{
  "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": 16.47992515563965,
      "learning_rate": 9.998973021172564e-06,
      "loss": 1.7834,
      "num_input_tokens_seen": 43648,
      "step": 10,
      "train_runtime": 64.3878,
      "train_tokens_per_second": 677.892
    },
    {
      "epoch": 0.04305705059203445,
      "grad_norm": 16.372161865234375,
      "learning_rate": 9.995423512524277e-06,
      "loss": 0.7676,
      "num_input_tokens_seen": 87072,
      "step": 20,
      "train_runtime": 95.1979,
      "train_tokens_per_second": 914.642
    },
    {
      "epoch": 0.06458557588805167,
      "grad_norm": 11.686817169189453,
      "learning_rate": 9.98934059499448e-06,
      "loss": 0.5817,
      "num_input_tokens_seen": 131520,
      "step": 30,
      "train_runtime": 126.5065,
      "train_tokens_per_second": 1039.63
    },
    {
      "epoch": 0.0861141011840689,
      "grad_norm": 3.2975926399230957,
      "learning_rate": 9.980727353510257e-06,
      "loss": 0.3531,
      "num_input_tokens_seen": 175680,
      "step": 40,
      "train_runtime": 157.6702,
      "train_tokens_per_second": 1114.225
    },
    {
      "epoch": 0.10764262648008611,
      "grad_norm": 3.4886627197265625,
      "learning_rate": 9.969588156242282e-06,
      "loss": 0.3352,
      "num_input_tokens_seen": 219584,
      "step": 50,
      "train_runtime": 188.7104,
      "train_tokens_per_second": 1163.603
    },
    {
      "epoch": 0.12917115177610333,
      "grad_norm": 3.484487295150757,
      "learning_rate": 9.95592865238951e-06,
      "loss": 0.2298,
      "num_input_tokens_seen": 263392,
      "step": 60,
      "train_runtime": 219.7077,
      "train_tokens_per_second": 1198.829
    },
    {
      "epoch": 0.15069967707212056,
      "grad_norm": 4.422971725463867,
      "learning_rate": 9.939755769314215e-06,
      "loss": 0.1981,
      "num_input_tokens_seen": 306912,
      "step": 70,
      "train_runtime": 250.6069,
      "train_tokens_per_second": 1224.675
    },
    {
      "epoch": 0.1722282023681378,
      "grad_norm": 1.4185198545455933,
      "learning_rate": 9.9210777090288e-06,
      "loss": 0.1755,
      "num_input_tokens_seen": 349824,
      "step": 80,
      "train_runtime": 281.2405,
      "train_tokens_per_second": 1243.86
    },
    {
      "epoch": 0.193756727664155,
      "grad_norm": 1.1935056447982788,
      "learning_rate": 9.899903944036185e-06,
      "loss": 0.1791,
      "num_input_tokens_seen": 393664,
      "step": 90,
      "train_runtime": 312.2075,
      "train_tokens_per_second": 1260.905
    },
    {
      "epoch": 0.21528525296017223,
      "grad_norm": 2.463397741317749,
      "learning_rate": 9.87624521252587e-06,
      "loss": 0.2051,
      "num_input_tokens_seen": 437952,
      "step": 100,
      "train_runtime": 343.3508,
      "train_tokens_per_second": 1275.523
    },
    {
      "epoch": 0.23681377825618946,
      "grad_norm": 3.8650355339050293,
      "learning_rate": 9.850113512928094e-06,
      "loss": 0.1912,
      "num_input_tokens_seen": 481472,
      "step": 110,
      "train_runtime": 374.2539,
      "train_tokens_per_second": 1286.485
    },
    {
      "epoch": 0.25834230355220666,
      "grad_norm": 4.704444885253906,
      "learning_rate": 9.821522097828884e-06,
      "loss": 0.1637,
      "num_input_tokens_seen": 525472,
      "step": 120,
      "train_runtime": 405.4374,
      "train_tokens_per_second": 1296.062
    },
    {
      "epoch": 0.2798708288482239,
      "grad_norm": 2.7267754077911377,
      "learning_rate": 9.790485467249065e-06,
      "loss": 0.1537,
      "num_input_tokens_seen": 569696,
      "step": 130,
      "train_runtime": 436.4834,
      "train_tokens_per_second": 1305.195
    },
    {
      "epoch": 0.3013993541442411,
      "grad_norm": 3.512826681137085,
      "learning_rate": 9.757019361290621e-06,
      "loss": 0.1846,
      "num_input_tokens_seen": 613760,
      "step": 140,
      "train_runtime": 467.5968,
      "train_tokens_per_second": 1312.584
    },
    {
      "epoch": 0.32292787944025836,
      "grad_norm": 2.484517812728882,
      "learning_rate": 9.721140752154182e-06,
      "loss": 0.1872,
      "num_input_tokens_seen": 657824,
      "step": 150,
      "train_runtime": 498.6389,
      "train_tokens_per_second": 1319.239
    },
    {
      "epoch": 0.3444564047362756,
      "grad_norm": 2.296607255935669,
      "learning_rate": 9.682867835531624e-06,
      "loss": 0.1372,
      "num_input_tokens_seen": 701760,
      "step": 160,
      "train_runtime": 529.7256,
      "train_tokens_per_second": 1324.761
    },
    {
      "epoch": 0.36598493003229277,
      "grad_norm": 3.887946128845215,
      "learning_rate": 9.642220021378212e-06,
      "loss": 0.2028,
      "num_input_tokens_seen": 745184,
      "step": 170,
      "train_runtime": 560.5239,
      "train_tokens_per_second": 1329.442
    },
    {
      "epoch": 0.38751345532831,
      "grad_norm": 2.1227307319641113,
      "learning_rate": 9.59921792406891e-06,
      "loss": 0.1765,
      "num_input_tokens_seen": 789216,
      "step": 180,
