{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004,
      "grad_norm": 0.07388347883476193,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.2514,
      "step": 1
    },
    {
      "epoch": 0.008,
      "grad_norm": 0.09795980573968203,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.3381,
      "step": 2
    },
    {
      "epoch": 0.012,
      "grad_norm": 0.06413810724110217,
      "learning_rate": 6e-06,
      "loss": 0.2392,
      "step": 3
    },
    {
      "epoch": 0.016,
      "grad_norm": 0.034123085192894025,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.0721,
      "step": 4
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.06232259003940912,
      "learning_rate": 1e-05,
      "loss": 0.1172,
      "step": 5
    },
    {
      "epoch": 0.024,
      "grad_norm": 0.12103187755930987,
      "learning_rate": 1.2e-05,
      "loss": 0.4085,
      "step": 6
    },
    {
      "epoch": 0.028,
      "grad_norm": 0.05752155590324857,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 0.1006,
      "step": 7
    },
    {
      "epoch": 0.032,
      "grad_norm": 0.08425524502344583,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.1664,
      "step": 8
    },
    {
      "epoch": 0.036,
      "grad_norm": 0.06561945132362518,
      "learning_rate": 1.8e-05,
      "loss": 0.2454,
      "step": 9
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.08051637521228767,
      "learning_rate": 2e-05,
      "loss": 0.132,
      "step": 10
    },
    {
      "epoch": 0.044,
      "grad_norm": 0.0986120083414008,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 0.2402,
      "step": 11
    },
    {
      "epoch": 0.048,
      "grad_norm": 0.06720936521726577,
      "learning_rate": 2.4e-05,
      "loss": 0.2361,
      "step": 12
    },
    {
      "epoch": 0.052,
      "grad_norm": 0.02162346097698542,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 0.0438,
      "step": 13
    },
    {
      "epoch": 0.056,
      "grad_norm": 0.03395487061118089,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 0.0815,
      "step": 14
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.03969125485464094,
      "learning_rate": 3e-05,
      "loss": 0.0903,
      "step": 15
    },
    {
      "epoch": 0.064,
      "grad_norm": 0.03385707574706431,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 0.0898,
      "step": 16
    },
    {
      "epoch": 0.068,
      "grad_norm": 0.04023068586555507,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 0.09,
      "step": 17
    },
    {
      "epoch": 0.072,
      "grad_norm": 0.060004002153418515,
      "learning_rate": 3.6e-05,
      "loss": 0.1244,
      "step": 18
    },
    {
      "epoch": 0.076,
      "grad_norm": 0.058641311800235046,
      "learning_rate": 3.8e-05,
      "loss": 0.1401,
      "step": 19
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.0926522345509661,
      "learning_rate": 4e-05,
      "loss": 0.3034,
      "step": 20
    },
    {
      "epoch": 0.084,
      "grad_norm": 0.03916584350916536,
      "learning_rate": 4.2e-05,
      "loss": 0.1165,
      "step": 21
    },
    {
      "epoch": 0.088,
      "grad_norm": 0.022892860374495977,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.0332,
      "step": 22
    },
    {
      "epoch": 0.092,
      "grad_norm": 0.02961058793805,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.052,
      "step": 23
    },
    {
      "epoch": 0.096,
      "grad_norm": 0.18559102630265306,
      "learning_rate": 4.8e-05,
      "loss": 0.353,
      "step": 24
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.10935023597670036,
      "learning_rate": 5e-05,
      "loss": 0.1428,
      "step": 25
    },
    {
      "epoch": 0.104,
      "grad_norm": 0.10779842778193618,
      "learning_rate": 5.2000000000000004e-05,
      "loss": 0.2226,
      "step": 26
    },
    {
      "epoch": 0.108,
      "grad_norm": 0.06224075016286207,
      "learning_rate": 5.4000000000000005e-05,
      "loss": 0.1137,
      "step": 27
    },
    {
      "epoch": 0.112,
      "grad_norm": 0.12050646042873905,
      "learning_rate": 5.6000000000000006e-05,
      "loss": 0.2841,
      "step": 28
    },
    {
      "epoch": 0.116,
      "grad_norm": 0.04934056170124901,
      "learning_rate": 5.8e-05,
      "loss": 0.0713,
      "step": 29
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.09345665116546086,
      "learning_rate": 6e-05,
      "loss": 0.1204,
      "step": 30
    },
    {
      "epoch": 0.124,
      "grad_norm": 0.13945388282552793,
      "learning_rate": 6.2e-05,
      "loss": 0.236,
      "step": 31
    },
    {
      "epoch": 0.128,
      "grad_norm": 0.0741124910365894,
      "learning_rate": 6.400000000000001e-05,
      "loss": 0.1773,
      "step": 32
    },
    {
      "epoch": 0.132,
      "grad_norm": 0.07278452345871356,
      "learning_rate": 6.6e-05,
      "loss": 0.1796,
      "step": 33
    },
    {
      "epoch": 0.136,
      "grad_norm": 0.08325646458072568,
      "learning_rate": 6.800000000000001e-05,
      "loss": 0.1304,
      "step": 34
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.04032605866332654,
      "learning_rate": 7e-05,
      "loss": 0.0386,
      "step": 35
    },
    {
      "epoch": 0.144,
      "grad_norm": 0.06366607819671799,
      "learning_rate": 7.2e-05,
      "loss": 0.1,
      "step": 36
    },
    {
      "epoch": 0.148,
      "grad_norm": 0.11303249373466372,
      "learning_rate": 7.4e-05,
      "loss": 0.1897,
      "step": 37
    },
    {
      "epoch": 0.152,
      "grad_norm": 0.09974436096894448,
      "learning_rate": 7.6e-05,
      "loss": 0.084,
      "step": 38
    },
    {
      "epoch": 0.156,
      "grad_norm": 0.03308908234015684,
      "learning_rate": 7.800000000000001e-05,
      "loss": 0.0414,
      "step": 39
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.03817872481868544,
      "learning_rate": 8e-05,
      "loss": 0.0354,
      "step": 40
    },
    {
      "epoch": 0.164,
      "grad_norm": 0.03146673818340689,
      "learning_rate": 8.2e-05,
      "loss": 0.036,
      "step": 41
    },
    {
      "epoch": 0.168,
      "grad_norm": 0.0321719633892723,
      "learning_rate": 8.4e-05,
      "loss": 0.0447,
      "step": 42
    },
    {
      "epoch": 0.172,
      "grad_norm": 0.04455164585905991,
      "learning_rate": 8.6e-05,
      "loss": 0.0586,
      "step": 43
    },
    {
      "epoch": 0.176,
      "grad_norm": 0.1373034173451929,
      "learning_rate": 8.800000000000001e-05,
      "loss": 0.1312,
      "step": 44
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.09869772142863122,
      "learning_rate": 9e-05,
      "loss": 0.1693,
      "step": 45
    },
    {
      "epoch": 0.184,
      "grad_norm": 0.06480039892231607,
      "learning_rate": 9.200000000000001e-05,
      "loss": 0.0676,
      "step": 46
    },
    {
      "epoch": 0.188,
      "grad_norm": 0.049858214936935374,
      "learning_rate": 9.4e-05,
      "loss": 0.0577,
      "step": 47
    },
    {
      "epoch": 0.192,
      "grad_norm": 0.0742291390645976,
      "learning_rate": 9.6e-05,
      "loss": 0.1237,
      "step": 48
    },
    {
      "epoch": 0.196,
      "grad_norm": 0.0875555392778545,
      "learning_rate": 9.8e-05,
      "loss": 0.1261,
      "step": 49
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.06625561698629966,
      "learning_rate": 0.0001,
      "loss": 0.0817,
      "step": 50
    },
    {
      "epoch": 0.204,
      "grad_norm": 0.17993726193994958,
      "learning_rate": 9.999878153526974e-05,
      "loss": 0.2117,
      "step": 51
    },
    {
      "epoch": 0.208,
      "grad_norm": 0.034174415723349505,
      "learning_rate": 9.999512620046522e-05,
      "loss": 0.0332,
      "step": 52
    },
    {
      "epoch": 0.212,
      "grad_norm": 0.09735494735921241,
      "learning_rate": 9.998903417374228e-05,
      "loss": 0.1324,
      "step": 53
    },
    {
      "epoch": 0.216,
      "grad_norm": 0.056421074758998144,
      "learning_rate": 9.998050575201771e-05,
      "loss": 0.031,
      "step": 54
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.05375373761271103,
      "learning_rate": 9.99695413509548e-05,
      "loss": 0.0207,
      "step": 55
    },
    {
      "epoch": 0.224,
      "grad_norm": 0.060294137961844574,
      "learning_rate": 9.995614150494293e-05,
      "loss": 0.0796,
      "step": 56
    },
    {
      "epoch": 0.228,
      "grad_norm": 0.08551982478332096,
      "learning_rate": 9.99403068670717e-05,
      "loss": 0.0919,
      "step": 57
    },
    {
      "epoch": 0.232,
      "grad_norm": 0.13255704119467507,
      "learning_rate": 9.992203820909906e-05,
      "loss": 0.118,
      "step": 58
    },
    {
      "epoch": 0.236,
      "grad_norm": 0.11556760833895804,
      "learning_rate": 9.990133642141359e-05,
      "loss": 0.0896,
      "step": 59
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.03235216107094882,
      "learning_rate": 9.987820251299122e-05,
      "loss": 0.0145,
      "step": 60
    },
    {
      "epoch": 0.244,
      "grad_norm": 0.11521891043881091,
      "learning_rate": 9.985263761134602e-05,
      "loss": 0.0342,
      "step": 61
    },
    {
      "epoch": 0.248,
      "grad_norm": 0.14309399783766089,
      "learning_rate": 9.982464296247522e-05,
      "loss": 0.0818,
      "step": 62
    },
    {
      "epoch": 0.252,
      "grad_norm": 0.0867395302232182,
      "learning_rate": 9.979421993079852e-05,
      "loss": 0.0436,
      "step": 63
    },
    {
      "epoch": 0.256,
      "grad_norm": 0.12566567183658883,
      "learning_rate": 9.976136999909156e-05,
      "loss": 0.0493,
      "step": 64
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.05723937513458756,
      "learning_rate": 9.972609476841367e-05,
      "loss": 0.0223,
      "step": 65
    },
    {
      "epoch": 0.264,
      "grad_norm": 0.1467099131268422,
      "learning_rate": 9.968839595802982e-05,
      "loss": 0.09,
      "step": 66
    },
    {
      "epoch": 0.268,
      "grad_norm": 0.17320601521009465,
      "learning_rate": 9.964827540532685e-05,
      "loss": 0.0505,
      "step": 67
    },
    {
      "epoch": 0.272,
