{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 50.0,
  "eval_steps": 500,
  "global_step": 800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0625,
      "grad_norm": 4.823320729485639,
      "learning_rate": 8.333333333333334e-06,
      "loss": 1.7211,
      "step": 1
    },
    {
      "epoch": 0.125,
      "grad_norm": 4.751505642660104,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 1.7397,
      "step": 2
    },
    {
      "epoch": 0.1875,
      "grad_norm": 4.487357353479602,
      "learning_rate": 2.5e-05,
      "loss": 1.5714,
      "step": 3
    },
    {
      "epoch": 0.25,
      "grad_norm": 4.036784685130891,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 1.4755,
      "step": 4
    },
    {
      "epoch": 0.3125,
      "grad_norm": 2.7951107515008022,
      "learning_rate": 4.166666666666667e-05,
      "loss": 1.0114,
      "step": 5
    },
    {
      "epoch": 0.375,
      "grad_norm": 3.3636166511290284,
      "learning_rate": 5e-05,
      "loss": 0.866,
      "step": 6
    },
    {
      "epoch": 0.4375,
      "grad_norm": 2.077830674223075,
      "learning_rate": 5.833333333333334e-05,
      "loss": 0.5604,
      "step": 7
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.1666324696634576,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.339,
      "step": 8
    },
    {
      "epoch": 0.5625,
      "grad_norm": 0.8956349805282244,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.2396,
      "step": 9
    },
    {
      "epoch": 0.625,
      "grad_norm": 1.6517703519389952,
      "learning_rate": 8.333333333333334e-05,
      "loss": 0.2824,
      "step": 10
    },
    {
      "epoch": 0.6875,
      "grad_norm": 1.4629478660244202,
      "learning_rate": 9.166666666666667e-05,
      "loss": 0.247,
      "step": 11
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.8768061602490982,
      "learning_rate": 0.0001,
      "loss": 0.1893,
      "step": 12
    },
    {
      "epoch": 0.8125,
      "grad_norm": 0.8390983969336053,
      "learning_rate": 0.00010833333333333333,
      "loss": 0.2698,
      "step": 13
    },
    {
      "epoch": 0.875,
      "grad_norm": 0.9480328834691413,
      "learning_rate": 0.00011666666666666668,
      "loss": 0.1832,
      "step": 14
    },
    {
      "epoch": 0.9375,
      "grad_norm": 0.9607185561225253,
      "learning_rate": 0.000125,
      "loss": 0.1928,
      "step": 15
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.4976206492811293,
      "learning_rate": 0.00013333333333333334,
      "loss": 0.1599,
      "step": 16
    },
    {
      "epoch": 1.0625,
      "grad_norm": 0.6230231499462036,
      "learning_rate": 0.00014166666666666668,
      "loss": 0.2803,
      "step": 17
    },
    {
      "epoch": 1.125,
      "grad_norm": 0.4748261671538313,
      "learning_rate": 0.00015000000000000001,
      "loss": 0.1435,
      "step": 18
    },
    {
      "epoch": 1.1875,
      "grad_norm": 0.41444527359590055,
      "learning_rate": 0.00015833333333333332,
      "loss": 0.1798,
      "step": 19
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.458788407239296,
      "learning_rate": 0.0001666666666666667,
      "loss": 0.1989,
      "step": 20
    },
    {
      "epoch": 1.3125,
      "grad_norm": 0.4563665244618309,
      "learning_rate": 0.000175,
      "loss": 0.1647,
      "step": 21
    },
    {
      "epoch": 1.375,
      "grad_norm": 0.4105054285151717,
      "learning_rate": 0.00018333333333333334,
      "loss": 0.1164,
      "step": 22
    },
    {
      "epoch": 1.4375,
      "grad_norm": 0.3562784794883969,
      "learning_rate": 0.00019166666666666667,
      "loss": 0.1172,
      "step": 23
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.38610424041542285,
      "learning_rate": 0.0002,
      "loss": 0.1056,
      "step": 24
    },
    {
      "epoch": 1.5625,
      "grad_norm": 0.35527586407983786,
      "learning_rate": 0.00019999918050612108,
      "loss": 0.1217,
      "step": 25
    },
    {
      "epoch": 1.625,
      "grad_norm": 0.3633401519435876,
      "learning_rate": 0.00019999672203791565,
      "loss": 0.1458,
      "step": 26
    },
    {
      "epoch": 1.6875,
      "grad_norm": 0.3603552448206864,
      "learning_rate": 0.00019999262463567773,
      "loss": 0.1261,
      "step": 27
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.3388971002892635,
      "learning_rate": 0.00019998688836656323,
      "loss": 0.1309,
      "step": 28
    },
    {
      "epoch": 1.8125,
      "grad_norm": 0.4061212900739158,
      "learning_rate": 0.0001999795133245889,
      "loss": 0.1091,
      "step": 29
    },
    {
      "epoch": 1.875,
      "grad_norm": 0.39345006056668225,
      "learning_rate": 0.0001999704996306308,
      "loss": 0.1627,
      "step": 30
    },
    {
      "epoch": 1.9375,
      "grad_norm": 0.38142281255979,
      "learning_rate": 0.00019995984743242226,
      "loss": 0.1327,
      "step": 31
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.5305995721603599,
      "learning_rate": 0.00019994755690455152,
      "loss": 0.156,
      "step": 32
    },
    {
      "epoch": 2.0625,
      "grad_norm": 0.2546721306857309,
      "learning_rate": 0.00019993362824845875,
      "loss": 0.0744,
      "step": 33
    },
    {
      "epoch": 2.125,
      "grad_norm": 0.3997709948362388,
      "learning_rate": 0.000199918061692433,
      "loss": 0.1605,
      "step": 34
    },
    {
      "epoch": 2.1875,
      "grad_norm": 0.39616508988629523,
      "learning_rate": 0.00019990085749160822,
      "loss": 0.1306,
      "step": 35
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.4619577933372822,
      "learning_rate": 0.0001998820159279591,
      "loss": 0.1318,
      "step": 36
    },
    {
      "epoch": 2.3125,
      "grad_norm": 0.2968290985862986,
      "learning_rate": 0.00019986153731029656,
      "loss": 0.0788,
      "step": 37
    },
    {
      "epoch": 2.375,
      "grad_norm": 0.4062269625672033,
      "learning_rate": 0.0001998394219742627,
      "loss": 0.1207,
      "step": 38
    },
    {
      "epoch": 2.4375,
      "grad_norm": 0.36908209897646804,
      "learning_rate": 0.00019981567028232514,
      "loss": 0.1006,
      "step": 39
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.23625961054729414,
      "learning_rate": 0.00019979028262377118,
      "loss": 0.1083,
      "step": 40
    },
    {
      "epoch": 2.5625,
      "grad_norm": 0.3069294649987226,
      "learning_rate": 0.00019976325941470146,
      "loss": 0.0962,
      "step": 41
    },
    {
      "epoch": 2.625,
      "grad_norm": 0.23363462936052565,
      "learning_rate": 0.00019973460109802305,
      "loss": 0.096,
      "step": 42
    },
    {
      "epoch": 2.6875,
      "grad_norm": 0.29106062929023385,
      "learning_rate": 0.0001997043081434423,
      "loss": 0.134,
      "step": 43
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.18087001511209616,
      "learning_rate": 0.00019967238104745696,
      "loss": 0.1031,
      "step": 44
    },
    {
      "epoch": 2.8125,
      "grad_norm": 0.24123960365724528,
      "learning_rate": 0.00019963882033334826,
      "loss": 0.1082,
      "step": 45
    },
    {
      "epoch": 2.875,
      "grad_norm": 0.25054721428161647,
      "learning_rate": 0.00019960362655117218,
      "loss": 0.0798,
      "step": 46
    },
    {
      "epoch": 2.9375,
      "grad_norm": 0.5422539067423491,
      "learning_rate": 0.00019956680027775051,
      "loss": 0.1254,
      "step": 47
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.2565364589386019,
      "learning_rate": 0.0001995283421166614,
      "loss": 0.0979,
      "step": 48
    },
    {
      "epoch": 3.0625,
      "grad_norm": 0.215553765023481,
      "learning_rate": 0.00019948825269822934,
      "loss": 0.0656,
      "step": 49
    },
    {
      "epoch": 3.125,
      "grad_norm": 0.20810732914954436,
      "learning_rate": 0.00019944653267951504,
      "loss": 0.0793,
      "step": 50
    },
    {
      "epoch": 3.1875,
      "grad_norm": 0.28614832461281026,
      "learning_rate": 0.00019940318274430449,
      "loss": 0.099,
      "step": 51
    },
    {
      "epoch": 3.25,
      "grad_norm": 0.32282441297094233,
      "learning_rate": 0.00019935820360309777,
      "loss": 0.0958,
      "step": 52
    },
    {
      "epoch": 3.3125,
      "grad_norm": 0.1946365159456168,
      "learning_rate": 0.00019931159599309757,
      "loss": 0.0808,
      "step": 53
    },
    {
      "epoch": 3.375,
      "grad_norm": 0.327313439667935,
      "learning_rate": 0.00019926336067819684,
      "loss": 0.081,
      "step": 54
    },
    {
      "epoch": 3.4375,
      "grad_norm": 0.25559584990032447,
      "learning_rate": 0.00019921349844896654,
      "loss": 0.0855,
      "step": 55
    },
    {
      "epoch": 3.5,
      "grad_norm": 0.42835063024229314,
      "learning_rate": 0.00019916201012264254,
      "loss": 0.0863,
      "step": 56
    },
    {
      "epoch": 3.5625,
      "grad_norm": 0.4079608215806712,
      "learning_rate": 0.00019910889654311208,
      "loss": 0.0749,
      "step": 57
    },
    {
      "epoch": 3.625,
      "grad_norm": 0.2658074683405381,
      "learning_rate": 0.00019905415858090036,
      "loss": 0.0829,
      "step": 58
    },
    {
      "epoch": 3.6875,
      "grad_norm": 0.26176221704098207,
      "learning_rate": 0.00019899779713315575,
      "loss": 0.0711,
      "step": 59
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.4107999604774971,
      "learning_rate": 0.00019893981312363562,
      "loss": 0.0925,
      "step": 60
    },
    {
      "epoch": 3.8125,
      "grad_norm": 0.28972083293965484,
      "learning_rate": 0.00019888020750269067,
      "loss": 0.1033,
      "step": 61
    },
    {
      "epoch": 3.875,
      "grad_norm": 0.22772110434236376,
      "learning_rate": 0.00019881898124724981,
      "loss": 0.0858,
      "step": 62
    },
    {
      "epoch": 3.9375,
      "grad_norm": 0.27554810738804436,
      "learning_rate": 0.0001987561353608038,
      "loss": 0.0988,
      "step": 63
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.33177229863886515,
      "learning_rate": 0.00019869167087338907,
      "loss": 0.0558,
      "step": 64
    },
    {
      "epoch": 4.0625,
      "grad_norm": 0.307474933124812,
      "learning_rate": 0.00019862558884157068,
      "loss": 0.099,
      "step": 65
    },
    {
      "epoch": 4.125,
      "grad_norm": 0.2608601940862889,
      "learning_rate": 0.00019855789034842504,
      "loss": 0.0633,
      "step": 66
    },
    {
      "epoch": 4.1875,
      "grad_norm": 0.17638905908873861,
      "learning_rate": 0.00019848857650352214,
      "loss": 0.0646,
      "step": 67
    },
    {
      "epoch": 4.25,
      "grad_norm": 0.29040712597623347,
      "learning_rate": 0.00019841764844290744,
      "loss": 0.1065,
      "step": 68
    },
    {
      "epoch": 4.3125,
      "grad_norm": 0.2712830927822363,
      "learning_rate": 0.00019834510732908315,
      "loss": 0.0829,
      "step": 69
    },
    {
      "epoch": 4.375,
      "grad_norm": 0.23863314085657264,
      "learning_rate": 0.00019827095435098925,
      "loss": 0.0745,
      "step": 70
    },
    {
      "epoch": 4.4375,
      "grad_norm": 0.1793388758656021,
      "learning_rate": 0.000198195190723984,
      "loss": 0.072,
      "step": 71
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.23980978867337854,
      "learning_rate": 0.0001981178176898239,
      "loss": 0.0838,
      "step": 72
    },
    {
      "epoch": 4.5625,
      "grad_norm": 0.2043201014604402,
      "learning_rate": 0.0001980388365166436,
      "loss": 0.0632,
      "step": 73
    },
    {
      "epoch": 4.625,
      "grad_norm": 0.15180530742317308,
      "learning_rate": 0.0001979582484989348,
      "loss": 0.0424,
      "step": 74
    },
    {
      "epoch": 4.6875,
      "grad_norm": 0.2547988574530317,
      "learning_rate": 0.00019787605495752528,
      "loss": 0.0851,
      "step": 75
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.1869380541972453,
      "learning_rate": 0.00019779225723955707,
      "loss": 0.0565,
      "step": 76
    },
    {
      "epoch": 4.8125,
      "grad_norm": 0.2716796793411313,
      "learning_rate": 0.00019770685671846456,
      "loss": 0.0772,
      "step": 77
    },
    {
      "epoch": 4.875,
      "grad_norm": 0.17433079754490768,
      "learning_rate": 0.0001976198547939518,
      "loss": 0.0482,
      "step": 78
    },
    {
      "epoch": 4.9375,
      "grad_norm": 0.18011278074792128,
      "learning_rate": 0.0001975312528919697,
      "loss": 0.0518,
      "step": 79
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.31968800498614336,
      "learning_rate": 0.00019744105246469263,
      "loss": 0.083,
      "step": 80
    },
    {
      "epoch": 5.0625,
      "grad_norm": 0.2357655152058512,
      "learning_rate": 0.00019734925499049447,
      "loss": 0.069,
      "step": 81
    },
    {
      "epoch": 5.125,
      "grad_norm": 0.34723380394306,
      "learning_rate": 0.0001972558619739246,
      "loss": 0.0514,
      "step": 82
    },
    {
      "epoch": 5.1875,
      "grad_norm": 0.21060705394218132,
      "learning_rate": 0.00019716087494568317,
      "loss": 0.0587,
      "step": 83
    },
    {
      "epoch": 5.25,
      "grad_norm": 0.24633955772510746,
      "learning_rate": 0.00019706429546259593,
      "loss": 0.0534,
      "step": 84
    },
    {
      "epoch": 5.3125,
      "grad_norm": 0.2306490127122376,
      "learning_rate": 0.00019696612510758876,
      "loss": 0.0585,
      "step": 85
    },
    {
      "epoch": 5.375,
      "grad_norm": 0.21690646083214937,
      "learning_rate": 0.00019686636548966178,
      "loss": 0.0545,
      "step": 86
    },
    {
      "epoch": 5.4375,
      "grad_norm": 0.25251567374450357,
      "learning_rate": 0.00019676501824386294,
      "loss": 0.049,
      "step": 87
    },
    {
      "epoch": 5.5,
      "grad_norm": 0.19640965019118795,
      "learning_rate": 0.00019666208503126112,
      "loss": 0.0419,
      "step": 88
    },
    {
      "epoch": 5.5625,
      "grad_norm": 0.268693556790637,
      "learning_rate": 0.00019655756753891916,
      "loss": 0.0619,
      "step": 89
    },
    {
      "epoch": 5.625,
      "grad_norm": 0.18291834239390523,
      "learning_rate": 0.0001964514674798659,
      "loss": 0.0295,
      "step": 90
    },
    {
      "epoch": 5.6875,
      "grad_norm": 0.24981548084974023,
      "learning_rate": 0.00019634378659306832,
      "loss": 0.0699,
      "step": 91
    },
    {
      "epoch": 5.75,
      "grad_norm": 0.2086985689956262,
      "learning_rate": 0.00019623452664340306,
      "loss": 0.0496,
      "step": 92
    },
    {
      "epoch": 5.8125,
      "grad_norm": 0.2043098976035677,
      "learning_rate": 0.0001961236894216272,
      "loss": 0.0407,
      "step": 93
    },
    {
      "epoch": 5.875,
      "grad_norm": 0.2730615592143012,
      "learning_rate": 0.00019601127674434928,
      "loss": 0.0631,
      "step": 94
    },
    {
      "epoch": 5.9375,
      "grad_norm": 0.3501805095305464,
      "learning_rate": 0.00019589729045399934,
      "loss": 0.071,
      "step": 95
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.20729413014232384,
      "learning_rate": 0.00019578173241879872,
      "loss": 0.045,
      "step": 96
    },
    {
      "epoch": 6.0625,
      "grad_norm": 0.15750376698735663,
      "learning_rate": 0.00019566460453272945,
      "loss": 0.0346,
      "step": 97
    },
    {
      "epoch": 6.125,
      "grad_norm": 0.3137680018534623,
      "learning_rate": 0.0001955459087155033,
      "loss": 0.0442,
      "step": 98
    },
    {
      "epoch": 6.1875,
      "grad_norm": 0.24917821149174516,
      "learning_rate": 0.0001954256469125301,
      "loss": 0.0425,
      "step": 99
    },
    {
      "epoch": 6.25,
      "grad_norm": 0.15986754420947197,
      "learning_rate": 0.0001953038210948861,
      "loss": 0.0468,
      "step": 100
    },
    {
      "epoch": 6.3125,
      "grad_norm": 0.22286136536824913,
      "learning_rate": 0.00019518043325928157,
      "loss": 0.0506,
      "step": 101
    },
    {
      "epoch": 6.375,
      "grad_norm": 0.24914639465088984,
      "learning_rate": 0.00019505548542802804,