      "train_runtime": 591.5958,
      "train_tokens_per_second": 1334.046
    },
    {
      "epoch": 0.40904198062432723,
      "grad_norm": 3.007882595062256,
      "learning_rate": 9.553883351943882e-06,
      "loss": 0.1882,
      "num_input_tokens_seen": 832960,
      "step": 190,
      "train_runtime": 622.6094,
      "train_tokens_per_second": 1337.853
    },
    {
      "epoch": 0.43057050592034446,
      "grad_norm": 1.8637208938598633,
      "learning_rate": 9.506239296248497e-06,
      "loss": 0.1336,
      "num_input_tokens_seen": 877664,
      "step": 200,
      "train_runtime": 654.0867,
      "train_tokens_per_second": 1341.816
    },
    {
      "epoch": 0.4520990312163617,
      "grad_norm": 1.1324244737625122,
      "learning_rate": 9.456309919473384e-06,
      "loss": 0.1521,
      "num_input_tokens_seen": 921536,
      "step": 210,
      "train_runtime": 685.133,
      "train_tokens_per_second": 1345.047
    },
    {
      "epoch": 0.4736275565123789,
      "grad_norm": 3.536208391189575,
      "learning_rate": 9.404120543100553e-06,
      "loss": 0.1864,
      "num_input_tokens_seen": 965344,
      "step": 220,
      "train_runtime": 716.7323,
      "train_tokens_per_second": 1346.868
    },
    {
      "epoch": 0.4951560818083961,
      "grad_norm": 1.73074209690094,
      "learning_rate": 9.34969763476168e-06,
      "loss": 0.1625,
      "num_input_tokens_seen": 1008672,
      "step": 230,
      "train_runtime": 747.5266,
      "train_tokens_per_second": 1349.346
    },
    {
      "epoch": 0.5166846071044133,
      "grad_norm": 1.4830386638641357,
      "learning_rate": 9.293068794815175e-06,
      "loss": 0.1456,
      "num_input_tokens_seen": 1051936,
      "step": 240,
      "train_runtime": 778.3038,
      "train_tokens_per_second": 1351.575
    },
    {
      "epoch": 0.5382131324004306,
      "grad_norm": 2.7361512184143066,
      "learning_rate": 9.234262742348764e-06,
      "loss": 0.165,
      "num_input_tokens_seen": 1095904,
      "step": 250,
      "train_runtime": 809.3485,
      "train_tokens_per_second": 1354.057
    },
    {
      "epoch": 0.5597416576964478,
      "grad_norm": 1.9473741054534912,
      "learning_rate": 9.17330930061471e-06,
      "loss": 0.1487,
      "num_input_tokens_seen": 1140064,
      "step": 260,
      "train_runtime": 840.5464,
      "train_tokens_per_second": 1356.337
    },
    {
      "epoch": 0.581270182992465,
      "grad_norm": 2.0610098838806152,
      "learning_rate": 9.11023938190509e-06,
      "loss": 0.1687,
      "num_input_tokens_seen": 1183872,
      "step": 270,
      "train_runtime": 871.5774,
      "train_tokens_per_second": 1358.31
    },
    {
      "epoch": 0.6027987082884823,
      "grad_norm": 3.2189598083496094,
      "learning_rate": 9.045084971874738e-06,
      "loss": 0.1511,
      "num_input_tokens_seen": 1228544,
      "step": 280,
      "train_runtime": 903.1191,
      "train_tokens_per_second": 1360.334
    },
    {
      "epoch": 0.6243272335844995,
      "grad_norm": 3.902757406234741,
      "learning_rate": 8.977879113319847e-06,
      "loss": 0.184,
      "num_input_tokens_seen": 1271968,
      "step": 290,
      "train_runtime": 934.0522,
      "train_tokens_per_second": 1361.774
    },
    {
      "epoch": 0.6458557588805167,
      "grad_norm": 4.2983012199401855,
      "learning_rate": 8.90865588942046e-06,
      "loss": 0.1305,
      "num_input_tokens_seen": 1316096,
      "step": 300,
      "train_runtime": 965.1778,
      "train_tokens_per_second": 1363.579
    },
    {
      "epoch": 0.667384284176534,
      "grad_norm": 2.6156394481658936,
      "learning_rate": 8.83745040645531e-06,
      "loss": 0.1405,
      "num_input_tokens_seen": 1360256,
      "step": 310,
      "train_runtime": 996.4077,
      "train_tokens_per_second": 1365.16
    },
    {
      "epoch": 0.6889128094725512,
      "grad_norm": 1.3506444692611694,
      "learning_rate": 8.764298775997823e-06,
      "loss": 0.1651,
      "num_input_tokens_seen": 1404064,
      "step": 320,
      "train_runtime": 1027.5151,
      "train_tokens_per_second": 1366.466
    },
    {
      "epoch": 0.7104413347685683,
      "grad_norm": 1.5117799043655396,
      "learning_rate": 8.68923809660227e-06,
      "loss": 0.1519,
      "num_input_tokens_seen": 1448256,
      "step": 330,
      "train_runtime": 1058.74,
      "train_tokens_per_second": 1367.905
    },
    {
      "epoch": 0.7319698600645855,
      "grad_norm": 1.2659231424331665,
      "learning_rate": 8.612306434989395e-06,
      "loss": 0.1574,
      "num_input_tokens_seen": 1492448,
      "step": 340,
      "train_runtime": 1090.0052,
      "train_tokens_per_second": 1369.212
    },
    {
      "epoch": 0.7534983853606028,
      "grad_norm": 2.6010894775390625,
      "learning_rate": 8.53354280674102e-06,
      "loss": 0.1587,
      "num_input_tokens_seen": 1537440,
      "step": 350,
      "train_runtime": 1121.55,
      "train_tokens_per_second": 1370.817
    },
    {
      "epoch": 0.77502691065662,
      "grad_norm": 1.6968097686767578,
      "learning_rate": 8.452987156513457e-06,
