{
  "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": 0.000117536197649315,
      "learning_rate": 2.9996919063517692e-05,
      "loss": 0.8869,
      "num_input_tokens_seen": 36480,
      "step": 10,
      "train_runtime": 21.2665,
      "train_tokens_per_second": 1715.375
    },
    {
      "epoch": 0.04305705059203445,
      "grad_norm": 0.00022027654631529003,
      "learning_rate": 2.9986270537572827e-05,
      "loss": 0.199,
      "num_input_tokens_seen": 72672,
      "step": 20,
      "train_runtime": 33.9024,
      "train_tokens_per_second": 2143.565
    },
    {
      "epoch": 0.06458557588805167,
      "grad_norm": 5.31423429492861e-05,
      "learning_rate": 2.9968021784983436e-05,
      "loss": 0.1809,
      "num_input_tokens_seen": 109472,
      "step": 30,
      "train_runtime": 47.2118,
      "train_tokens_per_second": 2318.741
    },
    {
      "epoch": 0.0861141011840689,
      "grad_norm": 5.581648292718455e-05,
      "learning_rate": 2.994218206053077e-05,
      "loss": 0.1766,
      "num_input_tokens_seen": 146656,
      "step": 40,
      "train_runtime": 60.0165,
      "train_tokens_per_second": 2443.597
    },
    {
      "epoch": 0.10764262648008611,
      "grad_norm": 4.6867833589203656e-05,
      "learning_rate": 2.9908764468726846e-05,
      "loss": 0.1784,
      "num_input_tokens_seen": 183200,
      "step": 50,
      "train_runtime": 72.7117,
      "train_tokens_per_second": 2519.54
    },
    {
      "epoch": 0.12917115177610333,
      "grad_norm": 7.248102338053286e-05,
      "learning_rate": 2.986778595716853e-05,
      "loss": 0.165,
      "num_input_tokens_seen": 220352,
      "step": 60,
      "train_runtime": 85.5372,
      "train_tokens_per_second": 2576.096
    },
    {
      "epoch": 0.15069967707212056,
      "grad_norm": 6.31138464086689e-05,
      "learning_rate": 2.981926730794264e-05,
      "loss": 0.1884,
      "num_input_tokens_seen": 256864,
      "step": 70,
      "train_runtime": 98.2829,
      "train_tokens_per_second": 2613.516
    },
    {
      "epoch": 0.1722282023681378,
      "grad_norm": 8.531038474757224e-05,
      "learning_rate": 2.9763233127086394e-05,
      "loss": 0.1668,
      "num_input_tokens_seen": 293408,
      "step": 80,
      "train_runtime": 110.9453,
      "train_tokens_per_second": 2644.619
    },
    {
      "epoch": 0.193756727664155,
      "grad_norm": 3.399947308935225e-05,
      "learning_rate": 2.9699711832108555e-05,
      "loss": 0.1538,
      "num_input_tokens_seen": 329152,
      "step": 90,
      "train_runtime": 123.3823,
      "train_tokens_per_second": 2667.742
    },
    {
      "epoch": 0.21528525296017223,
      "grad_norm": 6.389995542122051e-05,
      "learning_rate": 2.9628735637577612e-05,
      "loss": 0.1729,
      "num_input_tokens_seen": 365632,
      "step": 100,
      "train_runtime": 136.0062,
      "train_tokens_per_second": 2688.348
    },
    {
      "epoch": 0.23681377825618946,
      "grad_norm": 8.456028444925323e-05,
      "learning_rate": 2.955034053878428e-05,
      "loss": 0.1346,
      "num_input_tokens_seen": 401888,
      "step": 110,
      "train_runtime": 148.598,
      "train_tokens_per_second": 2704.531
    },
    {
      "epoch": 0.25834230355220666,
      "grad_norm": 6.639358616666868e-05,
      "learning_rate": 2.9464566293486653e-05,
      "loss": 0.1622,
      "num_input_tokens_seen": 438816,
      "step": 120,
      "train_runtime": 161.3801,
      "train_tokens_per_second": 2719.146
    },
    {
      "epoch": 0.2798708288482239,
      "grad_norm": 6.005055911373347e-05,
      "learning_rate": 2.9371456401747193e-05,
      "loss": 0.1357,
      "num_input_tokens_seen": 475648,
      "step": 130,
      "train_runtime": 174.163,
      "train_tokens_per_second": 2731.051
    },
    {
      "epoch": 0.3013993541442411,
      "grad_norm": 8.286784577649087e-05,
      "learning_rate": 2.9271058083871864e-05,
      "loss": 0.1537,
      "num_input_tokens_seen": 512256,
      "step": 140,
      "train_runtime": 186.8563,
      "train_tokens_per_second": 2741.444
    },
    {
      "epoch": 0.32292787944025836,
      "grad_norm": 4.636685480363667e-05,
      "learning_rate": 2.916342225646254e-05,
      "loss": 0.1183,
      "num_input_tokens_seen": 548992,
      "step": 150,
      "train_runtime": 199.6287,
      "train_tokens_per_second": 2750.065
    },
    {
      "epoch": 0.3444564047362756,
      "grad_norm": 4.919446655549109e-05,
      "learning_rate": 2.904860350659487e-05,
      "loss": 0.1336,
      "num_input_tokens_seen": 585792,
      "step": 160,
      "train_runtime": 212.4096,
      "train_tokens_per_second": 2757.841
    },
    {
      "epoch": 0.36598493003229277,
      "grad_norm": 8.222900942200795e-05,
      "learning_rate": 2.8926660064134637e-05,
      "loss": 0.1661,
      "num_input_tokens_seen": 621952,
      "step": 170,
      "train_runtime": 225.0588,
      "train_tokens_per_second": 2763.509
    },
    {
      "epoch": 0.38751345532831,
      "grad_norm": 3.515497155603953e-05,
      "learning_rate": 2.8797653772206726e-05,
      "loss": 0.1405,
      "num_input_tokens_seen": 658240,
      "step": 180,
