": 0.00012587646076794658,
      "loss": 0.9828,
      "num_input_tokens_seen": 3108546,
      "step": 1116,
      "train_runtime": 2856.5363,
      "train_tokens_per_second": 1088.222
    },
    {
      "epoch": 0.08975672472327688,
      "grad_norm": 0.30814117193222046,
      "learning_rate": 0.00012580968280467446,
      "loss": 1.1445,
      "num_input_tokens_seen": 3110768,
      "step": 1117,
      "train_runtime": 2858.492,
      "train_tokens_per_second": 1088.255
    },
    {
      "epoch": 0.08983707989312763,
      "grad_norm": 0.2673361599445343,
      "learning_rate": 0.00012574290484140233,
      "loss": 0.8873,
      "num_input_tokens_seen": 3114964,
      "step": 1118,
      "train_runtime": 2861.5538,
      "train_tokens_per_second": 1088.557
    },
    {
      "epoch": 0.08991743506297836,
      "grad_norm": 0.25939151644706726,
      "learning_rate": 0.00012567612687813023,
      "loss": 1.0144,
      "num_input_tokens_seen": 3117966,
      "step": 1119,
      "train_runtime": 2863.873,
      "train_tokens_per_second": 1088.724
    },
    {
      "epoch": 0.08999779023282911,
      "grad_norm": 0.39517298340797424,
      "learning_rate": 0.0001256093489148581,
      "loss": 1.1092,
      "num_input_tokens_seen": 3120626,
      "step": 1120,
      "train_runtime": 2865.9633,
      "train_tokens_per_second": 1088.858
    },
    {
      "epoch": 0.09007814540267985,
      "grad_norm": 0.32868483662605286,
      "learning_rate": 0.000125542570951586,
      "loss": 0.8212,
      "num_input_tokens_seen": 3122954,
      "step": 1121,
      "train_runtime": 2868.0347,
      "train_tokens_per_second": 1088.883
    },
    {
      "epoch": 0.09015850057253058,
      "grad_norm": 0.23024609684944153,
      "learning_rate": 0.00012547579298831388,
      "loss": 0.6645,
      "num_input_tokens_seen": 3125558,
      "step": 1122,
      "train_runtime": 2870.1293,
      "train_tokens_per_second": 1088.996
    },
    {
      "epoch": 0.09023885574238133,
      "grad_norm": 0.3265283405780792,
      "learning_rate": 0.00012540901502504175,
      "loss": 0.8031,
      "num_input_tokens_seen": 3129850,
      "step": 1123,
      "train_runtime": 2873.0987,
      "train_tokens_per_second": 1089.364
    },
    {
      "epoch": 0.09031921091223206,
      "grad_norm": 0.3203762173652649,
      "learning_rate": 0.00012534223706176962,
      "loss": 0.8992,
      "num_input_tokens_seen": 3131932,
      "step": 1124,
      "train_runtime": 2874.9894,
      "train_tokens_per_second": 1089.372
    },
    {
      "epoch": 0.0903995660820828,
      "grad_norm": 0.2812959849834442,
      "learning_rate": 0.0001252754590984975,
      "loss": 0.9781,
      "num_input_tokens_seen": 3134200,
      "step": 1125,
      "train_runtime": 2876.8935,
      "train_tokens_per_second": 1089.439
    },
    {
      "epoch": 0.09047992125193355,
      "grad_norm": 0.2835489809513092,
      "learning_rate": 0.0001252086811352254,
      "loss": 0.8796,
      "num_input_tokens_seen": 3137694,
      "step": 1126,
      "train_runtime": 2879.5734,
      "train_tokens_per_second": 1089.638
    },
    {
      "epoch": 0.09056027642178428,
      "grad_norm": 0.31469953060150146,
      "learning_rate": 0.00012514190317195327,
      "loss": 1.0935,
      "num_input_tokens_seen": 3139166,
      "step": 1127,
      "train_runtime": 2881.1656,
      "train_tokens_per_second": 1089.547
    },
    {
      "epoch": 0.09064063159163503,
      "grad_norm": 0.3524523377418518,
      "learning_rate": 0.00012507512520868114,
      "loss": 1.1163,
      "num_input_tokens_seen": 3142166,
      "step": 1128,
      "train_runtime": 2883.5085,
      "train_tokens_per_second": 1089.702
    },
    {
      "epoch": 0.09072098676148577,
      "grad_norm": 0.30217206478118896,
      "learning_rate": 0.00012500834724540902,
      "loss": 1.1273,
      "num_input_tokens_seen": 3145102,
      "step": 1129,
      "train_runtime": 2885.7967,
      "train_tokens_per_second": 1089.856
    },
    {
      "epoch": 0.0908013419313365,
      "grad_norm": 0.3433888256549835,
      "learning_rate": 0.0001249415692821369,
      "loss": 0.7967,
      "num_input_tokens_seen": 3147396,
      "step": 1130,
      "train_runtime": 2887.8093,
      "train_tokens_per_second": 1089.891
    },
    {
      "epoch": 0.09088169710118725,
      "grad_norm": 0.3130668103694916,
      "learning_rate": 0.00012487479131886476,
      "loss": 0.9809,
      "num_input_tokens_seen": 3149078,
      "step": 1131,
      "train_runtime": 2889.4744,
      "train_tokens_per_second": 1089.845
    },
    {
      "epoch": 0.09096205227103799,
      "grad_norm": 0.2751938998699188,
      "learning_rate": 0.00012480801335559266,
      "loss": 0.766,
      "num_input_tokens_seen": 3152012,
      "step": 1132,
      "train_runtime": 2891.7653,
      "train_tokens_per_second": 1089.996
    },
    {
      "epoch": 0.09104240744088873,
      "grad_norm": 0.30756300687789917,
      "learning_rate": 0.00012474123539232053,
      "loss": 1.0362,
      "num_input_tokens_seen": 3153830,
      "step": 1133,
      "train_runtime": 2893.5276,
      "train_tokens_per_second": 1089.96
    },
    {
      "epoch": 0.09112276261073947,
      "grad_norm": 0.3334805965423584,
      "learning_rate": 0.0001246744574290484,
      "loss": 0.8753,
      "num_input_tokens_seen": 3155538,
      "step": 1134,
      "train_runtime": 2895.1942,
      "train_tokens_per_second": 1089.923
    },
    {
      "epoch": 0.0912031177805902,
      "grad_norm": 0.27908143401145935,
      "learning_rate": 0.0001246076794657763,
      "loss": 0.6604,
      "num_input_tokens_seen": 3158164,
      "step": 1135,
      "train_runtime": 2897.2174,
      "train_tokens_per_second": 1090.068
    },
    {
      "epoch": 0.09128347295044095,
      "grad_norm": 0.3932334780693054,
      "learning_rate": 0.00012454090150250418,
      "loss": 1.0836,
      "num_input_tokens_seen": 3163376,
      "step": 1136,
      "train_runtime": 2900.8607,
      "train_tokens_per_second": 1090.496
    },
    {
      "epoch": 0.09136382812029169,
      "grad_norm": 0.33885687589645386,
      "learning_rate": 0.00012447412353923208,
      "loss": 0.9214,
      "num_input_tokens_seen": 3166084,
      "step": 1137,
      "train_runtime": 2903.015,
      "train_tokens_per_second": 1090.619
    },
    {
      "epoch": 0.09144418329014244,
      "grad_norm": 0.2946315407752991,
      "learning_rate": 0.00012440734557595995,
      "loss": 0.9546,
      "num_input_tokens_seen": 3167780,
      "step": 1138,
      "train_runtime": 2904.6595,
      "train_tokens_per_second": 1090.586
    },
    {
      "epoch": 0.09152453845999317,
      "grad_norm": 0.31270208954811096,
      "learning_rate": 0.00012434056761268783,
      "loss": 1.0801,
      "num_input_tokens_seen": 3170926,
      "step": 1139,
      "train_runtime": 2907.1033,
      "train_tokens_per_second": 1090.751
    },
    {
      "epoch": 0.0916048936298439,
      "grad_norm": 0.3032757341861725,
      "learning_rate": 0.0001242737896494157,
      "loss": 0.6802,
      "num_input_tokens_seen": 3175594,
      "step": 1140,
      "train_runtime": 2910.2688,
      "train_tokens_per_second": 1091.169
    },
    {
      "epoch": 0.09168524879969465,
      "grad_norm": 0.328593909740448,
      "learning_rate": 0.00012420701168614357,
      "loss": 0.8727,
      "num_input_tokens_seen": 3178492,
      "step": 1141,
      "train_runtime": 2914.6077,
      "train_tokens_per_second": 1090.539
    },
    {
      "epoch": 0.09176560396954539,
      "grad_norm": 0.32067033648490906,
      "learning_rate": 0.00012414023372287147,
      "loss": 1.3166,
      "num_input_tokens_seen": 3181490,
      "step": 1142,
      "train_runtime": 2916.9055,
      "train_tokens_per_second": 1090.707
    },
    {
      "epoch": 0.09184595913939612,
      "grad_norm": 0.27059289813041687,
      "learning_rate": 0.00012407345575959935,
      "loss": 0.9869,
      "num_input_tokens_seen": 3185200,
      "step": 1143,
      "train_runtime": 2919.5893,
      "train_tokens_per_second": 1090.975
    },
    {
      "epoch": 0.09192631430924687,
      "grad_norm": 0.3421857953071594,
      "learning_rate": 0.00012400667779632722,
      "loss": 1.0266,
      "num_input_tokens_seen": 3188268,
      "step": 1144,
      "train_runtime": 2921.9862,
      "train_tokens_per_second": 1091.13
    },
    {
      "epoch": 0.09200666947909761,
      "grad_norm": 0.34387344121932983,
      "learning_rate": 0.0001239398998330551,
      "loss": 1.0565,
      "num_input_tokens_seen": 3191514,
      "step": 1145,
      "train_runtime": 2924.3616,
      "train_tokens_per_second": 1091.354
    },
    {
      "epoch": 0.09208702464894836,
      "grad_norm": 0.45765259861946106,
      "learning_rate": 0.00012387312186978297,
      "loss": 1.2379,
      "num_input_tokens_seen": 3193618,
      "step": 1146,
      "train_runtime": 2926.2304,
      "train_tokens_per_second": 1091.376
    },
    {
      "epoch": 0.09216737981879909,
      "grad_norm": 0.27509790658950806,
      "learning_rate": 0.00012380634390651084,
      "loss": 0.9037,
      "num_input_tokens_seen": 3196160,
      "step": 1147,
      "train_runtime": 2928.2053,
      "train_tokens_per_second": 1091.508
    },
    {
      "epoch": 0.09224773498864983,
      "grad_norm": 0.28337612748146057,
      "learning_rate": 0.00012373956594323874,
      "loss": 0.8756,
      "num_input_tokens_seen": 3198826,
      "step": 1148,
      "train_runtime": 2930.3393,
      "train_tokens_per_second": 1091.623
    },
    {
      "epoch": 0.09232809015850058,
      "grad_norm": 0.3089130222797394,
      "learning_rate": 0.00012367278797996661,
      "loss": 1.1017,
      "num_input_tokens_seen": 3201486,
      "step": 1149,
      "train_runtime": 2932.5295,
      "train_tokens_per_second": 1091.715
    },
    {
      "epoch": 0.09240844532835131,
      "grad_norm": 0.31819039583206177,
      "learning_rate": 0.0001236060100166945,
      "loss": 0.8012,
      "num_input_tokens_seen": 3203882,
      "step": 1150,
      "train_runtime": 2934.4923,
      "train_tokens_per_second": 1091.801
    },
    {
      "epoch": 0.09248880049820206,
      "grad_norm": 0.3065102994441986,
      "learning_rate": 0.00012353923205342236,
      "loss": 0.9448,
      "num_input_tokens_seen": 3206016,
      "step": 1151,
      "train_runtime": 2936.2829,
      "train_tokens_per_second": 1091.862
    },
    {
      "epoch": 0.0925691556680528,
      "grad_norm": 0.3837001323699951,
      "learning_rate": 0.00012347245409015026,
      "loss": 0.6354,
      "num_input_tokens_seen": 3208838,
      "step": 1152,
      "train_runtime": 2938.5387,
      "train_tokens_per_second": 1091.984
    },
    {
      "epoch": 0.09264951083790353,
      "grad_norm": 0.3397212624549866,
      "learning_rate": 0.00012340567612687813,
      "loss": 1.0924,
      "num_input_tokens_seen": 3212190,
      "step": 1153,
      "train_runtime": 2941.0359,
      "train_tokens_per_second": 1092.197
    },
    {
      "epoch": 0.09272986600775428,
      "grad_norm": 0.30270108580589294,
      "learning_rate": 0.00012333889816360603,
      "loss": 0.8581,
      "num_input_tokens_seen": 3214572,
      "step": 1154,
      "train_runtime": 2943.0509,
      "train_tokens_per_second": 1092.258
    },
    {
      "epoch": 0.09281022117760501,
      "grad_norm": 0.2992537319660187,
      "learning_rate": 0.0001232721202003339,
      "loss": 0.8548,
      "num_input_tokens_seen": 3216486,
      "step": 1155,
      "train_runtime": 2944.734,
      "train_tokens_per_second": 1092.284
    },
    {
      "epoch": 0.09289057634745576,
      "grad_norm": 0.33950281143188477,
      "learning_rate": 0.00012320534223706178,
      "loss": 1.09,
      "num_input_tokens_seen": 3218280,
      "step": 1156,
      "train_runtime": 2946.3632,
      "train_tokens_per_second": 1092.289
    },
    {
      "epoch": 0.0929709315173065,
      "grad_norm": 0.315149188041687,
      "learning_rate": 0.00012313856427378965,
      "loss": 1.2139,
      "num_input_tokens_seen": 3221044,
      "step": 1157,
      "train_runtime": 2948.5947,
      "train_tokens_per_second": 1092.4
    },
    {
      "epoch": 0.09305128668715723,
      "grad_norm": 0.26932209730148315,
      "learning_rate": 0.00012307178631051755,
      "loss": 0.9881,
      "num_input_tokens_seen": 3225038,
      "step": 1158,
      "train_runtime": 2951.4962,
      "train_tokens_per_second": 1092.679
    },
    {
      "epoch": 0.09313164185700798,
      "grad_norm": 0.31834033131599426,
      "learning_rate": 0.00012300500834724543,
      "loss": 1.2673,
      "num_input_tokens_seen": 3227596,
      "step": 1159,
      "train_runtime": 2953.5391,
      "train_tokens_per_second": 1092.789
    },
    {
      "epoch": 0.09321199702685871,
      "grad_norm": 0.3501078188419342,
      "learning_rate": 0.0001229382303839733,
      "loss": 1.4483,
      "num_input_tokens_seen": 3230014,
      "step": 1160,
      "train_runtime": 2955.5747,
      "train_tokens_per_second": 1092.855
    },
    {
      "epoch": 0.09329235219670945,
      "grad_norm": 0.3090665638446808,
      "learning_rate": 0.00012287145242070117,
      "loss": 0.9208,
      "num_input_tokens_seen": 3232256,
      "step": 1161,
      "train_runtime": 2957.5276,
      "train_tokens_per_second": 1092.891
    },
    {
      "epoch": 0.0933727073665602,
      "grad_norm": 0.3182087540626526,
      "learning_rate": 0.00012280467445742905,
      "loss": 0.9703,
      "num_input_tokens_seen": 3234172,
      "step": 1162,
      "train_runtime": 2959.2768,
      "train_tokens_per_second": 1092.893
    },
    {
      "epoch": 0.09345306253641093,
      "grad_norm": 0.3346424698829651,
      "learning_rate": 0.00012273789649415692,
      "loss": 1.0238,
      "num_input_tokens_seen": 3236354,
      "step": 1163,
      "train_runtime": 2961.1973,
      "train_tokens_per_second": 1092.921
    },
    {
      "epoch": 0.09353341770626168,
      "grad_norm": 0.304471880197525,
      "learning_rate": 0.00012267111853088482,
      "loss": 1.1515,
      "num_input_tokens_seen": 3239612,
      "step": 1164,
      "train_runtime": 2963.5918,
      "train_tokens_per_second": 1093.137
    },
    {
      "epoch": 0.09361377287611242,
      "grad_norm": 0.26479142904281616,
      "learning_rate": 0.0001226043405676127,
      "loss": 0.8122,
      "num_input_tokens_seen": 3242874,
      "step": 1165,
      "train_runtime": 2966.052,
      "train_tokens_per_second": 1093.33
    },
    {
      "epoch": 0.09369412804596315,
      "grad_norm": 0.3055541515350342,
      "learning_rate": 0.00012253756260434057,
      "loss": 0.8506,
      "num_input_tokens_seen": 3245036,
      "step": 1166,
      "train_runtime": 2967.9365,
      "train_tokens_per_second": 1093.364
    },
    {
      "epoch": 0.0937744832158139,
      "grad_norm": 0.3619381785392761,
      "learning_rate": 0.00012247078464106844,
      "loss": 0.9579,
      "num_input_tokens_seen": 3246832,
      "step": 1167,
      "train_runtime": 2969.7046,
      "train_tokens_per_second": 1093.318
    },
    {
      "epoch": 0.09385483838566464,
      "grad_norm": 0.28716981410980225,
      "learning_rate": 0.0001224040066777963,
      "loss": 1.1342,
      "num_input_tokens_seen": 3249492,
      "step": 1168,
      "train_runtime": 2971.8211,
      "train_tokens_per_second": 1093.435
    },
    {
      "epoch": 0.09393519355551538,
      "grad_norm": 0.2521206736564636,
      "learning_rate": 0.0001223372287145242,
      "loss": 0.9544,
      "num_input_tokens_seen": 3251466,
      "step": 1169,
      "train_runtime": 2973.5833,
      "train_tokens_per_second": 1093.45
    },
    {
      "epoch": 0.09401554872536612,
      "grad_norm": 0.3139553964138031,
      "learning_rate": 0.00012227045075125209,
      "loss": 1.1726,
      "num_input_tokens_seen": 3255194,
      "step": 1170,
      "train_runtime": 2976.2718,
      "train_tokens_per_second": 1093.715
    },
    {
      "epoch": 0.09409590389521685,
      "grad_norm": 0.3215081989765167,
      "learning_rate": 0.00012220367278797999,
      "loss": 1.093,
      "num_input_tokens_seen": 3257140,
      "step": 1171,
      "train_runtime": 2980.0717,
      "train_tokens_per_second": 1092.974
    },
    {
      "epoch": 0.0941762590650676,
      "grad_norm": 0.25101542472839355,
      "learning_rate": 0.00012213689482470786,
      "loss": 0.6596,
      "num_input_tokens_seen": 3260770,
      "step": 1172,
      "train_runtime": 2982.74,
      "train_tokens_per_second": 1093.213
    },
    {
      "epoch": 0.09425661423491834,
      "grad_norm": 0.32819992303848267,
      "learning_rate": 0.00012207011686143572,
      "loss": 0.9306,
      "num_input_tokens_seen": 3263358,
      "step": 1173,
      "train_runtime": 2984.8253,
      "train_tokens_per_second": 1093.316
    },
    {
      "epoch": 0.09433696940476909,
      "grad_norm": 0.3013783097267151,
      "learning_rate": 0.00012200333889816362,
      "loss": 0.8609,
      "num_input_tokens_seen": 3265994,
      "step": 1174,
      "train_runtime": 2986.9099,
      "train_tokens_per_second": 1093.436
    },
    {
      "epoch": 0.09441732457461982,
      "grad_norm": 0.4124008119106293,
      "learning_rate": 0.00012193656093489149,
      "loss": 0.9928,
      "num_input_tokens_seen": 3267986,
      "step": 1175,
      "train_runtime": 2988.7424,
      "train_tokens_per_second": 1093.432
    },
    {
      "epoch": 0.09449767974447056,
      "grad_norm": 0.31742510199546814,
      "learning_rate": 0.00012186978297161938,
      "loss": 1.0538,
      "num_input_tokens_seen": 3270978,
      "step": 1176,
      "train_runtime": 2991.1303,
      "train_tokens_per_second": 1093.559
    },
    {
      "epoch": 0.0945780349143213,
      "grad_norm": 0.2772507667541504,
      "learning_rate": 0.00012180300500834725,
      "loss": 0.6553,
      "num_input_tokens_seen": 3273460,
      "step": 1177,
      "train_runtime": 2993.1726,
      "train_tokens_per_second": 1093.642
    },
    {
      "epoch": 0.09465839008417204,
      "grad_norm": 0.32513031363487244,
      "learning_rate": 0.00012173622704507512,
      "loss": 0.9216,
      "num_input_tokens_seen": 3275222,
      "step": 1178,
      "train_runtime": 2994.8408,
      "train_tokens_per_second": 1093.621
    },
    {
      "epoch": 0.09473874525402277,
      "grad_norm": 0.30071112513542175,
      "learning_rate": 0.00012166944908180303,
      "loss": 0.9744,
      "num_input_tokens_seen": 3277610,
      "step": 1179,
      "train_runtime": 2996.866,
      "train_tokens_per_second": 1093.679
    },
    {
      "epoch": 0.09481910042387352,
      "grad_norm": 0.4937950074672699,
      "learning_rate": 0.0001216026711185309,
      "loss": 0.684,
      "num_input_tokens_seen": 3279686,
      "step": 1180,
      "train_runtime": 2998.6405,
      "train_tokens_per_second": 1093.724
    },
    {
      "epoch": 0.09489945559372426,
      "grad_norm": 0.35838666558265686,
      "learning_rate": 0.00012153589315525877,
      "loss": 1.1709,
      "num_input_tokens_seen": 3282104,
      "step": 1181,
      "train_runtime": 3000.6778,
      "train_tokens_per_second": 1093.788
    },
    {
      "epoch": 0.094979810763575,
      "grad_norm": 0.3639245629310608,
      "learning_rate": 0.00012146911519198664,
      "loss": 0.8092,
      "num_input_tokens_seen": 3284068,
      "step": 1182,
      "train_runtime": 3002.4917,
      "train_tokens_per_second": 1093.781
    },
    {
      "epoch": 0.09506016593342574,
      "grad_norm": 0.31621286273002625,
      "learning_rate": 0.00012140233722871452,
      "loss": 0.9947,
      "num_input_tokens_seen": 3286962,
      "step": 1183,
      "train_runtime": 3004.727,
      "train_tokens_per_second": 1093.93
    },
    {
      "epoch": 0.09514052110327648,
      "grad_norm": 0.4309380054473877,
      "learning_rate": 0.0001213355592654424,
      "loss": 0.9539,
      "num_input_tokens_seen": 3290742,
      "step": 1184,
      "train_runtime": 3007.5212,
      "train_tokens_per_second": 1094.171
    },
    {
      "epoch": 0.09522087627312723,
      "grad_norm": 0.2795886993408203,
      "learning_rate": 0.00012126878130217029,
      "loss": 0.8429,
      "num_input_tokens_seen": 3293040,
      "step": 1185,
      "train_runtime": 3009.4995,
      "train_tokens_per_second": 1094.215
    },
    {
      "epoch": 0.09530123144297796,
      "grad_norm": 0.40882760286331177,
      "learning_rate": 0.00012120200333889818,
      "loss": 1.0339,
      "num_input_tokens_seen": 3296526,
      "step": 1186,
      "train_runtime": 3012.2903,
      "train_tokens_per_second": 1094.359
    },
    {
      "epoch": 0.09538158661282871,
      "grad_norm": 0.4669959247112274,
      "learning_rate": 0.00012113522537562605,
      "loss": 1.0336,
      "num_input_tokens_seen": 3299012,
      "step": 1187,
      "train_runtime": 3014.4237,
      "train_tokens_per_second": 1094.409
    },
    {
      "epoch": 0.09546194178267944,
      "grad_norm": 0.3601592779159546,
      "learning_rate": 0.00012106844741235392,
      "loss": 0.9862,
      "num_input_tokens_seen": 3301084,
      "step": 1188,
      "train_runtime": 3016.3675,
      "train_tokens_per_second": 1094.391
    },
    {
      "epoch": 0.09554229695253018,
      "grad_norm": 0.25590991973876953,
      "learning_rate": 0.0001210016694490818,
      "loss": 0.8542,
      "num_input_tokens_seen": 3302880,
      "step": 1189,
      "train_runtime": 3018.1605,
      "train_tokens_per_second": 1094.335
    },
    {
      "epoch": 0.09562265212238093,
      "grad_norm": 0.3337498903274536,
      "learning_rate": 0.0001209348914858097,
      "loss": 1.0826,
      "num_input_tokens_seen": 3304580,
      "step": 1190,
      "train_runtime": 3019.7608,
      "train_tokens_per_second": 1094.318
    },
    {
      "epoch": 0.09570300729223166,
      "grad_norm": 0.31814974546432495,
      "learning_rate": 0.00012086811352253757,
      "loss": 0.9773,
      "num_input_tokens_seen": 3307130,
      "step": 1191,
      "train_runtime": 3021.8885,
      "train_tokens_per_second": 1094.392
    },
    {
      "epoch": 0.09578336246208241,
      "grad_norm": 0.27574920654296875,
      "learning_rate": 0.00012080133555926544,
      "loss": 0.9745,
      "num_input_tokens_seen": 3310210,
      "step": 1192,
      "train_runtime": 3024.1884,
      "train_tokens_per_second": 1094.578
    },
    {
      "epoch": 0.09586371763193315,
      "grad_norm": 0.2402028739452362,
      "learning_rate": 0.00012073455759599333,
      "loss": 0.607,
      "num_input_tokens_seen": 3312534,
      "step": 1193,
      "train_runtime": 3026.1177,
      "train_tokens_per_second": 1094.648
    },
    {
      "epoch": 0.09594407280178388,
      "grad_norm": 0.3079510033130646,
      "learning_rate": 0.0001206677796327212,
      "loss": 1.1701,
      "num_input_tokens_seen": 3315400,
      "step": 1194,
      "train_runtime": 3028.4559,
      "train_tokens_per_second": 1094.749
    },
    {
      "epoch": 0.09602442797163463,
      "grad_norm": 0.30498263239860535,
      "learning_rate": 0.0001206010016694491,
      "loss": 1.2102,
      "num_input_tokens_seen": 3318620,
      "step": 1195,
      "train_runtime": 3030.936,
      "train_tokens_per_second": 1094.916
    },
    {
      "epoch": 0.09610478314148536,
      "grad_norm": 0.34362462162971497,
      "learning_rate": 0.00012053422370617698,
      "loss": 0.9796,
      "num_input_tokens_seen": 3321292,
      "step": 1196,
      "train_runtime": 3033.1273,
      "train_tokens_per_second": 1095.006
    },
    {
      "epoch": 0.0961851383113361,
      "grad_norm": 0.28773069381713867,
      "learning_rate": 0.00012046744574290485,
      "loss": 0.8244,
      "num_input_tokens_seen": 3323442,
      "step": 1197,
      "train_runtime": 3035.1383,
      "train_tokens_per_second": 1094.989
    },
    {
      "epoch": 0.09626549348118685,
      "grad_norm": 0.28585824370384216,
      "learning_rate": 0.00012040066777963272,
      "loss": 1.1634,
      "num_input_tokens_seen": 3325440,
      "step": 1198,
      "train_runtime": 3036.917,
      "train_tokens_per_second": 1095.005
    },
    {
      "epoch": 0.09634584865103758,
      "grad_norm": 0.2801036238670349,
      "learning_rate": 0.0001203338898163606,
      "loss": 0.883,
      "num_input_tokens_seen": 3327700,
      "step": 1199,
      "train_runtime": 3038.8898,
      "train_tokens_per_second": 1095.038
    },
    {
      "epoch": 0.09642620382088833,
      "grad_norm": 0.2651750147342682,
      "learning_rate": 0.00012026711185308848,
      "loss": 0.9353,
      "num_input_tokens_seen": 3329912,
      "step": 1200,
      "train_runtime": 3040.8113,
      "train_tokens_per_second": 1095.074
    },
    {
      "epoch": 0.09642620382088833,
      "eval_loss": 0.9765015244483948,
      "eval_runtime": 93.3343,
      "eval_samples_per_second": 10.65,
      "eval_steps_per_second": 5.325,
      "num_input_tokens_seen": 3329912,
      "step": 1200
    },
    {
      "epoch": 0.09650655899073907,
      "grad_norm": 0.314842164516449,
      "learning_rate": 0.00012020033388981637,
      "loss": 1.0329,
      "num_input_tokens_seen": 3332024,
      "step": 1201,
      "train_runtime": 3138.0196,
      "train_tokens_per_second": 1061.824
    },
    {
      "epoch": 0.0965869141605898,
      "grad_norm": 0.32207608222961426,
      "learning_rate": 0.00012013355592654426,
      "loss": 0.9658,
      "num_input_tokens_seen": 3336034,
      "step": 1202,
      "train_runtime": 3141.0618,
      "train_tokens_per_second": 1062.072
    },
    {
      "epoch": 0.09666726933044055,
      "grad_norm": 0.3103851079940796,
      "learning_rate": 0.00012006677796327213,
      "loss": 0.869,
      "num_input_tokens_seen": 3337826,
      "step": 1203,
      "train_runtime": 3142.8775,
      "train_tokens_per_second": 1062.029
    },
    {
      "epoch": 0.09674762450029128,
      "grad_norm": 0.3051537573337555,
      "learning_rate": 0.00012,
      "loss": 0.9135,
      "num_input_tokens_seen": 3339950,
      "step": 1204,
      "train_runtime": 3144.7626,
      "train_tokens_per_second": 1062.067
    },
    {
      "epoch": 0.09682797967014203,
      "grad_norm": 0.33335167169570923,
      "learning_rate": 0.00011993322203672788,
      "loss": 1.0979,
      "num_input_tokens_seen": 3343030,
      "step": 1205,
      "train_runtime": 3147.2014,
      "train_tokens_per_second": 1062.223
    },
    {
      "epoch": 0.09690833483999277,
      "grad_norm": 0.3052538335323334,
      "learning_rate": 0.00011986644407345578,
      "loss": 1.234,
      "num_input_tokens_seen": 3345764,
      "step": 1206,
      "train_runtime": 3149.4487,
      "train_tokens_per_second": 1062.333
    },
    {
      "epoch": 0.0969886900098435,
      "grad_norm": 0.29944851994514465,
      "learning_rate": 0.00011979966611018365,
      "loss": 1.0247,
      "num_input_tokens_seen": 3347902,
      "step": 1207,
      "train_runtime": 3151.3077,
      "train_tokens_per_second": 1062.385
    },
    {
      "epoch": 0.09706904517969425,
      "grad_norm": 0.32125842571258545,
      "learning_rate": 0.00011973288814691152,
      "loss": 0.9658,
      "num_input_tokens_seen": 3350306,
      "step": 1208,
      "train_runtime": 3153.2804,
      "train_tokens_per_second": 1062.483
    },
    {
      "epoch": 0.09714940034954499,
      "grad_norm": 0.3880539536476135,
      "learning_rate": 0.0001196661101836394,
      "loss": 1.0051,
      "num_input_tokens_seen": 3354422,
      "step": 1209,
      "train_runtime": 3156.2662,
      "train_tokens_per_second": 1062.782
    },
    {
      "epoch": 0.09722975551939574,
      "grad_norm": 0.3151746988296509,
      "learning_rate": 0.00011959933222036728,
      "loss": 0.9279,
      "num_input_tokens_seen": 3357024,
      "step": 1210,
      "train_runtime": 3158.4457,
      "train_tokens_per_second": 1062.872
    },
    {
      "epoch": 0.09731011068924647,
      "grad_norm": 0.2788081169128418,
      "learning_rate": 0.00011953255425709517,
      "loss": 0.7704,
      "num_input_tokens_seen": 3359668,
      "step": 1211,
      "train_runtime": 3160.6625,
      "train_tokens_per_second": 1062.963
    },
    {
      "epoch": 0.0973904658590972,
      "grad_norm": 0.2628996670246124,
      "learning_rate": 0.00011946577629382306,
      "loss": 0.8496,
      "num_input_tokens_seen": 3362928,
      "step": 1212,
      "train_runtime": 3163.2719,
      "train_tokens_per_second": 1063.117
    },
    {
      "epoch": 0.09747082102894795,
      "grad_norm": 0.3539094924926758,
      "learning_rate": 0.00011939899833055093,
      "loss": 1.2361,
      "num_input_tokens_seen": 3365638,
      "step": 1213,
      "train_runtime": 3165.4906,
      "train_tokens_per_second": 1063.228
    },
    {
      "epoch": 0.09755117619879869,
      "grad_norm": 0.3183014690876007,
      "learning_rate": 0.0001193322203672788,
      "loss": 0.8117,
      "num_input_tokens_seen": 3367574,
      "step": 1214,
      "train_runtime": 3167.3193,
      "train_tokens_per_second": 1063.225
    },
    {
      "epoch": 0.09763153136864942,
      "grad_norm": 0.29052481055259705,
      "learning_rate": 0.00011926544240400668,
      "loss": 0.8508,
      "num_input_tokens_seen": 3369940,
      "step": 1215,
      "train_runtime": 3169.3579,
      "train_tokens_per_second": 1063.288
    },
    {
      "epoch": 0.09771188653850017,
      "grad_norm": 0.27354535460472107,
      "learning_rate": 0.00011919866444073455,
      "loss": 0.6888,
      "num_input_tokens_seen": 3372702,
      "step": 1216,
      "train_runtime": 3171.5886,
      "train_tokens_per_second": 1063.411
    },
    {
      "epoch": 0.09779224170835091,
      "grad_norm": 0.2888067066669464,
      "learning_rate": 0.00011913188647746245,
      "loss": 0.8659,
      "num_input_tokens_seen": 3374686,
      "step": 1217,
      "train_runtime": 3173.3758,
      "train_tokens_per_second": 1063.437
    },
    {
      "epoch": 0.09787259687820166,
      "grad_norm": 0.51072758436203,
      "learning_rate": 0.00011906510851419032,
      "loss": 1.2043,
      "num_input_tokens_seen": 3378102,
      "step": 1218,
      "train_runtime": 3175.984,
      "train_tokens_per_second": 1063.639
    },
    {
      "epoch": 0.09795295204805239,
      "grad_norm": 0.2898387908935547,
      "learning_rate": 0.00011899833055091821,
      "loss": 0.7763,
      "num_input_tokens_seen": 3380130,
      "step": 1219,
      "train_runtime": 3177.7603,
      "train_tokens_per_second": 1063.683
    },
    {
      "epoch": 0.09803330721790313,
      "grad_norm": 0.3145882785320282,
      "learning_rate": 0.00011893155258764608,
      "loss": 1.2572,
      "num_input_tokens_seen": 3386286,
      "step": 1220,
      "train_runtime": 3181.9788,
      "train_tokens_per_second": 1064.208
    },
    {
      "epoch": 0.09811366238775387,
      "grad_norm": 0.24348506331443787,
      "learning_rate": 0.00011886477462437396,
      "loss": 0.7086,
      "num_input_tokens_seen": 3388292,
      "step": 1221,
      "train_runtime": 3183.8526,
      "train_tokens_per_second": 1064.211
    },
    {
      "epoch": 0.09819401755760461,
      "grad_norm": 0.32285764813423157,
      "learning_rate": 0.00011879799666110186,
      "loss": 1.1369,
      "num_input_tokens_seen": 3390588,
      "step": 1222,
      "train_runtime": 3185.8309,
      "train_tokens_per_second": 1064.271
    },
    {
      "epoch": 0.09827437272745536,
      "grad_norm": 2.257875680923462,
      "learning_rate": 0.00011873121869782973,
      "loss": 0.9062,
      "num_input_tokens_seen": 3392844,
      "step": 1223,
      "train_runtime": 3187.7494,
      "train_tokens_per_second": 1064.338
    },
    {
      "epoch": 0.0983547278973061,
      "grad_norm": 0.29244640469551086,
      "learning_rate": 0.0001186644407345576,
      "loss": 0.9722,
      "num_input_tokens_seen": 3394832,
      "step": 1224,
      "train_runtime": 3189.5074,
      "train_tokens_per_second": 1064.375
    },
    {
      "epoch": 0.09843508306715683,
      "grad_norm": 0.3648703694343567,
      "learning_rate": 0.00011859766277128547,
      "loss": 1.085,
      "num_input_tokens_seen": 3397904,
      "step": 1225,
      "train_runtime": 3191.9959,
      "train_tokens_per_second": 1064.508
    },
    {
      "epoch": 0.09851543823700758,
      "grad_norm": 0.3223254382610321,
      "learning_rate": 0.00011853088480801335,
      "loss": 1.059,
      "num_input_tokens_seen": 3401394,
      "step": 1226,
      "train_runtime": 3194.6813,
      "train_tokens_per_second": 1064.705
    },
    {
      "epoch": 0.09859579340685831,
      "grad_norm": 0.30533918738365173,
      "learning_rate": 0.00011846410684474125,
      "loss": 1.1062,
      "num_input_tokens_seen": 3404168,
      "step": 1227,
      "train_runtime": 3196.9597,
      "train_tokens_per_second": 1064.814
    },
    {
      "epoch": 0.09867614857670906,
      "grad_norm": 0.2948809266090393,
      "learning_rate": 0.00011839732888146912,
      "loss": 1.0735,
      "num_input_tokens_seen": 3406594,
      "step": 1228,
      "train_runtime": 3198.9441,
      "train_tokens_per_second": 1064.912
    },
    {
      "epoch": 0.0987565037465598,
      "grad_norm": 0.42436307668685913,
      "learning_rate": 0.00011833055091819701,
      "loss": 1.0877,
      "num_input_tokens_seen": 3409922,
      "step": 1229,
      "train_runtime": 3201.5479,
      "train_tokens_per_second": 1065.085
    },
    {
      "epoch": 0.09883685891641053,
      "grad_norm": 0.43277955055236816,
      "learning_rate": 0.00011826377295492488,
      "loss": 0.8141,
      "num_input_tokens_seen": 3411954,
      "step": 1230,
      "train_runtime": 3203.5064,
      "train_tokens_per_second": 1065.069
    },
    {
      "epoch": 0.09891721408626128,
      "grad_norm": 0.3280401825904846,
      "learning_rate": 0.00011819699499165275,
      "loss": 1.1611,
      "num_input_tokens_seen": 3415216,
      "step": 1231,
      "train_runtime": 3208.063,
      "train_tokens_per_second": 1064.573
    },
    {
      "epoch": 0.09899756925611201,
      "grad_norm": 0.35693255066871643,
      "learning_rate": 0.00011813021702838063,
      "loss": 0.8101,
      "num_input_tokens_seen": 3418256,
      "step": 1232,
      "train_runtime": 3210.416,
      "train_tokens_per_second": 1064.739
    },
    {
      "epoch": 0.09907792442596275,
      "grad_norm": 0.3006307780742645,
      "learning_rate": 0.00011806343906510853,
      "loss": 0.8263,
      "num_input_tokens_seen": 3421074,
      "step": 1233,
      "train_runtime": 3212.6835,
      "train_tokens_per_second": 1064.865
    },
    {
      "epoch": 0.0991582795958135,
      "grad_norm": 0.39713171124458313,
      "learning_rate": 0.0001179966611018364,
      "loss": 1.295,
      "num_input_tokens_seen": 3423054,
      "step": 1234,
      "train_runtime": 3214.5737,
      "train_tokens_per_second": 1064.855
    },
    {
      "epoch": 0.09923863476566423,
      "grad_norm": 0.30702802538871765,
      "learning_rate": 0.00011792988313856427,
      "loss": 0.9079,
      "num_input_tokens_seen": 3425640,
      "step": 1235,
      "train_runtime": 3216.7137,
      "train_tokens_per_second": 1064.95
    },
    {
      "epoch": 0.09931898993551498,
      "grad_norm": 0.3540534973144531,
      "learning_rate": 0.00011786310517529216,
      "loss": 1.2356,
      "num_input_tokens_seen": 3429086,
      "step": 1236,
      "train_runtime": 3219.2915,
      "train_tokens_per_second": 1065.168
    },
    {
      "epoch": 0.09939934510536572,
      "grad_norm": 0.26566973328590393,
      "learning_rate": 0.00011779632721202003,
      "loss": 0.6904,
      "num_input_tokens_seen": 3431732,
      "step": 1237,
      "train_runtime": 3221.4314,
      "train_tokens_per_second": 1065.282
    },
    {
      "epoch": 0.09947970027521645,
      "grad_norm": 0.27577558159828186,
      "learning_rate": 0.00011772954924874793,
      "loss": 0.9845,
      "num_input_tokens_seen": 3433848,
      "step": 1238,
      "train_runtime": 3223.349,
      "train_tokens_per_second": 1065.304
    },
    {
      "epoch": 0.0995600554450672,
      "grad_norm": 0.314525842666626,
      "learning_rate": 0.00011766277128547581,
      "loss": 1.0063,
      "num_input_tokens_seen": 3436860,
      "step": 1239,
      "train_runtime": 3225.6969,
      "train_tokens_per_second": 1065.463
    },
    {
      "epoch": 0.09964041061491793,
      "grad_norm": 0.3086790144443512,
      "learning_rate": 0.00011759599332220368,
      "loss": 1.2317,
      "num_input_tokens_seen": 3439382,
      "step": 1240,
      "train_runtime": 3227.76,
      "train_tokens_per_second": 1065.563
    },
    {
      "epoch": 0.09972076578476868,
      "grad_norm": 0.31807294487953186,
      "learning_rate": 0.00011752921535893155,
      "loss": 1.0096,
      "num_input_tokens_seen": 3442102,
      "step": 1241,
      "train_runtime": 3229.9903,
      "train_tokens_per_second": 1065.669
    },
    {
      "epoch": 0.09980112095461942,
      "grad_norm": 0.3649135231971741,
      "learning_rate": 0.00011746243739565943,
      "loss": 1.1959,
      "num_input_tokens_seen": 3445372,
      "step": 1242,
      "train_runtime": 3232.4229,
      "train_tokens_per_second": 1065.879
    },
    {
      "epoch": 0.09988147612447015,
      "grad_norm": 0.3410540819168091,
      "learning_rate": 0.00011739565943238733,
      "loss": 1.2128,
      "num_input_tokens_seen": 3450264,
      "step": 1243,
      "train_runtime": 3235.9695,
      "train_tokens_per_second": 1066.223
    },
    {
      "epoch": 0.0999618312943209,
      "grad_norm": 0.31123611330986023,
      "learning_rate": 0.0001173288814691152,
      "loss": 1.1788,
      "num_input_tokens_seen": 3453118,
      "step": 1244,
      "train_runtime": 3238.2566,
      "train_tokens_per_second": 1066.351
    },
    {
      "epoch": 0.10004218646417164,
      "grad_norm": 0.5428243279457092,
      "learning_rate": 0.00011726210350584307,
      "loss": 0.7929,
      "num_input_tokens_seen": 3454734,
      "step": 1245,
      "train_runtime": 3239.9068,
      "train_tokens_per_second": 1066.307
    },
    {
      "epoch": 0.10012254163402239,
      "grad_norm": 0.3110898435115814,
      "learning_rate": 0.00011719532554257096,
      "loss": 0.9553,
      "num_input_tokens_seen": 3457338,
      "step": 1246,
      "train_runtime": 3241.9796,
      "train_tokens_per_second": 1066.428
    },
    {
      "epoch": 0.10020289680387312,
      "grad_norm": 0.29573866724967957,
      "learning_rate": 0.00011712854757929883,
      "loss": 0.9578,
      "num_input_tokens_seen": 3464012,
      "step": 1247,
      "train_runtime": 3246.4569,
      "train_tokens_per_second": 1067.013
    },
    {
      "epoch": 0.10028325197372386,
      "grad_norm": 0.3642765283584595,
      "learning_rate": 0.0001170617696160267,
      "loss": 0.9672,
      "num_input_tokens_seen": 3466478,
      "step": 1248,
      "train_runtime": 3248.516,
      "train_tokens_per_second": 1067.096
    },
    {
      "epoch": 0.1003636071435746,
      "grad_norm": 0.36443278193473816,
      "learning_rate": 0.0001169949916527546,
      "loss": 0.9012,
      "num_input_tokens_seen": 3468728,
      "step": 1249,
      "train_runtime": 3250.4518,
      "train_tokens_per_second": 1067.153
    },
    {
      "epoch": 0.10044396231342534,
      "grad_norm": 0.3295373320579529,
      "learning_rate": 0.00011692821368948248,
      "loss": 0.8551,
      "num_input_tokens_seen": 3470494,
      "step": 1250,
      "train_runtime": 3252.1408,
      "train_tokens_per_second": 1067.141
    },
    {
      "epoch": 0.10052431748327607,
      "grad_norm": 0.2924972176551819,
      "learning_rate": 0.00011686143572621035,
      "loss": 0.9447,
      "num_input_tokens_seen": 3473070,
      "step": 1251,
      "train_runtime": 3254.2408,
      "train_tokens_per_second": 1067.244
    },
    {
      "epoch": 0.10060467265312682,
      "grad_norm": 0.3317801058292389,
      "learning_rate": 0.00011679465776293823,
      "loss": 1.1554,
      "num_input_tokens_seen": 3476870,
      "step": 1252,
      "train_runtime": 3257.0631,
      "train_tokens_per_second": 1067.486
    },
    {
      "epoch": 0.10068502782297756,
      "grad_norm": 0.29733848571777344,
      "learning_rate": 0.00011672787979966611,
      "loss": 1.2392,
      "num_input_tokens_seen": 3481962,
      "step": 1253,
      "train_runtime": 3260.6026,
      "train_tokens_per_second": 1067.889
    },
    {
      "epoch": 0.1007653829928283,
      "grad_norm": 0.39476510882377625,
      "learning_rate": 0.000116661101836394,
      "loss": 1.059,
      "num_input_tokens_seen": 3483844,
      "step": 1254,
      "train_runtime": 3262.4423,
      "train_tokens_per_second": 1067.864
    },
    {
      "epoch": 0.10084573816267904,
      "grad_norm": 0.34786123037338257,
      "learning_rate": 0.00011659432387312189,
      "loss": 0.8964,
      "num_input_tokens_seen": 3486148,
      "step": 1255,
      "train_runtime": 3264.3911,
      "train_tokens_per_second": 1067.932
    },
    {
      "epoch": 0.10092609333252978,
      "grad_norm": 0.3131040036678314,
      "learning_rate": 0.00011652754590984976,
      "loss": 1.064,
      "num_input_tokens_seen": 3489286,
      "step": 1256,
      "train_runtime": 3266.81,
      "train_tokens_per_second": 1068.102
    },
    {
      "epoch": 0.10100644850238052,
      "grad_norm": 0.3158160150051117,
      "learning_rate": 0.00011646076794657763,
      "loss": 0.8569,
      "num_input_tokens_seen": 3490964,
      "step": 1257,
      "train_runtime": 3268.4627,
      "train_tokens_per_second": 1068.075
    },
    {
      "epoch": 0.10108680367223126,
      "grad_norm": 0.3509765863418579,
      "learning_rate": 0.0001163939899833055,
      "loss": 1.1789,
      "num_input_tokens_seen": 3494808,
      "step": 1258,
      "train_runtime": 3271.2276,
      "train_tokens_per_second": 1068.348
    },
    {
      "epoch": 0.10116715884208201,
      "grad_norm": 0.3125106990337372,
      "learning_rate": 0.0001163272120200334,
      "loss": 0.9231,
      "num_input_tokens_seen": 3498294,
      "step": 1259,
      "train_runtime": 3273.8426,
      "train_tokens_per_second": 1068.559
    },
    {
      "epoch": 0.10124751401193274,
      "grad_norm": 0.3093608319759369,
      "learning_rate": 0.00011626043405676128,
      "loss": 0.6292,
      "num_input_tokens_seen": 3500254,
      "step": 1260,
      "train_runtime": 3275.6331,
      "train_tokens_per_second": 1068.573
    },
    {
      "epoch": 0.10132786918178348,
      "grad_norm": 0.36007463932037354,
      "learning_rate": 0.00011619365609348915,
      "loss": 0.7856,
      "num_input_tokens_seen": 3503508,
      "step": 1261,
      "train_runtime": 3280.1542,
      "train_tokens_per_second": 1068.092
    },
    {
      "epoch": 0.10140822435163423,
      "grad_norm": 0.28798437118530273,
      "learning_rate": 0.00011612687813021703,
      "loss": 0.8916,
      "num_input_tokens_seen": 3505852,
      "step": 1262,
      "train_runtime": 3282.1114,
      "train_tokens_per_second": 1068.17
    },
    {
      "epoch": 0.10148857952148496,
      "grad_norm": 0.2930155098438263,
      "learning_rate": 0.00011606010016694491,
      "loss": 1.3215,
      "num_input_tokens_seen": 3508346,
      "step": 1263,
      "train_runtime": 3284.1582,
      "train_tokens_per_second": 1068.263
    },
    {
      "epoch": 0.10156893469133571,
      "grad_norm": 0.311412513256073,
      "learning_rate": 0.00011599332220367279,
      "loss": 1.1918,
      "num_input_tokens_seen": 3513318,
      "step": 1264,
      "train_runtime": 3287.5156,
      "train_tokens_per_second": 1068.685
    },
    {
      "epoch": 0.10164928986118645,
      "grad_norm": 0.3272484540939331,
      "learning_rate": 0.00011592654424040069,
      "loss": 0.8194,
      "num_input_tokens_seen": 3516784,
      "step": 1265,
      "train_runtime": 3289.9868,
      "train_tokens_per_second": 1068.936
    },
    {
      "epoch": 0.10172964503103718,
      "grad_norm": 0.2773434817790985,
      "learning_rate": 0.00011585976627712856,
      "loss": 1.1294,
      "num_input_tokens_seen": 3519300,
      "step": 1266,
      "train_runtime": 3291.9888,
      "train_tokens_per_second": 1069.05
    },
    {
      "epoch": 0.10181000020088793,
      "grad_norm": 0.33304062485694885,
      "learning_rate": 0.00011579298831385643,
      "loss": 1.092,
      "num_input_tokens_seen": 3521494,
      "step": 1267,
      "train_runtime": 3293.8491,
      "train_tokens_per_second": 1069.112
    },
    {
      "epoch": 0.10189035537073866,
      "grad_norm": 0.2839994430541992,
      "learning_rate": 0.0001157262103505843,
      "loss": 1.1517,
      "num_input_tokens_seen": 3527394,
      "step": 1268,
      "train_runtime": 3297.807,
      "train_tokens_per_second": 1069.618
    },
    {
      "epoch": 0.1019707105405894,
      "grad_norm": 0.2699480950832367,
      "learning_rate": 0.00011565943238731218,
      "loss": 0.9478,
      "num_input_tokens_seen": 3533516,
      "step": 1269,
      "train_runtime": 3301.821,
      "train_tokens_per_second": 1070.172
    },
    {
      "epoch": 0.10205106571044015,
      "grad_norm": 0.3100241422653198,
      "learning_rate": 0.00011559265442404008,
      "loss": 1.0893,
      "num_input_tokens_seen": 3535846,
      "step": 1270,
      "train_runtime": 3303.8417,
      "train_tokens_per_second": 1070.223
    },
    {
      "epoch": 0.10213142088029088,
      "grad_norm": 0.3002788722515106,
      "learning_rate": 0.00011552587646076795,
      "loss": 0.8683,
      "num_input_tokens_seen": 3538008,
      "step": 1271,
      "train_runtime": 3305.749,
      "train_tokens_per_second": 1070.259
    },
    {
      "epoch": 0.10221177605014163,
      "grad_norm": 0.3277871012687683,
      "learning_rate": 0.00011545909849749584,
      "loss": 0.9769,
      "num_input_tokens_seen": 3539180,
      "step": 1272,
      "train_runtime": 3307.2377,
      "train_tokens_per_second": 1070.132
    },
    {
      "epoch": 0.10229213121999237,
      "grad_norm": 0.2547849714756012,
      "learning_rate": 0.00011539232053422371,
      "loss": 0.6298,
      "num_input_tokens_seen": 3541598,
      "step": 1273,
      "train_runtime": 3309.3054,
      "train_tokens_per_second": 1070.194
    },
    {
      "epoch": 0.1023724863898431,
      "grad_norm": 0.28834977746009827,
      "learning_rate": 0.00011532554257095158,
      "loss": 0.9077,
      "num_input_tokens_seen": 3543802,
      "step": 1274,
      "train_runtime": 3311.2111,
      "train_tokens_per_second": 1070.243
    },
    {
      "epoch": 0.10245284155969385,
      "grad_norm": 0.2694496214389801,
      "learning_rate": 0.00011525876460767948,
      "loss": 0.6518,
      "num_input_tokens_seen": 3546678,
      "step": 1275,
      "train_runtime": 3313.4455,
      "train_tokens_per_second": 1070.39
    },
    {
      "epoch": 0.10253319672954458,
      "grad_norm": 0.3871023654937744,
      "learning_rate": 0.00011519198664440736,
      "loss": 1.0278,
      "num_input_tokens_seen": 3548778,
      "step": 1276,
      "train_runtime": 3315.268,
      "train_tokens_per_second": 1070.435
    },
    {
      "epoch": 0.10261355189939533,
      "grad_norm": 0.3189085125923157,
      "learning_rate": 0.00011512520868113523,
      "loss": 0.6991,
      "num_input_tokens_seen": 3551284,
      "step": 1277,
      "train_runtime": 3317.2689,
      "train_tokens_per_second": 1070.545
    },
    {
      "epoch": 0.10269390706924607,
      "grad_norm": 0.2518240511417389,
      "learning_rate": 0.0001150584307178631,
      "loss": 0.9118,
      "num_input_tokens_seen": 3553470,
      "step": 1278,
      "train_runtime": 3319.1756,
      "train_tokens_per_second": 1070.588
    },
    {
      "epoch": 0.1027742622390968,
      "grad_norm": 0.39815038442611694,
      "learning_rate": 0.00011499165275459098,
      "loss": 1.1411,
      "num_input_tokens_seen": 3555536,
      "step": 1279,
      "train_runtime": 3321.0262,
      "train_tokens_per_second": 1070.614
    },
    {
      "epoch": 0.10285461740894755,
      "grad_norm": 0.3301222026348114,
      "learning_rate": 0.00011492487479131886,
      "loss": 0.986,
      "num_input_tokens_seen": 3557820,
      "step": 1280,
      "train_runtime": 3322.9737,
      "train_tokens_per_second": 1070.674
    },
    {
      "epoch": 0.10293497257879829,
      "grad_norm": 0.28770166635513306,
      "learning_rate": 0.00011485809682804675,
      "loss": 0.9974,
      "num_input_tokens_seen": 3560260,
      "step": 1281,
      "train_runtime": 3325.0176,
      "train_tokens_per_second": 1070.749
    },
    {
      "epoch": 0.10301532774864904,
      "grad_norm": 0.2879880964756012,
      "learning_rate": 0.00011479131886477464,
      "loss": 0.562,
      "num_input_tokens_seen": 3562704,
      "step": 1282,
      "train_runtime": 3327.0281,
      "train_tokens_per_second": 1070.837
    },
    {
      "epoch": 0.10309568291849977,
      "grad_norm": 0.3472915291786194,
      "learning_rate": 0.00011472454090150251,
      "loss": 1.2714,
      "num_input_tokens_seen": 3564448,
      "step": 1283,
      "train_runtime": 3328.7455,
      "train_tokens_per_second": 1070.808
    },
    {
      "epoch": 0.1031760380883505,
      "grad_norm": 0.3218541741371155,
      "learning_rate": 0.00011465776293823038,
      "loss": 0.7804,
      "num_input_tokens_seen": 3566114,
      "step": 1284,
      "train_runtime": 3330.3697,
      "train_tokens_per_second": 1070.786
    },
    {
      "epoch": 0.10325639325820125,
      "grad_norm": 0.26079198718070984,
      "learning_rate": 0.00011459098497495826,
      "loss": 0.7276,
      "num_input_tokens_seen": 3569880,
      "step": 1285,
      "train_runtime": 3333.0734,
      "train_tokens_per_second": 1071.048
    },
    {
      "epoch": 0.10333674842805199,
      "grad_norm": 0.32714328169822693,
      "learning_rate": 0.00011452420701168616,
      "loss": 1.0029,
      "num_input_tokens_seen": 3572704,
      "step": 1286,
      "train_runtime": 3335.2715,
      "train_tokens_per_second": 1071.188
    },
    {
      "epoch": 0.10341710359790272,
      "grad_norm": 0.30337613821029663,
      "learning_rate": 0.00011445742904841403,
      "loss": 0.817,
      "num_input_tokens_seen": 3574476,
      "step": 1287,
      "train_runtime": 3336.9605,
      "train_tokens_per_second": 1071.177
    },
    {
      "epoch": 0.10349745876775347,
      "grad_norm": 0.28457576036453247,
      "learning_rate": 0.0001143906510851419,
      "loss": 0.9901,
      "num_input_tokens_seen": 3577610,
      "step": 1288,
      "train_runtime": 3339.2978,
      "train_tokens_per_second": 1071.366
    },
    {
      "epoch": 0.10357781393760421,
      "grad_norm": 0.3802544176578522,
      "learning_rate": 0.00011432387312186979,
      "loss": 1.233,
      "num_input_tokens_seen": 3579984,
      "step": 1289,
      "train_runtime": 3341.3605,
      "train_tokens_per_second": 1071.415
    },
    {
      "epoch": 0.10365816910745496,
      "grad_norm": 0.4489324688911438,
      "learning_rate": 0.00011425709515859766,
      "loss": 1.1114,
      "num_input_tokens_seen": 3582412,
      "step": 1290,
      "train_runtime": 3343.2872,
      "train_tokens_per_second": 1071.524
    },
    {
      "epoch": 0.10373852427730569,
      "grad_norm": 0.2681956887245178,
      "learning_rate": 0.00011419031719532556,
      "loss": 0.6312,
      "num_input_tokens_seen": 3584672,
      "step": 1291,
      "train_runtime": 3347.2626,
      "train_tokens_per_second": 1070.926
    },
    {
      "epoch": 0.10381887944715643,
      "grad_norm": 0.3160945177078247,
      "learning_rate": 0.00011412353923205344,
      "loss": 0.8936,
      "num_input_tokens_seen": 3587328,
      "step": 1292,
      "train_runtime": 3349.3113,
      "train_tokens_per_second": 1071.064
    },
    {
      "epoch": 0.10389923461700717,
      "grad_norm": 0.3134382963180542,
      "learning_rate": 0.00011405676126878131,
      "loss": 1.1508,
      "num_input_tokens_seen": 3592166,
      "step": 1293,
      "train_runtime": 3352.6762,
      "train_tokens_per_second": 1071.432
    },
    {
      "epoch": 0.10397958978685791,
      "grad_norm": 0.32309138774871826,
      "learning_rate": 0.00011398998330550918,
      "loss": 1.1359,
      "num_input_tokens_seen": 3593964,
      "step": 1294,
      "train_runtime": 3354.4147,
      "train_tokens_per_second": 1071.413
    },
    {
      "epoch": 0.10405994495670866,
      "grad_norm": 0.32670146226882935,
      "learning_rate": 0.00011392320534223706,
      "loss": 1.0211,
      "num_input_tokens_seen": 3595924,
      "step": 1295,
      "train_runtime": 3356.2253,
      "train_tokens_per_second": 1071.419
    },
    {
      "epoch": 0.10414030012655939,
      "grad_norm": 0.2928754389286041,
      "learning_rate": 0.00011385642737896493,
      "loss": 1.0102,
      "num_input_tokens_seen": 3598434,
      "step": 1296,
      "train_runtime": 3358.2748,
      "train_tokens_per_second": 1071.513
    },
    {
      "epoch": 0.10422065529641013,
      "grad_norm": 0.39643630385398865,
      "learning_rate": 0.00011378964941569283,
      "loss": 0.968,
      "num_input_tokens_seen": 3601478,
      "step": 1297,
      "train_runtime": 3360.6429,
      "train_tokens_per_second": 1071.663
    },
    {
      "epoch": 0.10430101046626088,
      "grad_norm": 0.5213744044303894,
      "learning_rate": 0.0001137228714524207,
      "loss": 1.014,
      "num_input_tokens_seen": 3603918,
      "step": 1298,
      "train_runtime": 3362.7224,
      "train_tokens_per_second": 1071.726
    },
    {
      "epoch": 0.10438136563611161,
      "grad_norm": 0.23302114009857178,
      "learning_rate": 0.00011365609348914859,
      "loss": 0.8334,
      "num_input_tokens_seen": 3607438,
      "step": 1299,
      "train_runtime": 3365.3521,
      "train_tokens_per_second": 1071.935
    },
    {
      "epoch": 0.10446172080596236,
      "grad_norm": 0.39406678080558777,
      "learning_rate": 0.00011358931552587646,
      "loss": 1.1717,
      "num_input_tokens_seen": 3609654,
      "step": 1300,
      "train_runtime": 3367.2601,
      "train_tokens_per_second": 1071.986
    },
    {
      "epoch": 0.1045420759758131,
      "grad_norm": 0.2761419117450714,
      "learning_rate": 0.00011352253756260434,
      "loss": 1.004,
      "num_input_tokens_seen": 3613216,
      "step": 1301,
      "train_runtime": 3369.8443,
      "train_tokens_per_second": 1072.22
    },
    {
      "epoch": 0.10462243114566383,
      "grad_norm": 0.3041739761829376,
      "learning_rate": 0.00011345575959933224,
      "loss": 0.8539,
      "num_input_tokens_seen": 3615684,
      "step": 1302,
      "train_runtime": 3371.9426,
      "train_tokens_per_second": 1072.285
    },
    {
      "epoch": 0.10470278631551458,
      "grad_norm": 0.24297115206718445,
      "learning_rate": 0.00011338898163606011,
      "loss": 0.9024,
      "num_input_tokens_seen": 3618762,
      "step": 1303,
      "train_runtime": 3374.1865,
      "train_tokens_per_second": 1072.484
    },
    {
      "epoch": 0.10478314148536531,
      "grad_norm": 0.30591264367103577,
      "learning_rate": 0.00011332220367278798,
      "loss": 1.3905,
      "num_input_tokens_seen": 3620824,
      "step": 1304,
      "train_runtime": 3375.9352,
      "train_tokens_per_second": 1072.54
    },
    {
      "epoch": 0.10486349665521605,
      "grad_norm": 0.306010365486145,
      "learning_rate": 0.00011325542570951586,
      "loss": 0.8382,
      "num_input_tokens_seen": 3623304,
      "step": 1305,
      "train_runtime": 3378.0611,
      "train_tokens_per_second": 1072.599
    },
    {
      "epoch": 0.1049438518250668,
      "grad_norm": 0.3153892755508423,
      "learning_rate": 0.00011318864774624374,
      "loss": 1.1346,
      "num_input_tokens_seen": 3624926,
      "step": 1306,
      "train_runtime": 3379.6914,
      "train_tokens_per_second": 1072.561
    },
    {
      "epoch": 0.10502420699491753,
      "grad_norm": 0.35114431381225586,
      "learning_rate": 0.00011312186978297163,
      "loss": 1.0101,
      "num_input_tokens_seen": 3627426,
      "step": 1307,
      "train_runtime": 3381.8135,
      "train_tokens_per_second": 1072.627
    },
    {
      "epoch": 0.10510456216476828,
      "grad_norm": 0.3883284628391266,
      "learning_rate": 0.00011305509181969952,
      "loss": 1.0112,
      "num_input_tokens_seen": 3629386,
      "step": 1308,
      "train_runtime": 3383.6357,
      "train_tokens_per_second": 1072.629
    },
    {
      "epoch": 0.10518491733461902,
      "grad_norm": 0.41548824310302734,
      "learning_rate": 0.00011298831385642739,
      "loss": 0.9458,
      "num_input_tokens_seen": 3631680,
      "step": 1309,
      "train_runtime": 3385.5591,
      "train_tokens_per_second": 1072.697
    },
    {
      "epoch": 0.10526527250446975,
      "grad_norm": 0.31654754281044006,
      "learning_rate": 0.00011292153589315526,
      "loss": 0.986,
      "num_input_tokens_seen": 3633942,
      "step": 1310,
      "train_runtime": 3387.4288,
      "train_tokens_per_second": 1072.773
    },
    {
      "epoch": 0.1053456276743205,
      "grad_norm": 0.2752566933631897,
      "learning_rate": 0.00011285475792988314,
      "loss": 0.7523,
      "num_input_tokens_seen": 3635718,
      "step": 1311,
      "train_runtime": 3389.1385,
      "train_tokens_per_second": 1072.756
    },
    {
      "epoch": 0.10542598284417123,
      "grad_norm": 0.32447850704193115,
      "learning_rate": 0.00011278797996661104,
      "loss": 1.0836,
      "num_input_tokens_seen": 3637488,
      "step": 1312,
      "train_runtime": 3390.8198,
      "train_tokens_per_second": 1072.746
    },
    {
      "epoch": 0.10550633801402198,
      "grad_norm": 0.22222858667373657,
      "learning_rate": 0.00011272120200333891,
      "loss": 0.5156,
      "num_input_tokens_seen": 3639824,
      "step": 1313,
      "train_runtime": 3392.6896,
      "train_tokens_per_second": 1072.843
    },
    {
      "epoch": 0.10558669318387272,
      "grad_norm": 0.27509626746177673,
      "learning_rate": 0.00011265442404006678,
      "loss": 1.0833,
      "num_input_tokens_seen": 3642780,
      "step": 1314,
      "train_runtime": 3395.0259,
      "train_tokens_per_second": 1072.976
    },
    {
      "epoch": 0.10566704835372345,
      "grad_norm": 0.3218839764595032,
      "learning_rate": 0.00011258764607679465,
      "loss": 1.1377,
      "num_input_tokens_seen": 3645864,
      "step": 1315,
      "train_runtime": 3397.3436,
      "train_tokens_per_second": 1073.151
    },
    {
      "epoch": 0.1057474035235742,
      "grad_norm": 0.3075161576271057,
      "learning_rate": 0.00011252086811352254,
      "loss": 1.0003,
      "num_input_tokens_seen": 3647748,
      "step": 1316,
      "train_runtime": 3399.1268,
      "train_tokens_per_second": 1073.143
    },
    {
      "epoch": 0.10582775869342494,
      "grad_norm": 0.33633488416671753,
      "learning_rate": 0.00011245409015025041,
      "loss": 0.9582,
      "num_input_tokens_seen": 3649650,
      "step": 1317,
      "train_runtime": 3400.9247,
      "train_tokens_per_second": 1073.135
    },
    {
      "epoch": 0.10590811386327569,
      "grad_norm": 0.28925010561943054,
      "learning_rate": 0.00011238731218697832,
      "loss": 0.7256,
      "num_input_tokens_seen": 3651120,
      "step": 1318,
      "train_runtime": 3402.5544,
      "train_tokens_per_second": 1073.053
    },
    {
      "epoch": 0.10598846903312642,
      "grad_norm": 0.24827858805656433,
      "learning_rate": 0.00011232053422370619,
      "loss": 0.599,
      "num_input_tokens_seen": 3652890,
      "step": 1319,
      "train_runtime": 3404.2269,
      "train_tokens_per_second": 1073.045
    },
    {
      "epoch": 0.10606882420297715,
      "grad_norm": 0.384735643863678,
      "learning_rate": 0.00011225375626043406,
      "loss": 0.9334,
      "num_input_tokens_seen": 3654918,
      "step": 1320,
      "train_runtime": 3406.0906,
      "train_tokens_per_second": 1073.054
    },
    {
      "epoch": 0.1061491793728279,
      "grad_norm": 0.28056448698043823,
      "learning_rate": 0.00011218697829716193,
      "loss": 1.1756,
      "num_input_tokens_seen": 3657194,
      "step": 1321,
      "train_runtime": 3410.0353,
      "train_tokens_per_second": 1072.48
    },
    {
      "epoch": 0.10622953454267864,
      "grad_norm": 0.33696311712265015,
      "learning_rate": 0.00011212020033388981,
      "loss": 0.977,
      "num_input_tokens_seen": 3658668,
      "step": 1322,
      "train_runtime": 3411.6571,
      "train_tokens_per_second": 1072.402
    },
    {
      "epoch": 0.10630988971252937,
      "grad_norm": 0.34870147705078125,
      "learning_rate": 0.00011205342237061771,
      "loss": 1.141,
      "num_input_tokens_seen": 3660450,
      "step": 1323,
      "train_runtime": 3413.4146,
      "train_tokens_per_second": 1072.372
    },
    {
      "epoch": 0.10639024488238012,
      "grad_norm": 0.27133920788764954,
      "learning_rate": 0.00011198664440734558,
      "loss": 0.8748,
      "num_input_tokens_seen": 3662052,
      "step": 1324,
      "train_runtime": 3415.0366,
      "train_tokens_per_second": 1072.332
    },
    {
      "epoch": 0.10647060005223086,
      "grad_norm": 0.41314199566841125,
      "learning_rate": 0.00011191986644407347,
      "loss": 1.1214,
      "num_input_tokens_seen": 3664276,
      "step": 1325,
      "train_runtime": 3416.9565,
      "train_tokens_per_second": 1072.38
    },
    {
      "epoch": 0.1065509552220816,
      "grad_norm": 0.3371661305427551,
      "learning_rate": 0.00011185308848080134,
      "loss": 1.1208,
      "num_input_tokens_seen": 3667412,
      "step": 1326,
      "train_runtime": 3419.4142,
      "train_tokens_per_second": 1072.526
    },
    {
      "epoch": 0.10663131039193234,
      "grad_norm": 0.3550303876399994,
      "learning_rate": 0.00011178631051752921,
      "loss": 0.9646,
      "num_input_tokens_seen": 3668990,
      "step": 1327,
      "train_runtime": 3421.0603,
      "train_tokens_per_second": 1072.472
    },
    {
      "epoch": 0.10671166556178308,
      "grad_norm": 0.31260353326797485,
      "learning_rate": 0.00011171953255425711,
      "loss": 1.01,
      "num_input_tokens_seen": 3670524,
      "step": 1328,
      "train_runtime": 3422.6944,
      "train_tokens_per_second": 1072.408
    },
    {
      "epoch": 0.10679202073163382,
      "grad_norm": 0.2847209870815277,
      "learning_rate": 0.00011165275459098499,
      "loss": 1.0297,
      "num_input_tokens_seen": 3672938,
      "step": 1329,
      "train_runtime": 3424.769,
      "train_tokens_per_second": 1072.463
    },
    {
      "epoch": 0.10687237590148456,
      "grad_norm": 0.3199746012687683,
      "learning_rate": 0.00011158597662771286,
      "loss": 1.0204,
      "num_input_tokens_seen": 3675080,
      "step": 1330,
      "train_runtime": 3426.7408,
      "train_tokens_per_second": 1072.471
    },
    {
      "epoch": 0.10695273107133531,
      "grad_norm": 0.3620395064353943,
      "learning_rate": 0.00011151919866444073,
      "loss": 0.7482,
      "num_input_tokens_seen": 3676924,
      "step": 1331,
      "train_runtime": 3428.4531,
      "train_tokens_per_second": 1072.473
    },
    {
      "epoch": 0.10703308624118604,
      "grad_norm": 0.453721284866333,
      "learning_rate": 0.00011145242070116862,
      "loss": 1.1147,
      "num_input_tokens_seen": 3679230,
      "step": 1332,
      "train_runtime": 3430.5278,
      "train_tokens_per_second": 1072.497
    },
    {
      "epoch": 0.10711344141103678,
      "grad_norm": 0.278247594833374,
      "learning_rate": 0.0001113856427378965,
      "loss": 0.7915,
      "num_input_tokens_seen": 3682504,
      "step": 1333,
      "train_runtime": 3433.3317,
      "train_tokens_per_second": 1072.575
    },
    {
      "epoch": 0.10719379658088753,
      "grad_norm": 0.2761545181274414,
      "learning_rate": 0.0001113188647746244,
      "loss": 0.7856,
      "num_input_tokens_seen": 3684416,
      "step": 1334,
      "train_runtime": 3435.074,
      "train_tokens_per_second": 1072.587
    },
    {
      "epoch": 0.10727415175073826,
      "grad_norm": 0.2801143229007721,
      "learning_rate": 0.00011125208681135227,
      "loss": 1.1772,
      "num_input_tokens_seen": 3687104,
      "step": 1335,
      "train_runtime": 3437.1997,
      "train_tokens_per_second": 1072.706
    },
    {
      "epoch": 0.10735450692058901,
      "grad_norm": 0.393561452627182,
      "learning_rate": 0.00011118530884808014,
      "loss": 1.2837,
      "num_input_tokens_seen": 3689000,
      "step": 1336,
      "train_runtime": 3438.9737,
      "train_tokens_per_second": 1072.704
    },
    {
      "epoch": 0.10743486209043974,
      "grad_norm": 0.34230685234069824,
      "learning_rate": 0.00011111853088480801,
      "loss": 1.2398,
      "num_input_tokens_seen": 3692920,
      "step": 1337,
      "train_runtime": 3441.8685,
      "train_tokens_per_second": 1072.94
    },
    {
      "epoch": 0.10751521726029048,
      "grad_norm": 0.20867261290550232,
      "learning_rate": 0.00011105175292153589,
      "loss": 0.6112,
      "num_input_tokens_seen": 3695068,
      "step": 1338,
      "train_runtime": 3443.8284,
      "train_tokens_per_second": 1072.954
    },
    {
      "epoch": 0.10759557243014123,
      "grad_norm": 0.3125673234462738,
      "learning_rate": 0.00011098497495826379,
      "loss": 0.9997,
      "num_input_tokens_seen": 3697470,
      "step": 1339,
      "train_runtime": 3445.8147,
      "train_tokens_per_second": 1073.032
    },
    {
      "epoch": 0.10767592759999196,
      "grad_norm": 0.29269737005233765,
      "learning_rate": 0.00011091819699499166,
      "loss": 1.2169,
      "num_input_tokens_seen": 3699864,
      "step": 1340,
      "train_runtime": 3447.7964,
      "train_tokens_per_second": 1073.11
    },
    {
      "epoch": 0.1077562827698427,
      "grad_norm": 0.37325942516326904,
      "learning_rate": 0.00011085141903171953,
      "loss": 1.3217,
      "num_input_tokens_seen": 3703614,
      "step": 1341,
      "train_runtime": 3450.5096,
      "train_tokens_per_second": 1073.353
    },
    {
      "epoch": 0.10783663793969345,
      "grad_norm": 0.29524070024490356,
      "learning_rate": 0.00011078464106844742,
      "loss": 0.8587,
      "num_input_tokens_seen": 3705338,
      "step": 1342,
      "train_runtime": 3452.2589,
      "train_tokens_per_second": 1073.308
    },
    {
      "epoch": 0.10791699310954418,
      "grad_norm": 0.40680551528930664,
      "learning_rate": 0.00011071786310517529,
      "loss": 1.1864,
      "num_input_tokens_seen": 3708598,
      "step": 1343,
      "train_runtime": 3454.7802,
      "train_tokens_per_second": 1073.469
    },
    {
      "epoch": 0.10799734827939493,
      "grad_norm": 0.3449890911579132,
      "learning_rate": 0.0001106510851419032,
      "loss": 1.0204,
      "num_input_tokens_seen": 3712782,
      "step": 1344,
      "train_runtime": 3457.7119,
      "train_tokens_per_second": 1073.768
    },
    {
      "epoch": 0.10807770344924567,
      "grad_norm": 0.26540616154670715,
      "learning_rate": 0.00011058430717863107,
      "loss": 0.8519,
      "num_input_tokens_seen": 3715000,
      "step": 1345,
      "train_runtime": 3459.6408,
      "train_tokens_per_second": 1073.811
    },
    {
      "epoch": 0.1081580586190964,
      "grad_norm": 0.3037562966346741,
      "learning_rate": 0.00011051752921535894,
      "loss": 1.0007,
      "num_input_tokens_seen": 3718230,
      "step": 1346,
      "train_runtime": 3462.0935,
      "train_tokens_per_second": 1073.983
    },
    {
      "epoch": 0.10823841378894715,
      "grad_norm": 0.33242589235305786,
      "learning_rate": 0.00011045075125208681,
      "loss": 1.0404,
      "num_input_tokens_seen": 3720778,
      "step": 1347,
      "train_runtime": 3464.1848,
      "train_tokens_per_second": 1074.07
    },
    {
      "epoch": 0.10831876895879788,
      "grad_norm": 0.32828032970428467,
      "learning_rate": 0.00011038397328881469,
      "loss": 1.1201,
      "num_input_tokens_seen": 3722980,
      "step": 1348,
      "train_runtime": 3466.0775,
      "train_tokens_per_second": 1074.119
    },
    {
      "epoch": 0.10839912412864863,
      "grad_norm": 0.3383139967918396,
      "learning_rate": 0.00011031719532554257,
      "loss": 1.1004,
      "num_input_tokens_seen": 3726764,
      "step": 1349,
      "train_runtime": 3468.925,
      "train_tokens_per_second": 1074.328
    },
    {
      "epoch": 0.10847947929849937,
      "grad_norm": 0.26566532254219055,
      "learning_rate": 0.00011025041736227046,
      "loss": 0.9865,
      "num_input_tokens_seen": 3731318,
      "step": 1350,
      "train_runtime": 3472.0228,
      "train_tokens_per_second": 1074.681
    },
    {
      "epoch": 0.1085598344683501,
      "grad_norm": 0.27251118421554565,
      "learning_rate": 0.00011018363939899835,
      "loss": 0.9114,