      "loss": 0.1513,
      "num_input_tokens_seen": 1581280,
      "step": 360,
      "train_runtime": 1152.6858,
      "train_tokens_per_second": 1371.822
    },
    {
      "epoch": 0.7965554359526372,
      "grad_norm": 2.0298380851745605,
      "learning_rate": 8.370680337779737e-06,
      "loss": 0.1401,
      "num_input_tokens_seen": 1625824,
      "step": 370,
      "train_runtime": 1184.3507,
      "train_tokens_per_second": 1372.756
    },
    {
      "epoch": 0.8180839612486545,
      "grad_norm": 2.7614028453826904,
      "learning_rate": 8.286664092110935e-06,
      "loss": 0.1135,
      "num_input_tokens_seen": 1669472,
      "step": 380,
      "train_runtime": 1215.2935,
      "train_tokens_per_second": 1373.719
    },
    {
      "epoch": 0.8396124865446717,
      "grad_norm": 1.6297581195831299,
      "learning_rate": 8.200981028007095e-06,
      "loss": 0.1645,
      "num_input_tokens_seen": 1713312,
      "step": 390,
      "train_runtime": 1246.1254,
      "train_tokens_per_second": 1374.911
    },
    {
      "epoch": 0.8611410118406889,
      "grad_norm": 2.32612943649292,
      "learning_rate": 8.11367459928851e-06,
      "loss": 0.1129,
      "num_input_tokens_seen": 1757056,
      "step": 400,
      "train_runtime": 1277.0951,
      "train_tokens_per_second": 1375.822
    },
    {
      "epoch": 0.8826695371367062,
      "grad_norm": 1.5576270818710327,
      "learning_rate": 8.024789083058289e-06,
      "loss": 0.1333,
      "num_input_tokens_seen": 1801632,
      "step": 410,
      "train_runtime": 1308.5556,
      "train_tokens_per_second": 1376.81
    },
    {
      "epoch": 0.9041980624327234,
      "grad_norm": 8.420344352722168,
      "learning_rate": 7.934369557247397e-06,
      "loss": 0.1311,
      "num_input_tokens_seen": 1845280,
      "step": 420,
      "train_runtime": 1339.4668,
      "train_tokens_per_second": 1377.623
    },
    {
      "epoch": 0.9257265877287406,
      "grad_norm": 1.7291479110717773,
      "learning_rate": 7.842461877753575e-06,
      "loss": 0.1395,
      "num_input_tokens_seen": 1889472,
      "step": 430,
      "train_runtime": 1370.6369,
      "train_tokens_per_second": 1378.536
    },
    {
      "epoch": 0.9472551130247578,
      "grad_norm": 2.441693067550659,
      "learning_rate": 7.7491126551857e-06,
      "loss": 0.1175,
      "num_input_tokens_seen": 1932832,
      "step": 440,
      "train_runtime": 1401.3047,
      "train_tokens_per_second": 1379.309
    },
    {
      "epoch": 0.9687836383207751,
      "grad_norm": 1.3606727123260498,
      "learning_rate": 7.654369231225398e-06,
      "loss": 0.1154,
      "num_input_tokens_seen": 1976000,
      "step": 450,
      "train_runtime": 1431.9402,
      "train_tokens_per_second": 1379.946
    },
    {
      "epoch": 0.9903121636167922,
      "grad_norm": 2.911600351333618,
      "learning_rate": 7.5582796546179125e-06,
      "loss": 0.1408,
      "num_input_tokens_seen": 2019968,
      "step": 460,
      "train_runtime": 1463.035,
      "train_tokens_per_second": 1380.67
    },
    {
      "epoch": 1.0107642626480087,
      "grad_norm": 2.967047929763794,
      "learning_rate": 7.460892656804366e-06,
      "loss": 0.1126,
      "num_input_tokens_seen": 2061440,
      "step": 470,
      "train_runtime": 1492.4083,
      "train_tokens_per_second": 1381.284
    },
    {
      "epoch": 1.0322927879440258,
      "grad_norm": 1.90776789188385,
      "learning_rate": 7.362257627207818e-06,
      "loss": 0.1329,
      "num_input_tokens_seen": 2105216,
      "step": 480,
      "train_runtime": 1523.2979,
      "train_tokens_per_second": 1382.012
    },
    {
      "epoch": 1.0538213132400431,
      "grad_norm": 2.7770872116088867,
      "learning_rate": 7.2624245881856094e-06,
      "loss": 0.111,
      "num_input_tokens_seen": 2149888,
      "step": 490,
      "train_runtime": 1554.9287,
      "train_tokens_per_second": 1382.628
    },
    {
      "epoch": 1.0753498385360603,
      "grad_norm": 1.9737987518310547,
      "learning_rate": 7.161444169660723e-06,
      "loss": 0.113,
      "num_input_tokens_seen": 2194304,
      "step": 500,
      "train_runtime": 1586.2216,
      "train_tokens_per_second": 1383.353
    },
    {
      "epoch": 1.0968783638320776,
      "grad_norm": 1.2335457801818848,
      "learning_rate": 7.059367583445034e-06,
      "loss": 0.1303,
      "num_input_tokens_seen": 2238080,
      "step": 510,
      "train_runtime": 1617.0801,
      "train_tokens_per_second": 1384.025
    },
    {
      "epoch": 1.1184068891280947,
      "grad_norm": 2.525258779525757,
      "learning_rate": 6.956246597267438e-06,
      "loss": 0.1404,
      "num_input_tokens_seen": 2282368,
      "step": 520,
      "train_runtime": 1648.4109,
      "train_tokens_per_second": 1384.587
    },
    {
      "epoch": 1.1399354144241118,
      "grad_norm": 4.016520977020264,
      "learning_rate": 6.852133508520057e-06,
      "loss": 0.0972,
      "num_input_tokens_seen": 2325984,
      "step": 530,
      "train_runtime": 1679.3164,
      "train_tokens_per_second": 1385.078
