{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 8.807339449541285,
  "eval_steps": 500,
  "global_step": 1080,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.08154943934760449,
      "grad_norm": 12.287543296813965,
      "learning_rate": 9.259259259259259e-07,
      "loss": 1.1833,
      "step": 10
    },
    {
      "epoch": 0.16309887869520898,
      "grad_norm": 11.78052043914795,
      "learning_rate": 1.8518518518518519e-06,
      "loss": 1.0856,
      "step": 20
    },
    {
      "epoch": 0.24464831804281345,
      "grad_norm": 15.526741981506348,
      "learning_rate": 2.7777777777777783e-06,
      "loss": 1.0896,
      "step": 30
    },
    {
      "epoch": 0.32619775739041795,
      "grad_norm": 8.213525772094727,
      "learning_rate": 3.7037037037037037e-06,
      "loss": 0.9945,
      "step": 40
    },
    {
      "epoch": 0.4077471967380224,
      "grad_norm": 12.257412910461426,
      "learning_rate": 4.62962962962963e-06,
      "loss": 1.0562,
      "step": 50
    },
    {
      "epoch": 0.4892966360856269,
      "grad_norm": 8.546279907226562,
      "learning_rate": 5.555555555555557e-06,
      "loss": 1.0866,
      "step": 60
    },
    {
      "epoch": 0.5708460754332314,
      "grad_norm": 13.490253448486328,
      "learning_rate": 6.481481481481482e-06,
      "loss": 1.151,
      "step": 70
    },
    {
      "epoch": 0.6523955147808359,
      "grad_norm": 19.235937118530273,
      "learning_rate": 7.4074074074074075e-06,
      "loss": 0.9591,
      "step": 80
    },
    {
      "epoch": 0.7339449541284404,
      "grad_norm": 6.614363193511963,
      "learning_rate": 8.333333333333334e-06,
      "loss": 1.1466,
      "step": 90
    },
    {
      "epoch": 0.8154943934760448,
      "grad_norm": 9.026179313659668,
      "learning_rate": 9.25925925925926e-06,
      "loss": 1.0284,
      "step": 100
    },
    {
      "epoch": 0.8970438328236493,
      "grad_norm": 10.798940658569336,
      "learning_rate": 9.999895536228031e-06,
      "loss": 1.2601,
      "step": 110
    },
    {
      "epoch": 0.9785932721712538,
      "grad_norm": 9.451533317565918,
      "learning_rate": 9.996239762521152e-06,
      "loss": 1.1091,
      "step": 120
    },
    {
      "epoch": 1.0601427115188584,
      "grad_norm": 6.230576992034912,
      "learning_rate": 9.987365164467767e-06,
      "loss": 0.9699,
      "step": 130
    },
    {
      "epoch": 1.1416921508664628,
      "grad_norm": 7.345333099365234,
      "learning_rate": 9.973281012033009e-06,
      "loss": 0.8974,
      "step": 140
    },
    {
      "epoch": 1.2232415902140672,
      "grad_norm": 8.226116180419922,
      "learning_rate": 9.954002016824226e-06,
      "loss": 0.9152,
      "step": 150
    },
    {
      "epoch": 1.3047910295616718,
      "grad_norm": 5.226821422576904,
      "learning_rate": 9.929548316723983e-06,
      "loss": 0.8205,
      "step": 160
    },
    {
      "epoch": 1.3863404689092762,
      "grad_norm": 14.963723182678223,
      "learning_rate": 9.899945454855007e-06,
      "loss": 0.852,
      "step": 170
    },
    {
      "epoch": 1.4678899082568808,
      "grad_norm": 6.257207870483398,
      "learning_rate": 9.86522435289912e-06,
      "loss": 0.7172,
      "step": 180
    },
    {
      "epoch": 1.5494393476044852,
      "grad_norm": 5.380673885345459,
      "learning_rate": 9.825421278797984e-06,
      "loss": 0.7586,
      "step": 190
    },
    {
      "epoch": 1.6309887869520896,
      "grad_norm": 6.626143455505371,
      "learning_rate": 9.7805778088694e-06,
      "loss": 0.802,
      "step": 200
    },
    {
      "epoch": 1.7125382262996942,
      "grad_norm": 5.537201881408691,
      "learning_rate": 9.730740784378755e-06,
      "loss": 0.8788,
      "step": 210
    },
    {
      "epoch": 1.7940876656472988,
      "grad_norm": 14.645645141601562,
      "learning_rate": 9.67596226261095e-06,
      "loss": 0.8143,