      "grad_norm": 0.039574218905831716,
      "learning_rate": 9.96057350657239e-05,
      "loss": 0.0169,
      "step": 68
    },
    {
      "epoch": 0.276,
      "grad_norm": 0.0354483093246312,
      "learning_rate": 9.956077701257709e-05,
      "loss": 0.0109,
      "step": 69
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.03457133259811366,
      "learning_rate": 9.951340343707852e-05,
      "loss": 0.0113,
      "step": 70
    },
    {
      "epoch": 0.284,
      "grad_norm": 0.09796594149093318,
      "learning_rate": 9.946361664814943e-05,
      "loss": 0.0483,
      "step": 71
    },
    {
      "epoch": 0.288,
      "grad_norm": 0.0841886963166831,
      "learning_rate": 9.941141907232765e-05,
      "loss": 0.0338,
      "step": 72
    },
    {
      "epoch": 0.292,
      "grad_norm": 0.07754903697632377,
      "learning_rate": 9.93568132536494e-05,
      "loss": 0.0423,
      "step": 73
    },
    {
      "epoch": 0.296,
      "grad_norm": 0.07398761608452874,
      "learning_rate": 9.929980185352526e-05,
      "loss": 0.0475,
      "step": 74
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.23933165389400574,
      "learning_rate": 9.924038765061042e-05,
      "loss": 0.1911,
      "step": 75
    },
    {
      "epoch": 0.304,
      "grad_norm": 0.16849923275637996,
      "learning_rate": 9.917857354066931e-05,
      "loss": 0.0257,
      "step": 76
    },
    {
      "epoch": 0.308,
      "grad_norm": 0.01840716826653097,
      "learning_rate": 9.911436253643445e-05,
      "loss": 0.0045,
      "step": 77
    },
    {
      "epoch": 0.312,
      "grad_norm": 0.06121982260459926,
      "learning_rate": 9.904775776745958e-05,
      "loss": 0.0175,
      "step": 78
    },
    {
      "epoch": 0.316,
      "grad_norm": 0.07369645779979246,
      "learning_rate": 9.89787624799672e-05,
      "loss": 0.0166,
      "step": 79
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.02897618672900009,
      "learning_rate": 9.890738003669029e-05,
      "loss": 0.0071,
      "step": 80
    },
    {
      "epoch": 0.324,
      "grad_norm": 0.1395280733375726,
      "learning_rate": 9.88336139167084e-05,
      "loss": 0.0511,
      "step": 81
    },
    {
      "epoch": 0.328,
      "grad_norm": 0.04932771225152733,
      "learning_rate": 9.875746771527816e-05,
      "loss": 0.0068,
      "step": 82
    },
    {
      "epoch": 0.332,
      "grad_norm": 0.0882869650008996,
      "learning_rate": 9.867894514365802e-05,
      "loss": 0.0527,
      "step": 83
    },
    {
      "epoch": 0.336,
      "grad_norm": 0.195095539440849,
      "learning_rate": 9.859805002892732e-05,
      "loss": 0.0467,
      "step": 84
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.01187867042981877,
      "learning_rate": 9.851478631379982e-05,
      "loss": 0.0025,
      "step": 85
    },
    {
      "epoch": 0.344,
      "grad_norm": 0.16355086146898856,
      "learning_rate": 9.842915805643155e-05,
      "loss": 0.0356,
      "step": 86
    },
    {
      "epoch": 0.348,
      "grad_norm": 0.03558673326569455,
      "learning_rate": 9.834116943022298e-05,
      "loss": 0.0064,
      "step": 87
    },
    {
      "epoch": 0.352,
      "grad_norm": 0.017896258587265748,
      "learning_rate": 9.825082472361557e-05,
      "loss": 0.0048,
      "step": 88
    },
    {
      "epoch": 0.356,
      "grad_norm": 0.23790851142116018,
      "learning_rate": 9.815812833988291e-05,
      "loss": 0.1288,
      "step": 89
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.011634413387924678,
      "learning_rate": 9.806308479691595e-05,
      "loss": 0.0041,
      "step": 90
    },
    {
      "epoch": 0.364,
      "grad_norm": 0.00609763628083542,
      "learning_rate": 9.796569872700288e-05,
      "loss": 0.0018,
      "step": 91
    },
    {
      "epoch": 0.368,
      "grad_norm": 0.05459767320460325,
      "learning_rate": 9.786597487660337e-05,
      "loss": 0.0078,
      "step": 92
    },
    {
      "epoch": 0.372,
      "grad_norm": 0.06042886323579011,
      "learning_rate": 9.776391810611718e-05,
      "loss": 0.0125,
      "step": 93
    },
    {
      "epoch": 0.376,
      "grad_norm": 0.039059542767049996,
      "learning_rate": 9.765953338964735e-05,
      "loss": 0.0077,
      "step": 94
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.10327207978008529,
      "learning_rate": 9.755282581475769e-05,
      "loss": 0.0268,
      "step": 95
    },
    {
      "epoch": 0.384,
      "grad_norm": 0.04864708392391147,
      "learning_rate": 9.744380058222483e-05,
      "loss": 0.0091,
      "step": 96
    },
    {
      "epoch": 0.388,
      "grad_norm": 0.06499020044440326,
      "learning_rate": 9.733246300578483e-05,
      "loss": 0.039,
      "step": 97
    },
    {
      "epoch": 0.392,
      "grad_norm": 0.011139183561155766,
      "learning_rate": 9.721881851187406e-05,
      "loss": 0.0021,
      "step": 98
    },
    {
      "epoch": 0.396,
      "grad_norm": 0.005775447940199373,
      "learning_rate": 9.710287263936484e-05,
      "loss": 0.0015,
      "step": 99
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.10193594952480603,
      "learning_rate": 9.698463103929542e-05,
      "loss": 0.0174,
      "step": 100
    },
    {
      "epoch": 0.404,
      "grad_norm": 0.05856344174771734,
      "learning_rate": 9.686409947459458e-05,
      "loss": 0.0475,
      "step": 101
    },
    {
      "epoch": 0.408,
      "grad_norm": 0.02017970044315856,
      "learning_rate": 9.674128381980072e-05,
      "loss": 0.007,
      "step": 102
    },
    {
      "epoch": 0.412,
      "grad_norm": 0.11943609264859856,
      "learning_rate": 9.661619006077561e-05,
      "loss": 0.0839,
      "step": 103
    },
    {
      "epoch": 0.416,
      "grad_norm": 0.18905832726333935,
      "learning_rate": 9.648882429441257e-05,
      "loss": 0.0207,
      "step": 104
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.08889213679319669,
      "learning_rate": 9.635919272833938e-05,
      "loss": 0.0191,
      "step": 105
    },
    {
      "epoch": 0.424,
      "grad_norm": 0.17562704002698695,
      "learning_rate": 9.622730168061567e-05,
      "loss": 0.069,
      "step": 106
    },
    {
      "epoch": 0.428,
      "grad_norm": 0.028700964663471582,
      "learning_rate": 9.609315757942503e-05,
      "loss": 0.004,
      "step": 107
    },
    {
      "epoch": 0.432,
      "grad_norm": 0.009738063807270578,
      "learning_rate": 9.595676696276172e-05,
      "loss": 0.0022,
      "step": 108
    },
    {
      "epoch": 0.436,
      "grad_norm": 0.010547938125797419,
      "learning_rate": 9.581813647811198e-05,
      "loss": 0.0016,
      "step": 109
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.009364562549044194,
      "learning_rate": 9.567727288213005e-05,
      "loss": 0.0019,
      "step": 110
    },
    {
      "epoch": 0.444,
      "grad_norm": 0.07757018747323255,
      "learning_rate": 9.553418304030886e-05,
      "loss": 0.0093,
      "step": 111
    },
    {
      "epoch": 0.448,
      "grad_norm": 0.14787807375555986,
      "learning_rate": 9.538887392664544e-05,
      "loss": 0.0626,
      "step": 112
    },
    {
      "epoch": 0.452,
      "grad_norm": 0.14286737205454403,
      "learning_rate": 9.524135262330098e-05,
      "loss": 0.0285,
      "step": 113
    },
    {
      "epoch": 0.456,
      "grad_norm": 0.07472727437377708,
      "learning_rate": 9.50916263202557e-05,
      "loss": 0.0263,
      "step": 114
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.025763087930580957,
      "learning_rate": 9.493970231495835e-05,
      "loss": 0.0066,
      "step": 115
    },
    {
      "epoch": 0.464,
      "grad_norm": 0.002687665301546454,
      "learning_rate": 9.478558801197065e-05,
      "loss": 0.001,
      "step": 116
    },
    {
      "epoch": 0.468,
      "grad_norm": 0.007755775115883136,
      "learning_rate": 9.462929092260628e-05,
      "loss": 0.0028,
      "step": 117
    },
    {
      "epoch": 0.472,
      "grad_norm": 0.04949815093883928,
      "learning_rate": 9.447081866456489e-05,
      "loss": 0.0125,
      "step": 118
    },
    {
      "epoch": 0.476,
      "grad_norm": 0.06887173879132054,
      "learning_rate": 9.431017896156074e-05,
      "loss": 0.0217,
      "step": 119
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.03733399679427583,
      "learning_rate": 9.414737964294636e-05,
      "loss": 0.0165,
      "step": 120
    },
    {
      "epoch": 0.484,
      "grad_norm": 0.04691985961692329,
      "learning_rate": 9.398242864333083e-05,
      "loss": 0.0306,
      "step": 121
    },
    {
      "epoch": 0.488,
      "grad_norm": 0.04181397738491031,
      "learning_rate": 9.381533400219318e-05,
      "loss": 0.0099,
      "step": 122
    },
    {
      "epoch": 0.492,
      "grad_norm": 0.061248036707334724,
      "learning_rate": 9.364610386349049e-05,
      "loss": 0.0261,
      "step": 123
    },
    {
      "epoch": 0.496,
      "grad_norm": 0.09176252539997068,
      "learning_rate": 9.347474647526095e-05,
      "loss": 0.0236,
      "step": 124
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.0850583285355193,
      "learning_rate": 9.330127018922194e-05,
      "loss": 0.1214,
      "step": 125
    },
    {
      "epoch": 0.504,
      "grad_norm": 0.0056071374269381045,
      "learning_rate": 9.312568346036288e-05,
      "loss": 0.0024,
      "step": 126
    },
    {
      "epoch": 0.508,
      "grad_norm": 0.08172868685511792,
      "learning_rate": 9.294799484653323e-05,
      "loss": 0.0288,
      "step": 127
    },
    {
      "epoch": 0.512,
      "grad_norm": 0.007651371191605807,
      "learning_rate": 9.276821300802534e-05,
      "loss": 0.0026,
      "step": 128
    },
    {
      "epoch": 0.516,
      "grad_norm": 0.1002478400588997,
      "learning_rate": 9.258634670715238e-05,
      "loss": 0.0672,
      "step": 129
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.007231775374589401,
      "learning_rate": 9.24024048078213e-05,
      "loss": 0.0015,
      "step": 130
    },
    {
      "epoch": 0.524,
      "grad_norm": 0.06815148847401276,
      "learning_rate": 9.221639627510076e-05,