    },
    {
      "epoch": 1.1614639397201292,
      "grad_norm": 1.405586838722229,
      "learning_rate": 6.747081117735829e-06,
      "loss": 0.1092,
      "num_input_tokens_seen": 2369664,
      "step": 540,
      "train_runtime": 1710.2475,
      "train_tokens_per_second": 1385.568
    },
    {
      "epoch": 1.1829924650161463,
      "grad_norm": 2.151442766189575,
      "learning_rate": 6.641142701810932e-06,
      "loss": 0.1186,
      "num_input_tokens_seen": 2413312,
      "step": 550,
      "train_runtime": 1741.213,
      "train_tokens_per_second": 1385.995
    },
    {
      "epoch": 1.2045209903121636,
      "grad_norm": 1.9706578254699707,
      "learning_rate": 6.534371986985618e-06,
      "loss": 0.1332,
      "num_input_tokens_seen": 2457120,
      "step": 560,
      "train_runtime": 1772.1204,
      "train_tokens_per_second": 1386.542
    },
    {
      "epoch": 1.2260495156081808,
      "grad_norm": 2.3035449981689453,
      "learning_rate": 6.426823121597169e-06,
      "loss": 0.1481,
      "num_input_tokens_seen": 2500736,
      "step": 570,
      "train_runtime": 1803.0834,
      "train_tokens_per_second": 1386.922
    },
    {
      "epoch": 1.247578040904198,
      "grad_norm": 3.713632106781006,
      "learning_rate": 6.318550648618785e-06,
      "loss": 0.1332,
      "num_input_tokens_seen": 2545056,
      "step": 580,
      "train_runtime": 1834.1935,
      "train_tokens_per_second": 1387.561
    },
    {
      "epoch": 1.2691065662002152,
      "grad_norm": 1.8667478561401367,
      "learning_rate": 6.209609477998339e-06,
      "loss": 0.0978,
      "num_input_tokens_seen": 2588416,
      "step": 590,
      "train_runtime": 1864.9784,
      "train_tokens_per_second": 1387.907
    },
    {
      "epoch": 1.2906350914962326,
      "grad_norm": 2.295342445373535,
      "learning_rate": 6.100054858811012e-06,
      "loss": 0.1176,
      "num_input_tokens_seen": 2632352,
      "step": 600,
      "train_runtime": 1896.108,
      "train_tokens_per_second": 1388.292
    },
    {
      "epoch": 1.3121636167922497,
      "grad_norm": 2.2867720127105713,
      "learning_rate": 5.989942351239954e-06,
      "loss": 0.1279,
      "num_input_tokens_seen": 2676640,
      "step": 610,
      "train_runtime": 1927.5126,
      "train_tokens_per_second": 1388.65
    },
    {
      "epoch": 1.333692142088267,
      "grad_norm": 2.1158223152160645,
      "learning_rate": 5.879327798399155e-06,
      "loss": 0.1407,
      "num_input_tokens_seen": 2719968,
      "step": 620,
      "train_runtime": 1958.2576,
      "train_tokens_per_second": 1388.974
    },
    {
      "epoch": 1.3552206673842842,
      "grad_norm": 4.076441287994385,
      "learning_rate": 5.768267298012841e-06,
      "loss": 0.1168,
      "num_input_tokens_seen": 2764352,
      "step": 630,
      "train_runtime": 1989.6781,
      "train_tokens_per_second": 1389.346
    },
    {
      "epoch": 1.3767491926803013,
      "grad_norm": 4.354236602783203,
      "learning_rate": 5.656817173965733e-06,
      "loss": 0.1188,
      "num_input_tokens_seen": 2808832,
      "step": 640,
      "train_runtime": 2021.0224,
      "train_tokens_per_second": 1389.807
    },
    {
      "epoch": 1.3982777179763186,
      "grad_norm": 4.40167236328125,
      "learning_rate": 5.545033947738624e-06,
      "loss": 0.1326,
      "num_input_tokens_seen": 2852192,
      "step": 650,
      "train_runtime": 2051.72,
      "train_tokens_per_second": 1390.147
    },
    {
      "epoch": 1.419806243272336,
      "grad_norm": 2.4845104217529297,
      "learning_rate": 5.4329743097437316e-06,
      "loss": 0.1331,
      "num_input_tokens_seen": 2896256,
      "step": 660,
      "train_runtime": 2082.7361,
      "train_tokens_per_second": 1390.602
    },
    {
      "epoch": 1.441334768568353,
      "grad_norm": 2.7233705520629883,
      "learning_rate": 5.320695090574386e-06,
      "loss": 0.1082,
      "num_input_tokens_seen": 2939552,
      "step": 670,
      "train_runtime": 2113.5367,
      "train_tokens_per_second": 1390.821
    },
    {
      "epoch": 1.4628632938643702,
      "grad_norm": 3.289949893951416,
      "learning_rate": 5.208253232183625e-06,
      "loss": 0.1184,
      "num_input_tokens_seen": 2984000,
      "step": 680,
      "train_runtime": 2144.7931,
      "train_tokens_per_second": 1391.276
    },
    {
      "epoch": 1.4843918191603875,
      "grad_norm": 4.015010833740234,
      "learning_rate": 5.095705759006311e-06,
      "loss": 0.0942,
      "num_input_tokens_seen": 3028192,
      "step": 690,
      "train_runtime": 2176.0229,
      "train_tokens_per_second": 1391.618
    },
    {
      "epoch": 1.5059203444564049,
      "grad_norm": 5.474874973297119,
      "learning_rate": 4.9831097490394195e-06,
      "loss": 0.1204,
      "num_input_tokens_seen": 3071392,
      "step": 700,
      "train_runtime": 2206.8218,
      "train_tokens_per_second": 1391.772
    },
    {
      "epoch": 1.527448869752422,
      "grad_norm": 2.4211018085479736,
      "learning_rate": 4.870522304895164e-06,
      "loss": 0.1358,
