{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 1395,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.021528525296017224,
      "grad_norm": 0.00015049862849991769,
      "learning_rate": 2.9996919063517692e-05,
      "loss": 1.8642,
      "num_input_tokens_seen": 25888,
      "step": 10,
      "train_runtime": 69.7555,
      "train_tokens_per_second": 371.125
    },
    {
      "epoch": 0.04305705059203445,
      "grad_norm": 0.00011216237908229232,
      "learning_rate": 2.9986270537572827e-05,
      "loss": 0.4561,
      "num_input_tokens_seen": 51424,
      "step": 20,
      "train_runtime": 82.9938,
      "train_tokens_per_second": 619.613
    },
    {
      "epoch": 0.06458557588805167,
      "grad_norm": 4.08398118452169e-05,
      "learning_rate": 2.9968021784983436e-05,
      "loss": 0.3234,
      "num_input_tokens_seen": 77440,
      "step": 30,
      "train_runtime": 96.2791,
      "train_tokens_per_second": 804.328
    },
    {
      "epoch": 0.0861141011840689,
      "grad_norm": 6.656615732936189e-05,
      "learning_rate": 2.994218206053077e-05,
      "loss": 0.2637,
      "num_input_tokens_seen": 103904,
      "step": 40,
      "train_runtime": 109.6408,
      "train_tokens_per_second": 947.676
    },
    {
      "epoch": 0.10764262648008611,
      "grad_norm": 5.515288648894057e-05,
      "learning_rate": 2.9908764468726846e-05,
      "loss": 0.2632,
      "num_input_tokens_seen": 129856,
      "step": 50,
      "train_runtime": 122.7754,
      "train_tokens_per_second": 1057.671
    },
    {
      "epoch": 0.12917115177610333,
      "grad_norm": 7.325380283873528e-05,
      "learning_rate": 2.986778595716853e-05,
      "loss": 0.2319,
      "num_input_tokens_seen": 156064,
      "step": 60,
      "train_runtime": 135.9143,
      "train_tokens_per_second": 1148.253
    },
    {
      "epoch": 0.15069967707212056,
      "grad_norm": 5.39551219844725e-05,
      "learning_rate": 2.981926730794264e-05,
      "loss": 0.2421,
      "num_input_tokens_seen": 182112,
      "step": 70,
      "train_runtime": 149.1813,
      "train_tokens_per_second": 1220.742
    },
    {
      "epoch": 0.1722282023681378,
      "grad_norm": 4.94522500957828e-05,
      "learning_rate": 2.9763233127086394e-05,
      "loss": 0.2189,
      "num_input_tokens_seen": 208096,
      "step": 80,
      "train_runtime": 162.4355,
      "train_tokens_per_second": 1281.1
    },
    {
      "epoch": 0.193756727664155,
      "grad_norm": 2.3115093426895328e-05,
      "learning_rate": 2.9699711832108555e-05,
      "loss": 0.209,
      "num_input_tokens_seen": 233568,
      "step": 90,
      "train_runtime": 175.5666,
      "train_tokens_per_second": 1330.367
    },
    {
      "epoch": 0.21528525296017223,
      "grad_norm": 4.4237815018277615e-05,
      "learning_rate": 2.9628735637577612e-05,
      "loss": 0.2181,
      "num_input_tokens_seen": 259456,
      "step": 100,
      "train_runtime": 188.7096,
      "train_tokens_per_second": 1374.896
    },
    {
      "epoch": 0.23681377825618946,
      "grad_norm": 4.797423753188923e-05,
      "learning_rate": 2.955034053878428e-05,
      "loss": 0.2014,
      "num_input_tokens_seen": 285248,
      "step": 110,
      "train_runtime": 201.8209,
      "train_tokens_per_second": 1413.372
    },
    {
      "epoch": 0.25834230355220666,
      "grad_norm": 4.1381998016731814e-05,
      "learning_rate": 2.9464566293486653e-05,
      "loss": 0.2103,
      "num_input_tokens_seen": 311456,
      "step": 120,
      "train_runtime": 215.01,
      "train_tokens_per_second": 1448.565
    },
    {
      "epoch": 0.2798708288482239,
      "grad_norm": 4.077726771356538e-05,
      "learning_rate": 2.9371456401747193e-05,
      "loss": 0.208,
      "num_input_tokens_seen": 337728,
      "step": 130,
      "train_runtime": 228.2354,
      "train_tokens_per_second": 1479.735
    },
    {
      "epoch": 0.3013993541442411,
      "grad_norm": 5.619543662760407e-05,
      "learning_rate": 2.9271058083871864e-05,
      "loss": 0.2098,
      "num_input_tokens_seen": 363776,
      "step": 140,
      "train_runtime": 241.3694,
      "train_tokens_per_second": 1507.134
    },
    {
      "epoch": 0.32292787944025836,
      "grad_norm": 3.2092542824102566e-05,
      "learning_rate": 2.916342225646254e-05,
      "loss": 0.2094,
      "num_input_tokens_seen": 389504,
      "step": 150,
      "train_runtime": 254.4796,
      "train_tokens_per_second": 1530.59
    },
    {
      "epoch": 0.3444564047362756,
      "grad_norm": 7.814487616997212e-05,
      "learning_rate": 2.904860350659487e-05,
      "loss": 0.1712,
      "num_input_tokens_seen": 415776,
      "step": 160,
      "train_runtime": 267.6294,
      "train_tokens_per_second": 1553.552
    },
    {
      "epoch": 0.36598493003229277,
      "grad_norm": 6.528062658617273e-05,
      "learning_rate": 2.8926660064134637e-05,
      "loss": 0.2264,
      "num_input_tokens_seen": 441248,
      "step": 170,
      "train_runtime": 280.6664,
      "train_tokens_per_second": 1572.144
    },
    {
      "epoch": 0.38751345532831,
      "grad_norm": 2.0589804989867844e-05,
      "learning_rate": 2.8797653772206726e-05,
