{
  "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": 7.928458944661543e-05,
      "learning_rate": 2.9996919063517692e-05,
      "loss": 1.6735,
      "num_input_tokens_seen": 29856,
      "step": 10,
      "train_runtime": 19.5226,
      "train_tokens_per_second": 1529.306
    },
    {
      "epoch": 0.04305705059203445,
      "grad_norm": 7.366821955656633e-05,
      "learning_rate": 2.9986270537572827e-05,
      "loss": 0.4336,
      "num_input_tokens_seen": 59488,
      "step": 20,
      "train_runtime": 30.7633,
      "train_tokens_per_second": 1933.73
    },
    {
      "epoch": 0.06458557588805167,
      "grad_norm": 4.6334236685652286e-05,
      "learning_rate": 2.9968021784983436e-05,
      "loss": 0.2851,
      "num_input_tokens_seen": 89536,
      "step": 30,
      "train_runtime": 42.1628,
      "train_tokens_per_second": 2123.579
    },
    {
      "epoch": 0.0861141011840689,
      "grad_norm": 3.5027369449380785e-05,
      "learning_rate": 2.994218206053077e-05,
      "loss": 0.2813,
      "num_input_tokens_seen": 119904,
      "step": 40,
      "train_runtime": 53.5972,
      "train_tokens_per_second": 2237.133
    },
    {
      "epoch": 0.10764262648008611,
      "grad_norm": 2.6238762075081468e-05,
      "learning_rate": 2.9908764468726846e-05,
      "loss": 0.2555,
      "num_input_tokens_seen": 149888,
      "step": 50,
      "train_runtime": 64.9959,
      "train_tokens_per_second": 2306.116
    },
    {
      "epoch": 0.12917115177610333,
      "grad_norm": 5.1311333663761616e-05,
      "learning_rate": 2.986778595716853e-05,
      "loss": 0.2413,
      "num_input_tokens_seen": 180288,
      "step": 60,
      "train_runtime": 76.4086,
      "train_tokens_per_second": 2359.525
    },
    {
      "epoch": 0.15069967707212056,
      "grad_norm": 5.559088822337799e-05,
      "learning_rate": 2.981926730794264e-05,
      "loss": 0.277,
      "num_input_tokens_seen": 210272,
      "step": 70,
      "train_runtime": 87.7283,
      "train_tokens_per_second": 2396.856
    },
    {
      "epoch": 0.1722282023681378,
      "grad_norm": 2.377848250034731e-05,
      "learning_rate": 2.9763233127086394e-05,
      "loss": 0.2471,
      "num_input_tokens_seen": 240320,
      "step": 80,
      "train_runtime": 99.133,
      "train_tokens_per_second": 2424.217
    },
    {
      "epoch": 0.193756727664155,
      "grad_norm": 1.6515690731466748e-05,
      "learning_rate": 2.9699711832108555e-05,
      "loss": 0.2636,
      "num_input_tokens_seen": 269632,
      "step": 90,
      "train_runtime": 110.38,
      "train_tokens_per_second": 2442.761
    },
    {
      "epoch": 0.21528525296017223,
      "grad_norm": 2.6908706786343828e-05,
      "learning_rate": 2.9628735637577612e-05,
      "loss": 0.2545,
      "num_input_tokens_seen": 299616,
      "step": 100,
      "train_runtime": 121.7055,
      "train_tokens_per_second": 2461.812
    },
    {
      "epoch": 0.23681377825618946,
      "grad_norm": 3.175454548909329e-05,
      "learning_rate": 2.955034053878428e-05,
      "loss": 0.2389,
      "num_input_tokens_seen": 329248,
      "step": 110,
      "train_runtime": 133.0055,
      "train_tokens_per_second": 2475.446
    },
    {
      "epoch": 0.25834230355220666,
      "grad_norm": 4.9838836275739595e-05,
      "learning_rate": 2.9464566293486653e-05,
      "loss": 0.2285,
      "num_input_tokens_seen": 359456,
      "step": 120,
      "train_runtime": 144.4921,
      "train_tokens_per_second": 2487.72
    },
    {
      "epoch": 0.2798708288482239,
      "grad_norm": 7.634468784090132e-05,
      "learning_rate": 2.9371456401747193e-05,
      "loss": 0.2472,
      "num_input_tokens_seen": 389792,
      "step": 130,
      "train_runtime": 155.9528,
      "train_tokens_per_second": 2499.423
    },
    {
      "epoch": 0.3013993541442411,
      "grad_norm": 4.957841883879155e-05,
      "learning_rate": 2.9271058083871864e-05,
      "loss": 0.2582,
      "num_input_tokens_seen": 419712,
      "step": 140,
      "train_runtime": 167.4393,
      "train_tokens_per_second": 2506.651
    },
    {
      "epoch": 0.32292787944025836,
      "grad_norm": 3.682490569190122e-05,
      "learning_rate": 2.916342225646254e-05,
      "loss": 0.2157,
      "num_input_tokens_seen": 449952,
      "step": 150,
      "train_runtime": 179.0086,
      "train_tokens_per_second": 2513.577
    },
    {
      "epoch": 0.3444564047362756,
      "grad_norm": 2.852646866813302e-05,
      "learning_rate": 2.904860350659487e-05,
      "loss": 0.1913,
      "num_input_tokens_seen": 480224,
      "step": 160,
      "train_runtime": 190.3763,
      "train_tokens_per_second": 2522.499
    },
    {
      "epoch": 0.36598493003229277,
      "grad_norm": 6.110664980951697e-05,
      "learning_rate": 2.8926660064134637e-05,
      "loss": 0.2648,
      "num_input_tokens_seen": 509920,
      "step": 170,
      "train_runtime": 201.7302,
      "train_tokens_per_second": 2527.732
    },
    {
      "epoch": 0.38751345532831,
      "grad_norm": 2.867744115064852e-05,
      "learning_rate": 2.8797653772206726e-05,
      "loss": 0.235,
      "num_input_tokens_seen": 539776,