      "num_input_tokens_seen": 3114496,
      "step": 710,
      "train_runtime": 2237.5127,
      "train_tokens_per_second": 1391.946
    },
    {
      "epoch": 1.5489773950484391,
      "grad_norm": 2.5446665287017822,
      "learning_rate": 4.758000524841632e-06,
      "loss": 0.1313,
      "num_input_tokens_seen": 3158432,
      "step": 720,
      "train_runtime": 2268.6496,
      "train_tokens_per_second": 1392.208
    },
    {
      "epoch": 1.5705059203444565,
      "grad_norm": 5.643487453460693,
      "learning_rate": 4.645601473845636e-06,
      "loss": 0.0879,
      "num_input_tokens_seen": 3201888,
      "step": 730,
      "train_runtime": 2299.4823,
      "train_tokens_per_second": 1392.439
    },
    {
      "epoch": 1.5920344456404736,
      "grad_norm": 1.8654100894927979,
      "learning_rate": 4.533382154632442e-06,
      "loss": 0.0977,
      "num_input_tokens_seen": 3245856,
      "step": 740,
      "train_runtime": 2330.5679,
      "train_tokens_per_second": 1392.732
    },
    {
      "epoch": 1.6135629709364907,
      "grad_norm": 2.2264926433563232,
      "learning_rate": 4.421399478777064e-06,
      "loss": 0.1483,
      "num_input_tokens_seen": 3290208,
      "step": 750,
      "train_runtime": 2361.9169,
      "train_tokens_per_second": 1393.024
    },
    {
      "epoch": 1.635091496232508,
      "grad_norm": 2.8280999660491943,
      "learning_rate": 4.3097102378417985e-06,
      "loss": 0.1285,
      "num_input_tokens_seen": 3333152,
      "step": 760,
      "train_runtime": 2392.5926,
      "train_tokens_per_second": 1393.113
    },
    {
      "epoch": 1.6566200215285254,
      "grad_norm": 3.5213840007781982,
      "learning_rate": 4.198371074574597e-06,
      "loss": 0.1475,
      "num_input_tokens_seen": 3377664,
      "step": 770,
      "train_runtime": 2423.9491,
      "train_tokens_per_second": 1393.455
    },
    {
      "epoch": 1.6781485468245425,
      "grad_norm": 3.323622226715088,
      "learning_rate": 4.087438454182942e-06,
      "loss": 0.0904,
      "num_input_tokens_seen": 3422400,
      "step": 780,
      "train_runtime": 2455.5818,
      "train_tokens_per_second": 1393.723
    },
    {
      "epoch": 1.6996770721205596,
      "grad_norm": 4.368185997009277,
      "learning_rate": 3.976968635697732e-06,
      "loss": 0.1199,
      "num_input_tokens_seen": 3465760,
      "step": 790,
      "train_runtime": 2486.3362,
      "train_tokens_per_second": 1393.922
    },
    {
      "epoch": 1.721205597416577,
      "grad_norm": 3.0081822872161865,
      "learning_rate": 3.867017643441746e-06,
      "loss": 0.102,
      "num_input_tokens_seen": 3509760,
      "step": 800,
      "train_runtime": 2517.434,
      "train_tokens_per_second": 1394.182
    },
    {
      "epoch": 1.7427341227125943,
      "grad_norm": 3.7257721424102783,
      "learning_rate": 3.757641238617137e-06,
      "loss": 0.1194,
      "num_input_tokens_seen": 3554560,
      "step": 810,
      "train_runtime": 2549.4732,
      "train_tokens_per_second": 1394.233
    },
    {
      "epoch": 1.7642626480086114,
      "grad_norm": 3.8365535736083984,
      "learning_rate": 3.648894891026358e-06,
      "loss": 0.1507,
      "num_input_tokens_seen": 3599488,
      "step": 820,
      "train_runtime": 2581.0241,
      "train_tokens_per_second": 1394.597
    },
    {
      "epoch": 1.7857911733046286,
      "grad_norm": 2.1309561729431152,
      "learning_rate": 3.5408337509408764e-06,
      "loss": 0.1015,
      "num_input_tokens_seen": 3643680,
      "step": 830,
      "train_runtime": 2612.3216,
      "train_tokens_per_second": 1394.805
    },
    {
      "epoch": 1.807319698600646,
      "grad_norm": 2.5082457065582275,
      "learning_rate": 3.4335126211319412e-06,
      "loss": 0.1165,
      "num_input_tokens_seen": 3688160,
      "step": 840,
      "train_runtime": 2643.6717,
      "train_tokens_per_second": 1395.09
    },
    {
      "epoch": 1.8288482238966632,
      "grad_norm": 1.4419660568237305,
      "learning_rate": 3.32698592907757e-06,
      "loss": 0.0958,
      "num_input_tokens_seen": 3731392,
      "step": 850,
      "train_runtime": 2674.4286,
      "train_tokens_per_second": 1395.211
    },
    {
      "epoch": 1.8503767491926801,
      "grad_norm": 2.402513027191162,
      "learning_rate": 3.2213076993598857e-06,
      "loss": 0.1505,
      "num_input_tokens_seen": 3776128,
      "step": 860,
      "train_runtime": 2706.116,
      "train_tokens_per_second": 1395.405
    },
    {
      "epoch": 1.8719052744886975,
      "grad_norm": 0.9779609441757202,
      "learning_rate": 3.1165315262667535e-06,
      "loss": 0.1003,
      "num_input_tokens_seen": 3820896,
      "step": 870,
      "train_runtime": 2737.6181,
      "train_tokens_per_second": 1395.701
    },
    {
      "epoch": 1.8934337997847148,
      "grad_norm": 3.727255344390869,
      "learning_rate": 3.012710546611659e-06,
      "loss": 0.1483,
      "num_input_tokens_seen": 3864704,
      "step": 880,
      "train_runtime": 2768.6885,
      "train_tokens_per_second": 1395.861
    },
    {
      "epoch": 1.914962325080732,