      "loss": 0.1992,
      "num_input_tokens_seen": 467104,
      "step": 180,
      "train_runtime": 293.7363,
      "train_tokens_per_second": 1590.215
    },
    {
      "epoch": 0.40904198062432723,
      "grad_norm": 4.028606053907424e-05,
      "learning_rate": 2.8661650055831645e-05,
      "loss": 0.1989,
      "num_input_tokens_seen": 493120,
      "step": 190,
      "train_runtime": 306.8119,
      "train_tokens_per_second": 1607.239
    },
    {
      "epoch": 0.43057050592034446,
      "grad_norm": 2.2895075744600035e-05,
      "learning_rate": 2.8518717888745486e-05,
      "loss": 0.1539,
      "num_input_tokens_seen": 519296,
      "step": 200,
      "train_runtime": 319.9991,
      "train_tokens_per_second": 1622.804
    },
    {
      "epoch": 0.4520990312163617,
      "grad_norm": 3.252026726840995e-05,
      "learning_rate": 2.836892975842015e-05,
      "loss": 0.1814,
      "num_input_tokens_seen": 545216,
      "step": 210,
      "train_runtime": 333.1272,
      "train_tokens_per_second": 1636.66
    },
    {
      "epoch": 0.4736275565123789,
      "grad_norm": 6.744113488821313e-05,
      "learning_rate": 2.8212361629301655e-05,
      "loss": 0.1912,
      "num_input_tokens_seen": 571520,
      "step": 220,
      "train_runtime": 346.2915,
      "train_tokens_per_second": 1650.401
    },
    {
      "epoch": 0.4951560818083961,
      "grad_norm": 2.4845583538990468e-05,
      "learning_rate": 2.8049092904285033e-05,
      "loss": 0.1907,
      "num_input_tokens_seen": 597152,
      "step": 230,
      "train_runtime": 359.3902,
      "train_tokens_per_second": 1661.57
    },
    {
      "epoch": 0.5166846071044133,
      "grad_norm": 2.6988314857590012e-05,
      "learning_rate": 2.7879206384445526e-05,
      "loss": 0.1745,
      "num_input_tokens_seen": 623648,
      "step": 240,
      "train_runtime": 372.7872,
      "train_tokens_per_second": 1672.933
    },
    {
      "epoch": 0.5382131324004306,
      "grad_norm": 2.764669079624582e-05,
      "learning_rate": 2.7702788227046286e-05,
      "loss": 0.1695,
      "num_input_tokens_seen": 649312,
      "step": 250,
      "train_runtime": 385.8219,
      "train_tokens_per_second": 1682.932
    },
    {
      "epoch": 0.5597416576964478,
      "grad_norm": 4.1578186937840655e-05,
      "learning_rate": 2.751992790184413e-05,
      "loss": 0.1779,
      "num_input_tokens_seen": 676192,
      "step": 260,
      "train_runtime": 399.1149,
      "train_tokens_per_second": 1694.229
    },
    {
      "epoch": 0.581270182992465,
      "grad_norm": 4.988139698980376e-05,
      "learning_rate": 2.733071814571527e-05,
      "loss": 0.2117,
      "num_input_tokens_seen": 702272,
      "step": 270,
      "train_runtime": 412.2512,
      "train_tokens_per_second": 1703.505
    },
    {
      "epoch": 0.6027987082884823,
      "grad_norm": 2.9431823350023478e-05,
      "learning_rate": 2.7135254915624213e-05,
      "loss": 0.1857,
      "num_input_tokens_seen": 728928,
      "step": 280,
      "train_runtime": 425.464,
      "train_tokens_per_second": 1713.254
    },
    {
      "epoch": 0.6243272335844995,
      "grad_norm": 4.608207746059634e-05,
      "learning_rate": 2.693363733995954e-05,
      "loss": 0.1738,
      "num_input_tokens_seen": 755680,
      "step": 290,
      "train_runtime": 438.7509,
      "train_tokens_per_second": 1722.344
    },
    {
      "epoch": 0.6458557588805167,
      "grad_norm": 4.074648313689977e-05,
      "learning_rate": 2.672596766826138e-05,
      "loss": 0.1581,
      "num_input_tokens_seen": 781504,
      "step": 300,
      "train_runtime": 451.9281,
      "train_tokens_per_second": 1729.266
    },
    {
      "epoch": 0.667384284176534,
      "grad_norm": 4.662120409193449e-05,
      "learning_rate": 2.6512351219365927e-05,
      "loss": 0.1649,
      "num_input_tokens_seen": 807200,
      "step": 310,
      "train_runtime": 465.0346,
      "train_tokens_per_second": 1735.785
    },
    {
      "epoch": 0.6889128094725512,
      "grad_norm": 2.2527565306518227e-05,
      "learning_rate": 2.6292896327993467e-05,
      "loss": 0.1771,
      "num_input_tokens_seen": 833312,
      "step": 320,
      "train_runtime": 478.1705,
      "train_tokens_per_second": 1742.709
    },
    {
      "epoch": 0.7104413347685683,
      "grad_norm": 3.702789399540052e-05,
      "learning_rate": 2.606771428980681e-05,
      "loss": 0.1554,
      "num_input_tokens_seen": 859616,
      "step": 330,
      "train_runtime": 491.4008,
      "train_tokens_per_second": 1749.318
    },
    {
      "epoch": 0.7319698600645855,
      "grad_norm": 2.9329910830711015e-05,
      "learning_rate": 2.5836919304968184e-05,
      "loss": 0.1654,
      "num_input_tokens_seen": 885632,
      "step": 340,
      "train_runtime": 504.5115,
      "train_tokens_per_second": 1755.425
    },
    {
      "epoch": 0.7534983853606028,
      "grad_norm": 3.071613173233345e-05,
      "learning_rate": 2.5600628420223057e-05,
      "loss": 0.1534,
      "num_input_tokens_seen": 912896,
      "step": 350,
      "train_runtime": 517.9846,
      "train_tokens_per_second": 1762.4
    },
    {
      "epoch": 0.77502691065662,