      "loss": 0.0706,
      "step": 102
    },
    {
      "epoch": 6.4375,
      "grad_norm": 0.1752134277500243,
      "learning_rate": 0.00019492897964900512,
      "loss": 0.0468,
      "step": 103
    },
    {
      "epoch": 6.5,
      "grad_norm": 0.1927795196511069,
      "learning_rate": 0.00019480091799562704,
      "loss": 0.0561,
      "step": 104
    },
    {
      "epoch": 6.5625,
      "grad_norm": 0.1700835087697781,
      "learning_rate": 0.00019467130256680868,
      "loss": 0.0381,
      "step": 105
    },
    {
      "epoch": 6.625,
      "grad_norm": 0.2497630067036982,
      "learning_rate": 0.00019454013548693102,
      "loss": 0.0437,
      "step": 106
    },
    {
      "epoch": 6.6875,
      "grad_norm": 0.18837543811238555,
      "learning_rate": 0.00019440741890580643,
      "loss": 0.0526,
      "step": 107
    },
    {
      "epoch": 6.75,
      "grad_norm": 0.22518955186910106,
      "learning_rate": 0.00019427315499864344,
      "loss": 0.0424,
      "step": 108
    },
    {
      "epoch": 6.8125,
      "grad_norm": 0.20500175172697058,
      "learning_rate": 0.00019413734596601104,
      "loss": 0.052,
      "step": 109
    },
    {
      "epoch": 6.875,
      "grad_norm": 0.12045773353666347,
      "learning_rate": 0.00019399999403380266,
      "loss": 0.0189,
      "step": 110
    },
    {
      "epoch": 6.9375,
      "grad_norm": 0.2148224667509155,
      "learning_rate": 0.00019386110145319963,
      "loss": 0.0461,
      "step": 111
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.23216240017196932,
      "learning_rate": 0.00019372067050063438,
      "loss": 0.0628,
      "step": 112
    },
    {
      "epoch": 7.0625,
      "grad_norm": 0.27496419377253983,
      "learning_rate": 0.000193578703477753,
      "loss": 0.0504,
      "step": 113
    },
    {
      "epoch": 7.125,
      "grad_norm": 0.18897397234008034,
      "learning_rate": 0.00019343520271137763,
      "loss": 0.0418,
      "step": 114
    },
    {
      "epoch": 7.1875,
      "grad_norm": 0.17029862867910886,
      "learning_rate": 0.0001932901705534683,
      "loss": 0.0364,
      "step": 115
    },
    {
      "epoch": 7.25,
      "grad_norm": 0.1963658278569445,
      "learning_rate": 0.00019314360938108425,
      "loss": 0.0495,
      "step": 116
    },
    {
      "epoch": 7.3125,
      "grad_norm": 0.24460981520111755,
      "learning_rate": 0.00019299552159634517,
      "loss": 0.052,
      "step": 117
    },
    {
      "epoch": 7.375,
      "grad_norm": 0.2841837300999824,
      "learning_rate": 0.00019284590962639176,
      "loss": 0.0251,
      "step": 118
    },
    {
      "epoch": 7.4375,
      "grad_norm": 0.28136757069268153,
      "learning_rate": 0.0001926947759233459,
      "loss": 0.0583,
      "step": 119
    },
    {
      "epoch": 7.5,
      "grad_norm": 0.14123250014906416,
      "learning_rate": 0.00019254212296427044,
      "loss": 0.0247,
      "step": 120
    },
    {
      "epoch": 7.5625,
      "grad_norm": 0.24291209764375074,
      "learning_rate": 0.0001923879532511287,
      "loss": 0.041,
      "step": 121
    },
    {
      "epoch": 7.625,
      "grad_norm": 0.16313867008770622,
      "learning_rate": 0.0001922322693107434,
      "loss": 0.0267,
      "step": 122
    },
    {
      "epoch": 7.6875,
      "grad_norm": 0.20009496777109764,
      "learning_rate": 0.0001920750736947553,
      "loss": 0.0451,
      "step": 123
    },
    {
      "epoch": 7.75,
      "grad_norm": 0.33391229054639826,
      "learning_rate": 0.00019191636897958122,
      "loss": 0.0659,
      "step": 124
    },
    {
      "epoch": 7.8125,
      "grad_norm": 0.20026703724460174,
      "learning_rate": 0.0001917561577663721,
      "loss": 0.0309,
      "step": 125
    },
    {
      "epoch": 7.875,
      "grad_norm": 0.22455588585805783,
      "learning_rate": 0.00019159444268097012,
      "loss": 0.0396,
      "step": 126
    },
    {
      "epoch": 7.9375,
      "grad_norm": 0.26799981180801946,
      "learning_rate": 0.00019143122637386566,
      "loss": 0.0541,
      "step": 127
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.19547469523965652,
      "learning_rate": 0.00019126651152015403,
      "loss": 0.0551,
      "step": 128
    },
    {
      "epoch": 8.0625,
      "grad_norm": 0.1544644919932735,
      "learning_rate": 0.00019110030081949156,
      "loss": 0.025,
      "step": 129
    },
    {
      "epoch": 8.125,
      "grad_norm": 0.19630976778261258,
      "learning_rate": 0.00019093259699605125,
      "loss": 0.0301,
      "step": 130
    },
    {
      "epoch": 8.1875,
      "grad_norm": 0.17570512618519246,
      "learning_rate": 0.0001907634027984782,
      "loss": 0.0284,
      "step": 131
    },
    {
      "epoch": 8.25,
      "grad_norm": 0.15656419342775668,
      "learning_rate": 0.0001905927209998447,
      "loss": 0.0316,
      "step": 132
    },
    {
      "epoch": 8.3125,
      "grad_norm": 0.20086086826122584,
      "learning_rate": 0.00019042055439760444,
      "loss": 0.0406,
      "step": 133
    },
    {
      "epoch": 8.375,
      "grad_norm": 0.17929729489809332,
      "learning_rate": 0.000190246905813547,
      "loss": 0.0306,
      "step": 134
    },
    {
      "epoch": 8.4375,
      "grad_norm": 0.17408633598619777,
      "learning_rate": 0.0001900717780937514,
      "loss": 0.0331,
      "step": 135
    },
    {
      "epoch": 8.5,
      "grad_norm": 0.22245736981289757,
      "learning_rate": 0.00018989517410853955,
      "loss": 0.0296,
      "step": 136
    },
    {
      "epoch": 8.5625,
      "grad_norm": 0.15106374249798415,
      "learning_rate": 0.0001897170967524291,
      "loss": 0.0195,
      "step": 137
    },
    {
      "epoch": 8.625,
      "grad_norm": 0.19583907014296853,
      "learning_rate": 0.00018953754894408616,
      "loss": 0.034,
      "step": 138
    },
    {
      "epoch": 8.6875,
      "grad_norm": 0.2516341266285177,
      "learning_rate": 0.0001893565336262773,
      "loss": 0.0397,
      "step": 139
    },
    {
      "epoch": 8.75,
      "grad_norm": 0.21858233256386844,
      "learning_rate": 0.00018917405376582145,
      "loss": 0.0413,
      "step": 140
    },
    {
      "epoch": 8.8125,
      "grad_norm": 0.2198644417350149,
      "learning_rate": 0.00018899011235354115,
      "loss": 0.037,
      "step": 141
    },
    {
      "epoch": 8.875,
      "grad_norm": 0.14391410038600197,
      "learning_rate": 0.00018880471240421365,
      "loss": 0.0243,
      "step": 142
    },
    {
      "epoch": 8.9375,
      "grad_norm": 0.19142537938933432,
      "learning_rate": 0.00018861785695652142,
      "loss": 0.0378,
      "step": 143
    },
    {
      "epoch": 9.0,
      "grad_norm": 0.273105976253114,
      "learning_rate": 0.00018842954907300236,
      "loss": 0.0335,
      "step": 144
    },
    {
      "epoch": 9.0625,
      "grad_norm": 0.1764202905014519,
      "learning_rate": 0.00018823979183999964,
      "loss": 0.0269,
      "step": 145
    },
    {
      "epoch": 9.125,
      "grad_norm": 0.16768199512274265,
      "learning_rate": 0.00018804858836761107,
      "loss": 0.0274,
      "step": 146
    },
    {
      "epoch": 9.1875,
      "grad_norm": 0.11993797625263519,
      "learning_rate": 0.0001878559417896382,
      "loss": 0.0197,
      "step": 147
    },
    {
      "epoch": 9.25,
      "grad_norm": 0.1113638157849448,
      "learning_rate": 0.0001876618552635348,
      "loss": 0.0144,
      "step": 148
    },
    {
      "epoch": 9.3125,
      "grad_norm": 0.404272192796613,
      "learning_rate": 0.00018746633197035527,
      "loss": 0.0623,
      "step": 149
    },
    {
      "epoch": 9.375,
      "grad_norm": 0.21941870781755787,
      "learning_rate": 0.00018726937511470246,
      "loss": 0.0353,
      "step": 150
    },
    {
      "epoch": 9.4375,
      "grad_norm": 0.15013935972184475,
      "learning_rate": 0.00018707098792467515,
      "loss": 0.0212,
      "step": 151
    },
    {
      "epoch": 9.5,
      "grad_norm": 0.15110805919842035,
      "learning_rate": 0.00018687117365181512,
      "loss": 0.0218,
      "step": 152
    },
    {
      "epoch": 9.5625,
      "grad_norm": 0.2279637543691053,
      "learning_rate": 0.00018666993557105377,
      "loss": 0.0384,
      "step": 153
    },
    {
      "epoch": 9.625,
      "grad_norm": 0.22049558442795594,
      "learning_rate": 0.00018646727698065865,
      "loss": 0.0386,
      "step": 154
    },
    {
      "epoch": 9.6875,
      "grad_norm": 0.19273043393336428,
      "learning_rate": 0.00018626320120217923,
      "loss": 0.0261,
      "step": 155
    },
    {
      "epoch": 9.75,
      "grad_norm": 0.28617400835044193,
      "learning_rate": 0.00018605771158039253,
      "loss": 0.0366,
      "step": 156
    },
    {
      "epoch": 9.8125,
      "grad_norm": 0.18709984153605747,
      "learning_rate": 0.00018585081148324832,
      "loss": 0.0291,
      "step": 157
    },
    {
      "epoch": 9.875,
      "grad_norm": 0.1292782602493134,
      "learning_rate": 0.00018564250430181387,
      "loss": 0.0199,
      "step": 158
    },
    {
      "epoch": 9.9375,
      "grad_norm": 0.13979049475637031,
      "learning_rate": 0.00018543279345021834,
      "loss": 0.0157,
      "step": 159
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.19965128862936724,
      "learning_rate": 0.00018522168236559695,
      "loss": 0.0323,
      "step": 160
    },
    {
      "epoch": 10.0625,
      "grad_norm": 0.15854406411462987,
      "learning_rate": 0.0001850091745080345,
      "loss": 0.029,
      "step": 161
    },
    {
      "epoch": 10.125,
      "grad_norm": 0.21210803120758442,
      "learning_rate": 0.00018479527336050878,
      "loss": 0.0275,
      "step": 162
    },
    {
      "epoch": 10.1875,
      "grad_norm": 0.14570318547786973,
      "learning_rate": 0.00018457998242883344,
      "loss": 0.0198,
      "step": 163
    },
    {
      "epoch": 10.25,
      "grad_norm": 0.1365650856121692,
      "learning_rate": 0.00018436330524160047,
      "loss": 0.0187,
      "step": 164
    },
    {
      "epoch": 10.3125,
      "grad_norm": 0.14366514107884812,
      "learning_rate": 0.00018414524535012244,
      "loss": 0.0201,
      "step": 165
    },
    {
      "epoch": 10.375,
      "grad_norm": 0.11689977724032004,
      "learning_rate": 0.00018392580632837423,
      "loss": 0.0127,
      "step": 166
    },
    {
      "epoch": 10.4375,
      "grad_norm": 0.14547947591736377,
      "learning_rate": 0.00018370499177293464,
      "loss": 0.021,
      "step": 167
    },
    {
      "epoch": 10.5,
      "grad_norm": 0.13962090636283936,
      "learning_rate": 0.00018348280530292713,
      "loss": 0.0198,
      "step": 168
    },
    {
      "epoch": 10.5625,
      "grad_norm": 0.16513279241014592,
      "learning_rate": 0.00018325925055996076,
      "loss": 0.0292,
      "step": 169
    },
    {
      "epoch": 10.625,
      "grad_norm": 0.09476674044103778,
      "learning_rate": 0.0001830343312080704,
      "loss": 0.0163,
      "step": 170
    },
    {
      "epoch": 10.6875,
      "grad_norm": 0.2189319278786043,
      "learning_rate": 0.00018280805093365672,
      "loss": 0.0267,
      "step": 171
    },
    {
      "epoch": 10.75,
      "grad_norm": 0.11950416301901717,
      "learning_rate": 0.00018258041344542566,
      "loss": 0.0162,
      "step": 172
    },
    {
      "epoch": 10.8125,
      "grad_norm": 0.21271468962636456,
      "learning_rate": 0.00018235142247432782,
      "loss": 0.0341,
      "step": 173
    },
    {
      "epoch": 10.875,
      "grad_norm": 0.26291334746224676,
      "learning_rate": 0.0001821210817734972,
      "loss": 0.0183,
      "step": 174
    },
    {
      "epoch": 10.9375,
      "grad_norm": 0.7541232207929679,
      "learning_rate": 0.00018188939511818965,
      "loss": 0.0341,
      "step": 175
    },
    {
      "epoch": 11.0,
      "grad_norm": 0.2668657582109521,
      "learning_rate": 0.0001816563663057211,
      "loss": 0.0248,
      "step": 176
    },
    {
      "epoch": 11.0625,
      "grad_norm": 0.188760844502015,
      "learning_rate": 0.00018142199915540527,
      "loss": 0.0167,
      "step": 177
    },
    {
      "epoch": 11.125,
      "grad_norm": 0.13049188810156132,
      "learning_rate": 0.00018118629750849105,
      "loss": 0.0111,
      "step": 178
    },
    {
      "epoch": 11.1875,
      "grad_norm": 0.12147377234191868,
      "learning_rate": 0.0001809492652280996,
      "loss": 0.0132,
      "step": 179
    },
    {
      "epoch": 11.25,
      "grad_norm": 0.1561909420624479,
      "learning_rate": 0.00018071090619916093,
      "loss": 0.017,
      "step": 180
    },
    {
      "epoch": 11.3125,
      "grad_norm": 0.2761203872951007,
      "learning_rate": 0.00018047122432835038,
      "loss": 0.0242,
      "step": 181
    },
    {
      "epoch": 11.375,
      "grad_norm": 0.13770345739451892,
      "learning_rate": 0.0001802302235440245,
      "loss": 0.0168,
      "step": 182
    },
    {
      "epoch": 11.4375,
      "grad_norm": 0.18722163720923593,
      "learning_rate": 0.0001799879077961566,
      "loss": 0.0301,
      "step": 183
    },
    {
      "epoch": 11.5,
      "grad_norm": 0.16157498504579615,
      "learning_rate": 0.00017974428105627208,
      "loss": 0.0188,
      "step": 184
    },
    {
      "epoch": 11.5625,
      "grad_norm": 0.11914193260548754,
      "learning_rate": 0.00017949934731738347,
      "loss": 0.0167,
      "step": 185
    },
    {
      "epoch": 11.625,
      "grad_norm": 0.21366231459710572,
      "learning_rate": 0.0001792531105939247,
      "loss": 0.0307,
      "step": 186
    },
    {
      "epoch": 11.6875,
      "grad_norm": 0.12591549609993827,
      "learning_rate": 0.0001790055749216856,
      "loss": 0.0156,
      "step": 187
    },
    {
      "epoch": 11.75,
      "grad_norm": 0.17099000320728966,
      "learning_rate": 0.00017875674435774547,
      "loss": 0.0213,
      "step": 188
    },
    {
      "epoch": 11.8125,
      "grad_norm": 0.21223314589021008,
      "learning_rate": 0.00017850662298040678,
      "loss": 0.0303,
      "step": 189
    },
    {
      "epoch": 11.875,
      "grad_norm": 0.18967657804115237,
      "learning_rate": 0.0001782552148891283,
      "loss": 0.0147,
      "step": 190
    },
    {
      "epoch": 11.9375,
      "grad_norm": 0.2411319114734683,
      "learning_rate": 0.00017800252420445788,
      "loss": 0.0356,
      "step": 191
    },
    {
      "epoch": 12.0,
      "grad_norm": 0.19592901070523805,
      "learning_rate": 0.00017774855506796496,
      "loss": 0.0361,
      "step": 192
    },
    {
      "epoch": 12.0625,
      "grad_norm": 0.1922774431645892,
      "learning_rate": 0.0001774933116421725,
      "loss": 0.0163,
      "step": 193
    },
    {
      "epoch": 12.125,
      "grad_norm": 0.1336099627438813,
      "learning_rate": 0.00017723679811048904,
      "loss": 0.016,
      "step": 194
    },
    {
      "epoch": 12.1875,
      "grad_norm": 0.11902506990463049,
      "learning_rate": 0.00017697901867713995,
      "loss": 0.0128,
      "step": 195
    },
    {
      "epoch": 12.25,
      "grad_norm": 0.15265650075862036,
      "learning_rate": 0.00017671997756709863,
      "loss": 0.0158,
      "step": 196
    },
    {
      "epoch": 12.3125,
      "grad_norm": 0.10409493403901945,
      "learning_rate": 0.0001764596790260171,
      "loss": 0.01,
      "step": 197
    },
    {
      "epoch": 12.375,
      "grad_norm": 0.12592944148803245,
      "learning_rate": 0.00017619812732015664,
      "loss": 0.0081,
      "step": 198
    },
    {
      "epoch": 12.4375,
      "grad_norm": 0.15892735182879292,
      "learning_rate": 0.00017593532673631766,
      "loss": 0.0226,
      "step": 199
    },
    {
      "epoch": 12.5,
      "grad_norm": 0.16370083252883327,
      "learning_rate": 0.00017567128158176953,
      "loss": 0.0156,
      "step": 200
    },
    {
      "epoch": 12.5625,
      "grad_norm": 0.1926319403468188,
      "learning_rate": 0.00017540599618418007,
      "loss": 0.0198,
      "step": 201
    },
    {
      "epoch": 12.625,
      "grad_norm": 0.1361218264727559,
      "learning_rate": 0.00017513947489154443,
      "loss": 0.014,
      "step": 202
    },
    {
      "epoch": 12.6875,