      "step": 180,
      "train_runtime": 213.0917,
      "train_tokens_per_second": 2533.07
    },
    {
      "epoch": 0.40904198062432723,
      "grad_norm": 4.131464083911851e-05,
      "learning_rate": 2.8661650055831645e-05,
      "loss": 0.2457,
      "num_input_tokens_seen": 569760,
      "step": 190,
      "train_runtime": 224.4635,
      "train_tokens_per_second": 2538.32
    },
    {
      "epoch": 0.43057050592034446,
      "grad_norm": 1.611412335478235e-05,
      "learning_rate": 2.8518717888745486e-05,
      "loss": 0.2162,
      "num_input_tokens_seen": 600256,
      "step": 200,
      "train_runtime": 235.9692,
      "train_tokens_per_second": 2543.789
    },
    {
      "epoch": 0.4520990312163617,
      "grad_norm": 2.7770569431595504e-05,
      "learning_rate": 2.836892975842015e-05,
      "loss": 0.2229,
      "num_input_tokens_seen": 630048,
      "step": 210,
      "train_runtime": 247.2964,
      "train_tokens_per_second": 2547.744
    },
    {
      "epoch": 0.4736275565123789,
      "grad_norm": 3.672212551464327e-05,
      "learning_rate": 2.8212361629301655e-05,
      "loss": 0.2091,
      "num_input_tokens_seen": 660320,
      "step": 220,
      "train_runtime": 258.7843,
      "train_tokens_per_second": 2551.624
    },
    {
      "epoch": 0.4951560818083961,
      "grad_norm": 4.9590093112783507e-05,
      "learning_rate": 2.8049092904285033e-05,
      "loss": 0.2062,
      "num_input_tokens_seen": 690016,
      "step": 230,
      "train_runtime": 270.1544,
      "train_tokens_per_second": 2554.154
    },
    {
      "epoch": 0.5166846071044133,
      "grad_norm": 3.8173635402927175e-05,
      "learning_rate": 2.7879206384445526e-05,
      "loss": 0.1952,
      "num_input_tokens_seen": 720512,
      "step": 240,
      "train_runtime": 281.5373,
      "train_tokens_per_second": 2559.206
    },
    {
      "epoch": 0.5382131324004306,
      "grad_norm": 2.2552463633473963e-05,
      "learning_rate": 2.7702788227046286e-05,
      "loss": 0.2243,
      "num_input_tokens_seen": 750112,
      "step": 250,
      "train_runtime": 292.8276,
      "train_tokens_per_second": 2561.617
    },
    {
      "epoch": 0.5597416576964478,
      "grad_norm": 2.7334826881997287e-05,
      "learning_rate": 2.751992790184413e-05,
      "loss": 0.2057,
      "num_input_tokens_seen": 781280,
      "step": 260,
      "train_runtime": 304.3534,
      "train_tokens_per_second": 2567.016
    },
    {
      "epoch": 0.581270182992465,
      "grad_norm": 2.950379166577477e-05,
      "learning_rate": 2.733071814571527e-05,
      "loss": 0.2191,
      "num_input_tokens_seen": 811232,
      "step": 270,
      "train_runtime": 315.6789,
      "train_tokens_per_second": 2569.801
    },
    {
      "epoch": 0.6027987082884823,
      "grad_norm": 3.076158100157045e-05,
      "learning_rate": 2.7135254915624213e-05,
      "loss": 0.2026,
      "num_input_tokens_seen": 841856,
      "step": 280,
      "train_runtime": 327.1838,
      "train_tokens_per_second": 2573.037
    },
    {
      "epoch": 0.6243272335844995,
      "grad_norm": 5.050536128692329e-05,
      "learning_rate": 2.693363733995954e-05,
      "loss": 0.2283,
      "num_input_tokens_seen": 872608,
      "step": 290,
      "train_runtime": 338.65,
      "train_tokens_per_second": 2576.725
    },
    {
      "epoch": 0.6458557588805167,
      "grad_norm": 6.511939136544243e-05,
      "learning_rate": 2.672596766826138e-05,
      "loss": 0.1934,
      "num_input_tokens_seen": 902336,
      "step": 300,
      "train_runtime": 349.9311,
      "train_tokens_per_second": 2578.61
    },
    {
      "epoch": 0.667384284176534,
      "grad_norm": 4.5949171180836856e-05,
      "learning_rate": 2.6512351219365927e-05,
      "loss": 0.2067,
      "num_input_tokens_seen": 931808,
      "step": 310,
      "train_runtime": 361.2079,
      "train_tokens_per_second": 2579.7
    },
    {
      "epoch": 0.6889128094725512,
      "grad_norm": 2.2227935914997943e-05,
      "learning_rate": 2.6292896327993467e-05,
      "loss": 0.2105,
      "num_input_tokens_seen": 961952,
      "step": 320,
      "train_runtime": 372.6107,
      "train_tokens_per_second": 2581.654
    },
    {
      "epoch": 0.7104413347685683,
      "grad_norm": 2.7524683900992386e-05,
      "learning_rate": 2.606771428980681e-05,
      "loss": 0.1937,
      "num_input_tokens_seen": 992416,
      "step": 330,
      "train_runtime": 384.1399,
      "train_tokens_per_second": 2583.475
    },
    {
      "epoch": 0.7319698600645855,
      "grad_norm": 3.250837107771076e-05,
      "learning_rate": 2.5836919304968184e-05,
      "loss": 0.1783,
      "num_input_tokens_seen": 1022176,
      "step": 340,
      "train_runtime": 395.4628,
      "train_tokens_per_second": 2584.759
    },
    {
      "epoch": 0.7534983853606028,
      "grad_norm": 4.1476250771665946e-05,
      "learning_rate": 2.5600628420223057e-05,
      "loss": 0.226,
      "num_input_tokens_seen": 1054016,
      "step": 350,
      "train_runtime": 407.6582,
      "train_tokens_per_second": 2585.538
    },
    {
      "epoch": 0.77502691065662,
      "grad_norm": 2.3837223125156015e-05,
