{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9989550679205852,
  "eval_steps": 500,
  "global_step": 239,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0041797283176593526,
      "grad_norm": 3.793109431652244,
      "learning_rate": 2.5e-06,
      "loss": 1.3773,
      "step": 1
    },
    {
      "epoch": 0.008359456635318705,
      "grad_norm": 3.860176928138404,
      "learning_rate": 5e-06,
      "loss": 1.3697,
      "step": 2
    },
    {
      "epoch": 0.012539184952978056,
      "grad_norm": 3.472574916996383,
      "learning_rate": 7.500000000000001e-06,
      "loss": 1.3685,
      "step": 3
    },
    {
      "epoch": 0.01671891327063741,
      "grad_norm": 2.426017601116389,
      "learning_rate": 1e-05,
      "loss": 1.2106,
      "step": 4
    },
    {
      "epoch": 0.02089864158829676,
      "grad_norm": 2.369903447078516,
      "learning_rate": 1.25e-05,
      "loss": 1.2334,
      "step": 5
    },
    {
      "epoch": 0.025078369905956112,
      "grad_norm": 6.10573157991006,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 1.1659,
      "step": 6
    },
    {
      "epoch": 0.029258098223615466,
      "grad_norm": 2.828089970728912,
      "learning_rate": 1.7500000000000002e-05,
      "loss": 1.1133,
      "step": 7
    },
    {
      "epoch": 0.03343782654127482,
      "grad_norm": 3.7326281687198226,
      "learning_rate": 2e-05,
      "loss": 1.031,
      "step": 8
    },
    {
      "epoch": 0.03761755485893417,
      "grad_norm": 2.8226722768014127,
      "learning_rate": 1.9999075218579184e-05,
      "loss": 1.0627,
      "step": 9
    },
    {
      "epoch": 0.04179728317659352,
      "grad_norm": 2.3941660915552196,
      "learning_rate": 1.9996301045360874e-05,
      "loss": 0.8942,
      "step": 10
    },
    {
      "epoch": 0.04597701149425287,
      "grad_norm": 2.333154808520554,
      "learning_rate": 1.9991677993445832e-05,
      "loss": 0.8475,
      "step": 11
    },
    {
      "epoch": 0.050156739811912224,
      "grad_norm": 2.1073095025076003,
      "learning_rate": 1.9985206917896563e-05,
      "loss": 0.7987,
      "step": 12
    },
    {
      "epoch": 0.054336468129571575,
      "grad_norm": 2.0147413946396506,
      "learning_rate": 1.9976889015579167e-05,
      "loss": 0.7122,
      "step": 13
    },
    {
      "epoch": 0.05851619644723093,
      "grad_norm": 1.7762710304793987,
      "learning_rate": 1.9966725824941933e-05,
      "loss": 0.6907,
      "step": 14
    },
    {
      "epoch": 0.06269592476489028,
      "grad_norm": 1.854790476944044,
      "learning_rate": 1.9954719225730847e-05,
      "loss": 0.583,
      "step": 15
    },
    {
      "epoch": 0.06687565308254964,
      "grad_norm": 1.6431498121799215,
      "learning_rate": 1.994087143864188e-05,
      "loss": 0.5144,
      "step": 16
    },
    {
      "epoch": 0.07105538140020899,
      "grad_norm": 1.5849345385422475,
      "learning_rate": 1.992518502491028e-05,
      "loss": 0.4662,
      "step": 17
    },
    {
      "epoch": 0.07523510971786834,
      "grad_norm": 2.0976691689662488,
      "learning_rate": 1.9907662885836836e-05,
      "loss": 0.5641,
      "step": 18
    },
    {
      "epoch": 0.0794148380355277,
      "grad_norm": 1.8591436531727428,
      "learning_rate": 1.9888308262251286e-05,
      "loss": 0.5513,
      "step": 19
    },
    {
      "epoch": 0.08359456635318704,
      "grad_norm": 1.6077371820716417,
      "learning_rate": 1.986712473391289e-05,
      "loss": 0.514,
      "step": 20
    },
    {
      "epoch": 0.0877742946708464,
      "grad_norm": 1.6369595572017508,
      "learning_rate": 1.9844116218848335e-05,
      "loss": 0.4729,
      "step": 21
    },
    {
      "epoch": 0.09195402298850575,
      "grad_norm": 1.4143725432673173,
      "learning_rate": 1.9819286972627066e-05,
      "loss": 0.4404,
      "step": 22
    },
    {
      "epoch": 0.0961337513061651,
      "grad_norm": 1.9394850939365578,
      "learning_rate": 1.9792641587574212e-05,
      "loss": 0.4529,
      "step": 23
    },
    {
      "epoch": 0.10031347962382445,
      "grad_norm": 1.4086541686064722,
      "learning_rate": 1.9764184991921178e-05,
      "loss": 0.4101,
      "step": 24
    },
    {
      "epoch": 0.1044932079414838,
      "grad_norm": 2.563989330153671,
      "learning_rate": 1.973392244889415e-05,
      "loss": 0.3327,
      "step": 25
    },
    {
      "epoch": 0.10867293625914315,
      "grad_norm": 1.4054095165182117,
      "learning_rate": 1.9701859555740647e-05,
      "loss": 0.3673,
      "step": 26
    },
    {
      "epoch": 0.11285266457680251,
      "grad_norm": 1.3347083701406377,
      "learning_rate": 1.966800224269424e-05,
      "loss": 0.4013,
      "step": 27
    },
    {
      "epoch": 0.11703239289446186,
      "grad_norm": 1.3599751397794146,
      "learning_rate": 1.9632356771877735e-05,
      "loss": 0.3765,
      "step": 28
    },
    {
      "epoch": 0.12121212121212122,
      "grad_norm": 1.2994011828520489,
      "learning_rate": 1.9594929736144978e-05,
      "loss": 0.3092,
      "step": 29
    },
    {
      "epoch": 0.12539184952978055,
      "grad_norm": 1.4307464855692031,
      "learning_rate": 1.955572805786141e-05,
      "loss": 0.352,