      "train_runtime": 237.6697,
      "train_tokens_per_second": 2769.557
    },
    {
      "epoch": 0.40904198062432723,
      "grad_norm": 4.577370418701321e-05,
      "learning_rate": 2.8661650055831645e-05,
      "loss": 0.1456,
      "num_input_tokens_seen": 694880,
      "step": 190,
      "train_runtime": 250.3071,
      "train_tokens_per_second": 2776.11
    },
    {
      "epoch": 0.43057050592034446,
      "grad_norm": 5.2699797379318625e-05,
      "learning_rate": 2.8518717888745486e-05,
      "loss": 0.1158,
      "num_input_tokens_seen": 731968,
      "step": 200,
      "train_runtime": 263.1248,
      "train_tokens_per_second": 2781.828
    },
    {
      "epoch": 0.4520990312163617,
      "grad_norm": 4.86937933601439e-05,
      "learning_rate": 2.836892975842015e-05,
      "loss": 0.1293,
      "num_input_tokens_seen": 768416,
      "step": 210,
      "train_runtime": 275.7314,
      "train_tokens_per_second": 2786.828
    },
    {
      "epoch": 0.4736275565123789,
      "grad_norm": 5.121620051795617e-05,
      "learning_rate": 2.8212361629301655e-05,
      "loss": 0.1456,
      "num_input_tokens_seen": 805376,
      "step": 220,
      "train_runtime": 288.4976,
      "train_tokens_per_second": 2791.621
    },
    {
      "epoch": 0.4951560818083961,
      "grad_norm": 3.388872210052796e-05,
      "learning_rate": 2.8049092904285033e-05,
      "loss": 0.1428,
      "num_input_tokens_seen": 841472,
      "step": 230,
      "train_runtime": 301.0113,
      "train_tokens_per_second": 2795.483
    },
    {
      "epoch": 0.5166846071044133,
      "grad_norm": 5.026933649787679e-05,
      "learning_rate": 2.7879206384445526e-05,
      "loss": 0.1334,
      "num_input_tokens_seen": 878464,
      "step": 240,
      "train_runtime": 313.7308,
      "train_tokens_per_second": 2800.057
    },
    {
      "epoch": 0.5382131324004306,
      "grad_norm": 2.826541276590433e-05,
      "learning_rate": 2.7702788227046286e-05,
      "loss": 0.1414,
      "num_input_tokens_seen": 914432,
      "step": 250,
      "train_runtime": 326.1923,
      "train_tokens_per_second": 2803.352
    },
    {
      "epoch": 0.5597416576964478,
      "grad_norm": 5.1211100071668625e-05,
      "learning_rate": 2.751992790184413e-05,
      "loss": 0.1098,
      "num_input_tokens_seen": 952320,
      "step": 260,
      "train_runtime": 339.127,
      "train_tokens_per_second": 2808.151
    },
    {
      "epoch": 0.581270182992465,
      "grad_norm": 4.860994886257686e-05,
      "learning_rate": 2.733071814571527e-05,
      "loss": 0.1643,
      "num_input_tokens_seen": 988800,
      "step": 270,
      "train_runtime": 351.788,
      "train_tokens_per_second": 2810.784
    },
    {
      "epoch": 0.6027987082884823,
      "grad_norm": 3.192482836311683e-05,
      "learning_rate": 2.7135254915624213e-05,
      "loss": 0.1222,
      "num_input_tokens_seen": 1025952,
      "step": 280,
      "train_runtime": 364.6184,
      "train_tokens_per_second": 2813.769
    },
    {
      "epoch": 0.6243272335844995,
      "grad_norm": 6.410479545593262e-05,
      "learning_rate": 2.693363733995954e-05,
      "loss": 0.1489,
      "num_input_tokens_seen": 1063360,
      "step": 290,
      "train_runtime": 377.4609,
      "train_tokens_per_second": 2817.139
    },
    {
      "epoch": 0.6458557588805167,
      "grad_norm": 7.877861935412511e-05,
      "learning_rate": 2.672596766826138e-05,
      "loss": 0.1207,
      "num_input_tokens_seen": 1099648,
      "step": 300,
      "train_runtime": 390.138,
      "train_tokens_per_second": 2818.613
    },
    {
      "epoch": 0.667384284176534,
      "grad_norm": 5.999449058435857e-05,
      "learning_rate": 2.6512351219365927e-05,
      "loss": 0.1319,
      "num_input_tokens_seen": 1135712,
      "step": 310,
      "train_runtime": 402.7639,
      "train_tokens_per_second": 2819.796
    },
    {
      "epoch": 0.6889128094725512,
      "grad_norm": 2.934955591626931e-05,
      "learning_rate": 2.6292896327993467e-05,
      "loss": 0.1453,
      "num_input_tokens_seen": 1172384,
      "step": 320,
      "train_runtime": 415.4411,
      "train_tokens_per_second": 2822.022
    },
    {
      "epoch": 0.7104413347685683,
      "grad_norm": 4.606367292581126e-05,
      "learning_rate": 2.606771428980681e-05,
      "loss": 0.1202,
      "num_input_tokens_seen": 1209440,
      "step": 330,
      "train_runtime": 428.2678,
      "train_tokens_per_second": 2824.027
    },
    {
      "epoch": 0.7319698600645855,
      "grad_norm": 3.709704469656572e-05,
      "learning_rate": 2.5836919304968184e-05,
      "loss": 0.1283,
      "num_input_tokens_seen": 1245920,
      "step": 340,
      "train_runtime": 440.939,
      "train_tokens_per_second": 2825.606
    },
    {
      "epoch": 0.7534983853606028,
      "grad_norm": 4.9612910515861586e-05,
      "learning_rate": 2.5600628420223057e-05,
      "loss": 0.1374,
      "num_input_tokens_seen": 1284288,
      "step": 350,
      "train_runtime": 454.0391,
      "train_tokens_per_second": 2828.585
    },
    {
      "epoch": 0.77502691065662,
      "grad_norm": 3.349883991177194e-05,
