{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 4118,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00024283632831471587,
      "grad_norm": 1.154969573020935,
      "learning_rate": 0.0001,
      "loss": 2.074,
      "step": 1
    },
    {
      "epoch": 0.00048567265662943174,
      "grad_norm": 0.5944371223449707,
      "learning_rate": 0.0001,
      "loss": 1.9917,
      "step": 2
    },
    {
      "epoch": 0.0007285089849441476,
      "grad_norm": 2.7935380935668945,
      "learning_rate": 0.0001,
      "loss": 2.0706,
      "step": 3
    },
    {
      "epoch": 0.0009713453132588635,
      "grad_norm": 1.0198408365249634,
      "learning_rate": 0.0001,
      "loss": 1.9375,
      "step": 4
    },
    {
      "epoch": 0.0012141816415735794,
      "grad_norm": 0.4888165295124054,
      "learning_rate": 0.0001,
      "loss": 1.8088,
      "step": 5
    },
    {
      "epoch": 0.0014570179698882952,
      "grad_norm": 0.5396842956542969,
      "learning_rate": 0.0001,
      "loss": 1.889,
      "step": 6
    },
    {
      "epoch": 0.001699854298203011,
      "grad_norm": 0.5880922079086304,
      "learning_rate": 0.0001,
      "loss": 2.0318,
      "step": 7
    },
    {
      "epoch": 0.001942690626517727,
      "grad_norm": 0.48738348484039307,
      "learning_rate": 0.0001,
      "loss": 2.0962,
      "step": 8
    },
    {
      "epoch": 0.002185526954832443,
      "grad_norm": 0.37848764657974243,
      "learning_rate": 0.0001,
      "loss": 1.8828,
      "step": 9
    },
    {
      "epoch": 0.0024283632831471587,
      "grad_norm": 0.47919952869415283,
      "learning_rate": 0.0001,
      "loss": 1.8523,
      "step": 10
    },
    {
      "epoch": 0.0026711996114618746,
      "grad_norm": 0.45462319254875183,
      "learning_rate": 0.0001,
      "loss": 1.6941,
      "step": 11
    },
    {
      "epoch": 0.0029140359397765905,
      "grad_norm": 0.4438392221927643,
      "learning_rate": 0.0001,
      "loss": 1.9583,
      "step": 12
    },
    {
      "epoch": 0.0031568722680913063,
      "grad_norm": 0.35694992542266846,
      "learning_rate": 0.0001,
      "loss": 1.8645,
      "step": 13
    },
    {
      "epoch": 0.003399708596406022,
      "grad_norm": 0.37169331312179565,
      "learning_rate": 0.0001,
      "loss": 1.8335,
      "step": 14
    },
    {
      "epoch": 0.003642544924720738,
      "grad_norm": 0.39208415150642395,
      "learning_rate": 0.0001,
      "loss": 1.9256,
      "step": 15
    },
    {
      "epoch": 0.003885381253035454,
      "grad_norm": 0.39922136068344116,
      "learning_rate": 0.0001,
      "loss": 1.921,
      "step": 16
    },
    {
      "epoch": 0.00412821758135017,
      "grad_norm": 0.360889732837677,
      "learning_rate": 0.0001,
      "loss": 1.772,
      "step": 17
    },
    {
      "epoch": 0.004371053909664886,
      "grad_norm": 0.3539641499519348,
      "learning_rate": 0.0001,
      "loss": 1.8403,
      "step": 18
    },
    {
      "epoch": 0.004613890237979602,
      "grad_norm": 0.3484066128730774,
      "learning_rate": 0.0001,
      "loss": 1.877,
      "step": 19
    },
    {
      "epoch": 0.0048567265662943174,
      "grad_norm": 0.36553072929382324,
      "learning_rate": 0.0001,
      "loss": 1.867,
      "step": 20
    },
    {
      "epoch": 0.005099562894609034,
      "grad_norm": 0.38572415709495544,
      "learning_rate": 0.0001,
      "loss": 2.0538,
      "step": 21
    },
    {
      "epoch": 0.005342399222923749,
      "grad_norm": 0.38361066579818726,
      "learning_rate": 0.0001,
      "loss": 1.9438,
      "step": 22
    },
    {
      "epoch": 0.0055852355512384655,
      "grad_norm": 0.38020607829093933,
      "learning_rate": 0.0001,
      "loss": 1.8955,
      "step": 23
    },
    {
      "epoch": 0.005828071879553181,
      "grad_norm": 0.37575146555900574,
      "learning_rate": 0.0001,
      "loss": 1.6813,
      "step": 24
    },
    {
      "epoch": 0.006070908207867897,
      "grad_norm": 0.3370920419692993,
      "learning_rate": 0.0001,
      "loss": 1.7244,
      "step": 25
    },
    {
      "epoch": 0.006313744536182613,
      "grad_norm": 0.38152262568473816,
      "learning_rate": 0.0001,
      "loss": 1.8836,
      "step": 26
    },
    {
      "epoch": 0.006556580864497329,
      "grad_norm": 0.3809744119644165,
      "learning_rate": 0.0001,
      "loss": 1.8553,
      "step": 27
    },
    {
      "epoch": 0.006799417192812044,
      "grad_norm": 0.3849543035030365,
      "learning_rate": 0.0001,
      "loss": 1.9847,
      "step": 28
    },
    {
      "epoch": 0.007042253521126761,
      "grad_norm": 0.3533986806869507,
      "learning_rate": 0.0001,
      "loss": 1.8386,
      "step": 29
    },
    {
      "epoch": 0.007285089849441476,
      "grad_norm": 0.364499568939209,
      "learning_rate": 0.0001,
      "loss": 1.9516,
      "step": 30
    },
    {
      "epoch": 0.0075279261777561925,
      "grad_norm": 0.3528093099594116,
      "learning_rate": 0.0001,
      "loss": 1.8593,
      "step": 31
    },
    {
      "epoch": 0.007770762506070908,
      "grad_norm": 0.3829546570777893,
      "learning_rate": 0.0001,
      "loss": 1.7068,
      "step": 32
    },
    {
      "epoch": 0.008013598834385623,
      "grad_norm": 0.3647037148475647,
      "learning_rate": 0.0001,
      "loss": 1.6717,
      "step": 33
    },
    {
      "epoch": 0.00825643516270034,
      "grad_norm": 0.35392895340919495,
      "learning_rate": 0.0001,
      "loss": 1.9651,
      "step": 34
    },
    {
      "epoch": 0.008499271491015056,
      "grad_norm": 0.3464745879173279,
      "learning_rate": 0.0001,
      "loss": 1.7915,
      "step": 35
    },
    {
      "epoch": 0.008742107819329771,
      "grad_norm": 0.34323614835739136,
      "learning_rate": 0.0001,
      "loss": 1.8433,
      "step": 36
    },
    {
      "epoch": 0.008984944147644487,
      "grad_norm": 0.3553076982498169,
      "learning_rate": 0.0001,
      "loss": 1.863,
      "step": 37
    },
    {
      "epoch": 0.009227780475959204,
      "grad_norm": 0.39482808113098145,
      "learning_rate": 0.0001,
      "loss": 2.0614,
      "step": 38
    },
    {
      "epoch": 0.00947061680427392,
      "grad_norm": 0.34027889370918274,
      "learning_rate": 0.0001,
      "loss": 1.7918,
      "step": 39
    },
    {
      "epoch": 0.009713453132588635,
      "grad_norm": 0.3214944005012512,
      "learning_rate": 0.0001,
      "loss": 1.8169,
      "step": 40
    },
    {
      "epoch": 0.00995628946090335,
      "grad_norm": 0.3573228716850281,
      "learning_rate": 0.0001,
      "loss": 1.8388,
      "step": 41
    },
    {
      "epoch": 0.010199125789218067,
      "grad_norm": 0.34565800428390503,
      "learning_rate": 0.0001,
      "loss": 1.6319,
      "step": 42
    },
    {
      "epoch": 0.010441962117532783,
      "grad_norm": 0.3430301249027252,
      "learning_rate": 0.0001,
      "loss": 1.9434,
      "step": 43
    },
    {
      "epoch": 0.010684798445847498,
      "grad_norm": 0.3680773675441742,
      "learning_rate": 0.0001,
      "loss": 1.9171,
      "step": 44
    },
    {
      "epoch": 0.010927634774162214,
      "grad_norm": 0.29709112644195557,
      "learning_rate": 0.0001,
      "loss": 1.6346,
      "step": 45
    },
    {
      "epoch": 0.011170471102476931,
      "grad_norm": 0.31409093737602234,
      "learning_rate": 0.0001,
      "loss": 1.8279,
      "step": 46
    },
    {
      "epoch": 0.011413307430791646,
      "grad_norm": 0.3398358225822449,
      "learning_rate": 0.0001,
      "loss": 2.0094,
      "step": 47
    },
    {
      "epoch": 0.011656143759106362,
      "grad_norm": 0.35688671469688416,
      "learning_rate": 0.0001,
      "loss": 1.9384,
      "step": 48
    },
    {
      "epoch": 0.011898980087421079,
      "grad_norm": 0.33628571033477783,
      "learning_rate": 0.0001,
      "loss": 1.8191,
      "step": 49
    },
    {
      "epoch": 0.012141816415735794,
      "grad_norm": 0.3309676945209503,
      "learning_rate": 0.0001,
      "loss": 1.8202,
      "step": 50
    },
    {
      "epoch": 0.01238465274405051,
      "grad_norm": 0.3198463022708893,
      "learning_rate": 0.0001,
      "loss": 1.7509,
      "step": 51
    },
    {
      "epoch": 0.012627489072365225,
      "grad_norm": 0.3255227208137512,
      "learning_rate": 0.0001,
      "loss": 1.7996,
      "step": 52
    },
    {
      "epoch": 0.012870325400679943,
      "grad_norm": 0.3298192322254181,
      "learning_rate": 0.0001,
      "loss": 1.8607,
      "step": 53
    },
    {
      "epoch": 0.013113161728994658,
      "grad_norm": 0.33618688583374023,
      "learning_rate": 0.0001,
      "loss": 1.7293,
      "step": 54
    },
    {
      "epoch": 0.013355998057309373,
      "grad_norm": 0.37747520208358765,
      "learning_rate": 0.0001,
      "loss": 2.0701,
      "step": 55
    },
    {
      "epoch": 0.013598834385624089,
      "grad_norm": 0.3369693458080292,
      "learning_rate": 0.0001,
      "loss": 1.728,
      "step": 56
    },
    {
      "epoch": 0.013841670713938806,
      "grad_norm": 0.34572532773017883,
      "learning_rate": 0.0001,
      "loss": 1.9147,
      "step": 57
    },
    {
      "epoch": 0.014084507042253521,
      "grad_norm": 0.3507564067840576,
      "learning_rate": 0.0001,
      "loss": 1.802,
      "step": 58
    },
    {
      "epoch": 0.014327343370568237,
      "grad_norm": 0.32496514916419983,
      "learning_rate": 0.0001,
      "loss": 1.8054,
      "step": 59
    },
    {
      "epoch": 0.014570179698882952,
      "grad_norm": 0.31435415148735046,
      "learning_rate": 0.0001,
      "loss": 1.764,
      "step": 60
    },
    {
      "epoch": 0.01481301602719767,
      "grad_norm": 0.3256989121437073,
      "learning_rate": 0.0001,
      "loss": 1.76,
      "step": 61
    },
    {
      "epoch": 0.015055852355512385,
      "grad_norm": 0.346720814704895,
      "learning_rate": 0.0001,
      "loss": 1.8155,
      "step": 62
    },
    {
      "epoch": 0.0152986886838271,
      "grad_norm": 0.3564370274543762,
      "learning_rate": 0.0001,
      "loss": 1.8645,
      "step": 63
    },
    {
      "epoch": 0.015541525012141816,
      "grad_norm": 0.3128412663936615,
      "learning_rate": 0.0001,
      "loss": 1.7537,
      "step": 64
    },
    {
      "epoch": 0.01578436134045653,
      "grad_norm": 0.3623558580875397,
      "learning_rate": 0.0001,
      "loss": 1.8776,
      "step": 65
    },
    {
      "epoch": 0.016027197668771247,
      "grad_norm": 0.34936532378196716,
      "learning_rate": 0.0001,
      "loss": 1.7606,
      "step": 66
    },
    {
      "epoch": 0.016270033997085966,
      "grad_norm": 0.32261672616004944,
      "learning_rate": 0.0001,
      "loss": 1.8787,
      "step": 67
    },
    {
      "epoch": 0.01651287032540068,
      "grad_norm": 0.3463958203792572,
      "learning_rate": 0.0001,
      "loss": 1.8079,
      "step": 68
    },
    {
      "epoch": 0.016755706653715396,
      "grad_norm": 0.3851228654384613,
      "learning_rate": 0.0001,
      "loss": 2.0075,
      "step": 69
    },
    {
      "epoch": 0.016998542982030112,
      "grad_norm": 0.347501277923584,
      "learning_rate": 0.0001,
      "loss": 1.8561,
      "step": 70
    },
    {
      "epoch": 0.017241379310344827,
      "grad_norm": 0.33600863814353943,
      "learning_rate": 0.0001,
      "loss": 1.9005,
      "step": 71
    },
    {
      "epoch": 0.017484215638659543,
      "grad_norm": 0.3174665868282318,
      "learning_rate": 0.0001,
      "loss": 1.7522,
      "step": 72
    },
    {
      "epoch": 0.017727051966974258,
      "grad_norm": 0.3470480740070343,
      "learning_rate": 0.0001,
      "loss": 1.7866,
      "step": 73
    },
    {
      "epoch": 0.017969888295288974,
      "grad_norm": 0.35113289952278137,
      "learning_rate": 0.0001,
      "loss": 1.9659,
      "step": 74
    },
    {
      "epoch": 0.018212724623603693,
      "grad_norm": 0.35351213812828064,
      "learning_rate": 0.0001,
      "loss": 2.0076,
      "step": 75
    },
    {
      "epoch": 0.018455560951918408,
      "grad_norm": 0.3192991614341736,
      "learning_rate": 0.0001,
      "loss": 1.7576,
      "step": 76
    },
    {
      "epoch": 0.018698397280233123,
      "grad_norm": 0.3431519567966461,
      "learning_rate": 0.0001,
      "loss": 1.7957,
      "step": 77
    },
    {
      "epoch": 0.01894123360854784,
      "grad_norm": 0.3162384033203125,
      "learning_rate": 0.0001,
      "loss": 1.6776,
      "step": 78
    },
    {
      "epoch": 0.019184069936862554,
      "grad_norm": 0.31073257327079773,
      "learning_rate": 0.0001,
      "loss": 1.7227,
      "step": 79
    },
    {
      "epoch": 0.01942690626517727,
      "grad_norm": 0.319278359413147,
      "learning_rate": 0.0001,
      "loss": 1.7953,
      "step": 80
    },
    {
      "epoch": 0.019669742593491985,
      "grad_norm": 0.34543249011039734,
      "learning_rate": 0.0001,
      "loss": 1.8845,
      "step": 81
    },
    {
      "epoch": 0.0199125789218067,
      "grad_norm": 0.35296866297721863,
      "learning_rate": 0.0001,
      "loss": 2.0548,
      "step": 82
    },
    {
      "epoch": 0.02015541525012142,
      "grad_norm": 0.31435465812683105,
      "learning_rate": 0.0001,
      "loss": 1.7889,
      "step": 83
    },
    {
      "epoch": 0.020398251578436135,
      "grad_norm": 0.34530073404312134,
      "learning_rate": 0.0001,
      "loss": 1.8978,
      "step": 84
    },
    {
      "epoch": 0.02064108790675085,
      "grad_norm": 0.34861063957214355,
      "learning_rate": 0.0001,
      "loss": 1.7865,
      "step": 85
    },
    {
      "epoch": 0.020883924235065566,
      "grad_norm": 0.3257801830768585,
      "learning_rate": 0.0001,
      "loss": 1.8877,
      "step": 86
    },
    {
      "epoch": 0.02112676056338028,
      "grad_norm": 0.33134183287620544,
      "learning_rate": 0.0001,
      "loss": 1.9007,
      "step": 87
    },
    {
      "epoch": 0.021369596891694997,
      "grad_norm": 0.3427002727985382,
      "learning_rate": 0.0001,
      "loss": 1.8455,
      "step": 88
    },
    {
      "epoch": 0.021612433220009712,
      "grad_norm": 0.3223351836204529,
      "learning_rate": 0.0001,
      "loss": 1.8392,
      "step": 89
    },
    {
      "epoch": 0.021855269548324428,
      "grad_norm": 0.31952205300331116,
      "learning_rate": 0.0001,
      "loss": 1.805,
      "step": 90
    },
    {
      "epoch": 0.022098105876639147,
      "grad_norm": 0.33756670355796814,
      "learning_rate": 0.0001,
      "loss": 1.7672,
      "step": 91
    },
    {
      "epoch": 0.022340942204953862,
      "grad_norm": 0.32219502329826355,
      "learning_rate": 0.0001,
      "loss": 1.7457,
      "step": 92
    },
    {
      "epoch": 0.022583778533268577,
      "grad_norm": 0.3273763060569763,
      "learning_rate": 0.0001,
      "loss": 1.9236,
      "step": 93
    },
    {
      "epoch": 0.022826614861583293,
      "grad_norm": 0.34903284907341003,
      "learning_rate": 0.0001,
      "loss": 1.9005,
      "step": 94
    },
    {
      "epoch": 0.02306945118989801,
      "grad_norm": 0.31824183464050293,
      "learning_rate": 0.0001,
      "loss": 1.7147,
      "step": 95
    },
    {
      "epoch": 0.023312287518212724,
      "grad_norm": 0.3329814076423645,
      "learning_rate": 0.0001,
      "loss": 1.8056,
      "step": 96
    },
    {
      "epoch": 0.02355512384652744,
      "grad_norm": 0.3619345426559448,
      "learning_rate": 0.0001,
      "loss": 1.9174,
      "step": 97
    },
    {
      "epoch": 0.023797960174842158,
      "grad_norm": 0.32598811388015747,
      "learning_rate": 0.0001,
      "loss": 1.6829,
      "step": 98
    },
    {
      "epoch": 0.024040796503156873,
      "grad_norm": 0.33037275075912476,
      "learning_rate": 0.0001,
      "loss": 1.8287,
      "step": 99
    },
    {
      "epoch": 0.02428363283147159,
      "grad_norm": 0.3756055533885956,
      "learning_rate": 0.0001,
      "loss": 1.907,
      "step": 100
    },
    {
      "epoch": 0.024526469159786304,
      "grad_norm": 0.34237828850746155,
      "learning_rate": 0.0001,
      "loss": 1.8935,
      "step": 101
    },
    {
      "epoch": 0.02476930548810102,
      "grad_norm": 0.3579355478286743,
      "learning_rate": 0.0001,
      "loss": 1.8596,
      "step": 102
    },
    {
      "epoch": 0.025012141816415735,
      "grad_norm": 0.3444015085697174,
      "learning_rate": 0.0001,
      "loss": 1.8065,
      "step": 103
    },
    {
      "epoch": 0.02525497814473045,
      "grad_norm": 0.3132593333721161,
      "learning_rate": 0.0001,
      "loss": 1.4983,
      "step": 104
    },
    {
      "epoch": 0.025497814473045166,
      "grad_norm": 0.3273251950740814,
      "learning_rate": 0.0001,
      "loss": 1.7559,
      "step": 105
    },
    {
      "epoch": 0.025740650801359885,
      "grad_norm": 0.34056222438812256,
      "learning_rate": 0.0001,
      "loss": 1.7862,
      "step": 106
    },
    {
      "epoch": 0.0259834871296746,
      "grad_norm": 0.36062437295913696,
      "learning_rate": 0.0001,
      "loss": 2.0101,
      "step": 107
    },
    {
      "epoch": 0.026226323457989316,
      "grad_norm": 0.30914780497550964,
      "learning_rate": 0.0001,
      "loss": 1.7174,
      "step": 108
    },
    {
      "epoch": 0.02646915978630403,
      "grad_norm": 0.34357085824012756,
      "learning_rate": 0.0001,
      "loss": 1.9291,
      "step": 109
    },
    {
      "epoch": 0.026711996114618747,
      "grad_norm": 0.32527053356170654,
      "learning_rate": 0.0001,
      "loss": 1.7997,
      "step": 110
    },
    {
      "epoch": 0.026954832442933462,
      "grad_norm": 0.3142812252044678,
      "learning_rate": 0.0001,
      "loss": 1.697,
      "step": 111
    },
    {
      "epoch": 0.027197668771248178,
      "grad_norm": 0.32341161370277405,
      "learning_rate": 0.0001,
      "loss": 1.8289,
      "step": 112
    },
    {
      "epoch": 0.027440505099562893,
      "grad_norm": 0.3451857566833496,
      "learning_rate": 0.0001,
      "loss": 1.8117,
      "step": 113
    },
    {
      "epoch": 0.027683341427877612,
      "grad_norm": 0.3279164433479309,
      "learning_rate": 0.0001,
      "loss": 1.6973,
      "step": 114
    },
    {
      "epoch": 0.027926177756192327,
      "grad_norm": 0.31405824422836304,
      "learning_rate": 0.0001,
      "loss": 1.8133,
      "step": 115
    },
    {
      "epoch": 0.028169014084507043,
      "grad_norm": 0.3456348478794098,
      "learning_rate": 0.0001,
      "loss": 1.7479,
      "step": 116
    },
    {
      "epoch": 0.02841185041282176,
      "grad_norm": 0.31980592012405396,
      "learning_rate": 0.0001,
      "loss": 1.6188,
      "step": 117
    },
    {
      "epoch": 0.028654686741136474,
      "grad_norm": 0.3357499837875366,
      "learning_rate": 0.0001,
      "loss": 1.8281,
      "step": 118
    },
    {
      "epoch": 0.02889752306945119,
      "grad_norm": 0.32338961958885193,
      "learning_rate": 0.0001,
      "loss": 1.6977,
      "step": 119
    },
    {
      "epoch": 0.029140359397765905,
      "grad_norm": 0.3575379550457001,
      "learning_rate": 0.0001,
      "loss": 1.8371,
      "step": 120
    },
    {
      "epoch": 0.02938319572608062,
      "grad_norm": 0.35191610455513,
      "learning_rate": 0.0001,
      "loss": 1.8743,
      "step": 121
    },
    {
      "epoch": 0.02962603205439534,
      "grad_norm": 0.3267911374568939,
      "learning_rate": 0.0001,
      "loss": 1.8122,
      "step": 122
    },
    {
      "epoch": 0.029868868382710054,
      "grad_norm": 0.37161776423454285,
      "learning_rate": 0.0001,
      "loss": 1.8069,
      "step": 123
    },
    {
      "epoch": 0.03011170471102477,
      "grad_norm": 0.3311915397644043,
      "learning_rate": 0.0001,
      "loss": 1.8613,
      "step": 124
    },
    {
      "epoch": 0.030354541039339485,
      "grad_norm": 0.34188058972358704,
      "learning_rate": 0.0001,
      "loss": 1.9314,
      "step": 125
    },
    {
      "epoch": 0.0305973773676542,
      "grad_norm": 0.3299490213394165,
      "learning_rate": 0.0001,
      "loss": 1.737,
      "step": 126
    },
    {
      "epoch": 0.030840213695968916,
      "grad_norm": 0.3847339451313019,
      "learning_rate": 0.0001,
      "loss": 1.8562,
      "step": 127
    },
    {
      "epoch": 0.03108305002428363,
      "grad_norm": 0.34246835112571716,
      "learning_rate": 0.0001,
      "loss": 1.9026,
      "step": 128
    },
    {
      "epoch": 0.03132588635259835,
      "grad_norm": 0.33350276947021484,
      "learning_rate": 0.0001,
      "loss": 1.7956,
      "step": 129
    },
    {
      "epoch": 0.03156872268091306,
      "grad_norm": 0.3377183675765991,
      "learning_rate": 0.0001,
      "loss": 1.9385,
      "step": 130
    },
    {
      "epoch": 0.03181155900922778,
      "grad_norm": 0.36835625767707825,
      "learning_rate": 0.0001,
      "loss": 1.9134,
      "step": 131
    },
    {
      "epoch": 0.03205439533754249,
      "grad_norm": 0.3150362968444824,
      "learning_rate": 0.0001,
      "loss": 1.8216,
      "step": 132
    },
    {
      "epoch": 0.03229723166585721,
      "grad_norm": 0.31782564520835876,
      "learning_rate": 0.0001,
      "loss": 1.7913,
      "step": 133
    },
    {
      "epoch": 0.03254006799417193,
      "grad_norm": 0.3307948112487793,
      "learning_rate": 0.0001,
      "loss": 1.9733,
      "step": 134
    },
    {
      "epoch": 0.03278290432248664,
      "grad_norm": 0.3569931089878082,
      "learning_rate": 0.0001,
      "loss": 1.7413,
      "step": 135
    },
    {
      "epoch": 0.03302574065080136,
      "grad_norm": 0.325895756483078,
      "learning_rate": 0.0001,
      "loss": 1.73,
      "step": 136
    },
    {
      "epoch": 0.033268576979116074,
      "grad_norm": 0.33818817138671875,
      "learning_rate": 0.0001,
      "loss": 1.9272,
      "step": 137
    },
    {
      "epoch": 0.03351141330743079,
      "grad_norm": 0.3144405782222748,
      "learning_rate": 0.0001,
      "loss": 1.6437,
      "step": 138
    },
    {
      "epoch": 0.033754249635745505,
      "grad_norm": 0.3207385540008545,
      "learning_rate": 0.0001,
      "loss": 1.7101,
      "step": 139
    },
    {
      "epoch": 0.033997085964060224,
      "grad_norm": 0.34677934646606445,
      "learning_rate": 0.0001,
      "loss": 1.7878,
      "step": 140
    },
    {
      "epoch": 0.03423992229237494,
      "grad_norm": 0.3334146738052368,
      "learning_rate": 0.0001,
      "loss": 1.9245,
      "step": 141
    },
    {
      "epoch": 0.034482758620689655,
      "grad_norm": 0.3302057087421417,
      "learning_rate": 0.0001,
      "loss": 1.8471,
      "step": 142
    },
    {
      "epoch": 0.034725594949004374,
      "grad_norm": 0.3368750214576721,
      "learning_rate": 0.0001,
      "loss": 1.8592,
      "step": 143
    },
    {
      "epoch": 0.034968431277319086,
      "grad_norm": 0.3442443311214447,
      "learning_rate": 0.0001,
      "loss": 1.7756,
      "step": 144
    },
    {
      "epoch": 0.035211267605633804,
      "grad_norm": 0.311424195766449,
      "learning_rate": 0.0001,
      "loss": 1.6777,
      "step": 145
    },
    {
      "epoch": 0.035454103933948516,
      "grad_norm": 0.3602541387081146,
      "learning_rate": 0.0001,
      "loss": 1.9142,
      "step": 146
    },
    {
      "epoch": 0.035696940262263235,
      "grad_norm": 0.34334802627563477,
      "learning_rate": 0.0001,
      "loss": 1.8417,
      "step": 147
    },
    {
      "epoch": 0.03593977659057795,
      "grad_norm": 0.3241439461708069,
      "learning_rate": 0.0001,
      "loss": 1.7917,
      "step": 148
    },
    {
      "epoch": 0.036182612918892666,
      "grad_norm": 0.3447269797325134,
      "learning_rate": 0.0001,
      "loss": 1.9037,
      "step": 149
    },
    {
      "epoch": 0.036425449247207385,
      "grad_norm": 0.3451700508594513,
      "learning_rate": 0.0001,
      "loss": 1.9233,
      "step": 150
    },
    {
      "epoch": 0.0366682855755221,
      "grad_norm": 0.3239392638206482,
      "learning_rate": 0.0001,
      "loss": 1.7476,
      "step": 151
    },
    {
      "epoch": 0.036911121903836816,
      "grad_norm": 0.3280225694179535,
      "learning_rate": 0.0001,
      "loss": 1.6909,
      "step": 152
    },
    {
      "epoch": 0.03715395823215153,
      "grad_norm": 0.304781436920166,
      "learning_rate": 0.0001,
      "loss": 1.7175,
      "step": 153
    },
    {
      "epoch": 0.03739679456046625,
      "grad_norm": 0.325508177280426,
      "learning_rate": 0.0001,
      "loss": 1.8113,
      "step": 154
    },
    {
      "epoch": 0.03763963088878096,
      "grad_norm": 0.35371872782707214,
      "learning_rate": 0.0001,
      "loss": 1.7723,
      "step": 155
    },
    {
      "epoch": 0.03788246721709568,
      "grad_norm": 0.3424810469150543,
      "learning_rate": 0.0001,
      "loss": 1.8947,
      "step": 156
    },
    {
      "epoch": 0.0381253035454104,
      "grad_norm": 0.32752588391304016,
      "learning_rate": 0.0001,
      "loss": 1.6501,
      "step": 157
    },
    {
      "epoch": 0.03836813987372511,
      "grad_norm": 0.345795214176178,
      "learning_rate": 0.0001,
      "loss": 1.9681,
      "step": 158
    },
    {
      "epoch": 0.03861097620203983,
      "grad_norm": 0.33940884470939636,
      "learning_rate": 0.0001,
      "loss": 1.7907,
      "step": 159
    },
    {
      "epoch": 0.03885381253035454,
      "grad_norm": 0.3449257016181946,
      "learning_rate": 0.0001,
      "loss": 1.8764,
      "step": 160
    },
    {
      "epoch": 0.03909664885866926,
      "grad_norm": 0.3274487257003784,
      "learning_rate": 0.0001,
      "loss": 1.7455,
      "step": 161
    },
    {
      "epoch": 0.03933948518698397,
      "grad_norm": 0.3211442828178406,
      "learning_rate": 0.0001,
      "loss": 1.7798,
      "step": 162
    },
    {
      "epoch": 0.03958232151529869,
      "grad_norm": 0.3389868140220642,
      "learning_rate": 0.0001,
      "loss": 1.9089,
      "step": 163
    },
    {
      "epoch": 0.0398251578436134,
      "grad_norm": 0.3453965485095978,
      "learning_rate": 0.0001,
      "loss": 1.7192,
      "step": 164
    },
    {
      "epoch": 0.04006799417192812,
      "grad_norm": 0.3375605344772339,
      "learning_rate": 0.0001,
      "loss": 1.8576,
      "step": 165
    },
    {
      "epoch": 0.04031083050024284,
      "grad_norm": 0.3315996527671814,
      "learning_rate": 0.0001,
      "loss": 1.8289,
      "step": 166
    },
    {
      "epoch": 0.04055366682855755,
      "grad_norm": 0.3440003991127014,
      "learning_rate": 0.0001,
      "loss": 1.8337,
      "step": 167
    },
    {
      "epoch": 0.04079650315687227,
      "grad_norm": 0.3501895070075989,
      "learning_rate": 0.0001,
      "loss": 1.8386,
      "step": 168
    },
    {
      "epoch": 0.04103933948518698,
      "grad_norm": 0.3258925974369049,
      "learning_rate": 0.0001,
      "loss": 1.8062,
      "step": 169
    },
    {
      "epoch": 0.0412821758135017,
      "grad_norm": 0.3586982488632202,
      "learning_rate": 0.0001,
      "loss": 1.7095,
      "step": 170
    },
    {
      "epoch": 0.04152501214181641,
      "grad_norm": 0.33566269278526306,
      "learning_rate": 0.0001,
      "loss": 1.8377,
      "step": 171
    },
    {
      "epoch": 0.04176784847013113,
      "grad_norm": 0.3345046639442444,
      "learning_rate": 0.0001,
      "loss": 1.7932,
      "step": 172
    },
    {
      "epoch": 0.04201068479844585,
      "grad_norm": 0.35495448112487793,
      "learning_rate": 0.0001,
      "loss": 1.971,
      "step": 173
    },
    {
      "epoch": 0.04225352112676056,
      "grad_norm": 0.3234632611274719,
      "learning_rate": 0.0001,
      "loss": 1.6855,
      "step": 174
    },
    {
      "epoch": 0.04249635745507528,
      "grad_norm": 0.33416181802749634,
      "learning_rate": 0.0001,
      "loss": 1.7332,
      "step": 175
    },
    {
      "epoch": 0.042739193783389993,
      "grad_norm": 0.3306339979171753,
      "learning_rate": 0.0001,
      "loss": 1.8605,
      "step": 176
    },
    {
      "epoch": 0.04298203011170471,
      "grad_norm": 0.3303837478160858,
      "learning_rate": 0.0001,
      "loss": 1.7574,
      "step": 177
    },
    {
      "epoch": 0.043224866440019424,
      "grad_norm": 0.3273058235645294,
      "learning_rate": 0.0001,
      "loss": 1.7745,
      "step": 178
    },
    {
      "epoch": 0.04346770276833414,
      "grad_norm": 0.3233434557914734,
      "learning_rate": 0.0001,
      "loss": 1.6972,
      "step": 179
    },
    {
      "epoch": 0.043710539096648855,
      "grad_norm": 0.3248269855976105,
      "learning_rate": 0.0001,
      "loss": 1.7851,
      "step": 180
    },
    {
      "epoch": 0.043953375424963574,
      "grad_norm": 0.3396579921245575,
      "learning_rate": 0.0001,
      "loss": 1.85,
      "step": 181
    },
    {
      "epoch": 0.04419621175327829,
      "grad_norm": 0.32688796520233154,
      "learning_rate": 0.0001,
      "loss": 1.8739,
      "step": 182
    },
    {
      "epoch": 0.044439048081593005,
      "grad_norm": 0.3261704444885254,
      "learning_rate": 0.0001,
      "loss": 1.8259,
      "step": 183
    },
    {
      "epoch": 0.044681884409907724,
      "grad_norm": 0.31057843565940857,
      "learning_rate": 0.0001,
      "loss": 1.4949,
      "step": 184
    },
    {
      "epoch": 0.044924720738222436,
      "grad_norm": 0.32633113861083984,
      "learning_rate": 0.0001,
      "loss": 1.8141,
      "step": 185
    },
    {
      "epoch": 0.045167557066537155,
      "grad_norm": 0.33019405603408813,
      "learning_rate": 0.0001,
      "loss": 1.9112,
      "step": 186
    },
    {
      "epoch": 0.04541039339485187,
      "grad_norm": 0.3250356614589691,
      "learning_rate": 0.0001,
      "loss": 1.6517,
      "step": 187
    },
    {
      "epoch": 0.045653229723166586,
      "grad_norm": 0.3314146101474762,
      "learning_rate": 0.0001,
      "loss": 1.774,
      "step": 188
    },
    {
      "epoch": 0.045896066051481305,
      "grad_norm": 0.31466880440711975,
      "learning_rate": 0.0001,
      "loss": 1.6895,
      "step": 189
    },
    {
      "epoch": 0.04613890237979602,
      "grad_norm": 0.32345980405807495,
      "learning_rate": 0.0001,
      "loss": 1.6928,
      "step": 190
    },
    {
      "epoch": 0.046381738708110735,
      "grad_norm": 0.3433418273925781,
      "learning_rate": 0.0001,
      "loss": 1.8835,
      "step": 191
    },
    {
      "epoch": 0.04662457503642545,
      "grad_norm": 0.31912872195243835,
      "learning_rate": 0.0001,
      "loss": 1.5716,
      "step": 192
    },
    {
      "epoch": 0.046867411364740166,
      "grad_norm": 0.32511550188064575,
      "learning_rate": 0.0001,
      "loss": 1.7293,
      "step": 193
    },
    {
      "epoch": 0.04711024769305488,
      "grad_norm": 0.35534897446632385,
      "learning_rate": 0.0001,
      "loss": 1.8179,
      "step": 194
    },
    {
      "epoch": 0.0473530840213696,
      "grad_norm": 0.408071368932724,
      "learning_rate": 0.0001,
      "loss": 1.6752,
      "step": 195
    },
    {
      "epoch": 0.047595920349684316,
      "grad_norm": 0.349654883146286,
      "learning_rate": 0.0001,
      "loss": 1.805,
      "step": 196
    },
    {
      "epoch": 0.04783875667799903,
      "grad_norm": 0.351043701171875,
      "learning_rate": 0.0001,
      "loss": 1.7478,
      "step": 197
    },
    {
      "epoch": 0.04808159300631375,
      "grad_norm": 0.33979853987693787,
      "learning_rate": 0.0001,
      "loss": 1.778,
      "step": 198
    },
    {
      "epoch": 0.04832442933462846,
      "grad_norm": 0.32497379183769226,
      "learning_rate": 0.0001,
      "loss": 1.7174,
      "step": 199
    },
    {
      "epoch": 0.04856726566294318,
      "grad_norm": 0.3525506854057312,
      "learning_rate": 0.0001,
      "loss": 1.8249,
      "step": 200
    },
    {
      "epoch": 0.04881010199125789,
      "grad_norm": 0.347342312335968,
      "learning_rate": 0.0001,
      "loss": 1.7451,
      "step": 201
    },
    {
      "epoch": 0.04905293831957261,
      "grad_norm": 0.3493291139602661,
      "learning_rate": 0.0001,
      "loss": 1.9292,
      "step": 202
    },
    {
      "epoch": 0.04929577464788732,
      "grad_norm": 0.33886533975601196,
      "learning_rate": 0.0001,
      "loss": 1.7692,
      "step": 203
    },
    {
      "epoch": 0.04953861097620204,
      "grad_norm": 0.3711925148963928,
      "learning_rate": 0.0001,
      "loss": 2.0593,
      "step": 204
    },
    {
      "epoch": 0.04978144730451676,
      "grad_norm": 0.35785242915153503,
      "learning_rate": 0.0001,
      "loss": 1.8335,
      "step": 205
    },
    {
      "epoch": 0.05002428363283147,
      "grad_norm": 0.33297452330589294,
      "learning_rate": 0.0001,
      "loss": 1.6759,
      "step": 206
    },
    {
      "epoch": 0.05026711996114619,
      "grad_norm": 0.3229871988296509,
      "learning_rate": 0.0001,
      "loss": 1.767,
      "step": 207
    },
    {
      "epoch": 0.0505099562894609,
      "grad_norm": 0.328980952501297,
      "learning_rate": 0.0001,
      "loss": 1.7328,
      "step": 208
    },
    {
      "epoch": 0.05075279261777562,
      "grad_norm": 0.324479341506958,
      "learning_rate": 0.0001,
      "loss": 1.7687,
      "step": 209
    },
    {
      "epoch": 0.05099562894609033,
      "grad_norm": 0.33051010966300964,
      "learning_rate": 0.0001,
      "loss": 1.7712,
      "step": 210
    },
    {
      "epoch": 0.05123846527440505,
      "grad_norm": 0.32872438430786133,
      "learning_rate": 0.0001,
      "loss": 1.9356,
      "step": 211
    },
    {
      "epoch": 0.05148130160271977,
      "grad_norm": 0.314836710691452,
      "learning_rate": 0.0001,
      "loss": 1.6447,
      "step": 212
    },
    {
      "epoch": 0.05172413793103448,
      "grad_norm": 0.3571523129940033,
      "learning_rate": 0.0001,
      "loss": 1.8033,
      "step": 213
    },
    {
      "epoch": 0.0519669742593492,
      "grad_norm": 0.34860020875930786,
      "learning_rate": 0.0001,
      "loss": 1.7296,
      "step": 214
    },
    {
      "epoch": 0.05220981058766391,
      "grad_norm": 0.33273938298225403,
      "learning_rate": 0.0001,
      "loss": 1.8285,
      "step": 215
    },
    {
      "epoch": 0.05245264691597863,
      "grad_norm": 0.3613104522228241,
      "learning_rate": 0.0001,
      "loss": 1.9188,
      "step": 216
    },
    {
      "epoch": 0.052695483244293344,
      "grad_norm": 0.33187010884284973,
      "learning_rate": 0.0001,
      "loss": 1.8428,
      "step": 217
    },
    {
      "epoch": 0.05293831957260806,
      "grad_norm": 0.34256044030189514,
      "learning_rate": 0.0001,
      "loss": 1.9037,
      "step": 218
    },
    {
      "epoch": 0.053181155900922775,
      "grad_norm": 0.34337806701660156,
      "learning_rate": 0.0001,
      "loss": 1.7928,
      "step": 219
    },
    {
      "epoch": 0.053423992229237494,
      "grad_norm": 0.3422665596008301,
      "learning_rate": 0.0001,
      "loss": 1.9033,
      "step": 220
    },
    {
      "epoch": 0.05366682855755221,
      "grad_norm": 0.3296110928058624,
      "learning_rate": 0.0001,
      "loss": 1.8822,
      "step": 221
    },
    {
      "epoch": 0.053909664885866924,
      "grad_norm": 0.3197755515575409,
      "learning_rate": 0.0001,
      "loss": 1.6152,
      "step": 222
    },
    {
      "epoch": 0.05415250121418164,
      "grad_norm": 0.3136097490787506,
      "learning_rate": 0.0001,
      "loss": 1.6212,
      "step": 223
    },
    {
      "epoch": 0.054395337542496355,
      "grad_norm": 0.32114914059638977,
      "learning_rate": 0.0001,
      "loss": 1.8609,
      "step": 224
    },
    {
      "epoch": 0.054638173870811074,
      "grad_norm": 0.3307587206363678,
      "learning_rate": 0.0001,
      "loss": 1.6285,
      "step": 225
    },
    {
      "epoch": 0.054881010199125786,
      "grad_norm": 0.3631536364555359,
      "learning_rate": 0.0001,
      "loss": 1.8162,
      "step": 226
    },
    {
      "epoch": 0.055123846527440505,
      "grad_norm": 0.3194067180156708,
      "learning_rate": 0.0001,
      "loss": 1.7193,
      "step": 227
    },
    {
      "epoch": 0.055366682855755224,
      "grad_norm": 0.32269614934921265,
      "learning_rate": 0.0001,
      "loss": 1.6614,
      "step": 228
    },
    {
      "epoch": 0.055609519184069936,
      "grad_norm": 0.3524293303489685,
      "learning_rate": 0.0001,
      "loss": 1.7887,
      "step": 229
    },
    {
      "epoch": 0.055852355512384655,
      "grad_norm": 0.346718966960907,
      "learning_rate": 0.0001,
      "loss": 1.837,
      "step": 230
    },
    {
      "epoch": 0.05609519184069937,
      "grad_norm": 0.33732110261917114,
      "learning_rate": 0.0001,
      "loss": 1.7503,
      "step": 231
    },
    {
      "epoch": 0.056338028169014086,
      "grad_norm": 0.3578680753707886,
      "learning_rate": 0.0001,
      "loss": 1.9632,
      "step": 232
    },
    {
      "epoch": 0.0565808644973288,
      "grad_norm": 0.29416367411613464,
      "learning_rate": 0.0001,
      "loss": 1.5471,
      "step": 233
    },
    {
      "epoch": 0.05682370082564352,
      "grad_norm": 0.33240172266960144,
      "learning_rate": 0.0001,
      "loss": 1.8694,
      "step": 234
    },
    {
      "epoch": 0.057066537153958236,
      "grad_norm": 0.2979382574558258,
      "learning_rate": 0.0001,
      "loss": 1.5838,
      "step": 235
    },
    {
      "epoch": 0.05730937348227295,
      "grad_norm": 0.36220940947532654,
      "learning_rate": 0.0001,
      "loss": 1.9537,
      "step": 236
    },
    {
      "epoch": 0.057552209810587666,
      "grad_norm": 0.3136692941188812,
      "learning_rate": 0.0001,
      "loss": 1.8313,
      "step": 237
    },
    {
      "epoch": 0.05779504613890238,
      "grad_norm": 0.323866069316864,
      "learning_rate": 0.0001,
      "loss": 1.881,
      "step": 238
    },
    {
      "epoch": 0.0580378824672171,
      "grad_norm": 0.32461974024772644,
      "learning_rate": 0.0001,
      "loss": 1.6583,
      "step": 239
    },
    {
      "epoch": 0.05828071879553181,
      "grad_norm": 0.34341827034950256,
      "learning_rate": 0.0001,
      "loss": 1.8666,
      "step": 240
    },
    {
      "epoch": 0.05852355512384653,
      "grad_norm": 0.3396712839603424,
      "learning_rate": 0.0001,
      "loss": 1.8707,
      "step": 241
    },
    {
      "epoch": 0.05876639145216124,
      "grad_norm": 0.32624468207359314,
      "learning_rate": 0.0001,
      "loss": 1.8068,
      "step": 242
    },
    {
      "epoch": 0.05900922778047596,
      "grad_norm": 0.36357343196868896,
      "learning_rate": 0.0001,
      "loss": 1.9051,
      "step": 243
    },
    {
      "epoch": 0.05925206410879068,
      "grad_norm": 0.3347633481025696,
      "learning_rate": 0.0001,
      "loss": 1.7729,
      "step": 244
    },
    {
      "epoch": 0.05949490043710539,
      "grad_norm": 0.33824750781059265,
      "learning_rate": 0.0001,
      "loss": 1.8523,
      "step": 245
    },
    {
      "epoch": 0.05973773676542011,
      "grad_norm": 0.3478322923183441,
      "learning_rate": 0.0001,
      "loss": 1.8405,
      "step": 246
    },
    {
      "epoch": 0.05998057309373482,
      "grad_norm": 0.33303165435791016,
      "learning_rate": 0.0001,
      "loss": 1.78,
      "step": 247
    },
    {
      "epoch": 0.06022340942204954,
      "grad_norm": 0.319014310836792,
      "learning_rate": 0.0001,
      "loss": 1.7309,
      "step": 248
    },
    {
      "epoch": 0.06046624575036425,
      "grad_norm": 0.3302060067653656,
      "learning_rate": 0.0001,
      "loss": 1.7848,
      "step": 249
    },
    {
      "epoch": 0.06070908207867897,
      "grad_norm": 0.3306678831577301,
      "learning_rate": 0.0001,
      "loss": 1.7027,
      "step": 250
    },
    {
      "epoch": 0.06095191840699369,
      "grad_norm": 0.3561919629573822,
      "learning_rate": 0.0001,
      "loss": 1.8096,
      "step": 251
    },
    {
      "epoch": 0.0611947547353084,
      "grad_norm": 0.3397621214389801,
      "learning_rate": 0.0001,
      "loss": 1.8675,
      "step": 252
    },
    {
      "epoch": 0.06143759106362312,
      "grad_norm": 0.3315463066101074,
      "learning_rate": 0.0001,
      "loss": 1.6153,
      "step": 253
    },
    {
      "epoch": 0.06168042739193783,
      "grad_norm": 0.33696696162223816,
      "learning_rate": 0.0001,
      "loss": 1.8535,
      "step": 254
    },
    {
      "epoch": 0.06192326372025255,
      "grad_norm": 0.33258795738220215,
      "learning_rate": 0.0001,
      "loss": 1.92,
      "step": 255
    },
    {
      "epoch": 0.06216610004856726,
      "grad_norm": 0.3105964660644531,
      "learning_rate": 0.0001,
      "loss": 1.6666,
      "step": 256
    },
    {
      "epoch": 0.06240893637688198,
      "grad_norm": 0.3291695713996887,
      "learning_rate": 0.0001,
      "loss": 1.7214,
      "step": 257
    },
    {
      "epoch": 0.0626517727051967,
      "grad_norm": 0.3137649893760681,
      "learning_rate": 0.0001,
      "loss": 1.7208,
      "step": 258
    },
    {
      "epoch": 0.06289460903351142,
      "grad_norm": 0.32082751393318176,
      "learning_rate": 0.0001,
      "loss": 1.6783,
      "step": 259
    },
    {
      "epoch": 0.06313744536182612,
      "grad_norm": 0.33842983841896057,
      "learning_rate": 0.0001,
      "loss": 1.8637,
      "step": 260
    },
    {
      "epoch": 0.06338028169014084,
      "grad_norm": 0.31122681498527527,
      "learning_rate": 0.0001,
      "loss": 1.6713,
      "step": 261
    },
    {
      "epoch": 0.06362311801845556,
      "grad_norm": 0.3500661253929138,
      "learning_rate": 0.0001,
      "loss": 1.9031,
      "step": 262
    },
    {
      "epoch": 0.06386595434677028,
      "grad_norm": 0.3374018371105194,
      "learning_rate": 0.0001,
      "loss": 1.7129,
      "step": 263
    },
    {
      "epoch": 0.06410879067508499,
      "grad_norm": 0.3464416563510895,
      "learning_rate": 0.0001,
      "loss": 1.8661,
      "step": 264
    },
    {
      "epoch": 0.0643516270033997,
      "grad_norm": 0.34645119309425354,
      "learning_rate": 0.0001,
      "loss": 1.7762,
      "step": 265
    },
    {
      "epoch": 0.06459446333171442,
      "grad_norm": 0.3312506079673767,
      "learning_rate": 0.0001,
      "loss": 1.7213,
      "step": 266
    },
    {
      "epoch": 0.06483729966002914,
      "grad_norm": 0.33370789885520935,
      "learning_rate": 0.0001,
      "loss": 1.8174,
      "step": 267
    },
    {
      "epoch": 0.06508013598834386,
      "grad_norm": 0.36095327138900757,
      "learning_rate": 0.0001,
      "loss": 1.864,
      "step": 268
    },
    {
      "epoch": 0.06532297231665857,
      "grad_norm": 0.35194554924964905,
      "learning_rate": 0.0001,
      "loss": 1.9681,
      "step": 269
    },
    {
      "epoch": 0.06556580864497329,
      "grad_norm": 0.321267306804657,
      "learning_rate": 0.0001,
      "loss": 1.8207,
      "step": 270
    },
    {
      "epoch": 0.065808644973288,
      "grad_norm": 0.3664778769016266,
      "learning_rate": 0.0001,
      "loss": 1.8354,
      "step": 271
    },
    {
      "epoch": 0.06605148130160272,
      "grad_norm": 0.37435176968574524,
      "learning_rate": 0.0001,
      "loss": 1.8364,
      "step": 272
    },
    {
      "epoch": 0.06629431762991743,
      "grad_norm": 0.352040559053421,
      "learning_rate": 0.0001,
      "loss": 1.7374,
      "step": 273
    },
    {
      "epoch": 0.06653715395823215,
      "grad_norm": 0.3306851089000702,
      "learning_rate": 0.0001,
      "loss": 1.6247,
      "step": 274
    },
    {
      "epoch": 0.06677999028654687,
      "grad_norm": 0.3537105321884155,
      "learning_rate": 0.0001,
      "loss": 2.0477,
      "step": 275
    },
    {
      "epoch": 0.06702282661486159,
      "grad_norm": 0.33372071385383606,
      "learning_rate": 0.0001,
      "loss": 1.6669,
      "step": 276
    },
    {
      "epoch": 0.0672656629431763,
      "grad_norm": 0.3090078830718994,
      "learning_rate": 0.0001,
      "loss": 1.6595,
      "step": 277
    },
    {
      "epoch": 0.06750849927149101,
      "grad_norm": 0.3390367031097412,
      "learning_rate": 0.0001,
      "loss": 1.8347,
      "step": 278
    },
    {
      "epoch": 0.06775133559980573,
      "grad_norm": 0.34644871950149536,
      "learning_rate": 0.0001,
      "loss": 1.7536,
      "step": 279
    },
    {
      "epoch": 0.06799417192812045,
      "grad_norm": 0.3659220337867737,
      "learning_rate": 0.0001,
      "loss": 1.9978,
      "step": 280
    },
    {
      "epoch": 0.06823700825643517,
      "grad_norm": 0.3907443881034851,
      "learning_rate": 0.0001,
      "loss": 1.7997,
      "step": 281
    },
    {
      "epoch": 0.06847984458474989,
      "grad_norm": 0.3244320750236511,
      "learning_rate": 0.0001,
      "loss": 1.7634,
      "step": 282
    },
    {
      "epoch": 0.06872268091306459,
      "grad_norm": 0.3205203115940094,
      "learning_rate": 0.0001,
      "loss": 1.7509,
      "step": 283
    },
    {
      "epoch": 0.06896551724137931,
      "grad_norm": 0.3349047303199768,
      "learning_rate": 0.0001,
      "loss": 1.8051,
      "step": 284
    },
    {
      "epoch": 0.06920835356969403,
      "grad_norm": 0.3221355080604553,
      "learning_rate": 0.0001,
      "loss": 1.7895,
      "step": 285
    },
    {
      "epoch": 0.06945118989800875,
      "grad_norm": 0.3335476517677307,
      "learning_rate": 0.0001,
      "loss": 1.6604,
      "step": 286
    },
    {
      "epoch": 0.06969402622632345,
      "grad_norm": 0.34389203786849976,
      "learning_rate": 0.0001,
      "loss": 1.8544,
      "step": 287
    },
    {
      "epoch": 0.06993686255463817,
      "grad_norm": 0.3776070773601532,
      "learning_rate": 0.0001,
      "loss": 2.0003,
      "step": 288
    },
    {
      "epoch": 0.07017969888295289,
      "grad_norm": 0.3268282115459442,
      "learning_rate": 0.0001,
      "loss": 1.8411,
      "step": 289
    },
    {
      "epoch": 0.07042253521126761,
      "grad_norm": 0.33194541931152344,
      "learning_rate": 0.0001,
      "loss": 1.6502,
      "step": 290
    },
    {
      "epoch": 0.07066537153958233,
      "grad_norm": 0.3324213922023773,
      "learning_rate": 0.0001,
      "loss": 1.8558,
      "step": 291
    },
    {
      "epoch": 0.07090820786789703,
      "grad_norm": 0.3163861930370331,
      "learning_rate": 0.0001,
      "loss": 1.7196,
      "step": 292
    },
    {
      "epoch": 0.07115104419621175,
      "grad_norm": 0.3342863917350769,
      "learning_rate": 0.0001,
      "loss": 1.7527,
      "step": 293
    },
    {
      "epoch": 0.07139388052452647,
      "grad_norm": 0.32636377215385437,
      "learning_rate": 0.0001,
      "loss": 1.7561,
      "step": 294
    },
    {
      "epoch": 0.07163671685284119,
      "grad_norm": 0.3220379650592804,
      "learning_rate": 0.0001,
      "loss": 1.8233,
      "step": 295
    },
    {
      "epoch": 0.0718795531811559,
      "grad_norm": 0.3426109552383423,
      "learning_rate": 0.0001,
      "loss": 1.772,
      "step": 296
    },
    {
      "epoch": 0.07212238950947061,
      "grad_norm": 0.31494760513305664,
      "learning_rate": 0.0001,
      "loss": 1.71,
      "step": 297
    },
    {
      "epoch": 0.07236522583778533,
      "grad_norm": 0.3238487243652344,
      "learning_rate": 0.0001,
      "loss": 1.7168,
      "step": 298
    },
    {
      "epoch": 0.07260806216610005,
      "grad_norm": 0.34287282824516296,
      "learning_rate": 0.0001,
      "loss": 1.7304,
      "step": 299
    },
    {
      "epoch": 0.07285089849441477,
      "grad_norm": 0.3240461051464081,
      "learning_rate": 0.0001,
      "loss": 1.7774,
      "step": 300
    },
    {
      "epoch": 0.07309373482272948,
      "grad_norm": 0.33182549476623535,
      "learning_rate": 0.0001,
      "loss": 1.8504,
      "step": 301
    },
    {
      "epoch": 0.0733365711510442,
      "grad_norm": 0.3246839940547943,
      "learning_rate": 0.0001,
      "loss": 1.6562,
      "step": 302
    },
    {
      "epoch": 0.07357940747935891,
      "grad_norm": 0.33326035737991333,
      "learning_rate": 0.0001,
      "loss": 1.7738,
      "step": 303
    },
    {
      "epoch": 0.07382224380767363,
      "grad_norm": 0.3405746817588806,
      "learning_rate": 0.0001,
      "loss": 1.8336,
      "step": 304
    },
    {
      "epoch": 0.07406508013598834,
      "grad_norm": 0.328939825296402,
      "learning_rate": 0.0001,
      "loss": 1.5373,
      "step": 305
    },
    {
      "epoch": 0.07430791646430306,
      "grad_norm": 0.3274134397506714,
      "learning_rate": 0.0001,
      "loss": 1.834,
      "step": 306
    },
    {
      "epoch": 0.07455075279261777,
      "grad_norm": 0.3510563373565674,
      "learning_rate": 0.0001,
      "loss": 1.8107,
      "step": 307
    },
    {
      "epoch": 0.0747935891209325,
      "grad_norm": 0.3266802728176117,
      "learning_rate": 0.0001,
      "loss": 1.8049,
      "step": 308
    },
    {
      "epoch": 0.07503642544924721,
      "grad_norm": 0.3214014172554016,
      "learning_rate": 0.0001,
      "loss": 1.7304,
      "step": 309
    },
    {
      "epoch": 0.07527926177756192,
      "grad_norm": 0.3484552800655365,
      "learning_rate": 0.0001,
      "loss": 1.8983,
      "step": 310
    },
    {
      "epoch": 0.07552209810587664,
      "grad_norm": 0.32560989260673523,
      "learning_rate": 0.0001,
      "loss": 1.7182,
      "step": 311
    },
    {
      "epoch": 0.07576493443419136,
      "grad_norm": 0.31363043189048767,
      "learning_rate": 0.0001,
      "loss": 1.6791,
      "step": 312
    },
    {
      "epoch": 0.07600777076250607,
      "grad_norm": 0.30790412425994873,
      "learning_rate": 0.0001,
      "loss": 1.646,
      "step": 313
    },
    {
      "epoch": 0.0762506070908208,
      "grad_norm": 0.35224685072898865,
      "learning_rate": 0.0001,
      "loss": 1.7997,
      "step": 314
    },
    {
      "epoch": 0.0764934434191355,
      "grad_norm": 0.3509777784347534,
      "learning_rate": 0.0001,
      "loss": 1.8086,
      "step": 315
    },
    {
      "epoch": 0.07673627974745022,
      "grad_norm": 0.45471304655075073,
      "learning_rate": 0.0001,
      "loss": 1.7234,
      "step": 316
    },
    {
      "epoch": 0.07697911607576494,
      "grad_norm": 0.3238165080547333,
      "learning_rate": 0.0001,
      "loss": 1.8179,
      "step": 317
    },
    {
      "epoch": 0.07722195240407966,
      "grad_norm": 0.31785327196121216,
      "learning_rate": 0.0001,
      "loss": 1.6927,
      "step": 318
    },
    {
      "epoch": 0.07746478873239436,
      "grad_norm": 0.33723047375679016,
      "learning_rate": 0.0001,
      "loss": 1.807,