      "step": 30
    },
    {
      "epoch": 0.12957157784743992,
      "grad_norm": 1.1156432595068917,
      "learning_rate": 1.9514758987623784e-05,
      "loss": 0.2338,
      "step": 31
    },
    {
      "epoch": 0.13375130616509928,
      "grad_norm": 1.2383396612081838,
      "learning_rate": 1.9472030102919102e-05,
      "loss": 0.3058,
      "step": 32
    },
    {
      "epoch": 0.13793103448275862,
      "grad_norm": 1.3484241033815327,
      "learning_rate": 1.94275493067231e-05,
      "loss": 0.3032,
      "step": 33
    },
    {
      "epoch": 0.14211076280041798,
      "grad_norm": 1.1814763693141177,
      "learning_rate": 1.938132482603856e-05,
      "loss": 0.2606,
      "step": 34
    },
    {
      "epoch": 0.14629049111807732,
      "grad_norm": 1.1152475267807018,
      "learning_rate": 1.9333365210373668e-05,
      "loss": 0.2655,
      "step": 35
    },
    {
      "epoch": 0.15047021943573669,
      "grad_norm": 1.0274673205489633,
      "learning_rate": 1.9283679330160726e-05,
      "loss": 0.2393,
      "step": 36
    },
    {
      "epoch": 0.15464994775339602,
      "grad_norm": 1.1446059182649253,
      "learning_rate": 1.9232276375115517e-05,
      "loss": 0.3053,
      "step": 37
    },
    {
      "epoch": 0.1588296760710554,
      "grad_norm": 1.2083159051790577,
      "learning_rate": 1.9179165852537596e-05,
      "loss": 0.2887,
      "step": 38
    },
    {
      "epoch": 0.16300940438871472,
      "grad_norm": 1.5658760108502938,
      "learning_rate": 1.9124357585551872e-05,
      "loss": 0.3527,
      "step": 39
    },
    {
      "epoch": 0.1671891327063741,
      "grad_norm": 1.3279092492462565,
      "learning_rate": 1.9067861711291744e-05,
      "loss": 0.3117,
      "step": 40
    },
    {
      "epoch": 0.17136886102403343,
      "grad_norm": 1.177122725805587,
      "learning_rate": 1.900968867902419e-05,
      "loss": 0.2835,
      "step": 41
    },
    {
      "epoch": 0.1755485893416928,
      "grad_norm": 1.0091524997429624,
      "learning_rate": 1.89498492482171e-05,
      "loss": 0.2474,
      "step": 42
    },
    {
      "epoch": 0.17972831765935215,
      "grad_norm": 0.993604546498234,
      "learning_rate": 1.8888354486549238e-05,
      "loss": 0.2289,
      "step": 43
    },
    {
      "epoch": 0.1839080459770115,
      "grad_norm": 1.3521657755483074,
      "learning_rate": 1.8825215767863215e-05,
      "loss": 0.3526,
      "step": 44
    },
    {
      "epoch": 0.18808777429467086,
      "grad_norm": 1.1636752310323113,
      "learning_rate": 1.876044477006183e-05,
      "loss": 0.3034,
      "step": 45
    },
    {
      "epoch": 0.1922675026123302,
      "grad_norm": 0.9201202294799887,
      "learning_rate": 1.8694053472948154e-05,
      "loss": 0.2322,
      "step": 46
    },
    {
      "epoch": 0.19644723092998956,
      "grad_norm": 1.0591767636272813,
      "learning_rate": 1.8626054156009807e-05,
      "loss": 0.1968,
      "step": 47
    },
    {
      "epoch": 0.2006269592476489,
      "grad_norm": 1.1802357695207026,
      "learning_rate": 1.8556459396147777e-05,
      "loss": 0.248,
      "step": 48
    },
    {
      "epoch": 0.20480668756530826,
      "grad_norm": 1.05844609547146,
      "learning_rate": 1.8485282065350237e-05,
      "loss": 0.2361,
      "step": 49
    },
    {
      "epoch": 0.2089864158829676,
      "grad_norm": 0.8493872312825937,
      "learning_rate": 1.8412535328311813e-05,
      "loss": 0.163,
      "step": 50
    },
    {
      "epoch": 0.21316614420062696,
      "grad_norm": 1.1938936546063497,
      "learning_rate": 1.8338232639998672e-05,
      "loss": 0.2011,
      "step": 51
    },
    {
      "epoch": 0.2173458725182863,
      "grad_norm": 0.9821596876965665,
      "learning_rate": 1.826238774315995e-05,
      "loss": 0.1797,
      "step": 52
    },
    {
      "epoch": 0.22152560083594566,
      "grad_norm": 1.0890806155634258,
      "learning_rate": 1.8185014665785936e-05,
      "loss": 0.2356,
      "step": 53
    },
    {
      "epoch": 0.22570532915360503,
      "grad_norm": 3.1623969781776875,
      "learning_rate": 1.810612771851352e-05,
      "loss": 0.2341,
      "step": 54
    },
    {
      "epoch": 0.22988505747126436,
      "grad_norm": 1.0495429246657253,
      "learning_rate": 1.8025741491979326e-05,
      "loss": 0.2286,
      "step": 55
    },
    {
      "epoch": 0.23406478578892373,
      "grad_norm": 1.0998745172943698,
      "learning_rate": 1.7943870854121126e-05,
      "loss": 0.228,
      "step": 56
    },
    {
      "epoch": 0.23824451410658307,
      "grad_norm": 1.0611257780697756,
      "learning_rate": 1.7860530947427878e-05,
      "loss": 0.1926,
      "step": 57
    },
    {
      "epoch": 0.24242424242424243,
      "grad_norm": 1.3963883396571588,
      "learning_rate": 1.777573718613904e-05,
      "loss": 0.2455,
      "step": 58
    },
    {
      "epoch": 0.24660397074190177,
      "grad_norm": 0.8817393289325651,
      "learning_rate": 1.768950525339362e-05,
      "loss": 0.186,
      "step": 59
    },
    {
      "epoch": 0.2507836990595611,
      "grad_norm": 1.004709836073224,
      "learning_rate": 1.7601851098329484e-05,
      "loss": 0.2003,
      "step": 60
    },
    {