      "learning_rate": 2.535896146954037e-05,
      "loss": 0.1699,
      "num_input_tokens_seen": 1084096,
      "step": 360,
      "train_runtime": 419.0167,
      "train_tokens_per_second": 2587.238
    },
    {
      "epoch": 0.7965554359526372,
      "grad_norm": 3.133831705781631e-05,
      "learning_rate": 2.5112041013339212e-05,
      "loss": 0.189,
      "num_input_tokens_seen": 1115776,
      "step": 370,
      "train_runtime": 430.6973,
      "train_tokens_per_second": 2590.627
    },
    {
      "epoch": 0.8180839612486545,
      "grad_norm": 3.1692266929894686e-05,
      "learning_rate": 2.48599922763328e-05,
      "loss": 0.1536,
      "num_input_tokens_seen": 1145856,
      "step": 380,
      "train_runtime": 442.0573,
      "train_tokens_per_second": 2592.098
    },
    {
      "epoch": 0.8396124865446717,
      "grad_norm": 2.1839789042132907e-05,
      "learning_rate": 2.4602943084021286e-05,
      "loss": 0.2303,
      "num_input_tokens_seen": 1175488,
      "step": 390,
      "train_runtime": 453.3867,
      "train_tokens_per_second": 2592.683
    },
    {
      "epoch": 0.8611410118406889,
      "grad_norm": 3.287403524154797e-05,
      "learning_rate": 2.434102379786553e-05,
      "loss": 0.1807,
      "num_input_tokens_seen": 1205088,
      "step": 400,
      "train_runtime": 464.6651,
      "train_tokens_per_second": 2593.455
    },
    {
      "epoch": 0.8826695371367062,
      "grad_norm": 2.3363501895801164e-05,
      "learning_rate": 2.4074367249174864e-05,
      "loss": 0.1666,
      "num_input_tokens_seen": 1235296,
      "step": 410,
      "train_runtime": 476.0075,
      "train_tokens_per_second": 2595.119
    },
    {
      "epoch": 0.9041980624327234,
      "grad_norm": 3.2339172321371734e-05,
      "learning_rate": 2.380310867174219e-05,
      "loss": 0.1611,
      "num_input_tokens_seen": 1265440,
      "step": 420,
      "train_runtime": 487.3811,
      "train_tokens_per_second": 2596.408
    },
    {
      "epoch": 0.9257265877287406,
      "grad_norm": 2.5274217477999628e-05,
      "learning_rate": 2.3527385633260724e-05,
      "loss": 0.1765,
      "num_input_tokens_seen": 1295488,
      "step": 430,
      "train_runtime": 498.6324,
      "train_tokens_per_second": 2598.082
    },
    {
      "epoch": 0.9472551130247578,
      "grad_norm": 2.8673657652689144e-05,
      "learning_rate": 2.3247337965557097e-05,
      "loss": 0.1454,
      "num_input_tokens_seen": 1324096,
      "step": 440,
      "train_runtime": 509.775,
      "train_tokens_per_second": 2597.413
    },
    {
      "epoch": 0.9687836383207751,
      "grad_norm": 2.2292393623501994e-05,
      "learning_rate": 2.2963107693676192e-05,
      "loss": 0.1707,
      "num_input_tokens_seen": 1353920,
      "step": 450,
      "train_runtime": 521.0411,
      "train_tokens_per_second": 2598.49
    },
    {
      "epoch": 0.9903121636167922,
      "grad_norm": 3.275413473602384e-05,
      "learning_rate": 2.2674838963853738e-05,
      "loss": 0.1911,
      "num_input_tokens_seen": 1383680,
      "step": 460,
      "train_runtime": 532.324,
      "train_tokens_per_second": 2599.319
    },
    {
      "epoch": 1.0107642626480087,
      "grad_norm": 6.84974729665555e-05,
      "learning_rate": 2.2382677970413097e-05,
      "loss": 0.1576,
      "num_input_tokens_seen": 1412480,
      "step": 470,
      "train_runtime": 543.1133,
      "train_tokens_per_second": 2600.709
    },
    {
      "epoch": 1.0322927879440258,
      "grad_norm": 2.6211033400613815e-05,
      "learning_rate": 2.2086772881623455e-05,
      "loss": 0.1992,
      "num_input_tokens_seen": 1442272,
      "step": 480,
      "train_runtime": 554.3504,
      "train_tokens_per_second": 2601.734
    },
    {
      "epoch": 1.0538213132400431,
      "grad_norm": 4.215672015561722e-05,
      "learning_rate": 2.1787273764556827e-05,
      "loss": 0.1314,
      "num_input_tokens_seen": 1472992,
      "step": 490,
      "train_runtime": 565.8235,
      "train_tokens_per_second": 2603.271
    },
    {
      "epoch": 1.0753498385360603,
      "grad_norm": 4.263213122612797e-05,
      "learning_rate": 2.1484332508982167e-05,
      "loss": 0.1508,
      "num_input_tokens_seen": 1503680,
      "step": 500,
      "train_runtime": 577.2742,
      "train_tokens_per_second": 2604.793
    },
    {
      "epoch": 1.0968783638320776,
      "grad_norm": 3.428611307754181e-05,
      "learning_rate": 2.11781027503351e-05,
      "loss": 0.1804,
      "num_input_tokens_seen": 1533664,
      "step": 510,
      "train_runtime": 588.6267,
      "train_tokens_per_second": 2605.495
    },
    {
      "epoch": 1.1184068891280947,
      "grad_norm": 1.910557875817176e-05,
      "learning_rate": 2.0868739791802312e-05,
      "loss": 0.1632,
      "num_input_tokens_seen": 1563936,
      "step": 520,
      "train_runtime": 599.9649,
      "train_tokens_per_second": 2606.713
    },
    {
      "epoch": 1.1399354144241118,
      "grad_norm": 6.435342947952449e-05,
      "learning_rate": 2.0556400525560168e-05,
      "loss": 0.1382,
      "num_input_tokens_seen": 1593376,
      "step": 530,
      "train_runtime": 611.1362,