      "grad_norm": 0.15346622202020466,
      "learning_rate": 0.00017487172207211396,
      "loss": 0.0149,
      "step": 203
    },
    {
      "epoch": 12.75,
      "grad_norm": 0.1582659958129119,
      "learning_rate": 0.0001746027421143246,
      "loss": 0.0209,
      "step": 204
    },
    {
      "epoch": 12.8125,
      "grad_norm": 0.08422642994909509,
      "learning_rate": 0.00017433253942672496,
      "loss": 0.0107,
      "step": 205
    },
    {
      "epoch": 12.875,
      "grad_norm": 0.14151191719126865,
      "learning_rate": 0.000174061118437904,
      "loss": 0.0163,
      "step": 206
    },
    {
      "epoch": 12.9375,
      "grad_norm": 0.21170441141264745,
      "learning_rate": 0.00017378848359641847,
      "loss": 0.0248,
      "step": 207
    },
    {
      "epoch": 13.0,
      "grad_norm": 0.21854641357858184,
      "learning_rate": 0.00017351463937072004,
      "loss": 0.0314,
      "step": 208
    },
    {
      "epoch": 13.0625,
      "grad_norm": 0.1460880269124164,
      "learning_rate": 0.00017323959024908209,
      "loss": 0.01,
      "step": 209
    },
    {
      "epoch": 13.125,
      "grad_norm": 0.2314775735978235,
      "learning_rate": 0.00017296334073952605,
      "loss": 0.0158,
      "step": 210
    },
    {
      "epoch": 13.1875,
      "grad_norm": 0.25609276829659133,
      "learning_rate": 0.0001726858953697475,
      "loss": 0.0228,
      "step": 211
    },
    {
      "epoch": 13.25,
      "grad_norm": 0.05892147625450794,
      "learning_rate": 0.00017240725868704218,
      "loss": 0.0054,
      "step": 212
    },
    {
      "epoch": 13.3125,
      "grad_norm": 0.09265829411728602,
      "learning_rate": 0.00017212743525823112,
      "loss": 0.0121,
      "step": 213
    },
    {
      "epoch": 13.375,
      "grad_norm": 0.28766545739529126,
      "learning_rate": 0.0001718464296695861,
      "loss": 0.0166,
      "step": 214
    },
    {
      "epoch": 13.4375,
      "grad_norm": 0.22056507132944547,
      "learning_rate": 0.0001715642465267543,
      "loss": 0.0189,
      "step": 215
    },
    {
      "epoch": 13.5,
      "grad_norm": 0.20425585831942672,
      "learning_rate": 0.00017128089045468294,
      "loss": 0.0226,
      "step": 216
    },
    {
      "epoch": 13.5625,
      "grad_norm": 0.08920463828133202,
      "learning_rate": 0.00017099636609754329,
      "loss": 0.0111,
      "step": 217
    },
    {
      "epoch": 13.625,
      "grad_norm": 0.07587770729105793,
      "learning_rate": 0.00017071067811865476,
      "loss": 0.0107,
      "step": 218
    },
    {
      "epoch": 13.6875,
      "grad_norm": 0.2436581509018354,
      "learning_rate": 0.00017042383120040834,
      "loss": 0.0122,
      "step": 219
    },
    {
      "epoch": 13.75,
      "grad_norm": 0.2776351805516073,
      "learning_rate": 0.00017013583004418993,
      "loss": 0.0338,
      "step": 220
    },
    {
      "epoch": 13.8125,
      "grad_norm": 0.1500695552486999,
      "learning_rate": 0.00016984667937030318,
      "loss": 0.0128,
      "step": 221
    },
    {
      "epoch": 13.875,
      "grad_norm": 0.19022226617986523,
      "learning_rate": 0.00016955638391789228,
      "loss": 0.0182,
      "step": 222
    },
    {
      "epoch": 13.9375,
      "grad_norm": 0.07668703300214032,
      "learning_rate": 0.00016926494844486412,
      "loss": 0.0081,
      "step": 223
    },
    {
      "epoch": 14.0,
      "grad_norm": 0.09791844205901491,
      "learning_rate": 0.00016897237772781044,
      "loss": 0.0102,
      "step": 224
    },
    {
      "epoch": 14.0625,
      "grad_norm": 0.13334638868212356,
      "learning_rate": 0.00016867867656192946,
      "loss": 0.0154,
      "step": 225
    },
    {
      "epoch": 14.125,
      "grad_norm": 0.0443847074815828,
      "learning_rate": 0.00016838384976094738,
      "loss": 0.0038,
      "step": 226
    },
    {
      "epoch": 14.1875,
      "grad_norm": 0.22900928937804826,
      "learning_rate": 0.00016808790215703935,
      "loss": 0.0146,
      "step": 227
    },
    {
      "epoch": 14.25,
      "grad_norm": 0.09820050963868261,
      "learning_rate": 0.00016779083860075033,
      "loss": 0.0139,
      "step": 228
    },
    {
      "epoch": 14.3125,
      "grad_norm": 0.12261110863547245,
      "learning_rate": 0.0001674926639609157,
      "loss": 0.0081,
      "step": 229
    },
    {
      "epoch": 14.375,
      "grad_norm": 0.4737769261848427,
      "learning_rate": 0.00016719338312458124,
      "loss": 0.0196,
      "step": 230
    },
    {
      "epoch": 14.4375,
      "grad_norm": 0.0719988039842083,
      "learning_rate": 0.00016689300099692332,
      "loss": 0.0075,
      "step": 231
    },
    {
      "epoch": 14.5,
      "grad_norm": 0.24824987963252254,
      "learning_rate": 0.00016659152250116812,
      "loss": 0.0095,
      "step": 232
    },
    {
      "epoch": 14.5625,
      "grad_norm": 0.1806409419450783,
      "learning_rate": 0.00016628895257851135,
      "loss": 0.0177,
      "step": 233
    },
    {
      "epoch": 14.625,
      "grad_norm": 0.12022842892878163,
      "learning_rate": 0.000165985296188037,
      "loss": 0.0108,
      "step": 234
    },
    {
      "epoch": 14.6875,
      "grad_norm": 0.27573651025583323,
      "learning_rate": 0.0001656805583066361,
      "loss": 0.0408,
      "step": 235
    },
    {
      "epoch": 14.75,
      "grad_norm": 0.13027830763516066,
      "learning_rate": 0.00016537474392892528,
      "loss": 0.0164,
      "step": 236
    },
    {
      "epoch": 14.8125,
      "grad_norm": 0.2190711849455461,
      "learning_rate": 0.00016506785806716465,
      "loss": 0.0381,
      "step": 237
    },
    {
      "epoch": 14.875,
      "grad_norm": 0.18889153886713622,
      "learning_rate": 0.00016475990575117605,
      "loss": 0.0137,
      "step": 238
    },
    {
      "epoch": 14.9375,
      "grad_norm": 0.17427492795979294,
      "learning_rate": 0.0001644508920282601,
      "loss": 0.0259,
      "step": 239
    },
    {
      "epoch": 15.0,
      "grad_norm": 0.13217744356726124,
      "learning_rate": 0.000164140821963114,
      "loss": 0.0099,
      "step": 240
    },
    {
      "epoch": 15.0625,
      "grad_norm": 0.08062959617570911,
      "learning_rate": 0.0001638297006377481,
      "loss": 0.0065,
      "step": 241
    },
    {
      "epoch": 15.125,
      "grad_norm": 0.11874477325134665,
      "learning_rate": 0.00016351753315140287,
      "loss": 0.0132,
      "step": 242
    },
    {
      "epoch": 15.1875,
      "grad_norm": 0.08293002909973335,
      "learning_rate": 0.00016320432462046516,
      "loss": 0.0093,
      "step": 243
    },
    {
      "epoch": 15.25,
      "grad_norm": 0.08700808300439221,
      "learning_rate": 0.00016289008017838445,
      "loss": 0.0077,
      "step": 244
    },
    {
      "epoch": 15.3125,
      "grad_norm": 0.12073296585647869,
      "learning_rate": 0.00016257480497558873,
      "loss": 0.0096,
      "step": 245
    },
    {
      "epoch": 15.375,
      "grad_norm": 0.07654303999452532,
      "learning_rate": 0.0001622585041793999,
      "loss": 0.0059,
      "step": 246
    },
    {
      "epoch": 15.4375,
      "grad_norm": 0.2562520960634689,
      "learning_rate": 0.00016194118297394936,
      "loss": 0.0263,
      "step": 247
    },
    {
      "epoch": 15.5,
      "grad_norm": 0.08068310167444095,
      "learning_rate": 0.00016162284656009274,
      "loss": 0.0062,
      "step": 248
    },
    {
      "epoch": 15.5625,
      "grad_norm": 0.2090301776269612,
      "learning_rate": 0.00016130350015532496,
      "loss": 0.0201,
      "step": 249
    },
    {
      "epoch": 15.625,
      "grad_norm": 0.18851005491544473,
      "learning_rate": 0.00016098314899369446,
      "loss": 0.0129,
      "step": 250
    },
    {
      "epoch": 15.6875,
      "grad_norm": 0.10484133416084232,
      "learning_rate": 0.0001606617983257176,
      "loss": 0.0058,
      "step": 251
    },
    {
      "epoch": 15.75,
      "grad_norm": 0.17883267452407117,
      "learning_rate": 0.00016033945341829248,
      "loss": 0.0194,
      "step": 252
    },
    {
      "epoch": 15.8125,
      "grad_norm": 0.15051127763163427,
      "learning_rate": 0.00016001611955461265,
      "loss": 0.011,
      "step": 253
    },
    {
      "epoch": 15.875,
      "grad_norm": 0.09161429352244004,
      "learning_rate": 0.0001596918020340805,
      "loss": 0.0045,
      "step": 254
    },
    {
      "epoch": 15.9375,
      "grad_norm": 0.11734353010884248,
      "learning_rate": 0.00015936650617222063,
      "loss": 0.007,
      "step": 255
    },
    {
      "epoch": 16.0,
      "grad_norm": 0.1956323556889479,
      "learning_rate": 0.00015904023730059228,
      "loss": 0.0165,
      "step": 256
    },
    {
      "epoch": 16.0625,
      "grad_norm": 0.14747898179025032,
      "learning_rate": 0.00015871300076670234,
      "loss": 0.0146,
      "step": 257
    },
    {
      "epoch": 16.125,
      "grad_norm": 0.18498598991778836,
      "learning_rate": 0.00015838480193391754,
      "loss": 0.0102,
      "step": 258
    },
    {
      "epoch": 16.1875,
      "grad_norm": 0.2419734729440462,
      "learning_rate": 0.0001580556461813766,
      "loss": 0.02,
      "step": 259
    },
    {
      "epoch": 16.25,
      "grad_norm": 0.13549389704091608,
      "learning_rate": 0.00015772553890390197,
      "loss": 0.0096,
      "step": 260
    },
    {
      "epoch": 16.3125,
      "grad_norm": 0.09488023406511628,
      "learning_rate": 0.0001573944855119115,
      "loss": 0.0142,
      "step": 261
    },
    {
      "epoch": 16.375,
      "grad_norm": 0.08589054588641899,
      "learning_rate": 0.00015706249143132982,
      "loss": 0.0086,
      "step": 262
    },
    {
      "epoch": 16.4375,
      "grad_norm": 0.18599116361768675,
      "learning_rate": 0.00015672956210349923,
      "loss": 0.0158,
      "step": 263
    },
    {
      "epoch": 16.5,
      "grad_norm": 0.13790931505797482,
      "learning_rate": 0.00015639570298509064,
      "loss": 0.0076,
      "step": 264
    },
    {
      "epoch": 16.5625,
      "grad_norm": 0.14746745744271494,
      "learning_rate": 0.0001560609195480142,
      "loss": 0.0144,
      "step": 265
    },
    {
      "epoch": 16.625,
      "grad_norm": 0.1747110630319595,
      "learning_rate": 0.00015572521727932935,
      "loss": 0.0209,
      "step": 266
    },
    {
      "epoch": 16.6875,
      "grad_norm": 0.12294368607669189,
      "learning_rate": 0.00015538860168115527,
      "loss": 0.0076,
      "step": 267
    },
    {
      "epoch": 16.75,
      "grad_norm": 0.20591709989710746,
      "learning_rate": 0.00015505107827058036,
      "loss": 0.0109,
      "step": 268
    },
    {
      "epoch": 16.8125,
      "grad_norm": 0.12788905057548938,
      "learning_rate": 0.00015471265257957202,
      "loss": 0.0137,
      "step": 269
    },
    {
      "epoch": 16.875,
      "grad_norm": 0.0907643332122881,
      "learning_rate": 0.00015437333015488587,
      "loss": 0.004,
      "step": 270
    },
    {
      "epoch": 16.9375,
      "grad_norm": 0.11598695941727767,
      "learning_rate": 0.00015403311655797492,
      "loss": 0.0173,
      "step": 271
    },
    {
      "epoch": 17.0,
      "grad_norm": 0.1274596211332466,
      "learning_rate": 0.0001536920173648984,
      "loss": 0.0137,
      "step": 272
    },
    {
      "epoch": 17.0625,
      "grad_norm": 0.11710054449245828,
      "learning_rate": 0.00015335003816623028,
      "loss": 0.0114,
      "step": 273
    },
    {
      "epoch": 17.125,
      "grad_norm": 0.10810173543943251,
      "learning_rate": 0.00015300718456696778,
      "loss": 0.0083,
      "step": 274
    },
    {
      "epoch": 17.1875,
      "grad_norm": 0.11490194551290545,
      "learning_rate": 0.00015266346218643947,
      "loss": 0.0104,
      "step": 275
    },
    {
      "epoch": 17.25,
      "grad_norm": 0.10496790427956344,
      "learning_rate": 0.000152318876658213,
      "loss": 0.0117,
      "step": 276
    },
    {
      "epoch": 17.3125,
      "grad_norm": 0.13303819260520472,
      "learning_rate": 0.00015197343363000307,
      "loss": 0.0172,
      "step": 277
    },
    {
      "epoch": 17.375,
      "grad_norm": 0.07976187691489159,
      "learning_rate": 0.00015162713876357858,
      "loss": 0.0118,
      "step": 278
    },
    {
      "epoch": 17.4375,
      "grad_norm": 0.020200494852139077,
      "learning_rate": 0.00015127999773467002,
      "loss": 0.0018,
      "step": 279
    },
    {
      "epoch": 17.5,
      "grad_norm": 0.08302979036485983,
      "learning_rate": 0.00015093201623287631,
      "loss": 0.0074,
      "step": 280
    },
    {
      "epoch": 17.5625,
      "grad_norm": 0.09709353670094575,
      "learning_rate": 0.00015058319996157172,
      "loss": 0.0141,
      "step": 281
    },
    {
      "epoch": 17.625,
      "grad_norm": 0.10500205069763988,
      "learning_rate": 0.0001502335546378122,
      "loss": 0.0082,
      "step": 282
    },
    {
      "epoch": 17.6875,
      "grad_norm": 0.045653846262314286,
      "learning_rate": 0.00014988308599224183,
      "loss": 0.0037,
      "step": 283
    },
    {
      "epoch": 17.75,
      "grad_norm": 0.07350542815715672,
      "learning_rate": 0.00014953179976899878,
      "loss": 0.007,
      "step": 284
    },
    {
      "epoch": 17.8125,
      "grad_norm": 0.0740842955766792,
      "learning_rate": 0.0001491797017256212,
      "loss": 0.0041,
      "step": 285
    },
    {
      "epoch": 17.875,
      "grad_norm": 0.11610913098575787,
      "learning_rate": 0.00014882679763295306,
      "loss": 0.0177,
      "step": 286
    },
    {
      "epoch": 17.9375,
      "grad_norm": 0.11006673147506568,
      "learning_rate": 0.0001484730932750491,
      "loss": 0.0124,
      "step": 287
    },
    {
      "epoch": 18.0,
      "grad_norm": 0.17310654116013702,
      "learning_rate": 0.00014811859444908052,
      "loss": 0.0174,
      "step": 288
    },
    {
      "epoch": 18.0625,
      "grad_norm": 0.05397642477813828,
      "learning_rate": 0.00014776330696523963,
      "loss": 0.0047,
      "step": 289
    },
    {
      "epoch": 18.125,
      "grad_norm": 0.07510434928834142,
      "learning_rate": 0.00014740723664664483,
      "loss": 0.0084,
      "step": 290
    },
    {
      "epoch": 18.1875,
      "grad_norm": 0.15571308024563857,
      "learning_rate": 0.00014705038932924503,
      "loss": 0.0061,
      "step": 291
    },
    {
      "epoch": 18.25,
      "grad_norm": 0.04276361043100037,
      "learning_rate": 0.00014669277086172406,
      "loss": 0.0052,
      "step": 292
    },
    {
      "epoch": 18.3125,
      "grad_norm": 0.10566723422506075,
      "learning_rate": 0.00014633438710540489,
      "loss": 0.0095,
      "step": 293
    },
    {
      "epoch": 18.375,
      "grad_norm": 0.12332021633992264,
      "learning_rate": 0.00014597524393415335,
      "loss": 0.0125,
      "step": 294
    },
    {
      "epoch": 18.4375,
      "grad_norm": 0.06085873636051086,
      "learning_rate": 0.00014561534723428205,
      "loss": 0.0036,
      "step": 295
    },
    {
      "epoch": 18.5,
      "grad_norm": 0.08889363982250575,
      "learning_rate": 0.00014525470290445392,
      "loss": 0.0062,
      "step": 296
    },
    {
      "epoch": 18.5625,
      "grad_norm": 0.03473335852451447,
      "learning_rate": 0.00014489331685558525,
      "loss": 0.0026,
      "step": 297
    },
    {
      "epoch": 18.625,
      "grad_norm": 0.08852650667058536,
      "learning_rate": 0.00014453119501074924,
      "loss": 0.011,
      "step": 298
    },
    {
      "epoch": 18.6875,
      "grad_norm": 0.03288743535270982,
      "learning_rate": 0.00014416834330507856,
      "loss": 0.0031,
      "step": 299
    },
    {
      "epoch": 18.75,
      "grad_norm": 0.16831306636699117,
      "learning_rate": 0.00014380476768566824,
      "loss": 0.0142,
      "step": 300
    },
    {
      "epoch": 18.8125,
      "grad_norm": 0.060653415545579646,
      "learning_rate": 0.00014344047411147818,
      "loss": 0.0084,
      "step": 301
    },
    {
      "epoch": 18.875,
      "grad_norm": 0.0379853861449074,
      "learning_rate": 0.00014307546855323549,
      "loss": 0.0022,
      "step": 302
    },
    {
      "epoch": 18.9375,