      "step": 319
    },
    {
      "epoch": 0.07770762506070908,
      "grad_norm": 0.3171544373035431,
      "learning_rate": 0.0001,
      "loss": 1.6861,
      "step": 320
    },
    {
      "epoch": 0.0779504613890238,
      "grad_norm": 0.31941697001457214,
      "learning_rate": 0.0001,
      "loss": 1.7365,
      "step": 321
    },
    {
      "epoch": 0.07819329771733852,
      "grad_norm": 0.348278284072876,
      "learning_rate": 0.0001,
      "loss": 1.7495,
      "step": 322
    },
    {
      "epoch": 0.07843613404565324,
      "grad_norm": 0.33012667298316956,
      "learning_rate": 0.0001,
      "loss": 1.6431,
      "step": 323
    },
    {
      "epoch": 0.07867897037396794,
      "grad_norm": 0.3364087641239166,
      "learning_rate": 0.0001,
      "loss": 1.8088,
      "step": 324
    },
    {
      "epoch": 0.07892180670228266,
      "grad_norm": 0.31294387578964233,
      "learning_rate": 0.0001,
      "loss": 1.7033,
      "step": 325
    },
    {
      "epoch": 0.07916464303059738,
      "grad_norm": 0.32259300351142883,
      "learning_rate": 0.0001,
      "loss": 1.7142,
      "step": 326
    },
    {
      "epoch": 0.0794074793589121,
      "grad_norm": 0.34282100200653076,
      "learning_rate": 0.0001,
      "loss": 1.8833,
      "step": 327
    },
    {
      "epoch": 0.0796503156872268,
      "grad_norm": 0.34852755069732666,
      "learning_rate": 0.0001,
      "loss": 1.8127,
      "step": 328
    },
    {
      "epoch": 0.07989315201554152,
      "grad_norm": 0.31135955452919006,
      "learning_rate": 0.0001,
      "loss": 1.6054,
      "step": 329
    },
    {
      "epoch": 0.08013598834385624,
      "grad_norm": 0.34507012367248535,
      "learning_rate": 0.0001,
      "loss": 1.8801,
      "step": 330
    },
    {
      "epoch": 0.08037882467217096,
      "grad_norm": 0.33243077993392944,
      "learning_rate": 0.0001,
      "loss": 1.7688,
      "step": 331
    },
    {
      "epoch": 0.08062166100048568,
      "grad_norm": 0.32714489102363586,
      "learning_rate": 0.0001,
      "loss": 1.7063,
      "step": 332
    },
    {
      "epoch": 0.08086449732880038,
      "grad_norm": 0.337786465883255,
      "learning_rate": 0.0001,
      "loss": 1.8207,
      "step": 333
    },
    {
      "epoch": 0.0811073336571151,
      "grad_norm": 0.3262454569339752,
      "learning_rate": 0.0001,
      "loss": 1.7529,
      "step": 334
    },
    {
      "epoch": 0.08135016998542982,
      "grad_norm": 0.29085323214530945,
      "learning_rate": 0.0001,
      "loss": 1.5068,
      "step": 335
    },
    {
      "epoch": 0.08159300631374454,
      "grad_norm": 0.32306885719299316,
      "learning_rate": 0.0001,
      "loss": 1.7703,
      "step": 336
    },
    {
      "epoch": 0.08183584264205926,
      "grad_norm": 0.3417051434516907,
      "learning_rate": 0.0001,
      "loss": 1.8377,
      "step": 337
    },
    {
      "epoch": 0.08207867897037396,
      "grad_norm": 0.3171888589859009,
      "learning_rate": 0.0001,
      "loss": 1.7178,
      "step": 338
    },
    {
      "epoch": 0.08232151529868868,
      "grad_norm": 0.31271278858184814,
      "learning_rate": 0.0001,
      "loss": 1.6236,
      "step": 339
    },
    {
      "epoch": 0.0825643516270034,
      "grad_norm": 0.3355056047439575,
      "learning_rate": 0.0001,
      "loss": 1.744,
      "step": 340
    },
    {
      "epoch": 0.08280718795531812,
      "grad_norm": 0.3089101016521454,
      "learning_rate": 0.0001,
      "loss": 1.5724,
      "step": 341
    },
    {
      "epoch": 0.08305002428363283,
      "grad_norm": 0.3490867614746094,
      "learning_rate": 0.0001,
      "loss": 1.8589,
      "step": 342
    },
    {
      "epoch": 0.08329286061194754,
      "grad_norm": 0.3519854247570038,
      "learning_rate": 0.0001,
      "loss": 1.8846,
      "step": 343
    },
    {
      "epoch": 0.08353569694026226,
      "grad_norm": 0.3247239291667938,
      "learning_rate": 0.0001,
      "loss": 1.8309,
      "step": 344
    },
    {
      "epoch": 0.08377853326857698,
      "grad_norm": 0.3213523030281067,
      "learning_rate": 0.0001,
      "loss": 1.652,
      "step": 345
    },
    {
      "epoch": 0.0840213695968917,
      "grad_norm": 0.32256901264190674,
      "learning_rate": 0.0001,
      "loss": 1.6963,
      "step": 346
    },
    {
      "epoch": 0.0842642059252064,
      "grad_norm": 0.3471747934818268,
      "learning_rate": 0.0001,
      "loss": 1.786,
      "step": 347
    },
    {
      "epoch": 0.08450704225352113,
      "grad_norm": 0.2998027503490448,
      "learning_rate": 0.0001,
      "loss": 1.5939,
      "step": 348
    },
    {
      "epoch": 0.08474987858183584,
      "grad_norm": 0.3406505286693573,
      "learning_rate": 0.0001,
      "loss": 1.7973,
      "step": 349
    },
    {
      "epoch": 0.08499271491015056,
      "grad_norm": 0.3450128734111786,
      "learning_rate": 0.0001,
      "loss": 1.7727,
      "step": 350
    },
    {
      "epoch": 0.08523555123846527,
      "grad_norm": 0.33610254526138306,
      "learning_rate": 0.0001,
      "loss": 1.8183,
      "step": 351
    },
    {
      "epoch": 0.08547838756677999,
      "grad_norm": 0.3284231126308441,
      "learning_rate": 0.0001,
      "loss": 1.773,
      "step": 352
    },
    {
      "epoch": 0.0857212238950947,
      "grad_norm": 0.31227028369903564,
      "learning_rate": 0.0001,
      "loss": 1.7105,
      "step": 353
    },
    {
      "epoch": 0.08596406022340942,
      "grad_norm": 0.32455405592918396,
      "learning_rate": 0.0001,
      "loss": 1.6372,
      "step": 354
    },
    {
      "epoch": 0.08620689655172414,
      "grad_norm": 0.3156833052635193,
      "learning_rate": 0.0001,
      "loss": 1.8633,
      "step": 355
    },
    {
      "epoch": 0.08644973288003885,
      "grad_norm": 0.33389779925346375,
      "learning_rate": 0.0001,
      "loss": 1.6658,
      "step": 356
    },
    {
      "epoch": 0.08669256920835357,
      "grad_norm": 0.3363843560218811,
      "learning_rate": 0.0001,
      "loss": 1.751,
      "step": 357
    },
    {
      "epoch": 0.08693540553666829,
      "grad_norm": 0.33338844776153564,
      "learning_rate": 0.0001,
      "loss": 1.6921,
      "step": 358
    },
    {
      "epoch": 0.087178241864983,
      "grad_norm": 0.33218914270401,
      "learning_rate": 0.0001,
      "loss": 1.9268,
      "step": 359
    },
    {
      "epoch": 0.08742107819329771,
      "grad_norm": 0.32392585277557373,
      "learning_rate": 0.0001,
      "loss": 1.7296,
      "step": 360
    },
    {
      "epoch": 0.08766391452161243,
      "grad_norm": 0.33016619086265564,
      "learning_rate": 0.0001,
      "loss": 1.5462,
      "step": 361
    },
    {
      "epoch": 0.08790675084992715,
      "grad_norm": 0.34675028920173645,
      "learning_rate": 0.0001,
      "loss": 1.8096,
      "step": 362
    },
    {
      "epoch": 0.08814958717824187,
      "grad_norm": 0.36310189962387085,
      "learning_rate": 0.0001,
      "loss": 1.8494,
      "step": 363
    },
    {
      "epoch": 0.08839242350655659,
      "grad_norm": 0.31934189796447754,
      "learning_rate": 0.0001,
      "loss": 1.7754,
      "step": 364
    },
    {
      "epoch": 0.08863525983487129,
      "grad_norm": 0.3338501453399658,
      "learning_rate": 0.0001,
      "loss": 1.7632,
      "step": 365
    },
    {
      "epoch": 0.08887809616318601,
      "grad_norm": 0.361849308013916,
      "learning_rate": 0.0001,
      "loss": 1.9469,
      "step": 366
    },
    {
      "epoch": 0.08912093249150073,
      "grad_norm": 0.3035111129283905,
      "learning_rate": 0.0001,
      "loss": 1.6553,
      "step": 367
    },
    {
      "epoch": 0.08936376881981545,
      "grad_norm": 0.3231661021709442,
      "learning_rate": 0.0001,
      "loss": 1.8295,
      "step": 368
    },
    {
      "epoch": 0.08960660514813017,
      "grad_norm": 0.33550918102264404,
      "learning_rate": 0.0001,
      "loss": 1.8573,
      "step": 369
    },
    {
      "epoch": 0.08984944147644487,
      "grad_norm": 0.3429529666900635,
      "learning_rate": 0.0001,
      "loss": 1.8223,
      "step": 370
    },
    {
      "epoch": 0.09009227780475959,
      "grad_norm": 0.31798985600471497,
      "learning_rate": 0.0001,
      "loss": 1.6636,
      "step": 371
    },
    {
      "epoch": 0.09033511413307431,
      "grad_norm": 0.3305976986885071,
      "learning_rate": 0.0001,
      "loss": 1.7622,
      "step": 372
    },
    {
      "epoch": 0.09057795046138903,
      "grad_norm": 0.32804372906684875,
      "learning_rate": 0.0001,
      "loss": 1.7525,
      "step": 373
    },
    {
      "epoch": 0.09082078678970373,
      "grad_norm": 0.33984291553497314,
      "learning_rate": 0.0001,
      "loss": 1.9339,
      "step": 374
    },
    {
      "epoch": 0.09106362311801845,
      "grad_norm": 0.29461875557899475,
      "learning_rate": 0.0001,
      "loss": 1.7179,
      "step": 375
    },
    {
      "epoch": 0.09130645944633317,
      "grad_norm": 0.3483908474445343,
      "learning_rate": 0.0001,
      "loss": 1.8206,
      "step": 376
    },
    {
      "epoch": 0.09154929577464789,
      "grad_norm": 0.30330803990364075,
      "learning_rate": 0.0001,
      "loss": 1.6415,
      "step": 377
    },
    {
      "epoch": 0.09179213210296261,
      "grad_norm": 0.34878110885620117,
      "learning_rate": 0.0001,
      "loss": 1.7985,
      "step": 378
    },
    {
      "epoch": 0.09203496843127731,
      "grad_norm": 0.32693809270858765,
      "learning_rate": 0.0001,
      "loss": 1.7828,
      "step": 379
    },
    {
      "epoch": 0.09227780475959203,
      "grad_norm": 0.3255230486392975,
      "learning_rate": 0.0001,
      "loss": 1.7238,
      "step": 380
    },
    {
      "epoch": 0.09252064108790675,
      "grad_norm": 0.3401491045951843,
      "learning_rate": 0.0001,
      "loss": 1.7669,
      "step": 381
    },
    {
      "epoch": 0.09276347741622147,
      "grad_norm": 0.3755502998828888,
      "learning_rate": 0.0001,
      "loss": 1.8206,
      "step": 382
    },
    {
      "epoch": 0.09300631374453618,
      "grad_norm": 0.3082270920276642,
      "learning_rate": 0.0001,
      "loss": 1.7214,
      "step": 383
    },
    {
      "epoch": 0.0932491500728509,
      "grad_norm": 0.3387890160083771,
      "learning_rate": 0.0001,
      "loss": 1.8306,
      "step": 384
    },
    {
      "epoch": 0.09349198640116561,
      "grad_norm": 0.3410017192363739,
      "learning_rate": 0.0001,
      "loss": 1.7629,
      "step": 385
    },
    {
      "epoch": 0.09373482272948033,
      "grad_norm": 0.32617995142936707,
      "learning_rate": 0.0001,
      "loss": 1.6484,
      "step": 386
    },
    {
      "epoch": 0.09397765905779505,
      "grad_norm": 0.3287573754787445,
      "learning_rate": 0.0001,
      "loss": 1.8165,
      "step": 387
    },
    {
      "epoch": 0.09422049538610976,
      "grad_norm": 0.35360467433929443,
      "learning_rate": 0.0001,
      "loss": 1.8817,
      "step": 388
    },
    {
      "epoch": 0.09446333171442448,
      "grad_norm": 0.3364380896091461,
      "learning_rate": 0.0001,
      "loss": 1.8327,
      "step": 389
    },
    {
      "epoch": 0.0947061680427392,
      "grad_norm": 0.3223830461502075,
      "learning_rate": 0.0001,
      "loss": 1.7869,
      "step": 390
    },
    {
      "epoch": 0.09494900437105391,
      "grad_norm": 0.34865981340408325,
      "learning_rate": 0.0001,
      "loss": 1.7537,
      "step": 391
    },
    {
      "epoch": 0.09519184069936863,
      "grad_norm": 0.32730600237846375,
      "learning_rate": 0.0001,
      "loss": 1.7977,
      "step": 392
    },
    {
      "epoch": 0.09543467702768334,
      "grad_norm": 0.3286747336387634,
      "learning_rate": 0.0001,
      "loss": 1.6173,
      "step": 393
    },
    {
      "epoch": 0.09567751335599806,
      "grad_norm": 0.3481329679489136,
      "learning_rate": 0.0001,
      "loss": 1.837,
      "step": 394
    },
    {
      "epoch": 0.09592034968431278,
      "grad_norm": 0.3496640920639038,
      "learning_rate": 0.0001,
      "loss": 1.7007,
      "step": 395
    },
    {
      "epoch": 0.0961631860126275,
      "grad_norm": 0.35474106669425964,
      "learning_rate": 0.0001,
      "loss": 1.9123,
      "step": 396
    },
    {
      "epoch": 0.0964060223409422,
      "grad_norm": 0.3372008204460144,
      "learning_rate": 0.0001,
      "loss": 1.814,
      "step": 397
    },
    {
      "epoch": 0.09664885866925692,
      "grad_norm": 0.33549004793167114,
      "learning_rate": 0.0001,
      "loss": 1.8799,
      "step": 398
    },
    {
      "epoch": 0.09689169499757164,
      "grad_norm": 0.3251035213470459,
      "learning_rate": 0.0001,
      "loss": 1.6462,
      "step": 399
    },
    {
      "epoch": 0.09713453132588636,
      "grad_norm": 0.3387112021446228,
      "learning_rate": 0.0001,
      "loss": 1.8511,
      "step": 400
    },
    {
      "epoch": 0.09737736765420107,
      "grad_norm": 0.32067248225212097,
      "learning_rate": 0.0001,
      "loss": 1.922,
      "step": 401
    },
    {
      "epoch": 0.09762020398251578,
      "grad_norm": 0.3748277425765991,
      "learning_rate": 0.0001,
      "loss": 1.7758,
      "step": 402
    },
    {
      "epoch": 0.0978630403108305,
      "grad_norm": 0.3434027135372162,
      "learning_rate": 0.0001,
      "loss": 1.7802,
      "step": 403
    },
    {
      "epoch": 0.09810587663914522,
      "grad_norm": 0.31411343812942505,
      "learning_rate": 0.0001,
      "loss": 1.5578,
      "step": 404
    },
    {
      "epoch": 0.09834871296745994,
      "grad_norm": 0.3480266332626343,
      "learning_rate": 0.0001,
      "loss": 1.801,
      "step": 405
    },
    {
      "epoch": 0.09859154929577464,
      "grad_norm": 0.3487769663333893,
      "learning_rate": 0.0001,
      "loss": 1.8862,
      "step": 406
    },
    {
      "epoch": 0.09883438562408936,
      "grad_norm": 0.3437488079071045,
      "learning_rate": 0.0001,
      "loss": 1.9915,
      "step": 407
    },
    {
      "epoch": 0.09907722195240408,
      "grad_norm": 0.3372859060764313,
      "learning_rate": 0.0001,
      "loss": 1.8603,
      "step": 408
    },
    {
      "epoch": 0.0993200582807188,
      "grad_norm": 0.3287096321582794,
      "learning_rate": 0.0001,
      "loss": 1.7998,
      "step": 409
    },
    {
      "epoch": 0.09956289460903352,
      "grad_norm": 0.35017016530036926,
      "learning_rate": 0.0001,
      "loss": 1.9084,
      "step": 410
    },
    {
      "epoch": 0.09980573093734822,
      "grad_norm": 0.3446860909461975,
      "learning_rate": 0.0001,
      "loss": 1.9251,
      "step": 411
    },
    {
      "epoch": 0.10004856726566294,
      "grad_norm": 0.3437003493309021,
      "learning_rate": 0.0001,
      "loss": 1.9018,
      "step": 412
    },
    {
      "epoch": 0.10029140359397766,
      "grad_norm": 0.33558472990989685,
      "learning_rate": 0.0001,
      "loss": 1.7663,
      "step": 413
    },
    {
      "epoch": 0.10053423992229238,
      "grad_norm": 0.3409295082092285,
      "learning_rate": 0.0001,
      "loss": 1.9402,
      "step": 414
    },
    {
      "epoch": 0.1007770762506071,
      "grad_norm": 0.3394022285938263,
      "learning_rate": 0.0001,
      "loss": 1.8369,
      "step": 415
    },
    {
      "epoch": 0.1010199125789218,
      "grad_norm": 0.33328884840011597,
      "learning_rate": 0.0001,
      "loss": 1.8291,
      "step": 416
    },
    {
      "epoch": 0.10126274890723652,
      "grad_norm": 0.35661789774894714,
      "learning_rate": 0.0001,
      "loss": 1.8364,
      "step": 417
    },
    {
      "epoch": 0.10150558523555124,
      "grad_norm": 0.34863409399986267,
      "learning_rate": 0.0001,
      "loss": 1.9184,
      "step": 418
    },
    {
      "epoch": 0.10174842156386596,
      "grad_norm": 0.3370678424835205,
      "learning_rate": 0.0001,
      "loss": 1.728,
      "step": 419
    },
    {
      "epoch": 0.10199125789218066,
      "grad_norm": 0.3543989360332489,
      "learning_rate": 0.0001,
      "loss": 1.7346,
      "step": 420
    },
    {
      "epoch": 0.10223409422049538,
      "grad_norm": 0.3625810444355011,
      "learning_rate": 0.0001,
      "loss": 1.8968,
      "step": 421
    },
    {
      "epoch": 0.1024769305488101,
      "grad_norm": 0.31931373476982117,
      "learning_rate": 0.0001,
      "loss": 1.6817,
      "step": 422
    },
    {
      "epoch": 0.10271976687712482,
      "grad_norm": 0.33884134888648987,
      "learning_rate": 0.0001,
      "loss": 1.7368,
      "step": 423
    },
    {
      "epoch": 0.10296260320543954,
      "grad_norm": 0.32916051149368286,
      "learning_rate": 0.0001,
      "loss": 1.7307,
      "step": 424
    },
    {
      "epoch": 0.10320543953375425,
      "grad_norm": 0.3354704976081848,
      "learning_rate": 0.0001,
      "loss": 1.8176,
      "step": 425
    },
    {
      "epoch": 0.10344827586206896,
      "grad_norm": 0.32648178935050964,
      "learning_rate": 0.0001,
      "loss": 1.7642,
      "step": 426
    },
    {
      "epoch": 0.10369111219038368,
      "grad_norm": 0.3041088283061981,
      "learning_rate": 0.0001,
      "loss": 1.5416,
      "step": 427
    },
    {
      "epoch": 0.1039339485186984,
      "grad_norm": 0.34351372718811035,
      "learning_rate": 0.0001,
      "loss": 1.8235,
      "step": 428
    },
    {
      "epoch": 0.1041767848470131,
      "grad_norm": 0.33838582038879395,
      "learning_rate": 0.0001,
      "loss": 1.7042,
      "step": 429
    },
    {
      "epoch": 0.10441962117532783,
      "grad_norm": 0.34839949011802673,
      "learning_rate": 0.0001,
      "loss": 1.8394,
      "step": 430
    },
    {
      "epoch": 0.10466245750364254,
      "grad_norm": 0.32552745938301086,
      "learning_rate": 0.0001,
      "loss": 1.548,
      "step": 431
    },
    {
      "epoch": 0.10490529383195726,
      "grad_norm": 0.32809004187583923,
      "learning_rate": 0.0001,
      "loss": 1.861,
      "step": 432
    },
    {
      "epoch": 0.10514813016027198,
      "grad_norm": 0.35891664028167725,
      "learning_rate": 0.0001,
      "loss": 1.7159,
      "step": 433
    },
    {
      "epoch": 0.10539096648858669,
      "grad_norm": 0.3640296757221222,
      "learning_rate": 0.0001,
      "loss": 1.7317,
      "step": 434
    },
    {
      "epoch": 0.1056338028169014,
      "grad_norm": 0.3368436396121979,
      "learning_rate": 0.0001,
      "loss": 1.7596,
      "step": 435
    },
    {
      "epoch": 0.10587663914521613,
      "grad_norm": 0.31572410464286804,
      "learning_rate": 0.0001,
      "loss": 1.6735,
      "step": 436
    },
    {
      "epoch": 0.10611947547353084,
      "grad_norm": 0.341825395822525,
      "learning_rate": 0.0001,
      "loss": 1.8139,
      "step": 437
    },
    {
      "epoch": 0.10636231180184555,
      "grad_norm": 0.3547295928001404,
      "learning_rate": 0.0001,
      "loss": 1.8928,
      "step": 438
    },
    {
      "epoch": 0.10660514813016027,
      "grad_norm": 0.3295591175556183,
      "learning_rate": 0.0001,
      "loss": 1.6168,
      "step": 439
    },
    {
      "epoch": 0.10684798445847499,
      "grad_norm": 0.3370591104030609,
      "learning_rate": 0.0001,
      "loss": 1.9244,
      "step": 440
    },
    {
      "epoch": 0.1070908207867897,
      "grad_norm": 0.33818763494491577,
      "learning_rate": 0.0001,
      "loss": 1.7455,
      "step": 441
    },
    {
      "epoch": 0.10733365711510442,
      "grad_norm": 0.35193806886672974,
      "learning_rate": 0.0001,
      "loss": 1.7823,
      "step": 442
    },
    {
      "epoch": 0.10757649344341913,
      "grad_norm": 0.3400358557701111,
      "learning_rate": 0.0001,
      "loss": 1.8871,
      "step": 443
    },
    {
      "epoch": 0.10781932977173385,
      "grad_norm": 0.3593761622905731,
      "learning_rate": 0.0001,
      "loss": 1.8103,
      "step": 444
    },
    {
      "epoch": 0.10806216610004857,
      "grad_norm": 0.3300077021121979,
      "learning_rate": 0.0001,
      "loss": 1.786,
      "step": 445
    },
    {
      "epoch": 0.10830500242836329,
      "grad_norm": 0.3168260455131531,
      "learning_rate": 0.0001,
      "loss": 1.6937,
      "step": 446
    },
    {
      "epoch": 0.108547838756678,
      "grad_norm": 0.36123761534690857,
      "learning_rate": 0.0001,
      "loss": 1.9772,
      "step": 447
    },
    {
      "epoch": 0.10879067508499271,
      "grad_norm": 0.3371562659740448,
      "learning_rate": 0.0001,
      "loss": 1.8531,
      "step": 448
    },
    {
      "epoch": 0.10903351141330743,
      "grad_norm": 0.34795111417770386,
      "learning_rate": 0.0001,
      "loss": 1.7316,
      "step": 449
    },
    {
      "epoch": 0.10927634774162215,
      "grad_norm": 0.3352266252040863,
      "learning_rate": 0.0001,
      "loss": 1.799,
      "step": 450
    },
    {
      "epoch": 0.10951918406993687,
      "grad_norm": 0.31838861107826233,
      "learning_rate": 0.0001,
      "loss": 1.7578,
      "step": 451
    },
    {
      "epoch": 0.10976202039825157,
      "grad_norm": 0.3576313555240631,
      "learning_rate": 0.0001,
      "loss": 1.9083,
      "step": 452
    },
    {
      "epoch": 0.11000485672656629,
      "grad_norm": 0.3273816406726837,
      "learning_rate": 0.0001,
      "loss": 1.8002,
      "step": 453
    },
    {
      "epoch": 0.11024769305488101,
      "grad_norm": 0.3398890197277069,
      "learning_rate": 0.0001,
      "loss": 1.7914,
      "step": 454
    },
    {
      "epoch": 0.11049052938319573,
      "grad_norm": 0.3422471582889557,
      "learning_rate": 0.0001,
      "loss": 1.7997,
      "step": 455
    },
    {
      "epoch": 0.11073336571151045,
      "grad_norm": 0.3258894681930542,
      "learning_rate": 0.0001,
      "loss": 1.6239,
      "step": 456
    },
    {
      "epoch": 0.11097620203982515,
      "grad_norm": 0.3137560486793518,
      "learning_rate": 0.0001,
      "loss": 1.6991,
      "step": 457
    },
    {
      "epoch": 0.11121903836813987,
      "grad_norm": 0.3078104555606842,
      "learning_rate": 0.0001,
      "loss": 1.6844,
      "step": 458
    },
    {
      "epoch": 0.11146187469645459,
      "grad_norm": 0.3332054615020752,
      "learning_rate": 0.0001,
      "loss": 1.6455,
      "step": 459
    },
    {
      "epoch": 0.11170471102476931,
      "grad_norm": 0.3348306119441986,
      "learning_rate": 0.0001,
      "loss": 1.8087,
      "step": 460
    },
    {
      "epoch": 0.11194754735308401,
      "grad_norm": 0.3279060423374176,
      "learning_rate": 0.0001,
      "loss": 1.7843,
      "step": 461
    },
    {
      "epoch": 0.11219038368139873,
      "grad_norm": 0.330898255109787,
      "learning_rate": 0.0001,
      "loss": 1.7168,
      "step": 462
    },
    {
      "epoch": 0.11243322000971345,
      "grad_norm": 0.3327697813510895,
      "learning_rate": 0.0001,
      "loss": 1.7686,
      "step": 463
    },
    {
      "epoch": 0.11267605633802817,
      "grad_norm": 0.32823678851127625,
      "learning_rate": 0.0001,
      "loss": 1.6963,
      "step": 464
    },
    {
      "epoch": 0.11291889266634289,
      "grad_norm": 0.3418535590171814,
      "learning_rate": 0.0001,
      "loss": 1.794,
      "step": 465
    },
    {
      "epoch": 0.1131617289946576,
      "grad_norm": 0.34460851550102234,
      "learning_rate": 0.0001,
      "loss": 1.6965,
      "step": 466
    },
    {
      "epoch": 0.11340456532297231,
      "grad_norm": 0.32411473989486694,
      "learning_rate": 0.0001,
      "loss": 1.6109,
      "step": 467
    },
    {
      "epoch": 0.11364740165128703,
      "grad_norm": 0.3328358232975006,
      "learning_rate": 0.0001,
      "loss": 1.6898,
      "step": 468
    },
    {
      "epoch": 0.11389023797960175,
      "grad_norm": 0.35975319147109985,
      "learning_rate": 0.0001,
      "loss": 1.8466,
      "step": 469
    },
    {
      "epoch": 0.11413307430791647,
      "grad_norm": 0.32818377017974854,
      "learning_rate": 0.0001,
      "loss": 1.7855,
      "step": 470
    },
    {
      "epoch": 0.11437591063623118,
      "grad_norm": 0.29983869194984436,
      "learning_rate": 0.0001,
      "loss": 1.7306,
      "step": 471
    },
    {
      "epoch": 0.1146187469645459,
      "grad_norm": 0.3202539384365082,
      "learning_rate": 0.0001,
      "loss": 1.7312,
      "step": 472
    },
    {
      "epoch": 0.11486158329286061,
      "grad_norm": 0.3432933986186981,
      "learning_rate": 0.0001,
      "loss": 1.6656,
      "step": 473
    },
    {
      "epoch": 0.11510441962117533,
      "grad_norm": 0.32516351342201233,
      "learning_rate": 0.0001,
      "loss": 1.5996,
      "step": 474
    },
    {
      "epoch": 0.11534725594949004,
      "grad_norm": 0.3278566598892212,
      "learning_rate": 0.0001,
      "loss": 1.7985,
      "step": 475
    },
    {
      "epoch": 0.11559009227780476,
      "grad_norm": 0.32134944200515747,
      "learning_rate": 0.0001,
      "loss": 1.7525,
      "step": 476
    },
    {
      "epoch": 0.11583292860611948,
      "grad_norm": 0.3515169024467468,
      "learning_rate": 0.0001,
      "loss": 1.831,
      "step": 477
    },
    {
      "epoch": 0.1160757649344342,
      "grad_norm": 0.31824618577957153,
      "learning_rate": 0.0001,
      "loss": 1.6206,
      "step": 478
    },
    {
      "epoch": 0.11631860126274891,
      "grad_norm": 0.3707296550273895,
      "learning_rate": 0.0001,
      "loss": 1.9,
      "step": 479
    },
    {
      "epoch": 0.11656143759106362,
      "grad_norm": 0.33282679319381714,
      "learning_rate": 0.0001,
      "loss": 1.6946,
      "step": 480
    },
    {
      "epoch": 0.11680427391937834,
      "grad_norm": 0.33055081963539124,
      "learning_rate": 0.0001,
      "loss": 1.785,
      "step": 481
    },
    {
      "epoch": 0.11704711024769306,
      "grad_norm": 0.324970543384552,
      "learning_rate": 0.0001,
      "loss": 1.6858,
      "step": 482
    },
    {
      "epoch": 0.11728994657600778,
      "grad_norm": 0.3231464922428131,
      "learning_rate": 0.0001,
      "loss": 1.6393,
      "step": 483
    },
    {
      "epoch": 0.11753278290432248,
      "grad_norm": 0.34346282482147217,
      "learning_rate": 0.0001,
      "loss": 1.8243,
      "step": 484
    },
    {
      "epoch": 0.1177756192326372,
      "grad_norm": 0.3255658447742462,
      "learning_rate": 0.0001,
      "loss": 1.6712,
      "step": 485
    },
    {
      "epoch": 0.11801845556095192,
      "grad_norm": 0.3029419779777527,
      "learning_rate": 0.0001,
      "loss": 1.6564,
      "step": 486
    },
    {
      "epoch": 0.11826129188926664,
      "grad_norm": 0.3497994542121887,
      "learning_rate": 0.0001,
      "loss": 1.6322,
      "step": 487
    },
    {
      "epoch": 0.11850412821758136,
      "grad_norm": 0.3377457559108734,
      "learning_rate": 0.0001,
      "loss": 1.7269,
      "step": 488
    },
    {
      "epoch": 0.11874696454589606,
      "grad_norm": 0.3266313970088959,
      "learning_rate": 0.0001,
      "loss": 1.7791,
      "step": 489
    },
    {
      "epoch": 0.11898980087421078,
      "grad_norm": 0.352076917886734,
      "learning_rate": 0.0001,
      "loss": 1.9166,
      "step": 490
    },
    {
      "epoch": 0.1192326372025255,
      "grad_norm": 0.33107295632362366,
      "learning_rate": 0.0001,
      "loss": 1.8014,
      "step": 491
    },
    {
      "epoch": 0.11947547353084022,
      "grad_norm": 0.3481456935405731,
      "learning_rate": 0.0001,
      "loss": 1.8638,
      "step": 492
    },
    {
      "epoch": 0.11971830985915492,
      "grad_norm": 0.3252372741699219,
      "learning_rate": 0.0001,
      "loss": 1.7377,
      "step": 493
    },
    {
      "epoch": 0.11996114618746964,
      "grad_norm": 0.3195580542087555,
      "learning_rate": 0.0001,
      "loss": 1.7241,
      "step": 494
    },
    {
      "epoch": 0.12020398251578436,
      "grad_norm": 0.34002894163131714,
      "learning_rate": 0.0001,
      "loss": 1.6945,
      "step": 495
    },
    {
      "epoch": 0.12044681884409908,
      "grad_norm": 0.31778785586357117,
      "learning_rate": 0.0001,
      "loss": 1.6256,
      "step": 496
    },
    {
      "epoch": 0.1206896551724138,
      "grad_norm": 0.34070226550102234,
      "learning_rate": 0.0001,
      "loss": 1.8871,
      "step": 497
    },
    {
      "epoch": 0.1209324915007285,
      "grad_norm": 0.34664463996887207,
      "learning_rate": 0.0001,
      "loss": 1.9494,
      "step": 498
    },
    {
      "epoch": 0.12117532782904322,
      "grad_norm": 0.32989418506622314,
      "learning_rate": 0.0001,
      "loss": 1.7978,
      "step": 499
    },
    {
      "epoch": 0.12141816415735794,
      "grad_norm": 0.34744179248809814,
      "learning_rate": 0.0001,
      "loss": 1.8161,
      "step": 500
    },
    {
      "epoch": 0.12166100048567266,
      "grad_norm": 0.329497367143631,
      "learning_rate": 0.0001,
      "loss": 1.6066,
      "step": 501
    },
    {
      "epoch": 0.12190383681398738,
      "grad_norm": 0.3245505392551422,
      "learning_rate": 0.0001,
      "loss": 1.6534,
      "step": 502
    },
    {
      "epoch": 0.12214667314230208,
      "grad_norm": 0.35946330428123474,
      "learning_rate": 0.0001,
      "loss": 1.9062,
      "step": 503
    },
    {
      "epoch": 0.1223895094706168,
      "grad_norm": 0.36018314957618713,
      "learning_rate": 0.0001,
      "loss": 1.6777,
      "step": 504
    },
    {
      "epoch": 0.12263234579893152,
      "grad_norm": 0.35960716009140015,
      "learning_rate": 0.0001,
      "loss": 1.7391,
      "step": 505
    },
    {
      "epoch": 0.12287518212724624,
      "grad_norm": 0.32001397013664246,
      "learning_rate": 0.0001,
      "loss": 1.7721,
      "step": 506
    },
    {
      "epoch": 0.12311801845556095,
      "grad_norm": 0.31397107243537903,
      "learning_rate": 0.0001,
      "loss": 1.5642,
      "step": 507
    },
    {
      "epoch": 0.12336085478387566,
      "grad_norm": 0.358695924282074,
      "learning_rate": 0.0001,
      "loss": 1.8194,
      "step": 508
    },
    {
      "epoch": 0.12360369111219038,
      "grad_norm": 0.3431941270828247,
      "learning_rate": 0.0001,
      "loss": 1.8609,
      "step": 509
    },
    {
      "epoch": 0.1238465274405051,
      "grad_norm": 0.3243958652019501,
      "learning_rate": 0.0001,
      "loss": 1.844,
      "step": 510
    },
    {
      "epoch": 0.12408936376881982,
      "grad_norm": 0.3306693136692047,
      "learning_rate": 0.0001,
      "loss": 1.6731,
      "step": 511
    },
    {
      "epoch": 0.12433220009713453,
      "grad_norm": 0.31046029925346375,
      "learning_rate": 0.0001,
      "loss": 1.6751,
      "step": 512
    },
    {
      "epoch": 0.12457503642544925,
      "grad_norm": 0.3342955708503723,
      "learning_rate": 0.0001,
      "loss": 1.701,
      "step": 513
    },
    {
      "epoch": 0.12481787275376396,
      "grad_norm": 0.3202895224094391,
      "learning_rate": 0.0001,
      "loss": 1.6878,
      "step": 514
    },
    {
      "epoch": 0.12506070908207867,
      "grad_norm": 0.37030142545700073,
      "learning_rate": 0.0001,
      "loss": 1.9408,
      "step": 515
    },
    {
      "epoch": 0.1253035454103934,
      "grad_norm": 0.3309297561645508,
      "learning_rate": 0.0001,
      "loss": 1.7087,
      "step": 516
    },
    {
      "epoch": 0.1255463817387081,
      "grad_norm": 0.3122987151145935,
      "learning_rate": 0.0001,
      "loss": 1.7112,
      "step": 517
    },
    {
      "epoch": 0.12578921806702284,
      "grad_norm": 0.32288306951522827,
      "learning_rate": 0.0001,
      "loss": 1.6103,
      "step": 518
    },
    {
      "epoch": 0.12603205439533754,
      "grad_norm": 0.34724926948547363,
      "learning_rate": 0.0001,
      "loss": 1.9673,
      "step": 519
    },
    {
      "epoch": 0.12627489072365225,
      "grad_norm": 0.3623920679092407,
      "learning_rate": 0.0001,
      "loss": 1.8321,
      "step": 520
    },
    {
      "epoch": 0.12651772705196698,
      "grad_norm": 0.3387831151485443,
      "learning_rate": 0.0001,
      "loss": 1.8256,
      "step": 521
    },
    {
      "epoch": 0.1267605633802817,
      "grad_norm": 0.330636590719223,
      "learning_rate": 0.0001,
      "loss": 1.5437,
      "step": 522
    },
    {
      "epoch": 0.1270033997085964,
      "grad_norm": 0.36128905415534973,
      "learning_rate": 0.0001,
      "loss": 1.8782,
      "step": 523
    },
    {
      "epoch": 0.12724623603691113,
      "grad_norm": 0.33911311626434326,
      "learning_rate": 0.0001,
      "loss": 1.8102,
      "step": 524
    },
    {
      "epoch": 0.12748907236522583,
      "grad_norm": 0.3348335921764374,
      "learning_rate": 0.0001,
      "loss": 1.7712,
      "step": 525
    },
    {
      "epoch": 0.12773190869354056,
      "grad_norm": 0.3538147807121277,
      "learning_rate": 0.0001,
      "loss": 1.9288,
      "step": 526
    },
    {
      "epoch": 0.12797474502185527,
      "grad_norm": 0.3452605605125427,
      "learning_rate": 0.0001,
      "loss": 1.7673,
      "step": 527
    },
    {
      "epoch": 0.12821758135016997,
      "grad_norm": 0.31932786107063293,
      "learning_rate": 0.0001,
      "loss": 1.7969,
      "step": 528
    },
    {
      "epoch": 0.1284604176784847,
      "grad_norm": 0.34146761894226074,
      "learning_rate": 0.0001,
      "loss": 1.8897,
      "step": 529
    },
    {
      "epoch": 0.1287032540067994,
      "grad_norm": 0.333464652299881,
      "learning_rate": 0.0001,
      "loss": 1.8103,
      "step": 530
    },
    {
      "epoch": 0.12894609033511414,
      "grad_norm": 0.33564871549606323,
      "learning_rate": 0.0001,
      "loss": 1.7477,
      "step": 531
    },
    {
      "epoch": 0.12918892666342885,
      "grad_norm": 0.33123332262039185,
      "learning_rate": 0.0001,
      "loss": 1.8184,
      "step": 532
    },
    {
      "epoch": 0.12943176299174355,
      "grad_norm": 0.33338358998298645,
      "learning_rate": 0.0001,
      "loss": 1.7733,
      "step": 533
    },
    {
      "epoch": 0.1296745993200583,
      "grad_norm": 0.3367150127887726,
      "learning_rate": 0.0001,
      "loss": 1.8073,
      "step": 534
    },
    {
      "epoch": 0.129917435648373,
      "grad_norm": 0.3337629735469818,
      "learning_rate": 0.0001,
      "loss": 1.6832,
      "step": 535
    },
    {
      "epoch": 0.13016027197668772,
      "grad_norm": 0.33553630113601685,
      "learning_rate": 0.0001,
      "loss": 1.7301,
      "step": 536
    },
    {
      "epoch": 0.13040310830500243,
      "grad_norm": 0.32271620631217957,
      "learning_rate": 0.0001,
      "loss": 1.7196,
      "step": 537
    },
    {
      "epoch": 0.13064594463331713,
      "grad_norm": 0.3418119251728058,
      "learning_rate": 0.0001,
      "loss": 1.7169,
      "step": 538
    },
    {
      "epoch": 0.13088878096163187,
      "grad_norm": 0.3349496126174927,
      "learning_rate": 0.0001,
      "loss": 1.7522,
      "step": 539
    },
    {
      "epoch": 0.13113161728994657,
      "grad_norm": 0.35157325863838196,
      "learning_rate": 0.0001,
      "loss": 1.7772,
      "step": 540
    },
    {
      "epoch": 0.1313744536182613,
      "grad_norm": 0.3331843316555023,
      "learning_rate": 0.0001,
      "loss": 1.624,
      "step": 541
    },
    {
      "epoch": 0.131617289946576,
      "grad_norm": 0.34814342856407166,
      "learning_rate": 0.0001,
      "loss": 1.8239,
      "step": 542
    },
    {
      "epoch": 0.13186012627489072,
      "grad_norm": 0.326874703168869,
      "learning_rate": 0.0001,
      "loss": 1.5981,
      "step": 543
    },
    {
      "epoch": 0.13210296260320545,
      "grad_norm": 0.32875895500183105,
      "learning_rate": 0.0001,
      "loss": 1.6496,
      "step": 544
    },
    {
      "epoch": 0.13234579893152015,
      "grad_norm": 0.36154043674468994,
      "learning_rate": 0.0001,
      "loss": 1.8287,
      "step": 545
    },
    {
      "epoch": 0.13258863525983486,
      "grad_norm": 0.3320012092590332,
      "learning_rate": 0.0001,
      "loss": 1.8379,
      "step": 546
    },
    {
      "epoch": 0.1328314715881496,
      "grad_norm": 0.3641113340854645,
      "learning_rate": 0.0001,
      "loss": 1.8627,
      "step": 547
    },
    {
      "epoch": 0.1330743079164643,
      "grad_norm": 0.33732619881629944,
      "learning_rate": 0.0001,
      "loss": 1.755,
      "step": 548
    },
    {
      "epoch": 0.13331714424477903,
      "grad_norm": 0.32615527510643005,
      "learning_rate": 0.0001,
      "loss": 1.637,
      "step": 549
    },
    {
      "epoch": 0.13355998057309373,
      "grad_norm": 0.32284268736839294,
      "learning_rate": 0.0001,
      "loss": 1.8013,
      "step": 550
    },
    {
      "epoch": 0.13380281690140844,
      "grad_norm": 0.3423888683319092,
      "learning_rate": 0.0001,
      "loss": 1.7242,
      "step": 551
    },
    {
      "epoch": 0.13404565322972317,
      "grad_norm": 0.3580892086029053,
      "learning_rate": 0.0001,
      "loss": 1.9369,
      "step": 552
    },
    {
      "epoch": 0.13428848955803788,
      "grad_norm": 0.32854658365249634,
      "learning_rate": 0.0001,
      "loss": 1.6249,
      "step": 553
    },
    {
      "epoch": 0.1345313258863526,
      "grad_norm": 0.3406185507774353,
      "learning_rate": 0.0001,
      "loss": 1.9058,
      "step": 554
    },
    {
      "epoch": 0.13477416221466731,
      "grad_norm": 0.34051644802093506,
      "learning_rate": 0.0001,
      "loss": 1.8254,
      "step": 555
    },
    {
      "epoch": 0.13501699854298202,
      "grad_norm": 0.33127209544181824,
      "learning_rate": 0.0001,
      "loss": 1.6908,
      "step": 556
    },
    {
      "epoch": 0.13525983487129675,
      "grad_norm": 0.3405444622039795,
      "learning_rate": 0.0001,
      "loss": 1.9559,
      "step": 557
    },
    {
      "epoch": 0.13550267119961146,
      "grad_norm": 0.33380869030952454,
      "learning_rate": 0.0001,
      "loss": 1.8753,
      "step": 558
    },
    {
      "epoch": 0.1357455075279262,
      "grad_norm": 0.3485771119594574,
      "learning_rate": 0.0001,
      "loss": 1.9057,
      "step": 559
    },
    {
      "epoch": 0.1359883438562409,
      "grad_norm": 0.36507877707481384,
      "learning_rate": 0.0001,
      "loss": 1.8183,
      "step": 560
    },
    {
      "epoch": 0.1362311801845556,
      "grad_norm": 0.3692066967487335,
      "learning_rate": 0.0001,
      "loss": 1.7619,
      "step": 561
    },
    {
      "epoch": 0.13647401651287033,
      "grad_norm": 0.32849472761154175,
      "learning_rate": 0.0001,
      "loss": 1.5954,
      "step": 562
    },
    {
      "epoch": 0.13671685284118504,
      "grad_norm": 0.3511491119861603,
      "learning_rate": 0.0001,
      "loss": 1.7778,
      "step": 563
    },
    {
      "epoch": 0.13695968916949977,
      "grad_norm": 0.3083047866821289,
      "learning_rate": 0.0001,
      "loss": 1.5886,
      "step": 564
    },
    {
      "epoch": 0.13720252549781448,
      "grad_norm": 0.3148990571498871,
      "learning_rate": 0.0001,
      "loss": 1.6174,
      "step": 565
    },
    {
      "epoch": 0.13744536182612918,
      "grad_norm": 0.34316617250442505,
      "learning_rate": 0.0001,
      "loss": 1.7766,
      "step": 566
    },
    {
      "epoch": 0.1376881981544439,
      "grad_norm": 0.34978803992271423,
      "learning_rate": 0.0001,
      "loss": 1.8194,
      "step": 567
    },
    {
      "epoch": 0.13793103448275862,
      "grad_norm": 0.3379075229167938,
      "learning_rate": 0.0001,
      "loss": 1.8286,
      "step": 568
    },
    {
      "epoch": 0.13817387081107332,
      "grad_norm": 0.3451768755912781,
      "learning_rate": 0.0001,
      "loss": 1.7955,
      "step": 569
    },
    {
      "epoch": 0.13841670713938806,
      "grad_norm": 0.3590940535068512,
      "learning_rate": 0.0001,
      "loss": 1.87,
      "step": 570
    },
    {
      "epoch": 0.13865954346770276,
      "grad_norm": 0.3292856216430664,
      "learning_rate": 0.0001,
      "loss": 1.8424,
      "step": 571
    },
    {
      "epoch": 0.1389023797960175,
      "grad_norm": 0.3355911672115326,
      "learning_rate": 0.0001,
      "loss": 1.7929,
      "step": 572
    },
    {
      "epoch": 0.1391452161243322,
      "grad_norm": 0.33305859565734863,
      "learning_rate": 0.0001,
      "loss": 1.8061,
      "step": 573
    },
    {
      "epoch": 0.1393880524526469,
      "grad_norm": 0.3276538550853729,
      "learning_rate": 0.0001,
      "loss": 1.8168,
      "step": 574
    },
    {
      "epoch": 0.13963088878096164,
      "grad_norm": 0.32917794585227966,
      "learning_rate": 0.0001,
      "loss": 1.7679,
      "step": 575
    },
    {
      "epoch": 0.13987372510927634,
      "grad_norm": 0.32212531566619873,
      "learning_rate": 0.0001,
      "loss": 1.8593,
      "step": 576
    },
    {
      "epoch": 0.14011656143759108,
      "grad_norm": 0.3472840487957001,
      "learning_rate": 0.0001,
      "loss": 1.7646,
      "step": 577
    },
    {
      "epoch": 0.14035939776590578,
      "grad_norm": 0.3321049213409424,
      "learning_rate": 0.0001,
      "loss": 1.7385,
      "step": 578
    },
    {
      "epoch": 0.14060223409422049,
      "grad_norm": 0.316599041223526,
      "learning_rate": 0.0001,
      "loss": 1.7995,
      "step": 579
    },
    {
      "epoch": 0.14084507042253522,
      "grad_norm": 0.3480801284313202,
      "learning_rate": 0.0001,
      "loss": 1.7401,
      "step": 580
    },
    {
      "epoch": 0.14108790675084992,
      "grad_norm": 0.3251743018627167,
      "learning_rate": 0.0001,
      "loss": 1.6566,
      "step": 581
    },
    {
      "epoch": 0.14133074307916466,
      "grad_norm": 0.3258562982082367,
      "learning_rate": 0.0001,
      "loss": 1.7591,
      "step": 582
    },
    {
      "epoch": 0.14157357940747936,
      "grad_norm": 0.3337818682193756,
      "learning_rate": 0.0001,
      "loss": 1.7909,
      "step": 583
    },
    {
      "epoch": 0.14181641573579407,
      "grad_norm": 0.32916077971458435,
      "learning_rate": 0.0001,
      "loss": 1.768,
      "step": 584
    },
    {
      "epoch": 0.1420592520641088,
      "grad_norm": 0.3424016833305359,
      "learning_rate": 0.0001,
      "loss": 1.7864,
      "step": 585
    },
    {
      "epoch": 0.1423020883924235,
      "grad_norm": 0.3284737765789032,
      "learning_rate": 0.0001,
      "loss": 1.7983,
      "step": 586
    },
    {
      "epoch": 0.1425449247207382,
      "grad_norm": 0.3327319920063019,
      "learning_rate": 0.0001,
      "loss": 1.8206,
      "step": 587
    },
    {
      "epoch": 0.14278776104905294,
      "grad_norm": 0.34315967559814453,
      "learning_rate": 0.0001,
      "loss": 1.8132,
      "step": 588
    },
    {
      "epoch": 0.14303059737736765,
      "grad_norm": 0.31191959977149963,
      "learning_rate": 0.0001,
      "loss": 1.5604,
      "step": 589
    },
    {
      "epoch": 0.14327343370568238,
      "grad_norm": 0.35059061646461487,
      "learning_rate": 0.0001,
      "loss": 1.9002,
      "step": 590
    },
    {
      "epoch": 0.14351627003399708,
      "grad_norm": 0.3173618018627167,
      "learning_rate": 0.0001,
      "loss": 1.7253,
      "step": 591
    },
    {
      "epoch": 0.1437591063623118,
      "grad_norm": 0.3325091302394867,
      "learning_rate": 0.0001,
      "loss": 1.7917,
      "step": 592
    },
    {
      "epoch": 0.14400194269062652,
      "grad_norm": 0.3475317656993866,
      "learning_rate": 0.0001,
      "loss": 1.7804,
      "step": 593
    },
    {
      "epoch": 0.14424477901894123,
      "grad_norm": 0.34290921688079834,
      "learning_rate": 0.0001,
      "loss": 1.9093,
      "step": 594
    },
    {
      "epoch": 0.14448761534725596,
      "grad_norm": 0.32887551188468933,
      "learning_rate": 0.0001,
      "loss": 1.848,
      "step": 595
    },
    {
      "epoch": 0.14473045167557066,
      "grad_norm": 0.307198166847229,
      "learning_rate": 0.0001,
      "loss": 1.5295,
      "step": 596
    },
    {
      "epoch": 0.14497328800388537,
      "grad_norm": 0.31939375400543213,
      "learning_rate": 0.0001,
      "loss": 1.7991,
      "step": 597
    },
    {
      "epoch": 0.1452161243322001,
      "grad_norm": 0.3266233205795288,
      "learning_rate": 0.0001,
      "loss": 1.7458,
      "step": 598
    },
    {
      "epoch": 0.1454589606605148,
      "grad_norm": 0.3055732846260071,
      "learning_rate": 0.0001,
      "loss": 1.7317,
      "step": 599
    },
    {
      "epoch": 0.14570179698882954,
      "grad_norm": 0.31025564670562744,
      "learning_rate": 0.0001,
      "loss": 1.6654,
      "step": 600
    },
    {
      "epoch": 0.14594463331714425,
      "grad_norm": 0.34017419815063477,
      "learning_rate": 0.0001,
      "loss": 1.8332,
      "step": 601
    },
    {
      "epoch": 0.14618746964545895,
      "grad_norm": 0.32183635234832764,
      "learning_rate": 0.0001,
      "loss": 1.7224,
      "step": 602
    },
    {
      "epoch": 0.14643030597377368,
      "grad_norm": 0.35014277696609497,
      "learning_rate": 0.0001,
      "loss": 1.7805,
      "step": 603
    },
    {
      "epoch": 0.1466731423020884,
      "grad_norm": 0.35021066665649414,
      "learning_rate": 0.0001,
      "loss": 1.8568,
      "step": 604
    },
    {
      "epoch": 0.14691597863040312,
      "grad_norm": 0.3157992660999298,
      "learning_rate": 0.0001,
      "loss": 1.69,
      "step": 605
    },
    {
      "epoch": 0.14715881495871783,
      "grad_norm": 0.351871132850647,
      "learning_rate": 0.0001,
      "loss": 1.7699,
      "step": 606
    },
    {
      "epoch": 0.14740165128703253,
      "grad_norm": 0.3164587914943695,
      "learning_rate": 0.0001,
      "loss": 1.5622,
      "step": 607
    },
    {
      "epoch": 0.14764448761534726,
      "grad_norm": 0.35028359293937683,
      "learning_rate": 0.0001,
      "loss": 1.7516,
      "step": 608
    },
    {
      "epoch": 0.14788732394366197,
      "grad_norm": 0.3479732871055603,
      "learning_rate": 0.0001,
      "loss": 1.7743,
      "step": 609
    },
    {
      "epoch": 0.14813016027197667,
      "grad_norm": 0.3294665515422821,
      "learning_rate": 0.0001,
      "loss": 1.8481,
      "step": 610
    },
    {
      "epoch": 0.1483729966002914,
      "grad_norm": 0.319428026676178,
      "learning_rate": 0.0001,
      "loss": 1.8482,
      "step": 611
    },
    {
      "epoch": 0.1486158329286061,
      "grad_norm": 0.3437284827232361,
      "learning_rate": 0.0001,
      "loss": 1.8229,
      "step": 612
    },
    {
      "epoch": 0.14885866925692084,
      "grad_norm": 0.3444206118583679,
      "learning_rate": 0.0001,
      "loss": 1.958,
      "step": 613
    },
    {
      "epoch": 0.14910150558523555,
      "grad_norm": 0.3380897343158722,
      "learning_rate": 0.0001,
      "loss": 1.7921,
      "step": 614
    },
    {
      "epoch": 0.14934434191355025,
      "grad_norm": 0.33966124057769775,
      "learning_rate": 0.0001,
      "loss": 1.8157,
      "step": 615
    },
    {
      "epoch": 0.149587178241865,
      "grad_norm": 0.32820945978164673,
      "learning_rate": 0.0001,
      "loss": 1.805,
      "step": 616
    },
    {
      "epoch": 0.1498300145701797,
      "grad_norm": 0.31519755721092224,
      "learning_rate": 0.0001,
      "loss": 1.6555,
      "step": 617
    },
    {
      "epoch": 0.15007285089849443,
      "grad_norm": 0.3371162414550781,
      "learning_rate": 0.0001,
      "loss": 1.8202,
      "step": 618
    },
    {
      "epoch": 0.15031568722680913,
      "grad_norm": 0.33624976873397827,
      "learning_rate": 0.0001,
      "loss": 1.8528,
      "step": 619
    },
    {
      "epoch": 0.15055852355512384,
      "grad_norm": 0.3345712423324585,
      "learning_rate": 0.0001,
      "loss": 1.6205,
      "step": 620
    },
    {
      "epoch": 0.15080135988343857,
      "grad_norm": 0.32715481519699097,
      "learning_rate": 0.0001,
      "loss": 1.7635,
      "step": 621
    },
    {
      "epoch": 0.15104419621175327,
      "grad_norm": 0.35149484872817993,
      "learning_rate": 0.0001,
      "loss": 1.7891,
      "step": 622
    },
    {
      "epoch": 0.151287032540068,
      "grad_norm": 0.3212732672691345,
      "learning_rate": 0.0001,
      "loss": 1.7351,
      "step": 623
    },
    {
      "epoch": 0.1515298688683827,
      "grad_norm": 0.33642488718032837,
      "learning_rate": 0.0001,
      "loss": 1.8604,
      "step": 624
    },
    {
      "epoch": 0.15177270519669742,
      "grad_norm": 0.35052698850631714,
      "learning_rate": 0.0001,
      "loss": 1.9189,
      "step": 625
    },
    {
      "epoch": 0.15201554152501215,
      "grad_norm": 0.32550784945487976,
      "learning_rate": 0.0001,
      "loss": 1.5858,
      "step": 626
    },
    {
      "epoch": 0.15225837785332685,
      "grad_norm": 0.3275245428085327,
      "learning_rate": 0.0001,
      "loss": 1.6605,
      "step": 627
    },
    {
      "epoch": 0.1525012141816416,
      "grad_norm": 0.3372390568256378,
      "learning_rate": 0.0001,
      "loss": 1.7338,
      "step": 628
    },
    {
      "epoch": 0.1527440505099563,
      "grad_norm": 0.32684433460235596,
      "learning_rate": 0.0001,
      "loss": 1.6609,
      "step": 629
    },
    {
      "epoch": 0.152986886838271,
      "grad_norm": 0.3422621786594391,
      "learning_rate": 0.0001,
      "loss": 1.7199,
      "step": 630
    },
    {
      "epoch": 0.15322972316658573,
      "grad_norm": 0.3409489095211029,
      "learning_rate": 0.0001,
      "loss": 1.8106,
      "step": 631
    },
    {
      "epoch": 0.15347255949490043,
      "grad_norm": 0.3328961730003357,
      "learning_rate": 0.0001,
      "loss": 1.7863,
      "step": 632
    },
    {
      "epoch": 0.15371539582321514,
      "grad_norm": 0.32431545853614807,
      "learning_rate": 0.0001,
      "loss": 1.7251,
      "step": 633
    },
    {
      "epoch": 0.15395823215152987,
      "grad_norm": 0.3159632384777069,
      "learning_rate": 0.0001,
      "loss": 1.581,
      "step": 634
    },
    {
      "epoch": 0.15420106847984458,
      "grad_norm": 0.325759619474411,
      "learning_rate": 0.0001,
      "loss": 1.7943,
      "step": 635
    },
    {
      "epoch": 0.1544439048081593,
      "grad_norm": 0.34994059801101685,
      "learning_rate": 0.0001,
      "loss": 1.9051,
      "step": 636
    },
    {
      "epoch": 0.15468674113647402,
      "grad_norm": 0.33518439531326294,
      "learning_rate": 0.0001,
      "loss": 1.7541,
      "step": 637
    },
    {
      "epoch": 0.15492957746478872,
      "grad_norm": 0.32423776388168335,
      "learning_rate": 0.0001,
      "loss": 1.753,
      "step": 638
    },
    {
      "epoch": 0.15517241379310345,
      "grad_norm": 0.3157038688659668,
      "learning_rate": 0.0001,
      "loss": 1.7187,