      "num_input_tokens_seen": 3734166,
      "step": 1351,
      "train_runtime": 3476.1917,
      "train_tokens_per_second": 1074.212
    },
    {
      "epoch": 0.10864018963820085,
      "grad_norm": 0.3204699456691742,
      "learning_rate": 0.00011011686143572622,
      "loss": 1.0144,
      "num_input_tokens_seen": 3736466,
      "step": 1352,
      "train_runtime": 3478.162,
      "train_tokens_per_second": 1074.264
    },
    {
      "epoch": 0.10872054480805159,
      "grad_norm": 0.37815171480178833,
      "learning_rate": 0.00011005008347245409,
      "loss": 1.2469,
      "num_input_tokens_seen": 3738370,
      "step": 1353,
      "train_runtime": 3479.8996,
      "train_tokens_per_second": 1074.275
    },
    {
      "epoch": 0.10880089997790234,
      "grad_norm": 0.3219520151615143,
      "learning_rate": 0.00010998330550918197,
      "loss": 1.2015,
      "num_input_tokens_seen": 3741902,
      "step": 1354,
      "train_runtime": 3482.5231,
      "train_tokens_per_second": 1074.48
    },
    {
      "epoch": 0.10888125514775307,
      "grad_norm": 0.27995145320892334,
      "learning_rate": 0.00010991652754590987,
      "loss": 1.1808,
      "num_input_tokens_seen": 3745614,
      "step": 1355,
      "train_runtime": 3485.2325,
      "train_tokens_per_second": 1074.71
    },
    {
      "epoch": 0.1089616103176038,
      "grad_norm": 0.3314412236213684,
      "learning_rate": 0.00010984974958263774,
      "loss": 0.995,
      "num_input_tokens_seen": 3749444,
      "step": 1356,
      "train_runtime": 3487.9662,
      "train_tokens_per_second": 1074.966
    },
    {
      "epoch": 0.10904196548745455,
      "grad_norm": 0.39481425285339355,
      "learning_rate": 0.00010978297161936561,
      "loss": 1.0378,
      "num_input_tokens_seen": 3750840,
      "step": 1357,
      "train_runtime": 3489.5382,
      "train_tokens_per_second": 1074.881
    },
    {
      "epoch": 0.10912232065730529,
      "grad_norm": 0.3497794270515442,
      "learning_rate": 0.00010971619365609349,
      "loss": 1.1416,
      "num_input_tokens_seen": 3752454,
      "step": 1358,
      "train_runtime": 3491.1985,
      "train_tokens_per_second": 1074.833
    },
    {
      "epoch": 0.10920267582715602,
      "grad_norm": 0.3231332004070282,
      "learning_rate": 0.00010964941569282137,
      "loss": 0.8696,
      "num_input_tokens_seen": 3753972,
      "step": 1359,
      "train_runtime": 3492.8057,
      "train_tokens_per_second": 1074.773
    },
    {
      "epoch": 0.10928303099700677,
      "grad_norm": 0.250036358833313,
      "learning_rate": 0.00010958263772954926,
      "loss": 1.001,
      "num_input_tokens_seen": 3756846,
      "step": 1360,
      "train_runtime": 3495.0693,
      "train_tokens_per_second": 1074.899
    },
    {
      "epoch": 0.1093633861668575,
      "grad_norm": 0.29247865080833435,
      "learning_rate": 0.00010951585976627715,
      "loss": 0.9141,
      "num_input_tokens_seen": 3759354,
      "step": 1361,
      "train_runtime": 3497.0523,
      "train_tokens_per_second": 1075.007
    },
    {
      "epoch": 0.10944374133670826,
      "grad_norm": 0.2691769003868103,
      "learning_rate": 0.00010944908180300502,
      "loss": 0.996,
      "num_input_tokens_seen": 3762132,
      "step": 1362,
      "train_runtime": 3499.2741,
      "train_tokens_per_second": 1075.118
    },
    {
      "epoch": 0.10952409650655899,
      "grad_norm": 0.42689990997314453,
      "learning_rate": 0.00010938230383973289,
      "loss": 0.9984,
      "num_input_tokens_seen": 3765012,
      "step": 1363,
      "train_runtime": 3501.5444,
      "train_tokens_per_second": 1075.243
    },
    {
      "epoch": 0.10960445167640973,
      "grad_norm": 0.28335949778556824,
      "learning_rate": 0.00010931552587646076,
      "loss": 0.7242,
      "num_input_tokens_seen": 3766120,
      "step": 1364,
      "train_runtime": 3503.0292,
      "train_tokens_per_second": 1075.104
    },
    {
      "epoch": 0.10968480684626047,
      "grad_norm": 0.29798826575279236,
      "learning_rate": 0.00010924874791318864,
      "loss": 0.774,
      "num_input_tokens_seen": 3768304,
      "step": 1365,
      "train_runtime": 3504.8183,
      "train_tokens_per_second": 1075.178
    },
    {
      "epoch": 0.10976516201611121,
      "grad_norm": 0.2864566445350647,
      "learning_rate": 0.00010918196994991654,
      "loss": 0.8877,
      "num_input_tokens_seen": 3772924,
      "step": 1366,
      "train_runtime": 3508.0897,
      "train_tokens_per_second": 1075.492
    },
    {
      "epoch": 0.10984551718596196,
      "grad_norm": 0.3073413372039795,
      "learning_rate": 0.00010911519198664441,
      "loss": 1.0691,
      "num_input_tokens_seen": 3775548,
      "step": 1367,
      "train_runtime": 3510.2155,
      "train_tokens_per_second": 1075.589
    },
    {
      "epoch": 0.10992587235581269,
      "grad_norm": 0.29954490065574646,
      "learning_rate": 0.0001090484140233723,
      "loss": 1.0633,
      "num_input_tokens_seen": 3777364,
      "step": 1368,
      "train_runtime": 3511.9896,
      "train_tokens_per_second": 1075.562
    },
    {
      "epoch": 0.11000622752566343,
      "grad_norm": 0.28042760491371155,
      "learning_rate": 0.00010898163606010017,
      "loss": 1.0094,
      "num_input_tokens_seen": 3779544,
      "step": 1369,
      "train_runtime": 3513.9366,
      "train_tokens_per_second": 1075.587
    },
    {
      "epoch": 0.11008658269551418,
      "grad_norm": 0.3415967524051666,
      "learning_rate": 0.00010891485809682804,
      "loss": 1.0676,
      "num_input_tokens_seen": 3782884,
      "step": 1370,
      "train_runtime": 3516.366,
      "train_tokens_per_second": 1075.794
    },
    {
      "epoch": 0.11016693786536491,
      "grad_norm": 0.2940909266471863,
      "learning_rate": 0.00010884808013355594,
      "loss": 1.3355,
      "num_input_tokens_seen": 3785772,
      "step": 1371,
      "train_runtime": 3518.5452,
      "train_tokens_per_second": 1075.948
    },
    {
      "epoch": 0.11024729303521566,
      "grad_norm": 0.29695481061935425,
      "learning_rate": 0.00010878130217028382,
      "loss": 0.8361,
      "num_input_tokens_seen": 3787272,
      "step": 1372,
      "train_runtime": 3520.1921,
      "train_tokens_per_second": 1075.871
    },
    {
      "epoch": 0.1103276482050664,
      "grad_norm": 0.3094223141670227,
      "learning_rate": 0.00010871452420701169,
      "loss": 1.0499,
      "num_input_tokens_seen": 3790462,
      "step": 1373,
      "train_runtime": 3522.6757,
      "train_tokens_per_second": 1076.018
    },
    {
      "epoch": 0.11040800337491713,
      "grad_norm": 0.2748089134693146,
      "learning_rate": 0.00010864774624373956,
      "loss": 0.7391,
      "num_input_tokens_seen": 3792374,
      "step": 1374,
      "train_runtime": 3524.4285,
      "train_tokens_per_second": 1076.025
    },
    {
      "epoch": 0.11048835854476788,
      "grad_norm": 0.4110908806324005,
      "learning_rate": 0.00010858096828046744,
      "loss": 1.0267,
      "num_input_tokens_seen": 3793806,
      "step": 1375,
      "train_runtime": 3525.9834,
      "train_tokens_per_second": 1075.957
    },
    {
      "epoch": 0.11056871371461861,
      "grad_norm": 0.4627639055252075,
      "learning_rate": 0.00010851419031719534,
      "loss": 0.9407,
      "num_input_tokens_seen": 3797124,
      "step": 1376,
      "train_runtime": 3528.5315,
      "train_tokens_per_second": 1076.12
    },
    {
      "epoch": 0.11064906888446935,
      "grad_norm": 0.3095523416996002,
      "learning_rate": 0.00010844741235392321,
      "loss": 1.2738,
      "num_input_tokens_seen": 3799874,
      "step": 1377,
      "train_runtime": 3530.5832,
      "train_tokens_per_second": 1076.274
    },
    {
      "epoch": 0.1107294240543201,
      "grad_norm": 0.2851976454257965,
      "learning_rate": 0.0001083806343906511,
      "loss": 0.9193,
      "num_input_tokens_seen": 3803560,
      "step": 1378,
      "train_runtime": 3533.2815,
      "train_tokens_per_second": 1076.495
    },
    {
      "epoch": 0.11080977922417083,
      "grad_norm": 0.33301135897636414,
      "learning_rate": 0.00010831385642737897,
      "loss": 0.859,
      "num_input_tokens_seen": 3809396,
      "step": 1379,
      "train_runtime": 3537.2705,
      "train_tokens_per_second": 1076.931
    },
    {
      "epoch": 0.11089013439402158,
      "grad_norm": 0.33650487661361694,
      "learning_rate": 0.00010824707846410684,
      "loss": 1.0981,
      "num_input_tokens_seen": 3811146,
      "step": 1380,
      "train_runtime": 3539.0327,
      "train_tokens_per_second": 1076.889
    },
    {
      "epoch": 0.11097048956387232,
      "grad_norm": 0.35976412892341614,
      "learning_rate": 0.00010818030050083472,
      "loss": 0.8029,
      "num_input_tokens_seen": 3812980,
      "step": 1381,
      "train_runtime": 3542.4727,
      "train_tokens_per_second": 1076.361
    },
    {
      "epoch": 0.11105084473372305,
      "grad_norm": 0.31813323497772217,
      "learning_rate": 0.00010811352253756262,
      "loss": 0.7812,
      "num_input_tokens_seen": 3815602,
      "step": 1382,
      "train_runtime": 3544.5753,
      "train_tokens_per_second": 1076.462
    },
    {
      "epoch": 0.1111311999035738,
      "grad_norm": 0.34927958250045776,
      "learning_rate": 0.00010804674457429049,
      "loss": 0.8661,
      "num_input_tokens_seen": 3817980,
      "step": 1383,
      "train_runtime": 3546.5334,
      "train_tokens_per_second": 1076.539
    },
    {
      "epoch": 0.11121155507342453,
      "grad_norm": 0.23879820108413696,
      "learning_rate": 0.00010797996661101836,
      "loss": 0.8932,
      "num_input_tokens_seen": 3821382,
      "step": 1384,
      "train_runtime": 3549.0931,
      "train_tokens_per_second": 1076.721
    },
    {
      "epoch": 0.11129191024327528,
      "grad_norm": 0.3335447311401367,
      "learning_rate": 0.00010791318864774625,
      "loss": 1.2005,
      "num_input_tokens_seen": 3824918,
      "step": 1385,
      "train_runtime": 3551.658,
      "train_tokens_per_second": 1076.939
    },
    {
      "epoch": 0.11137226541312602,
      "grad_norm": 0.2883974015712738,
      "learning_rate": 0.00010784641068447412,
      "loss": 1.1426,
      "num_input_tokens_seen": 3827574,
      "step": 1386,
      "train_runtime": 3553.7373,
      "train_tokens_per_second": 1077.056
    },
    {
      "epoch": 0.11145262058297675,
      "grad_norm": 0.2518156170845032,
      "learning_rate": 0.00010777963272120202,
      "loss": 0.7465,
      "num_input_tokens_seen": 3829382,
      "step": 1387,
      "train_runtime": 3555.4441,
      "train_tokens_per_second": 1077.047
    },
    {
      "epoch": 0.1115329757528275,
      "grad_norm": 0.3041207194328308,
      "learning_rate": 0.0001077128547579299,
      "loss": 1.0229,
      "num_input_tokens_seen": 3833380,
      "step": 1388,
      "train_runtime": 3558.254,
      "train_tokens_per_second": 1077.321
    },
    {
      "epoch": 0.11161333092267824,
      "grad_norm": 0.217087060213089,
      "learning_rate": 0.00010764607679465777,
      "loss": 0.5558,
      "num_input_tokens_seen": 3836120,
      "step": 1389,
      "train_runtime": 3560.364,
      "train_tokens_per_second": 1077.452
    },
    {
      "epoch": 0.11169368609252898,
      "grad_norm": 0.2774825096130371,
      "learning_rate": 0.00010757929883138564,
      "loss": 1.1133,
      "num_input_tokens_seen": 3839662,
      "step": 1390,
      "train_runtime": 3562.8997,
      "train_tokens_per_second": 1077.679
    },
    {
      "epoch": 0.11177404126237972,
      "grad_norm": 0.3661069869995117,
      "learning_rate": 0.00010751252086811352,
      "loss": 1.2426,
      "num_input_tokens_seen": 3842038,
      "step": 1391,
      "train_runtime": 3564.8522,
      "train_tokens_per_second": 1077.755
    },
    {
      "epoch": 0.11185439643223045,
      "grad_norm": 0.29887625575065613,
      "learning_rate": 0.00010744574290484142,
      "loss": 0.8912,
      "num_input_tokens_seen": 3846600,
      "step": 1392,
      "train_runtime": 3568.1821,
      "train_tokens_per_second": 1078.028
    },
    {
      "epoch": 0.1119347516020812,
      "grad_norm": 0.2846955955028534,
      "learning_rate": 0.00010737896494156929,
      "loss": 0.9515,
      "num_input_tokens_seen": 3849338,
      "step": 1393,
      "train_runtime": 3570.3948,
      "train_tokens_per_second": 1078.127
    },
    {
      "epoch": 0.11201510677193194,
      "grad_norm": 0.30389881134033203,
      "learning_rate": 0.00010731218697829716,
      "loss": 0.9121,
      "num_input_tokens_seen": 3851292,
      "step": 1394,
      "train_runtime": 3572.2378,
      "train_tokens_per_second": 1078.117
    },
    {
      "epoch": 0.11209546194178267,
      "grad_norm": 0.2863844633102417,
      "learning_rate": 0.00010724540901502505,
      "loss": 1.0296,
      "num_input_tokens_seen": 3854908,
      "step": 1395,
      "train_runtime": 3574.8935,
      "train_tokens_per_second": 1078.328
    },
    {
      "epoch": 0.11217581711163342,
      "grad_norm": 0.314884215593338,
      "learning_rate": 0.00010717863105175292,
      "loss": 0.836,
      "num_input_tokens_seen": 3858098,
      "step": 1396,
      "train_runtime": 3577.2419,
      "train_tokens_per_second": 1078.512
    },
    {
      "epoch": 0.11225617228148416,
      "grad_norm": 0.28291910886764526,
      "learning_rate": 0.0001071118530884808,
      "loss": 1.1745,
      "num_input_tokens_seen": 3860180,
      "step": 1397,
      "train_runtime": 3579.0546,
      "train_tokens_per_second": 1078.547
    },
    {
      "epoch": 0.1123365274513349,
      "grad_norm": 0.29328039288520813,
      "learning_rate": 0.0001070450751252087,
      "loss": 1.0608,
      "num_input_tokens_seen": 3862534,
      "step": 1398,
      "train_runtime": 3580.9789,
      "train_tokens_per_second": 1078.625
    },
    {
      "epoch": 0.11241688262118564,
      "grad_norm": 0.3768022358417511,
      "learning_rate": 0.00010697829716193657,
      "loss": 0.9091,
      "num_input_tokens_seen": 3864172,
      "step": 1399,
      "train_runtime": 3582.6296,
      "train_tokens_per_second": 1078.585
    },
    {
      "epoch": 0.11249723779103638,
      "grad_norm": 0.33518317341804504,
      "learning_rate": 0.00010691151919866444,
      "loss": 0.906,
      "num_input_tokens_seen": 3866808,
      "step": 1400,
      "train_runtime": 3584.7371,
      "train_tokens_per_second": 1078.687
    },
    {
      "epoch": 0.11257759296088712,
      "grad_norm": 0.2500336766242981,
      "learning_rate": 0.00010684474123539232,
      "loss": 0.7949,
      "num_input_tokens_seen": 3869754,
      "step": 1401,
      "train_runtime": 3587.0515,
      "train_tokens_per_second": 1078.812
    },
    {
      "epoch": 0.11265794813073786,
      "grad_norm": 0.3795786201953888,
      "learning_rate": 0.0001067779632721202,
      "loss": 1.1297,
      "num_input_tokens_seen": 3871500,
      "step": 1402,
      "train_runtime": 3588.6919,
      "train_tokens_per_second": 1078.805
    },
    {
      "epoch": 0.11273830330058861,
      "grad_norm": 0.2931092381477356,
      "learning_rate": 0.00010671118530884809,
      "loss": 1.1258,
      "num_input_tokens_seen": 3874924,
      "step": 1403,
      "train_runtime": 3591.2505,
      "train_tokens_per_second": 1078.99
    },
    {
      "epoch": 0.11281865847043934,
      "grad_norm": 0.29299911856651306,
      "learning_rate": 0.00010664440734557598,
      "loss": 0.8513,
      "num_input_tokens_seen": 3876736,
      "step": 1404,
      "train_runtime": 3592.8513,
      "train_tokens_per_second": 1079.014
    },
    {
      "epoch": 0.11289901364029008,
      "grad_norm": 0.3434504568576813,
      "learning_rate": 0.00010657762938230385,
      "loss": 1.0769,
      "num_input_tokens_seen": 3879800,
      "step": 1405,
      "train_runtime": 3595.2112,
      "train_tokens_per_second": 1079.158
    },
    {
      "epoch": 0.11297936881014083,
      "grad_norm": 0.307935893535614,
      "learning_rate": 0.00010651085141903172,
      "loss": 1.0917,
      "num_input_tokens_seen": 3883250,
      "step": 1406,
      "train_runtime": 3597.689,
      "train_tokens_per_second": 1079.373
    },
    {
      "epoch": 0.11305972397999156,
      "grad_norm": 0.24430879950523376,
      "learning_rate": 0.0001064440734557596,
      "loss": 0.7913,
      "num_input_tokens_seen": 3885800,
      "step": 1407,
      "train_runtime": 3599.8448,
      "train_tokens_per_second": 1079.435
    },
    {
      "epoch": 0.11314007914984231,
      "grad_norm": 0.28974589705467224,
      "learning_rate": 0.0001063772954924875,
      "loss": 1.0021,
      "num_input_tokens_seen": 3887698,
      "step": 1408,
      "train_runtime": 3601.5295,
      "train_tokens_per_second": 1079.457
    },
    {
      "epoch": 0.11322043431969304,
      "grad_norm": 0.31558918952941895,
      "learning_rate": 0.00010631051752921537,
      "loss": 0.9754,
      "num_input_tokens_seen": 3891894,
      "step": 1409,
      "train_runtime": 3604.5385,
      "train_tokens_per_second": 1079.72
    },
    {
      "epoch": 0.11330078948954378,
      "grad_norm": 0.32462847232818604,
      "learning_rate": 0.00010624373956594324,
      "loss": 0.9992,
      "num_input_tokens_seen": 3894568,
      "step": 1410,
      "train_runtime": 3606.7005,
      "train_tokens_per_second": 1079.815
    },
    {
      "epoch": 0.11338114465939453,
      "grad_norm": 0.34621408581733704,
      "learning_rate": 0.00010617696160267111,
      "loss": 1.2976,
      "num_input_tokens_seen": 3896366,
      "step": 1411,
      "train_runtime": 3610.4554,
      "train_tokens_per_second": 1079.19
    },
    {
      "epoch": 0.11346149982924526,
      "grad_norm": 0.2981546223163605,
      "learning_rate": 0.000106110183639399,
      "loss": 0.8862,
      "num_input_tokens_seen": 3900532,
      "step": 1412,
      "train_runtime": 3613.396,
      "train_tokens_per_second": 1079.464
    },
    {
      "epoch": 0.113541854999096,
      "grad_norm": 0.40481603145599365,
      "learning_rate": 0.00010604340567612687,
      "loss": 1.1212,
      "num_input_tokens_seen": 3903224,
      "step": 1413,
      "train_runtime": 3615.5408,
      "train_tokens_per_second": 1079.569
    },
    {
      "epoch": 0.11362221016894675,
      "grad_norm": 0.35557135939598083,
      "learning_rate": 0.00010597662771285477,
      "loss": 1.0561,
      "num_input_tokens_seen": 3905938,
      "step": 1414,
      "train_runtime": 3617.8326,
      "train_tokens_per_second": 1079.635
    },
    {
      "epoch": 0.11370256533879748,
      "grad_norm": 0.302202969789505,
      "learning_rate": 0.00010590984974958265,
      "loss": 0.9211,
      "num_input_tokens_seen": 3908128,
      "step": 1415,
      "train_runtime": 3619.6637,
      "train_tokens_per_second": 1079.694
    },
    {
      "epoch": 0.11378292050864823,
      "grad_norm": 0.2821248769760132,
      "learning_rate": 0.00010584307178631052,
      "loss": 0.7784,
      "num_input_tokens_seen": 3910714,
      "step": 1416,
      "train_runtime": 3621.802,
      "train_tokens_per_second": 1079.77
    },
    {
      "epoch": 0.11386327567849897,
      "grad_norm": 0.3250631093978882,
      "learning_rate": 0.0001057762938230384,
      "loss": 0.8794,
      "num_input_tokens_seen": 3912254,
      "step": 1417,
      "train_runtime": 3623.4543,
      "train_tokens_per_second": 1079.703
    },
    {
      "epoch": 0.1139436308483497,
      "grad_norm": 0.34978342056274414,
      "learning_rate": 0.00010570951585976627,
      "loss": 1.1143,
      "num_input_tokens_seen": 3915770,
      "step": 1418,
      "train_runtime": 3625.9582,
      "train_tokens_per_second": 1079.927
    },
    {
      "epoch": 0.11402398601820045,
      "grad_norm": 0.34009402990341187,
      "learning_rate": 0.00010564273789649417,
      "loss": 0.9026,
      "num_input_tokens_seen": 3918878,
      "step": 1419,
      "train_runtime": 3628.3053,
      "train_tokens_per_second": 1080.085
    },
    {
      "epoch": 0.11410434118805118,
      "grad_norm": 0.33979278802871704,
      "learning_rate": 0.00010557595993322204,
      "loss": 0.9767,
      "num_input_tokens_seen": 3919988,
      "step": 1420,
      "train_runtime": 3629.8122,
      "train_tokens_per_second": 1079.942
    },
    {
      "epoch": 0.11418469635790193,
      "grad_norm": 0.34232306480407715,
      "learning_rate": 0.00010550918196994993,
      "loss": 1.1845,
      "num_input_tokens_seen": 3921820,
      "step": 1421,
      "train_runtime": 3631.4929,
      "train_tokens_per_second": 1079.947
    },
    {
      "epoch": 0.11426505152775267,
      "grad_norm": 0.33989453315734863,
      "learning_rate": 0.0001054424040066778,
      "loss": 0.8203,
      "num_input_tokens_seen": 3925958,
      "step": 1422,
      "train_runtime": 3634.6185,
      "train_tokens_per_second": 1080.157
    },
    {
      "epoch": 0.1143454066976034,
      "grad_norm": 0.3074391782283783,
      "learning_rate": 0.00010537562604340567,
      "loss": 1.0615,
      "num_input_tokens_seen": 3928358,
      "step": 1423,
      "train_runtime": 3636.5942,
      "train_tokens_per_second": 1080.23
    },
    {
      "epoch": 0.11442576186745415,
      "grad_norm": 0.3119485676288605,
      "learning_rate": 0.00010530884808013357,
      "loss": 0.958,
      "num_input_tokens_seen": 3931152,
      "step": 1424,
      "train_runtime": 3638.8064,
      "train_tokens_per_second": 1080.341
    },
    {
      "epoch": 0.11450611703730489,
      "grad_norm": 0.38976985216140747,
      "learning_rate": 0.00010524207011686145,
      "loss": 1.041,
      "num_input_tokens_seen": 3932784,
      "step": 1425,
      "train_runtime": 3640.4249,
      "train_tokens_per_second": 1080.309
    },
    {
      "epoch": 0.11458647220715563,
      "grad_norm": 0.3487476110458374,
      "learning_rate": 0.00010517529215358932,
      "loss": 1.1562,
      "num_input_tokens_seen": 3934768,
      "step": 1426,
      "train_runtime": 3642.1234,
      "train_tokens_per_second": 1080.35
    },
    {
      "epoch": 0.11466682737700637,
      "grad_norm": 0.3184686601161957,
      "learning_rate": 0.0001051085141903172,
      "loss": 1.0298,
      "num_input_tokens_seen": 3937902,
      "step": 1427,
      "train_runtime": 3644.5798,
      "train_tokens_per_second": 1080.482
    },
    {
      "epoch": 0.1147471825468571,
      "grad_norm": 0.4373387396335602,
      "learning_rate": 0.00010504173622704507,
      "loss": 1.304,
      "num_input_tokens_seen": 3942896,
      "step": 1428,
      "train_runtime": 3648.0381,
      "train_tokens_per_second": 1080.826
    },
    {
      "epoch": 0.11482753771670785,
      "grad_norm": 0.3348543643951416,
      "learning_rate": 0.00010497495826377295,
      "loss": 1.1626,
      "num_input_tokens_seen": 3945662,
      "step": 1429,
      "train_runtime": 3650.1428,
      "train_tokens_per_second": 1080.961
    },
    {
      "epoch": 0.11490789288655859,
      "grad_norm": 0.33286574482917786,
      "learning_rate": 0.00010490818030050084,
      "loss": 0.8599,
      "num_input_tokens_seen": 3948736,
      "step": 1430,
      "train_runtime": 3652.4788,
      "train_tokens_per_second": 1081.111
    },
    {
      "epoch": 0.11498824805640932,
      "grad_norm": 0.3033377528190613,
      "learning_rate": 0.00010484140233722873,
      "loss": 1.3113,
      "num_input_tokens_seen": 3951594,
      "step": 1431,
      "train_runtime": 3654.6938,
      "train_tokens_per_second": 1081.238
    },
    {
      "epoch": 0.11506860322626007,
      "grad_norm": 0.3794889748096466,
      "learning_rate": 0.0001047746243739566,
      "loss": 1.3646,
      "num_input_tokens_seen": 3953800,
      "step": 1432,
      "train_runtime": 3656.5296,
      "train_tokens_per_second": 1081.299
    },
    {
      "epoch": 0.1151489583961108,
      "grad_norm": 0.3114030063152313,
      "learning_rate": 0.00010470784641068447,
      "loss": 1.1194,
      "num_input_tokens_seen": 3956818,
      "step": 1433,
      "train_runtime": 3658.8008,
      "train_tokens_per_second": 1081.452
    },
    {
      "epoch": 0.11522931356596156,
      "grad_norm": 0.27211475372314453,
      "learning_rate": 0.00010464106844741235,
      "loss": 0.8474,
      "num_input_tokens_seen": 3959178,
      "step": 1434,
      "train_runtime": 3660.8197,
      "train_tokens_per_second": 1081.5
    },
    {
      "epoch": 0.11530966873581229,
      "grad_norm": 0.2972334921360016,
      "learning_rate": 0.00010457429048414025,
      "loss": 0.8336,
      "num_input_tokens_seen": 3960544,
      "step": 1435,
      "train_runtime": 3662.3874,
      "train_tokens_per_second": 1081.41
    },
    {
      "epoch": 0.11539002390566303,
      "grad_norm": 0.29183486104011536,
      "learning_rate": 0.00010450751252086812,
      "loss": 0.9665,
      "num_input_tokens_seen": 3962588,
      "step": 1436,
      "train_runtime": 3664.2521,
      "train_tokens_per_second": 1081.418
    },
    {
      "epoch": 0.11547037907551377,
      "grad_norm": 0.30539897084236145,
      "learning_rate": 0.00010444073455759599,
      "loss": 1.0487,
      "num_input_tokens_seen": 3965300,
      "step": 1437,
      "train_runtime": 3666.4031,
      "train_tokens_per_second": 1081.523
    },
    {
      "epoch": 0.11555073424536451,
      "grad_norm": 0.3198235034942627,
      "learning_rate": 0.00010437395659432388,
      "loss": 1.0284,
      "num_input_tokens_seen": 3967034,
      "step": 1438,
      "train_runtime": 3668.0817,
      "train_tokens_per_second": 1081.501
    },
    {
      "epoch": 0.11563108941521526,
      "grad_norm": 0.3200485110282898,
      "learning_rate": 0.00010430717863105175,
      "loss": 1.1556,
      "num_input_tokens_seen": 3970296,
      "step": 1439,
      "train_runtime": 3670.5023,
      "train_tokens_per_second": 1081.676
    },
    {
      "epoch": 0.11571144458506599,
      "grad_norm": 0.2981796860694885,
      "learning_rate": 0.00010424040066777965,
      "loss": 1.0596,
      "num_input_tokens_seen": 3972926,
      "step": 1440,
      "train_runtime": 3672.6602,
      "train_tokens_per_second": 1081.757
    },
    {
      "epoch": 0.11579179975491673,
      "grad_norm": 0.28679391741752625,
      "learning_rate": 0.00010417362270450753,
      "loss": 0.9033,
      "num_input_tokens_seen": 3976014,
      "step": 1441,
      "train_runtime": 3677.0176,
      "train_tokens_per_second": 1081.315
    },
    {
      "epoch": 0.11587215492476748,
      "grad_norm": 0.3157092332839966,
      "learning_rate": 0.0001041068447412354,
      "loss": 0.9198,
      "num_input_tokens_seen": 3979538,
      "step": 1442,
      "train_runtime": 3679.559,
      "train_tokens_per_second": 1081.526
    },
    {
      "epoch": 0.11595251009461821,
      "grad_norm": 0.3371581435203552,
      "learning_rate": 0.00010404006677796327,
      "loss": 1.0642,
      "num_input_tokens_seen": 3981386,
      "step": 1443,
      "train_runtime": 3681.3359,
      "train_tokens_per_second": 1081.506
    },
    {
      "epoch": 0.11603286526446896,
      "grad_norm": 0.2808104157447815,
      "learning_rate": 0.00010397328881469115,
      "loss": 1.4665,
      "num_input_tokens_seen": 3985510,
      "step": 1444,
      "train_runtime": 3684.3932,
      "train_tokens_per_second": 1081.728
    },
    {
      "epoch": 0.1161132204343197,
      "grad_norm": 0.31633618474006653,
      "learning_rate": 0.00010390651085141905,
      "loss": 1.0221,
      "num_input_tokens_seen": 3987956,
      "step": 1445,
      "train_runtime": 3686.4853,
      "train_tokens_per_second": 1081.777
    },
    {
      "epoch": 0.11619357560417043,
      "grad_norm": 0.29563218355178833,
      "learning_rate": 0.00010383973288814692,
      "loss": 0.9477,
      "num_input_tokens_seen": 3990216,
      "step": 1446,
      "train_runtime": 3688.3936,
      "train_tokens_per_second": 1081.83
    },
    {
      "epoch": 0.11627393077402118,
      "grad_norm": 0.3616282045841217,
      "learning_rate": 0.0001037729549248748,
      "loss": 0.9883,
      "num_input_tokens_seen": 3992298,
      "step": 1447,
      "train_runtime": 3690.2458,
      "train_tokens_per_second": 1081.852
    },
    {
      "epoch": 0.11635428594387191,
      "grad_norm": 0.3951462209224701,
      "learning_rate": 0.00010370617696160268,
      "loss": 1.3897,
      "num_input_tokens_seen": 3996792,
      "step": 1448,
      "train_runtime": 3693.5612,
      "train_tokens_per_second": 1082.097
    },
    {
      "epoch": 0.11643464111372265,
      "grad_norm": 0.3162688612937927,
      "learning_rate": 0.00010363939899833055,
      "loss": 0.8743,
      "num_input_tokens_seen": 3998988,
      "step": 1449,
      "train_runtime": 3695.4694,
      "train_tokens_per_second": 1082.133
    },
    {
      "epoch": 0.1165149962835734,
      "grad_norm": 0.3188450336456299,
      "learning_rate": 0.00010357262103505843,
      "loss": 1.0401,
      "num_input_tokens_seen": 4000944,
      "step": 1450,
      "train_runtime": 3697.2579,
      "train_tokens_per_second": 1082.138
    },
    {
      "epoch": 0.11659535145342413,
      "grad_norm": 0.2858012914657593,
      "learning_rate": 0.00010350584307178633,
      "loss": 1.0443,
      "num_input_tokens_seen": 4003162,
      "step": 1451,
      "train_runtime": 3699.0944,
      "train_tokens_per_second": 1082.201
    },
    {
      "epoch": 0.11667570662327488,
      "grad_norm": 0.3404531180858612,
      "learning_rate": 0.0001034390651085142,
      "loss": 0.6921,
      "num_input_tokens_seen": 4004826,
      "step": 1452,
      "train_runtime": 3700.7609,
      "train_tokens_per_second": 1082.163
    },
    {
      "epoch": 0.11675606179312562,
      "grad_norm": 0.31202077865600586,
      "learning_rate": 0.00010337228714524207,
      "loss": 0.9364,
      "num_input_tokens_seen": 4007318,
      "step": 1453,
      "train_runtime": 3702.8423,
      "train_tokens_per_second": 1082.228
    },
    {
      "epoch": 0.11683641696297635,
      "grad_norm": 0.3742130398750305,
      "learning_rate": 0.00010330550918196994,
      "loss": 0.8273,
      "num_input_tokens_seen": 4012332,
      "step": 1454,
      "train_runtime": 3706.531,
      "train_tokens_per_second": 1082.503
    },
    {
      "epoch": 0.1169167721328271,
      "grad_norm": 0.3088212311267853,
      "learning_rate": 0.00010323873121869783,
      "loss": 0.8714,
      "num_input_tokens_seen": 4014652,
      "step": 1455,
      "train_runtime": 3708.5804,
      "train_tokens_per_second": 1082.531
    },
    {
      "epoch": 0.11699712730267783,
      "grad_norm": 0.3392530679702759,
      "learning_rate": 0.00010317195325542572,
      "loss": 0.9011,
      "num_input_tokens_seen": 4016462,
      "step": 1456,
      "train_runtime": 3710.3192,
      "train_tokens_per_second": 1082.511
    },
    {
      "epoch": 0.11707748247252858,
      "grad_norm": 0.3812410533428192,
      "learning_rate": 0.0001031051752921536,
      "loss": 1.1298,
      "num_input_tokens_seen": 4018634,
      "step": 1457,
      "train_runtime": 3712.2635,
      "train_tokens_per_second": 1082.529
    },
    {
      "epoch": 0.11715783764237932,
      "grad_norm": 0.35789114236831665,
      "learning_rate": 0.00010303839732888148,
      "loss": 0.9295,
      "num_input_tokens_seen": 4020642,
      "step": 1458,
      "train_runtime": 3714.1219,
      "train_tokens_per_second": 1082.528
    },
    {
      "epoch": 0.11723819281223005,
      "grad_norm": 0.2593947649002075,
      "learning_rate": 0.00010297161936560935,
      "loss": 1.1925,
      "num_input_tokens_seen": 4023940,
      "step": 1459,
      "train_runtime": 3716.6038,
      "train_tokens_per_second": 1082.693
    },
    {
      "epoch": 0.1173185479820808,
      "grad_norm": 0.2739175260066986,
      "learning_rate": 0.00010290484140233722,
      "loss": 1.0771,
      "num_input_tokens_seen": 4026284,
      "step": 1460,
      "train_runtime": 3718.5772,
      "train_tokens_per_second": 1082.749
    },
    {
      "epoch": 0.11739890315193154,
      "grad_norm": 0.31637459993362427,
      "learning_rate": 0.00010283806343906512,
      "loss": 0.772,
      "num_input_tokens_seen": 4027962,
      "step": 1461,
      "train_runtime": 3720.2325,
      "train_tokens_per_second": 1082.718
    },
    {
      "epoch": 0.11747925832178228,
      "grad_norm": 0.2679910957813263,
      "learning_rate": 0.000102771285475793,
      "loss": 0.7238,
      "num_input_tokens_seen": 4032768,
      "step": 1462,
      "train_runtime": 3723.4886,
      "train_tokens_per_second": 1083.062
    },
    {
      "epoch": 0.11755961349163302,
      "grad_norm": 0.283014714717865,
      "learning_rate": 0.00010270450751252087,
      "loss": 0.8421,
      "num_input_tokens_seen": 4034896,
      "step": 1463,
      "train_runtime": 3725.3309,
      "train_tokens_per_second": 1083.097
    },
    {
      "epoch": 0.11763996866148375,
      "grad_norm": 0.31199800968170166,
      "learning_rate": 0.00010263772954924876,
      "loss": 1.0831,
      "num_input_tokens_seen": 4036852,
      "step": 1464,
      "train_runtime": 3727.0819,
      "train_tokens_per_second": 1083.113
    },
    {
      "epoch": 0.1177203238313345,
      "grad_norm": 0.29866108298301697,
      "learning_rate": 0.00010257095158597663,
      "loss": 0.9584,
      "num_input_tokens_seen": 4039966,
      "step": 1465,
      "train_runtime": 3729.5273,
      "train_tokens_per_second": 1083.238
    },
    {
      "epoch": 0.11780067900118524,
      "grad_norm": 0.28818005323410034,
      "learning_rate": 0.0001025041736227045,
      "loss": 1.0084,
      "num_input_tokens_seen": 4042228,
      "step": 1466,
      "train_runtime": 3731.3739,
      "train_tokens_per_second": 1083.308
    },
    {
      "epoch": 0.11788103417103597,
      "grad_norm": 0.36868417263031006,
      "learning_rate": 0.0001024373956594324,
      "loss": 1.0072,
      "num_input_tokens_seen": 4043934,
      "step": 1467,
      "train_runtime": 3733.0507,
      "train_tokens_per_second": 1083.279
    },
    {
      "epoch": 0.11796138934088672,
      "grad_norm": 0.4026516377925873,
      "learning_rate": 0.00010237061769616028,
      "loss": 0.9238,
      "num_input_tokens_seen": 4045512,
      "step": 1468,
      "train_runtime": 3734.6428,
      "train_tokens_per_second": 1083.239
    },
    {
      "epoch": 0.11804174451073746,
      "grad_norm": 0.3591594398021698,
      "learning_rate": 0.00010230383973288815,
      "loss": 1.1237,
      "num_input_tokens_seen": 4046834,
      "step": 1469,
      "train_runtime": 3736.1764,
      "train_tokens_per_second": 1083.149
    },
    {
      "epoch": 0.1181220996805882,
      "grad_norm": 0.32499995827674866,
      "learning_rate": 0.00010223706176961602,
      "loss": 0.9187,
      "num_input_tokens_seen": 4049092,
      "step": 1470,
      "train_runtime": 3738.2093,
      "train_tokens_per_second": 1083.164
    },
    {
      "epoch": 0.11820245485043894,
      "grad_norm": 0.2623390555381775,
      "learning_rate": 0.0001021702838063439,
      "loss": 0.9589,
      "num_input_tokens_seen": 4053266,
      "step": 1471,
      "train_runtime": 3742.8531,
      "train_tokens_per_second": 1082.935
    },
    {
      "epoch": 0.11828281002028967,
      "grad_norm": 0.3267577588558197,
      "learning_rate": 0.0001021035058430718,
      "loss": 0.7474,
      "num_input_tokens_seen": 4055404,
      "step": 1472,
      "train_runtime": 3744.7093,
      "train_tokens_per_second": 1082.969
    },
    {
      "epoch": 0.11836316519014042,
      "grad_norm": 0.3571934998035431,
      "learning_rate": 0.00010203672787979967,
      "loss": 1.4617,
      "num_input_tokens_seen": 4058278,
      "step": 1473,
      "train_runtime": 3746.958,
      "train_tokens_per_second": 1083.086
    },
    {
      "epoch": 0.11844352035999116,
      "grad_norm": 0.38490167260169983,
      "learning_rate": 0.00010196994991652756,
      "loss": 1.2963,
      "num_input_tokens_seen": 4060432,
      "step": 1474,
      "train_runtime": 3748.846,
      "train_tokens_per_second": 1083.115
    },
    {
      "epoch": 0.11852387552984191,
      "grad_norm": 0.2769409418106079,
      "learning_rate": 0.00010190317195325543,
      "loss": 0.901,
      "num_input_tokens_seen": 4063322,
      "step": 1475,
      "train_runtime": 3751.1721,
      "train_tokens_per_second": 1083.214
    },
    {
      "epoch": 0.11860423069969264,
      "grad_norm": 0.3213483691215515,
      "learning_rate": 0.0001018363939899833,
      "loss": 0.6892,
      "num_input_tokens_seen": 4066616,
      "step": 1476,
      "train_runtime": 3753.6915,
      "train_tokens_per_second": 1083.364
    },
    {
      "epoch": 0.11868458586954338,
      "grad_norm": 0.29150086641311646,
      "learning_rate": 0.0001017696160267112,
      "loss": 0.9867,
      "num_input_tokens_seen": 4069198,
      "step": 1477,
      "train_runtime": 3755.7526,
      "train_tokens_per_second": 1083.457
    },
    {
      "epoch": 0.11876494103939413,
      "grad_norm": 0.3302112817764282,
      "learning_rate": 0.00010170283806343908,
      "loss": 1.3474,
      "num_input_tokens_seen": 4071784,
      "step": 1478,
      "train_runtime": 3757.8616,
      "train_tokens_per_second": 1083.538
    },
    {
      "epoch": 0.11884529620924486,
      "grad_norm": 0.319048136472702,
      "learning_rate": 0.00010163606010016695,
      "loss": 1.1293,
      "num_input_tokens_seen": 4074594,
      "step": 1479,
      "train_runtime": 3760.0506,
      "train_tokens_per_second": 1083.654
    },
    {
      "epoch": 0.11892565137909561,
      "grad_norm": 0.4086817800998688,
      "learning_rate": 0.00010156928213689482,
      "loss": 0.9525,
      "num_input_tokens_seen": 4077978,
      "step": 1480,
      "train_runtime": 3762.6566,
      "train_tokens_per_second": 1083.803
    },
    {
      "epoch": 0.11900600654894634,
      "grad_norm": 0.3641253113746643,
      "learning_rate": 0.00010150250417362271,
      "loss": 1.2776,
      "num_input_tokens_seen": 4080702,
      "step": 1481,
      "train_runtime": 3764.7547,
      "train_tokens_per_second": 1083.922
    },
    {
      "epoch": 0.11908636171879708,
      "grad_norm": 0.29269173741340637,
      "learning_rate": 0.00010143572621035058,
      "loss": 1.2271,
      "num_input_tokens_seen": 4083220,
      "step": 1482,
      "train_runtime": 3766.8052,
      "train_tokens_per_second": 1084.001
    },
    {
      "epoch": 0.11916671688864783,
      "grad_norm": 0.3291515111923218,
      "learning_rate": 0.00010136894824707848,
      "loss": 1.1588,
      "num_input_tokens_seen": 4085380,
      "step": 1483,
      "train_runtime": 3768.7048,
      "train_tokens_per_second": 1084.028
    },
    {
      "epoch": 0.11924707205849856,
      "grad_norm": 0.2971457839012146,
      "learning_rate": 0.00010130217028380636,
      "loss": 1.0273,
      "num_input_tokens_seen": 4087444,
      "step": 1484,
      "train_runtime": 3770.5371,
      "train_tokens_per_second": 1084.048
    },
    {
      "epoch": 0.1193274272283493,
      "grad_norm": 0.32142123579978943,
      "learning_rate": 0.00010123539232053423,
      "loss": 0.9107,
      "num_input_tokens_seen": 4089092,
      "step": 1485,
      "train_runtime": 3772.1972,
      "train_tokens_per_second": 1084.008
    },
    {
      "epoch": 0.11940778239820005,
      "grad_norm": 0.3527226150035858,
      "learning_rate": 0.0001011686143572621,
      "loss": 1.1797,
      "num_input_tokens_seen": 4090958,
      "step": 1486,
      "train_runtime": 3773.8854,
      "train_tokens_per_second": 1084.018
    },
    {
      "epoch": 0.11948813756805078,
      "grad_norm": 0.3108879625797272,
      "learning_rate": 0.00010110183639398998,
      "loss": 0.8897,
      "num_input_tokens_seen": 4093624,
      "step": 1487,
      "train_runtime": 3776.0624,
      "train_tokens_per_second": 1084.099
    },
    {
      "epoch": 0.11956849273790153,
      "grad_norm": 0.3651483356952667,
      "learning_rate": 0.00010103505843071788,
      "loss": 0.8955,
      "num_input_tokens_seen": 4096158,
      "step": 1488,
      "train_runtime": 3778.1789,
      "train_tokens_per_second": 1084.162
    },
    {
      "epoch": 0.11964884790775226,
      "grad_norm": 0.3421587646007538,
      "learning_rate": 0.00010096828046744575,
      "loss": 1.2061,
      "num_input_tokens_seen": 4098476,
      "step": 1489,
      "train_runtime": 3780.11,
      "train_tokens_per_second": 1084.221
    },
    {
      "epoch": 0.119729203077603,
      "grad_norm": 0.27953073382377625,
      "learning_rate": 0.00010090150250417362,
      "loss": 0.8189,
      "num_input_tokens_seen": 4100152,
      "step": 1490,
      "train_runtime": 3781.7738,
      "train_tokens_per_second": 1084.188
    },
    {
      "epoch": 0.11980955824745375,
      "grad_norm": 0.29624682664871216,
      "learning_rate": 0.00010083472454090151,
      "loss": 0.7177,
      "num_input_tokens_seen": 4102288,
      "step": 1491,
      "train_runtime": 3783.6621,
      "train_tokens_per_second": 1084.211
    },
    {
      "epoch": 0.11988991341730448,
      "grad_norm": 0.3120693862438202,
      "learning_rate": 0.00010076794657762938,
      "loss": 0.903,
      "num_input_tokens_seen": 4104320,
      "step": 1492,
      "train_runtime": 3785.4741,
      "train_tokens_per_second": 1084.229
    },
    {
      "epoch": 0.11997026858715523,
      "grad_norm": 0.34112513065338135,
      "learning_rate": 0.00010070116861435728,
      "loss": 1.2143,
      "num_input_tokens_seen": 4106598,
      "step": 1493,
      "train_runtime": 3787.4906,
      "train_tokens_per_second": 1084.253
    },
    {
      "epoch": 0.12005062375700597,
      "grad_norm": 0.3413991928100586,
      "learning_rate": 0.00010063439065108516,
      "loss": 1.043,
      "num_input_tokens_seen": 4109652,
      "step": 1494,
      "train_runtime": 3789.903,
      "train_tokens_per_second": 1084.369
    },
    {
      "epoch": 0.1201309789268567,
      "grad_norm": 0.28197675943374634,
      "learning_rate": 0.00010056761268781303,
      "loss": 0.8596,
      "num_input_tokens_seen": 4112258,
      "step": 1495,
      "train_runtime": 3792.0023,
      "train_tokens_per_second": 1084.456
    },
    {
      "epoch": 0.12021133409670745,
      "grad_norm": 0.345485121011734,
      "learning_rate": 0.0001005008347245409,
      "loss": 1.0525,
      "num_input_tokens_seen": 4114188,
      "step": 1496,
      "train_runtime": 3793.737,
      "train_tokens_per_second": 1084.468
    },
    {
      "epoch": 0.12029168926655819,
      "grad_norm": 0.3160560429096222,
      "learning_rate": 0.00010043405676126878,
      "loss": 1.0691,
      "num_input_tokens_seen": 4118016,
      "step": 1497,
      "train_runtime": 3796.4987,
      "train_tokens_per_second": 1084.688
    },
    {
      "epoch": 0.12037204443640893,
      "grad_norm": 0.35005122423171997,
      "learning_rate": 0.00010036727879799666,
      "loss": 0.908,
      "num_input_tokens_seen": 4119490,
      "step": 1498,
      "train_runtime": 3798.0739,
      "train_tokens_per_second": 1084.626
    },
    {
      "epoch": 0.12045239960625967,
      "grad_norm": 0.293682724237442,
      "learning_rate": 0.00010030050083472455,
      "loss": 0.6331,
      "num_input_tokens_seen": 4122572,
      "step": 1499,
      "train_runtime": 3800.4721,
      "train_tokens_per_second": 1084.753
    },
    {
      "epoch": 0.1205327547761104,
      "grad_norm": 0.2874637544155121,
      "learning_rate": 0.00010023372287145244,
      "loss": 0.8274,
      "num_input_tokens_seen": 4125028,
      "step": 1500,
      "train_runtime": 3802.5022,
      "train_tokens_per_second": 1084.819
    },
    {
      "epoch": 0.1205327547761104,
      "eval_loss": 0.9687259197235107,
      "eval_runtime": 91.9069,
      "eval_samples_per_second": 10.815,
      "eval_steps_per_second": 5.408,
      "num_input_tokens_seen": 4125028,
      "step": 1500
    },
    {
      "epoch": 0.12061310994596115,
      "grad_norm": 0.4220530092716217,
      "learning_rate": 0.00010016694490818031,
      "loss": 1.0931,
      "num_input_tokens_seen": 4126924,
      "step": 1501,
      "train_runtime": 3898.2509,
      "train_tokens_per_second": 1058.66
    },
    {
      "epoch": 0.12069346511581189,
      "grad_norm": 0.3375208079814911,
      "learning_rate": 0.00010010016694490818,
      "loss": 1.026,
      "num_input_tokens_seen": 4128474,
      "step": 1502,
      "train_runtime": 3899.8855,
      "train_tokens_per_second": 1058.614
    },
    {
      "epoch": 0.12077382028566262,
      "grad_norm": 0.3334346115589142,
      "learning_rate": 0.00010003338898163605,
      "loss": 1.1606,
      "num_input_tokens_seen": 4130818,
      "step": 1503,
      "train_runtime": 3901.7898,
      "train_tokens_per_second": 1058.698
    },
    {
      "epoch": 0.12085417545551337,
      "grad_norm": 0.24376749992370605,
      "learning_rate": 9.996661101836394e-05,
      "loss": 0.8696,
      "num_input_tokens_seen": 4134226,
      "step": 1504,
      "train_runtime": 3904.3643,
      "train_tokens_per_second": 1058.873
    },
    {
      "epoch": 0.1209345306253641,
      "grad_norm": 0.33671870827674866,
      "learning_rate": 9.989983305509183e-05,
      "loss": 1.0061,
      "num_input_tokens_seen": 4136524,
      "step": 1505,
      "train_runtime": 3906.3474,
      "train_tokens_per_second": 1058.924
    },
    {
      "epoch": 0.12101488579521485,
      "grad_norm": 0.3007793724536896,
      "learning_rate": 9.98330550918197e-05,
      "loss": 1.0107,
      "num_input_tokens_seen": 4142152,
      "step": 1506,
      "train_runtime": 3910.1659,
      "train_tokens_per_second": 1059.329
    },
    {
      "epoch": 0.12109524096506559,
      "grad_norm": 0.38502055406570435,
      "learning_rate": 9.976627712854757e-05,
      "loss": 0.6883,
      "num_input_tokens_seen": 4145772,
      "step": 1507,
      "train_runtime": 3912.821,
      "train_tokens_per_second": 1059.535
    },
    {
      "epoch": 0.12117559613491632,
      "grad_norm": 0.3022374212741852,
      "learning_rate": 9.969949916527546e-05,
      "loss": 0.8321,
      "num_input_tokens_seen": 4148734,
      "step": 1508,
      "train_runtime": 3915.115,
      "train_tokens_per_second": 1059.671
    },
    {
      "epoch": 0.12125595130476707,
      "grad_norm": 0.3609469532966614,
      "learning_rate": 9.963272120200335e-05,
      "loss": 0.926,
      "num_input_tokens_seen": 4149902,
      "step": 1509,
      "train_runtime": 3916.5995,
      "train_tokens_per_second": 1059.568
    },
    {
      "epoch": 0.12133630647461781,
      "grad_norm": 0.36960041522979736,
      "learning_rate": 9.956594323873122e-05,
      "loss": 0.8551,
      "num_input_tokens_seen": 4151634,
      "step": 1510,
      "train_runtime": 3918.2656,
      "train_tokens_per_second": 1059.559
    },
    {
      "epoch": 0.12141666164446856,
      "grad_norm": 0.3294745087623596,
      "learning_rate": 9.949916527545911e-05,
      "loss": 0.9627,
      "num_input_tokens_seen": 4154556,
      "step": 1511,
      "train_runtime": 3920.5899,
      "train_tokens_per_second": 1059.676
    },
    {
      "epoch": 0.12149701681431929,
      "grad_norm": 0.34972280263900757,
      "learning_rate": 9.943238731218698e-05,
      "loss": 0.8768,
      "num_input_tokens_seen": 4156844,
      "step": 1512,
      "train_runtime": 3922.489,
      "train_tokens_per_second": 1059.746
    },
    {
      "epoch": 0.12157737198417003,
      "grad_norm": 0.29665154218673706,
      "learning_rate": 9.936560934891487e-05,
      "loss": 1.1868,
      "num_input_tokens_seen": 4158776,
      "step": 1513,
      "train_runtime": 3924.2701,
      "train_tokens_per_second": 1059.758
    },
    {
      "epoch": 0.12165772715402078,
      "grad_norm": 0.3763272762298584,
      "learning_rate": 9.929883138564274e-05,
      "loss": 1.1702,
      "num_input_tokens_seen": 4161048,
      "step": 1514,
      "train_runtime": 3926.2345,
      "train_tokens_per_second": 1059.806
    },
    {
      "epoch": 0.12173808232387151,
      "grad_norm": 0.30107128620147705,
      "learning_rate": 9.923205342237061e-05,
      "loss": 0.8206,
      "num_input_tokens_seen": 4163580,
      "step": 1515,
      "train_runtime": 3928.3509,
      "train_tokens_per_second": 1059.88
    },
    {
      "epoch": 0.12181843749372226,
      "grad_norm": 0.3155030906200409,
      "learning_rate": 9.91652754590985e-05,
      "loss": 1.0903,
      "num_input_tokens_seen": 4168808,
      "step": 1516,
      "train_runtime": 3932.0118,
      "train_tokens_per_second": 1060.223
    },
    {
      "epoch": 0.121898792663573,
      "grad_norm": 0.3096657395362854,
      "learning_rate": 9.909849749582639e-05,
      "loss": 0.8853,
      "num_input_tokens_seen": 4171406,
      "step": 1517,
      "train_runtime": 3934.1106,
      "train_tokens_per_second": 1060.317
    },
    {
      "epoch": 0.12197914783342373,
      "grad_norm": 0.29947328567504883,
      "learning_rate": 9.903171953255426e-05,
      "loss": 0.8163,
      "num_input_tokens_seen": 4174348,
      "step": 1518,
      "train_runtime": 3936.49,
      "train_tokens_per_second": 1060.424
    },
    {
      "epoch": 0.12205950300327448,
      "grad_norm": 0.4084034860134125,
      "learning_rate": 9.896494156928215e-05,
      "loss": 1.4146,
      "num_input_tokens_seen": 4178092,
      "step": 1519,
      "train_runtime": 3939.2342,
      "train_tokens_per_second": 1060.636
    },
    {
      "epoch": 0.12213985817312521,
      "grad_norm": 0.3082602620124817,
      "learning_rate": 9.889816360601002e-05,
      "loss": 0.9939,
      "num_input_tokens_seen": 4180770,
      "step": 1520,
      "train_runtime": 3941.3942,
      "train_tokens_per_second": 1060.734
    },
    {
      "epoch": 0.12222021334297595,
      "grad_norm": 0.29626932740211487,
      "learning_rate": 9.883138564273791e-05,
      "loss": 1.0401,
      "num_input_tokens_seen": 4183738,
      "step": 1521,
      "train_runtime": 3943.6946,
      "train_tokens_per_second": 1060.868
    },
    {
      "epoch": 0.1223005685128267,
      "grad_norm": 0.3405837416648865,
      "learning_rate": 9.876460767946578e-05,
      "loss": 1.0216,
      "num_input_tokens_seen": 4187014,
      "step": 1522,
      "train_runtime": 3946.1954,
      "train_tokens_per_second": 1061.026
    },
    {
      "epoch": 0.12238092368267743,
      "grad_norm": 0.2592165768146515,
      "learning_rate": 9.869782971619365e-05,
      "loss": 0.5401,
      "num_input_tokens_seen": 4188814,
      "step": 1523,
      "train_runtime": 3947.9318,
      "train_tokens_per_second": 1061.015
    },
    {
      "epoch": 0.12246127885252818,
      "grad_norm": 0.30895012617111206,
      "learning_rate": 9.863105175292154e-05,
      "loss": 1.1475,
      "num_input_tokens_seen": 4190888,
      "step": 1524,
      "train_runtime": 3949.8037,
      "train_tokens_per_second": 1061.037
    },
    {
      "epoch": 0.12254163402237891,
      "grad_norm": 0.26299959421157837,
      "learning_rate": 9.856427378964941e-05,
      "loss": 0.7644,
      "num_input_tokens_seen": 4193516,
      "step": 1525,
      "train_runtime": 3951.9798,
      "train_tokens_per_second": 1061.118
    },
    {
      "epoch": 0.12262198919222965,
      "grad_norm": 0.2840707302093506,
      "learning_rate": 9.84974958263773e-05,
      "loss": 0.8117,
      "num_input_tokens_seen": 4197604,
      "step": 1526,
      "train_runtime": 3955.0031,
      "train_tokens_per_second": 1061.34
    },
    {
      "epoch": 0.1227023443620804,
      "grad_norm": 0.27579811215400696,
      "learning_rate": 9.843071786310519e-05,
      "loss": 1.0271,
      "num_input_tokens_seen": 4201056,
      "step": 1527,
      "train_runtime": 3957.551,
      "train_tokens_per_second": 1061.529
    },
    {
      "epoch": 0.12278269953193113,
      "grad_norm": 0.30672600865364075,
      "learning_rate": 9.836393989983306e-05,
      "loss": 0.8594,
      "num_input_tokens_seen": 4204054,
      "step": 1528,
      "train_runtime": 3959.819,
      "train_tokens_per_second": 1061.678
    },
    {
      "epoch": 0.12286305470178188,
      "grad_norm": 0.3525024950504303,
      "learning_rate": 9.829716193656095e-05,
      "loss": 1.216,
      "num_input_tokens_seen": 4206328,
      "step": 1529,
      "train_runtime": 3961.8489,
      "train_tokens_per_second": 1061.708
    },
    {
      "epoch": 0.12294340987163262,
      "grad_norm": 0.29691240191459656,
      "learning_rate": 9.823038397328882e-05,
      "loss": 1.1783,
      "num_input_tokens_seen": 4209390,
      "step": 1530,
      "train_runtime": 3964.1874,
      "train_tokens_per_second": 1061.854
    },
    {
      "epoch": 0.12302376504148335,
      "grad_norm": 0.2713939845561981,
      "learning_rate": 9.816360601001669e-05,
      "loss": 0.8498,
      "num_input_tokens_seen": 4212364,
      "step": 1531,
      "train_runtime": 3968.5436,
      "train_tokens_per_second": 1061.438
    },
    {
      "epoch": 0.1231041202113341,
      "grad_norm": 0.2757306694984436,
      "learning_rate": 9.809682804674458e-05,
      "loss": 0.9648,
      "num_input_tokens_seen": 4214222,
      "step": 1532,
      "train_runtime": 3970.2781,
      "train_tokens_per_second": 1061.443
    },
    {
      "epoch": 0.12318447538118484,
      "grad_norm": 0.32179951667785645,
      "learning_rate": 9.803005008347245e-05,
      "loss": 0.9963,
      "num_input_tokens_seen": 4215616,
      "step": 1533,
      "train_runtime": 3971.8362,
      "train_tokens_per_second": 1061.377
    },
    {
      "epoch": 0.12326483055103558,
      "grad_norm": 0.29909610748291016,
      "learning_rate": 9.796327212020034e-05,
      "loss": 0.9778,
      "num_input_tokens_seen": 4218242,
      "step": 1534,
      "train_runtime": 3973.9604,
      "train_tokens_per_second": 1061.471
    },
    {
      "epoch": 0.12334518572088632,
      "grad_norm": 0.30370089411735535,
      "learning_rate": 9.789649415692823e-05,
      "loss": 1.0526,
      "num_input_tokens_seen": 4222488,
      "step": 1535,
      "train_runtime": 3977.1252,
      "train_tokens_per_second": 1061.694
    },
    {
      "epoch": 0.12342554089073705,
      "grad_norm": 0.33886227011680603,
      "learning_rate": 9.78297161936561e-05,
      "loss": 0.9511,
      "num_input_tokens_seen": 4224014,
      "step": 1536,
      "train_runtime": 3978.7271,
      "train_tokens_per_second": 1061.65
    },
    {
      "epoch": 0.1235058960605878,
      "grad_norm": 0.2827256917953491,
      "learning_rate": 9.776293823038399e-05,
      "loss": 1.2194,
      "num_input_tokens_seen": 4227360,
      "step": 1537,
      "train_runtime": 3981.2885,
      "train_tokens_per_second": 1061.807
    },
    {
      "epoch": 0.12358625123043854,
      "grad_norm": 0.29998552799224854,
      "learning_rate": 9.769616026711186e-05,
      "loss": 1.0784,
      "num_input_tokens_seen": 4230310,
      "step": 1538,
      "train_runtime": 3983.5407,
      "train_tokens_per_second": 1061.947
    },
    {
      "epoch": 0.12366660640028927,
      "grad_norm": 0.26519492268562317,
      "learning_rate": 9.762938230383973e-05,
      "loss": 0.9632,
      "num_input_tokens_seen": 4232910,
      "step": 1539,
      "train_runtime": 3985.6123,
      "train_tokens_per_second": 1062.048
    },
    {
      "epoch": 0.12374696157014002,
      "grad_norm": 0.33117997646331787,
      "learning_rate": 9.756260434056762e-05,
      "loss": 1.2052,
      "num_input_tokens_seen": 4236572,
      "step": 1540,
      "train_runtime": 3988.3582,
      "train_tokens_per_second": 1062.235
    },
    {
      "epoch": 0.12382731673999076,
      "grad_norm": 0.32519131898880005,
      "learning_rate": 9.749582637729549e-05,
      "loss": 1.0382,
      "num_input_tokens_seen": 4238426,
      "step": 1541,
      "train_runtime": 3990.093,
      "train_tokens_per_second": 1062.237
    },
    {
      "epoch": 0.1239076719098415,
      "grad_norm": 0.29621413350105286,
      "learning_rate": 9.742904841402337e-05,
      "loss": 0.7077,
      "num_input_tokens_seen": 4241042,
      "step": 1542,
      "train_runtime": 3992.137,
      "train_tokens_per_second": 1062.349
    },
    {
      "epoch": 0.12398802707969224,
      "grad_norm": 0.4213193356990814,
      "learning_rate": 9.736227045075125e-05,
      "loss": 1.0655,
      "num_input_tokens_seen": 4243640,
      "step": 1543,
      "train_runtime": 3994.2365,
      "train_tokens_per_second": 1062.441
    },
    {
      "epoch": 0.12406838224954297,
      "grad_norm": 0.3207896649837494,
      "learning_rate": 9.729549248747914e-05,
      "loss": 1.0782,
      "num_input_tokens_seen": 4246446,
      "step": 1544,
      "train_runtime": 3996.5018,
      "train_tokens_per_second": 1062.541
    },
    {
      "epoch": 0.12414873741939372,
      "grad_norm": 0.3392288386821747,
      "learning_rate": 9.722871452420703e-05,
      "loss": 0.9714,
      "num_input_tokens_seen": 4248688,
      "step": 1545,
      "train_runtime": 3998.4739,
      "train_tokens_per_second": 1062.577
    },
    {
      "epoch": 0.12422909258924446,
      "grad_norm": 0.3342682719230652,
      "learning_rate": 9.71619365609349e-05,
      "loss": 0.785,
      "num_input_tokens_seen": 4250868,
      "step": 1546,
      "train_runtime": 4000.3503,
      "train_tokens_per_second": 1062.624
    },
    {
      "epoch": 0.1243094477590952,
      "grad_norm": 0.46963852643966675,
      "learning_rate": 9.709515859766277e-05,
      "loss": 0.8838,
      "num_input_tokens_seen": 4254280,
      "step": 1547,
      "train_runtime": 4002.8892,
      "train_tokens_per_second": 1062.802
    },
    {
      "epoch": 0.12438980292894594,
      "grad_norm": 0.2585029602050781,
      "learning_rate": 9.702838063439066e-05,
      "loss": 0.716,
      "num_input_tokens_seen": 4257168,
      "step": 1548,
      "train_runtime": 4005.1704,
      "train_tokens_per_second": 1062.918
    },
    {
      "epoch": 0.12447015809879668,
      "grad_norm": 0.355210542678833,
      "learning_rate": 9.696160267111853e-05,
      "loss": 1.0856,
      "num_input_tokens_seen": 4260700,
      "step": 1549,
      "train_runtime": 4007.8139,
      "train_tokens_per_second": 1063.098
    },
    {
      "epoch": 0.12455051326864743,
      "grad_norm": 0.3496839106082916,
      "learning_rate": 9.68948247078464e-05,
      "loss": 1.2003,
      "num_input_tokens_seen": 4263468,
      "step": 1550,
      "train_runtime": 4010.0487,
      "train_tokens_per_second": 1063.196
    },
    {
      "epoch": 0.12463086843849816,
      "grad_norm": 0.35232529044151306,
      "learning_rate": 9.682804674457429e-05,
      "loss": 0.846,
      "num_input_tokens_seen": 4266430,
      "step": 1551,
      "train_runtime": 4012.3354,
      "train_tokens_per_second": 1063.328
    },
    {
      "epoch": 0.12471122360834891,
      "grad_norm": 0.5223502516746521,
      "learning_rate": 9.676126878130218e-05,
      "loss": 0.9476,
      "num_input_tokens_seen": 4268414,
      "step": 1552,
      "train_runtime": 4014.1462,
      "train_tokens_per_second": 1063.343
    },
    {
      "epoch": 0.12479157877819964,
      "grad_norm": 0.352175772190094,
      "learning_rate": 9.669449081803006e-05,
      "loss": 1.2018,
      "num_input_tokens_seen": 4272096,
      "step": 1553,
      "train_runtime": 4016.8707,
      "train_tokens_per_second": 1063.538
    },
    {
      "epoch": 0.12487193394805038,
      "grad_norm": 0.2820923626422882,
      "learning_rate": 9.662771285475794e-05,
      "loss": 0.8292,
      "num_input_tokens_seen": 4275040,
      "step": 1554,
      "train_runtime": 4019.155,
      "train_tokens_per_second": 1063.666
    },
    {
      "epoch": 0.12495228911790113,
      "grad_norm": 0.35349148511886597,
      "learning_rate": 9.656093489148581e-05,
      "loss": 0.8889,
      "num_input_tokens_seen": 4277406,
      "step": 1555,
      "train_runtime": 4021.1961,
      "train_tokens_per_second": 1063.715
    },
    {
      "epoch": 0.12503264428775188,
      "grad_norm": 0.28008055686950684,
      "learning_rate": 9.64941569282137e-05,
      "loss": 1.0218,
      "num_input_tokens_seen": 4279672,
      "step": 1556,
      "train_runtime": 4023.1201,
      "train_tokens_per_second": 1063.769
    },
    {
      "epoch": 0.1251129994576026,
      "grad_norm": 0.25264522433280945,
      "learning_rate": 9.642737896494157e-05,
      "loss": 0.7857,
      "num_input_tokens_seen": 4281902,
      "step": 1557,
      "train_runtime": 4025.038,
      "train_tokens_per_second": 1063.817
    },
    {
      "epoch": 0.12519335462745335,
      "grad_norm": 0.34692656993865967,
      "learning_rate": 9.636060100166944e-05,
      "loss": 0.9851,
      "num_input_tokens_seen": 4283642,
      "step": 1558,
      "train_runtime": 4026.8277,
      "train_tokens_per_second": 1063.776
    },
    {
      "epoch": 0.1252737097973041,
      "grad_norm": 0.36077600717544556,
      "learning_rate": 9.629382303839733e-05,
      "loss": 0.9562,
      "num_input_tokens_seen": 4286324,
      "step": 1559,
      "train_runtime": 4029.0136,
      "train_tokens_per_second": 1063.864
    },
    {
      "epoch": 0.12535406496715482,
      "grad_norm": 0.29979103803634644,
      "learning_rate": 9.62270450751252e-05,
      "loss": 0.9474,
      "num_input_tokens_seen": 4289516,
      "step": 1560,
      "train_runtime": 4031.5425,
      "train_tokens_per_second": 1063.989
    },
    {
      "epoch": 0.12543442013700556,
      "grad_norm": 0.4771215617656708,
      "learning_rate": 9.616026711185309e-05,
      "loss": 0.7319,
      "num_input_tokens_seen": 4291630,
      "step": 1561,
      "train_runtime": 4035.3936,
      "train_tokens_per_second": 1063.497
    },
    {
      "epoch": 0.1255147753068563,
      "grad_norm": 0.30186399817466736,
      "learning_rate": 9.609348914858098e-05,
      "loss": 0.9793,
      "num_input_tokens_seen": 4295206,
      "step": 1562,
      "train_runtime": 4038.0338,
      "train_tokens_per_second": 1063.687
    },
    {
      "epoch": 0.12559513047670703,
      "grad_norm": 0.3582748472690582,
      "learning_rate": 9.602671118530885e-05,
      "loss": 1.001,
      "num_input_tokens_seen": 4297150,
      "step": 1563,
      "train_runtime": 4039.8707,
      "train_tokens_per_second": 1063.685
    },
    {
      "epoch": 0.12567548564655778,
      "grad_norm": 0.2658465504646301,
      "learning_rate": 9.595993322203674e-05,
      "loss": 1.0441,
      "num_input_tokens_seen": 4299646,
      "step": 1564,
      "train_runtime": 4041.947,
      "train_tokens_per_second": 1063.756
    },
    {
      "epoch": 0.12575584081640853,
      "grad_norm": 0.3602330684661865,
      "learning_rate": 9.589315525876461e-05,
      "loss": 0.7958,
      "num_input_tokens_seen": 4304910,
      "step": 1565,
      "train_runtime": 4045.4956,
      "train_tokens_per_second": 1064.124
    },
    {
      "epoch": 0.12583619598625925,
      "grad_norm": 0.47481727600097656,
      "learning_rate": 9.582637729549248e-05,
      "loss": 1.2609,
      "num_input_tokens_seen": 4309734,
      "step": 1566,
      "train_runtime": 4049.116,
      "train_tokens_per_second": 1064.364
    },
    {
      "epoch": 0.12591655115611,
      "grad_norm": 0.4152240753173828,
      "learning_rate": 9.575959933222037e-05,
      "loss": 1.0946,
      "num_input_tokens_seen": 4312310,
      "step": 1567,
      "train_runtime": 4051.2582,
      "train_tokens_per_second": 1064.437
    },
    {
      "epoch": 0.12599690632596075,
      "grad_norm": 0.34455329179763794,
      "learning_rate": 9.569282136894824e-05,
      "loss": 1.1691,
      "num_input_tokens_seen": 4314016,
      "step": 1568,
      "train_runtime": 4052.9792,
      "train_tokens_per_second": 1064.406
    },
    {
      "epoch": 0.1260772614958115,
      "grad_norm": 0.36066463589668274,
      "learning_rate": 9.562604340567613e-05,
      "loss": 1.1051,
      "num_input_tokens_seen": 4315980,
      "step": 1569,
      "train_runtime": 4054.8256,
      "train_tokens_per_second": 1064.406
    },
    {
      "epoch": 0.12615761666566222,
      "grad_norm": 0.3376621901988983,
      "learning_rate": 9.555926544240402e-05,
      "loss": 1.3598,
      "num_input_tokens_seen": 4318890,
      "step": 1570,
      "train_runtime": 4057.0943,
      "train_tokens_per_second": 1064.528
    },
    {
      "epoch": 0.12623797183551297,
      "grad_norm": 0.29116007685661316,
      "learning_rate": 9.549248747913189e-05,
      "loss": 1.21,
      "num_input_tokens_seen": 4322766,
      "step": 1571,
      "train_runtime": 4059.8215,
      "train_tokens_per_second": 1064.768
    },
    {
      "epoch": 0.12631832700536372,
      "grad_norm": 0.3354508876800537,
      "learning_rate": 9.542570951585978e-05,
      "loss": 1.1107,
      "num_input_tokens_seen": 4325462,
      "step": 1572,
      "train_runtime": 4061.9668,
      "train_tokens_per_second": 1064.869
    },
    {
      "epoch": 0.12639868217521444,
      "grad_norm": 0.3176094591617584,
      "learning_rate": 9.535893155258765e-05,
      "loss": 1.1147,
      "num_input_tokens_seen": 4328720,
      "step": 1573,
      "train_runtime": 4064.3823,
      "train_tokens_per_second": 1065.038
    },
    {
      "epoch": 0.1264790373450652,
      "grad_norm": 0.2785077691078186,
      "learning_rate": 9.529215358931554e-05,
      "loss": 0.9175,
      "num_input_tokens_seen": 4332450,
      "step": 1574,
      "train_runtime": 4067.1878,
      "train_tokens_per_second": 1065.22
    },
    {
      "epoch": 0.12655939251491594,
      "grad_norm": 0.3080126643180847,
      "learning_rate": 9.522537562604341e-05,
      "loss": 0.8276,
      "num_input_tokens_seen": 4334602,
      "step": 1575,
      "train_runtime": 4069.0806,
      "train_tokens_per_second": 1065.253
    },
    {
      "epoch": 0.12663974768476666,
      "grad_norm": 0.25135132670402527,
      "learning_rate": 9.515859766277128e-05,
      "loss": 0.7839,
      "num_input_tokens_seen": 4338854,
      "step": 1576,
      "train_runtime": 4072.0152,
      "train_tokens_per_second": 1065.53
    },
    {
      "epoch": 0.1267201028546174,
      "grad_norm": 0.34842756390571594,
      "learning_rate": 9.509181969949917e-05,
      "loss": 0.7685,
      "num_input_tokens_seen": 4341122,
      "step": 1577,
      "train_runtime": 4074.0358,
      "train_tokens_per_second": 1065.558
    },
    {
      "epoch": 0.12680045802446815,
      "grad_norm": 0.29723262786865234,
      "learning_rate": 9.502504173622706e-05,
      "loss": 0.9885,
      "num_input_tokens_seen": 4345372,
      "step": 1578,
      "train_runtime": 4077.1368,
      "train_tokens_per_second": 1065.79
    },
    {
      "epoch": 0.1268808131943189,
      "grad_norm": 0.3278983235359192,
      "learning_rate": 9.495826377295493e-05,
      "loss": 1.1951,
      "num_input_tokens_seen": 4348502,
      "step": 1579,
      "train_runtime": 4079.5335,
      "train_tokens_per_second": 1065.931
    },
    {
      "epoch": 0.12696116836416962,
      "grad_norm": 0.3214225172996521,
      "learning_rate": 9.489148580968282e-05,
      "loss": 1.1698,
      "num_input_tokens_seen": 4350354,
      "step": 1580,
      "train_runtime": 4081.2724,
      "train_tokens_per_second": 1065.931
    },
    {
      "epoch": 0.12704152353402037,
      "grad_norm": 0.31389468908309937,
      "learning_rate": 9.482470784641069e-05,
      "loss": 0.6864,
      "num_input_tokens_seen": 4351910,
      "step": 1581,
      "train_runtime": 4082.8667,
      "train_tokens_per_second": 1065.896
    },
    {
      "epoch": 0.12712187870387112,
      "grad_norm": 0.35068169236183167,
      "learning_rate": 9.475792988313858e-05,
      "loss": 0.9628,
      "num_input_tokens_seen": 4353630,
      "step": 1582,
      "train_runtime": 4084.641,
      "train_tokens_per_second": 1065.854
    },
    {
      "epoch": 0.12720223387372184,
      "grad_norm": 0.2965669333934784,
      "learning_rate": 9.469115191986645e-05,
      "loss": 0.8182,
      "num_input_tokens_seen": 4356496,
      "step": 1583,
      "train_runtime": 4087.0429,
      "train_tokens_per_second": 1065.929
    },
    {
      "epoch": 0.1272825890435726,
      "grad_norm": 0.34938281774520874,
      "learning_rate": 9.462437395659432e-05,
      "loss": 0.8695,
      "num_input_tokens_seen": 4359048,
      "step": 1584,
      "train_runtime": 4089.2205,
      "train_tokens_per_second": 1065.985
    },
    {
      "epoch": 0.12736294421342334,
      "grad_norm": 0.33864614367485046,
      "learning_rate": 9.455759599332221e-05,
      "loss": 1.0555,
      "num_input_tokens_seen": 4361124,
      "step": 1585,
      "train_runtime": 4091.2194,
      "train_tokens_per_second": 1065.972
    },
    {
      "epoch": 0.12744329938327406,
      "grad_norm": 0.32404622435569763,
      "learning_rate": 9.449081803005008e-05,
      "loss": 1.0601,
      "num_input_tokens_seen": 4364340,
      "step": 1586,
      "train_runtime": 4093.6635,
      "train_tokens_per_second": 1066.121