      "grad_norm": 2.746630707406439e-05,
      "learning_rate": 2.535896146954037e-05,
      "loss": 0.1753,
      "num_input_tokens_seen": 938976,
      "step": 360,
      "train_runtime": 531.1381,
      "train_tokens_per_second": 1767.857
    },
    {
      "epoch": 0.7965554359526372,
      "grad_norm": 2.5821629606070928e-05,
      "learning_rate": 2.5112041013339212e-05,
      "loss": 0.1578,
      "num_input_tokens_seen": 966368,
      "step": 370,
      "train_runtime": 544.6424,
      "train_tokens_per_second": 1774.316
    },
    {
      "epoch": 0.8180839612486545,
      "grad_norm": 1.6281868738587946e-05,
      "learning_rate": 2.48599922763328e-05,
      "loss": 0.1339,
      "num_input_tokens_seen": 992544,
      "step": 380,
      "train_runtime": 557.8787,
      "train_tokens_per_second": 1779.14
    },
    {
      "epoch": 0.8396124865446717,
      "grad_norm": 2.4033668523770757e-05,
      "learning_rate": 2.4602943084021286e-05,
      "loss": 0.1895,
      "num_input_tokens_seen": 1018304,
      "step": 390,
      "train_runtime": 570.9772,
      "train_tokens_per_second": 1783.441
    },
    {
      "epoch": 0.8611410118406889,
      "grad_norm": 3.4451768442522734e-05,
      "learning_rate": 2.434102379786553e-05,
      "loss": 0.123,
      "num_input_tokens_seen": 1043936,
      "step": 400,
      "train_runtime": 584.0975,
      "train_tokens_per_second": 1787.263
    },
    {
      "epoch": 0.8826695371367062,
      "grad_norm": 3.287021172582172e-05,
      "learning_rate": 2.4074367249174864e-05,
      "loss": 0.157,
      "num_input_tokens_seen": 1069952,
      "step": 410,
      "train_runtime": 597.2774,
      "train_tokens_per_second": 1791.382
    },
    {
      "epoch": 0.9041980624327234,
      "grad_norm": 4.3571850255830213e-05,
      "learning_rate": 2.380310867174219e-05,
      "loss": 0.1308,
      "num_input_tokens_seen": 1095968,
      "step": 420,
      "train_runtime": 610.3435,
      "train_tokens_per_second": 1795.658
    },
    {
      "epoch": 0.9257265877287406,
      "grad_norm": 1.7651471353019588e-05,
      "learning_rate": 2.3527385633260724e-05,
      "loss": 0.1601,
      "num_input_tokens_seen": 1121920,
      "step": 430,
      "train_runtime": 624.0223,
      "train_tokens_per_second": 1797.885
    },
    {
      "epoch": 0.9472551130247578,
      "grad_norm": 3.0382463592104614e-05,
      "learning_rate": 2.3247337965557097e-05,
      "loss": 0.137,
      "num_input_tokens_seen": 1146656,
      "step": 440,
      "train_runtime": 637.0767,
      "train_tokens_per_second": 1799.871
    },
    {
      "epoch": 0.9687836383207751,
      "grad_norm": 1.7847232811618596e-05,
      "learning_rate": 2.2963107693676192e-05,
      "loss": 0.1404,
      "num_input_tokens_seen": 1172608,
      "step": 450,
      "train_runtime": 650.2332,
      "train_tokens_per_second": 1803.365
    },
    {
      "epoch": 0.9903121636167922,
      "grad_norm": 3.490272501949221e-05,
      "learning_rate": 2.2674838963853738e-05,
      "loss": 0.1319,
      "num_input_tokens_seen": 1198432,
      "step": 460,
      "train_runtime": 663.3441,
      "train_tokens_per_second": 1806.652
    },
    {
      "epoch": 1.0107642626480087,
      "grad_norm": 7.943774835439399e-05,
      "learning_rate": 2.2382677970413097e-05,
      "loss": 0.1274,
      "num_input_tokens_seen": 1223488,
      "step": 470,
      "train_runtime": 675.996,
      "train_tokens_per_second": 1809.904
    },
    {
      "epoch": 1.0322927879440258,
      "grad_norm": 2.8264159482205287e-05,
      "learning_rate": 2.2086772881623455e-05,
      "loss": 0.1546,
      "num_input_tokens_seen": 1249280,
      "step": 480,
      "train_runtime": 689.1037,
      "train_tokens_per_second": 1812.906
    },
    {
      "epoch": 1.0538213132400431,
      "grad_norm": 6.062540705897845e-05,
      "learning_rate": 2.1787273764556827e-05,
      "loss": 0.1137,
      "num_input_tokens_seen": 1275520,
      "step": 490,
      "train_runtime": 702.2966,
      "train_tokens_per_second": 1816.213
    },
    {
      "epoch": 1.0753498385360603,
      "grad_norm": 3.901288073393516e-05,
      "learning_rate": 2.1484332508982167e-05,
      "loss": 0.1057,
      "num_input_tokens_seen": 1302080,
      "step": 500,
      "train_runtime": 715.609,
      "train_tokens_per_second": 1819.541
    },
    {
      "epoch": 1.0968783638320776,
      "grad_norm": 2.3066615540301427e-05,
      "learning_rate": 2.11781027503351e-05,
      "loss": 0.1577,
      "num_input_tokens_seen": 1327936,
      "step": 510,
      "train_runtime": 728.7255,
      "train_tokens_per_second": 1822.272
    },
    {
      "epoch": 1.1184068891280947,
      "grad_norm": 3.826987085631117e-05,
      "learning_rate": 2.0868739791802312e-05,
      "loss": 0.136,
      "num_input_tokens_seen": 1354016,
      "step": 520,
      "train_runtime": 741.8823,
      "train_tokens_per_second": 1825.109
    },
    {
      "epoch": 1.1399354144241118,
      "grad_norm": 4.135877316002734e-05,
      "learning_rate": 2.0556400525560168e-05,
      "loss": 0.1125,
      "num_input_tokens_seen": 1379520,
      "step": 530,