      "learning_rate": 2.535896146954037e-05,
      "loss": 0.1355,
      "num_input_tokens_seen": 1320896,
      "step": 360,
      "train_runtime": 466.7256,
      "train_tokens_per_second": 2830.134
    },
    {
      "epoch": 0.7965554359526372,
      "grad_norm": 3.364904478075914e-05,
      "learning_rate": 2.5112041013339212e-05,
      "loss": 0.131,
      "num_input_tokens_seen": 1359264,
      "step": 370,
      "train_runtime": 479.8532,
      "train_tokens_per_second": 2832.667
    },
    {
      "epoch": 0.8180839612486545,
      "grad_norm": 3.689135701279156e-05,
      "learning_rate": 2.48599922763328e-05,
      "loss": 0.1034,
      "num_input_tokens_seen": 1396032,
      "step": 380,
      "train_runtime": 492.5082,
      "train_tokens_per_second": 2834.535
    },
    {
      "epoch": 0.8396124865446717,
      "grad_norm": 3.208614361938089e-05,
      "learning_rate": 2.4602943084021286e-05,
      "loss": 0.1465,
      "num_input_tokens_seen": 1432224,
      "step": 390,
      "train_runtime": 505.1852,
      "train_tokens_per_second": 2835.047
    },
    {
      "epoch": 0.8611410118406889,
      "grad_norm": 5.1129703933838755e-05,
      "learning_rate": 2.434102379786553e-05,
      "loss": 0.1049,
      "num_input_tokens_seen": 1468224,
      "step": 400,
      "train_runtime": 517.7561,
      "train_tokens_per_second": 2835.744
    },
    {
      "epoch": 0.8826695371367062,
      "grad_norm": 5.446456270874478e-05,
      "learning_rate": 2.4074367249174864e-05,
      "loss": 0.1024,
      "num_input_tokens_seen": 1505088,
      "step": 410,
      "train_runtime": 530.4568,
      "train_tokens_per_second": 2837.343
    },
    {
      "epoch": 0.9041980624327234,
      "grad_norm": 8.551870996598154e-05,
      "learning_rate": 2.380310867174219e-05,
      "loss": 0.1155,
      "num_input_tokens_seen": 1542048,
      "step": 420,
      "train_runtime": 543.2418,
      "train_tokens_per_second": 2838.603
    },
    {
      "epoch": 0.9257265877287406,
      "grad_norm": 2.7210369808017276e-05,
      "learning_rate": 2.3527385633260724e-05,
      "loss": 0.117,
      "num_input_tokens_seen": 1578496,
      "step": 430,
      "train_runtime": 555.7658,
      "train_tokens_per_second": 2840.218
    },
    {
      "epoch": 0.9472551130247578,
      "grad_norm": 3.511784234433435e-05,
      "learning_rate": 2.3247337965557097e-05,
      "loss": 0.1023,
      "num_input_tokens_seen": 1613440,
      "step": 440,
      "train_runtime": 568.0483,
      "train_tokens_per_second": 2840.322
    },
    {
      "epoch": 0.9687836383207751,
      "grad_norm": 2.4012464564293623e-05,
      "learning_rate": 2.2963107693676192e-05,
      "loss": 0.1046,
      "num_input_tokens_seen": 1649760,
      "step": 450,
      "train_runtime": 580.7665,
      "train_tokens_per_second": 2840.66
    },
    {
      "epoch": 0.9903121636167922,
      "grad_norm": 4.6500768803525716e-05,
      "learning_rate": 2.2674838963853738e-05,
      "loss": 0.1106,
      "num_input_tokens_seen": 1686016,
      "step": 460,
      "train_runtime": 593.3342,
      "train_tokens_per_second": 2841.596
    },
    {
      "epoch": 1.0107642626480087,
      "grad_norm": 0.00010867212404264137,
      "learning_rate": 2.2382677970413097e-05,
      "loss": 0.0859,
      "num_input_tokens_seen": 1721152,
      "step": 470,
      "train_runtime": 605.4653,
      "train_tokens_per_second": 2842.693
    },
    {
      "epoch": 1.0322927879440258,
      "grad_norm": 3.351278428453952e-05,
      "learning_rate": 2.2086772881623455e-05,
      "loss": 0.1089,
      "num_input_tokens_seen": 1757568,
      "step": 480,
      "train_runtime": 618.0558,
      "train_tokens_per_second": 2843.705
    },
    {
      "epoch": 1.0538213132400431,
      "grad_norm": 4.3039726733695716e-05,
      "learning_rate": 2.1787273764556827e-05,
      "loss": 0.0852,
      "num_input_tokens_seen": 1794848,
      "step": 490,
      "train_runtime": 630.8577,
      "train_tokens_per_second": 2845.092
    },
    {
      "epoch": 1.0753498385360603,
      "grad_norm": 4.8467773012816906e-05,
      "learning_rate": 2.1484332508982167e-05,
      "loss": 0.0988,
      "num_input_tokens_seen": 1832256,
      "step": 500,
      "train_runtime": 643.6632,
      "train_tokens_per_second": 2846.607
    },
    {
      "epoch": 1.0968783638320776,
      "grad_norm": 3.706678398884833e-05,
      "learning_rate": 2.11781027503351e-05,
      "loss": 0.1001,
      "num_input_tokens_seen": 1868928,
      "step": 510,
      "train_runtime": 656.3448,
      "train_tokens_per_second": 2847.479
    },
    {
      "epoch": 1.1184068891280947,
      "grad_norm": 4.0160564822144806e-05,
      "learning_rate": 2.0868739791802312e-05,
      "loss": 0.1107,
      "num_input_tokens_seen": 1905824,
      "step": 520,
      "train_runtime": 669.0663,
      "train_tokens_per_second": 2848.483
    },
    {
      "epoch": 1.1399354144241118,
      "grad_norm": 7.405655196635053e-05,
      "learning_rate": 2.0556400525560168e-05,
      "loss": 0.0894,
      "num_input_tokens_seen": 1941760,
      "step": 530,
      "train_runtime": 681.5699,
      "train_tokens_per_second": 2848.952