    },
    {
      "epoch": 0.1275236545531248,
      "grad_norm": 0.3412817418575287,
      "learning_rate": 9.442404006677797e-05,
      "loss": 1.1029,
      "num_input_tokens_seen": 4365894,
      "step": 1587,
      "train_runtime": 4095.2502,
      "train_tokens_per_second": 1066.087
    },
    {
      "epoch": 0.12760400972297556,
      "grad_norm": 0.38739249110221863,
      "learning_rate": 9.435726210350586e-05,
      "loss": 1.0414,
      "num_input_tokens_seen": 4369104,
      "step": 1588,
      "train_runtime": 4097.6883,
      "train_tokens_per_second": 1066.236
    },
    {
      "epoch": 0.12768436489282628,
      "grad_norm": 0.3215208649635315,
      "learning_rate": 9.429048414023373e-05,
      "loss": 0.9363,
      "num_input_tokens_seen": 4371096,
      "step": 1589,
      "train_runtime": 4099.5478,
      "train_tokens_per_second": 1066.239
    },
    {
      "epoch": 0.12776472006267703,
      "grad_norm": 0.249175563454628,
      "learning_rate": 9.422370617696162e-05,
      "loss": 0.7218,
      "num_input_tokens_seen": 4374214,
      "step": 1590,
      "train_runtime": 4101.9081,
      "train_tokens_per_second": 1066.385
    },
    {
      "epoch": 0.12784507523252778,
      "grad_norm": 0.26424458622932434,
      "learning_rate": 9.415692821368949e-05,
      "loss": 0.8172,
      "num_input_tokens_seen": 4376924,
      "step": 1591,
      "train_runtime": 4105.9222,
      "train_tokens_per_second": 1066.003
    },
    {
      "epoch": 0.12792543040237853,
      "grad_norm": 0.2841605842113495,
      "learning_rate": 9.409015025041736e-05,
      "loss": 0.8139,
      "num_input_tokens_seen": 4380006,
      "step": 1592,
      "train_runtime": 4108.3375,
      "train_tokens_per_second": 1066.126
    },
    {
      "epoch": 0.12800578557222925,
      "grad_norm": 0.33027318120002747,
      "learning_rate": 9.402337228714525e-05,
      "loss": 0.8235,
      "num_input_tokens_seen": 4381974,
      "step": 1593,
      "train_runtime": 4110.1363,
      "train_tokens_per_second": 1066.138
    },
    {
      "epoch": 0.12808614074208,
      "grad_norm": 0.3513645529747009,
      "learning_rate": 9.395659432387312e-05,
      "loss": 0.9179,
      "num_input_tokens_seen": 4384918,
      "step": 1594,
      "train_runtime": 4112.326,
      "train_tokens_per_second": 1066.287
    },
    {
      "epoch": 0.12816649591193074,
      "grad_norm": 0.32416418194770813,
      "learning_rate": 9.388981636060101e-05,
      "loss": 1.1617,
      "num_input_tokens_seen": 4386724,
      "step": 1595,
      "train_runtime": 4114.0026,
      "train_tokens_per_second": 1066.291
    },
    {
      "epoch": 0.12824685108178147,
      "grad_norm": 0.39823058247566223,
      "learning_rate": 9.38230383973289e-05,
      "loss": 1.329,
      "num_input_tokens_seen": 4391324,
      "step": 1596,
      "train_runtime": 4117.4051,
      "train_tokens_per_second": 1066.527
    },
    {
      "epoch": 0.12832720625163221,
      "grad_norm": 0.3567141890525818,
      "learning_rate": 9.375626043405677e-05,
      "loss": 1.0886,
      "num_input_tokens_seen": 4394018,
      "step": 1597,
      "train_runtime": 4119.5702,
      "train_tokens_per_second": 1066.62
    },
    {
      "epoch": 0.12840756142148296,
      "grad_norm": 0.32280534505844116,
      "learning_rate": 9.368948247078465e-05,
      "loss": 1.2491,
      "num_input_tokens_seen": 4396968,
      "step": 1598,
      "train_runtime": 4121.8578,
      "train_tokens_per_second": 1066.744
    },
    {
      "epoch": 0.12848791659133368,
      "grad_norm": 0.33652475476264954,
      "learning_rate": 9.362270450751253e-05,
      "loss": 0.9289,
      "num_input_tokens_seen": 4400246,
      "step": 1599,
      "train_runtime": 4124.3122,
      "train_tokens_per_second": 1066.904
    },
    {
      "epoch": 0.12856827176118443,
      "grad_norm": 0.3057388961315155,
      "learning_rate": 9.35559265442404e-05,
      "loss": 0.9814,
      "num_input_tokens_seen": 4403088,
      "step": 1600,
      "train_runtime": 4126.5194,
      "train_tokens_per_second": 1067.022
    },
    {
      "epoch": 0.12864862693103518,
      "grad_norm": 0.3203054666519165,
      "learning_rate": 9.348914858096829e-05,
      "loss": 0.9271,
      "num_input_tokens_seen": 4407068,
      "step": 1601,
      "train_runtime": 4129.3721,
      "train_tokens_per_second": 1067.249
    },
    {
      "epoch": 0.1287289821008859,
      "grad_norm": 0.3386048376560211,
      "learning_rate": 9.342237061769616e-05,
      "loss": 0.987,
      "num_input_tokens_seen": 4408798,
      "step": 1602,
      "train_runtime": 4131.0341,
      "train_tokens_per_second": 1067.238
    },
    {
      "epoch": 0.12880933727073665,
      "grad_norm": 0.25215280055999756,
      "learning_rate": 9.335559265442403e-05,
      "loss": 0.7593,
      "num_input_tokens_seen": 4413646,
      "step": 1603,
      "train_runtime": 4134.3417,
      "train_tokens_per_second": 1067.557
    },
    {
      "epoch": 0.1288896924405874,
      "grad_norm": 0.26832783222198486,
      "learning_rate": 9.328881469115192e-05,
      "loss": 1.0153,
      "num_input_tokens_seen": 4415762,
      "step": 1604,
      "train_runtime": 4136.1728,
      "train_tokens_per_second": 1067.596
    },
    {
      "epoch": 0.12897004761043815,
      "grad_norm": 0.42433273792266846,
      "learning_rate": 9.322203672787981e-05,
      "loss": 0.9614,
      "num_input_tokens_seen": 4418652,
      "step": 1605,
      "train_runtime": 4138.4507,
      "train_tokens_per_second": 1067.707
    },
    {
      "epoch": 0.12905040278028887,
      "grad_norm": 0.2514105439186096,
      "learning_rate": 9.31552587646077e-05,
      "loss": 0.8276,
      "num_input_tokens_seen": 4420204,
      "step": 1606,
      "train_runtime": 4140.0388,
      "train_tokens_per_second": 1067.672
    },
    {
      "epoch": 0.12913075795013962,
      "grad_norm": 0.2524038851261139,
      "learning_rate": 9.308848080133557e-05,
      "loss": 1.2064,
      "num_input_tokens_seen": 4422934,
      "step": 1607,
      "train_runtime": 4142.2247,
      "train_tokens_per_second": 1067.768
    },
    {
      "epoch": 0.12921111311999037,
      "grad_norm": 0.29408514499664307,
      "learning_rate": 9.302170283806344e-05,
      "loss": 1.0602,
      "num_input_tokens_seen": 4425688,
      "step": 1608,
      "train_runtime": 4144.3247,
      "train_tokens_per_second": 1067.891
    },
    {
      "epoch": 0.1292914682898411,
      "grad_norm": 0.2821078896522522,
      "learning_rate": 9.295492487479133e-05,
      "loss": 0.8052,
      "num_input_tokens_seen": 4428300,
      "step": 1609,
      "train_runtime": 4146.572,
      "train_tokens_per_second": 1067.942
    },
    {
      "epoch": 0.12937182345969184,
      "grad_norm": 0.25574159622192383,
      "learning_rate": 9.28881469115192e-05,
      "loss": 0.7596,
      "num_input_tokens_seen": 4430626,
      "step": 1610,
      "train_runtime": 4148.5215,
      "train_tokens_per_second": 1068.001
    },
    {
      "epoch": 0.12945217862954259,
      "grad_norm": 0.28250545263290405,
      "learning_rate": 9.282136894824707e-05,
      "loss": 0.955,
      "num_input_tokens_seen": 4432484,
      "step": 1611,
      "train_runtime": 4150.2126,
      "train_tokens_per_second": 1068.014
    },
    {
      "epoch": 0.1295325337993933,
      "grad_norm": 0.27344927191734314,
      "learning_rate": 9.275459098497496e-05,
      "loss": 0.5724,
      "num_input_tokens_seen": 4437890,
      "step": 1612,
      "train_runtime": 4153.9689,
      "train_tokens_per_second": 1068.349
    },
    {
      "epoch": 0.12961288896924406,
      "grad_norm": 0.3148863613605499,
      "learning_rate": 9.268781302170285e-05,
      "loss": 0.9932,
      "num_input_tokens_seen": 4440388,
      "step": 1613,
      "train_runtime": 4156.051,
      "train_tokens_per_second": 1068.415
    },
    {
      "epoch": 0.1296932441390948,
      "grad_norm": 0.3240756690502167,
      "learning_rate": 9.262103505843073e-05,
      "loss": 0.8635,
      "num_input_tokens_seen": 4442402,
      "step": 1614,
      "train_runtime": 4157.9847,
      "train_tokens_per_second": 1068.403
    },
    {
      "epoch": 0.12977359930894555,
      "grad_norm": 0.3345417082309723,
      "learning_rate": 9.255425709515861e-05,
      "loss": 1.1861,
      "num_input_tokens_seen": 4444962,
      "step": 1615,
      "train_runtime": 4160.0953,
      "train_tokens_per_second": 1068.476
    },
    {
      "epoch": 0.12985395447879627,
      "grad_norm": 0.33316949009895325,
      "learning_rate": 9.248747913188648e-05,
      "loss": 1.028,
      "num_input_tokens_seen": 4449454,
      "step": 1616,
      "train_runtime": 4163.2201,
      "train_tokens_per_second": 1068.753
    },
    {
      "epoch": 0.12993430964864702,
      "grad_norm": 0.49215489625930786,
      "learning_rate": 9.242070116861437e-05,
      "loss": 0.9807,
      "num_input_tokens_seen": 4451866,
      "step": 1617,
      "train_runtime": 4165.1562,
      "train_tokens_per_second": 1068.835
    },
    {
      "epoch": 0.13001466481849777,
      "grad_norm": 0.3640338182449341,
      "learning_rate": 9.235392320534224e-05,
      "loss": 0.9479,
      "num_input_tokens_seen": 4453300,
      "step": 1618,
      "train_runtime": 4166.7527,
      "train_tokens_per_second": 1068.77
    },
    {
      "epoch": 0.1300950199883485,
      "grad_norm": 0.24760620296001434,
      "learning_rate": 9.228714524207011e-05,
      "loss": 0.6054,
      "num_input_tokens_seen": 4456352,
      "step": 1619,
      "train_runtime": 4169.1115,
      "train_tokens_per_second": 1068.897
    },
    {
      "epoch": 0.13017537515819924,
      "grad_norm": 0.37118616700172424,
      "learning_rate": 9.2220367278798e-05,
      "loss": 0.9708,
      "num_input_tokens_seen": 4459332,
      "step": 1620,
      "train_runtime": 4171.557,
      "train_tokens_per_second": 1068.985
    },
    {
      "epoch": 0.13025573032805,
      "grad_norm": 0.30011922121047974,
      "learning_rate": 9.215358931552587e-05,
      "loss": 1.076,
      "num_input_tokens_seen": 4461230,
      "step": 1621,
      "train_runtime": 4174.9637,
      "train_tokens_per_second": 1068.567
    },
    {
      "epoch": 0.1303360854979007,
      "grad_norm": 0.35796383023262024,
      "learning_rate": 9.208681135225376e-05,
      "loss": 0.8814,
      "num_input_tokens_seen": 4463488,
      "step": 1622,
      "train_runtime": 4176.9112,
      "train_tokens_per_second": 1068.61
    },
    {
      "epoch": 0.13041644066775146,
      "grad_norm": 0.3439129889011383,
      "learning_rate": 9.202003338898165e-05,
      "loss": 0.987,
      "num_input_tokens_seen": 4467164,
      "step": 1623,
      "train_runtime": 4179.5879,
      "train_tokens_per_second": 1068.805
    },
    {
      "epoch": 0.1304967958376022,
      "grad_norm": 0.337166965007782,
      "learning_rate": 9.195325542570952e-05,
      "loss": 1.0371,
      "num_input_tokens_seen": 4470520,
      "step": 1624,
      "train_runtime": 4182.1445,
      "train_tokens_per_second": 1068.954
    },
    {
      "epoch": 0.13057715100745293,
      "grad_norm": 0.3797534108161926,
      "learning_rate": 9.18864774624374e-05,
      "loss": 0.846,
      "num_input_tokens_seen": 4471852,
      "step": 1625,
      "train_runtime": 4183.6934,
      "train_tokens_per_second": 1068.877
    },
    {
      "epoch": 0.13065750617730368,
      "grad_norm": 0.25332966446876526,
      "learning_rate": 9.181969949916528e-05,
      "loss": 0.8772,
      "num_input_tokens_seen": 4475116,
      "step": 1626,
      "train_runtime": 4186.1983,
      "train_tokens_per_second": 1069.017
    },
    {
      "epoch": 0.13073786134715443,
      "grad_norm": 0.31641215085983276,
      "learning_rate": 9.175292153589315e-05,
      "loss": 0.852,
      "num_input_tokens_seen": 4476870,
      "step": 1627,
      "train_runtime": 4187.8897,
      "train_tokens_per_second": 1069.004
    },
    {
      "epoch": 0.13081821651700518,
      "grad_norm": 0.30552247166633606,
      "learning_rate": 9.168614357262104e-05,
      "loss": 1.04,
      "num_input_tokens_seen": 4479824,
      "step": 1628,
      "train_runtime": 4190.0846,
      "train_tokens_per_second": 1069.149
    },
    {
      "epoch": 0.1308985716868559,
      "grad_norm": 0.28019553422927856,
      "learning_rate": 9.161936560934891e-05,
      "loss": 1.0666,
      "num_input_tokens_seen": 4482186,
      "step": 1629,
      "train_runtime": 4192.0882,
      "train_tokens_per_second": 1069.201
    },
    {
      "epoch": 0.13097892685670665,
      "grad_norm": 0.28002801537513733,
      "learning_rate": 9.15525876460768e-05,
      "loss": 0.9712,
      "num_input_tokens_seen": 4484060,
      "step": 1630,
      "train_runtime": 4193.8702,
      "train_tokens_per_second": 1069.194
    },
    {
      "epoch": 0.1310592820265574,
      "grad_norm": 0.25342047214508057,
      "learning_rate": 9.148580968280469e-05,
      "loss": 0.7143,
      "num_input_tokens_seen": 4487294,
      "step": 1631,
      "train_runtime": 4196.3134,
      "train_tokens_per_second": 1069.342
    },
    {
      "epoch": 0.13113963719640812,
      "grad_norm": 0.2778625786304474,
      "learning_rate": 9.141903171953256e-05,
      "loss": 0.7966,
      "num_input_tokens_seen": 4489094,
      "step": 1632,
      "train_runtime": 4198.1966,
      "train_tokens_per_second": 1069.291
    },
    {
      "epoch": 0.13121999236625886,
      "grad_norm": 0.33734869956970215,
      "learning_rate": 9.135225375626045e-05,
      "loss": 0.8386,
      "num_input_tokens_seen": 4491332,
      "step": 1633,
      "train_runtime": 4200.204,
      "train_tokens_per_second": 1069.313
    },
    {
      "epoch": 0.1313003475361096,
      "grad_norm": 0.3175203800201416,
      "learning_rate": 9.128547579298832e-05,
      "loss": 0.7832,
      "num_input_tokens_seen": 4493968,
      "step": 1634,
      "train_runtime": 4202.4361,
      "train_tokens_per_second": 1069.372
    },
    {
      "epoch": 0.13138070270596033,
      "grad_norm": 0.3065336346626282,
      "learning_rate": 9.121869782971619e-05,
      "loss": 0.8444,
      "num_input_tokens_seen": 4495802,
      "step": 1635,
      "train_runtime": 4204.2043,
      "train_tokens_per_second": 1069.359
    },
    {
      "epoch": 0.13146105787581108,
      "grad_norm": 0.3828697204589844,
      "learning_rate": 9.115191986644408e-05,
      "loss": 1.0292,
      "num_input_tokens_seen": 4497860,
      "step": 1636,
      "train_runtime": 4205.9624,
      "train_tokens_per_second": 1069.401
    },
    {
      "epoch": 0.13154141304566183,
      "grad_norm": 0.2931973338127136,
      "learning_rate": 9.108514190317195e-05,
      "loss": 1.1776,
      "num_input_tokens_seen": 4500282,
      "step": 1637,
      "train_runtime": 4207.9363,
      "train_tokens_per_second": 1069.475
    },
    {
      "epoch": 0.13162176821551255,
      "grad_norm": 0.2747010886669159,
      "learning_rate": 9.101836393989984e-05,
      "loss": 0.8025,
      "num_input_tokens_seen": 4503798,
      "step": 1638,
      "train_runtime": 4210.496,
      "train_tokens_per_second": 1069.66
    },
    {
      "epoch": 0.1317021233853633,
      "grad_norm": 0.3100590407848358,
      "learning_rate": 9.095158597662771e-05,
      "loss": 0.8297,
      "num_input_tokens_seen": 4506974,
      "step": 1639,
      "train_runtime": 4212.9528,
      "train_tokens_per_second": 1069.79
    },
    {
      "epoch": 0.13178247855521405,
      "grad_norm": 0.30835211277008057,
      "learning_rate": 9.08848080133556e-05,
      "loss": 1.0038,
      "num_input_tokens_seen": 4509138,
      "step": 1640,
      "train_runtime": 4214.8822,
      "train_tokens_per_second": 1069.814
    },
    {
      "epoch": 0.1318628337250648,
      "grad_norm": 0.20513121783733368,
      "learning_rate": 9.081803005008348e-05,
      "loss": 0.3474,
      "num_input_tokens_seen": 4511390,
      "step": 1641,
      "train_runtime": 4216.8515,
      "train_tokens_per_second": 1069.848
    },
    {
      "epoch": 0.13194318889491552,
      "grad_norm": 0.31580573320388794,
      "learning_rate": 9.075125208681136e-05,
      "loss": 0.8266,
      "num_input_tokens_seen": 4513928,
      "step": 1642,
      "train_runtime": 4218.9634,
      "train_tokens_per_second": 1069.914
    },
    {
      "epoch": 0.13202354406476627,
      "grad_norm": 0.4065244197845459,
      "learning_rate": 9.068447412353923e-05,
      "loss": 1.0224,
      "num_input_tokens_seen": 4515578,
      "step": 1643,
      "train_runtime": 4220.645,
      "train_tokens_per_second": 1069.879
    },
    {
      "epoch": 0.13210389923461702,
      "grad_norm": 0.3030647039413452,
      "learning_rate": 9.061769616026712e-05,
      "loss": 1.0523,
      "num_input_tokens_seen": 4517324,
      "step": 1644,
      "train_runtime": 4222.3246,
      "train_tokens_per_second": 1069.867
    },
    {
      "epoch": 0.13218425440446774,
      "grad_norm": 0.36461055278778076,
      "learning_rate": 9.055091819699499e-05,
      "loss": 0.8365,
      "num_input_tokens_seen": 4519522,
      "step": 1645,
      "train_runtime": 4224.2701,
      "train_tokens_per_second": 1069.894
    },
    {
      "epoch": 0.1322646095743185,
      "grad_norm": 0.42367351055145264,
      "learning_rate": 9.048414023372288e-05,
      "loss": 0.9529,
      "num_input_tokens_seen": 4520928,
      "step": 1646,
      "train_runtime": 4225.8589,
      "train_tokens_per_second": 1069.825
    },
    {
      "epoch": 0.13234496474416924,
      "grad_norm": 0.25916787981987,
      "learning_rate": 9.041736227045075e-05,
      "loss": 0.8301,
      "num_input_tokens_seen": 4523560,
      "step": 1647,
      "train_runtime": 4228.0031,
      "train_tokens_per_second": 1069.905
    },
    {
      "epoch": 0.13242531991401996,
      "grad_norm": 0.2737148702144623,
      "learning_rate": 9.035058430717864e-05,
      "loss": 0.9342,
      "num_input_tokens_seen": 4526918,
      "step": 1648,
      "train_runtime": 4230.5098,
      "train_tokens_per_second": 1070.064
    },
    {
      "epoch": 0.1325056750838707,
      "grad_norm": 0.29724356532096863,
      "learning_rate": 9.028380634390652e-05,
      "loss": 0.7756,
      "num_input_tokens_seen": 4530596,
      "step": 1649,
      "train_runtime": 4233.2695,
      "train_tokens_per_second": 1070.236
    },
    {
      "epoch": 0.13258603025372145,
      "grad_norm": 0.31504523754119873,
      "learning_rate": 9.02170283806344e-05,
      "loss": 0.969,
      "num_input_tokens_seen": 4531860,
      "step": 1650,
      "train_runtime": 4234.7892,
      "train_tokens_per_second": 1070.15
    },
    {
      "epoch": 0.1326663854235722,
      "grad_norm": 0.2676130533218384,
      "learning_rate": 9.015025041736227e-05,
      "loss": 0.8469,
      "num_input_tokens_seen": 4535260,
      "step": 1651,
      "train_runtime": 4239.0719,
      "train_tokens_per_second": 1069.871
    },
    {
      "epoch": 0.13274674059342292,
      "grad_norm": 0.3291208744049072,
      "learning_rate": 9.008347245409016e-05,
      "loss": 1.3977,
      "num_input_tokens_seen": 4538326,
      "step": 1652,
      "train_runtime": 4241.4146,
      "train_tokens_per_second": 1070.003
    },
    {
      "epoch": 0.13282709576327367,
      "grad_norm": 0.3381328284740448,
      "learning_rate": 9.001669449081803e-05,
      "loss": 1.0963,
      "num_input_tokens_seen": 4541152,
      "step": 1653,
      "train_runtime": 4243.6372,
      "train_tokens_per_second": 1070.108
    },
    {
      "epoch": 0.13290745093312442,
      "grad_norm": 0.33428165316581726,
      "learning_rate": 8.994991652754592e-05,
      "loss": 1.0073,
      "num_input_tokens_seen": 4543384,
      "step": 1654,
      "train_runtime": 4245.6022,
      "train_tokens_per_second": 1070.139
    },
    {
      "epoch": 0.13298780610297514,
      "grad_norm": 0.45007333159446716,
      "learning_rate": 8.988313856427379e-05,
      "loss": 1.1264,
      "num_input_tokens_seen": 4545610,
      "step": 1655,
      "train_runtime": 4247.5942,
      "train_tokens_per_second": 1070.161
    },
    {
      "epoch": 0.1330681612728259,
      "grad_norm": 0.24271412193775177,
      "learning_rate": 8.981636060100166e-05,
      "loss": 0.7438,
      "num_input_tokens_seen": 4548880,
      "step": 1656,
      "train_runtime": 4250.0792,
      "train_tokens_per_second": 1070.305
    },
    {
      "epoch": 0.13314851644267664,
      "grad_norm": 0.3081810772418976,
      "learning_rate": 8.974958263772955e-05,
      "loss": 1.0669,
      "num_input_tokens_seen": 4554630,
      "step": 1657,
      "train_runtime": 4254.0066,
      "train_tokens_per_second": 1070.668
    },
    {
      "epoch": 0.13322887161252736,
      "grad_norm": 0.3170272707939148,
      "learning_rate": 8.968280467445744e-05,
      "loss": 1.0973,
      "num_input_tokens_seen": 4558384,
      "step": 1658,
      "train_runtime": 4256.7557,
      "train_tokens_per_second": 1070.859
    },
    {
      "epoch": 0.1333092267823781,
      "grad_norm": 0.28447824716567993,
      "learning_rate": 8.961602671118531e-05,
      "loss": 0.6721,
      "num_input_tokens_seen": 4560850,
      "step": 1659,
      "train_runtime": 4258.8203,
      "train_tokens_per_second": 1070.919
    },
    {
      "epoch": 0.13338958195222886,
      "grad_norm": 0.24304047226905823,
      "learning_rate": 8.95492487479132e-05,
      "loss": 0.8053,
      "num_input_tokens_seen": 4563652,
      "step": 1660,
      "train_runtime": 4261.0141,
      "train_tokens_per_second": 1071.025
    },
    {
      "epoch": 0.13346993712207958,
      "grad_norm": 0.3356163799762726,
      "learning_rate": 8.948247078464107e-05,
      "loss": 0.9594,
      "num_input_tokens_seen": 4566372,
      "step": 1661,
      "train_runtime": 4263.1376,
      "train_tokens_per_second": 1071.129
    },
    {
      "epoch": 0.13355029229193033,
      "grad_norm": 0.35904183983802795,
      "learning_rate": 8.941569282136896e-05,
      "loss": 1.0483,
      "num_input_tokens_seen": 4569306,
      "step": 1662,
      "train_runtime": 4265.3941,
      "train_tokens_per_second": 1071.251
    },
    {
      "epoch": 0.13363064746178108,
      "grad_norm": 0.350241094827652,
      "learning_rate": 8.934891485809683e-05,
      "loss": 0.9494,
      "num_input_tokens_seen": 4571786,
      "step": 1663,
      "train_runtime": 4267.4207,
      "train_tokens_per_second": 1071.323
    },
    {
      "epoch": 0.13371100263163183,
      "grad_norm": 0.2899452745914459,
      "learning_rate": 8.92821368948247e-05,
      "loss": 0.9646,
      "num_input_tokens_seen": 4573962,
      "step": 1664,
      "train_runtime": 4269.4142,
      "train_tokens_per_second": 1071.332
    },
    {
      "epoch": 0.13379135780148255,
      "grad_norm": 0.3941836357116699,
      "learning_rate": 8.921535893155259e-05,
      "loss": 1.0085,
      "num_input_tokens_seen": 4576108,
      "step": 1665,
      "train_runtime": 4271.3959,
      "train_tokens_per_second": 1071.338
    },
    {
      "epoch": 0.1338717129713333,
      "grad_norm": 0.3039100468158722,
      "learning_rate": 8.914858096828048e-05,
      "loss": 1.0543,
      "num_input_tokens_seen": 4579370,
      "step": 1666,
      "train_runtime": 4273.8943,
      "train_tokens_per_second": 1071.475
    },
    {
      "epoch": 0.13395206814118404,
      "grad_norm": 0.3556115925312042,
      "learning_rate": 8.908180300500835e-05,
      "loss": 1.0549,
      "num_input_tokens_seen": 4581778,
      "step": 1667,
      "train_runtime": 4275.9314,
      "train_tokens_per_second": 1071.527
    },
    {
      "epoch": 0.13403242331103477,
      "grad_norm": 0.32420018315315247,
      "learning_rate": 8.901502504173624e-05,
      "loss": 0.8315,
      "num_input_tokens_seen": 4584206,
      "step": 1668,
      "train_runtime": 4277.9082,
      "train_tokens_per_second": 1071.6
    },
    {
      "epoch": 0.13411277848088551,
      "grad_norm": 0.397775262594223,
      "learning_rate": 8.894824707846411e-05,
      "loss": 0.8873,
      "num_input_tokens_seen": 4587772,
      "step": 1669,
      "train_runtime": 4280.5771,
      "train_tokens_per_second": 1071.765
    },
    {
      "epoch": 0.13419313365073626,
      "grad_norm": 0.3347060978412628,
      "learning_rate": 8.8881469115192e-05,
      "loss": 0.8835,
      "num_input_tokens_seen": 4589850,
      "step": 1670,
      "train_runtime": 4282.4087,
      "train_tokens_per_second": 1071.792
    },
    {
      "epoch": 0.13427348882058698,
      "grad_norm": 0.3155038058757782,
      "learning_rate": 8.881469115191987e-05,
      "loss": 0.9874,
      "num_input_tokens_seen": 4594972,
      "step": 1671,
      "train_runtime": 4286.0552,
      "train_tokens_per_second": 1072.075
    },
    {
      "epoch": 0.13435384399043773,
      "grad_norm": 0.41316670179367065,
      "learning_rate": 8.874791318864774e-05,
      "loss": 1.0672,
      "num_input_tokens_seen": 4597130,
      "step": 1672,
      "train_runtime": 4288.0242,
      "train_tokens_per_second": 1072.086
    },
    {
      "epoch": 0.13443419916028848,
      "grad_norm": 0.3476656973361969,
      "learning_rate": 8.868113522537563e-05,
      "loss": 1.159,
      "num_input_tokens_seen": 4600874,
      "step": 1673,
      "train_runtime": 4290.758,
      "train_tokens_per_second": 1072.275
    },
    {
      "epoch": 0.1345145543301392,
      "grad_norm": 0.33450794219970703,
      "learning_rate": 8.86143572621035e-05,
      "loss": 1.158,
      "num_input_tokens_seen": 4602716,
      "step": 1674,
      "train_runtime": 4292.4431,
      "train_tokens_per_second": 1072.284
    },
    {
      "epoch": 0.13459490949998995,
      "grad_norm": 0.3329657018184662,
      "learning_rate": 8.854757929883139e-05,
      "loss": 1.1793,
      "num_input_tokens_seen": 4605052,
      "step": 1675,
      "train_runtime": 4294.4282,
      "train_tokens_per_second": 1072.332
    },
    {
      "epoch": 0.1346752646698407,
      "grad_norm": 0.37308019399642944,
      "learning_rate": 8.848080133555928e-05,
      "loss": 1.1019,
      "num_input_tokens_seen": 4609106,
      "step": 1676,
      "train_runtime": 4297.3189,
      "train_tokens_per_second": 1072.554
    },
    {
      "epoch": 0.13475561983969145,
      "grad_norm": 0.30606627464294434,
      "learning_rate": 8.841402337228715e-05,
      "loss": 0.8794,
      "num_input_tokens_seen": 4611216,
      "step": 1677,
      "train_runtime": 4299.1549,
      "train_tokens_per_second": 1072.587
    },
    {
      "epoch": 0.13483597500954217,
      "grad_norm": 0.3290914297103882,
      "learning_rate": 8.834724540901504e-05,
      "loss": 1.0011,
      "num_input_tokens_seen": 4613654,
      "step": 1678,
      "train_runtime": 4301.1824,
      "train_tokens_per_second": 1072.648
    },
    {
      "epoch": 0.13491633017939292,
      "grad_norm": 0.33814892172813416,
      "learning_rate": 8.828046744574291e-05,
      "loss": 0.8095,
      "num_input_tokens_seen": 4615598,
      "step": 1679,
      "train_runtime": 4303.0239,
      "train_tokens_per_second": 1072.641
    },
    {
      "epoch": 0.13499668534924367,
      "grad_norm": 0.3479299545288086,
      "learning_rate": 8.821368948247078e-05,
      "loss": 0.8497,
      "num_input_tokens_seen": 4618324,
      "step": 1680,
      "train_runtime": 4305.2634,
      "train_tokens_per_second": 1072.716
    },
    {
      "epoch": 0.1350770405190944,
      "grad_norm": 0.3423509895801544,
      "learning_rate": 8.814691151919867e-05,
      "loss": 1.1246,
      "num_input_tokens_seen": 4620580,
      "step": 1681,
      "train_runtime": 4309.372,
      "train_tokens_per_second": 1072.217
    },
    {
      "epoch": 0.13515739568894514,
      "grad_norm": 0.3519035875797272,
      "learning_rate": 8.808013355592654e-05,
      "loss": 1.0326,
      "num_input_tokens_seen": 4622674,
      "step": 1682,
      "train_runtime": 4311.2041,
      "train_tokens_per_second": 1072.247
    },
    {
      "epoch": 0.13523775085879589,
      "grad_norm": 0.31621867418289185,
      "learning_rate": 8.801335559265443e-05,
      "loss": 1.0432,
      "num_input_tokens_seen": 4625648,
      "step": 1683,
      "train_runtime": 4313.5557,
      "train_tokens_per_second": 1072.352
    },
    {
      "epoch": 0.1353181060286466,
      "grad_norm": 0.4220649302005768,
      "learning_rate": 8.794657762938232e-05,
      "loss": 0.9385,
      "num_input_tokens_seen": 4629020,
      "step": 1684,
      "train_runtime": 4316.0764,
      "train_tokens_per_second": 1072.507
    },
    {
      "epoch": 0.13539846119849736,
      "grad_norm": 0.2913580536842346,
      "learning_rate": 8.787979966611019e-05,
      "loss": 0.8173,
      "num_input_tokens_seen": 4631074,
      "step": 1685,
      "train_runtime": 4317.9804,
      "train_tokens_per_second": 1072.509
    },
    {
      "epoch": 0.1354788163683481,
      "grad_norm": 0.2683812975883484,
      "learning_rate": 8.781302170283808e-05,
      "loss": 1.08,
      "num_input_tokens_seen": 4635102,
      "step": 1686,
      "train_runtime": 4320.904,
      "train_tokens_per_second": 1072.716
    },
    {
      "epoch": 0.13555917153819885,
      "grad_norm": 0.3216564655303955,
      "learning_rate": 8.774624373956595e-05,
      "loss": 0.8944,
      "num_input_tokens_seen": 4637268,
      "step": 1687,
      "train_runtime": 4322.7587,
      "train_tokens_per_second": 1072.757
    },
    {
      "epoch": 0.13563952670804957,
      "grad_norm": 0.2993555963039398,
      "learning_rate": 8.767946577629382e-05,
      "loss": 0.8588,
      "num_input_tokens_seen": 4640094,
      "step": 1688,
      "train_runtime": 4325.0756,
      "train_tokens_per_second": 1072.835
    },
    {
      "epoch": 0.13571988187790032,
      "grad_norm": 0.3338256776332855,
      "learning_rate": 8.761268781302171e-05,
      "loss": 0.9787,
      "num_input_tokens_seen": 4642262,
      "step": 1689,
      "train_runtime": 4326.918,
      "train_tokens_per_second": 1072.88
    },
    {
      "epoch": 0.13580023704775107,
      "grad_norm": 0.3616194427013397,
      "learning_rate": 8.754590984974958e-05,
      "loss": 1.0568,
      "num_input_tokens_seen": 4644614,
      "step": 1690,
      "train_runtime": 4328.9495,
      "train_tokens_per_second": 1072.919
    },
    {
      "epoch": 0.1358805922176018,
      "grad_norm": 0.2871371805667877,
      "learning_rate": 8.747913188647745e-05,
      "loss": 0.8855,
      "num_input_tokens_seen": 4646220,
      "step": 1691,
      "train_runtime": 4330.6302,
      "train_tokens_per_second": 1072.874
    },
    {
      "epoch": 0.13596094738745254,
      "grad_norm": 0.3189316689968109,
      "learning_rate": 8.741235392320535e-05,
      "loss": 1.1744,
      "num_input_tokens_seen": 4649310,
      "step": 1692,
      "train_runtime": 4333.021,
      "train_tokens_per_second": 1072.995
    },
    {
      "epoch": 0.1360413025573033,
      "grad_norm": 0.3624874949455261,
      "learning_rate": 8.734557595993323e-05,
      "loss": 0.8644,
      "num_input_tokens_seen": 4651880,
      "step": 1693,
      "train_runtime": 4335.0983,
      "train_tokens_per_second": 1073.074
    },
    {
      "epoch": 0.136121657727154,
      "grad_norm": 0.27284377813339233,
      "learning_rate": 8.727879799666111e-05,
      "loss": 0.6774,
      "num_input_tokens_seen": 4653500,
      "step": 1694,
      "train_runtime": 4336.6965,
      "train_tokens_per_second": 1073.052
    },
    {
      "epoch": 0.13620201289700476,
      "grad_norm": 0.24099037051200867,
      "learning_rate": 8.721202003338899e-05,
      "loss": 0.8722,
      "num_input_tokens_seen": 4656814,
      "step": 1695,
      "train_runtime": 4339.1886,
      "train_tokens_per_second": 1073.199
    },
    {
      "epoch": 0.1362823680668555,
      "grad_norm": 0.23552261292934418,
      "learning_rate": 8.714524207011686e-05,
      "loss": 0.4232,
      "num_input_tokens_seen": 4659278,
      "step": 1696,
      "train_runtime": 4341.3146,
      "train_tokens_per_second": 1073.241
    },
    {
      "epoch": 0.13636272323670623,
      "grad_norm": 0.3358173072338104,
      "learning_rate": 8.707846410684475e-05,
      "loss": 0.9733,
      "num_input_tokens_seen": 4661962,
      "step": 1697,
      "train_runtime": 4343.4613,
      "train_tokens_per_second": 1073.329
    },
    {
      "epoch": 0.13644307840655698,
      "grad_norm": 0.5365708470344543,
      "learning_rate": 8.701168614357262e-05,
      "loss": 1.0802,
      "num_input_tokens_seen": 4666956,
      "step": 1698,
      "train_runtime": 4346.9631,
      "train_tokens_per_second": 1073.613
    },
    {
      "epoch": 0.13652343357640773,
      "grad_norm": 0.3458505868911743,
      "learning_rate": 8.694490818030051e-05,
      "loss": 0.8993,
      "num_input_tokens_seen": 4668718,
      "step": 1699,
      "train_runtime": 4348.6378,
      "train_tokens_per_second": 1073.605
    },
    {
      "epoch": 0.13660378874625848,
      "grad_norm": 0.30510833859443665,
      "learning_rate": 8.687813021702838e-05,
      "loss": 1.0165,
      "num_input_tokens_seen": 4671682,
      "step": 1700,
      "train_runtime": 4350.8413,
      "train_tokens_per_second": 1073.742
    },
    {
      "epoch": 0.1366841439161092,
      "grad_norm": 0.3102036118507385,
      "learning_rate": 8.681135225375627e-05,
      "loss": 1.0339,
      "num_input_tokens_seen": 4673562,
      "step": 1701,
      "train_runtime": 4352.6165,
      "train_tokens_per_second": 1073.736
    },
    {
      "epoch": 0.13676449908595995,
      "grad_norm": 0.35591694712638855,
      "learning_rate": 8.674457429048415e-05,
      "loss": 0.8726,
      "num_input_tokens_seen": 4675602,
      "step": 1702,
      "train_runtime": 4354.4034,
      "train_tokens_per_second": 1073.764
    },
    {
      "epoch": 0.1368448542558107,
      "grad_norm": 0.325438916683197,
      "learning_rate": 8.667779632721203e-05,
      "loss": 1.001,
      "num_input_tokens_seen": 4677726,
      "step": 1703,
      "train_runtime": 4356.4042,
      "train_tokens_per_second": 1073.759
    },
    {
      "epoch": 0.13692520942566141,
      "grad_norm": 0.2606280446052551,
      "learning_rate": 8.66110183639399e-05,
      "loss": 0.8853,
      "num_input_tokens_seen": 4680866,
      "step": 1704,
      "train_runtime": 4358.7825,
      "train_tokens_per_second": 1073.893
    },
    {
      "epoch": 0.13700556459551216,
      "grad_norm": 0.26266175508499146,
      "learning_rate": 8.654424040066779e-05,
      "loss": 0.6576,
      "num_input_tokens_seen": 4683616,
      "step": 1705,
      "train_runtime": 4360.9884,
      "train_tokens_per_second": 1073.98
    },
    {
      "epoch": 0.1370859197653629,
      "grad_norm": 0.28277552127838135,
      "learning_rate": 8.647746243739566e-05,
      "loss": 1.3001,
      "num_input_tokens_seen": 4686820,
      "step": 1706,
      "train_runtime": 4363.3592,
      "train_tokens_per_second": 1074.131
    },
    {
      "epoch": 0.13716627493521363,
      "grad_norm": 0.3710005283355713,
      "learning_rate": 8.641068447412355e-05,
      "loss": 1.1717,
      "num_input_tokens_seen": 4689866,
      "step": 1707,
      "train_runtime": 4365.6024,
      "train_tokens_per_second": 1074.277
    },
    {
      "epoch": 0.13724663010506438,
      "grad_norm": 0.3213399350643158,
      "learning_rate": 8.634390651085142e-05,
      "loss": 1.0501,
      "num_input_tokens_seen": 4693132,
      "step": 1708,
      "train_runtime": 4368.0512,
      "train_tokens_per_second": 1074.422
    },
    {
      "epoch": 0.13732698527491513,
      "grad_norm": 0.4787834882736206,
      "learning_rate": 8.62771285475793e-05,
      "loss": 1.445,
      "num_input_tokens_seen": 4695088,
      "step": 1709,
      "train_runtime": 4369.9405,
      "train_tokens_per_second": 1074.405
    },
    {
      "epoch": 0.13740734044476585,
      "grad_norm": 0.29735395312309265,
      "learning_rate": 8.62103505843072e-05,
      "loss": 1.044,
      "num_input_tokens_seen": 4698032,
      "step": 1710,
      "train_runtime": 4372.2716,
      "train_tokens_per_second": 1074.506
    },
    {
      "epoch": 0.1374876956146166,
      "grad_norm": 0.3056807816028595,
      "learning_rate": 8.614357262103507e-05,
      "loss": 1.0037,
      "num_input_tokens_seen": 4700646,
      "step": 1711,
      "train_runtime": 4376.3613,
      "train_tokens_per_second": 1074.099
    },
    {
      "epoch": 0.13756805078446735,
      "grad_norm": 0.38576528429985046,
      "learning_rate": 8.607679465776294e-05,
      "loss": 1.1108,
      "num_input_tokens_seen": 4703298,
      "step": 1712,
      "train_runtime": 4378.5549,
      "train_tokens_per_second": 1074.167
    },
    {
      "epoch": 0.1376484059543181,
      "grad_norm": 0.33567014336586,
      "learning_rate": 8.601001669449083e-05,
      "loss": 1.1805,
      "num_input_tokens_seen": 4705322,
      "step": 1713,
      "train_runtime": 4380.3166,
      "train_tokens_per_second": 1074.197
    },
    {
      "epoch": 0.13772876112416882,
      "grad_norm": 0.26364225149154663,
      "learning_rate": 8.59432387312187e-05,
      "loss": 1.0343,
      "num_input_tokens_seen": 4707596,
      "step": 1714,
      "train_runtime": 4382.2285,
      "train_tokens_per_second": 1074.247
    },
    {
      "epoch": 0.13780911629401957,
      "grad_norm": 0.32710427045822144,
      "learning_rate": 8.587646076794659e-05,
      "loss": 0.9873,
      "num_input_tokens_seen": 4709850,
      "step": 1715,
      "train_runtime": 4384.2031,
      "train_tokens_per_second": 1074.277
    },
    {
      "epoch": 0.13788947146387032,
      "grad_norm": 0.9376202821731567,
      "learning_rate": 8.580968280467446e-05,
      "loss": 1.0839,
      "num_input_tokens_seen": 4713218,
      "step": 1716,
      "train_runtime": 4386.7057,
      "train_tokens_per_second": 1074.432
    },
    {
      "epoch": 0.13796982663372104,
      "grad_norm": 0.33692240715026855,
      "learning_rate": 8.574290484140233e-05,
      "loss": 1.1751,
      "num_input_tokens_seen": 4716474,
      "step": 1717,
      "train_runtime": 4389.1871,
      "train_tokens_per_second": 1074.567
    },
    {
      "epoch": 0.1380501818035718,
      "grad_norm": 0.29505711793899536,
      "learning_rate": 8.567612687813022e-05,
      "loss": 1.0749,
      "num_input_tokens_seen": 4718104,
      "step": 1718,
      "train_runtime": 4390.8474,
      "train_tokens_per_second": 1074.532
    },
    {
      "epoch": 0.13813053697342254,
      "grad_norm": 0.2989121675491333,
      "learning_rate": 8.56093489148581e-05,
      "loss": 0.9142,
      "num_input_tokens_seen": 4720348,
      "step": 1719,
      "train_runtime": 4392.7474,
      "train_tokens_per_second": 1074.578
    },
    {
      "epoch": 0.13821089214327326,
      "grad_norm": 0.35670560598373413,
      "learning_rate": 8.554257095158598e-05,
      "loss": 1.2714,
      "num_input_tokens_seen": 4722706,
      "step": 1720,
      "train_runtime": 4394.7899,
      "train_tokens_per_second": 1074.615
    },
    {
      "epoch": 0.138291247313124,
      "grad_norm": 0.28659960627555847,
      "learning_rate": 8.547579298831387e-05,
      "loss": 1.0231,
      "num_input_tokens_seen": 4727842,
      "step": 1721,
      "train_runtime": 4398.3018,
      "train_tokens_per_second": 1074.924
    },
    {
      "epoch": 0.13837160248297475,
      "grad_norm": 0.2947317957878113,
      "learning_rate": 8.540901502504174e-05,
      "loss": 0.9436,
      "num_input_tokens_seen": 4730294,
      "step": 1722,
      "train_runtime": 4400.3013,
      "train_tokens_per_second": 1074.993
    },
    {
      "epoch": 0.1384519576528255,
      "grad_norm": 0.3691924512386322,
      "learning_rate": 8.534223706176963e-05,
      "loss": 0.9322,
      "num_input_tokens_seen": 4732194,
      "step": 1723,
      "train_runtime": 4402.0304,
      "train_tokens_per_second": 1075.003
    },
    {
      "epoch": 0.13853231282267622,
      "grad_norm": 0.2930741012096405,
      "learning_rate": 8.52754590984975e-05,
      "loss": 0.9929,
      "num_input_tokens_seen": 4734300,
      "step": 1724,
      "train_runtime": 4403.8539,
      "train_tokens_per_second": 1075.036
    },
    {
      "epoch": 0.13861266799252697,
      "grad_norm": 0.3534330725669861,
      "learning_rate": 8.520868113522537e-05,
      "loss": 0.8906,
      "num_input_tokens_seen": 4736670,
      "step": 1725,
      "train_runtime": 4405.8647,
      "train_tokens_per_second": 1075.083
    },
    {
      "epoch": 0.13869302316237772,
      "grad_norm": 0.2637708783149719,
      "learning_rate": 8.514190317195326e-05,
      "loss": 0.7433,
      "num_input_tokens_seen": 4738966,
      "step": 1726,
      "train_runtime": 4407.8285,
      "train_tokens_per_second": 1075.125
    },
    {
      "epoch": 0.13877337833222844,
      "grad_norm": 0.34065011143684387,
      "learning_rate": 8.507512520868115e-05,
      "loss": 0.8965,
      "num_input_tokens_seen": 4741414,
      "step": 1727,
      "train_runtime": 4409.8232,
      "train_tokens_per_second": 1075.194
    },
    {
      "epoch": 0.1388537335020792,
      "grad_norm": 0.37217527627944946,
      "learning_rate": 8.500834724540902e-05,
      "loss": 1.343,
      "num_input_tokens_seen": 4743398,
      "step": 1728,
      "train_runtime": 4411.6143,
      "train_tokens_per_second": 1075.207
    },
    {
      "epoch": 0.13893408867192994,
      "grad_norm": 0.34970638155937195,
      "learning_rate": 8.49415692821369e-05,
      "loss": 0.8375,
      "num_input_tokens_seen": 4744930,
      "step": 1729,
      "train_runtime": 4413.1785,
      "train_tokens_per_second": 1075.173
    },
    {
      "epoch": 0.13901444384178066,
      "grad_norm": 0.3551918566226959,
      "learning_rate": 8.487479131886478e-05,
      "loss": 1.2019,
      "num_input_tokens_seen": 4747054,
      "step": 1730,
      "train_runtime": 4415.0128,
      "train_tokens_per_second": 1075.207
    },
    {
      "epoch": 0.1390947990116314,
      "grad_norm": 0.27151378989219666,
      "learning_rate": 8.480801335559267e-05,
      "loss": 0.6937,
      "num_input_tokens_seen": 4749468,
      "step": 1731,
      "train_runtime": 4416.9781,
      "train_tokens_per_second": 1075.275
    },
    {
      "epoch": 0.13917515418148216,
      "grad_norm": 0.32051900029182434,
      "learning_rate": 8.474123539232054e-05,
      "loss": 0.9629,
      "num_input_tokens_seen": 4751234,
      "step": 1732,
      "train_runtime": 4418.69,
      "train_tokens_per_second": 1075.259
    },
    {
      "epoch": 0.13925550935133288,
      "grad_norm": 0.32169175148010254,
      "learning_rate": 8.467445742904841e-05,
      "loss": 0.8266,
      "num_input_tokens_seen": 4754098,
      "step": 1733,
      "train_runtime": 4421.0239,
      "train_tokens_per_second": 1075.339
    },
    {
      "epoch": 0.13933586452118363,
      "grad_norm": 0.26191744208335876,
      "learning_rate": 8.46076794657763e-05,
      "loss": 0.5521,
      "num_input_tokens_seen": 4757408,
      "step": 1734,
      "train_runtime": 4423.5117,
      "train_tokens_per_second": 1075.482
    },
    {
      "epoch": 0.13941621969103438,
      "grad_norm": 0.3420778512954712,
      "learning_rate": 8.454090150250417e-05,
      "loss": 1.0926,
      "num_input_tokens_seen": 4759564,
      "step": 1735,
      "train_runtime": 4425.3401,
      "train_tokens_per_second": 1075.525
    },
    {
      "epoch": 0.13949657486088513,
      "grad_norm": 0.367861270904541,
      "learning_rate": 8.447412353923206e-05,
      "loss": 0.7912,
      "num_input_tokens_seen": 4762880,
      "step": 1736,
      "train_runtime": 4427.8488,
      "train_tokens_per_second": 1075.665
    },
    {
      "epoch": 0.13957693003073585,
      "grad_norm": 0.2986046075820923,
      "learning_rate": 8.440734557595994e-05,
      "loss": 0.7877,
      "num_input_tokens_seen": 4765186,
      "step": 1737,
      "train_runtime": 4429.9083,
      "train_tokens_per_second": 1075.685
    },
    {
      "epoch": 0.1396572852005866,
      "grad_norm": 0.37660762667655945,
      "learning_rate": 8.434056761268782e-05,
      "loss": 1.1873,
      "num_input_tokens_seen": 4767038,
      "step": 1738,
      "train_runtime": 4431.7161,
      "train_tokens_per_second": 1075.664
    },
    {
      "epoch": 0.13973764037043734,
      "grad_norm": 0.33014607429504395,
      "learning_rate": 8.42737896494157e-05,
      "loss": 1.203,
      "num_input_tokens_seen": 4768766,
      "step": 1739,
      "train_runtime": 4433.3984,
      "train_tokens_per_second": 1075.646
    },
    {
      "epoch": 0.13981799554028806,
      "grad_norm": 0.32945960760116577,
      "learning_rate": 8.420701168614358e-05,
      "loss": 0.875,
      "num_input_tokens_seen": 4771482,
      "step": 1740,
      "train_runtime": 4435.5963,
      "train_tokens_per_second": 1075.725
    },
    {
      "epoch": 0.1398983507101388,
      "grad_norm": 0.25348135828971863,
      "learning_rate": 8.414023372287145e-05,
      "loss": 0.699,
      "num_input_tokens_seen": 4773254,
      "step": 1741,
      "train_runtime": 4439.3401,
      "train_tokens_per_second": 1075.217
    },
    {
      "epoch": 0.13997870587998956,
      "grad_norm": 0.3237021565437317,
      "learning_rate": 8.407345575959934e-05,
      "loss": 0.934,
      "num_input_tokens_seen": 4775682,
      "step": 1742,
      "train_runtime": 4441.4177,
      "train_tokens_per_second": 1075.261
    },
    {
      "epoch": 0.14005906104984028,
      "grad_norm": 0.26454129815101624,
      "learning_rate": 8.400667779632721e-05,
      "loss": 0.931,
      "num_input_tokens_seen": 4778738,
      "step": 1743,
      "train_runtime": 4443.7954,
      "train_tokens_per_second": 1075.373
    },
    {
      "epoch": 0.14013941621969103,
      "grad_norm": 0.4479910731315613,
      "learning_rate": 8.39398998330551e-05,
      "loss": 0.9491,
      "num_input_tokens_seen": 4781942,
      "step": 1744,
      "train_runtime": 4446.353,
      "train_tokens_per_second": 1075.475
    },
    {
      "epoch": 0.14021977138954178,
      "grad_norm": 0.3604801297187805,
      "learning_rate": 8.387312186978298e-05,
      "loss": 1.2846,
      "num_input_tokens_seen": 4784942,
      "step": 1745,
      "train_runtime": 4448.6614,
      "train_tokens_per_second": 1075.591
    },
    {
      "epoch": 0.1403001265593925,
      "grad_norm": 0.28335997462272644,
      "learning_rate": 8.380634390651086e-05,
      "loss": 1.0659,
      "num_input_tokens_seen": 4787498,
      "step": 1746,
      "train_runtime": 4450.7215,
      "train_tokens_per_second": 1075.668
    },
    {
      "epoch": 0.14038048172924325,
      "grad_norm": 0.2960796654224396,
      "learning_rate": 8.373956594323874e-05,
      "loss": 0.8629,
      "num_input_tokens_seen": 4789710,
      "step": 1747,
      "train_runtime": 4452.6173,
      "train_tokens_per_second": 1075.707
    },
    {
      "epoch": 0.140460836899094,
      "grad_norm": 0.2844523787498474,
      "learning_rate": 8.367278797996662e-05,
      "loss": 0.9802,
      "num_input_tokens_seen": 4793382,
      "step": 1748,
      "train_runtime": 4455.2906,
      "train_tokens_per_second": 1075.885
    },
    {
      "epoch": 0.14054119206894475,
      "grad_norm": 0.3542284071445465,
      "learning_rate": 8.360601001669449e-05,
      "loss": 0.9333,
      "num_input_tokens_seen": 4795896,
      "step": 1749,
      "train_runtime": 4457.3781,
      "train_tokens_per_second": 1075.946
    },
    {
      "epoch": 0.14062154723879547,
      "grad_norm": 0.37776631116867065,
      "learning_rate": 8.353923205342238e-05,
      "loss": 0.9041,
      "num_input_tokens_seen": 4798136,
      "step": 1750,
      "train_runtime": 4459.3767,
      "train_tokens_per_second": 1075.966
    },
    {
      "epoch": 0.14070190240864622,
      "grad_norm": 0.31492453813552856,
      "learning_rate": 8.347245409015025e-05,
      "loss": 0.9638,
      "num_input_tokens_seen": 4801146,
      "step": 1751,
      "train_runtime": 4461.7929,
      "train_tokens_per_second": 1076.058
    },
    {
      "epoch": 0.14078225757849697,
      "grad_norm": 0.3911091983318329,
      "learning_rate": 8.340567612687812e-05,
      "loss": 1.0243,
      "num_input_tokens_seen": 4803740,
      "step": 1752,
      "train_runtime": 4463.8089,
      "train_tokens_per_second": 1076.153
    },
    {
      "epoch": 0.1408626127483477,
      "grad_norm": 0.36505600810050964,
      "learning_rate": 8.333889816360601e-05,
      "loss": 1.0655,
      "num_input_tokens_seen": 4805340,
      "step": 1753,
      "train_runtime": 4465.4437,
      "train_tokens_per_second": 1076.117
    },
    {
      "epoch": 0.14094296791819844,
      "grad_norm": 0.3208926320075989,
      "learning_rate": 8.32721202003339e-05,
      "loss": 1.1023,
      "num_input_tokens_seen": 4807752,
      "step": 1754,
      "train_runtime": 4467.47,
      "train_tokens_per_second": 1076.169
    },
    {
      "epoch": 0.14102332308804919,
      "grad_norm": 0.3210667073726654,
      "learning_rate": 8.320534223706178e-05,
      "loss": 1.0684,
      "num_input_tokens_seen": 4810732,
      "step": 1755,
      "train_runtime": 4469.735,
      "train_tokens_per_second": 1076.29
    },
    {
      "epoch": 0.1411036782578999,
      "grad_norm": 0.3312217891216278,
      "learning_rate": 8.313856427378966e-05,
      "loss": 0.969,
      "num_input_tokens_seen": 4813090,
      "step": 1756,
      "train_runtime": 4471.6076,
      "train_tokens_per_second": 1076.367
    },
    {
      "epoch": 0.14118403342775065,
      "grad_norm": 0.35422152280807495,
      "learning_rate": 8.307178631051753e-05,
      "loss": 0.5602,
      "num_input_tokens_seen": 4815772,
      "step": 1757,
      "train_runtime": 4473.749,
      "train_tokens_per_second": 1076.451
    },
    {
      "epoch": 0.1412643885976014,
      "grad_norm": 0.3862323462963104,
      "learning_rate": 8.300500834724542e-05,
      "loss": 1.0834,
      "num_input_tokens_seen": 4818168,
      "step": 1758,
      "train_runtime": 4475.8021,
      "train_tokens_per_second": 1076.493
    },
    {
      "epoch": 0.14134474376745215,
      "grad_norm": 0.28599441051483154,
      "learning_rate": 8.293823038397329e-05,
      "loss": 0.7657,
      "num_input_tokens_seen": 4820946,
      "step": 1759,
      "train_runtime": 4477.9166,
      "train_tokens_per_second": 1076.605
    },
    {
      "epoch": 0.14142509893730287,
      "grad_norm": 0.25712910294532776,
      "learning_rate": 8.287145242070116e-05,
      "loss": 0.9001,
      "num_input_tokens_seen": 4823344,
      "step": 1760,
      "train_runtime": 4479.8492,
      "train_tokens_per_second": 1076.676
    },
    {
      "epoch": 0.14150545410715362,
      "grad_norm": 0.24105006456375122,
      "learning_rate": 8.280467445742905e-05,
      "loss": 1.2903,
      "num_input_tokens_seen": 4828096,
      "step": 1761,
      "train_runtime": 4483.2322,
      "train_tokens_per_second": 1076.923
    },
    {
      "epoch": 0.14158580927700437,
      "grad_norm": 0.42079001665115356,
      "learning_rate": 8.273789649415694e-05,
      "loss": 1.2758,
      "num_input_tokens_seen": 4829848,
      "step": 1762,
      "train_runtime": 4484.9746,
      "train_tokens_per_second": 1076.895
    },
    {
      "epoch": 0.1416661644468551,
      "grad_norm": 0.2710529565811157,
      "learning_rate": 8.267111853088482e-05,
      "loss": 1.0162,
      "num_input_tokens_seen": 4832462,
      "step": 1763,
      "train_runtime": 4487.0989,
      "train_tokens_per_second": 1076.968
    },
    {
      "epoch": 0.14174651961670584,
      "grad_norm": 0.27971163392066956,
      "learning_rate": 8.26043405676127e-05,
      "loss": 1.0044,
      "num_input_tokens_seen": 4834896,
      "step": 1764,
      "train_runtime": 4489.1525,
      "train_tokens_per_second": 1077.018
    },
    {
      "epoch": 0.1418268747865566,
      "grad_norm": 0.3546024560928345,
      "learning_rate": 8.253756260434057e-05,
      "loss": 1.0111,
      "num_input_tokens_seen": 4836672,
      "step": 1765,
      "train_runtime": 4490.9071,
      "train_tokens_per_second": 1076.992
    },
    {
      "epoch": 0.1419072299564073,
      "grad_norm": 0.3018425703048706,
      "learning_rate": 8.247078464106846e-05,
      "loss": 0.9991,
      "num_input_tokens_seen": 4838946,
      "step": 1766,
      "train_runtime": 4492.8314,
      "train_tokens_per_second": 1077.037
    },
    {
      "epoch": 0.14198758512625806,
      "grad_norm": 0.25082850456237793,
      "learning_rate": 8.240400667779633e-05,
      "loss": 0.7473,
      "num_input_tokens_seen": 4841088,
      "step": 1767,
      "train_runtime": 4494.7616,
      "train_tokens_per_second": 1077.051
    },
    {
      "epoch": 0.1420679402961088,
      "grad_norm": 0.29914844036102295,
      "learning_rate": 8.23372287145242e-05,
      "loss": 1.209,
      "num_input_tokens_seen": 4844314,
      "step": 1768,
      "train_runtime": 4497.1685,
      "train_tokens_per_second": 1077.192
    },
    {
      "epoch": 0.14214829546595953,
      "grad_norm": 0.29812729358673096,
      "learning_rate": 8.227045075125209e-05,
      "loss": 1.0831,
      "num_input_tokens_seen": 4845794,
      "step": 1769,
      "train_runtime": 4498.8471,
      "train_tokens_per_second": 1077.119
    },
    {
      "epoch": 0.14222865063581028,
      "grad_norm": 0.30985888838768005,
      "learning_rate": 8.220367278797996e-05,
      "loss": 1.0373,
      "num_input_tokens_seen": 4849466,
      "step": 1770,
      "train_runtime": 4501.6246,
      "train_tokens_per_second": 1077.27
    },
    {
      "epoch": 0.14230900580566103,
      "grad_norm": 0.2536572813987732,
      "learning_rate": 8.213689482470785e-05,
      "loss": 0.9653,
      "num_input_tokens_seen": 4852044,
      "step": 1771,
      "train_runtime": 4506.0357,
      "train_tokens_per_second": 1076.788
    },
    {
      "epoch": 0.14238936097551178,
      "grad_norm": 0.4140840172767639,
      "learning_rate": 8.207011686143574e-05,
      "loss": 0.8622,
      "num_input_tokens_seen": 4853508,
      "step": 1772,
      "train_runtime": 4507.621,
      "train_tokens_per_second": 1076.734
    },
    {
      "epoch": 0.1424697161453625,
      "grad_norm": 0.391434907913208,
      "learning_rate": 8.200333889816361e-05,
      "loss": 1.3303,
      "num_input_tokens_seen": 4854818,
      "step": 1773,
      "train_runtime": 4509.1771,
      "train_tokens_per_second": 1076.653
    },
    {
      "epoch": 0.14255007131521324,
      "grad_norm": 0.29283902049064636,
      "learning_rate": 8.19365609348915e-05,
      "loss": 1.0958,
      "num_input_tokens_seen": 4856784,
      "step": 1774,
      "train_runtime": 4510.9736,
      "train_tokens_per_second": 1076.66
    },
    {
      "epoch": 0.142630426485064,
      "grad_norm": 0.30271488428115845,
      "learning_rate": 8.186978297161937e-05,
      "loss": 0.9557,
      "num_input_tokens_seen": 4859568,
      "step": 1775,
      "train_runtime": 4513.2247,
      "train_tokens_per_second": 1076.74
    },
    {
      "epoch": 0.14271078165491471,
      "grad_norm": 0.31910306215286255,
      "learning_rate": 8.180300500834724e-05,
      "loss": 1.0811,
      "num_input_tokens_seen": 4860940,
      "step": 1776,
      "train_runtime": 4514.7818,
      "train_tokens_per_second": 1076.672
    },
    {
      "epoch": 0.14279113682476546,
      "grad_norm": 0.290634423494339,
      "learning_rate": 8.173622704507513e-05,
      "loss": 1.0577,
      "num_input_tokens_seen": 4863640,
      "step": 1777,
      "train_runtime": 4516.9902,
      "train_tokens_per_second": 1076.744
    },
    {
      "epoch": 0.1428714919946162,
      "grad_norm": 0.3327615559101105,
      "learning_rate": 8.1669449081803e-05,
      "loss": 0.7974,
      "num_input_tokens_seen": 4866062,
      "step": 1778,
      "train_runtime": 4518.9905,
      "train_tokens_per_second": 1076.803
    },
    {
      "epoch": 0.14295184716446693,
      "grad_norm": 0.2692924737930298,
      "learning_rate": 8.160267111853089e-05,
      "loss": 0.8588,
      "num_input_tokens_seen": 4868934,
      "step": 1779,
      "train_runtime": 4521.2411,
      "train_tokens_per_second": 1076.902
    },
    {
      "epoch": 0.14303220233431768,
      "grad_norm": 0.3325542211532593,
      "learning_rate": 8.153589315525877e-05,
      "loss": 1.0313,
      "num_input_tokens_seen": 4871290,
      "step": 1780,
      "train_runtime": 4523.2753,
      "train_tokens_per_second": 1076.939
    },
    {
      "epoch": 0.14311255750416843,
      "grad_norm": 0.32254281640052795,
      "learning_rate": 8.146911519198665e-05,
      "loss": 0.9361,
      "num_input_tokens_seen": 4873974,
      "step": 1781,
      "train_runtime": 4525.3971,
      "train_tokens_per_second": 1077.027
    },
    {
      "epoch": 0.14319291267401915,
      "grad_norm": 0.31965771317481995,
      "learning_rate": 8.140233722871453e-05,
      "loss": 0.5854,
      "num_input_tokens_seen": 4876116,
      "step": 1782,
      "train_runtime": 4527.3632,
      "train_tokens_per_second": 1077.032
    },
    {
      "epoch": 0.1432732678438699,
      "grad_norm": 0.2537826597690582,
      "learning_rate": 8.133555926544241e-05,
      "loss": 0.7571,
      "num_input_tokens_seen": 4878490,
      "step": 1783,
      "train_runtime": 4529.4171,
      "train_tokens_per_second": 1077.068
    },
    {
      "epoch": 0.14335362301372065,
      "grad_norm": 0.3414912521839142,
      "learning_rate": 8.126878130217028e-05,
      "loss": 0.8181,
      "num_input_tokens_seen": 4880758,
      "step": 1784,
      "train_runtime": 4531.3433,
      "train_tokens_per_second": 1077.111
    },
    {
      "epoch": 0.1434339781835714,
      "grad_norm": 0.3281090259552002,
      "learning_rate": 8.120200333889817e-05,
      "loss": 0.8966,
      "num_input_tokens_seen": 4882458,
      "step": 1785,
      "train_runtime": 4533.0136,
      "train_tokens_per_second": 1077.089
    },
    {
      "epoch": 0.14351433335342212,
      "grad_norm": 0.33726608753204346,
      "learning_rate": 8.113522537562604e-05,
      "loss": 0.7388,
      "num_input_tokens_seen": 4884796,
      "step": 1786,
      "train_runtime": 4535.0516,
      "train_tokens_per_second": 1077.12
    },
    {
      "epoch": 0.14359468852327287,
      "grad_norm": 0.3633520007133484,
      "learning_rate": 8.106844741235393e-05,
      "loss": 0.9775,
      "num_input_tokens_seen": 4887166,
      "step": 1787,
      "train_runtime": 4537.061,
      "train_tokens_per_second": 1077.166
    },
    {
      "epoch": 0.14367504369312362,
      "grad_norm": 0.3352885842323303,
      "learning_rate": 8.10016694490818e-05,
      "loss": 0.7796,
      "num_input_tokens_seen": 4888628,
      "step": 1788,
      "train_runtime": 4538.6449,
      "train_tokens_per_second": 1077.112
    },
    {
      "epoch": 0.14375539886297434,
      "grad_norm": 0.39273911714553833,
      "learning_rate": 8.093489148580969e-05,
      "loss": 1.0027,
      "num_input_tokens_seen": 4893330,
      "step": 1789,
      "train_runtime": 4541.9119,
      "train_tokens_per_second": 1077.372
    },
    {
      "epoch": 0.14383575403282509,
      "grad_norm": 0.293765664100647,
      "learning_rate": 8.086811352253757e-05,
      "loss": 1.0126,
      "num_input_tokens_seen": 4895868,
      "step": 1790,
      "train_runtime": 4544.0354,
      "train_tokens_per_second": 1077.427
    },
    {
      "epoch": 0.14391610920267583,
      "grad_norm": 0.4194832444190979,
      "learning_rate": 8.080133555926545e-05,
      "loss": 0.8635,
      "num_input_tokens_seen": 4897834,
      "step": 1791,
      "train_runtime": 4545.8162,
      "train_tokens_per_second": 1077.438
    },
    {
      "epoch": 0.14399646437252656,
      "grad_norm": 0.3623499572277069,
      "learning_rate": 8.073455759599332e-05,
      "loss": 0.8487,
      "num_input_tokens_seen": 4900016,
      "step": 1792,
      "train_runtime": 4547.7089,
      "train_tokens_per_second": 1077.469
    },
    {
      "epoch": 0.1440768195423773,
      "grad_norm": 0.32666507363319397,
      "learning_rate": 8.066777963272121e-05,
      "loss": 0.9051,
      "num_input_tokens_seen": 4903478,
      "step": 1793,
      "train_runtime": 4550.281,
      "train_tokens_per_second": 1077.621
    },
    {
      "epoch": 0.14415717471222805,
      "grad_norm": 0.33514103293418884,
      "learning_rate": 8.060100166944908e-05,
      "loss": 1.1268,
      "num_input_tokens_seen": 4906908,
      "step": 1794,
      "train_runtime": 4552.9085,
      "train_tokens_per_second": 1077.752
    },
    {
      "epoch": 0.1442375298820788,
      "grad_norm": 0.3593701422214508,
      "learning_rate": 8.053422370617697e-05,
      "loss": 0.8101,
      "num_input_tokens_seen": 4909098,
      "step": 1795,
      "train_runtime": 4554.9556,
      "train_tokens_per_second": 1077.749
    },
    {
      "epoch": 0.14431788505192952,
      "grad_norm": 0.687896192073822,
      "learning_rate": 8.046744574290484e-05,
      "loss": 0.723,
      "num_input_tokens_seen": 4910916,
      "step": 1796,
      "train_runtime": 4556.7513,
      "train_tokens_per_second": 1077.723
    },
    {
      "epoch": 0.14439824022178027,
      "grad_norm": 0.4509037733078003,
      "learning_rate": 8.040066777963273e-05,
      "loss": 0.6966,
      "num_input_tokens_seen": 4912680,
      "step": 1797,
      "train_runtime": 4558.4128,
      "train_tokens_per_second": 1077.717
    },
    {
      "epoch": 0.14447859539163102,
      "grad_norm": 0.3475772440433502,
      "learning_rate": 8.033388981636061e-05,
      "loss": 1.0005,
      "num_input_tokens_seen": 4915064,
      "step": 1798,
      "train_runtime": 4560.4283,
      "train_tokens_per_second": 1077.764
    },
    {
      "epoch": 0.14455895056148174,
      "grad_norm": 0.33785468339920044,
      "learning_rate": 8.026711185308849e-05,
      "loss": 1.1092,
      "num_input_tokens_seen": 4916788,
      "step": 1799,
      "train_runtime": 4562.1033,
      "train_tokens_per_second": 1077.746
    },
    {
      "epoch": 0.1446393057313325,
      "grad_norm": 0.3122391998767853,
      "learning_rate": 8.020033388981636e-05,
      "loss": 0.9636,
      "num_input_tokens_seen": 4918674,
      "step": 1800,
      "train_runtime": 4563.8036,
      "train_tokens_per_second": 1077.758
    },
    {
      "epoch": 0.1446393057313325,
      "eval_loss": 0.9616109728813171,
      "eval_runtime": 92.1343,
      "eval_samples_per_second": 10.789,
      "eval_steps_per_second": 5.394,
      "num_input_tokens_seen": 4918674,
      "step": 1800
    },
    {
      "epoch": 0.14471966090118324,
      "grad_norm": 0.41370221972465515,
      "learning_rate": 8.013355592654425e-05,
      "loss": 1.2713,
      "num_input_tokens_seen": 4921512,
      "step": 1801,
      "train_runtime": 4660.2677,
      "train_tokens_per_second": 1056.058
    },
    {
      "epoch": 0.14480001607103396,
      "grad_norm": 0.2980494499206543,
      "learning_rate": 8.006677796327212e-05,
      "loss": 0.6827,
      "num_input_tokens_seen": 4924550,
      "step": 1802,
      "train_runtime": 4662.5942,
      "train_tokens_per_second": 1056.182
    },
    {
      "epoch": 0.1448803712408847,
      "grad_norm": 0.33906808495521545,
      "learning_rate": 8e-05,
      "loss": 1.1137,
      "num_input_tokens_seen": 4928058,
      "step": 1803,
      "train_runtime": 4665.1979,
      "train_tokens_per_second": 1056.345
    },
    {
      "epoch": 0.14496072641073546,
      "grad_norm": 0.31733813881874084,
      "learning_rate": 7.993322203672788e-05,
      "loss": 0.8569,
      "num_input_tokens_seen": 4930166,
      "step": 1804,
      "train_runtime": 4667.0501,
      "train_tokens_per_second": 1056.377
    },
    {
      "epoch": 0.14504108158058618,
      "grad_norm": 0.33354100584983826,
      "learning_rate": 7.986644407345575e-05,
      "loss": 0.8129,
      "num_input_tokens_seen": 4933748,
      "step": 1805,
      "train_runtime": 4669.7004,
      "train_tokens_per_second": 1056.545
    },
    {
      "epoch": 0.14512143675043693,
      "grad_norm": 0.2890450358390808,
      "learning_rate": 7.979966611018364e-05,
      "loss": 1.1081,
      "num_input_tokens_seen": 4936598,
      "step": 1806,
      "train_runtime": 4671.9897,
      "train_tokens_per_second": 1056.637
    },
    {
      "epoch": 0.14520179192028768,
      "grad_norm": 0.3265434205532074,
      "learning_rate": 7.973288814691153e-05,
      "loss": 0.9386,
      "num_input_tokens_seen": 4939290,
      "step": 1807,
      "train_runtime": 4674.1162,
      "train_tokens_per_second": 1056.732
    },
    {
      "epoch": 0.14528214709013842,
      "grad_norm": 0.31261536478996277,
      "learning_rate": 7.96661101836394e-05,
      "loss": 0.8162,
      "num_input_tokens_seen": 4942706,
      "step": 1808,
      "train_runtime": 4676.5737,
      "train_tokens_per_second": 1056.908
    },
    {
      "epoch": 0.14536250225998915,
      "grad_norm": 0.38988253474235535,
      "learning_rate": 7.959933222036729e-05,
      "loss": 0.9918,
      "num_input_tokens_seen": 4944108,
      "step": 1809,
      "train_runtime": 4678.1623,
      "train_tokens_per_second": 1056.848
    },
    {
      "epoch": 0.1454428574298399,
      "grad_norm": 0.30823296308517456,
      "learning_rate": 7.953255425709516e-05,
      "loss": 1.1804,
      "num_input_tokens_seen": 4947164,
      "step": 1810,
      "train_runtime": 4680.5276,
      "train_tokens_per_second": 1056.967
    },
    {
      "epoch": 0.14552321259969064,
      "grad_norm": 0.33548760414123535,
      "learning_rate": 7.946577629382305e-05,
      "loss": 0.9968,
      "num_input_tokens_seen": 4950030,
      "step": 1811,
      "train_runtime": 4682.8149,
      "train_tokens_per_second": 1057.063
    },
    {
      "epoch": 0.14560356776954136,
      "grad_norm": 0.32266977429389954,
      "learning_rate": 7.939899833055092e-05,
      "loss": 0.7696,
      "num_input_tokens_seen": 4951992,
      "step": 1812,
      "train_runtime": 4684.5912,
      "train_tokens_per_second": 1057.081
    },
    {
      "epoch": 0.1456839229393921,
      "grad_norm": 0.33458128571510315,
      "learning_rate": 7.933222036727879e-05,
      "loss": 0.8766,
      "num_input_tokens_seen": 4953388,
      "step": 1813,
      "train_runtime": 4686.1652,
      "train_tokens_per_second": 1057.024
    },
    {
      "epoch": 0.14576427810924286,
      "grad_norm": 0.34393540024757385,
      "learning_rate": 7.926544240400668e-05,
      "loss": 1.0121,
      "num_input_tokens_seen": 4955168,
      "step": 1814,
      "train_runtime": 4687.8651,
      "train_tokens_per_second": 1057.02
    },
    {
      "epoch": 0.14584463327909358,
      "grad_norm": 0.38021495938301086,
      "learning_rate": 7.919866444073457e-05,
      "loss": 0.9994,
      "num_input_tokens_seen": 4957354,
      "step": 1815,
      "train_runtime": 4689.8823,
      "train_tokens_per_second": 1057.032
    },
    {
      "epoch": 0.14592498844894433,
      "grad_norm": 0.27749598026275635,
      "learning_rate": 7.913188647746244e-05,
      "loss": 0.8157,
      "num_input_tokens_seen": 4961082,
      "step": 1816,
      "train_runtime": 4692.5458,
      "train_tokens_per_second": 1057.226
    },
    {
      "epoch": 0.14600534361879508,
      "grad_norm": 0.3151441812515259,
      "learning_rate": 7.906510851419033e-05,
      "loss": 0.793,
      "num_input_tokens_seen": 4962698,
      "step": 1817,
      "train_runtime": 4694.1284,
      "train_tokens_per_second": 1057.214
    },
    {
      "epoch": 0.1460856987886458,
      "grad_norm": 0.27059224247932434,
      "learning_rate": 7.89983305509182e-05,
      "loss": 0.892,
      "num_input_tokens_seen": 4964792,
      "step": 1818,
      "train_runtime": 4695.8982,
      "train_tokens_per_second": 1057.261
    },
    {
      "epoch": 0.14616605395849655,
      "grad_norm": 0.2897622287273407,
      "learning_rate": 7.893155258764609e-05,
      "loss": 0.7819,
      "num_input_tokens_seen": 4966272,
      "step": 1819,
      "train_runtime": 4697.4694,
      "train_tokens_per_second": 1057.223
    },
    {
      "epoch": 0.1462464091283473,
      "grad_norm": 0.27906838059425354,
      "learning_rate": 7.886477462437396e-05,
      "loss": 0.7947,
      "num_input_tokens_seen": 4969124,
      "step": 1820,
      "train_runtime": 4699.7433,
      "train_tokens_per_second": 1057.318
    },
    {
      "epoch": 0.14632676429819805,
      "grad_norm": 0.28144511580467224,
      "learning_rate": 7.879799666110183e-05,
      "loss": 0.5762,
      "num_input_tokens_seen": 4971258,
      "step": 1821,
      "train_runtime": 4701.5203,
      "train_tokens_per_second": 1057.372
    },
    {
      "epoch": 0.14640711946804877,
      "grad_norm": 0.2370944321155548,
      "learning_rate": 7.873121869782972e-05,
      "loss": 0.8168,
      "num_input_tokens_seen": 4973386,
      "step": 1822,
      "train_runtime": 4703.4156,
      "train_tokens_per_second": 1057.399