      "step": 639
    },
    {
      "epoch": 0.15541525012141816,
      "grad_norm": 0.33567070960998535,
      "learning_rate": 0.0001,
      "loss": 1.6209,
      "step": 640
    },
    {
      "epoch": 0.1556580864497329,
      "grad_norm": 0.3170723617076874,
      "learning_rate": 0.0001,
      "loss": 1.7539,
      "step": 641
    },
    {
      "epoch": 0.1559009227780476,
      "grad_norm": 0.32920214533805847,
      "learning_rate": 0.0001,
      "loss": 1.7689,
      "step": 642
    },
    {
      "epoch": 0.1561437591063623,
      "grad_norm": 0.3828733265399933,
      "learning_rate": 0.0001,
      "loss": 1.7785,
      "step": 643
    },
    {
      "epoch": 0.15638659543467703,
      "grad_norm": 0.35701796412467957,
      "learning_rate": 0.0001,
      "loss": 1.7291,
      "step": 644
    },
    {
      "epoch": 0.15662943176299174,
      "grad_norm": 0.3379092812538147,
      "learning_rate": 0.0001,
      "loss": 1.9058,
      "step": 645
    },
    {
      "epoch": 0.15687226809130647,
      "grad_norm": 0.35852211713790894,
      "learning_rate": 0.0001,
      "loss": 1.979,
      "step": 646
    },
    {
      "epoch": 0.15711510441962118,
      "grad_norm": 0.344271183013916,
      "learning_rate": 0.0001,
      "loss": 1.7638,
      "step": 647
    },
    {
      "epoch": 0.15735794074793588,
      "grad_norm": 0.34935322403907776,
      "learning_rate": 0.0001,
      "loss": 1.8539,
      "step": 648
    },
    {
      "epoch": 0.15760077707625061,
      "grad_norm": 0.3357618451118469,
      "learning_rate": 0.0001,
      "loss": 1.9117,
      "step": 649
    },
    {
      "epoch": 0.15784361340456532,
      "grad_norm": 0.3315989673137665,
      "learning_rate": 0.0001,
      "loss": 1.7258,
      "step": 650
    },
    {
      "epoch": 0.15808644973288005,
      "grad_norm": 0.33593153953552246,
      "learning_rate": 0.0001,
      "loss": 1.734,
      "step": 651
    },
    {
      "epoch": 0.15832928606119476,
      "grad_norm": 0.352849543094635,
      "learning_rate": 0.0001,
      "loss": 1.9499,
      "step": 652
    },
    {
      "epoch": 0.15857212238950946,
      "grad_norm": 0.3458506762981415,
      "learning_rate": 0.0001,
      "loss": 1.7657,
      "step": 653
    },
    {
      "epoch": 0.1588149587178242,
      "grad_norm": 0.33314961194992065,
      "learning_rate": 0.0001,
      "loss": 1.8456,
      "step": 654
    },
    {
      "epoch": 0.1590577950461389,
      "grad_norm": 0.3287748098373413,
      "learning_rate": 0.0001,
      "loss": 1.7065,
      "step": 655
    },
    {
      "epoch": 0.1593006313744536,
      "grad_norm": 0.3667800724506378,
      "learning_rate": 0.0001,
      "loss": 1.8753,
      "step": 656
    },
    {
      "epoch": 0.15954346770276834,
      "grad_norm": 0.3313208520412445,
      "learning_rate": 0.0001,
      "loss": 1.7463,
      "step": 657
    },
    {
      "epoch": 0.15978630403108304,
      "grad_norm": 0.33500197529792786,
      "learning_rate": 0.0001,
      "loss": 1.8661,
      "step": 658
    },
    {
      "epoch": 0.16002914035939778,
      "grad_norm": 0.3511989414691925,
      "learning_rate": 0.0001,
      "loss": 1.8231,
      "step": 659
    },
    {
      "epoch": 0.16027197668771248,
      "grad_norm": 0.343047559261322,
      "learning_rate": 0.0001,
      "loss": 1.7863,
      "step": 660
    },
    {
      "epoch": 0.16051481301602719,
      "grad_norm": 0.32878488302230835,
      "learning_rate": 0.0001,
      "loss": 1.749,
      "step": 661
    },
    {
      "epoch": 0.16075764934434192,
      "grad_norm": 0.3422974646091461,
      "learning_rate": 0.0001,
      "loss": 1.8309,
      "step": 662
    },
    {
      "epoch": 0.16100048567265662,
      "grad_norm": 0.32340022921562195,
      "learning_rate": 0.0001,
      "loss": 1.7192,
      "step": 663
    },
    {
      "epoch": 0.16124332200097136,
      "grad_norm": 0.35579779744148254,
      "learning_rate": 0.0001,
      "loss": 1.7947,
      "step": 664
    },
    {
      "epoch": 0.16148615832928606,
      "grad_norm": 0.34640470147132874,
      "learning_rate": 0.0001,
      "loss": 1.7191,
      "step": 665
    },
    {
      "epoch": 0.16172899465760077,
      "grad_norm": 0.3360711932182312,
      "learning_rate": 0.0001,
      "loss": 1.7657,
      "step": 666
    },
    {
      "epoch": 0.1619718309859155,
      "grad_norm": 0.3053452968597412,
      "learning_rate": 0.0001,
      "loss": 1.5486,
      "step": 667
    },
    {
      "epoch": 0.1622146673142302,
      "grad_norm": 0.34008896350860596,
      "learning_rate": 0.0001,
      "loss": 1.7507,
      "step": 668
    },
    {
      "epoch": 0.16245750364254494,
      "grad_norm": 0.374798059463501,
      "learning_rate": 0.0001,
      "loss": 1.8685,
      "step": 669
    },
    {
      "epoch": 0.16270033997085964,
      "grad_norm": 0.32496216893196106,
      "learning_rate": 0.0001,
      "loss": 1.7575,
      "step": 670
    },
    {
      "epoch": 0.16294317629917435,
      "grad_norm": 0.3341841697692871,
      "learning_rate": 0.0001,
      "loss": 1.6965,
      "step": 671
    },
    {
      "epoch": 0.16318601262748908,
      "grad_norm": 0.33429092168807983,
      "learning_rate": 0.0001,
      "loss": 1.702,
      "step": 672
    },
    {
      "epoch": 0.16342884895580378,
      "grad_norm": 0.3410471975803375,
      "learning_rate": 0.0001,
      "loss": 1.6771,
      "step": 673
    },
    {
      "epoch": 0.16367168528411852,
      "grad_norm": 0.35043495893478394,
      "learning_rate": 0.0001,
      "loss": 1.8804,
      "step": 674
    },
    {
      "epoch": 0.16391452161243322,
      "grad_norm": 0.37732118368148804,
      "learning_rate": 0.0001,
      "loss": 2.0228,
      "step": 675
    },
    {
      "epoch": 0.16415735794074793,
      "grad_norm": 0.36240872740745544,
      "learning_rate": 0.0001,
      "loss": 1.8744,
      "step": 676
    },
    {
      "epoch": 0.16440019426906266,
      "grad_norm": 0.37208640575408936,
      "learning_rate": 0.0001,
      "loss": 1.7277,
      "step": 677
    },
    {
      "epoch": 0.16464303059737737,
      "grad_norm": 0.3265897035598755,
      "learning_rate": 0.0001,
      "loss": 1.7531,
      "step": 678
    },
    {
      "epoch": 0.16488586692569207,
      "grad_norm": 0.3237488865852356,
      "learning_rate": 0.0001,
      "loss": 1.8174,
      "step": 679
    },
    {
      "epoch": 0.1651287032540068,
      "grad_norm": 0.33787935972213745,
      "learning_rate": 0.0001,
      "loss": 1.7959,
      "step": 680
    },
    {
      "epoch": 0.1653715395823215,
      "grad_norm": 0.3404181897640228,
      "learning_rate": 0.0001,
      "loss": 1.7234,
      "step": 681
    },
    {
      "epoch": 0.16561437591063624,
      "grad_norm": 0.3305708169937134,
      "learning_rate": 0.0001,
      "loss": 1.768,
      "step": 682
    },
    {
      "epoch": 0.16585721223895095,
      "grad_norm": 0.305209755897522,
      "learning_rate": 0.0001,
      "loss": 1.6544,
      "step": 683
    },
    {
      "epoch": 0.16610004856726565,
      "grad_norm": 0.3397432267665863,
      "learning_rate": 0.0001,
      "loss": 1.6999,
      "step": 684
    },
    {
      "epoch": 0.16634288489558038,
      "grad_norm": 0.3438527286052704,
      "learning_rate": 0.0001,
      "loss": 1.8724,
      "step": 685
    },
    {
      "epoch": 0.1665857212238951,
      "grad_norm": 0.34619447588920593,
      "learning_rate": 0.0001,
      "loss": 1.8039,
      "step": 686
    },
    {
      "epoch": 0.16682855755220982,
      "grad_norm": 0.33126625418663025,
      "learning_rate": 0.0001,
      "loss": 1.6389,
      "step": 687
    },
    {
      "epoch": 0.16707139388052453,
      "grad_norm": 0.34118935465812683,
      "learning_rate": 0.0001,
      "loss": 1.8495,
      "step": 688
    },
    {
      "epoch": 0.16731423020883923,
      "grad_norm": 0.3436391353607178,
      "learning_rate": 0.0001,
      "loss": 1.8543,
      "step": 689
    },
    {
      "epoch": 0.16755706653715396,
      "grad_norm": 0.3279227912425995,
      "learning_rate": 0.0001,
      "loss": 1.8061,
      "step": 690
    },
    {
      "epoch": 0.16779990286546867,
      "grad_norm": 0.32945412397384644,
      "learning_rate": 0.0001,
      "loss": 1.7128,
      "step": 691
    },
    {
      "epoch": 0.1680427391937834,
      "grad_norm": 0.32346343994140625,
      "learning_rate": 0.0001,
      "loss": 1.7311,
      "step": 692
    },
    {
      "epoch": 0.1682855755220981,
      "grad_norm": 0.31251630187034607,
      "learning_rate": 0.0001,
      "loss": 1.4493,
      "step": 693
    },
    {
      "epoch": 0.1685284118504128,
      "grad_norm": 0.3643507957458496,
      "learning_rate": 0.0001,
      "loss": 1.9172,
      "step": 694
    },
    {
      "epoch": 0.16877124817872755,
      "grad_norm": 0.34988000988960266,
      "learning_rate": 0.0001,
      "loss": 1.7222,
      "step": 695
    },
    {
      "epoch": 0.16901408450704225,
      "grad_norm": 0.3611491322517395,
      "learning_rate": 0.0001,
      "loss": 1.8329,
      "step": 696
    },
    {
      "epoch": 0.16925692083535698,
      "grad_norm": 0.32318976521492004,
      "learning_rate": 0.0001,
      "loss": 1.6282,
      "step": 697
    },
    {
      "epoch": 0.1694997571636717,
      "grad_norm": 0.34611061215400696,
      "learning_rate": 0.0001,
      "loss": 1.756,
      "step": 698
    },
    {
      "epoch": 0.1697425934919864,
      "grad_norm": 0.33741113543510437,
      "learning_rate": 0.0001,
      "loss": 1.8066,
      "step": 699
    },
    {
      "epoch": 0.16998542982030113,
      "grad_norm": 0.3311440944671631,
      "learning_rate": 0.0001,
      "loss": 1.6861,
      "step": 700
    },
    {
      "epoch": 0.17022826614861583,
      "grad_norm": 0.3328585624694824,
      "learning_rate": 0.0001,
      "loss": 1.7464,
      "step": 701
    },
    {
      "epoch": 0.17047110247693054,
      "grad_norm": 0.32861262559890747,
      "learning_rate": 0.0001,
      "loss": 1.7738,
      "step": 702
    },
    {
      "epoch": 0.17071393880524527,
      "grad_norm": 0.3296881318092346,
      "learning_rate": 0.0001,
      "loss": 1.6663,
      "step": 703
    },
    {
      "epoch": 0.17095677513355997,
      "grad_norm": 0.3431016206741333,
      "learning_rate": 0.0001,
      "loss": 1.8015,
      "step": 704
    },
    {
      "epoch": 0.1711996114618747,
      "grad_norm": 0.32219427824020386,
      "learning_rate": 0.0001,
      "loss": 1.6279,
      "step": 705
    },
    {
      "epoch": 0.1714424477901894,
      "grad_norm": 0.32654130458831787,
      "learning_rate": 0.0001,
      "loss": 1.6762,
      "step": 706
    },
    {
      "epoch": 0.17168528411850412,
      "grad_norm": 0.33777257800102234,
      "learning_rate": 0.0001,
      "loss": 1.9124,
      "step": 707
    },
    {
      "epoch": 0.17192812044681885,
      "grad_norm": 0.3405590057373047,
      "learning_rate": 0.0001,
      "loss": 1.7402,
      "step": 708
    },
    {
      "epoch": 0.17217095677513355,
      "grad_norm": 0.36250537633895874,
      "learning_rate": 0.0001,
      "loss": 1.8756,
      "step": 709
    },
    {
      "epoch": 0.1724137931034483,
      "grad_norm": 0.33841243386268616,
      "learning_rate": 0.0001,
      "loss": 1.8326,
      "step": 710
    },
    {
      "epoch": 0.172656629431763,
      "grad_norm": 0.33011090755462646,
      "learning_rate": 0.0001,
      "loss": 1.7608,
      "step": 711
    },
    {
      "epoch": 0.1728994657600777,
      "grad_norm": 0.3336141109466553,
      "learning_rate": 0.0001,
      "loss": 1.7569,
      "step": 712
    },
    {
      "epoch": 0.17314230208839243,
      "grad_norm": 0.33510497212409973,
      "learning_rate": 0.0001,
      "loss": 1.6263,
      "step": 713
    },
    {
      "epoch": 0.17338513841670714,
      "grad_norm": 0.32303619384765625,
      "learning_rate": 0.0001,
      "loss": 1.7252,
      "step": 714
    },
    {
      "epoch": 0.17362797474502187,
      "grad_norm": 0.34422940015792847,
      "learning_rate": 0.0001,
      "loss": 1.8453,
      "step": 715
    },
    {
      "epoch": 0.17387081107333657,
      "grad_norm": 0.3252314329147339,
      "learning_rate": 0.0001,
      "loss": 1.7176,
      "step": 716
    },
    {
      "epoch": 0.17411364740165128,
      "grad_norm": 0.34678858518600464,
      "learning_rate": 0.0001,
      "loss": 1.798,
      "step": 717
    },
    {
      "epoch": 0.174356483729966,
      "grad_norm": 0.3389412760734558,
      "learning_rate": 0.0001,
      "loss": 1.8228,
      "step": 718
    },
    {
      "epoch": 0.17459932005828072,
      "grad_norm": 0.33434489369392395,
      "learning_rate": 0.0001,
      "loss": 1.809,
      "step": 719
    },
    {
      "epoch": 0.17484215638659542,
      "grad_norm": 0.31604427099227905,
      "learning_rate": 0.0001,
      "loss": 1.3931,
      "step": 720
    },
    {
      "epoch": 0.17508499271491015,
      "grad_norm": 0.3255927264690399,
      "learning_rate": 0.0001,
      "loss": 1.6887,
      "step": 721
    },
    {
      "epoch": 0.17532782904322486,
      "grad_norm": 0.35983723402023315,
      "learning_rate": 0.0001,
      "loss": 1.994,
      "step": 722
    },
    {
      "epoch": 0.1755706653715396,
      "grad_norm": 0.32813572883605957,
      "learning_rate": 0.0001,
      "loss": 1.6857,
      "step": 723
    },
    {
      "epoch": 0.1758135016998543,
      "grad_norm": 0.3567933440208435,
      "learning_rate": 0.0001,
      "loss": 1.887,
      "step": 724
    },
    {
      "epoch": 0.176056338028169,
      "grad_norm": 0.34803661704063416,
      "learning_rate": 0.0001,
      "loss": 1.9376,
      "step": 725
    },
    {
      "epoch": 0.17629917435648373,
      "grad_norm": 0.34169384837150574,
      "learning_rate": 0.0001,
      "loss": 1.7396,
      "step": 726
    },
    {
      "epoch": 0.17654201068479844,
      "grad_norm": 0.3188161551952362,
      "learning_rate": 0.0001,
      "loss": 1.6764,
      "step": 727
    },
    {
      "epoch": 0.17678484701311317,
      "grad_norm": 0.33624914288520813,
      "learning_rate": 0.0001,
      "loss": 1.8037,
      "step": 728
    },
    {
      "epoch": 0.17702768334142788,
      "grad_norm": 0.3575475811958313,
      "learning_rate": 0.0001,
      "loss": 1.8988,
      "step": 729
    },
    {
      "epoch": 0.17727051966974258,
      "grad_norm": 0.3442639708518982,
      "learning_rate": 0.0001,
      "loss": 1.6926,
      "step": 730
    },
    {
      "epoch": 0.17751335599805732,
      "grad_norm": 0.327396959066391,
      "learning_rate": 0.0001,
      "loss": 1.7721,
      "step": 731
    },
    {
      "epoch": 0.17775619232637202,
      "grad_norm": 0.3187220096588135,
      "learning_rate": 0.0001,
      "loss": 1.6799,
      "step": 732
    },
    {
      "epoch": 0.17799902865468675,
      "grad_norm": 0.3227187395095825,
      "learning_rate": 0.0001,
      "loss": 1.6265,
      "step": 733
    },
    {
      "epoch": 0.17824186498300146,
      "grad_norm": 0.3068956732749939,
      "learning_rate": 0.0001,
      "loss": 1.6813,
      "step": 734
    },
    {
      "epoch": 0.17848470131131616,
      "grad_norm": 0.3299641013145447,
      "learning_rate": 0.0001,
      "loss": 1.7025,
      "step": 735
    },
    {
      "epoch": 0.1787275376396309,
      "grad_norm": 0.34794217348098755,
      "learning_rate": 0.0001,
      "loss": 1.7394,
      "step": 736
    },
    {
      "epoch": 0.1789703739679456,
      "grad_norm": 0.361322283744812,
      "learning_rate": 0.0001,
      "loss": 1.9326,
      "step": 737
    },
    {
      "epoch": 0.17921321029626033,
      "grad_norm": 0.32667848467826843,
      "learning_rate": 0.0001,
      "loss": 1.7336,
      "step": 738
    },
    {
      "epoch": 0.17945604662457504,
      "grad_norm": 0.32748880982398987,
      "learning_rate": 0.0001,
      "loss": 1.7288,
      "step": 739
    },
    {
      "epoch": 0.17969888295288974,
      "grad_norm": 0.30917081236839294,
      "learning_rate": 0.0001,
      "loss": 1.5652,
      "step": 740
    },
    {
      "epoch": 0.17994171928120448,
      "grad_norm": 0.32722023129463196,
      "learning_rate": 0.0001,
      "loss": 1.8459,
      "step": 741
    },
    {
      "epoch": 0.18018455560951918,
      "grad_norm": 0.346816748380661,
      "learning_rate": 0.0001,
      "loss": 1.8984,
      "step": 742
    },
    {
      "epoch": 0.1804273919378339,
      "grad_norm": 0.32882779836654663,
      "learning_rate": 0.0001,
      "loss": 1.8174,
      "step": 743
    },
    {
      "epoch": 0.18067022826614862,
      "grad_norm": 0.3457486629486084,
      "learning_rate": 0.0001,
      "loss": 1.848,
      "step": 744
    },
    {
      "epoch": 0.18091306459446332,
      "grad_norm": 0.3878709673881531,
      "learning_rate": 0.0001,
      "loss": 1.9496,
      "step": 745
    },
    {
      "epoch": 0.18115590092277806,
      "grad_norm": 0.3353903889656067,
      "learning_rate": 0.0001,
      "loss": 1.868,
      "step": 746
    },
    {
      "epoch": 0.18139873725109276,
      "grad_norm": 0.3408859372138977,
      "learning_rate": 0.0001,
      "loss": 1.6315,
      "step": 747
    },
    {
      "epoch": 0.18164157357940747,
      "grad_norm": 0.34258410334587097,
      "learning_rate": 0.0001,
      "loss": 1.7475,
      "step": 748
    },
    {
      "epoch": 0.1818844099077222,
      "grad_norm": 0.3541860580444336,
      "learning_rate": 0.0001,
      "loss": 1.7701,
      "step": 749
    },
    {
      "epoch": 0.1821272462360369,
      "grad_norm": 0.3331660330295563,
      "learning_rate": 0.0001,
      "loss": 1.8313,
      "step": 750
    },
    {
      "epoch": 0.18237008256435164,
      "grad_norm": 0.33861926198005676,
      "learning_rate": 0.0001,
      "loss": 1.6838,
      "step": 751
    },
    {
      "epoch": 0.18261291889266634,
      "grad_norm": 0.3338991105556488,
      "learning_rate": 0.0001,
      "loss": 1.7838,
      "step": 752
    },
    {
      "epoch": 0.18285575522098105,
      "grad_norm": 0.342155784368515,
      "learning_rate": 0.0001,
      "loss": 1.8657,
      "step": 753
    },
    {
      "epoch": 0.18309859154929578,
      "grad_norm": 0.3422032594680786,
      "learning_rate": 0.0001,
      "loss": 1.5988,
      "step": 754
    },
    {
      "epoch": 0.18334142787761049,
      "grad_norm": 0.3455490469932556,
      "learning_rate": 0.0001,
      "loss": 1.858,
      "step": 755
    },
    {
      "epoch": 0.18358426420592522,
      "grad_norm": 0.3422670364379883,
      "learning_rate": 0.0001,
      "loss": 1.7285,
      "step": 756
    },
    {
      "epoch": 0.18382710053423992,
      "grad_norm": 0.3451416790485382,
      "learning_rate": 0.0001,
      "loss": 1.8276,
      "step": 757
    },
    {
      "epoch": 0.18406993686255463,
      "grad_norm": 0.3365596532821655,
      "learning_rate": 0.0001,
      "loss": 1.7377,
      "step": 758
    },
    {
      "epoch": 0.18431277319086936,
      "grad_norm": 0.3370806872844696,
      "learning_rate": 0.0001,
      "loss": 1.763,
      "step": 759
    },
    {
      "epoch": 0.18455560951918407,
      "grad_norm": 0.3182312846183777,
      "learning_rate": 0.0001,
      "loss": 1.6808,
      "step": 760
    },
    {
      "epoch": 0.1847984458474988,
      "grad_norm": 0.3098803460597992,
      "learning_rate": 0.0001,
      "loss": 1.5653,
      "step": 761
    },
    {
      "epoch": 0.1850412821758135,
      "grad_norm": 0.34688398241996765,
      "learning_rate": 0.0001,
      "loss": 1.8465,
      "step": 762
    },
    {
      "epoch": 0.1852841185041282,
      "grad_norm": 0.3338069021701813,
      "learning_rate": 0.0001,
      "loss": 1.5843,
      "step": 763
    },
    {
      "epoch": 0.18552695483244294,
      "grad_norm": 0.3415437340736389,
      "learning_rate": 0.0001,
      "loss": 1.6496,
      "step": 764
    },
    {
      "epoch": 0.18576979116075765,
      "grad_norm": 0.33651795983314514,
      "learning_rate": 0.0001,
      "loss": 1.7268,
      "step": 765
    },
    {
      "epoch": 0.18601262748907235,
      "grad_norm": 0.31493595242500305,
      "learning_rate": 0.0001,
      "loss": 1.6041,
      "step": 766
    },
    {
      "epoch": 0.18625546381738708,
      "grad_norm": 0.35351628065109253,
      "learning_rate": 0.0001,
      "loss": 1.9251,
      "step": 767
    },
    {
      "epoch": 0.1864983001457018,
      "grad_norm": 0.3334658741950989,
      "learning_rate": 0.0001,
      "loss": 1.6445,
      "step": 768
    },
    {
      "epoch": 0.18674113647401652,
      "grad_norm": 0.33200061321258545,
      "learning_rate": 0.0001,
      "loss": 1.8252,
      "step": 769
    },
    {
      "epoch": 0.18698397280233123,
      "grad_norm": 0.32640960812568665,
      "learning_rate": 0.0001,
      "loss": 1.7991,
      "step": 770
    },
    {
      "epoch": 0.18722680913064593,
      "grad_norm": 0.3249363601207733,
      "learning_rate": 0.0001,
      "loss": 1.8871,
      "step": 771
    },
    {
      "epoch": 0.18746964545896067,
      "grad_norm": 0.33192652463912964,
      "learning_rate": 0.0001,
      "loss": 1.7089,
      "step": 772
    },
    {
      "epoch": 0.18771248178727537,
      "grad_norm": 0.3350042700767517,
      "learning_rate": 0.0001,
      "loss": 1.7861,
      "step": 773
    },
    {
      "epoch": 0.1879553181155901,
      "grad_norm": 0.34054598212242126,
      "learning_rate": 0.0001,
      "loss": 1.6476,
      "step": 774
    },
    {
      "epoch": 0.1881981544439048,
      "grad_norm": 0.3331514894962311,
      "learning_rate": 0.0001,
      "loss": 1.6713,
      "step": 775
    },
    {
      "epoch": 0.1884409907722195,
      "grad_norm": 0.34734639525413513,
      "learning_rate": 0.0001,
      "loss": 1.6703,
      "step": 776
    },
    {
      "epoch": 0.18868382710053425,
      "grad_norm": 0.3209059238433838,
      "learning_rate": 0.0001,
      "loss": 1.6813,
      "step": 777
    },
    {
      "epoch": 0.18892666342884895,
      "grad_norm": 0.33769097924232483,
      "learning_rate": 0.0001,
      "loss": 1.6145,
      "step": 778
    },
    {
      "epoch": 0.18916949975716368,
      "grad_norm": 0.3314269483089447,
      "learning_rate": 0.0001,
      "loss": 1.891,
      "step": 779
    },
    {
      "epoch": 0.1894123360854784,
      "grad_norm": 0.3271227478981018,
      "learning_rate": 0.0001,
      "loss": 1.7157,
      "step": 780
    },
    {
      "epoch": 0.1896551724137931,
      "grad_norm": 0.33472874760627747,
      "learning_rate": 0.0001,
      "loss": 1.8272,
      "step": 781
    },
    {
      "epoch": 0.18989800874210783,
      "grad_norm": 0.3343239426612854,
      "learning_rate": 0.0001,
      "loss": 1.8228,
      "step": 782
    },
    {
      "epoch": 0.19014084507042253,
      "grad_norm": 0.32455193996429443,
      "learning_rate": 0.0001,
      "loss": 1.8218,
      "step": 783
    },
    {
      "epoch": 0.19038368139873726,
      "grad_norm": 0.3229330778121948,
      "learning_rate": 0.0001,
      "loss": 1.6711,
      "step": 784
    },
    {
      "epoch": 0.19062651772705197,
      "grad_norm": 0.32274356484413147,
      "learning_rate": 0.0001,
      "loss": 1.6694,
      "step": 785
    },
    {
      "epoch": 0.19086935405536667,
      "grad_norm": 0.34533029794692993,
      "learning_rate": 0.0001,
      "loss": 1.9482,
      "step": 786
    },
    {
      "epoch": 0.1911121903836814,
      "grad_norm": 0.33702874183654785,
      "learning_rate": 0.0001,
      "loss": 1.6929,
      "step": 787
    },
    {
      "epoch": 0.1913550267119961,
      "grad_norm": 0.3319770097732544,
      "learning_rate": 0.0001,
      "loss": 1.74,
      "step": 788
    },
    {
      "epoch": 0.19159786304031082,
      "grad_norm": 0.35082754492759705,
      "learning_rate": 0.0001,
      "loss": 1.8279,
      "step": 789
    },
    {
      "epoch": 0.19184069936862555,
      "grad_norm": 0.36771881580352783,
      "learning_rate": 0.0001,
      "loss": 1.9524,
      "step": 790
    },
    {
      "epoch": 0.19208353569694026,
      "grad_norm": 0.3544480502605438,
      "learning_rate": 0.0001,
      "loss": 1.7698,
      "step": 791
    },
    {
      "epoch": 0.192326372025255,
      "grad_norm": 0.3526579737663269,
      "learning_rate": 0.0001,
      "loss": 1.7691,
      "step": 792
    },
    {
      "epoch": 0.1925692083535697,
      "grad_norm": 0.34120672941207886,
      "learning_rate": 0.0001,
      "loss": 1.5646,
      "step": 793
    },
    {
      "epoch": 0.1928120446818844,
      "grad_norm": 0.31328442692756653,
      "learning_rate": 0.0001,
      "loss": 1.6347,
      "step": 794
    },
    {
      "epoch": 0.19305488101019913,
      "grad_norm": 0.3437521159648895,
      "learning_rate": 0.0001,
      "loss": 1.7563,
      "step": 795
    },
    {
      "epoch": 0.19329771733851384,
      "grad_norm": 0.35751140117645264,
      "learning_rate": 0.0001,
      "loss": 1.8915,
      "step": 796
    },
    {
      "epoch": 0.19354055366682857,
      "grad_norm": 0.3591373562812805,
      "learning_rate": 0.0001,
      "loss": 1.9232,
      "step": 797
    },
    {
      "epoch": 0.19378338999514327,
      "grad_norm": 0.3579035997390747,
      "learning_rate": 0.0001,
      "loss": 1.7625,
      "step": 798
    },
    {
      "epoch": 0.19402622632345798,
      "grad_norm": 0.32526394724845886,
      "learning_rate": 0.0001,
      "loss": 1.6856,
      "step": 799
    },
    {
      "epoch": 0.1942690626517727,
      "grad_norm": 0.3462944030761719,
      "learning_rate": 0.0001,
      "loss": 1.9237,
      "step": 800
    },
    {
      "epoch": 0.19451189898008742,
      "grad_norm": 0.3462536633014679,
      "learning_rate": 0.0001,
      "loss": 1.7484,
      "step": 801
    },
    {
      "epoch": 0.19475473530840215,
      "grad_norm": 0.343323290348053,
      "learning_rate": 0.0001,
      "loss": 1.8536,
      "step": 802
    },
    {
      "epoch": 0.19499757163671685,
      "grad_norm": 0.3693440556526184,
      "learning_rate": 0.0001,
      "loss": 1.931,
      "step": 803
    },
    {
      "epoch": 0.19524040796503156,
      "grad_norm": 0.3456905782222748,
      "learning_rate": 0.0001,
      "loss": 1.8112,
      "step": 804
    },
    {
      "epoch": 0.1954832442933463,
      "grad_norm": 0.3245755136013031,
      "learning_rate": 0.0001,
      "loss": 1.8371,
      "step": 805
    },
    {
      "epoch": 0.195726080621661,
      "grad_norm": 0.3217308223247528,
      "learning_rate": 0.0001,
      "loss": 1.6225,
      "step": 806
    },
    {
      "epoch": 0.19596891694997573,
      "grad_norm": 0.3111560642719269,
      "learning_rate": 0.0001,
      "loss": 1.6079,
      "step": 807
    },
    {
      "epoch": 0.19621175327829043,
      "grad_norm": 0.33693355321884155,
      "learning_rate": 0.0001,
      "loss": 1.5574,
      "step": 808
    },
    {
      "epoch": 0.19645458960660514,
      "grad_norm": 0.3521810472011566,
      "learning_rate": 0.0001,
      "loss": 1.7933,
      "step": 809
    },
    {
      "epoch": 0.19669742593491987,
      "grad_norm": 0.3548601269721985,
      "learning_rate": 0.0001,
      "loss": 1.8576,
      "step": 810
    },
    {
      "epoch": 0.19694026226323458,
      "grad_norm": 0.33542320132255554,
      "learning_rate": 0.0001,
      "loss": 1.8073,
      "step": 811
    },
    {
      "epoch": 0.19718309859154928,
      "grad_norm": 0.3425523638725281,
      "learning_rate": 0.0001,
      "loss": 1.7489,
      "step": 812
    },
    {
      "epoch": 0.19742593491986402,
      "grad_norm": 0.34044787287712097,
      "learning_rate": 0.0001,
      "loss": 1.8246,
      "step": 813
    },
    {
      "epoch": 0.19766877124817872,
      "grad_norm": 0.33918073773384094,
      "learning_rate": 0.0001,
      "loss": 1.8108,
      "step": 814
    },
    {
      "epoch": 0.19791160757649345,
      "grad_norm": 0.35202568769454956,
      "learning_rate": 0.0001,
      "loss": 1.6453,
      "step": 815
    },
    {
      "epoch": 0.19815444390480816,
      "grad_norm": 0.33459073305130005,
      "learning_rate": 0.0001,
      "loss": 1.6348,
      "step": 816
    },
    {
      "epoch": 0.19839728023312286,
      "grad_norm": 0.3267429769039154,
      "learning_rate": 0.0001,
      "loss": 1.7359,
      "step": 817
    },
    {
      "epoch": 0.1986401165614376,
      "grad_norm": 0.3398960530757904,
      "learning_rate": 0.0001,
      "loss": 1.9159,
      "step": 818
    },
    {
      "epoch": 0.1988829528897523,
      "grad_norm": 0.35579246282577515,
      "learning_rate": 0.0001,
      "loss": 1.8049,
      "step": 819
    },
    {
      "epoch": 0.19912578921806703,
      "grad_norm": 0.34332728385925293,
      "learning_rate": 0.0001,
      "loss": 1.7352,
      "step": 820
    },
    {
      "epoch": 0.19936862554638174,
      "grad_norm": 0.3549124598503113,
      "learning_rate": 0.0001,
      "loss": 1.844,
      "step": 821
    },
    {
      "epoch": 0.19961146187469644,
      "grad_norm": 0.35614240169525146,
      "learning_rate": 0.0001,
      "loss": 1.768,
      "step": 822
    },
    {
      "epoch": 0.19985429820301118,
      "grad_norm": 0.32354140281677246,
      "learning_rate": 0.0001,
      "loss": 1.7093,
      "step": 823
    },
    {
      "epoch": 0.20009713453132588,
      "grad_norm": 0.3101747930049896,
      "learning_rate": 0.0001,
      "loss": 1.604,
      "step": 824
    },
    {
      "epoch": 0.20033997085964061,
      "grad_norm": 0.340231329202652,
      "learning_rate": 0.0001,
      "loss": 1.7756,
      "step": 825
    },
    {
      "epoch": 0.20058280718795532,
      "grad_norm": 0.3541657030582428,
      "learning_rate": 0.0001,
      "loss": 1.9198,
      "step": 826
    },
    {
      "epoch": 0.20082564351627002,
      "grad_norm": 0.34510529041290283,
      "learning_rate": 0.0001,
      "loss": 1.7255,
      "step": 827
    },
    {
      "epoch": 0.20106847984458476,
      "grad_norm": 0.31408369541168213,
      "learning_rate": 0.0001,
      "loss": 1.6342,
      "step": 828
    },
    {
      "epoch": 0.20131131617289946,
      "grad_norm": 0.33251187205314636,
      "learning_rate": 0.0001,
      "loss": 1.6463,
      "step": 829
    },
    {
      "epoch": 0.2015541525012142,
      "grad_norm": 0.368576318025589,
      "learning_rate": 0.0001,
      "loss": 1.8156,
      "step": 830
    },
    {
      "epoch": 0.2017969888295289,
      "grad_norm": 0.3493247926235199,
      "learning_rate": 0.0001,
      "loss": 1.8595,
      "step": 831
    },
    {
      "epoch": 0.2020398251578436,
      "grad_norm": 0.3393223285675049,
      "learning_rate": 0.0001,
      "loss": 1.7177,
      "step": 832
    },
    {
      "epoch": 0.20228266148615834,
      "grad_norm": 0.33512645959854126,
      "learning_rate": 0.0001,
      "loss": 1.7787,
      "step": 833
    },
    {
      "epoch": 0.20252549781447304,
      "grad_norm": 0.3425748646259308,
      "learning_rate": 0.0001,
      "loss": 1.7336,
      "step": 834
    },
    {
      "epoch": 0.20276833414278775,
      "grad_norm": 0.3310282826423645,
      "learning_rate": 0.0001,
      "loss": 1.67,
      "step": 835
    },
    {
      "epoch": 0.20301117047110248,
      "grad_norm": 0.34667834639549255,
      "learning_rate": 0.0001,
      "loss": 1.7295,
      "step": 836
    },
    {
      "epoch": 0.2032540067994172,
      "grad_norm": 0.3334478437900543,
      "learning_rate": 0.0001,
      "loss": 1.7254,
      "step": 837
    },
    {
      "epoch": 0.20349684312773192,
      "grad_norm": 0.3700467646121979,
      "learning_rate": 0.0001,
      "loss": 1.7012,
      "step": 838
    },
    {
      "epoch": 0.20373967945604662,
      "grad_norm": 0.36770233511924744,
      "learning_rate": 0.0001,
      "loss": 1.8803,
      "step": 839
    },
    {
      "epoch": 0.20398251578436133,
      "grad_norm": 0.38123154640197754,
      "learning_rate": 0.0001,
      "loss": 1.8847,
      "step": 840
    },
    {
      "epoch": 0.20422535211267606,
      "grad_norm": 0.35769400000572205,
      "learning_rate": 0.0001,
      "loss": 1.8712,
      "step": 841
    },
    {
      "epoch": 0.20446818844099077,
      "grad_norm": 0.3599579632282257,
      "learning_rate": 0.0001,
      "loss": 1.7323,
      "step": 842
    },
    {
      "epoch": 0.2047110247693055,
      "grad_norm": 0.3479319214820862,
      "learning_rate": 0.0001,
      "loss": 1.6642,
      "step": 843
    },
    {
      "epoch": 0.2049538610976202,
      "grad_norm": 0.33099737763404846,
      "learning_rate": 0.0001,
      "loss": 1.6858,
      "step": 844
    },
    {
      "epoch": 0.2051966974259349,
      "grad_norm": 0.3572336733341217,
      "learning_rate": 0.0001,
      "loss": 1.7512,
      "step": 845
    },
    {
      "epoch": 0.20543953375424964,
      "grad_norm": 0.36039644479751587,
      "learning_rate": 0.0001,
      "loss": 1.6811,
      "step": 846
    },
    {
      "epoch": 0.20568237008256435,
      "grad_norm": 0.33296865224838257,
      "learning_rate": 0.0001,
      "loss": 1.7238,
      "step": 847
    },
    {
      "epoch": 0.20592520641087908,
      "grad_norm": 0.32468274235725403,
      "learning_rate": 0.0001,
      "loss": 1.7885,
      "step": 848
    },
    {
      "epoch": 0.20616804273919379,
      "grad_norm": 0.34729287028312683,
      "learning_rate": 0.0001,
      "loss": 1.7283,
      "step": 849
    },
    {
      "epoch": 0.2064108790675085,
      "grad_norm": 0.32223427295684814,
      "learning_rate": 0.0001,
      "loss": 1.8602,
      "step": 850
    },
    {
      "epoch": 0.20665371539582322,
      "grad_norm": 0.33634892106056213,
      "learning_rate": 0.0001,
      "loss": 1.762,
      "step": 851
    },
    {
      "epoch": 0.20689655172413793,
      "grad_norm": 0.3353589177131653,
      "learning_rate": 0.0001,
      "loss": 1.7168,
      "step": 852
    },
    {
      "epoch": 0.20713938805245263,
      "grad_norm": 0.3339422047138214,
      "learning_rate": 0.0001,
      "loss": 1.6908,
      "step": 853
    },
    {
      "epoch": 0.20738222438076737,
      "grad_norm": 0.3292504847049713,
      "learning_rate": 0.0001,
      "loss": 1.6529,
      "step": 854
    },
    {
      "epoch": 0.20762506070908207,
      "grad_norm": 0.33124348521232605,
      "learning_rate": 0.0001,
      "loss": 1.666,
      "step": 855
    },
    {
      "epoch": 0.2078678970373968,
      "grad_norm": 0.3586646616458893,
      "learning_rate": 0.0001,
      "loss": 1.8378,
      "step": 856
    },
    {
      "epoch": 0.2081107333657115,
      "grad_norm": 0.3264204263687134,
      "learning_rate": 0.0001,
      "loss": 1.649,
      "step": 857
    },
    {
      "epoch": 0.2083535696940262,
      "grad_norm": 0.3210090100765228,
      "learning_rate": 0.0001,
      "loss": 1.7189,
      "step": 858
    },
    {
      "epoch": 0.20859640602234095,
      "grad_norm": 0.33976536989212036,
      "learning_rate": 0.0001,
      "loss": 1.8222,
      "step": 859
    },
    {
      "epoch": 0.20883924235065565,
      "grad_norm": 0.3441348373889923,
      "learning_rate": 0.0001,
      "loss": 1.834,
      "step": 860
    },
    {
      "epoch": 0.20908207867897038,
      "grad_norm": 0.356228768825531,
      "learning_rate": 0.0001,
      "loss": 1.7859,
      "step": 861
    },
    {
      "epoch": 0.2093249150072851,
      "grad_norm": 0.3454718589782715,
      "learning_rate": 0.0001,
      "loss": 1.7108,
      "step": 862
    },
    {
      "epoch": 0.2095677513355998,
      "grad_norm": 0.3440695106983185,
      "learning_rate": 0.0001,
      "loss": 1.757,
      "step": 863
    },
    {
      "epoch": 0.20981058766391453,
      "grad_norm": 0.368131548166275,
      "learning_rate": 0.0001,
      "loss": 1.8494,
      "step": 864
    },
    {
      "epoch": 0.21005342399222923,
      "grad_norm": 0.33047816157341003,
      "learning_rate": 0.0001,
      "loss": 1.7694,
      "step": 865
    },
    {
      "epoch": 0.21029626032054397,
      "grad_norm": 0.4003385603427887,
      "learning_rate": 0.0001,
      "loss": 1.7922,
      "step": 866
    },
    {
      "epoch": 0.21053909664885867,
      "grad_norm": 0.3227761387825012,
      "learning_rate": 0.0001,
      "loss": 1.6887,
      "step": 867
    },
    {
      "epoch": 0.21078193297717338,
      "grad_norm": 0.3393837511539459,
      "learning_rate": 0.0001,
      "loss": 1.7138,
      "step": 868
    },
    {
      "epoch": 0.2110247693054881,
      "grad_norm": 0.33335569500923157,
      "learning_rate": 0.0001,
      "loss": 1.7631,
      "step": 869
    },
    {
      "epoch": 0.2112676056338028,
      "grad_norm": 0.3336202800273895,
      "learning_rate": 0.0001,
      "loss": 1.6655,
      "step": 870
    },
    {
      "epoch": 0.21151044196211755,
      "grad_norm": 0.3336716592311859,
      "learning_rate": 0.0001,
      "loss": 1.6685,
      "step": 871
    },
    {
      "epoch": 0.21175327829043225,
      "grad_norm": 0.34575212001800537,
      "learning_rate": 0.0001,
      "loss": 1.8437,
      "step": 872
    },
    {
      "epoch": 0.21199611461874696,
      "grad_norm": 0.3509778678417206,
      "learning_rate": 0.0001,
      "loss": 1.674,
      "step": 873
    },
    {
      "epoch": 0.2122389509470617,
      "grad_norm": 0.32758817076683044,
      "learning_rate": 0.0001,
      "loss": 1.6476,
      "step": 874
    },
    {
      "epoch": 0.2124817872753764,
      "grad_norm": 0.3272348940372467,
      "learning_rate": 0.0001,
      "loss": 1.5082,
      "step": 875
    },
    {
      "epoch": 0.2127246236036911,
      "grad_norm": 0.3567332625389099,
      "learning_rate": 0.0001,
      "loss": 1.8242,
      "step": 876
    },
    {
      "epoch": 0.21296745993200583,
      "grad_norm": 0.36553505063056946,
      "learning_rate": 0.0001,
      "loss": 1.7268,
      "step": 877
    },
    {
      "epoch": 0.21321029626032054,
      "grad_norm": 0.3332120180130005,
      "learning_rate": 0.0001,
      "loss": 1.7061,
      "step": 878
    },
    {
      "epoch": 0.21345313258863527,
      "grad_norm": 0.3399730920791626,
      "learning_rate": 0.0001,
      "loss": 1.8343,
      "step": 879
    },
    {
      "epoch": 0.21369596891694997,
      "grad_norm": 0.346131831407547,
      "learning_rate": 0.0001,
      "loss": 1.7754,
      "step": 880
    },
    {
      "epoch": 0.21393880524526468,
      "grad_norm": 0.35501691699028015,
      "learning_rate": 0.0001,
      "loss": 1.7037,
      "step": 881
    },
    {
      "epoch": 0.2141816415735794,
      "grad_norm": 0.3296474516391754,
      "learning_rate": 0.0001,
      "loss": 1.7668,
      "step": 882
    },
    {
      "epoch": 0.21442447790189412,
      "grad_norm": 0.34373652935028076,
      "learning_rate": 0.0001,
      "loss": 1.9377,
      "step": 883
    },
    {
      "epoch": 0.21466731423020885,
      "grad_norm": 0.36046624183654785,
      "learning_rate": 0.0001,
      "loss": 1.6914,
      "step": 884
    },
    {
      "epoch": 0.21491015055852355,
      "grad_norm": 0.3706079125404358,
      "learning_rate": 0.0001,
      "loss": 1.835,
      "step": 885
    },
    {
      "epoch": 0.21515298688683826,
      "grad_norm": 0.3273414969444275,
      "learning_rate": 0.0001,
      "loss": 1.5998,
      "step": 886
    },
    {
      "epoch": 0.215395823215153,
      "grad_norm": 0.3641795814037323,
      "learning_rate": 0.0001,
      "loss": 1.7866,
      "step": 887
    },
    {
      "epoch": 0.2156386595434677,
      "grad_norm": 0.3411569595336914,
      "learning_rate": 0.0001,
      "loss": 1.5151,
      "step": 888
    },
    {
      "epoch": 0.21588149587178243,
      "grad_norm": 0.3366343080997467,
      "learning_rate": 0.0001,
      "loss": 1.8079,
      "step": 889
    },
    {
      "epoch": 0.21612433220009714,
      "grad_norm": 0.3335423171520233,
      "learning_rate": 0.0001,
      "loss": 1.7633,
      "step": 890
    },
    {
      "epoch": 0.21636716852841184,
      "grad_norm": 0.3207727074623108,
      "learning_rate": 0.0001,
      "loss": 1.5874,
      "step": 891
    },
    {
      "epoch": 0.21661000485672657,
      "grad_norm": 0.3799630403518677,
      "learning_rate": 0.0001,
      "loss": 1.6903,
      "step": 892
    },
    {
      "epoch": 0.21685284118504128,
      "grad_norm": 0.3363555669784546,
      "learning_rate": 0.0001,
      "loss": 1.7002,
      "step": 893
    },
    {
      "epoch": 0.217095677513356,
      "grad_norm": 0.3288339376449585,
      "learning_rate": 0.0001,
      "loss": 1.7877,
      "step": 894
    },
    {
      "epoch": 0.21733851384167072,
      "grad_norm": 0.3369339406490326,
      "learning_rate": 0.0001,
      "loss": 1.6823,
      "step": 895
    },
    {
      "epoch": 0.21758135016998542,
      "grad_norm": 0.3292815387248993,
      "learning_rate": 0.0001,
      "loss": 1.6129,
      "step": 896
    },
    {
      "epoch": 0.21782418649830015,
      "grad_norm": 0.3598136901855469,
      "learning_rate": 0.0001,
      "loss": 1.9099,
      "step": 897
    },
    {
      "epoch": 0.21806702282661486,
      "grad_norm": 0.3366025984287262,
      "learning_rate": 0.0001,
      "loss": 1.7813,
      "step": 898
    },
    {
      "epoch": 0.21830985915492956,
      "grad_norm": 0.33517763018608093,
      "learning_rate": 0.0001,
      "loss": 1.7882,
      "step": 899
    },
    {
      "epoch": 0.2185526954832443,
      "grad_norm": 0.36312445998191833,
      "learning_rate": 0.0001,
      "loss": 1.7446,
      "step": 900
    },
    {
      "epoch": 0.218795531811559,
      "grad_norm": 0.32939261198043823,
      "learning_rate": 0.0001,
      "loss": 1.7117,
      "step": 901
    },
    {
      "epoch": 0.21903836813987373,
      "grad_norm": 0.32495957612991333,
      "learning_rate": 0.0001,
      "loss": 1.5321,
      "step": 902
    },
    {
      "epoch": 0.21928120446818844,
      "grad_norm": 0.32330915331840515,
      "learning_rate": 0.0001,
      "loss": 1.7465,
      "step": 903
    },
    {
      "epoch": 0.21952404079650314,
      "grad_norm": 0.3653425872325897,
      "learning_rate": 0.0001,
      "loss": 1.9813,
      "step": 904
    },
    {
      "epoch": 0.21976687712481788,
      "grad_norm": 0.3383382260799408,
      "learning_rate": 0.0001,
      "loss": 1.706,
      "step": 905
    },
    {
      "epoch": 0.22000971345313258,
      "grad_norm": 0.31804466247558594,
      "learning_rate": 0.0001,
      "loss": 1.752,
      "step": 906
    },
    {
      "epoch": 0.22025254978144732,
      "grad_norm": 0.3489072322845459,
      "learning_rate": 0.0001,
      "loss": 1.9311,
      "step": 907
    },
    {
      "epoch": 0.22049538610976202,
      "grad_norm": 0.34647881984710693,
      "learning_rate": 0.0001,
      "loss": 1.9223,
      "step": 908
    },
    {
      "epoch": 0.22073822243807673,
      "grad_norm": 0.35893869400024414,
      "learning_rate": 0.0001,
      "loss": 1.7646,
      "step": 909
    },
    {
      "epoch": 0.22098105876639146,
      "grad_norm": 0.34356799721717834,
      "learning_rate": 0.0001,
      "loss": 1.7663,
      "step": 910
    },
    {
      "epoch": 0.22122389509470616,
      "grad_norm": 0.3178933262825012,
      "learning_rate": 0.0001,
      "loss": 1.5849,
      "step": 911
    },
    {
      "epoch": 0.2214667314230209,
      "grad_norm": 0.3426609933376312,
      "learning_rate": 0.0001,
      "loss": 1.7133,
      "step": 912
    },
    {
      "epoch": 0.2217095677513356,
      "grad_norm": 0.3565383851528168,
      "learning_rate": 0.0001,
      "loss": 1.7553,
      "step": 913
    },
    {
      "epoch": 0.2219524040796503,
      "grad_norm": 0.3379584550857544,
      "learning_rate": 0.0001,
      "loss": 1.6644,
      "step": 914
    },
    {
      "epoch": 0.22219524040796504,
      "grad_norm": 0.33079883456230164,
      "learning_rate": 0.0001,
      "loss": 1.609,
      "step": 915
    },
    {
      "epoch": 0.22243807673627974,
      "grad_norm": 0.32773423194885254,
      "learning_rate": 0.0001,
      "loss": 1.7287,
      "step": 916
    },
    {
      "epoch": 0.22268091306459448,
      "grad_norm": 0.3350023329257965,
      "learning_rate": 0.0001,
      "loss": 1.7385,
      "step": 917
    },
    {
      "epoch": 0.22292374939290918,
      "grad_norm": 0.34006428718566895,
      "learning_rate": 0.0001,
      "loss": 1.6834,
      "step": 918
    },
    {
      "epoch": 0.2231665857212239,
      "grad_norm": 0.33220747113227844,
      "learning_rate": 0.0001,
      "loss": 1.6562,
      "step": 919
    },
    {
      "epoch": 0.22340942204953862,
      "grad_norm": 0.3405970335006714,
      "learning_rate": 0.0001,
      "loss": 1.7838,
      "step": 920
    },
    {
      "epoch": 0.22365225837785332,
      "grad_norm": 0.32862991094589233,
      "learning_rate": 0.0001,
      "loss": 1.5924,
      "step": 921
    },
    {
      "epoch": 0.22389509470616803,
      "grad_norm": 0.33819305896759033,
      "learning_rate": 0.0001,
      "loss": 1.7711,
      "step": 922
    },
    {
      "epoch": 0.22413793103448276,
      "grad_norm": 0.3736564815044403,
      "learning_rate": 0.0001,
      "loss": 1.808,
      "step": 923
    },
    {
      "epoch": 0.22438076736279747,
      "grad_norm": 0.37374764680862427,
      "learning_rate": 0.0001,
      "loss": 1.9576,
      "step": 924
    },
    {
      "epoch": 0.2246236036911122,
      "grad_norm": 0.35444918274879456,
      "learning_rate": 0.0001,
      "loss": 1.896,
      "step": 925
    },
    {
      "epoch": 0.2248664400194269,
      "grad_norm": 0.34131696820259094,
      "learning_rate": 0.0001,
      "loss": 1.7908,
      "step": 926
    },
    {
      "epoch": 0.2251092763477416,
      "grad_norm": 0.3766651451587677,
      "learning_rate": 0.0001,
      "loss": 2.0439,
      "step": 927
    },
    {
      "epoch": 0.22535211267605634,
      "grad_norm": 0.343290239572525,
      "learning_rate": 0.0001,
      "loss": 1.8129,
      "step": 928
    },
    {
      "epoch": 0.22559494900437105,
      "grad_norm": 0.32844051718711853,
      "learning_rate": 0.0001,
      "loss": 1.6883,
      "step": 929
    },
    {
      "epoch": 0.22583778533268578,
      "grad_norm": 0.3543758988380432,
      "learning_rate": 0.0001,
      "loss": 1.8604,
      "step": 930
    },
    {
      "epoch": 0.22608062166100049,
      "grad_norm": 0.3468215763568878,
      "learning_rate": 0.0001,
      "loss": 1.8882,
      "step": 931
    },
    {
      "epoch": 0.2263234579893152,
      "grad_norm": 0.3537542521953583,
      "learning_rate": 0.0001,
      "loss": 1.8822,
      "step": 932
    },
    {
      "epoch": 0.22656629431762992,
      "grad_norm": 0.3319475054740906,
      "learning_rate": 0.0001,
      "loss": 1.6875,
      "step": 933
    },
    {
      "epoch": 0.22680913064594463,
      "grad_norm": 0.31763777136802673,
      "learning_rate": 0.0001,
      "loss": 1.7159,
      "step": 934
    },
    {
      "epoch": 0.22705196697425936,
      "grad_norm": 0.34334397315979004,
      "learning_rate": 0.0001,
      "loss": 1.6944,
      "step": 935
    },
    {
      "epoch": 0.22729480330257407,
      "grad_norm": 0.3653299808502197,
      "learning_rate": 0.0001,
      "loss": 1.8486,
      "step": 936
    },
    {
      "epoch": 0.22753763963088877,
      "grad_norm": 0.3533567786216736,
      "learning_rate": 0.0001,
      "loss": 1.8319,
      "step": 937
    },
    {
      "epoch": 0.2277804759592035,
      "grad_norm": 0.3710135221481323,
      "learning_rate": 0.0001,
      "loss": 1.7908,
      "step": 938
    },
    {
      "epoch": 0.2280233122875182,
      "grad_norm": 0.3512992262840271,
      "learning_rate": 0.0001,
      "loss": 1.891,
      "step": 939
    },
    {
      "epoch": 0.22826614861583294,
      "grad_norm": 0.34045422077178955,
      "learning_rate": 0.0001,
      "loss": 1.8874,
      "step": 940
    },
    {
      "epoch": 0.22850898494414765,
      "grad_norm": 0.30804723501205444,
      "learning_rate": 0.0001,
      "loss": 1.5742,
      "step": 941
    },
    {
      "epoch": 0.22875182127246235,
      "grad_norm": 0.3212561011314392,
      "learning_rate": 0.0001,
      "loss": 1.682,
      "step": 942
    },
    {
      "epoch": 0.22899465760077709,
      "grad_norm": 0.30441156029701233,
      "learning_rate": 0.0001,
      "loss": 1.6751,
      "step": 943
    },
    {
      "epoch": 0.2292374939290918,
      "grad_norm": 0.3184438645839691,
      "learning_rate": 0.0001,
      "loss": 1.6578,
      "step": 944
    },
    {
      "epoch": 0.2294803302574065,
      "grad_norm": 0.35863617062568665,
      "learning_rate": 0.0001,
      "loss": 1.7332,
      "step": 945
    },
    {
      "epoch": 0.22972316658572123,
      "grad_norm": 0.3359723687171936,
      "learning_rate": 0.0001,
      "loss": 1.7791,
      "step": 946
    },
    {
      "epoch": 0.22996600291403593,
      "grad_norm": 0.32070714235305786,
      "learning_rate": 0.0001,
      "loss": 1.6934,
      "step": 947
    },
    {
      "epoch": 0.23020883924235067,
      "grad_norm": 0.34434184432029724,
      "learning_rate": 0.0001,
      "loss": 1.7682,
      "step": 948
    },
    {
      "epoch": 0.23045167557066537,
      "grad_norm": 0.3323972225189209,
      "learning_rate": 0.0001,
      "loss": 1.8066,
      "step": 949
    },
    {
      "epoch": 0.23069451189898008,
      "grad_norm": 0.34174755215644836,
      "learning_rate": 0.0001,
      "loss": 1.6917,
      "step": 950
    },
    {
      "epoch": 0.2309373482272948,
      "grad_norm": 0.34504199028015137,
      "learning_rate": 0.0001,
      "loss": 1.9412,
      "step": 951
    },
    {
      "epoch": 0.2311801845556095,
      "grad_norm": 0.36647096276283264,
      "learning_rate": 0.0001,
      "loss": 1.8285,
      "step": 952
    },
    {
      "epoch": 0.23142302088392425,
      "grad_norm": 0.34093159437179565,
      "learning_rate": 0.0001,
      "loss": 1.8101,
      "step": 953
    },
    {
      "epoch": 0.23166585721223895,
      "grad_norm": 0.33534058928489685,
      "learning_rate": 0.0001,
      "loss": 1.6945,
      "step": 954
    },
    {
      "epoch": 0.23190869354055366,
      "grad_norm": 0.3567858934402466,
      "learning_rate": 0.0001,
      "loss": 1.9115,
      "step": 955
    },
    {
      "epoch": 0.2321515298688684,
      "grad_norm": 0.3529800474643707,
      "learning_rate": 0.0001,
      "loss": 1.9022,
      "step": 956
    },
    {
      "epoch": 0.2323943661971831,
      "grad_norm": 0.345431923866272,
      "learning_rate": 0.0001,
      "loss": 1.7276,
      "step": 957
    },
    {
      "epoch": 0.23263720252549783,
      "grad_norm": 0.3385120630264282,
      "learning_rate": 0.0001,
      "loss": 1.8453,
      "step": 958
    },
    {
      "epoch": 0.23288003885381253,
      "grad_norm": 0.3333473205566406,
      "learning_rate": 0.0001,
      "loss": 1.702,