    },
    {
      "epoch": 1.1614639397201292,
      "grad_norm": 4.077348785358481e-05,
      "learning_rate": 2.0241243353207484e-05,
      "loss": 0.1016,
      "num_input_tokens_seen": 1978208,
      "step": 540,
      "train_runtime": 694.297,
      "train_tokens_per_second": 2849.224
    },
    {
      "epoch": 1.1829924650161463,
      "grad_norm": 4.100809746887535e-05,
      "learning_rate": 1.9923428105432796e-05,
      "loss": 0.0932,
      "num_input_tokens_seen": 2015904,
      "step": 550,
      "train_runtime": 707.2242,
      "train_tokens_per_second": 2850.445
    },
    {
      "epoch": 1.2045209903121636,
      "grad_norm": 8.349807467311621e-05,
      "learning_rate": 1.9603115960956852e-05,
      "loss": 0.1056,
      "num_input_tokens_seen": 2052672,
      "step": 560,
      "train_runtime": 719.9875,
      "train_tokens_per_second": 2850.983
    },
    {
      "epoch": 1.2260495156081808,
      "grad_norm": 2.9793711291858926e-05,
      "learning_rate": 1.9280469364791505e-05,
      "loss": 0.1033,
      "num_input_tokens_seen": 2088736,
      "step": 570,
      "train_runtime": 732.5942,
      "train_tokens_per_second": 2851.15
    },
    {
      "epoch": 1.247578040904198,
      "grad_norm": 6.373477663146332e-05,
      "learning_rate": 1.8955651945856354e-05,
      "loss": 0.1035,
      "num_input_tokens_seen": 2125600,
      "step": 580,
      "train_runtime": 745.4583,
      "train_tokens_per_second": 2851.4
    },
    {
      "epoch": 1.2691065662002152,
      "grad_norm": 3.1842911994317546e-05,
      "learning_rate": 1.8628828433995013e-05,
      "loss": 0.0811,
      "num_input_tokens_seen": 2161792,
      "step": 590,
      "train_runtime": 758.1079,
      "train_tokens_per_second": 2851.562
    },
    {
      "epoch": 1.2906350914962326,
      "grad_norm": 4.18422459915746e-05,
      "learning_rate": 1.8300164576433035e-05,
      "loss": 0.0888,
      "num_input_tokens_seen": 2197920,
      "step": 600,
      "train_runtime": 770.7916,
      "train_tokens_per_second": 2851.51
    },
    {
      "epoch": 1.3121636167922497,
      "grad_norm": 4.866390736424364e-05,
      "learning_rate": 1.7969827053719862e-05,
      "loss": 0.0968,
      "num_input_tokens_seen": 2235168,
      "step": 610,
      "train_runtime": 783.669,
      "train_tokens_per_second": 2852.184
    },
    {
      "epoch": 1.333692142088267,
      "grad_norm": 4.293969686841592e-05,
      "learning_rate": 1.7637983395197465e-05,
      "loss": 0.1311,
      "num_input_tokens_seen": 2270816,
      "step": 620,
      "train_runtime": 796.2193,
      "train_tokens_per_second": 2851.998
    },
    {
      "epoch": 1.3552206673842842,
      "grad_norm": 6.0659429436782375e-05,
      "learning_rate": 1.7304801894038522e-05,
      "loss": 0.1001,
      "num_input_tokens_seen": 2308128,
      "step": 630,
      "train_runtime": 809.1462,
      "train_tokens_per_second": 2852.547
    },
    {
      "epoch": 1.3767491926803013,
      "grad_norm": 3.0738956411369145e-05,
      "learning_rate": 1.6970451521897198e-05,
      "loss": 0.0887,
      "num_input_tokens_seen": 2344992,
      "step": 640,
      "train_runtime": 822.0044,
      "train_tokens_per_second": 2852.773
    },
    {
      "epoch": 1.3982777179763186,
      "grad_norm": 8.285656076623127e-05,
      "learning_rate": 1.663510184321587e-05,
      "loss": 0.1041,
      "num_input_tokens_seen": 2381056,
      "step": 650,
      "train_runtime": 834.6005,
      "train_tokens_per_second": 2852.929
    },
    {
      "epoch": 1.419806243272336,
      "grad_norm": 5.0590566388564184e-05,
      "learning_rate": 1.6298922929231194e-05,
      "loss": 0.1022,
      "num_input_tokens_seen": 2417152,
      "step": 660,
      "train_runtime": 847.302,
      "train_tokens_per_second": 2852.763
    },
    {
      "epoch": 1.441334768568353,
      "grad_norm": 7.397153240162879e-05,
      "learning_rate": 1.5962085271723158e-05,
      "loss": 0.0831,
      "num_input_tokens_seen": 2454048,
      "step": 670,
      "train_runtime": 860.0949,
      "train_tokens_per_second": 2853.229
    },
    {
      "epoch": 1.4628632938643702,
      "grad_norm": 4.698023258242756e-05,
      "learning_rate": 1.5624759696550874e-05,
      "loss": 0.1055,
      "num_input_tokens_seen": 2491552,
      "step": 680,
      "train_runtime": 873.1207,
      "train_tokens_per_second": 2853.617
    },
    {
      "epoch": 1.4843918191603875,
      "grad_norm": 4.3750838813139126e-05,
      "learning_rate": 1.528711727701893e-05,
      "loss": 0.0812,
      "num_input_tokens_seen": 2528128,
      "step": 690,
      "train_runtime": 885.9021,
      "train_tokens_per_second": 2853.733
    },
    {
      "epoch": 1.5059203444564049,
      "grad_norm": 6.37572884443216e-05,
      "learning_rate": 1.4949329247118258e-05,
      "loss": 0.1016,
      "num_input_tokens_seen": 2564448,
      "step": 700,
      "train_runtime": 898.5373,
      "train_tokens_per_second": 2854.025
    },
    {
      "epoch": 1.527448869752422,
      "grad_norm": 4.437049574335106e-05,
      "learning_rate": 1.4611566914685489e-05,
      "loss": 0.0956,