      "grad_norm": 0.09664333466138222,
      "learning_rate": 0.00014270975699333654,
      "loss": 0.0057,
      "step": 303
    },
    {
      "epoch": 19.0,
      "grad_norm": 0.09719048721891094,
      "learning_rate": 0.00014234334542574906,
      "loss": 0.0092,
      "step": 304
    },
    {
      "epoch": 19.0625,
      "grad_norm": 0.04572304214406104,
      "learning_rate": 0.00014197623985591373,
      "loss": 0.0018,
      "step": 305
    },
    {
      "epoch": 19.125,
      "grad_norm": 0.05097257049081179,
      "learning_rate": 0.00014160844630064595,
      "loss": 0.0057,
      "step": 306
    },
    {
      "epoch": 19.1875,
      "grad_norm": 0.030553512605281084,
      "learning_rate": 0.00014123997078803707,
      "loss": 0.0026,
      "step": 307
    },
    {
      "epoch": 19.25,
      "grad_norm": 0.05445407477912121,
      "learning_rate": 0.00014087081935735564,
      "loss": 0.0057,
      "step": 308
    },
    {
      "epoch": 19.3125,
      "grad_norm": 0.06371575221400023,
      "learning_rate": 0.00014050099805894837,
      "loss": 0.0084,
      "step": 309
    },
    {
      "epoch": 19.375,
      "grad_norm": 0.06261401628069665,
      "learning_rate": 0.00014013051295414108,
      "loss": 0.0081,
      "step": 310
    },
    {
      "epoch": 19.4375,
      "grad_norm": 0.035834190161274884,
      "learning_rate": 0.00013975937011513932,
      "loss": 0.0045,
      "step": 311
    },
    {
      "epoch": 19.5,
      "grad_norm": 0.0524760162484543,
      "learning_rate": 0.00013938757562492873,
      "loss": 0.0062,
      "step": 312
    },
    {
      "epoch": 19.5625,
      "grad_norm": 0.10667039377919106,
      "learning_rate": 0.00013901513557717553,
      "loss": 0.0041,
      "step": 313
    },
    {
      "epoch": 19.625,
      "grad_norm": 0.05041143161792446,
      "learning_rate": 0.00013864205607612648,
      "loss": 0.0052,
      "step": 314
    },
    {
      "epoch": 19.6875,
      "grad_norm": 0.0424817284436791,
      "learning_rate": 0.000138268343236509,
      "loss": 0.0055,
      "step": 315
    },
    {
      "epoch": 19.75,
      "grad_norm": 0.05668813950953166,
      "learning_rate": 0.00013789400318343068,
      "loss": 0.0071,
      "step": 316
    },
    {
      "epoch": 19.8125,
      "grad_norm": 0.028939867545089322,
      "learning_rate": 0.0001375190420522792,
      "loss": 0.003,
      "step": 317
    },
    {
      "epoch": 19.875,
      "grad_norm": 0.06412293116062714,
      "learning_rate": 0.00013714346598862166,
      "loss": 0.0067,
      "step": 318
    },
    {
      "epoch": 19.9375,
      "grad_norm": 0.06349552195339572,
      "learning_rate": 0.00013676728114810367,
      "loss": 0.0068,
      "step": 319
    },
    {
      "epoch": 20.0,
      "grad_norm": 0.09285652351669382,
      "learning_rate": 0.00013639049369634876,
      "loss": 0.0108,
      "step": 320
    },
    {
      "epoch": 20.0625,
      "grad_norm": 0.038426113694616765,
      "learning_rate": 0.00013601310980885714,
      "loss": 0.0039,
      "step": 321
    },
    {
      "epoch": 20.125,
      "grad_norm": 0.03317557018896503,
      "learning_rate": 0.0001356351356709045,
      "loss": 0.003,
      "step": 322
    },
    {
      "epoch": 20.1875,
      "grad_norm": 0.059248532620137406,
      "learning_rate": 0.00013525657747744072,
      "loss": 0.0059,
      "step": 323
    },
    {
      "epoch": 20.25,
      "grad_norm": 0.05514984219384242,
      "learning_rate": 0.00013487744143298822,
      "loss": 0.004,
      "step": 324
    },
    {
      "epoch": 20.3125,
      "grad_norm": 0.041791984793344325,
      "learning_rate": 0.0001344977337515404,
      "loss": 0.0039,
      "step": 325
    },
    {
      "epoch": 20.375,
      "grad_norm": 0.06346255904291057,
      "learning_rate": 0.0001341174606564596,
      "loss": 0.0078,
      "step": 326
    },
    {
      "epoch": 20.4375,
      "grad_norm": 0.029692961959192875,
      "learning_rate": 0.00013373662838037537,
      "loss": 0.0027,
      "step": 327
    },
    {
      "epoch": 20.5,
      "grad_norm": 0.0481477761395951,
      "learning_rate": 0.00013335524316508208,
      "loss": 0.0072,
      "step": 328
    },
    {
      "epoch": 20.5625,
      "grad_norm": 0.07552298978231338,
      "learning_rate": 0.00013297331126143667,
      "loss": 0.0042,
      "step": 329
    },
    {
      "epoch": 20.625,
      "grad_norm": 0.03033536526572307,
      "learning_rate": 0.00013259083892925633,
      "loss": 0.0018,
      "step": 330
    },
    {
      "epoch": 20.6875,
      "grad_norm": 0.020436451514801952,
      "learning_rate": 0.00013220783243721572,
      "loss": 0.0018,
      "step": 331
    },
    {
      "epoch": 20.75,
      "grad_norm": 0.05066849555968109,
      "learning_rate": 0.0001318242980627444,
      "loss": 0.0068,
      "step": 332
    },
    {
      "epoch": 20.8125,
      "grad_norm": 0.062213499192457326,
      "learning_rate": 0.0001314402420919238,
      "loss": 0.0071,
      "step": 333
    },
    {
      "epoch": 20.875,
      "grad_norm": 0.0602651350512265,
      "learning_rate": 0.00013105567081938424,
      "loss": 0.0057,
      "step": 334
    },
    {
      "epoch": 20.9375,
      "grad_norm": 0.07142806102643208,
      "learning_rate": 0.00013067059054820183,
      "loss": 0.011,
      "step": 335
    },
    {
      "epoch": 21.0,
      "grad_norm": 0.11306410649612093,
      "learning_rate": 0.00013028500758979506,
      "loss": 0.0061,
      "step": 336
    },
    {
      "epoch": 21.0625,
      "grad_norm": 0.04077734941537789,
      "learning_rate": 0.00012989892826382145,
      "loss": 0.0047,
      "step": 337
    },
    {
      "epoch": 21.125,
      "grad_norm": 0.025438845626494507,
      "learning_rate": 0.00012951235889807386,
      "loss": 0.0024,
      "step": 338
    },
    {
      "epoch": 21.1875,
      "grad_norm": 0.04068069801179701,
      "learning_rate": 0.00012912530582837682,
      "loss": 0.0057,
      "step": 339
    },
    {
      "epoch": 21.25,
      "grad_norm": 0.06229646373645838,
      "learning_rate": 0.00012873777539848283,
      "loss": 0.0058,
      "step": 340
    },
    {
      "epoch": 21.3125,
      "grad_norm": 0.05047183477805069,
      "learning_rate": 0.00012834977395996818,
      "loss": 0.0073,
      "step": 341
    },
    {
      "epoch": 21.375,
      "grad_norm": 0.03503202770379194,
      "learning_rate": 0.0001279613078721289,
      "loss": 0.0035,
      "step": 342
    },
    {
      "epoch": 21.4375,
      "grad_norm": 0.02025118966232832,
      "learning_rate": 0.0001275723835018767,
      "loss": 0.0011,
      "step": 343
    },
    {
      "epoch": 21.5,
      "grad_norm": 0.04264241354679371,
      "learning_rate": 0.0001271830072236343,
      "loss": 0.0048,
      "step": 344
    },
    {
      "epoch": 21.5625,
      "grad_norm": 0.059381097002962034,
      "learning_rate": 0.0001267931854192313,
      "loss": 0.0065,
      "step": 345
    },
    {
      "epoch": 21.625,
      "grad_norm": 0.03970878770263772,
      "learning_rate": 0.0001264029244777993,
      "loss": 0.0052,
      "step": 346
    },
    {
      "epoch": 21.6875,
      "grad_norm": 0.04854204024259956,
      "learning_rate": 0.00012601223079566743,
      "loss": 0.0036,
      "step": 347
    },
    {
      "epoch": 21.75,
      "grad_norm": 0.06683236405906955,
      "learning_rate": 0.00012562111077625722,
      "loss": 0.0078,
      "step": 348
    },
    {
      "epoch": 21.8125,
      "grad_norm": 0.030148690304593224,
      "learning_rate": 0.000125229570829978,
      "loss": 0.0044,
      "step": 349
    },
    {
      "epoch": 21.875,
      "grad_norm": 0.043501391695934324,
      "learning_rate": 0.0001248376173741215,
      "loss": 0.0044,
      "step": 350
    },
    {
      "epoch": 21.9375,
      "grad_norm": 0.03558605264264846,
      "learning_rate": 0.00012444525683275688,
      "loss": 0.0045,
      "step": 351
    },
    {
      "epoch": 22.0,
      "grad_norm": 0.04991180242909837,
      "learning_rate": 0.00012405249563662537,
      "loss": 0.0067,
      "step": 352
    },
    {
      "epoch": 22.0625,
      "grad_norm": 0.037891384595762335,
      "learning_rate": 0.00012365934022303491,
      "loss": 0.006,
      "step": 353
    },
    {
      "epoch": 22.125,
      "grad_norm": 0.029393462316367327,
      "learning_rate": 0.00012326579703575462,
      "loss": 0.0059,
      "step": 354
    },
    {
      "epoch": 22.1875,
      "grad_norm": 0.02302636254784467,
      "learning_rate": 0.00012287187252490913,
      "loss": 0.002,
      "step": 355
    },
    {
      "epoch": 22.25,
      "grad_norm": 0.027267824953848163,
      "learning_rate": 0.00012247757314687297,
      "loss": 0.004,
      "step": 356
    },
    {
      "epoch": 22.3125,
      "grad_norm": 0.018193429790619855,
      "learning_rate": 0.00012208290536416463,
      "loss": 0.0016,
      "step": 357
    },
    {
      "epoch": 22.375,
      "grad_norm": 0.03843454617816289,
      "learning_rate": 0.00012168787564534078,
      "loss": 0.0044,
      "step": 358
    },
    {
      "epoch": 22.4375,
      "grad_norm": 0.033400388055823266,
      "learning_rate": 0.0001212924904648902,
      "loss": 0.0047,
      "step": 359
    },
    {
      "epoch": 22.5,
      "grad_norm": 0.026909120960616175,
      "learning_rate": 0.00012089675630312754,
      "loss": 0.0039,
      "step": 360
    },
    {
      "epoch": 22.5625,
      "grad_norm": 0.023045776166397722,
      "learning_rate": 0.00012050067964608724,
      "loss": 0.0022,
      "step": 361
    },
    {
      "epoch": 22.625,
      "grad_norm": 0.03806155810130501,
      "learning_rate": 0.00012010426698541728,
      "loss": 0.0052,
      "step": 362
    },
    {
      "epoch": 22.6875,
      "grad_norm": 0.018336649301123106,
      "learning_rate": 0.0001197075248182726,
      "loss": 0.0015,
      "step": 363
    },
    {
      "epoch": 22.75,
      "grad_norm": 0.04461178752361949,
      "learning_rate": 0.00011931045964720881,
      "loss": 0.0049,
      "step": 364
    },
    {
      "epoch": 22.8125,
      "grad_norm": 0.03716629926558807,
      "learning_rate": 0.00011891307798007536,
      "loss": 0.0051,
      "step": 365
    },
    {
      "epoch": 22.875,
      "grad_norm": 0.10941698413421153,
      "learning_rate": 0.00011851538632990921,
      "loss": 0.0061,
      "step": 366
    },
    {
      "epoch": 22.9375,
      "grad_norm": 0.031543122025977796,
      "learning_rate": 0.00011811739121482777,
      "loss": 0.0032,
      "step": 367
    },
    {
      "epoch": 23.0,
      "grad_norm": 0.03807850455755241,
      "learning_rate": 0.0001177190991579223,
      "loss": 0.0054,
      "step": 368
    },
    {
      "epoch": 23.0625,
      "grad_norm": 0.0691977966708894,
      "learning_rate": 0.00011732051668715081,
      "loss": 0.0077,
      "step": 369
    },
    {
      "epoch": 23.125,
      "grad_norm": 0.04515689836142195,
      "learning_rate": 0.00011692165033523117,
      "loss": 0.0057,
      "step": 370
    },
    {
      "epoch": 23.1875,
      "grad_norm": 0.03928887922524319,
      "learning_rate": 0.00011652250663953415,
      "loss": 0.0055,
      "step": 371
    },
    {
      "epoch": 23.25,
      "grad_norm": 0.02118235648867606,
      "learning_rate": 0.00011612309214197599,
      "loss": 0.0019,
      "step": 372
    },
    {
      "epoch": 23.3125,
      "grad_norm": 0.014975510606827147,
      "learning_rate": 0.00011572341338891144,
      "loss": 0.0013,
      "step": 373
    },
    {
      "epoch": 23.375,
      "grad_norm": 0.026742366673437934,
      "learning_rate": 0.00011532347693102632,
      "loss": 0.002,
      "step": 374
    },
    {
      "epoch": 23.4375,
      "grad_norm": 0.13222981379526266,
      "learning_rate": 0.00011492328932323022,
      "loss": 0.0065,
      "step": 375
    },
    {
      "epoch": 23.5,
      "grad_norm": 0.030786902568145764,
      "learning_rate": 0.00011452285712454904,
      "loss": 0.0032,
      "step": 376
    },
    {
      "epoch": 23.5625,
      "grad_norm": 0.04300178598048733,
      "learning_rate": 0.00011412218689801748,
      "loss": 0.0061,
      "step": 377
    },
    {
      "epoch": 23.625,
      "grad_norm": 0.031601651854756344,
      "learning_rate": 0.00011372128521057155,
      "loss": 0.0037,
      "step": 378
    },
    {
      "epoch": 23.6875,
      "grad_norm": 0.07785976030991328,
      "learning_rate": 0.00011332015863294076,
      "loss": 0.0051,
      "step": 379
    },
    {
      "epoch": 23.75,
      "grad_norm": 0.02462026239901167,
      "learning_rate": 0.00011291881373954065,
      "loss": 0.0024,
      "step": 380
    },
    {
      "epoch": 23.8125,
      "grad_norm": 0.015040874194919356,
      "learning_rate": 0.00011251725710836489,
      "loss": 0.0011,
      "step": 381
    },
    {
      "epoch": 23.875,
      "grad_norm": 0.013797157919588528,
      "learning_rate": 0.00011211549532087749,
      "loss": 0.0012,
      "step": 382
    },
    {
      "epoch": 23.9375,
      "grad_norm": 0.028167968760020725,
      "learning_rate": 0.00011171353496190498,
      "loss": 0.0032,
      "step": 383
    },
    {
      "epoch": 24.0,
      "grad_norm": 0.05318582105382964,
      "learning_rate": 0.00011131138261952845,
      "loss": 0.0093,
      "step": 384
    },
    {
      "epoch": 24.0625,
      "grad_norm": 0.027391464358218053,
      "learning_rate": 0.00011090904488497549,
      "loss": 0.0031,
      "step": 385
    },
    {
      "epoch": 24.125,
      "grad_norm": 0.03229113172070876,
      "learning_rate": 0.0001105065283525124,
      "loss": 0.0037,
      "step": 386
    },
    {
      "epoch": 24.1875,
      "grad_norm": 0.030127780485989034,
      "learning_rate": 0.00011010383961933581,
      "loss": 0.0049,
      "step": 387
    },
    {
      "epoch": 24.25,
      "grad_norm": 0.013416894665755037,
      "learning_rate": 0.00010970098528546481,
      "loss": 0.0002,
      "step": 388
    },
    {
      "epoch": 24.3125,
      "grad_norm": 0.013218671166503056,
      "learning_rate": 0.00010929797195363259,
      "loss": 0.0012,
      "step": 389
    },
    {
      "epoch": 24.375,
      "grad_norm": 0.03798660609085839,
      "learning_rate": 0.0001088948062291783,
      "loss": 0.0066,
      "step": 390
    },
    {
      "epoch": 24.4375,
      "grad_norm": 0.01875425458258293,
      "learning_rate": 0.00010849149471993882,
      "loss": 0.0014,
      "step": 391
    },
    {
      "epoch": 24.5,
      "grad_norm": 0.03284107465970819,
      "learning_rate": 0.00010808804403614043,
      "loss": 0.0051,
      "step": 392
    },
    {
      "epoch": 24.5625,
      "grad_norm": 0.0318856663182744,
      "learning_rate": 0.00010768446079029044,
      "loss": 0.0041,
      "step": 393
    },
    {
      "epoch": 24.625,
      "grad_norm": 0.03886319257158687,
      "learning_rate": 0.0001072807515970688,
      "loss": 0.0055,
      "step": 394
    },
    {
      "epoch": 24.6875,
      "grad_norm": 0.0240690438810162,
      "learning_rate": 0.00010687692307321984,
      "loss": 0.0026,
      "step": 395
    },
    {
      "epoch": 24.75,
      "grad_norm": 0.03153859131907879,
      "learning_rate": 0.00010647298183744359,
      "loss": 0.0037,
      "step": 396
    },
    {
      "epoch": 24.8125,
      "grad_norm": 0.028754869900780036,
      "learning_rate": 0.00010606893451028743,
      "loss": 0.0046,
      "step": 397
    },
    {
      "epoch": 24.875,
      "grad_norm": 0.036835349609799346,
      "learning_rate": 0.00010566478771403763,
      "loss": 0.0059,
      "step": 398
    },
    {
      "epoch": 24.9375,
      "grad_norm": 0.03742416571548899,
      "learning_rate": 0.00010526054807261067,
      "loss": 0.0057,
      "step": 399
    },
    {
      "epoch": 25.0,
      "grad_norm": 0.019217319407847776,
      "learning_rate": 0.00010485622221144484,
      "loss": 0.0019,
      "step": 400
    },
    {
      "epoch": 25.0625,
      "grad_norm": 0.0181018400795553,
      "learning_rate": 0.00010445181675739144,
      "loss": 0.0019,
      "step": 401
    },
    {
      "epoch": 25.125,
      "grad_norm": 0.015082303122275808,
      "learning_rate": 0.00010404733833860639,
      "loss": 0.001,
      "step": 402