      "step": 959
    },
    {
      "epoch": 0.23312287518212724,
      "grad_norm": 0.33618485927581787,
      "learning_rate": 0.0001,
      "loss": 1.6041,
      "step": 960
    },
    {
      "epoch": 0.23336571151044197,
      "grad_norm": 0.33134040236473083,
      "learning_rate": 0.0001,
      "loss": 1.681,
      "step": 961
    },
    {
      "epoch": 0.23360854783875667,
      "grad_norm": 0.3481343984603882,
      "learning_rate": 0.0001,
      "loss": 1.7838,
      "step": 962
    },
    {
      "epoch": 0.2338513841670714,
      "grad_norm": 0.3489163815975189,
      "learning_rate": 0.0001,
      "loss": 1.6594,
      "step": 963
    },
    {
      "epoch": 0.2340942204953861,
      "grad_norm": 0.3307543992996216,
      "learning_rate": 0.0001,
      "loss": 1.5912,
      "step": 964
    },
    {
      "epoch": 0.23433705682370082,
      "grad_norm": 0.3228107690811157,
      "learning_rate": 0.0001,
      "loss": 1.6931,
      "step": 965
    },
    {
      "epoch": 0.23457989315201555,
      "grad_norm": 0.3382891118526459,
      "learning_rate": 0.0001,
      "loss": 1.7528,
      "step": 966
    },
    {
      "epoch": 0.23482272948033026,
      "grad_norm": 0.325954794883728,
      "learning_rate": 0.0001,
      "loss": 1.7212,
      "step": 967
    },
    {
      "epoch": 0.23506556580864496,
      "grad_norm": 0.3358007073402405,
      "learning_rate": 0.0001,
      "loss": 1.8042,
      "step": 968
    },
    {
      "epoch": 0.2353084021369597,
      "grad_norm": 0.32503238320350647,
      "learning_rate": 0.0001,
      "loss": 1.6813,
      "step": 969
    },
    {
      "epoch": 0.2355512384652744,
      "grad_norm": 0.32884976267814636,
      "learning_rate": 0.0001,
      "loss": 1.6773,
      "step": 970
    },
    {
      "epoch": 0.23579407479358913,
      "grad_norm": 0.3466799259185791,
      "learning_rate": 0.0001,
      "loss": 1.8319,
      "step": 971
    },
    {
      "epoch": 0.23603691112190384,
      "grad_norm": 0.36012017726898193,
      "learning_rate": 0.0001,
      "loss": 1.8824,
      "step": 972
    },
    {
      "epoch": 0.23627974745021854,
      "grad_norm": 0.3252657353878021,
      "learning_rate": 0.0001,
      "loss": 1.7845,
      "step": 973
    },
    {
      "epoch": 0.23652258377853327,
      "grad_norm": 0.36254772543907166,
      "learning_rate": 0.0001,
      "loss": 1.776,
      "step": 974
    },
    {
      "epoch": 0.23676542010684798,
      "grad_norm": 0.369875431060791,
      "learning_rate": 0.0001,
      "loss": 1.7605,
      "step": 975
    },
    {
      "epoch": 0.2370082564351627,
      "grad_norm": 0.36949196457862854,
      "learning_rate": 0.0001,
      "loss": 1.9715,
      "step": 976
    },
    {
      "epoch": 0.23725109276347742,
      "grad_norm": 0.35166341066360474,
      "learning_rate": 0.0001,
      "loss": 1.8055,
      "step": 977
    },
    {
      "epoch": 0.23749392909179212,
      "grad_norm": 0.364952027797699,
      "learning_rate": 0.0001,
      "loss": 1.8131,
      "step": 978
    },
    {
      "epoch": 0.23773676542010685,
      "grad_norm": 0.3355357348918915,
      "learning_rate": 0.0001,
      "loss": 1.8336,
      "step": 979
    },
    {
      "epoch": 0.23797960174842156,
      "grad_norm": 0.34965017437934875,
      "learning_rate": 0.0001,
      "loss": 1.6352,
      "step": 980
    },
    {
      "epoch": 0.2382224380767363,
      "grad_norm": 0.350437194108963,
      "learning_rate": 0.0001,
      "loss": 1.6983,
      "step": 981
    },
    {
      "epoch": 0.238465274405051,
      "grad_norm": 0.32104507088661194,
      "learning_rate": 0.0001,
      "loss": 1.6894,
      "step": 982
    },
    {
      "epoch": 0.2387081107333657,
      "grad_norm": 0.32047441601753235,
      "learning_rate": 0.0001,
      "loss": 1.4182,
      "step": 983
    },
    {
      "epoch": 0.23895094706168044,
      "grad_norm": 0.3423304855823517,
      "learning_rate": 0.0001,
      "loss": 1.7036,
      "step": 984
    },
    {
      "epoch": 0.23919378338999514,
      "grad_norm": 0.3440588712692261,
      "learning_rate": 0.0001,
      "loss": 1.8234,
      "step": 985
    },
    {
      "epoch": 0.23943661971830985,
      "grad_norm": 0.32514244318008423,
      "learning_rate": 0.0001,
      "loss": 1.7542,
      "step": 986
    },
    {
      "epoch": 0.23967945604662458,
      "grad_norm": 0.32740339636802673,
      "learning_rate": 0.0001,
      "loss": 1.572,
      "step": 987
    },
    {
      "epoch": 0.23992229237493928,
      "grad_norm": 0.4369216561317444,
      "learning_rate": 0.0001,
      "loss": 1.7198,
      "step": 988
    },
    {
      "epoch": 0.24016512870325402,
      "grad_norm": 0.35954874753952026,
      "learning_rate": 0.0001,
      "loss": 1.899,
      "step": 989
    },
    {
      "epoch": 0.24040796503156872,
      "grad_norm": 0.3517872393131256,
      "learning_rate": 0.0001,
      "loss": 1.8624,
      "step": 990
    },
    {
      "epoch": 0.24065080135988343,
      "grad_norm": 0.36518657207489014,
      "learning_rate": 0.0001,
      "loss": 1.9199,
      "step": 991
    },
    {
      "epoch": 0.24089363768819816,
      "grad_norm": 0.3197648525238037,
      "learning_rate": 0.0001,
      "loss": 1.5309,
      "step": 992
    },
    {
      "epoch": 0.24113647401651286,
      "grad_norm": 0.35127437114715576,
      "learning_rate": 0.0001,
      "loss": 1.7003,
      "step": 993
    },
    {
      "epoch": 0.2413793103448276,
      "grad_norm": 0.355819433927536,
      "learning_rate": 0.0001,
      "loss": 1.8029,
      "step": 994
    },
    {
      "epoch": 0.2416221466731423,
      "grad_norm": 0.350644588470459,
      "learning_rate": 0.0001,
      "loss": 1.8459,
      "step": 995
    },
    {
      "epoch": 0.241864983001457,
      "grad_norm": 0.3549037575721741,
      "learning_rate": 0.0001,
      "loss": 1.6761,
      "step": 996
    },
    {
      "epoch": 0.24210781932977174,
      "grad_norm": 0.335479199886322,
      "learning_rate": 0.0001,
      "loss": 1.6987,
      "step": 997
    },
    {
      "epoch": 0.24235065565808644,
      "grad_norm": 0.3091221749782562,
      "learning_rate": 0.0001,
      "loss": 1.7735,
      "step": 998
    },
    {
      "epoch": 0.24259349198640118,
      "grad_norm": 0.3324226140975952,
      "learning_rate": 0.0001,
      "loss": 1.7726,
      "step": 999
    },
    {
      "epoch": 0.24283632831471588,
      "grad_norm": 0.35145866870880127,
      "learning_rate": 0.0001,
      "loss": 1.8383,
      "step": 1000
    },
    {
      "epoch": 0.2430791646430306,
      "grad_norm": 0.6738609671592712,
      "learning_rate": 0.0001,
      "loss": 1.7831,
      "step": 1001
    },
    {
      "epoch": 0.24332200097134532,
      "grad_norm": 0.34129729866981506,
      "learning_rate": 0.0001,
      "loss": 1.6946,
      "step": 1002
    },
    {
      "epoch": 0.24356483729966003,
      "grad_norm": 0.34623393416404724,
      "learning_rate": 0.0001,
      "loss": 1.8287,
      "step": 1003
    },
    {
      "epoch": 0.24380767362797476,
      "grad_norm": 0.32859230041503906,
      "learning_rate": 0.0001,
      "loss": 1.6113,
      "step": 1004
    },
    {
      "epoch": 0.24405050995628946,
      "grad_norm": 0.3451049327850342,
      "learning_rate": 0.0001,
      "loss": 1.8371,
      "step": 1005
    },
    {
      "epoch": 0.24429334628460417,
      "grad_norm": 0.3516858220100403,
      "learning_rate": 0.0001,
      "loss": 1.7367,
      "step": 1006
    },
    {
      "epoch": 0.2445361826129189,
      "grad_norm": 0.35251277685165405,
      "learning_rate": 0.0001,
      "loss": 1.8181,
      "step": 1007
    },
    {
      "epoch": 0.2447790189412336,
      "grad_norm": 0.3534775972366333,
      "learning_rate": 0.0001,
      "loss": 1.779,
      "step": 1008
    },
    {
      "epoch": 0.2450218552695483,
      "grad_norm": 0.35041284561157227,
      "learning_rate": 0.0001,
      "loss": 1.6782,
      "step": 1009
    },
    {
      "epoch": 0.24526469159786304,
      "grad_norm": 0.34657296538352966,
      "learning_rate": 0.0001,
      "loss": 1.7771,
      "step": 1010
    },
    {
      "epoch": 0.24550752792617775,
      "grad_norm": 0.33760038018226624,
      "learning_rate": 0.0001,
      "loss": 1.6074,
      "step": 1011
    },
    {
      "epoch": 0.24575036425449248,
      "grad_norm": 0.33262863755226135,
      "learning_rate": 0.0001,
      "loss": 1.5741,
      "step": 1012
    },
    {
      "epoch": 0.2459932005828072,
      "grad_norm": 0.358273983001709,
      "learning_rate": 0.0001,
      "loss": 1.8846,
      "step": 1013
    },
    {
      "epoch": 0.2462360369111219,
      "grad_norm": 0.33746251463890076,
      "learning_rate": 0.0001,
      "loss": 1.7365,
      "step": 1014
    },
    {
      "epoch": 0.24647887323943662,
      "grad_norm": 0.3334433436393738,
      "learning_rate": 0.0001,
      "loss": 1.8094,
      "step": 1015
    },
    {
      "epoch": 0.24672170956775133,
      "grad_norm": 0.3502381443977356,
      "learning_rate": 0.0001,
      "loss": 1.8298,
      "step": 1016
    },
    {
      "epoch": 0.24696454589606606,
      "grad_norm": 0.3290548026561737,
      "learning_rate": 0.0001,
      "loss": 1.6413,
      "step": 1017
    },
    {
      "epoch": 0.24720738222438077,
      "grad_norm": 0.3583875000476837,
      "learning_rate": 0.0001,
      "loss": 1.7214,
      "step": 1018
    },
    {
      "epoch": 0.24745021855269547,
      "grad_norm": 0.3353625535964966,
      "learning_rate": 0.0001,
      "loss": 1.7246,
      "step": 1019
    },
    {
      "epoch": 0.2476930548810102,
      "grad_norm": 0.3170337975025177,
      "learning_rate": 0.0001,
      "loss": 1.6516,
      "step": 1020
    },
    {
      "epoch": 0.2479358912093249,
      "grad_norm": 0.3242403268814087,
      "learning_rate": 0.0001,
      "loss": 1.6552,
      "step": 1021
    },
    {
      "epoch": 0.24817872753763964,
      "grad_norm": 0.3511442542076111,
      "learning_rate": 0.0001,
      "loss": 1.5973,
      "step": 1022
    },
    {
      "epoch": 0.24842156386595435,
      "grad_norm": 0.33678367733955383,
      "learning_rate": 0.0001,
      "loss": 1.7889,
      "step": 1023
    },
    {
      "epoch": 0.24866440019426905,
      "grad_norm": 0.37609171867370605,
      "learning_rate": 0.0001,
      "loss": 1.9319,
      "step": 1024
    },
    {
      "epoch": 0.24890723652258379,
      "grad_norm": 0.33121398091316223,
      "learning_rate": 0.0001,
      "loss": 1.5948,
      "step": 1025
    },
    {
      "epoch": 0.2491500728508985,
      "grad_norm": 0.3493736684322357,
      "learning_rate": 0.0001,
      "loss": 1.8547,
      "step": 1026
    },
    {
      "epoch": 0.24939290917921322,
      "grad_norm": 0.35862937569618225,
      "learning_rate": 0.0001,
      "loss": 1.6668,
      "step": 1027
    },
    {
      "epoch": 0.24963574550752793,
      "grad_norm": 0.34165531396865845,
      "learning_rate": 0.0001,
      "loss": 1.6661,
      "step": 1028
    },
    {
      "epoch": 0.24987858183584263,
      "grad_norm": 0.33531108498573303,
      "learning_rate": 0.0001,
      "loss": 1.8937,
      "step": 1029
    },
    {
      "epoch": 0.25012141816415734,
      "grad_norm": 0.34681662917137146,
      "learning_rate": 0.0001,
      "loss": 1.7362,
      "step": 1030
    },
    {
      "epoch": 0.25036425449247207,
      "grad_norm": 0.3579552471637726,
      "learning_rate": 0.0001,
      "loss": 1.8967,
      "step": 1031
    },
    {
      "epoch": 0.2506070908207868,
      "grad_norm": 0.3542710542678833,
      "learning_rate": 0.0001,
      "loss": 1.6862,
      "step": 1032
    },
    {
      "epoch": 0.2508499271491015,
      "grad_norm": 0.3317181169986725,
      "learning_rate": 0.0001,
      "loss": 1.816,
      "step": 1033
    },
    {
      "epoch": 0.2510927634774162,
      "grad_norm": 0.34695062041282654,
      "learning_rate": 0.0001,
      "loss": 1.8554,
      "step": 1034
    },
    {
      "epoch": 0.25133559980573095,
      "grad_norm": 0.3395116329193115,
      "learning_rate": 0.0001,
      "loss": 1.8296,
      "step": 1035
    },
    {
      "epoch": 0.2515784361340457,
      "grad_norm": 0.3394494354724884,
      "learning_rate": 0.0001,
      "loss": 1.7659,
      "step": 1036
    },
    {
      "epoch": 0.25182127246236036,
      "grad_norm": 0.34419873356819153,
      "learning_rate": 0.0001,
      "loss": 1.8675,
      "step": 1037
    },
    {
      "epoch": 0.2520641087906751,
      "grad_norm": 0.32015788555145264,
      "learning_rate": 0.0001,
      "loss": 1.8231,
      "step": 1038
    },
    {
      "epoch": 0.2523069451189898,
      "grad_norm": 0.33248281478881836,
      "learning_rate": 0.0001,
      "loss": 1.6975,
      "step": 1039
    },
    {
      "epoch": 0.2525497814473045,
      "grad_norm": 0.3703489303588867,
      "learning_rate": 0.0001,
      "loss": 1.7294,
      "step": 1040
    },
    {
      "epoch": 0.25279261777561923,
      "grad_norm": 0.353352427482605,
      "learning_rate": 0.0001,
      "loss": 1.8938,
      "step": 1041
    },
    {
      "epoch": 0.25303545410393397,
      "grad_norm": 0.3542862832546234,
      "learning_rate": 0.0001,
      "loss": 1.744,
      "step": 1042
    },
    {
      "epoch": 0.25327829043224864,
      "grad_norm": 0.3363003134727478,
      "learning_rate": 0.0001,
      "loss": 1.6655,
      "step": 1043
    },
    {
      "epoch": 0.2535211267605634,
      "grad_norm": 0.32284411787986755,
      "learning_rate": 0.0001,
      "loss": 1.6319,
      "step": 1044
    },
    {
      "epoch": 0.2537639630888781,
      "grad_norm": 0.33836817741394043,
      "learning_rate": 0.0001,
      "loss": 1.6846,
      "step": 1045
    },
    {
      "epoch": 0.2540067994171928,
      "grad_norm": 0.34558039903640747,
      "learning_rate": 0.0001,
      "loss": 1.6283,
      "step": 1046
    },
    {
      "epoch": 0.2542496357455075,
      "grad_norm": 0.3438437879085541,
      "learning_rate": 0.0001,
      "loss": 1.7201,
      "step": 1047
    },
    {
      "epoch": 0.25449247207382225,
      "grad_norm": 0.3524024486541748,
      "learning_rate": 0.0001,
      "loss": 1.8669,
      "step": 1048
    },
    {
      "epoch": 0.254735308402137,
      "grad_norm": 0.34446337819099426,
      "learning_rate": 0.0001,
      "loss": 1.838,
      "step": 1049
    },
    {
      "epoch": 0.25497814473045166,
      "grad_norm": 0.3405302166938782,
      "learning_rate": 0.0001,
      "loss": 1.5848,
      "step": 1050
    },
    {
      "epoch": 0.2552209810587664,
      "grad_norm": 0.3391573429107666,
      "learning_rate": 0.0001,
      "loss": 1.7496,
      "step": 1051
    },
    {
      "epoch": 0.2554638173870811,
      "grad_norm": 0.37194111943244934,
      "learning_rate": 0.0001,
      "loss": 1.7892,
      "step": 1052
    },
    {
      "epoch": 0.2557066537153958,
      "grad_norm": 0.37367933988571167,
      "learning_rate": 0.0001,
      "loss": 1.8058,
      "step": 1053
    },
    {
      "epoch": 0.25594949004371054,
      "grad_norm": 0.3580639958381653,
      "learning_rate": 0.0001,
      "loss": 1.6933,
      "step": 1054
    },
    {
      "epoch": 0.25619232637202527,
      "grad_norm": 0.3281092941761017,
      "learning_rate": 0.0001,
      "loss": 1.5708,
      "step": 1055
    },
    {
      "epoch": 0.25643516270033995,
      "grad_norm": 0.3570118546485901,
      "learning_rate": 0.0001,
      "loss": 1.8818,
      "step": 1056
    },
    {
      "epoch": 0.2566779990286547,
      "grad_norm": 0.358441561460495,
      "learning_rate": 0.0001,
      "loss": 1.943,
      "step": 1057
    },
    {
      "epoch": 0.2569208353569694,
      "grad_norm": 0.3399358093738556,
      "learning_rate": 0.0001,
      "loss": 1.746,
      "step": 1058
    },
    {
      "epoch": 0.25716367168528415,
      "grad_norm": 0.34234362840652466,
      "learning_rate": 0.0001,
      "loss": 1.6475,
      "step": 1059
    },
    {
      "epoch": 0.2574065080135988,
      "grad_norm": 0.34568747878074646,
      "learning_rate": 0.0001,
      "loss": 1.6586,
      "step": 1060
    },
    {
      "epoch": 0.25764934434191356,
      "grad_norm": 0.38038501143455505,
      "learning_rate": 0.0001,
      "loss": 1.6381,
      "step": 1061
    },
    {
      "epoch": 0.2578921806702283,
      "grad_norm": 0.33781126141548157,
      "learning_rate": 0.0001,
      "loss": 1.7275,
      "step": 1062
    },
    {
      "epoch": 0.25813501699854297,
      "grad_norm": 0.36487317085266113,
      "learning_rate": 0.0001,
      "loss": 1.7277,
      "step": 1063
    },
    {
      "epoch": 0.2583778533268577,
      "grad_norm": 0.35598310828208923,
      "learning_rate": 0.0001,
      "loss": 1.7926,
      "step": 1064
    },
    {
      "epoch": 0.25862068965517243,
      "grad_norm": 0.3338935673236847,
      "learning_rate": 0.0001,
      "loss": 1.5767,
      "step": 1065
    },
    {
      "epoch": 0.2588635259834871,
      "grad_norm": 0.3760107159614563,
      "learning_rate": 0.0001,
      "loss": 2.0399,
      "step": 1066
    },
    {
      "epoch": 0.25910636231180184,
      "grad_norm": 0.3443123698234558,
      "learning_rate": 0.0001,
      "loss": 1.6412,
      "step": 1067
    },
    {
      "epoch": 0.2593491986401166,
      "grad_norm": 0.32958465814590454,
      "learning_rate": 0.0001,
      "loss": 1.6551,
      "step": 1068
    },
    {
      "epoch": 0.25959203496843125,
      "grad_norm": 0.36839255690574646,
      "learning_rate": 0.0001,
      "loss": 1.805,
      "step": 1069
    },
    {
      "epoch": 0.259834871296746,
      "grad_norm": 0.33118322491645813,
      "learning_rate": 0.0001,
      "loss": 1.811,
      "step": 1070
    },
    {
      "epoch": 0.2600777076250607,
      "grad_norm": 0.33004656434059143,
      "learning_rate": 0.0001,
      "loss": 1.7253,
      "step": 1071
    },
    {
      "epoch": 0.26032054395337545,
      "grad_norm": 0.34900596737861633,
      "learning_rate": 0.0001,
      "loss": 1.8114,
      "step": 1072
    },
    {
      "epoch": 0.2605633802816901,
      "grad_norm": 0.31525158882141113,
      "learning_rate": 0.0001,
      "loss": 1.5306,
      "step": 1073
    },
    {
      "epoch": 0.26080621661000486,
      "grad_norm": 0.3604126274585724,
      "learning_rate": 0.0001,
      "loss": 1.9126,
      "step": 1074
    },
    {
      "epoch": 0.2610490529383196,
      "grad_norm": 0.36580803990364075,
      "learning_rate": 0.0001,
      "loss": 1.7778,
      "step": 1075
    },
    {
      "epoch": 0.26129188926663427,
      "grad_norm": 0.32837146520614624,
      "learning_rate": 0.0001,
      "loss": 1.5917,
      "step": 1076
    },
    {
      "epoch": 0.261534725594949,
      "grad_norm": 0.35636448860168457,
      "learning_rate": 0.0001,
      "loss": 1.8045,
      "step": 1077
    },
    {
      "epoch": 0.26177756192326374,
      "grad_norm": 0.3386078476905823,
      "learning_rate": 0.0001,
      "loss": 1.8125,
      "step": 1078
    },
    {
      "epoch": 0.2620203982515784,
      "grad_norm": 0.32973551750183105,
      "learning_rate": 0.0001,
      "loss": 1.7973,
      "step": 1079
    },
    {
      "epoch": 0.26226323457989315,
      "grad_norm": 0.3724091947078705,
      "learning_rate": 0.0001,
      "loss": 1.9441,
      "step": 1080
    },
    {
      "epoch": 0.2625060709082079,
      "grad_norm": 0.345823734998703,
      "learning_rate": 0.0001,
      "loss": 1.7337,
      "step": 1081
    },
    {
      "epoch": 0.2627489072365226,
      "grad_norm": 0.369800329208374,
      "learning_rate": 0.0001,
      "loss": 1.7492,
      "step": 1082
    },
    {
      "epoch": 0.2629917435648373,
      "grad_norm": 0.35781586170196533,
      "learning_rate": 0.0001,
      "loss": 1.8698,
      "step": 1083
    },
    {
      "epoch": 0.263234579893152,
      "grad_norm": 0.3464943766593933,
      "learning_rate": 0.0001,
      "loss": 1.7151,
      "step": 1084
    },
    {
      "epoch": 0.26347741622146675,
      "grad_norm": 0.34492209553718567,
      "learning_rate": 0.0001,
      "loss": 1.8346,
      "step": 1085
    },
    {
      "epoch": 0.26372025254978143,
      "grad_norm": 0.32053568959236145,
      "learning_rate": 0.0001,
      "loss": 1.7688,
      "step": 1086
    },
    {
      "epoch": 0.26396308887809616,
      "grad_norm": 0.30503571033477783,
      "learning_rate": 0.0001,
      "loss": 1.6619,
      "step": 1087
    },
    {
      "epoch": 0.2642059252064109,
      "grad_norm": 0.3440854847431183,
      "learning_rate": 0.0001,
      "loss": 1.9203,
      "step": 1088
    },
    {
      "epoch": 0.2644487615347256,
      "grad_norm": 0.3316665589809418,
      "learning_rate": 0.0001,
      "loss": 1.833,
      "step": 1089
    },
    {
      "epoch": 0.2646915978630403,
      "grad_norm": 0.34309491515159607,
      "learning_rate": 0.0001,
      "loss": 1.9606,
      "step": 1090
    },
    {
      "epoch": 0.26493443419135504,
      "grad_norm": 0.3301270008087158,
      "learning_rate": 0.0001,
      "loss": 1.8049,
      "step": 1091
    },
    {
      "epoch": 0.2651772705196697,
      "grad_norm": 0.32701781392097473,
      "learning_rate": 0.0001,
      "loss": 1.5786,
      "step": 1092
    },
    {
      "epoch": 0.26542010684798445,
      "grad_norm": 0.3516250550746918,
      "learning_rate": 0.0001,
      "loss": 1.8085,
      "step": 1093
    },
    {
      "epoch": 0.2656629431762992,
      "grad_norm": 0.34311559796333313,
      "learning_rate": 0.0001,
      "loss": 1.8699,
      "step": 1094
    },
    {
      "epoch": 0.2659057795046139,
      "grad_norm": 0.3706408441066742,
      "learning_rate": 0.0001,
      "loss": 1.621,
      "step": 1095
    },
    {
      "epoch": 0.2661486158329286,
      "grad_norm": 0.43447747826576233,
      "learning_rate": 0.0001,
      "loss": 1.6379,
      "step": 1096
    },
    {
      "epoch": 0.2663914521612433,
      "grad_norm": 0.35497137904167175,
      "learning_rate": 0.0001,
      "loss": 1.8351,
      "step": 1097
    },
    {
      "epoch": 0.26663428848955806,
      "grad_norm": 0.3245193064212799,
      "learning_rate": 0.0001,
      "loss": 1.5895,
      "step": 1098
    },
    {
      "epoch": 0.26687712481787274,
      "grad_norm": 0.3037337362766266,
      "learning_rate": 0.0001,
      "loss": 1.5043,
      "step": 1099
    },
    {
      "epoch": 0.26711996114618747,
      "grad_norm": 0.35018908977508545,
      "learning_rate": 0.0001,
      "loss": 1.7701,
      "step": 1100
    },
    {
      "epoch": 0.2673627974745022,
      "grad_norm": 0.3398996889591217,
      "learning_rate": 0.0001,
      "loss": 1.6926,
      "step": 1101
    },
    {
      "epoch": 0.2676056338028169,
      "grad_norm": 0.36320433020591736,
      "learning_rate": 0.0001,
      "loss": 1.8357,
      "step": 1102
    },
    {
      "epoch": 0.2678484701311316,
      "grad_norm": 0.3703799843788147,
      "learning_rate": 0.0001,
      "loss": 1.7514,
      "step": 1103
    },
    {
      "epoch": 0.26809130645944634,
      "grad_norm": 0.34844735264778137,
      "learning_rate": 0.0001,
      "loss": 1.7834,
      "step": 1104
    },
    {
      "epoch": 0.2683341427877611,
      "grad_norm": 0.3272978961467743,
      "learning_rate": 0.0001,
      "loss": 1.7616,
      "step": 1105
    },
    {
      "epoch": 0.26857697911607575,
      "grad_norm": 0.33166933059692383,
      "learning_rate": 0.0001,
      "loss": 1.6827,
      "step": 1106
    },
    {
      "epoch": 0.2688198154443905,
      "grad_norm": 0.32959240674972534,
      "learning_rate": 0.0001,
      "loss": 1.8041,
      "step": 1107
    },
    {
      "epoch": 0.2690626517727052,
      "grad_norm": 0.3293261229991913,
      "learning_rate": 0.0001,
      "loss": 1.5557,
      "step": 1108
    },
    {
      "epoch": 0.2693054881010199,
      "grad_norm": 0.3458395004272461,
      "learning_rate": 0.0001,
      "loss": 1.864,
      "step": 1109
    },
    {
      "epoch": 0.26954832442933463,
      "grad_norm": 0.3345559537410736,
      "learning_rate": 0.0001,
      "loss": 1.8151,
      "step": 1110
    },
    {
      "epoch": 0.26979116075764936,
      "grad_norm": 0.3532974123954773,
      "learning_rate": 0.0001,
      "loss": 1.7696,
      "step": 1111
    },
    {
      "epoch": 0.27003399708596404,
      "grad_norm": 0.32688626646995544,
      "learning_rate": 0.0001,
      "loss": 1.6863,
      "step": 1112
    },
    {
      "epoch": 0.27027683341427877,
      "grad_norm": 0.35777220129966736,
      "learning_rate": 0.0001,
      "loss": 1.7081,
      "step": 1113
    },
    {
      "epoch": 0.2705196697425935,
      "grad_norm": 0.3583647906780243,
      "learning_rate": 0.0001,
      "loss": 1.7594,
      "step": 1114
    },
    {
      "epoch": 0.2707625060709082,
      "grad_norm": 0.34510111808776855,
      "learning_rate": 0.0001,
      "loss": 1.7195,
      "step": 1115
    },
    {
      "epoch": 0.2710053423992229,
      "grad_norm": 0.3345058858394623,
      "learning_rate": 0.0001,
      "loss": 1.8784,
      "step": 1116
    },
    {
      "epoch": 0.27124817872753765,
      "grad_norm": 0.34480851888656616,
      "learning_rate": 0.0001,
      "loss": 1.6358,
      "step": 1117
    },
    {
      "epoch": 0.2714910150558524,
      "grad_norm": 0.33593034744262695,
      "learning_rate": 0.0001,
      "loss": 1.7992,
      "step": 1118
    },
    {
      "epoch": 0.27173385138416706,
      "grad_norm": 0.34252050518989563,
      "learning_rate": 0.0001,
      "loss": 1.6939,
      "step": 1119
    },
    {
      "epoch": 0.2719766877124818,
      "grad_norm": 0.3233051598072052,
      "learning_rate": 0.0001,
      "loss": 1.5801,
      "step": 1120
    },
    {
      "epoch": 0.2722195240407965,
      "grad_norm": 0.32091420888900757,
      "learning_rate": 0.0001,
      "loss": 1.8013,
      "step": 1121
    },
    {
      "epoch": 0.2724623603691112,
      "grad_norm": 0.35031384229660034,
      "learning_rate": 0.0001,
      "loss": 1.7404,
      "step": 1122
    },
    {
      "epoch": 0.27270519669742593,
      "grad_norm": 0.33827584981918335,
      "learning_rate": 0.0001,
      "loss": 1.9034,
      "step": 1123
    },
    {
      "epoch": 0.27294803302574067,
      "grad_norm": 0.3441354036331177,
      "learning_rate": 0.0001,
      "loss": 1.7871,
      "step": 1124
    },
    {
      "epoch": 0.27319086935405534,
      "grad_norm": 0.33458590507507324,
      "learning_rate": 0.0001,
      "loss": 1.6163,
      "step": 1125
    },
    {
      "epoch": 0.2734337056823701,
      "grad_norm": 0.36066651344299316,
      "learning_rate": 0.0001,
      "loss": 1.9302,
      "step": 1126
    },
    {
      "epoch": 0.2736765420106848,
      "grad_norm": 0.3288789689540863,
      "learning_rate": 0.0001,
      "loss": 1.6241,
      "step": 1127
    },
    {
      "epoch": 0.27391937833899954,
      "grad_norm": 0.35188278555870056,
      "learning_rate": 0.0001,
      "loss": 1.8012,
      "step": 1128
    },
    {
      "epoch": 0.2741622146673142,
      "grad_norm": 0.33101344108581543,
      "learning_rate": 0.0001,
      "loss": 1.6815,
      "step": 1129
    },
    {
      "epoch": 0.27440505099562895,
      "grad_norm": 0.3569447100162506,
      "learning_rate": 0.0001,
      "loss": 1.7191,
      "step": 1130
    },
    {
      "epoch": 0.2746478873239437,
      "grad_norm": 0.35680562257766724,
      "learning_rate": 0.0001,
      "loss": 1.7218,
      "step": 1131
    },
    {
      "epoch": 0.27489072365225836,
      "grad_norm": 0.33136245608329773,
      "learning_rate": 0.0001,
      "loss": 1.6662,
      "step": 1132
    },
    {
      "epoch": 0.2751335599805731,
      "grad_norm": 0.3718142807483673,
      "learning_rate": 0.0001,
      "loss": 1.9081,
      "step": 1133
    },
    {
      "epoch": 0.2753763963088878,
      "grad_norm": 0.32091063261032104,
      "learning_rate": 0.0001,
      "loss": 1.7089,
      "step": 1134
    },
    {
      "epoch": 0.2756192326372025,
      "grad_norm": 0.3750685453414917,
      "learning_rate": 0.0001,
      "loss": 1.9496,
      "step": 1135
    },
    {
      "epoch": 0.27586206896551724,
      "grad_norm": 0.36933282017707825,
      "learning_rate": 0.0001,
      "loss": 1.7596,
      "step": 1136
    },
    {
      "epoch": 0.27610490529383197,
      "grad_norm": 0.32732218503952026,
      "learning_rate": 0.0001,
      "loss": 1.7492,
      "step": 1137
    },
    {
      "epoch": 0.27634774162214665,
      "grad_norm": 0.3513854444026947,
      "learning_rate": 0.0001,
      "loss": 1.9196,
      "step": 1138
    },
    {
      "epoch": 0.2765905779504614,
      "grad_norm": 0.3355165123939514,
      "learning_rate": 0.0001,
      "loss": 1.7641,
      "step": 1139
    },
    {
      "epoch": 0.2768334142787761,
      "grad_norm": 0.3600131869316101,
      "learning_rate": 0.0001,
      "loss": 1.8531,
      "step": 1140
    },
    {
      "epoch": 0.27707625060709085,
      "grad_norm": 0.34687429666519165,
      "learning_rate": 0.0001,
      "loss": 1.7288,
      "step": 1141
    },
    {
      "epoch": 0.2773190869354055,
      "grad_norm": 0.3468457758426666,
      "learning_rate": 0.0001,
      "loss": 1.8647,
      "step": 1142
    },
    {
      "epoch": 0.27756192326372026,
      "grad_norm": 0.36524146795272827,
      "learning_rate": 0.0001,
      "loss": 1.745,
      "step": 1143
    },
    {
      "epoch": 0.277804759592035,
      "grad_norm": 0.3697652220726013,
      "learning_rate": 0.0001,
      "loss": 1.7623,
      "step": 1144
    },
    {
      "epoch": 0.27804759592034967,
      "grad_norm": 0.3558076024055481,
      "learning_rate": 0.0001,
      "loss": 1.5631,
      "step": 1145
    },
    {
      "epoch": 0.2782904322486644,
      "grad_norm": 0.3579453229904175,
      "learning_rate": 0.0001,
      "loss": 1.7172,
      "step": 1146
    },
    {
      "epoch": 0.27853326857697913,
      "grad_norm": 0.35010629892349243,
      "learning_rate": 0.0001,
      "loss": 1.7867,
      "step": 1147
    },
    {
      "epoch": 0.2787761049052938,
      "grad_norm": 0.3663645386695862,
      "learning_rate": 0.0001,
      "loss": 1.8068,
      "step": 1148
    },
    {
      "epoch": 0.27901894123360854,
      "grad_norm": 0.34487295150756836,
      "learning_rate": 0.0001,
      "loss": 1.7174,
      "step": 1149
    },
    {
      "epoch": 0.2792617775619233,
      "grad_norm": 0.35026541352272034,
      "learning_rate": 0.0001,
      "loss": 1.6901,
      "step": 1150
    },
    {
      "epoch": 0.279504613890238,
      "grad_norm": 0.33784544467926025,
      "learning_rate": 0.0001,
      "loss": 1.8176,
      "step": 1151
    },
    {
      "epoch": 0.2797474502185527,
      "grad_norm": 0.35948890447616577,
      "learning_rate": 0.0001,
      "loss": 1.6484,
      "step": 1152
    },
    {
      "epoch": 0.2799902865468674,
      "grad_norm": 0.33784404397010803,
      "learning_rate": 0.0001,
      "loss": 1.829,
      "step": 1153
    },
    {
      "epoch": 0.28023312287518215,
      "grad_norm": 0.36918583512306213,
      "learning_rate": 0.0001,
      "loss": 1.691,
      "step": 1154
    },
    {
      "epoch": 0.2804759592034968,
      "grad_norm": 0.32872992753982544,
      "learning_rate": 0.0001,
      "loss": 1.4461,
      "step": 1155
    },
    {
      "epoch": 0.28071879553181156,
      "grad_norm": 0.3497970402240753,
      "learning_rate": 0.0001,
      "loss": 1.8181,
      "step": 1156
    },
    {
      "epoch": 0.2809616318601263,
      "grad_norm": 0.34661903977394104,
      "learning_rate": 0.0001,
      "loss": 1.9472,
      "step": 1157
    },
    {
      "epoch": 0.28120446818844097,
      "grad_norm": 0.34361758828163147,
      "learning_rate": 0.0001,
      "loss": 1.7651,
      "step": 1158
    },
    {
      "epoch": 0.2814473045167557,
      "grad_norm": 0.32696354389190674,
      "learning_rate": 0.0001,
      "loss": 1.4568,
      "step": 1159
    },
    {
      "epoch": 0.28169014084507044,
      "grad_norm": 0.3309074640274048,
      "learning_rate": 0.0001,
      "loss": 1.6149,
      "step": 1160
    },
    {
      "epoch": 0.2819329771733851,
      "grad_norm": 0.32438015937805176,
      "learning_rate": 0.0001,
      "loss": 1.5894,
      "step": 1161
    },
    {
      "epoch": 0.28217581350169985,
      "grad_norm": 0.36756324768066406,
      "learning_rate": 0.0001,
      "loss": 1.8005,
      "step": 1162
    },
    {
      "epoch": 0.2824186498300146,
      "grad_norm": 0.3359546661376953,
      "learning_rate": 0.0001,
      "loss": 1.6269,
      "step": 1163
    },
    {
      "epoch": 0.2826614861583293,
      "grad_norm": 0.3752913475036621,
      "learning_rate": 0.0001,
      "loss": 1.8105,
      "step": 1164
    },
    {
      "epoch": 0.282904322486644,
      "grad_norm": 0.33042916655540466,
      "learning_rate": 0.0001,
      "loss": 1.6781,
      "step": 1165
    },
    {
      "epoch": 0.2831471588149587,
      "grad_norm": 0.3419846296310425,
      "learning_rate": 0.0001,
      "loss": 1.7958,
      "step": 1166
    },
    {
      "epoch": 0.28338999514327345,
      "grad_norm": 0.354832261800766,
      "learning_rate": 0.0001,
      "loss": 1.8978,
      "step": 1167
    },
    {
      "epoch": 0.28363283147158813,
      "grad_norm": 0.3304416835308075,
      "learning_rate": 0.0001,
      "loss": 1.6946,
      "step": 1168
    },
    {
      "epoch": 0.28387566779990286,
      "grad_norm": 0.3565024435520172,
      "learning_rate": 0.0001,
      "loss": 1.7724,
      "step": 1169
    },
    {
      "epoch": 0.2841185041282176,
      "grad_norm": 0.341604083776474,
      "learning_rate": 0.0001,
      "loss": 1.764,
      "step": 1170
    },
    {
      "epoch": 0.2843613404565323,
      "grad_norm": 0.35261428356170654,
      "learning_rate": 0.0001,
      "loss": 1.7225,
      "step": 1171
    },
    {
      "epoch": 0.284604176784847,
      "grad_norm": 0.32757368683815,
      "learning_rate": 0.0001,
      "loss": 1.4222,
      "step": 1172
    },
    {
      "epoch": 0.28484701311316174,
      "grad_norm": 0.3484155237674713,
      "learning_rate": 0.0001,
      "loss": 1.7307,
      "step": 1173
    },
    {
      "epoch": 0.2850898494414764,
      "grad_norm": 0.3501245379447937,
      "learning_rate": 0.0001,
      "loss": 1.7643,
      "step": 1174
    },
    {
      "epoch": 0.28533268576979115,
      "grad_norm": 0.3286750614643097,
      "learning_rate": 0.0001,
      "loss": 1.6124,
      "step": 1175
    },
    {
      "epoch": 0.2855755220981059,
      "grad_norm": 0.3350943624973297,
      "learning_rate": 0.0001,
      "loss": 1.7167,
      "step": 1176
    },
    {
      "epoch": 0.2858183584264206,
      "grad_norm": 0.3862624764442444,
      "learning_rate": 0.0001,
      "loss": 1.8182,
      "step": 1177
    },
    {
      "epoch": 0.2860611947547353,
      "grad_norm": 0.4132660925388336,
      "learning_rate": 0.0001,
      "loss": 1.8159,
      "step": 1178
    },
    {
      "epoch": 0.28630403108305,
      "grad_norm": 0.3466343283653259,
      "learning_rate": 0.0001,
      "loss": 1.8844,
      "step": 1179
    },
    {
      "epoch": 0.28654686741136476,
      "grad_norm": 0.342422217130661,
      "learning_rate": 0.0001,
      "loss": 1.8049,
      "step": 1180
    },
    {
      "epoch": 0.28678970373967944,
      "grad_norm": 0.3274907171726227,
      "learning_rate": 0.0001,
      "loss": 1.8003,
      "step": 1181
    },
    {
      "epoch": 0.28703254006799417,
      "grad_norm": 0.34425753355026245,
      "learning_rate": 0.0001,
      "loss": 1.8901,
      "step": 1182
    },
    {
      "epoch": 0.2872753763963089,
      "grad_norm": 0.3420788645744324,
      "learning_rate": 0.0001,
      "loss": 1.8626,
      "step": 1183
    },
    {
      "epoch": 0.2875182127246236,
      "grad_norm": 0.37537306547164917,
      "learning_rate": 0.0001,
      "loss": 1.768,
      "step": 1184
    },
    {
      "epoch": 0.2877610490529383,
      "grad_norm": 0.3327960968017578,
      "learning_rate": 0.0001,
      "loss": 1.743,
      "step": 1185
    },
    {
      "epoch": 0.28800388538125304,
      "grad_norm": 0.3642200231552124,
      "learning_rate": 0.0001,
      "loss": 1.9741,
      "step": 1186
    },
    {
      "epoch": 0.2882467217095678,
      "grad_norm": 0.3397223949432373,
      "learning_rate": 0.0001,
      "loss": 1.8706,
      "step": 1187
    },
    {
      "epoch": 0.28848955803788245,
      "grad_norm": 0.32793283462524414,
      "learning_rate": 0.0001,
      "loss": 1.6044,
      "step": 1188
    },
    {
      "epoch": 0.2887323943661972,
      "grad_norm": 0.3629734516143799,
      "learning_rate": 0.0001,
      "loss": 1.7398,
      "step": 1189
    },
    {
      "epoch": 0.2889752306945119,
      "grad_norm": 0.34847310185432434,
      "learning_rate": 0.0001,
      "loss": 1.687,
      "step": 1190
    },
    {
      "epoch": 0.2892180670228266,
      "grad_norm": 0.34422287344932556,
      "learning_rate": 0.0001,
      "loss": 1.7657,
      "step": 1191
    },
    {
      "epoch": 0.28946090335114133,
      "grad_norm": 0.34734368324279785,
      "learning_rate": 0.0001,
      "loss": 1.8799,
      "step": 1192
    },
    {
      "epoch": 0.28970373967945606,
      "grad_norm": 0.3396027386188507,
      "learning_rate": 0.0001,
      "loss": 1.7647,
      "step": 1193
    },
    {
      "epoch": 0.28994657600777074,
      "grad_norm": 0.3526374399662018,
      "learning_rate": 0.0001,
      "loss": 1.9586,
      "step": 1194
    },
    {
      "epoch": 0.2901894123360855,
      "grad_norm": 0.34345927834510803,
      "learning_rate": 0.0001,
      "loss": 1.7944,
      "step": 1195
    },
    {
      "epoch": 0.2904322486644002,
      "grad_norm": 0.3783220052719116,
      "learning_rate": 0.0001,
      "loss": 1.681,
      "step": 1196
    },
    {
      "epoch": 0.2906750849927149,
      "grad_norm": 0.35073280334472656,
      "learning_rate": 0.0001,
      "loss": 1.5092,
      "step": 1197
    },
    {
      "epoch": 0.2909179213210296,
      "grad_norm": 0.352989137172699,
      "learning_rate": 0.0001,
      "loss": 1.7788,
      "step": 1198
    },
    {
      "epoch": 0.29116075764934435,
      "grad_norm": 0.346623033285141,
      "learning_rate": 0.0001,
      "loss": 1.6689,
      "step": 1199
    },
    {
      "epoch": 0.2914035939776591,
      "grad_norm": 0.3349616825580597,
      "learning_rate": 0.0001,
      "loss": 1.8613,
      "step": 1200
    },
    {
      "epoch": 0.29164643030597376,
      "grad_norm": 0.3176116943359375,
      "learning_rate": 0.0001,
      "loss": 1.5567,
      "step": 1201
    },
    {
      "epoch": 0.2918892666342885,
      "grad_norm": 0.316850870847702,
      "learning_rate": 0.0001,
      "loss": 1.6958,
      "step": 1202
    },
    {
      "epoch": 0.2921321029626032,
      "grad_norm": 0.36198094487190247,
      "learning_rate": 0.0001,
      "loss": 1.9053,
      "step": 1203
    },
    {
      "epoch": 0.2923749392909179,
      "grad_norm": 0.33802559971809387,
      "learning_rate": 0.0001,
      "loss": 1.6131,
      "step": 1204
    },
    {
      "epoch": 0.29261777561923263,
      "grad_norm": 0.353243887424469,
      "learning_rate": 0.0001,
      "loss": 1.6892,
      "step": 1205
    },
    {
      "epoch": 0.29286061194754737,
      "grad_norm": 0.3540343642234802,
      "learning_rate": 0.0001,
      "loss": 1.6646,
      "step": 1206
    },
    {
      "epoch": 0.29310344827586204,
      "grad_norm": 0.3387995958328247,
      "learning_rate": 0.0001,
      "loss": 1.7045,
      "step": 1207
    },
    {
      "epoch": 0.2933462846041768,
      "grad_norm": 0.3644068241119385,
      "learning_rate": 0.0001,
      "loss": 1.8486,
      "step": 1208
    },
    {
      "epoch": 0.2935891209324915,
      "grad_norm": 0.35250940918922424,
      "learning_rate": 0.0001,
      "loss": 1.7084,
      "step": 1209
    },
    {
      "epoch": 0.29383195726080624,
      "grad_norm": 0.3456801176071167,
      "learning_rate": 0.0001,
      "loss": 1.6556,
      "step": 1210
    },
    {
      "epoch": 0.2940747935891209,
      "grad_norm": 0.35365554690361023,
      "learning_rate": 0.0001,
      "loss": 1.8963,
      "step": 1211
    },
    {
      "epoch": 0.29431762991743565,
      "grad_norm": 0.3480612337589264,
      "learning_rate": 0.0001,
      "loss": 1.6345,
      "step": 1212
    },
    {
      "epoch": 0.2945604662457504,
      "grad_norm": 0.3324591815471649,
      "learning_rate": 0.0001,
      "loss": 1.7345,
      "step": 1213
    },
    {
      "epoch": 0.29480330257406506,
      "grad_norm": 0.36215439438819885,
      "learning_rate": 0.0001,
      "loss": 1.7907,
      "step": 1214
    },
    {
      "epoch": 0.2950461389023798,
      "grad_norm": 0.36731013655662537,
      "learning_rate": 0.0001,
      "loss": 1.7969,
      "step": 1215
    },
    {
      "epoch": 0.29528897523069453,
      "grad_norm": 0.3301365375518799,
      "learning_rate": 0.0001,
      "loss": 1.7202,
      "step": 1216
    },
    {
      "epoch": 0.2955318115590092,
      "grad_norm": 0.3361029624938965,
      "learning_rate": 0.0001,
      "loss": 1.6314,
      "step": 1217
    },
    {
      "epoch": 0.29577464788732394,
      "grad_norm": 0.3317160904407501,
      "learning_rate": 0.0001,
      "loss": 1.6849,
      "step": 1218
    },
    {
      "epoch": 0.29601748421563867,
      "grad_norm": 0.32502481341362,
      "learning_rate": 0.0001,
      "loss": 1.4891,
      "step": 1219
    },
    {
      "epoch": 0.29626032054395335,
      "grad_norm": 0.3280821740627289,
      "learning_rate": 0.0001,
      "loss": 1.6065,
      "step": 1220
    },
    {
      "epoch": 0.2965031568722681,
      "grad_norm": 0.3488386869430542,
      "learning_rate": 0.0001,
      "loss": 1.7172,
      "step": 1221
    },
    {
      "epoch": 0.2967459932005828,
      "grad_norm": 0.3545106053352356,
      "learning_rate": 0.0001,
      "loss": 1.7147,
      "step": 1222
    },
    {
      "epoch": 0.29698882952889755,
      "grad_norm": 0.34625324606895447,
      "learning_rate": 0.0001,
      "loss": 1.7451,
      "step": 1223
    },
    {
      "epoch": 0.2972316658572122,
      "grad_norm": 0.365535169839859,
      "learning_rate": 0.0001,
      "loss": 1.9306,
      "step": 1224
    },
    {
      "epoch": 0.29747450218552696,
      "grad_norm": 0.3703577518463135,
      "learning_rate": 0.0001,
      "loss": 1.7975,
      "step": 1225
    },
    {
      "epoch": 0.2977173385138417,
      "grad_norm": 0.4047102630138397,
      "learning_rate": 0.0001,
      "loss": 2.0506,
      "step": 1226
    },
    {
      "epoch": 0.29796017484215637,
      "grad_norm": 0.35949185490608215,
      "learning_rate": 0.0001,
      "loss": 1.9218,
      "step": 1227
    },
    {
      "epoch": 0.2982030111704711,
      "grad_norm": 0.34228986501693726,
      "learning_rate": 0.0001,
      "loss": 1.6128,
      "step": 1228
    },
    {
      "epoch": 0.29844584749878583,
      "grad_norm": 0.3485621511936188,
      "learning_rate": 0.0001,
      "loss": 1.8851,
      "step": 1229
    },
    {
      "epoch": 0.2986886838271005,
      "grad_norm": 0.3268779516220093,
      "learning_rate": 0.0001,
      "loss": 1.68,
      "step": 1230
    },
    {
      "epoch": 0.29893152015541524,
      "grad_norm": 0.3476763069629669,
      "learning_rate": 0.0001,
      "loss": 1.7868,
      "step": 1231
    },
    {
      "epoch": 0.29917435648373,
      "grad_norm": 0.34079062938690186,
      "learning_rate": 0.0001,
      "loss": 1.6513,
      "step": 1232
    },
    {
      "epoch": 0.2994171928120447,
      "grad_norm": 0.3342800736427307,
      "learning_rate": 0.0001,
      "loss": 1.6783,
      "step": 1233
    },
    {
      "epoch": 0.2996600291403594,
      "grad_norm": 0.3746899366378784,
      "learning_rate": 0.0001,
      "loss": 1.7811,
      "step": 1234
    },
    {
      "epoch": 0.2999028654686741,
      "grad_norm": 0.362958699464798,
      "learning_rate": 0.0001,
      "loss": 1.8097,
      "step": 1235
    },
    {
      "epoch": 0.30014570179698885,
      "grad_norm": 0.34721145033836365,
      "learning_rate": 0.0001,
      "loss": 1.8311,
      "step": 1236
    },
    {
      "epoch": 0.30038853812530353,
      "grad_norm": 0.3652319610118866,
      "learning_rate": 0.0001,
      "loss": 1.948,
      "step": 1237
    },
    {
      "epoch": 0.30063137445361826,
      "grad_norm": 0.331961452960968,
      "learning_rate": 0.0001,
      "loss": 1.6468,
      "step": 1238
    },
    {
      "epoch": 0.300874210781933,
      "grad_norm": 0.3217456638813019,
      "learning_rate": 0.0001,
      "loss": 1.7451,
      "step": 1239
    },
    {
      "epoch": 0.30111704711024767,
      "grad_norm": 0.3655998408794403,
      "learning_rate": 0.0001,
      "loss": 1.8643,
      "step": 1240
    },
    {
      "epoch": 0.3013598834385624,
      "grad_norm": 0.32531607151031494,
      "learning_rate": 0.0001,
      "loss": 1.713,
      "step": 1241
    },
    {
      "epoch": 0.30160271976687714,
      "grad_norm": 0.3473297357559204,
      "learning_rate": 0.0001,
      "loss": 1.7422,
      "step": 1242
    },
    {
      "epoch": 0.3018455560951918,
      "grad_norm": 0.33475473523139954,
      "learning_rate": 0.0001,
      "loss": 1.5828,
      "step": 1243
    },
    {
      "epoch": 0.30208839242350655,
      "grad_norm": 0.3307666480541229,
      "learning_rate": 0.0001,
      "loss": 1.6227,
      "step": 1244
    },
    {
      "epoch": 0.3023312287518213,
      "grad_norm": 0.345688134431839,
      "learning_rate": 0.0001,
      "loss": 1.8162,
      "step": 1245
    },
    {
      "epoch": 0.302574065080136,
      "grad_norm": 0.37213560938835144,
      "learning_rate": 0.0001,
      "loss": 1.777,
      "step": 1246
    },
    {
      "epoch": 0.3028169014084507,
      "grad_norm": 0.3556644320487976,
      "learning_rate": 0.0001,
      "loss": 1.7458,
      "step": 1247
    },
    {
      "epoch": 0.3030597377367654,
      "grad_norm": 0.31667470932006836,
      "learning_rate": 0.0001,
      "loss": 1.6664,
      "step": 1248
    },
    {
      "epoch": 0.30330257406508015,
      "grad_norm": 0.3247813284397125,
      "learning_rate": 0.0001,
      "loss": 1.5334,
      "step": 1249
    },
    {
      "epoch": 0.30354541039339483,
      "grad_norm": 0.3366319239139557,
      "learning_rate": 0.0001,
      "loss": 1.7826,
      "step": 1250
    },
    {
      "epoch": 0.30378824672170957,
      "grad_norm": 0.3648545444011688,
      "learning_rate": 0.0001,
      "loss": 1.8183,
      "step": 1251
    },
    {
      "epoch": 0.3040310830500243,
      "grad_norm": 0.33485549688339233,
      "learning_rate": 0.0001,
      "loss": 1.7598,
      "step": 1252
    },
    {
      "epoch": 0.304273919378339,
      "grad_norm": 0.3424319624900818,
      "learning_rate": 0.0001,
      "loss": 1.73,
      "step": 1253
    },
    {
      "epoch": 0.3045167557066537,
      "grad_norm": 0.3423166871070862,
      "learning_rate": 0.0001,
      "loss": 1.9582,
      "step": 1254
    },
    {
      "epoch": 0.30475959203496844,
      "grad_norm": 0.3264666795730591,
      "learning_rate": 0.0001,
      "loss": 1.6974,
      "step": 1255
    },
    {
      "epoch": 0.3050024283632832,
      "grad_norm": 0.3453938663005829,
      "learning_rate": 0.0001,
      "loss": 1.7855,
      "step": 1256
    },
    {
      "epoch": 0.30524526469159785,
      "grad_norm": 0.34164607524871826,
      "learning_rate": 0.0001,
      "loss": 1.7747,
      "step": 1257
    },
    {
      "epoch": 0.3054881010199126,
      "grad_norm": 0.3310208022594452,
      "learning_rate": 0.0001,
      "loss": 1.6425,
      "step": 1258
    },
    {
      "epoch": 0.3057309373482273,
      "grad_norm": 0.37676697969436646,
      "learning_rate": 0.0001,
      "loss": 1.7901,
      "step": 1259
    },
    {
      "epoch": 0.305973773676542,
      "grad_norm": 0.3380137085914612,
      "learning_rate": 0.0001,
      "loss": 1.7139,
      "step": 1260
    },
    {
      "epoch": 0.3062166100048567,
      "grad_norm": 0.3388557732105255,
      "learning_rate": 0.0001,
      "loss": 1.7235,
      "step": 1261
    },
    {
      "epoch": 0.30645944633317146,
      "grad_norm": 0.31976181268692017,
      "learning_rate": 0.0001,
      "loss": 1.6238,
      "step": 1262
    },
    {
      "epoch": 0.30670228266148614,
      "grad_norm": 0.3477676808834076,
      "learning_rate": 0.0001,
      "loss": 1.7641,
      "step": 1263
    },
    {
      "epoch": 0.30694511898980087,
      "grad_norm": 0.3569043278694153,
      "learning_rate": 0.0001,
      "loss": 1.852,
      "step": 1264
    },
    {
      "epoch": 0.3071879553181156,
      "grad_norm": 0.3672388792037964,
      "learning_rate": 0.0001,
      "loss": 1.9131,
      "step": 1265
    },
    {
      "epoch": 0.3074307916464303,
      "grad_norm": 0.3455599546432495,
      "learning_rate": 0.0001,
      "loss": 1.5969,
      "step": 1266
    },
    {
      "epoch": 0.307673627974745,
      "grad_norm": 0.33061739802360535,
      "learning_rate": 0.0001,
      "loss": 1.5682,
      "step": 1267
    },
    {
      "epoch": 0.30791646430305974,
      "grad_norm": 0.34152230620384216,
      "learning_rate": 0.0001,
      "loss": 1.5829,
      "step": 1268
    },
    {
      "epoch": 0.3081593006313745,
      "grad_norm": 0.36030837893486023,
      "learning_rate": 0.0001,
      "loss": 1.696,
      "step": 1269
    },
    {
      "epoch": 0.30840213695968915,
      "grad_norm": 0.3398350179195404,
      "learning_rate": 0.0001,
      "loss": 1.9147,
      "step": 1270
    },
    {
      "epoch": 0.3086449732880039,
      "grad_norm": 0.33502107858657837,
      "learning_rate": 0.0001,
      "loss": 1.696,
      "step": 1271
    },
    {
      "epoch": 0.3088878096163186,
      "grad_norm": 0.34701743721961975,
      "learning_rate": 0.0001,
      "loss": 1.625,
      "step": 1272
    },
    {
      "epoch": 0.3091306459446333,
      "grad_norm": 0.3735635578632355,
      "learning_rate": 0.0001,
      "loss": 1.7718,
      "step": 1273
    },
    {
      "epoch": 0.30937348227294803,
      "grad_norm": 0.35327258706092834,
      "learning_rate": 0.0001,
      "loss": 1.7464,
      "step": 1274
    },
    {
      "epoch": 0.30961631860126276,
      "grad_norm": 0.35081592202186584,
      "learning_rate": 0.0001,
      "loss": 1.7941,
      "step": 1275
    },
    {
      "epoch": 0.30985915492957744,
      "grad_norm": 0.32337236404418945,
      "learning_rate": 0.0001,
      "loss": 1.5163,
      "step": 1276
    },
    {
      "epoch": 0.3101019912578922,
      "grad_norm": 0.32143720984458923,
      "learning_rate": 0.0001,
      "loss": 1.5397,
      "step": 1277
    },
    {
      "epoch": 0.3103448275862069,
      "grad_norm": 0.3316914439201355,
      "learning_rate": 0.0001,