    },
    {
      "epoch": 0.14648747463789952,
      "grad_norm": 0.3134889304637909,
      "learning_rate": 7.86644407345576e-05,
      "loss": 0.8297,
      "num_input_tokens_seen": 4975456,
      "step": 1823,
      "train_runtime": 4705.2445,
      "train_tokens_per_second": 1057.428
    },
    {
      "epoch": 0.14656782980775027,
      "grad_norm": 0.3657674193382263,
      "learning_rate": 7.859766277128548e-05,
      "loss": 1.0269,
      "num_input_tokens_seen": 4978124,
      "step": 1824,
      "train_runtime": 4707.3684,
      "train_tokens_per_second": 1057.517
    },
    {
      "epoch": 0.146648184977601,
      "grad_norm": 0.3408898115158081,
      "learning_rate": 7.853088480801337e-05,
      "loss": 1.0257,
      "num_input_tokens_seen": 4980896,
      "step": 1825,
      "train_runtime": 4709.5561,
      "train_tokens_per_second": 1057.615
    },
    {
      "epoch": 0.14672854014745174,
      "grad_norm": 0.32894331216812134,
      "learning_rate": 7.846410684474124e-05,
      "loss": 1.1435,
      "num_input_tokens_seen": 4984224,
      "step": 1826,
      "train_runtime": 4712.0255,
      "train_tokens_per_second": 1057.767
    },
    {
      "epoch": 0.14680889531730248,
      "grad_norm": 0.4239885210990906,
      "learning_rate": 7.839732888146912e-05,
      "loss": 0.7581,
      "num_input_tokens_seen": 4986756,
      "step": 1827,
      "train_runtime": 4714.1954,
      "train_tokens_per_second": 1057.817
    },
    {
      "epoch": 0.1468892504871532,
      "grad_norm": 0.2986898720264435,
      "learning_rate": 7.8330550918197e-05,
      "loss": 0.8127,
      "num_input_tokens_seen": 4988648,
      "step": 1828,
      "train_runtime": 4716.011,
      "train_tokens_per_second": 1057.811
    },
    {
      "epoch": 0.14696960565700395,
      "grad_norm": 0.379794180393219,
      "learning_rate": 7.826377295492487e-05,
      "loss": 1.0461,
      "num_input_tokens_seen": 4991768,
      "step": 1829,
      "train_runtime": 4718.4877,
      "train_tokens_per_second": 1057.917
    },
    {
      "epoch": 0.1470499608268547,
      "grad_norm": 0.3045078217983246,
      "learning_rate": 7.819699499165276e-05,
      "loss": 0.9049,
      "num_input_tokens_seen": 4994264,
      "step": 1830,
      "train_runtime": 4720.4511,
      "train_tokens_per_second": 1058.006
    },
    {
      "epoch": 0.14713031599670545,
      "grad_norm": 0.2988998293876648,
      "learning_rate": 7.813021702838063e-05,
      "loss": 1.097,
      "num_input_tokens_seen": 4996932,
      "step": 1831,
      "train_runtime": 4724.7509,
      "train_tokens_per_second": 1057.608
    },
    {
      "epoch": 0.14721067116655617,
      "grad_norm": 0.3512405753135681,
      "learning_rate": 7.806343906510852e-05,
      "loss": 1.1671,
      "num_input_tokens_seen": 4999478,
      "step": 1832,
      "train_runtime": 4726.8514,
      "train_tokens_per_second": 1057.676
    },
    {
      "epoch": 0.14729102633640692,
      "grad_norm": 0.2384834587574005,
      "learning_rate": 7.79966611018364e-05,
      "loss": 0.6767,
      "num_input_tokens_seen": 5002526,
      "step": 1833,
      "train_runtime": 4729.108,
      "train_tokens_per_second": 1057.816
    },
    {
      "epoch": 0.14737138150625767,
      "grad_norm": 0.35167041420936584,
      "learning_rate": 7.792988313856428e-05,
      "loss": 0.8955,
      "num_input_tokens_seen": 5006018,
      "step": 1834,
      "train_runtime": 4731.8312,
      "train_tokens_per_second": 1057.945
    },
    {
      "epoch": 0.1474517366761084,
      "grad_norm": 0.31407415866851807,
      "learning_rate": 7.786310517529216e-05,
      "loss": 1.0368,
      "num_input_tokens_seen": 5008698,
      "step": 1835,
      "train_runtime": 4733.9481,
      "train_tokens_per_second": 1058.038
    },
    {
      "epoch": 0.14753209184595914,
      "grad_norm": 0.31917980313301086,
      "learning_rate": 7.779632721202004e-05,
      "loss": 1.2385,
      "num_input_tokens_seen": 5011016,
      "step": 1836,
      "train_runtime": 4735.9331,
      "train_tokens_per_second": 1058.084
    },
    {
      "epoch": 0.1476124470158099,
      "grad_norm": 0.36754271388053894,
      "learning_rate": 7.772954924874791e-05,
      "loss": 1.2063,
      "num_input_tokens_seen": 5013288,
      "step": 1837,
      "train_runtime": 4737.9122,
      "train_tokens_per_second": 1058.122
    },
    {
      "epoch": 0.1476928021856606,
      "grad_norm": 0.398705393075943,
      "learning_rate": 7.76627712854758e-05,
      "loss": 0.8671,
      "num_input_tokens_seen": 5016040,
      "step": 1838,
      "train_runtime": 4740.1781,
      "train_tokens_per_second": 1058.197
    },
    {
      "epoch": 0.14777315735551136,
      "grad_norm": 0.2788223624229431,
      "learning_rate": 7.759599332220367e-05,
      "loss": 0.8191,
      "num_input_tokens_seen": 5018418,
      "step": 1839,
      "train_runtime": 4742.2362,
      "train_tokens_per_second": 1058.239
    },
    {
      "epoch": 0.1478535125253621,
      "grad_norm": 0.2603667080402374,
      "learning_rate": 7.752921535893156e-05,
      "loss": 0.6136,
      "num_input_tokens_seen": 5021834,
      "step": 1840,
      "train_runtime": 4744.9004,
      "train_tokens_per_second": 1058.364
    },
    {
      "epoch": 0.14793386769521283,
      "grad_norm": 0.4340488016605377,
      "learning_rate": 7.746243739565944e-05,
      "loss": 1.17,
      "num_input_tokens_seen": 5023670,
      "step": 1841,
      "train_runtime": 4746.706,
      "train_tokens_per_second": 1058.349
    },
    {
      "epoch": 0.14801422286506358,
      "grad_norm": 0.3422788679599762,
      "learning_rate": 7.739565943238732e-05,
      "loss": 1.1022,
      "num_input_tokens_seen": 5027296,
      "step": 1842,
      "train_runtime": 4749.4105,
      "train_tokens_per_second": 1058.509
    },
    {
      "epoch": 0.14809457803491433,
      "grad_norm": 0.3044053018093109,
      "learning_rate": 7.73288814691152e-05,
      "loss": 0.7276,
      "num_input_tokens_seen": 5030706,
      "step": 1843,
      "train_runtime": 4752.0853,
      "train_tokens_per_second": 1058.631
    },
    {
      "epoch": 0.14817493320476507,
      "grad_norm": 0.3811112642288208,
      "learning_rate": 7.726210350584308e-05,
      "loss": 1.1837,
      "num_input_tokens_seen": 5032990,
      "step": 1844,
      "train_runtime": 4754.0788,
      "train_tokens_per_second": 1058.668
    },
    {
      "epoch": 0.1482552883746158,
      "grad_norm": 0.3198108375072479,
      "learning_rate": 7.719532554257095e-05,
      "loss": 0.9298,
      "num_input_tokens_seen": 5036188,
      "step": 1845,
      "train_runtime": 4756.5978,
      "train_tokens_per_second": 1058.779
    },
    {
      "epoch": 0.14833564354446654,
      "grad_norm": 0.3088493347167969,
      "learning_rate": 7.712854757929884e-05,
      "loss": 0.9046,
      "num_input_tokens_seen": 5038732,
      "step": 1846,
      "train_runtime": 4758.6419,
      "train_tokens_per_second": 1058.859
    },
    {
      "epoch": 0.1484159987143173,
      "grad_norm": 0.3104758858680725,
      "learning_rate": 7.706176961602671e-05,
      "loss": 0.9043,
      "num_input_tokens_seen": 5042912,
      "step": 1847,
      "train_runtime": 4761.5946,
      "train_tokens_per_second": 1059.08
    },
    {
      "epoch": 0.14849635388416801,
      "grad_norm": 0.39914751052856445,
      "learning_rate": 7.69949916527546e-05,
      "loss": 1.0854,
      "num_input_tokens_seen": 5045060,
      "step": 1848,
      "train_runtime": 4763.5631,
      "train_tokens_per_second": 1059.094
    },
    {
      "epoch": 0.14857670905401876,
      "grad_norm": 0.28599026799201965,
      "learning_rate": 7.692821368948247e-05,
      "loss": 0.6572,
      "num_input_tokens_seen": 5047270,
      "step": 1849,
      "train_runtime": 4765.5135,
      "train_tokens_per_second": 1059.124
    },
    {
      "epoch": 0.1486570642238695,
      "grad_norm": 0.3523062467575073,
      "learning_rate": 7.686143572621036e-05,
      "loss": 1.0561,
      "num_input_tokens_seen": 5049036,
      "step": 1850,
      "train_runtime": 4767.2338,
      "train_tokens_per_second": 1059.112
    },
    {
      "epoch": 0.14873741939372023,
      "grad_norm": 0.34514477849006653,
      "learning_rate": 7.679465776293824e-05,
      "loss": 0.9255,
      "num_input_tokens_seen": 5052864,
      "step": 1851,
      "train_runtime": 4770.0394,
      "train_tokens_per_second": 1059.292
    },
    {
      "epoch": 0.14881777456357098,
      "grad_norm": 0.27023229002952576,
      "learning_rate": 7.672787979966612e-05,
      "loss": 0.7405,
      "num_input_tokens_seen": 5055216,
      "step": 1852,
      "train_runtime": 4771.9484,
      "train_tokens_per_second": 1059.361
    },
    {
      "epoch": 0.14889812973342173,
      "grad_norm": 0.3530409038066864,
      "learning_rate": 7.666110183639399e-05,
      "loss": 0.934,
      "num_input_tokens_seen": 5057806,
      "step": 1853,
      "train_runtime": 4774.1141,
      "train_tokens_per_second": 1059.423
    },
    {
      "epoch": 0.14897848490327245,
      "grad_norm": 0.29033932089805603,
      "learning_rate": 7.659432387312188e-05,
      "loss": 0.6937,
      "num_input_tokens_seen": 5059462,
      "step": 1854,
      "train_runtime": 4775.712,
      "train_tokens_per_second": 1059.415
    },
    {
      "epoch": 0.1490588400731232,
      "grad_norm": 0.3162938058376312,
      "learning_rate": 7.652754590984975e-05,
      "loss": 1.2512,
      "num_input_tokens_seen": 5062448,
      "step": 1855,
      "train_runtime": 4778.0421,
      "train_tokens_per_second": 1059.524
    },
    {
      "epoch": 0.14913919524297395,
      "grad_norm": 0.33358120918273926,
      "learning_rate": 7.646076794657764e-05,
      "loss": 0.9231,
      "num_input_tokens_seen": 5065472,
      "step": 1856,
      "train_runtime": 4780.3905,
      "train_tokens_per_second": 1059.636
    },
    {
      "epoch": 0.1492195504128247,
      "grad_norm": 0.39617830514907837,
      "learning_rate": 7.639398998330551e-05,
      "loss": 1.2502,
      "num_input_tokens_seen": 5067392,
      "step": 1857,
      "train_runtime": 4782.0899,
      "train_tokens_per_second": 1059.661
    },
    {
      "epoch": 0.14929990558267542,
      "grad_norm": 0.3091368079185486,
      "learning_rate": 7.63272120200334e-05,
      "loss": 0.6895,
      "num_input_tokens_seen": 5070066,
      "step": 1858,
      "train_runtime": 4784.1645,
      "train_tokens_per_second": 1059.76
    },
    {
      "epoch": 0.14938026075252617,
      "grad_norm": 0.28823646903038025,
      "learning_rate": 7.626043405676128e-05,
      "loss": 0.914,
      "num_input_tokens_seen": 5072422,
      "step": 1859,
      "train_runtime": 4786.1703,
      "train_tokens_per_second": 1059.808
    },
    {
      "epoch": 0.14946061592237692,
      "grad_norm": 0.2921087145805359,
      "learning_rate": 7.619365609348916e-05,
      "loss": 0.8947,
      "num_input_tokens_seen": 5075142,
      "step": 1860,
      "train_runtime": 4788.3545,
      "train_tokens_per_second": 1059.893
    },
    {
      "epoch": 0.14954097109222764,
      "grad_norm": 0.31922945380210876,
      "learning_rate": 7.612687813021703e-05,
      "loss": 0.8345,
      "num_input_tokens_seen": 5077324,
      "step": 1861,
      "train_runtime": 4792.3665,
      "train_tokens_per_second": 1059.461
    },
    {
      "epoch": 0.14962132626207839,
      "grad_norm": 0.3315534293651581,
      "learning_rate": 7.606010016694492e-05,
      "loss": 0.8381,
      "num_input_tokens_seen": 5082268,
      "step": 1862,
      "train_runtime": 4795.816,
      "train_tokens_per_second": 1059.73
    },
    {
      "epoch": 0.14970168143192913,
      "grad_norm": 0.2970399558544159,
      "learning_rate": 7.599332220367279e-05,
      "loss": 0.8596,
      "num_input_tokens_seen": 5084770,
      "step": 1863,
      "train_runtime": 4797.9035,
      "train_tokens_per_second": 1059.79
    },
    {
      "epoch": 0.14978203660177986,
      "grad_norm": 0.30772262811660767,
      "learning_rate": 7.592654424040068e-05,
      "loss": 0.7521,
      "num_input_tokens_seen": 5088996,
      "step": 1864,
      "train_runtime": 4800.8487,
      "train_tokens_per_second": 1060.02
    },
    {
      "epoch": 0.1498623917716306,
      "grad_norm": 0.39119404554367065,
      "learning_rate": 7.585976627712855e-05,
      "loss": 0.8695,
      "num_input_tokens_seen": 5091114,
      "step": 1865,
      "train_runtime": 4802.7391,
      "train_tokens_per_second": 1060.044
    },
    {
      "epoch": 0.14994274694148135,
      "grad_norm": 0.3998778462409973,
      "learning_rate": 7.579298831385642e-05,
      "loss": 0.9175,
      "num_input_tokens_seen": 5095358,
      "step": 1866,
      "train_runtime": 4805.7097,
      "train_tokens_per_second": 1060.272
    },
    {
      "epoch": 0.15002310211133207,
      "grad_norm": 0.25680747628211975,
      "learning_rate": 7.572621035058431e-05,
      "loss": 0.9966,
      "num_input_tokens_seen": 5099042,
      "step": 1867,
      "train_runtime": 4808.3481,
      "train_tokens_per_second": 1060.456
    },
    {
      "epoch": 0.15010345728118282,
      "grad_norm": 0.37600386142730713,
      "learning_rate": 7.56594323873122e-05,
      "loss": 1.0055,
      "num_input_tokens_seen": 5100804,
      "step": 1868,
      "train_runtime": 4810.0694,
      "train_tokens_per_second": 1060.443
    },
    {
      "epoch": 0.15018381245103357,
      "grad_norm": 0.3055481016635895,
      "learning_rate": 7.559265442404007e-05,
      "loss": 0.8539,
      "num_input_tokens_seen": 5102828,
      "step": 1869,
      "train_runtime": 4811.9025,
      "train_tokens_per_second": 1060.46
    },
    {
      "epoch": 0.15026416762088432,
      "grad_norm": 0.39110270142555237,
      "learning_rate": 7.552587646076796e-05,
      "loss": 1.1043,
      "num_input_tokens_seen": 5104350,
      "step": 1870,
      "train_runtime": 4813.5109,
      "train_tokens_per_second": 1060.421
    },
    {
      "epoch": 0.15034452279073504,
      "grad_norm": 0.5098723769187927,
      "learning_rate": 7.545909849749583e-05,
      "loss": 0.7585,
      "num_input_tokens_seen": 5110098,
      "step": 1871,
      "train_runtime": 4817.4461,
      "train_tokens_per_second": 1060.748
    },
    {
      "epoch": 0.1504248779605858,
      "grad_norm": 0.316049724817276,
      "learning_rate": 7.539232053422371e-05,
      "loss": 1.0061,
      "num_input_tokens_seen": 5112320,
      "step": 1872,
      "train_runtime": 4819.3869,
      "train_tokens_per_second": 1060.782
    },
    {
      "epoch": 0.15050523313043654,
      "grad_norm": 0.33303219079971313,
      "learning_rate": 7.532554257095159e-05,
      "loss": 0.9716,
      "num_input_tokens_seen": 5114994,
      "step": 1873,
      "train_runtime": 4821.6076,
      "train_tokens_per_second": 1060.848
    },
    {
      "epoch": 0.15058558830028726,
      "grad_norm": 0.38001522421836853,
      "learning_rate": 7.525876460767946e-05,
      "loss": 1.1266,
      "num_input_tokens_seen": 5119332,
      "step": 1874,
      "train_runtime": 4824.6852,
      "train_tokens_per_second": 1061.071
    },
    {
      "epoch": 0.150665943470138,
      "grad_norm": 0.3607804477214813,
      "learning_rate": 7.519198664440735e-05,
      "loss": 0.9092,
      "num_input_tokens_seen": 5121648,
      "step": 1875,
      "train_runtime": 4826.657,
      "train_tokens_per_second": 1061.117
    },
    {
      "epoch": 0.15074629863998876,
      "grad_norm": 0.34840768575668335,
      "learning_rate": 7.512520868113523e-05,
      "loss": 1.2315,
      "num_input_tokens_seen": 5125280,
      "step": 1876,
      "train_runtime": 4829.3761,
      "train_tokens_per_second": 1061.272
    },
    {
      "epoch": 0.15082665380983948,
      "grad_norm": 0.31236961483955383,
      "learning_rate": 7.505843071786311e-05,
      "loss": 0.99,
      "num_input_tokens_seen": 5129068,
      "step": 1877,
      "train_runtime": 4832.1284,
      "train_tokens_per_second": 1061.451
    },
    {
      "epoch": 0.15090700897969023,
      "grad_norm": 0.32192865014076233,
      "learning_rate": 7.4991652754591e-05,
      "loss": 0.8144,
      "num_input_tokens_seen": 5130916,
      "step": 1878,
      "train_runtime": 4833.8107,
      "train_tokens_per_second": 1061.464
    },
    {
      "epoch": 0.15098736414954098,
      "grad_norm": 0.2932069003582001,
      "learning_rate": 7.492487479131887e-05,
      "loss": 0.8109,
      "num_input_tokens_seen": 5132514,
      "step": 1879,
      "train_runtime": 4835.5079,
      "train_tokens_per_second": 1061.422
    },
    {
      "epoch": 0.15106771931939172,
      "grad_norm": 0.3649126887321472,
      "learning_rate": 7.485809682804675e-05,
      "loss": 1.0694,
      "num_input_tokens_seen": 5134464,
      "step": 1880,
      "train_runtime": 4837.3481,
      "train_tokens_per_second": 1061.421
    },
    {
      "epoch": 0.15114807448924245,
      "grad_norm": 0.30628979206085205,
      "learning_rate": 7.479131886477463e-05,
      "loss": 0.7232,
      "num_input_tokens_seen": 5137726,
      "step": 1881,
      "train_runtime": 4839.9165,
      "train_tokens_per_second": 1061.532
    },
    {
      "epoch": 0.1512284296590932,
      "grad_norm": 0.33181437849998474,
      "learning_rate": 7.47245409015025e-05,
      "loss": 0.9539,
      "num_input_tokens_seen": 5143082,
      "step": 1882,
      "train_runtime": 4843.794,
      "train_tokens_per_second": 1061.788
    },
    {
      "epoch": 0.15130878482894394,
      "grad_norm": 0.3224644660949707,
      "learning_rate": 7.465776293823039e-05,
      "loss": 0.9683,
      "num_input_tokens_seen": 5145648,
      "step": 1883,
      "train_runtime": 4845.9106,
      "train_tokens_per_second": 1061.854
    },
    {
      "epoch": 0.15138913999879466,
      "grad_norm": 0.3094010651111603,
      "learning_rate": 7.459098497495826e-05,
      "loss": 0.9663,
      "num_input_tokens_seen": 5148430,
      "step": 1884,
      "train_runtime": 4848.1518,
      "train_tokens_per_second": 1061.937
    },
    {
      "epoch": 0.1514694951686454,
      "grad_norm": 0.37314093112945557,
      "learning_rate": 7.452420701168615e-05,
      "loss": 1.1559,
      "num_input_tokens_seen": 5152000,
      "step": 1885,
      "train_runtime": 4850.7413,
      "train_tokens_per_second": 1062.106
    },
    {
      "epoch": 0.15154985033849616,
      "grad_norm": 0.36039674282073975,
      "learning_rate": 7.445742904841403e-05,
      "loss": 1.208,
      "num_input_tokens_seen": 5153956,
      "step": 1886,
      "train_runtime": 4852.6567,
      "train_tokens_per_second": 1062.09
    },
    {
      "epoch": 0.15163020550834688,
      "grad_norm": 0.3279634118080139,
      "learning_rate": 7.439065108514191e-05,
      "loss": 1.1465,
      "num_input_tokens_seen": 5156578,
      "step": 1887,
      "train_runtime": 4854.7492,
      "train_tokens_per_second": 1062.172
    },
    {
      "epoch": 0.15171056067819763,
      "grad_norm": 0.40980371832847595,
      "learning_rate": 7.43238731218698e-05,
      "loss": 1.1245,
      "num_input_tokens_seen": 5158624,
      "step": 1888,
      "train_runtime": 4856.6932,
      "train_tokens_per_second": 1062.168
    },
    {
      "epoch": 0.15179091584804838,
      "grad_norm": 0.4229605495929718,
      "learning_rate": 7.425709515859767e-05,
      "loss": 0.9635,
      "num_input_tokens_seen": 5161528,
      "step": 1889,
      "train_runtime": 4858.962,
      "train_tokens_per_second": 1062.27
    },
    {
      "epoch": 0.1518712710178991,
      "grad_norm": 0.3345964848995209,
      "learning_rate": 7.419031719532554e-05,
      "loss": 0.9265,
      "num_input_tokens_seen": 5163202,
      "step": 1890,
      "train_runtime": 4860.6204,
      "train_tokens_per_second": 1062.252
    },
    {
      "epoch": 0.15195162618774985,
      "grad_norm": 0.3538210391998291,
      "learning_rate": 7.412353923205343e-05,
      "loss": 0.8929,
      "num_input_tokens_seen": 5165900,
      "step": 1891,
      "train_runtime": 4864.9363,
      "train_tokens_per_second": 1061.864
    },
    {
      "epoch": 0.1520319813576006,
      "grad_norm": 0.38495078682899475,
      "learning_rate": 7.40567612687813e-05,
      "loss": 0.8546,
      "num_input_tokens_seen": 5167566,
      "step": 1892,
      "train_runtime": 4866.5927,
      "train_tokens_per_second": 1061.845
    },
    {
      "epoch": 0.15211233652745135,
      "grad_norm": 0.349223256111145,
      "learning_rate": 7.398998330550919e-05,
      "loss": 1.1528,
      "num_input_tokens_seen": 5170206,
      "step": 1893,
      "train_runtime": 4868.7182,
      "train_tokens_per_second": 1061.923
    },
    {
      "epoch": 0.15219269169730207,
      "grad_norm": 0.3576130270957947,
      "learning_rate": 7.392320534223707e-05,
      "loss": 1.0046,
      "num_input_tokens_seen": 5172036,
      "step": 1894,
      "train_runtime": 4870.426,
      "train_tokens_per_second": 1061.927
    },
    {
      "epoch": 0.15227304686715282,
      "grad_norm": 1.4568307399749756,
      "learning_rate": 7.385642737896495e-05,
      "loss": 0.8733,
      "num_input_tokens_seen": 5175464,
      "step": 1895,
      "train_runtime": 4872.9803,
      "train_tokens_per_second": 1062.074
    },
    {
      "epoch": 0.15235340203700357,
      "grad_norm": 0.29496410489082336,
      "learning_rate": 7.378964941569283e-05,
      "loss": 1.0438,
      "num_input_tokens_seen": 5178876,
      "step": 1896,
      "train_runtime": 4875.5744,
      "train_tokens_per_second": 1062.208
    },
    {
      "epoch": 0.1524337572068543,
      "grad_norm": 0.2935889661312103,
      "learning_rate": 7.37228714524207e-05,
      "loss": 1.1221,
      "num_input_tokens_seen": 5181552,
      "step": 1897,
      "train_runtime": 4877.7336,
      "train_tokens_per_second": 1062.287
    },
    {
      "epoch": 0.15251411237670504,
      "grad_norm": 0.3043937087059021,
      "learning_rate": 7.365609348914858e-05,
      "loss": 0.5566,
      "num_input_tokens_seen": 5184480,
      "step": 1898,
      "train_runtime": 4879.9705,
      "train_tokens_per_second": 1062.4
    },
    {
      "epoch": 0.15259446754655578,
      "grad_norm": 0.2773970067501068,
      "learning_rate": 7.358931552587647e-05,
      "loss": 0.8711,
      "num_input_tokens_seen": 5186758,
      "step": 1899,
      "train_runtime": 4881.9367,
      "train_tokens_per_second": 1062.439
    },
    {
      "epoch": 0.1526748227164065,
      "grad_norm": 0.3039838671684265,
      "learning_rate": 7.352253756260434e-05,
      "loss": 0.793,
      "num_input_tokens_seen": 5189176,
      "step": 1900,
      "train_runtime": 4883.9847,
      "train_tokens_per_second": 1062.488
    },
    {
      "epoch": 0.15275517788625725,
      "grad_norm": 0.25697827339172363,
      "learning_rate": 7.345575959933221e-05,
      "loss": 1.0684,
      "num_input_tokens_seen": 5192512,
      "step": 1901,
      "train_runtime": 4886.5191,
      "train_tokens_per_second": 1062.62
    },
    {
      "epoch": 0.152835533056108,
      "grad_norm": 0.35505983233451843,
      "learning_rate": 7.33889816360601e-05,
      "loss": 1.3142,
      "num_input_tokens_seen": 5193996,
      "step": 1902,
      "train_runtime": 4888.0521,
      "train_tokens_per_second": 1062.59
    },
    {
      "epoch": 0.15291588822595872,
      "grad_norm": 0.3592718243598938,
      "learning_rate": 7.332220367278799e-05,
      "loss": 0.7646,
      "num_input_tokens_seen": 5195492,
      "step": 1903,
      "train_runtime": 4889.6754,
      "train_tokens_per_second": 1062.543
    },
    {
      "epoch": 0.15299624339580947,
      "grad_norm": 0.3807624876499176,
      "learning_rate": 7.325542570951587e-05,
      "loss": 0.7792,
      "num_input_tokens_seen": 5197704,
      "step": 1904,
      "train_runtime": 4891.5146,
      "train_tokens_per_second": 1062.596
    },
    {
      "epoch": 0.15307659856566022,
      "grad_norm": 0.2654176950454712,
      "learning_rate": 7.318864774624375e-05,
      "loss": 0.795,
      "num_input_tokens_seen": 5200306,
      "step": 1905,
      "train_runtime": 4893.5224,
      "train_tokens_per_second": 1062.692
    },
    {
      "epoch": 0.15315695373551097,
      "grad_norm": 0.41271671652793884,
      "learning_rate": 7.312186978297162e-05,
      "loss": 1.1301,
      "num_input_tokens_seen": 5202424,
      "step": 1906,
      "train_runtime": 4895.3515,
      "train_tokens_per_second": 1062.727
    },
    {
      "epoch": 0.1532373089053617,
      "grad_norm": 0.3771877586841583,
      "learning_rate": 7.30550918196995e-05,
      "loss": 1.1185,
      "num_input_tokens_seen": 5204678,
      "step": 1907,
      "train_runtime": 4897.2459,
      "train_tokens_per_second": 1062.777
    },
    {
      "epoch": 0.15331766407521244,
      "grad_norm": 0.3171680271625519,
      "learning_rate": 7.298831385642738e-05,
      "loss": 1.1087,
      "num_input_tokens_seen": 5207316,
      "step": 1908,
      "train_runtime": 4899.3389,
      "train_tokens_per_second": 1062.861
    },
    {
      "epoch": 0.1533980192450632,
      "grad_norm": 0.31542688608169556,
      "learning_rate": 7.292153589315525e-05,
      "loss": 0.893,
      "num_input_tokens_seen": 5210150,
      "step": 1909,
      "train_runtime": 4901.5634,
      "train_tokens_per_second": 1062.957
    },
    {
      "epoch": 0.1534783744149139,
      "grad_norm": 0.343566358089447,
      "learning_rate": 7.285475792988314e-05,
      "loss": 0.9828,
      "num_input_tokens_seen": 5213138,
      "step": 1910,
      "train_runtime": 4903.8576,
      "train_tokens_per_second": 1063.069
    },
    {
      "epoch": 0.15355872958476466,
      "grad_norm": 0.3639136552810669,
      "learning_rate": 7.278797996661103e-05,
      "loss": 1.1935,
      "num_input_tokens_seen": 5216152,
      "step": 1911,
      "train_runtime": 4906.1572,
      "train_tokens_per_second": 1063.185
    },
    {
      "epoch": 0.1536390847546154,
      "grad_norm": 0.3396984040737152,
      "learning_rate": 7.272120200333891e-05,
      "loss": 0.9106,
      "num_input_tokens_seen": 5218510,
      "step": 1912,
      "train_runtime": 4908.2123,
      "train_tokens_per_second": 1063.22
    },
    {
      "epoch": 0.15371943992446613,
      "grad_norm": 0.2760705053806305,
      "learning_rate": 7.265442404006679e-05,
      "loss": 1.6664,
      "num_input_tokens_seen": 5223978,
      "step": 1913,
      "train_runtime": 4911.9747,
      "train_tokens_per_second": 1063.519
    },
    {
      "epoch": 0.15379979509431688,
      "grad_norm": 0.39068683981895447,
      "learning_rate": 7.258764607679466e-05,
      "loss": 1.0712,
      "num_input_tokens_seen": 5228494,
      "step": 1914,
      "train_runtime": 4915.1275,
      "train_tokens_per_second": 1063.756
    },
    {
      "epoch": 0.15388015026416763,
      "grad_norm": 0.36092981696128845,
      "learning_rate": 7.252086811352255e-05,
      "loss": 0.9338,
      "num_input_tokens_seen": 5230158,
      "step": 1915,
      "train_runtime": 4916.7702,
      "train_tokens_per_second": 1063.739
    },
    {
      "epoch": 0.15396050543401837,
      "grad_norm": 0.3354611098766327,
      "learning_rate": 7.245409015025042e-05,
      "loss": 0.7635,
      "num_input_tokens_seen": 5233528,
      "step": 1916,
      "train_runtime": 4919.2822,
      "train_tokens_per_second": 1063.88
    },
    {
      "epoch": 0.1540408606038691,
      "grad_norm": 0.37053558230400085,
      "learning_rate": 7.238731218697829e-05,
      "loss": 0.8299,
      "num_input_tokens_seen": 5236698,
      "step": 1917,
      "train_runtime": 4921.8174,
      "train_tokens_per_second": 1063.976
    },
    {
      "epoch": 0.15412121577371984,
      "grad_norm": 0.28253820538520813,
      "learning_rate": 7.232053422370618e-05,
      "loss": 0.8143,
      "num_input_tokens_seen": 5238578,
      "step": 1918,
      "train_runtime": 4923.6118,
      "train_tokens_per_second": 1063.971
    },
    {
      "epoch": 0.1542015709435706,
      "grad_norm": 0.2893552780151367,
      "learning_rate": 7.225375626043405e-05,
      "loss": 0.6788,
      "num_input_tokens_seen": 5241272,
      "step": 1919,
      "train_runtime": 4925.8895,
      "train_tokens_per_second": 1064.025
    },
    {
      "epoch": 0.1542819261134213,
      "grad_norm": 0.36532264947891235,
      "learning_rate": 7.218697829716194e-05,
      "loss": 0.9269,
      "num_input_tokens_seen": 5242960,
      "step": 1920,
      "train_runtime": 4927.6228,
      "train_tokens_per_second": 1063.994
    },
    {
      "epoch": 0.15436228128327206,
      "grad_norm": 0.28941330313682556,
      "learning_rate": 7.212020033388982e-05,
      "loss": 0.8461,
      "num_input_tokens_seen": 5246010,
      "step": 1921,
      "train_runtime": 4931.9946,
      "train_tokens_per_second": 1063.669
    },
    {
      "epoch": 0.1544426364531228,
      "grad_norm": 0.3160305321216583,
      "learning_rate": 7.20534223706177e-05,
      "loss": 0.8282,
      "num_input_tokens_seen": 5249440,
      "step": 1922,
      "train_runtime": 4934.4495,
      "train_tokens_per_second": 1063.835
    },
    {
      "epoch": 0.15452299162297353,
      "grad_norm": 0.30684584379196167,
      "learning_rate": 7.198664440734558e-05,
      "loss": 1.0146,
      "num_input_tokens_seen": 5254612,
      "step": 1923,
      "train_runtime": 4938.1082,
      "train_tokens_per_second": 1064.094
    },
    {
      "epoch": 0.15460334679282428,
      "grad_norm": 0.36664050817489624,
      "learning_rate": 7.191986644407346e-05,
      "loss": 0.8535,
      "num_input_tokens_seen": 5257486,
      "step": 1924,
      "train_runtime": 4940.3908,
      "train_tokens_per_second": 1064.184
    },
    {
      "epoch": 0.15468370196267503,
      "grad_norm": 0.3335873782634735,
      "learning_rate": 7.185308848080133e-05,
      "loss": 1.3193,
      "num_input_tokens_seen": 5260324,
      "step": 1925,
      "train_runtime": 4942.6239,
      "train_tokens_per_second": 1064.278
    },
    {
      "epoch": 0.15476405713252575,
      "grad_norm": 0.35470297932624817,
      "learning_rate": 7.178631051752922e-05,
      "loss": 1.101,
      "num_input_tokens_seen": 5262270,
      "step": 1926,
      "train_runtime": 4944.434,
      "train_tokens_per_second": 1064.282
    },
    {
      "epoch": 0.1548444123023765,
      "grad_norm": 0.3678680658340454,
      "learning_rate": 7.171953255425709e-05,
      "loss": 0.9404,
      "num_input_tokens_seen": 5264188,
      "step": 1927,
      "train_runtime": 4946.2821,
      "train_tokens_per_second": 1064.272
    },
    {
      "epoch": 0.15492476747222725,
      "grad_norm": 0.3112119734287262,
      "learning_rate": 7.165275459098498e-05,
      "loss": 1.0872,
      "num_input_tokens_seen": 5267080,
      "step": 1928,
      "train_runtime": 4948.5653,
      "train_tokens_per_second": 1064.365
    },
    {
      "epoch": 0.155005122642078,
      "grad_norm": 0.3068605959415436,
      "learning_rate": 7.158597662771286e-05,
      "loss": 0.6879,
      "num_input_tokens_seen": 5270234,
      "step": 1929,
      "train_runtime": 4950.9807,
      "train_tokens_per_second": 1064.483
    },
    {
      "epoch": 0.15508547781192872,
      "grad_norm": 0.34481146931648254,
      "learning_rate": 7.151919866444074e-05,
      "loss": 1.0135,
      "num_input_tokens_seen": 5272392,
      "step": 1930,
      "train_runtime": 4952.8855,
      "train_tokens_per_second": 1064.509
    },
    {
      "epoch": 0.15516583298177947,
      "grad_norm": 0.3026423454284668,
      "learning_rate": 7.145242070116862e-05,
      "loss": 0.7957,
      "num_input_tokens_seen": 5274920,
      "step": 1931,
      "train_runtime": 4954.9971,
      "train_tokens_per_second": 1064.566
    },
    {
      "epoch": 0.15524618815163022,
      "grad_norm": 0.2699041962623596,
      "learning_rate": 7.13856427378965e-05,
      "loss": 0.7993,
      "num_input_tokens_seen": 5279748,
      "step": 1932,
      "train_runtime": 4958.319,
      "train_tokens_per_second": 1064.826
    },
    {
      "epoch": 0.15532654332148094,
      "grad_norm": 0.3555600941181183,
      "learning_rate": 7.131886477462437e-05,
      "loss": 0.9359,
      "num_input_tokens_seen": 5282704,
      "step": 1933,
      "train_runtime": 4960.6019,
      "train_tokens_per_second": 1064.932
    },
    {
      "epoch": 0.15540689849133169,
      "grad_norm": 0.25394558906555176,
      "learning_rate": 7.125208681135226e-05,
      "loss": 0.6316,
      "num_input_tokens_seen": 5285336,
      "step": 1934,
      "train_runtime": 4962.7909,
      "train_tokens_per_second": 1064.993
    },
    {
      "epoch": 0.15548725366118243,
      "grad_norm": 0.3657974600791931,
      "learning_rate": 7.118530884808013e-05,
      "loss": 1.1242,
      "num_input_tokens_seen": 5288346,
      "step": 1935,
      "train_runtime": 4964.963,
      "train_tokens_per_second": 1065.133
    },
    {
      "epoch": 0.15556760883103316,
      "grad_norm": 0.32395562529563904,
      "learning_rate": 7.111853088480802e-05,
      "loss": 0.9284,
      "num_input_tokens_seen": 5291148,
      "step": 1936,
      "train_runtime": 4967.2269,
      "train_tokens_per_second": 1065.212
    },
    {
      "epoch": 0.1556479640008839,
      "grad_norm": 0.3251456022262573,
      "learning_rate": 7.105175292153589e-05,
      "loss": 0.8826,
      "num_input_tokens_seen": 5294288,
      "step": 1937,
      "train_runtime": 4969.5641,
      "train_tokens_per_second": 1065.343
    },
    {
      "epoch": 0.15572831917073465,
      "grad_norm": 0.25326696038246155,
      "learning_rate": 7.098497495826378e-05,
      "loss": 0.9146,
      "num_input_tokens_seen": 5299438,
      "step": 1938,
      "train_runtime": 4973.0897,
      "train_tokens_per_second": 1065.623
    },
    {
      "epoch": 0.15580867434058537,
      "grad_norm": 0.27798202633857727,
      "learning_rate": 7.091819699499166e-05,
      "loss": 1.0162,
      "num_input_tokens_seen": 5301834,
      "step": 1939,
      "train_runtime": 4975.0732,
      "train_tokens_per_second": 1065.68
    },
    {
      "epoch": 0.15588902951043612,
      "grad_norm": 0.37339428067207336,
      "learning_rate": 7.085141903171954e-05,
      "loss": 0.9491,
      "num_input_tokens_seen": 5303752,
      "step": 1940,
      "train_runtime": 4976.8114,
      "train_tokens_per_second": 1065.693
    },
    {
      "epoch": 0.15596938468028687,
      "grad_norm": 0.2563164234161377,
      "learning_rate": 7.078464106844741e-05,
      "loss": 0.5693,
      "num_input_tokens_seen": 5306322,
      "step": 1941,
      "train_runtime": 4978.9217,
      "train_tokens_per_second": 1065.757
    },
    {
      "epoch": 0.15604973985013762,
      "grad_norm": 0.28995391726493835,
      "learning_rate": 7.07178631051753e-05,
      "loss": 0.6531,
      "num_input_tokens_seen": 5310170,
      "step": 1942,
      "train_runtime": 4981.758,
      "train_tokens_per_second": 1065.923
    },
    {
      "epoch": 0.15613009501998834,
      "grad_norm": 0.370501309633255,
      "learning_rate": 7.065108514190317e-05,
      "loss": 0.8652,
      "num_input_tokens_seen": 5313770,
      "step": 1943,
      "train_runtime": 4984.492,
      "train_tokens_per_second": 1066.06
    },
    {
      "epoch": 0.1562104501898391,
      "grad_norm": 0.31541895866394043,
      "learning_rate": 7.058430717863106e-05,
      "loss": 0.8789,
      "num_input_tokens_seen": 5315990,
      "step": 1944,
      "train_runtime": 4986.3345,
      "train_tokens_per_second": 1066.112
    },
    {
      "epoch": 0.15629080535968984,
      "grad_norm": 0.2401241511106491,
      "learning_rate": 7.051752921535893e-05,
      "loss": 1.0005,
      "num_input_tokens_seen": 5318204,
      "step": 1945,
      "train_runtime": 4988.2485,
      "train_tokens_per_second": 1066.147
    },
    {
      "epoch": 0.15637116052954056,
      "grad_norm": 0.3356855511665344,
      "learning_rate": 7.045075125208682e-05,
      "loss": 0.7906,
      "num_input_tokens_seen": 5319620,
      "step": 1946,
      "train_runtime": 4989.8284,
      "train_tokens_per_second": 1066.093
    },
    {
      "epoch": 0.1564515156993913,
      "grad_norm": 0.3931116759777069,
      "learning_rate": 7.03839732888147e-05,
      "loss": 1.0376,
      "num_input_tokens_seen": 5323762,
      "step": 1947,
      "train_runtime": 4992.6268,
      "train_tokens_per_second": 1066.325
    },
    {
      "epoch": 0.15653187086924206,
      "grad_norm": 0.3554634749889374,
      "learning_rate": 7.031719532554258e-05,
      "loss": 0.7906,
      "num_input_tokens_seen": 5327932,
      "step": 1948,
      "train_runtime": 4995.5836,
      "train_tokens_per_second": 1066.528
    },
    {
      "epoch": 0.15661222603909278,
      "grad_norm": 0.3552740514278412,
      "learning_rate": 7.025041736227045e-05,
      "loss": 1.1707,
      "num_input_tokens_seen": 5329492,
      "step": 1949,
      "train_runtime": 4997.1604,
      "train_tokens_per_second": 1066.504
    },
    {
      "epoch": 0.15669258120894353,
      "grad_norm": 0.28638574481010437,
      "learning_rate": 7.018363939899834e-05,
      "loss": 1.0422,
      "num_input_tokens_seen": 5332904,
      "step": 1950,
      "train_runtime": 4999.7415,
      "train_tokens_per_second": 1066.636
    },
    {
      "epoch": 0.15677293637879428,
      "grad_norm": 0.2969324290752411,
      "learning_rate": 7.011686143572621e-05,
      "loss": 0.9198,
      "num_input_tokens_seen": 5335286,
      "step": 1951,
      "train_runtime": 5003.4018,
      "train_tokens_per_second": 1066.332
    },
    {
      "epoch": 0.15685329154864502,
      "grad_norm": 0.29404449462890625,
      "learning_rate": 7.00500834724541e-05,
      "loss": 0.9238,
      "num_input_tokens_seen": 5337290,
      "step": 1952,
      "train_runtime": 5005.2396,
      "train_tokens_per_second": 1066.341
    },
    {
      "epoch": 0.15693364671849575,
      "grad_norm": 0.35833194851875305,
      "learning_rate": 6.998330550918197e-05,
      "loss": 0.8637,
      "num_input_tokens_seen": 5340164,
      "step": 1953,
      "train_runtime": 5007.5406,
      "train_tokens_per_second": 1066.424
    },
    {
      "epoch": 0.1570140018883465,
      "grad_norm": 0.3525274693965912,
      "learning_rate": 6.991652754590986e-05,
      "loss": 0.9701,
      "num_input_tokens_seen": 5342766,
      "step": 1954,
      "train_runtime": 5009.6672,
      "train_tokens_per_second": 1066.491
    },
    {
      "epoch": 0.15709435705819724,
      "grad_norm": 0.262835830450058,
      "learning_rate": 6.984974958263774e-05,
      "loss": 0.8071,
      "num_input_tokens_seen": 5345382,
      "step": 1955,
      "train_runtime": 5011.8178,
      "train_tokens_per_second": 1066.556
    },
    {
      "epoch": 0.15717471222804796,
      "grad_norm": 0.29403117299079895,
      "learning_rate": 6.978297161936562e-05,
      "loss": 0.8564,
      "num_input_tokens_seen": 5347578,
      "step": 1956,
      "train_runtime": 5013.6731,
      "train_tokens_per_second": 1066.599
    },
    {
      "epoch": 0.1572550673978987,
      "grad_norm": 0.29831740260124207,
      "learning_rate": 6.971619365609349e-05,
      "loss": 0.91,
      "num_input_tokens_seen": 5350144,
      "step": 1957,
      "train_runtime": 5015.725,
      "train_tokens_per_second": 1066.674
    },
    {
      "epoch": 0.15733542256774946,
      "grad_norm": 0.3791574239730835,
      "learning_rate": 6.964941569282138e-05,
      "loss": 0.7161,
      "num_input_tokens_seen": 5352736,
      "step": 1958,
      "train_runtime": 5017.9306,
      "train_tokens_per_second": 1066.722
    },
    {
      "epoch": 0.15741577773760018,
      "grad_norm": 0.2974484860897064,
      "learning_rate": 6.958263772954925e-05,
      "loss": 0.7892,
      "num_input_tokens_seen": 5355908,
      "step": 1959,
      "train_runtime": 5020.3733,
      "train_tokens_per_second": 1066.835
    },
    {
      "epoch": 0.15749613290745093,
      "grad_norm": 0.37496134638786316,
      "learning_rate": 6.951585976627714e-05,
      "loss": 0.9776,
      "num_input_tokens_seen": 5358350,
      "step": 1960,
      "train_runtime": 5022.4329,
      "train_tokens_per_second": 1066.883
    },
    {
      "epoch": 0.15757648807730168,
      "grad_norm": 0.32669517397880554,
      "learning_rate": 6.944908180300501e-05,
      "loss": 1.1729,
      "num_input_tokens_seen": 5360426,
      "step": 1961,
      "train_runtime": 5024.251,
      "train_tokens_per_second": 1066.91
    },
    {
      "epoch": 0.1576568432471524,
      "grad_norm": 0.33246564865112305,
      "learning_rate": 6.938230383973288e-05,
      "loss": 1.1655,
      "num_input_tokens_seen": 5362906,
      "step": 1962,
      "train_runtime": 5026.2137,
      "train_tokens_per_second": 1066.987
    },
    {
      "epoch": 0.15773719841700315,
      "grad_norm": 0.29764726758003235,
      "learning_rate": 6.931552587646077e-05,
      "loss": 0.8907,
      "num_input_tokens_seen": 5366248,
      "step": 1963,
      "train_runtime": 5028.7282,
      "train_tokens_per_second": 1067.118
    },
    {
      "epoch": 0.1578175535868539,
      "grad_norm": 0.29955893754959106,
      "learning_rate": 6.924874791318865e-05,
      "loss": 1.1362,
      "num_input_tokens_seen": 5368114,
      "step": 1964,
      "train_runtime": 5030.4758,
      "train_tokens_per_second": 1067.119
    },
    {
      "epoch": 0.15789790875670465,
      "grad_norm": 0.41301074624061584,
      "learning_rate": 6.918196994991654e-05,
      "loss": 0.8787,
      "num_input_tokens_seen": 5370326,
      "step": 1965,
      "train_runtime": 5032.501,
      "train_tokens_per_second": 1067.129
    },
    {
      "epoch": 0.15797826392655537,
      "grad_norm": 0.3436276614665985,
      "learning_rate": 6.911519198664441e-05,
      "loss": 0.9528,
      "num_input_tokens_seen": 5373510,
      "step": 1966,
      "train_runtime": 5034.8354,
      "train_tokens_per_second": 1067.266
    },
    {
      "epoch": 0.15805861909640612,
      "grad_norm": 0.4939381778240204,
      "learning_rate": 6.904841402337229e-05,
      "loss": 0.9795,
      "num_input_tokens_seen": 5377876,
      "step": 1967,
      "train_runtime": 5038.0103,
      "train_tokens_per_second": 1067.46
    },
    {
      "epoch": 0.15813897426625687,
      "grad_norm": 0.41390731930732727,
      "learning_rate": 6.898163606010017e-05,
      "loss": 0.8077,
      "num_input_tokens_seen": 5379832,
      "step": 1968,
      "train_runtime": 5039.8319,
      "train_tokens_per_second": 1067.463
    },
    {
      "epoch": 0.1582193294361076,
      "grad_norm": 0.2849280536174774,
      "learning_rate": 6.891485809682805e-05,
      "loss": 1.0216,
      "num_input_tokens_seen": 5385832,
      "step": 1969,
      "train_runtime": 5043.919,
      "train_tokens_per_second": 1067.787
    },
    {
      "epoch": 0.15829968460595834,
      "grad_norm": 0.29219865798950195,
      "learning_rate": 6.884808013355592e-05,
      "loss": 0.8279,
      "num_input_tokens_seen": 5388922,
      "step": 1970,
      "train_runtime": 5046.3134,
      "train_tokens_per_second": 1067.893
    },
    {
      "epoch": 0.15838003977580908,
      "grad_norm": 0.3463384211063385,
      "learning_rate": 6.878130217028381e-05,
      "loss": 1.1366,
      "num_input_tokens_seen": 5391822,
      "step": 1971,
      "train_runtime": 5048.778,
      "train_tokens_per_second": 1067.946
    },
    {
      "epoch": 0.1584603949456598,
      "grad_norm": 0.248994380235672,
      "learning_rate": 6.87145242070117e-05,
      "loss": 0.8292,
      "num_input_tokens_seen": 5398074,
      "step": 1972,
      "train_runtime": 5053.109,
      "train_tokens_per_second": 1068.268
    },
    {
      "epoch": 0.15854075011551055,
      "grad_norm": 0.3142145574092865,
      "learning_rate": 6.864774624373958e-05,
      "loss": 0.829,
      "num_input_tokens_seen": 5399646,
      "step": 1973,
      "train_runtime": 5054.7784,
      "train_tokens_per_second": 1068.226
    },
    {
      "epoch": 0.1586211052853613,
      "grad_norm": 0.35669538378715515,
      "learning_rate": 6.858096828046745e-05,
      "loss": 1.1861,
      "num_input_tokens_seen": 5402010,
      "step": 1974,
      "train_runtime": 5056.8084,
      "train_tokens_per_second": 1068.265
    },
    {
      "epoch": 0.15870146045521202,
      "grad_norm": 0.3485780358314514,
      "learning_rate": 6.851419031719533e-05,
      "loss": 0.6812,
      "num_input_tokens_seen": 5404206,
      "step": 1975,
      "train_runtime": 5058.6761,
      "train_tokens_per_second": 1068.304
    },
    {
      "epoch": 0.15878181562506277,
      "grad_norm": 0.3071657121181488,
      "learning_rate": 6.844741235392321e-05,
      "loss": 0.7835,
      "num_input_tokens_seen": 5405990,
      "step": 1976,
      "train_runtime": 5060.4081,
      "train_tokens_per_second": 1068.291
    },
    {
      "epoch": 0.15886217079491352,
      "grad_norm": 0.2993569076061249,
      "learning_rate": 6.838063439065109e-05,
      "loss": 1.1551,
      "num_input_tokens_seen": 5408148,
      "step": 1977,
      "train_runtime": 5062.2414,
      "train_tokens_per_second": 1068.331
    },
    {
      "epoch": 0.15894252596476427,
      "grad_norm": 0.34315863251686096,
      "learning_rate": 6.831385642737896e-05,
      "loss": 1.2995,
      "num_input_tokens_seen": 5409808,
      "step": 1978,
      "train_runtime": 5063.944,
      "train_tokens_per_second": 1068.299
    },
    {
      "epoch": 0.159022881134615,
      "grad_norm": 0.395594984292984,
      "learning_rate": 6.824707846410685e-05,
      "loss": 1.3062,
      "num_input_tokens_seen": 5411408,
      "step": 1979,
      "train_runtime": 5065.5757,
      "train_tokens_per_second": 1068.271
    },
    {
      "epoch": 0.15910323630446574,
      "grad_norm": 0.3101098835468292,
      "learning_rate": 6.818030050083472e-05,
      "loss": 1.0669,
      "num_input_tokens_seen": 5414612,
      "step": 1980,
      "train_runtime": 5067.9197,
      "train_tokens_per_second": 1068.409
    },
    {
      "epoch": 0.1591835914743165,
      "grad_norm": 0.28457948565483093,
      "learning_rate": 6.811352253756261e-05,
      "loss": 1.1173,
      "num_input_tokens_seen": 5418072,
      "step": 1981,
      "train_runtime": 5072.0984,
      "train_tokens_per_second": 1068.211
    },
    {
      "epoch": 0.1592639466441672,
      "grad_norm": 0.2898434102535248,
      "learning_rate": 6.80467445742905e-05,
      "loss": 0.8002,
      "num_input_tokens_seen": 5419758,
      "step": 1982,
      "train_runtime": 5073.886,
      "train_tokens_per_second": 1068.167
    },
    {
      "epoch": 0.15934430181401796,
      "grad_norm": 0.27834630012512207,
      "learning_rate": 6.797996661101837e-05,
      "loss": 0.763,
      "num_input_tokens_seen": 5423198,
      "step": 1983,
      "train_runtime": 5076.4303,
      "train_tokens_per_second": 1068.309
    },
    {
      "epoch": 0.1594246569838687,
      "grad_norm": 0.3703613579273224,
      "learning_rate": 6.791318864774625e-05,
      "loss": 0.9988,
      "num_input_tokens_seen": 5425062,
      "step": 1984,
      "train_runtime": 5078.181,
      "train_tokens_per_second": 1068.308
    },
    {
      "epoch": 0.15950501215371943,
      "grad_norm": 0.2970855236053467,
      "learning_rate": 6.784641068447413e-05,
      "loss": 1.0776,
      "num_input_tokens_seen": 5427042,
      "step": 1985,
      "train_runtime": 5080.045,
      "train_tokens_per_second": 1068.306
    },
    {
      "epoch": 0.15958536732357018,
      "grad_norm": 0.34390586614608765,
      "learning_rate": 6.7779632721202e-05,
      "loss": 0.978,
      "num_input_tokens_seen": 5429028,
      "step": 1986,
      "train_runtime": 5081.9533,
      "train_tokens_per_second": 1068.296
    },
    {
      "epoch": 0.15966572249342093,
      "grad_norm": 0.26503974199295044,
      "learning_rate": 6.771285475792989e-05,
      "loss": 0.8988,
      "num_input_tokens_seen": 5432860,
      "step": 1987,
      "train_runtime": 5084.6806,
      "train_tokens_per_second": 1068.476
    },
    {
      "epoch": 0.15974607766327167,
      "grad_norm": 0.27086111903190613,
      "learning_rate": 6.764607679465776e-05,
      "loss": 0.9022,
      "num_input_tokens_seen": 5435918,
      "step": 1988,
      "train_runtime": 5087.1247,
      "train_tokens_per_second": 1068.564
    },
    {
      "epoch": 0.1598264328331224,
      "grad_norm": 0.4578646719455719,
      "learning_rate": 6.757929883138565e-05,
      "loss": 0.893,
      "num_input_tokens_seen": 5437828,
      "step": 1989,
      "train_runtime": 5088.8072,
      "train_tokens_per_second": 1068.586
    },
    {
      "epoch": 0.15990678800297314,
      "grad_norm": 0.3436759412288666,
      "learning_rate": 6.751252086811353e-05,
      "loss": 0.9503,
      "num_input_tokens_seen": 5440438,
      "step": 1990,
      "train_runtime": 5090.8326,
      "train_tokens_per_second": 1068.674
    },
    {
      "epoch": 0.1599871431728239,
      "grad_norm": 0.3114043176174164,
      "learning_rate": 6.74457429048414e-05,
      "loss": 0.9616,
      "num_input_tokens_seen": 5446342,
      "step": 1991,
      "train_runtime": 5094.9021,
      "train_tokens_per_second": 1068.979
    },
    {
      "epoch": 0.1600674983426746,
      "grad_norm": 0.3067304491996765,
      "learning_rate": 6.737896494156929e-05,
      "loss": 0.7178,
      "num_input_tokens_seen": 5448296,
      "step": 1992,
      "train_runtime": 5096.7909,
      "train_tokens_per_second": 1068.966
    },
    {
      "epoch": 0.16014785351252536,
      "grad_norm": 0.2988907992839813,
      "learning_rate": 6.731218697829717e-05,
      "loss": 1.0715,
      "num_input_tokens_seen": 5454086,
      "step": 1993,
      "train_runtime": 5100.8145,
      "train_tokens_per_second": 1069.258
    },
    {
      "epoch": 0.1602282086823761,
      "grad_norm": 0.2712036371231079,
      "learning_rate": 6.724540901502504e-05,
      "loss": 1.1776,
      "num_input_tokens_seen": 5459316,
      "step": 1994,
      "train_runtime": 5104.4322,
      "train_tokens_per_second": 1069.525
    },
    {
      "epoch": 0.16030856385222683,
      "grad_norm": 0.5102654695510864,
      "learning_rate": 6.717863105175293e-05,
      "loss": 0.7674,
      "num_input_tokens_seen": 5463292,
      "step": 1995,
      "train_runtime": 5107.4182,
      "train_tokens_per_second": 1069.678
    },
    {
      "epoch": 0.16038891902207758,
      "grad_norm": 0.37160399556159973,
      "learning_rate": 6.71118530884808e-05,
      "loss": 1.031,
      "num_input_tokens_seen": 5466880,
      "step": 1996,
      "train_runtime": 5110.0017,
      "train_tokens_per_second": 1069.839
    },
    {
      "epoch": 0.16046927419192833,
      "grad_norm": 0.2616192400455475,
      "learning_rate": 6.704507512520869e-05,
      "loss": 1.0917,
      "num_input_tokens_seen": 5469918,
      "step": 1997,
      "train_runtime": 5112.3784,
      "train_tokens_per_second": 1069.936
    },
    {
      "epoch": 0.16054962936177905,
      "grad_norm": 0.3706570863723755,
      "learning_rate": 6.697829716193656e-05,
      "loss": 0.7708,
      "num_input_tokens_seen": 5471192,
      "step": 1998,
      "train_runtime": 5113.8874,
      "train_tokens_per_second": 1069.869
    },
    {
      "epoch": 0.1606299845316298,
      "grad_norm": 0.412650465965271,
      "learning_rate": 6.691151919866445e-05,
      "loss": 0.4588,
      "num_input_tokens_seen": 5475472,
      "step": 1999,
      "train_runtime": 5117.0738,
      "train_tokens_per_second": 1070.04
    },
    {
      "epoch": 0.16071033970148055,
      "grad_norm": 0.37960755825042725,
      "learning_rate": 6.684474123539233e-05,
      "loss": 1.0389,
      "num_input_tokens_seen": 5478570,
      "step": 2000,
      "train_runtime": 5119.3993,
      "train_tokens_per_second": 1070.159
    },
    {
      "epoch": 0.1607906948713313,
      "grad_norm": 0.29763177037239075,
      "learning_rate": 6.67779632721202e-05,
      "loss": 0.8047,
      "num_input_tokens_seen": 5480588,
      "step": 2001,
      "train_runtime": 5121.2035,
      "train_tokens_per_second": 1070.176
    },
    {
      "epoch": 0.16087105004118202,
      "grad_norm": 0.2461473047733307,
      "learning_rate": 6.671118530884808e-05,
      "loss": 0.878,
      "num_input_tokens_seen": 5484472,
      "step": 2002,
      "train_runtime": 5124.0742,
      "train_tokens_per_second": 1070.334
    },
    {
      "epoch": 0.16095140521103277,
      "grad_norm": 0.38381364941596985,
      "learning_rate": 6.664440734557597e-05,
      "loss": 1.2021,
      "num_input_tokens_seen": 5487396,
      "step": 2003,
      "train_runtime": 5126.3374,
      "train_tokens_per_second": 1070.432
    },
    {
      "epoch": 0.16103176038088352,
      "grad_norm": 0.36581820249557495,
      "learning_rate": 6.657762938230384e-05,
      "loss": 1.1594,
      "num_input_tokens_seen": 5488868,
      "step": 2004,
      "train_runtime": 5127.9142,
      "train_tokens_per_second": 1070.39
    },
    {
      "epoch": 0.16111211555073424,
      "grad_norm": 0.3146212100982666,
      "learning_rate": 6.651085141903173e-05,
      "loss": 1.0656,
      "num_input_tokens_seen": 5491252,
      "step": 2005,
      "train_runtime": 5129.9666,
      "train_tokens_per_second": 1070.426
    },
    {
      "epoch": 0.16119247072058498,
      "grad_norm": 0.3928852379322052,
      "learning_rate": 6.64440734557596e-05,
      "loss": 1.1531,
      "num_input_tokens_seen": 5493438,
      "step": 2006,
      "train_runtime": 5131.9773,
      "train_tokens_per_second": 1070.433
    },
    {
      "epoch": 0.16127282589043573,
      "grad_norm": 0.29733264446258545,
      "learning_rate": 6.637729549248749e-05,
      "loss": 1.0532,
      "num_input_tokens_seen": 5495702,
      "step": 2007,
      "train_runtime": 5133.846,
      "train_tokens_per_second": 1070.484
    },
    {
      "epoch": 0.16135318106028645,
      "grad_norm": 0.28483888506889343,
      "learning_rate": 6.631051752921537e-05,
      "loss": 1.1196,
      "num_input_tokens_seen": 5498840,
      "step": 2008,
      "train_runtime": 5136.282,
      "train_tokens_per_second": 1070.588
    },
    {
      "epoch": 0.1614335362301372,
      "grad_norm": 0.31708404421806335,
      "learning_rate": 6.624373956594325e-05,
      "loss": 0.9999,
      "num_input_tokens_seen": 5501512,
      "step": 2009,
      "train_runtime": 5138.4136,
      "train_tokens_per_second": 1070.664
    },
    {
      "epoch": 0.16151389139998795,
      "grad_norm": 0.30248215794563293,
      "learning_rate": 6.617696160267112e-05,
      "loss": 1.0217,
      "num_input_tokens_seen": 5504706,
      "step": 2010,
      "train_runtime": 5140.9301,
      "train_tokens_per_second": 1070.761
    },
    {
      "epoch": 0.16159424656983867,
      "grad_norm": 0.3167038857936859,
      "learning_rate": 6.6110183639399e-05,
      "loss": 0.6999,
      "num_input_tokens_seen": 5507808,
      "step": 2011,
      "train_runtime": 5145.091,
      "train_tokens_per_second": 1070.498
    },
    {
      "epoch": 0.16167460173968942,
      "grad_norm": 0.26810508966445923,
      "learning_rate": 6.604340567612688e-05,
      "loss": 0.7864,
      "num_input_tokens_seen": 5511200,
      "step": 2012,
      "train_runtime": 5147.6391,
      "train_tokens_per_second": 1070.627
    },
    {
      "epoch": 0.16175495690954017,
      "grad_norm": 0.3418722152709961,
      "learning_rate": 6.597662771285476e-05,
      "loss": 1.0623,
      "num_input_tokens_seen": 5514692,
      "step": 2013,
      "train_runtime": 5150.319,
      "train_tokens_per_second": 1070.748
    },
    {
      "epoch": 0.16183531207939092,
      "grad_norm": 0.3534741699695587,
      "learning_rate": 6.590984974958264e-05,
      "loss": 1.0377,
      "num_input_tokens_seen": 5517002,
      "step": 2014,
      "train_runtime": 5152.3375,
      "train_tokens_per_second": 1070.777
    },
    {
      "epoch": 0.16191566724924164,
      "grad_norm": 0.2800027132034302,
      "learning_rate": 6.584307178631051e-05,
      "loss": 0.8989,
      "num_input_tokens_seen": 5520926,
      "step": 2015,
      "train_runtime": 5155.1256,
      "train_tokens_per_second": 1070.959
    },
    {
      "epoch": 0.1619960224190924,
      "grad_norm": 0.31767845153808594,
      "learning_rate": 6.57762938230384e-05,
      "loss": 1.0594,
      "num_input_tokens_seen": 5522890,
      "step": 2016,
      "train_runtime": 5156.8531,
      "train_tokens_per_second": 1070.981
    },
    {
      "epoch": 0.16207637758894314,
      "grad_norm": 0.23786990344524384,
      "learning_rate": 6.570951585976628e-05,
      "loss": 0.6891,
      "num_input_tokens_seen": 5526156,
      "step": 2017,
      "train_runtime": 5159.2947,
      "train_tokens_per_second": 1071.107
    },
    {
      "epoch": 0.16215673275879386,
      "grad_norm": 0.3688257336616516,
      "learning_rate": 6.564273789649416e-05,
      "loss": 1.0869,
      "num_input_tokens_seen": 5530634,
      "step": 2018,
      "train_runtime": 5162.6587,
      "train_tokens_per_second": 1071.276
    },
    {
      "epoch": 0.1622370879286446,
      "grad_norm": 0.3014098107814789,
      "learning_rate": 6.557595993322204e-05,
      "loss": 1.0303,
      "num_input_tokens_seen": 5533082,
      "step": 2019,
      "train_runtime": 5164.7146,
      "train_tokens_per_second": 1071.324
    },
    {
      "epoch": 0.16231744309849536,
      "grad_norm": 0.34500762820243835,
      "learning_rate": 6.550918196994992e-05,
      "loss": 0.9926,
      "num_input_tokens_seen": 5535222,
      "step": 2020,
      "train_runtime": 5166.621,
      "train_tokens_per_second": 1071.343
    },
    {
      "epoch": 0.16239779826834608,
      "grad_norm": 0.24485503137111664,
      "learning_rate": 6.54424040066778e-05,
      "loss": 0.5423,
      "num_input_tokens_seen": 5537210,
      "step": 2021,
      "train_runtime": 5168.4414,
      "train_tokens_per_second": 1071.35
    },
    {
      "epoch": 0.16247815343819683,
      "grad_norm": 0.2860889136791229,
      "learning_rate": 6.537562604340568e-05,
      "loss": 0.9626,
      "num_input_tokens_seen": 5540068,
      "step": 2022,
      "train_runtime": 5170.7148,
      "train_tokens_per_second": 1071.432
    },
    {
      "epoch": 0.16255850860804757,
      "grad_norm": 0.3342204988002777,
      "learning_rate": 6.530884808013355e-05,
      "loss": 0.8224,
      "num_input_tokens_seen": 5542534,
      "step": 2023,
      "train_runtime": 5172.7647,
      "train_tokens_per_second": 1071.484
    },
    {
      "epoch": 0.16263886377789832,
      "grad_norm": 0.4009633958339691,
      "learning_rate": 6.524207011686144e-05,
      "loss": 0.6614,
      "num_input_tokens_seen": 5544744,
      "step": 2024,
      "train_runtime": 5174.5959,
      "train_tokens_per_second": 1071.532
    },
    {
      "epoch": 0.16271921894774904,
      "grad_norm": 0.42543843388557434,
      "learning_rate": 6.517529215358932e-05,
      "loss": 0.9491,
      "num_input_tokens_seen": 5546498,
      "step": 2025,
      "train_runtime": 5176.2583,
      "train_tokens_per_second": 1071.527
    },
    {
      "epoch": 0.1627995741175998,
      "grad_norm": 0.2977692186832428,
      "learning_rate": 6.51085141903172e-05,
      "loss": 0.8044,
      "num_input_tokens_seen": 5548960,
      "step": 2026,
      "train_runtime": 5178.2467,
      "train_tokens_per_second": 1071.591
    },
    {
      "epoch": 0.16287992928745054,
      "grad_norm": 0.37556591629981995,
      "learning_rate": 6.504173622704508e-05,
      "loss": 1.2746,
      "num_input_tokens_seen": 5551680,
      "step": 2027,
      "train_runtime": 5180.4911,
      "train_tokens_per_second": 1071.651
    },
    {
      "epoch": 0.16296028445730126,
      "grad_norm": 0.30354321002960205,
      "learning_rate": 6.497495826377296e-05,
      "loss": 2.0511,
      "num_input_tokens_seen": 5557666,
      "step": 2028,
      "train_runtime": 5184.4913,
      "train_tokens_per_second": 1071.979
    },
    {
      "epoch": 0.163040639627152,
      "grad_norm": 0.288390576839447,
      "learning_rate": 6.490818030050084e-05,
      "loss": 1.2055,
      "num_input_tokens_seen": 5561890,
      "step": 2029,
      "train_runtime": 5187.4586,
      "train_tokens_per_second": 1072.18
    },
    {
      "epoch": 0.16312099479700276,
      "grad_norm": 0.31022995710372925,
      "learning_rate": 6.484140233722872e-05,
      "loss": 1.4456,
      "num_input_tokens_seen": 5565122,
      "step": 2030,
      "train_runtime": 5189.9589,
      "train_tokens_per_second": 1072.286
    },
    {
      "epoch": 0.16320134996685348,
      "grad_norm": 0.30416494607925415,
      "learning_rate": 6.477462437395659e-05,
      "loss": 1.0686,
      "num_input_tokens_seen": 5568044,
      "step": 2031,
      "train_runtime": 5192.1884,
      "train_tokens_per_second": 1072.389
    },
    {
      "epoch": 0.16328170513670423,
      "grad_norm": 0.3376562297344208,
      "learning_rate": 6.470784641068448e-05,
      "loss": 0.8914,
      "num_input_tokens_seen": 5574294,
      "step": 2032,
      "train_runtime": 5196.367,
      "train_tokens_per_second": 1072.729
    },
    {
      "epoch": 0.16336206030655498,
      "grad_norm": 0.2965356409549713,
      "learning_rate": 6.464106844741235e-05,
      "loss": 0.682,
      "num_input_tokens_seen": 5576298,
      "step": 2033,
      "train_runtime": 5198.2066,
      "train_tokens_per_second": 1072.735
    },
    {
      "epoch": 0.1634424154764057,
      "grad_norm": 0.33282458782196045,
      "learning_rate": 6.457429048414024e-05,
      "loss": 1.1757,
      "num_input_tokens_seen": 5579398,
      "step": 2034,
      "train_runtime": 5200.5787,
      "train_tokens_per_second": 1072.842
    },
    {
      "epoch": 0.16352277064625645,
      "grad_norm": 0.30046728253364563,
      "learning_rate": 6.450751252086812e-05,
      "loss": 1.1992,
      "num_input_tokens_seen": 5582100,
      "step": 2035,
      "train_runtime": 5202.7948,
      "train_tokens_per_second": 1072.904
    },
    {
      "epoch": 0.1636031258161072,
      "grad_norm": 0.3563089966773987,
      "learning_rate": 6.4440734557596e-05,
      "loss": 0.9134,
      "num_input_tokens_seen": 5585168,
      "step": 2036,
      "train_runtime": 5205.1611,
      "train_tokens_per_second": 1073.006
    },
    {
      "epoch": 0.16368348098595795,
      "grad_norm": 0.2713526487350464,
      "learning_rate": 6.437395659432388e-05,
      "loss": 1.0813,
      "num_input_tokens_seen": 5588818,
      "step": 2037,
      "train_runtime": 5207.7872,
      "train_tokens_per_second": 1073.166
    },
    {
      "epoch": 0.16376383615580867,
      "grad_norm": 0.3220246136188507,
      "learning_rate": 6.430717863105176e-05,
      "loss": 1.0672,
      "num_input_tokens_seen": 5592964,
      "step": 2038,
      "train_runtime": 5210.7337,
      "train_tokens_per_second": 1073.354
    },
    {
      "epoch": 0.16384419132565942,
      "grad_norm": 0.3526294231414795,
      "learning_rate": 6.424040066777963e-05,
      "loss": 1.1681,
      "num_input_tokens_seen": 5596184,
      "step": 2039,
      "train_runtime": 5213.0688,
      "train_tokens_per_second": 1073.491
    },
    {
      "epoch": 0.16392454649551016,
      "grad_norm": 0.4200699031352997,
      "learning_rate": 6.417362270450752e-05,
      "loss": 1.2005,
      "num_input_tokens_seen": 5599934,
      "step": 2040,
      "train_runtime": 5215.8182,
      "train_tokens_per_second": 1073.644
    },
    {
      "epoch": 0.16400490166536089,
      "grad_norm": 0.276871919631958,
      "learning_rate": 6.410684474123539e-05,
      "loss": 0.7408,
      "num_input_tokens_seen": 5601698,
      "step": 2041,
      "train_runtime": 5219.1761,
      "train_tokens_per_second": 1073.292
    },
    {
      "epoch": 0.16408525683521163,
      "grad_norm": 0.2654786705970764,
      "learning_rate": 6.404006677796328e-05,
      "loss": 0.8731,
      "num_input_tokens_seen": 5605180,
      "step": 2042,
      "train_runtime": 5221.7471,
      "train_tokens_per_second": 1073.43
    },
    {
      "epoch": 0.16416561200506238,
      "grad_norm": 0.3190358281135559,
      "learning_rate": 6.397328881469116e-05,
      "loss": 1.0788,
      "num_input_tokens_seen": 5608582,
      "step": 2043,
      "train_runtime": 5224.374,
      "train_tokens_per_second": 1073.541
    },
    {
      "epoch": 0.1642459671749131,
      "grad_norm": 0.31441324949264526,
      "learning_rate": 6.390651085141904e-05,
      "loss": 1.1984,
      "num_input_tokens_seen": 5610288,
      "step": 2044,
      "train_runtime": 5225.9917,
      "train_tokens_per_second": 1073.536
    },
    {
      "epoch": 0.16432632234476385,
      "grad_norm": 0.2650690972805023,
      "learning_rate": 6.383973288814692e-05,
      "loss": 0.7757,
      "num_input_tokens_seen": 5613108,
      "step": 2045,
      "train_runtime": 5228.2134,
      "train_tokens_per_second": 1073.619
    },
    {
      "epoch": 0.1644066775146146,
      "grad_norm": 0.3159467875957489,
      "learning_rate": 6.37729549248748e-05,
      "loss": 1.0578,
      "num_input_tokens_seen": 5614898,
      "step": 2046,
      "train_runtime": 5229.9417,
      "train_tokens_per_second": 1073.606
    },
    {
      "epoch": 0.16448703268446532,
      "grad_norm": 0.30501702427864075,
      "learning_rate": 6.370617696160267e-05,
      "loss": 0.8563,
      "num_input_tokens_seen": 5618320,
      "step": 2047,
      "train_runtime": 5232.6038,
      "train_tokens_per_second": 1073.714
    },
    {
      "epoch": 0.16456738785431607,
      "grad_norm": 0.28293177485466003,
      "learning_rate": 6.363939899833056e-05,
      "loss": 0.9986,
      "num_input_tokens_seen": 5621794,
      "step": 2048,
      "train_runtime": 5235.1232,
      "train_tokens_per_second": 1073.861
    },
    {
      "epoch": 0.16464774302416682,
      "grad_norm": 0.2597467303276062,
      "learning_rate": 6.357262103505843e-05,
      "loss": 0.8053,
      "num_input_tokens_seen": 5624258,
      "step": 2049,
      "train_runtime": 5237.1675,
      "train_tokens_per_second": 1073.912
    },
    {
      "epoch": 0.16472809819401757,
      "grad_norm": 0.47438162565231323,
      "learning_rate": 6.35058430717863e-05,
      "loss": 0.64,
      "num_input_tokens_seen": 5626210,
      "step": 2050,
      "train_runtime": 5238.957,
      "train_tokens_per_second": 1073.918
    },
    {
      "epoch": 0.1648084533638683,
      "grad_norm": 0.3228159248828888,
      "learning_rate": 6.343906510851419e-05,
      "loss": 1.0785,
      "num_input_tokens_seen": 5628302,
      "step": 2051,
      "train_runtime": 5240.7838,
      "train_tokens_per_second": 1073.943
    },
    {
      "epoch": 0.16488880853371904,
      "grad_norm": 0.4315949082374573,
      "learning_rate": 6.337228714524208e-05,
      "loss": 1.1422,
      "num_input_tokens_seen": 5631272,
      "step": 2052,
      "train_runtime": 5242.9917,
      "train_tokens_per_second": 1074.057
    },
    {
      "epoch": 0.1649691637035698,
      "grad_norm": 0.31984058022499084,
      "learning_rate": 6.330550918196996e-05,
      "loss": 1.2722,
      "num_input_tokens_seen": 5633816,
      "step": 2053,
      "train_runtime": 5245.0384,
      "train_tokens_per_second": 1074.123
    },
    {
      "epoch": 0.1650495188734205,
      "grad_norm": 0.31677675247192383,
      "learning_rate": 6.323873121869784e-05,
      "loss": 1.0599,
      "num_input_tokens_seen": 5636902,
      "step": 2054,
      "train_runtime": 5247.4159,
      "train_tokens_per_second": 1074.224
    },
    {
      "epoch": 0.16512987404327126,
      "grad_norm": 0.3054615557193756,
      "learning_rate": 6.317195325542571e-05,
      "loss": 0.9121,
      "num_input_tokens_seen": 5638898,
      "step": 2055,
      "train_runtime": 5249.2337,
      "train_tokens_per_second": 1074.233
    },
    {
      "epoch": 0.165210229213122,
      "grad_norm": 0.3366914987564087,
      "learning_rate": 6.31051752921536e-05,
      "loss": 1.083,
      "num_input_tokens_seen": 5642702,
      "step": 2056,
      "train_runtime": 5252.099,
      "train_tokens_per_second": 1074.371
    },
    {
      "epoch": 0.16529058438297273,
      "grad_norm": 0.3531474769115448,
      "learning_rate": 6.303839732888147e-05,
      "loss": 1.2471,
      "num_input_tokens_seen": 5645056,
      "step": 2057,
      "train_runtime": 5254.0375,
      "train_tokens_per_second": 1074.422
    },
    {
      "epoch": 0.16537093955282348,
      "grad_norm": 0.31711310148239136,
      "learning_rate": 6.297161936560934e-05,
      "loss": 0.7763,
      "num_input_tokens_seen": 5646962,
      "step": 2058,
      "train_runtime": 5255.9512,
      "train_tokens_per_second": 1074.394
    },
    {
      "epoch": 0.16545129472267422,
      "grad_norm": 0.34467220306396484,
      "learning_rate": 6.290484140233723e-05,
      "loss": 1.2711,
      "num_input_tokens_seen": 5650330,
      "step": 2059,
      "train_runtime": 5258.4124,
      "train_tokens_per_second": 1074.532