      "loss": 0.0202,
      "step": 131
    },
    {
      "epoch": 0.528,
      "grad_norm": 0.036802086357260254,
      "learning_rate": 9.202833017478422e-05,
      "loss": 0.0068,
      "step": 132
    },
    {
      "epoch": 0.532,
      "grad_norm": 0.03836929209135655,
      "learning_rate": 9.183821567294809e-05,
      "loss": 0.0123,
      "step": 133
    },
    {
      "epoch": 0.536,
      "grad_norm": 0.029377306593278314,
      "learning_rate": 9.164606203550497e-05,
      "loss": 0.0082,
      "step": 134
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.07500567200684409,
      "learning_rate": 9.145187862775209e-05,
      "loss": 0.0239,
      "step": 135
    },
    {
      "epoch": 0.544,
      "grad_norm": 0.06588323188308597,
      "learning_rate": 9.125567491391476e-05,
      "loss": 0.0435,
      "step": 136
    },
    {
      "epoch": 0.548,
      "grad_norm": 0.10713463485686177,
      "learning_rate": 9.105746045668521e-05,
      "loss": 0.035,
      "step": 137
    },
    {
      "epoch": 0.552,
      "grad_norm": 0.04037178641572994,
      "learning_rate": 9.085724491675642e-05,
      "loss": 0.011,
      "step": 138
    },
    {
      "epoch": 0.556,
      "grad_norm": 0.23400404142175088,
      "learning_rate": 9.065503805235138e-05,
      "loss": 0.1519,
      "step": 139
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.01103001350969286,
      "learning_rate": 9.045084971874738e-05,
      "loss": 0.0015,
      "step": 140
    },
    {
      "epoch": 0.564,
      "grad_norm": 0.00516542968995078,
      "learning_rate": 9.02446898677957e-05,
      "loss": 0.0018,
      "step": 141
    },
    {
      "epoch": 0.568,
      "grad_norm": 0.03516354355432361,
      "learning_rate": 9.003656854743667e-05,
      "loss": 0.0112,
      "step": 142
    },
    {
      "epoch": 0.572,
      "grad_norm": 0.006244510925825045,
      "learning_rate": 8.982649590120982e-05,
      "loss": 0.0011,
      "step": 143
    },
    {
      "epoch": 0.576,
      "grad_norm": 0.0135727079846708,
      "learning_rate": 8.961448216775954e-05,
      "loss": 0.0056,
      "step": 144
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.06110460884476186,
      "learning_rate": 8.940053768033609e-05,
      "loss": 0.0164,
      "step": 145
    },
    {
      "epoch": 0.584,
      "grad_norm": 0.04054705347154667,
      "learning_rate": 8.9184672866292e-05,
      "loss": 0.0085,
      "step": 146
    },
    {
      "epoch": 0.588,
      "grad_norm": 0.07360145966650167,
      "learning_rate": 8.896689824657372e-05,
      "loss": 0.0224,
      "step": 147
    },
    {
      "epoch": 0.592,
      "grad_norm": 0.005445824819210158,
      "learning_rate": 8.874722443520899e-05,
      "loss": 0.0011,
      "step": 148
    },
    {
      "epoch": 0.596,
      "grad_norm": 0.12447600517779284,
      "learning_rate": 8.852566213878947e-05,
      "loss": 0.0229,
      "step": 149
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.04235187506627168,
      "learning_rate": 8.83022221559489e-05,
      "loss": 0.0222,
      "step": 150
    },
    {
      "epoch": 0.604,
      "grad_norm": 0.005951434577412209,
      "learning_rate": 8.807691537683685e-05,
      "loss": 0.0015,
      "step": 151
    },
    {
      "epoch": 0.608,
      "grad_norm": 0.003541729111439659,
      "learning_rate": 8.784975278258783e-05,
      "loss": 0.0009,
      "step": 152
    },
    {
      "epoch": 0.612,
      "grad_norm": 0.05949780656802584,
      "learning_rate": 8.762074544478623e-05,
      "loss": 0.0058,
      "step": 153
    },
    {
      "epoch": 0.616,
      "grad_norm": 0.07257850393588734,
      "learning_rate": 8.73899045249266e-05,
      "loss": 0.0232,
      "step": 154
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.03250766262947475,
      "learning_rate": 8.715724127386972e-05,
      "loss": 0.0035,
      "step": 155
    },
    {
      "epoch": 0.624,
      "grad_norm": 0.017801959499827148,
      "learning_rate": 8.692276703129421e-05,
      "loss": 0.0022,
      "step": 156
    },
    {
      "epoch": 0.628,
      "grad_norm": 0.030093108040630024,
      "learning_rate": 8.668649322514382e-05,
      "loss": 0.0061,
      "step": 157
    },
    {
      "epoch": 0.632,
      "grad_norm": 0.05373501497480199,
      "learning_rate": 8.644843137107059e-05,
      "loss": 0.0068,
      "step": 158
    },
    {
      "epoch": 0.636,
      "grad_norm": 0.0022477256178792464,
      "learning_rate": 8.620859307187339e-05,
      "loss": 0.0005,
      "step": 159
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.010452967530973674,
      "learning_rate": 8.596699001693255e-05,
      "loss": 0.0034,
      "step": 160
    },
    {
      "epoch": 0.644,
      "grad_norm": 0.11564755181571154,
      "learning_rate": 8.572363398164017e-05,
      "loss": 0.0309,
      "step": 161
    },
    {
      "epoch": 0.648,
      "grad_norm": 0.017670822514925792,
      "learning_rate": 8.547853682682604e-05,
      "loss": 0.0032,
      "step": 162
    },
    {
      "epoch": 0.652,
      "grad_norm": 0.02008211048745366,
      "learning_rate": 8.523171049817974e-05,
      "loss": 0.0075,
      "step": 163
    },
    {
      "epoch": 0.656,
      "grad_norm": 0.07082781334762206,
      "learning_rate": 8.498316702566828e-05,
      "loss": 0.0267,
      "step": 164
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.06564438227587992,
      "learning_rate": 8.473291852294987e-05,
      "loss": 0.0207,
      "step": 165
    },
    {
      "epoch": 0.664,
      "grad_norm": 0.016621328678439844,
      "learning_rate": 8.44809771867835e-05,
      "loss": 0.0034,
      "step": 166
    },
    {
      "epoch": 0.668,
      "grad_norm": 0.07994176151133796,
      "learning_rate": 8.422735529643444e-05,
      "loss": 0.0174,
      "step": 167
    },
    {
      "epoch": 0.672,
      "grad_norm": 0.004495333605561698,
      "learning_rate": 8.397206521307584e-05,
      "loss": 0.0008,
      "step": 168
    },
    {
      "epoch": 0.676,
      "grad_norm": 0.0006016810735263043,
      "learning_rate": 8.371511937918616e-05,
      "loss": 0.0002,
      "step": 169
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.0605700090477828,
      "learning_rate": 8.345653031794292e-05,
      "loss": 0.0074,
      "step": 170
    },
    {
      "epoch": 0.684,
      "grad_norm": 0.07127712389727038,
      "learning_rate": 8.319631063261209e-05,
      "loss": 0.0303,
      "step": 171
    },
    {
      "epoch": 0.688,
      "grad_norm": 0.0038939147483986313,
      "learning_rate": 8.293447300593402e-05,
      "loss": 0.0011,
      "step": 172
    },
    {
      "epoch": 0.692,
      "grad_norm": 0.01167767178414831,
      "learning_rate": 8.267103019950529e-05,
      "loss": 0.0014,
      "step": 173
    },
    {
      "epoch": 0.696,
      "grad_norm": 0.03515059087305607,
      "learning_rate": 8.240599505315655e-05,
      "loss": 0.0041,
      "step": 174
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.006182524418573478,
      "learning_rate": 8.213938048432697e-05,
      "loss": 0.0017,
      "step": 175
    },
    {
      "epoch": 0.704,
      "grad_norm": 0.007679699822707157,
      "learning_rate": 8.18711994874345e-05,
      "loss": 0.0013,
      "step": 176
    },
    {
      "epoch": 0.708,
      "grad_norm": 0.01871388757789054,
      "learning_rate": 8.160146513324254e-05,
      "loss": 0.0026,
      "step": 177
    },
    {
      "epoch": 0.712,
      "grad_norm": 0.0029928386833481223,
      "learning_rate": 8.133019056822304e-05,
      "loss": 0.0004,
      "step": 178
    },
    {
      "epoch": 0.716,
      "grad_norm": 0.007848092747321551,
      "learning_rate": 8.105738901391552e-05,
      "loss": 0.0014,
      "step": 179
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.0006029176643066531,
      "learning_rate": 8.07830737662829e-05,
      "loss": 0.0001,
      "step": 180
    },
    {
      "epoch": 0.724,
      "grad_norm": 0.0029733206643785216,
      "learning_rate": 8.05072581950634e-05,
      "loss": 0.0006,
      "step": 181
    },
    {
      "epoch": 0.728,
      "grad_norm": 0.0912510175347323,
      "learning_rate": 8.022995574311876e-05,
      "loss": 0.036,
      "step": 182
    },
    {
      "epoch": 0.732,
      "grad_norm": 0.0009894161873779859,
      "learning_rate": 7.99511799257793e-05,
      "loss": 0.0002,
      "step": 183
    },
    {
      "epoch": 0.736,
      "grad_norm": 0.08289637667968551,
      "learning_rate": 7.967094433018508e-05,
      "loss": 0.0315,
      "step": 184
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.016248126015123914,
      "learning_rate": 7.938926261462366e-05,
      "loss": 0.0017,
      "step": 185
    },
    {
      "epoch": 0.744,
      "grad_norm": 0.009231571145695913,
      "learning_rate": 7.910614850786448e-05,
      "loss": 0.001,
      "step": 186
    },
    {
      "epoch": 0.748,
      "grad_norm": 0.0008493559079045681,
      "learning_rate": 7.882161580848967e-05,
      "loss": 0.0001,
      "step": 187
    },
    {
      "epoch": 0.752,
      "grad_norm": 0.004930762547579106,
      "learning_rate": 7.85356783842216e-05,
      "loss": 0.0008,
      "step": 188
    },
    {
      "epoch": 0.756,
      "grad_norm": 0.0008400133582123328,
      "learning_rate": 7.82483501712469e-05,
      "loss": 0.0002,
      "step": 189
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.0007546350821365371,
      "learning_rate": 7.795964517353735e-05,
      "loss": 0.0001,
      "step": 190
    },
    {
      "epoch": 0.764,
      "grad_norm": 0.05825647322081714,
      "learning_rate": 7.766957746216721e-05,
      "loss": 0.0039,
      "step": 191
    },
    {
      "epoch": 0.768,
      "grad_norm": 0.05363968854598717,
      "learning_rate": 7.737816117462752e-05,
      "loss": 0.0117,
      "step": 192
    },
    {
      "epoch": 0.772,
      "grad_norm": 0.000517209698196293,
      "learning_rate": 7.7085410514137e-05,
      "loss": 0.0001,
      "step": 193
    },
    {
      "epoch": 0.776,
      "grad_norm": 0.05506548824613901,
      "learning_rate": 7.679133974894983e-05,
      "loss": 0.0077,
      "step": 194