      "loss": 1.7304,
      "step": 1278
    },
    {
      "epoch": 0.31058766391452164,
      "grad_norm": 0.336727112531662,
      "learning_rate": 0.0001,
      "loss": 1.6108,
      "step": 1279
    },
    {
      "epoch": 0.3108305002428363,
      "grad_norm": 0.33665451407432556,
      "learning_rate": 0.0001,
      "loss": 1.6865,
      "step": 1280
    },
    {
      "epoch": 0.31107333657115105,
      "grad_norm": 0.34888777136802673,
      "learning_rate": 0.0001,
      "loss": 1.707,
      "step": 1281
    },
    {
      "epoch": 0.3113161728994658,
      "grad_norm": 0.3641833961009979,
      "learning_rate": 0.0001,
      "loss": 1.9235,
      "step": 1282
    },
    {
      "epoch": 0.31155900922778046,
      "grad_norm": 0.3506174087524414,
      "learning_rate": 0.0001,
      "loss": 1.7019,
      "step": 1283
    },
    {
      "epoch": 0.3118018455560952,
      "grad_norm": 0.33513206243515015,
      "learning_rate": 0.0001,
      "loss": 1.7835,
      "step": 1284
    },
    {
      "epoch": 0.3120446818844099,
      "grad_norm": 0.31959009170532227,
      "learning_rate": 0.0001,
      "loss": 1.6274,
      "step": 1285
    },
    {
      "epoch": 0.3122875182127246,
      "grad_norm": 0.33899250626564026,
      "learning_rate": 0.0001,
      "loss": 1.6284,
      "step": 1286
    },
    {
      "epoch": 0.31253035454103933,
      "grad_norm": 0.32353276014328003,
      "learning_rate": 0.0001,
      "loss": 1.7145,
      "step": 1287
    },
    {
      "epoch": 0.31277319086935407,
      "grad_norm": 0.3442525565624237,
      "learning_rate": 0.0001,
      "loss": 1.8363,
      "step": 1288
    },
    {
      "epoch": 0.31301602719766874,
      "grad_norm": 0.3439512550830841,
      "learning_rate": 0.0001,
      "loss": 1.76,
      "step": 1289
    },
    {
      "epoch": 0.3132588635259835,
      "grad_norm": 0.32078948616981506,
      "learning_rate": 0.0001,
      "loss": 1.7175,
      "step": 1290
    },
    {
      "epoch": 0.3135016998542982,
      "grad_norm": 0.33406755328178406,
      "learning_rate": 0.0001,
      "loss": 1.8142,
      "step": 1291
    },
    {
      "epoch": 0.31374453618261294,
      "grad_norm": 0.36002907156944275,
      "learning_rate": 0.0001,
      "loss": 1.7645,
      "step": 1292
    },
    {
      "epoch": 0.3139873725109276,
      "grad_norm": 0.33843985199928284,
      "learning_rate": 0.0001,
      "loss": 1.6883,
      "step": 1293
    },
    {
      "epoch": 0.31423020883924235,
      "grad_norm": 0.33737096190452576,
      "learning_rate": 0.0001,
      "loss": 1.7249,
      "step": 1294
    },
    {
      "epoch": 0.3144730451675571,
      "grad_norm": 0.36410626769065857,
      "learning_rate": 0.0001,
      "loss": 1.7059,
      "step": 1295
    },
    {
      "epoch": 0.31471588149587176,
      "grad_norm": 0.35024207830429077,
      "learning_rate": 0.0001,
      "loss": 1.8031,
      "step": 1296
    },
    {
      "epoch": 0.3149587178241865,
      "grad_norm": 0.35997697710990906,
      "learning_rate": 0.0001,
      "loss": 1.8436,
      "step": 1297
    },
    {
      "epoch": 0.31520155415250123,
      "grad_norm": 0.34690675139427185,
      "learning_rate": 0.0001,
      "loss": 1.8562,
      "step": 1298
    },
    {
      "epoch": 0.3154443904808159,
      "grad_norm": 0.3096066415309906,
      "learning_rate": 0.0001,
      "loss": 1.3673,
      "step": 1299
    },
    {
      "epoch": 0.31568722680913064,
      "grad_norm": 0.3499063551425934,
      "learning_rate": 0.0001,
      "loss": 1.7165,
      "step": 1300
    },
    {
      "epoch": 0.31593006313744537,
      "grad_norm": 0.3726784288883209,
      "learning_rate": 0.0001,
      "loss": 1.9799,
      "step": 1301
    },
    {
      "epoch": 0.3161728994657601,
      "grad_norm": 0.32575857639312744,
      "learning_rate": 0.0001,
      "loss": 1.6414,
      "step": 1302
    },
    {
      "epoch": 0.3164157357940748,
      "grad_norm": 0.3620957136154175,
      "learning_rate": 0.0001,
      "loss": 1.7306,
      "step": 1303
    },
    {
      "epoch": 0.3166585721223895,
      "grad_norm": 0.3882763683795929,
      "learning_rate": 0.0001,
      "loss": 1.8216,
      "step": 1304
    },
    {
      "epoch": 0.31690140845070425,
      "grad_norm": 0.37228524684906006,
      "learning_rate": 0.0001,
      "loss": 1.9724,
      "step": 1305
    },
    {
      "epoch": 0.3171442447790189,
      "grad_norm": 0.3398188650608063,
      "learning_rate": 0.0001,
      "loss": 1.7181,
      "step": 1306
    },
    {
      "epoch": 0.31738708110733366,
      "grad_norm": 0.3388833701610565,
      "learning_rate": 0.0001,
      "loss": 1.6542,
      "step": 1307
    },
    {
      "epoch": 0.3176299174356484,
      "grad_norm": 0.3429461717605591,
      "learning_rate": 0.0001,
      "loss": 1.6822,
      "step": 1308
    },
    {
      "epoch": 0.31787275376396307,
      "grad_norm": 0.3804921805858612,
      "learning_rate": 0.0001,
      "loss": 2.0971,
      "step": 1309
    },
    {
      "epoch": 0.3181155900922778,
      "grad_norm": 0.3579254448413849,
      "learning_rate": 0.0001,
      "loss": 1.7412,
      "step": 1310
    },
    {
      "epoch": 0.31835842642059253,
      "grad_norm": 0.3545834422111511,
      "learning_rate": 0.0001,
      "loss": 1.7221,
      "step": 1311
    },
    {
      "epoch": 0.3186012627489072,
      "grad_norm": 0.3544105887413025,
      "learning_rate": 0.0001,
      "loss": 1.8662,
      "step": 1312
    },
    {
      "epoch": 0.31884409907722194,
      "grad_norm": 0.35376471281051636,
      "learning_rate": 0.0001,
      "loss": 1.7822,
      "step": 1313
    },
    {
      "epoch": 0.3190869354055367,
      "grad_norm": 0.34648221731185913,
      "learning_rate": 0.0001,
      "loss": 1.6543,
      "step": 1314
    },
    {
      "epoch": 0.3193297717338514,
      "grad_norm": 0.35739588737487793,
      "learning_rate": 0.0001,
      "loss": 1.7637,
      "step": 1315
    },
    {
      "epoch": 0.3195726080621661,
      "grad_norm": 0.3418010473251343,
      "learning_rate": 0.0001,
      "loss": 1.7071,
      "step": 1316
    },
    {
      "epoch": 0.3198154443904808,
      "grad_norm": 0.34581372141838074,
      "learning_rate": 0.0001,
      "loss": 1.7706,
      "step": 1317
    },
    {
      "epoch": 0.32005828071879555,
      "grad_norm": 0.3538050651550293,
      "learning_rate": 0.0001,
      "loss": 1.6932,
      "step": 1318
    },
    {
      "epoch": 0.32030111704711023,
      "grad_norm": 0.3454187214374542,
      "learning_rate": 0.0001,
      "loss": 1.719,
      "step": 1319
    },
    {
      "epoch": 0.32054395337542496,
      "grad_norm": 0.35107696056365967,
      "learning_rate": 0.0001,
      "loss": 1.7574,
      "step": 1320
    },
    {
      "epoch": 0.3207867897037397,
      "grad_norm": 0.3261638879776001,
      "learning_rate": 0.0001,
      "loss": 1.6096,
      "step": 1321
    },
    {
      "epoch": 0.32102962603205437,
      "grad_norm": 0.36752843856811523,
      "learning_rate": 0.0001,
      "loss": 1.8988,
      "step": 1322
    },
    {
      "epoch": 0.3212724623603691,
      "grad_norm": 0.3847307562828064,
      "learning_rate": 0.0001,
      "loss": 1.9061,
      "step": 1323
    },
    {
      "epoch": 0.32151529868868384,
      "grad_norm": 0.3572913110256195,
      "learning_rate": 0.0001,
      "loss": 1.7083,
      "step": 1324
    },
    {
      "epoch": 0.32175813501699857,
      "grad_norm": 0.34111347794532776,
      "learning_rate": 0.0001,
      "loss": 1.6437,
      "step": 1325
    },
    {
      "epoch": 0.32200097134531325,
      "grad_norm": 0.346483439207077,
      "learning_rate": 0.0001,
      "loss": 1.6687,
      "step": 1326
    },
    {
      "epoch": 0.322243807673628,
      "grad_norm": 0.35447895526885986,
      "learning_rate": 0.0001,
      "loss": 1.83,
      "step": 1327
    },
    {
      "epoch": 0.3224866440019427,
      "grad_norm": 0.3246619999408722,
      "learning_rate": 0.0001,
      "loss": 1.5981,
      "step": 1328
    },
    {
      "epoch": 0.3227294803302574,
      "grad_norm": 0.3411337435245514,
      "learning_rate": 0.0001,
      "loss": 1.8093,
      "step": 1329
    },
    {
      "epoch": 0.3229723166585721,
      "grad_norm": 0.33755388855934143,
      "learning_rate": 0.0001,
      "loss": 1.7882,
      "step": 1330
    },
    {
      "epoch": 0.32321515298688686,
      "grad_norm": 0.3567180335521698,
      "learning_rate": 0.0001,
      "loss": 1.6274,
      "step": 1331
    },
    {
      "epoch": 0.32345798931520153,
      "grad_norm": 0.3355996906757355,
      "learning_rate": 0.0001,
      "loss": 1.6726,
      "step": 1332
    },
    {
      "epoch": 0.32370082564351627,
      "grad_norm": 0.34801527857780457,
      "learning_rate": 0.0001,
      "loss": 1.651,
      "step": 1333
    },
    {
      "epoch": 0.323943661971831,
      "grad_norm": 0.34723374247550964,
      "learning_rate": 0.0001,
      "loss": 1.8538,
      "step": 1334
    },
    {
      "epoch": 0.3241864983001457,
      "grad_norm": 0.33615392446517944,
      "learning_rate": 0.0001,
      "loss": 1.5849,
      "step": 1335
    },
    {
      "epoch": 0.3244293346284604,
      "grad_norm": 0.3725658655166626,
      "learning_rate": 0.0001,
      "loss": 1.8235,
      "step": 1336
    },
    {
      "epoch": 0.32467217095677514,
      "grad_norm": 0.3329296410083771,
      "learning_rate": 0.0001,
      "loss": 1.7168,
      "step": 1337
    },
    {
      "epoch": 0.3249150072850899,
      "grad_norm": 0.3577701449394226,
      "learning_rate": 0.0001,
      "loss": 1.7251,
      "step": 1338
    },
    {
      "epoch": 0.32515784361340455,
      "grad_norm": 0.33400166034698486,
      "learning_rate": 0.0001,
      "loss": 1.7042,
      "step": 1339
    },
    {
      "epoch": 0.3254006799417193,
      "grad_norm": 0.3208601474761963,
      "learning_rate": 0.0001,
      "loss": 1.5577,
      "step": 1340
    },
    {
      "epoch": 0.325643516270034,
      "grad_norm": 0.35561156272888184,
      "learning_rate": 0.0001,
      "loss": 1.7421,
      "step": 1341
    },
    {
      "epoch": 0.3258863525983487,
      "grad_norm": 0.3534224331378937,
      "learning_rate": 0.0001,
      "loss": 1.6784,
      "step": 1342
    },
    {
      "epoch": 0.3261291889266634,
      "grad_norm": 0.34640082716941833,
      "learning_rate": 0.0001,
      "loss": 1.7314,
      "step": 1343
    },
    {
      "epoch": 0.32637202525497816,
      "grad_norm": 0.35193905234336853,
      "learning_rate": 0.0001,
      "loss": 1.7245,
      "step": 1344
    },
    {
      "epoch": 0.32661486158329284,
      "grad_norm": 0.3425352871417999,
      "learning_rate": 0.0001,
      "loss": 1.7848,
      "step": 1345
    },
    {
      "epoch": 0.32685769791160757,
      "grad_norm": 0.343161016702652,
      "learning_rate": 0.0001,
      "loss": 1.6993,
      "step": 1346
    },
    {
      "epoch": 0.3271005342399223,
      "grad_norm": 0.36429017782211304,
      "learning_rate": 0.0001,
      "loss": 1.7049,
      "step": 1347
    },
    {
      "epoch": 0.32734337056823704,
      "grad_norm": 0.3147446811199188,
      "learning_rate": 0.0001,
      "loss": 1.6569,
      "step": 1348
    },
    {
      "epoch": 0.3275862068965517,
      "grad_norm": 0.3437577188014984,
      "learning_rate": 0.0001,
      "loss": 1.6792,
      "step": 1349
    },
    {
      "epoch": 0.32782904322486645,
      "grad_norm": 0.37101826071739197,
      "learning_rate": 0.0001,
      "loss": 1.6529,
      "step": 1350
    },
    {
      "epoch": 0.3280718795531812,
      "grad_norm": 0.36280542612075806,
      "learning_rate": 0.0001,
      "loss": 1.7888,
      "step": 1351
    },
    {
      "epoch": 0.32831471588149586,
      "grad_norm": 0.3654901087284088,
      "learning_rate": 0.0001,
      "loss": 1.9293,
      "step": 1352
    },
    {
      "epoch": 0.3285575522098106,
      "grad_norm": 0.3405821621417999,
      "learning_rate": 0.0001,
      "loss": 1.5597,
      "step": 1353
    },
    {
      "epoch": 0.3288003885381253,
      "grad_norm": 0.33608195185661316,
      "learning_rate": 0.0001,
      "loss": 1.7485,
      "step": 1354
    },
    {
      "epoch": 0.32904322486644,
      "grad_norm": 0.3530266284942627,
      "learning_rate": 0.0001,
      "loss": 1.7163,
      "step": 1355
    },
    {
      "epoch": 0.32928606119475473,
      "grad_norm": 0.35584160685539246,
      "learning_rate": 0.0001,
      "loss": 1.6821,
      "step": 1356
    },
    {
      "epoch": 0.32952889752306946,
      "grad_norm": 0.33496901392936707,
      "learning_rate": 0.0001,
      "loss": 1.5494,
      "step": 1357
    },
    {
      "epoch": 0.32977173385138414,
      "grad_norm": 0.33359116315841675,
      "learning_rate": 0.0001,
      "loss": 1.7098,
      "step": 1358
    },
    {
      "epoch": 0.3300145701796989,
      "grad_norm": 0.34813758730888367,
      "learning_rate": 0.0001,
      "loss": 1.8438,
      "step": 1359
    },
    {
      "epoch": 0.3302574065080136,
      "grad_norm": 0.37358787655830383,
      "learning_rate": 0.0001,
      "loss": 1.8214,
      "step": 1360
    },
    {
      "epoch": 0.33050024283632834,
      "grad_norm": 0.3805057406425476,
      "learning_rate": 0.0001,
      "loss": 1.8504,
      "step": 1361
    },
    {
      "epoch": 0.330743079164643,
      "grad_norm": 0.35961398482322693,
      "learning_rate": 0.0001,
      "loss": 1.8235,
      "step": 1362
    },
    {
      "epoch": 0.33098591549295775,
      "grad_norm": 0.37374040484428406,
      "learning_rate": 0.0001,
      "loss": 1.8868,
      "step": 1363
    },
    {
      "epoch": 0.3312287518212725,
      "grad_norm": 0.33554673194885254,
      "learning_rate": 0.0001,
      "loss": 1.6591,
      "step": 1364
    },
    {
      "epoch": 0.33147158814958716,
      "grad_norm": 0.3829343318939209,
      "learning_rate": 0.0001,
      "loss": 1.9936,
      "step": 1365
    },
    {
      "epoch": 0.3317144244779019,
      "grad_norm": 0.35701921582221985,
      "learning_rate": 0.0001,
      "loss": 1.7347,
      "step": 1366
    },
    {
      "epoch": 0.3319572608062166,
      "grad_norm": 0.35286015272140503,
      "learning_rate": 0.0001,
      "loss": 1.7225,
      "step": 1367
    },
    {
      "epoch": 0.3322000971345313,
      "grad_norm": 0.3747299015522003,
      "learning_rate": 0.0001,
      "loss": 1.9376,
      "step": 1368
    },
    {
      "epoch": 0.33244293346284604,
      "grad_norm": 0.3595302402973175,
      "learning_rate": 0.0001,
      "loss": 1.7511,
      "step": 1369
    },
    {
      "epoch": 0.33268576979116077,
      "grad_norm": 0.36371472477912903,
      "learning_rate": 0.0001,
      "loss": 1.6936,
      "step": 1370
    },
    {
      "epoch": 0.3329286061194755,
      "grad_norm": 0.3578799068927765,
      "learning_rate": 0.0001,
      "loss": 1.7735,
      "step": 1371
    },
    {
      "epoch": 0.3331714424477902,
      "grad_norm": 0.3342965245246887,
      "learning_rate": 0.0001,
      "loss": 1.6468,
      "step": 1372
    },
    {
      "epoch": 0.3334142787761049,
      "grad_norm": 0.41231048107147217,
      "learning_rate": 0.0001,
      "loss": 1.7679,
      "step": 1373
    },
    {
      "epoch": 0.33365711510441964,
      "grad_norm": 0.3757196068763733,
      "learning_rate": 0.0001,
      "loss": 1.8346,
      "step": 1374
    },
    {
      "epoch": 0.3338999514327343,
      "grad_norm": 0.3343157470226288,
      "learning_rate": 0.0001,
      "loss": 1.5956,
      "step": 1375
    },
    {
      "epoch": 0.33414278776104905,
      "grad_norm": 0.3424040675163269,
      "learning_rate": 0.0001,
      "loss": 1.7438,
      "step": 1376
    },
    {
      "epoch": 0.3343856240893638,
      "grad_norm": 0.35042518377304077,
      "learning_rate": 0.0001,
      "loss": 1.6813,
      "step": 1377
    },
    {
      "epoch": 0.33462846041767846,
      "grad_norm": 0.3496730327606201,
      "learning_rate": 0.0001,
      "loss": 1.7828,
      "step": 1378
    },
    {
      "epoch": 0.3348712967459932,
      "grad_norm": 0.3514433801174164,
      "learning_rate": 0.0001,
      "loss": 1.6792,
      "step": 1379
    },
    {
      "epoch": 0.33511413307430793,
      "grad_norm": 0.37553372979164124,
      "learning_rate": 0.0001,
      "loss": 1.8507,
      "step": 1380
    },
    {
      "epoch": 0.3353569694026226,
      "grad_norm": 0.3467164933681488,
      "learning_rate": 0.0001,
      "loss": 1.7598,
      "step": 1381
    },
    {
      "epoch": 0.33559980573093734,
      "grad_norm": 0.3462772071361542,
      "learning_rate": 0.0001,
      "loss": 1.8121,
      "step": 1382
    },
    {
      "epoch": 0.3358426420592521,
      "grad_norm": 0.3283761441707611,
      "learning_rate": 0.0001,
      "loss": 1.6598,
      "step": 1383
    },
    {
      "epoch": 0.3360854783875668,
      "grad_norm": 0.34071117639541626,
      "learning_rate": 0.0001,
      "loss": 1.6984,
      "step": 1384
    },
    {
      "epoch": 0.3363283147158815,
      "grad_norm": 0.3350028395652771,
      "learning_rate": 0.0001,
      "loss": 1.7327,
      "step": 1385
    },
    {
      "epoch": 0.3365711510441962,
      "grad_norm": 0.33372920751571655,
      "learning_rate": 0.0001,
      "loss": 1.7262,
      "step": 1386
    },
    {
      "epoch": 0.33681398737251095,
      "grad_norm": 0.36559152603149414,
      "learning_rate": 0.0001,
      "loss": 1.6677,
      "step": 1387
    },
    {
      "epoch": 0.3370568237008256,
      "grad_norm": 0.3441181480884552,
      "learning_rate": 0.0001,
      "loss": 1.6399,
      "step": 1388
    },
    {
      "epoch": 0.33729966002914036,
      "grad_norm": 0.37654390931129456,
      "learning_rate": 0.0001,
      "loss": 1.8389,
      "step": 1389
    },
    {
      "epoch": 0.3375424963574551,
      "grad_norm": 0.3329363465309143,
      "learning_rate": 0.0001,
      "loss": 1.6735,
      "step": 1390
    },
    {
      "epoch": 0.33778533268576977,
      "grad_norm": 0.34903109073638916,
      "learning_rate": 0.0001,
      "loss": 1.7471,
      "step": 1391
    },
    {
      "epoch": 0.3380281690140845,
      "grad_norm": 0.31249791383743286,
      "learning_rate": 0.0001,
      "loss": 1.659,
      "step": 1392
    },
    {
      "epoch": 0.33827100534239923,
      "grad_norm": 0.3493087887763977,
      "learning_rate": 0.0001,
      "loss": 1.7833,
      "step": 1393
    },
    {
      "epoch": 0.33851384167071397,
      "grad_norm": 0.3338823616504669,
      "learning_rate": 0.0001,
      "loss": 1.7842,
      "step": 1394
    },
    {
      "epoch": 0.33875667799902864,
      "grad_norm": 0.33419299125671387,
      "learning_rate": 0.0001,
      "loss": 1.7272,
      "step": 1395
    },
    {
      "epoch": 0.3389995143273434,
      "grad_norm": 0.35470160841941833,
      "learning_rate": 0.0001,
      "loss": 1.5258,
      "step": 1396
    },
    {
      "epoch": 0.3392423506556581,
      "grad_norm": 0.33839255571365356,
      "learning_rate": 0.0001,
      "loss": 1.7123,
      "step": 1397
    },
    {
      "epoch": 0.3394851869839728,
      "grad_norm": 0.36841094493865967,
      "learning_rate": 0.0001,
      "loss": 1.8869,
      "step": 1398
    },
    {
      "epoch": 0.3397280233122875,
      "grad_norm": 0.3372727334499359,
      "learning_rate": 0.0001,
      "loss": 1.6797,
      "step": 1399
    },
    {
      "epoch": 0.33997085964060225,
      "grad_norm": 0.30915823578834534,
      "learning_rate": 0.0001,
      "loss": 1.5637,
      "step": 1400
    },
    {
      "epoch": 0.34021369596891693,
      "grad_norm": 0.31683653593063354,
      "learning_rate": 0.0001,
      "loss": 1.6456,
      "step": 1401
    },
    {
      "epoch": 0.34045653229723166,
      "grad_norm": 0.32861605286598206,
      "learning_rate": 0.0001,
      "loss": 1.7809,
      "step": 1402
    },
    {
      "epoch": 0.3406993686255464,
      "grad_norm": 0.3321719765663147,
      "learning_rate": 0.0001,
      "loss": 1.7957,
      "step": 1403
    },
    {
      "epoch": 0.34094220495386107,
      "grad_norm": 0.34554940462112427,
      "learning_rate": 0.0001,
      "loss": 1.6703,
      "step": 1404
    },
    {
      "epoch": 0.3411850412821758,
      "grad_norm": 0.4025746285915375,
      "learning_rate": 0.0001,
      "loss": 1.8275,
      "step": 1405
    },
    {
      "epoch": 0.34142787761049054,
      "grad_norm": 0.34001925587654114,
      "learning_rate": 0.0001,
      "loss": 1.6531,
      "step": 1406
    },
    {
      "epoch": 0.34167071393880527,
      "grad_norm": 0.38130107522010803,
      "learning_rate": 0.0001,
      "loss": 1.8164,
      "step": 1407
    },
    {
      "epoch": 0.34191355026711995,
      "grad_norm": 0.3359420597553253,
      "learning_rate": 0.0001,
      "loss": 1.5757,
      "step": 1408
    },
    {
      "epoch": 0.3421563865954347,
      "grad_norm": 0.32789063453674316,
      "learning_rate": 0.0001,
      "loss": 1.5708,
      "step": 1409
    },
    {
      "epoch": 0.3423992229237494,
      "grad_norm": 0.3676958680152893,
      "learning_rate": 0.0001,
      "loss": 1.7915,
      "step": 1410
    },
    {
      "epoch": 0.3426420592520641,
      "grad_norm": 0.376945436000824,
      "learning_rate": 0.0001,
      "loss": 1.8536,
      "step": 1411
    },
    {
      "epoch": 0.3428848955803788,
      "grad_norm": 0.32787764072418213,
      "learning_rate": 0.0001,
      "loss": 1.7462,
      "step": 1412
    },
    {
      "epoch": 0.34312773190869356,
      "grad_norm": 0.40259850025177,
      "learning_rate": 0.0001,
      "loss": 1.8247,
      "step": 1413
    },
    {
      "epoch": 0.34337056823700823,
      "grad_norm": 0.36955440044403076,
      "learning_rate": 0.0001,
      "loss": 1.8181,
      "step": 1414
    },
    {
      "epoch": 0.34361340456532297,
      "grad_norm": 0.336770623922348,
      "learning_rate": 0.0001,
      "loss": 1.7475,
      "step": 1415
    },
    {
      "epoch": 0.3438562408936377,
      "grad_norm": 0.360385537147522,
      "learning_rate": 0.0001,
      "loss": 1.7349,
      "step": 1416
    },
    {
      "epoch": 0.3440990772219524,
      "grad_norm": 0.36137905716896057,
      "learning_rate": 0.0001,
      "loss": 1.7219,
      "step": 1417
    },
    {
      "epoch": 0.3443419135502671,
      "grad_norm": 0.3530389368534088,
      "learning_rate": 0.0001,
      "loss": 1.7545,
      "step": 1418
    },
    {
      "epoch": 0.34458474987858184,
      "grad_norm": 0.3292929232120514,
      "learning_rate": 0.0001,
      "loss": 1.7318,
      "step": 1419
    },
    {
      "epoch": 0.3448275862068966,
      "grad_norm": 0.3354680836200714,
      "learning_rate": 0.0001,
      "loss": 1.7757,
      "step": 1420
    },
    {
      "epoch": 0.34507042253521125,
      "grad_norm": 0.3158077895641327,
      "learning_rate": 0.0001,
      "loss": 1.4464,
      "step": 1421
    },
    {
      "epoch": 0.345313258863526,
      "grad_norm": 0.3723140060901642,
      "learning_rate": 0.0001,
      "loss": 1.7779,
      "step": 1422
    },
    {
      "epoch": 0.3455560951918407,
      "grad_norm": 0.3340955078601837,
      "learning_rate": 0.0001,
      "loss": 1.8542,
      "step": 1423
    },
    {
      "epoch": 0.3457989315201554,
      "grad_norm": 0.3777885437011719,
      "learning_rate": 0.0001,
      "loss": 1.7496,
      "step": 1424
    },
    {
      "epoch": 0.3460417678484701,
      "grad_norm": 0.352083295583725,
      "learning_rate": 0.0001,
      "loss": 1.8658,
      "step": 1425
    },
    {
      "epoch": 0.34628460417678486,
      "grad_norm": 0.3653123378753662,
      "learning_rate": 0.0001,
      "loss": 1.7476,
      "step": 1426
    },
    {
      "epoch": 0.34652744050509954,
      "grad_norm": 0.35643982887268066,
      "learning_rate": 0.0001,
      "loss": 1.6245,
      "step": 1427
    },
    {
      "epoch": 0.34677027683341427,
      "grad_norm": 0.3376123309135437,
      "learning_rate": 0.0001,
      "loss": 1.7228,
      "step": 1428
    },
    {
      "epoch": 0.347013113161729,
      "grad_norm": 0.3336902856826782,
      "learning_rate": 0.0001,
      "loss": 1.5102,
      "step": 1429
    },
    {
      "epoch": 0.34725594949004374,
      "grad_norm": 0.35413724184036255,
      "learning_rate": 0.0001,
      "loss": 1.7126,
      "step": 1430
    },
    {
      "epoch": 0.3474987858183584,
      "grad_norm": 0.3520013093948364,
      "learning_rate": 0.0001,
      "loss": 1.7388,
      "step": 1431
    },
    {
      "epoch": 0.34774162214667315,
      "grad_norm": 0.3354269564151764,
      "learning_rate": 0.0001,
      "loss": 1.7037,
      "step": 1432
    },
    {
      "epoch": 0.3479844584749879,
      "grad_norm": 0.37913069128990173,
      "learning_rate": 0.0001,
      "loss": 1.615,
      "step": 1433
    },
    {
      "epoch": 0.34822729480330256,
      "grad_norm": 0.35012054443359375,
      "learning_rate": 0.0001,
      "loss": 1.7384,
      "step": 1434
    },
    {
      "epoch": 0.3484701311316173,
      "grad_norm": 0.32427114248275757,
      "learning_rate": 0.0001,
      "loss": 1.5869,
      "step": 1435
    },
    {
      "epoch": 0.348712967459932,
      "grad_norm": 0.35008156299591064,
      "learning_rate": 0.0001,
      "loss": 1.7277,
      "step": 1436
    },
    {
      "epoch": 0.3489558037882467,
      "grad_norm": 0.3610100746154785,
      "learning_rate": 0.0001,
      "loss": 1.654,
      "step": 1437
    },
    {
      "epoch": 0.34919864011656143,
      "grad_norm": 0.34857314825057983,
      "learning_rate": 0.0001,
      "loss": 1.748,
      "step": 1438
    },
    {
      "epoch": 0.34944147644487616,
      "grad_norm": 0.3715338706970215,
      "learning_rate": 0.0001,
      "loss": 1.5845,
      "step": 1439
    },
    {
      "epoch": 0.34968431277319084,
      "grad_norm": 0.3498457074165344,
      "learning_rate": 0.0001,
      "loss": 1.78,
      "step": 1440
    },
    {
      "epoch": 0.3499271491015056,
      "grad_norm": 0.33495819568634033,
      "learning_rate": 0.0001,
      "loss": 1.6412,
      "step": 1441
    },
    {
      "epoch": 0.3501699854298203,
      "grad_norm": 0.38084495067596436,
      "learning_rate": 0.0001,
      "loss": 1.5973,
      "step": 1442
    },
    {
      "epoch": 0.35041282175813504,
      "grad_norm": 0.348736047744751,
      "learning_rate": 0.0001,
      "loss": 1.739,
      "step": 1443
    },
    {
      "epoch": 0.3506556580864497,
      "grad_norm": 0.33308863639831543,
      "learning_rate": 0.0001,
      "loss": 1.7768,
      "step": 1444
    },
    {
      "epoch": 0.35089849441476445,
      "grad_norm": 0.3505055606365204,
      "learning_rate": 0.0001,
      "loss": 1.7172,
      "step": 1445
    },
    {
      "epoch": 0.3511413307430792,
      "grad_norm": 0.3615930676460266,
      "learning_rate": 0.0001,
      "loss": 1.8188,
      "step": 1446
    },
    {
      "epoch": 0.35138416707139386,
      "grad_norm": 0.33312997221946716,
      "learning_rate": 0.0001,
      "loss": 1.6119,
      "step": 1447
    },
    {
      "epoch": 0.3516270033997086,
      "grad_norm": 0.34861356019973755,
      "learning_rate": 0.0001,
      "loss": 1.7229,
      "step": 1448
    },
    {
      "epoch": 0.3518698397280233,
      "grad_norm": 0.34396806359291077,
      "learning_rate": 0.0001,
      "loss": 1.6834,
      "step": 1449
    },
    {
      "epoch": 0.352112676056338,
      "grad_norm": 0.3472231924533844,
      "learning_rate": 0.0001,
      "loss": 1.6941,
      "step": 1450
    },
    {
      "epoch": 0.35235551238465274,
      "grad_norm": 0.33803194761276245,
      "learning_rate": 0.0001,
      "loss": 1.663,
      "step": 1451
    },
    {
      "epoch": 0.35259834871296747,
      "grad_norm": 0.3511747419834137,
      "learning_rate": 0.0001,
      "loss": 1.7896,
      "step": 1452
    },
    {
      "epoch": 0.3528411850412822,
      "grad_norm": 0.3400138318538666,
      "learning_rate": 0.0001,
      "loss": 1.7776,
      "step": 1453
    },
    {
      "epoch": 0.3530840213695969,
      "grad_norm": 0.34329378604888916,
      "learning_rate": 0.0001,
      "loss": 1.7485,
      "step": 1454
    },
    {
      "epoch": 0.3533268576979116,
      "grad_norm": 0.3581506311893463,
      "learning_rate": 0.0001,
      "loss": 1.8759,
      "step": 1455
    },
    {
      "epoch": 0.35356969402622634,
      "grad_norm": 0.3372110426425934,
      "learning_rate": 0.0001,
      "loss": 1.7721,
      "step": 1456
    },
    {
      "epoch": 0.353812530354541,
      "grad_norm": 0.33393704891204834,
      "learning_rate": 0.0001,
      "loss": 1.7497,
      "step": 1457
    },
    {
      "epoch": 0.35405536668285575,
      "grad_norm": 0.3570740818977356,
      "learning_rate": 0.0001,
      "loss": 1.8485,
      "step": 1458
    },
    {
      "epoch": 0.3542982030111705,
      "grad_norm": 0.3288159668445587,
      "learning_rate": 0.0001,
      "loss": 1.7145,
      "step": 1459
    },
    {
      "epoch": 0.35454103933948516,
      "grad_norm": 0.352760910987854,
      "learning_rate": 0.0001,
      "loss": 1.7081,
      "step": 1460
    },
    {
      "epoch": 0.3547838756677999,
      "grad_norm": 0.33334290981292725,
      "learning_rate": 0.0001,
      "loss": 1.6774,
      "step": 1461
    },
    {
      "epoch": 0.35502671199611463,
      "grad_norm": 0.36578717827796936,
      "learning_rate": 0.0001,
      "loss": 1.6655,
      "step": 1462
    },
    {
      "epoch": 0.3552695483244293,
      "grad_norm": 0.34803271293640137,
      "learning_rate": 0.0001,
      "loss": 1.8048,
      "step": 1463
    },
    {
      "epoch": 0.35551238465274404,
      "grad_norm": 0.34706413745880127,
      "learning_rate": 0.0001,
      "loss": 1.666,
      "step": 1464
    },
    {
      "epoch": 0.3557552209810588,
      "grad_norm": 0.3548772633075714,
      "learning_rate": 0.0001,
      "loss": 1.7261,
      "step": 1465
    },
    {
      "epoch": 0.3559980573093735,
      "grad_norm": 0.34797123074531555,
      "learning_rate": 0.0001,
      "loss": 1.7126,
      "step": 1466
    },
    {
      "epoch": 0.3562408936376882,
      "grad_norm": 0.3482211232185364,
      "learning_rate": 0.0001,
      "loss": 1.703,
      "step": 1467
    },
    {
      "epoch": 0.3564837299660029,
      "grad_norm": 0.380947083234787,
      "learning_rate": 0.0001,
      "loss": 1.8435,
      "step": 1468
    },
    {
      "epoch": 0.35672656629431765,
      "grad_norm": 0.33425045013427734,
      "learning_rate": 0.0001,
      "loss": 1.5974,
      "step": 1469
    },
    {
      "epoch": 0.3569694026226323,
      "grad_norm": 0.33053338527679443,
      "learning_rate": 0.0001,
      "loss": 1.5485,
      "step": 1470
    },
    {
      "epoch": 0.35721223895094706,
      "grad_norm": 0.33164921402931213,
      "learning_rate": 0.0001,
      "loss": 1.7375,
      "step": 1471
    },
    {
      "epoch": 0.3574550752792618,
      "grad_norm": 0.3542589843273163,
      "learning_rate": 0.0001,
      "loss": 1.8596,
      "step": 1472
    },
    {
      "epoch": 0.35769791160757647,
      "grad_norm": 0.36848315596580505,
      "learning_rate": 0.0001,
      "loss": 1.8971,
      "step": 1473
    },
    {
      "epoch": 0.3579407479358912,
      "grad_norm": 0.3263801336288452,
      "learning_rate": 0.0001,
      "loss": 1.6098,
      "step": 1474
    },
    {
      "epoch": 0.35818358426420593,
      "grad_norm": 0.3408321142196655,
      "learning_rate": 0.0001,
      "loss": 1.7054,
      "step": 1475
    },
    {
      "epoch": 0.35842642059252067,
      "grad_norm": 0.3380805552005768,
      "learning_rate": 0.0001,
      "loss": 1.7244,
      "step": 1476
    },
    {
      "epoch": 0.35866925692083534,
      "grad_norm": 0.3343592882156372,
      "learning_rate": 0.0001,
      "loss": 1.6569,
      "step": 1477
    },
    {
      "epoch": 0.3589120932491501,
      "grad_norm": 0.3570973873138428,
      "learning_rate": 0.0001,
      "loss": 1.7114,
      "step": 1478
    },
    {
      "epoch": 0.3591549295774648,
      "grad_norm": 0.3254868984222412,
      "learning_rate": 0.0001,
      "loss": 1.7063,
      "step": 1479
    },
    {
      "epoch": 0.3593977659057795,
      "grad_norm": 0.33075040578842163,
      "learning_rate": 0.0001,
      "loss": 1.6969,
      "step": 1480
    },
    {
      "epoch": 0.3596406022340942,
      "grad_norm": 0.3524262011051178,
      "learning_rate": 0.0001,
      "loss": 1.7907,
      "step": 1481
    },
    {
      "epoch": 0.35988343856240895,
      "grad_norm": 0.36520183086395264,
      "learning_rate": 0.0001,
      "loss": 1.8654,
      "step": 1482
    },
    {
      "epoch": 0.36012627489072363,
      "grad_norm": 0.3641956150531769,
      "learning_rate": 0.0001,
      "loss": 1.8671,
      "step": 1483
    },
    {
      "epoch": 0.36036911121903836,
      "grad_norm": 0.3406789004802704,
      "learning_rate": 0.0001,
      "loss": 1.7531,
      "step": 1484
    },
    {
      "epoch": 0.3606119475473531,
      "grad_norm": 0.33503258228302,
      "learning_rate": 0.0001,
      "loss": 1.6198,
      "step": 1485
    },
    {
      "epoch": 0.3608547838756678,
      "grad_norm": 0.3479795753955841,
      "learning_rate": 0.0001,
      "loss": 1.8708,
      "step": 1486
    },
    {
      "epoch": 0.3610976202039825,
      "grad_norm": 0.33785268664360046,
      "learning_rate": 0.0001,
      "loss": 1.772,
      "step": 1487
    },
    {
      "epoch": 0.36134045653229724,
      "grad_norm": 0.37516918778419495,
      "learning_rate": 0.0001,
      "loss": 1.7923,
      "step": 1488
    },
    {
      "epoch": 0.36158329286061197,
      "grad_norm": 0.35505229234695435,
      "learning_rate": 0.0001,
      "loss": 1.8019,
      "step": 1489
    },
    {
      "epoch": 0.36182612918892665,
      "grad_norm": 0.3542248606681824,
      "learning_rate": 0.0001,
      "loss": 1.7801,
      "step": 1490
    },
    {
      "epoch": 0.3620689655172414,
      "grad_norm": 0.36539554595947266,
      "learning_rate": 0.0001,
      "loss": 1.8357,
      "step": 1491
    },
    {
      "epoch": 0.3623118018455561,
      "grad_norm": 0.33330610394477844,
      "learning_rate": 0.0001,
      "loss": 1.6645,
      "step": 1492
    },
    {
      "epoch": 0.3625546381738708,
      "grad_norm": 0.331411749124527,
      "learning_rate": 0.0001,
      "loss": 1.6139,
      "step": 1493
    },
    {
      "epoch": 0.3627974745021855,
      "grad_norm": 0.35043078660964966,
      "learning_rate": 0.0001,
      "loss": 1.7627,
      "step": 1494
    },
    {
      "epoch": 0.36304031083050026,
      "grad_norm": 0.33841216564178467,
      "learning_rate": 0.0001,
      "loss": 1.6763,
      "step": 1495
    },
    {
      "epoch": 0.36328314715881493,
      "grad_norm": 0.3518892526626587,
      "learning_rate": 0.0001,
      "loss": 1.6667,
      "step": 1496
    },
    {
      "epoch": 0.36352598348712967,
      "grad_norm": 0.3356791138648987,
      "learning_rate": 0.0001,
      "loss": 1.7648,
      "step": 1497
    },
    {
      "epoch": 0.3637688198154444,
      "grad_norm": 0.33415839076042175,
      "learning_rate": 0.0001,
      "loss": 1.6838,
      "step": 1498
    },
    {
      "epoch": 0.36401165614375913,
      "grad_norm": 0.35371869802474976,
      "learning_rate": 0.0001,
      "loss": 1.7046,
      "step": 1499
    },
    {
      "epoch": 0.3642544924720738,
      "grad_norm": 0.3273749053478241,
      "learning_rate": 0.0001,
      "loss": 1.5889,
      "step": 1500
    },
    {
      "epoch": 0.36449732880038854,
      "grad_norm": 0.39279237389564514,
      "learning_rate": 0.0001,
      "loss": 1.8744,
      "step": 1501
    },
    {
      "epoch": 0.3647401651287033,
      "grad_norm": 0.3533165752887726,
      "learning_rate": 0.0001,
      "loss": 1.7483,
      "step": 1502
    },
    {
      "epoch": 0.36498300145701795,
      "grad_norm": 0.34811070561408997,
      "learning_rate": 0.0001,
      "loss": 1.624,
      "step": 1503
    },
    {
      "epoch": 0.3652258377853327,
      "grad_norm": 0.36167362332344055,
      "learning_rate": 0.0001,
      "loss": 1.6942,
      "step": 1504
    },
    {
      "epoch": 0.3654686741136474,
      "grad_norm": 0.33047184348106384,
      "learning_rate": 0.0001,
      "loss": 1.5787,
      "step": 1505
    },
    {
      "epoch": 0.3657115104419621,
      "grad_norm": 0.36502954363822937,
      "learning_rate": 0.0001,
      "loss": 1.76,
      "step": 1506
    },
    {
      "epoch": 0.36595434677027683,
      "grad_norm": 0.34943050146102905,
      "learning_rate": 0.0001,
      "loss": 1.7061,
      "step": 1507
    },
    {
      "epoch": 0.36619718309859156,
      "grad_norm": 0.3463122546672821,
      "learning_rate": 0.0001,
      "loss": 1.8641,
      "step": 1508
    },
    {
      "epoch": 0.36644001942690624,
      "grad_norm": 0.3423214256763458,
      "learning_rate": 0.0001,
      "loss": 1.6902,
      "step": 1509
    },
    {
      "epoch": 0.36668285575522097,
      "grad_norm": 0.3398582339286804,
      "learning_rate": 0.0001,
      "loss": 1.569,
      "step": 1510
    },
    {
      "epoch": 0.3669256920835357,
      "grad_norm": 0.3707105815410614,
      "learning_rate": 0.0001,
      "loss": 1.8673,
      "step": 1511
    },
    {
      "epoch": 0.36716852841185044,
      "grad_norm": 0.32776641845703125,
      "learning_rate": 0.0001,
      "loss": 1.5771,
      "step": 1512
    },
    {
      "epoch": 0.3674113647401651,
      "grad_norm": 0.3616585433483124,
      "learning_rate": 0.0001,
      "loss": 1.7499,
      "step": 1513
    },
    {
      "epoch": 0.36765420106847985,
      "grad_norm": 0.3802650272846222,
      "learning_rate": 0.0001,
      "loss": 1.8077,
      "step": 1514
    },
    {
      "epoch": 0.3678970373967946,
      "grad_norm": 0.3210430145263672,
      "learning_rate": 0.0001,
      "loss": 1.5099,
      "step": 1515
    },
    {
      "epoch": 0.36813987372510926,
      "grad_norm": 0.33786076307296753,
      "learning_rate": 0.0001,
      "loss": 1.7593,
      "step": 1516
    },
    {
      "epoch": 0.368382710053424,
      "grad_norm": 0.3465520441532135,
      "learning_rate": 0.0001,
      "loss": 1.5437,
      "step": 1517
    },
    {
      "epoch": 0.3686255463817387,
      "grad_norm": 0.3694959580898285,
      "learning_rate": 0.0001,
      "loss": 1.6638,
      "step": 1518
    },
    {
      "epoch": 0.3688683827100534,
      "grad_norm": 0.3485603332519531,
      "learning_rate": 0.0001,
      "loss": 1.6426,
      "step": 1519
    },
    {
      "epoch": 0.36911121903836813,
      "grad_norm": 0.3477795422077179,
      "learning_rate": 0.0001,
      "loss": 1.851,
      "step": 1520
    },
    {
      "epoch": 0.36935405536668287,
      "grad_norm": 0.3689527213573456,
      "learning_rate": 0.0001,
      "loss": 1.8298,
      "step": 1521
    },
    {
      "epoch": 0.3695968916949976,
      "grad_norm": 0.337512344121933,
      "learning_rate": 0.0001,
      "loss": 1.7631,
      "step": 1522
    },
    {
      "epoch": 0.3698397280233123,
      "grad_norm": 0.3260522484779358,
      "learning_rate": 0.0001,
      "loss": 1.5334,
      "step": 1523
    },
    {
      "epoch": 0.370082564351627,
      "grad_norm": 0.3651309013366699,
      "learning_rate": 0.0001,
      "loss": 1.766,
      "step": 1524
    },
    {
      "epoch": 0.37032540067994174,
      "grad_norm": 0.345079630613327,
      "learning_rate": 0.0001,
      "loss": 1.7777,
      "step": 1525
    },
    {
      "epoch": 0.3705682370082564,
      "grad_norm": 0.36292821168899536,
      "learning_rate": 0.0001,
      "loss": 1.7766,
      "step": 1526
    },
    {
      "epoch": 0.37081107333657115,
      "grad_norm": 0.34273114800453186,
      "learning_rate": 0.0001,
      "loss": 1.8465,
      "step": 1527
    },
    {
      "epoch": 0.3710539096648859,
      "grad_norm": 0.3660784363746643,
      "learning_rate": 0.0001,
      "loss": 1.7597,
      "step": 1528
    },
    {
      "epoch": 0.37129674599320056,
      "grad_norm": 0.3469104468822479,
      "learning_rate": 0.0001,
      "loss": 1.6156,
      "step": 1529
    },
    {
      "epoch": 0.3715395823215153,
      "grad_norm": 0.3330024778842926,
      "learning_rate": 0.0001,
      "loss": 1.6671,
      "step": 1530
    },
    {
      "epoch": 0.37178241864983,
      "grad_norm": 0.3564825654029846,
      "learning_rate": 0.0001,
      "loss": 1.814,
      "step": 1531
    },
    {
      "epoch": 0.3720252549781447,
      "grad_norm": 0.3567167818546295,
      "learning_rate": 0.0001,
      "loss": 1.7433,
      "step": 1532
    },
    {
      "epoch": 0.37226809130645944,
      "grad_norm": 0.3620092272758484,
      "learning_rate": 0.0001,
      "loss": 1.7187,
      "step": 1533
    },
    {
      "epoch": 0.37251092763477417,
      "grad_norm": 0.36428067088127136,
      "learning_rate": 0.0001,
      "loss": 1.7305,
      "step": 1534
    },
    {
      "epoch": 0.3727537639630889,
      "grad_norm": 0.3668808043003082,
      "learning_rate": 0.0001,
      "loss": 1.7448,
      "step": 1535
    },
    {
      "epoch": 0.3729966002914036,
      "grad_norm": 0.35790756344795227,
      "learning_rate": 0.0001,
      "loss": 1.6841,
      "step": 1536
    },
    {
      "epoch": 0.3732394366197183,
      "grad_norm": 0.3803383409976959,
      "learning_rate": 0.0001,
      "loss": 1.6986,
      "step": 1537
    },
    {
      "epoch": 0.37348227294803304,
      "grad_norm": 0.35111626982688904,
      "learning_rate": 0.0001,
      "loss": 1.6785,
      "step": 1538
    },
    {
      "epoch": 0.3737251092763477,
      "grad_norm": 0.3548724055290222,
      "learning_rate": 0.0001,
      "loss": 1.6832,
      "step": 1539
    },
    {
      "epoch": 0.37396794560466246,
      "grad_norm": 0.3986164629459381,
      "learning_rate": 0.0001,
      "loss": 1.8182,
      "step": 1540
    },
    {
      "epoch": 0.3742107819329772,
      "grad_norm": 0.35997113585472107,
      "learning_rate": 0.0001,
      "loss": 1.8724,
      "step": 1541
    },
    {
      "epoch": 0.37445361826129187,
      "grad_norm": 0.35795292258262634,
      "learning_rate": 0.0001,
      "loss": 1.7193,
      "step": 1542
    },
    {
      "epoch": 0.3746964545896066,
      "grad_norm": 0.35861730575561523,
      "learning_rate": 0.0001,
      "loss": 1.7895,
      "step": 1543
    },
    {
      "epoch": 0.37493929091792133,
      "grad_norm": 0.3538043200969696,
      "learning_rate": 0.0001,
      "loss": 1.7324,
      "step": 1544
    },
    {
      "epoch": 0.37518212724623606,
      "grad_norm": 0.3179115355014801,
      "learning_rate": 0.0001,
      "loss": 1.728,
      "step": 1545
    },
    {
      "epoch": 0.37542496357455074,
      "grad_norm": 0.3643108606338501,
      "learning_rate": 0.0001,
      "loss": 1.7634,
      "step": 1546
    },
    {
      "epoch": 0.3756677999028655,
      "grad_norm": 0.342984676361084,
      "learning_rate": 0.0001,
      "loss": 1.7601,
      "step": 1547
    },
    {
      "epoch": 0.3759106362311802,
      "grad_norm": 0.3417559266090393,
      "learning_rate": 0.0001,
      "loss": 1.7127,
      "step": 1548
    },
    {
      "epoch": 0.3761534725594949,
      "grad_norm": 0.3582463264465332,
      "learning_rate": 0.0001,
      "loss": 1.8364,
      "step": 1549
    },
    {
      "epoch": 0.3763963088878096,
      "grad_norm": 0.3490729331970215,
      "learning_rate": 0.0001,
      "loss": 1.6524,
      "step": 1550
    },
    {
      "epoch": 0.37663914521612435,
      "grad_norm": 0.33228304982185364,
      "learning_rate": 0.0001,
      "loss": 1.6648,
      "step": 1551
    },
    {
      "epoch": 0.376881981544439,
      "grad_norm": 0.3435888886451721,
      "learning_rate": 0.0001,
      "loss": 1.6654,
      "step": 1552
    },
    {
      "epoch": 0.37712481787275376,
      "grad_norm": 0.34257256984710693,
      "learning_rate": 0.0001,
      "loss": 1.6772,
      "step": 1553
    },
    {
      "epoch": 0.3773676542010685,
      "grad_norm": 0.3472994267940521,
      "learning_rate": 0.0001,
      "loss": 1.6576,
      "step": 1554
    },
    {
      "epoch": 0.37761049052938317,
      "grad_norm": 0.36370810866355896,
      "learning_rate": 0.0001,
      "loss": 1.7062,
      "step": 1555
    },
    {
      "epoch": 0.3778533268576979,
      "grad_norm": 0.33490875363349915,
      "learning_rate": 0.0001,
      "loss": 1.7911,
      "step": 1556
    },
    {
      "epoch": 0.37809616318601263,
      "grad_norm": 0.3443796932697296,
      "learning_rate": 0.0001,
      "loss": 1.6962,
      "step": 1557
    },
    {
      "epoch": 0.37833899951432737,
      "grad_norm": 0.3875202536582947,
      "learning_rate": 0.0001,
      "loss": 1.7595,
      "step": 1558
    },
    {
      "epoch": 0.37858183584264204,
      "grad_norm": 0.3398173451423645,
      "learning_rate": 0.0001,
      "loss": 1.7226,
      "step": 1559
    },
    {
      "epoch": 0.3788246721709568,
      "grad_norm": 0.3591778576374054,
      "learning_rate": 0.0001,
      "loss": 1.7631,
      "step": 1560
    },
    {
      "epoch": 0.3790675084992715,
      "grad_norm": 0.3375695049762726,
      "learning_rate": 0.0001,
      "loss": 1.7873,
      "step": 1561
    },
    {
      "epoch": 0.3793103448275862,
      "grad_norm": 0.3466924726963043,
      "learning_rate": 0.0001,
      "loss": 1.8187,
      "step": 1562
    },
    {
      "epoch": 0.3795531811559009,
      "grad_norm": 0.3472690284252167,
      "learning_rate": 0.0001,
      "loss": 1.7919,
      "step": 1563
    },
    {
      "epoch": 0.37979601748421565,
      "grad_norm": 0.3553646206855774,
      "learning_rate": 0.0001,
      "loss": 1.7024,
      "step": 1564
    },
    {
      "epoch": 0.38003885381253033,
      "grad_norm": 0.36639171838760376,
      "learning_rate": 0.0001,
      "loss": 1.7116,
      "step": 1565
    },
    {
      "epoch": 0.38028169014084506,
      "grad_norm": 0.3448989689350128,
      "learning_rate": 0.0001,
      "loss": 1.8521,
      "step": 1566
    },
    {
      "epoch": 0.3805245264691598,
      "grad_norm": 0.3462245762348175,
      "learning_rate": 0.0001,
      "loss": 1.7795,
      "step": 1567
    },
    {
      "epoch": 0.38076736279747453,
      "grad_norm": 0.3502362370491028,
      "learning_rate": 0.0001,
      "loss": 1.8843,
      "step": 1568
    },
    {
      "epoch": 0.3810101991257892,
      "grad_norm": 0.36486515402793884,
      "learning_rate": 0.0001,
      "loss": 1.7465,
      "step": 1569
    },
    {
      "epoch": 0.38125303545410394,
      "grad_norm": 0.3245135545730591,
      "learning_rate": 0.0001,
      "loss": 1.5685,
      "step": 1570
    },
    {
      "epoch": 0.38149587178241867,
      "grad_norm": 0.33357447385787964,
      "learning_rate": 0.0001,
      "loss": 1.849,
      "step": 1571
    },
    {
      "epoch": 0.38173870811073335,
      "grad_norm": 0.3470303416252136,
      "learning_rate": 0.0001,
      "loss": 1.7443,
      "step": 1572
    },
    {
      "epoch": 0.3819815444390481,
      "grad_norm": 0.3471367061138153,
      "learning_rate": 0.0001,
      "loss": 1.6293,
      "step": 1573
    },
    {
      "epoch": 0.3822243807673628,
      "grad_norm": 0.32362157106399536,
      "learning_rate": 0.0001,
      "loss": 1.557,
      "step": 1574
    },
    {
      "epoch": 0.3824672170956775,
      "grad_norm": 0.33684155344963074,
      "learning_rate": 0.0001,
      "loss": 1.7009,
      "step": 1575
    },
    {
      "epoch": 0.3827100534239922,
      "grad_norm": 0.3579930067062378,
      "learning_rate": 0.0001,
      "loss": 1.8257,
      "step": 1576
    },
    {
      "epoch": 0.38295288975230696,
      "grad_norm": 0.3533737063407898,
      "learning_rate": 0.0001,
      "loss": 1.8456,
      "step": 1577
    },
    {
      "epoch": 0.38319572608062163,
      "grad_norm": 0.36287230253219604,
      "learning_rate": 0.0001,
      "loss": 1.7177,
      "step": 1578
    },
    {
      "epoch": 0.38343856240893637,
      "grad_norm": 0.35856160521507263,
      "learning_rate": 0.0001,
      "loss": 1.7155,
      "step": 1579
    },
    {
      "epoch": 0.3836813987372511,
      "grad_norm": 0.3617909550666809,
      "learning_rate": 0.0001,
      "loss": 1.7741,
      "step": 1580
    },
    {
      "epoch": 0.38392423506556583,
      "grad_norm": 0.35343724489212036,
      "learning_rate": 0.0001,
      "loss": 1.7358,
      "step": 1581
    },
    {
      "epoch": 0.3841670713938805,
      "grad_norm": 0.3512588441371918,
      "learning_rate": 0.0001,
      "loss": 1.8715,
      "step": 1582
    },
    {
      "epoch": 0.38440990772219524,
      "grad_norm": 0.3507743775844574,
      "learning_rate": 0.0001,
      "loss": 1.7062,
      "step": 1583
    },
    {
      "epoch": 0.38465274405051,
      "grad_norm": 0.36896374821662903,
      "learning_rate": 0.0001,
      "loss": 1.7494,
      "step": 1584
    },
    {
      "epoch": 0.38489558037882465,
      "grad_norm": 0.3897750675678253,
      "learning_rate": 0.0001,
      "loss": 1.7953,
      "step": 1585
    },
    {
      "epoch": 0.3851384167071394,
      "grad_norm": 0.3406773805618286,
      "learning_rate": 0.0001,
      "loss": 1.7495,
      "step": 1586
    },
    {
      "epoch": 0.3853812530354541,
      "grad_norm": 0.35307011008262634,
      "learning_rate": 0.0001,
      "loss": 1.6683,
      "step": 1587
    },
    {
      "epoch": 0.3856240893637688,
      "grad_norm": 0.32022756338119507,
      "learning_rate": 0.0001,
      "loss": 1.5555,
      "step": 1588
    },
    {
      "epoch": 0.38586692569208353,
      "grad_norm": 0.3569686710834503,
      "learning_rate": 0.0001,
      "loss": 1.8626,
      "step": 1589
    },
    {
      "epoch": 0.38610976202039826,
      "grad_norm": 0.3682452142238617,
      "learning_rate": 0.0001,
      "loss": 1.8699,
      "step": 1590
    },
    {
      "epoch": 0.386352598348713,
      "grad_norm": 0.3267800211906433,
      "learning_rate": 0.0001,
      "loss": 1.5346,
      "step": 1591
    },
    {
      "epoch": 0.38659543467702767,
      "grad_norm": 0.34441307187080383,
      "learning_rate": 0.0001,
      "loss": 1.7764,
      "step": 1592
    },
    {
      "epoch": 0.3868382710053424,
      "grad_norm": 0.33361029624938965,
      "learning_rate": 0.0001,
      "loss": 1.6749,
      "step": 1593
    },
    {
      "epoch": 0.38708110733365714,
      "grad_norm": 0.34326493740081787,
      "learning_rate": 0.0001,
      "loss": 1.783,
      "step": 1594
    },
    {
      "epoch": 0.3873239436619718,
      "grad_norm": 0.3545917868614197,
      "learning_rate": 0.0001,
      "loss": 1.7397,
      "step": 1595
    },
    {
      "epoch": 0.38756677999028655,
      "grad_norm": 0.3923596143722534,
      "learning_rate": 0.0001,
      "loss": 1.7354,
      "step": 1596
    },
    {
      "epoch": 0.3878096163186013,