    },
    {
      "epoch": 0.16553164989252497,
      "grad_norm": 0.30267950892448425,
      "learning_rate": 6.283806343906511e-05,
      "loss": 0.9442,
      "num_input_tokens_seen": 5652646,
      "step": 2060,
      "train_runtime": 5260.383,
      "train_tokens_per_second": 1074.569
    },
    {
      "epoch": 0.1656120050623757,
      "grad_norm": 0.28625982999801636,
      "learning_rate": 6.2771285475793e-05,
      "loss": 0.907,
      "num_input_tokens_seen": 5656780,
      "step": 2061,
      "train_runtime": 5263.2694,
      "train_tokens_per_second": 1074.765
    },
    {
      "epoch": 0.16569236023222644,
      "grad_norm": 0.37177157402038574,
      "learning_rate": 6.270450751252087e-05,
      "loss": 1.2762,
      "num_input_tokens_seen": 5659600,
      "step": 2062,
      "train_runtime": 5265.4257,
      "train_tokens_per_second": 1074.861
    },
    {
      "epoch": 0.1657727154020772,
      "grad_norm": 0.34112870693206787,
      "learning_rate": 6.263772954924875e-05,
      "loss": 0.864,
      "num_input_tokens_seen": 5662498,
      "step": 2063,
      "train_runtime": 5267.6289,
      "train_tokens_per_second": 1074.961
    },
    {
      "epoch": 0.1658530705719279,
      "grad_norm": 0.3134649693965912,
      "learning_rate": 6.257095158597663e-05,
      "loss": 0.6045,
      "num_input_tokens_seen": 5665668,
      "step": 2064,
      "train_runtime": 5269.9908,
      "train_tokens_per_second": 1075.081
    },
    {
      "epoch": 0.16593342574177866,
      "grad_norm": 0.3043017089366913,
      "learning_rate": 6.250417362270451e-05,
      "loss": 1.0348,
      "num_input_tokens_seen": 5668906,
      "step": 2065,
      "train_runtime": 5272.3977,
      "train_tokens_per_second": 1075.205
    },
    {
      "epoch": 0.1660137809116294,
      "grad_norm": 0.32345324754714966,
      "learning_rate": 6.243739565943238e-05,
      "loss": 0.8914,
      "num_input_tokens_seen": 5672164,
      "step": 2066,
      "train_runtime": 5274.8868,
      "train_tokens_per_second": 1075.315
    },
    {
      "epoch": 0.16609413608148013,
      "grad_norm": 0.368843674659729,
      "learning_rate": 6.237061769616027e-05,
      "loss": 1.2069,
      "num_input_tokens_seen": 5674750,
      "step": 2067,
      "train_runtime": 5277.0032,
      "train_tokens_per_second": 1075.374
    },
    {
      "epoch": 0.16617449125133088,
      "grad_norm": 0.35405561327934265,
      "learning_rate": 6.230383973288815e-05,
      "loss": 1.1689,
      "num_input_tokens_seen": 5676472,
      "step": 2068,
      "train_runtime": 5278.7059,
      "train_tokens_per_second": 1075.353
    },
    {
      "epoch": 0.16625484642118163,
      "grad_norm": 0.28620272874832153,
      "learning_rate": 6.223706176961604e-05,
      "loss": 0.984,
      "num_input_tokens_seen": 5679514,
      "step": 2069,
      "train_runtime": 5281.048,
      "train_tokens_per_second": 1075.452
    },
    {
      "epoch": 0.16633520159103235,
      "grad_norm": 0.3701074719429016,
      "learning_rate": 6.217028380634391e-05,
      "loss": 1.0772,
      "num_input_tokens_seen": 5681956,
      "step": 2070,
      "train_runtime": 5283.0865,
      "train_tokens_per_second": 1075.499
    },
    {
      "epoch": 0.1664155567608831,
      "grad_norm": 0.46768495440483093,
      "learning_rate": 6.210350584307179e-05,
      "loss": 1.0691,
      "num_input_tokens_seen": 5685598,
      "step": 2071,
      "train_runtime": 5287.3326,
      "train_tokens_per_second": 1075.324
    },
    {
      "epoch": 0.16649591193073385,
      "grad_norm": 0.32651495933532715,
      "learning_rate": 6.203672787979967e-05,
      "loss": 0.8737,
      "num_input_tokens_seen": 5687910,
      "step": 2072,
      "train_runtime": 5289.3436,
      "train_tokens_per_second": 1075.353
    },
    {
      "epoch": 0.1665762671005846,
      "grad_norm": 0.3948456943035126,
      "learning_rate": 6.196994991652755e-05,
      "loss": 1.0487,
      "num_input_tokens_seen": 5693982,
      "step": 2073,
      "train_runtime": 5293.3662,
      "train_tokens_per_second": 1075.683
    },
    {
      "epoch": 0.16665662227043532,
      "grad_norm": 0.386605441570282,
      "learning_rate": 6.190317195325542e-05,
      "loss": 1.089,
      "num_input_tokens_seen": 5697036,
      "step": 2074,
      "train_runtime": 5295.7537,
      "train_tokens_per_second": 1075.774
    },
    {
      "epoch": 0.16673697744028607,
      "grad_norm": 0.31722506880760193,
      "learning_rate": 6.183639398998331e-05,
      "loss": 1.1406,
      "num_input_tokens_seen": 5700596,
      "step": 2075,
      "train_runtime": 5298.4093,
      "train_tokens_per_second": 1075.907
    },
    {
      "epoch": 0.16681733261013681,
      "grad_norm": 0.31338486075401306,
      "learning_rate": 6.176961602671118e-05,
      "loss": 0.685,
      "num_input_tokens_seen": 5704598,
      "step": 2076,
      "train_runtime": 5301.298,
      "train_tokens_per_second": 1076.076
    },
    {
      "epoch": 0.16689768777998754,
      "grad_norm": 0.31885960698127747,
      "learning_rate": 6.170283806343907e-05,
      "loss": 0.8387,
      "num_input_tokens_seen": 5707762,
      "step": 2077,
      "train_runtime": 5303.7714,
      "train_tokens_per_second": 1076.17
    },
    {
      "epoch": 0.16697804294983828,
      "grad_norm": 0.3406186103820801,
      "learning_rate": 6.163606010016695e-05,
      "loss": 1.0031,
      "num_input_tokens_seen": 5709990,
      "step": 2078,
      "train_runtime": 5305.7598,
      "train_tokens_per_second": 1076.187
    },
    {
      "epoch": 0.16705839811968903,
      "grad_norm": 0.32604217529296875,
      "learning_rate": 6.156928213689483e-05,
      "loss": 0.8163,
      "num_input_tokens_seen": 5712006,
      "step": 2079,
      "train_runtime": 5307.6824,
      "train_tokens_per_second": 1076.177
    },
    {
      "epoch": 0.16713875328953975,
      "grad_norm": 0.3653429448604584,
      "learning_rate": 6.150250417362271e-05,
      "loss": 0.9424,
      "num_input_tokens_seen": 5714072,
      "step": 2080,
      "train_runtime": 5309.5057,
      "train_tokens_per_second": 1076.197
    },
    {
      "epoch": 0.1672191084593905,
      "grad_norm": 0.3046902120113373,
      "learning_rate": 6.143572621035059e-05,
      "loss": 0.8268,
      "num_input_tokens_seen": 5717166,
      "step": 2081,
      "train_runtime": 5311.8409,
      "train_tokens_per_second": 1076.306
    },
    {
      "epoch": 0.16729946362924125,
      "grad_norm": 0.3862263858318329,
      "learning_rate": 6.136894824707846e-05,
      "loss": 0.9915,
      "num_input_tokens_seen": 5718692,
      "step": 2082,
      "train_runtime": 5313.4471,
      "train_tokens_per_second": 1076.268
    },
    {
      "epoch": 0.16737981879909197,
      "grad_norm": 0.289603054523468,
      "learning_rate": 6.130217028380635e-05,
      "loss": 0.8169,
      "num_input_tokens_seen": 5721638,
      "step": 2083,
      "train_runtime": 5315.623,
      "train_tokens_per_second": 1076.381
    },
    {
      "epoch": 0.16746017396894272,
      "grad_norm": 0.2774498462677002,
      "learning_rate": 6.123539232053422e-05,
      "loss": 0.7985,
      "num_input_tokens_seen": 5725192,
      "step": 2084,
      "train_runtime": 5318.2295,
      "train_tokens_per_second": 1076.522
    },
    {
      "epoch": 0.16754052913879347,
      "grad_norm": 0.31284478306770325,
      "learning_rate": 6.11686143572621e-05,
      "loss": 0.6903,
      "num_input_tokens_seen": 5727052,
      "step": 2085,
      "train_runtime": 5319.9579,
      "train_tokens_per_second": 1076.522
    },
    {
      "epoch": 0.16762088430864422,
      "grad_norm": 0.3342660367488861,
      "learning_rate": 6.110183639398999e-05,
      "loss": 0.9719,
      "num_input_tokens_seen": 5732260,
      "step": 2086,
      "train_runtime": 5323.4947,
      "train_tokens_per_second": 1076.785
    },
    {
      "epoch": 0.16770123947849494,
      "grad_norm": 0.304916650056839,
      "learning_rate": 6.103505843071786e-05,
      "loss": 0.8482,
      "num_input_tokens_seen": 5734388,
      "step": 2087,
      "train_runtime": 5325.3133,
      "train_tokens_per_second": 1076.817
    },
    {
      "epoch": 0.1677815946483457,
      "grad_norm": 0.32424411177635193,
      "learning_rate": 6.0968280467445746e-05,
      "loss": 0.9961,
      "num_input_tokens_seen": 5737374,
      "step": 2088,
      "train_runtime": 5327.6219,
      "train_tokens_per_second": 1076.911
    },
    {
      "epoch": 0.16786194981819644,
      "grad_norm": 0.4448409378528595,
      "learning_rate": 6.0901502504173626e-05,
      "loss": 1.0527,
      "num_input_tokens_seen": 5739696,
      "step": 2089,
      "train_runtime": 5329.524,
      "train_tokens_per_second": 1076.962
    },
    {
      "epoch": 0.16794230498804716,
      "grad_norm": 0.3010452091693878,
      "learning_rate": 6.083472454090151e-05,
      "loss": 1.08,
      "num_input_tokens_seen": 5745940,
      "step": 2090,
      "train_runtime": 5333.6509,
      "train_tokens_per_second": 1077.3
    },
    {
      "epoch": 0.1680226601578979,
      "grad_norm": 0.3194243907928467,
      "learning_rate": 6.0767946577629386e-05,
      "loss": 0.8631,
      "num_input_tokens_seen": 5747926,
      "step": 2091,
      "train_runtime": 5335.4773,
      "train_tokens_per_second": 1077.303
    },
    {
      "epoch": 0.16810301532774866,
      "grad_norm": 0.33902600407600403,
      "learning_rate": 6.070116861435726e-05,
      "loss": 1.2106,
      "num_input_tokens_seen": 5752358,
      "step": 2092,
      "train_runtime": 5338.5443,
      "train_tokens_per_second": 1077.514
    },
    {
      "epoch": 0.16818337049759938,
      "grad_norm": 0.31348052620887756,
      "learning_rate": 6.0634390651085146e-05,
      "loss": 1.0063,
      "num_input_tokens_seen": 5755212,
      "step": 2093,
      "train_runtime": 5340.715,
      "train_tokens_per_second": 1077.611
    },
    {
      "epoch": 0.16826372566745013,
      "grad_norm": 0.3678327798843384,
      "learning_rate": 6.0567612687813026e-05,
      "loss": 1.0293,
      "num_input_tokens_seen": 5759412,
      "step": 2094,
      "train_runtime": 5343.8144,
      "train_tokens_per_second": 1077.772
    },
    {
      "epoch": 0.16834408083730087,
      "grad_norm": 0.25327804684638977,
      "learning_rate": 6.05008347245409e-05,
      "loss": 0.8533,
      "num_input_tokens_seen": 5762774,
      "step": 2095,
      "train_runtime": 5346.213,
      "train_tokens_per_second": 1077.917
    },
    {
      "epoch": 0.16842443600715162,
      "grad_norm": 0.31719061732292175,
      "learning_rate": 6.0434056761268785e-05,
      "loss": 1.2534,
      "num_input_tokens_seen": 5766262,
      "step": 2096,
      "train_runtime": 5348.7901,
      "train_tokens_per_second": 1078.05
    },
    {
      "epoch": 0.16850479117700234,
      "grad_norm": 0.28992027044296265,
      "learning_rate": 6.0367278797996665e-05,
      "loss": 0.869,
      "num_input_tokens_seen": 5770454,
      "step": 2097,
      "train_runtime": 5351.8107,
      "train_tokens_per_second": 1078.225
    },
    {
      "epoch": 0.1685851463468531,
      "grad_norm": 0.28021690249443054,
      "learning_rate": 6.030050083472455e-05,
      "loss": 1.0398,
      "num_input_tokens_seen": 5774530,
      "step": 2098,
      "train_runtime": 5354.7263,
      "train_tokens_per_second": 1078.399
    },
    {
      "epoch": 0.16866550151670384,
      "grad_norm": 0.34704574942588806,
      "learning_rate": 6.0233722871452425e-05,
      "loss": 0.8232,
      "num_input_tokens_seen": 5777106,
      "step": 2099,
      "train_runtime": 5356.7587,
      "train_tokens_per_second": 1078.47
    },
    {
      "epoch": 0.16874585668655456,
      "grad_norm": 0.35061630606651306,
      "learning_rate": 6.01669449081803e-05,
      "loss": 0.9711,
      "num_input_tokens_seen": 5779562,
      "step": 2100,
      "train_runtime": 5358.8311,
      "train_tokens_per_second": 1078.512
    },
    {
      "epoch": 0.16874585668655456,
      "eval_loss": 0.9555262327194214,
      "eval_runtime": 91.7902,
      "eval_samples_per_second": 10.829,
      "eval_steps_per_second": 5.415,
      "num_input_tokens_seen": 5779562,
      "step": 2100
    },
    {
      "epoch": 0.1688262118564053,
      "grad_norm": 0.29205822944641113,
      "learning_rate": 6.0100166944908185e-05,
      "loss": 1.0588,
      "num_input_tokens_seen": 5782154,
      "step": 2101,
      "train_runtime": 5454.7148,
      "train_tokens_per_second": 1060.029
    },
    {
      "epoch": 0.16890656702625606,
      "grad_norm": 0.37343764305114746,
      "learning_rate": 6.0033388981636065e-05,
      "loss": 1.1664,
      "num_input_tokens_seen": 5784936,
      "step": 2102,
      "train_runtime": 5456.8828,
      "train_tokens_per_second": 1060.117
    },
    {
      "epoch": 0.16898692219610678,
      "grad_norm": 0.2603963613510132,
      "learning_rate": 5.996661101836394e-05,
      "loss": 1.057,
      "num_input_tokens_seen": 5788276,
      "step": 2103,
      "train_runtime": 5459.425,
      "train_tokens_per_second": 1060.235
    },
    {
      "epoch": 0.16906727736595753,
      "grad_norm": 0.3672252297401428,
      "learning_rate": 5.9899833055091825e-05,
      "loss": 1.218,
      "num_input_tokens_seen": 5790588,
      "step": 2104,
      "train_runtime": 5461.4335,
      "train_tokens_per_second": 1060.269
    },
    {
      "epoch": 0.16914763253580828,
      "grad_norm": 0.3165772557258606,
      "learning_rate": 5.98330550918197e-05,
      "loss": 0.7407,
      "num_input_tokens_seen": 5793890,
      "step": 2105,
      "train_runtime": 5463.9333,
      "train_tokens_per_second": 1060.388
    },
    {
      "epoch": 0.169227987705659,
      "grad_norm": 0.3723558485507965,
      "learning_rate": 5.9766277128547585e-05,
      "loss": 1.043,
      "num_input_tokens_seen": 5796290,
      "step": 2106,
      "train_runtime": 5466.069,
      "train_tokens_per_second": 1060.413
    },
    {
      "epoch": 0.16930834287550975,
      "grad_norm": 0.3076302111148834,
      "learning_rate": 5.9699499165275465e-05,
      "loss": 0.9268,
      "num_input_tokens_seen": 5798656,
      "step": 2107,
      "train_runtime": 5468.0054,
      "train_tokens_per_second": 1060.47
    },
    {
      "epoch": 0.1693886980453605,
      "grad_norm": 0.3354271352291107,
      "learning_rate": 5.963272120200334e-05,
      "loss": 1.1592,
      "num_input_tokens_seen": 5801386,
      "step": 2108,
      "train_runtime": 5470.2001,
      "train_tokens_per_second": 1060.544
    },
    {
      "epoch": 0.16946905321521125,
      "grad_norm": 0.2791424095630646,
      "learning_rate": 5.9565943238731224e-05,
      "loss": 0.9955,
      "num_input_tokens_seen": 5803808,
      "step": 2109,
      "train_runtime": 5472.253,
      "train_tokens_per_second": 1060.588
    },
    {
      "epoch": 0.16954940838506197,
      "grad_norm": 0.3554924726486206,
      "learning_rate": 5.9499165275459104e-05,
      "loss": 0.9301,
      "num_input_tokens_seen": 5805480,
      "step": 2110,
      "train_runtime": 5473.88,
      "train_tokens_per_second": 1060.579
    },
    {
      "epoch": 0.16962976355491272,
      "grad_norm": 0.36179855465888977,
      "learning_rate": 5.943238731218698e-05,
      "loss": 1.238,
      "num_input_tokens_seen": 5806822,
      "step": 2111,
      "train_runtime": 5475.4525,
      "train_tokens_per_second": 1060.519
    },
    {
      "epoch": 0.16971011872476346,
      "grad_norm": 0.33809080719947815,
      "learning_rate": 5.9365609348914864e-05,
      "loss": 0.9749,
      "num_input_tokens_seen": 5809482,
      "step": 2112,
      "train_runtime": 5477.6398,
      "train_tokens_per_second": 1060.581
    },
    {
      "epoch": 0.16979047389461419,
      "grad_norm": 0.35085058212280273,
      "learning_rate": 5.929883138564274e-05,
      "loss": 0.9903,
      "num_input_tokens_seen": 5812106,
      "step": 2113,
      "train_runtime": 5479.785,
      "train_tokens_per_second": 1060.645
    },
    {
      "epoch": 0.16987082906446493,
      "grad_norm": 0.5488579869270325,
      "learning_rate": 5.9232053422370624e-05,
      "loss": 0.9423,
      "num_input_tokens_seen": 5814802,
      "step": 2114,
      "train_runtime": 5481.9787,
      "train_tokens_per_second": 1060.712
    },
    {
      "epoch": 0.16995118423431568,
      "grad_norm": 0.250041127204895,
      "learning_rate": 5.9165275459098504e-05,
      "loss": 0.9151,
      "num_input_tokens_seen": 5817138,
      "step": 2115,
      "train_runtime": 5483.8562,
      "train_tokens_per_second": 1060.775
    },
    {
      "epoch": 0.1700315394041664,
      "grad_norm": 0.2918633818626404,
      "learning_rate": 5.909849749582638e-05,
      "loss": 0.8644,
      "num_input_tokens_seen": 5821178,
      "step": 2116,
      "train_runtime": 5486.6776,
      "train_tokens_per_second": 1060.966
    },
    {
      "epoch": 0.17011189457401715,
      "grad_norm": 0.3103772699832916,
      "learning_rate": 5.9031719532554264e-05,
      "loss": 0.7843,
      "num_input_tokens_seen": 5823280,
      "step": 2117,
      "train_runtime": 5488.652,
      "train_tokens_per_second": 1060.967
    },
    {
      "epoch": 0.1701922497438679,
      "grad_norm": 0.4383689761161804,
      "learning_rate": 5.896494156928214e-05,
      "loss": 0.9798,
      "num_input_tokens_seen": 5827230,
      "step": 2118,
      "train_runtime": 5491.5274,
      "train_tokens_per_second": 1061.131
    },
    {
      "epoch": 0.17027260491371862,
      "grad_norm": 0.31260019540786743,
      "learning_rate": 5.889816360601002e-05,
      "loss": 0.9507,
      "num_input_tokens_seen": 5829018,
      "step": 2119,
      "train_runtime": 5493.187,
      "train_tokens_per_second": 1061.136
    },
    {
      "epoch": 0.17035296008356937,
      "grad_norm": 0.3282405734062195,
      "learning_rate": 5.8831385642737904e-05,
      "loss": 0.7941,
      "num_input_tokens_seen": 5831650,
      "step": 2120,
      "train_runtime": 5495.2362,
      "train_tokens_per_second": 1061.219
    },
    {
      "epoch": 0.17043331525342012,
      "grad_norm": 0.4777829647064209,
      "learning_rate": 5.876460767946578e-05,
      "loss": 1.209,
      "num_input_tokens_seen": 5833540,
      "step": 2121,
      "train_runtime": 5497.0435,
      "train_tokens_per_second": 1061.214
    },
    {
      "epoch": 0.17051367042327087,
      "grad_norm": 0.2787031829357147,
      "learning_rate": 5.8697829716193664e-05,
      "loss": 1.0741,
      "num_input_tokens_seen": 5836860,
      "step": 2122,
      "train_runtime": 5499.5386,
      "train_tokens_per_second": 1061.336
    },
    {
      "epoch": 0.1705940255931216,
      "grad_norm": 0.25041425228118896,
      "learning_rate": 5.863105175292154e-05,
      "loss": 0.5899,
      "num_input_tokens_seen": 5839532,
      "step": 2123,
      "train_runtime": 5501.8578,
      "train_tokens_per_second": 1061.375
    },
    {
      "epoch": 0.17067438076297234,
      "grad_norm": 0.27655819058418274,
      "learning_rate": 5.856427378964942e-05,
      "loss": 0.8819,
      "num_input_tokens_seen": 5842316,
      "step": 2124,
      "train_runtime": 5504.0998,
      "train_tokens_per_second": 1061.448
    },
    {
      "epoch": 0.1707547359328231,
      "grad_norm": 0.2906401753425598,
      "learning_rate": 5.84974958263773e-05,
      "loss": 0.7467,
      "num_input_tokens_seen": 5845314,
      "step": 2125,
      "train_runtime": 5506.4001,
      "train_tokens_per_second": 1061.549
    },
    {
      "epoch": 0.1708350911026738,
      "grad_norm": 0.3474268615245819,
      "learning_rate": 5.8430717863105176e-05,
      "loss": 1.0808,
      "num_input_tokens_seen": 5849972,
      "step": 2126,
      "train_runtime": 5509.7259,
      "train_tokens_per_second": 1061.754
    },
    {
      "epoch": 0.17091544627252456,
      "grad_norm": 0.2856059968471527,
      "learning_rate": 5.8363939899833056e-05,
      "loss": 0.7916,
      "num_input_tokens_seen": 5852764,
      "step": 2127,
      "train_runtime": 5511.945,
      "train_tokens_per_second": 1061.833
    },
    {
      "epoch": 0.1709958014423753,
      "grad_norm": 0.3286946713924408,
      "learning_rate": 5.829716193656094e-05,
      "loss": 1.1438,
      "num_input_tokens_seen": 5856044,
      "step": 2128,
      "train_runtime": 5514.4225,
      "train_tokens_per_second": 1061.951
    },
    {
      "epoch": 0.17107615661222603,
      "grad_norm": 0.24851734936237335,
      "learning_rate": 5.8230383973288816e-05,
      "loss": 0.8662,
      "num_input_tokens_seen": 5859522,
      "step": 2129,
      "train_runtime": 5516.9343,
      "train_tokens_per_second": 1062.097
    },
    {
      "epoch": 0.17115651178207678,
      "grad_norm": 0.3044750392436981,
      "learning_rate": 5.81636060100167e-05,
      "loss": 0.8745,
      "num_input_tokens_seen": 5861638,
      "step": 2130,
      "train_runtime": 5518.7387,
      "train_tokens_per_second": 1062.134
    },
    {
      "epoch": 0.17123686695192752,
      "grad_norm": 0.33549076318740845,
      "learning_rate": 5.8096828046744576e-05,
      "loss": 0.9702,
      "num_input_tokens_seen": 5864052,
      "step": 2131,
      "train_runtime": 5522.7266,
      "train_tokens_per_second": 1061.804
    },
    {
      "epoch": 0.17131722212177827,
      "grad_norm": 0.3433608412742615,
      "learning_rate": 5.8030050083472456e-05,
      "loss": 1.2358,
      "num_input_tokens_seen": 5867970,
      "step": 2132,
      "train_runtime": 5525.467,
      "train_tokens_per_second": 1061.986
    },
    {
      "epoch": 0.171397577291629,
      "grad_norm": 0.31587350368499756,
      "learning_rate": 5.796327212020034e-05,
      "loss": 1.0999,
      "num_input_tokens_seen": 5871172,
      "step": 2133,
      "train_runtime": 5527.9147,
      "train_tokens_per_second": 1062.095
    },
    {
      "epoch": 0.17147793246147974,
      "grad_norm": 0.29557958245277405,
      "learning_rate": 5.7896494156928216e-05,
      "loss": 0.9023,
      "num_input_tokens_seen": 5874522,
      "step": 2134,
      "train_runtime": 5530.494,
      "train_tokens_per_second": 1062.206
    },
    {
      "epoch": 0.1715582876313305,
      "grad_norm": 0.3485921621322632,
      "learning_rate": 5.782971619365609e-05,
      "loss": 1.4575,
      "num_input_tokens_seen": 5877366,
      "step": 2135,
      "train_runtime": 5532.6758,
      "train_tokens_per_second": 1062.301
    },
    {
      "epoch": 0.1716386428011812,
      "grad_norm": 0.328443706035614,
      "learning_rate": 5.7762938230383976e-05,
      "loss": 0.9593,
      "num_input_tokens_seen": 5879546,
      "step": 2136,
      "train_runtime": 5534.6312,
      "train_tokens_per_second": 1062.319
    },
    {
      "epoch": 0.17171899797103196,
      "grad_norm": 0.30363771319389343,
      "learning_rate": 5.7696160267111856e-05,
      "loss": 0.9198,
      "num_input_tokens_seen": 5883364,
      "step": 2137,
      "train_runtime": 5537.4513,
      "train_tokens_per_second": 1062.468
    },
    {
      "epoch": 0.1717993531408827,
      "grad_norm": 0.34566518664360046,
      "learning_rate": 5.762938230383974e-05,
      "loss": 0.9268,
      "num_input_tokens_seen": 5886500,
      "step": 2138,
      "train_runtime": 5539.9076,
      "train_tokens_per_second": 1062.563
    },
    {
      "epoch": 0.17187970831073343,
      "grad_norm": 0.26693105697631836,
      "learning_rate": 5.7562604340567616e-05,
      "loss": 0.6212,
      "num_input_tokens_seen": 5888414,
      "step": 2139,
      "train_runtime": 5541.6587,
      "train_tokens_per_second": 1062.572
    },
    {
      "epoch": 0.17196006348058418,
      "grad_norm": 0.22253353893756866,
      "learning_rate": 5.749582637729549e-05,
      "loss": 0.5865,
      "num_input_tokens_seen": 5891498,
      "step": 2140,
      "train_runtime": 5544.0344,
      "train_tokens_per_second": 1062.673
    },
    {
      "epoch": 0.17204041865043493,
      "grad_norm": 0.3196808099746704,
      "learning_rate": 5.7429048414023375e-05,
      "loss": 1.2904,
      "num_input_tokens_seen": 5895136,
      "step": 2141,
      "train_runtime": 5546.7029,
      "train_tokens_per_second": 1062.818
    },
    {
      "epoch": 0.17212077382028565,
      "grad_norm": 0.2977009415626526,
      "learning_rate": 5.7362270450751255e-05,
      "loss": 1.0414,
      "num_input_tokens_seen": 5896872,
      "step": 2142,
      "train_runtime": 5548.3192,
      "train_tokens_per_second": 1062.821
    },
    {
      "epoch": 0.1722011289901364,
      "grad_norm": 0.603030264377594,
      "learning_rate": 5.729549248747913e-05,
      "loss": 1.3422,
      "num_input_tokens_seen": 5899618,
      "step": 2143,
      "train_runtime": 5550.4526,
      "train_tokens_per_second": 1062.908
    },
    {
      "epoch": 0.17228148415998715,
      "grad_norm": 0.3160606324672699,
      "learning_rate": 5.7228714524207015e-05,
      "loss": 1.1736,
      "num_input_tokens_seen": 5902676,
      "step": 2144,
      "train_runtime": 5552.8535,
      "train_tokens_per_second": 1062.999
    },
    {
      "epoch": 0.1723618393298379,
      "grad_norm": 0.33425402641296387,
      "learning_rate": 5.7161936560934895e-05,
      "loss": 1.0781,
      "num_input_tokens_seen": 5904826,
      "step": 2145,
      "train_runtime": 5554.7041,
      "train_tokens_per_second": 1063.032
    },
    {
      "epoch": 0.17244219449968862,
      "grad_norm": 0.3198067843914032,
      "learning_rate": 5.709515859766278e-05,
      "loss": 1.0518,
      "num_input_tokens_seen": 5911678,
      "step": 2146,
      "train_runtime": 5559.2346,
      "train_tokens_per_second": 1063.398
    },
    {
      "epoch": 0.17252254966953937,
      "grad_norm": 0.3363427221775055,
      "learning_rate": 5.7028380634390655e-05,
      "loss": 0.6258,
      "num_input_tokens_seen": 5914912,
      "step": 2147,
      "train_runtime": 5561.6859,
      "train_tokens_per_second": 1063.511
    },
    {
      "epoch": 0.17260290483939011,
      "grad_norm": 0.40328025817871094,
      "learning_rate": 5.696160267111853e-05,
      "loss": 0.8357,
      "num_input_tokens_seen": 5918354,
      "step": 2148,
      "train_runtime": 5564.2906,
      "train_tokens_per_second": 1063.631
    },
    {
      "epoch": 0.17268326000924084,
      "grad_norm": 0.2903386354446411,
      "learning_rate": 5.6894824707846415e-05,
      "loss": 0.9364,
      "num_input_tokens_seen": 5921040,
      "step": 2149,
      "train_runtime": 5566.3747,
      "train_tokens_per_second": 1063.716
    },
    {
      "epoch": 0.17276361517909158,
      "grad_norm": 0.31180712580680847,
      "learning_rate": 5.6828046744574295e-05,
      "loss": 0.9142,
      "num_input_tokens_seen": 5923840,
      "step": 2150,
      "train_runtime": 5568.6209,
      "train_tokens_per_second": 1063.789
    },
    {
      "epoch": 0.17284397034894233,
      "grad_norm": 0.3400718569755554,
      "learning_rate": 5.676126878130217e-05,
      "loss": 0.7969,
      "num_input_tokens_seen": 5926654,
      "step": 2151,
      "train_runtime": 5570.7788,
      "train_tokens_per_second": 1063.882
    },
    {
      "epoch": 0.17292432551879305,
      "grad_norm": 0.32087886333465576,
      "learning_rate": 5.6694490818030055e-05,
      "loss": 1.1123,
      "num_input_tokens_seen": 5933130,
      "step": 2152,
      "train_runtime": 5575.1339,
      "train_tokens_per_second": 1064.213
    },
    {
      "epoch": 0.1730046806886438,
      "grad_norm": 0.26884207129478455,
      "learning_rate": 5.662771285475793e-05,
      "loss": 0.908,
      "num_input_tokens_seen": 5937852,
      "step": 2153,
      "train_runtime": 5578.5348,
      "train_tokens_per_second": 1064.411
    },
    {
      "epoch": 0.17308503585849455,
      "grad_norm": 0.37048181891441345,
      "learning_rate": 5.6560934891485815e-05,
      "loss": 0.7005,
      "num_input_tokens_seen": 5940014,
      "step": 2154,
      "train_runtime": 5580.3666,
      "train_tokens_per_second": 1064.449
    },
    {
      "epoch": 0.17316539102834527,
      "grad_norm": 0.3180694580078125,
      "learning_rate": 5.6494156928213694e-05,
      "loss": 0.9271,
      "num_input_tokens_seen": 5943994,
      "step": 2155,
      "train_runtime": 5583.3319,
      "train_tokens_per_second": 1064.596
    },
    {
      "epoch": 0.17324574619819602,
      "grad_norm": 0.2703879475593567,
      "learning_rate": 5.642737896494157e-05,
      "loss": 0.6581,
      "num_input_tokens_seen": 5947668,
      "step": 2156,
      "train_runtime": 5585.9431,
      "train_tokens_per_second": 1064.756
    },
    {
      "epoch": 0.17332610136804677,
      "grad_norm": 0.2960408329963684,
      "learning_rate": 5.6360601001669454e-05,
      "loss": 0.9948,
      "num_input_tokens_seen": 5949860,
      "step": 2157,
      "train_runtime": 5587.8377,
      "train_tokens_per_second": 1064.788
    },
    {
      "epoch": 0.17340645653789752,
      "grad_norm": 0.371141642332077,
      "learning_rate": 5.629382303839733e-05,
      "loss": 1.0421,
      "num_input_tokens_seen": 5952540,
      "step": 2158,
      "train_runtime": 5590.0524,
      "train_tokens_per_second": 1064.845
    },
    {
      "epoch": 0.17348681170774824,
      "grad_norm": 0.4357776343822479,
      "learning_rate": 5.622704507512521e-05,
      "loss": 0.8462,
      "num_input_tokens_seen": 5954938,
      "step": 2159,
      "train_runtime": 5592.0527,
      "train_tokens_per_second": 1064.893
    },
    {
      "epoch": 0.173567166877599,
      "grad_norm": 0.356094092130661,
      "learning_rate": 5.6160267111853094e-05,
      "loss": 1.0046,
      "num_input_tokens_seen": 5956392,
      "step": 2160,
      "train_runtime": 5593.652,
      "train_tokens_per_second": 1064.849
    },
    {
      "epoch": 0.17364752204744974,
      "grad_norm": 0.23461705446243286,
      "learning_rate": 5.609348914858097e-05,
      "loss": 0.759,
      "num_input_tokens_seen": 5959500,
      "step": 2161,
      "train_runtime": 5598.6446,
      "train_tokens_per_second": 1064.454
    },
    {
      "epoch": 0.17372787721730046,
      "grad_norm": 0.31692934036254883,
      "learning_rate": 5.6026711185308854e-05,
      "loss": 1.0124,
      "num_input_tokens_seen": 5961598,
      "step": 2162,
      "train_runtime": 5600.4215,
      "train_tokens_per_second": 1064.491
    },
    {
      "epoch": 0.1738082323871512,
      "grad_norm": 0.40736809372901917,
      "learning_rate": 5.5959933222036734e-05,
      "loss": 0.9999,
      "num_input_tokens_seen": 5963460,
      "step": 2163,
      "train_runtime": 5602.1499,
      "train_tokens_per_second": 1064.495
    },
    {
      "epoch": 0.17388858755700196,
      "grad_norm": 0.2923264801502228,
      "learning_rate": 5.589315525876461e-05,
      "loss": 0.8445,
      "num_input_tokens_seen": 5965208,
      "step": 2164,
      "train_runtime": 5603.87,
      "train_tokens_per_second": 1064.48
    },
    {
      "epoch": 0.17396894272685268,
      "grad_norm": 0.31024986505508423,
      "learning_rate": 5.5826377295492494e-05,
      "loss": 1.2309,
      "num_input_tokens_seen": 5968326,
      "step": 2165,
      "train_runtime": 5606.2184,
      "train_tokens_per_second": 1064.59
    },
    {
      "epoch": 0.17404929789670343,
      "grad_norm": 0.382074773311615,
      "learning_rate": 5.575959933222037e-05,
      "loss": 1.0833,
      "num_input_tokens_seen": 5970660,
      "step": 2166,
      "train_runtime": 5608.2012,
      "train_tokens_per_second": 1064.63
    },
    {
      "epoch": 0.17412965306655417,
      "grad_norm": 0.32277077436447144,
      "learning_rate": 5.569282136894825e-05,
      "loss": 0.8775,
      "num_input_tokens_seen": 5972126,
      "step": 2167,
      "train_runtime": 5609.7733,
      "train_tokens_per_second": 1064.593
    },
    {
      "epoch": 0.17421000823640492,
      "grad_norm": 0.3061690330505371,
      "learning_rate": 5.5626043405676134e-05,
      "loss": 0.7693,
      "num_input_tokens_seen": 5973952,
      "step": 2168,
      "train_runtime": 5611.6301,
      "train_tokens_per_second": 1064.566
    },
    {
      "epoch": 0.17429036340625564,
      "grad_norm": 0.4055381119251251,
      "learning_rate": 5.555926544240401e-05,
      "loss": 1.2553,
      "num_input_tokens_seen": 5975970,
      "step": 2169,
      "train_runtime": 5613.4791,
      "train_tokens_per_second": 1064.575
    },
    {
      "epoch": 0.1743707185761064,
      "grad_norm": 0.38947683572769165,
      "learning_rate": 5.5492487479131893e-05,
      "loss": 1.0117,
      "num_input_tokens_seen": 5978158,
      "step": 2170,
      "train_runtime": 5615.4511,
      "train_tokens_per_second": 1064.591
    },
    {
      "epoch": 0.17445107374595714,
      "grad_norm": 0.2549003064632416,
      "learning_rate": 5.5425709515859767e-05,
      "loss": 0.5143,
      "num_input_tokens_seen": 5980528,
      "step": 2171,
      "train_runtime": 5617.437,
      "train_tokens_per_second": 1064.636
    },
    {
      "epoch": 0.17453142891580786,
      "grad_norm": 0.3197709321975708,
      "learning_rate": 5.5358931552587646e-05,
      "loss": 0.8939,
      "num_input_tokens_seen": 5983044,
      "step": 2172,
      "train_runtime": 5619.566,
      "train_tokens_per_second": 1064.681
    },
    {
      "epoch": 0.1746117840856586,
      "grad_norm": 0.32079893350601196,
      "learning_rate": 5.529215358931553e-05,
      "loss": 1.0111,
      "num_input_tokens_seen": 5986838,
      "step": 2173,
      "train_runtime": 5622.3213,
      "train_tokens_per_second": 1064.834
    },
    {
      "epoch": 0.17469213925550936,
      "grad_norm": 0.3871116638183594,
      "learning_rate": 5.5225375626043406e-05,
      "loss": 1.0913,
      "num_input_tokens_seen": 5988310,
      "step": 2174,
      "train_runtime": 5623.8987,
      "train_tokens_per_second": 1064.797
    },
    {
      "epoch": 0.17477249442536008,
      "grad_norm": 0.39901208877563477,
      "learning_rate": 5.5158597662771286e-05,
      "loss": 1.3411,
      "num_input_tokens_seen": 5990742,
      "step": 2175,
      "train_runtime": 5625.9333,
      "train_tokens_per_second": 1064.844
    },
    {
      "epoch": 0.17485284959521083,
      "grad_norm": 0.2963050305843353,
      "learning_rate": 5.509181969949917e-05,
      "loss": 0.8994,
      "num_input_tokens_seen": 5993482,
      "step": 2176,
      "train_runtime": 5628.0454,
      "train_tokens_per_second": 1064.931
    },
    {
      "epoch": 0.17493320476506158,
      "grad_norm": 0.3584640324115753,
      "learning_rate": 5.5025041736227046e-05,
      "loss": 1.1943,
      "num_input_tokens_seen": 5995374,
      "step": 2177,
      "train_runtime": 5629.8282,
      "train_tokens_per_second": 1064.93
    },
    {
      "epoch": 0.1750135599349123,
      "grad_norm": 0.30105990171432495,
      "learning_rate": 5.495826377295493e-05,
      "loss": 0.9561,
      "num_input_tokens_seen": 5998136,
      "step": 2178,
      "train_runtime": 5631.9653,
      "train_tokens_per_second": 1065.017
    },
    {
      "epoch": 0.17509391510476305,
      "grad_norm": 0.3401104509830475,
      "learning_rate": 5.4891485809682806e-05,
      "loss": 1.2485,
      "num_input_tokens_seen": 6000356,
      "step": 2179,
      "train_runtime": 5633.8603,
      "train_tokens_per_second": 1065.052
    },
    {
      "epoch": 0.1751742702746138,
      "grad_norm": 0.3163185119628906,
      "learning_rate": 5.4824707846410686e-05,
      "loss": 1.0947,
      "num_input_tokens_seen": 6003166,
      "step": 2180,
      "train_runtime": 5636.1367,
      "train_tokens_per_second": 1065.121
    },
    {
      "epoch": 0.17525462544446455,
      "grad_norm": 0.3311319053173065,
      "learning_rate": 5.475792988313857e-05,
      "loss": 1.0749,
      "num_input_tokens_seen": 6005678,
      "step": 2181,
      "train_runtime": 5638.2072,
      "train_tokens_per_second": 1065.175
    },
    {
      "epoch": 0.17533498061431527,
      "grad_norm": 0.30566927790641785,
      "learning_rate": 5.4691151919866446e-05,
      "loss": 1.0406,
      "num_input_tokens_seen": 6009646,
      "step": 2182,
      "train_runtime": 5641.1383,
      "train_tokens_per_second": 1065.325
    },
    {
      "epoch": 0.17541533578416602,
      "grad_norm": 0.2625481188297272,
      "learning_rate": 5.462437395659432e-05,
      "loss": 0.8298,
      "num_input_tokens_seen": 6013836,
      "step": 2183,
      "train_runtime": 5644.0808,
      "train_tokens_per_second": 1065.512
    },
    {
      "epoch": 0.17549569095401676,
      "grad_norm": 0.29297178983688354,
      "learning_rate": 5.4557595993322206e-05,
      "loss": 0.9246,
      "num_input_tokens_seen": 6016636,
      "step": 2184,
      "train_runtime": 5646.2712,
      "train_tokens_per_second": 1065.595
    },
    {
      "epoch": 0.17557604612386749,
      "grad_norm": 0.2942541837692261,
      "learning_rate": 5.4490818030050086e-05,
      "loss": 0.9006,
      "num_input_tokens_seen": 6021172,
      "step": 2185,
      "train_runtime": 5649.4554,
      "train_tokens_per_second": 1065.797
    },
    {
      "epoch": 0.17565640129371823,
      "grad_norm": 0.33823662996292114,
      "learning_rate": 5.442404006677797e-05,
      "loss": 1.0208,
      "num_input_tokens_seen": 6024588,
      "step": 2186,
      "train_runtime": 5651.9796,
      "train_tokens_per_second": 1065.925
    },
    {
      "epoch": 0.17573675646356898,
      "grad_norm": 0.3025861382484436,
      "learning_rate": 5.4357262103505845e-05,
      "loss": 0.7754,
      "num_input_tokens_seen": 6026486,
      "step": 2187,
      "train_runtime": 5653.716,
      "train_tokens_per_second": 1065.934
    },
    {
      "epoch": 0.1758171116334197,
      "grad_norm": 0.34149742126464844,
      "learning_rate": 5.429048414023372e-05,
      "loss": 0.8844,
      "num_input_tokens_seen": 6028020,
      "step": 2188,
      "train_runtime": 5655.3211,
      "train_tokens_per_second": 1065.902
    },
    {
      "epoch": 0.17589746680327045,
      "grad_norm": 0.30632808804512024,
      "learning_rate": 5.4223706176961605e-05,
      "loss": 0.9067,
      "num_input_tokens_seen": 6031950,
      "step": 2189,
      "train_runtime": 5658.2195,
      "train_tokens_per_second": 1066.051
    },
    {
      "epoch": 0.1759778219731212,
      "grad_norm": 0.35990339517593384,
      "learning_rate": 5.4156928213689485e-05,
      "loss": 0.8391,
      "num_input_tokens_seen": 6033782,
      "step": 2190,
      "train_runtime": 5660.0219,
      "train_tokens_per_second": 1066.035
    },
    {
      "epoch": 0.17605817714297192,
      "grad_norm": 0.32285207509994507,
      "learning_rate": 5.409015025041736e-05,
      "loss": 1.0469,
      "num_input_tokens_seen": 6035710,
      "step": 2191,
      "train_runtime": 5663.5833,
      "train_tokens_per_second": 1065.705
    },
    {
      "epoch": 0.17613853231282267,
      "grad_norm": 0.3809260129928589,
      "learning_rate": 5.4023372287145245e-05,
      "loss": 1.1226,
      "num_input_tokens_seen": 6037460,
      "step": 2192,
      "train_runtime": 5665.2713,
      "train_tokens_per_second": 1065.697
    },
    {
      "epoch": 0.17621888748267342,
      "grad_norm": 0.28848832845687866,
      "learning_rate": 5.3956594323873125e-05,
      "loss": 0.912,
      "num_input_tokens_seen": 6040066,
      "step": 2193,
      "train_runtime": 5667.4408,
      "train_tokens_per_second": 1065.748
    },
    {
      "epoch": 0.17629924265252417,
      "grad_norm": 0.2915174961090088,
      "learning_rate": 5.388981636060101e-05,
      "loss": 0.7903,
      "num_input_tokens_seen": 6042672,
      "step": 2194,
      "train_runtime": 5669.5079,
      "train_tokens_per_second": 1065.819
    },
    {
      "epoch": 0.1763795978223749,
      "grad_norm": 0.25477486848831177,
      "learning_rate": 5.3823038397328885e-05,
      "loss": 0.7922,
      "num_input_tokens_seen": 6044656,
      "step": 2195,
      "train_runtime": 5671.3154,
      "train_tokens_per_second": 1065.83
    },
    {
      "epoch": 0.17645995299222564,
      "grad_norm": 0.3283901810646057,
      "learning_rate": 5.375626043405676e-05,
      "loss": 0.9957,
      "num_input_tokens_seen": 6047492,
      "step": 2196,
      "train_runtime": 5673.5133,
      "train_tokens_per_second": 1065.917
    },
    {
      "epoch": 0.1765403081620764,
      "grad_norm": 0.33062899112701416,
      "learning_rate": 5.3689482470784645e-05,
      "loss": 0.9998,
      "num_input_tokens_seen": 6049486,
      "step": 2197,
      "train_runtime": 5675.3365,
      "train_tokens_per_second": 1065.926
    },
    {
      "epoch": 0.1766206633319271,
      "grad_norm": 0.27277809381484985,
      "learning_rate": 5.3622704507512525e-05,
      "loss": 1.0771,
      "num_input_tokens_seen": 6054772,
      "step": 2198,
      "train_runtime": 5678.9557,
      "train_tokens_per_second": 1066.177
    },
    {
      "epoch": 0.17670101850177786,
      "grad_norm": 0.331997811794281,
      "learning_rate": 5.35559265442404e-05,
      "loss": 1.0936,
      "num_input_tokens_seen": 6057470,
      "step": 2199,
      "train_runtime": 5681.2069,
      "train_tokens_per_second": 1066.229
    },
    {
      "epoch": 0.1767813736716286,
      "grad_norm": 0.30002301931381226,
      "learning_rate": 5.3489148580968285e-05,
      "loss": 0.728,
      "num_input_tokens_seen": 6059150,
      "step": 2200,
      "train_runtime": 5682.8509,
      "train_tokens_per_second": 1066.217
    },
    {
      "epoch": 0.17686172884147933,
      "grad_norm": 0.3315964341163635,
      "learning_rate": 5.342237061769616e-05,
      "loss": 0.7822,
      "num_input_tokens_seen": 6065192,
      "step": 2201,
      "train_runtime": 5687.092,
      "train_tokens_per_second": 1066.484
    },
    {
      "epoch": 0.17694208401133008,
      "grad_norm": 0.3553049862384796,
      "learning_rate": 5.3355592654424044e-05,
      "loss": 1.2104,
      "num_input_tokens_seen": 6067588,
      "step": 2202,
      "train_runtime": 5689.2827,
      "train_tokens_per_second": 1066.494
    },
    {
      "epoch": 0.17702243918118082,
      "grad_norm": 0.27361515164375305,
      "learning_rate": 5.3288814691151924e-05,
      "loss": 1.0252,
      "num_input_tokens_seen": 6070078,
      "step": 2203,
      "train_runtime": 5691.3186,
      "train_tokens_per_second": 1066.55
    },
    {
      "epoch": 0.17710279435103157,
      "grad_norm": 0.30246469378471375,
      "learning_rate": 5.32220367278798e-05,
      "loss": 1.1775,
      "num_input_tokens_seen": 6072462,
      "step": 2204,
      "train_runtime": 5693.2606,
      "train_tokens_per_second": 1066.605
    },
    {
      "epoch": 0.1771831495208823,
      "grad_norm": 0.2991829216480255,
      "learning_rate": 5.3155258764607684e-05,
      "loss": 0.9873,
      "num_input_tokens_seen": 6074842,
      "step": 2205,
      "train_runtime": 5695.2927,
      "train_tokens_per_second": 1066.643
    },
    {
      "epoch": 0.17726350469073304,
      "grad_norm": 0.3418027460575104,
      "learning_rate": 5.308848080133556e-05,
      "loss": 0.8657,
      "num_input_tokens_seen": 6076696,
      "step": 2206,
      "train_runtime": 5697.1007,
      "train_tokens_per_second": 1066.63
    },
    {
      "epoch": 0.1773438598605838,
      "grad_norm": 0.3597825765609741,
      "learning_rate": 5.302170283806344e-05,
      "loss": 1.199,
      "num_input_tokens_seen": 6079590,
      "step": 2207,
      "train_runtime": 5699.3843,
      "train_tokens_per_second": 1066.71
    },
    {
      "epoch": 0.1774242150304345,
      "grad_norm": 0.29662024974823,
      "learning_rate": 5.2954924874791324e-05,
      "loss": 0.8954,
      "num_input_tokens_seen": 6083138,
      "step": 2208,
      "train_runtime": 5702.1327,
      "train_tokens_per_second": 1066.818
    },
    {
      "epoch": 0.17750457020028526,
      "grad_norm": 0.3632822036743164,
      "learning_rate": 5.28881469115192e-05,
      "loss": 0.9905,
      "num_input_tokens_seen": 6086604,
      "step": 2209,
      "train_runtime": 5704.691,
      "train_tokens_per_second": 1066.947
    },
    {
      "epoch": 0.177584925370136,
      "grad_norm": 0.4715336263179779,
      "learning_rate": 5.2821368948247084e-05,
      "loss": 0.8444,
      "num_input_tokens_seen": 6089566,
      "step": 2210,
      "train_runtime": 5706.9798,
      "train_tokens_per_second": 1067.038
    },
    {
      "epoch": 0.17766528053998673,
      "grad_norm": 0.3637067973613739,
      "learning_rate": 5.2754590984974964e-05,
      "loss": 1.1719,
      "num_input_tokens_seen": 6091198,
      "step": 2211,
      "train_runtime": 5708.601,
      "train_tokens_per_second": 1067.021
    },
    {
      "epoch": 0.17774563570983748,
      "grad_norm": 0.3065914213657379,
      "learning_rate": 5.268781302170284e-05,
      "loss": 1.0216,
      "num_input_tokens_seen": 6093536,
      "step": 2212,
      "train_runtime": 5710.6367,
      "train_tokens_per_second": 1067.05
    },
    {
      "epoch": 0.17782599087968823,
      "grad_norm": 0.28996747732162476,
      "learning_rate": 5.2621035058430724e-05,
      "loss": 1.1535,
      "num_input_tokens_seen": 6096170,
      "step": 2213,
      "train_runtime": 5712.7073,
      "train_tokens_per_second": 1067.125
    },
    {
      "epoch": 0.17790634604953895,
      "grad_norm": 0.33675292134284973,
      "learning_rate": 5.25542570951586e-05,
      "loss": 0.9864,
      "num_input_tokens_seen": 6098406,
      "step": 2214,
      "train_runtime": 5714.6553,
      "train_tokens_per_second": 1067.152
    },
    {
      "epoch": 0.1779867012193897,
      "grad_norm": 0.35894355177879333,
      "learning_rate": 5.248747913188648e-05,
      "loss": 1.29,
      "num_input_tokens_seen": 6101070,
      "step": 2215,
      "train_runtime": 5716.7987,
      "train_tokens_per_second": 1067.218
    },
    {
      "epoch": 0.17806705638924045,
      "grad_norm": 0.29682108759880066,
      "learning_rate": 5.2420701168614363e-05,
      "loss": 0.9371,
      "num_input_tokens_seen": 6104104,
      "step": 2216,
      "train_runtime": 5719.2471,
      "train_tokens_per_second": 1067.292
    },
    {
      "epoch": 0.1781474115590912,
      "grad_norm": 0.34291523694992065,
      "learning_rate": 5.2353923205342237e-05,
      "loss": 1.1021,
      "num_input_tokens_seen": 6109426,
      "step": 2217,
      "train_runtime": 5722.9345,
      "train_tokens_per_second": 1067.534
    },
    {
      "epoch": 0.17822776672894192,
      "grad_norm": 0.27323198318481445,
      "learning_rate": 5.228714524207012e-05,
      "loss": 1.0301,
      "num_input_tokens_seen": 6112258,
      "step": 2218,
      "train_runtime": 5725.1115,
      "train_tokens_per_second": 1067.623
    },
    {
      "epoch": 0.17830812189879267,
      "grad_norm": 0.32020869851112366,
      "learning_rate": 5.2220367278797996e-05,
      "loss": 1.0896,
      "num_input_tokens_seen": 6113846,
      "step": 2219,
      "train_runtime": 5726.7018,
      "train_tokens_per_second": 1067.603
    },
    {
      "epoch": 0.17838847706864341,
      "grad_norm": 0.352998286485672,
      "learning_rate": 5.2153589315525876e-05,
      "loss": 0.9862,
      "num_input_tokens_seen": 6116594,
      "step": 2220,
      "train_runtime": 5728.9033,
      "train_tokens_per_second": 1067.673
    },
    {
      "epoch": 0.17846883223849414,
      "grad_norm": 0.3459780514240265,
      "learning_rate": 5.208681135225376e-05,
      "loss": 0.9316,
      "num_input_tokens_seen": 6119074,
      "step": 2221,
      "train_runtime": 5733.0229,
      "train_tokens_per_second": 1067.338
    },
    {
      "epoch": 0.17854918740834488,
      "grad_norm": 0.367080420255661,
      "learning_rate": 5.2020033388981636e-05,
      "loss": 0.9275,
      "num_input_tokens_seen": 6120440,
      "step": 2222,
      "train_runtime": 5734.5782,
      "train_tokens_per_second": 1067.287
    },
    {
      "epoch": 0.17862954257819563,
      "grad_norm": 0.2654534876346588,
      "learning_rate": 5.195325542570952e-05,
      "loss": 0.6607,
      "num_input_tokens_seen": 6122078,
      "step": 2223,
      "train_runtime": 5736.2811,
      "train_tokens_per_second": 1067.256
    },
    {
      "epoch": 0.17870989774804635,
      "grad_norm": 0.2552686929702759,
      "learning_rate": 5.18864774624374e-05,
      "loss": 0.8954,
      "num_input_tokens_seen": 6126014,
      "step": 2224,
      "train_runtime": 5739.2236,
      "train_tokens_per_second": 1067.394
    },
    {
      "epoch": 0.1787902529178971,
      "grad_norm": 0.31892409920692444,
      "learning_rate": 5.1819699499165276e-05,
      "loss": 1.163,
      "num_input_tokens_seen": 6129302,
      "step": 2225,
      "train_runtime": 5741.822,
      "train_tokens_per_second": 1067.484
    },
    {
      "epoch": 0.17887060808774785,
      "grad_norm": 0.31341803073883057,
      "learning_rate": 5.175292153589316e-05,
      "loss": 1.2026,
      "num_input_tokens_seen": 6132292,
      "step": 2226,
      "train_runtime": 5744.1113,
      "train_tokens_per_second": 1067.579
    },
    {
      "epoch": 0.17895096325759857,
      "grad_norm": 0.34947773814201355,
      "learning_rate": 5.1686143572621036e-05,
      "loss": 0.9215,
      "num_input_tokens_seen": 6134406,
      "step": 2227,
      "train_runtime": 5746.062,
      "train_tokens_per_second": 1067.584
    },
    {
      "epoch": 0.17903131842744932,
      "grad_norm": 0.2981410622596741,
      "learning_rate": 5.1619365609348916e-05,
      "loss": 0.9967,
      "num_input_tokens_seen": 6136788,
      "step": 2228,
      "train_runtime": 5748.0223,
      "train_tokens_per_second": 1067.635
    },
    {
      "epoch": 0.17911167359730007,
      "grad_norm": 0.322319895029068,
      "learning_rate": 5.15525876460768e-05,
      "loss": 0.9858,
      "num_input_tokens_seen": 6138824,
      "step": 2229,
      "train_runtime": 5749.9562,
      "train_tokens_per_second": 1067.63
    },
    {
      "epoch": 0.17919202876715082,
      "grad_norm": 0.3060888946056366,
      "learning_rate": 5.1485809682804676e-05,
      "loss": 0.9619,
      "num_input_tokens_seen": 6141296,
      "step": 2230,
      "train_runtime": 5752.0445,
      "train_tokens_per_second": 1067.672
    },
    {
      "epoch": 0.17927238393700154,
      "grad_norm": 0.2686266005039215,
      "learning_rate": 5.141903171953256e-05,
      "loss": 0.91,
      "num_input_tokens_seen": 6143836,
      "step": 2231,
      "train_runtime": 5754.1701,
      "train_tokens_per_second": 1067.719
    },
    {
      "epoch": 0.1793527391068523,
      "grad_norm": 0.3219556212425232,
      "learning_rate": 5.1352253756260436e-05,
      "loss": 1.0496,
      "num_input_tokens_seen": 6145532,
      "step": 2232,
      "train_runtime": 5755.8004,
      "train_tokens_per_second": 1067.711
    },
    {
      "epoch": 0.17943309427670304,
      "grad_norm": 0.30096307396888733,
      "learning_rate": 5.1285475792988315e-05,
      "loss": 0.7293,
      "num_input_tokens_seen": 6147114,
      "step": 2233,
      "train_runtime": 5757.4422,
      "train_tokens_per_second": 1067.681
    },
    {
      "epoch": 0.17951344944655376,
      "grad_norm": 0.42000851035118103,
      "learning_rate": 5.12186978297162e-05,
      "loss": 1.2816,
      "num_input_tokens_seen": 6149128,
      "step": 2234,
      "train_runtime": 5759.2989,
      "train_tokens_per_second": 1067.687
    },
    {
      "epoch": 0.1795938046164045,
      "grad_norm": 0.30302050709724426,
      "learning_rate": 5.1151919866444075e-05,
      "loss": 1.0725,
      "num_input_tokens_seen": 6152618,
      "step": 2235,
      "train_runtime": 5761.9761,
      "train_tokens_per_second": 1067.797
    },
    {
      "epoch": 0.17967415978625526,
      "grad_norm": 0.34265613555908203,
      "learning_rate": 5.108514190317195e-05,
      "loss": 1.109,
      "num_input_tokens_seen": 6155284,
      "step": 2236,
      "train_runtime": 5764.1388,
      "train_tokens_per_second": 1067.858
    },
    {
      "epoch": 0.17975451495610598,
      "grad_norm": 0.3333062529563904,
      "learning_rate": 5.1018363939899835e-05,
      "loss": 1.0783,
      "num_input_tokens_seen": 6157294,
      "step": 2237,
      "train_runtime": 5765.9684,
      "train_tokens_per_second": 1067.868
    },
    {
      "epoch": 0.17983487012595673,
      "grad_norm": 0.29393231868743896,
      "learning_rate": 5.0951585976627715e-05,
      "loss": 0.9709,
      "num_input_tokens_seen": 6160650,
      "step": 2238,
      "train_runtime": 5768.4622,
      "train_tokens_per_second": 1067.988
    },
    {
      "epoch": 0.17991522529580747,
      "grad_norm": 0.35577812790870667,
      "learning_rate": 5.08848080133556e-05,
      "loss": 1.0123,
      "num_input_tokens_seen": 6163518,
      "step": 2239,
      "train_runtime": 5770.7793,
      "train_tokens_per_second": 1068.056
    },
    {
      "epoch": 0.17999558046565822,
      "grad_norm": 0.3715101182460785,
      "learning_rate": 5.0818030050083475e-05,
      "loss": 1.0461,
      "num_input_tokens_seen": 6164740,
      "step": 2240,
      "train_runtime": 5772.2902,
      "train_tokens_per_second": 1067.989
    },
    {
      "epoch": 0.18007593563550894,
      "grad_norm": 0.3006369173526764,
      "learning_rate": 5.0751252086811355e-05,
      "loss": 0.9726,
      "num_input_tokens_seen": 6166876,
      "step": 2241,
      "train_runtime": 5774.1876,
      "train_tokens_per_second": 1068.008
    },
    {
      "epoch": 0.1801562908053597,
      "grad_norm": 0.27124038338661194,
      "learning_rate": 5.068447412353924e-05,
      "loss": 1.0184,
      "num_input_tokens_seen": 6168608,
      "step": 2242,
      "train_runtime": 5775.8722,
      "train_tokens_per_second": 1067.996
    },
    {
      "epoch": 0.18023664597521044,
      "grad_norm": 0.33870163559913635,
      "learning_rate": 5.0617696160267115e-05,
      "loss": 1.0748,
      "num_input_tokens_seen": 6171164,
      "step": 2243,
      "train_runtime": 5777.9868,
      "train_tokens_per_second": 1068.047
    },
    {
      "epoch": 0.18031700114506116,
      "grad_norm": 0.3271868824958801,
      "learning_rate": 5.055091819699499e-05,
      "loss": 0.9215,
      "num_input_tokens_seen": 6173436,
      "step": 2244,
      "train_runtime": 5779.9888,
      "train_tokens_per_second": 1068.071
    },
    {
      "epoch": 0.1803973563149119,
      "grad_norm": 0.3185940682888031,
      "learning_rate": 5.0484140233722875e-05,
      "loss": 1.1283,
      "num_input_tokens_seen": 6175948,
      "step": 2245,
      "train_runtime": 5782.0879,
      "train_tokens_per_second": 1068.117
    },
    {
      "epoch": 0.18047771148476266,
      "grad_norm": 0.3163585364818573,
      "learning_rate": 5.0417362270450755e-05,
      "loss": 0.9332,
      "num_input_tokens_seen": 6178670,
      "step": 2246,
      "train_runtime": 5784.2214,
      "train_tokens_per_second": 1068.194
    },
    {
      "epoch": 0.18055806665461338,
      "grad_norm": 0.3045605719089508,
      "learning_rate": 5.035058430717864e-05,
      "loss": 1.5736,
      "num_input_tokens_seen": 6182328,
      "step": 2247,
      "train_runtime": 5786.8281,
      "train_tokens_per_second": 1068.345
    },
    {
      "epoch": 0.18063842182446413,
      "grad_norm": 0.28818830847740173,
      "learning_rate": 5.0283806343906514e-05,
      "loss": 1.2267,
      "num_input_tokens_seen": 6184828,
      "step": 2248,
      "train_runtime": 5788.8769,
      "train_tokens_per_second": 1068.399
    },
    {
      "epoch": 0.18071877699431488,
      "grad_norm": 0.35958895087242126,
      "learning_rate": 5.021702838063439e-05,
      "loss": 0.7491,
      "num_input_tokens_seen": 6186976,
      "step": 2249,
      "train_runtime": 5790.8594,
      "train_tokens_per_second": 1068.404
    },
    {
      "epoch": 0.1807991321641656,
      "grad_norm": 0.2572154402732849,
      "learning_rate": 5.0150250417362274e-05,
      "loss": 0.8362,
      "num_input_tokens_seen": 6188974,
      "step": 2250,
      "train_runtime": 5792.6095,
      "train_tokens_per_second": 1068.426
    },
    {
      "epoch": 0.18087948733401635,
      "grad_norm": 0.2994706630706787,
      "learning_rate": 5.0083472454090154e-05,
      "loss": 1.1099,
      "num_input_tokens_seen": 6193522,
      "step": 2251,
      "train_runtime": 5797.717,
      "train_tokens_per_second": 1068.269
    },
    {
      "epoch": 0.1809598425038671,
      "grad_norm": 0.2824646234512329,
      "learning_rate": 5.001669449081803e-05,
      "loss": 0.7092,
      "num_input_tokens_seen": 6195504,
      "step": 2252,
      "train_runtime": 5799.4813,
      "train_tokens_per_second": 1068.286
    },
    {
      "epoch": 0.18104019767371785,
      "grad_norm": 0.32443007826805115,
      "learning_rate": 4.9949916527545914e-05,
      "loss": 0.8126,
      "num_input_tokens_seen": 6197450,
      "step": 2253,
      "train_runtime": 5801.2977,
      "train_tokens_per_second": 1068.287
    },
    {
      "epoch": 0.18112055284356857,
      "grad_norm": 0.33935296535491943,
      "learning_rate": 4.988313856427379e-05,
      "loss": 0.5973,
      "num_input_tokens_seen": 6200882,
      "step": 2254,
      "train_runtime": 5803.9293,
      "train_tokens_per_second": 1068.394
    },
    {
      "epoch": 0.18120090801341932,
      "grad_norm": 0.3202540874481201,
      "learning_rate": 4.9816360601001674e-05,
      "loss": 0.882,
      "num_input_tokens_seen": 6203588,
      "step": 2255,
      "train_runtime": 5806.1281,
      "train_tokens_per_second": 1068.455
    },
    {
      "epoch": 0.18128126318327006,
      "grad_norm": 0.3738347589969635,
      "learning_rate": 4.9749582637729554e-05,
      "loss": 1.0906,
      "num_input_tokens_seen": 6204936,
      "step": 2256,
      "train_runtime": 5807.6338,
      "train_tokens_per_second": 1068.41
    },
    {
      "epoch": 0.18136161835312078,
      "grad_norm": 0.3130134642124176,
      "learning_rate": 4.9682804674457434e-05,
      "loss": 0.9808,
      "num_input_tokens_seen": 6207264,
      "step": 2257,
      "train_runtime": 5809.6093,
      "train_tokens_per_second": 1068.448
    },
    {
      "epoch": 0.18144197352297153,
      "grad_norm": 0.35142961144447327,
      "learning_rate": 4.961602671118531e-05,
      "loss": 0.8468,
      "num_input_tokens_seen": 6208718,
      "step": 2258,
      "train_runtime": 5811.2405,
      "train_tokens_per_second": 1068.398
    },
    {
      "epoch": 0.18152232869282228,
      "grad_norm": 0.3048270344734192,
      "learning_rate": 4.9549248747913194e-05,
      "loss": 0.6867,
      "num_input_tokens_seen": 6210984,
      "step": 2259,
      "train_runtime": 5813.1614,
      "train_tokens_per_second": 1068.435
    },
    {
      "epoch": 0.181602683862673,
      "grad_norm": 0.2752719521522522,
      "learning_rate": 4.9482470784641074e-05,
      "loss": 0.9185,
      "num_input_tokens_seen": 6213820,
      "step": 2260,
      "train_runtime": 5815.355,
      "train_tokens_per_second": 1068.519
    },
    {
      "epoch": 0.18168303903252375,
      "grad_norm": 0.40631142258644104,
      "learning_rate": 4.9415692821368953e-05,
      "loss": 0.8205,
      "num_input_tokens_seen": 6216630,
      "step": 2261,
      "train_runtime": 5817.5943,
      "train_tokens_per_second": 1068.591
    },
    {
      "epoch": 0.1817633942023745,
      "grad_norm": 0.3239935040473938,
      "learning_rate": 4.934891485809683e-05,
      "loss": 1.1012,
      "num_input_tokens_seen": 6218662,
      "step": 2262,
      "train_runtime": 5819.5233,
      "train_tokens_per_second": 1068.586
    },
    {
      "epoch": 0.18184374937222522,
      "grad_norm": 0.39451807737350464,
      "learning_rate": 4.9282136894824707e-05,
      "loss": 0.7311,
      "num_input_tokens_seen": 6221182,
      "step": 2263,
      "train_runtime": 5821.5694,
      "train_tokens_per_second": 1068.643
    },
    {
      "epoch": 0.18192410454207597,
      "grad_norm": 0.36721745133399963,
      "learning_rate": 4.921535893155259e-05,
      "loss": 0.7233,
      "num_input_tokens_seen": 6223126,
      "step": 2264,
      "train_runtime": 5823.3943,
      "train_tokens_per_second": 1068.642
    },
    {
      "epoch": 0.18200445971192672,
      "grad_norm": 0.24146652221679688,
      "learning_rate": 4.914858096828047e-05,
      "loss": 0.7028,
      "num_input_tokens_seen": 6226154,
      "step": 2265,
      "train_runtime": 5825.7432,
      "train_tokens_per_second": 1068.731
    },
    {
      "epoch": 0.18208481488177747,
      "grad_norm": 0.322639524936676,
      "learning_rate": 4.9081803005008346e-05,
      "loss": 0.7768,
      "num_input_tokens_seen": 6229964,
      "step": 2266,
      "train_runtime": 5828.5276,
      "train_tokens_per_second": 1068.874
    },
    {
      "epoch": 0.1821651700516282,
      "grad_norm": 0.32387587428092957,
      "learning_rate": 4.9015025041736226e-05,
      "loss": 0.917,
      "num_input_tokens_seen": 6232198,
      "step": 2267,
      "train_runtime": 5830.5358,
      "train_tokens_per_second": 1068.889
    },
    {
      "epoch": 0.18224552522147894,
      "grad_norm": 0.37525859475135803,
      "learning_rate": 4.894824707846411e-05,
      "loss": 0.8793,
      "num_input_tokens_seen": 6234590,
      "step": 2268,
      "train_runtime": 5832.5035,
      "train_tokens_per_second": 1068.939
    },
    {
      "epoch": 0.1823258803913297,
      "grad_norm": 0.28123602271080017,
      "learning_rate": 4.888146911519199e-05,
      "loss": 0.8465,
      "num_input_tokens_seen": 6236906,
      "step": 2269,
      "train_runtime": 5834.4655,
      "train_tokens_per_second": 1068.976
    },
    {
      "epoch": 0.1824062355611804,
      "grad_norm": 0.3211343586444855,
      "learning_rate": 4.8814691151919866e-05,
      "loss": 0.7852,
      "num_input_tokens_seen": 6243846,
      "step": 2270,
      "train_runtime": 5839.2002,
      "train_tokens_per_second": 1069.298
    },
    {
      "epoch": 0.18248659073103116,
      "grad_norm": 0.390897661447525,
      "learning_rate": 4.8747913188647746e-05,
      "loss": 0.7559,
      "num_input_tokens_seen": 6245240,
      "step": 2271,
      "train_runtime": 5840.7348,
      "train_tokens_per_second": 1069.256
    },
    {
      "epoch": 0.1825669459008819,
      "grad_norm": 0.3211084008216858,
      "learning_rate": 4.8681135225375626e-05,
      "loss": 0.8246,
      "num_input_tokens_seen": 6247572,
      "step": 2272,
      "train_runtime": 5842.6495,
      "train_tokens_per_second": 1069.305
    },
    {
      "epoch": 0.18264730107073263,
      "grad_norm": 0.3849484622478485,
      "learning_rate": 4.861435726210351e-05,
      "loss": 1.0931,
      "num_input_tokens_seen": 6249328,
      "step": 2273,
      "train_runtime": 5844.3815,
      "train_tokens_per_second": 1069.288
    },
    {
      "epoch": 0.18272765624058337,
      "grad_norm": 0.32431185245513916,
      "learning_rate": 4.8547579298831386e-05,
      "loss": 1.034,
      "num_input_tokens_seen": 6253152,
      "step": 2274,
      "train_runtime": 5847.1441,
      "train_tokens_per_second": 1069.437
    },
    {
      "epoch": 0.18280801141043412,
      "grad_norm": 0.3326425552368164,
      "learning_rate": 4.8480801335559266e-05,
      "loss": 0.9919,
      "num_input_tokens_seen": 6256082,
      "step": 2275,
      "train_runtime": 5849.4119,
      "train_tokens_per_second": 1069.523
    },
    {
      "epoch": 0.18288836658028487,
      "grad_norm": 0.29046377539634705,
      "learning_rate": 4.8414023372287146e-05,
      "loss": 0.8487,
      "num_input_tokens_seen": 6258048,
      "step": 2276,
      "train_runtime": 5851.2065,
      "train_tokens_per_second": 1069.531
    },
    {
      "epoch": 0.1829687217501356,
      "grad_norm": 0.4438607096672058,
      "learning_rate": 4.834724540901503e-05,
      "loss": 0.9298,
      "num_input_tokens_seen": 6259816,
      "step": 2277,
      "train_runtime": 5852.9488,
      "train_tokens_per_second": 1069.515
    },
    {
      "epoch": 0.18304907691998634,
      "grad_norm": 0.3096592426300049,
      "learning_rate": 4.8280467445742906e-05,
      "loss": 0.8273,
      "num_input_tokens_seen": 6262084,
      "step": 2278,
      "train_runtime": 5854.9615,
      "train_tokens_per_second": 1069.535
    },
    {
      "epoch": 0.1831294320898371,
      "grad_norm": 0.3317062556743622,
      "learning_rate": 4.8213689482470785e-05,
      "loss": 1.1396,
      "num_input_tokens_seen": 6263846,
      "step": 2279,
      "train_runtime": 5856.7215,
      "train_tokens_per_second": 1069.514
    },
    {
      "epoch": 0.1832097872596878,
      "grad_norm": 0.3365456461906433,
      "learning_rate": 4.8146911519198665e-05,
      "loss": 1.1601,
      "num_input_tokens_seen": 6265866,
      "step": 2280,
      "train_runtime": 5858.6592,
      "train_tokens_per_second": 1069.505
    },
    {
      "epoch": 0.18329014242953856,
      "grad_norm": 0.32816290855407715,
      "learning_rate": 4.8080133555926545e-05,
      "loss": 0.9689,
      "num_input_tokens_seen": 6270664,
      "step": 2281,
      "train_runtime": 5863.6826,
      "train_tokens_per_second": 1069.407
    },
    {
      "epoch": 0.1833704975993893,
      "grad_norm": 0.3837275207042694,
      "learning_rate": 4.8013355592654425e-05,
      "loss": 1.3614,
      "num_input_tokens_seen": 6275480,
      "step": 2282,
      "train_runtime": 5867.0544,
      "train_tokens_per_second": 1069.613
    },
    {
      "epoch": 0.18345085276924003,
      "grad_norm": 0.3201889395713806,
      "learning_rate": 4.7946577629382305e-05,
      "loss": 0.8872,
      "num_input_tokens_seen": 6277180,
      "step": 2283,
      "train_runtime": 5868.6885,
      "train_tokens_per_second": 1069.605
    },
    {
      "epoch": 0.18353120793909078,
      "grad_norm": 0.26214835047721863,
      "learning_rate": 4.7879799666110185e-05,
      "loss": 0.6135,
      "num_input_tokens_seen": 6280390,
      "step": 2284,
      "train_runtime": 5871.0665,
      "train_tokens_per_second": 1069.719
    },
    {
      "epoch": 0.18361156310894153,
      "grad_norm": 0.3387789726257324,
      "learning_rate": 4.7813021702838065e-05,
      "loss": 0.9593,
      "num_input_tokens_seen": 6285522,
      "step": 2285,
      "train_runtime": 5874.683,
      "train_tokens_per_second": 1069.934
    },
    {
      "epoch": 0.18369191827879225,
      "grad_norm": 0.4115189015865326,
      "learning_rate": 4.7746243739565945e-05,
      "loss": 0.9443,
      "num_input_tokens_seen": 6287382,
      "step": 2286,
      "train_runtime": 5876.4711,
      "train_tokens_per_second": 1069.925
    },
    {
      "epoch": 0.183772273448643,
      "grad_norm": 0.2629833519458771,
      "learning_rate": 4.7679465776293825e-05,
      "loss": 1.1946,
      "num_input_tokens_seen": 6290098,
      "step": 2287,
      "train_runtime": 5878.5678,
      "train_tokens_per_second": 1070.005
    },
    {
      "epoch": 0.18385262861849375,
      "grad_norm": 0.3172478973865509,
      "learning_rate": 4.7612687813021705e-05,
      "loss": 1.0908,
      "num_input_tokens_seen": 6292668,
      "step": 2288,
      "train_runtime": 5880.6891,
      "train_tokens_per_second": 1070.056
    },
    {
      "epoch": 0.1839329837883445,
      "grad_norm": 0.3685589134693146,
      "learning_rate": 4.7545909849749585e-05,
      "loss": 0.9134,
      "num_input_tokens_seen": 6294796,
      "step": 2289,
      "train_runtime": 5882.6749,
      "train_tokens_per_second": 1070.057
    },
    {
      "epoch": 0.18401333895819522,
      "grad_norm": 0.3100009560585022,
      "learning_rate": 4.7479131886477465e-05,
      "loss": 0.8958,
      "num_input_tokens_seen": 6297902,
      "step": 2290,
      "train_runtime": 5885.0113,
      "train_tokens_per_second": 1070.16
    },
    {
      "epoch": 0.18409369412804596,
      "grad_norm": 0.34115126729011536,
      "learning_rate": 4.7412353923205345e-05,
      "loss": 1.0444,
      "num_input_tokens_seen": 6300448,
      "step": 2291,
      "train_runtime": 5887.0616,
      "train_tokens_per_second": 1070.219
    },
    {
      "epoch": 0.1841740492978967,
      "grad_norm": 0.3575955927371979,
      "learning_rate": 4.7345575959933225e-05,
      "loss": 1.0956,
      "num_input_tokens_seen": 6301970,
      "step": 2292,
      "train_runtime": 5888.6741,
      "train_tokens_per_second": 1070.185
    },
    {
      "epoch": 0.18425440446774743,
      "grad_norm": 0.3460978865623474,
      "learning_rate": 4.7278797996661104e-05,
      "loss": 1.1619,
      "num_input_tokens_seen": 6306278,
      "step": 2293,
      "train_runtime": 5891.7978,
      "train_tokens_per_second": 1070.349
    },
    {
      "epoch": 0.18433475963759818,
      "grad_norm": 0.3807061016559601,
      "learning_rate": 4.7212020033388984e-05,
      "loss": 1.2679,
      "num_input_tokens_seen": 6308894,
      "step": 2294,
      "train_runtime": 5893.8825,
      "train_tokens_per_second": 1070.414
    },
    {
      "epoch": 0.18441511480744893,
      "grad_norm": 0.3346267640590668,
      "learning_rate": 4.7145242070116864e-05,