    },
    {
      "epoch": 25.1875,
      "grad_norm": 0.024352538732525977,
      "learning_rate": 0.00010364279358444144,
      "loss": 0.0031,
      "step": 403
    },
    {
      "epoch": 25.25,
      "grad_norm": 0.03063940290250688,
      "learning_rate": 0.00010323818912533561,
      "loss": 0.0041,
      "step": 404
    },
    {
      "epoch": 25.3125,
      "grad_norm": 0.02851015433219716,
      "learning_rate": 0.00010283353159270643,
      "loss": 0.0046,
      "step": 405
    },
    {
      "epoch": 25.375,
      "grad_norm": 0.017078363597566818,
      "learning_rate": 0.00010242882761884131,
      "loss": 0.0017,
      "step": 406
    },
    {
      "epoch": 25.4375,
      "grad_norm": 0.015453958897260991,
      "learning_rate": 0.00010202408383678888,
      "loss": 0.0011,
      "step": 407
    },
    {
      "epoch": 25.5,
      "grad_norm": 0.028055877963511246,
      "learning_rate": 0.00010161930688025017,
      "loss": 0.0029,
      "step": 408
    },
    {
      "epoch": 25.5625,
      "grad_norm": 0.003095926902821122,
      "learning_rate": 0.0001012145033834699,
      "loss": 0.0001,
      "step": 409
    },
    {
      "epoch": 25.625,
      "grad_norm": 0.04032729610564608,
      "learning_rate": 0.00010080967998112787,
      "loss": 0.0084,
      "step": 410
    },
    {
      "epoch": 25.6875,
      "grad_norm": 0.03694941225075249,
      "learning_rate": 0.00010040484330823006,
      "loss": 0.0058,
      "step": 411
    },
    {
      "epoch": 25.75,
      "grad_norm": 0.040456452234492964,
      "learning_rate": 0.0001,
      "loss": 0.0067,
      "step": 412
    },
    {
      "epoch": 25.8125,
      "grad_norm": 0.02482728072796654,
      "learning_rate": 9.959515669176996e-05,
      "loss": 0.0027,
      "step": 413
    },
    {
      "epoch": 25.875,
      "grad_norm": 0.046729978524535094,
      "learning_rate": 9.919032001887215e-05,
      "loss": 0.0088,
      "step": 414
    },
    {
      "epoch": 25.9375,
      "grad_norm": 0.03304233160758997,
      "learning_rate": 9.878549661653012e-05,
      "loss": 0.0043,
      "step": 415
    },
    {
      "epoch": 26.0,
      "grad_norm": 0.03375904734753774,
      "learning_rate": 9.838069311974986e-05,
      "loss": 0.0044,
      "step": 416
    },
    {
      "epoch": 26.0625,
      "grad_norm": 0.01611151201817884,
      "learning_rate": 9.797591616321114e-05,
      "loss": 0.0016,
      "step": 417
    },
    {
      "epoch": 26.125,
      "grad_norm": 0.01947904547375526,
      "learning_rate": 9.757117238115871e-05,
      "loss": 0.0017,
      "step": 418
    },
    {
      "epoch": 26.1875,
      "grad_norm": 0.020613065991069244,
      "learning_rate": 9.716646840729361e-05,
      "loss": 0.0022,
      "step": 419
    },
    {
      "epoch": 26.25,
      "grad_norm": 0.028039730983769738,
      "learning_rate": 9.676181087466444e-05,
      "loss": 0.006,
      "step": 420
    },
    {
      "epoch": 26.3125,
      "grad_norm": 0.03218256334320815,
      "learning_rate": 9.635720641555858e-05,
      "loss": 0.0047,
      "step": 421
    },
    {
      "epoch": 26.375,
      "grad_norm": 0.017143091500272017,
      "learning_rate": 9.595266166139366e-05,
      "loss": 0.0016,
      "step": 422
    },
    {
      "epoch": 26.4375,
      "grad_norm": 0.024547894378123793,
      "learning_rate": 9.554818324260859e-05,
      "loss": 0.0024,
      "step": 423
    },
    {
      "epoch": 26.5,
      "grad_norm": 0.0017051489220323973,
      "learning_rate": 9.514377778855521e-05,
      "loss": 0.0001,
      "step": 424
    },
    {
      "epoch": 26.5625,
      "grad_norm": 0.03211276051099752,
      "learning_rate": 9.473945192738933e-05,
      "loss": 0.0043,
      "step": 425
    },
    {
      "epoch": 26.625,
      "grad_norm": 0.029672554162065486,
      "learning_rate": 9.433521228596237e-05,
      "loss": 0.0049,
      "step": 426
    },
    {
      "epoch": 26.6875,
      "grad_norm": 0.02974958108819204,
      "learning_rate": 9.393106548971256e-05,
      "loss": 0.0046,
      "step": 427
    },
    {
      "epoch": 26.75,
      "grad_norm": 0.02994562158825566,
      "learning_rate": 9.352701816255643e-05,
      "loss": 0.0032,
      "step": 428
    },
    {
      "epoch": 26.8125,
      "grad_norm": 0.021945113737497645,
      "learning_rate": 9.312307692678017e-05,
      "loss": 0.0025,
      "step": 429
    },
    {
      "epoch": 26.875,
      "grad_norm": 0.039440593653657555,
      "learning_rate": 9.27192484029312e-05,
      "loss": 0.0064,
      "step": 430
    },
    {
      "epoch": 26.9375,
      "grad_norm": 0.034294640588989456,
      "learning_rate": 9.231553920970958e-05,
      "loss": 0.0067,
      "step": 431
    },
    {
      "epoch": 27.0,
      "grad_norm": 0.045874360009063225,
      "learning_rate": 9.19119559638596e-05,
      "loss": 0.0091,
      "step": 432
    },
    {
      "epoch": 27.0625,
      "grad_norm": 0.029633783585542237,
      "learning_rate": 9.150850528006119e-05,
      "loss": 0.0052,
      "step": 433
    },
    {
      "epoch": 27.125,
      "grad_norm": 0.01916779629257807,
      "learning_rate": 9.110519377082172e-05,
      "loss": 0.0023,
      "step": 434
    },
    {
      "epoch": 27.1875,
      "grad_norm": 0.029144898881918332,
      "learning_rate": 9.070202804636745e-05,
      "loss": 0.0041,
      "step": 435
    },
    {
      "epoch": 27.25,
      "grad_norm": 0.02042965784384683,
      "learning_rate": 9.02990147145352e-05,
      "loss": 0.0036,
      "step": 436
    },
    {
      "epoch": 27.3125,
      "grad_norm": 0.03052607215886322,
      "learning_rate": 8.98961603806642e-05,
      "loss": 0.0041,
      "step": 437
    },
    {
      "epoch": 27.375,
      "grad_norm": 0.011632711843821695,
      "learning_rate": 8.949347164748762e-05,
      "loss": 0.0009,
      "step": 438
    },
    {
      "epoch": 27.4375,
      "grad_norm": 0.02552057287290442,
      "learning_rate": 8.909095511502452e-05,
      "loss": 0.0049,
      "step": 439
    },
    {
      "epoch": 27.5,
      "grad_norm": 0.022469807056043515,
      "learning_rate": 8.868861738047158e-05,
      "loss": 0.0031,
      "step": 440
    },
    {
      "epoch": 27.5625,
      "grad_norm": 0.023608406072661937,
      "learning_rate": 8.828646503809504e-05,
      "loss": 0.0022,
      "step": 441
    },
    {
      "epoch": 27.625,
      "grad_norm": 0.03344361764249166,
      "learning_rate": 8.788450467912255e-05,
      "loss": 0.006,
      "step": 442
    },
    {
      "epoch": 27.6875,
      "grad_norm": 0.021786698400709355,
      "learning_rate": 8.748274289163514e-05,
      "loss": 0.0024,
      "step": 443
    },
    {
      "epoch": 27.75,
      "grad_norm": 0.030927337267203,
      "learning_rate": 8.70811862604594e-05,
      "loss": 0.0044,
      "step": 444
    },
    {
      "epoch": 27.8125,
      "grad_norm": 0.022165776937843076,
      "learning_rate": 8.667984136705928e-05,
      "loss": 0.0046,
      "step": 445
    },
    {
      "epoch": 27.875,
      "grad_norm": 0.01725502880403943,
      "learning_rate": 8.627871478942851e-05,
      "loss": 0.0028,
      "step": 446
    },
    {
      "epoch": 27.9375,
      "grad_norm": 0.020259168330017313,
      "learning_rate": 8.587781310198255e-05,
      "loss": 0.0025,
      "step": 447
    },
    {
      "epoch": 28.0,
      "grad_norm": 0.022633252879775623,
      "learning_rate": 8.5477142875451e-05,
      "loss": 0.0027,
      "step": 448
    },
    {
      "epoch": 28.0625,
      "grad_norm": 0.019201479355893957,
      "learning_rate": 8.507671067676979e-05,
      "loss": 0.0031,
      "step": 449
    },
    {
      "epoch": 28.125,
      "grad_norm": 0.0009559951000656043,
      "learning_rate": 8.467652306897369e-05,
      "loss": 0.0,
      "step": 450
    },
    {
      "epoch": 28.1875,
      "grad_norm": 0.023486468299252494,
      "learning_rate": 8.427658661108857e-05,
      "loss": 0.0031,
      "step": 451
    },
    {
      "epoch": 28.25,
      "grad_norm": 0.015284012764583016,
      "learning_rate": 8.387690785802402e-05,
      "loss": 0.0013,
      "step": 452
    },
    {
      "epoch": 28.3125,
      "grad_norm": 0.021772765753157987,
      "learning_rate": 8.347749336046586e-05,
      "loss": 0.0035,
      "step": 453
    },
    {
      "epoch": 28.375,
      "grad_norm": 0.011499608138854372,
      "learning_rate": 8.307834966476884e-05,
      "loss": 0.0009,
      "step": 454
    },
    {
      "epoch": 28.4375,
      "grad_norm": 0.014877751008139171,
      "learning_rate": 8.267948331284923e-05,
      "loss": 0.0032,
      "step": 455
    },
    {
      "epoch": 28.5,
      "grad_norm": 0.020317448663989896,
      "learning_rate": 8.228090084207774e-05,
      "loss": 0.0051,
      "step": 456
    },
    {
      "epoch": 28.5625,
      "grad_norm": 0.011464514826041898,
      "learning_rate": 8.188260878517224e-05,
      "loss": 0.0027,
      "step": 457
    },
    {
      "epoch": 28.625,
      "grad_norm": 0.02729549690762567,
      "learning_rate": 8.14846136700908e-05,
      "loss": 0.0037,
      "step": 458
    },
    {
      "epoch": 28.6875,
      "grad_norm": 0.016332439503834673,
      "learning_rate": 8.108692201992465e-05,
      "loss": 0.0019,
      "step": 459
    },
    {
      "epoch": 28.75,
      "grad_norm": 0.020328088646763758,
      "learning_rate": 8.068954035279121e-05,
      "loss": 0.0045,
      "step": 460
    },
    {
      "epoch": 28.8125,
      "grad_norm": 0.04010266322110663,
      "learning_rate": 8.02924751817274e-05,
      "loss": 0.0085,
      "step": 461
    },
    {
      "epoch": 28.875,
      "grad_norm": 0.024097139912002265,
      "learning_rate": 7.989573301458273e-05,
      "loss": 0.0042,
      "step": 462
    },
    {
      "epoch": 28.9375,
      "grad_norm": 0.036490969223195995,
      "learning_rate": 7.949932035391278e-05,
      "loss": 0.0053,
      "step": 463
    },
    {
      "epoch": 29.0,
      "grad_norm": 0.03282570248481929,
      "learning_rate": 7.91032436968725e-05,
      "loss": 0.0045,
      "step": 464
    },
    {
      "epoch": 29.0625,
      "grad_norm": 0.01039285670497238,
      "learning_rate": 7.870750953510984e-05,
      "loss": 0.0006,
      "step": 465
    },
    {
      "epoch": 29.125,
      "grad_norm": 0.01787195379230809,
      "learning_rate": 7.831212435465924e-05,
      "loss": 0.0022,
      "step": 466
    },
    {
      "epoch": 29.1875,
      "grad_norm": 0.02385197046230436,
      "learning_rate": 7.79170946358354e-05,
      "loss": 0.0039,
      "step": 467
    },
    {
      "epoch": 29.25,
      "grad_norm": 0.018022930648381282,
      "learning_rate": 7.75224268531271e-05,
      "loss": 0.0017,
      "step": 468
    },
    {
      "epoch": 29.3125,
      "grad_norm": 0.027503021332335603,
      "learning_rate": 7.71281274750909e-05,
      "loss": 0.0036,
      "step": 469
    },
    {
      "epoch": 29.375,
      "grad_norm": 0.009996199602644211,
      "learning_rate": 7.673420296424541e-05,
      "loss": 0.0009,
      "step": 470
    },
    {
      "epoch": 29.4375,
      "grad_norm": 0.018797312474179668,
      "learning_rate": 7.634065977696511e-05,
      "loss": 0.0026,
      "step": 471
    },
    {
      "epoch": 29.5,
      "grad_norm": 0.028684614773920647,
      "learning_rate": 7.594750436337467e-05,
      "loss": 0.004,
      "step": 472
    },
    {
      "epoch": 29.5625,
      "grad_norm": 0.020479906972316966,
      "learning_rate": 7.555474316724313e-05,
      "loss": 0.0013,
      "step": 473
    },
    {
      "epoch": 29.625,
      "grad_norm": 0.025765673045083433,
      "learning_rate": 7.516238262587851e-05,
      "loss": 0.0033,
      "step": 474
    },
    {
      "epoch": 29.6875,
      "grad_norm": 0.045631738026105505,
      "learning_rate": 7.4770429170022e-05,
      "loss": 0.0109,
      "step": 475
    },
    {
      "epoch": 29.75,
      "grad_norm": 0.02215455016152165,
      "learning_rate": 7.437888922374276e-05,
      "loss": 0.003,
      "step": 476
    },
    {
      "epoch": 29.8125,
      "grad_norm": 0.03072966417955352,
      "learning_rate": 7.398776920433258e-05,
      "loss": 0.0055,
      "step": 477
    },
    {
      "epoch": 29.875,
      "grad_norm": 0.017660307308308684,
      "learning_rate": 7.35970755222007e-05,
      "loss": 0.0016,
      "step": 478
    },
    {
      "epoch": 29.9375,
      "grad_norm": 0.03283219124830058,
      "learning_rate": 7.320681458076871e-05,
      "loss": 0.0052,
      "step": 479
    },
    {
      "epoch": 30.0,
      "grad_norm": 0.03754750336262435,
      "learning_rate": 7.281699277636572e-05,
      "loss": 0.0059,
      "step": 480
    },
    {
      "epoch": 30.0625,
      "grad_norm": 0.0232647542444026,
      "learning_rate": 7.242761649812335e-05,
      "loss": 0.0026,
      "step": 481
    },
    {
      "epoch": 30.125,
      "grad_norm": 0.010280990700798192,
      "learning_rate": 7.20386921278711e-05,
      "loss": 0.0008,
      "step": 482
    },
    {
      "epoch": 30.1875,
      "grad_norm": 0.023701325742362663,
      "learning_rate": 7.165022604003186e-05,
      "loss": 0.0045,
      "step": 483
    },
    {
      "epoch": 30.25,
      "grad_norm": 0.02723103090348779,
      "learning_rate": 7.126222460151719e-05,
      "loss": 0.0036,
      "step": 484
    },
    {
      "epoch": 30.3125,
      "grad_norm": 0.022573066749760014,
      "learning_rate": 7.08746941716232e-05,
      "loss": 0.0028,
      "step": 485
    },
    {
      "epoch": 30.375,
      "grad_norm": 0.015662563973923037,
      "learning_rate": 7.048764110192618e-05,
      "loss": 0.0027,
      "step": 486
    },
    {
      "epoch": 30.4375,
      "grad_norm": 0.03142186630699031,
      "learning_rate": 7.010107173617857e-05,
      "loss": 0.0046,
      "step": 487
    },
    {
      "epoch": 30.5,
      "grad_norm": 0.010153995879187198,
      "learning_rate": 6.971499241020495e-05,
      "loss": 0.001,
      "step": 488
    },
    {
      "epoch": 30.5625,
      "grad_norm": 0.027760410207358414,
      "learning_rate": 6.932940945179818e-05,
      "loss": 0.0035,
      "step": 489
    },
    {
      "epoch": 30.625,
      "grad_norm": 0.02569158532689841,
      "learning_rate": 6.894432918061579e-05,
      "loss": 0.0042,
      "step": 490
    },
    {
      "epoch": 30.6875,
      "grad_norm": 0.03360414388929754,
      "learning_rate": 6.855975790807623e-05,
      "loss": 0.0052,
      "step": 491
    },
    {
      "epoch": 30.75,
      "grad_norm": 0.03809908186634646,
      "learning_rate": 6.817570193725564e-05,
      "loss": 0.0046,
      "step": 492
    },
    {
      "epoch": 30.8125,
      "grad_norm": 0.01683175407924854,
      "learning_rate": 6.77921675627843e-05,
      "loss": 0.0018,
      "step": 493
    },
    {
      "epoch": 30.875,
      "grad_norm": 0.025360976440516984,
      "learning_rate": 6.740916107074372e-05,
      "loss": 0.0037,
      "step": 494
    },
    {
      "epoch": 30.9375,
      "grad_norm": 0.023768876753983793,
      "learning_rate": 6.702668873856338e-05,
      "loss": 0.0031,
      "step": 495
    },
    {
      "epoch": 31.0,
      "grad_norm": 0.03989426250367044,
      "learning_rate": 6.664475683491796e-05,
      "loss": 0.008,
      "step": 496
    },
    {
      "epoch": 31.0625,
      "grad_norm": 0.010352984240734544,
      "learning_rate": 6.626337161962461e-05,
      "loss": 0.0008,
      "step": 497
    },
    {
      "epoch": 31.125,
      "grad_norm": 0.021953465861397056,
      "learning_rate": 6.588253934354039e-05,
      "loss": 0.0031,
      "step": 498
    },
    {
      "epoch": 31.1875,
      "grad_norm": 0.019549539753814318,
      "learning_rate": 6.550226624845961e-05,
      "loss": 0.0019,
      "step": 499
    },
    {
      "epoch": 31.25,
      "grad_norm": 0.02490729225984056,
      "learning_rate": 6.512255856701177e-05,
      "loss": 0.0027,
      "step": 500
    },
    {
      "epoch": 31.3125,
      "grad_norm": 0.019764190511534645,
      "learning_rate": 6.474342252255927e-05,
      "loss": 0.0027,
      "step": 501
    },
    {
      "epoch": 31.375,
      "grad_norm": 0.03215845958993517,
      "learning_rate": 6.43648643290955e-05,
      "loss": 0.0069,
      "step": 502
    },
    {
      "epoch": 31.4375,