      "grad_norm": 0.35300305485725403,
      "learning_rate": 0.0001,
      "loss": 1.7267,
      "step": 1597
    },
    {
      "epoch": 0.38805245264691596,
      "grad_norm": 0.42850735783576965,
      "learning_rate": 0.0001,
      "loss": 1.8924,
      "step": 1598
    },
    {
      "epoch": 0.3882952889752307,
      "grad_norm": 0.361499160528183,
      "learning_rate": 0.0001,
      "loss": 1.7558,
      "step": 1599
    },
    {
      "epoch": 0.3885381253035454,
      "grad_norm": 0.35589104890823364,
      "learning_rate": 0.0001,
      "loss": 1.634,
      "step": 1600
    },
    {
      "epoch": 0.3887809616318601,
      "grad_norm": 0.37668150663375854,
      "learning_rate": 0.0001,
      "loss": 1.799,
      "step": 1601
    },
    {
      "epoch": 0.38902379796017483,
      "grad_norm": 0.34165313839912415,
      "learning_rate": 0.0001,
      "loss": 1.7319,
      "step": 1602
    },
    {
      "epoch": 0.38926663428848957,
      "grad_norm": 0.341299831867218,
      "learning_rate": 0.0001,
      "loss": 1.7136,
      "step": 1603
    },
    {
      "epoch": 0.3895094706168043,
      "grad_norm": 0.34204524755477905,
      "learning_rate": 0.0001,
      "loss": 1.5748,
      "step": 1604
    },
    {
      "epoch": 0.389752306945119,
      "grad_norm": 0.3674558401107788,
      "learning_rate": 0.0001,
      "loss": 1.7721,
      "step": 1605
    },
    {
      "epoch": 0.3899951432734337,
      "grad_norm": 0.33253777027130127,
      "learning_rate": 0.0001,
      "loss": 1.6459,
      "step": 1606
    },
    {
      "epoch": 0.39023797960174844,
      "grad_norm": 0.36387547850608826,
      "learning_rate": 0.0001,
      "loss": 1.6895,
      "step": 1607
    },
    {
      "epoch": 0.3904808159300631,
      "grad_norm": 0.3665448725223541,
      "learning_rate": 0.0001,
      "loss": 1.7362,
      "step": 1608
    },
    {
      "epoch": 0.39072365225837785,
      "grad_norm": 0.3469473123550415,
      "learning_rate": 0.0001,
      "loss": 1.8904,
      "step": 1609
    },
    {
      "epoch": 0.3909664885866926,
      "grad_norm": 0.365561306476593,
      "learning_rate": 0.0001,
      "loss": 1.661,
      "step": 1610
    },
    {
      "epoch": 0.39120932491500726,
      "grad_norm": 0.340475469827652,
      "learning_rate": 0.0001,
      "loss": 1.7433,
      "step": 1611
    },
    {
      "epoch": 0.391452161243322,
      "grad_norm": 0.3633958697319031,
      "learning_rate": 0.0001,
      "loss": 1.6913,
      "step": 1612
    },
    {
      "epoch": 0.3916949975716367,
      "grad_norm": 0.3879851996898651,
      "learning_rate": 0.0001,
      "loss": 1.7817,
      "step": 1613
    },
    {
      "epoch": 0.39193783389995146,
      "grad_norm": 0.3605000376701355,
      "learning_rate": 0.0001,
      "loss": 1.7143,
      "step": 1614
    },
    {
      "epoch": 0.39218067022826614,
      "grad_norm": 0.3635120689868927,
      "learning_rate": 0.0001,
      "loss": 1.8388,
      "step": 1615
    },
    {
      "epoch": 0.39242350655658087,
      "grad_norm": 0.3307245969772339,
      "learning_rate": 0.0001,
      "loss": 1.6376,
      "step": 1616
    },
    {
      "epoch": 0.3926663428848956,
      "grad_norm": 0.3344554603099823,
      "learning_rate": 0.0001,
      "loss": 1.7015,
      "step": 1617
    },
    {
      "epoch": 0.3929091792132103,
      "grad_norm": 0.3535119891166687,
      "learning_rate": 0.0001,
      "loss": 1.6508,
      "step": 1618
    },
    {
      "epoch": 0.393152015541525,
      "grad_norm": 0.35171249508857727,
      "learning_rate": 0.0001,
      "loss": 1.8599,
      "step": 1619
    },
    {
      "epoch": 0.39339485186983975,
      "grad_norm": 0.34228235483169556,
      "learning_rate": 0.0001,
      "loss": 1.7377,
      "step": 1620
    },
    {
      "epoch": 0.3936376881981544,
      "grad_norm": 0.3834402859210968,
      "learning_rate": 0.0001,
      "loss": 1.7709,
      "step": 1621
    },
    {
      "epoch": 0.39388052452646916,
      "grad_norm": 0.34480172395706177,
      "learning_rate": 0.0001,
      "loss": 1.7076,
      "step": 1622
    },
    {
      "epoch": 0.3941233608547839,
      "grad_norm": 0.35638877749443054,
      "learning_rate": 0.0001,
      "loss": 1.7495,
      "step": 1623
    },
    {
      "epoch": 0.39436619718309857,
      "grad_norm": 0.3370180130004883,
      "learning_rate": 0.0001,
      "loss": 1.6121,
      "step": 1624
    },
    {
      "epoch": 0.3946090335114133,
      "grad_norm": 0.37028077244758606,
      "learning_rate": 0.0001,
      "loss": 1.9731,
      "step": 1625
    },
    {
      "epoch": 0.39485186983972803,
      "grad_norm": 0.3528951108455658,
      "learning_rate": 0.0001,
      "loss": 1.6877,
      "step": 1626
    },
    {
      "epoch": 0.39509470616804276,
      "grad_norm": 0.3634454905986786,
      "learning_rate": 0.0001,
      "loss": 1.7671,
      "step": 1627
    },
    {
      "epoch": 0.39533754249635744,
      "grad_norm": 0.3391231894493103,
      "learning_rate": 0.0001,
      "loss": 1.6827,
      "step": 1628
    },
    {
      "epoch": 0.3955803788246722,
      "grad_norm": 0.3546386957168579,
      "learning_rate": 0.0001,
      "loss": 1.6595,
      "step": 1629
    },
    {
      "epoch": 0.3958232151529869,
      "grad_norm": 0.34772059321403503,
      "learning_rate": 0.0001,
      "loss": 1.7565,
      "step": 1630
    },
    {
      "epoch": 0.3960660514813016,
      "grad_norm": 0.3234846591949463,
      "learning_rate": 0.0001,
      "loss": 1.6029,
      "step": 1631
    },
    {
      "epoch": 0.3963088878096163,
      "grad_norm": 0.34120452404022217,
      "learning_rate": 0.0001,
      "loss": 1.612,
      "step": 1632
    },
    {
      "epoch": 0.39655172413793105,
      "grad_norm": 0.34159693121910095,
      "learning_rate": 0.0001,
      "loss": 1.7595,
      "step": 1633
    },
    {
      "epoch": 0.3967945604662457,
      "grad_norm": 0.35023289918899536,
      "learning_rate": 0.0001,
      "loss": 1.7803,
      "step": 1634
    },
    {
      "epoch": 0.39703739679456046,
      "grad_norm": 0.3789989650249481,
      "learning_rate": 0.0001,
      "loss": 1.7713,
      "step": 1635
    },
    {
      "epoch": 0.3972802331228752,
      "grad_norm": 0.36430492997169495,
      "learning_rate": 0.0001,
      "loss": 1.8014,
      "step": 1636
    },
    {
      "epoch": 0.3975230694511899,
      "grad_norm": 0.3651634156703949,
      "learning_rate": 0.0001,
      "loss": 1.9387,
      "step": 1637
    },
    {
      "epoch": 0.3977659057795046,
      "grad_norm": 0.33149656653404236,
      "learning_rate": 0.0001,
      "loss": 1.7681,
      "step": 1638
    },
    {
      "epoch": 0.39800874210781934,
      "grad_norm": 0.3321326971054077,
      "learning_rate": 0.0001,
      "loss": 1.7215,
      "step": 1639
    },
    {
      "epoch": 0.39825157843613407,
      "grad_norm": 0.36206716299057007,
      "learning_rate": 0.0001,
      "loss": 1.7122,
      "step": 1640
    },
    {
      "epoch": 0.39849441476444875,
      "grad_norm": 0.3373028039932251,
      "learning_rate": 0.0001,
      "loss": 1.696,
      "step": 1641
    },
    {
      "epoch": 0.3987372510927635,
      "grad_norm": 0.34143638610839844,
      "learning_rate": 0.0001,
      "loss": 1.7125,
      "step": 1642
    },
    {
      "epoch": 0.3989800874210782,
      "grad_norm": 0.3449946343898773,
      "learning_rate": 0.0001,
      "loss": 1.7054,
      "step": 1643
    },
    {
      "epoch": 0.3992229237493929,
      "grad_norm": 0.36091458797454834,
      "learning_rate": 0.0001,
      "loss": 1.6785,
      "step": 1644
    },
    {
      "epoch": 0.3994657600777076,
      "grad_norm": 0.35519731044769287,
      "learning_rate": 0.0001,
      "loss": 1.7946,
      "step": 1645
    },
    {
      "epoch": 0.39970859640602235,
      "grad_norm": 0.34838569164276123,
      "learning_rate": 0.0001,
      "loss": 1.7462,
      "step": 1646
    },
    {
      "epoch": 0.39995143273433703,
      "grad_norm": 0.3532315194606781,
      "learning_rate": 0.0001,
      "loss": 1.7446,
      "step": 1647
    },
    {
      "epoch": 0.40019426906265176,
      "grad_norm": 0.34514230489730835,
      "learning_rate": 0.0001,
      "loss": 1.6328,
      "step": 1648
    },
    {
      "epoch": 0.4004371053909665,
      "grad_norm": 0.3498397767543793,
      "learning_rate": 0.0001,
      "loss": 1.9101,
      "step": 1649
    },
    {
      "epoch": 0.40067994171928123,
      "grad_norm": 0.36845964193344116,
      "learning_rate": 0.0001,
      "loss": 1.7796,
      "step": 1650
    },
    {
      "epoch": 0.4009227780475959,
      "grad_norm": 0.3510598838329315,
      "learning_rate": 0.0001,
      "loss": 1.6093,
      "step": 1651
    },
    {
      "epoch": 0.40116561437591064,
      "grad_norm": 0.33339422941207886,
      "learning_rate": 0.0001,
      "loss": 1.6679,
      "step": 1652
    },
    {
      "epoch": 0.4014084507042254,
      "grad_norm": 0.33257877826690674,
      "learning_rate": 0.0001,
      "loss": 1.7101,
      "step": 1653
    },
    {
      "epoch": 0.40165128703254005,
      "grad_norm": 0.3461006283760071,
      "learning_rate": 0.0001,
      "loss": 1.6459,
      "step": 1654
    },
    {
      "epoch": 0.4018941233608548,
      "grad_norm": 0.3384184241294861,
      "learning_rate": 0.0001,
      "loss": 1.6637,
      "step": 1655
    },
    {
      "epoch": 0.4021369596891695,
      "grad_norm": 0.34670189023017883,
      "learning_rate": 0.0001,
      "loss": 1.6808,
      "step": 1656
    },
    {
      "epoch": 0.4023797960174842,
      "grad_norm": 0.34890082478523254,
      "learning_rate": 0.0001,
      "loss": 1.7313,
      "step": 1657
    },
    {
      "epoch": 0.4026226323457989,
      "grad_norm": 0.3604244887828827,
      "learning_rate": 0.0001,
      "loss": 1.92,
      "step": 1658
    },
    {
      "epoch": 0.40286546867411366,
      "grad_norm": 0.3499077260494232,
      "learning_rate": 0.0001,
      "loss": 1.7897,
      "step": 1659
    },
    {
      "epoch": 0.4031083050024284,
      "grad_norm": 0.3570781648159027,
      "learning_rate": 0.0001,
      "loss": 1.7975,
      "step": 1660
    },
    {
      "epoch": 0.40335114133074307,
      "grad_norm": 0.39222121238708496,
      "learning_rate": 0.0001,
      "loss": 1.6846,
      "step": 1661
    },
    {
      "epoch": 0.4035939776590578,
      "grad_norm": 0.34657740592956543,
      "learning_rate": 0.0001,
      "loss": 1.6149,
      "step": 1662
    },
    {
      "epoch": 0.40383681398737253,
      "grad_norm": 0.34491151571273804,
      "learning_rate": 0.0001,
      "loss": 1.85,
      "step": 1663
    },
    {
      "epoch": 0.4040796503156872,
      "grad_norm": 0.3567986488342285,
      "learning_rate": 0.0001,
      "loss": 1.8325,
      "step": 1664
    },
    {
      "epoch": 0.40432248664400194,
      "grad_norm": 0.34347212314605713,
      "learning_rate": 0.0001,
      "loss": 1.7163,
      "step": 1665
    },
    {
      "epoch": 0.4045653229723167,
      "grad_norm": 0.36440059542655945,
      "learning_rate": 0.0001,
      "loss": 1.8832,
      "step": 1666
    },
    {
      "epoch": 0.40480815930063135,
      "grad_norm": 0.3367612063884735,
      "learning_rate": 0.0001,
      "loss": 1.7437,
      "step": 1667
    },
    {
      "epoch": 0.4050509956289461,
      "grad_norm": 0.3793525993824005,
      "learning_rate": 0.0001,
      "loss": 1.6954,
      "step": 1668
    },
    {
      "epoch": 0.4052938319572608,
      "grad_norm": 0.34889304637908936,
      "learning_rate": 0.0001,
      "loss": 1.6213,
      "step": 1669
    },
    {
      "epoch": 0.4055366682855755,
      "grad_norm": 0.33768993616104126,
      "learning_rate": 0.0001,
      "loss": 1.6736,
      "step": 1670
    },
    {
      "epoch": 0.40577950461389023,
      "grad_norm": 0.349176824092865,
      "learning_rate": 0.0001,
      "loss": 1.8238,
      "step": 1671
    },
    {
      "epoch": 0.40602234094220496,
      "grad_norm": 0.35951730608940125,
      "learning_rate": 0.0001,
      "loss": 1.8022,
      "step": 1672
    },
    {
      "epoch": 0.4062651772705197,
      "grad_norm": 0.36430323123931885,
      "learning_rate": 0.0001,
      "loss": 1.8402,
      "step": 1673
    },
    {
      "epoch": 0.4065080135988344,
      "grad_norm": 0.3360004723072052,
      "learning_rate": 0.0001,
      "loss": 1.5745,
      "step": 1674
    },
    {
      "epoch": 0.4067508499271491,
      "grad_norm": 0.3629210293292999,
      "learning_rate": 0.0001,
      "loss": 1.825,
      "step": 1675
    },
    {
      "epoch": 0.40699368625546384,
      "grad_norm": 0.35018420219421387,
      "learning_rate": 0.0001,
      "loss": 1.7295,
      "step": 1676
    },
    {
      "epoch": 0.4072365225837785,
      "grad_norm": 0.3400278687477112,
      "learning_rate": 0.0001,
      "loss": 1.6958,
      "step": 1677
    },
    {
      "epoch": 0.40747935891209325,
      "grad_norm": 0.34914901852607727,
      "learning_rate": 0.0001,
      "loss": 1.6115,
      "step": 1678
    },
    {
      "epoch": 0.407722195240408,
      "grad_norm": 0.3571421504020691,
      "learning_rate": 0.0001,
      "loss": 1.7384,
      "step": 1679
    },
    {
      "epoch": 0.40796503156872266,
      "grad_norm": 0.3605473041534424,
      "learning_rate": 0.0001,
      "loss": 1.8193,
      "step": 1680
    },
    {
      "epoch": 0.4082078678970374,
      "grad_norm": 0.37068045139312744,
      "learning_rate": 0.0001,
      "loss": 1.9287,
      "step": 1681
    },
    {
      "epoch": 0.4084507042253521,
      "grad_norm": 0.34394922852516174,
      "learning_rate": 0.0001,
      "loss": 1.7579,
      "step": 1682
    },
    {
      "epoch": 0.4086935405536668,
      "grad_norm": 0.3192780315876007,
      "learning_rate": 0.0001,
      "loss": 1.5888,
      "step": 1683
    },
    {
      "epoch": 0.40893637688198153,
      "grad_norm": 0.3368231952190399,
      "learning_rate": 0.0001,
      "loss": 1.7074,
      "step": 1684
    },
    {
      "epoch": 0.40917921321029627,
      "grad_norm": 0.3436574935913086,
      "learning_rate": 0.0001,
      "loss": 1.6942,
      "step": 1685
    },
    {
      "epoch": 0.409422049538611,
      "grad_norm": 0.3476155996322632,
      "learning_rate": 0.0001,
      "loss": 1.7639,
      "step": 1686
    },
    {
      "epoch": 0.4096648858669257,
      "grad_norm": 0.33362385630607605,
      "learning_rate": 0.0001,
      "loss": 1.6146,
      "step": 1687
    },
    {
      "epoch": 0.4099077221952404,
      "grad_norm": 0.3497028052806854,
      "learning_rate": 0.0001,
      "loss": 1.5816,
      "step": 1688
    },
    {
      "epoch": 0.41015055852355514,
      "grad_norm": 0.3652702569961548,
      "learning_rate": 0.0001,
      "loss": 1.8025,
      "step": 1689
    },
    {
      "epoch": 0.4103933948518698,
      "grad_norm": 0.3235238194465637,
      "learning_rate": 0.0001,
      "loss": 1.5762,
      "step": 1690
    },
    {
      "epoch": 0.41063623118018455,
      "grad_norm": 0.3403080403804779,
      "learning_rate": 0.0001,
      "loss": 1.7607,
      "step": 1691
    },
    {
      "epoch": 0.4108790675084993,
      "grad_norm": 0.3631891906261444,
      "learning_rate": 0.0001,
      "loss": 1.8404,
      "step": 1692
    },
    {
      "epoch": 0.41112190383681396,
      "grad_norm": 0.3438990116119385,
      "learning_rate": 0.0001,
      "loss": 1.7936,
      "step": 1693
    },
    {
      "epoch": 0.4113647401651287,
      "grad_norm": 0.35403940081596375,
      "learning_rate": 0.0001,
      "loss": 1.7144,
      "step": 1694
    },
    {
      "epoch": 0.41160757649344343,
      "grad_norm": 0.3484737277030945,
      "learning_rate": 0.0001,
      "loss": 1.6433,
      "step": 1695
    },
    {
      "epoch": 0.41185041282175816,
      "grad_norm": 0.3359359800815582,
      "learning_rate": 0.0001,
      "loss": 1.583,
      "step": 1696
    },
    {
      "epoch": 0.41209324915007284,
      "grad_norm": 0.35734885931015015,
      "learning_rate": 0.0001,
      "loss": 1.8943,
      "step": 1697
    },
    {
      "epoch": 0.41233608547838757,
      "grad_norm": 0.37979212403297424,
      "learning_rate": 0.0001,
      "loss": 1.7848,
      "step": 1698
    },
    {
      "epoch": 0.4125789218067023,
      "grad_norm": 0.3913626968860626,
      "learning_rate": 0.0001,
      "loss": 1.9229,
      "step": 1699
    },
    {
      "epoch": 0.412821758135017,
      "grad_norm": 0.38415777683258057,
      "learning_rate": 0.0001,
      "loss": 1.8476,
      "step": 1700
    },
    {
      "epoch": 0.4130645944633317,
      "grad_norm": 0.3729405403137207,
      "learning_rate": 0.0001,
      "loss": 1.8208,
      "step": 1701
    },
    {
      "epoch": 0.41330743079164645,
      "grad_norm": 0.36757805943489075,
      "learning_rate": 0.0001,
      "loss": 1.6945,
      "step": 1702
    },
    {
      "epoch": 0.4135502671199611,
      "grad_norm": 0.34475553035736084,
      "learning_rate": 0.0001,
      "loss": 1.6519,
      "step": 1703
    },
    {
      "epoch": 0.41379310344827586,
      "grad_norm": 0.34733209013938904,
      "learning_rate": 0.0001,
      "loss": 1.5978,
      "step": 1704
    },
    {
      "epoch": 0.4140359397765906,
      "grad_norm": 0.3799619674682617,
      "learning_rate": 0.0001,
      "loss": 1.677,
      "step": 1705
    },
    {
      "epoch": 0.41427877610490527,
      "grad_norm": 0.3415276110172272,
      "learning_rate": 0.0001,
      "loss": 1.736,
      "step": 1706
    },
    {
      "epoch": 0.41452161243322,
      "grad_norm": 0.35778936743736267,
      "learning_rate": 0.0001,
      "loss": 1.8407,
      "step": 1707
    },
    {
      "epoch": 0.41476444876153473,
      "grad_norm": 0.347223699092865,
      "learning_rate": 0.0001,
      "loss": 1.6321,
      "step": 1708
    },
    {
      "epoch": 0.41500728508984946,
      "grad_norm": 0.36570411920547485,
      "learning_rate": 0.0001,
      "loss": 1.8242,
      "step": 1709
    },
    {
      "epoch": 0.41525012141816414,
      "grad_norm": 0.3869484066963196,
      "learning_rate": 0.0001,
      "loss": 1.734,
      "step": 1710
    },
    {
      "epoch": 0.4154929577464789,
      "grad_norm": 0.36324143409729004,
      "learning_rate": 0.0001,
      "loss": 1.6839,
      "step": 1711
    },
    {
      "epoch": 0.4157357940747936,
      "grad_norm": 0.35620391368865967,
      "learning_rate": 0.0001,
      "loss": 1.7555,
      "step": 1712
    },
    {
      "epoch": 0.4159786304031083,
      "grad_norm": 0.3773002028465271,
      "learning_rate": 0.0001,
      "loss": 1.8965,
      "step": 1713
    },
    {
      "epoch": 0.416221466731423,
      "grad_norm": 0.3578970432281494,
      "learning_rate": 0.0001,
      "loss": 1.6807,
      "step": 1714
    },
    {
      "epoch": 0.41646430305973775,
      "grad_norm": 0.36377522349357605,
      "learning_rate": 0.0001,
      "loss": 1.9471,
      "step": 1715
    },
    {
      "epoch": 0.4167071393880524,
      "grad_norm": 0.3433545231819153,
      "learning_rate": 0.0001,
      "loss": 1.7947,
      "step": 1716
    },
    {
      "epoch": 0.41694997571636716,
      "grad_norm": 0.3755989968776703,
      "learning_rate": 0.0001,
      "loss": 1.8395,
      "step": 1717
    },
    {
      "epoch": 0.4171928120446819,
      "grad_norm": 0.33563217520713806,
      "learning_rate": 0.0001,
      "loss": 1.6544,
      "step": 1718
    },
    {
      "epoch": 0.4174356483729966,
      "grad_norm": 0.33854302763938904,
      "learning_rate": 0.0001,
      "loss": 1.6766,
      "step": 1719
    },
    {
      "epoch": 0.4176784847013113,
      "grad_norm": 0.3348889648914337,
      "learning_rate": 0.0001,
      "loss": 1.7056,
      "step": 1720
    },
    {
      "epoch": 0.41792132102962604,
      "grad_norm": 0.3639450669288635,
      "learning_rate": 0.0001,
      "loss": 1.7382,
      "step": 1721
    },
    {
      "epoch": 0.41816415735794077,
      "grad_norm": 0.34863460063934326,
      "learning_rate": 0.0001,
      "loss": 1.6771,
      "step": 1722
    },
    {
      "epoch": 0.41840699368625545,
      "grad_norm": 0.35017862915992737,
      "learning_rate": 0.0001,
      "loss": 1.7366,
      "step": 1723
    },
    {
      "epoch": 0.4186498300145702,
      "grad_norm": 0.3767487108707428,
      "learning_rate": 0.0001,
      "loss": 1.8036,
      "step": 1724
    },
    {
      "epoch": 0.4188926663428849,
      "grad_norm": 0.35246649384498596,
      "learning_rate": 0.0001,
      "loss": 1.7452,
      "step": 1725
    },
    {
      "epoch": 0.4191355026711996,
      "grad_norm": 0.33406075835227966,
      "learning_rate": 0.0001,
      "loss": 1.7114,
      "step": 1726
    },
    {
      "epoch": 0.4193783389995143,
      "grad_norm": 0.35864290595054626,
      "learning_rate": 0.0001,
      "loss": 1.7944,
      "step": 1727
    },
    {
      "epoch": 0.41962117532782905,
      "grad_norm": 0.37582850456237793,
      "learning_rate": 0.0001,
      "loss": 1.8609,
      "step": 1728
    },
    {
      "epoch": 0.41986401165614373,
      "grad_norm": 0.3527149558067322,
      "learning_rate": 0.0001,
      "loss": 1.6799,
      "step": 1729
    },
    {
      "epoch": 0.42010684798445846,
      "grad_norm": 0.3358190059661865,
      "learning_rate": 0.0001,
      "loss": 1.6489,
      "step": 1730
    },
    {
      "epoch": 0.4203496843127732,
      "grad_norm": 0.3604898750782013,
      "learning_rate": 0.0001,
      "loss": 1.6847,
      "step": 1731
    },
    {
      "epoch": 0.42059252064108793,
      "grad_norm": 0.3554806113243103,
      "learning_rate": 0.0001,
      "loss": 1.7477,
      "step": 1732
    },
    {
      "epoch": 0.4208353569694026,
      "grad_norm": 0.3845835328102112,
      "learning_rate": 0.0001,
      "loss": 1.8873,
      "step": 1733
    },
    {
      "epoch": 0.42107819329771734,
      "grad_norm": 0.37965497374534607,
      "learning_rate": 0.0001,
      "loss": 1.8029,
      "step": 1734
    },
    {
      "epoch": 0.4213210296260321,
      "grad_norm": 0.3325018882751465,
      "learning_rate": 0.0001,
      "loss": 1.6127,
      "step": 1735
    },
    {
      "epoch": 0.42156386595434675,
      "grad_norm": 0.3592410981655121,
      "learning_rate": 0.0001,
      "loss": 1.7491,
      "step": 1736
    },
    {
      "epoch": 0.4218067022826615,
      "grad_norm": 0.3251824378967285,
      "learning_rate": 0.0001,
      "loss": 1.5902,
      "step": 1737
    },
    {
      "epoch": 0.4220495386109762,
      "grad_norm": 0.36237943172454834,
      "learning_rate": 0.0001,
      "loss": 1.7964,
      "step": 1738
    },
    {
      "epoch": 0.4222923749392909,
      "grad_norm": 0.320538729429245,
      "learning_rate": 0.0001,
      "loss": 1.6364,
      "step": 1739
    },
    {
      "epoch": 0.4225352112676056,
      "grad_norm": 0.34662577509880066,
      "learning_rate": 0.0001,
      "loss": 1.6954,
      "step": 1740
    },
    {
      "epoch": 0.42277804759592036,
      "grad_norm": 0.35295093059539795,
      "learning_rate": 0.0001,
      "loss": 1.7166,
      "step": 1741
    },
    {
      "epoch": 0.4230208839242351,
      "grad_norm": 0.3445799648761749,
      "learning_rate": 0.0001,
      "loss": 1.6879,
      "step": 1742
    },
    {
      "epoch": 0.42326372025254977,
      "grad_norm": 0.368622362613678,
      "learning_rate": 0.0001,
      "loss": 1.8182,
      "step": 1743
    },
    {
      "epoch": 0.4235065565808645,
      "grad_norm": 0.3579779267311096,
      "learning_rate": 0.0001,
      "loss": 1.8737,
      "step": 1744
    },
    {
      "epoch": 0.42374939290917923,
      "grad_norm": 0.35230371356010437,
      "learning_rate": 0.0001,
      "loss": 1.6501,
      "step": 1745
    },
    {
      "epoch": 0.4239922292374939,
      "grad_norm": 0.3304905593395233,
      "learning_rate": 0.0001,
      "loss": 1.7073,
      "step": 1746
    },
    {
      "epoch": 0.42423506556580864,
      "grad_norm": 0.34176966547966003,
      "learning_rate": 0.0001,
      "loss": 1.6926,
      "step": 1747
    },
    {
      "epoch": 0.4244779018941234,
      "grad_norm": 0.3877265453338623,
      "learning_rate": 0.0001,
      "loss": 1.8788,
      "step": 1748
    },
    {
      "epoch": 0.42472073822243805,
      "grad_norm": 0.35358646512031555,
      "learning_rate": 0.0001,
      "loss": 1.7133,
      "step": 1749
    },
    {
      "epoch": 0.4249635745507528,
      "grad_norm": 0.3553861677646637,
      "learning_rate": 0.0001,
      "loss": 1.7243,
      "step": 1750
    },
    {
      "epoch": 0.4252064108790675,
      "grad_norm": 0.3517538607120514,
      "learning_rate": 0.0001,
      "loss": 1.751,
      "step": 1751
    },
    {
      "epoch": 0.4254492472073822,
      "grad_norm": 0.36939677596092224,
      "learning_rate": 0.0001,
      "loss": 1.8552,
      "step": 1752
    },
    {
      "epoch": 0.42569208353569693,
      "grad_norm": 0.33711880445480347,
      "learning_rate": 0.0001,
      "loss": 1.6337,
      "step": 1753
    },
    {
      "epoch": 0.42593491986401166,
      "grad_norm": 0.3469174802303314,
      "learning_rate": 0.0001,
      "loss": 1.6768,
      "step": 1754
    },
    {
      "epoch": 0.4261777561923264,
      "grad_norm": 0.4004037380218506,
      "learning_rate": 0.0001,
      "loss": 1.888,
      "step": 1755
    },
    {
      "epoch": 0.4264205925206411,
      "grad_norm": 0.378012478351593,
      "learning_rate": 0.0001,
      "loss": 1.8775,
      "step": 1756
    },
    {
      "epoch": 0.4266634288489558,
      "grad_norm": 0.3540249764919281,
      "learning_rate": 0.0001,
      "loss": 1.8041,
      "step": 1757
    },
    {
      "epoch": 0.42690626517727054,
      "grad_norm": 0.36658915877342224,
      "learning_rate": 0.0001,
      "loss": 1.6588,
      "step": 1758
    },
    {
      "epoch": 0.4271491015055852,
      "grad_norm": 0.36717551946640015,
      "learning_rate": 0.0001,
      "loss": 1.7617,
      "step": 1759
    },
    {
      "epoch": 0.42739193783389995,
      "grad_norm": 0.3298027217388153,
      "learning_rate": 0.0001,
      "loss": 1.5488,
      "step": 1760
    },
    {
      "epoch": 0.4276347741622147,
      "grad_norm": 0.34846270084381104,
      "learning_rate": 0.0001,
      "loss": 1.8168,
      "step": 1761
    },
    {
      "epoch": 0.42787761049052936,
      "grad_norm": 0.3526111841201782,
      "learning_rate": 0.0001,
      "loss": 1.7507,
      "step": 1762
    },
    {
      "epoch": 0.4281204468188441,
      "grad_norm": 0.35745999217033386,
      "learning_rate": 0.0001,
      "loss": 1.7467,
      "step": 1763
    },
    {
      "epoch": 0.4283632831471588,
      "grad_norm": 0.360914409160614,
      "learning_rate": 0.0001,
      "loss": 1.587,
      "step": 1764
    },
    {
      "epoch": 0.42860611947547356,
      "grad_norm": 0.3446374535560608,
      "learning_rate": 0.0001,
      "loss": 1.6978,
      "step": 1765
    },
    {
      "epoch": 0.42884895580378823,
      "grad_norm": 0.3785291910171509,
      "learning_rate": 0.0001,
      "loss": 1.8251,
      "step": 1766
    },
    {
      "epoch": 0.42909179213210297,
      "grad_norm": 0.35578465461730957,
      "learning_rate": 0.0001,
      "loss": 1.7313,
      "step": 1767
    },
    {
      "epoch": 0.4293346284604177,
      "grad_norm": 0.39511141180992126,
      "learning_rate": 0.0001,
      "loss": 1.8263,
      "step": 1768
    },
    {
      "epoch": 0.4295774647887324,
      "grad_norm": 0.36311453580856323,
      "learning_rate": 0.0001,
      "loss": 1.8702,
      "step": 1769
    },
    {
      "epoch": 0.4298203011170471,
      "grad_norm": 0.39010193943977356,
      "learning_rate": 0.0001,
      "loss": 1.868,
      "step": 1770
    },
    {
      "epoch": 0.43006313744536184,
      "grad_norm": 0.3557552099227905,
      "learning_rate": 0.0001,
      "loss": 1.738,
      "step": 1771
    },
    {
      "epoch": 0.4303059737736765,
      "grad_norm": 0.34697064757347107,
      "learning_rate": 0.0001,
      "loss": 1.7679,
      "step": 1772
    },
    {
      "epoch": 0.43054881010199125,
      "grad_norm": 0.3562696576118469,
      "learning_rate": 0.0001,
      "loss": 1.6916,
      "step": 1773
    },
    {
      "epoch": 0.430791646430306,
      "grad_norm": 0.3603959381580353,
      "learning_rate": 0.0001,
      "loss": 1.8988,
      "step": 1774
    },
    {
      "epoch": 0.43103448275862066,
      "grad_norm": 0.34189701080322266,
      "learning_rate": 0.0001,
      "loss": 1.628,
      "step": 1775
    },
    {
      "epoch": 0.4312773190869354,
      "grad_norm": 0.3767715394496918,
      "learning_rate": 0.0001,
      "loss": 1.7701,
      "step": 1776
    },
    {
      "epoch": 0.43152015541525013,
      "grad_norm": 0.3822302222251892,
      "learning_rate": 0.0001,
      "loss": 1.7353,
      "step": 1777
    },
    {
      "epoch": 0.43176299174356486,
      "grad_norm": 0.37838810682296753,
      "learning_rate": 0.0001,
      "loss": 1.866,
      "step": 1778
    },
    {
      "epoch": 0.43200582807187954,
      "grad_norm": 0.36003077030181885,
      "learning_rate": 0.0001,
      "loss": 1.8084,
      "step": 1779
    },
    {
      "epoch": 0.43224866440019427,
      "grad_norm": 0.3723047375679016,
      "learning_rate": 0.0001,
      "loss": 1.8549,
      "step": 1780
    },
    {
      "epoch": 0.432491500728509,
      "grad_norm": 0.3599883019924164,
      "learning_rate": 0.0001,
      "loss": 1.692,
      "step": 1781
    },
    {
      "epoch": 0.4327343370568237,
      "grad_norm": 0.35087719559669495,
      "learning_rate": 0.0001,
      "loss": 1.7804,
      "step": 1782
    },
    {
      "epoch": 0.4329771733851384,
      "grad_norm": 0.3767469823360443,
      "learning_rate": 0.0001,
      "loss": 1.7284,
      "step": 1783
    },
    {
      "epoch": 0.43322000971345315,
      "grad_norm": 0.33156484365463257,
      "learning_rate": 0.0001,
      "loss": 1.6055,
      "step": 1784
    },
    {
      "epoch": 0.4334628460417678,
      "grad_norm": 0.3367984890937805,
      "learning_rate": 0.0001,
      "loss": 1.5564,
      "step": 1785
    },
    {
      "epoch": 0.43370568237008256,
      "grad_norm": 0.3552202880382538,
      "learning_rate": 0.0001,
      "loss": 1.829,
      "step": 1786
    },
    {
      "epoch": 0.4339485186983973,
      "grad_norm": 0.353182852268219,
      "learning_rate": 0.0001,
      "loss": 1.6823,
      "step": 1787
    },
    {
      "epoch": 0.434191355026712,
      "grad_norm": 0.3680180013179779,
      "learning_rate": 0.0001,
      "loss": 1.7933,
      "step": 1788
    },
    {
      "epoch": 0.4344341913550267,
      "grad_norm": 0.338387668132782,
      "learning_rate": 0.0001,
      "loss": 1.7065,
      "step": 1789
    },
    {
      "epoch": 0.43467702768334143,
      "grad_norm": 0.36020609736442566,
      "learning_rate": 0.0001,
      "loss": 1.8273,
      "step": 1790
    },
    {
      "epoch": 0.43491986401165617,
      "grad_norm": 0.3294867277145386,
      "learning_rate": 0.0001,
      "loss": 1.6185,
      "step": 1791
    },
    {
      "epoch": 0.43516270033997084,
      "grad_norm": 0.36147022247314453,
      "learning_rate": 0.0001,
      "loss": 1.4078,
      "step": 1792
    },
    {
      "epoch": 0.4354055366682856,
      "grad_norm": 0.35484182834625244,
      "learning_rate": 0.0001,
      "loss": 1.6535,
      "step": 1793
    },
    {
      "epoch": 0.4356483729966003,
      "grad_norm": 0.33079761266708374,
      "learning_rate": 0.0001,
      "loss": 1.585,
      "step": 1794
    },
    {
      "epoch": 0.435891209324915,
      "grad_norm": 0.35023409128189087,
      "learning_rate": 0.0001,
      "loss": 1.73,
      "step": 1795
    },
    {
      "epoch": 0.4361340456532297,
      "grad_norm": 0.3546440601348877,
      "learning_rate": 0.0001,
      "loss": 1.7474,
      "step": 1796
    },
    {
      "epoch": 0.43637688198154445,
      "grad_norm": 0.35574740171432495,
      "learning_rate": 0.0001,
      "loss": 1.9602,
      "step": 1797
    },
    {
      "epoch": 0.43661971830985913,
      "grad_norm": 0.3341620862483978,
      "learning_rate": 0.0001,
      "loss": 1.6204,
      "step": 1798
    },
    {
      "epoch": 0.43686255463817386,
      "grad_norm": 0.3560903072357178,
      "learning_rate": 0.0001,
      "loss": 1.6419,
      "step": 1799
    },
    {
      "epoch": 0.4371053909664886,
      "grad_norm": 0.35073208808898926,
      "learning_rate": 0.0001,
      "loss": 1.6708,
      "step": 1800
    },
    {
      "epoch": 0.4373482272948033,
      "grad_norm": 0.35862451791763306,
      "learning_rate": 0.0001,
      "loss": 1.7134,
      "step": 1801
    },
    {
      "epoch": 0.437591063623118,
      "grad_norm": 0.36779993772506714,
      "learning_rate": 0.0001,
      "loss": 1.8379,
      "step": 1802
    },
    {
      "epoch": 0.43783389995143274,
      "grad_norm": 0.35280412435531616,
      "learning_rate": 0.0001,
      "loss": 1.9213,
      "step": 1803
    },
    {
      "epoch": 0.43807673627974747,
      "grad_norm": 0.3488386869430542,
      "learning_rate": 0.0001,
      "loss": 1.7665,
      "step": 1804
    },
    {
      "epoch": 0.43831957260806215,
      "grad_norm": 0.3537375330924988,
      "learning_rate": 0.0001,
      "loss": 1.6809,
      "step": 1805
    },
    {
      "epoch": 0.4385624089363769,
      "grad_norm": 0.3733062148094177,
      "learning_rate": 0.0001,
      "loss": 1.9652,
      "step": 1806
    },
    {
      "epoch": 0.4388052452646916,
      "grad_norm": 0.3548164367675781,
      "learning_rate": 0.0001,
      "loss": 1.8302,
      "step": 1807
    },
    {
      "epoch": 0.4390480815930063,
      "grad_norm": 0.35301345586776733,
      "learning_rate": 0.0001,
      "loss": 1.861,
      "step": 1808
    },
    {
      "epoch": 0.439290917921321,
      "grad_norm": 0.34159356355667114,
      "learning_rate": 0.0001,
      "loss": 1.6349,
      "step": 1809
    },
    {
      "epoch": 0.43953375424963576,
      "grad_norm": 0.36830708384513855,
      "learning_rate": 0.0001,
      "loss": 1.8702,
      "step": 1810
    },
    {
      "epoch": 0.4397765905779505,
      "grad_norm": 0.3883112370967865,
      "learning_rate": 0.0001,
      "loss": 1.9395,
      "step": 1811
    },
    {
      "epoch": 0.44001942690626517,
      "grad_norm": 0.34868156909942627,
      "learning_rate": 0.0001,
      "loss": 1.7817,
      "step": 1812
    },
    {
      "epoch": 0.4402622632345799,
      "grad_norm": 0.36234238743782043,
      "learning_rate": 0.0001,
      "loss": 1.8744,
      "step": 1813
    },
    {
      "epoch": 0.44050509956289463,
      "grad_norm": 0.35224321484565735,
      "learning_rate": 0.0001,
      "loss": 1.8057,
      "step": 1814
    },
    {
      "epoch": 0.4407479358912093,
      "grad_norm": 0.3585010766983032,
      "learning_rate": 0.0001,
      "loss": 1.7047,
      "step": 1815
    },
    {
      "epoch": 0.44099077221952404,
      "grad_norm": 0.3444727957248688,
      "learning_rate": 0.0001,
      "loss": 1.6929,
      "step": 1816
    },
    {
      "epoch": 0.4412336085478388,
      "grad_norm": 0.34820786118507385,
      "learning_rate": 0.0001,
      "loss": 1.7742,
      "step": 1817
    },
    {
      "epoch": 0.44147644487615345,
      "grad_norm": 0.32335302233695984,
      "learning_rate": 0.0001,
      "loss": 1.6456,
      "step": 1818
    },
    {
      "epoch": 0.4417192812044682,
      "grad_norm": 0.3473454713821411,
      "learning_rate": 0.0001,
      "loss": 1.6254,
      "step": 1819
    },
    {
      "epoch": 0.4419621175327829,
      "grad_norm": 0.32856178283691406,
      "learning_rate": 0.0001,
      "loss": 1.5016,
      "step": 1820
    },
    {
      "epoch": 0.4422049538610976,
      "grad_norm": 0.346571147441864,
      "learning_rate": 0.0001,
      "loss": 1.7544,
      "step": 1821
    },
    {
      "epoch": 0.4424477901894123,
      "grad_norm": 0.35231831669807434,
      "learning_rate": 0.0001,
      "loss": 1.6208,
      "step": 1822
    },
    {
      "epoch": 0.44269062651772706,
      "grad_norm": 0.35263708233833313,
      "learning_rate": 0.0001,
      "loss": 1.7064,
      "step": 1823
    },
    {
      "epoch": 0.4429334628460418,
      "grad_norm": 0.3441479206085205,
      "learning_rate": 0.0001,
      "loss": 1.7732,
      "step": 1824
    },
    {
      "epoch": 0.44317629917435647,
      "grad_norm": 0.3736101984977722,
      "learning_rate": 0.0001,
      "loss": 1.7838,
      "step": 1825
    },
    {
      "epoch": 0.4434191355026712,
      "grad_norm": 0.3775104880332947,
      "learning_rate": 0.0001,
      "loss": 1.6946,
      "step": 1826
    },
    {
      "epoch": 0.44366197183098594,
      "grad_norm": 0.36177685856819153,
      "learning_rate": 0.0001,
      "loss": 1.9272,
      "step": 1827
    },
    {
      "epoch": 0.4439048081593006,
      "grad_norm": 0.35995036363601685,
      "learning_rate": 0.0001,
      "loss": 1.8043,
      "step": 1828
    },
    {
      "epoch": 0.44414764448761535,
      "grad_norm": 0.36013609170913696,
      "learning_rate": 0.0001,
      "loss": 1.8527,
      "step": 1829
    },
    {
      "epoch": 0.4443904808159301,
      "grad_norm": 0.35395514965057373,
      "learning_rate": 0.0001,
      "loss": 1.7323,
      "step": 1830
    },
    {
      "epoch": 0.44463331714424476,
      "grad_norm": 0.3533592224121094,
      "learning_rate": 0.0001,
      "loss": 1.7319,
      "step": 1831
    },
    {
      "epoch": 0.4448761534725595,
      "grad_norm": 0.350333034992218,
      "learning_rate": 0.0001,
      "loss": 1.6835,
      "step": 1832
    },
    {
      "epoch": 0.4451189898008742,
      "grad_norm": 0.3550693094730377,
      "learning_rate": 0.0001,
      "loss": 1.8188,
      "step": 1833
    },
    {
      "epoch": 0.44536182612918895,
      "grad_norm": 0.3411439061164856,
      "learning_rate": 0.0001,
      "loss": 1.7458,
      "step": 1834
    },
    {
      "epoch": 0.44560466245750363,
      "grad_norm": 0.3670227825641632,
      "learning_rate": 0.0001,
      "loss": 1.7695,
      "step": 1835
    },
    {
      "epoch": 0.44584749878581836,
      "grad_norm": 0.333892822265625,
      "learning_rate": 0.0001,
      "loss": 1.6497,
      "step": 1836
    },
    {
      "epoch": 0.4460903351141331,
      "grad_norm": 0.3675907850265503,
      "learning_rate": 0.0001,
      "loss": 1.7669,
      "step": 1837
    },
    {
      "epoch": 0.4463331714424478,
      "grad_norm": 0.36830997467041016,
      "learning_rate": 0.0001,
      "loss": 1.9007,
      "step": 1838
    },
    {
      "epoch": 0.4465760077707625,
      "grad_norm": 0.34332361817359924,
      "learning_rate": 0.0001,
      "loss": 1.62,
      "step": 1839
    },
    {
      "epoch": 0.44681884409907724,
      "grad_norm": 0.3424611985683441,
      "learning_rate": 0.0001,
      "loss": 1.5369,
      "step": 1840
    },
    {
      "epoch": 0.4470616804273919,
      "grad_norm": 0.3539303243160248,
      "learning_rate": 0.0001,
      "loss": 1.7515,
      "step": 1841
    },
    {
      "epoch": 0.44730451675570665,
      "grad_norm": 0.3896627724170685,
      "learning_rate": 0.0001,
      "loss": 1.861,
      "step": 1842
    },
    {
      "epoch": 0.4475473530840214,
      "grad_norm": 0.3435671031475067,
      "learning_rate": 0.0001,
      "loss": 1.6936,
      "step": 1843
    },
    {
      "epoch": 0.44779018941233606,
      "grad_norm": 0.3572455942630768,
      "learning_rate": 0.0001,
      "loss": 1.7223,
      "step": 1844
    },
    {
      "epoch": 0.4480330257406508,
      "grad_norm": 0.3949750065803528,
      "learning_rate": 0.0001,
      "loss": 1.8012,
      "step": 1845
    },
    {
      "epoch": 0.4482758620689655,
      "grad_norm": 0.3492611050605774,
      "learning_rate": 0.0001,
      "loss": 1.645,
      "step": 1846
    },
    {
      "epoch": 0.44851869839728026,
      "grad_norm": 0.33171871304512024,
      "learning_rate": 0.0001,
      "loss": 1.5878,
      "step": 1847
    },
    {
      "epoch": 0.44876153472559493,
      "grad_norm": 0.3477494418621063,
      "learning_rate": 0.0001,
      "loss": 1.7476,
      "step": 1848
    },
    {
      "epoch": 0.44900437105390967,
      "grad_norm": 0.35885927081108093,
      "learning_rate": 0.0001,
      "loss": 1.8095,
      "step": 1849
    },
    {
      "epoch": 0.4492472073822244,
      "grad_norm": 0.35665690898895264,
      "learning_rate": 0.0001,
      "loss": 1.8063,
      "step": 1850
    },
    {
      "epoch": 0.4494900437105391,
      "grad_norm": 0.31958791613578796,
      "learning_rate": 0.0001,
      "loss": 1.5572,
      "step": 1851
    },
    {
      "epoch": 0.4497328800388538,
      "grad_norm": 0.34891462326049805,
      "learning_rate": 0.0001,
      "loss": 1.5507,
      "step": 1852
    },
    {
      "epoch": 0.44997571636716854,
      "grad_norm": 0.34152233600616455,
      "learning_rate": 0.0001,
      "loss": 1.7601,
      "step": 1853
    },
    {
      "epoch": 0.4502185526954832,
      "grad_norm": 0.37790095806121826,
      "learning_rate": 0.0001,
      "loss": 1.8899,
      "step": 1854
    },
    {
      "epoch": 0.45046138902379795,
      "grad_norm": 0.35623136162757874,
      "learning_rate": 0.0001,
      "loss": 1.8179,
      "step": 1855
    },
    {
      "epoch": 0.4507042253521127,
      "grad_norm": 0.3646155297756195,
      "learning_rate": 0.0001,
      "loss": 1.8371,
      "step": 1856
    },
    {
      "epoch": 0.4509470616804274,
      "grad_norm": 0.3449878692626953,
      "learning_rate": 0.0001,
      "loss": 1.7263,
      "step": 1857
    },
    {
      "epoch": 0.4511898980087421,
      "grad_norm": 0.3796631991863251,
      "learning_rate": 0.0001,
      "loss": 1.8453,
      "step": 1858
    },
    {
      "epoch": 0.45143273433705683,
      "grad_norm": 0.36037197709083557,
      "learning_rate": 0.0001,
      "loss": 1.7422,
      "step": 1859
    },
    {
      "epoch": 0.45167557066537156,
      "grad_norm": 0.3578833341598511,
      "learning_rate": 0.0001,
      "loss": 1.7522,
      "step": 1860
    },
    {
      "epoch": 0.45191840699368624,
      "grad_norm": 0.36234596371650696,
      "learning_rate": 0.0001,
      "loss": 1.9205,
      "step": 1861
    },
    {
      "epoch": 0.45216124332200097,
      "grad_norm": 0.34786903858184814,
      "learning_rate": 0.0001,
      "loss": 1.6425,
      "step": 1862
    },
    {
      "epoch": 0.4524040796503157,
      "grad_norm": 0.3553171455860138,
      "learning_rate": 0.0001,
      "loss": 1.7245,
      "step": 1863
    },
    {
      "epoch": 0.4526469159786304,
      "grad_norm": 0.3154301345348358,
      "learning_rate": 0.0001,
      "loss": 1.4654,
      "step": 1864
    },
    {
      "epoch": 0.4528897523069451,
      "grad_norm": 0.34884896874427795,
      "learning_rate": 0.0001,
      "loss": 1.6409,
      "step": 1865
    },
    {
      "epoch": 0.45313258863525985,
      "grad_norm": 0.353974848985672,
      "learning_rate": 0.0001,
      "loss": 1.6605,
      "step": 1866
    },
    {
      "epoch": 0.4533754249635745,
      "grad_norm": 0.3681231141090393,
      "learning_rate": 0.0001,
      "loss": 1.7232,
      "step": 1867
    },
    {
      "epoch": 0.45361826129188926,
      "grad_norm": 0.3452052175998688,
      "learning_rate": 0.0001,
      "loss": 1.6828,
      "step": 1868
    },
    {
      "epoch": 0.453861097620204,
      "grad_norm": 0.3314920663833618,
      "learning_rate": 0.0001,
      "loss": 1.609,
      "step": 1869
    },
    {
      "epoch": 0.4541039339485187,
      "grad_norm": 0.36491215229034424,
      "learning_rate": 0.0001,
      "loss": 1.683,
      "step": 1870
    },
    {
      "epoch": 0.4543467702768334,
      "grad_norm": 0.3672131896018982,
      "learning_rate": 0.0001,
      "loss": 1.7338,
      "step": 1871
    },
    {
      "epoch": 0.45458960660514813,
      "grad_norm": 0.3563331663608551,
      "learning_rate": 0.0001,
      "loss": 1.877,
      "step": 1872
    },
    {
      "epoch": 0.45483244293346287,
      "grad_norm": 0.35385245084762573,
      "learning_rate": 0.0001,
      "loss": 1.6971,
      "step": 1873
    },
    {
      "epoch": 0.45507527926177754,
      "grad_norm": 0.33963268995285034,
      "learning_rate": 0.0001,
      "loss": 1.5958,
      "step": 1874
    },
    {
      "epoch": 0.4553181155900923,
      "grad_norm": 0.3327486515045166,
      "learning_rate": 0.0001,
      "loss": 1.7109,
      "step": 1875
    },
    {
      "epoch": 0.455560951918407,
      "grad_norm": 0.35015425086021423,
      "learning_rate": 0.0001,
      "loss": 1.7014,
      "step": 1876
    },
    {
      "epoch": 0.4558037882467217,
      "grad_norm": 0.3478430211544037,
      "learning_rate": 0.0001,
      "loss": 1.7751,
      "step": 1877
    },
    {
      "epoch": 0.4560466245750364,
      "grad_norm": 0.3734671175479889,
      "learning_rate": 0.0001,
      "loss": 1.8478,
      "step": 1878
    },
    {
      "epoch": 0.45628946090335115,
      "grad_norm": 0.3410831391811371,
      "learning_rate": 0.0001,
      "loss": 1.5819,
      "step": 1879
    },
    {
      "epoch": 0.4565322972316659,
      "grad_norm": 0.3469892740249634,
      "learning_rate": 0.0001,
      "loss": 1.706,
      "step": 1880
    },
    {
      "epoch": 0.45677513355998056,
      "grad_norm": 0.35892051458358765,
      "learning_rate": 0.0001,
      "loss": 1.6773,
      "step": 1881
    },
    {
      "epoch": 0.4570179698882953,
      "grad_norm": 0.3807806372642517,
      "learning_rate": 0.0001,
      "loss": 1.8788,
      "step": 1882
    },
    {
      "epoch": 0.45726080621661,
      "grad_norm": 0.34584805369377136,
      "learning_rate": 0.0001,
      "loss": 1.7598,
      "step": 1883
    },
    {
      "epoch": 0.4575036425449247,
      "grad_norm": 0.4014298617839813,
      "learning_rate": 0.0001,
      "loss": 1.932,
      "step": 1884
    },
    {
      "epoch": 0.45774647887323944,
      "grad_norm": 0.36960092186927795,
      "learning_rate": 0.0001,
      "loss": 1.8015,
      "step": 1885
    },
    {
      "epoch": 0.45798931520155417,
      "grad_norm": 0.34163033962249756,
      "learning_rate": 0.0001,
      "loss": 1.593,
      "step": 1886
    },
    {
      "epoch": 0.45823215152986885,
      "grad_norm": 0.36526426672935486,
      "learning_rate": 0.0001,
      "loss": 1.8164,
      "step": 1887
    },
    {
      "epoch": 0.4584749878581836,
      "grad_norm": 0.31698787212371826,
      "learning_rate": 0.0001,
      "loss": 1.65,
      "step": 1888
    },
    {
      "epoch": 0.4587178241864983,
      "grad_norm": 0.3732241988182068,
      "learning_rate": 0.0001,
      "loss": 1.8166,
      "step": 1889
    },
    {
      "epoch": 0.458960660514813,
      "grad_norm": 0.38333624601364136,
      "learning_rate": 0.0001,
      "loss": 1.8435,
      "step": 1890
    },
    {
      "epoch": 0.4592034968431277,
      "grad_norm": 0.3748052716255188,
      "learning_rate": 0.0001,
      "loss": 1.6272,
      "step": 1891
    },
    {
      "epoch": 0.45944633317144246,
      "grad_norm": 0.34906819462776184,
      "learning_rate": 0.0001,
      "loss": 1.719,
      "step": 1892
    },
    {
      "epoch": 0.4596891694997572,
      "grad_norm": 0.3537774384021759,
      "learning_rate": 0.0001,
      "loss": 1.6984,
      "step": 1893
    },
    {
      "epoch": 0.45993200582807187,
      "grad_norm": 0.3544154167175293,
      "learning_rate": 0.0001,
      "loss": 1.8092,
      "step": 1894
    },
    {
      "epoch": 0.4601748421563866,
      "grad_norm": 0.3754320740699768,
      "learning_rate": 0.0001,
      "loss": 1.5766,
      "step": 1895
    },
    {
      "epoch": 0.46041767848470133,
      "grad_norm": 0.34848159551620483,
      "learning_rate": 0.0001,
      "loss": 1.6516,
      "step": 1896
    },
    {
      "epoch": 0.460660514813016,
      "grad_norm": 0.3676202893257141,
      "learning_rate": 0.0001,
      "loss": 1.6977,
      "step": 1897
    },
    {
      "epoch": 0.46090335114133074,
      "grad_norm": 0.323017954826355,
      "learning_rate": 0.0001,
      "loss": 1.605,
      "step": 1898
    },
    {
      "epoch": 0.4611461874696455,
      "grad_norm": 0.3825530707836151,
      "learning_rate": 0.0001,
      "loss": 1.7385,
      "step": 1899
    },
    {
      "epoch": 0.46138902379796015,
      "grad_norm": 0.3630683124065399,
      "learning_rate": 0.0001,
      "loss": 1.786,
      "step": 1900
    },
    {
      "epoch": 0.4616318601262749,
      "grad_norm": 0.33266085386276245,
      "learning_rate": 0.0001,
      "loss": 1.6726,
      "step": 1901
    },
    {
      "epoch": 0.4618746964545896,
      "grad_norm": 0.3453136384487152,
      "learning_rate": 0.0001,
      "loss": 1.5764,
      "step": 1902
    },
    {
      "epoch": 0.46211753278290435,
      "grad_norm": 0.3523610830307007,
      "learning_rate": 0.0001,
      "loss": 1.7086,
      "step": 1903
    },
    {
      "epoch": 0.462360369111219,
      "grad_norm": 0.34262746572494507,
      "learning_rate": 0.0001,
      "loss": 1.6824,
      "step": 1904
    },
    {
      "epoch": 0.46260320543953376,
      "grad_norm": 0.3551229238510132,
      "learning_rate": 0.0001,
      "loss": 1.7265,
      "step": 1905
    },
    {
      "epoch": 0.4628460417678485,
      "grad_norm": 0.3580963909626007,
      "learning_rate": 0.0001,
      "loss": 1.7887,
      "step": 1906
    },
    {
      "epoch": 0.46308887809616317,
      "grad_norm": 0.32605284452438354,
      "learning_rate": 0.0001,
      "loss": 1.5617,
      "step": 1907
    },
    {
      "epoch": 0.4633317144244779,
      "grad_norm": 0.3345804512500763,
      "learning_rate": 0.0001,
      "loss": 1.5542,
      "step": 1908
    },
    {
      "epoch": 0.46357455075279264,
      "grad_norm": 0.35500892996788025,
      "learning_rate": 0.0001,
      "loss": 1.6323,
      "step": 1909
    },
    {
      "epoch": 0.4638173870811073,
      "grad_norm": 0.3939019739627838,
      "learning_rate": 0.0001,
      "loss": 1.6667,
      "step": 1910
    },
    {
      "epoch": 0.46406022340942205,
      "grad_norm": 0.3847699463367462,
      "learning_rate": 0.0001,
      "loss": 1.869,
      "step": 1911
    },
    {
      "epoch": 0.4643030597377368,
      "grad_norm": 0.35265713930130005,
      "learning_rate": 0.0001,
      "loss": 1.8361,
      "step": 1912
    },
    {
      "epoch": 0.46454589606605146,
      "grad_norm": 0.3612956404685974,
      "learning_rate": 0.0001,
      "loss": 1.867,
      "step": 1913
    },
    {
      "epoch": 0.4647887323943662,
      "grad_norm": 0.31383344531059265,
      "learning_rate": 0.0001,
      "loss": 1.4885,
      "step": 1914
    },
    {
      "epoch": 0.4650315687226809,
      "grad_norm": 0.3553354740142822,
      "learning_rate": 0.0001,
      "loss": 1.7453,
      "step": 1915
    },