      "loss": 0.9922,
      "num_input_tokens_seen": 6310684,
      "step": 2295,
      "train_runtime": 5895.5551,
      "train_tokens_per_second": 1070.414
    },
    {
      "epoch": 0.18449546997729965,
      "grad_norm": 0.2876211404800415,
      "learning_rate": 4.7078464106844744e-05,
      "loss": 1.1814,
      "num_input_tokens_seen": 6313498,
      "step": 2296,
      "train_runtime": 5897.8163,
      "train_tokens_per_second": 1070.481
    },
    {
      "epoch": 0.1845758251471504,
      "grad_norm": 0.2939240634441376,
      "learning_rate": 4.7011686143572624e-05,
      "loss": 0.8314,
      "num_input_tokens_seen": 6315472,
      "step": 2297,
      "train_runtime": 5899.6443,
      "train_tokens_per_second": 1070.484
    },
    {
      "epoch": 0.18465618031700115,
      "grad_norm": 0.28995025157928467,
      "learning_rate": 4.6944908180300504e-05,
      "loss": 1.0929,
      "num_input_tokens_seen": 6318302,
      "step": 2298,
      "train_runtime": 5901.9067,
      "train_tokens_per_second": 1070.553
    },
    {
      "epoch": 0.18473653548685187,
      "grad_norm": 0.3637752830982208,
      "learning_rate": 4.6878130217028384e-05,
      "loss": 0.8314,
      "num_input_tokens_seen": 6319344,
      "step": 2299,
      "train_runtime": 5903.3586,
      "train_tokens_per_second": 1070.466
    },
    {
      "epoch": 0.18481689065670262,
      "grad_norm": 0.2879713475704193,
      "learning_rate": 4.6811352253756264e-05,
      "loss": 0.6484,
      "num_input_tokens_seen": 6323512,
      "step": 2300,
      "train_runtime": 5906.3927,
      "train_tokens_per_second": 1070.622
    },
    {
      "epoch": 0.18489724582655337,
      "grad_norm": 0.28264346718788147,
      "learning_rate": 4.6744574290484144e-05,
      "loss": 0.9958,
      "num_input_tokens_seen": 6325804,
      "step": 2301,
      "train_runtime": 5908.2474,
      "train_tokens_per_second": 1070.674
    },
    {
      "epoch": 0.18497760099640412,
      "grad_norm": 0.3375772535800934,
      "learning_rate": 4.667779632721202e-05,
      "loss": 0.8864,
      "num_input_tokens_seen": 6327514,
      "step": 2302,
      "train_runtime": 5909.975,
      "train_tokens_per_second": 1070.65
    },
    {
      "epoch": 0.18505795616625484,
      "grad_norm": 0.33063581585884094,
      "learning_rate": 4.6611018363939904e-05,
      "loss": 0.9557,
      "num_input_tokens_seen": 6329634,
      "step": 2303,
      "train_runtime": 5911.8692,
      "train_tokens_per_second": 1070.665
    },
    {
      "epoch": 0.1851383113361056,
      "grad_norm": 0.32154470682144165,
      "learning_rate": 4.6544240400667784e-05,
      "loss": 1.2514,
      "num_input_tokens_seen": 6332200,
      "step": 2304,
      "train_runtime": 5913.9963,
      "train_tokens_per_second": 1070.714
    },
    {
      "epoch": 0.18521866650595634,
      "grad_norm": 0.3089728355407715,
      "learning_rate": 4.6477462437395664e-05,
      "loss": 0.9419,
      "num_input_tokens_seen": 6334566,
      "step": 2305,
      "train_runtime": 5916.0117,
      "train_tokens_per_second": 1070.749
    },
    {
      "epoch": 0.18529902167580706,
      "grad_norm": 0.3115226626396179,
      "learning_rate": 4.641068447412354e-05,
      "loss": 1.2372,
      "num_input_tokens_seen": 6336808,
      "step": 2306,
      "train_runtime": 5917.9917,
      "train_tokens_per_second": 1070.77
    },
    {
      "epoch": 0.1853793768456578,
      "grad_norm": 0.27321842312812805,
      "learning_rate": 4.6343906510851423e-05,
      "loss": 1.0608,
      "num_input_tokens_seen": 6338128,
      "step": 2307,
      "train_runtime": 5919.5686,
      "train_tokens_per_second": 1070.708
    },
    {
      "epoch": 0.18545973201550855,
      "grad_norm": 0.2833107113838196,
      "learning_rate": 4.6277128547579303e-05,
      "loss": 0.7515,
      "num_input_tokens_seen": 6340670,
      "step": 2308,
      "train_runtime": 5921.6334,
      "train_tokens_per_second": 1070.764
    },
    {
      "epoch": 0.18554008718535928,
      "grad_norm": 0.3309327960014343,
      "learning_rate": 4.621035058430718e-05,
      "loss": 1.0604,
      "num_input_tokens_seen": 6342012,
      "step": 2309,
      "train_runtime": 5923.1966,
      "train_tokens_per_second": 1070.708
    },
    {
      "epoch": 0.18562044235521002,
      "grad_norm": 0.33111390471458435,
      "learning_rate": 4.6143572621035056e-05,
      "loss": 1.1106,
      "num_input_tokens_seen": 6344012,
      "step": 2310,
      "train_runtime": 5924.9883,
      "train_tokens_per_second": 1070.721
    },
    {
      "epoch": 0.18570079752506077,
      "grad_norm": 0.3710658550262451,
      "learning_rate": 4.6076794657762936e-05,
      "loss": 0.8322,
      "num_input_tokens_seen": 6346142,
      "step": 2311,
      "train_runtime": 5928.5952,
      "train_tokens_per_second": 1070.429
    },
    {
      "epoch": 0.18578115269491152,
      "grad_norm": 0.42098838090896606,
      "learning_rate": 4.601001669449082e-05,
      "loss": 0.9208,
      "num_input_tokens_seen": 6349276,
      "step": 2312,
      "train_runtime": 5931.0157,
      "train_tokens_per_second": 1070.521
    },
    {
      "epoch": 0.18586150786476224,
      "grad_norm": 0.379133939743042,
      "learning_rate": 4.59432387312187e-05,
      "loss": 1.2124,
      "num_input_tokens_seen": 6351484,
      "step": 2313,
      "train_runtime": 5933.056,
      "train_tokens_per_second": 1070.525
    },
    {
      "epoch": 0.185941863034613,
      "grad_norm": 0.2824403941631317,
      "learning_rate": 4.5876460767946576e-05,
      "loss": 0.8327,
      "num_input_tokens_seen": 6353198,
      "step": 2314,
      "train_runtime": 5934.7507,
      "train_tokens_per_second": 1070.508
    },
    {
      "epoch": 0.18602221820446374,
      "grad_norm": 0.2821195423603058,
      "learning_rate": 4.5809682804674456e-05,
      "loss": 0.7801,
      "num_input_tokens_seen": 6355340,
      "step": 2315,
      "train_runtime": 5936.6381,
      "train_tokens_per_second": 1070.528
    },
    {
      "epoch": 0.18610257337431446,
      "grad_norm": 0.3796484172344208,
      "learning_rate": 4.574290484140234e-05,
      "loss": 1.2639,
      "num_input_tokens_seen": 6357318,
      "step": 2316,
      "train_runtime": 5938.468,
      "train_tokens_per_second": 1070.532
    },
    {
      "epoch": 0.1861829285441652,
      "grad_norm": 0.29804062843322754,
      "learning_rate": 4.567612687813022e-05,
      "loss": 0.801,
      "num_input_tokens_seen": 6362200,
      "step": 2317,
      "train_runtime": 5941.8702,
      "train_tokens_per_second": 1070.74
    },
    {
      "epoch": 0.18626328371401596,
      "grad_norm": 0.34569695591926575,
      "learning_rate": 4.5609348914858096e-05,
      "loss": 1.0351,
      "num_input_tokens_seen": 6367904,
      "step": 2318,
      "train_runtime": 5945.7448,
      "train_tokens_per_second": 1071.002
    },
    {
      "epoch": 0.18634363888386668,
      "grad_norm": 0.3258000612258911,
      "learning_rate": 4.5542570951585976e-05,
      "loss": 0.8478,
      "num_input_tokens_seen": 6372234,
      "step": 2319,
      "train_runtime": 5948.8619,
      "train_tokens_per_second": 1071.169
    },
    {
      "epoch": 0.18642399405371743,
      "grad_norm": 0.5333878993988037,
      "learning_rate": 4.5475792988313856e-05,
      "loss": 0.7371,
      "num_input_tokens_seen": 6374028,
      "step": 2320,
      "train_runtime": 5950.5584,
      "train_tokens_per_second": 1071.165
    },
    {
      "epoch": 0.18650434922356818,
      "grad_norm": 0.34493404626846313,
      "learning_rate": 4.540901502504174e-05,
      "loss": 1.025,
      "num_input_tokens_seen": 6376504,
      "step": 2321,
      "train_runtime": 5952.5597,
      "train_tokens_per_second": 1071.221
    },
    {
      "epoch": 0.1865847043934189,
      "grad_norm": 0.35565921664237976,
      "learning_rate": 4.5342237061769616e-05,
      "loss": 0.8247,
      "num_input_tokens_seen": 6378186,
      "step": 2322,
      "train_runtime": 5954.2389,
      "train_tokens_per_second": 1071.201
    },
    {
      "epoch": 0.18666505956326965,
      "grad_norm": 0.2732880413532257,
      "learning_rate": 4.5275459098497496e-05,
      "loss": 1.4874,
      "num_input_tokens_seen": 6382574,
      "step": 2323,
      "train_runtime": 5957.5968,
      "train_tokens_per_second": 1071.334
    },
    {
      "epoch": 0.1867454147331204,
      "grad_norm": 0.28319212794303894,
      "learning_rate": 4.5208681135225376e-05,
      "loss": 0.9535,
      "num_input_tokens_seen": 6386954,
      "step": 2324,
      "train_runtime": 5960.7084,
      "train_tokens_per_second": 1071.509
    },
    {
      "epoch": 0.18682576990297114,
      "grad_norm": 0.3138498067855835,
      "learning_rate": 4.514190317195326e-05,
      "loss": 0.9758,
      "num_input_tokens_seen": 6389560,
      "step": 2325,
      "train_runtime": 5962.8551,
      "train_tokens_per_second": 1071.561
    },
    {
      "epoch": 0.18690612507282187,
      "grad_norm": 0.3402535915374756,
      "learning_rate": 4.5075125208681135e-05,
      "loss": 1.0171,
      "num_input_tokens_seen": 6392596,
      "step": 2326,
      "train_runtime": 5965.1924,
      "train_tokens_per_second": 1071.65
    },
    {
      "epoch": 0.18698648024267261,
      "grad_norm": 0.3673248887062073,
      "learning_rate": 4.5008347245409015e-05,
      "loss": 1.116,
      "num_input_tokens_seen": 6394160,
      "step": 2327,
      "train_runtime": 5966.868,
      "train_tokens_per_second": 1071.611
    },
    {
      "epoch": 0.18706683541252336,
      "grad_norm": 0.3493563234806061,
      "learning_rate": 4.4941569282136895e-05,
      "loss": 1.0953,
      "num_input_tokens_seen": 6397894,
      "step": 2328,
      "train_runtime": 5969.5821,
      "train_tokens_per_second": 1071.749
    },
    {
      "epoch": 0.18714719058237408,
      "grad_norm": 0.356676310300827,
      "learning_rate": 4.4874791318864775e-05,
      "loss": 1.0006,
      "num_input_tokens_seen": 6401426,
      "step": 2329,
      "train_runtime": 5972.2754,
      "train_tokens_per_second": 1071.857
    },
    {
      "epoch": 0.18722754575222483,
      "grad_norm": 0.34382879734039307,
      "learning_rate": 4.4808013355592655e-05,
      "loss": 1.2569,
      "num_input_tokens_seen": 6403920,
      "step": 2330,
      "train_runtime": 5974.3,
      "train_tokens_per_second": 1071.911
    },
    {
      "epoch": 0.18730790092207558,
      "grad_norm": 0.271439790725708,
      "learning_rate": 4.4741235392320535e-05,
      "loss": 1.0944,
      "num_input_tokens_seen": 6406330,
      "step": 2331,
      "train_runtime": 5976.2713,
      "train_tokens_per_second": 1071.961
    },
    {
      "epoch": 0.1873882560919263,
      "grad_norm": 0.3470132648944855,
      "learning_rate": 4.4674457429048415e-05,
      "loss": 1.2693,
      "num_input_tokens_seen": 6408936,
      "step": 2332,
      "train_runtime": 5978.475,
      "train_tokens_per_second": 1072.002
    },
    {
      "epoch": 0.18746861126177705,
      "grad_norm": 0.340411514043808,
      "learning_rate": 4.4607679465776295e-05,
      "loss": 0.99,
      "num_input_tokens_seen": 6410980,
      "step": 2333,
      "train_runtime": 5980.3985,
      "train_tokens_per_second": 1071.999
    },
    {
      "epoch": 0.1875489664316278,
      "grad_norm": 0.31622838973999023,
      "learning_rate": 4.4540901502504175e-05,
      "loss": 0.9647,
      "num_input_tokens_seen": 6414012,
      "step": 2334,
      "train_runtime": 5982.7868,
      "train_tokens_per_second": 1072.078
    },
    {
      "epoch": 0.18762932160147852,
      "grad_norm": 0.3553694188594818,
      "learning_rate": 4.4474123539232055e-05,
      "loss": 1.0953,
      "num_input_tokens_seen": 6417172,
      "step": 2335,
      "train_runtime": 5985.1355,
      "train_tokens_per_second": 1072.185
    },
    {
      "epoch": 0.18770967677132927,
      "grad_norm": 0.320942223072052,
      "learning_rate": 4.4407345575959935e-05,
      "loss": 0.7614,
      "num_input_tokens_seen": 6418362,
      "step": 2336,
      "train_runtime": 5986.6348,
      "train_tokens_per_second": 1072.115
    },
    {
      "epoch": 0.18779003194118002,
      "grad_norm": 0.3402732014656067,
      "learning_rate": 4.4340567612687815e-05,
      "loss": 0.9423,
      "num_input_tokens_seen": 6420638,
      "step": 2337,
      "train_runtime": 5988.5982,
      "train_tokens_per_second": 1072.144
    },
    {
      "epoch": 0.18787038711103077,
      "grad_norm": 0.29002895951271057,
      "learning_rate": 4.4273789649415695e-05,
      "loss": 1.2723,
      "num_input_tokens_seen": 6424186,
      "step": 2338,
      "train_runtime": 5991.2178,
      "train_tokens_per_second": 1072.267
    },
    {
      "epoch": 0.1879507422808815,
      "grad_norm": 0.3135744631290436,
      "learning_rate": 4.4207011686143574e-05,
      "loss": 0.7233,
      "num_input_tokens_seen": 6426764,
      "step": 2339,
      "train_runtime": 5993.3702,
      "train_tokens_per_second": 1072.312
    },
    {
      "epoch": 0.18803109745073224,
      "grad_norm": 0.32115882635116577,
      "learning_rate": 4.4140233722871454e-05,
      "loss": 1.0044,
      "num_input_tokens_seen": 6429174,
      "step": 2340,
      "train_runtime": 5995.4218,
      "train_tokens_per_second": 1072.347
    },
    {
      "epoch": 0.188111452620583,
      "grad_norm": 0.29237309098243713,
      "learning_rate": 4.4073455759599334e-05,
      "loss": 0.8649,
      "num_input_tokens_seen": 6431370,
      "step": 2341,
      "train_runtime": 5999.0424,
      "train_tokens_per_second": 1072.066
    },
    {
      "epoch": 0.1881918077904337,
      "grad_norm": 0.29760250449180603,
      "learning_rate": 4.4006677796327214e-05,
      "loss": 0.8849,
      "num_input_tokens_seen": 6434130,
      "step": 2342,
      "train_runtime": 6001.1684,
      "train_tokens_per_second": 1072.146
    },
    {
      "epoch": 0.18827216296028446,
      "grad_norm": 0.31768155097961426,
      "learning_rate": 4.3939899833055094e-05,
      "loss": 1.2445,
      "num_input_tokens_seen": 6437588,
      "step": 2343,
      "train_runtime": 6003.7344,
      "train_tokens_per_second": 1072.264
    },
    {
      "epoch": 0.1883525181301352,
      "grad_norm": 0.2507043778896332,
      "learning_rate": 4.3873121869782974e-05,
      "loss": 0.5031,
      "num_input_tokens_seen": 6445040,
      "step": 2344,
      "train_runtime": 6008.7274,
      "train_tokens_per_second": 1072.613
    },
    {
      "epoch": 0.18843287329998593,
      "grad_norm": 0.2699557840824127,
      "learning_rate": 4.3806343906510854e-05,
      "loss": 0.6923,
      "num_input_tokens_seen": 6446724,
      "step": 2345,
      "train_runtime": 6010.3856,
      "train_tokens_per_second": 1072.597
    },
    {
      "epoch": 0.18851322846983667,
      "grad_norm": 0.3183298707008362,
      "learning_rate": 4.373956594323873e-05,
      "loss": 0.4887,
      "num_input_tokens_seen": 6449376,
      "step": 2346,
      "train_runtime": 6012.5392,
      "train_tokens_per_second": 1072.654
    },
    {
      "epoch": 0.18859358363968742,
      "grad_norm": 0.2718125581741333,
      "learning_rate": 4.3672787979966614e-05,
      "loss": 0.8204,
      "num_input_tokens_seen": 6451784,
      "step": 2347,
      "train_runtime": 6014.5947,
      "train_tokens_per_second": 1072.688
    },
    {
      "epoch": 0.18867393880953817,
      "grad_norm": 0.3327609896659851,
      "learning_rate": 4.3606010016694494e-05,
      "loss": 0.9818,
      "num_input_tokens_seen": 6454174,
      "step": 2348,
      "train_runtime": 6016.5705,
      "train_tokens_per_second": 1072.733
    },
    {
      "epoch": 0.1887542939793889,
      "grad_norm": 0.2682245373725891,
      "learning_rate": 4.3539232053422374e-05,
      "loss": 0.764,
      "num_input_tokens_seen": 6457878,
      "step": 2349,
      "train_runtime": 6019.3973,
      "train_tokens_per_second": 1072.845
    },
    {
      "epoch": 0.18883464914923964,
      "grad_norm": 0.27816519141197205,
      "learning_rate": 4.3472454090150254e-05,
      "loss": 0.6938,
      "num_input_tokens_seen": 6460140,
      "step": 2350,
      "train_runtime": 6021.3789,
      "train_tokens_per_second": 1072.867
    },
    {
      "epoch": 0.1889150043190904,
      "grad_norm": 0.40029385685920715,
      "learning_rate": 4.3405676126878134e-05,
      "loss": 1.0792,
      "num_input_tokens_seen": 6462378,
      "step": 2351,
      "train_runtime": 6023.3642,
      "train_tokens_per_second": 1072.885
    },
    {
      "epoch": 0.1889953594889411,
      "grad_norm": 0.3159874975681305,
      "learning_rate": 4.3338898163606014e-05,
      "loss": 0.9472,
      "num_input_tokens_seen": 6464216,
      "step": 2352,
      "train_runtime": 6025.0806,
      "train_tokens_per_second": 1072.885
    },
    {
      "epoch": 0.18907571465879186,
      "grad_norm": 0.26435011625289917,
      "learning_rate": 4.3272120200333893e-05,
      "loss": 1.0529,
      "num_input_tokens_seen": 6467076,
      "step": 2353,
      "train_runtime": 6027.3932,
      "train_tokens_per_second": 1072.947
    },
    {
      "epoch": 0.1891560698286426,
      "grad_norm": 0.31573250889778137,
      "learning_rate": 4.3205342237061773e-05,
      "loss": 1.1073,
      "num_input_tokens_seen": 6469118,
      "step": 2354,
      "train_runtime": 6029.2498,
      "train_tokens_per_second": 1072.956
    },
    {
      "epoch": 0.18923642499849333,
      "grad_norm": 0.279494047164917,
      "learning_rate": 4.313856427378965e-05,
      "loss": 0.9573,
      "num_input_tokens_seen": 6472412,
      "step": 2355,
      "train_runtime": 6031.8175,
      "train_tokens_per_second": 1073.045
    },
    {
      "epoch": 0.18931678016834408,
      "grad_norm": 0.3572263717651367,
      "learning_rate": 4.307178631051753e-05,
      "loss": 0.9254,
      "num_input_tokens_seen": 6474856,
      "step": 2356,
      "train_runtime": 6033.7637,
      "train_tokens_per_second": 1073.104
    },
    {
      "epoch": 0.18939713533819483,
      "grad_norm": 0.3090859353542328,
      "learning_rate": 4.300500834724541e-05,
      "loss": 1.1394,
      "num_input_tokens_seen": 6477788,
      "step": 2357,
      "train_runtime": 6036.23,
      "train_tokens_per_second": 1073.151
    },
    {
      "epoch": 0.18947749050804555,
      "grad_norm": 0.4273018538951874,
      "learning_rate": 4.293823038397329e-05,
      "loss": 0.8537,
      "num_input_tokens_seen": 6480154,
      "step": 2358,
      "train_runtime": 6038.226,
      "train_tokens_per_second": 1073.188
    },
    {
      "epoch": 0.1895578456778963,
      "grad_norm": 0.282573401927948,
      "learning_rate": 4.2871452420701166e-05,
      "loss": 0.9571,
      "num_input_tokens_seen": 6483698,
      "step": 2359,
      "train_runtime": 6040.971,
      "train_tokens_per_second": 1073.287
    },
    {
      "epoch": 0.18963820084774705,
      "grad_norm": 0.3736294209957123,
      "learning_rate": 4.280467445742905e-05,
      "loss": 1.1787,
      "num_input_tokens_seen": 6485350,
      "step": 2360,
      "train_runtime": 6042.5952,
      "train_tokens_per_second": 1073.272
    },
    {
      "epoch": 0.1897185560175978,
      "grad_norm": 0.35731983184814453,
      "learning_rate": 4.273789649415693e-05,
      "loss": 1.2476,
      "num_input_tokens_seen": 6489322,
      "step": 2361,
      "train_runtime": 6045.4784,
      "train_tokens_per_second": 1073.417
    },
    {
      "epoch": 0.18979891118744852,
      "grad_norm": 0.27530911564826965,
      "learning_rate": 4.267111853088481e-05,
      "loss": 0.6291,
      "num_input_tokens_seen": 6491520,
      "step": 2362,
      "train_runtime": 6047.2803,
      "train_tokens_per_second": 1073.461
    },
    {
      "epoch": 0.18987926635729926,
      "grad_norm": 0.3326406180858612,
      "learning_rate": 4.2604340567612686e-05,
      "loss": 1.0815,
      "num_input_tokens_seen": 6494368,
      "step": 2363,
      "train_runtime": 6049.5072,
      "train_tokens_per_second": 1073.537
    },
    {
      "epoch": 0.18995962152715,
      "grad_norm": 0.3549513816833496,
      "learning_rate": 4.253756260434057e-05,
      "loss": 0.9236,
      "num_input_tokens_seen": 6496210,
      "step": 2364,
      "train_runtime": 6051.2104,
      "train_tokens_per_second": 1073.539
    },
    {
      "epoch": 0.19003997669700073,
      "grad_norm": 0.3189884424209595,
      "learning_rate": 4.247078464106845e-05,
      "loss": 1.0045,
      "num_input_tokens_seen": 6498918,
      "step": 2365,
      "train_runtime": 6053.4713,
      "train_tokens_per_second": 1073.585
    },
    {
      "epoch": 0.19012033186685148,
      "grad_norm": 0.4169856905937195,
      "learning_rate": 4.240400667779633e-05,
      "loss": 0.9863,
      "num_input_tokens_seen": 6500360,
      "step": 2366,
      "train_runtime": 6055.0558,
      "train_tokens_per_second": 1073.543
    },
    {
      "epoch": 0.19020068703670223,
      "grad_norm": 0.27368399500846863,
      "learning_rate": 4.2337228714524206e-05,
      "loss": 0.6798,
      "num_input_tokens_seen": 6503998,
      "step": 2367,
      "train_runtime": 6057.8511,
      "train_tokens_per_second": 1073.648
    },
    {
      "epoch": 0.19028104220655295,
      "grad_norm": 0.331129789352417,
      "learning_rate": 4.2270450751252086e-05,
      "loss": 0.8617,
      "num_input_tokens_seen": 6505764,
      "step": 2368,
      "train_runtime": 6059.5472,
      "train_tokens_per_second": 1073.639
    },
    {
      "epoch": 0.1903613973764037,
      "grad_norm": 0.4752596914768219,
      "learning_rate": 4.220367278797997e-05,
      "loss": 0.9673,
      "num_input_tokens_seen": 6507824,
      "step": 2369,
      "train_runtime": 6061.3396,
      "train_tokens_per_second": 1073.661
    },
    {
      "epoch": 0.19044175254625445,
      "grad_norm": 0.32049426436424255,
      "learning_rate": 4.213689482470785e-05,
      "loss": 0.9059,
      "num_input_tokens_seen": 6510262,
      "step": 2370,
      "train_runtime": 6063.4035,
      "train_tokens_per_second": 1073.698
    },
    {
      "epoch": 0.19052210771610517,
      "grad_norm": 0.3731323778629303,
      "learning_rate": 4.2070116861435725e-05,
      "loss": 1.1558,
      "num_input_tokens_seen": 6512640,
      "step": 2371,
      "train_runtime": 6067.1397,
      "train_tokens_per_second": 1073.428
    },
    {
      "epoch": 0.19060246288595592,
      "grad_norm": 0.38448357582092285,
      "learning_rate": 4.2003338898163605e-05,
      "loss": 1.1643,
      "num_input_tokens_seen": 6514484,
      "step": 2372,
      "train_runtime": 6068.9161,
      "train_tokens_per_second": 1073.418
    },
    {
      "epoch": 0.19068281805580667,
      "grad_norm": 0.3158486783504486,
      "learning_rate": 4.193656093489149e-05,
      "loss": 1.1535,
      "num_input_tokens_seen": 6518266,
      "step": 2373,
      "train_runtime": 6071.776,
      "train_tokens_per_second": 1073.535
    },
    {
      "epoch": 0.19076317322565742,
      "grad_norm": 0.36079153418540955,
      "learning_rate": 4.186978297161937e-05,
      "loss": 0.7358,
      "num_input_tokens_seen": 6521742,
      "step": 2374,
      "train_runtime": 6074.4449,
      "train_tokens_per_second": 1073.636
    },
    {
      "epoch": 0.19084352839550814,
      "grad_norm": 0.27630579471588135,
      "learning_rate": 4.1803005008347245e-05,
      "loss": 0.7908,
      "num_input_tokens_seen": 6523998,
      "step": 2375,
      "train_runtime": 6076.3599,
      "train_tokens_per_second": 1073.669
    },
    {
      "epoch": 0.1909238835653589,
      "grad_norm": 0.41513073444366455,
      "learning_rate": 4.1736227045075125e-05,
      "loss": 0.8549,
      "num_input_tokens_seen": 6525714,
      "step": 2376,
      "train_runtime": 6078.0685,
      "train_tokens_per_second": 1073.649
    },
    {
      "epoch": 0.19100423873520964,
      "grad_norm": 0.31493115425109863,
      "learning_rate": 4.1669449081803005e-05,
      "loss": 0.9587,
      "num_input_tokens_seen": 6527792,
      "step": 2377,
      "train_runtime": 6079.9809,
      "train_tokens_per_second": 1073.653
    },
    {
      "epoch": 0.19108459390506036,
      "grad_norm": 0.32812929153442383,
      "learning_rate": 4.160267111853089e-05,
      "loss": 0.849,
      "num_input_tokens_seen": 6529466,
      "step": 2378,
      "train_runtime": 6081.6239,
      "train_tokens_per_second": 1073.639
    },
    {
      "epoch": 0.1911649490749111,
      "grad_norm": 0.3051978349685669,
      "learning_rate": 4.1535893155258765e-05,
      "loss": 1.2035,
      "num_input_tokens_seen": 6531546,
      "step": 2379,
      "train_runtime": 6083.5258,
      "train_tokens_per_second": 1073.645
    },
    {
      "epoch": 0.19124530424476185,
      "grad_norm": 0.3230092227458954,
      "learning_rate": 4.1469115191986645e-05,
      "loss": 1.0275,
      "num_input_tokens_seen": 6534402,
      "step": 2380,
      "train_runtime": 6085.7987,
      "train_tokens_per_second": 1073.713
    },
    {
      "epoch": 0.19132565941461258,
      "grad_norm": 0.4046890139579773,
      "learning_rate": 4.1402337228714525e-05,
      "loss": 1.1253,
      "num_input_tokens_seen": 6536086,
      "step": 2381,
      "train_runtime": 6087.4224,
      "train_tokens_per_second": 1073.703
    },
    {
      "epoch": 0.19140601458446332,
      "grad_norm": 0.35911914706230164,
      "learning_rate": 4.133555926544241e-05,
      "loss": 0.8435,
      "num_input_tokens_seen": 6537544,
      "step": 2382,
      "train_runtime": 6089.0922,
      "train_tokens_per_second": 1073.648
    },
    {
      "epoch": 0.19148636975431407,
      "grad_norm": 0.2804661989212036,
      "learning_rate": 4.1268781302170285e-05,
      "loss": 0.8464,
      "num_input_tokens_seen": 6539790,
      "step": 2383,
      "train_runtime": 6091.0568,
      "train_tokens_per_second": 1073.671
    },
    {
      "epoch": 0.19156672492416482,
      "grad_norm": 0.3444625437259674,
      "learning_rate": 4.1202003338898165e-05,
      "loss": 1.2321,
      "num_input_tokens_seen": 6542590,
      "step": 2384,
      "train_runtime": 6093.2774,
      "train_tokens_per_second": 1073.739
    },
    {
      "epoch": 0.19164708009401554,
      "grad_norm": 0.3022491931915283,
      "learning_rate": 4.1135225375626044e-05,
      "loss": 0.7703,
      "num_input_tokens_seen": 6545400,
      "step": 2385,
      "train_runtime": 6095.4623,
      "train_tokens_per_second": 1073.815
    },
    {
      "epoch": 0.1917274352638663,
      "grad_norm": 0.38246074318885803,
      "learning_rate": 4.1068447412353924e-05,
      "loss": 0.7253,
      "num_input_tokens_seen": 6548834,
      "step": 2386,
      "train_runtime": 6097.9584,
      "train_tokens_per_second": 1073.939
    },
    {
      "epoch": 0.19180779043371704,
      "grad_norm": 0.3465690016746521,
      "learning_rate": 4.1001669449081804e-05,
      "loss": 0.8618,
      "num_input_tokens_seen": 6552600,
      "step": 2387,
      "train_runtime": 6100.9864,
      "train_tokens_per_second": 1074.023
    },
    {
      "epoch": 0.19188814560356776,
      "grad_norm": 0.2807244658470154,
      "learning_rate": 4.0934891485809684e-05,
      "loss": 1.0056,
      "num_input_tokens_seen": 6555216,
      "step": 2388,
      "train_runtime": 6103.1002,
      "train_tokens_per_second": 1074.08
    },
    {
      "epoch": 0.1919685007734185,
      "grad_norm": 0.28758811950683594,
      "learning_rate": 4.0868113522537564e-05,
      "loss": 0.7992,
      "num_input_tokens_seen": 6556970,
      "step": 2389,
      "train_runtime": 6104.822,
      "train_tokens_per_second": 1074.064
    },
    {
      "epoch": 0.19204885594326926,
      "grad_norm": 0.31233835220336914,
      "learning_rate": 4.0801335559265444e-05,
      "loss": 0.8729,
      "num_input_tokens_seen": 6559388,
      "step": 2390,
      "train_runtime": 6106.843,
      "train_tokens_per_second": 1074.105
    },
    {
      "epoch": 0.19212921111311998,
      "grad_norm": 0.2847444713115692,
      "learning_rate": 4.0734557595993324e-05,
      "loss": 1.0129,
      "num_input_tokens_seen": 6562582,
      "step": 2391,
      "train_runtime": 6109.2983,
      "train_tokens_per_second": 1074.196
    },
    {
      "epoch": 0.19220956628297073,
      "grad_norm": 0.30738258361816406,
      "learning_rate": 4.0667779632721204e-05,
      "loss": 0.8534,
      "num_input_tokens_seen": 6565360,
      "step": 2392,
      "train_runtime": 6111.4727,
      "train_tokens_per_second": 1074.268
    },
    {
      "epoch": 0.19228992145282148,
      "grad_norm": 0.34981194138526917,
      "learning_rate": 4.0601001669449084e-05,
      "loss": 0.9299,
      "num_input_tokens_seen": 6568706,
      "step": 2393,
      "train_runtime": 6113.9057,
      "train_tokens_per_second": 1074.388
    },
    {
      "epoch": 0.1923702766226722,
      "grad_norm": 0.33307093381881714,
      "learning_rate": 4.0534223706176964e-05,
      "loss": 0.8867,
      "num_input_tokens_seen": 6570248,
      "step": 2394,
      "train_runtime": 6115.5821,
      "train_tokens_per_second": 1074.345
    },
    {
      "epoch": 0.19245063179252295,
      "grad_norm": 0.28083500266075134,
      "learning_rate": 4.0467445742904844e-05,
      "loss": 1.0867,
      "num_input_tokens_seen": 6572680,
      "step": 2395,
      "train_runtime": 6117.6535,
      "train_tokens_per_second": 1074.379
    },
    {
      "epoch": 0.1925309869623737,
      "grad_norm": 0.25081828236579895,
      "learning_rate": 4.0400667779632724e-05,
      "loss": 0.6236,
      "num_input_tokens_seen": 6575098,
      "step": 2396,
      "train_runtime": 6119.6416,
      "train_tokens_per_second": 1074.425
    },
    {
      "epoch": 0.19261134213222444,
      "grad_norm": 0.3000496029853821,
      "learning_rate": 4.0333889816360604e-05,
      "loss": 0.8846,
      "num_input_tokens_seen": 6577948,
      "step": 2397,
      "train_runtime": 6121.8345,
      "train_tokens_per_second": 1074.506
    },
    {
      "epoch": 0.19269169730207517,
      "grad_norm": 0.33908024430274963,
      "learning_rate": 4.0267111853088484e-05,
      "loss": 0.9774,
      "num_input_tokens_seen": 6580748,
      "step": 2398,
      "train_runtime": 6124.0722,
      "train_tokens_per_second": 1074.571
    },
    {
      "epoch": 0.19277205247192591,
      "grad_norm": 0.3886743187904358,
      "learning_rate": 4.0200333889816363e-05,
      "loss": 1.2201,
      "num_input_tokens_seen": 6583026,
      "step": 2399,
      "train_runtime": 6126.0713,
      "train_tokens_per_second": 1074.592
    },
    {
      "epoch": 0.19285240764177666,
      "grad_norm": 0.33162257075309753,
      "learning_rate": 4.0133555926544243e-05,
      "loss": 0.9361,
      "num_input_tokens_seen": 6585668,
      "step": 2400,
      "train_runtime": 6128.1505,
      "train_tokens_per_second": 1074.658
    },
    {
      "epoch": 0.19285240764177666,
      "eval_loss": 0.9504203200340271,
      "eval_runtime": 91.9657,
      "eval_samples_per_second": 10.808,
      "eval_steps_per_second": 5.404,
      "num_input_tokens_seen": 6585668,
      "step": 2400
    },
    {
      "epoch": 0.19293276281162738,
      "grad_norm": 0.32750827074050903,
      "learning_rate": 4.006677796327212e-05,
      "loss": 0.9879,
      "num_input_tokens_seen": 6587752,
      "step": 2401,
      "train_runtime": 6223.9332,
      "train_tokens_per_second": 1058.455
    },
    {
      "epoch": 0.19301311798147813,
      "grad_norm": 0.3905124068260193,
      "learning_rate": 4e-05,
      "loss": 1.1807,
      "num_input_tokens_seen": 6589056,
      "step": 2402,
      "train_runtime": 6225.4982,
      "train_tokens_per_second": 1058.398
    },
    {
      "epoch": 0.19309347315132888,
      "grad_norm": 0.2537885904312134,
      "learning_rate": 3.9933222036727876e-05,
      "loss": 0.8074,
      "num_input_tokens_seen": 6591830,
      "step": 2403,
      "train_runtime": 6227.6576,
      "train_tokens_per_second": 1058.477
    },
    {
      "epoch": 0.1931738283211796,
      "grad_norm": 0.2999289035797119,
      "learning_rate": 3.986644407345576e-05,
      "loss": 0.8308,
      "num_input_tokens_seen": 6593994,
      "step": 2404,
      "train_runtime": 6229.5632,
      "train_tokens_per_second": 1058.5
    },
    {
      "epoch": 0.19325418349103035,
      "grad_norm": 0.3161380887031555,
      "learning_rate": 3.979966611018364e-05,
      "loss": 1.1495,
      "num_input_tokens_seen": 6596840,
      "step": 2405,
      "train_runtime": 6231.7913,
      "train_tokens_per_second": 1058.578
    },
    {
      "epoch": 0.1933345386608811,
      "grad_norm": 0.28440114855766296,
      "learning_rate": 3.973288814691152e-05,
      "loss": 0.9586,
      "num_input_tokens_seen": 6601202,
      "step": 2406,
      "train_runtime": 6234.8798,
      "train_tokens_per_second": 1058.754
    },
    {
      "epoch": 0.19341489383073182,
      "grad_norm": 0.3244627118110657,
      "learning_rate": 3.9666110183639396e-05,
      "loss": 0.6884,
      "num_input_tokens_seen": 6603970,
      "step": 2407,
      "train_runtime": 6237.1354,
      "train_tokens_per_second": 1058.815
    },
    {
      "epoch": 0.19349524900058257,
      "grad_norm": 0.31390276551246643,
      "learning_rate": 3.959933222036728e-05,
      "loss": 0.9528,
      "num_input_tokens_seen": 6606010,
      "step": 2408,
      "train_runtime": 6238.9019,
      "train_tokens_per_second": 1058.842
    },
    {
      "epoch": 0.19357560417043332,
      "grad_norm": 0.33451029658317566,
      "learning_rate": 3.953255425709516e-05,
      "loss": 0.8656,
      "num_input_tokens_seen": 6609482,
      "step": 2409,
      "train_runtime": 6241.5444,
      "train_tokens_per_second": 1058.95
    },
    {
      "epoch": 0.19365595934028407,
      "grad_norm": 0.2858388423919678,
      "learning_rate": 3.946577629382304e-05,
      "loss": 0.9274,
      "num_input_tokens_seen": 6612334,
      "step": 2410,
      "train_runtime": 6243.754,
      "train_tokens_per_second": 1059.032
    },
    {
      "epoch": 0.1937363145101348,
      "grad_norm": 0.27476608753204346,
      "learning_rate": 3.9398998330550916e-05,
      "loss": 0.8203,
      "num_input_tokens_seen": 6614684,
      "step": 2411,
      "train_runtime": 6245.7469,
      "train_tokens_per_second": 1059.07
    },
    {
      "epoch": 0.19381666967998554,
      "grad_norm": 0.33219289779663086,
      "learning_rate": 3.93322203672788e-05,
      "loss": 0.9426,
      "num_input_tokens_seen": 6619112,
      "step": 2412,
      "train_runtime": 6248.9617,
      "train_tokens_per_second": 1059.234
    },
    {
      "epoch": 0.19389702484983629,
      "grad_norm": 0.2844533920288086,
      "learning_rate": 3.926544240400668e-05,
      "loss": 0.87,
      "num_input_tokens_seen": 6621710,
      "step": 2413,
      "train_runtime": 6251.0844,
      "train_tokens_per_second": 1059.29
    },
    {
      "epoch": 0.193977380019687,
      "grad_norm": 0.2911604642868042,
      "learning_rate": 3.919866444073456e-05,
      "loss": 0.807,
      "num_input_tokens_seen": 6624134,
      "step": 2414,
      "train_runtime": 6253.1218,
      "train_tokens_per_second": 1059.332
    },
    {
      "epoch": 0.19405773518953776,
      "grad_norm": 0.37950772047042847,
      "learning_rate": 3.9131886477462436e-05,
      "loss": 1.2061,
      "num_input_tokens_seen": 6626408,
      "step": 2415,
      "train_runtime": 6255.1404,
      "train_tokens_per_second": 1059.354
    },
    {
      "epoch": 0.1941380903593885,
      "grad_norm": 0.3641248643398285,
      "learning_rate": 3.9065108514190316e-05,
      "loss": 1.0681,
      "num_input_tokens_seen": 6628788,
      "step": 2416,
      "train_runtime": 6257.1809,
      "train_tokens_per_second": 1059.389
    },
    {
      "epoch": 0.19421844552923923,
      "grad_norm": 0.31378281116485596,
      "learning_rate": 3.89983305509182e-05,
      "loss": 1.2204,
      "num_input_tokens_seen": 6631396,
      "step": 2417,
      "train_runtime": 6259.2567,
      "train_tokens_per_second": 1059.454
    },
    {
      "epoch": 0.19429880069908997,
      "grad_norm": 0.25066882371902466,
      "learning_rate": 3.893155258764608e-05,
      "loss": 0.7568,
      "num_input_tokens_seen": 6634142,
      "step": 2418,
      "train_runtime": 6261.4039,
      "train_tokens_per_second": 1059.529
    },
    {
      "epoch": 0.19437915586894072,
      "grad_norm": 0.3530522584915161,
      "learning_rate": 3.8864774624373955e-05,
      "loss": 0.861,
      "num_input_tokens_seen": 6637412,
      "step": 2419,
      "train_runtime": 6263.8928,
      "train_tokens_per_second": 1059.631
    },
    {
      "epoch": 0.19445951103879147,
      "grad_norm": 0.3219548165798187,
      "learning_rate": 3.8797996661101835e-05,
      "loss": 0.9204,
      "num_input_tokens_seen": 6639986,
      "step": 2420,
      "train_runtime": 6265.9187,
      "train_tokens_per_second": 1059.699
    },
    {
      "epoch": 0.1945398662086422,
      "grad_norm": 0.31155136227607727,
      "learning_rate": 3.873121869782972e-05,
      "loss": 1.1368,
      "num_input_tokens_seen": 6642074,
      "step": 2421,
      "train_runtime": 6267.7409,
      "train_tokens_per_second": 1059.724
    },
    {
      "epoch": 0.19462022137849294,
      "grad_norm": 0.3154776990413666,
      "learning_rate": 3.86644407345576e-05,
      "loss": 0.9429,
      "num_input_tokens_seen": 6644438,
      "step": 2422,
      "train_runtime": 6269.6787,
      "train_tokens_per_second": 1059.773
    },
    {
      "epoch": 0.1947005765483437,
      "grad_norm": 0.2579866945743561,
      "learning_rate": 3.8597662771285475e-05,
      "loss": 0.5127,
      "num_input_tokens_seen": 6647748,
      "step": 2423,
      "train_runtime": 6272.1858,
      "train_tokens_per_second": 1059.877
    },
    {
      "epoch": 0.1947809317181944,
      "grad_norm": 0.28103676438331604,
      "learning_rate": 3.8530884808013355e-05,
      "loss": 0.9606,
      "num_input_tokens_seen": 6650942,
      "step": 2424,
      "train_runtime": 6274.567,
      "train_tokens_per_second": 1059.984
    },
    {
      "epoch": 0.19486128688804516,
      "grad_norm": 0.31860002875328064,
      "learning_rate": 3.8464106844741235e-05,
      "loss": 1.0137,
      "num_input_tokens_seen": 6653204,
      "step": 2425,
      "train_runtime": 6276.4407,
      "train_tokens_per_second": 1060.028
    },
    {
      "epoch": 0.1949416420578959,
      "grad_norm": 0.32986995577812195,
      "learning_rate": 3.839732888146912e-05,
      "loss": 0.9636,
      "num_input_tokens_seen": 6657364,
      "step": 2426,
      "train_runtime": 6279.4053,
      "train_tokens_per_second": 1060.19
    },
    {
      "epoch": 0.19502199722774663,
      "grad_norm": 0.2918531894683838,
      "learning_rate": 3.8330550918196995e-05,
      "loss": 0.7397,
      "num_input_tokens_seen": 6663302,
      "step": 2427,
      "train_runtime": 6283.4524,
      "train_tokens_per_second": 1060.452
    },
    {
      "epoch": 0.19510235239759738,
      "grad_norm": 0.33364513516426086,
      "learning_rate": 3.8263772954924875e-05,
      "loss": 1.0251,
      "num_input_tokens_seen": 6668602,
      "step": 2428,
      "train_runtime": 6286.9381,
      "train_tokens_per_second": 1060.707
    },
    {
      "epoch": 0.19518270756744813,
      "grad_norm": 0.30675721168518066,
      "learning_rate": 3.8196994991652755e-05,
      "loss": 1.0195,
      "num_input_tokens_seen": 6670712,
      "step": 2429,
      "train_runtime": 6288.69,
      "train_tokens_per_second": 1060.747
    },
    {
      "epoch": 0.19526306273729885,
      "grad_norm": 0.2794005274772644,
      "learning_rate": 3.813021702838064e-05,
      "loss": 0.8031,
      "num_input_tokens_seen": 6673344,
      "step": 2430,
      "train_runtime": 6290.7487,
      "train_tokens_per_second": 1060.819
    },
    {
      "epoch": 0.1953434179071496,
      "grad_norm": 0.2989581823348999,
      "learning_rate": 3.8063439065108514e-05,
      "loss": 0.9371,
      "num_input_tokens_seen": 6677694,
      "step": 2431,
      "train_runtime": 6295.8081,
      "train_tokens_per_second": 1060.657
    },
    {
      "epoch": 0.19542377307700035,
      "grad_norm": 0.319682240486145,
      "learning_rate": 3.7996661101836394e-05,
      "loss": 1.1694,
      "num_input_tokens_seen": 6680228,
      "step": 2432,
      "train_runtime": 6297.7972,
      "train_tokens_per_second": 1060.725
    },
    {
      "epoch": 0.1955041282468511,
      "grad_norm": 0.34445488452911377,
      "learning_rate": 3.7929883138564274e-05,
      "loss": 1.1355,
      "num_input_tokens_seen": 6681880,
      "step": 2433,
      "train_runtime": 6299.4504,
      "train_tokens_per_second": 1060.708
    },
    {
      "epoch": 0.19558448341670182,
      "grad_norm": 0.36404046416282654,
      "learning_rate": 3.7863105175292154e-05,
      "loss": 0.8176,
      "num_input_tokens_seen": 6684426,
      "step": 2434,
      "train_runtime": 6301.6144,
      "train_tokens_per_second": 1060.748
    },
    {
      "epoch": 0.19566483858655256,
      "grad_norm": 0.307226300239563,
      "learning_rate": 3.7796327212020034e-05,
      "loss": 0.9641,
      "num_input_tokens_seen": 6686416,
      "step": 2435,
      "train_runtime": 6303.4884,
      "train_tokens_per_second": 1060.749
    },
    {
      "epoch": 0.1957451937564033,
      "grad_norm": 0.2776503264904022,
      "learning_rate": 3.7729549248747914e-05,
      "loss": 0.7334,
      "num_input_tokens_seen": 6687934,
      "step": 2436,
      "train_runtime": 6305.1103,
      "train_tokens_per_second": 1060.716
    },
    {
      "epoch": 0.19582554892625403,
      "grad_norm": 0.32679176330566406,
      "learning_rate": 3.7662771285475794e-05,
      "loss": 0.9927,
      "num_input_tokens_seen": 6690244,
      "step": 2437,
      "train_runtime": 6307.047,
      "train_tokens_per_second": 1060.757
    },
    {
      "epoch": 0.19590590409610478,
      "grad_norm": 0.36547166109085083,
      "learning_rate": 3.7595993322203674e-05,
      "loss": 1.18,
      "num_input_tokens_seen": 6693160,
      "step": 2438,
      "train_runtime": 6309.3187,
      "train_tokens_per_second": 1060.837
    },
    {
      "epoch": 0.19598625926595553,
      "grad_norm": 0.31803661584854126,
      "learning_rate": 3.7529215358931554e-05,
      "loss": 0.696,
      "num_input_tokens_seen": 6694940,
      "step": 2439,
      "train_runtime": 6311.1235,
      "train_tokens_per_second": 1060.816
    },
    {
      "epoch": 0.19606661443580625,
      "grad_norm": 0.34208938479423523,
      "learning_rate": 3.7462437395659434e-05,
      "loss": 0.8492,
      "num_input_tokens_seen": 6697220,
      "step": 2440,
      "train_runtime": 6313.1404,
      "train_tokens_per_second": 1060.838
    },
    {
      "epoch": 0.196146969605657,
      "grad_norm": 0.2791343629360199,
      "learning_rate": 3.7395659432387314e-05,
      "loss": 1.0493,
      "num_input_tokens_seen": 6702412,
      "step": 2441,
      "train_runtime": 6316.8527,
      "train_tokens_per_second": 1061.037
    },
    {
      "epoch": 0.19622732477550775,
      "grad_norm": 0.2930096983909607,
      "learning_rate": 3.7328881469115194e-05,
      "loss": 1.1102,
      "num_input_tokens_seen": 6706494,
      "step": 2442,
      "train_runtime": 6319.7433,
      "train_tokens_per_second": 1061.197
    },
    {
      "epoch": 0.19630767994535847,
      "grad_norm": 0.27030718326568604,
      "learning_rate": 3.7262103505843074e-05,
      "loss": 0.9448,
      "num_input_tokens_seen": 6708596,
      "step": 2443,
      "train_runtime": 6321.5725,
      "train_tokens_per_second": 1061.223
    },
    {
      "epoch": 0.19638803511520922,
      "grad_norm": 0.28284841775894165,
      "learning_rate": 3.7195325542570954e-05,
      "loss": 0.768,
      "num_input_tokens_seen": 6710428,
      "step": 2444,
      "train_runtime": 6323.3645,
      "train_tokens_per_second": 1061.212
    },
    {
      "epoch": 0.19646839028505997,
      "grad_norm": 0.24541085958480835,
      "learning_rate": 3.7128547579298833e-05,
      "loss": 0.8682,
      "num_input_tokens_seen": 6713602,
      "step": 2445,
      "train_runtime": 6325.7785,
      "train_tokens_per_second": 1061.308
    },
    {
      "epoch": 0.19654874545491072,
      "grad_norm": 0.26650404930114746,
      "learning_rate": 3.7061769616026713e-05,
      "loss": 0.8112,
      "num_input_tokens_seen": 6715268,
      "step": 2446,
      "train_runtime": 6327.4473,
      "train_tokens_per_second": 1061.292
    },
    {
      "epoch": 0.19662910062476144,
      "grad_norm": 0.2887314260005951,
      "learning_rate": 3.699499165275459e-05,
      "loss": 0.9284,
      "num_input_tokens_seen": 6717818,
      "step": 2447,
      "train_runtime": 6329.537,
      "train_tokens_per_second": 1061.344
    },
    {
      "epoch": 0.1967094557946122,
      "grad_norm": 0.27001363039016724,
      "learning_rate": 3.692821368948247e-05,
      "loss": 0.8024,
      "num_input_tokens_seen": 6720532,
      "step": 2448,
      "train_runtime": 6331.7748,
      "train_tokens_per_second": 1061.398
    },
    {
      "epoch": 0.19678981096446294,
      "grad_norm": 0.35359224677085876,
      "learning_rate": 3.686143572621035e-05,
      "loss": 1.2375,
      "num_input_tokens_seen": 6725752,
      "step": 2449,
      "train_runtime": 6335.3254,
      "train_tokens_per_second": 1061.627
    },
    {
      "epoch": 0.19687016613431366,
      "grad_norm": 0.2674376368522644,
      "learning_rate": 3.679465776293823e-05,
      "loss": 0.8288,
      "num_input_tokens_seen": 6728112,
      "step": 2450,
      "train_runtime": 6337.2923,
      "train_tokens_per_second": 1061.67
    },
    {
      "epoch": 0.1969505213041644,
      "grad_norm": 0.3892456591129303,
      "learning_rate": 3.6727879799666106e-05,
      "loss": 0.8965,
      "num_input_tokens_seen": 6730768,
      "step": 2451,
      "train_runtime": 6339.4992,
      "train_tokens_per_second": 1061.719
    },
    {
      "epoch": 0.19703087647401515,
      "grad_norm": 0.45840367674827576,
      "learning_rate": 3.666110183639399e-05,
      "loss": 1.5376,
      "num_input_tokens_seen": 6733332,
      "step": 2452,
      "train_runtime": 6341.5603,
      "train_tokens_per_second": 1061.778
    },
    {
      "epoch": 0.19711123164386588,
      "grad_norm": 0.31940752267837524,
      "learning_rate": 3.659432387312187e-05,
      "loss": 1.2081,
      "num_input_tokens_seen": 6736724,
      "step": 2453,
      "train_runtime": 6344.0347,
      "train_tokens_per_second": 1061.899
    },
    {
      "epoch": 0.19719158681371662,
      "grad_norm": 0.31960538029670715,
      "learning_rate": 3.652754590984975e-05,
      "loss": 0.7606,
      "num_input_tokens_seen": 6739186,
      "step": 2454,
      "train_runtime": 6346.1753,
      "train_tokens_per_second": 1061.929
    },
    {
      "epoch": 0.19727194198356737,
      "grad_norm": 0.2756505608558655,
      "learning_rate": 3.6460767946577626e-05,
      "loss": 1.1466,
      "num_input_tokens_seen": 6742722,
      "step": 2455,
      "train_runtime": 6348.7388,
      "train_tokens_per_second": 1062.057
    },
    {
      "epoch": 0.19735229715341812,
      "grad_norm": 0.293165385723114,
      "learning_rate": 3.639398998330551e-05,
      "loss": 1.3314,
      "num_input_tokens_seen": 6745248,
      "step": 2456,
      "train_runtime": 6350.8175,
      "train_tokens_per_second": 1062.107
    },
    {
      "epoch": 0.19743265232326884,
      "grad_norm": 0.3405793011188507,
      "learning_rate": 3.632721202003339e-05,
      "loss": 0.9363,
      "num_input_tokens_seen": 6747486,
      "step": 2457,
      "train_runtime": 6352.8011,
      "train_tokens_per_second": 1062.128
    },
    {
      "epoch": 0.1975130074931196,
      "grad_norm": 0.32479724287986755,
      "learning_rate": 3.626043405676127e-05,
      "loss": 1.1031,
      "num_input_tokens_seen": 6749332,
      "step": 2458,
      "train_runtime": 6354.5351,
      "train_tokens_per_second": 1062.128
    },
    {
      "epoch": 0.19759336266297034,
      "grad_norm": 0.30004751682281494,
      "learning_rate": 3.6193656093489146e-05,
      "loss": 1.1166,
      "num_input_tokens_seen": 6752850,
      "step": 2459,
      "train_runtime": 6357.1873,
      "train_tokens_per_second": 1062.239
    },
    {
      "epoch": 0.19767371783282106,
      "grad_norm": 0.32536259293556213,
      "learning_rate": 3.6126878130217026e-05,
      "loss": 0.8243,
      "num_input_tokens_seen": 6754952,
      "step": 2460,
      "train_runtime": 6359.0399,
      "train_tokens_per_second": 1062.26
    },
    {
      "epoch": 0.1977540730026718,
      "grad_norm": 0.21679791808128357,
      "learning_rate": 3.606010016694491e-05,
      "loss": 1.2679,
      "num_input_tokens_seen": 6757206,
      "step": 2461,
      "train_runtime": 6363.03,
      "train_tokens_per_second": 1061.948
    },
    {
      "epoch": 0.19783442817252256,
      "grad_norm": 0.3343377709388733,
      "learning_rate": 3.599332220367279e-05,
      "loss": 0.694,
      "num_input_tokens_seen": 6759614,
      "step": 2462,
      "train_runtime": 6365.0191,
      "train_tokens_per_second": 1061.994
    },
    {
      "epoch": 0.19791478334237328,
      "grad_norm": 0.34032630920410156,
      "learning_rate": 3.5926544240400665e-05,
      "loss": 1.2218,
      "num_input_tokens_seen": 6762950,
      "step": 2463,
      "train_runtime": 6367.5206,
      "train_tokens_per_second": 1062.101
    },
    {
      "epoch": 0.19799513851222403,
      "grad_norm": 0.3259252607822418,
      "learning_rate": 3.5859766277128545e-05,
      "loss": 0.864,
      "num_input_tokens_seen": 6764568,
      "step": 2464,
      "train_runtime": 6369.2484,
      "train_tokens_per_second": 1062.067
    },
    {
      "epoch": 0.19807549368207478,
      "grad_norm": 0.36383065581321716,
      "learning_rate": 3.579298831385643e-05,
      "loss": 0.7951,
      "num_input_tokens_seen": 6766578,
      "step": 2465,
      "train_runtime": 6371.0441,
      "train_tokens_per_second": 1062.083
    },
    {
      "epoch": 0.1981558488519255,
      "grad_norm": 0.36285698413848877,
      "learning_rate": 3.572621035058431e-05,
      "loss": 0.6244,
      "num_input_tokens_seen": 6769278,
      "step": 2466,
      "train_runtime": 6373.2886,
      "train_tokens_per_second": 1062.133
    },
    {
      "epoch": 0.19823620402177625,
      "grad_norm": 0.3284915089607239,
      "learning_rate": 3.5659432387312185e-05,
      "loss": 0.8725,
      "num_input_tokens_seen": 6771400,
      "step": 2467,
      "train_runtime": 6375.1712,
      "train_tokens_per_second": 1062.152
    },
    {
      "epoch": 0.198316559191627,
      "grad_norm": 0.34764623641967773,
      "learning_rate": 3.5592654424040065e-05,
      "loss": 1.0129,
      "num_input_tokens_seen": 6774194,
      "step": 2468,
      "train_runtime": 6377.4395,
      "train_tokens_per_second": 1062.212
    },
    {
      "epoch": 0.19839691436147774,
      "grad_norm": 0.314209520816803,
      "learning_rate": 3.5525876460767945e-05,
      "loss": 1.2032,
      "num_input_tokens_seen": 6777538,
      "step": 2469,
      "train_runtime": 6379.9604,
      "train_tokens_per_second": 1062.317
    },
    {
      "epoch": 0.19847726953132847,
      "grad_norm": 0.2474755197763443,
      "learning_rate": 3.545909849749583e-05,
      "loss": 0.5714,
      "num_input_tokens_seen": 6781134,
      "step": 2470,
      "train_runtime": 6382.6045,
      "train_tokens_per_second": 1062.44
    },
    {
      "epoch": 0.19855762470117921,
      "grad_norm": 0.415279358625412,
      "learning_rate": 3.5392320534223705e-05,
      "loss": 0.8931,
      "num_input_tokens_seen": 6784054,
      "step": 2471,
      "train_runtime": 6384.8581,
      "train_tokens_per_second": 1062.522
    },
    {
      "epoch": 0.19863797987102996,
      "grad_norm": 0.3680597245693207,
      "learning_rate": 3.5325542570951585e-05,
      "loss": 0.9366,
      "num_input_tokens_seen": 6785938,
      "step": 2472,
      "train_runtime": 6386.6441,
      "train_tokens_per_second": 1062.52
    },
    {
      "epoch": 0.19871833504088068,
      "grad_norm": 0.32916340231895447,
      "learning_rate": 3.5258764607679465e-05,
      "loss": 0.8794,
      "num_input_tokens_seen": 6787884,
      "step": 2473,
      "train_runtime": 6388.3378,
      "train_tokens_per_second": 1062.543
    },
    {
      "epoch": 0.19879869021073143,
      "grad_norm": 0.3731456398963928,
      "learning_rate": 3.519198664440735e-05,
      "loss": 1.212,
      "num_input_tokens_seen": 6790908,
      "step": 2474,
      "train_runtime": 6390.7884,
      "train_tokens_per_second": 1062.609
    },
    {
      "epoch": 0.19887904538058218,
      "grad_norm": 0.3627934753894806,
      "learning_rate": 3.5125208681135225e-05,
      "loss": 1.0878,
      "num_input_tokens_seen": 6794140,
      "step": 2475,
      "train_runtime": 6393.2212,
      "train_tokens_per_second": 1062.71
    },
    {
      "epoch": 0.1989594005504329,
      "grad_norm": 0.29463857412338257,
      "learning_rate": 3.5058430717863105e-05,
      "loss": 0.7593,
      "num_input_tokens_seen": 6797450,
      "step": 2476,
      "train_runtime": 6395.782,
      "train_tokens_per_second": 1062.802
    },
    {
      "epoch": 0.19903975572028365,
      "grad_norm": 0.31568753719329834,
      "learning_rate": 3.4991652754590984e-05,
      "loss": 1.3336,
      "num_input_tokens_seen": 6800806,
      "step": 2477,
      "train_runtime": 6398.3314,
      "train_tokens_per_second": 1062.903
    },
    {
      "epoch": 0.1991201108901344,
      "grad_norm": 0.3566600978374481,
      "learning_rate": 3.492487479131887e-05,
      "loss": 1.2366,
      "num_input_tokens_seen": 6803638,
      "step": 2478,
      "train_runtime": 6400.7014,
      "train_tokens_per_second": 1062.952
    },
    {
      "epoch": 0.19920046605998512,
      "grad_norm": 0.3126729726791382,
      "learning_rate": 3.4858096828046744e-05,
      "loss": 1.1001,
      "num_input_tokens_seen": 6806108,
      "step": 2479,
      "train_runtime": 6402.8289,
      "train_tokens_per_second": 1062.985
    },
    {
      "epoch": 0.19928082122983587,
      "grad_norm": 0.2759259343147278,
      "learning_rate": 3.4791318864774624e-05,
      "loss": 0.625,
      "num_input_tokens_seen": 6808676,
      "step": 2480,
      "train_runtime": 6405.0031,
      "train_tokens_per_second": 1063.025
    },
    {
      "epoch": 0.19936117639968662,
      "grad_norm": 0.298539936542511,
      "learning_rate": 3.4724540901502504e-05,
      "loss": 0.8877,
      "num_input_tokens_seen": 6811472,
      "step": 2481,
      "train_runtime": 6407.2884,
      "train_tokens_per_second": 1063.082
    },
    {
      "epoch": 0.19944153156953737,
      "grad_norm": 0.2916746139526367,
      "learning_rate": 3.4657762938230384e-05,
      "loss": 0.8358,
      "num_input_tokens_seen": 6814338,
      "step": 2482,
      "train_runtime": 6409.6097,
      "train_tokens_per_second": 1063.144
    },
    {
      "epoch": 0.1995218867393881,
      "grad_norm": 0.3314656913280487,
      "learning_rate": 3.459098497495827e-05,
      "loss": 1.0687,
      "num_input_tokens_seen": 6817654,
      "step": 2483,
      "train_runtime": 6412.1109,
      "train_tokens_per_second": 1063.246
    },
    {
      "epoch": 0.19960224190923884,
      "grad_norm": 0.3381015956401825,
      "learning_rate": 3.4524207011686144e-05,
      "loss": 1.1911,
      "num_input_tokens_seen": 6819462,
      "step": 2484,
      "train_runtime": 6413.8793,
      "train_tokens_per_second": 1063.235
    },
    {
      "epoch": 0.19968259707908959,
      "grad_norm": 0.3587280213832855,
      "learning_rate": 3.4457429048414024e-05,
      "loss": 0.8833,
      "num_input_tokens_seen": 6824428,
      "step": 2485,
      "train_runtime": 6417.4358,
      "train_tokens_per_second": 1063.42
    },
    {
      "epoch": 0.1997629522489403,
      "grad_norm": 0.29025721549987793,
      "learning_rate": 3.4390651085141904e-05,
      "loss": 0.8576,
      "num_input_tokens_seen": 6826822,
      "step": 2486,
      "train_runtime": 6419.3817,
      "train_tokens_per_second": 1063.47
    },
    {
      "epoch": 0.19984330741879106,
      "grad_norm": 0.26699090003967285,
      "learning_rate": 3.432387312186979e-05,
      "loss": 0.8747,
      "num_input_tokens_seen": 6828642,
      "step": 2487,
      "train_runtime": 6421.113,
      "train_tokens_per_second": 1063.467
    },
    {
      "epoch": 0.1999236625886418,
      "grad_norm": 0.3921565115451813,
      "learning_rate": 3.4257095158597664e-05,
      "loss": 0.833,
      "num_input_tokens_seen": 6829966,
      "step": 2488,
      "train_runtime": 6422.6846,
      "train_tokens_per_second": 1063.413
    },
    {
      "epoch": 0.20000401775849252,
      "grad_norm": 0.322794109582901,
      "learning_rate": 3.4190317195325544e-05,
      "loss": 1.1786,
      "num_input_tokens_seen": 6834044,
      "step": 2489,
      "train_runtime": 6425.6764,
      "train_tokens_per_second": 1063.552
    },
    {
      "epoch": 0.20008437292834327,
      "grad_norm": 0.3190370798110962,
      "learning_rate": 3.4123539232053424e-05,
      "loss": 1.005,
      "num_input_tokens_seen": 6837172,
      "step": 2490,
      "train_runtime": 6428.0826,
      "train_tokens_per_second": 1063.641
    },
    {
      "epoch": 0.20016472809819402,
      "grad_norm": 0.25128650665283203,
      "learning_rate": 3.4056761268781303e-05,
      "loss": 0.6746,
      "num_input_tokens_seen": 6839372,
      "step": 2491,
      "train_runtime": 6431.9319,
      "train_tokens_per_second": 1063.346
    },
    {
      "epoch": 0.20024508326804477,
      "grad_norm": 0.2743060886859894,
      "learning_rate": 3.3989983305509183e-05,
      "loss": 0.9751,
      "num_input_tokens_seen": 6841322,
      "step": 2492,
      "train_runtime": 6433.6852,
      "train_tokens_per_second": 1063.36
    },
    {
      "epoch": 0.2003254384378955,
      "grad_norm": 0.2684800922870636,
      "learning_rate": 3.392320534223706e-05,
      "loss": 0.9082,
      "num_input_tokens_seen": 6843902,
      "step": 2493,
      "train_runtime": 6435.745,
      "train_tokens_per_second": 1063.42
    },
    {
      "epoch": 0.20040579360774624,
      "grad_norm": 0.3017663061618805,
      "learning_rate": 3.385642737896494e-05,
      "loss": 1.0876,
      "num_input_tokens_seen": 6846730,
      "step": 2494,
      "train_runtime": 6437.941,
      "train_tokens_per_second": 1063.497
    },
    {
      "epoch": 0.200486148777597,
      "grad_norm": 0.24603429436683655,
      "learning_rate": 3.378964941569282e-05,
      "loss": 1.1759,
      "num_input_tokens_seen": 6849506,
      "step": 2495,
      "train_runtime": 6440.0774,
      "train_tokens_per_second": 1063.575
    },
    {
      "epoch": 0.2005665039474477,
      "grad_norm": 0.37041062116622925,
      "learning_rate": 3.37228714524207e-05,
      "loss": 0.8201,
      "num_input_tokens_seen": 6851176,
      "step": 2496,
      "train_runtime": 6441.7505,
      "train_tokens_per_second": 1063.558
    },
    {
      "epoch": 0.20064685911729846,
      "grad_norm": 0.3324015438556671,
      "learning_rate": 3.365609348914858e-05,
      "loss": 0.92,
      "num_input_tokens_seen": 6854918,
      "step": 2497,
      "train_runtime": 6444.5417,
      "train_tokens_per_second": 1063.678
    },
    {
      "epoch": 0.2007272142871492,
      "grad_norm": 0.3053337633609772,
      "learning_rate": 3.358931552587646e-05,
      "loss": 1.0511,
      "num_input_tokens_seen": 6857064,
      "step": 2498,
      "train_runtime": 6446.49,
      "train_tokens_per_second": 1063.69
    },
    {
      "epoch": 0.20080756945699993,
      "grad_norm": 0.37749725580215454,
      "learning_rate": 3.352253756260434e-05,
      "loss": 1.0809,
      "num_input_tokens_seen": 6859868,
      "step": 2499,
      "train_runtime": 6448.7584,
      "train_tokens_per_second": 1063.75
    },
    {
      "epoch": 0.20088792462685068,
      "grad_norm": 0.2786153256893158,
      "learning_rate": 3.345575959933222e-05,
      "loss": 0.7254,
      "num_input_tokens_seen": 6861820,
      "step": 2500,
      "train_runtime": 6450.5111,
      "train_tokens_per_second": 1063.764
    },
    {
      "epoch": 0.20096827979670143,
      "grad_norm": 0.3043036162853241,
      "learning_rate": 3.33889816360601e-05,
      "loss": 1.2435,
      "num_input_tokens_seen": 6864276,
      "step": 2501,
      "train_runtime": 6452.4659,
      "train_tokens_per_second": 1063.822
    },
    {
      "epoch": 0.20104863496655215,
      "grad_norm": 0.29842492938041687,
      "learning_rate": 3.332220367278798e-05,
      "loss": 0.9389,
      "num_input_tokens_seen": 6868786,
      "step": 2502,
      "train_runtime": 6455.7018,
      "train_tokens_per_second": 1063.987
    },
    {
      "epoch": 0.2011289901364029,
      "grad_norm": 0.3799628019332886,
      "learning_rate": 3.325542570951586e-05,
      "loss": 0.7088,
      "num_input_tokens_seen": 6869876,
      "step": 2503,
      "train_runtime": 6457.2087,
      "train_tokens_per_second": 1063.908
    },
    {
      "epoch": 0.20120934530625365,
      "grad_norm": 0.39191216230392456,
      "learning_rate": 3.318864774624374e-05,
      "loss": 0.8537,
      "num_input_tokens_seen": 6871742,
      "step": 2504,
      "train_runtime": 6459.0027,
      "train_tokens_per_second": 1063.901
    },
    {
      "epoch": 0.2012897004761044,
      "grad_norm": 0.2769601345062256,
      "learning_rate": 3.312186978297162e-05,
      "loss": 1.0407,
      "num_input_tokens_seen": 6873866,
      "step": 2505,
      "train_runtime": 6460.9644,
      "train_tokens_per_second": 1063.907
    },
    {
      "epoch": 0.20137005564595511,
      "grad_norm": 0.4551903307437897,
      "learning_rate": 3.30550918196995e-05,
      "loss": 0.7974,
      "num_input_tokens_seen": 6876014,
      "step": 2506,
      "train_runtime": 6462.9258,
      "train_tokens_per_second": 1063.917
    },
    {
      "epoch": 0.20145041081580586,
      "grad_norm": 0.2582300007343292,
      "learning_rate": 3.298831385642738e-05,
      "loss": 0.6837,
      "num_input_tokens_seen": 6878138,
      "step": 2507,
      "train_runtime": 6464.7705,
      "train_tokens_per_second": 1063.942
    },
    {
      "epoch": 0.2015307659856566,
      "grad_norm": 0.29647937417030334,
      "learning_rate": 3.2921535893155256e-05,
      "loss": 1.0044,
      "num_input_tokens_seen": 6880764,
      "step": 2508,
      "train_runtime": 6466.9028,
      "train_tokens_per_second": 1063.997
    },
    {
      "epoch": 0.20161112115550733,
      "grad_norm": 0.2961001396179199,
      "learning_rate": 3.285475792988314e-05,
      "loss": 1.2751,
      "num_input_tokens_seen": 6884026,
      "step": 2509,
      "train_runtime": 6469.462,
      "train_tokens_per_second": 1064.08
    },
    {
      "epoch": 0.20169147632535808,
      "grad_norm": 0.3233042061328888,
      "learning_rate": 3.278797996661102e-05,
      "loss": 0.8876,
      "num_input_tokens_seen": 6886030,
      "step": 2510,
      "train_runtime": 6471.2494,
      "train_tokens_per_second": 1064.096
    },
    {
      "epoch": 0.20177183149520883,
      "grad_norm": 0.31158211827278137,
      "learning_rate": 3.27212020033389e-05,
      "loss": 0.8903,
      "num_input_tokens_seen": 6888694,
      "step": 2511,
      "train_runtime": 6473.4197,
      "train_tokens_per_second": 1064.151
    },
    {
      "epoch": 0.20185218666505955,
      "grad_norm": 0.35895296931266785,
      "learning_rate": 3.2654424040066775e-05,
      "loss": 1.1882,
      "num_input_tokens_seen": 6892362,
      "step": 2512,
      "train_runtime": 6476.0984,
      "train_tokens_per_second": 1064.277
    },
    {
      "epoch": 0.2019325418349103,
      "grad_norm": 0.3267019987106323,
      "learning_rate": 3.258764607679466e-05,
      "loss": 0.828,
      "num_input_tokens_seen": 6895408,
      "step": 2513,
      "train_runtime": 6478.4322,
      "train_tokens_per_second": 1064.364
    },
    {
      "epoch": 0.20201289700476105,
      "grad_norm": 0.3304426670074463,
      "learning_rate": 3.252086811352254e-05,
      "loss": 0.5703,
      "num_input_tokens_seen": 6899018,
      "step": 2514,
      "train_runtime": 6481.0315,
      "train_tokens_per_second": 1064.494
    },
    {
      "epoch": 0.20209325217461177,
      "grad_norm": 0.30019691586494446,
      "learning_rate": 3.245409015025042e-05,
      "loss": 0.9755,
      "num_input_tokens_seen": 6901990,
      "step": 2515,
      "train_runtime": 6483.3692,
      "train_tokens_per_second": 1064.568
    },
    {
      "epoch": 0.20217360734446252,
      "grad_norm": 0.3224036991596222,
      "learning_rate": 3.2387312186978295e-05,
      "loss": 1.1182,
      "num_input_tokens_seen": 6906232,
      "step": 2516,
      "train_runtime": 6486.3679,
      "train_tokens_per_second": 1064.73
    },
    {
      "epoch": 0.20225396251431327,
      "grad_norm": 0.28178849816322327,
      "learning_rate": 3.2320534223706175e-05,
      "loss": 1.2554,
      "num_input_tokens_seen": 6908974,
      "step": 2517,
      "train_runtime": 6488.505,
      "train_tokens_per_second": 1064.802
    },
    {
      "epoch": 0.20233431768416402,
      "grad_norm": 0.3748055696487427,
      "learning_rate": 3.225375626043406e-05,
      "loss": 1.0476,
      "num_input_tokens_seen": 6911370,
      "step": 2518,
      "train_runtime": 6490.4681,
      "train_tokens_per_second": 1064.849
    },
    {
      "epoch": 0.20241467285401474,
      "grad_norm": 0.3252844214439392,
      "learning_rate": 3.218697829716194e-05,
      "loss": 0.849,
      "num_input_tokens_seen": 6913494,
      "step": 2519,
      "train_runtime": 6492.316,
      "train_tokens_per_second": 1064.873
    },
    {
      "epoch": 0.2024950280238655,
      "grad_norm": 0.23035064339637756,
      "learning_rate": 3.2120200333889815e-05,
      "loss": 0.6645,
      "num_input_tokens_seen": 6916264,
      "step": 2520,
      "train_runtime": 6494.5213,
      "train_tokens_per_second": 1064.938
    },
    {
      "epoch": 0.20257538319371624,
      "grad_norm": 0.22935079038143158,
      "learning_rate": 3.2053422370617695e-05,
      "loss": 0.7179,
      "num_input_tokens_seen": 6919738,
      "step": 2521,
      "train_runtime": 6499.1175,
      "train_tokens_per_second": 1064.72
    },
    {
      "epoch": 0.20265573836356696,
      "grad_norm": 0.3111288249492645,
      "learning_rate": 3.198664440734558e-05,
      "loss": 0.683,
      "num_input_tokens_seen": 6922012,
      "step": 2522,
      "train_runtime": 6501.012,
      "train_tokens_per_second": 1064.759
    },
    {
      "epoch": 0.2027360935334177,
      "grad_norm": 0.3371312618255615,
      "learning_rate": 3.191986644407346e-05,
      "loss": 0.5878,
      "num_input_tokens_seen": 6924750,
      "step": 2523,
      "train_runtime": 6503.1509,
      "train_tokens_per_second": 1064.83
    },
    {
      "epoch": 0.20281644870326845,
      "grad_norm": 0.28429242968559265,
      "learning_rate": 3.1853088480801334e-05,
      "loss": 0.8481,
      "num_input_tokens_seen": 6926450,
      "step": 2524,
      "train_runtime": 6504.7521,
      "train_tokens_per_second": 1064.829
    },
    {
      "epoch": 0.20289680387311917,
      "grad_norm": 0.28056249022483826,
      "learning_rate": 3.1786310517529214e-05,
      "loss": 1.0278,
      "num_input_tokens_seen": 6929038,
      "step": 2525,
      "train_runtime": 6506.8304,
      "train_tokens_per_second": 1064.887
    },
    {
      "epoch": 0.20297715904296992,
      "grad_norm": 0.343074232339859,
      "learning_rate": 3.1719532554257094e-05,
      "loss": 0.8376,
      "num_input_tokens_seen": 6932302,
      "step": 2526,
      "train_runtime": 6509.3591,
      "train_tokens_per_second": 1064.975
    },
    {
      "epoch": 0.20305751421282067,
      "grad_norm": 0.3657798171043396,
      "learning_rate": 3.165275459098498e-05,
      "loss": 1.0568,
      "num_input_tokens_seen": 6934786,
      "step": 2527,
      "train_runtime": 6511.4316,
      "train_tokens_per_second": 1065.017
    },
    {
      "epoch": 0.20313786938267142,
      "grad_norm": 0.39345574378967285,
      "learning_rate": 3.1585976627712854e-05,
      "loss": 1.1339,
      "num_input_tokens_seen": 6936608,
      "step": 2528,
      "train_runtime": 6513.2353,
      "train_tokens_per_second": 1065.002
    },
    {
      "epoch": 0.20321822455252214,
      "grad_norm": 0.2972884476184845,
      "learning_rate": 3.1519198664440734e-05,
      "loss": 1.0111,
      "num_input_tokens_seen": 6938784,
      "step": 2529,
      "train_runtime": 6515.1279,
      "train_tokens_per_second": 1065.027
    },
    {
      "epoch": 0.2032985797223729,
      "grad_norm": 0.29959917068481445,
      "learning_rate": 3.1452420701168614e-05,
      "loss": 0.9685,
      "num_input_tokens_seen": 6941472,
      "step": 2530,
      "train_runtime": 6517.3115,
      "train_tokens_per_second": 1065.082