      "train_runtime": 754.9858,
      "train_tokens_per_second": 1827.213
    },
    {
      "epoch": 1.1614639397201292,
      "grad_norm": 2.640750608406961e-05,
      "learning_rate": 2.0241243353207484e-05,
      "loss": 0.1171,
      "num_input_tokens_seen": 1405408,
      "step": 540,
      "train_runtime": 768.1512,
      "train_tokens_per_second": 1829.598
    },
    {
      "epoch": 1.1829924650161463,
      "grad_norm": 2.962505277537275e-05,
      "learning_rate": 1.9923428105432796e-05,
      "loss": 0.1134,
      "num_input_tokens_seen": 1432160,
      "step": 550,
      "train_runtime": 781.5035,
      "train_tokens_per_second": 1832.57
    },
    {
      "epoch": 1.2045209903121636,
      "grad_norm": 3.0228229661588557e-05,
      "learning_rate": 1.9603115960956852e-05,
      "loss": 0.1468,
      "num_input_tokens_seen": 1458304,
      "step": 560,
      "train_runtime": 794.6258,
      "train_tokens_per_second": 1835.209
    },
    {
      "epoch": 1.2260495156081808,
      "grad_norm": 2.415182279946748e-05,
      "learning_rate": 1.9280469364791505e-05,
      "loss": 0.1352,
      "num_input_tokens_seen": 1483840,
      "step": 570,
      "train_runtime": 807.6893,
      "train_tokens_per_second": 1837.142
    },
    {
      "epoch": 1.247578040904198,
      "grad_norm": 4.6627985284430906e-05,
      "learning_rate": 1.8955651945856354e-05,
      "loss": 0.1276,
      "num_input_tokens_seen": 1509888,
      "step": 580,
      "train_runtime": 820.8961,
      "train_tokens_per_second": 1839.317
    },
    {
      "epoch": 1.2691065662002152,
      "grad_norm": 3.416907929931767e-05,
      "learning_rate": 1.8628828433995013e-05,
      "loss": 0.113,
      "num_input_tokens_seen": 1535616,
      "step": 590,
      "train_runtime": 834.0516,
      "train_tokens_per_second": 1841.152
    },
    {
      "epoch": 1.2906350914962326,
      "grad_norm": 2.523140028642956e-05,
      "learning_rate": 1.8300164576433035e-05,
      "loss": 0.1236,
      "num_input_tokens_seen": 1561216,
      "step": 600,
      "train_runtime": 847.0655,
      "train_tokens_per_second": 1843.088
    },
    {
      "epoch": 1.3121636167922497,
      "grad_norm": 3.7026082281954587e-05,
      "learning_rate": 1.7969827053719862e-05,
      "loss": 0.1343,
      "num_input_tokens_seen": 1587776,
      "step": 610,
      "train_runtime": 860.3124,
      "train_tokens_per_second": 1845.581
    },
    {
      "epoch": 1.333692142088267,
      "grad_norm": 3.648900383268483e-05,
      "learning_rate": 1.7637983395197465e-05,
      "loss": 0.1611,
      "num_input_tokens_seen": 1613184,
      "step": 620,
      "train_runtime": 873.3346,
      "train_tokens_per_second": 1847.154
    },
    {
      "epoch": 1.3552206673842842,
      "grad_norm": 4.887824979959987e-05,
      "learning_rate": 1.7304801894038522e-05,
      "loss": 0.1318,
      "num_input_tokens_seen": 1639776,
      "step": 630,
      "train_runtime": 886.6116,
      "train_tokens_per_second": 1849.486
    },
    {
      "epoch": 1.3767491926803013,
      "grad_norm": 3.865800681523979e-05,
      "learning_rate": 1.6970451521897198e-05,
      "loss": 0.1193,
      "num_input_tokens_seen": 1666016,
      "step": 640,
      "train_runtime": 899.781,
      "train_tokens_per_second": 1851.579
    },
    {
      "epoch": 1.3982777179763186,
      "grad_norm": 8.725399675313383e-05,
      "learning_rate": 1.663510184321587e-05,
      "loss": 0.1691,
      "num_input_tokens_seen": 1691680,
      "step": 650,
      "train_runtime": 912.8286,
      "train_tokens_per_second": 1853.229
    },
    {
      "epoch": 1.419806243272336,
      "grad_norm": 3.235097756260075e-05,
      "learning_rate": 1.6298922929231194e-05,
      "loss": 0.1116,
      "num_input_tokens_seen": 1717472,
      "step": 660,
      "train_runtime": 925.9227,
      "train_tokens_per_second": 1854.876
    },
    {
      "epoch": 1.441334768568353,
      "grad_norm": 4.359897502581589e-05,
      "learning_rate": 1.5962085271723158e-05,
      "loss": 0.1194,
      "num_input_tokens_seen": 1743648,
      "step": 670,
      "train_runtime": 939.0604,
      "train_tokens_per_second": 1856.801
    },
    {
      "epoch": 1.4628632938643702,
      "grad_norm": 5.545392195926979e-05,
      "learning_rate": 1.5624759696550874e-05,
      "loss": 0.1481,
      "num_input_tokens_seen": 1770112,
      "step": 680,
      "train_runtime": 952.3362,
      "train_tokens_per_second": 1858.705
    },
    {
      "epoch": 1.4843918191603875,
      "grad_norm": 4.8013796913437545e-05,
      "learning_rate": 1.528711727701893e-05,
      "loss": 0.1086,
      "num_input_tokens_seen": 1796256,
      "step": 690,
      "train_runtime": 965.4677,
      "train_tokens_per_second": 1860.503
    },
    {
      "epoch": 1.5059203444564049,
      "grad_norm": 6.239710637601092e-05,
      "learning_rate": 1.4949329247118258e-05,
      "loss": 0.1263,
      "num_input_tokens_seen": 1821952,
      "step": 700,
      "train_runtime": 978.7232,
      "train_tokens_per_second": 1861.56
    },
    {
      "epoch": 1.527448869752422,
      "grad_norm": 5.263877392280847e-05,
      "learning_rate": 1.4611566914685489e-05,