      "train_tokens_per_second": 2607.235
    },
    {
      "epoch": 1.1614639397201292,
      "grad_norm": 1.7743157513905317e-05,
      "learning_rate": 2.0241243353207484e-05,
      "loss": 0.1389,
      "num_input_tokens_seen": 1623072,
      "step": 540,
      "train_runtime": 622.3963,
      "train_tokens_per_second": 2607.779
    },
    {
      "epoch": 1.1829924650161463,
      "grad_norm": 2.971130197693128e-05,
      "learning_rate": 1.9923428105432796e-05,
      "loss": 0.1196,
      "num_input_tokens_seen": 1653984,
      "step": 550,
      "train_runtime": 633.8771,
      "train_tokens_per_second": 2609.313
    },
    {
      "epoch": 1.2045209903121636,
      "grad_norm": 3.8723457691958174e-05,
      "learning_rate": 1.9603115960956852e-05,
      "loss": 0.1888,
      "num_input_tokens_seen": 1684320,
      "step": 560,
      "train_runtime": 645.2616,
      "train_tokens_per_second": 2610.29
    },
    {
      "epoch": 1.2260495156081808,
      "grad_norm": 2.192369538533967e-05,
      "learning_rate": 1.9280469364791505e-05,
      "loss": 0.1687,
      "num_input_tokens_seen": 1713920,
      "step": 570,
      "train_runtime": 656.4418,
      "train_tokens_per_second": 2610.924
    },
    {
      "epoch": 1.247578040904198,
      "grad_norm": 9.329665772384033e-05,
      "learning_rate": 1.8955651945856354e-05,
      "loss": 0.1797,
      "num_input_tokens_seen": 1744192,
      "step": 580,
      "train_runtime": 667.8491,
      "train_tokens_per_second": 2611.656
    },
    {
      "epoch": 1.2691065662002152,
      "grad_norm": 2.671034053491894e-05,
      "learning_rate": 1.8628828433995013e-05,
      "loss": 0.1488,
      "num_input_tokens_seen": 1773760,
      "step": 590,
      "train_runtime": 679.0241,
      "train_tokens_per_second": 2612.22
    },
    {
      "epoch": 1.2906350914962326,
      "grad_norm": 2.8991107683395967e-05,
      "learning_rate": 1.8300164576433035e-05,
      "loss": 0.1584,
      "num_input_tokens_seen": 1803424,
      "step": 600,
      "train_runtime": 690.3044,
      "train_tokens_per_second": 2612.505
    },
    {
      "epoch": 1.3121636167922497,
      "grad_norm": 3.321948315715417e-05,
      "learning_rate": 1.7969827053719862e-05,
      "loss": 0.1658,
      "num_input_tokens_seen": 1834048,
      "step": 610,
      "train_runtime": 701.7741,
      "train_tokens_per_second": 2613.445
    },
    {
      "epoch": 1.333692142088267,
      "grad_norm": 2.9109982278896496e-05,
      "learning_rate": 1.7637983395197465e-05,
      "loss": 0.1912,
      "num_input_tokens_seen": 1863328,
      "step": 620,
      "train_runtime": 712.9314,
      "train_tokens_per_second": 2613.615
    },
    {
      "epoch": 1.3552206673842842,
      "grad_norm": 5.012260953662917e-05,
      "learning_rate": 1.7304801894038522e-05,
      "loss": 0.1538,
      "num_input_tokens_seen": 1894080,
      "step": 630,
      "train_runtime": 724.275,
      "train_tokens_per_second": 2615.139
    },
    {
      "epoch": 1.3767491926803013,
      "grad_norm": 4.207729216432199e-05,
      "learning_rate": 1.6970451521897198e-05,
      "loss": 0.1397,
      "num_input_tokens_seen": 1924288,
      "step": 640,
      "train_runtime": 735.6149,
      "train_tokens_per_second": 2615.89
    },
    {
      "epoch": 1.3982777179763186,
      "grad_norm": 8.394964970648289e-05,
      "learning_rate": 1.663510184321587e-05,
      "loss": 0.1773,
      "num_input_tokens_seen": 1953824,
      "step": 650,
      "train_runtime": 746.7518,
      "train_tokens_per_second": 2616.43
    },
    {
      "epoch": 1.419806243272336,
      "grad_norm": 2.6376255846116692e-05,
      "learning_rate": 1.6298922929231194e-05,
      "loss": 0.153,
      "num_input_tokens_seen": 1983328,
      "step": 660,
      "train_runtime": 757.9924,
      "train_tokens_per_second": 2616.554
    },
    {
      "epoch": 1.441334768568353,
      "grad_norm": 4.21045879193116e-05,
      "learning_rate": 1.5962085271723158e-05,
      "loss": 0.1353,
      "num_input_tokens_seen": 2013504,
      "step": 670,
      "train_runtime": 769.3073,
      "train_tokens_per_second": 2617.295
    },
    {
      "epoch": 1.4628632938643702,
      "grad_norm": 4.316286867833696e-05,
      "learning_rate": 1.5624759696550874e-05,
      "loss": 0.1786,
      "num_input_tokens_seen": 2044480,
      "step": 680,
      "train_runtime": 780.8845,
      "train_tokens_per_second": 2618.159
    },
    {
      "epoch": 1.4843918191603875,
      "grad_norm": 3.7209880247246474e-05,
      "learning_rate": 1.528711727701893e-05,
      "loss": 0.1306,
      "num_input_tokens_seen": 2074656,
      "step": 690,
      "train_runtime": 792.2182,
      "train_tokens_per_second": 2618.794
    },
    {
      "epoch": 1.5059203444564049,
      "grad_norm": 5.088735997560434e-05,
      "learning_rate": 1.4949329247118258e-05,
      "loss": 0.1689,
      "num_input_tokens_seen": 2104672,
      "step": 700,
      "train_runtime": 803.5281,
      "train_tokens_per_second": 2619.289
    },
    {
      "epoch": 1.527448869752422,
      "grad_norm": 3.9965700125321746e-05,
      "learning_rate": 1.4611566914685489e-05,