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.0036760042648846985,
      "learning_rate": 7.649596321166024e-05,
      "loss": 0.0007,
      "step": 195
    },
    {
      "epoch": 0.784,
      "grad_norm": 0.0030167045961558385,
      "learning_rate": 7.619929529850397e-05,
      "loss": 0.0005,
      "step": 196
    },
    {
      "epoch": 0.788,
      "grad_norm": 0.005077444121552554,
      "learning_rate": 7.590135046865651e-05,
      "loss": 0.0008,
      "step": 197
    },
    {
      "epoch": 0.792,
      "grad_norm": 0.0020544233779263474,
      "learning_rate": 7.560214324352858e-05,
      "loss": 0.0002,
      "step": 198
    },
    {
      "epoch": 0.796,
      "grad_norm": 0.0029148632262192184,
      "learning_rate": 7.530168820605818e-05,
      "loss": 0.0006,
      "step": 199
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.004366112040524964,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.0009,
      "step": 200
    },
    {
      "epoch": 0.804,
      "grad_norm": 0.0006096117832762582,
      "learning_rate": 7.469709332921155e-05,
      "loss": 0.0001,
      "step": 201
    },
    {
      "epoch": 0.808,
      "grad_norm": 0.023663828638711353,
      "learning_rate": 7.439298295693665e-05,
      "loss": 0.0028,
      "step": 202
    },
    {
      "epoch": 0.812,
      "grad_norm": 0.0023111594948878956,
      "learning_rate": 7.408768370508576e-05,
      "loss": 0.0005,
      "step": 203
    },
    {
      "epoch": 0.816,
      "grad_norm": 0.03223914922784597,
      "learning_rate": 7.378121045351378e-05,
      "loss": 0.0039,
      "step": 204
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.0008703539699525897,
      "learning_rate": 7.347357813929454e-05,
      "loss": 0.0002,
      "step": 205
    },
    {
      "epoch": 0.824,
      "grad_norm": 0.0009475723772251561,
      "learning_rate": 7.316480175599309e-05,
      "loss": 0.0002,
      "step": 206
    },
    {
      "epoch": 0.828,
      "grad_norm": 0.023251472889037693,
      "learning_rate": 7.285489635293472e-05,
      "loss": 0.0027,
      "step": 207
    },
    {
      "epoch": 0.832,
      "grad_norm": 0.01119996825794998,
      "learning_rate": 7.254387703447154e-05,
      "loss": 0.0007,
      "step": 208
    },
    {
      "epoch": 0.836,
      "grad_norm": 0.0003485484288381587,
      "learning_rate": 7.223175895924638e-05,
      "loss": 0.0001,
      "step": 209
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.0003139308974783195,
      "learning_rate": 7.191855733945387e-05,
      "loss": 0.0001,
      "step": 210
    },
    {
      "epoch": 0.844,
      "grad_norm": 0.0015420754062894142,
      "learning_rate": 7.160428744009912e-05,
      "loss": 0.0002,
      "step": 211
    },
    {
      "epoch": 0.848,
      "grad_norm": 0.00213000435193029,
      "learning_rate": 7.128896457825364e-05,
      "loss": 0.0003,
      "step": 212
    },
    {
      "epoch": 0.852,
      "grad_norm": 0.0004042691040131959,
      "learning_rate": 7.097260412230886e-05,
      "loss": 0.0001,
      "step": 213
    },
    {
      "epoch": 0.856,
      "grad_norm": 0.15683957412100924,
      "learning_rate": 7.06552214912271e-05,
      "loss": 0.0608,
      "step": 214
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.013664144246845906,
      "learning_rate": 7.033683215379002e-05,
      "loss": 0.0029,
      "step": 215
    },
    {
      "epoch": 0.864,
      "grad_norm": 0.0034511349647422733,
      "learning_rate": 7.001745162784477e-05,
      "loss": 0.0005,
      "step": 216
    },
    {
      "epoch": 0.868,
      "grad_norm": 0.0003853741025041897,
      "learning_rate": 6.969709547954756e-05,
      "loss": 0.0,
      "step": 217
    },
    {
      "epoch": 0.872,
      "grad_norm": 0.00427845558636739,
      "learning_rate": 6.937577932260515e-05,
      "loss": 0.0004,
      "step": 218
    },
    {
      "epoch": 0.876,
      "grad_norm": 0.1549607507505506,
      "learning_rate": 6.905351881751372e-05,
      "loss": 0.0154,
      "step": 219
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.05918463653979411,
      "learning_rate": 6.873032967079561e-05,
      "loss": 0.0025,
      "step": 220
    },
    {
      "epoch": 0.884,
      "grad_norm": 0.00033321038755560714,
      "learning_rate": 6.840622763423391e-05,
      "loss": 0.0001,
      "step": 221
    },
    {
      "epoch": 0.888,
      "grad_norm": 0.003448744788930108,
      "learning_rate": 6.808122850410461e-05,
      "loss": 0.0009,
      "step": 222
    },
    {
      "epoch": 0.892,
      "grad_norm": 0.0007754559505008236,
      "learning_rate": 6.775534812040685e-05,
      "loss": 0.0002,
      "step": 223
    },
    {
      "epoch": 0.896,
      "grad_norm": 0.0013795470144753014,
      "learning_rate": 6.742860236609077e-05,
      "loss": 0.0004,
      "step": 224
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.025465973861298608,
      "learning_rate": 6.710100716628344e-05,
      "loss": 0.0041,
      "step": 225
    },
    {
      "epoch": 0.904,
      "grad_norm": 0.006517107079396501,
      "learning_rate": 6.677257848751277e-05,
      "loss": 0.0005,
      "step": 226
    },
    {
      "epoch": 0.908,
      "grad_norm": 0.004050151767509786,
      "learning_rate": 6.644333233692916e-05,
      "loss": 0.0007,
      "step": 227
    },
    {
      "epoch": 0.912,
      "grad_norm": 0.0013584681424563913,
      "learning_rate": 6.611328476152557e-05,
      "loss": 0.0001,
      "step": 228
    },
    {
      "epoch": 0.916,
      "grad_norm": 0.001046626650340696,
      "learning_rate": 6.578245184735513e-05,
      "loss": 0.0002,
      "step": 229
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.009091093240854583,
      "learning_rate": 6.545084971874738e-05,
      "loss": 0.0008,
      "step": 230
    },
    {
      "epoch": 0.924,
      "grad_norm": 0.006724895544637462,
      "learning_rate": 6.511849453752223e-05,
      "loss": 0.0007,
      "step": 231
    },
    {
      "epoch": 0.928,
      "grad_norm": 0.06525922108919285,
      "learning_rate": 6.478540250220234e-05,
      "loss": 0.0065,
      "step": 232
    },
    {
      "epoch": 0.932,
      "grad_norm": 0.0014997783892137838,
      "learning_rate": 6.445158984722358e-05,
      "loss": 0.0001,
      "step": 233
    },
    {
      "epoch": 0.936,
      "grad_norm": 0.005432162729234838,
      "learning_rate": 6.411707284214384e-05,
      "loss": 0.0005,
      "step": 234
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.09119558967509243,
      "learning_rate": 6.378186779084995e-05,
      "loss": 0.006,
      "step": 235
    },
    {
      "epoch": 0.944,
      "grad_norm": 0.005540056822887767,
      "learning_rate": 6.344599103076329e-05,
      "loss": 0.0011,
      "step": 236
    },
    {
      "epoch": 0.948,
      "grad_norm": 0.03352585445209951,
      "learning_rate": 6.310945893204324e-05,
      "loss": 0.0033,
      "step": 237
    },
    {
      "epoch": 0.952,
      "grad_norm": 0.015964700325608334,
      "learning_rate": 6.277228789678953e-05,
      "loss": 0.0009,
      "step": 238
    },
    {
      "epoch": 0.956,
      "grad_norm": 0.00137230578160471,
      "learning_rate": 6.243449435824276e-05,
      "loss": 0.0003,
      "step": 239
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.002999042059181582,
      "learning_rate": 6.209609477998338e-05,
      "loss": 0.0006,
      "step": 240
    },
    {
      "epoch": 0.964,
      "grad_norm": 0.010209860946894876,
      "learning_rate": 6.17571056551295e-05,
      "loss": 0.0007,
      "step": 241
    },
    {
      "epoch": 0.968,
      "grad_norm": 0.0003757646759679207,
      "learning_rate": 6.141754350553279e-05,
      "loss": 0.0001,
      "step": 242
    },
    {
      "epoch": 0.972,
      "grad_norm": 0.22399087631263928,
      "learning_rate": 6.107742488097338e-05,
      "loss": 0.0255,
      "step": 243
    },
    {
      "epoch": 0.976,
      "grad_norm": 0.018998145435118054,
      "learning_rate": 6.073676635835317e-05,
      "loss": 0.0014,
      "step": 244
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.00014909238789021518,
      "learning_rate": 6.0395584540887963e-05,
      "loss": 0.0,
      "step": 245
    },
    {
      "epoch": 0.984,
      "grad_norm": 0.12285426845667556,
      "learning_rate": 6.005389605729824e-05,
      "loss": 0.0096,
      "step": 246
    },
    {
      "epoch": 0.988,
      "grad_norm": 0.002036581647055449,
      "learning_rate": 5.97117175609986e-05,
      "loss": 0.0003,
      "step": 247
    },
    {
      "epoch": 0.992,
      "grad_norm": 0.0031904800033398387,
      "learning_rate": 5.9369065729286245e-05,
      "loss": 0.0007,
      "step": 248
    },
    {
      "epoch": 0.996,
      "grad_norm": 0.002608368677591945,
      "learning_rate": 5.902595726252801e-05,
      "loss": 0.0003,
      "step": 249
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.00013851869547249697,
      "learning_rate": 5.868240888334653e-05,
      "loss": 0.0,
      "step": 250
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.3860524892807007,
      "eval_runtime": 0.3992,
      "eval_samples_per_second": 2.505,
      "eval_steps_per_second": 2.505,
      "step": 250
    },
    {
      "epoch": 1.004,
      "grad_norm": 0.0012897378494725019,
      "learning_rate": 5.833843733580512e-05,
      "loss": 0.0001,
      "step": 251
    },
    {
      "epoch": 1.008,
      "grad_norm": 0.0014222749689371925,
      "learning_rate": 5.799405938459175e-05,
      "loss": 0.0001,
      "step": 252
    },
    {
      "epoch": 1.012,
      "grad_norm": 0.08269507382389912,
      "learning_rate": 5.764929181420191e-05,
      "loss": 0.051,
      "step": 253
    },
    {
      "epoch": 1.016,
      "grad_norm": 0.0015490062317750953,
      "learning_rate": 5.730415142812059e-05,
      "loss": 0.0002,
      "step": 254
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.027204151052306086,
      "learning_rate": 5.695865504800327e-05,
      "loss": 0.0009,
      "step": 255
    },
    {
      "epoch": 1.024,
      "grad_norm": 0.0005416810673326349,
      "learning_rate": 5.661281951285613e-05,
      "loss": 0.0001,
      "step": 256