      "loss": 0.138,
      "num_input_tokens_seen": 1848096,
      "step": 710,
      "train_runtime": 992.1376,
      "train_tokens_per_second": 1862.742
    },
    {
      "epoch": 1.5489773950484391,
      "grad_norm": 3.61037346010562e-05,
      "learning_rate": 1.4274001574524894e-05,
      "loss": 0.1202,
      "num_input_tokens_seen": 1874304,
      "step": 720,
      "train_runtime": 1005.4875,
      "train_tokens_per_second": 1864.075
    },
    {
      "epoch": 1.5705059203444565,
      "grad_norm": 3.5984608985017985e-05,
      "learning_rate": 1.3936804421536906e-05,
      "loss": 0.0985,
      "num_input_tokens_seen": 1899392,
      "step": 730,
      "train_runtime": 1018.7219,
      "train_tokens_per_second": 1864.485
    },
    {
      "epoch": 1.5920344456404736,
      "grad_norm": 1.9361184968147427e-05,
      "learning_rate": 1.3600146463897327e-05,
      "loss": 0.1293,
      "num_input_tokens_seen": 1925344,
      "step": 740,
      "train_runtime": 1032.1312,
      "train_tokens_per_second": 1865.406
    },
    {
      "epoch": 1.6135629709364907,
      "grad_norm": 2.6571358830551617e-05,
      "learning_rate": 1.3264198436331193e-05,
      "loss": 0.1432,
      "num_input_tokens_seen": 1951520,
      "step": 750,
      "train_runtime": 1045.5198,
      "train_tokens_per_second": 1866.555
    },
    {
      "epoch": 1.635091496232508,
      "grad_norm": 1.8363367416895926e-05,
      "learning_rate": 1.2929130713525396e-05,
      "loss": 0.1164,
      "num_input_tokens_seen": 1977216,
      "step": 760,
      "train_runtime": 1058.8676,
      "train_tokens_per_second": 1867.293
    },
    {
      "epoch": 1.6566200215285254,
      "grad_norm": 2.6945261197397485e-05,
      "learning_rate": 1.2595113223723788e-05,
      "loss": 0.1421,
      "num_input_tokens_seen": 2003648,
      "step": 770,
      "train_runtime": 1072.266,
      "train_tokens_per_second": 1868.611
    },
    {
      "epoch": 1.6781485468245425,
      "grad_norm": 5.403673276305199e-05,
      "learning_rate": 1.2262315362548825e-05,
      "loss": 0.1076,
      "num_input_tokens_seen": 2030432,
      "step": 780,
      "train_runtime": 1085.8231,
      "train_tokens_per_second": 1869.947
    },
    {
      "epoch": 1.6996770721205596,
      "grad_norm": 7.509243732783943e-05,
      "learning_rate": 1.1930905907093196e-05,
      "loss": 0.1346,
      "num_input_tokens_seen": 2055104,
      "step": 790,
      "train_runtime": 1099.0396,
      "train_tokens_per_second": 1869.909
    },
    {
      "epoch": 1.721205597416577,
      "grad_norm": 6.094072887208313e-05,
      "learning_rate": 1.1601052930325236e-05,
      "loss": 0.1251,
      "num_input_tokens_seen": 2081248,
      "step": 800,
      "train_runtime": 1112.404,
      "train_tokens_per_second": 1870.946
    },
    {
      "epoch": 1.7427341227125943,
      "grad_norm": 4.0481922042090446e-05,
      "learning_rate": 1.127292371585141e-05,
      "loss": 0.1267,
      "num_input_tokens_seen": 2107520,
      "step": 810,
      "train_runtime": 1125.7611,
      "train_tokens_per_second": 1872.085
    },
    {
      "epoch": 1.7642626480086114,
      "grad_norm": 4.967824861523695e-05,
      "learning_rate": 1.0946684673079074e-05,
      "loss": 0.1531,
      "num_input_tokens_seen": 2133856,
      "step": 820,
      "train_runtime": 1138.9801,
      "train_tokens_per_second": 1873.48
    },
    {
      "epoch": 1.7857911733046286,
      "grad_norm": 3.5426357499090955e-05,
      "learning_rate": 1.0622501252822628e-05,
      "loss": 0.1083,
      "num_input_tokens_seen": 2159904,
      "step": 830,
      "train_runtime": 1152.1924,
      "train_tokens_per_second": 1874.604
    },
    {
      "epoch": 1.807319698600646,
      "grad_norm": 3.692304380820133e-05,
      "learning_rate": 1.0300537863395823e-05,
      "loss": 0.1353,
      "num_input_tokens_seen": 2186368,
      "step": 840,
      "train_runtime": 1165.4112,
      "train_tokens_per_second": 1876.049
    },
    {
      "epoch": 1.8288482238966632,
      "grad_norm": 3.939384259865619e-05,
      "learning_rate": 9.98095778723271e-06,
      "loss": 0.1008,
      "num_input_tokens_seen": 2212256,
      "step": 850,
      "train_runtime": 1178.5802,
      "train_tokens_per_second": 1877.052
    },
    {
      "epoch": 1.8503767491926801,
      "grad_norm": 3.193708835169673e-05,
      "learning_rate": 9.663923098079656e-06,
      "loss": 0.1227,
      "num_input_tokens_seen": 2238432,
      "step": 860,
      "train_runtime": 1191.8467,
      "train_tokens_per_second": 1878.121
    },
    {
      "epoch": 1.8719052744886975,
      "grad_norm": 1.8965831259265542e-05,
      "learning_rate": 9.349594578800259e-06,
      "loss": 0.1169,
      "num_input_tokens_seen": 2264896,
      "step": 870,
      "train_runtime": 1205.1831,
      "train_tokens_per_second": 1879.296
    },
    {
      "epoch": 1.8934337997847148,
      "grad_norm": 4.2244591895723715e-05,
      "learning_rate": 9.038131639834977e-06,
      "loss": 0.1481,
      "num_input_tokens_seen": 2290976,
      "step": 880,
      "train_runtime": 1218.4291,
      "train_tokens_per_second": 1880.27