      "step": 220
    },
    {
      "epoch": 1.8756371049949032,
      "grad_norm": 5.771991729736328,
      "learning_rate": 9.616299462493952e-06,
      "loss": 0.8228,
      "step": 230
    },
    {
      "epoch": 1.9571865443425076,
      "grad_norm": 7.7364091873168945,
      "learning_rate": 9.551814704830734e-06,
      "loss": 0.7194,
      "step": 240
    },
    {
      "epoch": 2.038735983690112,
      "grad_norm": 7.201632499694824,
      "learning_rate": 9.482575347202047e-06,
      "loss": 0.6792,
      "step": 250
    },
    {
      "epoch": 2.120285423037717,
      "grad_norm": 10.6597318649292,
      "learning_rate": 9.40865371360804e-06,
      "loss": 0.5799,
      "step": 260
    },
    {
      "epoch": 2.2018348623853212,
      "grad_norm": 5.66484260559082,
      "learning_rate": 9.330127018922195e-06,
      "loss": 0.5307,
      "step": 270
    },
    {
      "epoch": 2.2833843017329256,
      "grad_norm": 11.373878479003906,
      "learning_rate": 9.247077288236488e-06,
      "loss": 0.5052,
      "step": 280
    },
    {
      "epoch": 2.36493374108053,
      "grad_norm": 7.9306864738464355,
      "learning_rate": 9.159591271182058e-06,
      "loss": 0.4944,
      "step": 290
    },
    {
      "epoch": 2.4464831804281344,
      "grad_norm": 4.888540267944336,
      "learning_rate": 9.067760351314838e-06,
      "loss": 0.5803,
      "step": 300
    },
    {
      "epoch": 2.528032619775739,
      "grad_norm": 5.445197582244873,
      "learning_rate": 8.97168045066082e-06,
      "loss": 0.4859,
      "step": 310
    },
    {
      "epoch": 2.6095820591233436,
      "grad_norm": 6.764598369598389,
      "learning_rate": 8.871451929520662e-06,
      "loss": 0.5261,
      "step": 320
    },
    {
      "epoch": 2.691131498470948,
      "grad_norm": 4.372009754180908,
      "learning_rate": 8.767179481638303e-06,
      "loss": 0.5705,
      "step": 330
    },
    {
      "epoch": 2.7726809378185524,
      "grad_norm": 5.150655746459961,
      "learning_rate": 8.658972024843063e-06,
      "loss": 0.5248,
      "step": 340
    },
    {
      "epoch": 2.8542303771661572,
      "grad_norm": 7.280877590179443,
      "learning_rate": 8.546942587279465e-06,
      "loss": 0.4764,
      "step": 350
    },
    {
      "epoch": 2.9357798165137616,
      "grad_norm": 6.407410621643066,
      "learning_rate": 8.43120818934367e-06,
      "loss": 0.5481,
      "step": 360
    },
    {
      "epoch": 3.017329255861366,
      "grad_norm": 3.3497812747955322,
      "learning_rate": 8.31188972144974e-06,
      "loss": 0.4215,
      "step": 370
    },
    {
      "epoch": 3.0988786952089704,
      "grad_norm": 8.869242668151855,
      "learning_rate": 8.18911181775353e-06,
      "loss": 0.3038,
      "step": 380
    },
    {
      "epoch": 3.180428134556575,
      "grad_norm": 7.462818622589111,
      "learning_rate": 8.063002725966014e-06,
      "loss": 0.3173,
      "step": 390
    },
    {
      "epoch": 3.261977573904179,
      "grad_norm": 4.446223258972168,
      "learning_rate": 7.93369417339209e-06,
      "loss": 0.3345,
      "step": 400
    },
    {
      "epoch": 3.343527013251784,
      "grad_norm": 4.583369255065918,
      "learning_rate": 7.801321229334764e-06,
      "loss": 0.3165,
      "step": 410
    },
    {
      "epoch": 3.4250764525993884,
      "grad_norm": 3.839564323425293,
      "learning_rate": 7.666022164008458e-06,
      "loss": 0.2921,
      "step": 420
    },
    {
      "epoch": 3.506625891946993,
      "grad_norm": 3.111229181289673,
      "learning_rate": 7.527938304108795e-06,
      "loss": 0.3062,
      "step": 430
    },
    {
      "epoch": 3.588175331294597,
      "grad_norm": 4.569026470184326,
      "learning_rate": 7.387213885189746e-06,
      "loss": 0.2924,
      "step": 440
    },
    {
      "epoch": 3.669724770642202,
      "grad_norm": 6.843806266784668,
      "learning_rate": 7.243995901002312e-06,