      "epoch": 0.2549634273772205,
      "grad_norm": 1.009032284413222,
      "learning_rate": 1.7512790933133435e-05,
      "loss": 0.2389,
      "step": 61
    },
    {
      "epoch": 0.25914315569487983,
      "grad_norm": 1.0842648970922986,
      "learning_rate": 1.74223412300427e-05,
      "loss": 0.2324,
      "step": 62
    },
    {
      "epoch": 0.26332288401253917,
      "grad_norm": 1.1362114248016657,
      "learning_rate": 1.7330518718298263e-05,
      "loss": 0.2636,
      "step": 63
    },
    {
      "epoch": 0.26750261233019856,
      "grad_norm": 1.2490219244647938,
      "learning_rate": 1.72373403810507e-05,
      "loss": 0.2392,
      "step": 64
    },
    {
      "epoch": 0.2716823406478579,
      "grad_norm": 0.9229560381350037,
      "learning_rate": 1.7142823452219036e-05,
      "loss": 0.2207,
      "step": 65
    },
    {
      "epoch": 0.27586206896551724,
      "grad_norm": 0.9194163633451438,
      "learning_rate": 1.7046985413303215e-05,
      "loss": 0.2208,
      "step": 66
    },
    {
      "epoch": 0.2800417972831766,
      "grad_norm": 0.9431299303955618,
      "learning_rate": 1.6949843990150798e-05,
      "loss": 0.1877,
      "step": 67
    },
    {
      "epoch": 0.28422152560083597,
      "grad_norm": 1.0695148700733295,
      "learning_rate": 1.6851417149678442e-05,
      "loss": 0.2431,
      "step": 68
    },
    {
      "epoch": 0.2884012539184953,
      "grad_norm": 0.9981918594046101,
      "learning_rate": 1.6751723096548834e-05,
      "loss": 0.2134,
      "step": 69
    },
    {
      "epoch": 0.29258098223615464,
      "grad_norm": 0.7779428741827196,
      "learning_rate": 1.6650780269803587e-05,
      "loss": 0.1772,
      "step": 70
    },
    {
      "epoch": 0.296760710553814,
      "grad_norm": 0.7017283200072758,
      "learning_rate": 1.6548607339452853e-05,
      "loss": 0.154,
      "step": 71
    },
    {
      "epoch": 0.30094043887147337,
      "grad_norm": 1.2657237042635219,
      "learning_rate": 1.644522320302217e-05,
      "loss": 0.1989,
      "step": 72
    },
    {
      "epoch": 0.3051201671891327,
      "grad_norm": 0.9141216726104325,
      "learning_rate": 1.634064698205725e-05,
      "loss": 0.1632,
      "step": 73
    },
    {
      "epoch": 0.30929989550679204,
      "grad_norm": 0.853915517926657,
      "learning_rate": 1.6234898018587336e-05,
      "loss": 0.2127,
      "step": 74
    },
    {
      "epoch": 0.31347962382445144,
      "grad_norm": 1.0323735847419797,
      "learning_rate": 1.612799587154777e-05,
      "loss": 0.2167,
      "step": 75
    },
    {
      "epoch": 0.3176593521421108,
      "grad_norm": 0.930984497217397,
      "learning_rate": 1.6019960313162436e-05,
      "loss": 0.1956,
      "step": 76
    },
    {
      "epoch": 0.3218390804597701,
      "grad_norm": 1.006790474010485,
      "learning_rate": 1.5910811325286768e-05,
      "loss": 0.2224,
      "step": 77
    },
    {
      "epoch": 0.32601880877742945,
      "grad_norm": 0.6839140410559078,
      "learning_rate": 1.5800569095711983e-05,
      "loss": 0.1534,
      "step": 78
    },
    {
      "epoch": 0.33019853709508884,
      "grad_norm": 0.9218639125486777,
      "learning_rate": 1.5689254014431225e-05,
      "loss": 0.2173,
      "step": 79
    },
    {
      "epoch": 0.3343782654127482,
      "grad_norm": 0.9628453893674432,
      "learning_rate": 1.5576886669868297e-05,
      "loss": 0.2144,
      "step": 80
    },
    {
      "epoch": 0.3385579937304075,
      "grad_norm": 1.0976706938526015,
      "learning_rate": 1.5463487845069708e-05,
      "loss": 0.2869,
      "step": 81
    },
    {
      "epoch": 0.34273772204806685,
      "grad_norm": 0.843877293720617,
      "learning_rate": 1.5349078513860728e-05,
      "loss": 0.1644,
      "step": 82
    },
    {
      "epoch": 0.34691745036572624,
      "grad_norm": 1.0395155619197012,
      "learning_rate": 1.5233679836966122e-05,
      "loss": 0.2048,
      "step": 83
    },
    {
      "epoch": 0.3510971786833856,
      "grad_norm": 0.62654878377622,
      "learning_rate": 1.5117313158096371e-05,
      "loss": 0.1011,
      "step": 84
    },
    {
      "epoch": 0.3552769070010449,
      "grad_norm": 0.8765627543230693,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.2282,
      "step": 85
    },
    {
      "epoch": 0.3594566353187043,
      "grad_norm": 0.8493012101164317,
      "learning_rate": 1.4881762060482814e-05,
      "loss": 0.2303,
      "step": 86
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 0.9061911256912499,
      "learning_rate": 1.476262120839475e-05,
      "loss": 0.1772,
      "step": 87
    },
    {
      "epoch": 0.367816091954023,
      "grad_norm": 0.7714054432768603,
      "learning_rate": 1.4642599479585106e-05,
      "loss": 0.1827,
      "step": 88
    },
    {
      "epoch": 0.3719958202716823,
      "grad_norm": 1.0302753021934954,
      "learning_rate": 1.4521719072826858e-05,
      "loss": 0.2556,
      "step": 89
    },
    {
      "epoch": 0.3761755485893417,
      "grad_norm": 0.8170194188293493,
      "learning_rate": 1.4400002345710871e-05,
      "loss": 0.2131,
      "step": 90
    },
    {
      "epoch": 0.38035527690700105,
      "grad_norm": 0.8353229514513874,