    },
    {
      "epoch": 1.914962325080732,
      "grad_norm": 1.6058927940321155e-05,
      "learning_rate": 8.7296922383568e-06,
      "loss": 0.1319,
      "num_input_tokens_seen": 2316608,
      "step": 890,
      "train_runtime": 1231.5483,
      "train_tokens_per_second": 1881.053
    },
    {
      "epoch": 1.936490850376749,
      "grad_norm": 3.993209975305945e-05,
      "learning_rate": 8.424432798163838e-06,
      "loss": 0.1133,
      "num_input_tokens_seen": 2342208,
      "step": 900,
      "train_runtime": 1244.6398,
      "train_tokens_per_second": 1881.836
    },
    {
      "epoch": 1.9580193756727664,
      "grad_norm": 2.757447145995684e-05,
      "learning_rate": 8.122508130349711e-06,
      "loss": 0.0905,
      "num_input_tokens_seen": 2367968,
      "step": 910,
      "train_runtime": 1257.7972,
      "train_tokens_per_second": 1882.631
    },
    {
      "epoch": 1.9795479009687837,
      "grad_norm": 3.641565490397625e-05,
      "learning_rate": 7.82407135479164e-06,
      "loss": 0.1374,
      "num_input_tokens_seen": 2393920,
      "step": 920,
      "train_runtime": 1270.9937,
      "train_tokens_per_second": 1883.503
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.0924720956827514e-05,
      "learning_rate": 7.529273822496379e-06,
      "loss": 0.0857,
      "num_input_tokens_seen": 2419136,
      "step": 930,
      "train_runtime": 1283.5615,
      "train_tokens_per_second": 1884.706
    },
    {
      "epoch": 2.0215285252960173,
      "grad_norm": 5.763619265053421e-05,
      "learning_rate": 7.238265038843143e-06,
      "loss": 0.129,
      "num_input_tokens_seen": 2444992,
      "step": 940,
      "train_runtime": 1296.665,
      "train_tokens_per_second": 1885.6
    },
    {
      "epoch": 2.0430570505920342,
      "grad_norm": 4.3982894567307085e-05,
      "learning_rate": 6.951192587762555e-06,
      "loss": 0.0711,
      "num_input_tokens_seen": 2471008,
      "step": 950,
      "train_runtime": 1309.7833,
      "train_tokens_per_second": 1886.578
    },
    {
      "epoch": 2.0645855758880516,
      "grad_norm": 1.84689779416658e-05,
      "learning_rate": 6.668202056890052e-06,
      "loss": 0.1049,
      "num_input_tokens_seen": 2497152,
      "step": 960,
      "train_runtime": 1322.9959,
      "train_tokens_per_second": 1887.498
    },
    {
      "epoch": 2.086114101184069,
      "grad_norm": 3.773774005821906e-05,
      "learning_rate": 6.389436963731707e-06,
      "loss": 0.0965,
      "num_input_tokens_seen": 2523328,
      "step": 970,
      "train_runtime": 1336.2038,
      "train_tokens_per_second": 1888.43
    },
    {
      "epoch": 2.1076426264800863,
      "grad_norm": 3.176877726218663e-05,
      "learning_rate": 6.115038682879904e-06,
      "loss": 0.1418,
      "num_input_tokens_seen": 2549792,
      "step": 980,
      "train_runtime": 1349.5257,
      "train_tokens_per_second": 1889.399
    },
    {
      "epoch": 2.129171151776103,
      "grad_norm": 2.900951585615985e-05,
      "learning_rate": 5.845146374315781e-06,
      "loss": 0.0946,
      "num_input_tokens_seen": 2575520,
      "step": 990,
      "train_runtime": 1362.6176,
      "train_tokens_per_second": 1890.127
    },
    {
      "epoch": 2.1506996770721205,
      "grad_norm": 4.064288077643141e-05,
      "learning_rate": 5.579896912834807e-06,
      "loss": 0.1227,
      "num_input_tokens_seen": 2601888,
      "step": 1000,
      "train_runtime": 1375.8117,
      "train_tokens_per_second": 1891.166
    },
    {
      "epoch": 2.172228202368138,
      "grad_norm": 6.559904431924224e-05,
      "learning_rate": 5.319424818631293e-06,
      "loss": 0.1208,
      "num_input_tokens_seen": 2628064,
      "step": 1010,
      "train_runtime": 1390.5019,
      "train_tokens_per_second": 1890.011
    },
    {
      "epoch": 2.193756727664155,
      "grad_norm": 2.7879113986273296e-05,
      "learning_rate": 5.063862189076988e-06,
      "loss": 0.112,
      "num_input_tokens_seen": 2654528,
      "step": 1020,
      "train_runtime": 1403.7726,
      "train_tokens_per_second": 1890.996
    },
    {
      "epoch": 2.215285252960172,
      "grad_norm": 4.046166213811375e-05,
      "learning_rate": 4.813338631728471e-06,
      "loss": 0.1082,
      "num_input_tokens_seen": 2680864,
      "step": 1030,
      "train_runtime": 1416.9915,
      "train_tokens_per_second": 1891.941
    },
    {
      "epoch": 2.2368137782561894,
      "grad_norm": 6.314706115517765e-05,
      "learning_rate": 4.567981198597168e-06,
      "loss": 0.1173,
      "num_input_tokens_seen": 2707328,
      "step": 1040,
      "train_runtime": 1430.247,
      "train_tokens_per_second": 1892.909
    },
    {
      "epoch": 2.2583423035522068,
      "grad_norm": 2.4318607756868005e-05,
      "learning_rate": 4.327914321715459e-06,
      "loss": 0.0901,
      "num_input_tokens_seen": 2733184,
      "step": 1050,
      "train_runtime": 1443.4569,
      "train_tokens_per_second": 1893.499
    },
    {
      "epoch": 2.2798708288482237,
      "grad_norm": 2.560779103077948e-05,
      "learning_rate": 4.09325975003146e-06,
      "loss": 0.1068,
      "num_input_tokens_seen": 2758912,