    },
    {
      "epoch": 0.20337893489222364,
      "grad_norm": 0.3754681944847107,
      "learning_rate": 3.13856427378965e-05,
      "loss": 1.0576,
      "num_input_tokens_seen": 6945244,
      "step": 2531,
      "train_runtime": 6520.0172,
      "train_tokens_per_second": 1065.219
    },
    {
      "epoch": 0.20345929006207436,
      "grad_norm": 0.30845966935157776,
      "learning_rate": 3.1318864774624374e-05,
      "loss": 0.7853,
      "num_input_tokens_seen": 6949136,
      "step": 2532,
      "train_runtime": 6522.803,
      "train_tokens_per_second": 1065.36
    },
    {
      "epoch": 0.2035396452319251,
      "grad_norm": 0.32504501938819885,
      "learning_rate": 3.1252086811352254e-05,
      "loss": 1.0266,
      "num_input_tokens_seen": 6951800,
      "step": 2533,
      "train_runtime": 6524.8404,
      "train_tokens_per_second": 1065.436
    },
    {
      "epoch": 0.20362000040177586,
      "grad_norm": 0.27619948983192444,
      "learning_rate": 3.1185308848080134e-05,
      "loss": 0.8903,
      "num_input_tokens_seen": 6955344,
      "step": 2534,
      "train_runtime": 6527.5022,
      "train_tokens_per_second": 1065.544
    },
    {
      "epoch": 0.20370035557162658,
      "grad_norm": 0.36974743008613586,
      "learning_rate": 3.111853088480802e-05,
      "loss": 1.168,
      "num_input_tokens_seen": 6958496,
      "step": 2535,
      "train_runtime": 6529.9366,
      "train_tokens_per_second": 1065.63
    },
    {
      "epoch": 0.20378071074147733,
      "grad_norm": 0.3140815198421478,
      "learning_rate": 3.1051752921535894e-05,
      "loss": 0.8182,
      "num_input_tokens_seen": 6961644,
      "step": 2536,
      "train_runtime": 6532.4645,
      "train_tokens_per_second": 1065.699
    },
    {
      "epoch": 0.20386106591132808,
      "grad_norm": 0.2374127209186554,
      "learning_rate": 3.0984974958263773e-05,
      "loss": 0.5429,
      "num_input_tokens_seen": 6963914,
      "step": 2537,
      "train_runtime": 6534.4251,
      "train_tokens_per_second": 1065.727
    },
    {
      "epoch": 0.2039414210811788,
      "grad_norm": 0.30528032779693604,
      "learning_rate": 3.0918196994991653e-05,
      "loss": 0.9504,
      "num_input_tokens_seen": 6967408,
      "step": 2538,
      "train_runtime": 6537.0058,
      "train_tokens_per_second": 1065.841
    },
    {
      "epoch": 0.20402177625102955,
      "grad_norm": 0.28372102975845337,
      "learning_rate": 3.085141903171953e-05,
      "loss": 1.0239,
      "num_input_tokens_seen": 6969096,
      "step": 2539,
      "train_runtime": 6538.6651,
      "train_tokens_per_second": 1065.829
    },
    {
      "epoch": 0.2041021314208803,
      "grad_norm": 0.2745300829410553,
      "learning_rate": 3.078464106844741e-05,
      "loss": 1.0085,
      "num_input_tokens_seen": 6970954,
      "step": 2540,
      "train_runtime": 6540.422,
      "train_tokens_per_second": 1065.826
    },
    {
      "epoch": 0.20418248659073104,
      "grad_norm": 0.4016948938369751,
      "learning_rate": 3.071786310517529e-05,
      "loss": 0.9755,
      "num_input_tokens_seen": 6973208,
      "step": 2541,
      "train_runtime": 6542.4188,
      "train_tokens_per_second": 1065.846
    },
    {
      "epoch": 0.20426284176058176,
      "grad_norm": 0.3127979040145874,
      "learning_rate": 3.065108514190317e-05,
      "loss": 0.8566,
      "num_input_tokens_seen": 6974720,
      "step": 2542,
      "train_runtime": 6544.0163,
      "train_tokens_per_second": 1065.816
    },
    {
      "epoch": 0.2043431969304325,
      "grad_norm": 0.2961050271987915,
      "learning_rate": 3.058430717863105e-05,
      "loss": 0.8619,
      "num_input_tokens_seen": 6976470,
      "step": 2543,
      "train_runtime": 6545.6652,
      "train_tokens_per_second": 1065.815
    },
    {
      "epoch": 0.20442355210028326,
      "grad_norm": 0.3067968785762787,
      "learning_rate": 3.051752921535893e-05,
      "loss": 0.9161,
      "num_input_tokens_seen": 6978776,
      "step": 2544,
      "train_runtime": 6547.6325,
      "train_tokens_per_second": 1065.847
    },
    {
      "epoch": 0.20450390727013398,
      "grad_norm": 0.29742324352264404,
      "learning_rate": 3.0450751252086813e-05,
      "loss": 1.2435,
      "num_input_tokens_seen": 6981800,
      "step": 2545,
      "train_runtime": 6549.961,
      "train_tokens_per_second": 1065.93
    },
    {
      "epoch": 0.20458426243998473,
      "grad_norm": 1.1142953634262085,
      "learning_rate": 3.0383973288814693e-05,
      "loss": 0.7926,
      "num_input_tokens_seen": 6983706,
      "step": 2546,
      "train_runtime": 6551.6653,
      "train_tokens_per_second": 1065.944
    },
    {
      "epoch": 0.20466461760983548,
      "grad_norm": 0.33697038888931274,
      "learning_rate": 3.0317195325542573e-05,
      "loss": 0.8405,
      "num_input_tokens_seen": 6986260,
      "step": 2547,
      "train_runtime": 6553.7381,
      "train_tokens_per_second": 1065.996
    },
    {
      "epoch": 0.2047449727796862,
      "grad_norm": 0.30040961503982544,
      "learning_rate": 3.025041736227045e-05,
      "loss": 0.7504,
      "num_input_tokens_seen": 6989338,
      "step": 2548,
      "train_runtime": 6556.1074,
      "train_tokens_per_second": 1066.08
    },
    {
      "epoch": 0.20482532794953695,
      "grad_norm": 0.28556397557258606,
      "learning_rate": 3.0183639398998333e-05,
      "loss": 0.8503,
      "num_input_tokens_seen": 6991216,
      "step": 2549,
      "train_runtime": 6557.8535,
      "train_tokens_per_second": 1066.083
    },
    {
      "epoch": 0.2049056831193877,
      "grad_norm": 0.42993345856666565,
      "learning_rate": 3.0116861435726213e-05,
      "loss": 0.975,
      "num_input_tokens_seen": 6994302,
      "step": 2550,
      "train_runtime": 6560.1569,
      "train_tokens_per_second": 1066.179
    },
    {
      "epoch": 0.20498603828923842,
      "grad_norm": 0.32767632603645325,
      "learning_rate": 3.0050083472454093e-05,
      "loss": 0.9443,
      "num_input_tokens_seen": 6996802,
      "step": 2551,
      "train_runtime": 6564.0704,
      "train_tokens_per_second": 1065.924
    },
    {
      "epoch": 0.20506639345908917,
      "grad_norm": 0.3201397657394409,
      "learning_rate": 2.998330550918197e-05,
      "loss": 0.8547,
      "num_input_tokens_seen": 6999148,
      "step": 2552,
      "train_runtime": 6566.0875,
      "train_tokens_per_second": 1065.954
    },
    {
      "epoch": 0.20514674862893992,
      "grad_norm": 0.2681242823600769,
      "learning_rate": 2.991652754590985e-05,
      "loss": 0.83,
      "num_input_tokens_seen": 7001700,
      "step": 2553,
      "train_runtime": 6568.2147,
      "train_tokens_per_second": 1065.997
    },
    {
      "epoch": 0.20522710379879067,
      "grad_norm": 0.281637579202652,
      "learning_rate": 2.9849749582637732e-05,
      "loss": 0.9952,
      "num_input_tokens_seen": 7005058,
      "step": 2554,
      "train_runtime": 6570.7853,
      "train_tokens_per_second": 1066.091
    },
    {
      "epoch": 0.2053074589686414,
      "grad_norm": 0.3066181242465973,
      "learning_rate": 2.9782971619365612e-05,
      "loss": 0.8452,
      "num_input_tokens_seen": 7006698,
      "step": 2555,
      "train_runtime": 6572.3758,
      "train_tokens_per_second": 1066.083
    },
    {
      "epoch": 0.20538781413849214,
      "grad_norm": 0.43828171491622925,
      "learning_rate": 2.971619365609349e-05,
      "loss": 1.0042,
      "num_input_tokens_seen": 7008880,
      "step": 2556,
      "train_runtime": 6574.282,
      "train_tokens_per_second": 1066.106
    },
    {
      "epoch": 0.20546816930834289,
      "grad_norm": 0.22972333431243896,
      "learning_rate": 2.964941569282137e-05,
      "loss": 0.9515,
      "num_input_tokens_seen": 7012440,
      "step": 2557,
      "train_runtime": 6576.882,
      "train_tokens_per_second": 1066.226
    },
    {
      "epoch": 0.2055485244781936,
      "grad_norm": 0.32284027338027954,
      "learning_rate": 2.9582637729549252e-05,
      "loss": 0.9505,
      "num_input_tokens_seen": 7015090,
      "step": 2558,
      "train_runtime": 6579.0679,
      "train_tokens_per_second": 1066.274
    },
    {
      "epoch": 0.20562887964804435,
      "grad_norm": 0.2418496161699295,
      "learning_rate": 2.9515859766277132e-05,
      "loss": 0.8734,
      "num_input_tokens_seen": 7018630,
      "step": 2559,
      "train_runtime": 6581.6306,
      "train_tokens_per_second": 1066.397
    },
    {
      "epoch": 0.2057092348178951,
      "grad_norm": 0.31431302428245544,
      "learning_rate": 2.944908180300501e-05,
      "loss": 1.0124,
      "num_input_tokens_seen": 7022046,
      "step": 2560,
      "train_runtime": 6584.2637,
      "train_tokens_per_second": 1066.489
    },
    {
      "epoch": 0.20578958998774582,
      "grad_norm": 0.32687124609947205,
      "learning_rate": 2.938230383973289e-05,
      "loss": 0.828,
      "num_input_tokens_seen": 7024366,
      "step": 2561,
      "train_runtime": 6586.2513,
      "train_tokens_per_second": 1066.52
    },
    {
      "epoch": 0.20586994515759657,
      "grad_norm": 0.2755314111709595,
      "learning_rate": 2.931552587646077e-05,
      "loss": 0.75,
      "num_input_tokens_seen": 7026568,
      "step": 2562,
      "train_runtime": 6588.1639,
      "train_tokens_per_second": 1066.544
    },
    {
      "epoch": 0.20595030032744732,
      "grad_norm": 0.311623215675354,
      "learning_rate": 2.924874791318865e-05,
      "loss": 1.0934,
      "num_input_tokens_seen": 7032308,
      "step": 2563,
      "train_runtime": 6592.2321,
      "train_tokens_per_second": 1066.757
    },
    {
      "epoch": 0.20603065549729807,
      "grad_norm": 0.2689889073371887,
      "learning_rate": 2.9181969949916528e-05,
      "loss": 0.9113,
      "num_input_tokens_seen": 7034958,
      "step": 2564,
      "train_runtime": 6594.4275,
      "train_tokens_per_second": 1066.803
    },
    {
      "epoch": 0.2061110106671488,
      "grad_norm": 0.338100403547287,
      "learning_rate": 2.9115191986644408e-05,
      "loss": 0.9602,
      "num_input_tokens_seen": 7037022,
      "step": 2565,
      "train_runtime": 6596.1829,
      "train_tokens_per_second": 1066.832
    },
    {
      "epoch": 0.20619136583699954,
      "grad_norm": 0.3333339989185333,
      "learning_rate": 2.9048414023372288e-05,
      "loss": 1.2499,
      "num_input_tokens_seen": 7039262,
      "step": 2566,
      "train_runtime": 6598.0852,
      "train_tokens_per_second": 1066.864
    },
    {
      "epoch": 0.2062717210068503,
      "grad_norm": 0.2797660231590271,
      "learning_rate": 2.898163606010017e-05,
      "loss": 1.0527,
      "num_input_tokens_seen": 7041886,
      "step": 2567,
      "train_runtime": 6600.2254,
      "train_tokens_per_second": 1066.916
    },
    {
      "epoch": 0.206352076176701,
      "grad_norm": 0.21665260195732117,
      "learning_rate": 2.8914858096828045e-05,
      "loss": 0.4836,
      "num_input_tokens_seen": 7048516,
      "step": 2568,
      "train_runtime": 6604.6541,
      "train_tokens_per_second": 1067.204
    },
    {
      "epoch": 0.20643243134655176,
      "grad_norm": 0.38138285279273987,
      "learning_rate": 2.8848080133555928e-05,
      "loss": 1.0912,
      "num_input_tokens_seen": 7051190,
      "step": 2569,
      "train_runtime": 6606.8892,
      "train_tokens_per_second": 1067.248
    },
    {
      "epoch": 0.2065127865164025,
      "grad_norm": 0.40003708004951477,
      "learning_rate": 2.8781302170283808e-05,
      "loss": 1.2075,
      "num_input_tokens_seen": 7052824,
      "step": 2570,
      "train_runtime": 6608.4814,
      "train_tokens_per_second": 1067.238
    },
    {
      "epoch": 0.20659314168625323,
      "grad_norm": 0.298174649477005,
      "learning_rate": 2.8714524207011688e-05,
      "loss": 0.8016,
      "num_input_tokens_seen": 7056034,
      "step": 2571,
      "train_runtime": 6611.0225,
      "train_tokens_per_second": 1067.314
    },
    {
      "epoch": 0.20667349685610398,
      "grad_norm": 0.2843630611896515,
      "learning_rate": 2.8647746243739564e-05,
      "loss": 0.7817,
      "num_input_tokens_seen": 7059210,
      "step": 2572,
      "train_runtime": 6613.5279,
      "train_tokens_per_second": 1067.389
    },
    {
      "epoch": 0.20675385202595473,
      "grad_norm": 0.3161008358001709,
      "learning_rate": 2.8580968280467448e-05,
      "loss": 0.9338,
      "num_input_tokens_seen": 7061704,
      "step": 2573,
      "train_runtime": 6615.5676,
      "train_tokens_per_second": 1067.437
    },
    {
      "epoch": 0.20683420719580545,
      "grad_norm": 0.3127140998840332,
      "learning_rate": 2.8514190317195328e-05,
      "loss": 0.65,
      "num_input_tokens_seen": 7063208,
      "step": 2574,
      "train_runtime": 6617.2057,
      "train_tokens_per_second": 1067.4
    },
    {
      "epoch": 0.2069145623656562,
      "grad_norm": 0.45175424218177795,
      "learning_rate": 2.8447412353923207e-05,
      "loss": 0.8879,
      "num_input_tokens_seen": 7066348,
      "step": 2575,
      "train_runtime": 6619.6717,
      "train_tokens_per_second": 1067.477
    },
    {
      "epoch": 0.20699491753550694,
      "grad_norm": 0.3343769311904907,
      "learning_rate": 2.8380634390651084e-05,
      "loss": 1.0982,
      "num_input_tokens_seen": 7067970,
      "step": 2576,
      "train_runtime": 6621.2544,
      "train_tokens_per_second": 1067.467
    },
    {
      "epoch": 0.2070752727053577,
      "grad_norm": 0.3058067560195923,
      "learning_rate": 2.8313856427378964e-05,
      "loss": 0.871,
      "num_input_tokens_seen": 7070704,
      "step": 2577,
      "train_runtime": 6623.3464,
      "train_tokens_per_second": 1067.543
    },
    {
      "epoch": 0.20715562787520841,
      "grad_norm": 0.31188082695007324,
      "learning_rate": 2.8247078464106847e-05,
      "loss": 0.8766,
      "num_input_tokens_seen": 7074054,
      "step": 2578,
      "train_runtime": 6625.8045,
      "train_tokens_per_second": 1067.652
    },
    {
      "epoch": 0.20723598304505916,
      "grad_norm": 0.33825647830963135,
      "learning_rate": 2.8180300500834727e-05,
      "loss": 0.875,
      "num_input_tokens_seen": 7076172,
      "step": 2579,
      "train_runtime": 6627.6876,
      "train_tokens_per_second": 1067.668
    },
    {
      "epoch": 0.2073163382149099,
      "grad_norm": 0.28344404697418213,
      "learning_rate": 2.8113522537562604e-05,
      "loss": 1.0178,
      "num_input_tokens_seen": 7078598,
      "step": 2580,
      "train_runtime": 6629.8123,
      "train_tokens_per_second": 1067.692
    },
    {
      "epoch": 0.20739669338476063,
      "grad_norm": 0.34717726707458496,
      "learning_rate": 2.8046744574290484e-05,
      "loss": 0.9004,
      "num_input_tokens_seen": 7081574,
      "step": 2581,
      "train_runtime": 6633.8341,
      "train_tokens_per_second": 1067.493
    },
    {
      "epoch": 0.20747704855461138,
      "grad_norm": 0.33153900504112244,
      "learning_rate": 2.7979966611018367e-05,
      "loss": 1.059,
      "num_input_tokens_seen": 7085034,
      "step": 2582,
      "train_runtime": 6636.4238,
      "train_tokens_per_second": 1067.598
    },
    {
      "epoch": 0.20755740372446213,
      "grad_norm": 0.24847878515720367,
      "learning_rate": 2.7913188647746247e-05,
      "loss": 0.9254,
      "num_input_tokens_seen": 7087960,
      "step": 2583,
      "train_runtime": 6638.6533,
      "train_tokens_per_second": 1067.68
    },
    {
      "epoch": 0.20763775889431285,
      "grad_norm": 0.3006184995174408,
      "learning_rate": 2.7846410684474123e-05,
      "loss": 0.9193,
      "num_input_tokens_seen": 7091708,
      "step": 2584,
      "train_runtime": 6641.4229,
      "train_tokens_per_second": 1067.799
    },
    {
      "epoch": 0.2077181140641636,
      "grad_norm": 0.24109326303005219,
      "learning_rate": 2.7779632721202003e-05,
      "loss": 0.811,
      "num_input_tokens_seen": 7097692,
      "step": 2585,
      "train_runtime": 6645.4647,
      "train_tokens_per_second": 1068.051
    },
    {
      "epoch": 0.20779846923401435,
      "grad_norm": 0.27665063738822937,
      "learning_rate": 2.7712854757929883e-05,
      "loss": 0.9955,
      "num_input_tokens_seen": 7102658,
      "step": 2586,
      "train_runtime": 6648.9607,
      "train_tokens_per_second": 1068.236
    },
    {
      "epoch": 0.20787882440386507,
      "grad_norm": 0.30551180243492126,
      "learning_rate": 2.7646076794657767e-05,
      "loss": 1.0939,
      "num_input_tokens_seen": 7105280,
      "step": 2587,
      "train_runtime": 6651.1024,
      "train_tokens_per_second": 1068.286
    },
    {
      "epoch": 0.20795917957371582,
      "grad_norm": 0.34263044595718384,
      "learning_rate": 2.7579298831385643e-05,
      "loss": 1.0195,
      "num_input_tokens_seen": 7108260,
      "step": 2588,
      "train_runtime": 6653.442,
      "train_tokens_per_second": 1068.358
    },
    {
      "epoch": 0.20803953474356657,
      "grad_norm": 0.31950539350509644,
      "learning_rate": 2.7512520868113523e-05,
      "loss": 1.1237,
      "num_input_tokens_seen": 7113474,
      "step": 2589,
      "train_runtime": 6657.0473,
      "train_tokens_per_second": 1068.563
    },
    {
      "epoch": 0.20811988991341732,
      "grad_norm": 0.3273375928401947,
      "learning_rate": 2.7445742904841403e-05,
      "loss": 1.044,
      "num_input_tokens_seen": 7116646,
      "step": 2590,
      "train_runtime": 6659.5073,
      "train_tokens_per_second": 1068.645
    },
    {
      "epoch": 0.20820024508326804,
      "grad_norm": 0.4363780617713928,
      "learning_rate": 2.7378964941569286e-05,
      "loss": 0.9577,
      "num_input_tokens_seen": 7118516,
      "step": 2591,
      "train_runtime": 6661.2619,
      "train_tokens_per_second": 1068.644
    },
    {
      "epoch": 0.20828060025311879,
      "grad_norm": 0.2938719391822815,
      "learning_rate": 2.731218697829716e-05,
      "loss": 1.2035,
      "num_input_tokens_seen": 7121542,
      "step": 2592,
      "train_runtime": 6663.57,
      "train_tokens_per_second": 1068.728
    },
    {
      "epoch": 0.20836095542296953,
      "grad_norm": 0.32434898614883423,
      "learning_rate": 2.7245409015025043e-05,
      "loss": 0.9278,
      "num_input_tokens_seen": 7124434,
      "step": 2593,
      "train_runtime": 6665.8409,
      "train_tokens_per_second": 1068.797
    },
    {
      "epoch": 0.20844131059282026,
      "grad_norm": 0.34501877427101135,
      "learning_rate": 2.7178631051752923e-05,
      "loss": 0.5575,
      "num_input_tokens_seen": 7127106,
      "step": 2594,
      "train_runtime": 6667.9953,
      "train_tokens_per_second": 1068.853
    },
    {
      "epoch": 0.208521665762671,
      "grad_norm": 0.2891291081905365,
      "learning_rate": 2.7111853088480803e-05,
      "loss": 0.7796,
      "num_input_tokens_seen": 7129588,
      "step": 2595,
      "train_runtime": 6670.0028,
      "train_tokens_per_second": 1068.903
    },
    {
      "epoch": 0.20860202093252175,
      "grad_norm": 0.3389528691768646,
      "learning_rate": 2.704507512520868e-05,
      "loss": 1.2186,
      "num_input_tokens_seen": 7131818,
      "step": 2596,
      "train_runtime": 6671.9101,
      "train_tokens_per_second": 1068.932
    },
    {
      "epoch": 0.20868237610237247,
      "grad_norm": 0.331019788980484,
      "learning_rate": 2.6978297161936563e-05,
      "loss": 0.7591,
      "num_input_tokens_seen": 7134266,
      "step": 2597,
      "train_runtime": 6673.9387,
      "train_tokens_per_second": 1068.974
    },
    {
      "epoch": 0.20876273127222322,
      "grad_norm": 0.5408254265785217,
      "learning_rate": 2.6911519198664442e-05,
      "loss": 0.74,
      "num_input_tokens_seen": 7137480,
      "step": 2598,
      "train_runtime": 6676.368,
      "train_tokens_per_second": 1069.066
    },
    {
      "epoch": 0.20884308644207397,
      "grad_norm": 0.4499052166938782,
      "learning_rate": 2.6844741235392322e-05,
      "loss": 0.8022,
      "num_input_tokens_seen": 7139864,
      "step": 2599,
      "train_runtime": 6678.4226,
      "train_tokens_per_second": 1069.094
    },
    {
      "epoch": 0.20892344161192472,
      "grad_norm": 0.3211537003517151,
      "learning_rate": 2.67779632721202e-05,
      "loss": 1.0694,
      "num_input_tokens_seen": 7143004,
      "step": 2600,
      "train_runtime": 6680.8664,
      "train_tokens_per_second": 1069.173
    },
    {
      "epoch": 0.20900379678177544,
      "grad_norm": 0.32943010330200195,
      "learning_rate": 2.671118530884808e-05,
      "loss": 1.2333,
      "num_input_tokens_seen": 7147510,
      "step": 2601,
      "train_runtime": 6684.1297,
      "train_tokens_per_second": 1069.325
    },
    {
      "epoch": 0.2090841519516262,
      "grad_norm": 0.29366013407707214,
      "learning_rate": 2.6644407345575962e-05,
      "loss": 0.97,
      "num_input_tokens_seen": 7150262,
      "step": 2602,
      "train_runtime": 6686.3203,
      "train_tokens_per_second": 1069.387
    },
    {
      "epoch": 0.20916450712147694,
      "grad_norm": 0.42908358573913574,
      "learning_rate": 2.6577629382303842e-05,
      "loss": 0.7171,
      "num_input_tokens_seen": 7153274,
      "step": 2603,
      "train_runtime": 6688.6324,
      "train_tokens_per_second": 1069.467
    },
    {
      "epoch": 0.20924486229132766,
      "grad_norm": 0.2905648946762085,
      "learning_rate": 2.651085141903172e-05,
      "loss": 0.6807,
      "num_input_tokens_seen": 7159390,
      "step": 2604,
      "train_runtime": 6692.8467,
      "train_tokens_per_second": 1069.708
    },
    {
      "epoch": 0.2093252174611784,
      "grad_norm": 0.36995115876197815,
      "learning_rate": 2.64440734557596e-05,
      "loss": 0.9172,
      "num_input_tokens_seen": 7161184,
      "step": 2605,
      "train_runtime": 6694.5556,
      "train_tokens_per_second": 1069.703
    },
    {
      "epoch": 0.20940557263102916,
      "grad_norm": 0.3697642683982849,
      "learning_rate": 2.6377295492487482e-05,
      "loss": 1.1028,
      "num_input_tokens_seen": 7163334,
      "step": 2606,
      "train_runtime": 6696.4025,
      "train_tokens_per_second": 1069.729
    },
    {
      "epoch": 0.20948592780087988,
      "grad_norm": 0.2805689573287964,
      "learning_rate": 2.6310517529215362e-05,
      "loss": 0.9916,
      "num_input_tokens_seen": 7165998,
      "step": 2607,
      "train_runtime": 6698.5314,
      "train_tokens_per_second": 1069.786
    },
    {
      "epoch": 0.20956628297073063,
      "grad_norm": 0.34024694561958313,
      "learning_rate": 2.624373956594324e-05,
      "loss": 0.9108,
      "num_input_tokens_seen": 7169338,
      "step": 2608,
      "train_runtime": 6700.9744,
      "train_tokens_per_second": 1069.895
    },
    {
      "epoch": 0.20964663814058138,
      "grad_norm": 0.34759029746055603,
      "learning_rate": 2.6176961602671118e-05,
      "loss": 0.8615,
      "num_input_tokens_seen": 7171878,
      "step": 2609,
      "train_runtime": 6703.0758,
      "train_tokens_per_second": 1069.938
    },
    {
      "epoch": 0.2097269933104321,
      "grad_norm": 0.29292887449264526,
      "learning_rate": 2.6110183639398998e-05,
      "loss": 0.9368,
      "num_input_tokens_seen": 7175876,
      "step": 2610,
      "train_runtime": 6705.9241,
      "train_tokens_per_second": 1070.08
    },
    {
      "epoch": 0.20980734848028285,
      "grad_norm": 0.35517358779907227,
      "learning_rate": 2.604340567612688e-05,
      "loss": 1.0702,
      "num_input_tokens_seen": 7178460,
      "step": 2611,
      "train_runtime": 6709.9519,
      "train_tokens_per_second": 1069.823
    },
    {
      "epoch": 0.2098877036501336,
      "grad_norm": 0.391550749540329,
      "learning_rate": 2.597662771285476e-05,
      "loss": 0.7596,
      "num_input_tokens_seen": 7180238,
      "step": 2612,
      "train_runtime": 6711.7046,
      "train_tokens_per_second": 1069.808
    },
    {
      "epoch": 0.20996805881998434,
      "grad_norm": 0.29763591289520264,
      "learning_rate": 2.5909849749582638e-05,
      "loss": 0.9017,
      "num_input_tokens_seen": 7183710,
      "step": 2613,
      "train_runtime": 6714.2835,
      "train_tokens_per_second": 1069.915
    },
    {
      "epoch": 0.21004841398983506,
      "grad_norm": 0.35687410831451416,
      "learning_rate": 2.5843071786310518e-05,
      "loss": 1.3214,
      "num_input_tokens_seen": 7187950,
      "step": 2614,
      "train_runtime": 6717.2198,
      "train_tokens_per_second": 1070.078
    },
    {
      "epoch": 0.2101287691596858,
      "grad_norm": 0.27503442764282227,
      "learning_rate": 2.57762938230384e-05,
      "loss": 0.8821,
      "num_input_tokens_seen": 7190992,
      "step": 2615,
      "train_runtime": 6719.5529,
      "train_tokens_per_second": 1070.159
    },
    {
      "epoch": 0.21020912432953656,
      "grad_norm": 0.33326447010040283,
      "learning_rate": 2.570951585976628e-05,
      "loss": 1.1317,
      "num_input_tokens_seen": 7193508,
      "step": 2616,
      "train_runtime": 6721.5633,
      "train_tokens_per_second": 1070.214
    },
    {
      "epoch": 0.21028947949938728,
      "grad_norm": 0.4501754641532898,
      "learning_rate": 2.5642737896494158e-05,
      "loss": 0.7386,
      "num_input_tokens_seen": 7195122,
      "step": 2617,
      "train_runtime": 6723.1615,
      "train_tokens_per_second": 1070.199
    },
    {
      "epoch": 0.21036983466923803,
      "grad_norm": 0.25743284821510315,
      "learning_rate": 2.5575959933222038e-05,
      "loss": 0.7228,
      "num_input_tokens_seen": 7199788,
      "step": 2618,
      "train_runtime": 6726.5604,
      "train_tokens_per_second": 1070.352
    },
    {
      "epoch": 0.21045018983908878,
      "grad_norm": 0.37954598665237427,
      "learning_rate": 2.5509181969949918e-05,
      "loss": 1.0888,
      "num_input_tokens_seen": 7201342,
      "step": 2619,
      "train_runtime": 6728.1832,
      "train_tokens_per_second": 1070.325
    },
    {
      "epoch": 0.2105305450089395,
      "grad_norm": 0.2765674293041229,
      "learning_rate": 2.54424040066778e-05,
      "loss": 0.8129,
      "num_input_tokens_seen": 7204266,
      "step": 2620,
      "train_runtime": 6730.4785,
      "train_tokens_per_second": 1070.394
    },
    {
      "epoch": 0.21061090017879025,
      "grad_norm": 0.3497758209705353,
      "learning_rate": 2.5375626043405677e-05,
      "loss": 1.0757,
      "num_input_tokens_seen": 7206274,
      "step": 2621,
      "train_runtime": 6732.3507,
      "train_tokens_per_second": 1070.395
    },
    {
      "epoch": 0.210691255348641,
      "grad_norm": 0.3260670304298401,
      "learning_rate": 2.5308848080133557e-05,
      "loss": 1.0409,
      "num_input_tokens_seen": 7208778,
      "step": 2622,
      "train_runtime": 6734.3594,
      "train_tokens_per_second": 1070.447
    },
    {
      "epoch": 0.21077161051849172,
      "grad_norm": 0.3261588513851166,
      "learning_rate": 2.5242070116861437e-05,
      "loss": 0.9688,
      "num_input_tokens_seen": 7211160,
      "step": 2623,
      "train_runtime": 6736.4843,
      "train_tokens_per_second": 1070.463
    },
    {
      "epoch": 0.21085196568834247,
      "grad_norm": 0.332161009311676,
      "learning_rate": 2.517529215358932e-05,
      "loss": 0.7956,
      "num_input_tokens_seen": 7213212,
      "step": 2624,
      "train_runtime": 6738.3343,
      "train_tokens_per_second": 1070.474
    },
    {
      "epoch": 0.21093232085819322,
      "grad_norm": 0.31648534536361694,
      "learning_rate": 2.5108514190317194e-05,
      "loss": 0.9231,
      "num_input_tokens_seen": 7216274,
      "step": 2625,
      "train_runtime": 6740.7297,
      "train_tokens_per_second": 1070.548
    },
    {
      "epoch": 0.21101267602804397,
      "grad_norm": 0.3199121654033661,
      "learning_rate": 2.5041736227045077e-05,
      "loss": 0.8654,
      "num_input_tokens_seen": 7219034,
      "step": 2626,
      "train_runtime": 6742.9141,
      "train_tokens_per_second": 1070.61
    },
    {
      "epoch": 0.2110930311978947,
      "grad_norm": 0.36725541949272156,
      "learning_rate": 2.4974958263772957e-05,
      "loss": 0.8796,
      "num_input_tokens_seen": 7221744,
      "step": 2627,
      "train_runtime": 6745.1631,
      "train_tokens_per_second": 1070.655
    },
    {
      "epoch": 0.21117338636774544,
      "grad_norm": 0.26133593916893005,
      "learning_rate": 2.4908180300500837e-05,
      "loss": 0.851,
      "num_input_tokens_seen": 7223944,
      "step": 2628,
      "train_runtime": 6747.1141,
      "train_tokens_per_second": 1070.672
    },
    {
      "epoch": 0.21125374153759618,
      "grad_norm": 0.24963918328285217,
      "learning_rate": 2.4841402337228717e-05,
      "loss": 1.0526,
      "num_input_tokens_seen": 7228004,
      "step": 2629,
      "train_runtime": 6749.9509,
      "train_tokens_per_second": 1070.823
    },
    {
      "epoch": 0.2113340967074469,
      "grad_norm": 0.44179970026016235,
      "learning_rate": 2.4774624373956597e-05,
      "loss": 0.8394,
      "num_input_tokens_seen": 7230254,
      "step": 2630,
      "train_runtime": 6751.9408,
      "train_tokens_per_second": 1070.841
    },
    {
      "epoch": 0.21141445187729765,
      "grad_norm": 0.45255523920059204,
      "learning_rate": 2.4707846410684477e-05,
      "loss": 1.1112,
      "num_input_tokens_seen": 7232764,
      "step": 2631,
      "train_runtime": 6753.985,
      "train_tokens_per_second": 1070.888
    },
    {
      "epoch": 0.2114948070471484,
      "grad_norm": 0.3996327221393585,
      "learning_rate": 2.4641068447412353e-05,
      "loss": 1.0531,
      "num_input_tokens_seen": 7235238,
      "step": 2632,
      "train_runtime": 6756.0367,
      "train_tokens_per_second": 1070.929
    },
    {
      "epoch": 0.21157516221699912,
      "grad_norm": 0.38754984736442566,
      "learning_rate": 2.4574290484140237e-05,
      "loss": 1.175,
      "num_input_tokens_seen": 7238556,
      "step": 2633,
      "train_runtime": 6758.5238,
      "train_tokens_per_second": 1071.026
    },
    {
      "epoch": 0.21165551738684987,
      "grad_norm": 0.2536124587059021,
      "learning_rate": 2.4507512520868113e-05,
      "loss": 0.8024,
      "num_input_tokens_seen": 7242192,
      "step": 2634,
      "train_runtime": 6761.2341,
      "train_tokens_per_second": 1071.135
    },
    {
      "epoch": 0.21173587255670062,
      "grad_norm": 0.35031622648239136,
      "learning_rate": 2.4440734557595996e-05,
      "loss": 1.1753,
      "num_input_tokens_seen": 7244398,
      "step": 2635,
      "train_runtime": 6763.1332,
      "train_tokens_per_second": 1071.16
    },
    {
      "epoch": 0.21181622772655137,
      "grad_norm": 0.3115178346633911,
      "learning_rate": 2.4373956594323873e-05,
      "loss": 1.0367,
      "num_input_tokens_seen": 7248448,
      "step": 2636,
      "train_runtime": 6766.02,
      "train_tokens_per_second": 1071.302
    },
    {
      "epoch": 0.2118965828964021,
      "grad_norm": 0.3466339707374573,
      "learning_rate": 2.4307178631051756e-05,
      "loss": 0.6242,
      "num_input_tokens_seen": 7251678,
      "step": 2637,
      "train_runtime": 6768.4347,
      "train_tokens_per_second": 1071.397
    },
    {
      "epoch": 0.21197693806625284,
      "grad_norm": 0.3892398774623871,
      "learning_rate": 2.4240400667779633e-05,
      "loss": 0.7594,
      "num_input_tokens_seen": 7255652,
      "step": 2638,
      "train_runtime": 6771.2308,
      "train_tokens_per_second": 1071.541
    },
    {
      "epoch": 0.2120572932361036,
      "grad_norm": 0.2880546450614929,
      "learning_rate": 2.4173622704507516e-05,
      "loss": 0.6689,
      "num_input_tokens_seen": 7258814,
      "step": 2639,
      "train_runtime": 6773.595,
      "train_tokens_per_second": 1071.634
    },
    {
      "epoch": 0.2121376484059543,
      "grad_norm": 0.33683693408966064,
      "learning_rate": 2.4106844741235393e-05,
      "loss": 0.9005,
      "num_input_tokens_seen": 7261264,
      "step": 2640,
      "train_runtime": 6775.73,
      "train_tokens_per_second": 1071.658
    },
    {
      "epoch": 0.21221800357580506,
      "grad_norm": 0.3389016389846802,
      "learning_rate": 2.4040066777963273e-05,
      "loss": 1.077,
      "num_input_tokens_seen": 7263852,
      "step": 2641,
      "train_runtime": 6779.7983,
      "train_tokens_per_second": 1071.396
    },
    {
      "epoch": 0.2122983587456558,
      "grad_norm": 0.35491853952407837,
      "learning_rate": 2.3973288814691153e-05,
      "loss": 0.9035,
      "num_input_tokens_seen": 7265538,
      "step": 2642,
      "train_runtime": 6781.5295,
      "train_tokens_per_second": 1071.372
    },
    {
      "epoch": 0.21237871391550653,
      "grad_norm": 0.38348373770713806,
      "learning_rate": 2.3906510851419033e-05,
      "loss": 0.935,
      "num_input_tokens_seen": 7268478,
      "step": 2643,
      "train_runtime": 6783.9147,
      "train_tokens_per_second": 1071.428
    },
    {
      "epoch": 0.21245906908535728,
      "grad_norm": 0.3012312352657318,
      "learning_rate": 2.3839732888146912e-05,
      "loss": 0.9347,
      "num_input_tokens_seen": 7270988,
      "step": 2644,
      "train_runtime": 6786.0063,
      "train_tokens_per_second": 1071.468
    },
    {
      "epoch": 0.21253942425520803,
      "grad_norm": 0.25290215015411377,
      "learning_rate": 2.3772954924874792e-05,
      "loss": 0.8887,
      "num_input_tokens_seen": 7273678,
      "step": 2645,
      "train_runtime": 6788.2374,
      "train_tokens_per_second": 1071.512
    },
    {
      "epoch": 0.21261977942505875,
      "grad_norm": 0.29149237275123596,
      "learning_rate": 2.3706176961602672e-05,
      "loss": 0.9653,
      "num_input_tokens_seen": 7275446,
      "step": 2646,
      "train_runtime": 6790.0114,
      "train_tokens_per_second": 1071.492
    },
    {
      "epoch": 0.2127001345949095,
      "grad_norm": 0.30114826560020447,
      "learning_rate": 2.3639398998330552e-05,
      "loss": 1.0161,
      "num_input_tokens_seen": 7277926,
      "step": 2647,
      "train_runtime": 6792.107,
      "train_tokens_per_second": 1071.527
    },
    {
      "epoch": 0.21278048976476024,
      "grad_norm": 0.3214522898197174,
      "learning_rate": 2.3572621035058432e-05,
      "loss": 1.1982,
      "num_input_tokens_seen": 7281208,
      "step": 2648,
      "train_runtime": 6794.5576,
      "train_tokens_per_second": 1071.624
    },
    {
      "epoch": 0.212860844934611,
      "grad_norm": 0.43071138858795166,
      "learning_rate": 2.3505843071786312e-05,
      "loss": 0.9475,
      "num_input_tokens_seen": 7283412,
      "step": 2649,
      "train_runtime": 6796.5115,
      "train_tokens_per_second": 1071.64
    },
    {
      "epoch": 0.21294120010446171,
      "grad_norm": 0.39023107290267944,
      "learning_rate": 2.3439065108514192e-05,
      "loss": 1.2034,
      "num_input_tokens_seen": 7285842,
      "step": 2650,
      "train_runtime": 6798.5606,
      "train_tokens_per_second": 1071.674
    },
    {
      "epoch": 0.21302155527431246,
      "grad_norm": 0.35178694128990173,
      "learning_rate": 2.3372287145242072e-05,
      "loss": 1.1156,
      "num_input_tokens_seen": 7288624,
      "step": 2651,
      "train_runtime": 6800.7516,
      "train_tokens_per_second": 1071.738
    },
    {
      "epoch": 0.2131019104441632,
      "grad_norm": 0.3110933005809784,
      "learning_rate": 2.3305509181969952e-05,
      "loss": 0.9604,
      "num_input_tokens_seen": 7292628,
      "step": 2652,
      "train_runtime": 6803.6143,
      "train_tokens_per_second": 1071.876
    },
    {
      "epoch": 0.21318226561401393,
      "grad_norm": 0.2999524176120758,
      "learning_rate": 2.3238731218697832e-05,
      "loss": 1.0385,
      "num_input_tokens_seen": 7295632,
      "step": 2653,
      "train_runtime": 6805.9416,
      "train_tokens_per_second": 1071.95
    },
    {
      "epoch": 0.21326262078386468,
      "grad_norm": 0.31076672673225403,
      "learning_rate": 2.3171953255425712e-05,
      "loss": 0.887,
      "num_input_tokens_seen": 7298492,
      "step": 2654,
      "train_runtime": 6808.302,
      "train_tokens_per_second": 1071.999
    },
    {
      "epoch": 0.21334297595371543,
      "grad_norm": 0.2822515070438385,
      "learning_rate": 2.310517529215359e-05,
      "loss": 0.7181,
      "num_input_tokens_seen": 7302234,
      "step": 2655,
      "train_runtime": 6811.0159,
      "train_tokens_per_second": 1072.121
    },
    {
      "epoch": 0.21342333112356615,
      "grad_norm": 0.29139649868011475,
      "learning_rate": 2.3038397328881468e-05,
      "loss": 0.7603,
      "num_input_tokens_seen": 7303838,
      "step": 2656,
      "train_runtime": 6812.6968,
      "train_tokens_per_second": 1072.092
    },
    {
      "epoch": 0.2135036862934169,
      "grad_norm": 0.34181493520736694,
      "learning_rate": 2.297161936560935e-05,
      "loss": 1.024,
      "num_input_tokens_seen": 7307208,
      "step": 2657,
      "train_runtime": 6815.1543,
      "train_tokens_per_second": 1072.2
    },
    {
      "epoch": 0.21358404146326765,
      "grad_norm": 0.30981960892677307,
      "learning_rate": 2.2904841402337228e-05,
      "loss": 0.9214,
      "num_input_tokens_seen": 7309716,
      "step": 2658,
      "train_runtime": 6817.2121,
      "train_tokens_per_second": 1072.244
    },
    {
      "epoch": 0.21366439663311837,
      "grad_norm": 0.3379795253276825,
      "learning_rate": 2.283806343906511e-05,
      "loss": 0.9486,
      "num_input_tokens_seen": 7315028,
      "step": 2659,
      "train_runtime": 6820.9329,
      "train_tokens_per_second": 1072.438
    },
    {
      "epoch": 0.21374475180296912,
      "grad_norm": 0.3424399495124817,
      "learning_rate": 2.2771285475792988e-05,
      "loss": 1.1573,
      "num_input_tokens_seen": 7316896,
      "step": 2660,
      "train_runtime": 6822.7476,
      "train_tokens_per_second": 1072.427
    },
    {
      "epoch": 0.21382510697281987,
      "grad_norm": 0.3007705509662628,
      "learning_rate": 2.270450751252087e-05,
      "loss": 1.3266,
      "num_input_tokens_seen": 7319546,
      "step": 2661,
      "train_runtime": 6824.8194,
      "train_tokens_per_second": 1072.489
    },
    {
      "epoch": 0.21390546214267062,
      "grad_norm": 0.22102737426757812,
      "learning_rate": 2.2637729549248748e-05,
      "loss": 0.4628,
      "num_input_tokens_seen": 7321902,
      "step": 2662,
      "train_runtime": 6826.7466,
      "train_tokens_per_second": 1072.532
    },
    {
      "epoch": 0.21398581731252134,
      "grad_norm": 0.32190918922424316,
      "learning_rate": 2.257095158597663e-05,
      "loss": 0.9612,
      "num_input_tokens_seen": 7324602,
      "step": 2663,
      "train_runtime": 6828.9039,
      "train_tokens_per_second": 1072.588
    },
    {
      "epoch": 0.21406617248237209,
      "grad_norm": 0.3150859773159027,
      "learning_rate": 2.2504173622704508e-05,
      "loss": 0.8958,
      "num_input_tokens_seen": 7326422,
      "step": 2664,
      "train_runtime": 6830.5725,
      "train_tokens_per_second": 1072.593
    },
    {
      "epoch": 0.21414652765222283,
      "grad_norm": 0.24605362117290497,
      "learning_rate": 2.2437395659432388e-05,
      "loss": 0.6223,
      "num_input_tokens_seen": 7329524,
      "step": 2665,
      "train_runtime": 6832.9372,
      "train_tokens_per_second": 1072.675
    },
    {
      "epoch": 0.21422688282207356,
      "grad_norm": 0.31505972146987915,
      "learning_rate": 2.2370617696160268e-05,
      "loss": 0.9699,
      "num_input_tokens_seen": 7332948,
      "step": 2666,
      "train_runtime": 6835.4707,
      "train_tokens_per_second": 1072.779
    },
    {
      "epoch": 0.2143072379919243,
      "grad_norm": 0.3338657319545746,
      "learning_rate": 2.2303839732888147e-05,
      "loss": 1.0288,
      "num_input_tokens_seen": 7334594,
      "step": 2667,
      "train_runtime": 6837.097,
      "train_tokens_per_second": 1072.764
    },
    {
      "epoch": 0.21438759316177505,
      "grad_norm": 0.30498605966567993,
      "learning_rate": 2.2237061769616027e-05,
      "loss": 0.8056,
      "num_input_tokens_seen": 7337100,
      "step": 2668,
      "train_runtime": 6839.0923,
      "train_tokens_per_second": 1072.818
    },
    {
      "epoch": 0.21446794833162577,
      "grad_norm": 0.26526975631713867,
      "learning_rate": 2.2170283806343907e-05,
      "loss": 0.5352,
      "num_input_tokens_seen": 7341566,
      "step": 2669,
      "train_runtime": 6842.3069,
      "train_tokens_per_second": 1072.966
    },
    {
      "epoch": 0.21454830350147652,
      "grad_norm": 0.29485267400741577,
      "learning_rate": 2.2103505843071787e-05,
      "loss": 1.0351,
      "num_input_tokens_seen": 7347226,
      "step": 2670,
      "train_runtime": 6846.0991,
      "train_tokens_per_second": 1073.199
    },
    {
      "epoch": 0.21462865867132727,
      "grad_norm": 0.3023536503314972,
      "learning_rate": 2.2036727879799667e-05,
      "loss": 1.0583,
      "num_input_tokens_seen": 7351488,
      "step": 2671,
      "train_runtime": 6851.2074,
      "train_tokens_per_second": 1073.021
    },
    {
      "epoch": 0.21470901384117802,
      "grad_norm": 0.3053301274776459,
      "learning_rate": 2.1969949916527547e-05,
      "loss": 0.991,
      "num_input_tokens_seen": 7354532,
      "step": 2672,
      "train_runtime": 6853.4359,
      "train_tokens_per_second": 1073.116
    },
    {
      "epoch": 0.21478936901102874,
      "grad_norm": 0.24701736867427826,
      "learning_rate": 2.1903171953255427e-05,
      "loss": 1.0048,
      "num_input_tokens_seen": 7356676,
      "step": 2673,
      "train_runtime": 6855.2744,
      "train_tokens_per_second": 1073.141
    },
    {
      "epoch": 0.2148697241808795,
      "grad_norm": 0.2527194917201996,
      "learning_rate": 2.1836393989983307e-05,
      "loss": 0.7104,
      "num_input_tokens_seen": 7360688,
      "step": 2674,
      "train_runtime": 6858.1179,
      "train_tokens_per_second": 1073.281
    },
    {
      "epoch": 0.21495007935073024,
      "grad_norm": 0.2915332615375519,
      "learning_rate": 2.1769616026711187e-05,
      "loss": 0.9118,
      "num_input_tokens_seen": 7363500,
      "step": 2675,
      "train_runtime": 6860.3455,
      "train_tokens_per_second": 1073.342
    },
    {
      "epoch": 0.21503043452058096,
      "grad_norm": 0.32805347442626953,
      "learning_rate": 2.1702838063439067e-05,
      "loss": 1.0744,
      "num_input_tokens_seen": 7366078,
      "step": 2676,
      "train_runtime": 6862.5163,
      "train_tokens_per_second": 1073.379
    },
    {
      "epoch": 0.2151107896904317,
      "grad_norm": 0.2782030403614044,
      "learning_rate": 2.1636060100166947e-05,
      "loss": 0.6709,
      "num_input_tokens_seen": 7368460,
      "step": 2677,
      "train_runtime": 6864.5478,
      "train_tokens_per_second": 1073.408
    },
    {
      "epoch": 0.21519114486028246,
      "grad_norm": 0.35958606004714966,
      "learning_rate": 2.1569282136894827e-05,
      "loss": 0.8976,
      "num_input_tokens_seen": 7370048,
      "step": 2678,
      "train_runtime": 6866.2331,
      "train_tokens_per_second": 1073.376
    },
    {
      "epoch": 0.21527150003013318,
      "grad_norm": 0.35285475850105286,
      "learning_rate": 2.1502504173622707e-05,
      "loss": 1.0993,
      "num_input_tokens_seen": 7371954,
      "step": 2679,
      "train_runtime": 6868.0118,
      "train_tokens_per_second": 1073.375
    },
    {
      "epoch": 0.21535185519998393,
      "grad_norm": 0.2908136546611786,
      "learning_rate": 2.1435726210350583e-05,
      "loss": 0.8416,
      "num_input_tokens_seen": 7374582,
      "step": 2680,
      "train_runtime": 6870.0835,
      "train_tokens_per_second": 1073.434
    },
    {
      "epoch": 0.21543221036983468,
      "grad_norm": 0.3031177520751953,
      "learning_rate": 2.1368948247078466e-05,
      "loss": 0.8274,
      "num_input_tokens_seen": 7376486,
      "step": 2681,
      "train_runtime": 6871.7699,
      "train_tokens_per_second": 1073.448
    },
    {
      "epoch": 0.2155125655396854,
      "grad_norm": 0.2759515643119812,
      "learning_rate": 2.1302170283806343e-05,
      "loss": 0.9795,
      "num_input_tokens_seen": 7378832,
      "step": 2682,
      "train_runtime": 6873.698,
      "train_tokens_per_second": 1073.488
    },
    {
      "epoch": 0.21559292070953615,
      "grad_norm": 0.35168221592903137,
      "learning_rate": 2.1235392320534226e-05,
      "loss": 1.2491,
      "num_input_tokens_seen": 7381410,
      "step": 2683,
      "train_runtime": 6875.7596,
      "train_tokens_per_second": 1073.541
    },
    {
      "epoch": 0.2156732758793869,
      "grad_norm": 0.3279869556427002,
      "learning_rate": 2.1168614357262103e-05,
      "loss": 0.8237,
      "num_input_tokens_seen": 7384458,
      "step": 2684,
      "train_runtime": 6878.1204,
      "train_tokens_per_second": 1073.616
    },
    {
      "epoch": 0.21575363104923764,
      "grad_norm": 0.2660094201564789,
      "learning_rate": 2.1101836393989986e-05,
      "loss": 0.8046,
      "num_input_tokens_seen": 7389552,
      "step": 2685,
      "train_runtime": 6881.7289,
      "train_tokens_per_second": 1073.793
    },
    {
      "epoch": 0.21583398621908836,
      "grad_norm": 0.31938111782073975,
      "learning_rate": 2.1035058430717863e-05,
      "loss": 0.8759,
      "num_input_tokens_seen": 7391530,
      "step": 2686,
      "train_runtime": 6883.4879,
      "train_tokens_per_second": 1073.806
    },
    {
      "epoch": 0.2159143413889391,
      "grad_norm": 0.3493044376373291,
      "learning_rate": 2.0968280467445746e-05,
      "loss": 0.6627,
      "num_input_tokens_seen": 7393370,
      "step": 2687,
      "train_runtime": 6885.1023,
      "train_tokens_per_second": 1073.821
    },
    {
      "epoch": 0.21599469655878986,
      "grad_norm": 0.33528077602386475,
      "learning_rate": 2.0901502504173623e-05,
      "loss": 0.9373,
      "num_input_tokens_seen": 7397214,
      "step": 2688,
      "train_runtime": 6887.931,
      "train_tokens_per_second": 1073.938
    },
    {
      "epoch": 0.21607505172864058,
      "grad_norm": 0.2748642861843109,
      "learning_rate": 2.0834724540901503e-05,
      "loss": 0.7841,
      "num_input_tokens_seen": 7400056,
      "step": 2689,
      "train_runtime": 6890.1277,
      "train_tokens_per_second": 1074.009
    },
    {
      "epoch": 0.21615540689849133,
      "grad_norm": 0.4419088065624237,
      "learning_rate": 2.0767946577629382e-05,
      "loss": 0.9283,
      "num_input_tokens_seen": 7402320,
      "step": 2690,
      "train_runtime": 6892.0847,
      "train_tokens_per_second": 1074.032
    },
    {
      "epoch": 0.21623576206834208,
      "grad_norm": 0.3332759737968445,
      "learning_rate": 2.0701168614357262e-05,
      "loss": 0.8147,
      "num_input_tokens_seen": 7404310,
      "step": 2691,
      "train_runtime": 6893.915,
      "train_tokens_per_second": 1074.036
    },
    {
      "epoch": 0.2163161172381928,
      "grad_norm": 0.35181930661201477,
      "learning_rate": 2.0634390651085142e-05,
      "loss": 0.6183,
      "num_input_tokens_seen": 7408904,
      "step": 2692,
      "train_runtime": 6897.1517,
      "train_tokens_per_second": 1074.198
    },
    {
      "epoch": 0.21639647240804355,
      "grad_norm": 0.2954927980899811,
      "learning_rate": 2.0567612687813022e-05,
      "loss": 0.7353,
      "num_input_tokens_seen": 7410632,
      "step": 2693,
      "train_runtime": 6898.8635,
      "train_tokens_per_second": 1074.182
    },
    {
      "epoch": 0.2164768275778943,
      "grad_norm": 0.3183957040309906,
      "learning_rate": 2.0500834724540902e-05,
      "loss": 1.1305,
      "num_input_tokens_seen": 7413232,
      "step": 2694,
      "train_runtime": 6901.0044,
      "train_tokens_per_second": 1074.225
    },
    {
      "epoch": 0.21655718274774502,
      "grad_norm": 0.26527586579322815,
      "learning_rate": 2.0434056761268782e-05,
      "loss": 0.6788,
      "num_input_tokens_seen": 7416196,
      "step": 2695,
      "train_runtime": 6903.2506,
      "train_tokens_per_second": 1074.305
    },
    {
      "epoch": 0.21663753791759577,
      "grad_norm": 0.2783952057361603,
      "learning_rate": 2.0367278797996662e-05,
      "loss": 0.9476,
      "num_input_tokens_seen": 7419270,
      "step": 2696,
      "train_runtime": 6905.6032,
      "train_tokens_per_second": 1074.384
    },
    {
      "epoch": 0.21671789308744652,
      "grad_norm": 0.3151261806488037,
      "learning_rate": 2.0300500834724542e-05,
      "loss": 0.739,
      "num_input_tokens_seen": 7420666,
      "step": 2697,
      "train_runtime": 6907.2119,
      "train_tokens_per_second": 1074.336
    },
    {
      "epoch": 0.21679824825729727,
      "grad_norm": 0.34722790122032166,
      "learning_rate": 2.0233722871452422e-05,
      "loss": 0.9448,
      "num_input_tokens_seen": 7422662,
      "step": 2698,
      "train_runtime": 6909.0533,
      "train_tokens_per_second": 1074.339
    },
    {
      "epoch": 0.216878603427148,
      "grad_norm": 0.3775460422039032,
      "learning_rate": 2.0166944908180302e-05,
      "loss": 1.0543,
      "num_input_tokens_seen": 7425192,
      "step": 2699,
      "train_runtime": 6911.1317,
      "train_tokens_per_second": 1074.381
    },
    {
      "epoch": 0.21695895859699874,
      "grad_norm": 0.3494967818260193,
      "learning_rate": 2.0100166944908182e-05,
      "loss": 0.9087,
      "num_input_tokens_seen": 7426926,
      "step": 2700,
      "train_runtime": 6912.7909,
      "train_tokens_per_second": 1074.374
    },
    {
      "epoch": 0.21695895859699874,
      "eval_loss": 0.9462488889694214,
      "eval_runtime": 91.8617,
      "eval_samples_per_second": 10.821,
      "eval_steps_per_second": 5.41,
      "num_input_tokens_seen": 7426926,
      "step": 2700
    },
    {
      "epoch": 0.21703931376684948,
      "grad_norm": 0.23258782923221588,
      "learning_rate": 2.003338898163606e-05,
      "loss": 0.6272,
      "num_input_tokens_seen": 7429876,
      "step": 2701,
      "train_runtime": 7008.9304,
      "train_tokens_per_second": 1060.058
    },
    {
      "epoch": 0.2171196689367002,
      "grad_norm": 0.32834675908088684,
      "learning_rate": 1.9966611018363938e-05,
      "loss": 1.1235,
      "num_input_tokens_seen": 7433088,
      "step": 2702,
      "train_runtime": 7011.5475,
      "train_tokens_per_second": 1060.121
    },
    {
      "epoch": 0.21720002410655095,
      "grad_norm": 0.30189961194992065,
      "learning_rate": 1.989983305509182e-05,
      "loss": 0.677,
      "num_input_tokens_seen": 7435490,
      "step": 2703,
      "train_runtime": 7013.6099,
      "train_tokens_per_second": 1060.152
    },
    {
      "epoch": 0.2172803792764017,
      "grad_norm": 0.4429466724395752,
      "learning_rate": 1.9833055091819698e-05,
      "loss": 0.8828,
      "num_input_tokens_seen": 7438714,
      "step": 2704,
      "train_runtime": 7016.0518,
      "train_tokens_per_second": 1060.242
    },
    {
      "epoch": 0.21736073444625242,
      "grad_norm": 0.2981514036655426,
      "learning_rate": 1.976627712854758e-05,
      "loss": 0.8641,
      "num_input_tokens_seen": 7441586,
      "step": 2705,
      "train_runtime": 7018.3498,
      "train_tokens_per_second": 1060.304
    },
    {
      "epoch": 0.21744108961610317,
      "grad_norm": 0.3092416226863861,
      "learning_rate": 1.9699499165275458e-05,
      "loss": 1.1596,
      "num_input_tokens_seen": 7444180,
      "step": 2706,
      "train_runtime": 7020.4766,
      "train_tokens_per_second": 1060.353
    },
    {
      "epoch": 0.21752144478595392,
      "grad_norm": 0.31896716356277466,
      "learning_rate": 1.963272120200334e-05,
      "loss": 1.0078,
      "num_input_tokens_seen": 7446328,
      "step": 2707,
      "train_runtime": 7022.5245,
      "train_tokens_per_second": 1060.349
    },
    {
      "epoch": 0.21760179995580467,
      "grad_norm": 0.38206470012664795,
      "learning_rate": 1.9565943238731218e-05,
      "loss": 0.9398,
      "num_input_tokens_seen": 7449000,
      "step": 2708,
      "train_runtime": 7024.8444,
      "train_tokens_per_second": 1060.379
    },
    {
      "epoch": 0.2176821551256554,
      "grad_norm": 0.3746073544025421,
      "learning_rate": 1.94991652754591e-05,
      "loss": 1.1794,
      "num_input_tokens_seen": 7452144,
      "step": 2709,
      "train_runtime": 7027.2468,
      "train_tokens_per_second": 1060.464
    },
    {
      "epoch": 0.21776251029550614,
      "grad_norm": 0.3502049744129181,
      "learning_rate": 1.9432387312186978e-05,
      "loss": 0.8753,
      "num_input_tokens_seen": 7458520,
      "step": 2710,
      "train_runtime": 7031.5253,
      "train_tokens_per_second": 1060.726
    },
    {
      "epoch": 0.2178428654653569,
      "grad_norm": 0.3128703832626343,
      "learning_rate": 1.936560934891486e-05,
      "loss": 0.7591,
      "num_input_tokens_seen": 7460892,
      "step": 2711,
      "train_runtime": 7033.6334,
      "train_tokens_per_second": 1060.745
    },
    {
      "epoch": 0.2179232206352076,
      "grad_norm": 0.2930780053138733,
      "learning_rate": 1.9298831385642738e-05,
      "loss": 0.8508,
      "num_input_tokens_seen": 7463138,
      "step": 2712,
      "train_runtime": 7035.5719,
      "train_tokens_per_second": 1060.772
    },
    {
      "epoch": 0.21800357580505836,
      "grad_norm": 0.3618297576904297,
      "learning_rate": 1.9232053422370617e-05,
      "loss": 0.8672,
      "num_input_tokens_seen": 7465172,
      "step": 2713,
      "train_runtime": 7037.566,
      "train_tokens_per_second": 1060.76
    },
    {
      "epoch": 0.2180839309749091,
      "grad_norm": 0.3108785152435303,
      "learning_rate": 1.9165275459098497e-05,
      "loss": 1.2028,
      "num_input_tokens_seen": 7468206,
      "step": 2714,
      "train_runtime": 7039.9215,
      "train_tokens_per_second": 1060.837
    },
    {
      "epoch": 0.21816428614475983,
      "grad_norm": 0.302286297082901,
      "learning_rate": 1.9098497495826377e-05,
      "loss": 0.9869,
      "num_input_tokens_seen": 7469672,
      "step": 2715,
      "train_runtime": 7041.6064,
      "train_tokens_per_second": 1060.791
    },
    {
      "epoch": 0.21824464131461058,
      "grad_norm": 0.3044348955154419,
      "learning_rate": 1.9031719532554257e-05,
      "loss": 1.1177,
      "num_input_tokens_seen": 7472026,
      "step": 2716,
      "train_runtime": 7043.6345,
      "train_tokens_per_second": 1060.82
    },
    {
      "epoch": 0.21832499648446133,
      "grad_norm": 0.3829231262207031,
      "learning_rate": 1.8964941569282137e-05,
      "loss": 1.1893,
      "num_input_tokens_seen": 7473822,
      "step": 2717,
      "train_runtime": 7045.3432,
      "train_tokens_per_second": 1060.817
    },
    {
      "epoch": 0.21840535165431205,
      "grad_norm": 0.33265331387519836,
      "learning_rate": 1.8898163606010017e-05,
      "loss": 0.869,
      "num_input_tokens_seen": 7475930,
      "step": 2718,
      "train_runtime": 7047.1866,
      "train_tokens_per_second": 1060.839
    },
    {
      "epoch": 0.2184857068241628,
      "grad_norm": 0.2811022996902466,
      "learning_rate": 1.8831385642737897e-05,
      "loss": 0.7001,
      "num_input_tokens_seen": 7480962,
      "step": 2719,
      "train_runtime": 7050.6597,
      "train_tokens_per_second": 1061.03
    },
    {
      "epoch": 0.21856606199401354,
      "grad_norm": 0.3676680028438568,
      "learning_rate": 1.8764607679465777e-05,
      "loss": 0.9166,
      "num_input_tokens_seen": 7483488,
      "step": 2720,
      "train_runtime": 7052.7755,
      "train_tokens_per_second": 1061.07
    },
    {
      "epoch": 0.2186464171638643,
      "grad_norm": 0.3773687481880188,
      "learning_rate": 1.8697829716193657e-05,
      "loss": 1.1516,
      "num_input_tokens_seen": 7487532,
      "step": 2721,
      "train_runtime": 7055.8931,
      "train_tokens_per_second": 1061.174
    },
    {
      "epoch": 0.218726772333715,
      "grad_norm": 0.33698898553848267,
      "learning_rate": 1.8631051752921537e-05,
      "loss": 1.0416,
      "num_input_tokens_seen": 7490190,
      "step": 2722,
      "train_runtime": 7058.0962,
      "train_tokens_per_second": 1061.22
    },
    {
      "epoch": 0.21880712750356576,
      "grad_norm": 0.3313373327255249,
      "learning_rate": 1.8564273789649417e-05,
      "loss": 1.0212,
      "num_input_tokens_seen": 7492270,
      "step": 2723,
      "train_runtime": 7059.9386,
      "train_tokens_per_second": 1061.237
    },
    {
      "epoch": 0.2188874826734165,
      "grad_norm": 0.41645348072052,
      "learning_rate": 1.8497495826377297e-05,
      "loss": 1.2292,
      "num_input_tokens_seen": 7494912,
      "step": 2724,
      "train_runtime": 7062.216,
      "train_tokens_per_second": 1061.269
    },
    {
      "epoch": 0.21896783784326723,
      "grad_norm": 0.3400404155254364,
      "learning_rate": 1.8430717863105177e-05,
      "loss": 0.9372,
      "num_input_tokens_seen": 7502262,
      "step": 2725,
      "train_runtime": 7067.224,
      "train_tokens_per_second": 1061.557
    },
    {
      "epoch": 0.21904819301311798,
      "grad_norm": 0.434825599193573,
      "learning_rate": 1.8363939899833053e-05,
      "loss": 0.9718,
      "num_input_tokens_seen": 7504992,
      "step": 2726,
      "train_runtime": 7069.4429,
      "train_tokens_per_second": 1061.61
    },
    {
      "epoch": 0.21912854818296873,
      "grad_norm": 0.3313480317592621,
      "learning_rate": 1.8297161936560936e-05,
      "loss": 0.9407,
      "num_input_tokens_seen": 7507210,
      "step": 2727,
      "train_runtime": 7071.3254,
      "train_tokens_per_second": 1061.641
    },
    {
      "epoch": 0.21920890335281945,
      "grad_norm": 0.3347714841365814,
      "learning_rate": 1.8230383973288813e-05,
      "loss": 0.9767,
      "num_input_tokens_seen": 7509756,
      "step": 2728,
      "train_runtime": 7073.4817,
      "train_tokens_per_second": 1061.677
    },
    {
      "epoch": 0.2192892585226702,
      "grad_norm": 0.3706495761871338,
      "learning_rate": 1.8163606010016696e-05,
      "loss": 0.9196,
      "num_input_tokens_seen": 7511502,
      "step": 2729,
      "train_runtime": 7075.1574,
      "train_tokens_per_second": 1061.673
    },
    {
      "epoch": 0.21936961369252095,
      "grad_norm": 0.2683712840080261,
      "learning_rate": 1.8096828046744573e-05,
      "loss": 0.9637,
      "num_input_tokens_seen": 7514504,
      "step": 2730,
      "train_runtime": 7077.4961,
      "train_tokens_per_second": 1061.746
    },
    {
      "epoch": 0.21944996886237167,
      "grad_norm": 0.3433111608028412,
      "learning_rate": 1.8030050083472456e-05,
      "loss": 0.9535,
      "num_input_tokens_seen": 7516332,
      "step": 2731,
      "train_runtime": 7081.2034,
      "train_tokens_per_second": 1061.448
    },
    {
      "epoch": 0.21953032403222242,
      "grad_norm": 0.3412112593650818,
      "learning_rate": 1.7963272120200333e-05,
      "loss": 0.847,
      "num_input_tokens_seen": 7518756,
      "step": 2732,
      "train_runtime": 7083.2854,
      "train_tokens_per_second": 1061.479
    },
    {
      "epoch": 0.21961067920207317,
      "grad_norm": 0.3159896731376648,
      "learning_rate": 1.7896494156928216e-05,
      "loss": 0.8763,
      "num_input_tokens_seen": 7521188,
      "step": 2733,
      "train_runtime": 7085.2766,
      "train_tokens_per_second": 1061.524
    },
    {
      "epoch": 0.21969103437192392,
      "grad_norm": 0.30528387427330017,
      "learning_rate": 1.7829716193656093e-05,
      "loss": 1.2493,
      "num_input_tokens_seen": 7523178,
      "step": 2734,
      "train_runtime": 7087.0701,
      "train_tokens_per_second": 1061.536
    },
    {
      "epoch": 0.21977138954177464,
      "grad_norm": 0.30628272891044617,
      "learning_rate": 1.7762938230383973e-05,
      "loss": 0.5852,
      "num_input_tokens_seen": 7524804,
      "step": 2735,
      "train_runtime": 7088.6891,
      "train_tokens_per_second": 1061.523
    },
    {
      "epoch": 0.21985174471162539,
      "grad_norm": 0.3449867367744446,
      "learning_rate": 1.7696160267111852e-05,
      "loss": 0.7541,
      "num_input_tokens_seen": 7528124,
      "step": 2736,
      "train_runtime": 7091.273,
      "train_tokens_per_second": 1061.604
    },
    {
      "epoch": 0.21993209988147613,
      "grad_norm": 0.3475934863090515,
      "learning_rate": 1.7629382303839732e-05,
      "loss": 0.843,
      "num_input_tokens_seen": 7530232,
      "step": 2737,
      "train_runtime": 7093.1845,
      "train_tokens_per_second": 1061.615
    },
    {
      "epoch": 0.22001245505132686,
      "grad_norm": 0.28976932168006897,
      "learning_rate": 1.7562604340567612e-05,
      "loss": 0.8691,
      "num_input_tokens_seen": 7532906,
      "step": 2738,
      "train_runtime": 7095.3236,
      "train_tokens_per_second": 1061.672
    },
    {
      "epoch": 0.2200928102211776,
      "grad_norm": 0.33810850977897644,
      "learning_rate": 1.7495826377295492e-05,
      "loss": 1.1956,
      "num_input_tokens_seen": 7534788,
      "step": 2739,
      "train_runtime": 7097.08,
      "train_tokens_per_second": 1061.674
    },
    {
      "epoch": 0.22017316539102835,
      "grad_norm": 0.2440091371536255,
      "learning_rate": 1.7429048414023372e-05,
      "loss": 0.7232,
      "num_input_tokens_seen": 7536906,
      "step": 2740,
      "train_runtime": 7098.973,
      "train_tokens_per_second": 1061.69
    },
    {
      "epoch": 0.22025352056087907,
      "grad_norm": 0.4699127674102783,
      "learning_rate": 1.7362270450751252e-05,
      "loss": 1.0435,
      "num_input_tokens_seen": 7539552,
      "step": 2741,
      "train_runtime": 7101.1193,
      "train_tokens_per_second": 1061.741
    },
    {
      "epoch": 0.22033387573072982,
      "grad_norm": 0.3182070553302765,
      "learning_rate": 1.7295492487479135e-05,
      "loss": 1.1047,
      "num_input_tokens_seen": 7542468,
      "step": 2742,
      "train_runtime": 7103.3859,
      "train_tokens_per_second": 1061.813
    },
    {
      "epoch": 0.22041423090058057,
      "grad_norm": 0.29897093772888184,
      "learning_rate": 1.7228714524207012e-05,
      "loss": 0.882,
      "num_input_tokens_seen": 7545064,
      "step": 2743,
      "train_runtime": 7105.455,
      "train_tokens_per_second": 1061.869
    },
    {
      "epoch": 0.22049458607043132,
      "grad_norm": 0.2339617758989334,
      "learning_rate": 1.7161936560934895e-05,
      "loss": 0.7417,
      "num_input_tokens_seen": 7547950,
      "step": 2744,
      "train_runtime": 7107.6417,
      "train_tokens_per_second": 1061.949
    },
    {
      "epoch": 0.22057494124028204,
      "grad_norm": 0.33375948667526245,
      "learning_rate": 1.7095158597662772e-05,
      "loss": 1.0305,
      "num_input_tokens_seen": 7550510,
      "step": 2745,
      "train_runtime": 7109.7677,
      "train_tokens_per_second": 1061.991
    },
    {
      "epoch": 0.2206552964101328,
      "grad_norm": 0.3424736261367798,
      "learning_rate": 1.7028380634390652e-05,
      "loss": 1.039,
      "num_input_tokens_seen": 7552700,
      "step": 2746,
      "train_runtime": 7111.6861,
      "train_tokens_per_second": 1062.013
    },
    {
      "epoch": 0.22073565157998354,
      "grad_norm": 0.26616615056991577,
      "learning_rate": 1.696160267111853e-05,
      "loss": 0.7395,
      "num_input_tokens_seen": 7555232,
      "step": 2747,
      "train_runtime": 7113.7743,
      "train_tokens_per_second": 1062.057
    },
    {
      "epoch": 0.22081600674983426,
      "grad_norm": 0.2896953225135803,
      "learning_rate": 1.689482470784641e-05,
      "loss": 0.971,
      "num_input_tokens_seen": 7557462,
      "step": 2748,
      "train_runtime": 7115.6848,
      "train_tokens_per_second": 1062.085
    },
    {
      "epoch": 0.220896361919685,
      "grad_norm": 0.3005113899707794,
      "learning_rate": 1.682804674457429e-05,
      "loss": 0.8776,
      "num_input_tokens_seen": 7560614,
      "step": 2749,
      "train_runtime": 7118.0764,
      "train_tokens_per_second": 1062.171
    },
    {
      "epoch": 0.22097671708953576,
      "grad_norm": 0.3157309591770172,
      "learning_rate": 1.676126878130217e-05,
      "loss": 1.0592,
      "num_input_tokens_seen": 7562062,
      "step": 2750,
      "train_runtime": 7119.7255,
      "train_tokens_per_second": 1062.128
    },
    {
      "epoch": 0.22105707225938648,
      "grad_norm": 0.30210593342781067,
      "learning_rate": 1.669449081803005e-05,
      "loss": 1.1242,
      "num_input_tokens_seen": 7565244,
      "step": 2751,
      "train_runtime": 7122.1925,
      "train_tokens_per_second": 1062.207
    },
    {
      "epoch": 0.22113742742923723,
      "grad_norm": 0.3235064446926117,
      "learning_rate": 1.662771285475793e-05,
      "loss": 0.9333,
      "num_input_tokens_seen": 7568790,
      "step": 2752,
      "train_runtime": 7124.8272,
      "train_tokens_per_second": 1062.312
    },
    {
      "epoch": 0.22121778259908798,
      "grad_norm": 0.3814135789871216,
      "learning_rate": 1.656093489148581e-05,
      "loss": 0.6731,
      "num_input_tokens_seen": 7571830,
      "step": 2753,
      "train_runtime": 7127.2932,
      "train_tokens_per_second": 1062.371
    },
    {
      "epoch": 0.2212981377689387,
      "grad_norm": 0.3221556842327118,
      "learning_rate": 1.649415692821369e-05,
      "loss": 1.0155,
      "num_input_tokens_seen": 7574554,
      "step": 2754,
      "train_runtime": 7129.5174,
      "train_tokens_per_second": 1062.422
    },
    {
      "epoch": 0.22137849293878945,
      "grad_norm": 0.29579707980155945,
      "learning_rate": 1.642737896494157e-05,
      "loss": 0.7501,
      "num_input_tokens_seen": 7577142,
      "step": 2755,
      "train_runtime": 7131.7048,
      "train_tokens_per_second": 1062.459
    },
    {
      "epoch": 0.2214588481086402,
      "grad_norm": 0.23488302528858185,
      "learning_rate": 1.636060100166945e-05,
      "loss": 0.6321,
      "num_input_tokens_seen": 7580082,
      "step": 2756,
      "train_runtime": 7134.0728,
      "train_tokens_per_second": 1062.518
    },
    {
      "epoch": 0.22153920327849094,
      "grad_norm": 0.2756893038749695,
      "learning_rate": 1.629382303839733e-05,
      "loss": 0.7717,
      "num_input_tokens_seen": 7582392,
      "step": 2757,
      "train_runtime": 7136.0213,
      "train_tokens_per_second": 1062.552
    },
    {
      "epoch": 0.22161955844834166,
      "grad_norm": 0.33371415734291077,
      "learning_rate": 1.622704507512521e-05,
      "loss": 1.2048,
      "num_input_tokens_seen": 7585716,
      "step": 2758,
      "train_runtime": 7138.4669,
      "train_tokens_per_second": 1062.653
    },
    {
      "epoch": 0.2216999136181924,
      "grad_norm": 0.3020438551902771,
      "learning_rate": 1.6160267111853087e-05,
      "loss": 0.9294,
      "num_input_tokens_seen": 7587408,
      "step": 2759,
      "train_runtime": 7140.1474,
      "train_tokens_per_second": 1062.64
    },
    {
      "epoch": 0.22178026878804316,
      "grad_norm": 0.3730298578739166,
      "learning_rate": 1.609348914858097e-05,
      "loss": 0.7758,
      "num_input_tokens_seen": 7590666,
      "step": 2760,
      "train_runtime": 7142.7447,
      "train_tokens_per_second": 1062.71
    },
    {
      "epoch": 0.22186062395789388,
      "grad_norm": 0.3398522734642029,
      "learning_rate": 1.6026711185308847e-05,
      "loss": 1.0886,
      "num_input_tokens_seen": 7593454,
      "step": 2761,
      "train_runtime": 7147.0288,
      "train_tokens_per_second": 1062.463
    },
    {
      "epoch": 0.22194097912774463,
      "grad_norm": 0.28900668025016785,
      "learning_rate": 1.595993322203673e-05,
      "loss": 1.1081,
      "num_input_tokens_seen": 7596162,
      "step": 2762,
      "train_runtime": 7149.1478,
      "train_tokens_per_second": 1062.527
    },
    {
      "epoch": 0.22202133429759538,
      "grad_norm": 0.26332908868789673,
      "learning_rate": 1.5893155258764607e-05,
      "loss": 1.0853,
      "num_input_tokens_seen": 7599620,
      "step": 2763,
      "train_runtime": 7151.8001,
      "train_tokens_per_second": 1062.616
    },
    {
      "epoch": 0.2221016894674461,
      "grad_norm": 0.3361031115055084,
      "learning_rate": 1.582637729549249e-05,
      "loss": 1.1744,
      "num_input_tokens_seen": 7602324,
      "step": 2764,
      "train_runtime": 7154.0234,
      "train_tokens_per_second": 1062.664
    },
    {
      "epoch": 0.22218204463729685,
      "grad_norm": 0.31317827105522156,
      "learning_rate": 1.5759599332220367e-05,
      "loss": 0.7566,
      "num_input_tokens_seen": 7604942,
      "step": 2765,
      "train_runtime": 7156.1307,
      "train_tokens_per_second": 1062.717
    },
    {
      "epoch": 0.2222623998071476,
      "grad_norm": 0.36296823620796204,