      "learning_rate": 1.427747181051071e-05,
      "loss": 0.1774,
      "step": 91
    },
    {
      "epoch": 0.3845350052246604,
      "grad_norm": 0.8996783507158662,
      "learning_rate": 1.4154150130018867e-05,
      "loss": 0.2074,
      "step": 92
    },
    {
      "epoch": 0.3887147335423197,
      "grad_norm": 1.0120263532766836,
      "learning_rate": 1.4030060113355118e-05,
      "loss": 0.2291,
      "step": 93
    },
    {
      "epoch": 0.3928944618599791,
      "grad_norm": 0.8009583936768994,
      "learning_rate": 1.3905224711747844e-05,
      "loss": 0.1793,
      "step": 94
    },
    {
      "epoch": 0.39707419017763845,
      "grad_norm": 0.8764394706115101,
      "learning_rate": 1.3779667014289067e-05,
      "loss": 0.1639,
      "step": 95
    },
    {
      "epoch": 0.4012539184952978,
      "grad_norm": 1.2665682983369955,
      "learning_rate": 1.3653410243663953e-05,
      "loss": 0.2469,
      "step": 96
    },
    {
      "epoch": 0.4054336468129572,
      "grad_norm": 2.02790558792064,
      "learning_rate": 1.3526477751855645e-05,
      "loss": 0.2458,
      "step": 97
    },
    {
      "epoch": 0.4096133751306165,
      "grad_norm": 0.8765418818205535,
      "learning_rate": 1.3398893015826166e-05,
      "loss": 0.2141,
      "step": 98
    },
    {
      "epoch": 0.41379310344827586,
      "grad_norm": 0.8402084285563245,
      "learning_rate": 1.3270679633174219e-05,
      "loss": 0.17,
      "step": 99
    },
    {
      "epoch": 0.4179728317659352,
      "grad_norm": 0.8927111902324383,
      "learning_rate": 1.3141861317770628e-05,
      "loss": 0.2419,
      "step": 100
    },
    {
      "epoch": 0.4221525600835946,
      "grad_norm": 0.9307548734337794,
      "learning_rate": 1.3012461895372343e-05,
      "loss": 0.2211,
      "step": 101
    },
    {
      "epoch": 0.4263322884012539,
      "grad_norm": 0.8182965551610069,
      "learning_rate": 1.2882505299215711e-05,
      "loss": 0.1546,
      "step": 102
    },
    {
      "epoch": 0.43051201671891326,
      "grad_norm": 0.5513459236136672,
      "learning_rate": 1.2752015565589852e-05,
      "loss": 0.1028,
      "step": 103
    },
    {
      "epoch": 0.4346917450365726,
      "grad_norm": 0.7895344671438952,
      "learning_rate": 1.2621016829391022e-05,
      "loss": 0.147,
      "step": 104
    },
    {
      "epoch": 0.438871473354232,
      "grad_norm": 0.7347726170918327,
      "learning_rate": 1.2489533319658703e-05,
      "loss": 0.1499,
      "step": 105
    },
    {
      "epoch": 0.4430512016718913,
      "grad_norm": 0.9930172706392331,
      "learning_rate": 1.2357589355094275e-05,
      "loss": 0.1999,
      "step": 106
    },
    {
      "epoch": 0.44723092998955066,
      "grad_norm": 0.700736172068482,
      "learning_rate": 1.2225209339563144e-05,
      "loss": 0.1727,
      "step": 107
    },
    {
      "epoch": 0.45141065830721006,
      "grad_norm": 0.9578584031611264,
      "learning_rate": 1.2092417757581085e-05,
      "loss": 0.2408,
      "step": 108
    },
    {
      "epoch": 0.4555903866248694,
      "grad_norm": 0.8203132426685221,
      "learning_rate": 1.1959239169785668e-05,
      "loss": 0.1951,
      "step": 109
    },
    {
      "epoch": 0.45977011494252873,
      "grad_norm": 0.9270712768052833,
      "learning_rate": 1.182569820839362e-05,
      "loss": 0.2222,
      "step": 110
    },
    {
      "epoch": 0.46394984326018807,
      "grad_norm": 0.9063435343774252,
      "learning_rate": 1.1691819572644941e-05,
      "loss": 0.1827,
      "step": 111
    },
    {
      "epoch": 0.46812957157784746,
      "grad_norm": 0.9742349956564444,
      "learning_rate": 1.155762802423463e-05,
      "loss": 0.2166,
      "step": 112
    },
    {
      "epoch": 0.4723092998955068,
      "grad_norm": 0.740191368731416,
      "learning_rate": 1.1423148382732854e-05,
      "loss": 0.162,
      "step": 113
    },
    {
      "epoch": 0.47648902821316613,
      "grad_norm": 2.152737189786801,
      "learning_rate": 1.128840552099439e-05,
      "loss": 0.2347,
      "step": 114
    },
    {
      "epoch": 0.48066875653082547,
      "grad_norm": 0.6837438960691173,
      "learning_rate": 1.1153424360558268e-05,
      "loss": 0.1344,
      "step": 115
    },
    {
      "epoch": 0.48484848484848486,
      "grad_norm": 0.8928037881326962,
      "learning_rate": 1.1018229867038358e-05,
      "loss": 0.1723,
      "step": 116
    },
    {
      "epoch": 0.4890282131661442,
      "grad_norm": 1.004544035476834,
      "learning_rate": 1.0882847045505809e-05,
      "loss": 0.1962,
      "step": 117
    },
    {
      "epoch": 0.49320794148380354,
      "grad_norm": 0.841348260273773,
      "learning_rate": 1.0747300935864245e-05,
      "loss": 0.1803,
      "step": 118
    },
    {
      "epoch": 0.49738766980146293,
      "grad_norm": 0.8496561849066544,
      "learning_rate": 1.0611616608218429e-05,
      "loss": 0.1905,
      "step": 119
    },
    {
      "epoch": 0.5015673981191222,
      "grad_norm": 1.1461318580041029,
      "learning_rate": 1.0475819158237426e-05,
      "loss": 0.2435,
      "step": 120
    },
    {
      "epoch": 0.5057471264367817,
      "grad_norm": 0.9171413538232467,
      "learning_rate": 1.0339933702512978e-05,