      "step": 1060,
      "train_runtime": 1456.6406,
      "train_tokens_per_second": 1894.024
    },
    {
      "epoch": 2.301399354144241,
      "grad_norm": 3.13112250296399e-05,
      "learning_rate": 3.864136487664538e-06,
      "loss": 0.0904,
      "num_input_tokens_seen": 2784480,
      "step": 1070,
      "train_runtime": 1469.7536,
      "train_tokens_per_second": 1894.522
    },
    {
      "epoch": 2.3229278794402584,
      "grad_norm": 2.460617179167457e-05,
      "learning_rate": 3.640660733552863e-06,
      "loss": 0.0633,
      "num_input_tokens_seen": 2810176,
      "step": 1080,
      "train_runtime": 1482.8577,
      "train_tokens_per_second": 1895.108
    },
    {
      "epoch": 2.3444564047362757,
      "grad_norm": 4.727970008389093e-05,
      "learning_rate": 3.4229458225235696e-06,
      "loss": 0.0847,
      "num_input_tokens_seen": 2836896,
      "step": 1090,
      "train_runtime": 1496.1489,
      "train_tokens_per_second": 1896.132
    },
    {
      "epoch": 2.3659849300322926,
      "grad_norm": 2.8999687856412493e-05,
      "learning_rate": 3.211102167815469e-06,
      "loss": 0.0959,
      "num_input_tokens_seen": 2862464,
      "step": 1100,
      "train_runtime": 1509.2497,
      "train_tokens_per_second": 1896.614
    },
    {
      "epoch": 2.38751345532831,
      "grad_norm": 8.412307215621695e-05,
      "learning_rate": 3.0052372050834025e-06,
      "loss": 0.1154,
      "num_input_tokens_seen": 2889056,
      "step": 1110,
      "train_runtime": 1522.5818,
      "train_tokens_per_second": 1897.472
    },
    {
      "epoch": 2.4090419806243273,
      "grad_norm": 4.532782259047963e-05,
      "learning_rate": 2.805455337912722e-06,
      "loss": 0.1063,
      "num_input_tokens_seen": 2915392,
      "step": 1120,
      "train_runtime": 1535.9475,
      "train_tokens_per_second": 1898.107
    },
    {
      "epoch": 2.4305705059203446,
      "grad_norm": 4.3450949306134135e-05,
      "learning_rate": 2.611857884871379e-06,
      "loss": 0.1067,
      "num_input_tokens_seen": 2941472,
      "step": 1130,
      "train_runtime": 1549.1638,
      "train_tokens_per_second": 1898.748
    },
    {
      "epoch": 2.4520990312163615,
      "grad_norm": 2.563163434388116e-05,
      "learning_rate": 2.424543028126661e-06,
      "loss": 0.1194,
      "num_input_tokens_seen": 2967424,
      "step": 1140,
      "train_runtime": 1562.3102,
      "train_tokens_per_second": 1899.382
    },
    {
      "epoch": 2.473627556512379,
      "grad_norm": 3.6229073884896934e-05,
      "learning_rate": 2.243605763652489e-06,
      "loss": 0.0986,
      "num_input_tokens_seen": 2993856,
      "step": 1150,
      "train_runtime": 1575.5085,
      "train_tokens_per_second": 1900.247
    },
    {
      "epoch": 2.495156081808396,
      "grad_norm": 2.7871834390680306e-05,
      "learning_rate": 2.0691378530525933e-06,
      "loss": 0.1069,
      "num_input_tokens_seen": 3019648,
      "step": 1160,
      "train_runtime": 1588.5936,
      "train_tokens_per_second": 1900.831
    },
    {
      "epoch": 2.5166846071044136,
      "grad_norm": 4.834825813304633e-05,
      "learning_rate": 1.9012277770239829e-06,
      "loss": 0.1124,
      "num_input_tokens_seen": 3045504,
      "step": 1170,
      "train_runtime": 1601.6858,
      "train_tokens_per_second": 1901.437
    },
    {
      "epoch": 2.5382131324004304,
      "grad_norm": 1.700134453130886e-05,
      "learning_rate": 1.7399606904843224e-06,
      "loss": 0.0751,
      "num_input_tokens_seen": 3071424,
      "step": 1180,
      "train_runtime": 1614.839,
      "train_tokens_per_second": 1902.0
    },
    {
      "epoch": 2.559741657696448,
      "grad_norm": 6.077123907743953e-05,
      "learning_rate": 1.5854183793859577e-06,
      "loss": 0.1124,
      "num_input_tokens_seen": 3098112,
      "step": 1190,
      "train_runtime": 1628.2435,
      "train_tokens_per_second": 1902.733
    },
    {
      "epoch": 2.581270182992465,
      "grad_norm": 2.4882172510842793e-05,
      "learning_rate": 1.437679219238498e-06,
      "loss": 0.0872,
      "num_input_tokens_seen": 3124128,
      "step": 1200,
      "train_runtime": 1641.5054,
      "train_tokens_per_second": 1903.209
    },
    {
      "epoch": 2.602798708288482,
      "grad_norm": 5.312330540618859e-05,
      "learning_rate": 1.2968181353609854e-06,
      "loss": 0.1071,
      "num_input_tokens_seen": 3150400,
      "step": 1210,
      "train_runtime": 1654.8347,
      "train_tokens_per_second": 1903.755
    },
    {
      "epoch": 2.6243272335844994,
      "grad_norm": 2.769859020190779e-05,
      "learning_rate": 1.1629065648838495e-06,
      "loss": 0.1054,
      "num_input_tokens_seen": 3175680,
      "step": 1220,
      "train_runtime": 1667.8812,
      "train_tokens_per_second": 1904.021
    },
    {
      "epoch": 2.6458557588805167,
      "grad_norm": 3.681317684822716e-05,
      "learning_rate": 1.0360124205198102e-06,
      "loss": 0.1029,
      "num_input_tokens_seen": 3201664,
      "step": 1230,
      "train_runtime": 1681.0555,
      "train_tokens_per_second": 1904.556
    },
    {
      "epoch": 2.667384284176534,
      "grad_norm": 7.651891792193055e-05,