      "loss": 0.1892,
      "num_input_tokens_seen": 2134720,
      "step": 710,
      "train_runtime": 814.7904,
      "train_tokens_per_second": 2619.962
    },
    {
      "epoch": 1.5489773950484391,
      "grad_norm": 6.375992961693555e-05,
      "learning_rate": 1.4274001574524894e-05,
      "loss": 0.1655,
      "num_input_tokens_seen": 2164960,
      "step": 720,
      "train_runtime": 826.0604,
      "train_tokens_per_second": 2620.825
    },
    {
      "epoch": 1.5705059203444565,
      "grad_norm": 2.914392462116666e-05,
      "learning_rate": 1.3936804421536906e-05,
      "loss": 0.134,
      "num_input_tokens_seen": 2193984,
      "step": 730,
      "train_runtime": 837.5845,
      "train_tokens_per_second": 2619.418
    },
    {
      "epoch": 1.5920344456404736,
      "grad_norm": 3.144361835438758e-05,
      "learning_rate": 1.3600146463897327e-05,
      "loss": 0.1549,
      "num_input_tokens_seen": 2223712,
      "step": 740,
      "train_runtime": 848.8344,
      "train_tokens_per_second": 2619.724
    },
    {
      "epoch": 1.6135629709364907,
      "grad_norm": 3.254741750424728e-05,
      "learning_rate": 1.3264198436331193e-05,
      "loss": 0.1806,
      "num_input_tokens_seen": 2253888,
      "step": 750,
      "train_runtime": 860.0879,
      "train_tokens_per_second": 2620.532
    },
    {
      "epoch": 1.635091496232508,
      "grad_norm": 2.9551278203143738e-05,
      "learning_rate": 1.2929130713525396e-05,
      "loss": 0.1738,
      "num_input_tokens_seen": 2283488,
      "step": 760,
      "train_runtime": 871.3384,
      "train_tokens_per_second": 2620.667
    },
    {
      "epoch": 1.6566200215285254,
      "grad_norm": 4.3200441723456606e-05,
      "learning_rate": 1.2595113223723788e-05,
      "loss": 0.16,
      "num_input_tokens_seen": 2314080,
      "step": 770,
      "train_runtime": 882.7871,
      "train_tokens_per_second": 2621.334
    },
    {
      "epoch": 1.6781485468245425,
      "grad_norm": 7.302625454030931e-05,
      "learning_rate": 1.2262315362548825e-05,
      "loss": 0.1389,
      "num_input_tokens_seen": 2344832,
      "step": 780,
      "train_runtime": 894.2092,
      "train_tokens_per_second": 2622.241
    },
    {
      "epoch": 1.6996770721205596,
      "grad_norm": 8.391755545744672e-05,
      "learning_rate": 1.1930905907093196e-05,
      "loss": 0.1663,
      "num_input_tokens_seen": 2373376,
      "step": 790,
      "train_runtime": 905.2972,
      "train_tokens_per_second": 2621.654
    },
    {
      "epoch": 1.721205597416577,
      "grad_norm": 3.845444371108897e-05,
      "learning_rate": 1.1601052930325236e-05,
      "loss": 0.1443,
      "num_input_tokens_seen": 2403456,
      "step": 800,
      "train_runtime": 916.6541,
      "train_tokens_per_second": 2621.988
    },
    {
      "epoch": 1.7427341227125943,
      "grad_norm": 5.4990796343190596e-05,
      "learning_rate": 1.127292371585141e-05,
      "loss": 0.1557,
      "num_input_tokens_seen": 2433760,
      "step": 810,
      "train_runtime": 928.0614,
      "train_tokens_per_second": 2622.413
    },
    {
      "epoch": 1.7642626480086114,
      "grad_norm": 4.249240373610519e-05,
      "learning_rate": 1.0946684673079074e-05,
      "loss": 0.1604,
      "num_input_tokens_seen": 2464032,
      "step": 820,
      "train_runtime": 939.3978,
      "train_tokens_per_second": 2622.991
    },
    {
      "epoch": 1.7857911733046286,
      "grad_norm": 3.500353341223672e-05,
      "learning_rate": 1.0622501252822628e-05,
      "loss": 0.1314,
      "num_input_tokens_seen": 2493920,
      "step": 830,
      "train_runtime": 950.6704,
      "train_tokens_per_second": 2623.328
    },
    {
      "epoch": 1.807319698600646,
      "grad_norm": 3.529454625095241e-05,
      "learning_rate": 1.0300537863395823e-05,
      "loss": 0.1494,
      "num_input_tokens_seen": 2524320,
      "step": 840,
      "train_runtime": 962.0766,
      "train_tokens_per_second": 2623.824
    },
    {
      "epoch": 1.8288482238966632,
      "grad_norm": 2.364149440836627e-05,
      "learning_rate": 9.98095778723271e-06,
      "loss": 0.1329,
      "num_input_tokens_seen": 2554112,
      "step": 850,
      "train_runtime": 973.349,
      "train_tokens_per_second": 2624.045
    },
    {
      "epoch": 1.8503767491926801,
      "grad_norm": 3.57489989255555e-05,
      "learning_rate": 9.663923098079656e-06,
      "loss": 0.1693,
      "num_input_tokens_seen": 2584512,
      "step": 860,
      "train_runtime": 984.7952,
      "train_tokens_per_second": 2624.416
    },
    {
      "epoch": 1.8719052744886975,
      "grad_norm": 2.4963468604255468e-05,
      "learning_rate": 9.349594578800259e-06,
      "loss": 0.1695,
      "num_input_tokens_seen": 2614944,
      "step": 870,
      "train_runtime": 996.2639,
      "train_tokens_per_second": 2624.75
    },
    {
      "epoch": 1.8934337997847148,
      "grad_norm": 6.0471880715340376e-05,
      "learning_rate": 9.038131639834977e-06,
      "loss": 0.1722,
      "num_input_tokens_seen": 2644928,
      "step": 880,
      "train_runtime": 1007.6479,
      "train_tokens_per_second": 2624.853
    },
    {