    },
    {
      "epoch": 1.028,
      "grad_norm": 0.0033640913603334654,
      "learning_rate": 5.6266661678215216e-05,
      "loss": 0.0003,
      "step": 257
    },
    {
      "epoch": 1.032,
      "grad_norm": 0.0030816533366147406,
      "learning_rate": 5.5920198415325064e-05,
      "loss": 0.0004,
      "step": 258
    },
    {
      "epoch": 1.036,
      "grad_norm": 0.00026894088434033746,
      "learning_rate": 5.557344661031627e-05,
      "loss": 0.0001,
      "step": 259
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.0005195845787863701,
      "learning_rate": 5.522642316338268e-05,
      "loss": 0.0001,
      "step": 260
    },
    {
      "epoch": 1.044,
      "grad_norm": 0.01600901905266607,
      "learning_rate": 5.487914498795747e-05,
      "loss": 0.0026,
      "step": 261
    },
    {
      "epoch": 1.048,
      "grad_norm": 0.002027752526150662,
      "learning_rate": 5.453162900988902e-05,
      "loss": 0.0003,
      "step": 262
    },
    {
      "epoch": 1.052,
      "grad_norm": 0.0007646703499707342,
      "learning_rate": 5.418389216661579e-05,
      "loss": 0.0001,
      "step": 263
    },
    {
      "epoch": 1.056,
      "grad_norm": 0.002401802660342866,
      "learning_rate": 5.383595140634093e-05,
      "loss": 0.0004,
      "step": 264
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.0006327852266118566,
      "learning_rate": 5.348782368720626e-05,
      "loss": 0.0001,
      "step": 265
    },
    {
      "epoch": 1.064,
      "grad_norm": 0.0008125363877252098,
      "learning_rate": 5.313952597646568e-05,
      "loss": 0.0001,
      "step": 266
    },
    {
      "epoch": 1.068,
      "grad_norm": 0.004729057296283605,
      "learning_rate": 5.279107524965819e-05,
      "loss": 0.0007,
      "step": 267
    },
    {
      "epoch": 1.072,
      "grad_norm": 0.0023936646886998808,
      "learning_rate": 5.244248848978067e-05,
      "loss": 0.0004,
      "step": 268
    },
    {
      "epoch": 1.076,
      "grad_norm": 0.006772549917315507,
      "learning_rate": 5.209378268645998e-05,
      "loss": 0.0009,
      "step": 269
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.0026089481169799624,
      "learning_rate": 5.174497483512506e-05,
      "loss": 0.0002,
      "step": 270
    },
    {
      "epoch": 1.084,
      "grad_norm": 0.09178542551364251,
      "learning_rate": 5.139608193617845e-05,
      "loss": 0.0071,
      "step": 271
    },
    {
      "epoch": 1.088,
      "grad_norm": 0.002242039574269421,
      "learning_rate": 5.104712099416785e-05,
      "loss": 0.0003,
      "step": 272
    },
    {
      "epoch": 1.092,
      "grad_norm": 0.05975063980807792,
      "learning_rate": 5.0698109016957274e-05,
      "loss": 0.0082,
      "step": 273
    },
    {
      "epoch": 1.096,
      "grad_norm": 0.02259349142578404,
      "learning_rate": 5.034906301489808e-05,
      "loss": 0.0015,
      "step": 274
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.0006097539721925062,
      "learning_rate": 5e-05,
      "loss": 0.0001,
      "step": 275
    },
    {
      "epoch": 1.104,
      "grad_norm": 0.0005818351561168097,
      "learning_rate": 4.965093698510193e-05,
      "loss": 0.0001,
      "step": 276
    },
    {
      "epoch": 1.108,
      "grad_norm": 0.16002050865805742,
      "learning_rate": 4.9301890983042744e-05,
      "loss": 0.0411,
      "step": 277
    },
    {
      "epoch": 1.112,
      "grad_norm": 0.004920141337228506,
      "learning_rate": 4.895287900583216e-05,
      "loss": 0.0007,
      "step": 278
    },
    {
      "epoch": 1.116,
      "grad_norm": 0.0028444560176627293,
      "learning_rate": 4.860391806382157e-05,
      "loss": 0.0004,
      "step": 279
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.000546566127549948,
      "learning_rate": 4.825502516487497e-05,
      "loss": 0.0,
      "step": 280
    },
    {
      "epoch": 1.124,
      "grad_norm": 0.0007815247630506817,
      "learning_rate": 4.790621731354003e-05,
      "loss": 0.0001,
      "step": 281
    },
    {
      "epoch": 1.1280000000000001,
      "grad_norm": 0.024077950638236954,
      "learning_rate": 4.755751151021934e-05,
      "loss": 0.004,
      "step": 282
    },
    {
      "epoch": 1.1320000000000001,
      "grad_norm": 0.0006944997754158855,
      "learning_rate": 4.720892475034181e-05,
      "loss": 0.0001,
      "step": 283
    },
    {
      "epoch": 1.1360000000000001,
      "grad_norm": 0.0004852945628594941,
      "learning_rate": 4.6860474023534335e-05,
      "loss": 0.0001,
      "step": 284
    },
    {
      "epoch": 1.1400000000000001,
      "grad_norm": 0.0004198274349150998,
      "learning_rate": 4.6512176312793736e-05,
      "loss": 0.0001,
      "step": 285
    },
    {
      "epoch": 1.144,
      "grad_norm": 0.02709978297166529,
      "learning_rate": 4.616404859365907e-05,
      "loss": 0.0013,
      "step": 286
    },
    {
      "epoch": 1.148,
      "grad_norm": 0.0008907847602245985,
      "learning_rate": 4.5816107833384234e-05,
      "loss": 0.0002,
      "step": 287
    },
    {
      "epoch": 1.152,
      "grad_norm": 0.0005685459987011892,
      "learning_rate": 4.5468370990111006e-05,
      "loss": 0.0001,
      "step": 288
    },
    {
      "epoch": 1.156,
      "grad_norm": 0.0011349301476918407,
      "learning_rate": 4.512085501204253e-05,
      "loss": 0.0002,
      "step": 289
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.002237969387449989,
      "learning_rate": 4.477357683661734e-05,
      "loss": 0.0003,
      "step": 290
    },
    {
      "epoch": 1.164,
      "grad_norm": 0.0006230805438419524,
      "learning_rate": 4.442655338968373e-05,
      "loss": 0.0001,
      "step": 291
    },
    {
      "epoch": 1.168,
      "grad_norm": 0.010244131125367525,
      "learning_rate": 4.407980158467495e-05,
      "loss": 0.0011,
      "step": 292
    },
    {
      "epoch": 1.172,
      "grad_norm": 0.01013947642327459,
      "learning_rate": 4.373333832178478e-05,
      "loss": 0.0009,
      "step": 293
    },
    {
      "epoch": 1.176,
      "grad_norm": 0.0057915673829967585,
      "learning_rate": 4.3387180487143876e-05,
      "loss": 0.0004,
      "step": 294
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.0010499193174197956,
      "learning_rate": 4.3041344951996746e-05,
      "loss": 0.0002,
      "step": 295
    },
    {
      "epoch": 1.184,
      "grad_norm": 0.000983205809325645,
      "learning_rate": 4.269584857187943e-05,
      "loss": 0.0001,
      "step": 296
    },
    {
      "epoch": 1.188,
      "grad_norm": 0.00307751200565114,
      "learning_rate": 4.23507081857981e-05,
      "loss": 0.0002,
      "step": 297
    },
    {
      "epoch": 1.192,
      "grad_norm": 0.0007766191943371171,
      "learning_rate": 4.2005940615408264e-05,
      "loss": 0.0002,
      "step": 298
    },
    {
      "epoch": 1.196,
      "grad_norm": 0.010789969966789846,
      "learning_rate": 4.166156266419489e-05,
      "loss": 0.0014,
      "step": 299
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.0022487238670015537,
      "learning_rate": 4.131759111665349e-05,
      "loss": 0.0003,
      "step": 300
    },
    {
      "epoch": 1.204,
      "grad_norm": 0.0018331589413666599,
      "learning_rate": 4.0974042737472006e-05,
      "loss": 0.0003,
      "step": 301
    },
    {
      "epoch": 1.208,
      "grad_norm": 0.009835403161860016,
      "learning_rate": 4.063093427071376e-05,
      "loss": 0.0008,
      "step": 302
    },
    {
      "epoch": 1.212,
      "grad_norm": 0.00040525918355532143,
      "learning_rate": 4.028828243900141e-05,
      "loss": 0.0001,
      "step": 303
    },
    {
      "epoch": 1.216,
      "grad_norm": 0.02139012926538603,
      "learning_rate": 3.9946103942701777e-05,
      "loss": 0.0013,
      "step": 304
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.0010336741767038196,
      "learning_rate": 3.960441545911204e-05,
      "loss": 0.0002,
      "step": 305
    },
    {
      "epoch": 1.224,
      "grad_norm": 0.0002629189777626845,
      "learning_rate": 3.926323364164684e-05,
      "loss": 0.0,
      "step": 306
    },
    {
      "epoch": 1.228,
      "grad_norm": 0.002336433904496914,
      "learning_rate": 3.892257511902664e-05,
      "loss": 0.0003,
      "step": 307
    },
    {
      "epoch": 1.232,
      "grad_norm": 0.0007155120624926272,
      "learning_rate": 3.858245649446721e-05,
      "loss": 0.0001,
      "step": 308
    },
    {
      "epoch": 1.236,
      "grad_norm": 0.0009616345052158773,
      "learning_rate": 3.82428943448705e-05,
      "loss": 0.0002,
      "step": 309
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.001087854753517169,
      "learning_rate": 3.790390522001662e-05,
      "loss": 0.0002,
      "step": 310
    },
    {
      "epoch": 1.244,
      "grad_norm": 0.003928812071717386,
      "learning_rate": 3.756550564175727e-05,
      "loss": 0.0006,
      "step": 311
    },
    {
      "epoch": 1.248,
      "grad_norm": 0.0006688944775370527,
      "learning_rate": 3.7227712103210486e-05,
      "loss": 0.0001,
      "step": 312
    },
    {
      "epoch": 1.252,
      "grad_norm": 0.02021068546553723,
      "learning_rate": 3.6890541067956776e-05,
      "loss": 0.0014,
      "step": 313
    },
    {
      "epoch": 1.256,
      "grad_norm": 0.008953151216484862,
      "learning_rate": 3.655400896923672e-05,
      "loss": 0.0008,
      "step": 314
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.0019131675996281415,
      "learning_rate": 3.6218132209150045e-05,
      "loss": 0.0002,
      "step": 315
    },
    {
      "epoch": 1.264,
      "grad_norm": 0.0009426544743443118,
      "learning_rate": 3.588292715785617e-05,
      "loss": 0.0001,
      "step": 316
    },
    {
      "epoch": 1.268,
      "grad_norm": 0.0009087154732827485,
      "learning_rate": 3.554841015277641e-05,
      "loss": 0.0001,
      "step": 317
    },
    {
      "epoch": 1.272,
      "grad_norm": 0.00010799511492406229,
      "learning_rate": 3.5214597497797684e-05,
      "loss": 0.0,
      "step": 318
    },
    {
      "epoch": 1.276,
      "grad_norm": 0.0007772691400646037,