      "grad_norm": 0.0244310478279658,
      "learning_rate": 6.398689019114289e-05,
      "loss": 0.0033,
      "step": 503
    },
    {
      "epoch": 31.5,
      "grad_norm": 0.018051312193521227,
      "learning_rate": 6.360950630365126e-05,
      "loss": 0.0025,
      "step": 504
    },
    {
      "epoch": 31.5625,
      "grad_norm": 0.00962728862133209,
      "learning_rate": 6.323271885189635e-05,
      "loss": 0.0007,
      "step": 505
    },
    {
      "epoch": 31.625,
      "grad_norm": 0.013047383716533472,
      "learning_rate": 6.285653401137837e-05,
      "loss": 0.0011,
      "step": 506
    },
    {
      "epoch": 31.6875,
      "grad_norm": 0.026493369355222587,
      "learning_rate": 6.248095794772079e-05,
      "loss": 0.0035,
      "step": 507
    },
    {
      "epoch": 31.75,
      "grad_norm": 0.023599070967015707,
      "learning_rate": 6.210599681656933e-05,
      "loss": 0.0039,
      "step": 508
    },
    {
      "epoch": 31.8125,
      "grad_norm": 0.02343395190062143,
      "learning_rate": 6.173165676349103e-05,
      "loss": 0.0037,
      "step": 509
    },
    {
      "epoch": 31.875,
      "grad_norm": 0.01983520736229701,
      "learning_rate": 6.135794392387353e-05,
      "loss": 0.0025,
      "step": 510
    },
    {
      "epoch": 31.9375,
      "grad_norm": 0.0318092460128266,
      "learning_rate": 6.0984864422824496e-05,
      "loss": 0.0045,
      "step": 511
    },
    {
      "epoch": 32.0,
      "grad_norm": 0.04242840537695658,
      "learning_rate": 6.061242437507131e-05,
      "loss": 0.009,
      "step": 512
    },
    {
      "epoch": 32.0625,
      "grad_norm": 0.02009402381973668,
      "learning_rate": 6.024062988486072e-05,
      "loss": 0.0032,
      "step": 513
    },
    {
      "epoch": 32.125,
      "grad_norm": 0.011445958109287144,
      "learning_rate": 5.986948704585895e-05,
      "loss": 0.0024,
      "step": 514
    },
    {
      "epoch": 32.1875,
      "grad_norm": 0.016981136262338714,
      "learning_rate": 5.949900194105167e-05,
      "loss": 0.0019,
      "step": 515
    },
    {
      "epoch": 32.25,
      "grad_norm": 0.03620971462935574,
      "learning_rate": 5.9129180642644414e-05,
      "loss": 0.0061,
      "step": 516
    },
    {
      "epoch": 32.3125,
      "grad_norm": 0.022511016593835396,
      "learning_rate": 5.8760029211962954e-05,
      "loss": 0.0045,
      "step": 517
    },
    {
      "epoch": 32.375,
      "grad_norm": 0.027254211809580164,
      "learning_rate": 5.839155369935407e-05,
      "loss": 0.0039,
      "step": 518
    },
    {
      "epoch": 32.4375,
      "grad_norm": 0.02011520557513548,
      "learning_rate": 5.802376014408632e-05,
      "loss": 0.002,
      "step": 519
    },
    {
      "epoch": 32.5,
      "grad_norm": 0.029228301710033645,
      "learning_rate": 5.765665457425102e-05,
      "loss": 0.0051,
      "step": 520
    },
    {
      "epoch": 32.5625,
      "grad_norm": 0.01687690442346917,
      "learning_rate": 5.729024300666349e-05,
      "loss": 0.0011,
      "step": 521
    },
    {
      "epoch": 32.625,
      "grad_norm": 0.02623446617433343,
      "learning_rate": 5.6924531446764504e-05,
      "loss": 0.0038,
      "step": 522
    },
    {
      "epoch": 32.6875,
      "grad_norm": 0.018653711010246068,
      "learning_rate": 5.6559525888521815e-05,
      "loss": 0.0019,
      "step": 523
    },
    {
      "epoch": 32.75,
      "grad_norm": 0.01760737201183457,
      "learning_rate": 5.6195232314331766e-05,
      "loss": 0.002,
      "step": 524
    },
    {
      "epoch": 32.8125,
      "grad_norm": 0.0142233875057376,
      "learning_rate": 5.5831656694921465e-05,
      "loss": 0.0017,
      "step": 525
    },
    {
      "epoch": 32.875,
      "grad_norm": 0.02523792596742103,
      "learning_rate": 5.5468804989250786e-05,
      "loss": 0.0034,
      "step": 526
    },
    {
      "epoch": 32.9375,
      "grad_norm": 0.017246202374674464,
      "learning_rate": 5.510668314441474e-05,
      "loss": 0.0024,
      "step": 527
    },
    {
      "epoch": 33.0,
      "grad_norm": 0.029238162207787115,
      "learning_rate": 5.474529709554612e-05,
      "loss": 0.0064,
      "step": 528
    },
    {
      "epoch": 33.0625,
      "grad_norm": 0.019900275674923293,
      "learning_rate": 5.438465276571796e-05,
      "loss": 0.0024,
      "step": 529
    },
    {
      "epoch": 33.125,
      "grad_norm": 0.022436670149349774,
      "learning_rate": 5.402475606584669e-05,
      "loss": 0.0024,
      "step": 530
    },
    {
      "epoch": 33.1875,
      "grad_norm": 0.026570662311707075,
      "learning_rate": 5.366561289459512e-05,
      "loss": 0.0038,
      "step": 531
    },
    {
      "epoch": 33.25,
      "grad_norm": 0.007873326312030244,
      "learning_rate": 5.3307229138275936e-05,
      "loss": 0.0007,
      "step": 532
    },
    {
      "epoch": 33.3125,
      "grad_norm": 0.010633636981874433,
      "learning_rate": 5.2949610670755e-05,
      "loss": 0.0007,
      "step": 533
    },
    {
      "epoch": 33.375,
      "grad_norm": 0.021848223370386288,
      "learning_rate": 5.259276335335521e-05,
      "loss": 0.0032,
      "step": 534
    },
    {
      "epoch": 33.4375,
      "grad_norm": 0.02671646079052259,
      "learning_rate": 5.223669303476041e-05,
      "loss": 0.004,
      "step": 535
    },
    {
      "epoch": 33.5,
      "grad_norm": 0.025191000069873352,
      "learning_rate": 5.1881405550919493e-05,
      "loss": 0.0038,
      "step": 536
    },
    {
      "epoch": 33.5625,
      "grad_norm": 0.02686634304764754,
      "learning_rate": 5.152690672495091e-05,
      "loss": 0.0036,
      "step": 537
    },
    {
      "epoch": 33.625,
      "grad_norm": 0.016683908355091075,
      "learning_rate": 5.117320236704697e-05,
      "loss": 0.0013,
      "step": 538
    },
    {
      "epoch": 33.6875,
      "grad_norm": 0.025955098003678897,
      "learning_rate": 5.08202982743788e-05,
      "loss": 0.0044,
      "step": 539
    },
    {
      "epoch": 33.75,
      "grad_norm": 0.03159147242134654,
      "learning_rate": 5.0468200231001286e-05,
      "loss": 0.0053,
      "step": 540
    },
    {
      "epoch": 33.8125,
      "grad_norm": 0.020310055743322283,
      "learning_rate": 5.01169140077582e-05,
      "loss": 0.0047,
      "step": 541
    },
    {
      "epoch": 33.875,
      "grad_norm": 0.02010895023321247,
      "learning_rate": 4.976644536218783e-05,
      "loss": 0.0022,
      "step": 542
    },
    {
      "epoch": 33.9375,
      "grad_norm": 0.030234549444115102,
      "learning_rate": 4.9416800038428324e-05,
      "loss": 0.007,
      "step": 543
    },
    {
      "epoch": 34.0,
      "grad_norm": 0.028337369877458187,
      "learning_rate": 4.9067983767123736e-05,
      "loss": 0.0037,
      "step": 544
    },
    {
      "epoch": 34.0625,
      "grad_norm": 0.016650841753729186,
      "learning_rate": 4.8720002265330015e-05,
      "loss": 0.0019,
      "step": 545
    },
    {
      "epoch": 34.125,
      "grad_norm": 0.024415205244997427,
      "learning_rate": 4.837286123642141e-05,
      "loss": 0.0039,
      "step": 546
    },
    {
      "epoch": 34.1875,
      "grad_norm": 0.030479700286681143,
      "learning_rate": 4.8026566369996926e-05,
      "loss": 0.0058,
      "step": 547
    },
    {
      "epoch": 34.25,
      "grad_norm": 0.015608780057852139,
      "learning_rate": 4.768112334178699e-05,
      "loss": 0.0024,
      "step": 548
    },
    {
      "epoch": 34.3125,
      "grad_norm": 0.015068945094923566,
      "learning_rate": 4.733653781356055e-05,
      "loss": 0.0022,
      "step": 549
    },
    {
      "epoch": 34.375,
      "grad_norm": 0.026394448783195724,
      "learning_rate": 4.699281543303222e-05,
      "loss": 0.0033,
      "step": 550
    },
    {
      "epoch": 34.4375,
      "grad_norm": 0.01722401756403534,
      "learning_rate": 4.6649961833769715e-05,
      "loss": 0.0019,
      "step": 551
    },
    {
      "epoch": 34.5,
      "grad_norm": 0.008965687726181431,
      "learning_rate": 4.630798263510162e-05,
      "loss": 0.0005,
      "step": 552
    },
    {
      "epoch": 34.5625,
      "grad_norm": 0.01778818504579463,
      "learning_rate": 4.596688344202509e-05,
      "loss": 0.0021,
      "step": 553
    },
    {
      "epoch": 34.625,
      "grad_norm": 0.025545188168124802,
      "learning_rate": 4.562666984511416e-05,
      "loss": 0.0029,
      "step": 554
    },
    {
      "epoch": 34.6875,
      "grad_norm": 0.022767194931364268,
      "learning_rate": 4.528734742042803e-05,
      "loss": 0.0026,
      "step": 555
    },
    {
      "epoch": 34.75,
      "grad_norm": 0.02949871036401744,
      "learning_rate": 4.494892172941965e-05,
      "loss": 0.005,
      "step": 556
    },
    {
      "epoch": 34.8125,
      "grad_norm": 0.027099445112904772,
      "learning_rate": 4.461139831884474e-05,
      "loss": 0.004,
      "step": 557
    },
    {
      "epoch": 34.875,
      "grad_norm": 0.02425594257821406,
      "learning_rate": 4.427478272067066e-05,
      "loss": 0.0038,
      "step": 558
    },
    {
      "epoch": 34.9375,
      "grad_norm": 0.03234728199101579,
      "learning_rate": 4.393908045198585e-05,
      "loss": 0.0051,
      "step": 559
    },
    {
      "epoch": 35.0,
      "grad_norm": 0.021712331665851,
      "learning_rate": 4.360429701490934e-05,
      "loss": 0.0031,
      "step": 560
    },
    {
      "epoch": 35.0625,
      "grad_norm": 0.030962991791000313,
      "learning_rate": 4.327043789650078e-05,
      "loss": 0.0061,
      "step": 561
    },
    {
      "epoch": 35.125,
      "grad_norm": 0.019615835902719432,
      "learning_rate": 4.2937508568670194e-05,
      "loss": 0.0025,
      "step": 562
    },
    {
      "epoch": 35.1875,
      "grad_norm": 0.019445691272185775,
      "learning_rate": 4.2605514488088515e-05,
      "loss": 0.0023,
      "step": 563
    },
    {
      "epoch": 35.25,
      "grad_norm": 0.025561602674203147,
      "learning_rate": 4.227446109609809e-05,
      "loss": 0.0043,
      "step": 564
    },
    {
      "epoch": 35.3125,
      "grad_norm": 0.021631887342721947,
      "learning_rate": 4.1944353818623424e-05,
      "loss": 0.0024,
      "step": 565
    },
    {
      "epoch": 35.375,
      "grad_norm": 0.023929421104664724,
      "learning_rate": 4.161519806608247e-05,
      "loss": 0.0053,
      "step": 566
    },
    {
      "epoch": 35.4375,
      "grad_norm": 0.026767345162646533,
      "learning_rate": 4.12869992332977e-05,
      "loss": 0.0033,
      "step": 567
    },
    {
      "epoch": 35.5,
      "grad_norm": 0.019265712551654487,
      "learning_rate": 4.0959762699407766e-05,
      "loss": 0.0019,
      "step": 568
    },
    {
      "epoch": 35.5625,
      "grad_norm": 0.024435718154567927,
      "learning_rate": 4.0633493827779425e-05,
      "loss": 0.0034,
      "step": 569
    },
    {
      "epoch": 35.625,
      "grad_norm": 0.01536328570946991,
      "learning_rate": 4.030819796591949e-05,
      "loss": 0.0024,
      "step": 570
    },
    {
      "epoch": 35.6875,
      "grad_norm": 0.016210675375350477,
      "learning_rate": 3.9983880445387366e-05,
      "loss": 0.0024,
      "step": 571
    },
    {
      "epoch": 35.75,
      "grad_norm": 0.01324379048342,
      "learning_rate": 3.966054658170754e-05,
      "loss": 0.0011,
      "step": 572
    },
    {
      "epoch": 35.8125,
      "grad_norm": 0.009268348242829393,
      "learning_rate": 3.9338201674282406e-05,
      "loss": 0.0008,
      "step": 573
    },
    {
      "epoch": 35.875,
      "grad_norm": 0.024688307625187852,
      "learning_rate": 3.9016851006305545e-05,
      "loss": 0.0029,
      "step": 574
    },
    {
      "epoch": 35.9375,
      "grad_norm": 0.04355703985093336,
      "learning_rate": 3.869649984467504e-05,
      "loss": 0.0094,
      "step": 575
    },
    {
      "epoch": 36.0,
      "grad_norm": 0.02001971316741883,
      "learning_rate": 3.8377153439907266e-05,
      "loss": 0.0021,
      "step": 576
    },
    {
      "epoch": 36.0625,
      "grad_norm": 0.025891577060147455,
      "learning_rate": 3.8058817026050677e-05,
      "loss": 0.0038,
      "step": 577
    },
    {
      "epoch": 36.125,
      "grad_norm": 0.022905458332606304,
      "learning_rate": 3.774149582060012e-05,
      "loss": 0.0029,
      "step": 578
    },
    {
      "epoch": 36.1875,
      "grad_norm": 0.020383820791147977,
      "learning_rate": 3.742519502441132e-05,
      "loss": 0.0042,
      "step": 579
    },
    {
      "epoch": 36.25,
      "grad_norm": 0.022297278719448848,
      "learning_rate": 3.710991982161555e-05,
      "loss": 0.0041,
      "step": 580
    },
    {
      "epoch": 36.3125,
      "grad_norm": 0.01800159818226978,
      "learning_rate": 3.679567537953485e-05,
      "loss": 0.002,
      "step": 581
    },
    {
      "epoch": 36.375,
      "grad_norm": 0.026472437396163635,
      "learning_rate": 3.648246684859716e-05,
      "loss": 0.004,
      "step": 582
    },
    {
      "epoch": 36.4375,
      "grad_norm": 0.034982086882290896,
      "learning_rate": 3.617029936225193e-05,
      "loss": 0.0044,
      "step": 583
    },
    {
      "epoch": 36.5,
      "grad_norm": 0.025380193547038656,
      "learning_rate": 3.585917803688603e-05,
      "loss": 0.004,
      "step": 584
    },
    {
      "epoch": 36.5625,
      "grad_norm": 0.024933823706227147,
      "learning_rate": 3.55491079717399e-05,
      "loss": 0.0034,
      "step": 585
    },
    {
      "epoch": 36.625,
      "grad_norm": 0.032395369351154765,
      "learning_rate": 3.5240094248824e-05,
      "loss": 0.0047,
      "step": 586
    },
    {
      "epoch": 36.6875,
      "grad_norm": 0.02426742332899305,
      "learning_rate": 3.493214193283536e-05,
      "loss": 0.0025,
      "step": 587
    },
    {
      "epoch": 36.75,
      "grad_norm": 0.016326134651246785,
      "learning_rate": 3.4625256071074773e-05,
      "loss": 0.0019,
      "step": 588
    },
    {
      "epoch": 36.8125,
      "grad_norm": 0.022068578736713734,
      "learning_rate": 3.4319441693363906e-05,
      "loss": 0.0026,
      "step": 589
    },
    {
      "epoch": 36.875,
      "grad_norm": 0.022624745551108428,
      "learning_rate": 3.4014703811963025e-05,
      "loss": 0.003,
      "step": 590
    },
    {
      "epoch": 36.9375,
      "grad_norm": 0.02155433442232093,
      "learning_rate": 3.3711047421488675e-05,
      "loss": 0.0031,
      "step": 591
    },
    {
      "epoch": 37.0,
      "grad_norm": 0.00020949892145012195,
      "learning_rate": 3.340847749883191e-05,
      "loss": 0.0,
      "step": 592
    },
    {
      "epoch": 37.0625,
      "grad_norm": 0.025941183284377157,
      "learning_rate": 3.3106999003076746e-05,
      "loss": 0.0042,
      "step": 593
    },
    {
      "epoch": 37.125,
      "grad_norm": 0.021731884621000654,
      "learning_rate": 3.280661687541876e-05,
      "loss": 0.0024,
      "step": 594
    },
    {
      "epoch": 37.1875,
      "grad_norm": 0.016413130007407132,
      "learning_rate": 3.2507336039084314e-05,
      "loss": 0.0033,
      "step": 595
    },
    {
      "epoch": 37.25,
      "grad_norm": 0.033791751348081316,
      "learning_rate": 3.2209161399249674e-05,
      "loss": 0.0055,
      "step": 596
    },
    {
      "epoch": 37.3125,
      "grad_norm": 0.007739198098749652,
      "learning_rate": 3.191209784296068e-05,
      "loss": 0.0006,
      "step": 597
    },
    {
      "epoch": 37.375,
      "grad_norm": 0.01698515372085753,
      "learning_rate": 3.161615023905265e-05,
      "loss": 0.0018,
      "step": 598
    },
    {
      "epoch": 37.4375,
      "grad_norm": 0.009237792411598502,
      "learning_rate": 3.132132343807056e-05,
      "loss": 0.002,
      "step": 599
    },
    {
      "epoch": 37.5,
      "grad_norm": 0.01638142995950318,
      "learning_rate": 3.102762227218957e-05,
      "loss": 0.0014,
      "step": 600
    },
    {
      "epoch": 37.5625,
      "grad_norm": 0.016279487432564795,
      "learning_rate": 3.073505155513591e-05,
      "loss": 0.0015,
      "step": 601
    },
    {
      "epoch": 37.625,
      "grad_norm": 0.011772194410100805,
      "learning_rate": 3.044361608210775e-05,
      "loss": 0.0009,
      "step": 602
    },