      "epoch": 1.914962325080732,
      "grad_norm": 3.468965587671846e-05,
      "learning_rate": 8.7296922383568e-06,
      "loss": 0.1674,
      "num_input_tokens_seen": 2674560,
      "step": 890,
      "train_runtime": 1019.0002,
      "train_tokens_per_second": 2624.69
    },
    {
      "epoch": 1.936490850376749,
      "grad_norm": 5.565357787418179e-05,
      "learning_rate": 8.424432798163838e-06,
      "loss": 0.1654,
      "num_input_tokens_seen": 2704224,
      "step": 900,
      "train_runtime": 1030.174,
      "train_tokens_per_second": 2625.017
    },
    {
      "epoch": 1.9580193756727664,
      "grad_norm": 2.5728633772814646e-05,
      "learning_rate": 8.122508130349711e-06,
      "loss": 0.1191,
      "num_input_tokens_seen": 2734048,
      "step": 910,
      "train_runtime": 1041.4,
      "train_tokens_per_second": 2625.358
    },
    {
      "epoch": 1.9795479009687837,
      "grad_norm": 4.2412902985233814e-05,
      "learning_rate": 7.82407135479164e-06,
      "loss": 0.1874,
      "num_input_tokens_seen": 2764064,
      "step": 920,
      "train_runtime": 1052.6791,
      "train_tokens_per_second": 2625.742
    },
    {
      "epoch": 2.0,
      "grad_norm": 7.404840289382264e-05,
      "learning_rate": 7.529273822496379e-06,
      "loss": 0.149,
      "num_input_tokens_seen": 2793184,
      "step": 930,
      "train_runtime": 1063.5331,
      "train_tokens_per_second": 2626.325
    },
    {
      "epoch": 2.0215285252960173,
      "grad_norm": 7.416486914735287e-05,
      "learning_rate": 7.238265038843143e-06,
      "loss": 0.1483,
      "num_input_tokens_seen": 2822880,
      "step": 940,
      "train_runtime": 1074.7444,
      "train_tokens_per_second": 2626.559
    },
    {
      "epoch": 2.0430570505920342,
      "grad_norm": 3.8171758205862716e-05,
      "learning_rate": 6.951192587762555e-06,
      "loss": 0.1234,
      "num_input_tokens_seen": 2852800,
      "step": 950,
      "train_runtime": 1086.0291,
      "train_tokens_per_second": 2626.817
    },
    {
      "epoch": 2.0645855758880516,
      "grad_norm": 4.259482375346124e-05,
      "learning_rate": 6.668202056890052e-06,
      "loss": 0.1319,
      "num_input_tokens_seen": 2882592,
      "step": 960,
      "train_runtime": 1097.2876,
      "train_tokens_per_second": 2627.016
    },
    {
      "epoch": 2.086114101184069,
      "grad_norm": 3.2160005503101274e-05,
      "learning_rate": 6.389436963731707e-06,
      "loss": 0.1361,
      "num_input_tokens_seen": 2912576,
      "step": 970,
      "train_runtime": 1108.6746,
      "train_tokens_per_second": 2627.079
    },
    {
      "epoch": 2.1076426264800863,
      "grad_norm": 4.2470204789424315e-05,
      "learning_rate": 6.115038682879904e-06,
      "loss": 0.1592,
      "num_input_tokens_seen": 2942976,
      "step": 980,
      "train_runtime": 1120.0755,
      "train_tokens_per_second": 2627.48
    },
    {
      "epoch": 2.129171151776103,
      "grad_norm": 5.224125925451517e-05,
      "learning_rate": 5.845146374315781e-06,
      "loss": 0.1256,
      "num_input_tokens_seen": 2972800,
      "step": 990,
      "train_runtime": 1131.3002,
      "train_tokens_per_second": 2627.773
    },
    {
      "epoch": 2.1506996770721205,
      "grad_norm": 4.592683399096131e-05,
      "learning_rate": 5.579896912834807e-06,
      "loss": 0.1648,
      "num_input_tokens_seen": 3003424,
      "step": 1000,
      "train_runtime": 1142.7014,
      "train_tokens_per_second": 2628.354
    },
    {
      "epoch": 2.172228202368138,
      "grad_norm": 3.322002521599643e-05,
      "learning_rate": 5.319424818631293e-06,
      "loss": 0.1586,
      "num_input_tokens_seen": 3033728,
      "step": 1010,
      "train_runtime": 1155.3218,
      "train_tokens_per_second": 2625.873
    },
    {
      "epoch": 2.193756727664155,
      "grad_norm": 3.684938928927295e-05,
      "learning_rate": 5.063862189076988e-06,
      "loss": 0.1566,
      "num_input_tokens_seen": 3064128,
      "step": 1020,
      "train_runtime": 1166.6004,
      "train_tokens_per_second": 2626.545
    },
    {
      "epoch": 2.215285252960172,
      "grad_norm": 4.779713344760239e-05,
      "learning_rate": 4.813338631728471e-06,
      "loss": 0.1442,
      "num_input_tokens_seen": 3094464,
      "step": 1030,
      "train_runtime": 1177.9038,
      "train_tokens_per_second": 2627.094
    },
    {
      "epoch": 2.2368137782561894,
      "grad_norm": 5.730693737859838e-05,
      "learning_rate": 4.567981198597168e-06,
      "loss": 0.1497,
      "num_input_tokens_seen": 3124960,
      "step": 1040,
      "train_runtime": 1189.2859,
      "train_tokens_per_second": 2627.594
    },
    {
      "epoch": 2.2583423035522068,
      "grad_norm": 4.8961104766931385e-05,
      "learning_rate": 4.327914321715459e-06,
      "loss": 0.1184,
      "num_input_tokens_seen": 3154752,
      "step": 1050,
      "train_runtime": 1200.5146,
      "train_tokens_per_second": 2627.833
    },
    {
      "epoch": 2.2798708288482237,
      "grad_norm": 6.034982288838364e-05,
      "learning_rate": 4.09325975003146e-06,
      "loss": 0.1312,
      "num_input_tokens_seen": 3184416,
      "step": 1060,