      "learning_rate": 9.162000561222672e-07,
      "loss": 0.1005,
      "num_input_tokens_seen": 3228192,
      "step": 1240,
      "train_runtime": 1694.3825,
      "train_tokens_per_second": 1905.232
    },
    {
      "epoch": 2.6889128094725514,
      "grad_norm": 3.385964373592287e-05,
      "learning_rate": 8.035302340484646e-07,
      "loss": 0.1126,
      "num_input_tokens_seen": 3253664,
      "step": 1250,
      "train_runtime": 1707.5029,
      "train_tokens_per_second": 1905.51
    },
    {
      "epoch": 2.7104413347685683,
      "grad_norm": 2.672344635357149e-05,
      "learning_rate": 6.980600943441273e-07,
      "loss": 0.0842,
      "num_input_tokens_seen": 3279552,
      "step": 1260,
      "train_runtime": 1720.6468,
      "train_tokens_per_second": 1906.0
    },
    {
      "epoch": 2.7319698600645856,
      "grad_norm": 4.816047294298187e-05,
      "learning_rate": 5.998431257651132e-07,
      "loss": 0.1228,
      "num_input_tokens_seen": 3305888,
      "step": 1270,
      "train_runtime": 1733.8693,
      "train_tokens_per_second": 1906.653
    },
    {
      "epoch": 2.7534983853606025,
      "grad_norm": 4.461811840883456e-05,
      "learning_rate": 5.089291386508055e-07,
      "loss": 0.0889,
      "num_input_tokens_seen": 3331840,
      "step": 1280,
      "train_runtime": 1747.1212,
      "train_tokens_per_second": 1907.046
    },
    {
      "epoch": 2.77502691065662,
      "grad_norm": 2.788665187836159e-05,
      "learning_rate": 4.2536423966299256e-07,
      "loss": 0.0862,
      "num_input_tokens_seen": 3357632,
      "step": 1290,
      "train_runtime": 1760.2352,
      "train_tokens_per_second": 1907.491
    },
    {
      "epoch": 2.7965554359526372,
      "grad_norm": 5.6285829487023875e-05,
      "learning_rate": 3.491908084030776e-07,
      "loss": 0.116,
      "num_input_tokens_seen": 3384000,
      "step": 1300,
      "train_runtime": 1773.5214,
      "train_tokens_per_second": 1908.068
    },
    {
      "epoch": 2.8180839612486546,
      "grad_norm": 4.5707623939961195e-05,
      "learning_rate": 2.804474759194176e-07,
      "loss": 0.1054,
      "num_input_tokens_seen": 3410368,
      "step": 1310,
      "train_runtime": 1786.7429,
      "train_tokens_per_second": 1908.707
    },
    {
      "epoch": 2.839612486544672,
      "grad_norm": 2.5656630896264687e-05,
      "learning_rate": 2.1916910511575627e-07,
      "loss": 0.1038,
      "num_input_tokens_seen": 3436224,
      "step": 1320,
      "train_runtime": 1799.8495,
      "train_tokens_per_second": 1909.173
    },
    {
      "epoch": 2.861141011840689,
      "grad_norm": 3.635259417933412e-05,
      "learning_rate": 1.6538677307062933e-07,
      "loss": 0.086,
      "num_input_tokens_seen": 3462304,
      "step": 1330,
      "train_runtime": 1813.0288,
      "train_tokens_per_second": 1909.68
    },
    {
      "epoch": 2.882669537136706,
      "grad_norm": 6.195970490807667e-05,
      "learning_rate": 1.1912775527676278e-07,
      "loss": 0.0935,
      "num_input_tokens_seen": 3488128,
      "step": 1340,
      "train_runtime": 1826.2193,
      "train_tokens_per_second": 1910.027
    },
    {
      "epoch": 2.9041980624327235,
      "grad_norm": 4.8543439334025607e-05,
      "learning_rate": 8.04155118084099e-08,
      "loss": 0.1117,
      "num_input_tokens_seen": 3513728,
      "step": 1350,
      "train_runtime": 1839.3642,
      "train_tokens_per_second": 1910.295
    },
    {
      "epoch": 2.9257265877287404,
      "grad_norm": 3.0174160201568156e-05,
      "learning_rate": 4.9269675423671556e-08,
      "loss": 0.1099,
      "num_input_tokens_seen": 3539936,
      "step": 1360,
      "train_runtime": 1852.5438,
      "train_tokens_per_second": 1910.851
    },
    {
      "epoch": 2.9472551130247577,
      "grad_norm": 5.9879115724470466e-05,
      "learning_rate": 2.570604160781753e-08,
      "loss": 0.087,
      "num_input_tokens_seen": 3565792,
      "step": 1370,
      "train_runtime": 1865.7413,
      "train_tokens_per_second": 1911.193
    },
    {
      "epoch": 2.968783638320775,
      "grad_norm": 3.549754183040932e-05,
      "learning_rate": 9.736560562675912e-09,
      "loss": 0.0705,
      "num_input_tokens_seen": 3591744,
      "step": 1380,
      "train_runtime": 1878.9623,
      "train_tokens_per_second": 1911.557
    },
    {
      "epoch": 2.9903121636167924,
      "grad_norm": 3.281807221355848e-05,
      "learning_rate": 1.3693311461326908e-09,
      "loss": 0.0982,
      "num_input_tokens_seen": 3618112,
      "step": 1390,
      "train_runtime": 1892.4676,
      "train_tokens_per_second": 1911.849
    },
    {
      "epoch": 3.0,
      "num_input_tokens_seen": 3629856,
      "step": 1395,
      "total_flos": 1.818821603079291e+17,
      "train_loss": 0.15204335844217662,
      "train_runtime": 1900.0282,
      "train_samples_per_second": 5.867,
      "train_steps_per_second": 0.734
    }
  ],
  "logging_steps": 10,
  "max_steps": 1395,
  "num_input_tokens_seen": 3629856,
  "num_train_epochs": 3,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.818821603079291e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}