      "loss": 0.253,
      "step": 121
    },
    {
      "epoch": 0.509926854754441,
      "grad_norm": 0.7285731725801685,
      "learning_rate": 1.0203985373914056e-05,
      "loss": 0.1548,
      "step": 122
    },
    {
      "epoch": 0.5141065830721003,
      "grad_norm": 0.8652478932288088,
      "learning_rate": 1.0067999316938348e-05,
      "loss": 0.204,
      "step": 123
    },
    {
      "epoch": 0.5182863113897597,
      "grad_norm": 1.38915125451696,
      "learning_rate": 9.932000683061654e-06,
      "loss": 0.2442,
      "step": 124
    },
    {
      "epoch": 0.522466039707419,
      "grad_norm": 0.9268619467285734,
      "learning_rate": 9.79601462608595e-06,
      "loss": 0.2297,
      "step": 125
    },
    {
      "epoch": 0.5266457680250783,
      "grad_norm": 0.7880807499102447,
      "learning_rate": 9.660066297487024e-06,
      "loss": 0.181,
      "step": 126
    },
    {
      "epoch": 0.5308254963427377,
      "grad_norm": 0.754337068341988,
      "learning_rate": 9.524180841762577e-06,
      "loss": 0.1641,
      "step": 127
    },
    {
      "epoch": 0.5350052246603971,
      "grad_norm": 1.1050970913178217,
      "learning_rate": 9.388383391781576e-06,
      "loss": 0.2412,
      "step": 128
    },
    {
      "epoch": 0.5391849529780565,
      "grad_norm": 0.911580865548536,
      "learning_rate": 9.252699064135759e-06,
      "loss": 0.2092,
      "step": 129
    },
    {
      "epoch": 0.5433646812957158,
      "grad_norm": 0.8679286773035769,
      "learning_rate": 9.117152954494195e-06,
      "loss": 0.1902,
      "step": 130
    },
    {
      "epoch": 0.5475444096133751,
      "grad_norm": 0.76185165053938,
      "learning_rate": 8.981770132961649e-06,
      "loss": 0.2015,
      "step": 131
    },
    {
      "epoch": 0.5517241379310345,
      "grad_norm": 0.6153250846939452,
      "learning_rate": 8.846575639441732e-06,
      "loss": 0.1301,
      "step": 132
    },
    {
      "epoch": 0.5559038662486938,
      "grad_norm": 0.8129974861108137,
      "learning_rate": 8.711594479005614e-06,
      "loss": 0.1688,
      "step": 133
    },
    {
      "epoch": 0.5600835945663531,
      "grad_norm": 0.8932210165728961,
      "learning_rate": 8.576851617267151e-06,
      "loss": 0.142,
      "step": 134
    },
    {
      "epoch": 0.5642633228840125,
      "grad_norm": 1.2480488477623177,
      "learning_rate": 8.442371975765368e-06,
      "loss": 0.2117,
      "step": 135
    },
    {
      "epoch": 0.5684430512016719,
      "grad_norm": 1.2436168746130591,
      "learning_rate": 8.308180427355062e-06,
      "loss": 0.2376,
      "step": 136
    },
    {
      "epoch": 0.5726227795193313,
      "grad_norm": 0.8002711459729407,
      "learning_rate": 8.174301791606384e-06,
      "loss": 0.1827,
      "step": 137
    },
    {
      "epoch": 0.5768025078369906,
      "grad_norm": 0.8475588556028547,
      "learning_rate": 8.040760830214334e-06,
      "loss": 0.1658,
      "step": 138
    },
    {
      "epoch": 0.58098223615465,
      "grad_norm": 0.8639184920307346,
      "learning_rate": 7.907582242418916e-06,
      "loss": 0.17,
      "step": 139
    },
    {
      "epoch": 0.5851619644723093,
      "grad_norm": 0.8598440958255673,
      "learning_rate": 7.774790660436857e-06,
      "loss": 0.2112,
      "step": 140
    },
    {
      "epoch": 0.5893416927899686,
      "grad_norm": 0.8792127016276399,
      "learning_rate": 7.642410644905726e-06,
      "loss": 0.2157,
      "step": 141
    },
    {
      "epoch": 0.593521421107628,
      "grad_norm": 0.5310794775676358,
      "learning_rate": 7.5104666803413015e-06,
      "loss": 0.0995,
      "step": 142
    },
    {
      "epoch": 0.5977011494252874,
      "grad_norm": 0.6031042659903776,
      "learning_rate": 7.378983170608982e-06,
      "loss": 0.1226,
      "step": 143
    },
    {
      "epoch": 0.6018808777429467,
      "grad_norm": 0.9322591026745743,
      "learning_rate": 7.24798443441015e-06,
      "loss": 0.2269,
      "step": 144
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 0.7650887090451826,
      "learning_rate": 7.117494700784292e-06,
      "loss": 0.1972,
      "step": 145
    },
    {
      "epoch": 0.6102403343782654,
      "grad_norm": 1.08702051590229,
      "learning_rate": 6.9875381046276605e-06,
      "loss": 0.2312,
      "step": 146
    },
    {
      "epoch": 0.6144200626959248,
      "grad_norm": 1.1938077188160876,
      "learning_rate": 6.8581386822293765e-06,
      "loss": 0.1944,
      "step": 147
    },
    {
      "epoch": 0.6185997910135841,
      "grad_norm": 0.8286648541039914,
      "learning_rate": 6.729320366825785e-06,
      "loss": 0.2012,
      "step": 148
    },
    {
      "epoch": 0.6227795193312434,
      "grad_norm": 1.0448385340552901,
      "learning_rate": 6.601106984173835e-06,
      "loss": 0.176,
      "step": 149
    },
    {
      "epoch": 0.6269592476489029,
      "grad_norm": 0.6431098205020193,
      "learning_rate": 6.473522248144359e-06,
      "loss": 0.1611,
      "step": 150
    },
    {
      "epoch": 0.6311389759665622,
      "grad_norm": 0.6902192399005279,
      "learning_rate": 6.34658975633605e-06,
      "loss": 0.1571,
      "step": 151
    },