      "num_input_tokens_seen": 2601056,
      "step": 710,
      "train_runtime": 911.3255,
      "train_tokens_per_second": 2854.146
    },
    {
      "epoch": 1.5489773950484391,
      "grad_norm": 5.1571758376667276e-05,
      "learning_rate": 1.4274001574524894e-05,
      "loss": 0.0936,
      "num_input_tokens_seen": 2637888,
      "step": 720,
      "train_runtime": 924.1173,
      "train_tokens_per_second": 2854.495
    },
    {
      "epoch": 1.5705059203444565,
      "grad_norm": 6.383092113537714e-05,
      "learning_rate": 1.3936804421536906e-05,
      "loss": 0.0728,
      "num_input_tokens_seen": 2673312,
      "step": 730,
      "train_runtime": 937.0636,
      "train_tokens_per_second": 2852.861
    },
    {
      "epoch": 1.5920344456404736,
      "grad_norm": 2.989832137245685e-05,
      "learning_rate": 1.3600146463897327e-05,
      "loss": 0.0839,
      "num_input_tokens_seen": 2709696,
      "step": 740,
      "train_runtime": 949.7913,
      "train_tokens_per_second": 2852.938
    },
    {
      "epoch": 1.6135629709364907,
      "grad_norm": 0.00010528949496801943,
      "learning_rate": 1.3264198436331193e-05,
      "loss": 0.1308,
      "num_input_tokens_seen": 2746528,
      "step": 750,
      "train_runtime": 962.4815,
      "train_tokens_per_second": 2853.591
    },
    {
      "epoch": 1.635091496232508,
      "grad_norm": 3.306402868474834e-05,
      "learning_rate": 1.2929130713525396e-05,
      "loss": 0.0927,
      "num_input_tokens_seen": 2782656,
      "step": 760,
      "train_runtime": 975.0888,
      "train_tokens_per_second": 2853.746
    },
    {
      "epoch": 1.6566200215285254,
      "grad_norm": 3.532276605255902e-05,
      "learning_rate": 1.2595113223723788e-05,
      "loss": 0.1135,
      "num_input_tokens_seen": 2819744,
      "step": 770,
      "train_runtime": 987.9762,
      "train_tokens_per_second": 2854.061
    },
    {
      "epoch": 1.6781485468245425,
      "grad_norm": 6.426728941733018e-05,
      "learning_rate": 1.2262315362548825e-05,
      "loss": 0.0813,
      "num_input_tokens_seen": 2857216,
      "step": 780,
      "train_runtime": 1001.0086,
      "train_tokens_per_second": 2854.337
    },
    {
      "epoch": 1.6996770721205596,
      "grad_norm": 8.145067113218829e-05,
      "learning_rate": 1.1930905907093196e-05,
      "loss": 0.0933,
      "num_input_tokens_seen": 2892000,
      "step": 790,
      "train_runtime": 1013.4209,
      "train_tokens_per_second": 2853.701
    },
    {
      "epoch": 1.721205597416577,
      "grad_norm": 0.00011963866563746706,
      "learning_rate": 1.1601052930325236e-05,
      "loss": 0.0881,
      "num_input_tokens_seen": 2928832,
      "step": 800,
      "train_runtime": 1026.3231,
      "train_tokens_per_second": 2853.714
    },
    {
      "epoch": 1.7427341227125943,
      "grad_norm": 5.128612610860728e-05,
      "learning_rate": 1.127292371585141e-05,
      "loss": 0.079,
      "num_input_tokens_seen": 2965632,
      "step": 810,
      "train_runtime": 1039.1265,
      "train_tokens_per_second": 2853.966
    },
    {
      "epoch": 1.7642626480086114,
      "grad_norm": 7.751452358206734e-05,
      "learning_rate": 1.0946684673079074e-05,
      "loss": 0.1132,
      "num_input_tokens_seen": 3002496,
      "step": 820,
      "train_runtime": 1051.9839,
      "train_tokens_per_second": 2854.127
    },
    {
      "epoch": 1.7857911733046286,
      "grad_norm": 4.0213129977928475e-05,
      "learning_rate": 1.0622501252822628e-05,
      "loss": 0.0793,
      "num_input_tokens_seen": 3038944,
      "step": 830,
      "train_runtime": 1064.6802,
      "train_tokens_per_second": 2854.326
    },
    {
      "epoch": 1.807319698600646,
      "grad_norm": 4.586371505865827e-05,
      "learning_rate": 1.0300537863395823e-05,
      "loss": 0.0918,
      "num_input_tokens_seen": 3076032,
      "step": 840,
      "train_runtime": 1077.5859,
      "train_tokens_per_second": 2854.559
    },
    {
      "epoch": 1.8288482238966632,
      "grad_norm": 3.2480071240570396e-05,
      "learning_rate": 9.98095778723271e-06,
      "loss": 0.093,
      "num_input_tokens_seen": 3112448,
      "step": 850,
      "train_runtime": 1090.3056,
      "train_tokens_per_second": 2854.656
    },
    {
      "epoch": 1.8503767491926801,
      "grad_norm": 5.541834980249405e-05,
      "learning_rate": 9.663923098079656e-06,
      "loss": 0.0973,
      "num_input_tokens_seen": 3149376,
      "step": 860,
      "train_runtime": 1103.1339,
      "train_tokens_per_second": 2854.935
    },
    {
      "epoch": 1.8719052744886975,
      "grad_norm": 3.055919296457432e-05,
      "learning_rate": 9.349594578800259e-06,
      "loss": 0.0995,
      "num_input_tokens_seen": 3186304,
      "step": 870,
      "train_runtime": 1116.0618,
      "train_tokens_per_second": 2854.953
    },
    {
      "epoch": 1.8934337997847148,
      "grad_norm": 4.6285113057820126e-05,
      "learning_rate": 9.038131639834977e-06,
      "loss": 0.1238,
      "num_input_tokens_seen": 3222752,
      "step": 880,
      "train_runtime": 1128.7812,
      "train_tokens_per_second": 2855.072
    },
    {