      "loss": 0.2929,
      "step": 450
    },
    {
      "epoch": 3.7512742099898064,
      "grad_norm": 3.574108839035034,
      "learning_rate": 7.098433949952146e-06,
      "loss": 0.3544,
      "step": 460
    },
    {
      "epoch": 3.832823649337411,
      "grad_norm": 3.691032886505127,
      "learning_rate": 6.950680078836475e-06,
      "loss": 0.3103,
      "step": 470
    },
    {
      "epoch": 3.914373088685015,
      "grad_norm": 3.9971020221710205,
      "learning_rate": 6.800888624023552e-06,
      "loss": 0.2742,
      "step": 480
    },
    {
      "epoch": 3.9959225280326196,
      "grad_norm": 2.556887149810791,
      "learning_rate": 6.649216050240539e-06,
      "loss": 0.3339,
      "step": 490
    },
    {
      "epoch": 4.077471967380224,
      "grad_norm": 2.1896212100982666,
      "learning_rate": 6.495820787138209e-06,
      "loss": 0.193,
      "step": 500
    },
    {
      "epoch": 4.077471967380224,
      "eval_loss": 1.6843377351760864,
      "eval_runtime": 4.0486,
      "eval_samples_per_second": 27.17,
      "eval_steps_per_second": 27.17,
      "step": 500
    },
    {
      "epoch": 4.159021406727828,
      "grad_norm": 4.995784282684326,
      "learning_rate": 6.340863063803187e-06,
      "loss": 0.2102,
      "step": 510
    },
    {
      "epoch": 4.240570846075434,
      "grad_norm": 3.022165536880493,
      "learning_rate": 6.184504741390596e-06,
      "loss": 0.1799,
      "step": 520
    },
    {
      "epoch": 4.322120285423038,
      "grad_norm": 3.73397159576416,
      "learning_rate": 6.02690914405191e-06,
      "loss": 0.1945,
      "step": 530
    },
    {
      "epoch": 4.4036697247706424,
      "grad_norm": 3.1206557750701904,
      "learning_rate": 5.8682408883346535e-06,
      "loss": 0.1186,
      "step": 540
    },
    {
      "epoch": 4.485219164118247,
      "grad_norm": 6.9254326820373535,
      "learning_rate": 5.708665711232103e-06,
      "loss": 0.1761,
      "step": 550
    },
    {
      "epoch": 4.566768603465851,
      "grad_norm": 1.7940961122512817,
      "learning_rate": 5.548350297062659e-06,
      "loss": 0.1747,
      "step": 560
    },
    {
      "epoch": 4.648318042813456,
      "grad_norm": 3.6487746238708496,
      "learning_rate": 5.387462103359655e-06,
      "loss": 0.205,
      "step": 570
    },
    {
      "epoch": 4.72986748216106,
      "grad_norm": 2.9038984775543213,
      "learning_rate": 5.2261691859535325e-06,
      "loss": 0.1384,
      "step": 580
    },
    {
      "epoch": 4.811416921508664,
      "grad_norm": 7.9458842277526855,
      "learning_rate": 5.064640023429042e-06,
      "loss": 0.1475,
      "step": 590
    },
    {
      "epoch": 4.892966360856269,
      "grad_norm": 3.393327236175537,
      "learning_rate": 4.903043341140879e-06,
      "loss": 0.1647,
      "step": 600
    },
    {
      "epoch": 4.974515800203873,
      "grad_norm": 5.027682781219482,
      "learning_rate": 4.741547934971528e-06,
      "loss": 0.1691,
      "step": 610
    },
    {
      "epoch": 5.0560652395514785,
      "grad_norm": 1.468493938446045,
      "learning_rate": 4.580322495015466e-06,
      "loss": 0.1022,
      "step": 620
    },
    {
      "epoch": 5.137614678899083,
      "grad_norm": 1.9386202096939087,
      "learning_rate": 4.4195354293738484e-06,
      "loss": 0.0677,
      "step": 630
    },
    {
      "epoch": 5.219164118246687,
      "grad_norm": 2.2345311641693115,
      "learning_rate": 4.259354688243758e-06,
      "loss": 0.0761,
      "step": 640
    },
    {
      "epoch": 5.300713557594292,
      "grad_norm": 2.380667209625244,
      "learning_rate": 4.099947588485744e-06,
      "loss": 0.0733,
      "step": 650
    },
    {
      "epoch": 5.382262996941896,
      "grad_norm": 1.2158432006835938,
      "learning_rate": 3.941480638852948e-06,
      "loss": 0.0841,
      "step": 660
    },
    {
      "epoch": 5.4638124362895,