      "grad_norm": 1.5679094791412354,
      "learning_rate": 2.9098974127856e-06,
      "loss": 0.117,
      "num_input_tokens_seen": 3908544,
      "step": 890,
      "train_runtime": 2799.6713,
      "train_tokens_per_second": 1396.072
    },
    {
      "epoch": 1.936490850376749,
      "grad_norm": 2.775408983230591,
      "learning_rate": 2.8081442660546126e-06,
      "loss": 0.1303,
      "num_input_tokens_seen": 3952576,
      "step": 900,
      "train_runtime": 2830.9247,
      "train_tokens_per_second": 1396.214
    },
    {
      "epoch": 1.9580193756727664,
      "grad_norm": 1.0387197732925415,
      "learning_rate": 2.7075027101165706e-06,
      "loss": 0.0971,
      "num_input_tokens_seen": 3996416,
      "step": 910,
      "train_runtime": 2862.0695,
      "train_tokens_per_second": 1396.338
    },
    {
      "epoch": 1.9795479009687837,
      "grad_norm": 3.794166326522827,
      "learning_rate": 2.6080237849305467e-06,
      "loss": 0.132,
      "num_input_tokens_seen": 4040736,
      "step": 920,
      "train_runtime": 2893.2903,
      "train_tokens_per_second": 1396.589
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.547866702079773,
      "learning_rate": 2.5097579408321264e-06,
      "loss": 0.093,
      "num_input_tokens_seen": 4083200,
      "step": 930,
      "train_runtime": 2923.2158,
      "train_tokens_per_second": 1396.818
    },
    {
      "epoch": 2.0215285252960173,
      "grad_norm": 3.5177085399627686,
      "learning_rate": 2.4127550129477145e-06,
      "loss": 0.1324,
      "num_input_tokens_seen": 4127040,
      "step": 940,
      "train_runtime": 2954.2784,
      "train_tokens_per_second": 1396.971
    },
    {
      "epoch": 2.0430570505920342,
      "grad_norm": 1.8717275857925415,
      "learning_rate": 2.317064195920852e-06,
      "loss": 0.0841,
      "num_input_tokens_seen": 4170816,
      "step": 950,
      "train_runtime": 2985.3333,
      "train_tokens_per_second": 1397.102
    },
    {
      "epoch": 2.0645855758880516,
      "grad_norm": 1.233929991722107,
      "learning_rate": 2.2227340189633508e-06,
      "loss": 0.1138,
      "num_input_tokens_seen": 4214272,
      "step": 960,
      "train_runtime": 3016.1559,
      "train_tokens_per_second": 1397.233
    },
    {
      "epoch": 2.086114101184069,
      "grad_norm": 1.6053682565689087,
      "learning_rate": 2.1298123212439028e-06,
      "loss": 0.0892,
      "num_input_tokens_seen": 4258592,
      "step": 970,
      "train_runtime": 3047.3824,
      "train_tokens_per_second": 1397.459
    },
    {
      "epoch": 2.1076426264800863,
      "grad_norm": 2.4293172359466553,
      "learning_rate": 2.0383462276266347e-06,
      "loss": 0.1277,
      "num_input_tokens_seen": 4302656,
      "step": 980,
      "train_runtime": 3078.5923,
      "train_tokens_per_second": 1397.605
    },
    {
      "epoch": 2.129171151776103,
      "grad_norm": 2.0637197494506836,
      "learning_rate": 1.948382124771927e-06,
      "loss": 0.0968,
      "num_input_tokens_seen": 4345888,
      "step": 990,
      "train_runtime": 3109.3083,
      "train_tokens_per_second": 1397.703
    },
    {
      "epoch": 2.1506996770721205,
      "grad_norm": 3.5659446716308594,
      "learning_rate": 1.8599656376116026e-06,
      "loss": 0.1226,
      "num_input_tokens_seen": 4390528,
      "step": 1000,
      "train_runtime": 3140.7259,
      "train_tokens_per_second": 1397.934
    },
    {
      "epoch": 2.172228202368138,
      "grad_norm": 3.0801503658294678,
      "learning_rate": 1.773141606210431e-06,
      "loss": 0.1201,
      "num_input_tokens_seen": 4434784,
      "step": 1010,
      "train_runtime": 3178.3674,
      "train_tokens_per_second": 1395.303
    },
    {
      "epoch": 2.193756727664155,
      "grad_norm": 2.0641636848449707,
      "learning_rate": 1.687954063025663e-06,
      "loss": 0.0966,
      "num_input_tokens_seen": 4478976,
      "step": 1020,
      "train_runtime": 3209.6721,
      "train_tokens_per_second": 1395.462
    },
    {
      "epoch": 2.215285252960172,
      "grad_norm": 2.951422929763794,
      "learning_rate": 1.604446210576157e-06,
      "loss": 0.1143,
      "num_input_tokens_seen": 4523616,
      "step": 1030,
      "train_runtime": 3241.2178,
      "train_tokens_per_second": 1395.653
    },
    {
      "epoch": 2.2368137782561894,
      "grad_norm": 5.046944618225098,
      "learning_rate": 1.5226603995323897e-06,
      "loss": 0.1114,
      "num_input_tokens_seen": 4567264,
      "step": 1040,
      "train_runtime": 3272.1686,
      "train_tokens_per_second": 1395.791
    },
    {
      "epoch": 2.2583423035522068,
      "grad_norm": 1.3710857629776,
      "learning_rate": 1.4426381072384866e-06,
      "loss": 0.0981,
      "num_input_tokens_seen": 4611104,
      "step": 1050,
      "train_runtime": 3303.3652,
      "train_tokens_per_second": 1395.881
    },
    {
      "epoch": 2.2798708288482237,
      "grad_norm": 2.5878543853759766,
      "learning_rate": 1.3644199166771531e-06,
      "loss": 0.1135,
      "num_input_tokens_seen": 4655040,