    {
      "epoch": 37.6875,
      "grad_norm": 0.030973278706032212,
      "learning_rate": 3.0153320629696846e-05,
      "loss": 0.0051,
      "step": 603
    },
    {
      "epoch": 37.75,
      "grad_norm": 0.026566291070872453,
      "learning_rate": 2.9864169955810084e-05,
      "loss": 0.006,
      "step": 604
    },
    {
      "epoch": 37.8125,
      "grad_norm": 0.01901261238275813,
      "learning_rate": 2.9576168799591664e-05,
      "loss": 0.0022,
      "step": 605
    },
    {
      "epoch": 37.875,
      "grad_norm": 0.022602205933752235,
      "learning_rate": 2.9289321881345254e-05,
      "loss": 0.0032,
      "step": 606
    },
    {
      "epoch": 37.9375,
      "grad_norm": 0.022472706538328845,
      "learning_rate": 2.900363390245674e-05,
      "loss": 0.0028,
      "step": 607
    },
    {
      "epoch": 38.0,
      "grad_norm": 0.02802199698579331,
      "learning_rate": 2.8719109545317103e-05,
      "loss": 0.0061,
      "step": 608
    },
    {
      "epoch": 38.0625,
      "grad_norm": 0.007286991170262413,
      "learning_rate": 2.8435753473245698e-05,
      "loss": 0.0005,
      "step": 609
    },
    {
      "epoch": 38.125,
      "grad_norm": 0.015507395834223437,
      "learning_rate": 2.8153570330413925e-05,
      "loss": 0.0017,
      "step": 610
    },
    {
      "epoch": 38.1875,
      "grad_norm": 0.020058482119342358,
      "learning_rate": 2.7872564741768913e-05,
      "loss": 0.0034,
      "step": 611
    },
    {
      "epoch": 38.25,
      "grad_norm": 0.01760996920462819,
      "learning_rate": 2.759274131295787e-05,
      "loss": 0.0034,
      "step": 612
    },
    {
      "epoch": 38.3125,
      "grad_norm": 0.029096639894173625,
      "learning_rate": 2.73141046302525e-05,
      "loss": 0.005,
      "step": 613
    },
    {
      "epoch": 38.375,
      "grad_norm": 0.013393983973238514,
      "learning_rate": 2.7036659260473974e-05,
      "loss": 0.0009,
      "step": 614
    },
    {
      "epoch": 38.4375,
      "grad_norm": 0.012945478401193455,
      "learning_rate": 2.6760409750917927e-05,
      "loss": 0.0019,
      "step": 615
    },
    {
      "epoch": 38.5,
      "grad_norm": 0.026715394934565314,
      "learning_rate": 2.6485360629279987e-05,
      "loss": 0.0038,
      "step": 616
    },
    {
      "epoch": 38.5625,
      "grad_norm": 0.017347535771603556,
      "learning_rate": 2.6211516403581582e-05,
      "loss": 0.002,
      "step": 617
    },
    {
      "epoch": 38.625,
      "grad_norm": 0.016971730219051892,
      "learning_rate": 2.593888156209603e-05,
      "loss": 0.0017,
      "step": 618
    },
    {
      "epoch": 38.6875,
      "grad_norm": 0.025742679582296178,
      "learning_rate": 2.5667460573275028e-05,
      "loss": 0.0048,
      "step": 619
    },
    {
      "epoch": 38.75,
      "grad_norm": 0.02333983935112299,
      "learning_rate": 2.5397257885675397e-05,
      "loss": 0.0029,
      "step": 620
    },
    {
      "epoch": 38.8125,
      "grad_norm": 0.010043103557887257,
      "learning_rate": 2.5128277927886055e-05,
      "loss": 0.0039,
      "step": 621
    },
    {
      "epoch": 38.875,
      "grad_norm": 0.030322831096171305,
      "learning_rate": 2.48605251084556e-05,
      "loss": 0.0046,
      "step": 622
    },
    {
      "epoch": 38.9375,
      "grad_norm": 0.027197588947040895,
      "learning_rate": 2.4594003815819966e-05,
      "loss": 0.006,
      "step": 623
    },
    {
      "epoch": 39.0,
      "grad_norm": 0.023325558051272913,
      "learning_rate": 2.432871841823047e-05,
      "loss": 0.0026,
      "step": 624
    },
    {
      "epoch": 39.0625,
      "grad_norm": 0.027132468126116745,
      "learning_rate": 2.406467326368237e-05,
      "loss": 0.005,
      "step": 625
    },
    {
      "epoch": 39.125,
      "grad_norm": 0.023365336051910712,
      "learning_rate": 2.3801872679843385e-05,
      "loss": 0.0047,
      "step": 626
    },
    {
      "epoch": 39.1875,
      "grad_norm": 0.019510221468659553,
      "learning_rate": 2.3540320973982922e-05,
      "loss": 0.0022,
      "step": 627
    },
    {
      "epoch": 39.25,
      "grad_norm": 0.016164965587762533,
      "learning_rate": 2.3280022432901383e-05,
      "loss": 0.0025,
      "step": 628
    },
    {
      "epoch": 39.3125,
      "grad_norm": 0.021451656557391466,
      "learning_rate": 2.302098132286006e-05,
      "loss": 0.003,
      "step": 629
    },
    {
      "epoch": 39.375,
      "grad_norm": 0.031011855175604085,
      "learning_rate": 2.2763201889510987e-05,
      "loss": 0.0064,
      "step": 630
    },
    {
      "epoch": 39.4375,
      "grad_norm": 0.0006149615487194747,
      "learning_rate": 2.2506688357827544e-05,
      "loss": 0.0,
      "step": 631
    },
    {
      "epoch": 39.5,
      "grad_norm": 0.02248522871311791,
      "learning_rate": 2.2251444932035094e-05,
      "loss": 0.0027,
      "step": 632
    },
    {
      "epoch": 39.5625,
      "grad_norm": 0.034693691996730266,
      "learning_rate": 2.199747579554211e-05,
      "loss": 0.0057,
      "step": 633
    },
    {
      "epoch": 39.625,
      "grad_norm": 0.02437060961249019,
      "learning_rate": 2.174478511087171e-05,
      "loss": 0.0032,
      "step": 634
    },
    {
      "epoch": 39.6875,
      "grad_norm": 0.015994793305714,
      "learning_rate": 2.149337701959325e-05,
      "loss": 0.0018,
      "step": 635
    },
    {
      "epoch": 39.75,
      "grad_norm": 0.0002863850040924982,
      "learning_rate": 2.1243255642254578e-05,
      "loss": 0.0,
      "step": 636
    },
    {
      "epoch": 39.8125,
      "grad_norm": 0.020151892270889014,
      "learning_rate": 2.099442507831444e-05,
      "loss": 0.002,
      "step": 637
    },
    {
      "epoch": 39.875,
      "grad_norm": 0.028381722329777846,
      "learning_rate": 2.074688940607529e-05,
      "loss": 0.0043,
      "step": 638
    },
    {
      "epoch": 39.9375,
      "grad_norm": 0.024166442271924462,
      "learning_rate": 2.050065268261655e-05,
      "loss": 0.0026,
      "step": 639
    },
    {
      "epoch": 40.0,
      "grad_norm": 0.02278485322740038,
      "learning_rate": 2.025571894372794e-05,
      "loss": 0.0028,
      "step": 640
    },
    {
      "epoch": 40.0625,
      "grad_norm": 0.013588204702480308,
      "learning_rate": 2.001209220384346e-05,
      "loss": 0.0008,
      "step": 641
    },
    {
      "epoch": 40.125,
      "grad_norm": 0.02222330407423336,
      "learning_rate": 1.976977645597552e-05,
      "loss": 0.0032,
      "step": 642
    },
    {
      "epoch": 40.1875,
      "grad_norm": 0.015083512876565244,
      "learning_rate": 1.9528775671649592e-05,
      "loss": 0.001,
      "step": 643
    },
    {
      "epoch": 40.25,
      "grad_norm": 0.025069080823856827,
      "learning_rate": 1.9289093800839066e-05,
      "loss": 0.0031,
      "step": 644
    },
    {
      "epoch": 40.3125,
      "grad_norm": 0.02025912896301005,
      "learning_rate": 1.9050734771900413e-05,
      "loss": 0.0025,
      "step": 645
    },
    {
      "epoch": 40.375,
      "grad_norm": 0.013091990888201397,
      "learning_rate": 1.8813702491508955e-05,
      "loss": 0.0015,
      "step": 646
    },
    {
      "epoch": 40.4375,
      "grad_norm": 0.021454823181291053,
      "learning_rate": 1.8578000844594747e-05,
      "loss": 0.0027,
      "step": 647
    },
    {
      "epoch": 40.5,
      "grad_norm": 0.02158593959614438,
      "learning_rate": 1.8343633694278895e-05,
      "loss": 0.0038,
      "step": 648
    },
    {
      "epoch": 40.5625,
      "grad_norm": 0.030736687344323808,
      "learning_rate": 1.8110604881810355e-05,
      "loss": 0.0064,
      "step": 649
    },
    {
      "epoch": 40.625,
      "grad_norm": 0.0186110157411597,
      "learning_rate": 1.7878918226502816e-05,
      "loss": 0.0029,
      "step": 650
    },
    {
      "epoch": 40.6875,
      "grad_norm": 0.021941314758172913,
      "learning_rate": 1.7648577525672194e-05,
      "loss": 0.0037,
      "step": 651
    },
    {
      "epoch": 40.75,
      "grad_norm": 0.021140994060252962,
      "learning_rate": 1.741958655457436e-05,
      "loss": 0.0042,
      "step": 652
    },
    {
      "epoch": 40.8125,
      "grad_norm": 0.022090278067811005,
      "learning_rate": 1.7191949066343305e-05,
      "loss": 0.0026,
      "step": 653
    },
    {
      "epoch": 40.875,
      "grad_norm": 0.020131252147946324,
      "learning_rate": 1.69656687919296e-05,
      "loss": 0.0034,
      "step": 654
    },
    {
      "epoch": 40.9375,
      "grad_norm": 0.015348430101637372,
      "learning_rate": 1.6740749440039262e-05,
      "loss": 0.0017,
      "step": 655
    },
    {
      "epoch": 41.0,
      "grad_norm": 0.029281984191582698,
      "learning_rate": 1.65171946970729e-05,
      "loss": 0.0049,
      "step": 656
    },
    {
      "epoch": 41.0625,
      "grad_norm": 0.024216109331731846,
      "learning_rate": 1.6295008227065367e-05,
      "loss": 0.0046,
      "step": 657
    },
    {
      "epoch": 41.125,
      "grad_norm": 0.021824879823787828,
      "learning_rate": 1.607419367162577e-05,
      "loss": 0.003,
      "step": 658
    },
    {
      "epoch": 41.1875,
      "grad_norm": 0.01662964866399249,
      "learning_rate": 1.5854754649877603e-05,
      "loss": 0.0017,
      "step": 659
    },
    {
      "epoch": 41.25,
      "grad_norm": 0.028069673310717295,
      "learning_rate": 1.563669475839956e-05,
      "loss": 0.005,
      "step": 660
    },
    {
      "epoch": 41.3125,
      "grad_norm": 0.02097056263367153,
      "learning_rate": 1.542001757116658e-05,
      "loss": 0.0021,
      "step": 661
    },
    {
      "epoch": 41.375,
      "grad_norm": 0.02751701662543221,
      "learning_rate": 1.5204726639491218e-05,
      "loss": 0.0043,
      "step": 662
    },
    {
      "epoch": 41.4375,
      "grad_norm": 0.02821649307326047,
      "learning_rate": 1.4990825491965522e-05,
      "loss": 0.0049,
      "step": 663
    },
    {
      "epoch": 41.5,
      "grad_norm": 0.0194640022449319,
      "learning_rate": 1.4778317634403083e-05,
      "loss": 0.0041,
      "step": 664
    },
    {
      "epoch": 41.5625,
      "grad_norm": 0.015127933678134073,
      "learning_rate": 1.4567206549781698e-05,
      "loss": 0.0009,
      "step": 665
    },
    {
      "epoch": 41.625,
      "grad_norm": 0.01576306460154544,
      "learning_rate": 1.4357495698186186e-05,
      "loss": 0.0016,
      "step": 666
    },
    {
      "epoch": 41.6875,
      "grad_norm": 0.022161541294332343,
      "learning_rate": 1.41491885167517e-05,
      "loss": 0.0032,
      "step": 667
    },
    {
      "epoch": 41.75,
      "grad_norm": 0.027773972314299573,
      "learning_rate": 1.3942288419607475e-05,
      "loss": 0.0048,
      "step": 668
    },
    {
      "epoch": 41.8125,
      "grad_norm": 0.018134086844876104,
      "learning_rate": 1.3736798797820782e-05,
      "loss": 0.0019,
      "step": 669
    },
    {
      "epoch": 41.875,
      "grad_norm": 0.010646599496077774,
      "learning_rate": 1.3532723019341375e-05,
      "loss": 0.0009,
      "step": 670
    },
    {
      "epoch": 41.9375,
      "grad_norm": 0.022768008186984896,
      "learning_rate": 1.3330064428946254e-05,
      "loss": 0.0034,
      "step": 671
    },
    {
      "epoch": 42.0,
      "grad_norm": 0.022539707407857736,
      "learning_rate": 1.3128826348184887e-05,
      "loss": 0.0026,
      "step": 672
    },
    {
      "epoch": 42.0625,
      "grad_norm": 0.017092620865285464,
      "learning_rate": 1.2929012075324831e-05,
      "loss": 0.002,
      "step": 673
    },
    {
      "epoch": 42.125,
      "grad_norm": 0.027898487565227375,
      "learning_rate": 1.2730624885297537e-05,
      "loss": 0.0036,
      "step": 674
    },
    {
      "epoch": 42.1875,
      "grad_norm": 0.01888830939120487,
      "learning_rate": 1.2533668029644751e-05,
      "loss": 0.0023,
      "step": 675
    },
    {
      "epoch": 42.25,
      "grad_norm": 0.034797071188688175,
      "learning_rate": 1.233814473646524e-05,
      "loss": 0.0047,
      "step": 676
    },
    {
      "epoch": 42.3125,
      "grad_norm": 0.029204011780076224,
      "learning_rate": 1.214405821036182e-05,
      "loss": 0.0045,
      "step": 677
    },
    {
      "epoch": 42.375,
      "grad_norm": 0.018166143974674018,
      "learning_rate": 1.195141163238892e-05,
      "loss": 0.0025,
      "step": 678
    },
    {
      "epoch": 42.4375,
      "grad_norm": 0.02110557824065673,
      "learning_rate": 1.1760208160000363e-05,
      "loss": 0.004,
      "step": 679
    },
    {
      "epoch": 42.5,
      "grad_norm": 0.025933927942001878,
      "learning_rate": 1.1570450926997655e-05,
      "loss": 0.0047,
      "step": 680
    },
    {
      "epoch": 42.5625,
      "grad_norm": 0.01153329092029045,
      "learning_rate": 1.13821430434786e-05,
      "loss": 0.0008,
      "step": 681
    },
    {
      "epoch": 42.625,
      "grad_norm": 0.02871409530487605,
      "learning_rate": 1.1195287595786352e-05,
      "loss": 0.0033,
      "step": 682
    },
    {
      "epoch": 42.6875,
      "grad_norm": 0.020158017546353265,
      "learning_rate": 1.1009887646458861e-05,
      "loss": 0.0038,
      "step": 683
    },
    {
      "epoch": 42.75,
      "grad_norm": 0.020818728203346376,
      "learning_rate": 1.0825946234178574e-05,
      "loss": 0.0035,
      "step": 684
    },
    {
      "epoch": 42.8125,
      "grad_norm": 0.019312470454732285,
      "learning_rate": 1.0643466373722711e-05,
      "loss": 0.0021,
      "step": 685
    },
    {
      "epoch": 42.875,
      "grad_norm": 0.01710427570336108,
      "learning_rate": 1.0462451055913847e-05,
      "loss": 0.002,
      "step": 686
    },
    {
      "epoch": 42.9375,
      "grad_norm": 0.024424739421454587,
      "learning_rate": 1.0282903247570908e-05,
      "loss": 0.0034,
      "step": 687
    },
    {
      "epoch": 43.0,
      "grad_norm": 0.014619911575136098,
      "learning_rate": 1.010482589146048e-05,
      "loss": 0.0016,
      "step": 688
    },
    {
      "epoch": 43.0625,
      "grad_norm": 0.018861495766737645,
      "learning_rate": 9.928221906248614e-06,
      "loss": 0.0018,
      "step": 689
    },
    {
      "epoch": 43.125,
      "grad_norm": 0.018818551498274123,
      "learning_rate": 9.753094186453026e-06,
      "loss": 0.0028,
      "step": 690
    },
    {
      "epoch": 43.1875,
      "grad_norm": 0.00016866140217513113,
      "learning_rate": 9.579445602395576e-06,
      "loss": 0.0,
      "step": 691
    },
    {
      "epoch": 43.25,
      "grad_norm": 0.02583319297082241,
      "learning_rate": 9.407279000155312e-06,
      "loss": 0.0036,
      "step": 692
    },
    {
      "epoch": 43.3125,
      "grad_norm": 0.023265334503398318,
      "learning_rate": 9.23659720152179e-06,
      "loss": 0.0031,
      "step": 693
    },
    {
      "epoch": 43.375,
      "grad_norm": 0.02996502207699758,
      "learning_rate": 9.067403003948782e-06,
      "loss": 0.0055,
      "step": 694
    },
    {
      "epoch": 43.4375,
      "grad_norm": 0.024920753877047955,
      "learning_rate": 8.89969918050847e-06,
      "loss": 0.0032,
      "step": 695
    },
    {
      "epoch": 43.5,
      "grad_norm": 0.020615629249821864,
      "learning_rate": 8.733488479845997e-06,
      "loss": 0.0023,
      "step": 696
    },
    {
      "epoch": 43.5625,
      "grad_norm": 0.02585899233303135,
      "learning_rate": 8.568773626134364e-06,
      "loss": 0.0039,
      "step": 697
    },
    {
      "epoch": 43.625,
      "grad_norm": 0.03234297820570824,
      "learning_rate": 8.405557319029912e-06,
      "loss": 0.0055,
      "step": 698
    },
    {
      "epoch": 43.6875,
      "grad_norm": 0.02361422744737024,
      "learning_rate": 8.243842233627896e-06,
      "loss": 0.004,
      "step": 699
    },
    {
      "epoch": 43.75,
      "grad_norm": 0.015111961904190768,
      "learning_rate": 8.083631020418791e-06,
      "loss": 0.0017,
      "step": 700
    },
    {
      "epoch": 43.8125,
      "grad_norm": 0.019087732617034506,
      "learning_rate": 7.924926305244728e-06,
      "loss": 0.0015,
      "step": 701
    },
    {
      "epoch": 43.875,
      "grad_norm": 0.019039580961815027,
      "learning_rate": 7.767730689256614e-06,