      "epoch": 1.914962325080732,
      "grad_norm": 4.483849261305295e-05,
      "learning_rate": 8.7296922383568e-06,
      "loss": 0.0912,
      "num_input_tokens_seen": 3258976,
      "step": 890,
      "train_runtime": 1141.5159,
      "train_tokens_per_second": 2854.954
    },
    {
      "epoch": 1.936490850376749,
      "grad_norm": 5.5885881010908633e-05,
      "learning_rate": 8.424432798163838e-06,
      "loss": 0.0935,
      "num_input_tokens_seen": 3295040,
      "step": 900,
      "train_runtime": 1154.1132,
      "train_tokens_per_second": 2855.041
    },
    {
      "epoch": 1.9580193756727664,
      "grad_norm": 3.8836213207105175e-05,
      "learning_rate": 8.122508130349711e-06,
      "loss": 0.0722,
      "num_input_tokens_seen": 3331392,
      "step": 910,
      "train_runtime": 1166.7365,
      "train_tokens_per_second": 2855.308
    },
    {
      "epoch": 1.9795479009687837,
      "grad_norm": 6.445301551138982e-05,
      "learning_rate": 7.82407135479164e-06,
      "loss": 0.1001,
      "num_input_tokens_seen": 3368032,
      "step": 920,
      "train_runtime": 1179.5494,
      "train_tokens_per_second": 2855.355
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.651374805078376e-05,
      "learning_rate": 7.529273822496379e-06,
      "loss": 0.0647,
      "num_input_tokens_seen": 3403584,
      "step": 930,
      "train_runtime": 1191.8451,
      "train_tokens_per_second": 2855.727
    },
    {
      "epoch": 2.0215285252960173,
      "grad_norm": 9.509055962553248e-05,
      "learning_rate": 7.238265038843143e-06,
      "loss": 0.0881,
      "num_input_tokens_seen": 3440000,
      "step": 940,
      "train_runtime": 1204.5181,
      "train_tokens_per_second": 2855.914
    },
    {
      "epoch": 2.0430570505920342,
      "grad_norm": 5.2039082220289856e-05,
      "learning_rate": 6.951192587762555e-06,
      "loss": 0.0573,
      "num_input_tokens_seen": 3476640,
      "step": 950,
      "train_runtime": 1217.2855,
      "train_tokens_per_second": 2856.06
    },
    {
      "epoch": 2.0645855758880516,
      "grad_norm": 3.832239599432796e-05,
      "learning_rate": 6.668202056890052e-06,
      "loss": 0.0646,
      "num_input_tokens_seen": 3512992,
      "step": 960,
      "train_runtime": 1229.9665,
      "train_tokens_per_second": 2856.169
    },
    {
      "epoch": 2.086114101184069,
      "grad_norm": 5.945537122897804e-05,
      "learning_rate": 6.389436963731707e-06,
      "loss": 0.0682,
      "num_input_tokens_seen": 3549504,
      "step": 970,
      "train_runtime": 1242.6763,
      "train_tokens_per_second": 2856.338
    },
    {
      "epoch": 2.1076426264800863,
      "grad_norm": 6.592880527023226e-05,
      "learning_rate": 6.115038682879904e-06,
      "loss": 0.0931,
      "num_input_tokens_seen": 3586592,
      "step": 980,
      "train_runtime": 1255.5495,
      "train_tokens_per_second": 2856.591
    },
    {
      "epoch": 2.129171151776103,
      "grad_norm": 2.9424012609524652e-05,
      "learning_rate": 5.845146374315781e-06,
      "loss": 0.0686,
      "num_input_tokens_seen": 3622944,
      "step": 990,
      "train_runtime": 1268.2309,
      "train_tokens_per_second": 2856.691
    },
    {
      "epoch": 2.1506996770721205,
      "grad_norm": 5.797382618766278e-05,
      "learning_rate": 5.579896912834807e-06,
      "loss": 0.0917,
      "num_input_tokens_seen": 3660096,
      "step": 1000,
      "train_runtime": 1281.1218,
      "train_tokens_per_second": 2856.946
    },
    {
      "epoch": 2.172228202368138,
      "grad_norm": 5.529261034098454e-05,
      "learning_rate": 5.319424818631293e-06,
      "loss": 0.0931,
      "num_input_tokens_seen": 3697024,
      "step": 1010,
      "train_runtime": 1294.9799,
      "train_tokens_per_second": 2854.889
    },
    {
      "epoch": 2.193756727664155,
      "grad_norm": 5.912356573389843e-05,
      "learning_rate": 5.063862189076988e-06,
      "loss": 0.0855,
      "num_input_tokens_seen": 3734240,
      "step": 1020,
      "train_runtime": 1307.7689,
      "train_tokens_per_second": 2855.428
    },
    {
      "epoch": 2.215285252960172,
      "grad_norm": 5.3946208936395124e-05,
      "learning_rate": 4.813338631728471e-06,
      "loss": 0.0909,
      "num_input_tokens_seen": 3771232,
      "step": 1030,
      "train_runtime": 1320.5384,
      "train_tokens_per_second": 2855.829
    },
    {
      "epoch": 2.2368137782561894,
      "grad_norm": 6.292902253335342e-05,
      "learning_rate": 4.567981198597168e-06,
      "loss": 0.0817,
      "num_input_tokens_seen": 3808416,
      "step": 1040,
      "train_runtime": 1333.3555,
      "train_tokens_per_second": 2856.264
    },
    {
      "epoch": 2.2583423035522068,
      "grad_norm": 4.050394636578858e-05,
      "learning_rate": 4.327914321715459e-06,
      "loss": 0.0599,
      "num_input_tokens_seen": 3844768,
      "step": 1050,
      "train_runtime": 1345.9495,
      "train_tokens_per_second": 2856.547
    },
    {
      "epoch": 2.2798708288482237,
      "grad_norm": 5.740273991250433e-05,
      "learning_rate": 4.09325975003146e-06,
      "loss": 0.0711,
      "num_input_tokens_seen": 3881088,