      "grad_norm": 2.382153034210205,
      "learning_rate": 3.784119366064293e-06,
      "loss": 0.0929,
      "step": 670
    },
    {
      "epoch": 5.545361875637105,
      "grad_norm": 2.7093493938446045,
      "learning_rate": 3.6280281419034934e-06,
      "loss": 0.0634,
      "step": 680
    },
    {
      "epoch": 5.626911314984709,
      "grad_norm": 1.597835659980774,
      "learning_rate": 3.473370011524435e-06,
      "loss": 0.1201,
      "step": 690
    },
    {
      "epoch": 5.708460754332314,
      "grad_norm": 2.351442575454712,
      "learning_rate": 3.3203065231422904e-06,
      "loss": 0.0811,
      "step": 700
    },
    {
      "epoch": 5.790010193679919,
      "grad_norm": 2.5688395500183105,
      "learning_rate": 3.1689975592882603e-06,
      "loss": 0.0708,
      "step": 710
    },
    {
      "epoch": 5.871559633027523,
      "grad_norm": 3.0511674880981445,
      "learning_rate": 3.019601169804216e-06,
      "loss": 0.0846,
      "step": 720
    },
    {
      "epoch": 5.953109072375128,
      "grad_norm": 5.8331298828125,
      "learning_rate": 2.8722734067516637e-06,
      "loss": 0.0782,
      "step": 730
    },
    {
      "epoch": 6.034658511722732,
      "grad_norm": 1.1434911489486694,
      "learning_rate": 2.7271681614074973e-06,
      "loss": 0.0624,
      "step": 740
    },
    {
      "epoch": 6.116207951070336,
      "grad_norm": 2.7258777618408203,
      "learning_rate": 2.5844370035168077e-06,
      "loss": 0.0373,
      "step": 750
    },
    {
      "epoch": 6.197757390417941,
      "grad_norm": 0.5002675652503967,
      "learning_rate": 2.4442290229706344e-06,
      "loss": 0.0331,
      "step": 760
    },
    {
      "epoch": 6.279306829765545,
      "grad_norm": 0.7639961242675781,
      "learning_rate": 2.3066906740740626e-06,
      "loss": 0.0232,
      "step": 770
    },
    {
      "epoch": 6.36085626911315,
      "grad_norm": 2.4113121032714844,
      "learning_rate": 2.171965622567308e-06,
      "loss": 0.0356,
      "step": 780
    },
    {
      "epoch": 6.442405708460754,
      "grad_norm": 1.4164212942123413,
      "learning_rate": 2.0401945955596206e-06,
      "loss": 0.0599,
      "step": 790
    },
    {
      "epoch": 6.523955147808358,
      "grad_norm": 1.6531766653060913,
      "learning_rate": 1.9115152345327154e-06,
      "loss": 0.0298,
      "step": 800
    },
    {
      "epoch": 6.605504587155964,
      "grad_norm": 1.0268425941467285,
      "learning_rate": 1.7860619515673034e-06,
      "loss": 0.029,
      "step": 810
    },
    {
      "epoch": 6.687054026503568,
      "grad_norm": 1.7151083946228027,
      "learning_rate": 1.6639657889429017e-06,
      "loss": 0.0322,
      "step": 820
    },
    {
      "epoch": 6.7686034658511725,
      "grad_norm": 1.3095351457595825,
      "learning_rate": 1.5453542822575624e-06,
      "loss": 0.0276,
      "step": 830
    },
    {
      "epoch": 6.850152905198777,
      "grad_norm": 2.322570323944092,
      "learning_rate": 1.4303513272105057e-06,
      "loss": 0.0307,
      "step": 840
    },
    {
      "epoch": 6.931702344546381,
      "grad_norm": 5.283768653869629,
      "learning_rate": 1.3190770501868243e-06,
      "loss": 0.0439,
      "step": 850
    },
    {
      "epoch": 7.013251783893986,
      "grad_norm": 2.021864414215088,
      "learning_rate": 1.2116476827794104e-06,
      "loss": 0.0349,
      "step": 860
    },
    {
      "epoch": 7.09480122324159,
      "grad_norm": 1.7383687496185303,
      "learning_rate": 1.1081754403792e-06,
      "loss": 0.0141,
      "step": 870
    },
    {
      "epoch": 7.176350662589194,
      "grad_norm": 0.9524487853050232,
      "learning_rate": 1.008768404960535e-06,
      "loss": 0.0131,
      "step": 880
    },
    {
      "epoch": 7.257900101936799,
      "grad_norm": 0.6132211089134216,
      "learning_rate": 9.135304121840976e-07,
      "loss": 0.0156,
      "step": 890
    },