    {
      "epoch": 0.46527440505099565,
      "grad_norm": 0.34411731362342834,
      "learning_rate": 0.0001,
      "loss": 1.5518,
      "step": 1916
    },
    {
      "epoch": 0.46551724137931033,
      "grad_norm": 0.3603088855743408,
      "learning_rate": 0.0001,
      "loss": 1.7127,
      "step": 1917
    },
    {
      "epoch": 0.46576007770762506,
      "grad_norm": 0.36315515637397766,
      "learning_rate": 0.0001,
      "loss": 1.7099,
      "step": 1918
    },
    {
      "epoch": 0.4660029140359398,
      "grad_norm": 0.33600983023643494,
      "learning_rate": 0.0001,
      "loss": 1.6182,
      "step": 1919
    },
    {
      "epoch": 0.4662457503642545,
      "grad_norm": 0.3534862995147705,
      "learning_rate": 0.0001,
      "loss": 1.6327,
      "step": 1920
    },
    {
      "epoch": 0.4664885866925692,
      "grad_norm": 0.3495807647705078,
      "learning_rate": 0.0001,
      "loss": 1.7932,
      "step": 1921
    },
    {
      "epoch": 0.46673142302088394,
      "grad_norm": 0.3636340796947479,
      "learning_rate": 0.0001,
      "loss": 1.7882,
      "step": 1922
    },
    {
      "epoch": 0.4669742593491986,
      "grad_norm": 0.3711027503013611,
      "learning_rate": 0.0001,
      "loss": 1.7182,
      "step": 1923
    },
    {
      "epoch": 0.46721709567751335,
      "grad_norm": 0.34208375215530396,
      "learning_rate": 0.0001,
      "loss": 1.6319,
      "step": 1924
    },
    {
      "epoch": 0.4674599320058281,
      "grad_norm": 0.3585227429866791,
      "learning_rate": 0.0001,
      "loss": 1.7672,
      "step": 1925
    },
    {
      "epoch": 0.4677027683341428,
      "grad_norm": 0.36455008387565613,
      "learning_rate": 0.0001,
      "loss": 1.7176,
      "step": 1926
    },
    {
      "epoch": 0.4679456046624575,
      "grad_norm": 0.3617648184299469,
      "learning_rate": 0.0001,
      "loss": 1.77,
      "step": 1927
    },
    {
      "epoch": 0.4681884409907722,
      "grad_norm": 0.3547668755054474,
      "learning_rate": 0.0001,
      "loss": 1.6885,
      "step": 1928
    },
    {
      "epoch": 0.46843127731908696,
      "grad_norm": 0.3823624849319458,
      "learning_rate": 0.0001,
      "loss": 1.8885,
      "step": 1929
    },
    {
      "epoch": 0.46867411364740164,
      "grad_norm": 0.36009302735328674,
      "learning_rate": 0.0001,
      "loss": 1.6773,
      "step": 1930
    },
    {
      "epoch": 0.46891694997571637,
      "grad_norm": 0.34541547298431396,
      "learning_rate": 0.0001,
      "loss": 1.7179,
      "step": 1931
    },
    {
      "epoch": 0.4691597863040311,
      "grad_norm": 0.3517388105392456,
      "learning_rate": 0.0001,
      "loss": 1.8771,
      "step": 1932
    },
    {
      "epoch": 0.4694026226323458,
      "grad_norm": 0.3726664185523987,
      "learning_rate": 0.0001,
      "loss": 1.9417,
      "step": 1933
    },
    {
      "epoch": 0.4696454589606605,
      "grad_norm": 0.3713761568069458,
      "learning_rate": 0.0001,
      "loss": 1.8633,
      "step": 1934
    },
    {
      "epoch": 0.46988829528897524,
      "grad_norm": 0.34845829010009766,
      "learning_rate": 0.0001,
      "loss": 1.6415,
      "step": 1935
    },
    {
      "epoch": 0.4701311316172899,
      "grad_norm": 0.3643852770328522,
      "learning_rate": 0.0001,
      "loss": 1.8709,
      "step": 1936
    },
    {
      "epoch": 0.47037396794560465,
      "grad_norm": 0.37239712476730347,
      "learning_rate": 0.0001,
      "loss": 1.8805,
      "step": 1937
    },
    {
      "epoch": 0.4706168042739194,
      "grad_norm": 0.34127670526504517,
      "learning_rate": 0.0001,
      "loss": 1.7415,
      "step": 1938
    },
    {
      "epoch": 0.4708596406022341,
      "grad_norm": 0.3482852876186371,
      "learning_rate": 0.0001,
      "loss": 1.8389,
      "step": 1939
    },
    {
      "epoch": 0.4711024769305488,
      "grad_norm": 0.35871124267578125,
      "learning_rate": 0.0001,
      "loss": 1.7189,
      "step": 1940
    },
    {
      "epoch": 0.47134531325886353,
      "grad_norm": 0.33642691373825073,
      "learning_rate": 0.0001,
      "loss": 1.5999,
      "step": 1941
    },
    {
      "epoch": 0.47158814958717826,
      "grad_norm": 0.36310839653015137,
      "learning_rate": 0.0001,
      "loss": 1.7654,
      "step": 1942
    },
    {
      "epoch": 0.47183098591549294,
      "grad_norm": 0.334359735250473,
      "learning_rate": 0.0001,
      "loss": 1.5615,
      "step": 1943
    },
    {
      "epoch": 0.4720738222438077,
      "grad_norm": 0.3224147856235504,
      "learning_rate": 0.0001,
      "loss": 1.5394,
      "step": 1944
    },
    {
      "epoch": 0.4723166585721224,
      "grad_norm": 0.34409651160240173,
      "learning_rate": 0.0001,
      "loss": 1.6985,
      "step": 1945
    },
    {
      "epoch": 0.4725594949004371,
      "grad_norm": 0.346697598695755,
      "learning_rate": 0.0001,
      "loss": 1.5183,
      "step": 1946
    },
    {
      "epoch": 0.4728023312287518,
      "grad_norm": 0.33633822202682495,
      "learning_rate": 0.0001,
      "loss": 1.6303,
      "step": 1947
    },
    {
      "epoch": 0.47304516755706655,
      "grad_norm": 0.3662882447242737,
      "learning_rate": 0.0001,
      "loss": 1.9724,
      "step": 1948
    },
    {
      "epoch": 0.4732880038853812,
      "grad_norm": 0.35439127683639526,
      "learning_rate": 0.0001,
      "loss": 1.5703,
      "step": 1949
    },
    {
      "epoch": 0.47353084021369596,
      "grad_norm": 0.34117916226387024,
      "learning_rate": 0.0001,
      "loss": 1.5617,
      "step": 1950
    },
    {
      "epoch": 0.4737736765420107,
      "grad_norm": 0.3378322124481201,
      "learning_rate": 0.0001,
      "loss": 1.704,
      "step": 1951
    },
    {
      "epoch": 0.4740165128703254,
      "grad_norm": 0.32980063557624817,
      "learning_rate": 0.0001,
      "loss": 1.631,
      "step": 1952
    },
    {
      "epoch": 0.4742593491986401,
      "grad_norm": 0.37976697087287903,
      "learning_rate": 0.0001,
      "loss": 1.8591,
      "step": 1953
    },
    {
      "epoch": 0.47450218552695483,
      "grad_norm": 0.35602810978889465,
      "learning_rate": 0.0001,
      "loss": 1.6447,
      "step": 1954
    },
    {
      "epoch": 0.47474502185526957,
      "grad_norm": 0.34376612305641174,
      "learning_rate": 0.0001,
      "loss": 1.7337,
      "step": 1955
    },
    {
      "epoch": 0.47498785818358424,
      "grad_norm": 0.3398648798465729,
      "learning_rate": 0.0001,
      "loss": 1.7641,
      "step": 1956
    },
    {
      "epoch": 0.475230694511899,
      "grad_norm": 0.3545699119567871,
      "learning_rate": 0.0001,
      "loss": 1.7535,
      "step": 1957
    },
    {
      "epoch": 0.4754735308402137,
      "grad_norm": 0.31818118691444397,
      "learning_rate": 0.0001,
      "loss": 1.6283,
      "step": 1958
    },
    {
      "epoch": 0.4757163671685284,
      "grad_norm": 0.3350203335285187,
      "learning_rate": 0.0001,
      "loss": 1.6672,
      "step": 1959
    },
    {
      "epoch": 0.4759592034968431,
      "grad_norm": 0.37390241026878357,
      "learning_rate": 0.0001,
      "loss": 1.7388,
      "step": 1960
    },
    {
      "epoch": 0.47620203982515785,
      "grad_norm": 0.35129526257514954,
      "learning_rate": 0.0001,
      "loss": 1.7303,
      "step": 1961
    },
    {
      "epoch": 0.4764448761534726,
      "grad_norm": 0.35428422689437866,
      "learning_rate": 0.0001,
      "loss": 1.8535,
      "step": 1962
    },
    {
      "epoch": 0.47668771248178726,
      "grad_norm": 0.3445446491241455,
      "learning_rate": 0.0001,
      "loss": 1.7052,
      "step": 1963
    },
    {
      "epoch": 0.476930548810102,
      "grad_norm": 0.35369494557380676,
      "learning_rate": 0.0001,
      "loss": 1.7234,
      "step": 1964
    },
    {
      "epoch": 0.47717338513841673,
      "grad_norm": 0.3897012770175934,
      "learning_rate": 0.0001,
      "loss": 1.7988,
      "step": 1965
    },
    {
      "epoch": 0.4774162214667314,
      "grad_norm": 0.3370416760444641,
      "learning_rate": 0.0001,
      "loss": 1.7065,
      "step": 1966
    },
    {
      "epoch": 0.47765905779504614,
      "grad_norm": 0.3592875301837921,
      "learning_rate": 0.0001,
      "loss": 1.7909,
      "step": 1967
    },
    {
      "epoch": 0.47790189412336087,
      "grad_norm": 0.3603609800338745,
      "learning_rate": 0.0001,
      "loss": 1.5402,
      "step": 1968
    },
    {
      "epoch": 0.47814473045167555,
      "grad_norm": 0.34685638546943665,
      "learning_rate": 0.0001,
      "loss": 1.734,
      "step": 1969
    },
    {
      "epoch": 0.4783875667799903,
      "grad_norm": 0.3325386643409729,
      "learning_rate": 0.0001,
      "loss": 1.4731,
      "step": 1970
    },
    {
      "epoch": 0.478630403108305,
      "grad_norm": 0.3731539249420166,
      "learning_rate": 0.0001,
      "loss": 1.7351,
      "step": 1971
    },
    {
      "epoch": 0.4788732394366197,
      "grad_norm": 0.34925469756126404,
      "learning_rate": 0.0001,
      "loss": 1.7128,
      "step": 1972
    },
    {
      "epoch": 0.4791160757649344,
      "grad_norm": 0.3686675429344177,
      "learning_rate": 0.0001,
      "loss": 1.7226,
      "step": 1973
    },
    {
      "epoch": 0.47935891209324916,
      "grad_norm": 0.3480817675590515,
      "learning_rate": 0.0001,
      "loss": 1.6505,
      "step": 1974
    },
    {
      "epoch": 0.4796017484215639,
      "grad_norm": 0.3474791944026947,
      "learning_rate": 0.0001,
      "loss": 1.5714,
      "step": 1975
    },
    {
      "epoch": 0.47984458474987857,
      "grad_norm": 0.36596643924713135,
      "learning_rate": 0.0001,
      "loss": 1.8255,
      "step": 1976
    },
    {
      "epoch": 0.4800874210781933,
      "grad_norm": 0.3744461238384247,
      "learning_rate": 0.0001,
      "loss": 1.8206,
      "step": 1977
    },
    {
      "epoch": 0.48033025740650803,
      "grad_norm": 0.3452211618423462,
      "learning_rate": 0.0001,
      "loss": 1.7232,
      "step": 1978
    },
    {
      "epoch": 0.4805730937348227,
      "grad_norm": 0.3522043824195862,
      "learning_rate": 0.0001,
      "loss": 1.5923,
      "step": 1979
    },
    {
      "epoch": 0.48081593006313744,
      "grad_norm": 0.36870819330215454,
      "learning_rate": 0.0001,
      "loss": 1.8372,
      "step": 1980
    },
    {
      "epoch": 0.4810587663914522,
      "grad_norm": 0.36815735697746277,
      "learning_rate": 0.0001,
      "loss": 1.7929,
      "step": 1981
    },
    {
      "epoch": 0.48130160271976685,
      "grad_norm": 0.3665809631347656,
      "learning_rate": 0.0001,
      "loss": 1.7355,
      "step": 1982
    },
    {
      "epoch": 0.4815444390480816,
      "grad_norm": 0.4022822678089142,
      "learning_rate": 0.0001,
      "loss": 2.1035,
      "step": 1983
    },
    {
      "epoch": 0.4817872753763963,
      "grad_norm": 0.336421400308609,
      "learning_rate": 0.0001,
      "loss": 1.6003,
      "step": 1984
    },
    {
      "epoch": 0.48203011170471105,
      "grad_norm": 0.35770168900489807,
      "learning_rate": 0.0001,
      "loss": 1.7644,
      "step": 1985
    },
    {
      "epoch": 0.48227294803302573,
      "grad_norm": 0.3650563657283783,
      "learning_rate": 0.0001,
      "loss": 1.8717,
      "step": 1986
    },
    {
      "epoch": 0.48251578436134046,
      "grad_norm": 0.39090093970298767,
      "learning_rate": 0.0001,
      "loss": 1.7926,
      "step": 1987
    },
    {
      "epoch": 0.4827586206896552,
      "grad_norm": 0.37530893087387085,
      "learning_rate": 0.0001,
      "loss": 1.6973,
      "step": 1988
    },
    {
      "epoch": 0.48300145701796987,
      "grad_norm": 0.3707762062549591,
      "learning_rate": 0.0001,
      "loss": 1.8737,
      "step": 1989
    },
    {
      "epoch": 0.4832442933462846,
      "grad_norm": 0.35839855670928955,
      "learning_rate": 0.0001,
      "loss": 1.5889,
      "step": 1990
    },
    {
      "epoch": 0.48348712967459934,
      "grad_norm": 0.3365665674209595,
      "learning_rate": 0.0001,
      "loss": 1.5985,
      "step": 1991
    },
    {
      "epoch": 0.483729966002914,
      "grad_norm": 0.3621251583099365,
      "learning_rate": 0.0001,
      "loss": 1.8714,
      "step": 1992
    },
    {
      "epoch": 0.48397280233122875,
      "grad_norm": 0.34295421838760376,
      "learning_rate": 0.0001,
      "loss": 1.6912,
      "step": 1993
    },
    {
      "epoch": 0.4842156386595435,
      "grad_norm": 0.3422091007232666,
      "learning_rate": 0.0001,
      "loss": 1.7885,
      "step": 1994
    },
    {
      "epoch": 0.48445847498785816,
      "grad_norm": 0.3517985939979553,
      "learning_rate": 0.0001,
      "loss": 1.7095,
      "step": 1995
    },
    {
      "epoch": 0.4847013113161729,
      "grad_norm": 0.36482927203178406,
      "learning_rate": 0.0001,
      "loss": 1.7971,
      "step": 1996
    },
    {
      "epoch": 0.4849441476444876,
      "grad_norm": 0.359244167804718,
      "learning_rate": 0.0001,
      "loss": 1.5207,
      "step": 1997
    },
    {
      "epoch": 0.48518698397280235,
      "grad_norm": 0.3884401321411133,
      "learning_rate": 0.0001,
      "loss": 1.8081,
      "step": 1998
    },
    {
      "epoch": 0.48542982030111703,
      "grad_norm": 0.3618896007537842,
      "learning_rate": 0.0001,
      "loss": 1.805,
      "step": 1999
    },
    {
      "epoch": 0.48567265662943176,
      "grad_norm": 0.3770894706249237,
      "learning_rate": 0.0001,
      "loss": 1.7803,
      "step": 2000
    },
    {
      "epoch": 0.4859154929577465,
      "grad_norm": 0.3363848030567169,
      "learning_rate": 0.0001,
      "loss": 1.5195,
      "step": 2001
    },
    {
      "epoch": 0.4861583292860612,
      "grad_norm": 0.3846726417541504,
      "learning_rate": 0.0001,
      "loss": 1.7128,
      "step": 2002
    },
    {
      "epoch": 0.4864011656143759,
      "grad_norm": 0.3797794282436371,
      "learning_rate": 0.0001,
      "loss": 1.769,
      "step": 2003
    },
    {
      "epoch": 0.48664400194269064,
      "grad_norm": 0.3585468530654907,
      "learning_rate": 0.0001,
      "loss": 1.8298,
      "step": 2004
    },
    {
      "epoch": 0.4868868382710053,
      "grad_norm": 0.34364134073257446,
      "learning_rate": 0.0001,
      "loss": 1.8856,
      "step": 2005
    },
    {
      "epoch": 0.48712967459932005,
      "grad_norm": 0.3312186896800995,
      "learning_rate": 0.0001,
      "loss": 1.5684,
      "step": 2006
    },
    {
      "epoch": 0.4873725109276348,
      "grad_norm": 0.3411174416542053,
      "learning_rate": 0.0001,
      "loss": 1.6713,
      "step": 2007
    },
    {
      "epoch": 0.4876153472559495,
      "grad_norm": 0.36895158886909485,
      "learning_rate": 0.0001,
      "loss": 1.8552,
      "step": 2008
    },
    {
      "epoch": 0.4878581835842642,
      "grad_norm": 0.3716360032558441,
      "learning_rate": 0.0001,
      "loss": 1.886,
      "step": 2009
    },
    {
      "epoch": 0.4881010199125789,
      "grad_norm": 0.3676500916481018,
      "learning_rate": 0.0001,
      "loss": 1.7994,
      "step": 2010
    },
    {
      "epoch": 0.48834385624089366,
      "grad_norm": 0.35117608308792114,
      "learning_rate": 0.0001,
      "loss": 1.6921,
      "step": 2011
    },
    {
      "epoch": 0.48858669256920834,
      "grad_norm": 0.3696017265319824,
      "learning_rate": 0.0001,
      "loss": 1.8051,
      "step": 2012
    },
    {
      "epoch": 0.48882952889752307,
      "grad_norm": 0.3984812796115875,
      "learning_rate": 0.0001,
      "loss": 1.8622,
      "step": 2013
    },
    {
      "epoch": 0.4890723652258378,
      "grad_norm": 0.3565909266471863,
      "learning_rate": 0.0001,
      "loss": 1.731,
      "step": 2014
    },
    {
      "epoch": 0.4893152015541525,
      "grad_norm": 0.35164588689804077,
      "learning_rate": 0.0001,
      "loss": 1.7458,
      "step": 2015
    },
    {
      "epoch": 0.4895580378824672,
      "grad_norm": 0.36790066957473755,
      "learning_rate": 0.0001,
      "loss": 1.9291,
      "step": 2016
    },
    {
      "epoch": 0.48980087421078194,
      "grad_norm": 0.36064937710762024,
      "learning_rate": 0.0001,
      "loss": 1.7073,
      "step": 2017
    },
    {
      "epoch": 0.4900437105390966,
      "grad_norm": 0.3749876916408539,
      "learning_rate": 0.0001,
      "loss": 1.7085,
      "step": 2018
    },
    {
      "epoch": 0.49028654686741135,
      "grad_norm": 0.3429160416126251,
      "learning_rate": 0.0001,
      "loss": 1.619,
      "step": 2019
    },
    {
      "epoch": 0.4905293831957261,
      "grad_norm": 0.3696604371070862,
      "learning_rate": 0.0001,
      "loss": 1.7821,
      "step": 2020
    },
    {
      "epoch": 0.4907722195240408,
      "grad_norm": 0.3488297760486603,
      "learning_rate": 0.0001,
      "loss": 1.7989,
      "step": 2021
    },
    {
      "epoch": 0.4910150558523555,
      "grad_norm": 0.3773786127567291,
      "learning_rate": 0.0001,
      "loss": 1.8005,
      "step": 2022
    },
    {
      "epoch": 0.49125789218067023,
      "grad_norm": 0.36527174711227417,
      "learning_rate": 0.0001,
      "loss": 1.8716,
      "step": 2023
    },
    {
      "epoch": 0.49150072850898496,
      "grad_norm": 0.3569713830947876,
      "learning_rate": 0.0001,
      "loss": 1.7227,
      "step": 2024
    },
    {
      "epoch": 0.49174356483729964,
      "grad_norm": 0.358445942401886,
      "learning_rate": 0.0001,
      "loss": 1.802,
      "step": 2025
    },
    {
      "epoch": 0.4919864011656144,
      "grad_norm": 0.34071841835975647,
      "learning_rate": 0.0001,
      "loss": 1.7493,
      "step": 2026
    },
    {
      "epoch": 0.4922292374939291,
      "grad_norm": 0.37970632314682007,
      "learning_rate": 0.0001,
      "loss": 1.7215,
      "step": 2027
    },
    {
      "epoch": 0.4924720738222438,
      "grad_norm": 0.3528176546096802,
      "learning_rate": 0.0001,
      "loss": 1.6993,
      "step": 2028
    },
    {
      "epoch": 0.4927149101505585,
      "grad_norm": 0.36992496252059937,
      "learning_rate": 0.0001,
      "loss": 1.8592,
      "step": 2029
    },
    {
      "epoch": 0.49295774647887325,
      "grad_norm": 0.36124253273010254,
      "learning_rate": 0.0001,
      "loss": 1.703,
      "step": 2030
    },
    {
      "epoch": 0.493200582807188,
      "grad_norm": 0.35745152831077576,
      "learning_rate": 0.0001,
      "loss": 1.7555,
      "step": 2031
    },
    {
      "epoch": 0.49344341913550266,
      "grad_norm": 0.36192139983177185,
      "learning_rate": 0.0001,
      "loss": 1.8643,
      "step": 2032
    },
    {
      "epoch": 0.4936862554638174,
      "grad_norm": 0.3453996777534485,
      "learning_rate": 0.0001,
      "loss": 1.4152,
      "step": 2033
    },
    {
      "epoch": 0.4939290917921321,
      "grad_norm": 0.36538147926330566,
      "learning_rate": 0.0001,
      "loss": 1.7344,
      "step": 2034
    },
    {
      "epoch": 0.4941719281204468,
      "grad_norm": 0.3792535066604614,
      "learning_rate": 0.0001,
      "loss": 1.8989,
      "step": 2035
    },
    {
      "epoch": 0.49441476444876153,
      "grad_norm": 0.3576485216617584,
      "learning_rate": 0.0001,
      "loss": 1.7443,
      "step": 2036
    },
    {
      "epoch": 0.49465760077707627,
      "grad_norm": 0.3794737756252289,
      "learning_rate": 0.0001,
      "loss": 1.8746,
      "step": 2037
    },
    {
      "epoch": 0.49490043710539094,
      "grad_norm": 0.35361653566360474,
      "learning_rate": 0.0001,
      "loss": 1.6875,
      "step": 2038
    },
    {
      "epoch": 0.4951432734337057,
      "grad_norm": 0.34948018193244934,
      "learning_rate": 0.0001,
      "loss": 1.7099,
      "step": 2039
    },
    {
      "epoch": 0.4953861097620204,
      "grad_norm": 0.3640398383140564,
      "learning_rate": 0.0001,
      "loss": 1.7508,
      "step": 2040
    },
    {
      "epoch": 0.4956289460903351,
      "grad_norm": 0.3670260012149811,
      "learning_rate": 0.0001,
      "loss": 1.7846,
      "step": 2041
    },
    {
      "epoch": 0.4958717824186498,
      "grad_norm": 0.35220327973365784,
      "learning_rate": 0.0001,
      "loss": 1.7981,
      "step": 2042
    },
    {
      "epoch": 0.49611461874696455,
      "grad_norm": 0.36040714383125305,
      "learning_rate": 0.0001,
      "loss": 1.756,
      "step": 2043
    },
    {
      "epoch": 0.4963574550752793,
      "grad_norm": 0.3941101133823395,
      "learning_rate": 0.0001,
      "loss": 1.7831,
      "step": 2044
    },
    {
      "epoch": 0.49660029140359396,
      "grad_norm": 0.37837153673171997,
      "learning_rate": 0.0001,
      "loss": 1.7308,
      "step": 2045
    },
    {
      "epoch": 0.4968431277319087,
      "grad_norm": 0.3587818145751953,
      "learning_rate": 0.0001,
      "loss": 1.6994,
      "step": 2046
    },
    {
      "epoch": 0.49708596406022343,
      "grad_norm": 0.33974432945251465,
      "learning_rate": 0.0001,
      "loss": 1.7277,
      "step": 2047
    },
    {
      "epoch": 0.4973288003885381,
      "grad_norm": 0.34887629747390747,
      "learning_rate": 0.0001,
      "loss": 1.6602,
      "step": 2048
    },
    {
      "epoch": 0.49757163671685284,
      "grad_norm": 0.348217636346817,
      "learning_rate": 0.0001,
      "loss": 1.7607,
      "step": 2049
    },
    {
      "epoch": 0.49781447304516757,
      "grad_norm": 0.38094449043273926,
      "learning_rate": 0.0001,
      "loss": 1.8629,
      "step": 2050
    },
    {
      "epoch": 0.49805730937348225,
      "grad_norm": 0.3522297739982605,
      "learning_rate": 0.0001,
      "loss": 1.8177,
      "step": 2051
    },
    {
      "epoch": 0.498300145701797,
      "grad_norm": 0.35744819045066833,
      "learning_rate": 0.0001,
      "loss": 1.637,
      "step": 2052
    },
    {
      "epoch": 0.4985429820301117,
      "grad_norm": 0.3388936221599579,
      "learning_rate": 0.0001,
      "loss": 1.5783,
      "step": 2053
    },
    {
      "epoch": 0.49878581835842645,
      "grad_norm": 0.35953521728515625,
      "learning_rate": 0.0001,
      "loss": 1.6525,
      "step": 2054
    },
    {
      "epoch": 0.4990286546867411,
      "grad_norm": 0.3640817105770111,
      "learning_rate": 0.0001,
      "loss": 1.676,
      "step": 2055
    },
    {
      "epoch": 0.49927149101505586,
      "grad_norm": 0.3549385666847229,
      "learning_rate": 0.0001,
      "loss": 1.6348,
      "step": 2056
    },
    {
      "epoch": 0.4995143273433706,
      "grad_norm": 0.36192211508750916,
      "learning_rate": 0.0001,
      "loss": 1.7436,
      "step": 2057
    },
    {
      "epoch": 0.49975716367168527,
      "grad_norm": 0.33218875527381897,
      "learning_rate": 0.0001,
      "loss": 1.7101,
      "step": 2058
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.3624739944934845,
      "learning_rate": 0.0001,
      "loss": 1.7806,
      "step": 2059
    },
    {
      "epoch": 0.5002428363283147,
      "grad_norm": 0.33389827609062195,
      "learning_rate": 0.0001,
      "loss": 1.6401,
      "step": 2060
    },
    {
      "epoch": 0.5004856726566295,
      "grad_norm": 0.3476710915565491,
      "learning_rate": 0.0001,
      "loss": 1.64,
      "step": 2061
    },
    {
      "epoch": 0.5007285089849441,
      "grad_norm": 0.37466832995414734,
      "learning_rate": 0.0001,
      "loss": 1.7737,
      "step": 2062
    },
    {
      "epoch": 0.5009713453132588,
      "grad_norm": 0.34958529472351074,
      "learning_rate": 0.0001,
      "loss": 1.7622,
      "step": 2063
    },
    {
      "epoch": 0.5012141816415736,
      "grad_norm": 0.3609834611415863,
      "learning_rate": 0.0001,
      "loss": 1.6627,
      "step": 2064
    },
    {
      "epoch": 0.5014570179698883,
      "grad_norm": 0.3426356315612793,
      "learning_rate": 0.0001,
      "loss": 1.7813,
      "step": 2065
    },
    {
      "epoch": 0.501699854298203,
      "grad_norm": 0.3439556360244751,
      "learning_rate": 0.0001,
      "loss": 1.6216,
      "step": 2066
    },
    {
      "epoch": 0.5019426906265178,
      "grad_norm": 0.34030354022979736,
      "learning_rate": 0.0001,
      "loss": 1.7757,
      "step": 2067
    },
    {
      "epoch": 0.5021855269548324,
      "grad_norm": 0.3406347632408142,
      "learning_rate": 0.0001,
      "loss": 1.776,
      "step": 2068
    },
    {
      "epoch": 0.5024283632831471,
      "grad_norm": 0.3400167226791382,
      "learning_rate": 0.0001,
      "loss": 1.6481,
      "step": 2069
    },
    {
      "epoch": 0.5026711996114619,
      "grad_norm": 0.3473275899887085,
      "learning_rate": 0.0001,
      "loss": 1.7927,
      "step": 2070
    },
    {
      "epoch": 0.5029140359397766,
      "grad_norm": 0.40642112493515015,
      "learning_rate": 0.0001,
      "loss": 1.7355,
      "step": 2071
    },
    {
      "epoch": 0.5031568722680914,
      "grad_norm": 0.3894907832145691,
      "learning_rate": 0.0001,
      "loss": 1.7494,
      "step": 2072
    },
    {
      "epoch": 0.503399708596406,
      "grad_norm": 0.34641486406326294,
      "learning_rate": 0.0001,
      "loss": 1.6575,
      "step": 2073
    },
    {
      "epoch": 0.5036425449247207,
      "grad_norm": 0.34368327260017395,
      "learning_rate": 0.0001,
      "loss": 1.706,
      "step": 2074
    },
    {
      "epoch": 0.5038853812530355,
      "grad_norm": 0.3774399161338806,
      "learning_rate": 0.0001,
      "loss": 1.809,
      "step": 2075
    },
    {
      "epoch": 0.5041282175813502,
      "grad_norm": 0.3743821084499359,
      "learning_rate": 0.0001,
      "loss": 1.6534,
      "step": 2076
    },
    {
      "epoch": 0.5043710539096649,
      "grad_norm": 0.3586447834968567,
      "learning_rate": 0.0001,
      "loss": 1.7862,
      "step": 2077
    },
    {
      "epoch": 0.5046138902379796,
      "grad_norm": 0.3688397705554962,
      "learning_rate": 0.0001,
      "loss": 1.6918,
      "step": 2078
    },
    {
      "epoch": 0.5048567265662943,
      "grad_norm": 0.3526496887207031,
      "learning_rate": 0.0001,
      "loss": 1.7641,
      "step": 2079
    },
    {
      "epoch": 0.505099562894609,
      "grad_norm": 0.35157665610313416,
      "learning_rate": 0.0001,
      "loss": 1.6627,
      "step": 2080
    },
    {
      "epoch": 0.5053423992229238,
      "grad_norm": 0.3329943120479584,
      "learning_rate": 0.0001,
      "loss": 1.5457,
      "step": 2081
    },
    {
      "epoch": 0.5055852355512385,
      "grad_norm": 0.34889695048332214,
      "learning_rate": 0.0001,
      "loss": 1.683,
      "step": 2082
    },
    {
      "epoch": 0.5058280718795531,
      "grad_norm": 0.34723854064941406,
      "learning_rate": 0.0001,
      "loss": 1.7536,
      "step": 2083
    },
    {
      "epoch": 0.5060709082078679,
      "grad_norm": 0.36458468437194824,
      "learning_rate": 0.0001,
      "loss": 1.7602,
      "step": 2084
    },
    {
      "epoch": 0.5063137445361826,
      "grad_norm": 0.369058221578598,
      "learning_rate": 0.0001,
      "loss": 1.7215,
      "step": 2085
    },
    {
      "epoch": 0.5065565808644973,
      "grad_norm": 0.35088416934013367,
      "learning_rate": 0.0001,
      "loss": 1.7839,
      "step": 2086
    },
    {
      "epoch": 0.5067994171928121,
      "grad_norm": 0.389716774225235,
      "learning_rate": 0.0001,
      "loss": 1.7872,
      "step": 2087
    },
    {
      "epoch": 0.5070422535211268,
      "grad_norm": 0.36617782711982727,
      "learning_rate": 0.0001,
      "loss": 1.7071,
      "step": 2088
    },
    {
      "epoch": 0.5072850898494414,
      "grad_norm": 0.36376142501831055,
      "learning_rate": 0.0001,
      "loss": 1.6022,
      "step": 2089
    },
    {
      "epoch": 0.5075279261777562,
      "grad_norm": 0.36511048674583435,
      "learning_rate": 0.0001,
      "loss": 1.7629,
      "step": 2090
    },
    {
      "epoch": 0.5077707625060709,
      "grad_norm": 0.3856488764286041,
      "learning_rate": 0.0001,
      "loss": 1.7168,
      "step": 2091
    },
    {
      "epoch": 0.5080135988343856,
      "grad_norm": 0.37251293659210205,
      "learning_rate": 0.0001,
      "loss": 1.6808,
      "step": 2092
    },
    {
      "epoch": 0.5082564351627004,
      "grad_norm": 0.35156142711639404,
      "learning_rate": 0.0001,
      "loss": 1.7066,
      "step": 2093
    },
    {
      "epoch": 0.508499271491015,
      "grad_norm": 0.32686617970466614,
      "learning_rate": 0.0001,
      "loss": 1.6424,
      "step": 2094
    },
    {
      "epoch": 0.5087421078193298,
      "grad_norm": 0.34049227833747864,
      "learning_rate": 0.0001,
      "loss": 1.5752,
      "step": 2095
    },
    {
      "epoch": 0.5089849441476445,
      "grad_norm": 0.3699936866760254,
      "learning_rate": 0.0001,
      "loss": 1.7793,
      "step": 2096
    },
    {
      "epoch": 0.5092277804759592,
      "grad_norm": 0.3387681543827057,
      "learning_rate": 0.0001,
      "loss": 1.762,
      "step": 2097
    },
    {
      "epoch": 0.509470616804274,
      "grad_norm": 0.34820428490638733,
      "learning_rate": 0.0001,
      "loss": 1.6655,
      "step": 2098
    },
    {
      "epoch": 0.5097134531325886,
      "grad_norm": 0.3199908435344696,
      "learning_rate": 0.0001,
      "loss": 1.5278,
      "step": 2099
    },
    {
      "epoch": 0.5099562894609033,
      "grad_norm": 0.33102068305015564,
      "learning_rate": 0.0001,
      "loss": 1.4953,
      "step": 2100
    },
    {
      "epoch": 0.5101991257892181,
      "grad_norm": 0.351879358291626,
      "learning_rate": 0.0001,
      "loss": 1.6847,
      "step": 2101
    },
    {
      "epoch": 0.5104419621175328,
      "grad_norm": 0.3446686863899231,
      "learning_rate": 0.0001,
      "loss": 1.6902,
      "step": 2102
    },
    {
      "epoch": 0.5106847984458475,
      "grad_norm": 0.3876822888851166,
      "learning_rate": 0.0001,
      "loss": 1.8998,
      "step": 2103
    },
    {
      "epoch": 0.5109276347741623,
      "grad_norm": 0.3358386158943176,
      "learning_rate": 0.0001,
      "loss": 1.5755,
      "step": 2104
    },
    {
      "epoch": 0.5111704711024769,
      "grad_norm": 0.3508923053741455,
      "learning_rate": 0.0001,
      "loss": 1.7499,
      "step": 2105
    },
    {
      "epoch": 0.5114133074307916,
      "grad_norm": 0.3613468110561371,
      "learning_rate": 0.0001,
      "loss": 1.8817,
      "step": 2106
    },
    {
      "epoch": 0.5116561437591064,
      "grad_norm": 0.3340017795562744,
      "learning_rate": 0.0001,
      "loss": 1.5897,
      "step": 2107
    },
    {
      "epoch": 0.5118989800874211,
      "grad_norm": 0.3561114966869354,
      "learning_rate": 0.0001,
      "loss": 1.6141,
      "step": 2108
    },
    {
      "epoch": 0.5121418164157358,
      "grad_norm": 0.3637521266937256,
      "learning_rate": 0.0001,
      "loss": 1.6148,
      "step": 2109
    },
    {
      "epoch": 0.5123846527440505,
      "grad_norm": 0.3636873662471771,
      "learning_rate": 0.0001,
      "loss": 1.8358,
      "step": 2110
    },
    {
      "epoch": 0.5126274890723652,
      "grad_norm": 0.3566911220550537,
      "learning_rate": 0.0001,
      "loss": 1.7366,
      "step": 2111
    },
    {
      "epoch": 0.5128703254006799,
      "grad_norm": 0.33937013149261475,
      "learning_rate": 0.0001,
      "loss": 1.7309,
      "step": 2112
    },
    {
      "epoch": 0.5131131617289947,
      "grad_norm": 0.36434876918792725,
      "learning_rate": 0.0001,
      "loss": 1.7042,
      "step": 2113
    },
    {
      "epoch": 0.5133559980573094,
      "grad_norm": 0.35941821336746216,
      "learning_rate": 0.0001,
      "loss": 1.8072,
      "step": 2114
    },
    {
      "epoch": 0.513598834385624,
      "grad_norm": 0.36316072940826416,
      "learning_rate": 0.0001,
      "loss": 1.5349,
      "step": 2115
    },
    {
      "epoch": 0.5138416707139388,
      "grad_norm": 0.37612438201904297,
      "learning_rate": 0.0001,
      "loss": 1.8388,
      "step": 2116
    },
    {
      "epoch": 0.5140845070422535,
      "grad_norm": 0.3560262620449066,
      "learning_rate": 0.0001,
      "loss": 1.663,
      "step": 2117
    },
    {
      "epoch": 0.5143273433705683,
      "grad_norm": 0.37767353653907776,
      "learning_rate": 0.0001,
      "loss": 1.9991,
      "step": 2118
    },
    {
      "epoch": 0.514570179698883,
      "grad_norm": 0.33369192481040955,
      "learning_rate": 0.0001,
      "loss": 1.6447,
      "step": 2119
    },
    {
      "epoch": 0.5148130160271976,
      "grad_norm": 0.38418659567832947,
      "learning_rate": 0.0001,
      "loss": 1.8411,
      "step": 2120
    },
    {
      "epoch": 0.5150558523555124,
      "grad_norm": 0.3417730927467346,
      "learning_rate": 0.0001,
      "loss": 1.701,
      "step": 2121
    },
    {
      "epoch": 0.5152986886838271,
      "grad_norm": 0.3553977608680725,
      "learning_rate": 0.0001,
      "loss": 1.6914,
      "step": 2122
    },
    {
      "epoch": 0.5155415250121418,
      "grad_norm": 0.3727787435054779,
      "learning_rate": 0.0001,
      "loss": 1.8149,
      "step": 2123
    },
    {
      "epoch": 0.5157843613404566,
      "grad_norm": 0.38117727637290955,
      "learning_rate": 0.0001,
      "loss": 1.7707,
      "step": 2124
    },
    {
      "epoch": 0.5160271976687713,
      "grad_norm": 0.378630131483078,
      "learning_rate": 0.0001,
      "loss": 1.7856,
      "step": 2125
    },
    {
      "epoch": 0.5162700339970859,
      "grad_norm": 0.36176884174346924,
      "learning_rate": 0.0001,
      "loss": 1.8372,
      "step": 2126
    },
    {
      "epoch": 0.5165128703254007,
      "grad_norm": 0.3552527129650116,
      "learning_rate": 0.0001,
      "loss": 1.6077,
      "step": 2127
    },
    {
      "epoch": 0.5167557066537154,
      "grad_norm": 0.3917563855648041,
      "learning_rate": 0.0001,
      "loss": 1.825,
      "step": 2128
    },
    {
      "epoch": 0.5169985429820301,
      "grad_norm": 0.34401583671569824,
      "learning_rate": 0.0001,
      "loss": 1.7408,
      "step": 2129
    },
    {
      "epoch": 0.5172413793103449,
      "grad_norm": 0.36782148480415344,
      "learning_rate": 0.0001,
      "loss": 1.7808,
      "step": 2130
    },
    {
      "epoch": 0.5174842156386595,
      "grad_norm": 0.342189759016037,
      "learning_rate": 0.0001,
      "loss": 1.6103,
      "step": 2131
    },
    {
      "epoch": 0.5177270519669742,
      "grad_norm": 0.353523313999176,
      "learning_rate": 0.0001,
      "loss": 1.7307,
      "step": 2132
    },
    {
      "epoch": 0.517969888295289,
      "grad_norm": 0.3676711320877075,
      "learning_rate": 0.0001,
      "loss": 1.8342,
      "step": 2133
    },
    {
      "epoch": 0.5182127246236037,
      "grad_norm": 0.32856231927871704,
      "learning_rate": 0.0001,
      "loss": 1.6015,
      "step": 2134
    },
    {
      "epoch": 0.5184555609519184,
      "grad_norm": 0.35762903094291687,
      "learning_rate": 0.0001,
      "loss": 1.7695,
      "step": 2135
    },
    {
      "epoch": 0.5186983972802331,
      "grad_norm": 0.3561696708202362,
      "learning_rate": 0.0001,
      "loss": 1.8157,
      "step": 2136
    },
    {
      "epoch": 0.5189412336085478,
      "grad_norm": 0.343925416469574,
      "learning_rate": 0.0001,
      "loss": 1.7213,
      "step": 2137
    },
    {
      "epoch": 0.5191840699368625,
      "grad_norm": 0.356314480304718,
      "learning_rate": 0.0001,
      "loss": 1.7979,
      "step": 2138
    },
    {
      "epoch": 0.5194269062651773,
      "grad_norm": 0.37433797121047974,
      "learning_rate": 0.0001,
      "loss": 1.8121,
      "step": 2139
    },
    {
      "epoch": 0.519669742593492,
      "grad_norm": 0.34445253014564514,
      "learning_rate": 0.0001,
      "loss": 1.7456,
      "step": 2140
    },
    {
      "epoch": 0.5199125789218068,
      "grad_norm": 0.36527112126350403,
      "learning_rate": 0.0001,
      "loss": 1.7099,
      "step": 2141
    },
    {
      "epoch": 0.5201554152501214,
      "grad_norm": 0.3757280707359314,
      "learning_rate": 0.0001,
      "loss": 1.7387,
      "step": 2142
    },
    {
      "epoch": 0.5203982515784361,
      "grad_norm": 0.3530278503894806,
      "learning_rate": 0.0001,
      "loss": 1.7446,
      "step": 2143
    },
    {
      "epoch": 0.5206410879067509,
      "grad_norm": 0.36654311418533325,
      "learning_rate": 0.0001,
      "loss": 1.6952,
      "step": 2144
    },
    {
      "epoch": 0.5208839242350656,
      "grad_norm": 0.3734758198261261,
      "learning_rate": 0.0001,
      "loss": 1.6797,
      "step": 2145
    },
    {
      "epoch": 0.5211267605633803,
      "grad_norm": 0.3684309422969818,
      "learning_rate": 0.0001,
      "loss": 1.746,
      "step": 2146
    },
    {
      "epoch": 0.521369596891695,
      "grad_norm": 0.3862002491950989,
      "learning_rate": 0.0001,
      "loss": 1.8234,
      "step": 2147
    },
    {
      "epoch": 0.5216124332200097,
      "grad_norm": 0.3541664183139801,
      "learning_rate": 0.0001,
      "loss": 1.6616,
      "step": 2148
    },
    {
      "epoch": 0.5218552695483244,
      "grad_norm": 0.35986196994781494,
      "learning_rate": 0.0001,
      "loss": 1.711,
      "step": 2149
    },
    {
      "epoch": 0.5220981058766392,
      "grad_norm": 0.3609998822212219,
      "learning_rate": 0.0001,
      "loss": 1.8106,
      "step": 2150
    },
    {
      "epoch": 0.5223409422049539,
      "grad_norm": 0.3694157004356384,
      "learning_rate": 0.0001,
      "loss": 1.7587,
      "step": 2151
    },
    {
      "epoch": 0.5225837785332685,
      "grad_norm": 0.3328161835670471,
      "learning_rate": 0.0001,
      "loss": 1.5532,
      "step": 2152
    },
    {
      "epoch": 0.5228266148615833,
      "grad_norm": 0.32815298438072205,
      "learning_rate": 0.0001,
      "loss": 1.5474,
      "step": 2153
    },
    {
      "epoch": 0.523069451189898,
      "grad_norm": 0.37450891733169556,
      "learning_rate": 0.0001,
      "loss": 1.6402,
      "step": 2154
    },
    {
      "epoch": 0.5233122875182127,
      "grad_norm": 0.3439551591873169,
      "learning_rate": 0.0001,
      "loss": 1.7877,
      "step": 2155
    },
    {
      "epoch": 0.5235551238465275,
      "grad_norm": 0.3464338183403015,
      "learning_rate": 0.0001,
      "loss": 1.6436,
      "step": 2156
    },
    {
      "epoch": 0.5237979601748421,
      "grad_norm": 0.3607020378112793,
      "learning_rate": 0.0001,
      "loss": 1.7033,
      "step": 2157
    },
    {
      "epoch": 0.5240407965031568,
      "grad_norm": 0.36788424849510193,
      "learning_rate": 0.0001,
      "loss": 1.8079,
      "step": 2158
    },
    {
      "epoch": 0.5242836328314716,
      "grad_norm": 0.36867842078208923,
      "learning_rate": 0.0001,
      "loss": 1.7583,
      "step": 2159
    },
    {
      "epoch": 0.5245264691597863,
      "grad_norm": 0.34436655044555664,
      "learning_rate": 0.0001,
      "loss": 1.6348,
      "step": 2160
    },
    {
      "epoch": 0.524769305488101,
      "grad_norm": 0.33278247714042664,
      "learning_rate": 0.0001,
      "loss": 1.6667,
      "step": 2161
    },
    {
      "epoch": 0.5250121418164158,
      "grad_norm": 0.34308114647865295,
      "learning_rate": 0.0001,
      "loss": 1.6579,
      "step": 2162
    },
    {
      "epoch": 0.5252549781447304,
      "grad_norm": 0.33665505051612854,
      "learning_rate": 0.0001,
      "loss": 1.5245,
      "step": 2163
    },
    {
      "epoch": 0.5254978144730452,
      "grad_norm": 0.34041884541511536,
      "learning_rate": 0.0001,
      "loss": 1.6513,
      "step": 2164
    },
    {
      "epoch": 0.5257406508013599,
      "grad_norm": 0.6450101137161255,
      "learning_rate": 0.0001,
      "loss": 1.8578,
      "step": 2165
    },
    {
      "epoch": 0.5259834871296746,
      "grad_norm": 0.3492946922779083,
      "learning_rate": 0.0001,
      "loss": 1.6815,
      "step": 2166
    },
    {
      "epoch": 0.5262263234579894,
      "grad_norm": 0.39884358644485474,
      "learning_rate": 0.0001,
      "loss": 1.8944,
      "step": 2167
    },
    {
      "epoch": 0.526469159786304,
      "grad_norm": 0.3553820848464966,
      "learning_rate": 0.0001,
      "loss": 1.7083,
      "step": 2168
    },
    {
      "epoch": 0.5267119961146187,
      "grad_norm": 0.35157763957977295,
      "learning_rate": 0.0001,
      "loss": 1.7117,
      "step": 2169
    },
    {
      "epoch": 0.5269548324429335,
      "grad_norm": 0.33850955963134766,
      "learning_rate": 0.0001,
      "loss": 1.7726,
      "step": 2170
    },
    {
      "epoch": 0.5271976687712482,
      "grad_norm": 0.3849356472492218,
      "learning_rate": 0.0001,
      "loss": 1.8159,
      "step": 2171
    },
    {
      "epoch": 0.5274405050995629,
      "grad_norm": 0.3549228012561798,
      "learning_rate": 0.0001,
      "loss": 1.7322,
      "step": 2172
    },
    {
      "epoch": 0.5276833414278777,
      "grad_norm": 0.36918267607688904,
      "learning_rate": 0.0001,
      "loss": 1.8431,
      "step": 2173
    },
    {
      "epoch": 0.5279261777561923,
      "grad_norm": 0.3418583273887634,
      "learning_rate": 0.0001,
      "loss": 1.6985,
      "step": 2174
    },
    {
      "epoch": 0.528169014084507,
      "grad_norm": 0.34361299872398376,
      "learning_rate": 0.0001,
      "loss": 1.632,
      "step": 2175
    },
    {
      "epoch": 0.5284118504128218,
      "grad_norm": 0.35774165391921997,
      "learning_rate": 0.0001,
      "loss": 1.7134,
      "step": 2176
    },
    {
      "epoch": 0.5286546867411365,
      "grad_norm": 0.35581788420677185,
      "learning_rate": 0.0001,
      "loss": 1.7574,
      "step": 2177
    },
    {
      "epoch": 0.5288975230694511,
      "grad_norm": 0.3620051443576813,
      "learning_rate": 0.0001,
      "loss": 1.6363,
      "step": 2178
    },
    {
      "epoch": 0.5291403593977659,
      "grad_norm": 0.3930107355117798,
      "learning_rate": 0.0001,
      "loss": 1.9798,
      "step": 2179
    },
    {
      "epoch": 0.5293831957260806,
      "grad_norm": 0.37771743535995483,
      "learning_rate": 0.0001,
      "loss": 1.93,
      "step": 2180
    },
    {
      "epoch": 0.5296260320543953,
      "grad_norm": 0.3509754240512848,
      "learning_rate": 0.0001,
      "loss": 1.6358,
      "step": 2181
    },
    {
      "epoch": 0.5298688683827101,
      "grad_norm": 0.3814009428024292,
      "learning_rate": 0.0001,
      "loss": 1.8188,
      "step": 2182
    },
    {
      "epoch": 0.5301117047110248,
      "grad_norm": 0.34286820888519287,
      "learning_rate": 0.0001,
      "loss": 1.7615,
      "step": 2183
    },
    {
      "epoch": 0.5303545410393394,
      "grad_norm": 0.3434714674949646,
      "learning_rate": 0.0001,
      "loss": 1.8686,
      "step": 2184
    },
    {
      "epoch": 0.5305973773676542,
      "grad_norm": 0.35580405592918396,
      "learning_rate": 0.0001,
      "loss": 1.7198,
      "step": 2185
    },
    {
      "epoch": 0.5308402136959689,
      "grad_norm": 0.34779301285743713,
      "learning_rate": 0.0001,
      "loss": 1.5524,
      "step": 2186
    },
    {
      "epoch": 0.5310830500242837,
      "grad_norm": 0.37721124291419983,
      "learning_rate": 0.0001,
      "loss": 1.77,
      "step": 2187
    },
    {
      "epoch": 0.5313258863525984,
      "grad_norm": 0.32382553815841675,
      "learning_rate": 0.0001,
      "loss": 1.5097,
      "step": 2188
    },
    {
      "epoch": 0.531568722680913,
      "grad_norm": 0.3677857518196106,
      "learning_rate": 0.0001,
      "loss": 1.8567,
      "step": 2189
    },
    {
      "epoch": 0.5318115590092278,
      "grad_norm": 0.3528325855731964,
      "learning_rate": 0.0001,
      "loss": 1.781,
      "step": 2190
    },
    {
      "epoch": 0.5320543953375425,
      "grad_norm": 0.35266974568367004,
      "learning_rate": 0.0001,
      "loss": 1.6727,
      "step": 2191
    },
    {
      "epoch": 0.5322972316658572,
      "grad_norm": 0.38058826327323914,
      "learning_rate": 0.0001,
      "loss": 1.9637,
      "step": 2192
    },
    {
      "epoch": 0.532540067994172,
      "grad_norm": 0.37474575638771057,
      "learning_rate": 0.0001,
      "loss": 1.8031,
      "step": 2193
    },
    {
      "epoch": 0.5327829043224867,
      "grad_norm": 0.3883814215660095,
      "learning_rate": 0.0001,
      "loss": 1.8517,
      "step": 2194
    },
    {
      "epoch": 0.5330257406508013,
      "grad_norm": 0.37679925560951233,
      "learning_rate": 0.0001,
      "loss": 1.8224,
      "step": 2195
    },
    {
      "epoch": 0.5332685769791161,
      "grad_norm": 0.3672942817211151,
      "learning_rate": 0.0001,
      "loss": 1.9133,
      "step": 2196
    },
    {
      "epoch": 0.5335114133074308,
      "grad_norm": 0.35203585028648376,
      "learning_rate": 0.0001,
      "loss": 1.782,
      "step": 2197
    },
    {
      "epoch": 0.5337542496357455,
      "grad_norm": 0.3779953420162201,
      "learning_rate": 0.0001,
      "loss": 1.7499,
      "step": 2198
    },
    {
      "epoch": 0.5339970859640603,
      "grad_norm": 0.34961748123168945,
      "learning_rate": 0.0001,
      "loss": 1.8082,
      "step": 2199
    },
    {
      "epoch": 0.5342399222923749,
      "grad_norm": 0.36361628770828247,
      "learning_rate": 0.0001,
      "loss": 1.9355,
      "step": 2200
    },
    {
      "epoch": 0.5344827586206896,
      "grad_norm": 0.3708609342575073,
      "learning_rate": 0.0001,
      "loss": 1.7396,
      "step": 2201
    },
    {
      "epoch": 0.5347255949490044,
      "grad_norm": 0.3699905276298523,
      "learning_rate": 0.0001,
      "loss": 1.8169,
      "step": 2202
    },
    {
      "epoch": 0.5349684312773191,
      "grad_norm": 0.34191349148750305,
      "learning_rate": 0.0001,
      "loss": 1.6469,
      "step": 2203
    },
    {
      "epoch": 0.5352112676056338,
      "grad_norm": 0.34473779797554016,
      "learning_rate": 0.0001,
      "loss": 1.6362,
      "step": 2204
    },
    {
      "epoch": 0.5354541039339485,
      "grad_norm": 0.34325653314590454,
      "learning_rate": 0.0001,
      "loss": 1.623,
      "step": 2205
    },
    {
      "epoch": 0.5356969402622632,
      "grad_norm": 0.386184424161911,
      "learning_rate": 0.0001,
      "loss": 1.8688,
      "step": 2206
    },
    {
      "epoch": 0.5359397765905779,
      "grad_norm": 0.33374670147895813,
      "learning_rate": 0.0001,
      "loss": 1.4897,
      "step": 2207
    },
    {
      "epoch": 0.5361826129188927,
      "grad_norm": 0.3500770330429077,
      "learning_rate": 0.0001,
      "loss": 1.6829,
      "step": 2208
    },
    {
      "epoch": 0.5364254492472074,
      "grad_norm": 0.35503390431404114,
      "learning_rate": 0.0001,
      "loss": 1.7536,
      "step": 2209
    },
    {
      "epoch": 0.5366682855755222,
      "grad_norm": 0.3908369243144989,
      "learning_rate": 0.0001,
      "loss": 1.9048,
      "step": 2210
    },
    {
      "epoch": 0.5369111219038368,
      "grad_norm": 0.400844544172287,
      "learning_rate": 0.0001,
      "loss": 1.8024,
      "step": 2211
    },
    {
      "epoch": 0.5371539582321515,
      "grad_norm": 0.3470640480518341,
      "learning_rate": 0.0001,
      "loss": 1.7288,
      "step": 2212
    },
    {
      "epoch": 0.5373967945604663,
      "grad_norm": 0.34582623839378357,
      "learning_rate": 0.0001,
      "loss": 1.668,
      "step": 2213
    },
    {
      "epoch": 0.537639630888781,
      "grad_norm": 0.3589247763156891,
      "learning_rate": 0.0001,
      "loss": 1.8759,
      "step": 2214
    },
    {
      "epoch": 0.5378824672170957,
      "grad_norm": 0.313040167093277,
      "learning_rate": 0.0001,
      "loss": 1.5255,
      "step": 2215
    },
    {
      "epoch": 0.5381253035454104,
      "grad_norm": 0.36275044083595276,
      "learning_rate": 0.0001,
      "loss": 1.765,
      "step": 2216
    },
    {
      "epoch": 0.5383681398737251,
      "grad_norm": 0.35852962732315063,
      "learning_rate": 0.0001,
      "loss": 1.7484,
      "step": 2217
    },
    {
      "epoch": 0.5386109762020398,
      "grad_norm": 0.3503633737564087,
      "learning_rate": 0.0001,
      "loss": 1.6479,
      "step": 2218
    },
    {
      "epoch": 0.5388538125303546,
      "grad_norm": 0.34514978528022766,
      "learning_rate": 0.0001,
      "loss": 1.6507,
      "step": 2219
    },
    {
      "epoch": 0.5390966488586693,
      "grad_norm": 0.3434945344924927,
      "learning_rate": 0.0001,
      "loss": 1.5982,
      "step": 2220
    },
    {
      "epoch": 0.5393394851869839,
      "grad_norm": 0.3565481901168823,
      "learning_rate": 0.0001,
      "loss": 1.6312,
      "step": 2221
    },
    {
      "epoch": 0.5395823215152987,
      "grad_norm": 0.34751904010772705,
      "learning_rate": 0.0001,
      "loss": 1.5949,
      "step": 2222
    },
    {
      "epoch": 0.5398251578436134,
      "grad_norm": 0.36038702726364136,
      "learning_rate": 0.0001,
      "loss": 1.8388,
      "step": 2223
    },
    {
      "epoch": 0.5400679941719281,
      "grad_norm": 0.3723470866680145,
      "learning_rate": 0.0001,
      "loss": 1.7968,
      "step": 2224
    },
    {
      "epoch": 0.5403108305002429,
      "grad_norm": 0.35585710406303406,
      "learning_rate": 0.0001,
      "loss": 1.6106,
      "step": 2225
    },
    {
      "epoch": 0.5405536668285575,
      "grad_norm": 0.48469194769859314,
      "learning_rate": 0.0001,
      "loss": 1.5693,
      "step": 2226
    },
    {
      "epoch": 0.5407965031568722,
      "grad_norm": 0.37182337045669556,
      "learning_rate": 0.0001,
      "loss": 1.8367,
      "step": 2227
    },
    {
      "epoch": 0.541039339485187,
      "grad_norm": 0.3512868285179138,
      "learning_rate": 0.0001,
      "loss": 1.7117,
      "step": 2228
    },
    {
      "epoch": 0.5412821758135017,
      "grad_norm": 0.37100744247436523,
      "learning_rate": 0.0001,
      "loss": 1.6873,
      "step": 2229
    },
    {
      "epoch": 0.5415250121418164,
      "grad_norm": 0.3717704117298126,
      "learning_rate": 0.0001,
      "loss": 1.815,
      "step": 2230
    },
    {
      "epoch": 0.5417678484701312,
      "grad_norm": 0.3247489333152771,
      "learning_rate": 0.0001,
      "loss": 1.5791,
      "step": 2231
    },
    {
      "epoch": 0.5420106847984458,
      "grad_norm": 0.3410359025001526,
      "learning_rate": 0.0001,
      "loss": 1.7587,
      "step": 2232
    },
    {
      "epoch": 0.5422535211267606,
      "grad_norm": 0.3632831275463104,
      "learning_rate": 0.0001,
      "loss": 1.7161,
      "step": 2233
    },
    {
      "epoch": 0.5424963574550753,
      "grad_norm": 0.34458988904953003,
      "learning_rate": 0.0001,
      "loss": 1.8121,
      "step": 2234
    },
    {
      "epoch": 0.54273919378339,