      "step": 1060,
      "train_runtime": 1358.6041,
      "train_tokens_per_second": 2856.673
    },
    {
      "epoch": 2.301399354144241,
      "grad_norm": 7.280943100340664e-05,
      "learning_rate": 3.864136487664538e-06,
      "loss": 0.0723,
      "num_input_tokens_seen": 3916928,
      "step": 1070,
      "train_runtime": 1371.1849,
      "train_tokens_per_second": 2856.601
    },
    {
      "epoch": 2.3229278794402584,
      "grad_norm": 4.93960760650225e-05,
      "learning_rate": 3.640660733552863e-06,
      "loss": 0.0524,
      "num_input_tokens_seen": 3952928,
      "step": 1080,
      "train_runtime": 1383.7869,
      "train_tokens_per_second": 2856.602
    },
    {
      "epoch": 2.3444564047362757,
      "grad_norm": 7.801376341376454e-05,
      "learning_rate": 3.4229458225235696e-06,
      "loss": 0.0513,
      "num_input_tokens_seen": 3990720,
      "step": 1090,
      "train_runtime": 1396.9725,
      "train_tokens_per_second": 2856.692
    },
    {
      "epoch": 2.3659849300322926,
      "grad_norm": 4.576309584081173e-05,
      "learning_rate": 3.211102167815469e-06,
      "loss": 0.0773,
      "num_input_tokens_seen": 4026656,
      "step": 1100,
      "train_runtime": 1409.5674,
      "train_tokens_per_second": 2856.661
    },
    {
      "epoch": 2.38751345532831,
      "grad_norm": 0.0001336225977865979,
      "learning_rate": 3.0052372050834025e-06,
      "loss": 0.0896,
      "num_input_tokens_seen": 4064192,
      "step": 1110,
      "train_runtime": 1422.5111,
      "train_tokens_per_second": 2857.055
    },
    {
      "epoch": 2.4090419806243273,
      "grad_norm": 8.807591075310484e-05,
      "learning_rate": 2.805455337912722e-06,
      "loss": 0.0719,
      "num_input_tokens_seen": 4101632,
      "step": 1120,
      "train_runtime": 1435.3867,
      "train_tokens_per_second": 2857.51
    },
    {
      "epoch": 2.4305705059203446,
      "grad_norm": 3.673565879580565e-05,
      "learning_rate": 2.611857884871379e-06,
      "loss": 0.0801,
      "num_input_tokens_seen": 4138208,
      "step": 1130,
      "train_runtime": 1448.0115,
      "train_tokens_per_second": 2857.856
    },
    {
      "epoch": 2.4520990312163615,
      "grad_norm": 2.5486542654107325e-05,
      "learning_rate": 2.424543028126661e-06,
      "loss": 0.0821,
      "num_input_tokens_seen": 4174592,
      "step": 1140,
      "train_runtime": 1460.6997,
      "train_tokens_per_second": 2857.94
    },
    {
      "epoch": 2.473627556512379,
      "grad_norm": 6.899920845171437e-05,
      "learning_rate": 2.243605763652489e-06,
      "loss": 0.0654,
      "num_input_tokens_seen": 4211616,
      "step": 1150,
      "train_runtime": 1473.5308,
      "train_tokens_per_second": 2858.18
    },
    {
      "epoch": 2.495156081808396,
      "grad_norm": 4.8038589739007875e-05,
      "learning_rate": 2.0691378530525933e-06,
      "loss": 0.0767,
      "num_input_tokens_seen": 4247744,
      "step": 1160,
      "train_runtime": 1486.1222,
      "train_tokens_per_second": 2858.274
    },
    {
      "epoch": 2.5166846071044136,
      "grad_norm": 7.49904866097495e-05,
      "learning_rate": 1.9012277770239829e-06,
      "loss": 0.0655,
      "num_input_tokens_seen": 4284224,
      "step": 1170,
      "train_runtime": 1498.696,
      "train_tokens_per_second": 2858.634
    },
    {
      "epoch": 2.5382131324004304,
      "grad_norm": 6.3829000282567e-05,
      "learning_rate": 1.7399606904843224e-06,
      "loss": 0.0584,
      "num_input_tokens_seen": 4320672,
      "step": 1180,
      "train_runtime": 1511.4477,
      "train_tokens_per_second": 2858.631
    },
    {
      "epoch": 2.559741657696448,
      "grad_norm": 5.6509288697270676e-05,
      "learning_rate": 1.5854183793859577e-06,
      "loss": 0.0601,
      "num_input_tokens_seen": 4358080,
      "step": 1190,
      "train_runtime": 1524.3591,
      "train_tokens_per_second": 2858.959
    },
    {
      "epoch": 2.581270182992465,
      "grad_norm": 4.297928171581589e-05,
      "learning_rate": 1.437679219238498e-06,
      "loss": 0.0665,
      "num_input_tokens_seen": 4394752,
      "step": 1200,
      "train_runtime": 1537.0429,
      "train_tokens_per_second": 2859.225
    },
    {
      "epoch": 2.602798708288482,
      "grad_norm": 8.681658800924197e-05,
      "learning_rate": 1.2968181353609854e-06,
      "loss": 0.0658,
      "num_input_tokens_seen": 4431808,
      "step": 1210,
      "train_runtime": 1549.8548,
      "train_tokens_per_second": 2859.499
    },
    {
      "epoch": 2.6243272335844994,
      "grad_norm": 3.503639891277999e-05,
      "learning_rate": 1.1629065648838495e-06,
      "loss": 0.0898,
      "num_input_tokens_seen": 4467328,
      "step": 1220,
      "train_runtime": 1562.3415,
      "train_tokens_per_second": 2859.38
    },
    {
      "epoch": 2.6458557588805167,
      "grad_norm": 3.0625513318227604e-05,
      "learning_rate": 1.0360124205198102e-06,
      "loss": 0.0619,
      "num_input_tokens_seen": 4503808,
      "step": 1230,
      "train_runtime": 1575.0206,
      "train_tokens_per_second": 2859.523
    },
    {
      "epoch": 2.667384284176534,