      "learning_rate": 3.488150546247778e-05,
      "loss": 0.0001,
      "step": 319
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.00808278270839332,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 0.0005,
      "step": 320
    },
    {
      "epoch": 1.284,
      "grad_norm": 0.00016073397319208503,
      "learning_rate": 3.4217548152644885e-05,
      "loss": 0.0,
      "step": 321
    },
    {
      "epoch": 1.288,
      "grad_norm": 0.0010341775273125917,
      "learning_rate": 3.388671523847445e-05,
      "loss": 0.0002,
      "step": 322
    },
    {
      "epoch": 1.292,
      "grad_norm": 0.00014697215996962055,
      "learning_rate": 3.355666766307084e-05,
      "loss": 0.0,
      "step": 323
    },
    {
      "epoch": 1.296,
      "grad_norm": 0.000991184058303414,
      "learning_rate": 3.322742151248725e-05,
      "loss": 0.0001,
      "step": 324
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.01901372117338796,
      "learning_rate": 3.289899283371657e-05,
      "loss": 0.0014,
      "step": 325
    },
    {
      "epoch": 1.304,
      "grad_norm": 0.001113881226145738,
      "learning_rate": 3.257139763390925e-05,
      "loss": 0.0001,
      "step": 326
    },
    {
      "epoch": 1.308,
      "grad_norm": 5.36684502601617e-05,
      "learning_rate": 3.224465187959316e-05,
      "loss": 0.0,
      "step": 327
    },
    {
      "epoch": 1.312,
      "grad_norm": 0.0020994303711139514,
      "learning_rate": 3.1918771495895396e-05,
      "loss": 0.0002,
      "step": 328
    },
    {
      "epoch": 1.316,
      "grad_norm": 0.008867258550497624,
      "learning_rate": 3.1593772365766105e-05,
      "loss": 0.0008,
      "step": 329
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.0026960740913966564,
      "learning_rate": 3.12696703292044e-05,
      "loss": 0.0003,
      "step": 330
    },
    {
      "epoch": 1.324,
      "grad_norm": 0.00960485868038758,
      "learning_rate": 3.09464811824863e-05,
      "loss": 0.0014,
      "step": 331
    },
    {
      "epoch": 1.328,
      "grad_norm": 0.0009125991241853596,
      "learning_rate": 3.062422067739485e-05,
      "loss": 0.0002,
      "step": 332
    },
    {
      "epoch": 1.332,
      "grad_norm": 0.07428456950208079,
      "learning_rate": 3.0302904520452447e-05,
      "loss": 0.0098,
      "step": 333
    },
    {
      "epoch": 1.336,
      "grad_norm": 0.02043274386822593,
      "learning_rate": 2.9982548372155263e-05,
      "loss": 0.0043,
      "step": 334
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.0010665717207919772,
      "learning_rate": 2.9663167846209998e-05,
      "loss": 0.0001,
      "step": 335
    },
    {
      "epoch": 1.3439999999999999,
      "grad_norm": 0.0001529066592978424,
      "learning_rate": 2.934477850877292e-05,
      "loss": 0.0,
      "step": 336
    },
    {
      "epoch": 1.3479999999999999,
      "grad_norm": 0.0020871547212612195,
      "learning_rate": 2.9027395877691144e-05,
      "loss": 0.0002,
      "step": 337
    },
    {
      "epoch": 1.3519999999999999,
      "grad_norm": 0.006276589864837327,
      "learning_rate": 2.8711035421746367e-05,
      "loss": 0.0005,
      "step": 338
    },
    {
      "epoch": 1.3559999999999999,
      "grad_norm": 0.0012169805558767542,
      "learning_rate": 2.8395712559900877e-05,
      "loss": 0.0001,
      "step": 339
    },
    {
      "epoch": 1.3599999999999999,
      "grad_norm": 0.0004225868160053743,
      "learning_rate": 2.8081442660546125e-05,
      "loss": 0.0001,
      "step": 340
    },
    {
      "epoch": 1.3639999999999999,
      "grad_norm": 0.002437420607153,
      "learning_rate": 2.776824104075364e-05,
      "loss": 0.0002,
      "step": 341
    },
    {
      "epoch": 1.3679999999999999,
      "grad_norm": 0.0002966468591098742,
      "learning_rate": 2.7456122965528475e-05,
      "loss": 0.0,
      "step": 342
    },
    {
      "epoch": 1.3719999999999999,
      "grad_norm": 0.0010532776746726787,
      "learning_rate": 2.7145103647065308e-05,
      "loss": 0.0001,
      "step": 343
    },
    {
      "epoch": 1.376,
      "grad_norm": 0.0006290539012404299,
      "learning_rate": 2.6835198244006927e-05,
      "loss": 0.0001,
      "step": 344
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.5932717856891564,
      "learning_rate": 2.6526421860705473e-05,
      "loss": 0.0587,
      "step": 345
    },
    {
      "epoch": 1.384,
      "grad_norm": 0.0014082511362570247,
      "learning_rate": 2.6218789546486234e-05,
      "loss": 0.0002,
      "step": 346
    },
    {
      "epoch": 1.388,
      "grad_norm": 0.005822416981433907,
      "learning_rate": 2.591231629491423e-05,
      "loss": 0.0005,
      "step": 347
    },
    {
      "epoch": 1.392,
      "grad_norm": 0.00020585322555278054,
      "learning_rate": 2.560701704306336e-05,
      "loss": 0.0,
      "step": 348
    },
    {
      "epoch": 1.396,
      "grad_norm": 0.0016532674147848431,
      "learning_rate": 2.5302906670788462e-05,
      "loss": 0.0001,
      "step": 349
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.0007241259965886648,
      "learning_rate": 2.500000000000001e-05,
      "loss": 0.0001,
      "step": 350
    },
    {
      "epoch": 1.404,
      "grad_norm": 0.001293894476063566,
      "learning_rate": 2.469831179394182e-05,
      "loss": 0.0001,
      "step": 351
    },
    {
      "epoch": 1.408,
      "grad_norm": 0.0004067955611343988,
      "learning_rate": 2.4397856756471432e-05,
      "loss": 0.0001,
      "step": 352
    },
    {
      "epoch": 1.412,
      "grad_norm": 0.0017044747117935404,
      "learning_rate": 2.4098649531343497e-05,
      "loss": 0.0003,
      "step": 353
    },
    {
      "epoch": 1.416,
      "grad_norm": 0.02762583080860224,
      "learning_rate": 2.3800704701496053e-05,
      "loss": 0.0015,
      "step": 354
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.0017951321136892772,
      "learning_rate": 2.350403678833976e-05,
      "loss": 0.0005,
      "step": 355
    },
    {
      "epoch": 1.424,
      "grad_norm": 0.0002868509794478107,
      "learning_rate": 2.3208660251050158e-05,
      "loss": 0.0001,
      "step": 356
    },
    {
      "epoch": 1.428,
      "grad_norm": 0.05254862600119911,
      "learning_rate": 2.2914589485863014e-05,
      "loss": 0.006,
      "step": 357
    },
    {
      "epoch": 1.432,
      "grad_norm": 0.0010504212204218654,
      "learning_rate": 2.2621838825372493e-05,
      "loss": 0.0001,
      "step": 358
    },
    {
      "epoch": 1.436,
      "grad_norm": 0.000967047271408532,
      "learning_rate": 2.23304225378328e-05,
      "loss": 0.0002,
      "step": 359
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.028882234240049383,
      "learning_rate": 2.2040354826462668e-05,
      "loss": 0.0022,
      "step": 360
    },
    {
      "epoch": 1.444,
      "grad_norm": 0.012630976215518492,
      "learning_rate": 2.1751649828753106e-05,
      "loss": 0.0007,
      "step": 361
    },
    {
      "epoch": 1.448,
      "grad_norm": 0.00011242375440854577,
      "learning_rate": 2.1464321615778422e-05,
      "loss": 0.0,
      "step": 362
    },
    {
      "epoch": 1.452,
      "grad_norm": 0.0012076070593356998,
      "learning_rate": 2.117838419151034e-05,
      "loss": 0.0002,
      "step": 363
    },
    {
      "epoch": 1.456,
      "grad_norm": 0.017855030623970147,
      "learning_rate": 2.0893851492135537e-05,
      "loss": 0.0018,
      "step": 364
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.0003117116616772664,
      "learning_rate": 2.061073738537635e-05,
      "loss": 0.0001,
      "step": 365
    },
    {
      "epoch": 1.464,
      "grad_norm": 0.0007254915284999582,
      "learning_rate": 2.0329055669814934e-05,
      "loss": 0.0001,
      "step": 366
    },
    {
      "epoch": 1.468,
      "grad_norm": 0.12932836634121367,
      "learning_rate": 2.0048820074220715e-05,
      "loss": 0.0148,
      "step": 367
    },
    {
      "epoch": 1.472,
      "grad_norm": 0.007646417837751074,
      "learning_rate": 1.977004425688126e-05,
      "loss": 0.0006,
      "step": 368
    },
    {
      "epoch": 1.476,
      "grad_norm": 0.022832970423344517,
      "learning_rate": 1.9492741804936622e-05,
      "loss": 0.002,
      "step": 369
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.00036420061704775823,
      "learning_rate": 1.9216926233717085e-05,
      "loss": 0.0001,
      "step": 370
    },
    {
      "epoch": 1.484,
      "grad_norm": 0.0013653058489299184,
      "learning_rate": 1.8942610986084486e-05,
      "loss": 0.0003,
      "step": 371
    },
    {
      "epoch": 1.488,
      "grad_norm": 0.0023013267706678154,
      "learning_rate": 1.866980943177699e-05,
      "loss": 0.0002,
      "step": 372
    },
    {
      "epoch": 1.492,
      "grad_norm": 0.0002498736387627182,
      "learning_rate": 1.8398534866757454e-05,
      "loss": 0.0,
      "step": 373
    },
    {
      "epoch": 1.496,
      "grad_norm": 0.017608269992493826,
      "learning_rate": 1.8128800512565513e-05,
      "loss": 0.0021,
      "step": 374
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.00045515011172825645,
      "learning_rate": 1.7860619515673033e-05,
      "loss": 0.0,
      "step": 375
    },
    {
      "epoch": 1.504,
      "grad_norm": 0.16447839725980104,
      "learning_rate": 1.7594004946843456e-05,
      "loss": 0.0091,
      "step": 376
    },
    {
      "epoch": 1.508,
      "grad_norm": 0.04246064300189182,
      "learning_rate": 1.7328969800494726e-05,
      "loss": 0.0036,
      "step": 377
    },
    {
      "epoch": 1.512,
      "grad_norm": 0.0008479062947195134,
      "learning_rate": 1.7065526994065973e-05,
      "loss": 0.0002,
      "step": 378
    },
    {
      "epoch": 1.516,
      "grad_norm": 0.0006798669051807401,
      "learning_rate": 1.680368936738792e-05,
      "loss": 0.0002,
      "step": 379
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.004065270303916747,
      "learning_rate": 1.6543469682057106e-05,
      "loss": 0.0006,
      "step": 380
    },
    {
      "epoch": 1.524,
      "grad_norm": 0.002878477109135299,
      "learning_rate": 1.6284880620813848e-05,