      "loss": 0.0032,
      "step": 702
    },
    {
      "epoch": 43.9375,
      "grad_norm": 0.01619120710381878,
      "learning_rate": 7.612046748871327e-06,
      "loss": 0.0022,
      "step": 703
    },
    {
      "epoch": 44.0,
      "grad_norm": 0.02963246039764041,
      "learning_rate": 7.457877035729588e-06,
      "loss": 0.0036,
      "step": 704
    },
    {
      "epoch": 44.0625,
      "grad_norm": 0.02289200860477135,
      "learning_rate": 7.305224076654127e-06,
      "loss": 0.0031,
      "step": 705
    },
    {
      "epoch": 44.125,
      "grad_norm": 0.022076965629593947,
      "learning_rate": 7.154090373608235e-06,
      "loss": 0.0025,
      "step": 706
    },
    {
      "epoch": 44.1875,
      "grad_norm": 0.017284412700166694,
      "learning_rate": 7.004478403654835e-06,
      "loss": 0.0028,
      "step": 707
    },
    {
      "epoch": 44.25,
      "grad_norm": 0.02994271330026706,
      "learning_rate": 6.856390618915775e-06,
      "loss": 0.0036,
      "step": 708
    },
    {
      "epoch": 44.3125,
      "grad_norm": 0.010544664066729796,
      "learning_rate": 6.709829446531734e-06,
      "loss": 0.0025,
      "step": 709
    },
    {
      "epoch": 44.375,
      "grad_norm": 0.01646307933490768,
      "learning_rate": 6.564797288622371e-06,
      "loss": 0.0017,
      "step": 710
    },
    {
      "epoch": 44.4375,
      "grad_norm": 0.02167552761136678,
      "learning_rate": 6.4212965222470115e-06,
      "loss": 0.0027,
      "step": 711
    },
    {
      "epoch": 44.5,
      "grad_norm": 0.011538751711468322,
      "learning_rate": 6.2793294993656494e-06,
      "loss": 0.001,
      "step": 712
    },
    {
      "epoch": 44.5625,
      "grad_norm": 0.026342695075247172,
      "learning_rate": 6.138898546800398e-06,
      "loss": 0.003,
      "step": 713
    },
    {
      "epoch": 44.625,
      "grad_norm": 0.01981392389008597,
      "learning_rate": 6.000005966197387e-06,
      "loss": 0.0023,
      "step": 714
    },
    {
      "epoch": 44.6875,
      "grad_norm": 0.025550367873884112,
      "learning_rate": 5.86265403398899e-06,
      "loss": 0.0046,
      "step": 715
    },
    {
      "epoch": 44.75,
      "grad_norm": 0.026765500781078307,
      "learning_rate": 5.726845001356573e-06,
      "loss": 0.0047,
      "step": 716
    },
    {
      "epoch": 44.8125,
      "grad_norm": 0.027059728322687618,
      "learning_rate": 5.592581094193583e-06,
      "loss": 0.0045,
      "step": 717
    },
    {
      "epoch": 44.875,
      "grad_norm": 0.02091280497177447,
      "learning_rate": 5.45986451306899e-06,
      "loss": 0.0031,
      "step": 718
    },
    {
      "epoch": 44.9375,
      "grad_norm": 0.023097246575829473,
      "learning_rate": 5.328697433191321e-06,
      "loss": 0.0035,
      "step": 719
    },
    {
      "epoch": 45.0,
      "grad_norm": 0.018080675383588967,
      "learning_rate": 5.199082004372957e-06,
      "loss": 0.0019,
      "step": 720
    },
    {
      "epoch": 45.0625,
      "grad_norm": 0.022599144419125274,
      "learning_rate": 5.0710203509948924e-06,
      "loss": 0.0041,
      "step": 721
    },
    {
      "epoch": 45.125,
      "grad_norm": 0.019069660618183874,
      "learning_rate": 4.944514571971981e-06,
      "loss": 0.0023,
      "step": 722
    },
    {
      "epoch": 45.1875,
      "grad_norm": 0.01673711361691797,
      "learning_rate": 4.819566740718439e-06,
      "loss": 0.0017,
      "step": 723
    },
    {
      "epoch": 45.25,
      "grad_norm": 0.017407725192059877,
      "learning_rate": 4.6961789051139124e-06,
      "loss": 0.0025,
      "step": 724
    },
    {
      "epoch": 45.3125,
      "grad_norm": 0.007214982705630053,
      "learning_rate": 4.574353087469929e-06,
      "loss": 0.0005,
      "step": 725
    },
    {
      "epoch": 45.375,
      "grad_norm": 0.02158166297132307,
      "learning_rate": 4.454091284496731e-06,
      "loss": 0.003,
      "step": 726
    },
    {
      "epoch": 45.4375,
      "grad_norm": 0.031535142634254686,
      "learning_rate": 4.335395467270553e-06,
      "loss": 0.0047,
      "step": 727
    },
    {
      "epoch": 45.5,
      "grad_norm": 0.030571739192627623,
      "learning_rate": 4.2182675812012965e-06,
      "loss": 0.0048,
      "step": 728
    },
    {
      "epoch": 45.5625,
      "grad_norm": 0.02876678712689496,
      "learning_rate": 4.102709546000671e-06,
      "loss": 0.0052,
      "step": 729
    },
    {
      "epoch": 45.625,
      "grad_norm": 0.0211381008090843,
      "learning_rate": 3.988723255650728e-06,
      "loss": 0.0036,
      "step": 730
    },
    {
      "epoch": 45.6875,
      "grad_norm": 0.02281744512369711,
      "learning_rate": 3.876310578372832e-06,
      "loss": 0.0027,
      "step": 731
    },
    {
      "epoch": 45.75,
      "grad_norm": 0.021897414013302245,
      "learning_rate": 3.7654733565969826e-06,
      "loss": 0.002,
      "step": 732
    },
    {
      "epoch": 45.8125,
      "grad_norm": 0.01919140132177964,
      "learning_rate": 3.6562134069316854e-06,
      "loss": 0.0017,
      "step": 733
    },
    {
      "epoch": 45.875,
      "grad_norm": 0.017425342650746963,
      "learning_rate": 3.548532520134129e-06,
      "loss": 0.0017,
      "step": 734
    },
    {
      "epoch": 45.9375,
      "grad_norm": 0.018230395206139328,
      "learning_rate": 3.442432461080858e-06,
      "loss": 0.0023,
      "step": 735
    },
    {
      "epoch": 46.0,
      "grad_norm": 0.024588454053071196,
      "learning_rate": 3.3379149687388867e-06,
      "loss": 0.0036,
      "step": 736
    },
    {
      "epoch": 46.0625,
      "grad_norm": 0.019822471128658536,
      "learning_rate": 3.23498175613709e-06,
      "loss": 0.0028,
      "step": 737
    },
    {
      "epoch": 46.125,
      "grad_norm": 0.02110858026864958,
      "learning_rate": 3.1336345103382346e-06,
      "loss": 0.0032,
      "step": 738
    },
    {
      "epoch": 46.1875,
      "grad_norm": 0.037579636108812396,
      "learning_rate": 3.0338748924112483e-06,
      "loss": 0.0046,
      "step": 739
    },
    {
      "epoch": 46.25,
      "grad_norm": 0.023073536580485558,
      "learning_rate": 2.9357045374040825e-06,
      "loss": 0.0024,
      "step": 740
    },
    {
      "epoch": 46.3125,
      "grad_norm": 0.019009441815773284,
      "learning_rate": 2.839125054316838e-06,
      "loss": 0.0018,
      "step": 741
    },
    {
      "epoch": 46.375,
      "grad_norm": 0.021241256500812482,
      "learning_rate": 2.7441380260754048e-06,
      "loss": 0.0032,
      "step": 742
    },
    {
      "epoch": 46.4375,
      "grad_norm": 0.02621426587903328,
      "learning_rate": 2.6507450095055618e-06,
      "loss": 0.0034,
      "step": 743
    },
    {
      "epoch": 46.5,
      "grad_norm": 0.03271609828252333,
      "learning_rate": 2.5589475353073988e-06,
      "loss": 0.0057,
      "step": 744
    },
    {
      "epoch": 46.5625,
      "grad_norm": 0.016395104429952994,
      "learning_rate": 2.4687471080302894e-06,
      "loss": 0.0014,
      "step": 745
    },
    {
      "epoch": 46.625,
      "grad_norm": 0.010491752550833084,
      "learning_rate": 2.380145206048201e-06,
      "loss": 0.0008,
      "step": 746
    },
    {
      "epoch": 46.6875,
      "grad_norm": 0.01997489278725962,
      "learning_rate": 2.2931432815354594e-06,
      "loss": 0.0028,
      "step": 747
    },
    {
      "epoch": 46.75,
      "grad_norm": 0.025314386929928892,
      "learning_rate": 2.2077427604429433e-06,
      "loss": 0.004,
      "step": 748
    },
    {
      "epoch": 46.8125,
      "grad_norm": 0.017939092098162242,
      "learning_rate": 2.1239450424747508e-06,
      "loss": 0.0032,
      "step": 749
    },
    {
      "epoch": 46.875,
      "grad_norm": 0.02330397906421762,
      "learning_rate": 2.041751501065203e-06,
      "loss": 0.0031,
      "step": 750
    },
    {
      "epoch": 46.9375,
      "grad_norm": 0.017815651039882807,
      "learning_rate": 1.9611634833564096e-06,
      "loss": 0.0023,
      "step": 751
    },
    {
      "epoch": 47.0,
      "grad_norm": 0.022593869860642213,
      "learning_rate": 1.882182310176095e-06,
      "loss": 0.0022,
      "step": 752
    },
    {
      "epoch": 47.0625,
      "grad_norm": 0.016063543491380557,
      "learning_rate": 1.8048092760160285e-06,
      "loss": 0.0021,
      "step": 753
    },
    {
      "epoch": 47.125,
      "grad_norm": 0.021873269229617312,
      "learning_rate": 1.729045649010752e-06,
      "loss": 0.0045,
      "step": 754
    },
    {
      "epoch": 47.1875,
      "grad_norm": 0.01569665609733916,
      "learning_rate": 1.6548926709168633e-06,
      "loss": 0.0009,
      "step": 755
    },
    {
      "epoch": 47.25,
      "grad_norm": 0.011413815570498292,
      "learning_rate": 1.5823515570925763e-06,
      "loss": 0.0009,
      "step": 756
    },
    {
      "epoch": 47.3125,
      "grad_norm": 0.029109810904750552,
      "learning_rate": 1.5114234964778707e-06,
      "loss": 0.0049,
      "step": 757
    },
    {
      "epoch": 47.375,
      "grad_norm": 0.024048997338655526,
      "learning_rate": 1.4421096515749855e-06,
      "loss": 0.0034,
      "step": 758
    },
    {
      "epoch": 47.4375,
      "grad_norm": 0.027172675685767346,
      "learning_rate": 1.3744111584293228e-06,
      "loss": 0.0037,
      "step": 759
    },
    {
      "epoch": 47.5,
      "grad_norm": 0.0035756082981969447,
      "learning_rate": 1.30832912661093e-06,
      "loss": 0.0017,
      "step": 760
    },
    {
      "epoch": 47.5625,
      "grad_norm": 0.02566413955432668,
      "learning_rate": 1.2438646391962129e-06,
      "loss": 0.0035,
      "step": 761
    },
    {
      "epoch": 47.625,
      "grad_norm": 0.032308326355990395,
      "learning_rate": 1.1810187527502182e-06,
      "loss": 0.0048,
      "step": 762
    },
    {
      "epoch": 47.6875,
      "grad_norm": 0.026149262681996552,
      "learning_rate": 1.1197924973093464e-06,
      "loss": 0.0034,
      "step": 763
    },
    {
      "epoch": 47.75,
      "grad_norm": 0.015097613235697617,
      "learning_rate": 1.0601868763643996e-06,
      "loss": 0.0015,
      "step": 764
    },
    {
      "epoch": 47.8125,
      "grad_norm": 0.019142453822043624,
      "learning_rate": 1.0022028668442375e-06,
      "loss": 0.002,
      "step": 765
    },
    {
      "epoch": 47.875,
      "grad_norm": 0.014333115422685562,
      "learning_rate": 9.458414190996689e-07,
      "loss": 0.0009,
      "step": 766
    },
    {
      "epoch": 47.9375,
      "grad_norm": 0.031884445051135024,
      "learning_rate": 8.911034568879207e-07,
      "loss": 0.0069,
      "step": 767
    },
    {
      "epoch": 48.0,
      "grad_norm": 0.017813313479240984,
      "learning_rate": 8.379898773574924e-07,
      "loss": 0.0017,
      "step": 768
    },
    {
      "epoch": 48.0625,
      "grad_norm": 0.021676976013209007,
      "learning_rate": 7.865015510334472e-07,
      "loss": 0.0052,
      "step": 769
    },
    {
      "epoch": 48.125,
      "grad_norm": 0.023872711159584236,
      "learning_rate": 7.366393218031564e-07,
      "loss": 0.0037,
      "step": 770
    },
    {
      "epoch": 48.1875,
      "grad_norm": 0.013560684928306489,
      "learning_rate": 6.884040069024434e-07,
      "loss": 0.0009,
      "step": 771
    },
    {
      "epoch": 48.25,
      "grad_norm": 0.027547746861488913,
      "learning_rate": 6.41796396902239e-07,
      "loss": 0.0041,
      "step": 772
    },
    {
      "epoch": 48.3125,
      "grad_norm": 0.02379869982898958,
      "learning_rate": 5.968172556955365e-07,
      "loss": 0.0025,
      "step": 773
    },
    {
      "epoch": 48.375,
      "grad_norm": 0.01470048963673093,
      "learning_rate": 5.534673204849572e-07,
      "loss": 0.0009,
      "step": 774
    },
    {
      "epoch": 48.4375,
      "grad_norm": 0.023911550191113507,
      "learning_rate": 5.117473017706486e-07,
      "loss": 0.0053,
      "step": 775
    },
    {
      "epoch": 48.5,
      "grad_norm": 0.022409728345083195,
      "learning_rate": 4.7165788333860536e-07,
      "loss": 0.0026,
      "step": 776
    },
    {
      "epoch": 48.5625,
      "grad_norm": 0.020732847008319438,
      "learning_rate": 4.331997222494777e-07,
      "loss": 0.0062,
      "step": 777
    },
    {
      "epoch": 48.625,
      "grad_norm": 0.017704460576832216,
      "learning_rate": 3.963734488278248e-07,
      "loss": 0.0016,
      "step": 778
    },
    {
      "epoch": 48.6875,
      "grad_norm": 0.019405071538624658,
      "learning_rate": 3.611796666517564e-07,
      "loss": 0.0029,
      "step": 779
    },
    {
      "epoch": 48.75,
      "grad_norm": 0.02358804433732502,
      "learning_rate": 3.2761895254306287e-07,
      "loss": 0.0033,
      "step": 780
    },
    {
      "epoch": 48.8125,
      "grad_norm": 0.00883061746234552,
      "learning_rate": 2.956918565577338e-07,
      "loss": 0.0005,
      "step": 781
    },
    {
      "epoch": 48.875,
      "grad_norm": 0.017159129222209444,
      "learning_rate": 2.6539890197695427e-07,
      "loss": 0.0016,
      "step": 782
    },
    {
      "epoch": 48.9375,
      "grad_norm": 0.027402061759511424,
      "learning_rate": 2.3674058529855602e-07,
      "loss": 0.0028,
      "step": 783
    },
    {
      "epoch": 49.0,
      "grad_norm": 0.02128952613240459,
      "learning_rate": 2.0971737622883515e-07,
      "loss": 0.0025,
      "step": 784
    },
    {
      "epoch": 49.0625,
      "grad_norm": 0.026668820402710704,
      "learning_rate": 1.843297176748804e-07,
      "loss": 0.0046,
      "step": 785
    },
    {
      "epoch": 49.125,
      "grad_norm": 0.022353610802999663,
      "learning_rate": 1.605780257373124e-07,
      "loss": 0.0039,
      "step": 786
    },
    {
      "epoch": 49.1875,
      "grad_norm": 0.010951321068786213,
      "learning_rate": 1.3846268970344466e-07,
      "loss": 0.0009,
      "step": 787
    },
    {
      "epoch": 49.25,
      "grad_norm": 0.014538946273216842,
      "learning_rate": 1.179840720409331e-07,
      "loss": 0.0014,
      "step": 788
    },
    {
      "epoch": 49.3125,
      "grad_norm": 0.01556854992979764,
      "learning_rate": 9.914250839180294e-08,
      "loss": 0.0015,
      "step": 789
    },
    {
      "epoch": 49.375,
      "grad_norm": 0.020104302093467665,
      "learning_rate": 8.193830756699772e-08,
      "loss": 0.0017,
      "step": 790
    },
    {
      "epoch": 49.4375,
      "grad_norm": 0.021695384943554476,
      "learning_rate": 6.637175154124986e-08,
      "loss": 0.0044,
      "step": 791
    },
    {
      "epoch": 49.5,
      "grad_norm": 0.018630437617792112,
      "learning_rate": 5.2443095448506674e-08,
      "loss": 0.0041,
      "step": 792
    },
    {
      "epoch": 49.5625,
      "grad_norm": 0.02382024881730757,
      "learning_rate": 4.015256757774477e-08,
      "loss": 0.0023,
      "step": 793
    },
    {
      "epoch": 49.625,
      "grad_norm": 0.03427139945682584,
      "learning_rate": 2.9500369369195312e-08,
      "loss": 0.0046,
      "step": 794
    },
    {
      "epoch": 49.6875,
      "grad_norm": 0.012461120044038388,
      "learning_rate": 2.0486675411102163e-08,
      "loss": 0.0011,
      "step": 795
    },
    {
      "epoch": 49.75,
      "grad_norm": 0.020179437876762667,
      "learning_rate": 1.3111633436779791e-08,
      "loss": 0.0017,
      "step": 796
    },
    {
      "epoch": 49.8125,
      "grad_norm": 0.02050009583056352,
      "learning_rate": 7.375364322292911e-09,
      "loss": 0.0021,
      "step": 797
    },
    {
      "epoch": 49.875,
      "grad_norm": 0.02353884448233026,
      "learning_rate": 3.2779620843692572e-09,
      "loss": 0.0039,
      "step": 798
    },
    {
      "epoch": 49.9375,
      "grad_norm": 0.02422128610193115,
      "learning_rate": 8.194938789451989e-10,
      "loss": 0.0033,
      "step": 799
    },
    {
      "epoch": 50.0,
      "grad_norm": 0.028631822991189302,
      "learning_rate": 0.0,
      "loss": 0.0048,
      "step": 800
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 800,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 80,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 9578559744000.0,
  "train_batch_size": 5,
  "trial_name": null,
  "trial_params": null
}