      "grad_norm": 6.452269735746086e-05,
      "learning_rate": 9.162000561222672e-07,
      "loss": 0.0763,
      "num_input_tokens_seen": 4541536,
      "step": 1240,
      "train_runtime": 1588.024,
      "train_tokens_per_second": 2859.866
    },
    {
      "epoch": 2.6889128094725514,
      "grad_norm": 2.9151902708690614e-05,
      "learning_rate": 8.035302340484646e-07,
      "loss": 0.0909,
      "num_input_tokens_seen": 4577664,
      "step": 1250,
      "train_runtime": 1600.6009,
      "train_tokens_per_second": 2859.966
    },
    {
      "epoch": 2.7104413347685683,
      "grad_norm": 6.271711754379794e-05,
      "learning_rate": 6.980600943441273e-07,
      "loss": 0.0683,
      "num_input_tokens_seen": 4614176,
      "step": 1260,
      "train_runtime": 1613.2327,
      "train_tokens_per_second": 2860.205
    },
    {
      "epoch": 2.7319698600645856,
      "grad_norm": 5.843062899657525e-05,
      "learning_rate": 5.998431257651132e-07,
      "loss": 0.0867,
      "num_input_tokens_seen": 4650848,
      "step": 1270,
      "train_runtime": 1625.9328,
      "train_tokens_per_second": 2860.418
    },
    {
      "epoch": 2.7534983853606025,
      "grad_norm": 5.628829603665508e-05,
      "learning_rate": 5.089291386508055e-07,
      "loss": 0.0648,
      "num_input_tokens_seen": 4687456,
      "step": 1280,
      "train_runtime": 1638.6467,
      "train_tokens_per_second": 2860.565
    },
    {
      "epoch": 2.77502691065662,
      "grad_norm": 4.340551095083356e-05,
      "learning_rate": 4.2536423966299256e-07,
      "loss": 0.0672,
      "num_input_tokens_seen": 4723840,
      "step": 1290,
      "train_runtime": 1651.315,
      "train_tokens_per_second": 2860.653
    },
    {
      "epoch": 2.7965554359526372,
      "grad_norm": 7.965038093971089e-05,
      "learning_rate": 3.491908084030776e-07,
      "loss": 0.089,
      "num_input_tokens_seen": 4761056,
      "step": 1300,
      "train_runtime": 1664.1238,
      "train_tokens_per_second": 2860.999
    },
    {
      "epoch": 2.8180839612486546,
      "grad_norm": 5.358497583074495e-05,
      "learning_rate": 2.804474759194176e-07,
      "loss": 0.07,
      "num_input_tokens_seen": 4798240,
      "step": 1310,
      "train_runtime": 1676.8916,
      "train_tokens_per_second": 2861.39
    },
    {
      "epoch": 2.839612486544672,
      "grad_norm": 3.2774096325738356e-05,
      "learning_rate": 2.1916910511575627e-07,
      "loss": 0.0807,
      "num_input_tokens_seen": 4834560,
      "step": 1320,
      "train_runtime": 1689.486,
      "train_tokens_per_second": 2861.557
    },
    {
      "epoch": 2.861141011840689,
      "grad_norm": 6.476557609857991e-05,
      "learning_rate": 1.6538677307062933e-07,
      "loss": 0.0669,
      "num_input_tokens_seen": 4871328,
      "step": 1330,
      "train_runtime": 1702.2714,
      "train_tokens_per_second": 2861.664
    },
    {
      "epoch": 2.882669537136706,
      "grad_norm": 8.176828123396263e-05,
      "learning_rate": 1.1912775527676278e-07,
      "loss": 0.0552,
      "num_input_tokens_seen": 4907744,
      "step": 1340,
      "train_runtime": 1714.8671,
      "train_tokens_per_second": 2861.88
    },
    {
      "epoch": 2.9041980624327235,
      "grad_norm": 7.269597699632868e-05,
      "learning_rate": 8.04155118084099e-08,
      "loss": 0.0917,
      "num_input_tokens_seen": 4943712,
      "step": 1350,
      "train_runtime": 1727.4271,
      "train_tokens_per_second": 2861.893
    },
    {
      "epoch": 2.9257265877287404,
      "grad_norm": 3.572047353372909e-05,
      "learning_rate": 4.9269675423671556e-08,
      "loss": 0.074,
      "num_input_tokens_seen": 4980480,
      "step": 1360,
      "train_runtime": 1740.2346,
      "train_tokens_per_second": 2861.959
    },
    {
      "epoch": 2.9472551130247577,
      "grad_norm": 8.610687655163929e-05,
      "learning_rate": 2.570604160781753e-08,
      "loss": 0.0599,
      "num_input_tokens_seen": 5016800,
      "step": 1370,
      "train_runtime": 1752.9074,
      "train_tokens_per_second": 2861.988
    },
    {
      "epoch": 2.968783638320775,
      "grad_norm": 5.934851287747733e-05,
      "learning_rate": 9.736560562675912e-09,
      "loss": 0.0593,
      "num_input_tokens_seen": 5053440,
      "step": 1380,
      "train_runtime": 1765.6959,
      "train_tokens_per_second": 2862.01
    },
    {
      "epoch": 2.9903121636167924,
      "grad_norm": 3.7563720979960635e-05,
      "learning_rate": 1.3693311461326908e-09,
      "loss": 0.0763,
      "num_input_tokens_seen": 5090592,
      "step": 1390,
      "train_runtime": 1778.5239,
      "train_tokens_per_second": 2862.257
    },
    {
      "epoch": 3.0,
      "num_input_tokens_seen": 5107072,
      "step": 1395,
      "total_flos": 2.186269295986606e+17,
      "train_loss": 0.10808860245144068,
      "train_runtime": 1785.3385,
      "train_samples_per_second": 6.244,
      "train_steps_per_second": 0.781
    }
  ],
  "logging_steps": 10,
  "max_steps": 1395,
  "num_input_tokens_seen": 5107072,
  "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": 2.186269295986606e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}