      "train_runtime": 1211.7687,
      "train_tokens_per_second": 2627.908
    },
    {
      "epoch": 2.301399354144241,
      "grad_norm": 4.687267210101709e-05,
      "learning_rate": 3.864136487664538e-06,
      "loss": 0.1339,
      "num_input_tokens_seen": 3213856,
      "step": 1070,
      "train_runtime": 1222.9391,
      "train_tokens_per_second": 2627.977
    },
    {
      "epoch": 2.3229278794402584,
      "grad_norm": 2.0478708393056877e-05,
      "learning_rate": 3.640660733552863e-06,
      "loss": 0.0975,
      "num_input_tokens_seen": 3243488,
      "step": 1080,
      "train_runtime": 1234.1558,
      "train_tokens_per_second": 2628.103
    },
    {
      "epoch": 2.3444564047362757,
      "grad_norm": 4.68106591142714e-05,
      "learning_rate": 3.4229458225235696e-06,
      "loss": 0.1083,
      "num_input_tokens_seen": 3274720,
      "step": 1090,
      "train_runtime": 1245.7361,
      "train_tokens_per_second": 2628.743
    },
    {
      "epoch": 2.3659849300322926,
      "grad_norm": 3.146538438159041e-05,
      "learning_rate": 3.211102167815469e-06,
      "loss": 0.1186,
      "num_input_tokens_seen": 3304320,
      "step": 1100,
      "train_runtime": 1256.915,
      "train_tokens_per_second": 2628.913
    },
    {
      "epoch": 2.38751345532831,
      "grad_norm": 5.921598494751379e-05,
      "learning_rate": 3.0052372050834025e-06,
      "loss": 0.1287,
      "num_input_tokens_seen": 3335264,
      "step": 1110,
      "train_runtime": 1268.3967,
      "train_tokens_per_second": 2629.512
    },
    {
      "epoch": 2.4090419806243273,
      "grad_norm": 3.6115765396971256e-05,
      "learning_rate": 2.805455337912722e-06,
      "loss": 0.1308,
      "num_input_tokens_seen": 3366144,
      "step": 1120,
      "train_runtime": 1279.8947,
      "train_tokens_per_second": 2630.016
    },
    {
      "epoch": 2.4305705059203446,
      "grad_norm": 5.221314859227277e-05,
      "learning_rate": 2.611857884871379e-06,
      "loss": 0.1617,
      "num_input_tokens_seen": 3396096,
      "step": 1130,
      "train_runtime": 1291.193,
      "train_tokens_per_second": 2630.2
    },
    {
      "epoch": 2.4520990312163615,
      "grad_norm": 4.589893069351092e-05,
      "learning_rate": 2.424543028126661e-06,
      "loss": 0.1538,
      "num_input_tokens_seen": 3426048,
      "step": 1140,
      "train_runtime": 1302.5244,
      "train_tokens_per_second": 2630.314
    },
    {
      "epoch": 2.473627556512379,
      "grad_norm": 3.316588481538929e-05,
      "learning_rate": 2.243605763652489e-06,
      "loss": 0.1326,
      "num_input_tokens_seen": 3456352,
      "step": 1150,
      "train_runtime": 1313.8974,
      "train_tokens_per_second": 2630.61
    },
    {
      "epoch": 2.495156081808396,
      "grad_norm": 2.757367110461928e-05,
      "learning_rate": 2.0691378530525933e-06,
      "loss": 0.1428,
      "num_input_tokens_seen": 3486048,
      "step": 1160,
      "train_runtime": 1325.1036,
      "train_tokens_per_second": 2630.774
    },
    {
      "epoch": 2.5166846071044136,
      "grad_norm": 5.3478455811273307e-05,
      "learning_rate": 1.9012277770239829e-06,
      "loss": 0.1429,
      "num_input_tokens_seen": 3515904,
      "step": 1170,
      "train_runtime": 1336.3712,
      "train_tokens_per_second": 2630.934
    },
    {
      "epoch": 2.5382131324004304,
      "grad_norm": 2.4738732463447377e-05,
      "learning_rate": 1.7399606904843224e-06,
      "loss": 0.1101,
      "num_input_tokens_seen": 3545792,
      "step": 1180,
      "train_runtime": 1347.7396,
      "train_tokens_per_second": 2630.918
    },
    {
      "epoch": 2.559741657696448,
      "grad_norm": 3.890502921422012e-05,
      "learning_rate": 1.5854183793859577e-06,
      "loss": 0.1437,
      "num_input_tokens_seen": 3576544,
      "step": 1190,
      "train_runtime": 1359.269,
      "train_tokens_per_second": 2631.226
    },
    {
      "epoch": 2.581270182992465,
      "grad_norm": 4.013535726699047e-05,
      "learning_rate": 1.437679219238498e-06,
      "loss": 0.1383,
      "num_input_tokens_seen": 3606592,
      "step": 1200,
      "train_runtime": 1370.4896,
      "train_tokens_per_second": 2631.608
    },
    {
      "epoch": 2.602798708288482,
      "grad_norm": 7.226341404020786e-05,
      "learning_rate": 1.2968181353609854e-06,
      "loss": 0.1266,
      "num_input_tokens_seen": 3636960,
      "step": 1210,
      "train_runtime": 1381.7996,
      "train_tokens_per_second": 2632.046
    },
    {
      "epoch": 2.6243272335844994,
      "grad_norm": 4.252417420502752e-05,
      "learning_rate": 1.1629065648838495e-06,
      "loss": 0.1577,
      "num_input_tokens_seen": 3666112,
      "step": 1220,
      "train_runtime": 1392.9941,
      "train_tokens_per_second": 2631.822
    },
    {
      "epoch": 2.6458557588805167,
      "grad_norm": 4.0854803955880925e-05,
      "learning_rate": 1.0360124205198102e-06,
      "loss": 0.1188,
      "num_input_tokens_seen": 3696064,
      "step": 1230,
      "train_runtime": 1404.3361,
      "train_tokens_per_second": 2631.894
    },
    {
      "epoch": 2.667384284176534,
      "grad_norm": 5.907309605390765e-05,