      "step": 1060,
      "train_runtime": 3334.5323,
      "train_tokens_per_second": 1396.01
    },
    {
      "epoch": 2.301399354144241,
      "grad_norm": 2.481158494949341,
      "learning_rate": 1.2880454958881794e-06,
      "loss": 0.1081,
      "num_input_tokens_seen": 4698432,
      "step": 1070,
      "train_runtime": 3365.4557,
      "train_tokens_per_second": 1396.076
    },
    {
      "epoch": 2.3229278794402584,
      "grad_norm": 1.2142502069473267,
      "learning_rate": 1.2135535778509545e-06,
      "loss": 0.0685,
      "num_input_tokens_seen": 4742848,
      "step": 1080,
      "train_runtime": 3396.7938,
      "train_tokens_per_second": 1396.272
    },
    {
      "epoch": 2.3444564047362757,
      "grad_norm": 3.040208339691162,
      "learning_rate": 1.1409819408411898e-06,
      "loss": 0.0857,
      "num_input_tokens_seen": 4786944,
      "step": 1090,
      "train_runtime": 3427.8943,
      "train_tokens_per_second": 1396.468
    },
    {
      "epoch": 2.3659849300322926,
      "grad_norm": 1.4514607191085815,
      "learning_rate": 1.070367389271823e-06,
      "loss": 0.1008,
      "num_input_tokens_seen": 4830624,
      "step": 1100,
      "train_runtime": 3458.8433,
      "train_tokens_per_second": 1396.601
    },
    {
      "epoch": 2.38751345532831,
      "grad_norm": 3.533973455429077,
      "learning_rate": 1.001745735027801e-06,
      "loss": 0.1137,
      "num_input_tokens_seen": 4874944,
      "step": 1110,
      "train_runtime": 3490.3116,
      "train_tokens_per_second": 1396.707
    },
    {
      "epoch": 2.4090419806243273,
      "grad_norm": 4.195890426635742,
      "learning_rate": 9.351517793042408e-07,
      "loss": 0.1038,
      "num_input_tokens_seen": 4918496,
      "step": 1120,
      "train_runtime": 3521.354,
      "train_tokens_per_second": 1396.763
    },
    {
      "epoch": 2.4305705059203446,
      "grad_norm": 4.139116287231445,
      "learning_rate": 8.706192949571262e-07,
      "loss": 0.1194,
      "num_input_tokens_seen": 4961920,
      "step": 1130,
      "train_runtime": 3552.2441,
      "train_tokens_per_second": 1396.841
    },
    {
      "epoch": 2.4520990312163615,
      "grad_norm": 3.0998311042785645,
      "learning_rate": 8.081810093755537e-07,
      "loss": 0.1161,
      "num_input_tokens_seen": 5005440,
      "step": 1140,
      "train_runtime": 3583.1755,
      "train_tokens_per_second": 1396.928
    },
    {
      "epoch": 2.473627556512379,
      "grad_norm": 2.2539584636688232,
      "learning_rate": 7.478685878841629e-07,
      "loss": 0.1,
      "num_input_tokens_seen": 5049344,
      "step": 1150,
      "train_runtime": 3614.3079,
      "train_tokens_per_second": 1397.043
    },
    {
      "epoch": 2.495156081808396,
      "grad_norm": 3.1125545501708984,
      "learning_rate": 6.897126176841978e-07,
      "loss": 0.0996,
      "num_input_tokens_seen": 5092896,
      "step": 1160,
      "train_runtime": 3645.2176,
      "train_tokens_per_second": 1397.145
    },
    {
      "epoch": 2.5166846071044136,
      "grad_norm": 3.5130155086517334,
      "learning_rate": 6.337425923413276e-07,
      "loss": 0.0944,
      "num_input_tokens_seen": 5136928,
      "step": 1170,
      "train_runtime": 3676.3104,
      "train_tokens_per_second": 1397.305
    },
    {
      "epoch": 2.5382131324004304,
      "grad_norm": 2.059572696685791,
      "learning_rate": 5.799868968281075e-07,
      "loss": 0.08,
      "num_input_tokens_seen": 5180960,
      "step": 1180,
      "train_runtime": 3707.5494,
      "train_tokens_per_second": 1397.408
    },
    {
      "epoch": 2.559741657696448,
      "grad_norm": 4.341704845428467,
      "learning_rate": 5.284727931286526e-07,
      "loss": 0.0962,
      "num_input_tokens_seen": 5225376,
      "step": 1190,
      "train_runtime": 3738.9161,
      "train_tokens_per_second": 1397.564
    },
    {
      "epoch": 2.581270182992465,
      "grad_norm": 2.8424837589263916,
      "learning_rate": 4.792264064128327e-07,
      "loss": 0.094,
      "num_input_tokens_seen": 5268992,
      "step": 1200,
      "train_runtime": 3769.9587,
      "train_tokens_per_second": 1397.626
    },
    {
      "epoch": 2.602798708288482,
      "grad_norm": 3.76309871673584,
      "learning_rate": 4.322727117869951e-07,
      "loss": 0.1005,
      "num_input_tokens_seen": 5312992,
      "step": 1210,
      "train_runtime": 3801.0547,
      "train_tokens_per_second": 1397.768
    },
    {
      "epoch": 2.6243272335844994,
      "grad_norm": 2.470759153366089,
      "learning_rate": 3.8763552162794983e-07,
      "loss": 0.1167,
      "num_input_tokens_seen": 5356448,
      "step": 1220,
      "train_runtime": 3831.9201,
      "train_tokens_per_second": 1397.85
    },
    {
      "epoch": 2.6458557588805167,
      "grad_norm": 2.948512315750122,
      "learning_rate": 3.453374735066034e-07,
      "loss": 0.0907,
      "num_input_tokens_seen": 5400672,
      "step": 1230,
      "train_runtime": 3863.178,
      "train_tokens_per_second": 1397.987
    },
    {
      "epoch": 2.667384284176534,
      "grad_norm": 2.082956552505493,