    {
      "epoch": 0.6353187042842215,
      "grad_norm": 1.4126319469067181,
      "learning_rate": 6.220332985710936e-06,
      "loss": 0.2253,
      "step": 152
    },
    {
      "epoch": 0.6394984326018809,
      "grad_norm": 0.9250333234508288,
      "learning_rate": 6.094775288252157e-06,
      "loss": 0.1659,
      "step": 153
    },
    {
      "epoch": 0.6436781609195402,
      "grad_norm": 1.0996139187104108,
      "learning_rate": 5.9699398866448846e-06,
      "loss": 0.1927,
      "step": 154
    },
    {
      "epoch": 0.6478578892371996,
      "grad_norm": 0.6600170395773419,
      "learning_rate": 5.845849869981137e-06,
      "loss": 0.1504,
      "step": 155
    },
    {
      "epoch": 0.6520376175548589,
      "grad_norm": 0.8091133996097659,
      "learning_rate": 5.722528189489294e-06,
      "loss": 0.1728,
      "step": 156
    },
    {
      "epoch": 0.6562173458725182,
      "grad_norm": 0.6787441434180406,
      "learning_rate": 5.599997654289129e-06,
      "loss": 0.1306,
      "step": 157
    },
    {
      "epoch": 0.6603970741901777,
      "grad_norm": 0.8602639720659184,
      "learning_rate": 5.478280927173145e-06,
      "loss": 0.168,
      "step": 158
    },
    {
      "epoch": 0.664576802507837,
      "grad_norm": 0.6826930291022159,
      "learning_rate": 5.357400520414898e-06,
      "loss": 0.1205,
      "step": 159
    },
    {
      "epoch": 0.6687565308254964,
      "grad_norm": 0.6701217983786535,
      "learning_rate": 5.237378791605249e-06,
      "loss": 0.1452,
      "step": 160
    },
    {
      "epoch": 0.6729362591431557,
      "grad_norm": 0.7578669897505922,
      "learning_rate": 5.11823793951719e-06,
      "loss": 0.1626,
      "step": 161
    },
    {
      "epoch": 0.677115987460815,
      "grad_norm": 0.8367106457596803,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.1981,
      "step": 162
    },
    {
      "epoch": 0.6812957157784744,
      "grad_norm": 0.8324385169673411,
      "learning_rate": 4.882686841903627e-06,
      "loss": 0.1945,
      "step": 163
    },
    {
      "epoch": 0.6854754440961337,
      "grad_norm": 0.8336564812001488,
      "learning_rate": 4.766320163033882e-06,
      "loss": 0.1497,
      "step": 164
    },
    {
      "epoch": 0.6896551724137931,
      "grad_norm": 0.9106572698316981,
      "learning_rate": 4.6509214861392785e-06,
      "loss": 0.2122,
      "step": 165
    },
    {
      "epoch": 0.6938349007314525,
      "grad_norm": 0.9899598996363259,
      "learning_rate": 4.5365121549302916e-06,
      "loss": 0.291,
      "step": 166
    },
    {
      "epoch": 0.6980146290491118,
      "grad_norm": 0.7526211809216892,
      "learning_rate": 4.423113330131708e-06,
      "loss": 0.1398,
      "step": 167
    },
    {
      "epoch": 0.7021943573667712,
      "grad_norm": 0.580108623854109,
      "learning_rate": 4.310745985568779e-06,
      "loss": 0.1228,
      "step": 168
    },
    {
      "epoch": 0.7063740856844305,
      "grad_norm": 0.5902701942035865,
      "learning_rate": 4.19943090428802e-06,
      "loss": 0.122,
      "step": 169
    },
    {
      "epoch": 0.7105538140020898,
      "grad_norm": 0.8908115982207944,
      "learning_rate": 4.0891886747132356e-06,
      "loss": 0.1679,
      "step": 170
    },
    {
      "epoch": 0.7147335423197492,
      "grad_norm": 0.7610727588905307,
      "learning_rate": 3.9800396868375675e-06,
      "loss": 0.1331,
      "step": 171
    },
    {
      "epoch": 0.7189132706374086,
      "grad_norm": 0.9084312218170052,
      "learning_rate": 3.872004128452231e-06,
      "loss": 0.2037,
      "step": 172
    },
    {
      "epoch": 0.723092998955068,
      "grad_norm": 0.7951214412363438,
      "learning_rate": 3.7651019814126656e-06,
      "loss": 0.152,
      "step": 173
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 1.0007664865334354,
      "learning_rate": 3.659353017942754e-06,
      "loss": 0.199,
      "step": 174
    },
    {
      "epoch": 0.7314524555903866,
      "grad_norm": 0.8000807402307464,
      "learning_rate": 3.5547767969778355e-06,
      "loss": 0.1874,
      "step": 175
    },
    {
      "epoch": 0.735632183908046,
      "grad_norm": 0.7056309894011483,
      "learning_rate": 3.4513926605471504e-06,
      "loss": 0.1408,
      "step": 176
    },
    {
      "epoch": 0.7398119122257053,
      "grad_norm": 1.0154810770219924,
      "learning_rate": 3.3492197301964145e-06,
      "loss": 0.1516,
      "step": 177
    },
    {
      "epoch": 0.7439916405433646,
      "grad_norm": 0.6774580884476967,
      "learning_rate": 3.248276903451171e-06,
      "loss": 0.1389,
      "step": 178
    },
    {
      "epoch": 0.7481713688610241,
      "grad_norm": 1.383727265413641,
      "learning_rate": 3.1485828503215588e-06,
      "loss": 0.1652,
      "step": 179
    },
    {
      "epoch": 0.7523510971786834,
      "grad_norm": 0.9084226985349156,
      "learning_rate": 3.0501560098492056e-06,
      "loss": 0.22,
      "step": 180
    },
    {
      "epoch": 0.7565308254963428,
      "grad_norm": 0.7904229164202767,
      "learning_rate": 2.9530145866967897e-06,
      "loss": 0.1688,
      "step": 181
    },
    {
      "epoch": 0.7607105538140021,
      "grad_norm": 0.9537175407913363,