      "learning_rate": 9.162000561222672e-07,
      "loss": 0.1446,
      "num_input_tokens_seen": 3727040,
      "step": 1240,
      "train_runtime": 1415.8346,
      "train_tokens_per_second": 2632.398
    },
    {
      "epoch": 2.6889128094725514,
      "grad_norm": 4.082039595232345e-05,
      "learning_rate": 8.035302340484646e-07,
      "loss": 0.1382,
      "num_input_tokens_seen": 3756768,
      "step": 1250,
      "train_runtime": 1427.0816,
      "train_tokens_per_second": 2632.483
    },
    {
      "epoch": 2.7104413347685683,
      "grad_norm": 4.765476842294447e-05,
      "learning_rate": 6.980600943441273e-07,
      "loss": 0.1281,
      "num_input_tokens_seen": 3786688,
      "step": 1260,
      "train_runtime": 1438.424,
      "train_tokens_per_second": 2632.526
    },
    {
      "epoch": 2.7319698600645856,
      "grad_norm": 3.876964547089301e-05,
      "learning_rate": 5.998431257651132e-07,
      "loss": 0.1494,
      "num_input_tokens_seen": 3816864,
      "step": 1270,
      "train_runtime": 1449.8712,
      "train_tokens_per_second": 2632.554
    },
    {
      "epoch": 2.7534983853606025,
      "grad_norm": 4.9277557991445065e-05,
      "learning_rate": 5.089291386508055e-07,
      "loss": 0.126,
      "num_input_tokens_seen": 3847040,
      "step": 1280,
      "train_runtime": 1461.3789,
      "train_tokens_per_second": 2632.473
    },
    {
      "epoch": 2.77502691065662,
      "grad_norm": 3.499957892927341e-05,
      "learning_rate": 4.2536423966299256e-07,
      "loss": 0.1189,
      "num_input_tokens_seen": 3876832,
      "step": 1290,
      "train_runtime": 1472.7417,
      "train_tokens_per_second": 2632.391
    },
    {
      "epoch": 2.7965554359526372,
      "grad_norm": 4.8889476602198556e-05,
      "learning_rate": 3.491908084030776e-07,
      "loss": 0.1345,
      "num_input_tokens_seen": 3907328,
      "step": 1300,
      "train_runtime": 1484.2866,
      "train_tokens_per_second": 2632.462
    },
    {
      "epoch": 2.8180839612486546,
      "grad_norm": 4.742259989143349e-05,
      "learning_rate": 2.804474759194176e-07,
      "loss": 0.1403,
      "num_input_tokens_seen": 3937984,
      "step": 1310,
      "train_runtime": 1495.8239,
      "train_tokens_per_second": 2632.652
    },
    {
      "epoch": 2.839612486544672,
      "grad_norm": 4.425349470693618e-05,
      "learning_rate": 2.1916910511575627e-07,
      "loss": 0.151,
      "num_input_tokens_seen": 3967840,
      "step": 1320,
      "train_runtime": 1507.2136,
      "train_tokens_per_second": 2632.566
    },
    {
      "epoch": 2.861141011840689,
      "grad_norm": 4.406994048622437e-05,
      "learning_rate": 1.6538677307062933e-07,
      "loss": 0.1459,
      "num_input_tokens_seen": 3998176,
      "step": 1330,
      "train_runtime": 1518.6021,
      "train_tokens_per_second": 2632.8
    },
    {
      "epoch": 2.882669537136706,
      "grad_norm": 7.402410119539127e-05,
      "learning_rate": 1.1912775527676278e-07,
      "loss": 0.1396,
      "num_input_tokens_seen": 4027936,
      "step": 1340,
      "train_runtime": 1529.9966,
      "train_tokens_per_second": 2632.644
    },
    {
      "epoch": 2.9041980624327235,
      "grad_norm": 5.9873302234336734e-05,
      "learning_rate": 8.04155118084099e-08,
      "loss": 0.1437,
      "num_input_tokens_seen": 4057504,
      "step": 1350,
      "train_runtime": 1541.1503,
      "train_tokens_per_second": 2632.776
    },
    {
      "epoch": 2.9257265877287404,
      "grad_norm": 4.384098429000005e-05,
      "learning_rate": 4.9269675423671556e-08,
      "loss": 0.1437,
      "num_input_tokens_seen": 4087680,
      "step": 1360,
      "train_runtime": 1552.5119,
      "train_tokens_per_second": 2632.946
    },
    {
      "epoch": 2.9472551130247577,
      "grad_norm": 5.602654709946364e-05,
      "learning_rate": 2.570604160781753e-08,
      "loss": 0.1067,
      "num_input_tokens_seen": 4117632,
      "step": 1370,
      "train_runtime": 1563.7521,
      "train_tokens_per_second": 2633.174
    },
    {
      "epoch": 2.968783638320775,
      "grad_norm": 5.69059957342688e-05,
      "learning_rate": 9.736560562675912e-09,
      "loss": 0.1254,
      "num_input_tokens_seen": 4147552,
      "step": 1380,
      "train_runtime": 1575.0768,
      "train_tokens_per_second": 2633.238
    },
    {
      "epoch": 2.9903121636167924,
      "grad_norm": 3.0387620427063666e-05,
      "learning_rate": 1.3693311461326908e-09,
      "loss": 0.1337,
      "num_input_tokens_seen": 4178016,
      "step": 1390,
      "train_runtime": 1586.5117,
      "train_tokens_per_second": 2633.461
    },
    {
      "epoch": 3.0,
      "num_input_tokens_seen": 4191488,
      "step": 1395,
      "total_flos": 1.8926822146808218e+17,
      "train_loss": 0.18203292623642953,
      "train_runtime": 1592.8694,
      "train_samples_per_second": 6.999,
      "train_steps_per_second": 0.876
    }
  ],
  "logging_steps": 10,
  "max_steps": 1395,
  "num_input_tokens_seen": 4191488,
  "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": 1.8926822146808218e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}