      "learning_rate": 1.569282136894825e-05,
      "loss": 0.9936,
      "num_input_tokens_seen": 7607210,
      "step": 2766,
      "train_runtime": 7158.0767,
      "train_tokens_per_second": 1062.745
    },
    {
      "epoch": 0.22234275497699832,
      "grad_norm": 0.3713575601577759,
      "learning_rate": 1.5626043405676127e-05,
      "loss": 0.7825,
      "num_input_tokens_seen": 7609224,
      "step": 2767,
      "train_runtime": 7159.941,
      "train_tokens_per_second": 1062.75
    },
    {
      "epoch": 0.22242311014684907,
      "grad_norm": 0.2904340624809265,
      "learning_rate": 1.555926544240401e-05,
      "loss": 0.7011,
      "num_input_tokens_seen": 7611612,
      "step": 2768,
      "train_runtime": 7162.011,
      "train_tokens_per_second": 1062.776
    },
    {
      "epoch": 0.22250346531669982,
      "grad_norm": 0.3913591206073761,
      "learning_rate": 1.5492487479131887e-05,
      "loss": 0.8902,
      "num_input_tokens_seen": 7615296,
      "step": 2769,
      "train_runtime": 7164.9327,
      "train_tokens_per_second": 1062.857
    },
    {
      "epoch": 0.22258382048655057,
      "grad_norm": 0.2655889391899109,
      "learning_rate": 1.5425709515859767e-05,
      "loss": 0.9361,
      "num_input_tokens_seen": 7617554,
      "step": 2770,
      "train_runtime": 7166.8552,
      "train_tokens_per_second": 1062.887
    },
    {
      "epoch": 0.2226641756564013,
      "grad_norm": 0.37378987669944763,
      "learning_rate": 1.5358931552587647e-05,
      "loss": 1.2193,
      "num_input_tokens_seen": 7621758,
      "step": 2771,
      "train_runtime": 7169.8825,
      "train_tokens_per_second": 1063.024
    },
    {
      "epoch": 0.22274453082625204,
      "grad_norm": 0.39028868079185486,
      "learning_rate": 1.5292153589315527e-05,
      "loss": 0.9422,
      "num_input_tokens_seen": 7624642,
      "step": 2772,
      "train_runtime": 7172.1362,
      "train_tokens_per_second": 1063.092
    },
    {
      "epoch": 0.22282488599610278,
      "grad_norm": 0.3148764371871948,
      "learning_rate": 1.5225375626043406e-05,
      "loss": 0.8326,
      "num_input_tokens_seen": 7627618,
      "step": 2773,
      "train_runtime": 7174.462,
      "train_tokens_per_second": 1063.162
    },
    {
      "epoch": 0.2229052411659535,
      "grad_norm": 0.3463549017906189,
      "learning_rate": 1.5158597662771286e-05,
      "loss": 0.7761,
      "num_input_tokens_seen": 7632144,
      "step": 2774,
      "train_runtime": 7177.9553,
      "train_tokens_per_second": 1063.276
    },
    {
      "epoch": 0.22298559633580425,
      "grad_norm": 0.2773972749710083,
      "learning_rate": 1.5091819699499166e-05,
      "loss": 0.8171,
      "num_input_tokens_seen": 7634462,
      "step": 2775,
      "train_runtime": 7179.9033,
      "train_tokens_per_second": 1063.31
    },
    {
      "epoch": 0.223065951505655,
      "grad_norm": 0.29489219188690186,
      "learning_rate": 1.5025041736227046e-05,
      "loss": 0.7423,
      "num_input_tokens_seen": 7637066,
      "step": 2776,
      "train_runtime": 7182.1101,
      "train_tokens_per_second": 1063.346
    },
    {
      "epoch": 0.22314630667550572,
      "grad_norm": 0.3101787269115448,
      "learning_rate": 1.4958263772954924e-05,
      "loss": 1.2891,
      "num_input_tokens_seen": 7639400,
      "step": 2777,
      "train_runtime": 7184.084,
      "train_tokens_per_second": 1063.378
    },
    {
      "epoch": 0.22322666184535647,
      "grad_norm": 0.3292182981967926,
      "learning_rate": 1.4891485809682806e-05,
      "loss": 1.0232,
      "num_input_tokens_seen": 7641670,
      "step": 2778,
      "train_runtime": 7186.082,
      "train_tokens_per_second": 1063.399
    },
    {
      "epoch": 0.22330701701520722,
      "grad_norm": 0.3355922996997833,
      "learning_rate": 1.4824707846410684e-05,
      "loss": 0.9747,
      "num_input_tokens_seen": 7644130,
      "step": 2779,
      "train_runtime": 7188.1654,
      "train_tokens_per_second": 1063.433
    },
    {
      "epoch": 0.22338737218505797,
      "grad_norm": 0.38050124049186707,
      "learning_rate": 1.4757929883138566e-05,
      "loss": 1.1495,
      "num_input_tokens_seen": 7646682,
      "step": 2780,
      "train_runtime": 7190.3454,
      "train_tokens_per_second": 1063.465
    },
    {
      "epoch": 0.2234677273549087,
      "grad_norm": 0.29221051931381226,
      "learning_rate": 1.4691151919866444e-05,
      "loss": 1.0444,
      "num_input_tokens_seen": 7649800,
      "step": 2781,
      "train_runtime": 7192.7621,
      "train_tokens_per_second": 1063.541
    },
    {
      "epoch": 0.22354808252475944,
      "grad_norm": 0.329502671957016,
      "learning_rate": 1.4624373956594326e-05,
      "loss": 1.2311,
      "num_input_tokens_seen": 7652364,
      "step": 2782,
      "train_runtime": 7194.8995,
      "train_tokens_per_second": 1063.582
    },
    {
      "epoch": 0.2236284376946102,
      "grad_norm": 0.5192813277244568,
      "learning_rate": 1.4557595993322204e-05,
      "loss": 1.0872,
      "num_input_tokens_seen": 7657690,
      "step": 2783,
      "train_runtime": 7198.7038,
      "train_tokens_per_second": 1063.76
    },
    {
      "epoch": 0.2237087928644609,
      "grad_norm": 0.2748900055885315,
      "learning_rate": 1.4490818030050086e-05,
      "loss": 1.121,
      "num_input_tokens_seen": 7661360,
      "step": 2784,
      "train_runtime": 7201.4884,
      "train_tokens_per_second": 1063.858
    },
    {
      "epoch": 0.22378914803431166,
      "grad_norm": 0.336200088262558,
      "learning_rate": 1.4424040066777964e-05,
      "loss": 0.8542,
      "num_input_tokens_seen": 7662968,
      "step": 2785,
      "train_runtime": 7203.1515,
      "train_tokens_per_second": 1063.835
    },
    {
      "epoch": 0.2238695032041624,
      "grad_norm": 0.3252107799053192,
      "learning_rate": 1.4357262103505844e-05,
      "loss": 0.6889,
      "num_input_tokens_seen": 7664384,
      "step": 2786,
      "train_runtime": 7204.7823,
      "train_tokens_per_second": 1063.791
    },
    {
      "epoch": 0.22394985837401313,
      "grad_norm": 0.3249092102050781,
      "learning_rate": 1.4290484140233724e-05,
      "loss": 0.9642,
      "num_input_tokens_seen": 7665848,
      "step": 2787,
      "train_runtime": 7206.4266,
      "train_tokens_per_second": 1063.752
    },
    {
      "epoch": 0.22403021354386388,
      "grad_norm": 0.30040207505226135,
      "learning_rate": 1.4223706176961604e-05,
      "loss": 1.1605,
      "num_input_tokens_seen": 7668158,
      "step": 2788,
      "train_runtime": 7208.3572,
      "train_tokens_per_second": 1063.787
    },
    {
      "epoch": 0.22411056871371463,
      "grad_norm": 0.2391747534275055,
      "learning_rate": 1.4156928213689482e-05,
      "loss": 0.9312,
      "num_input_tokens_seen": 7670494,
      "step": 2789,
      "train_runtime": 7210.359,
      "train_tokens_per_second": 1063.816
    },
    {
      "epoch": 0.22419092388356535,
      "grad_norm": 0.24021784961223602,
      "learning_rate": 1.4090150250417364e-05,
      "loss": 0.6008,
      "num_input_tokens_seen": 7672992,
      "step": 2790,
      "train_runtime": 7212.4287,
      "train_tokens_per_second": 1063.857
    },
    {
      "epoch": 0.2242712790534161,
      "grad_norm": 0.2623099088668823,
      "learning_rate": 1.4023372287145242e-05,
      "loss": 1.1019,
      "num_input_tokens_seen": 7676050,
      "step": 2791,
      "train_runtime": 7216.8544,
      "train_tokens_per_second": 1063.628
    },
    {
      "epoch": 0.22435163422326684,
      "grad_norm": 0.3006986379623413,
      "learning_rate": 1.3956594323873123e-05,
      "loss": 1.0244,
      "num_input_tokens_seen": 7680218,
      "step": 2792,
      "train_runtime": 7219.8453,
      "train_tokens_per_second": 1063.765
    },
    {
      "epoch": 0.2244319893931176,
      "grad_norm": 0.3254138231277466,
      "learning_rate": 1.3889816360601002e-05,
      "loss": 1.1955,
      "num_input_tokens_seen": 7684362,
      "step": 2793,
      "train_runtime": 7222.9531,
      "train_tokens_per_second": 1063.881
    },
    {
      "epoch": 0.2245123445629683,
      "grad_norm": 0.30785423517227173,
      "learning_rate": 1.3823038397328883e-05,
      "loss": 0.7092,
      "num_input_tokens_seen": 7686524,
      "step": 2794,
      "train_runtime": 7224.8765,
      "train_tokens_per_second": 1063.897
    },
    {
      "epoch": 0.22459269973281906,
      "grad_norm": 0.3039698600769043,
      "learning_rate": 1.3756260434056762e-05,
      "loss": 1.0811,
      "num_input_tokens_seen": 7688250,
      "step": 2795,
      "train_runtime": 7226.594,
      "train_tokens_per_second": 1063.883
    },
    {
      "epoch": 0.2246730549026698,
      "grad_norm": 0.34966331720352173,
      "learning_rate": 1.3689482470784643e-05,
      "loss": 1.0938,
      "num_input_tokens_seen": 7691452,
      "step": 2796,
      "train_runtime": 7228.9636,
      "train_tokens_per_second": 1063.977
    },
    {
      "epoch": 0.22475341007252053,
      "grad_norm": 0.2906704843044281,
      "learning_rate": 1.3622704507512521e-05,
      "loss": 0.92,
      "num_input_tokens_seen": 7694226,
      "step": 2797,
      "train_runtime": 7231.1877,
      "train_tokens_per_second": 1064.034
    },
    {
      "epoch": 0.22483376524237128,
      "grad_norm": 0.29575860500335693,
      "learning_rate": 1.3555926544240401e-05,
      "loss": 0.7041,
      "num_input_tokens_seen": 7698728,
      "step": 2798,
      "train_runtime": 7234.3646,
      "train_tokens_per_second": 1064.189
    },
    {
      "epoch": 0.22491412041222203,
      "grad_norm": 0.2830314040184021,
      "learning_rate": 1.3489148580968281e-05,
      "loss": 0.8335,
      "num_input_tokens_seen": 7701848,
      "step": 2799,
      "train_runtime": 7236.8206,
      "train_tokens_per_second": 1064.259
    },
    {
      "epoch": 0.22499447558207275,
      "grad_norm": 0.30267012119293213,
      "learning_rate": 1.3422370617696161e-05,
      "loss": 1.0311,
      "num_input_tokens_seen": 7704054,
      "step": 2800,
      "train_runtime": 7238.6722,
      "train_tokens_per_second": 1064.291
    },
    {
      "epoch": 0.2250748307519235,
      "grad_norm": 0.2815823554992676,
      "learning_rate": 1.335559265442404e-05,
      "loss": 0.9588,
      "num_input_tokens_seen": 7709760,
      "step": 2801,
      "train_runtime": 7242.709,
      "train_tokens_per_second": 1064.486
    },
    {
      "epoch": 0.22515518592177425,
      "grad_norm": 0.3106796443462372,
      "learning_rate": 1.3288814691151921e-05,
      "loss": 0.7551,
      "num_input_tokens_seen": 7712384,
      "step": 2802,
      "train_runtime": 7244.8426,
      "train_tokens_per_second": 1064.534
    },
    {
      "epoch": 0.22523554109162497,
      "grad_norm": 0.46703192591667175,
      "learning_rate": 1.32220367278798e-05,
      "loss": 1.188,
      "num_input_tokens_seen": 7714664,
      "step": 2803,
      "train_runtime": 7246.9056,
      "train_tokens_per_second": 1064.546
    },
    {
      "epoch": 0.22531589626147572,
      "grad_norm": 0.3115676939487457,
      "learning_rate": 1.3155258764607681e-05,
      "loss": 0.9955,
      "num_input_tokens_seen": 7717418,
      "step": 2804,
      "train_runtime": 7249.0705,
      "train_tokens_per_second": 1064.608
    },
    {
      "epoch": 0.22539625143132647,
      "grad_norm": 0.3923710286617279,
      "learning_rate": 1.3088480801335559e-05,
      "loss": 0.9465,
      "num_input_tokens_seen": 7720622,
      "step": 2805,
      "train_runtime": 7251.4792,
      "train_tokens_per_second": 1064.696
    },
    {
      "epoch": 0.22547660660117722,
      "grad_norm": 0.38434702157974243,
      "learning_rate": 1.302170283806344e-05,
      "loss": 0.8667,
      "num_input_tokens_seen": 7722630,
      "step": 2806,
      "train_runtime": 7253.4237,
      "train_tokens_per_second": 1064.688
    },
    {
      "epoch": 0.22555696177102794,
      "grad_norm": 0.3066710829734802,
      "learning_rate": 1.2954924874791319e-05,
      "loss": 0.8483,
      "num_input_tokens_seen": 7725104,
      "step": 2807,
      "train_runtime": 7255.4333,
      "train_tokens_per_second": 1064.734
    },
    {
      "epoch": 0.22563731694087868,
      "grad_norm": 0.24363166093826294,
      "learning_rate": 1.28881469115192e-05,
      "loss": 0.7707,
      "num_input_tokens_seen": 7727402,
      "step": 2808,
      "train_runtime": 7257.4068,
      "train_tokens_per_second": 1064.761
    },
    {
      "epoch": 0.22571767211072943,
      "grad_norm": 0.2831483483314514,
      "learning_rate": 1.2821368948247079e-05,
      "loss": 0.9463,
      "num_input_tokens_seen": 7731166,
      "step": 2809,
      "train_runtime": 7260.1278,
      "train_tokens_per_second": 1064.88
    },
    {
      "epoch": 0.22579802728058015,
      "grad_norm": 0.36960479617118835,
      "learning_rate": 1.2754590984974959e-05,
      "loss": 0.8618,
      "num_input_tokens_seen": 7733192,
      "step": 2810,
      "train_runtime": 7261.8793,
      "train_tokens_per_second": 1064.902
    },
    {
      "epoch": 0.2258783824504309,
      "grad_norm": 0.3136604428291321,
      "learning_rate": 1.2687813021702839e-05,
      "loss": 1.0546,
      "num_input_tokens_seen": 7735074,
      "step": 2811,
      "train_runtime": 7263.6456,
      "train_tokens_per_second": 1064.902
    },
    {
      "epoch": 0.22595873762028165,
      "grad_norm": 0.33273473381996155,
      "learning_rate": 1.2621035058430719e-05,
      "loss": 1.0375,
      "num_input_tokens_seen": 7737414,
      "step": 2812,
      "train_runtime": 7265.6748,
      "train_tokens_per_second": 1064.927
    },
    {
      "epoch": 0.22603909279013237,
      "grad_norm": 0.3242916166782379,
      "learning_rate": 1.2554257095158597e-05,
      "loss": 1.1273,
      "num_input_tokens_seen": 7740026,
      "step": 2813,
      "train_runtime": 7267.8137,
      "train_tokens_per_second": 1064.973
    },
    {
      "epoch": 0.22611944795998312,
      "grad_norm": 0.304437518119812,
      "learning_rate": 1.2487479131886479e-05,
      "loss": 0.626,
      "num_input_tokens_seen": 7743526,
      "step": 2814,
      "train_runtime": 7270.623,
      "train_tokens_per_second": 1065.043
    },
    {
      "epoch": 0.22619980312983387,
      "grad_norm": 0.33516693115234375,
      "learning_rate": 1.2420701168614358e-05,
      "loss": 0.8642,
      "num_input_tokens_seen": 7746918,
      "step": 2815,
      "train_runtime": 7273.2653,
      "train_tokens_per_second": 1065.122
    },
    {
      "epoch": 0.22628015829968462,
      "grad_norm": 0.3425694406032562,
      "learning_rate": 1.2353923205342238e-05,
      "loss": 0.9588,
      "num_input_tokens_seen": 7749070,
      "step": 2816,
      "train_runtime": 7275.104,
      "train_tokens_per_second": 1065.149
    },
    {
      "epoch": 0.22636051346953534,
      "grad_norm": 0.34973594546318054,
      "learning_rate": 1.2287145242070118e-05,
      "loss": 1.1461,
      "num_input_tokens_seen": 7751182,
      "step": 2817,
      "train_runtime": 7276.9547,
      "train_tokens_per_second": 1065.168
    },
    {
      "epoch": 0.2264408686393861,
      "grad_norm": 0.3162853419780731,
      "learning_rate": 1.2220367278797998e-05,
      "loss": 1.0239,
      "num_input_tokens_seen": 7753020,
      "step": 2818,
      "train_runtime": 7278.6801,
      "train_tokens_per_second": 1065.168
    },
    {
      "epoch": 0.22652122380923684,
      "grad_norm": 0.33880770206451416,
      "learning_rate": 1.2153589315525878e-05,
      "loss": 0.6929,
      "num_input_tokens_seen": 7755194,
      "step": 2819,
      "train_runtime": 7280.6995,
      "train_tokens_per_second": 1065.172
    },
    {
      "epoch": 0.22660157897908756,
      "grad_norm": 0.3280443251132965,
      "learning_rate": 1.2086811352253758e-05,
      "loss": 0.7346,
      "num_input_tokens_seen": 7757484,
      "step": 2820,
      "train_runtime": 7282.6962,
      "train_tokens_per_second": 1065.194
    },
    {
      "epoch": 0.2266819341489383,
      "grad_norm": 0.2715861201286316,
      "learning_rate": 1.2020033388981636e-05,
      "loss": 0.93,
      "num_input_tokens_seen": 7760738,
      "step": 2821,
      "train_runtime": 7287.0263,
      "train_tokens_per_second": 1065.008
    },
    {
      "epoch": 0.22676228931878906,
      "grad_norm": 0.33349916338920593,
      "learning_rate": 1.1953255425709516e-05,
      "loss": 0.8541,
      "num_input_tokens_seen": 7763130,
      "step": 2822,
      "train_runtime": 7289.0782,
      "train_tokens_per_second": 1065.036
    },
    {
      "epoch": 0.22684264448863978,
      "grad_norm": 0.2905218303203583,
      "learning_rate": 1.1886477462437396e-05,
      "loss": 1.2399,
      "num_input_tokens_seen": 7768128,
      "step": 2823,
      "train_runtime": 7292.4876,
      "train_tokens_per_second": 1065.223
    },
    {
      "epoch": 0.22692299965849053,
      "grad_norm": 0.2846352159976959,
      "learning_rate": 1.1819699499165276e-05,
      "loss": 1.0046,
      "num_input_tokens_seen": 7770704,
      "step": 2824,
      "train_runtime": 7294.6479,
      "train_tokens_per_second": 1065.261
    },
    {
      "epoch": 0.22700335482834127,
      "grad_norm": 0.3132615387439728,
      "learning_rate": 1.1752921535893156e-05,
      "loss": 0.8045,
      "num_input_tokens_seen": 7772632,
      "step": 2825,
      "train_runtime": 7296.423,
      "train_tokens_per_second": 1065.266
    },
    {
      "epoch": 0.227083709998192,
      "grad_norm": 0.284273236989975,
      "learning_rate": 1.1686143572621036e-05,
      "loss": 0.5296,
      "num_input_tokens_seen": 7775532,
      "step": 2826,
      "train_runtime": 7298.7683,
      "train_tokens_per_second": 1065.321
    },
    {
      "epoch": 0.22716406516804274,
      "grad_norm": 0.36679720878601074,
      "learning_rate": 1.1619365609348916e-05,
      "loss": 0.9873,
      "num_input_tokens_seen": 7777886,
      "step": 2827,
      "train_runtime": 7300.6926,
      "train_tokens_per_second": 1065.363
    },
    {
      "epoch": 0.2272444203378935,
      "grad_norm": 0.3023168444633484,
      "learning_rate": 1.1552587646076796e-05,
      "loss": 0.8889,
      "num_input_tokens_seen": 7780918,
      "step": 2828,
      "train_runtime": 7303.0048,
      "train_tokens_per_second": 1065.441
    },
    {
      "epoch": 0.22732477550774424,
      "grad_norm": 0.3108291029930115,
      "learning_rate": 1.1485809682804676e-05,
      "loss": 1.0372,
      "num_input_tokens_seen": 7783254,
      "step": 2829,
      "train_runtime": 7304.9604,
      "train_tokens_per_second": 1065.475
    },
    {
      "epoch": 0.22740513067759496,
      "grad_norm": 0.3111181855201721,
      "learning_rate": 1.1419031719532556e-05,
      "loss": 1.3196,
      "num_input_tokens_seen": 7786082,
      "step": 2830,
      "train_runtime": 7307.22,
      "train_tokens_per_second": 1065.533
    },
    {
      "epoch": 0.2274854858474457,
      "grad_norm": 0.29720649123191833,
      "learning_rate": 1.1352253756260436e-05,
      "loss": 1.0281,
      "num_input_tokens_seen": 7787492,
      "step": 2831,
      "train_runtime": 7308.8079,
      "train_tokens_per_second": 1065.494
    },
    {
      "epoch": 0.22756584101729646,
      "grad_norm": 0.4728342592716217,
      "learning_rate": 1.1285475792988316e-05,
      "loss": 0.8499,
      "num_input_tokens_seen": 7789354,
      "step": 2832,
      "train_runtime": 7310.6382,
      "train_tokens_per_second": 1065.482
    },
    {
      "epoch": 0.22764619618714718,
      "grad_norm": 0.3209225833415985,
      "learning_rate": 1.1218697829716194e-05,
      "loss": 0.876,
      "num_input_tokens_seen": 7791106,
      "step": 2833,
      "train_runtime": 7312.3845,
      "train_tokens_per_second": 1065.467
    },
    {
      "epoch": 0.22772655135699793,
      "grad_norm": 0.2421085387468338,
      "learning_rate": 1.1151919866444074e-05,
      "loss": 0.5991,
      "num_input_tokens_seen": 7796354,
      "step": 2834,
      "train_runtime": 7316.0531,
      "train_tokens_per_second": 1065.65
    },
    {
      "epoch": 0.22780690652684868,
      "grad_norm": 0.33176374435424805,
      "learning_rate": 1.1085141903171954e-05,
      "loss": 1.0241,
      "num_input_tokens_seen": 7797930,
      "step": 2835,
      "train_runtime": 7317.7093,
      "train_tokens_per_second": 1065.624
    },
    {
      "epoch": 0.2278872616966994,
      "grad_norm": 0.3912816345691681,
      "learning_rate": 1.1018363939899834e-05,
      "loss": 1.2112,
      "num_input_tokens_seen": 7803436,
      "step": 2836,
      "train_runtime": 7321.5706,
      "train_tokens_per_second": 1065.814
    },
    {
      "epoch": 0.22796761686655015,
      "grad_norm": 0.32537949085235596,
      "learning_rate": 1.0951585976627714e-05,
      "loss": 0.8654,
      "num_input_tokens_seen": 7806686,
      "step": 2837,
      "train_runtime": 7324.0952,
      "train_tokens_per_second": 1065.891
    },
    {
      "epoch": 0.2280479720364009,
      "grad_norm": 0.41164231300354004,
      "learning_rate": 1.0884808013355593e-05,
      "loss": 0.7822,
      "num_input_tokens_seen": 7808626,
      "step": 2838,
      "train_runtime": 7326.0177,
      "train_tokens_per_second": 1065.876
    },
    {
      "epoch": 0.22812832720625162,
      "grad_norm": 0.21988171339035034,
      "learning_rate": 1.0818030050083473e-05,
      "loss": 0.5962,
      "num_input_tokens_seen": 7813740,
      "step": 2839,
      "train_runtime": 7329.5564,
      "train_tokens_per_second": 1066.059
    },
    {
      "epoch": 0.22820868237610237,
      "grad_norm": 0.42122432589530945,
      "learning_rate": 1.0751252086811353e-05,
      "loss": 0.7253,
      "num_input_tokens_seen": 7815254,
      "step": 2840,
      "train_runtime": 7331.1427,
      "train_tokens_per_second": 1066.035
    },
    {
      "epoch": 0.22828903754595312,
      "grad_norm": 0.3596966862678528,
      "learning_rate": 1.0684474123539233e-05,
      "loss": 1.2278,
      "num_input_tokens_seen": 7820932,
      "step": 2841,
      "train_runtime": 7335.0457,
      "train_tokens_per_second": 1066.242
    },
    {
      "epoch": 0.22836939271580387,
      "grad_norm": 0.27379220724105835,
      "learning_rate": 1.0617696160267113e-05,
      "loss": 0.8762,
      "num_input_tokens_seen": 7824268,
      "step": 2842,
      "train_runtime": 7337.6548,
      "train_tokens_per_second": 1066.317
    },
    {
      "epoch": 0.22844974788565459,
      "grad_norm": 0.26335805654525757,
      "learning_rate": 1.0550918196994993e-05,
      "loss": 0.7938,
      "num_input_tokens_seen": 7827360,
      "step": 2843,
      "train_runtime": 7339.9795,
      "train_tokens_per_second": 1066.401
    },
    {
      "epoch": 0.22853010305550533,
      "grad_norm": 0.31241142749786377,
      "learning_rate": 1.0484140233722873e-05,
      "loss": 0.9654,
      "num_input_tokens_seen": 7830142,
      "step": 2844,
      "train_runtime": 7342.3408,
      "train_tokens_per_second": 1066.437
    },
    {
      "epoch": 0.22861045822535608,
      "grad_norm": 0.2643781900405884,
      "learning_rate": 1.0417362270450751e-05,
      "loss": 0.7209,
      "num_input_tokens_seen": 7834530,
      "step": 2845,
      "train_runtime": 7345.4189,
      "train_tokens_per_second": 1066.587
    },
    {
      "epoch": 0.2286908133952068,
      "grad_norm": 0.3357027471065521,
      "learning_rate": 1.0350584307178631e-05,
      "loss": 0.9132,
      "num_input_tokens_seen": 7837466,
      "step": 2846,
      "train_runtime": 7347.6646,
      "train_tokens_per_second": 1066.661
    },
    {
      "epoch": 0.22877116856505755,
      "grad_norm": 0.2800379991531372,
      "learning_rate": 1.0283806343906511e-05,
      "loss": 0.7535,
      "num_input_tokens_seen": 7839400,
      "step": 2847,
      "train_runtime": 7349.4723,
      "train_tokens_per_second": 1066.662
    },
    {
      "epoch": 0.2288515237349083,
      "grad_norm": 0.2625751495361328,
      "learning_rate": 1.0217028380634391e-05,
      "loss": 1.2553,
      "num_input_tokens_seen": 7842714,
      "step": 2848,
      "train_runtime": 7352.0432,
      "train_tokens_per_second": 1066.739
    },
    {
      "epoch": 0.22893187890475902,
      "grad_norm": 0.26711219549179077,
      "learning_rate": 1.0150250417362271e-05,
      "loss": 0.8721,
      "num_input_tokens_seen": 7846890,
      "step": 2849,
      "train_runtime": 7354.9753,
      "train_tokens_per_second": 1066.882
    },
    {
      "epoch": 0.22901223407460977,
      "grad_norm": 0.44203436374664307,
      "learning_rate": 1.0083472454090151e-05,
      "loss": 0.8652,
      "num_input_tokens_seen": 7849840,
      "step": 2850,
      "train_runtime": 7357.2167,
      "train_tokens_per_second": 1066.958
    },
    {
      "epoch": 0.22909258924446052,
      "grad_norm": 0.3775481581687927,
      "learning_rate": 1.001669449081803e-05,
      "loss": 1.0818,
      "num_input_tokens_seen": 7851864,
      "step": 2851,
      "train_runtime": 7361.0026,
      "train_tokens_per_second": 1066.684
    },
    {
      "epoch": 0.22917294441431127,
      "grad_norm": 0.3612276613712311,
      "learning_rate": 9.94991652754591e-06,
      "loss": 0.9377,
      "num_input_tokens_seen": 7854684,
      "step": 2852,
      "train_runtime": 7363.1545,
      "train_tokens_per_second": 1066.755
    },
    {
      "epoch": 0.229253299584162,
      "grad_norm": 0.3302251398563385,
      "learning_rate": 9.88313856427379e-06,
      "loss": 1.0098,
      "num_input_tokens_seen": 7857074,
      "step": 2853,
      "train_runtime": 7365.1826,
      "train_tokens_per_second": 1066.786
    },
    {
      "epoch": 0.22933365475401274,
      "grad_norm": 0.3119541108608246,
      "learning_rate": 9.81636060100167e-06,
      "loss": 1.0693,
      "num_input_tokens_seen": 7859954,
      "step": 2854,
      "train_runtime": 7367.434,
      "train_tokens_per_second": 1066.851
    },
    {
      "epoch": 0.2294140099238635,
      "grad_norm": 0.31083476543426514,
      "learning_rate": 9.74958263772955e-06,
      "loss": 0.9766,
      "num_input_tokens_seen": 7862176,
      "step": 2855,
      "train_runtime": 7369.3726,
      "train_tokens_per_second": 1066.872
    },
    {
      "epoch": 0.2294943650937142,
      "grad_norm": 0.3210439682006836,
      "learning_rate": 9.68280467445743e-06,
      "loss": 0.9786,
      "num_input_tokens_seen": 7864320,
      "step": 2856,
      "train_runtime": 7371.2841,
      "train_tokens_per_second": 1066.886
    },
    {
      "epoch": 0.22957472026356496,
      "grad_norm": 0.3303755819797516,
      "learning_rate": 9.616026711185309e-06,
      "loss": 1.245,
      "num_input_tokens_seen": 7865732,
      "step": 2857,
      "train_runtime": 7372.8087,
      "train_tokens_per_second": 1066.857
    },
    {
      "epoch": 0.2296550754334157,
      "grad_norm": 0.4303469657897949,
      "learning_rate": 9.549248747913189e-06,
      "loss": 0.7776,
      "num_input_tokens_seen": 7866982,
      "step": 2858,
      "train_runtime": 7374.3863,
      "train_tokens_per_second": 1066.798
    },
    {
      "epoch": 0.22973543060326643,
      "grad_norm": 0.3318173587322235,
      "learning_rate": 9.482470784641069e-06,
      "loss": 1.1827,
      "num_input_tokens_seen": 7869742,
      "step": 2859,
      "train_runtime": 7376.5709,
      "train_tokens_per_second": 1066.856
    },
    {
      "epoch": 0.22981578577311718,
      "grad_norm": 0.3349165618419647,
      "learning_rate": 9.415692821368949e-06,
      "loss": 1.2823,
      "num_input_tokens_seen": 7873914,
      "step": 2860,
      "train_runtime": 7379.4401,
      "train_tokens_per_second": 1067.007
    },
    {
      "epoch": 0.22989614094296792,
      "grad_norm": 0.44777733087539673,
      "learning_rate": 9.348914858096828e-06,
      "loss": 1.1932,
      "num_input_tokens_seen": 7877024,
      "step": 2861,
      "train_runtime": 7381.8517,
      "train_tokens_per_second": 1067.08
    },
    {
      "epoch": 0.22997649611281865,
      "grad_norm": 0.2634897232055664,
      "learning_rate": 9.282136894824708e-06,
      "loss": 0.6923,
      "num_input_tokens_seen": 7881064,
      "step": 2862,
      "train_runtime": 7384.7315,
      "train_tokens_per_second": 1067.211
    },
    {
      "epoch": 0.2300568512826694,
      "grad_norm": 0.32381150126457214,
      "learning_rate": 9.215358931552588e-06,
      "loss": 0.8596,
      "num_input_tokens_seen": 7882750,
      "step": 2863,
      "train_runtime": 7386.3447,
      "train_tokens_per_second": 1067.206
    },
    {
      "epoch": 0.23013720645252014,
      "grad_norm": 0.28828147053718567,
      "learning_rate": 9.148580968280468e-06,
      "loss": 0.7077,
      "num_input_tokens_seen": 7884816,
      "step": 2864,
      "train_runtime": 7388.1629,
      "train_tokens_per_second": 1067.223
    },
    {
      "epoch": 0.2302175616223709,
      "grad_norm": 0.311092346906662,
      "learning_rate": 9.081803005008348e-06,
      "loss": 0.8973,
      "num_input_tokens_seen": 7886860,
      "step": 2865,
      "train_runtime": 7389.9721,
      "train_tokens_per_second": 1067.238
    },
    {
      "epoch": 0.2302979167922216,
      "grad_norm": 0.3248305320739746,
      "learning_rate": 9.015025041736228e-06,
      "loss": 1.0069,
      "num_input_tokens_seen": 7889794,
      "step": 2866,
      "train_runtime": 7392.2721,
      "train_tokens_per_second": 1067.303
    },
    {
      "epoch": 0.23037827196207236,
      "grad_norm": 0.41866010427474976,
      "learning_rate": 8.948247078464108e-06,
      "loss": 0.6156,
      "num_input_tokens_seen": 7892032,
      "step": 2867,
      "train_runtime": 7394.1612,
      "train_tokens_per_second": 1067.333
    },
    {
      "epoch": 0.2304586271319231,
      "grad_norm": 0.2685219645500183,
      "learning_rate": 8.881469115191986e-06,
      "loss": 0.7219,
      "num_input_tokens_seen": 7893876,
      "step": 2868,
      "train_runtime": 7395.8705,
      "train_tokens_per_second": 1067.336
    },
    {
      "epoch": 0.23053898230177383,
      "grad_norm": 0.33013466000556946,
      "learning_rate": 8.814691151919866e-06,
      "loss": 1.0515,
      "num_input_tokens_seen": 7897892,
      "step": 2869,
      "train_runtime": 7398.7903,
      "train_tokens_per_second": 1067.457
    },
    {
      "epoch": 0.23061933747162458,
      "grad_norm": 0.3439158797264099,
      "learning_rate": 8.747913188647746e-06,
      "loss": 0.9476,
      "num_input_tokens_seen": 7900860,
      "step": 2870,
      "train_runtime": 7401.1311,
      "train_tokens_per_second": 1067.521
    },
    {
      "epoch": 0.23069969264147533,
      "grad_norm": 0.47075122594833374,
      "learning_rate": 8.681135225375626e-06,
      "loss": 0.818,
      "num_input_tokens_seen": 7903916,
      "step": 2871,
      "train_runtime": 7403.4968,
      "train_tokens_per_second": 1067.592
    },
    {
      "epoch": 0.23078004781132605,
      "grad_norm": 0.34985777735710144,
      "learning_rate": 8.614357262103506e-06,
      "loss": 1.2345,
      "num_input_tokens_seen": 7906336,
      "step": 2872,
      "train_runtime": 7405.5741,
      "train_tokens_per_second": 1067.62
    },
    {
      "epoch": 0.2308604029811768,
      "grad_norm": 0.3292609751224518,
      "learning_rate": 8.547579298831386e-06,
      "loss": 0.8414,
      "num_input_tokens_seen": 7908388,
      "step": 2873,
      "train_runtime": 7407.5154,
      "train_tokens_per_second": 1067.617
    },
    {
      "epoch": 0.23094075815102755,
      "grad_norm": 0.2691466808319092,
      "learning_rate": 8.480801335559266e-06,
      "loss": 0.5495,
      "num_input_tokens_seen": 7910158,
      "step": 2874,
      "train_runtime": 7409.2047,
      "train_tokens_per_second": 1067.612
    },
    {
      "epoch": 0.23102111332087827,
      "grad_norm": 0.48094266653060913,
      "learning_rate": 8.414023372287146e-06,
      "loss": 1.5238,
      "num_input_tokens_seen": 7913244,
      "step": 2875,
      "train_runtime": 7411.558,
      "train_tokens_per_second": 1067.69
    },
    {
      "epoch": 0.23110146849072902,
      "grad_norm": 0.26958349347114563,
      "learning_rate": 8.347245409015026e-06,
      "loss": 1.1002,
      "num_input_tokens_seen": 7915352,
      "step": 2876,
      "train_runtime": 7413.3457,
      "train_tokens_per_second": 1067.717
    },
    {
      "epoch": 0.23118182366057977,
      "grad_norm": 0.3663804233074188,
      "learning_rate": 8.280467445742906e-06,
      "loss": 0.8882,
      "num_input_tokens_seen": 7918284,
      "step": 2877,
      "train_runtime": 7415.6583,
      "train_tokens_per_second": 1067.779
    },
    {
      "epoch": 0.23126217883043051,
      "grad_norm": 0.30901092290878296,
      "learning_rate": 8.213689482470786e-06,
      "loss": 0.9035,
      "num_input_tokens_seen": 7921850,
      "step": 2878,
      "train_runtime": 7418.3172,
      "train_tokens_per_second": 1067.877
    },
    {
      "epoch": 0.23134253400028124,
      "grad_norm": 0.32220202684402466,
      "learning_rate": 8.146911519198665e-06,
      "loss": 0.8439,
      "num_input_tokens_seen": 7924204,
      "step": 2879,
      "train_runtime": 7420.3741,
      "train_tokens_per_second": 1067.898
    },
    {
      "epoch": 0.23142288917013198,
      "grad_norm": 0.33323851227760315,
      "learning_rate": 8.080133555926544e-06,
      "loss": 1.2147,
      "num_input_tokens_seen": 7927764,
      "step": 2880,
      "train_runtime": 7423.0877,
      "train_tokens_per_second": 1067.987
    },
    {
      "epoch": 0.23150324433998273,
      "grad_norm": 0.32072654366493225,
      "learning_rate": 8.013355592654424e-06,
      "loss": 0.7404,
      "num_input_tokens_seen": 7929586,
      "step": 2881,
      "train_runtime": 7426.6932,
      "train_tokens_per_second": 1067.714
    },
    {
      "epoch": 0.23158359950983345,
      "grad_norm": 0.25004059076309204,
      "learning_rate": 7.946577629382304e-06,
      "loss": 0.8091,
      "num_input_tokens_seen": 7932858,
      "step": 2882,
      "train_runtime": 7429.1247,
      "train_tokens_per_second": 1067.805
    },
    {
      "epoch": 0.2316639546796842,
      "grad_norm": 0.3316909074783325,
      "learning_rate": 7.879799666110184e-06,
      "loss": 1.0841,
      "num_input_tokens_seen": 7936180,
      "step": 2883,
      "train_runtime": 7431.5776,
      "train_tokens_per_second": 1067.9
    },
    {
      "epoch": 0.23174430984953495,
      "grad_norm": 0.31683462858200073,
      "learning_rate": 7.813021702838063e-06,
      "loss": 1.1515,
      "num_input_tokens_seen": 7939672,
      "step": 2884,
      "train_runtime": 7434.0602,
      "train_tokens_per_second": 1068.013
    },
    {
      "epoch": 0.23182466501938567,
      "grad_norm": 0.28647980093955994,
      "learning_rate": 7.746243739565943e-06,
      "loss": 0.856,
      "num_input_tokens_seen": 7942304,
      "step": 2885,
      "train_runtime": 7436.2105,
      "train_tokens_per_second": 1068.058
    },
    {
      "epoch": 0.23190502018923642,
      "grad_norm": 0.34173357486724854,
      "learning_rate": 7.679465776293823e-06,
      "loss": 0.8874,
      "num_input_tokens_seen": 7944342,
      "step": 2886,
      "train_runtime": 7438.0798,
      "train_tokens_per_second": 1068.064
    },
    {
      "epoch": 0.23198537535908717,
      "grad_norm": 0.4894322156906128,
      "learning_rate": 7.612687813021703e-06,
      "loss": 0.8204,
      "num_input_tokens_seen": 7946780,
      "step": 2887,
      "train_runtime": 7440.0672,
      "train_tokens_per_second": 1068.106
    },
    {
      "epoch": 0.23206573052893792,
      "grad_norm": 0.29378753900527954,
      "learning_rate": 7.545909849749583e-06,
      "loss": 0.8396,
      "num_input_tokens_seen": 7949056,
      "step": 2888,
      "train_runtime": 7442.0464,
      "train_tokens_per_second": 1068.128
    },
    {
      "epoch": 0.23214608569878864,
      "grad_norm": 0.4440731704235077,
      "learning_rate": 7.479131886477462e-06,
      "loss": 1.5238,
      "num_input_tokens_seen": 7950558,
      "step": 2889,
      "train_runtime": 7443.6071,
      "train_tokens_per_second": 1068.106
    },
    {
      "epoch": 0.2322264408686394,
      "grad_norm": 0.3225192725658417,
      "learning_rate": 7.412353923205342e-06,
      "loss": 1.0,
      "num_input_tokens_seen": 7953398,
      "step": 2890,
      "train_runtime": 7445.8224,
      "train_tokens_per_second": 1068.169
    },
    {
      "epoch": 0.23230679603849014,
      "grad_norm": 0.3230671286582947,
      "learning_rate": 7.345575959933222e-06,
      "loss": 0.9854,
      "num_input_tokens_seen": 7956788,
      "step": 2891,
      "train_runtime": 7448.3302,
      "train_tokens_per_second": 1068.265
    },
    {
      "epoch": 0.23238715120834086,
      "grad_norm": 0.28703269362449646,
      "learning_rate": 7.278797996661102e-06,
      "loss": 1.0658,
      "num_input_tokens_seen": 7958956,
      "step": 2892,
      "train_runtime": 7450.291,
      "train_tokens_per_second": 1068.275
    },
    {
      "epoch": 0.2324675063781916,
      "grad_norm": 0.3324466943740845,
      "learning_rate": 7.212020033388982e-06,
      "loss": 0.8563,
      "num_input_tokens_seen": 7960240,
      "step": 2893,
      "train_runtime": 7451.8441,
      "train_tokens_per_second": 1068.224
    },
    {
      "epoch": 0.23254786154804236,
      "grad_norm": 0.28836894035339355,
      "learning_rate": 7.145242070116862e-06,
      "loss": 0.9315,
      "num_input_tokens_seen": 7964606,
      "step": 2894,
      "train_runtime": 7455.0547,
      "train_tokens_per_second": 1068.35
    },
    {
      "epoch": 0.23262821671789308,
      "grad_norm": 0.32132986187934875,
      "learning_rate": 7.078464106844741e-06,
      "loss": 0.9448,
      "num_input_tokens_seen": 7967774,
      "step": 2895,
      "train_runtime": 7457.5341,
      "train_tokens_per_second": 1068.419
    },
    {
      "epoch": 0.23270857188774383,
      "grad_norm": 0.313288152217865,
      "learning_rate": 7.011686143572621e-06,
      "loss": 0.9615,
      "num_input_tokens_seen": 7969556,
      "step": 2896,
      "train_runtime": 7459.321,
      "train_tokens_per_second": 1068.402
    },
    {
      "epoch": 0.23278892705759457,
      "grad_norm": 0.3201817572116852,
      "learning_rate": 6.944908180300501e-06,
      "loss": 0.7358,
      "num_input_tokens_seen": 7972040,
      "step": 2897,
      "train_runtime": 7461.4026,
      "train_tokens_per_second": 1068.437
    },
    {
      "epoch": 0.2328692822274453,
      "grad_norm": 0.270537793636322,
      "learning_rate": 6.878130217028381e-06,
      "loss": 0.7319,
      "num_input_tokens_seen": 7973974,
      "step": 2898,
      "train_runtime": 7463.0965,
      "train_tokens_per_second": 1068.454
    },
    {
      "epoch": 0.23294963739729604,
      "grad_norm": 0.30717888474464417,
      "learning_rate": 6.811352253756261e-06,
      "loss": 0.6131,
      "num_input_tokens_seen": 7975220,
      "step": 2899,
      "train_runtime": 7464.6087,
      "train_tokens_per_second": 1068.404
    },
    {
      "epoch": 0.2330299925671468,
      "grad_norm": 0.3329704999923706,
      "learning_rate": 6.744574290484141e-06,
      "loss": 1.0444,
      "num_input_tokens_seen": 7978612,
      "step": 2900,
      "train_runtime": 7467.1717,
      "train_tokens_per_second": 1068.492
    },
    {
      "epoch": 0.23311034773699754,
      "grad_norm": 0.3216586709022522,
      "learning_rate": 6.67779632721202e-06,
      "loss": 0.9202,
      "num_input_tokens_seen": 7981040,
      "step": 2901,
      "train_runtime": 7469.1513,
      "train_tokens_per_second": 1068.534
    },
    {
      "epoch": 0.23319070290684826,
      "grad_norm": 0.3532184660434723,
      "learning_rate": 6.6110183639399e-06,
      "loss": 1.1332,
      "num_input_tokens_seen": 7984168,
      "step": 2902,
      "train_runtime": 7471.6063,
      "train_tokens_per_second": 1068.601
    },
    {
      "epoch": 0.233271058076699,
      "grad_norm": 0.3200591802597046,
      "learning_rate": 6.5442404006677796e-06,
      "loss": 1.0378,
      "num_input_tokens_seen": 7986090,
      "step": 2903,
      "train_runtime": 7473.4554,
      "train_tokens_per_second": 1068.594
    },
    {
      "epoch": 0.23335141324654976,
      "grad_norm": 0.34241044521331787,
      "learning_rate": 6.4774624373956595e-06,
      "loss": 1.0959,
      "num_input_tokens_seen": 7989238,
      "step": 2904,
      "train_runtime": 7475.8081,
      "train_tokens_per_second": 1068.679
    },
    {
      "epoch": 0.23343176841640048,
      "grad_norm": 0.34307292103767395,
      "learning_rate": 6.4106844741235394e-06,
      "loss": 1.1283,
      "num_input_tokens_seen": 7991828,
      "step": 2905,
      "train_runtime": 7477.9181,
      "train_tokens_per_second": 1068.724
    },
    {
      "epoch": 0.23351212358625123,
      "grad_norm": 0.3693479001522064,
      "learning_rate": 6.343906510851419e-06,
      "loss": 0.9773,
      "num_input_tokens_seen": 7994854,
      "step": 2906,
      "train_runtime": 7480.2302,
      "train_tokens_per_second": 1068.798
    },
    {
      "epoch": 0.23359247875610198,
      "grad_norm": 0.3348240554332733,
      "learning_rate": 6.2771285475792984e-06,
      "loss": 1.063,
      "num_input_tokens_seen": 7996460,
      "step": 2907,
      "train_runtime": 7481.8134,
      "train_tokens_per_second": 1068.786
    },
    {
      "epoch": 0.2336728339259527,
      "grad_norm": 0.426025390625,
      "learning_rate": 6.210350584307179e-06,
      "loss": 0.6624,
      "num_input_tokens_seen": 7998892,
      "step": 2908,
      "train_runtime": 7483.8507,
      "train_tokens_per_second": 1068.82
    },
    {
      "epoch": 0.23375318909580345,
      "grad_norm": 0.40392255783081055,
      "learning_rate": 6.143572621035059e-06,
      "loss": 0.7834,
      "num_input_tokens_seen": 8002782,
      "step": 2909,
      "train_runtime": 7486.6932,
      "train_tokens_per_second": 1068.934
    },
    {
      "epoch": 0.2338335442656542,
      "grad_norm": 0.3087678551673889,
      "learning_rate": 6.076794657762939e-06,
      "loss": 0.7977,
      "num_input_tokens_seen": 8005676,
      "step": 2910,
      "train_runtime": 7488.9499,
      "train_tokens_per_second": 1068.998
    },
    {
      "epoch": 0.23391389943550492,
      "grad_norm": 0.32012656331062317,
      "learning_rate": 6.010016694490818e-06,
      "loss": 0.9661,
      "num_input_tokens_seen": 8007100,
      "step": 2911,
      "train_runtime": 7492.5151,
      "train_tokens_per_second": 1068.68
    },
    {
      "epoch": 0.23399425460535567,
      "grad_norm": 0.2771163582801819,
      "learning_rate": 5.943238731218698e-06,
      "loss": 0.7767,
      "num_input_tokens_seen": 8010128,
      "step": 2912,
      "train_runtime": 7494.8461,
      "train_tokens_per_second": 1068.751
    },
    {
      "epoch": 0.23407460977520642,
      "grad_norm": 0.28324493765830994,
      "learning_rate": 5.876460767946578e-06,
      "loss": 0.7878,
      "num_input_tokens_seen": 8011928,
      "step": 2913,
      "train_runtime": 7496.6383,
      "train_tokens_per_second": 1068.736
    },
    {
      "epoch": 0.23415496494505716,
      "grad_norm": 0.42554348707199097,
      "learning_rate": 5.809682804674458e-06,
      "loss": 0.8004,
      "num_input_tokens_seen": 8013598,
      "step": 2914,
      "train_runtime": 7498.4033,
      "train_tokens_per_second": 1068.707
    },
    {
      "epoch": 0.23423532011490789,
      "grad_norm": 0.32236963510513306,
      "learning_rate": 5.742904841402338e-06,
      "loss": 1.0473,
      "num_input_tokens_seen": 8016122,
      "step": 2915,
      "train_runtime": 7500.4596,
      "train_tokens_per_second": 1068.751
    },
    {
      "epoch": 0.23431567528475863,
      "grad_norm": 0.29543444514274597,
      "learning_rate": 5.676126878130218e-06,
      "loss": 0.8251,
      "num_input_tokens_seen": 8017744,
      "step": 2916,
      "train_runtime": 7502.0766,
      "train_tokens_per_second": 1068.737
    },
    {
      "epoch": 0.23439603045460938,
      "grad_norm": 0.42115992307662964,
      "learning_rate": 5.609348914858097e-06,
      "loss": 0.8719,
      "num_input_tokens_seen": 8019734,
      "step": 2917,
      "train_runtime": 7503.9417,
      "train_tokens_per_second": 1068.736
    },
    {
      "epoch": 0.2344763856244601,
      "grad_norm": 0.31785961985588074,
      "learning_rate": 5.542570951585977e-06,
      "loss": 1.1883,
      "num_input_tokens_seen": 8022950,
      "step": 2918,
      "train_runtime": 7506.4343,
      "train_tokens_per_second": 1068.81
    },
    {
      "epoch": 0.23455674079431085,
      "grad_norm": 0.307235449552536,
      "learning_rate": 5.475792988313857e-06,
      "loss": 1.1506,
      "num_input_tokens_seen": 8027022,
      "step": 2919,
      "train_runtime": 7509.2837,
      "train_tokens_per_second": 1068.946
    },
    {
      "epoch": 0.2346370959641616,
      "grad_norm": 0.2724400460720062,
      "learning_rate": 5.409015025041737e-06,
      "loss": 0.7967,
      "num_input_tokens_seen": 8029058,
      "step": 2920,
      "train_runtime": 7511.1206,
      "train_tokens_per_second": 1068.956
    },
    {
      "epoch": 0.23471745113401232,
      "grad_norm": 0.34455740451812744,
      "learning_rate": 5.342237061769617e-06,
      "loss": 1.1755,
      "num_input_tokens_seen": 8032596,
      "step": 2921,
      "train_runtime": 7513.7369,
      "train_tokens_per_second": 1069.055
    },
    {
      "epoch": 0.23479780630386307,
      "grad_norm": 0.34917742013931274,
      "learning_rate": 5.2754590984974965e-06,
      "loss": 1.2025,
      "num_input_tokens_seen": 8035236,
      "step": 2922,
      "train_runtime": 7515.9475,
      "train_tokens_per_second": 1069.092
    },
    {
      "epoch": 0.23487816147371382,
      "grad_norm": 0.31094813346862793,
      "learning_rate": 5.208681135225376e-06,
      "loss": 0.8137,
      "num_input_tokens_seen": 8038310,
      "step": 2923,
      "train_runtime": 7518.2876,
      "train_tokens_per_second": 1069.168
    },
    {
      "epoch": 0.23495851664356457,
      "grad_norm": 0.3259291648864746,
      "learning_rate": 5.1419031719532556e-06,
      "loss": 1.2322,
      "num_input_tokens_seen": 8041862,
      "step": 2924,
      "train_runtime": 7520.9054,
      "train_tokens_per_second": 1069.268
    },
    {
      "epoch": 0.2350388718134153,
      "grad_norm": 0.3635278642177582,
      "learning_rate": 5.0751252086811355e-06,
      "loss": 0.9854,
      "num_input_tokens_seen": 8044136,
      "step": 2925,
      "train_runtime": 7522.8269,
      "train_tokens_per_second": 1069.297
    },
    {
      "epoch": 0.23511922698326604,
      "grad_norm": 0.33314600586891174,
      "learning_rate": 5.008347245409015e-06,
      "loss": 0.6917,
      "num_input_tokens_seen": 8046452,
      "step": 2926,
      "train_runtime": 7524.85,
      "train_tokens_per_second": 1069.317
    },
    {
      "epoch": 0.2351995821531168,
      "grad_norm": 0.3343305289745331,
      "learning_rate": 4.941569282136895e-06,
      "loss": 0.8613,
      "num_input_tokens_seen": 8050578,
      "step": 2927,
      "train_runtime": 7527.8506,
      "train_tokens_per_second": 1069.439
    },
    {
      "epoch": 0.2352799373229675,
      "grad_norm": 0.2475036084651947,
      "learning_rate": 4.874791318864775e-06,
      "loss": 0.7761,
      "num_input_tokens_seen": 8055048,
      "step": 2928,
      "train_runtime": 7531.0326,
      "train_tokens_per_second": 1069.581
    },
    {
      "epoch": 0.23536029249281826,
      "grad_norm": 0.3320787847042084,
      "learning_rate": 4.808013355592654e-06,
      "loss": 0.9233,
      "num_input_tokens_seen": 8062624,
      "step": 2929,
      "train_runtime": 7536.0394,
      "train_tokens_per_second": 1069.876
    },
    {
      "epoch": 0.235440647662669,
      "grad_norm": 0.3156088590621948,
      "learning_rate": 4.741235392320534e-06,
      "loss": 0.783,
      "num_input_tokens_seen": 8064260,
      "step": 2930,
      "train_runtime": 7537.7486,
      "train_tokens_per_second": 1069.85
    },
    {
      "epoch": 0.23552100283251973,
      "grad_norm": 0.3170846700668335,
      "learning_rate": 4.674457429048414e-06,
      "loss": 0.9894,
      "num_input_tokens_seen": 8067966,
      "step": 2931,
      "train_runtime": 7540.4163,
      "train_tokens_per_second": 1069.963
    },
    {
      "epoch": 0.23560135800237048,
      "grad_norm": 0.32855215668678284,
      "learning_rate": 4.607679465776294e-06,
      "loss": 1.1993,
      "num_input_tokens_seen": 8072422,
      "step": 2932,
      "train_runtime": 7543.6664,
      "train_tokens_per_second": 1070.093
    },
    {
      "epoch": 0.23568171317222122,
      "grad_norm": 0.30814823508262634,
      "learning_rate": 4.540901502504174e-06,
      "loss": 1.0975,
      "num_input_tokens_seen": 8074270,
      "step": 2933,
      "train_runtime": 7545.4666,
      "train_tokens_per_second": 1070.082
    },
    {
      "epoch": 0.23576206834207195,
      "grad_norm": 0.3340173363685608,
      "learning_rate": 4.474123539232054e-06,
      "loss": 0.927,
      "num_input_tokens_seen": 8076026,
      "step": 2934,
      "train_runtime": 7547.1305,
      "train_tokens_per_second": 1070.079
    },
    {
      "epoch": 0.2358424235119227,
      "grad_norm": 0.2794358432292938,
      "learning_rate": 4.407345575959933e-06,
      "loss": 0.8787,
      "num_input_tokens_seen": 8077654,
      "step": 2935,
      "train_runtime": 7548.7851,
      "train_tokens_per_second": 1070.06
    },
    {
      "epoch": 0.23592277868177344,
      "grad_norm": 0.3027331233024597,
      "learning_rate": 4.340567612687813e-06,
      "loss": 1.0477,
      "num_input_tokens_seen": 8079394,
      "step": 2936,
      "train_runtime": 7550.4611,
      "train_tokens_per_second": 1070.053
    },
    {
      "epoch": 0.2360031338516242,
      "grad_norm": 0.36117789149284363,
      "learning_rate": 4.273789649415693e-06,
      "loss": 1.0531,
      "num_input_tokens_seen": 8082796,
      "step": 2937,
      "train_runtime": 7553.1072,
      "train_tokens_per_second": 1070.129
    },
    {
      "epoch": 0.2360834890214749,
      "grad_norm": 0.33424028754234314,
      "learning_rate": 4.207011686143573e-06,
      "loss": 0.9032,
      "num_input_tokens_seen": 8084530,
      "step": 2938,
      "train_runtime": 7554.8794,
      "train_tokens_per_second": 1070.107
    },
    {
      "epoch": 0.23616384419132566,
      "grad_norm": 0.3893217444419861,
      "learning_rate": 4.140233722871453e-06,
      "loss": 1.0802,
      "num_input_tokens_seen": 8085684,
      "step": 2939,
      "train_runtime": 7556.3951,
      "train_tokens_per_second": 1070.045
    },
    {
      "epoch": 0.2362441993611764,
      "grad_norm": 0.34217315912246704,
      "learning_rate": 4.073455759599333e-06,
      "loss": 1.0972,
      "num_input_tokens_seen": 8087684,
      "step": 2940,
      "train_runtime": 7558.2903,
      "train_tokens_per_second": 1070.041
    },
    {
      "epoch": 0.23632455453102713,
      "grad_norm": 0.32231464982032776,
      "learning_rate": 4.006677796327212e-06,
      "loss": 0.8265,
      "num_input_tokens_seen": 8090178,
      "step": 2941,
      "train_runtime": 7562.2381,
      "train_tokens_per_second": 1069.813
    },
    {
      "epoch": 0.23640490970087788,
      "grad_norm": 0.3164416551589966,
      "learning_rate": 3.939899833055092e-06,
      "loss": 0.8601,
      "num_input_tokens_seen": 8091982,
      "step": 2942,
      "train_runtime": 7564.0161,
      "train_tokens_per_second": 1069.8
    },
    {
      "epoch": 0.23648526487072863,
      "grad_norm": 0.28371524810791016,
      "learning_rate": 3.873121869782972e-06,
      "loss": 0.704,
      "num_input_tokens_seen": 8094278,
      "step": 2943,
      "train_runtime": 7565.9239,
      "train_tokens_per_second": 1069.833
    },
    {
      "epoch": 0.23656562004057935,
      "grad_norm": 0.39141759276390076,
      "learning_rate": 3.8063439065108516e-06,
      "loss": 0.6815,
      "num_input_tokens_seen": 8096112,
      "step": 2944,
      "train_runtime": 7567.6515,
      "train_tokens_per_second": 1069.832
    },
    {
      "epoch": 0.2366459752104301,
      "grad_norm": 0.3549150228500366,
      "learning_rate": 3.739565943238731e-06,
      "loss": 1.1416,
      "num_input_tokens_seen": 8098100,
      "step": 2945,
      "train_runtime": 7569.4856,
      "train_tokens_per_second": 1069.835
    },
    {
      "epoch": 0.23672633038028085,
      "grad_norm": 0.33833369612693787,
      "learning_rate": 3.672787979966611e-06,
      "loss": 1.2012,
      "num_input_tokens_seen": 8099646,
      "step": 2946,
      "train_runtime": 7571.0573,
      "train_tokens_per_second": 1069.817
    },
    {
      "epoch": 0.23680668555013157,
      "grad_norm": 0.3133547902107239,
      "learning_rate": 3.606010016694491e-06,
      "loss": 1.1224,
      "num_input_tokens_seen": 8101918,
      "step": 2947,
      "train_runtime": 7572.982,
      "train_tokens_per_second": 1069.845
    },
    {
      "epoch": 0.23688704071998232,
      "grad_norm": 0.30431944131851196,
      "learning_rate": 3.5392320534223705e-06,
      "loss": 0.9183,
      "num_input_tokens_seen": 8107114,
      "step": 2948,
      "train_runtime": 7576.6564,
      "train_tokens_per_second": 1070.012
    },
    {
      "epoch": 0.23696739588983307,
      "grad_norm": 0.30427950620651245,
      "learning_rate": 3.4724540901502504e-06,
      "loss": 1.1034,
      "num_input_tokens_seen": 8108794,
      "step": 2949,
      "train_runtime": 7578.3324,
      "train_tokens_per_second": 1069.997
    },
    {
      "epoch": 0.23704775105968381,
      "grad_norm": 0.34604355692863464,
      "learning_rate": 3.4056761268781303e-06,
      "loss": 0.9188,
      "num_input_tokens_seen": 8111558,
      "step": 2950,
      "train_runtime": 7580.6007,
      "train_tokens_per_second": 1070.042
    },
    {
      "epoch": 0.23712810622953454,
      "grad_norm": 0.329996258020401,
      "learning_rate": 3.33889816360601e-06,
      "loss": 0.8895,
      "num_input_tokens_seen": 8113664,
      "step": 2951,
      "train_runtime": 7582.5268,
      "train_tokens_per_second": 1070.048
    },
    {
      "epoch": 0.23720846139938528,
      "grad_norm": 0.35375866293907166,
      "learning_rate": 3.2721202003338898e-06,
      "loss": 0.5765,
      "num_input_tokens_seen": 8115088,
      "step": 2952,
      "train_runtime": 7584.1142,
      "train_tokens_per_second": 1070.011
    },
    {
      "epoch": 0.23728881656923603,
      "grad_norm": 0.3138856887817383,
      "learning_rate": 3.2053422370617697e-06,
      "loss": 0.9511,
      "num_input_tokens_seen": 8118080,
      "step": 2953,
      "train_runtime": 7586.4993,
      "train_tokens_per_second": 1070.069
    },
    {
      "epoch": 0.23736917173908675,
      "grad_norm": 0.3197863698005676,
      "learning_rate": 3.1385642737896492e-06,
      "loss": 0.9779,
      "num_input_tokens_seen": 8121460,
      "step": 2954,
      "train_runtime": 7588.9963,
      "train_tokens_per_second": 1070.163
    },
    {
      "epoch": 0.2374495269089375,
      "grad_norm": 0.37144437432289124,
      "learning_rate": 3.0717863105175296e-06,
      "loss": 1.2845,
      "num_input_tokens_seen": 8124492,
      "step": 2955,
      "train_runtime": 7591.2886,
      "train_tokens_per_second": 1070.239
    },
    {
      "epoch": 0.23752988207878825,
      "grad_norm": 0.5184103846549988,
      "learning_rate": 3.005008347245409e-06,
      "loss": 1.414,
      "num_input_tokens_seen": 8126102,
      "step": 2956,
      "train_runtime": 7593.0321,
      "train_tokens_per_second": 1070.205
    },
    {
      "epoch": 0.23761023724863897,
      "grad_norm": 0.2918568551540375,
      "learning_rate": 2.938230383973289e-06,
      "loss": 1.0303,
      "num_input_tokens_seen": 8129702,
      "step": 2957,
      "train_runtime": 7595.6467,
      "train_tokens_per_second": 1070.311
    },
    {
      "epoch": 0.23769059241848972,
      "grad_norm": 0.33410003781318665,
      "learning_rate": 2.871452420701169e-06,
      "loss": 0.6886,
      "num_input_tokens_seen": 8132176,
      "step": 2958,
      "train_runtime": 7597.7294,
      "train_tokens_per_second": 1070.343
    },
    {
      "epoch": 0.23777094758834047,
      "grad_norm": 0.32255151867866516,
      "learning_rate": 2.8046744574290484e-06,
      "loss": 0.9506,
      "num_input_tokens_seen": 8134450,
      "step": 2959,
      "train_runtime": 7599.7847,
      "train_tokens_per_second": 1070.353
    },
    {
      "epoch": 0.23785130275819122,
      "grad_norm": 0.45826491713523865,
      "learning_rate": 2.7378964941569284e-06,
      "loss": 0.8267,
      "num_input_tokens_seen": 8139176,
      "step": 2960,
      "train_runtime": 7603.1713,
      "train_tokens_per_second": 1070.498
    },
    {
      "epoch": 0.23793165792804194,
      "grad_norm": 0.35797688364982605,
      "learning_rate": 2.6711185308848083e-06,
      "loss": 0.9327,
      "num_input_tokens_seen": 8142284,
      "step": 2961,
      "train_runtime": 7605.5174,
      "train_tokens_per_second": 1070.576
    },
    {
      "epoch": 0.2380120130978927,
      "grad_norm": 0.3346036672592163,
      "learning_rate": 2.604340567612688e-06,
      "loss": 0.9001,
      "num_input_tokens_seen": 8143858,
      "step": 2962,
      "train_runtime": 7607.1052,
      "train_tokens_per_second": 1070.559
    },
    {
      "epoch": 0.23809236826774344,
      "grad_norm": 0.26113417744636536,
      "learning_rate": 2.5375626043405677e-06,
      "loss": 0.8289,
      "num_input_tokens_seen": 8145894,
      "step": 2963,
      "train_runtime": 7608.9296,
      "train_tokens_per_second": 1070.57
    },
    {
      "epoch": 0.23817272343759416,
      "grad_norm": 0.37985503673553467,
      "learning_rate": 2.4707846410684477e-06,
      "loss": 0.7447,
      "num_input_tokens_seen": 8147672,
      "step": 2964,
      "train_runtime": 7610.62,
      "train_tokens_per_second": 1070.566
    },
    {
      "epoch": 0.2382530786074449,
      "grad_norm": 0.3743630349636078,
      "learning_rate": 2.404006677796327e-06,
      "loss": 0.9972,
      "num_input_tokens_seen": 8151302,
      "step": 2965,
      "train_runtime": 7613.342,
      "train_tokens_per_second": 1070.66
    },
    {
      "epoch": 0.23833343377729566,
      "grad_norm": 0.2937847077846527,
      "learning_rate": 2.337228714524207e-06,
      "loss": 1.2896,
      "num_input_tokens_seen": 8154100,
      "step": 2966,
      "train_runtime": 7615.6077,
      "train_tokens_per_second": 1070.709
    },
    {
      "epoch": 0.23841378894714638,
      "grad_norm": 0.3713630139827728,
      "learning_rate": 2.270450751252087e-06,
      "loss": 0.9951,
      "num_input_tokens_seen": 8155920,
      "step": 2967,
      "train_runtime": 7617.3183,
      "train_tokens_per_second": 1070.708
    },
    {
      "epoch": 0.23849414411699713,
      "grad_norm": 0.33164307475090027,
      "learning_rate": 2.2036727879799665e-06,
      "loss": 0.903,
      "num_input_tokens_seen": 8159650,
      "step": 2968,
      "train_runtime": 7619.9835,
      "train_tokens_per_second": 1070.823
    },
    {
      "epoch": 0.23857449928684787,
      "grad_norm": 0.48896369338035583,
      "learning_rate": 2.1368948247078465e-06,
      "loss": 0.8399,
      "num_input_tokens_seen": 8162546,
      "step": 2969,
      "train_runtime": 7622.3141,
      "train_tokens_per_second": 1070.875
    },
    {
      "epoch": 0.2386548544566986,
      "grad_norm": 0.2850237190723419,
      "learning_rate": 2.0701168614357264e-06,
      "loss": 0.7736,
      "num_input_tokens_seen": 8165076,
      "step": 2970,
      "train_runtime": 7624.4194,
      "train_tokens_per_second": 1070.911
    },
    {
      "epoch": 0.23873520962654934,
      "grad_norm": 0.3177092969417572,
      "learning_rate": 2.003338898163606e-06,
      "loss": 0.9229,
      "num_input_tokens_seen": 8169654,
      "step": 2971,
      "train_runtime": 7629.6834,
      "train_tokens_per_second": 1070.772
    },
    {
      "epoch": 0.2388155647964001,
      "grad_norm": 0.32074159383773804,
      "learning_rate": 1.936560934891486e-06,
      "loss": 1.2391,
      "num_input_tokens_seen": 8172598,
      "step": 2972,
      "train_runtime": 7632.0303,
      "train_tokens_per_second": 1070.829
    },
    {
      "epoch": 0.23889591996625084,
      "grad_norm": 0.3533782660961151,
      "learning_rate": 1.8697829716193656e-06,
      "loss": 0.9408,
      "num_input_tokens_seen": 8175678,
      "step": 2973,
      "train_runtime": 7634.3486,
      "train_tokens_per_second": 1070.907
    },
    {
      "epoch": 0.23897627513610156,
      "grad_norm": 0.17639051377773285,
      "learning_rate": 1.8030050083472455e-06,
      "loss": 0.5982,
      "num_input_tokens_seen": 8179426,
      "step": 2974,
      "train_runtime": 7637.0742,
      "train_tokens_per_second": 1071.016
    },
    {
      "epoch": 0.2390566303059523,
      "grad_norm": 0.3037920892238617,
      "learning_rate": 1.7362270450751252e-06,
      "loss": 0.9133,
      "num_input_tokens_seen": 8181330,
      "step": 2975,
      "train_runtime": 7638.8352,
      "train_tokens_per_second": 1071.018
    },
    {
      "epoch": 0.23913698547580306,
      "grad_norm": 0.29526981711387634,
      "learning_rate": 1.669449081803005e-06,
      "loss": 1.0546,
      "num_input_tokens_seen": 8184290,
      "step": 2976,
      "train_runtime": 7641.1425,
      "train_tokens_per_second": 1071.082
    },
    {
      "epoch": 0.23921734064565378,
      "grad_norm": 0.2659931182861328,
      "learning_rate": 1.6026711185308849e-06,
      "loss": 0.6936,
      "num_input_tokens_seen": 8185886,
      "step": 2977,
      "train_runtime": 7642.7435,
      "train_tokens_per_second": 1071.066
    },
    {
      "epoch": 0.23929769581550453,
      "grad_norm": 0.2962379455566406,
      "learning_rate": 1.5358931552587648e-06,
      "loss": 1.0793,
      "num_input_tokens_seen": 8189026,
      "step": 2978,
      "train_runtime": 7645.1237,
      "train_tokens_per_second": 1071.144
    },
    {
      "epoch": 0.23937805098535528,
      "grad_norm": 0.2869679033756256,
      "learning_rate": 1.4691151919866445e-06,
      "loss": 0.8109,
      "num_input_tokens_seen": 8191782,
      "step": 2979,
      "train_runtime": 7647.2669,
      "train_tokens_per_second": 1071.204
    },
    {
      "epoch": 0.239458406155206,
      "grad_norm": 0.3432118594646454,
      "learning_rate": 1.4023372287145242e-06,
      "loss": 0.812,
      "num_input_tokens_seen": 8194612,
      "step": 2980,
      "train_runtime": 7649.4137,
      "train_tokens_per_second": 1071.273
    },
    {
      "epoch": 0.23953876132505675,
      "grad_norm": 0.32874420285224915,
      "learning_rate": 1.3355592654424042e-06,
      "loss": 1.0621,
      "num_input_tokens_seen": 8197462,
      "step": 2981,
      "train_runtime": 7651.6827,
      "train_tokens_per_second": 1071.328
    },
    {
      "epoch": 0.2396191164949075,
      "grad_norm": 0.29024678468704224,
      "learning_rate": 1.2687813021702839e-06,
      "loss": 1.039,
      "num_input_tokens_seen": 8201198,
      "step": 2982,
      "train_runtime": 7654.4315,
      "train_tokens_per_second": 1071.431
    },
    {
      "epoch": 0.23969947166475822,
      "grad_norm": 0.4135216772556305,
      "learning_rate": 1.2020033388981636e-06,
      "loss": 0.7295,
      "num_input_tokens_seen": 8204784,
      "step": 2983,
      "train_runtime": 7656.9932,
      "train_tokens_per_second": 1071.541
    },
    {
      "epoch": 0.23977982683460897,
      "grad_norm": 0.2965007722377777,
      "learning_rate": 1.1352253756260435e-06,
      "loss": 0.8961,
      "num_input_tokens_seen": 8208730,
      "step": 2984,
      "train_runtime": 7659.9785,
      "train_tokens_per_second": 1071.639
    },
    {
      "epoch": 0.23986018200445972,
      "grad_norm": 0.3110213577747345,
      "learning_rate": 1.0684474123539232e-06,
      "loss": 0.9759,
      "num_input_tokens_seen": 8212202,
      "step": 2985,
      "train_runtime": 7662.4615,
      "train_tokens_per_second": 1071.745
    },
    {
      "epoch": 0.23994053717431046,
      "grad_norm": 0.3023281693458557,
      "learning_rate": 1.001669449081803e-06,
      "loss": 0.8993,
      "num_input_tokens_seen": 8216986,
      "step": 2986,
      "train_runtime": 7665.8781,
      "train_tokens_per_second": 1071.891
    },
    {
      "epoch": 0.24002089234416119,
      "grad_norm": 0.2991240620613098,
      "learning_rate": 9.348914858096828e-07,
      "loss": 0.9214,
      "num_input_tokens_seen": 8219304,
      "step": 2987,
      "train_runtime": 7667.8302,
      "train_tokens_per_second": 1071.92
    },
    {
      "epoch": 0.24010124751401193,
      "grad_norm": 0.28949102759361267,
      "learning_rate": 8.681135225375626e-07,
      "loss": 0.829,
      "num_input_tokens_seen": 8222132,
      "step": 2988,
      "train_runtime": 7670.056,
      "train_tokens_per_second": 1071.978
    },
    {
      "epoch": 0.24018160268386268,
      "grad_norm": 0.27209827303886414,
      "learning_rate": 8.013355592654424e-07,
      "loss": 0.8092,
      "num_input_tokens_seen": 8224578,
      "step": 2989,
      "train_runtime": 7672.1561,
      "train_tokens_per_second": 1072.003
    },
    {
      "epoch": 0.2402619578537134,
      "grad_norm": 0.5143537521362305,
      "learning_rate": 7.345575959933223e-07,
      "loss": 1.0663,
      "num_input_tokens_seen": 8228546,
      "step": 2990,
      "train_runtime": 7674.9998,
      "train_tokens_per_second": 1072.123
    },
    {
      "epoch": 0.24034231302356415,
      "grad_norm": 0.3822253942489624,
      "learning_rate": 6.677796327212021e-07,
      "loss": 0.8202,
      "num_input_tokens_seen": 8231618,
      "step": 2991,
      "train_runtime": 7677.3573,
      "train_tokens_per_second": 1072.194
    },
    {
      "epoch": 0.2404226681934149,
      "grad_norm": 0.3168606162071228,
      "learning_rate": 6.010016694490818e-07,
      "loss": 0.9721,
      "num_input_tokens_seen": 8233942,
      "step": 2992,
      "train_runtime": 7679.363,
      "train_tokens_per_second": 1072.217
    },
    {
      "epoch": 0.24050302336326562,
      "grad_norm": 0.2644681930541992,
      "learning_rate": 5.342237061769616e-07,
      "loss": 1.1678,
      "num_input_tokens_seen": 8237552,
      "step": 2993,
      "train_runtime": 7681.9391,
      "train_tokens_per_second": 1072.327
    },
    {
      "epoch": 0.24058337853311637,
      "grad_norm": 0.3312743902206421,
      "learning_rate": 4.674457429048414e-07,
      "loss": 0.9957,
      "num_input_tokens_seen": 8239090,
      "step": 2994,
      "train_runtime": 7683.5754,
      "train_tokens_per_second": 1072.299
    },
    {
      "epoch": 0.24066373370296712,
      "grad_norm": 0.33308568596839905,
      "learning_rate": 4.006677796327212e-07,
      "loss": 0.8272,
      "num_input_tokens_seen": 8243064,
      "step": 2995,
      "train_runtime": 7686.5121,
      "train_tokens_per_second": 1072.406
    },
    {
      "epoch": 0.24074408887281787,
      "grad_norm": 0.3265121877193451,
      "learning_rate": 3.3388981636060104e-07,
      "loss": 0.8331,
      "num_input_tokens_seen": 8245414,
      "step": 2996,
      "train_runtime": 7688.5347,
      "train_tokens_per_second": 1072.43
    },
    {
      "epoch": 0.2408244440426686,
      "grad_norm": 0.33722981810569763,
      "learning_rate": 2.671118530884808e-07,
      "loss": 0.9718,
      "num_input_tokens_seen": 8248748,
      "step": 2997,
      "train_runtime": 7691.0796,
      "train_tokens_per_second": 1072.508
    },
    {
      "epoch": 0.24090479921251934,
      "grad_norm": 0.29454463720321655,
      "learning_rate": 2.003338898163606e-07,
      "loss": 0.9714,
      "num_input_tokens_seen": 8250556,
      "step": 2998,
      "train_runtime": 7692.8277,
      "train_tokens_per_second": 1072.5
    },
    {
      "epoch": 0.2409851543823701,
      "grad_norm": 0.33806341886520386,
      "learning_rate": 1.335559265442404e-07,
      "loss": 1.148,
      "num_input_tokens_seen": 8253372,
      "step": 2999,
      "train_runtime": 7695.0414,
      "train_tokens_per_second": 1072.557
    },
    {
      "epoch": 0.2410655095522208,
      "grad_norm": 0.28387361764907837,
      "learning_rate": 6.67779632721202e-08,
      "loss": 0.7471,
      "num_input_tokens_seen": 8256702,
      "step": 3000,
      "train_runtime": 7697.5834,
      "train_tokens_per_second": 1072.636
    },
    {
      "epoch": 0.2410655095522208,
      "eval_loss": 0.9442114233970642,
      "eval_runtime": 91.5536,
      "eval_samples_per_second": 10.857,
      "eval_steps_per_second": 5.429,
      "num_input_tokens_seen": 8256702,
      "step": 3000
    }
  ],
  "logging_steps": 1,
  "max_steps": 3000,
  "num_input_tokens_seen": 8256702,
  "num_train_epochs": 1,
  "save_steps": 30,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.5978029402170368e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}