      "learning_rate": 2.8571765477809645e-06,
      "loss": 0.2203,
      "step": 182
    },
    {
      "epoch": 0.7648902821316614,
      "grad_norm": 0.9045213538470349,
      "learning_rate": 2.7626596189492983e-06,
      "loss": 0.2235,
      "step": 183
    },
    {
      "epoch": 0.7690700104493208,
      "grad_norm": 0.8437761162278906,
      "learning_rate": 2.669481281701739e-06,
      "loss": 0.2405,
      "step": 184
    },
    {
      "epoch": 0.7732497387669801,
      "grad_norm": 0.9170366332680835,
      "learning_rate": 2.5776587699573007e-06,
      "loss": 0.2356,
      "step": 185
    },
    {
      "epoch": 0.7774294670846394,
      "grad_norm": 0.859807092495294,
      "learning_rate": 2.487209066866565e-06,
      "loss": 0.1581,
      "step": 186
    },
    {
      "epoch": 0.7816091954022989,
      "grad_norm": 0.7541718313410231,
      "learning_rate": 2.398148901670521e-06,
      "loss": 0.1735,
      "step": 187
    },
    {
      "epoch": 0.7857889237199582,
      "grad_norm": 0.5298691809105464,
      "learning_rate": 2.3104947466063785e-06,
      "loss": 0.1247,
      "step": 188
    },
    {
      "epoch": 0.7899686520376176,
      "grad_norm": 0.5509131740535222,
      "learning_rate": 2.224262813860962e-06,
      "loss": 0.1135,
      "step": 189
    },
    {
      "epoch": 0.7941483803552769,
      "grad_norm": 1.211014988349712,
      "learning_rate": 2.1394690525721275e-06,
      "loss": 0.2396,
      "step": 190
    },
    {
      "epoch": 0.7983281086729362,
      "grad_norm": 0.8304883104248448,
      "learning_rate": 2.0561291458788736e-06,
      "loss": 0.1495,
      "step": 191
    },
    {
      "epoch": 0.8025078369905956,
      "grad_norm": 0.6575918108532034,
      "learning_rate": 1.9742585080206754e-06,
      "loss": 0.1705,
      "step": 192
    },
    {
      "epoch": 0.8066875653082549,
      "grad_norm": 0.6483828691185768,
      "learning_rate": 1.8938722814864863e-06,
      "loss": 0.1179,
      "step": 193
    },
    {
      "epoch": 0.8108672936259144,
      "grad_norm": 0.7874326139345168,
      "learning_rate": 1.8149853342140644e-06,
      "loss": 0.1911,
      "step": 194
    },
    {
      "epoch": 0.8150470219435737,
      "grad_norm": 0.9396325288901851,
      "learning_rate": 1.7376122568400533e-06,
      "loss": 0.2313,
      "step": 195
    },
    {
      "epoch": 0.819226750261233,
      "grad_norm": 0.8980422324440791,
      "learning_rate": 1.6617673600013295e-06,
      "loss": 0.2637,
      "step": 196
    },
    {
      "epoch": 0.8234064785788924,
      "grad_norm": 0.8393671732445989,
      "learning_rate": 1.587464671688187e-06,
      "loss": 0.2008,
      "step": 197
    },
    {
      "epoch": 0.8275862068965517,
      "grad_norm": 0.9628013968768121,
      "learning_rate": 1.5147179346497665e-06,
      "loss": 0.1803,
      "step": 198
    },
    {
      "epoch": 0.831765935214211,
      "grad_norm": 0.8055867006150553,
      "learning_rate": 1.443540603852227e-06,
      "loss": 0.2058,
      "step": 199
    },
    {
      "epoch": 0.8359456635318704,
      "grad_norm": 0.7402053267386335,
      "learning_rate": 1.373945843990192e-06,
      "loss": 0.1678,
      "step": 200
    },
    {
      "epoch": 0.8401253918495298,
      "grad_norm": 0.6345259422046625,
      "learning_rate": 1.3059465270518469e-06,
      "loss": 0.1499,
      "step": 201
    },
    {
      "epoch": 0.8443051201671892,
      "grad_norm": 0.9278886576312714,
      "learning_rate": 1.2395552299381742e-06,
      "loss": 0.1376,
      "step": 202
    },
    {
      "epoch": 0.8484848484848485,
      "grad_norm": 0.7305237779713204,
      "learning_rate": 1.1747842321367886e-06,
      "loss": 0.1689,
      "step": 203
    },
    {
      "epoch": 0.8526645768025078,
      "grad_norm": 0.785339323897186,
      "learning_rate": 1.1116455134507665e-06,
      "loss": 0.1515,
      "step": 204
    },
    {
      "epoch": 0.8568443051201672,
      "grad_norm": 0.8232817926726907,
      "learning_rate": 1.0501507517829012e-06,
      "loss": 0.2148,
      "step": 205
    },
    {
      "epoch": 0.8610240334378265,
      "grad_norm": 0.920951874634867,
      "learning_rate": 9.903113209758098e-07,
      "loss": 0.1981,
      "step": 206
    },
    {
      "epoch": 0.8652037617554859,
      "grad_norm": 0.778697980092621,
      "learning_rate": 9.321382887082564e-07,
      "loss": 0.1697,
      "step": 207
    },
    {
      "epoch": 0.8693834900731452,
      "grad_norm": 0.552864168373498,
      "learning_rate": 8.756424144481313e-07,
      "loss": 0.1197,
      "step": 208
    },
    {
      "epoch": 0.8735632183908046,
      "grad_norm": 0.7721394559949915,
      "learning_rate": 8.208341474624071e-07,
      "loss": 0.1636,
      "step": 209
    },
    {
      "epoch": 0.877742946708464,
      "grad_norm": 0.5887750742792935,
      "learning_rate": 7.677236248844855e-07,
      "loss": 0.1238,
      "step": 210
    },
    {
      "epoch": 0.8819226750261233,
      "grad_norm": 1.1706298717117896,
      "learning_rate": 7.163206698392744e-07,
      "loss": 0.2685,
      "step": 211
    },
    {
      "epoch": 0.8861024033437827,
      "grad_norm": 0.7156481787290525,