      "loss": 0.0004,
      "step": 381
    },
    {
      "epoch": 1.528,
      "grad_norm": 0.00032536588425509356,
      "learning_rate": 1.602793478692419e-05,
      "loss": 0.0001,
      "step": 382
    },
    {
      "epoch": 1.532,
      "grad_norm": 0.0006798847551303467,
      "learning_rate": 1.5772644703565565e-05,
      "loss": 0.0001,
      "step": 383
    },
    {
      "epoch": 1.536,
      "grad_norm": 0.000209164785652081,
      "learning_rate": 1.551902281321651e-05,
      "loss": 0.0,
      "step": 384
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.0006440547836359067,
      "learning_rate": 1.526708147705013e-05,
      "loss": 0.0001,
      "step": 385
    },
    {
      "epoch": 1.544,
      "grad_norm": 0.001240720438368086,
      "learning_rate": 1.5016832974331724e-05,
      "loss": 0.0003,
      "step": 386
    },
    {
      "epoch": 1.548,
      "grad_norm": 0.0009680784239302523,
      "learning_rate": 1.4768289501820265e-05,
      "loss": 0.0002,
      "step": 387
    },
    {
      "epoch": 1.552,
      "grad_norm": 0.0003246862345354507,
      "learning_rate": 1.4521463173173965e-05,
      "loss": 0.0001,
      "step": 388
    },
    {
      "epoch": 1.556,
      "grad_norm": 0.00016800551667232755,
      "learning_rate": 1.4276366018359844e-05,
      "loss": 0.0,
      "step": 389
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.0008258487655385937,
      "learning_rate": 1.4033009983067452e-05,
      "loss": 0.0002,
      "step": 390
    },
    {
      "epoch": 1.564,
      "grad_norm": 0.0010973522117693247,
      "learning_rate": 1.3791406928126638e-05,
      "loss": 0.0001,
      "step": 391
    },
    {
      "epoch": 1.568,
      "grad_norm": 0.0010675610999394077,
      "learning_rate": 1.3551568628929434e-05,
      "loss": 0.0001,
      "step": 392
    },
    {
      "epoch": 1.572,
      "grad_norm": 0.0006597386198841764,
      "learning_rate": 1.3313506774856177e-05,
      "loss": 0.0002,
      "step": 393
    },
    {
      "epoch": 1.576,
      "grad_norm": 0.0003231437783471342,
      "learning_rate": 1.3077232968705805e-05,
      "loss": 0.0,
      "step": 394
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.0016008151800119834,
      "learning_rate": 1.2842758726130283e-05,
      "loss": 0.0003,
      "step": 395
    },
    {
      "epoch": 1.584,
      "grad_norm": 0.0012018868664600438,
      "learning_rate": 1.2610095475073414e-05,
      "loss": 0.0002,
      "step": 396
    },
    {
      "epoch": 1.588,
      "grad_norm": 0.002922858301019376,
      "learning_rate": 1.2379254555213788e-05,
      "loss": 0.0005,
      "step": 397
    },
    {
      "epoch": 1.592,
      "grad_norm": 0.0005116916420474867,
      "learning_rate": 1.2150247217412186e-05,
      "loss": 0.0001,
      "step": 398
    },
    {
      "epoch": 1.596,
      "grad_norm": 0.0007843529021223258,
      "learning_rate": 1.1923084623163172e-05,
      "loss": 0.0002,
      "step": 399
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.00027688649093886346,
      "learning_rate": 1.1697777844051105e-05,
      "loss": 0.0,
      "step": 400
    },
    {
      "epoch": 1.604,
      "grad_norm": 0.00034691612949939646,
      "learning_rate": 1.1474337861210543e-05,
      "loss": 0.0001,
      "step": 401
    },
    {
      "epoch": 1.608,
      "grad_norm": 0.0001941425459437523,
      "learning_rate": 1.1252775564791024e-05,
      "loss": 0.0,
      "step": 402
    },
    {
      "epoch": 1.612,
      "grad_norm": 0.0009222769953851323,
      "learning_rate": 1.1033101753426283e-05,
      "loss": 0.0001,
      "step": 403
    },
    {
      "epoch": 1.616,
      "grad_norm": 0.0012688540687766055,
      "learning_rate": 1.0815327133708015e-05,
      "loss": 0.0002,
      "step": 404
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.00011921359762988645,
      "learning_rate": 1.0599462319663905e-05,
      "loss": 0.0,
      "step": 405
    },
    {
      "epoch": 1.624,
      "grad_norm": 0.00023221320702237518,
      "learning_rate": 1.0385517832240471e-05,
      "loss": 0.0,
      "step": 406
    },
    {
      "epoch": 1.6280000000000001,
      "grad_norm": 0.0003007915332286273,
      "learning_rate": 1.0173504098790187e-05,
      "loss": 0.0001,
      "step": 407
    },
    {
      "epoch": 1.6320000000000001,
      "grad_norm": 0.0027946879346035674,
      "learning_rate": 9.963431452563332e-06,
      "loss": 0.0003,
      "step": 408
    },
    {
      "epoch": 1.6360000000000001,
      "grad_norm": 0.0005725182515345473,
      "learning_rate": 9.755310132204298e-06,
      "loss": 0.0001,
      "step": 409
    },
    {
      "epoch": 1.6400000000000001,
      "grad_norm": 0.0030550767480791497,
      "learning_rate": 9.549150281252633e-06,
      "loss": 0.0006,
      "step": 410
    },
    {
      "epoch": 1.6440000000000001,
      "grad_norm": 0.0018643410552353268,
      "learning_rate": 9.344961947648623e-06,
      "loss": 0.0003,
      "step": 411
    },
    {
      "epoch": 1.6480000000000001,
      "grad_norm": 0.00016720194425784736,
      "learning_rate": 9.142755083243576e-06,
      "loss": 0.0,
      "step": 412
    },
    {
      "epoch": 1.6520000000000001,
      "grad_norm": 0.0003728235777896952,
      "learning_rate": 8.9425395433148e-06,
      "loss": 0.0001,
      "step": 413
    },
    {
      "epoch": 1.6560000000000001,
      "grad_norm": 0.001785954016586295,
      "learning_rate": 8.744325086085248e-06,
      "loss": 0.0002,
      "step": 414
    },
    {
      "epoch": 1.6600000000000001,
      "grad_norm": 0.0010621159399172206,
      "learning_rate": 8.548121372247918e-06,
      "loss": 0.0001,
      "step": 415
    },
    {
      "epoch": 1.6640000000000001,
      "grad_norm": 0.0012625962686358663,
      "learning_rate": 8.353937964495029e-06,
      "loss": 0.0002,
      "step": 416
    },
    {
      "epoch": 1.6680000000000001,
      "grad_norm": 0.14235561129405266,
      "learning_rate": 8.16178432705192e-06,
      "loss": 0.0093,
      "step": 417
    },
    {
      "epoch": 1.6720000000000002,
      "grad_norm": 0.0008384034296007557,
      "learning_rate": 7.971669825215788e-06,
      "loss": 0.0002,
      "step": 418
    },
    {
      "epoch": 1.6760000000000002,
      "grad_norm": 0.00043048513028285274,
      "learning_rate": 7.783603724899257e-06,
      "loss": 0.0001,
      "step": 419
    },
    {
      "epoch": 1.6800000000000002,
      "grad_norm": 0.0005196799637667099,
      "learning_rate": 7.597595192178702e-06,
      "loss": 0.0001,
      "step": 420
    },
    {
      "epoch": 1.6840000000000002,
      "grad_norm": 0.0011757663197855717,
      "learning_rate": 7.413653292847617e-06,
      "loss": 0.0002,
      "step": 421
    },
    {
      "epoch": 1.688,
      "grad_norm": 0.0007305995469050543,
      "learning_rate": 7.2317869919746705e-06,
      "loss": 0.0002,
      "step": 422
    },
    {
      "epoch": 1.692,
      "grad_norm": 0.00016462438193801725,
      "learning_rate": 7.052005153466779e-06,
      "loss": 0.0,
      "step": 423
    },
    {
      "epoch": 1.696,
      "grad_norm": 0.00029412944565010646,
      "learning_rate": 6.874316539637127e-06,
      "loss": 0.0,
      "step": 424
    },
    {
      "epoch": 1.7,
      "grad_norm": 0.0031592277256137047,
      "learning_rate": 6.698729810778065e-06,
      "loss": 0.0006,
      "step": 425
    },
    {
      "epoch": 1.704,
      "grad_norm": 0.0010701566254851449,
      "learning_rate": 6.52525352473905e-06,
      "loss": 0.0001,
      "step": 426
    },
    {
      "epoch": 1.708,
      "grad_norm": 9.139042738159152e-05,
      "learning_rate": 6.353896136509524e-06,
      "loss": 0.0,
      "step": 427
    },
    {
      "epoch": 1.712,
      "grad_norm": 0.00014677141426081507,
      "learning_rate": 6.184665997806832e-06,
      "loss": 0.0,
      "step": 428
    },
    {
      "epoch": 1.716,
      "grad_norm": 0.00025525657408586173,
      "learning_rate": 6.017571356669183e-06,
      "loss": 0.0001,
      "step": 429
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.0003314172552667404,
      "learning_rate": 5.852620357053651e-06,
      "loss": 0.0001,
      "step": 430
    },
    {
      "epoch": 1.724,
      "grad_norm": 0.0019127977098360484,
      "learning_rate": 5.689821038439263e-06,
      "loss": 0.0003,
      "step": 431
    },
    {
      "epoch": 1.728,
      "grad_norm": 0.0013027787657261668,
      "learning_rate": 5.529181335435124e-06,
      "loss": 0.0004,
      "step": 432
    },
    {
      "epoch": 1.732,
      "grad_norm": 0.004100115497888855,
      "learning_rate": 5.370709077393721e-06,
      "loss": 0.0005,
      "step": 433
    },
    {
      "epoch": 1.736,
      "grad_norm": 0.00039457553002234983,
      "learning_rate": 5.214411988029355e-06,
      "loss": 0.0,
      "step": 434
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.0005762472494917307,
      "learning_rate": 5.060297685041659e-06,
      "loss": 0.0002,
      "step": 435
    },
    {
      "epoch": 1.744,
      "grad_norm": 0.0020538289732796727,
      "learning_rate": 4.908373679744316e-06,
      "loss": 0.0005,
      "step": 436
    },
    {
      "epoch": 1.748,
      "grad_norm": 0.003344409423183302,
      "learning_rate": 4.758647376699032e-06,
      "loss": 0.0005,
      "step": 437
    },
    {
      "epoch": 1.752,
      "grad_norm": 0.00036404148573838073,
      "learning_rate": 4.611126073354571e-06,
      "loss": 0.0001,
      "step": 438
    },
    {
      "epoch": 1.756,
      "grad_norm": 0.0008138688033199093,
      "learning_rate": 4.465816959691149e-06,
      "loss": 0.0002,
      "step": 439
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.004314126629255425,
      "learning_rate": 4.322727117869951e-06,
      "loss": 0.0009,
      "step": 440
    },
    {
      "epoch": 1.764,
      "grad_norm": 0.0016494410034327352,
      "learning_rate": 4.181863521888019e-06,
      "loss": 0.0003,
      "step": 441
    },
    {
      "epoch": 1.768,
      "grad_norm": 0.00011057454546572829,
      "learning_rate": 4.043233037238281e-06,
      "loss": 0.0,
      "step": 442
    },
    {
      "epoch": 1.772,