    {
      "epoch": 7.339449541284404,
      "grad_norm": 0.16791853308677673,
      "learning_rate": 8.225609429353187e-07,
      "loss": 0.0129,
      "step": 900
    },
    {
      "epoch": 7.4209989806320085,
      "grad_norm": 0.24683955311775208,
      "learning_rate": 7.35955019411585e-07,
      "loss": 0.0142,
      "step": 910
    },
    {
      "epoch": 7.502548419979613,
      "grad_norm": 0.177597314119339,
      "learning_rate": 6.53803105866761e-07,
      "loss": 0.0103,
      "step": 920
    },
    {
      "epoch": 7.584097859327217,
      "grad_norm": 1.1235181093215942,
      "learning_rate": 5.76191014116711e-07,
      "loss": 0.0166,
      "step": 930
    },
    {
      "epoch": 7.665647298674822,
      "grad_norm": 0.7732178568840027,
      "learning_rate": 5.031998139045352e-07,
      "loss": 0.0144,
      "step": 940
    },
    {
      "epoch": 7.747196738022426,
      "grad_norm": 1.0395855903625488,
      "learning_rate": 4.349057482191299e-07,
      "loss": 0.0165,
      "step": 950
    },
    {
      "epoch": 7.82874617737003,
      "grad_norm": 0.6988322734832764,
      "learning_rate": 3.7138015365554834e-07,
      "loss": 0.0118,
      "step": 960
    },
    {
      "epoch": 7.910295616717635,
      "grad_norm": 0.8800845742225647,
      "learning_rate": 3.1268938590032495e-07,
      "loss": 0.0141,
      "step": 970
    },
    {
      "epoch": 7.991845056065239,
      "grad_norm": 0.32935312390327454,
      "learning_rate": 2.5889475041961767e-07,
      "loss": 0.0096,
      "step": 980
    },
    {
      "epoch": 8.073394495412844,
      "grad_norm": 0.8593177199363708,
      "learning_rate": 2.1005243842255552e-07,
      "loss": 0.0207,
      "step": 990
    },
    {
      "epoch": 8.154943934760448,
      "grad_norm": 0.561518132686615,
      "learning_rate": 1.6621346816668993e-07,
      "loss": 0.0083,
      "step": 1000
    },
    {
      "epoch": 8.154943934760448,
      "eval_loss": 2.4749834537506104,
      "eval_runtime": 3.9418,
      "eval_samples_per_second": 27.906,
      "eval_steps_per_second": 27.906,
      "step": 1000
    },
    {
      "epoch": 8.236493374108052,
      "grad_norm": 0.1318514049053192,
      "learning_rate": 1.2742363166685035e-07,
      "loss": 0.0084,
      "step": 1010
    },
    {
      "epoch": 8.318042813455657,
      "grad_norm": 0.09888464212417603,
      "learning_rate": 9.372344686307655e-08,
      "loss": 0.0065,
      "step": 1020
    },
    {
      "epoch": 8.399592252803261,
      "grad_norm": 0.2378782033920288,
      "learning_rate": 6.514811529758747e-08,
      "loss": 0.0116,
      "step": 1030
    },
    {
      "epoch": 8.481141692150867,
      "grad_norm": 0.2287997156381607,
      "learning_rate": 4.172748534499449e-08,
      "loss": 0.0108,
      "step": 1040
    },
    {
      "epoch": 8.562691131498472,
      "grad_norm": 0.5172834992408752,
      "learning_rate": 2.3486021034170857e-08,
      "loss": 0.0088,
      "step": 1050
    },
    {
      "epoch": 8.644240570846076,
      "grad_norm": 0.272684246301651,
      "learning_rate": 1.044277649433989e-08,
      "loss": 0.0086,
      "step": 1060
    },
    {
      "epoch": 8.72579001019368,
      "grad_norm": 0.2911706864833832,
      "learning_rate": 2.611376052073511e-09,
      "loss": 0.0086,
      "step": 1070
    },
    {
      "epoch": 8.807339449541285,
      "grad_norm": 0.5013861060142517,
      "learning_rate": 0.0,
      "loss": 0.0063,
      "step": 1080
    },
    {
      "epoch": 8.807339449541285,
      "step": 1080,
      "total_flos": 1.2056852760477696e+16,
      "train_loss": 0.3479360826589443,
      "train_runtime": 1723.5404,
      "train_samples_per_second": 5.013,
      "train_steps_per_second": 0.627
    }
  ],
  "logging_steps": 10,
  "max_steps": 1080,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.2056852760477696e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}