      "learning_rate": 6.666347896263326e-07,
      "loss": 0.1272,
      "step": 212
    },
    {
      "epoch": 0.890282131661442,
      "grad_norm": 0.9148297160556185,
      "learning_rate": 6.186751739614405e-07,
      "loss": 0.1878,
      "step": 213
    },
    {
      "epoch": 0.8944618599791013,
      "grad_norm": 0.8443233125008259,
      "learning_rate": 5.724506932769014e-07,
      "loss": 0.1872,
      "step": 214
    },
    {
      "epoch": 0.8986415882967607,
      "grad_norm": 0.7906265849992707,
      "learning_rate": 5.279698970809011e-07,
      "loss": 0.1672,
      "step": 215
    },
    {
      "epoch": 0.9028213166144201,
      "grad_norm": 0.8000045890229907,
      "learning_rate": 4.852410123762164e-07,
      "loss": 0.1357,
      "step": 216
    },
    {
      "epoch": 0.9070010449320794,
      "grad_norm": 0.7692285277924558,
      "learning_rate": 4.4427194213859216e-07,
      "loss": 0.1733,
      "step": 217
    },
    {
      "epoch": 0.9111807732497388,
      "grad_norm": 0.6828595942343869,
      "learning_rate": 4.0507026385502747e-07,
      "loss": 0.1377,
      "step": 218
    },
    {
      "epoch": 0.9153605015673981,
      "grad_norm": 0.7616205791512801,
      "learning_rate": 3.6764322812226416e-07,
      "loss": 0.1743,
      "step": 219
    },
    {
      "epoch": 0.9195402298850575,
      "grad_norm": 0.6920015184607431,
      "learning_rate": 3.319977573057642e-07,
      "loss": 0.178,
      "step": 220
    },
    {
      "epoch": 0.9237199582027168,
      "grad_norm": 0.8599083603091006,
      "learning_rate": 2.9814044425935605e-07,
      "loss": 0.211,
      "step": 221
    },
    {
      "epoch": 0.9278996865203761,
      "grad_norm": 0.8372366140139196,
      "learning_rate": 2.6607755110584886e-07,
      "loss": 0.1693,
      "step": 222
    },
    {
      "epoch": 0.9320794148380356,
      "grad_norm": 0.740216908527879,
      "learning_rate": 2.3581500807882462e-07,
      "loss": 0.1677,
      "step": 223
    },
    {
      "epoch": 0.9362591431556949,
      "grad_norm": 0.7486951708296801,
      "learning_rate": 2.0735841242578992e-07,
      "loss": 0.1367,
      "step": 224
    },
    {
      "epoch": 0.9404388714733543,
      "grad_norm": 0.6059460975199559,
      "learning_rate": 1.8071302737293294e-07,
      "loss": 0.1287,
      "step": 225
    },
    {
      "epoch": 0.9446185997910136,
      "grad_norm": 0.6725131427516452,
      "learning_rate": 1.558837811516667e-07,
      "loss": 0.1355,
      "step": 226
    },
    {
      "epoch": 0.9487983281086729,
      "grad_norm": 0.89103460281586,
      "learning_rate": 1.3287526608711132e-07,
      "loss": 0.2224,
      "step": 227
    },
    {
      "epoch": 0.9529780564263323,
      "grad_norm": 0.7294258012671165,
      "learning_rate": 1.1169173774871478e-07,
      "loss": 0.1247,
      "step": 228
    },
    {
      "epoch": 0.9571577847439916,
      "grad_norm": 0.5563178146719318,
      "learning_rate": 9.233711416316571e-08,
      "loss": 0.1229,
      "step": 229
    },
    {
      "epoch": 0.9613375130616509,
      "grad_norm": 1.2052590543331665,
      "learning_rate": 7.481497508972313e-08,
      "loss": 0.1689,
      "step": 230
    },
    {
      "epoch": 0.9655172413793104,
      "grad_norm": 0.8195585207091408,
      "learning_rate": 5.912856135812051e-08,
      "loss": 0.1942,
      "step": 231
    },
    {
      "epoch": 0.9696969696969697,
      "grad_norm": 2.0024730819109773,
      "learning_rate": 4.528077426915412e-08,
      "loss": 0.1719,
      "step": 232
    },
    {
      "epoch": 0.9738766980146291,
      "grad_norm": 0.6177697556146066,
      "learning_rate": 3.327417505806785e-08,
      "loss": 0.1286,
      "step": 233
    },
    {
      "epoch": 0.9780564263322884,
      "grad_norm": 2.319773776301439,
      "learning_rate": 2.311098442083659e-08,
      "loss": 0.2415,
      "step": 234
    },
    {
      "epoch": 0.9822361546499477,
      "grad_norm": 1.1389478132400794,
      "learning_rate": 1.4793082103435885e-08,
      "loss": 0.1913,
      "step": 235
    },
    {
      "epoch": 0.9864158829676071,
      "grad_norm": 0.8823489571757573,
      "learning_rate": 8.322006554171147e-09,
      "loss": 0.2037,
      "step": 236
    },
    {
      "epoch": 0.9905956112852664,
      "grad_norm": 0.6469295501938342,
      "learning_rate": 3.698954639129726e-09,
      "loss": 0.1321,
      "step": 237
    },
    {
      "epoch": 0.9947753396029259,
      "grad_norm": 0.8129747939307076,
      "learning_rate": 9.24781420816867e-10,
      "loss": 0.194,
      "step": 238
    },
    {
      "epoch": 0.9989550679205852,
      "grad_norm": 0.9261130483396426,
      "learning_rate": 0.0,
      "loss": 0.1793,
      "step": 239
    },
    {
      "epoch": 0.9989550679205852,
      "step": 239,
      "total_flos": 449784335171584.0,
      "train_loss": 0.26206765498451606,
      "train_runtime": 2412.5124,
      "train_samples_per_second": 12.688,
      "train_steps_per_second": 0.099
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 239,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 50000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 449784335171584.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}