      "grad_norm": 0.001890762581192377,
      "learning_rate": 3.90684242057498e-06,
      "loss": 0.0004,
      "step": 443
    },
    {
      "epoch": 1.776,
      "grad_norm": 0.003548286973224894,
      "learning_rate": 3.772698319384349e-06,
      "loss": 0.0005,
      "step": 444
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.061014194731111415,
      "learning_rate": 3.6408072716606346e-06,
      "loss": 0.0028,
      "step": 445
    },
    {
      "epoch": 1.784,
      "grad_norm": 0.001929609007443394,
      "learning_rate": 3.511175705587433e-06,
      "loss": 0.0004,
      "step": 446
    },
    {
      "epoch": 1.788,
      "grad_norm": 0.0013446004970147306,
      "learning_rate": 3.3838099392243916e-06,
      "loss": 0.0001,
      "step": 447
    },
    {
      "epoch": 1.792,
      "grad_norm": 0.001447852724568891,
      "learning_rate": 3.258716180199278e-06,
      "loss": 0.0002,
      "step": 448
    },
    {
      "epoch": 1.796,
      "grad_norm": 0.0001229407069279758,
      "learning_rate": 3.1359005254054273e-06,
      "loss": 0.0,
      "step": 449
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.0013982650814133373,
      "learning_rate": 3.0153689607045845e-06,
      "loss": 0.0002,
      "step": 450
    },
    {
      "epoch": 1.804,
      "grad_norm": 0.0016975874138212552,
      "learning_rate": 2.8971273606351658e-06,
      "loss": 0.0003,
      "step": 451
    },
    {
      "epoch": 1.808,
      "grad_norm": 0.0005882169927293892,
      "learning_rate": 2.7811814881259503e-06,
      "loss": 0.0002,
      "step": 452
    },
    {
      "epoch": 1.812,
      "grad_norm": 0.014926744545613373,
      "learning_rate": 2.667536994215186e-06,
      "loss": 0.0008,
      "step": 453
    },
    {
      "epoch": 1.8159999999999998,
      "grad_norm": 0.0019071593879477857,
      "learning_rate": 2.5561994177751737e-06,
      "loss": 0.0002,
      "step": 454
    },
    {
      "epoch": 1.8199999999999998,
      "grad_norm": 0.005540325563473482,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 0.0013,
      "step": 455
    },
    {
      "epoch": 1.8239999999999998,
      "grad_norm": 0.0022672906118384726,
      "learning_rate": 2.340466610352654e-06,
      "loss": 0.0005,
      "step": 456
    },
    {
      "epoch": 1.8279999999999998,
      "grad_norm": 0.0013242104112468145,
      "learning_rate": 2.2360818938828187e-06,
      "loss": 0.0001,
      "step": 457
    },
    {
      "epoch": 1.8319999999999999,
      "grad_norm": 0.003681568518013968,
      "learning_rate": 2.134025123396638e-06,
      "loss": 0.0008,
      "step": 458
    },
    {
      "epoch": 1.8359999999999999,
      "grad_norm": 0.0008894729391345621,
      "learning_rate": 2.0343012729971243e-06,
      "loss": 0.0001,
      "step": 459
    },
    {
      "epoch": 1.8399999999999999,
      "grad_norm": 0.0010339730166608447,
      "learning_rate": 1.9369152030840556e-06,
      "loss": 0.0002,
      "step": 460
    },
    {
      "epoch": 1.8439999999999999,
      "grad_norm": 0.0018192332431137726,
      "learning_rate": 1.841871660117095e-06,
      "loss": 0.0001,
      "step": 461
    },
    {
      "epoch": 1.8479999999999999,
      "grad_norm": 0.00023824264005219514,
      "learning_rate": 1.7491752763844293e-06,
      "loss": 0.0,
      "step": 462
    },
    {
      "epoch": 1.8519999999999999,
      "grad_norm": 0.0017042925583684405,
      "learning_rate": 1.6588305697770313e-06,
      "loss": 0.0003,
      "step": 463
    },
    {
      "epoch": 1.8559999999999999,
      "grad_norm": 0.0019443299757462937,
      "learning_rate": 1.5708419435684462e-06,
      "loss": 0.0003,
      "step": 464
    },
    {
      "epoch": 1.8599999999999999,
      "grad_norm": 0.00015146844368129546,
      "learning_rate": 1.4852136862001764e-06,
      "loss": 0.0,
      "step": 465
    },
    {
      "epoch": 1.8639999999999999,
      "grad_norm": 0.10571800851140933,
      "learning_rate": 1.4019499710726913e-06,
      "loss": 0.0078,
      "step": 466
    },
    {
      "epoch": 1.8679999999999999,
      "grad_norm": 0.0004953841170492576,
      "learning_rate": 1.3210548563419856e-06,
      "loss": 0.0001,
      "step": 467
    },
    {
      "epoch": 1.8719999999999999,
      "grad_norm": 0.004082453004537698,
      "learning_rate": 1.2425322847218368e-06,
      "loss": 0.001,
      "step": 468
    },
    {
      "epoch": 1.876,
      "grad_norm": 0.0001458282006386659,
      "learning_rate": 1.166386083291604e-06,
      "loss": 0.0,
      "step": 469
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.0005004550846422535,
      "learning_rate": 1.0926199633097157e-06,
      "loss": 0.0001,
      "step": 470
    },
    {
      "epoch": 1.884,
      "grad_norm": 0.0032491478194401754,
      "learning_rate": 1.0212375200327973e-06,
      "loss": 0.0003,
      "step": 471
    },
    {
      "epoch": 1.888,
      "grad_norm": 0.000750771783136979,
      "learning_rate": 9.522422325404235e-07,
      "loss": 0.0001,
      "step": 472
    },
    {
      "epoch": 1.892,
      "grad_norm": 0.0002160044713151201,
      "learning_rate": 8.856374635655695e-07,
      "loss": 0.0,
      "step": 473
    },
    {
      "epoch": 1.896,
      "grad_norm": 0.0038119756721294263,
      "learning_rate": 8.214264593307098e-07,
      "loss": 0.001,
      "step": 474
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.0005707442440346829,
      "learning_rate": 7.596123493895991e-07,
      "loss": 0.0001,
      "step": 475
    },
    {
      "epoch": 1.904,
      "grad_norm": 0.0004801129433504118,
      "learning_rate": 7.001981464747565e-07,
      "loss": 0.0001,
      "step": 476
    },
    {
      "epoch": 1.908,
      "grad_norm": 0.0002040638200331393,
      "learning_rate": 6.431867463506048e-07,
      "loss": 0.0,
      "step": 477
    },
    {
      "epoch": 1.912,
      "grad_norm": 0.0005499121710829063,
      "learning_rate": 5.885809276723608e-07,
      "loss": 0.0001,
      "step": 478
    },
    {
      "epoch": 1.916,
      "grad_norm": 0.0007844410277565046,
      "learning_rate": 5.363833518505834e-07,
      "loss": 0.0001,
      "step": 479
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.0009603581214355816,
      "learning_rate": 4.865965629214819e-07,
      "loss": 0.0001,
      "step": 480
    },
    {
      "epoch": 1.924,
      "grad_norm": 0.0006210384693467429,
      "learning_rate": 4.392229874229159e-07,
      "loss": 0.0001,
      "step": 481
    },
    {
      "epoch": 1.928,
      "grad_norm": 0.0010150384600477855,
      "learning_rate": 3.9426493427611177e-07,
      "loss": 0.0001,
      "step": 482
    },
    {
      "epoch": 1.932,
      "grad_norm": 0.0004526524770126363,
      "learning_rate": 3.517245946731529e-07,
      "loss": 0.0001,
      "step": 483
    },
    {
      "epoch": 1.936,
      "grad_norm": 0.0007936833990619904,
      "learning_rate": 3.1160404197018154e-07,
      "loss": 0.0001,
      "step": 484
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.0005009468116182134,
      "learning_rate": 2.7390523158633554e-07,
      "loss": 0.0,
      "step": 485
    },
    {
      "epoch": 1.944,
      "grad_norm": 0.00018071438082195023,
      "learning_rate": 2.386300009084408e-07,
      "loss": 0.0001,
      "step": 486
    },
    {
      "epoch": 1.948,
      "grad_norm": 0.0016963082895478617,
      "learning_rate": 2.057800692014833e-07,
      "loss": 0.0003,
      "step": 487
    },
    {
      "epoch": 1.952,
      "grad_norm": 0.0005970417033062099,
      "learning_rate": 1.753570375247815e-07,
      "loss": 0.0001,
      "step": 488
    },
    {
      "epoch": 1.956,
      "grad_norm": 0.004250799594702139,
      "learning_rate": 1.4736238865398765e-07,
      "loss": 0.0004,
      "step": 489
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.002555295045055843,
      "learning_rate": 1.2179748700879012e-07,
      "loss": 0.0004,
      "step": 490
    },
    {
      "epoch": 1.964,
      "grad_norm": 0.0004137486589034219,
      "learning_rate": 9.866357858642205e-08,
      "loss": 0.0001,
      "step": 491
    },
    {
      "epoch": 1.968,
      "grad_norm": 0.0005713848566592968,
      "learning_rate": 7.796179090094891e-08,
      "loss": 0.0001,
      "step": 492
    },
    {
      "epoch": 1.972,
      "grad_norm": 0.0018112016790910615,
      "learning_rate": 5.969313292830125e-08,
      "loss": 0.0003,
      "step": 493
    },
    {
      "epoch": 1.976,
      "grad_norm": 0.0012953439504586927,
      "learning_rate": 4.385849505708084e-08,
      "loss": 0.0001,
      "step": 494
    },
    {
      "epoch": 1.98,
      "grad_norm": 0.00928875469833967,
      "learning_rate": 3.04586490452119e-08,
      "loss": 0.0012,
      "step": 495
    },
    {
      "epoch": 1.984,
      "grad_norm": 0.00042175783414109385,
      "learning_rate": 1.949424798228239e-08,
      "loss": 0.0001,
      "step": 496
    },
    {
      "epoch": 1.988,
      "grad_norm": 0.0008624143257451655,
      "learning_rate": 1.096582625772502e-08,
      "loss": 0.0001,
      "step": 497
    },
    {
      "epoch": 1.992,
      "grad_norm": 0.000337374320551604,
      "learning_rate": 4.873799534788059e-09,
      "loss": 0.0,
      "step": 498
    },
    {
      "epoch": 1.996,
      "grad_norm": 0.051591742195231494,
      "learning_rate": 1.2184647302626583e-09,
      "loss": 0.0091,
      "step": 499
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.0007316878786147284,
      "learning_rate": 0.0,
      "loss": 0.0001,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.3350621461868286,
      "eval_runtime": 1.4629,
      "eval_samples_per_second": 0.684,
      "eval_steps_per_second": 0.684,
      "step": 500
    },
    {
      "epoch": 2.0,
      "step": 500,
      "total_flos": 6106581245952.0,
      "train_loss": 0.0,
      "train_runtime": 1.3476,
      "train_samples_per_second": 742.074,
      "train_steps_per_second": 371.037
    }
  ],
  "logging_steps": 1,
  "max_steps": 500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "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": 6106581245952.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}