      "learning_rate": 3.054000187074224e-07,
      "loss": 0.1074,
      "num_input_tokens_seen": 5444576,
      "step": 1240,
      "train_runtime": 3894.1783,
      "train_tokens_per_second": 1398.132
    },
    {
      "epoch": 2.6889128094725514,
      "grad_norm": 2.035034656524658,
      "learning_rate": 2.678434113494882e-07,
      "loss": 0.1128,
      "num_input_tokens_seen": 5488160,
      "step": 1250,
      "train_runtime": 3925.0458,
      "train_tokens_per_second": 1398.241
    },
    {
      "epoch": 2.7104413347685683,
      "grad_norm": 3.7168209552764893,
      "learning_rate": 2.326866981147091e-07,
      "loss": 0.1016,
      "num_input_tokens_seen": 5532000,
      "step": 1260,
      "train_runtime": 3956.0804,
      "train_tokens_per_second": 1398.354
    },
    {
      "epoch": 2.7319698600645856,
      "grad_norm": 4.349425315856934,
      "learning_rate": 1.999477085883711e-07,
      "loss": 0.1052,
      "num_input_tokens_seen": 5575808,
      "step": 1270,
      "train_runtime": 3987.1129,
      "train_tokens_per_second": 1398.458
    },
    {
      "epoch": 2.7534983853606025,
      "grad_norm": 1.8204060792922974,
      "learning_rate": 1.6964304621693516e-07,
      "loss": 0.0918,
      "num_input_tokens_seen": 5619200,
      "step": 1280,
      "train_runtime": 4017.8653,
      "train_tokens_per_second": 1398.554
    },
    {
      "epoch": 2.77502691065662,
      "grad_norm": 3.752577066421509,
      "learning_rate": 1.4178807988766419e-07,
      "loss": 0.109,
      "num_input_tokens_seen": 5662656,
      "step": 1290,
      "train_runtime": 4048.795,
      "train_tokens_per_second": 1398.603
    },
    {
      "epoch": 2.7965554359526372,
      "grad_norm": 3.749866247177124,
      "learning_rate": 1.1639693613435921e-07,
      "loss": 0.1224,
      "num_input_tokens_seen": 5706656,
      "step": 1300,
      "train_runtime": 4079.9279,
      "train_tokens_per_second": 1398.715
    },
    {
      "epoch": 2.8180839612486546,
      "grad_norm": 2.628767251968384,
      "learning_rate": 9.348249197313918e-08,
      "loss": 0.0992,
      "num_input_tokens_seen": 5751232,
      "step": 1310,
      "train_runtime": 4111.4798,
      "train_tokens_per_second": 1398.823
    },
    {
      "epoch": 2.839612486544672,
      "grad_norm": 0.7992174625396729,
      "learning_rate": 7.305636837191876e-08,
      "loss": 0.107,
      "num_input_tokens_seen": 5794976,
      "step": 1320,
      "train_runtime": 4142.503,
      "train_tokens_per_second": 1398.907
    },
    {
      "epoch": 2.861141011840689,
      "grad_norm": 3.0042312145233154,
      "learning_rate": 5.512892435687645e-08,
      "loss": 0.0944,
      "num_input_tokens_seen": 5838368,
      "step": 1330,
      "train_runtime": 4173.35,
      "train_tokens_per_second": 1398.964
    },
    {
      "epoch": 2.882669537136706,
      "grad_norm": 3.208315134048462,
      "learning_rate": 3.970925175892093e-08,
      "loss": 0.0867,
      "num_input_tokens_seen": 5881984,
      "step": 1340,
      "train_runtime": 4204.3268,
      "train_tokens_per_second": 1399.031
    },
    {
      "epoch": 2.9041980624327235,
      "grad_norm": 3.5772690773010254,
      "learning_rate": 2.6805170602803297e-08,
      "loss": 0.1075,
      "num_input_tokens_seen": 5925664,
      "step": 1350,
      "train_runtime": 4235.1869,
      "train_tokens_per_second": 1399.151
    },
    {
      "epoch": 2.9257265877287404,
      "grad_norm": 3.261751413345337,
      "learning_rate": 1.6423225141223854e-08,
      "loss": 0.1042,
      "num_input_tokens_seen": 5969408,
      "step": 1360,
      "train_runtime": 4266.2069,
      "train_tokens_per_second": 1399.231
    },
    {
      "epoch": 2.9472551130247577,
      "grad_norm": 4.744147300720215,
      "learning_rate": 8.568680535939177e-09,
      "loss": 0.0964,
      "num_input_tokens_seen": 6012576,
      "step": 1370,
      "train_runtime": 4297.0229,
      "train_tokens_per_second": 1399.242
    },
    {
      "epoch": 2.968783638320775,
      "grad_norm": 2.3966517448425293,
      "learning_rate": 3.2455201875586372e-09,
      "loss": 0.0922,
      "num_input_tokens_seen": 6056000,
      "step": 1380,
      "train_runtime": 4328.462,
      "train_tokens_per_second": 1399.111
    },
    {
      "epoch": 2.9903121636167924,
      "grad_norm": 1.2950575351715088,
      "learning_rate": 4.5644371537756363e-10,
      "loss": 0.115,
      "num_input_tokens_seen": 6100192,
      "step": 1390,
      "train_runtime": 4359.7222,
      "train_tokens_per_second": 1399.216
    },
    {
      "epoch": 3.0,
      "num_input_tokens_seen": 6120032,
      "step": 1395,
      "total_flos": 4.115769119160883e+17,
      "train_loss": 0.1491297289889346,
      "train_runtime": 4379.932,
      "train_samples_per_second": 2.545,
      "train_steps_per_second": 0.318
    }
  ],
  "logging_steps": 10,
  "max_steps": 1395,
  "num_input_tokens_seen": 6120032,
  "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": 4.115769119160883e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}