      "grad_norm": 0.3627568185329437,
      "learning_rate": 0.0001,
      "loss": 1.729,
      "step": 2235
    },
    {
      "epoch": 0.5429820301117048,
      "grad_norm": 0.3488233983516693,
      "learning_rate": 0.0001,
      "loss": 1.8708,
      "step": 2236
    },
    {
      "epoch": 0.5432248664400194,
      "grad_norm": 0.37403398752212524,
      "learning_rate": 0.0001,
      "loss": 1.8789,
      "step": 2237
    },
    {
      "epoch": 0.5434677027683341,
      "grad_norm": 0.34006038308143616,
      "learning_rate": 0.0001,
      "loss": 1.6338,
      "step": 2238
    },
    {
      "epoch": 0.5437105390966489,
      "grad_norm": 0.3680495321750641,
      "learning_rate": 0.0001,
      "loss": 1.8293,
      "step": 2239
    },
    {
      "epoch": 0.5439533754249636,
      "grad_norm": 0.33197692036628723,
      "learning_rate": 0.0001,
      "loss": 1.6022,
      "step": 2240
    },
    {
      "epoch": 0.5441962117532783,
      "grad_norm": 0.3618168532848358,
      "learning_rate": 0.0001,
      "loss": 1.7324,
      "step": 2241
    },
    {
      "epoch": 0.544439048081593,
      "grad_norm": 0.357052206993103,
      "learning_rate": 0.0001,
      "loss": 1.7062,
      "step": 2242
    },
    {
      "epoch": 0.5446818844099077,
      "grad_norm": 0.346254825592041,
      "learning_rate": 0.0001,
      "loss": 1.6179,
      "step": 2243
    },
    {
      "epoch": 0.5449247207382224,
      "grad_norm": 0.37603044509887695,
      "learning_rate": 0.0001,
      "loss": 1.8327,
      "step": 2244
    },
    {
      "epoch": 0.5451675570665372,
      "grad_norm": 0.3336375951766968,
      "learning_rate": 0.0001,
      "loss": 1.6047,
      "step": 2245
    },
    {
      "epoch": 0.5454103933948519,
      "grad_norm": 0.354708194732666,
      "learning_rate": 0.0001,
      "loss": 1.7201,
      "step": 2246
    },
    {
      "epoch": 0.5456532297231665,
      "grad_norm": 0.35244420170783997,
      "learning_rate": 0.0001,
      "loss": 1.6719,
      "step": 2247
    },
    {
      "epoch": 0.5458960660514813,
      "grad_norm": 0.3440207839012146,
      "learning_rate": 0.0001,
      "loss": 1.5432,
      "step": 2248
    },
    {
      "epoch": 0.546138902379796,
      "grad_norm": 0.36056414246559143,
      "learning_rate": 0.0001,
      "loss": 1.8697,
      "step": 2249
    },
    {
      "epoch": 0.5463817387081107,
      "grad_norm": 0.32609304785728455,
      "learning_rate": 0.0001,
      "loss": 1.5132,
      "step": 2250
    },
    {
      "epoch": 0.5466245750364255,
      "grad_norm": 0.3519289493560791,
      "learning_rate": 0.0001,
      "loss": 1.4774,
      "step": 2251
    },
    {
      "epoch": 0.5468674113647402,
      "grad_norm": 0.35858359932899475,
      "learning_rate": 0.0001,
      "loss": 1.6757,
      "step": 2252
    },
    {
      "epoch": 0.5471102476930548,
      "grad_norm": 0.3339728116989136,
      "learning_rate": 0.0001,
      "loss": 1.5052,
      "step": 2253
    },
    {
      "epoch": 0.5473530840213696,
      "grad_norm": 0.36226099729537964,
      "learning_rate": 0.0001,
      "loss": 1.7865,
      "step": 2254
    },
    {
      "epoch": 0.5475959203496843,
      "grad_norm": 0.3570326864719391,
      "learning_rate": 0.0001,
      "loss": 1.7404,
      "step": 2255
    },
    {
      "epoch": 0.5478387566779991,
      "grad_norm": 0.376078724861145,
      "learning_rate": 0.0001,
      "loss": 1.8135,
      "step": 2256
    },
    {
      "epoch": 0.5480815930063138,
      "grad_norm": 0.3843255937099457,
      "learning_rate": 0.0001,
      "loss": 1.6103,
      "step": 2257
    },
    {
      "epoch": 0.5483244293346284,
      "grad_norm": 0.3383480906486511,
      "learning_rate": 0.0001,
      "loss": 1.6364,
      "step": 2258
    },
    {
      "epoch": 0.5485672656629432,
      "grad_norm": 0.36217546463012695,
      "learning_rate": 0.0001,
      "loss": 1.7194,
      "step": 2259
    },
    {
      "epoch": 0.5488101019912579,
      "grad_norm": 0.36972519755363464,
      "learning_rate": 0.0001,
      "loss": 1.7844,
      "step": 2260
    },
    {
      "epoch": 0.5490529383195726,
      "grad_norm": 0.3472967743873596,
      "learning_rate": 0.0001,
      "loss": 1.6209,
      "step": 2261
    },
    {
      "epoch": 0.5492957746478874,
      "grad_norm": 0.3597005009651184,
      "learning_rate": 0.0001,
      "loss": 1.6316,
      "step": 2262
    },
    {
      "epoch": 0.549538610976202,
      "grad_norm": 0.34558573365211487,
      "learning_rate": 0.0001,
      "loss": 1.6492,
      "step": 2263
    },
    {
      "epoch": 0.5497814473045167,
      "grad_norm": 0.3785961866378784,
      "learning_rate": 0.0001,
      "loss": 1.8447,
      "step": 2264
    },
    {
      "epoch": 0.5500242836328315,
      "grad_norm": 0.3740163743495941,
      "learning_rate": 0.0001,
      "loss": 1.7282,
      "step": 2265
    },
    {
      "epoch": 0.5502671199611462,
      "grad_norm": 0.37089261412620544,
      "learning_rate": 0.0001,
      "loss": 1.7876,
      "step": 2266
    },
    {
      "epoch": 0.5505099562894609,
      "grad_norm": 0.36666280031204224,
      "learning_rate": 0.0001,
      "loss": 1.7471,
      "step": 2267
    },
    {
      "epoch": 0.5507527926177757,
      "grad_norm": 0.36859360337257385,
      "learning_rate": 0.0001,
      "loss": 1.7612,
      "step": 2268
    },
    {
      "epoch": 0.5509956289460903,
      "grad_norm": 0.3185975253582001,
      "learning_rate": 0.0001,
      "loss": 1.5243,
      "step": 2269
    },
    {
      "epoch": 0.551238465274405,
      "grad_norm": 0.35294508934020996,
      "learning_rate": 0.0001,
      "loss": 1.7157,
      "step": 2270
    },
    {
      "epoch": 0.5514813016027198,
      "grad_norm": 0.3579111099243164,
      "learning_rate": 0.0001,
      "loss": 1.7531,
      "step": 2271
    },
    {
      "epoch": 0.5517241379310345,
      "grad_norm": 0.33986896276474,
      "learning_rate": 0.0001,
      "loss": 1.5978,
      "step": 2272
    },
    {
      "epoch": 0.5519669742593492,
      "grad_norm": 0.3544631600379944,
      "learning_rate": 0.0001,
      "loss": 1.7566,
      "step": 2273
    },
    {
      "epoch": 0.5522098105876639,
      "grad_norm": 0.3550451099872589,
      "learning_rate": 0.0001,
      "loss": 1.6645,
      "step": 2274
    },
    {
      "epoch": 0.5524526469159786,
      "grad_norm": 0.3410367965698242,
      "learning_rate": 0.0001,
      "loss": 1.7098,
      "step": 2275
    },
    {
      "epoch": 0.5526954832442933,
      "grad_norm": 0.37646061182022095,
      "learning_rate": 0.0001,
      "loss": 1.6585,
      "step": 2276
    },
    {
      "epoch": 0.5529383195726081,
      "grad_norm": 0.3732985854148865,
      "learning_rate": 0.0001,
      "loss": 1.7198,
      "step": 2277
    },
    {
      "epoch": 0.5531811559009228,
      "grad_norm": 0.3641611337661743,
      "learning_rate": 0.0001,
      "loss": 1.8386,
      "step": 2278
    },
    {
      "epoch": 0.5534239922292375,
      "grad_norm": 0.3658776879310608,
      "learning_rate": 0.0001,
      "loss": 1.7795,
      "step": 2279
    },
    {
      "epoch": 0.5536668285575522,
      "grad_norm": 0.37467020750045776,
      "learning_rate": 0.0001,
      "loss": 1.7865,
      "step": 2280
    },
    {
      "epoch": 0.5539096648858669,
      "grad_norm": 0.3677930533885956,
      "learning_rate": 0.0001,
      "loss": 1.7656,
      "step": 2281
    },
    {
      "epoch": 0.5541525012141817,
      "grad_norm": 0.3741012513637543,
      "learning_rate": 0.0001,
      "loss": 1.6091,
      "step": 2282
    },
    {
      "epoch": 0.5543953375424964,
      "grad_norm": 0.34404319524765015,
      "learning_rate": 0.0001,
      "loss": 1.6146,
      "step": 2283
    },
    {
      "epoch": 0.554638173870811,
      "grad_norm": 0.363204687833786,
      "learning_rate": 0.0001,
      "loss": 1.7769,
      "step": 2284
    },
    {
      "epoch": 0.5548810101991258,
      "grad_norm": 0.3678099811077118,
      "learning_rate": 0.0001,
      "loss": 1.7809,
      "step": 2285
    },
    {
      "epoch": 0.5551238465274405,
      "grad_norm": 0.3772222697734833,
      "learning_rate": 0.0001,
      "loss": 1.7406,
      "step": 2286
    },
    {
      "epoch": 0.5553666828557552,
      "grad_norm": 0.3508082330226898,
      "learning_rate": 0.0001,
      "loss": 1.772,
      "step": 2287
    },
    {
      "epoch": 0.55560951918407,
      "grad_norm": 0.35846176743507385,
      "learning_rate": 0.0001,
      "loss": 1.8131,
      "step": 2288
    },
    {
      "epoch": 0.5558523555123847,
      "grad_norm": 0.35966917872428894,
      "learning_rate": 0.0001,
      "loss": 1.677,
      "step": 2289
    },
    {
      "epoch": 0.5560951918406993,
      "grad_norm": 0.35907042026519775,
      "learning_rate": 0.0001,
      "loss": 1.7196,
      "step": 2290
    },
    {
      "epoch": 0.5563380281690141,
      "grad_norm": 0.36267662048339844,
      "learning_rate": 0.0001,
      "loss": 1.8167,
      "step": 2291
    },
    {
      "epoch": 0.5565808644973288,
      "grad_norm": 0.3560062050819397,
      "learning_rate": 0.0001,
      "loss": 1.7889,
      "step": 2292
    },
    {
      "epoch": 0.5568237008256435,
      "grad_norm": 0.3562415540218353,
      "learning_rate": 0.0001,
      "loss": 1.6888,
      "step": 2293
    },
    {
      "epoch": 0.5570665371539583,
      "grad_norm": 0.365655779838562,
      "learning_rate": 0.0001,
      "loss": 1.8064,
      "step": 2294
    },
    {
      "epoch": 0.5573093734822729,
      "grad_norm": 0.35320165753364563,
      "learning_rate": 0.0001,
      "loss": 1.7139,
      "step": 2295
    },
    {
      "epoch": 0.5575522098105876,
      "grad_norm": 0.39178183674812317,
      "learning_rate": 0.0001,
      "loss": 1.6381,
      "step": 2296
    },
    {
      "epoch": 0.5577950461389024,
      "grad_norm": 0.36920660734176636,
      "learning_rate": 0.0001,
      "loss": 1.81,
      "step": 2297
    },
    {
      "epoch": 0.5580378824672171,
      "grad_norm": 0.362719863653183,
      "learning_rate": 0.0001,
      "loss": 1.6541,
      "step": 2298
    },
    {
      "epoch": 0.5582807187955318,
      "grad_norm": 0.3680153489112854,
      "learning_rate": 0.0001,
      "loss": 1.9173,
      "step": 2299
    },
    {
      "epoch": 0.5585235551238465,
      "grad_norm": 0.37122082710266113,
      "learning_rate": 0.0001,
      "loss": 1.8223,
      "step": 2300
    },
    {
      "epoch": 0.5587663914521612,
      "grad_norm": 0.37359800934791565,
      "learning_rate": 0.0001,
      "loss": 1.7213,
      "step": 2301
    },
    {
      "epoch": 0.559009227780476,
      "grad_norm": 0.36098548769950867,
      "learning_rate": 0.0001,
      "loss": 1.7291,
      "step": 2302
    },
    {
      "epoch": 0.5592520641087907,
      "grad_norm": 0.3613625466823578,
      "learning_rate": 0.0001,
      "loss": 1.7256,
      "step": 2303
    },
    {
      "epoch": 0.5594949004371054,
      "grad_norm": 0.35531091690063477,
      "learning_rate": 0.0001,
      "loss": 1.7711,
      "step": 2304
    },
    {
      "epoch": 0.5597377367654202,
      "grad_norm": 0.37719884514808655,
      "learning_rate": 0.0001,
      "loss": 1.7064,
      "step": 2305
    },
    {
      "epoch": 0.5599805730937348,
      "grad_norm": 0.33986160159111023,
      "learning_rate": 0.0001,
      "loss": 1.6904,
      "step": 2306
    },
    {
      "epoch": 0.5602234094220495,
      "grad_norm": 0.3476993143558502,
      "learning_rate": 0.0001,
      "loss": 1.757,
      "step": 2307
    },
    {
      "epoch": 0.5604662457503643,
      "grad_norm": 0.3476622402667999,
      "learning_rate": 0.0001,
      "loss": 1.6747,
      "step": 2308
    },
    {
      "epoch": 0.560709082078679,
      "grad_norm": 0.33027929067611694,
      "learning_rate": 0.0001,
      "loss": 1.4717,
      "step": 2309
    },
    {
      "epoch": 0.5609519184069937,
      "grad_norm": 0.3475174307823181,
      "learning_rate": 0.0001,
      "loss": 1.7306,
      "step": 2310
    },
    {
      "epoch": 0.5611947547353084,
      "grad_norm": 0.3623664081096649,
      "learning_rate": 0.0001,
      "loss": 1.6337,
      "step": 2311
    },
    {
      "epoch": 0.5614375910636231,
      "grad_norm": 0.3583698272705078,
      "learning_rate": 0.0001,
      "loss": 1.7394,
      "step": 2312
    },
    {
      "epoch": 0.5616804273919378,
      "grad_norm": 0.3368013799190521,
      "learning_rate": 0.0001,
      "loss": 1.7616,
      "step": 2313
    },
    {
      "epoch": 0.5619232637202526,
      "grad_norm": 0.3758639395236969,
      "learning_rate": 0.0001,
      "loss": 1.6414,
      "step": 2314
    },
    {
      "epoch": 0.5621661000485673,
      "grad_norm": 0.3569642901420593,
      "learning_rate": 0.0001,
      "loss": 1.7582,
      "step": 2315
    },
    {
      "epoch": 0.5624089363768819,
      "grad_norm": 0.38806402683258057,
      "learning_rate": 0.0001,
      "loss": 1.9369,
      "step": 2316
    },
    {
      "epoch": 0.5626517727051967,
      "grad_norm": 0.3777295649051666,
      "learning_rate": 0.0001,
      "loss": 1.8661,
      "step": 2317
    },
    {
      "epoch": 0.5628946090335114,
      "grad_norm": 0.33787399530410767,
      "learning_rate": 0.0001,
      "loss": 1.6169,
      "step": 2318
    },
    {
      "epoch": 0.5631374453618261,
      "grad_norm": 0.3719368875026703,
      "learning_rate": 0.0001,
      "loss": 1.7745,
      "step": 2319
    },
    {
      "epoch": 0.5633802816901409,
      "grad_norm": 0.3340829610824585,
      "learning_rate": 0.0001,
      "loss": 1.5761,
      "step": 2320
    },
    {
      "epoch": 0.5636231180184555,
      "grad_norm": 0.3609652519226074,
      "learning_rate": 0.0001,
      "loss": 1.7269,
      "step": 2321
    },
    {
      "epoch": 0.5638659543467702,
      "grad_norm": 0.34608790278434753,
      "learning_rate": 0.0001,
      "loss": 1.5749,
      "step": 2322
    },
    {
      "epoch": 0.564108790675085,
      "grad_norm": 0.3639295697212219,
      "learning_rate": 0.0001,
      "loss": 1.7738,
      "step": 2323
    },
    {
      "epoch": 0.5643516270033997,
      "grad_norm": 0.3458191454410553,
      "learning_rate": 0.0001,
      "loss": 1.7874,
      "step": 2324
    },
    {
      "epoch": 0.5645944633317144,
      "grad_norm": 0.36620959639549255,
      "learning_rate": 0.0001,
      "loss": 1.7537,
      "step": 2325
    },
    {
      "epoch": 0.5648372996600292,
      "grad_norm": 0.3745834231376648,
      "learning_rate": 0.0001,
      "loss": 1.7993,
      "step": 2326
    },
    {
      "epoch": 0.5650801359883438,
      "grad_norm": 0.3734908401966095,
      "learning_rate": 0.0001,
      "loss": 1.6732,
      "step": 2327
    },
    {
      "epoch": 0.5653229723166586,
      "grad_norm": 0.35607555508613586,
      "learning_rate": 0.0001,
      "loss": 1.831,
      "step": 2328
    },
    {
      "epoch": 0.5655658086449733,
      "grad_norm": 0.3680264949798584,
      "learning_rate": 0.0001,
      "loss": 1.671,
      "step": 2329
    },
    {
      "epoch": 0.565808644973288,
      "grad_norm": 0.33606141805648804,
      "learning_rate": 0.0001,
      "loss": 1.4986,
      "step": 2330
    },
    {
      "epoch": 0.5660514813016028,
      "grad_norm": 0.3809346854686737,
      "learning_rate": 0.0001,
      "loss": 1.8307,
      "step": 2331
    },
    {
      "epoch": 0.5662943176299174,
      "grad_norm": 0.3508315682411194,
      "learning_rate": 0.0001,
      "loss": 1.6756,
      "step": 2332
    },
    {
      "epoch": 0.5665371539582321,
      "grad_norm": 0.34755823016166687,
      "learning_rate": 0.0001,
      "loss": 1.6317,
      "step": 2333
    },
    {
      "epoch": 0.5667799902865469,
      "grad_norm": 0.3529258966445923,
      "learning_rate": 0.0001,
      "loss": 1.6157,
      "step": 2334
    },
    {
      "epoch": 0.5670228266148616,
      "grad_norm": 0.3624809980392456,
      "learning_rate": 0.0001,
      "loss": 1.7348,
      "step": 2335
    },
    {
      "epoch": 0.5672656629431763,
      "grad_norm": 0.36164456605911255,
      "learning_rate": 0.0001,
      "loss": 1.7053,
      "step": 2336
    },
    {
      "epoch": 0.567508499271491,
      "grad_norm": 0.3604152500629425,
      "learning_rate": 0.0001,
      "loss": 1.7097,
      "step": 2337
    },
    {
      "epoch": 0.5677513355998057,
      "grad_norm": 0.3611051142215729,
      "learning_rate": 0.0001,
      "loss": 1.7406,
      "step": 2338
    },
    {
      "epoch": 0.5679941719281204,
      "grad_norm": 0.3456195294857025,
      "learning_rate": 0.0001,
      "loss": 1.7376,
      "step": 2339
    },
    {
      "epoch": 0.5682370082564352,
      "grad_norm": 0.34690234065055847,
      "learning_rate": 0.0001,
      "loss": 1.7485,
      "step": 2340
    },
    {
      "epoch": 0.5684798445847499,
      "grad_norm": 0.39296525716781616,
      "learning_rate": 0.0001,
      "loss": 1.7647,
      "step": 2341
    },
    {
      "epoch": 0.5687226809130645,
      "grad_norm": 0.36980095505714417,
      "learning_rate": 0.0001,
      "loss": 1.8579,
      "step": 2342
    },
    {
      "epoch": 0.5689655172413793,
      "grad_norm": 0.36739233136177063,
      "learning_rate": 0.0001,
      "loss": 1.7161,
      "step": 2343
    },
    {
      "epoch": 0.569208353569694,
      "grad_norm": 0.3408131003379822,
      "learning_rate": 0.0001,
      "loss": 1.5564,
      "step": 2344
    },
    {
      "epoch": 0.5694511898980087,
      "grad_norm": 0.3714952766895294,
      "learning_rate": 0.0001,
      "loss": 1.9206,
      "step": 2345
    },
    {
      "epoch": 0.5696940262263235,
      "grad_norm": 0.34982210397720337,
      "learning_rate": 0.0001,
      "loss": 1.5884,
      "step": 2346
    },
    {
      "epoch": 0.5699368625546382,
      "grad_norm": 0.35678350925445557,
      "learning_rate": 0.0001,
      "loss": 1.7792,
      "step": 2347
    },
    {
      "epoch": 0.5701796988829528,
      "grad_norm": 0.35591959953308105,
      "learning_rate": 0.0001,
      "loss": 1.7195,
      "step": 2348
    },
    {
      "epoch": 0.5704225352112676,
      "grad_norm": 0.3432258665561676,
      "learning_rate": 0.0001,
      "loss": 1.7125,
      "step": 2349
    },
    {
      "epoch": 0.5706653715395823,
      "grad_norm": 0.37227383255958557,
      "learning_rate": 0.0001,
      "loss": 1.7048,
      "step": 2350
    },
    {
      "epoch": 0.5709082078678971,
      "grad_norm": 0.34763336181640625,
      "learning_rate": 0.0001,
      "loss": 1.6562,
      "step": 2351
    },
    {
      "epoch": 0.5711510441962118,
      "grad_norm": 0.3854842185974121,
      "learning_rate": 0.0001,
      "loss": 1.5957,
      "step": 2352
    },
    {
      "epoch": 0.5713938805245264,
      "grad_norm": 0.364199697971344,
      "learning_rate": 0.0001,
      "loss": 1.7444,
      "step": 2353
    },
    {
      "epoch": 0.5716367168528412,
      "grad_norm": 0.35069018602371216,
      "learning_rate": 0.0001,
      "loss": 1.6804,
      "step": 2354
    },
    {
      "epoch": 0.5718795531811559,
      "grad_norm": 0.34375110268592834,
      "learning_rate": 0.0001,
      "loss": 1.6428,
      "step": 2355
    },
    {
      "epoch": 0.5721223895094706,
      "grad_norm": 0.3363782465457916,
      "learning_rate": 0.0001,
      "loss": 1.6488,
      "step": 2356
    },
    {
      "epoch": 0.5723652258377854,
      "grad_norm": 0.3884119987487793,
      "learning_rate": 0.0001,
      "loss": 1.7942,
      "step": 2357
    },
    {
      "epoch": 0.5726080621661,
      "grad_norm": 0.3595485985279083,
      "learning_rate": 0.0001,
      "loss": 1.7423,
      "step": 2358
    },
    {
      "epoch": 0.5728508984944147,
      "grad_norm": 0.3650807738304138,
      "learning_rate": 0.0001,
      "loss": 1.751,
      "step": 2359
    },
    {
      "epoch": 0.5730937348227295,
      "grad_norm": 0.3601040244102478,
      "learning_rate": 0.0001,
      "loss": 1.6485,
      "step": 2360
    },
    {
      "epoch": 0.5733365711510442,
      "grad_norm": 0.3422304689884186,
      "learning_rate": 0.0001,
      "loss": 1.6074,
      "step": 2361
    },
    {
      "epoch": 0.5735794074793589,
      "grad_norm": 0.35458236932754517,
      "learning_rate": 0.0001,
      "loss": 1.6944,
      "step": 2362
    },
    {
      "epoch": 0.5738222438076737,
      "grad_norm": 0.35751640796661377,
      "learning_rate": 0.0001,
      "loss": 1.7291,
      "step": 2363
    },
    {
      "epoch": 0.5740650801359883,
      "grad_norm": 0.34842342138290405,
      "learning_rate": 0.0001,
      "loss": 1.6863,
      "step": 2364
    },
    {
      "epoch": 0.574307916464303,
      "grad_norm": 0.352594256401062,
      "learning_rate": 0.0001,
      "loss": 1.7502,
      "step": 2365
    },
    {
      "epoch": 0.5745507527926178,
      "grad_norm": 0.36940956115722656,
      "learning_rate": 0.0001,
      "loss": 1.723,
      "step": 2366
    },
    {
      "epoch": 0.5747935891209325,
      "grad_norm": 0.34755411744117737,
      "learning_rate": 0.0001,
      "loss": 1.7728,
      "step": 2367
    },
    {
      "epoch": 0.5750364254492472,
      "grad_norm": 0.34088075160980225,
      "learning_rate": 0.0001,
      "loss": 1.6711,
      "step": 2368
    },
    {
      "epoch": 0.575279261777562,
      "grad_norm": 0.36846232414245605,
      "learning_rate": 0.0001,
      "loss": 1.695,
      "step": 2369
    },
    {
      "epoch": 0.5755220981058766,
      "grad_norm": 0.3828352391719818,
      "learning_rate": 0.0001,
      "loss": 1.8301,
      "step": 2370
    },
    {
      "epoch": 0.5757649344341913,
      "grad_norm": 0.3610253930091858,
      "learning_rate": 0.0001,
      "loss": 1.6749,
      "step": 2371
    },
    {
      "epoch": 0.5760077707625061,
      "grad_norm": 0.34908774495124817,
      "learning_rate": 0.0001,
      "loss": 1.7736,
      "step": 2372
    },
    {
      "epoch": 0.5762506070908208,
      "grad_norm": 0.3399523198604584,
      "learning_rate": 0.0001,
      "loss": 1.7515,
      "step": 2373
    },
    {
      "epoch": 0.5764934434191356,
      "grad_norm": 0.3800559341907501,
      "learning_rate": 0.0001,
      "loss": 1.828,
      "step": 2374
    },
    {
      "epoch": 0.5767362797474502,
      "grad_norm": 0.33364370465278625,
      "learning_rate": 0.0001,
      "loss": 1.6143,
      "step": 2375
    },
    {
      "epoch": 0.5769791160757649,
      "grad_norm": 0.3978668451309204,
      "learning_rate": 0.0001,
      "loss": 1.8787,
      "step": 2376
    },
    {
      "epoch": 0.5772219524040797,
      "grad_norm": 0.3593222498893738,
      "learning_rate": 0.0001,
      "loss": 1.734,
      "step": 2377
    },
    {
      "epoch": 0.5774647887323944,
      "grad_norm": 0.3668678104877472,
      "learning_rate": 0.0001,
      "loss": 1.7804,
      "step": 2378
    },
    {
      "epoch": 0.577707625060709,
      "grad_norm": 0.36095866560935974,
      "learning_rate": 0.0001,
      "loss": 1.6346,
      "step": 2379
    },
    {
      "epoch": 0.5779504613890238,
      "grad_norm": 0.33825570344924927,
      "learning_rate": 0.0001,
      "loss": 1.6229,
      "step": 2380
    },
    {
      "epoch": 0.5781932977173385,
      "grad_norm": 0.3309857249259949,
      "learning_rate": 0.0001,
      "loss": 1.4445,
      "step": 2381
    },
    {
      "epoch": 0.5784361340456532,
      "grad_norm": 0.3851270079612732,
      "learning_rate": 0.0001,
      "loss": 1.8312,
      "step": 2382
    },
    {
      "epoch": 0.578678970373968,
      "grad_norm": 0.3448592722415924,
      "learning_rate": 0.0001,
      "loss": 1.7032,
      "step": 2383
    },
    {
      "epoch": 0.5789218067022827,
      "grad_norm": 0.3687342405319214,
      "learning_rate": 0.0001,
      "loss": 1.7347,
      "step": 2384
    },
    {
      "epoch": 0.5791646430305973,
      "grad_norm": 0.3448163568973541,
      "learning_rate": 0.0001,
      "loss": 1.6357,
      "step": 2385
    },
    {
      "epoch": 0.5794074793589121,
      "grad_norm": 0.3667164742946625,
      "learning_rate": 0.0001,
      "loss": 1.7147,
      "step": 2386
    },
    {
      "epoch": 0.5796503156872268,
      "grad_norm": 0.3618122935295105,
      "learning_rate": 0.0001,
      "loss": 1.746,
      "step": 2387
    },
    {
      "epoch": 0.5798931520155415,
      "grad_norm": 0.3543204963207245,
      "learning_rate": 0.0001,
      "loss": 1.7053,
      "step": 2388
    },
    {
      "epoch": 0.5801359883438563,
      "grad_norm": 0.3682529032230377,
      "learning_rate": 0.0001,
      "loss": 1.7884,
      "step": 2389
    },
    {
      "epoch": 0.580378824672171,
      "grad_norm": 0.3413553237915039,
      "learning_rate": 0.0001,
      "loss": 1.619,
      "step": 2390
    },
    {
      "epoch": 0.5806216610004856,
      "grad_norm": 0.3471716642379761,
      "learning_rate": 0.0001,
      "loss": 1.6995,
      "step": 2391
    },
    {
      "epoch": 0.5808644973288004,
      "grad_norm": 0.34798210859298706,
      "learning_rate": 0.0001,
      "loss": 1.6615,
      "step": 2392
    },
    {
      "epoch": 0.5811073336571151,
      "grad_norm": 0.3592934012413025,
      "learning_rate": 0.0001,
      "loss": 1.603,
      "step": 2393
    },
    {
      "epoch": 0.5813501699854298,
      "grad_norm": 0.3641080856323242,
      "learning_rate": 0.0001,
      "loss": 1.8342,
      "step": 2394
    },
    {
      "epoch": 0.5815930063137446,
      "grad_norm": 0.3546813726425171,
      "learning_rate": 0.0001,
      "loss": 1.7738,
      "step": 2395
    },
    {
      "epoch": 0.5818358426420592,
      "grad_norm": 0.3424903452396393,
      "learning_rate": 0.0001,
      "loss": 1.6683,
      "step": 2396
    },
    {
      "epoch": 0.582078678970374,
      "grad_norm": 0.3579918444156647,
      "learning_rate": 0.0001,
      "loss": 1.6985,
      "step": 2397
    },
    {
      "epoch": 0.5823215152986887,
      "grad_norm": 0.35348811745643616,
      "learning_rate": 0.0001,
      "loss": 1.6904,
      "step": 2398
    },
    {
      "epoch": 0.5825643516270034,
      "grad_norm": 0.3623894155025482,
      "learning_rate": 0.0001,
      "loss": 1.7517,
      "step": 2399
    },
    {
      "epoch": 0.5828071879553182,
      "grad_norm": 0.36576738953590393,
      "learning_rate": 0.0001,
      "loss": 1.7106,
      "step": 2400
    },
    {
      "epoch": 0.5830500242836328,
      "grad_norm": 0.3525370657444,
      "learning_rate": 0.0001,
      "loss": 1.5246,
      "step": 2401
    },
    {
      "epoch": 0.5832928606119475,
      "grad_norm": 0.3752543330192566,
      "learning_rate": 0.0001,
      "loss": 1.758,
      "step": 2402
    },
    {
      "epoch": 0.5835356969402623,
      "grad_norm": 0.364324152469635,
      "learning_rate": 0.0001,
      "loss": 1.7315,
      "step": 2403
    },
    {
      "epoch": 0.583778533268577,
      "grad_norm": 0.3631533086299896,
      "learning_rate": 0.0001,
      "loss": 1.6388,
      "step": 2404
    },
    {
      "epoch": 0.5840213695968917,
      "grad_norm": 0.3875806927680969,
      "learning_rate": 0.0001,
      "loss": 1.9683,
      "step": 2405
    },
    {
      "epoch": 0.5842642059252064,
      "grad_norm": 0.35269179940223694,
      "learning_rate": 0.0001,
      "loss": 1.6662,
      "step": 2406
    },
    {
      "epoch": 0.5845070422535211,
      "grad_norm": 0.3396052122116089,
      "learning_rate": 0.0001,
      "loss": 1.6706,
      "step": 2407
    },
    {
      "epoch": 0.5847498785818358,
      "grad_norm": 0.3488730490207672,
      "learning_rate": 0.0001,
      "loss": 1.5859,
      "step": 2408
    },
    {
      "epoch": 0.5849927149101506,
      "grad_norm": 0.356308251619339,
      "learning_rate": 0.0001,
      "loss": 1.6471,
      "step": 2409
    },
    {
      "epoch": 0.5852355512384653,
      "grad_norm": 0.366447389125824,
      "learning_rate": 0.0001,
      "loss": 1.6691,
      "step": 2410
    },
    {
      "epoch": 0.58547838756678,
      "grad_norm": 0.35855424404144287,
      "learning_rate": 0.0001,
      "loss": 1.7121,
      "step": 2411
    },
    {
      "epoch": 0.5857212238950947,
      "grad_norm": 0.3608049154281616,
      "learning_rate": 0.0001,
      "loss": 1.6389,
      "step": 2412
    },
    {
      "epoch": 0.5859640602234094,
      "grad_norm": 0.3465087115764618,
      "learning_rate": 0.0001,
      "loss": 1.7342,
      "step": 2413
    },
    {
      "epoch": 0.5862068965517241,
      "grad_norm": 0.38191014528274536,
      "learning_rate": 0.0001,
      "loss": 1.7077,
      "step": 2414
    },
    {
      "epoch": 0.5864497328800389,
      "grad_norm": 0.36607179045677185,
      "learning_rate": 0.0001,
      "loss": 1.6006,
      "step": 2415
    },
    {
      "epoch": 0.5866925692083536,
      "grad_norm": 0.3555096983909607,
      "learning_rate": 0.0001,
      "loss": 1.6222,
      "step": 2416
    },
    {
      "epoch": 0.5869354055366682,
      "grad_norm": 0.36466822028160095,
      "learning_rate": 0.0001,
      "loss": 1.8408,
      "step": 2417
    },
    {
      "epoch": 0.587178241864983,
      "grad_norm": 0.36148613691329956,
      "learning_rate": 0.0001,
      "loss": 1.8149,
      "step": 2418
    },
    {
      "epoch": 0.5874210781932977,
      "grad_norm": 0.36628925800323486,
      "learning_rate": 0.0001,
      "loss": 1.7473,
      "step": 2419
    },
    {
      "epoch": 0.5876639145216125,
      "grad_norm": 0.32230299711227417,
      "learning_rate": 0.0001,
      "loss": 1.5559,
      "step": 2420
    },
    {
      "epoch": 0.5879067508499272,
      "grad_norm": 0.3491017818450928,
      "learning_rate": 0.0001,
      "loss": 1.6424,
      "step": 2421
    },
    {
      "epoch": 0.5881495871782418,
      "grad_norm": 0.3417803645133972,
      "learning_rate": 0.0001,
      "loss": 1.5939,
      "step": 2422
    },
    {
      "epoch": 0.5883924235065566,
      "grad_norm": 0.3433781862258911,
      "learning_rate": 0.0001,
      "loss": 1.6713,
      "step": 2423
    },
    {
      "epoch": 0.5886352598348713,
      "grad_norm": 0.35483452677726746,
      "learning_rate": 0.0001,
      "loss": 1.6194,
      "step": 2424
    },
    {
      "epoch": 0.588878096163186,
      "grad_norm": 0.3255854547023773,
      "learning_rate": 0.0001,
      "loss": 1.6343,
      "step": 2425
    },
    {
      "epoch": 0.5891209324915008,
      "grad_norm": 0.3660036325454712,
      "learning_rate": 0.0001,
      "loss": 1.7423,
      "step": 2426
    },
    {
      "epoch": 0.5893637688198154,
      "grad_norm": 0.36366453766822815,
      "learning_rate": 0.0001,
      "loss": 1.7891,
      "step": 2427
    },
    {
      "epoch": 0.5896066051481301,
      "grad_norm": 0.36128878593444824,
      "learning_rate": 0.0001,
      "loss": 1.7371,
      "step": 2428
    },
    {
      "epoch": 0.5898494414764449,
      "grad_norm": 0.3683059811592102,
      "learning_rate": 0.0001,
      "loss": 1.7501,
      "step": 2429
    },
    {
      "epoch": 0.5900922778047596,
      "grad_norm": 0.37489473819732666,
      "learning_rate": 0.0001,
      "loss": 1.7181,
      "step": 2430
    },
    {
      "epoch": 0.5903351141330743,
      "grad_norm": 0.34531599283218384,
      "learning_rate": 0.0001,
      "loss": 1.541,
      "step": 2431
    },
    {
      "epoch": 0.5905779504613891,
      "grad_norm": 0.37469252943992615,
      "learning_rate": 0.0001,
      "loss": 1.8827,
      "step": 2432
    },
    {
      "epoch": 0.5908207867897037,
      "grad_norm": 0.36275172233581543,
      "learning_rate": 0.0001,
      "loss": 1.7681,
      "step": 2433
    },
    {
      "epoch": 0.5910636231180184,
      "grad_norm": 0.35241881012916565,
      "learning_rate": 0.0001,
      "loss": 1.6363,
      "step": 2434
    },
    {
      "epoch": 0.5913064594463332,
      "grad_norm": 0.36884981393814087,
      "learning_rate": 0.0001,
      "loss": 1.6707,
      "step": 2435
    },
    {
      "epoch": 0.5915492957746479,
      "grad_norm": 0.353599488735199,
      "learning_rate": 0.0001,
      "loss": 1.7469,
      "step": 2436
    },
    {
      "epoch": 0.5917921321029626,
      "grad_norm": 0.3728369176387787,
      "learning_rate": 0.0001,
      "loss": 1.8387,
      "step": 2437
    },
    {
      "epoch": 0.5920349684312773,
      "grad_norm": 0.3472503125667572,
      "learning_rate": 0.0001,
      "loss": 1.6978,
      "step": 2438
    },
    {
      "epoch": 0.592277804759592,
      "grad_norm": 0.3670949339866638,
      "learning_rate": 0.0001,
      "loss": 1.7349,
      "step": 2439
    },
    {
      "epoch": 0.5925206410879067,
      "grad_norm": 0.3487583100795746,
      "learning_rate": 0.0001,
      "loss": 1.8164,
      "step": 2440
    },
    {
      "epoch": 0.5927634774162215,
      "grad_norm": 0.37606680393218994,
      "learning_rate": 0.0001,
      "loss": 1.7293,
      "step": 2441
    },
    {
      "epoch": 0.5930063137445362,
      "grad_norm": 0.3815697729587555,
      "learning_rate": 0.0001,
      "loss": 1.6828,
      "step": 2442
    },
    {
      "epoch": 0.593249150072851,
      "grad_norm": 0.35163789987564087,
      "learning_rate": 0.0001,
      "loss": 1.7363,
      "step": 2443
    },
    {
      "epoch": 0.5934919864011656,
      "grad_norm": 0.3514600396156311,
      "learning_rate": 0.0001,
      "loss": 1.7021,
      "step": 2444
    },
    {
      "epoch": 0.5937348227294803,
      "grad_norm": 0.3635815978050232,
      "learning_rate": 0.0001,
      "loss": 1.6672,
      "step": 2445
    },
    {
      "epoch": 0.5939776590577951,
      "grad_norm": 0.37155693769454956,
      "learning_rate": 0.0001,
      "loss": 1.6893,
      "step": 2446
    },
    {
      "epoch": 0.5942204953861098,
      "grad_norm": 0.3761202394962311,
      "learning_rate": 0.0001,
      "loss": 1.777,
      "step": 2447
    },
    {
      "epoch": 0.5944633317144244,
      "grad_norm": 0.3590002954006195,
      "learning_rate": 0.0001,
      "loss": 1.7537,
      "step": 2448
    },
    {
      "epoch": 0.5947061680427392,
      "grad_norm": 0.35956037044525146,
      "learning_rate": 0.0001,
      "loss": 1.6303,
      "step": 2449
    },
    {
      "epoch": 0.5949490043710539,
      "grad_norm": 0.345742791891098,
      "learning_rate": 0.0001,
      "loss": 1.8315,
      "step": 2450
    },
    {
      "epoch": 0.5951918406993686,
      "grad_norm": 0.35536545515060425,
      "learning_rate": 0.0001,
      "loss": 1.7572,
      "step": 2451
    },
    {
      "epoch": 0.5954346770276834,
      "grad_norm": 0.35663557052612305,
      "learning_rate": 0.0001,
      "loss": 1.8523,
      "step": 2452
    },
    {
      "epoch": 0.5956775133559981,
      "grad_norm": 0.341354101896286,
      "learning_rate": 0.0001,
      "loss": 1.7301,
      "step": 2453
    },
    {
      "epoch": 0.5959203496843127,
      "grad_norm": 0.3717452883720398,
      "learning_rate": 0.0001,
      "loss": 1.7591,
      "step": 2454
    },
    {
      "epoch": 0.5961631860126275,
      "grad_norm": 0.3513641357421875,
      "learning_rate": 0.0001,
      "loss": 1.565,
      "step": 2455
    },
    {
      "epoch": 0.5964060223409422,
      "grad_norm": 0.35649535059928894,
      "learning_rate": 0.0001,
      "loss": 1.6757,
      "step": 2456
    },
    {
      "epoch": 0.5966488586692569,
      "grad_norm": 0.375552237033844,
      "learning_rate": 0.0001,
      "loss": 1.7401,
      "step": 2457
    },
    {
      "epoch": 0.5968916949975717,
      "grad_norm": 0.3812007009983063,
      "learning_rate": 0.0001,
      "loss": 1.7399,
      "step": 2458
    },
    {
      "epoch": 0.5971345313258863,
      "grad_norm": 0.3755456805229187,
      "learning_rate": 0.0001,
      "loss": 1.7869,
      "step": 2459
    },
    {
      "epoch": 0.597377367654201,
      "grad_norm": 0.357443243265152,
      "learning_rate": 0.0001,
      "loss": 1.7365,
      "step": 2460
    },
    {
      "epoch": 0.5976202039825158,
      "grad_norm": 0.3731152415275574,
      "learning_rate": 0.0001,
      "loss": 1.677,
      "step": 2461
    },
    {
      "epoch": 0.5978630403108305,
      "grad_norm": 0.3395928144454956,
      "learning_rate": 0.0001,
      "loss": 1.4929,
      "step": 2462
    },
    {
      "epoch": 0.5981058766391452,
      "grad_norm": 0.36976638436317444,
      "learning_rate": 0.0001,
      "loss": 1.864,
      "step": 2463
    },
    {
      "epoch": 0.59834871296746,
      "grad_norm": 0.3605227470397949,
      "learning_rate": 0.0001,
      "loss": 1.733,
      "step": 2464
    },
    {
      "epoch": 0.5985915492957746,
      "grad_norm": 0.3721519112586975,
      "learning_rate": 0.0001,
      "loss": 1.5462,
      "step": 2465
    },
    {
      "epoch": 0.5988343856240894,
      "grad_norm": 0.3864923119544983,
      "learning_rate": 0.0001,
      "loss": 1.7985,
      "step": 2466
    },
    {
      "epoch": 0.5990772219524041,
      "grad_norm": 0.35037946701049805,
      "learning_rate": 0.0001,
      "loss": 1.8543,
      "step": 2467
    },
    {
      "epoch": 0.5993200582807188,
      "grad_norm": 0.3450808525085449,
      "learning_rate": 0.0001,
      "loss": 1.6291,
      "step": 2468
    },
    {
      "epoch": 0.5995628946090336,
      "grad_norm": 0.3674449920654297,
      "learning_rate": 0.0001,
      "loss": 1.7048,
      "step": 2469
    },
    {
      "epoch": 0.5998057309373482,
      "grad_norm": 0.3634595274925232,
      "learning_rate": 0.0001,
      "loss": 1.8068,
      "step": 2470
    },
    {
      "epoch": 0.6000485672656629,
      "grad_norm": 0.36641907691955566,
      "learning_rate": 0.0001,
      "loss": 1.7807,
      "step": 2471
    },
    {
      "epoch": 0.6002914035939777,
      "grad_norm": 0.3526882231235504,
      "learning_rate": 0.0001,
      "loss": 1.6572,
      "step": 2472
    },
    {
      "epoch": 0.6005342399222924,
      "grad_norm": 0.33994728326797485,
      "learning_rate": 0.0001,
      "loss": 1.6133,
      "step": 2473
    },
    {
      "epoch": 0.6007770762506071,
      "grad_norm": 0.35702425241470337,
      "learning_rate": 0.0001,
      "loss": 1.75,
      "step": 2474
    },
    {
      "epoch": 0.6010199125789218,
      "grad_norm": 0.3425130546092987,
      "learning_rate": 0.0001,
      "loss": 1.651,
      "step": 2475
    },
    {
      "epoch": 0.6012627489072365,
      "grad_norm": 0.37431252002716064,
      "learning_rate": 0.0001,
      "loss": 1.5988,
      "step": 2476
    },
    {
      "epoch": 0.6015055852355512,
      "grad_norm": 0.3774125874042511,
      "learning_rate": 0.0001,
      "loss": 1.8422,
      "step": 2477
    },
    {
      "epoch": 0.601748421563866,
      "grad_norm": 0.36197513341903687,
      "learning_rate": 0.0001,
      "loss": 1.7569,
      "step": 2478
    },
    {
      "epoch": 0.6019912578921807,
      "grad_norm": 0.37086841464042664,
      "learning_rate": 0.0001,
      "loss": 1.6805,
      "step": 2479
    },
    {
      "epoch": 0.6022340942204953,
      "grad_norm": 0.3471476137638092,
      "learning_rate": 0.0001,
      "loss": 1.7082,
      "step": 2480
    },
    {
      "epoch": 0.6024769305488101,
      "grad_norm": 0.34941521286964417,
      "learning_rate": 0.0001,
      "loss": 1.6494,
      "step": 2481
    },
    {
      "epoch": 0.6027197668771248,
      "grad_norm": 0.3552822768688202,
      "learning_rate": 0.0001,
      "loss": 1.6597,
      "step": 2482
    },
    {
      "epoch": 0.6029626032054395,
      "grad_norm": 0.3908756673336029,
      "learning_rate": 0.0001,
      "loss": 1.7776,
      "step": 2483
    },
    {
      "epoch": 0.6032054395337543,
      "grad_norm": 0.3638925850391388,
      "learning_rate": 0.0001,
      "loss": 1.6662,
      "step": 2484
    },
    {
      "epoch": 0.603448275862069,
      "grad_norm": 0.35275381803512573,
      "learning_rate": 0.0001,
      "loss": 1.6222,
      "step": 2485
    },
    {
      "epoch": 0.6036911121903836,
      "grad_norm": 0.34850969910621643,
      "learning_rate": 0.0001,
      "loss": 1.7403,
      "step": 2486
    },
    {
      "epoch": 0.6039339485186984,
      "grad_norm": 0.3468134105205536,
      "learning_rate": 0.0001,
      "loss": 1.7265,
      "step": 2487
    },
    {
      "epoch": 0.6041767848470131,
      "grad_norm": 0.36274734139442444,
      "learning_rate": 0.0001,
      "loss": 1.6525,
      "step": 2488
    },
    {
      "epoch": 0.6044196211753279,
      "grad_norm": 0.3470844328403473,
      "learning_rate": 0.0001,
      "loss": 1.699,
      "step": 2489
    },
    {
      "epoch": 0.6046624575036426,
      "grad_norm": 0.3809892535209656,
      "learning_rate": 0.0001,
      "loss": 1.8901,
      "step": 2490
    },
    {
      "epoch": 0.6049052938319572,
      "grad_norm": 0.35343703627586365,
      "learning_rate": 0.0001,
      "loss": 1.6325,
      "step": 2491
    },
    {
      "epoch": 0.605148130160272,
      "grad_norm": 0.345963716506958,
      "learning_rate": 0.0001,
      "loss": 1.7031,
      "step": 2492
    },
    {
      "epoch": 0.6053909664885867,
      "grad_norm": 0.37025344371795654,
      "learning_rate": 0.0001,
      "loss": 1.8492,
      "step": 2493
    },
    {
      "epoch": 0.6056338028169014,
      "grad_norm": 0.3486054539680481,
      "learning_rate": 0.0001,
      "loss": 1.6296,
      "step": 2494
    },
    {
      "epoch": 0.6058766391452162,
      "grad_norm": 0.34594064950942993,
      "learning_rate": 0.0001,
      "loss": 1.5821,
      "step": 2495
    },
    {
      "epoch": 0.6061194754735308,
      "grad_norm": 0.37135642766952515,
      "learning_rate": 0.0001,
      "loss": 1.7481,
      "step": 2496
    },
    {
      "epoch": 0.6063623118018455,
      "grad_norm": 0.3677951693534851,
      "learning_rate": 0.0001,
      "loss": 1.7435,
      "step": 2497
    },
    {
      "epoch": 0.6066051481301603,
      "grad_norm": 0.34797558188438416,
      "learning_rate": 0.0001,
      "loss": 1.6978,
      "step": 2498
    },
    {
      "epoch": 0.606847984458475,
      "grad_norm": 0.37647685408592224,
      "learning_rate": 0.0001,
      "loss": 1.7424,
      "step": 2499
    },
    {
      "epoch": 0.6070908207867897,
      "grad_norm": 0.35586169362068176,
      "learning_rate": 0.0001,
      "loss": 1.6601,
      "step": 2500
    },
    {
      "epoch": 0.6073336571151045,
      "grad_norm": 0.37027135491371155,
      "learning_rate": 0.0001,
      "loss": 1.7841,
      "step": 2501
    },
    {
      "epoch": 0.6075764934434191,
      "grad_norm": 0.3504107892513275,
      "learning_rate": 0.0001,
      "loss": 1.6539,
      "step": 2502
    },
    {
      "epoch": 0.6078193297717338,
      "grad_norm": 0.3658817708492279,
      "learning_rate": 0.0001,
      "loss": 1.6151,
      "step": 2503
    },
    {
      "epoch": 0.6080621661000486,
      "grad_norm": 0.3606242835521698,
      "learning_rate": 0.0001,
      "loss": 1.7228,
      "step": 2504
    },
    {
      "epoch": 0.6083050024283633,
      "grad_norm": 0.3518098294734955,
      "learning_rate": 0.0001,
      "loss": 1.7997,
      "step": 2505
    },
    {
      "epoch": 0.608547838756678,
      "grad_norm": 0.3385467827320099,
      "learning_rate": 0.0001,
      "loss": 1.6134,
      "step": 2506
    },
    {
      "epoch": 0.6087906750849927,
      "grad_norm": 0.35489702224731445,
      "learning_rate": 0.0001,
      "loss": 1.6014,
      "step": 2507
    },
    {
      "epoch": 0.6090335114133074,
      "grad_norm": 0.341196209192276,
      "learning_rate": 0.0001,
      "loss": 1.6125,
      "step": 2508
    },
    {
      "epoch": 0.6092763477416221,
      "grad_norm": 0.3764464557170868,
      "learning_rate": 0.0001,
      "loss": 1.7118,
      "step": 2509
    },
    {
      "epoch": 0.6095191840699369,
      "grad_norm": 0.3787265121936798,
      "learning_rate": 0.0001,
      "loss": 1.8545,
      "step": 2510
    },
    {
      "epoch": 0.6097620203982516,
      "grad_norm": 0.3660854399204254,
      "learning_rate": 0.0001,
      "loss": 1.7869,
      "step": 2511
    },
    {
      "epoch": 0.6100048567265663,
      "grad_norm": 0.32687124609947205,
      "learning_rate": 0.0001,
      "loss": 1.5415,
      "step": 2512
    },
    {
      "epoch": 0.610247693054881,
      "grad_norm": 0.3694194257259369,
      "learning_rate": 0.0001,
      "loss": 1.7856,
      "step": 2513
    },
    {
      "epoch": 0.6104905293831957,
      "grad_norm": 0.37333932518959045,
      "learning_rate": 0.0001,
      "loss": 1.6478,
      "step": 2514
    },
    {
      "epoch": 0.6107333657115105,
      "grad_norm": 0.363366961479187,
      "learning_rate": 0.0001,
      "loss": 1.8028,
      "step": 2515
    },
    {
      "epoch": 0.6109762020398252,
      "grad_norm": 0.3533819615840912,
      "learning_rate": 0.0001,
      "loss": 1.8264,
      "step": 2516
    },
    {
      "epoch": 0.6112190383681398,
      "grad_norm": 0.3544665575027466,
      "learning_rate": 0.0001,
      "loss": 1.6421,
      "step": 2517
    },
    {
      "epoch": 0.6114618746964546,
      "grad_norm": 0.358873575925827,
      "learning_rate": 0.0001,
      "loss": 1.5992,
      "step": 2518
    },
    {
      "epoch": 0.6117047110247693,
      "grad_norm": 0.3534395098686218,
      "learning_rate": 0.0001,
      "loss": 1.5811,
      "step": 2519
    },
    {
      "epoch": 0.611947547353084,
      "grad_norm": 0.3571525812149048,
      "learning_rate": 0.0001,
      "loss": 1.6819,
      "step": 2520
    },
    {
      "epoch": 0.6121903836813988,
      "grad_norm": 0.35596469044685364,
      "learning_rate": 0.0001,
      "loss": 1.7233,
      "step": 2521
    },
    {
      "epoch": 0.6124332200097135,
      "grad_norm": 0.36680924892425537,
      "learning_rate": 0.0001,
      "loss": 1.653,
      "step": 2522
    },
    {
      "epoch": 0.6126760563380281,
      "grad_norm": 0.3670637905597687,
      "learning_rate": 0.0001,
      "loss": 1.7633,
      "step": 2523
    },
    {
      "epoch": 0.6129188926663429,
      "grad_norm": 0.3535623550415039,
      "learning_rate": 0.0001,
      "loss": 1.7664,
      "step": 2524
    },
    {
      "epoch": 0.6131617289946576,
      "grad_norm": 0.36691170930862427,
      "learning_rate": 0.0001,
      "loss": 1.811,
      "step": 2525
    },
    {
      "epoch": 0.6134045653229723,
      "grad_norm": 0.3531697690486908,
      "learning_rate": 0.0001,
      "loss": 1.7234,
      "step": 2526
    },
    {
      "epoch": 0.6136474016512871,
      "grad_norm": 0.3551914691925049,
      "learning_rate": 0.0001,
      "loss": 1.6729,
      "step": 2527
    },
    {
      "epoch": 0.6138902379796017,
      "grad_norm": 0.36252561211586,
      "learning_rate": 0.0001,
      "loss": 1.7676,
      "step": 2528
    },
    {
      "epoch": 0.6141330743079164,
      "grad_norm": 0.3431050181388855,
      "learning_rate": 0.0001,
      "loss": 1.7429,
      "step": 2529
    },
    {
      "epoch": 0.6143759106362312,
      "grad_norm": 0.34209996461868286,
      "learning_rate": 0.0001,
      "loss": 1.6347,
      "step": 2530
    },
    {
      "epoch": 0.6146187469645459,
      "grad_norm": 0.34176865220069885,
      "learning_rate": 0.0001,
      "loss": 1.5931,
      "step": 2531
    },
    {
      "epoch": 0.6148615832928606,
      "grad_norm": 0.3619667887687683,
      "learning_rate": 0.0001,
      "loss": 1.6407,
      "step": 2532
    },
    {
      "epoch": 0.6151044196211753,
      "grad_norm": 0.3568732738494873,
      "learning_rate": 0.0001,
      "loss": 1.5506,
      "step": 2533
    },
    {
      "epoch": 0.61534725594949,
      "grad_norm": 0.3575170040130615,
      "learning_rate": 0.0001,
      "loss": 1.6636,
      "step": 2534
    },
    {
      "epoch": 0.6155900922778048,
      "grad_norm": 0.3435920774936676,
      "learning_rate": 0.0001,
      "loss": 1.6788,
      "step": 2535
    },
    {
      "epoch": 0.6158329286061195,
      "grad_norm": 0.3486897945404053,
      "learning_rate": 0.0001,
      "loss": 1.6519,
      "step": 2536
    },
    {
      "epoch": 0.6160757649344342,
      "grad_norm": 0.36233004927635193,
      "learning_rate": 0.0001,
      "loss": 1.7229,
      "step": 2537
    },
    {
      "epoch": 0.616318601262749,
      "grad_norm": 0.36419281363487244,
      "learning_rate": 0.0001,
      "loss": 1.7536,
      "step": 2538
    },
    {
      "epoch": 0.6165614375910636,
      "grad_norm": 0.36152634024620056,
      "learning_rate": 0.0001,
      "loss": 1.7739,
      "step": 2539
    },
    {
      "epoch": 0.6168042739193783,
      "grad_norm": 0.3492983877658844,
      "learning_rate": 0.0001,
      "loss": 1.6995,
      "step": 2540
    },
    {
      "epoch": 0.6170471102476931,
      "grad_norm": 0.35183224081993103,
      "learning_rate": 0.0001,
      "loss": 1.5924,
      "step": 2541
    },
    {
      "epoch": 0.6172899465760078,
      "grad_norm": 0.3499293029308319,
      "learning_rate": 0.0001,
      "loss": 1.6334,
      "step": 2542
    },
    {
      "epoch": 0.6175327829043225,
      "grad_norm": 0.36725538969039917,
      "learning_rate": 0.0001,
      "loss": 1.8583,
      "step": 2543
    },
    {
      "epoch": 0.6177756192326372,
      "grad_norm": 0.3790731430053711,
      "learning_rate": 0.0001,
      "loss": 1.729,
      "step": 2544
    },
    {
      "epoch": 0.6180184555609519,
      "grad_norm": 0.3718978464603424,
      "learning_rate": 0.0001,
      "loss": 1.7639,
      "step": 2545
    },
    {
      "epoch": 0.6182612918892666,
      "grad_norm": 0.3763367831707001,
      "learning_rate": 0.0001,
      "loss": 1.8659,
      "step": 2546
    },
    {
      "epoch": 0.6185041282175814,
      "grad_norm": 0.36928942799568176,
      "learning_rate": 0.0001,
      "loss": 1.8138,
      "step": 2547
    },
    {
      "epoch": 0.6187469645458961,
      "grad_norm": 0.36250531673431396,
      "learning_rate": 0.0001,
      "loss": 1.623,
      "step": 2548
    },
    {
      "epoch": 0.6189898008742107,
      "grad_norm": 0.36561158299446106,
      "learning_rate": 0.0001,
      "loss": 1.6913,
      "step": 2549
    },
    {
      "epoch": 0.6192326372025255,
      "grad_norm": 0.3352079689502716,
      "learning_rate": 0.0001,
      "loss": 1.5639,
      "step": 2550
    },
    {
      "epoch": 0.6194754735308402,
      "grad_norm": 0.3718976378440857,
      "learning_rate": 0.0001,
      "loss": 1.6785,
      "step": 2551
    },
    {
      "epoch": 0.6197183098591549,
      "grad_norm": 0.33391445875167847,
      "learning_rate": 0.0001,
      "loss": 1.4878,
      "step": 2552
    },
    {
      "epoch": 0.6199611461874697,
      "grad_norm": 0.34767550230026245,
      "learning_rate": 0.0001,
      "loss": 1.6781,
      "step": 2553
    },
    {
      "epoch": 0.6202039825157843,
      "grad_norm": 0.41571366786956787,
      "learning_rate": 0.0001,
      "loss": 2.0179,
      "step": 2554
    },
    {
      "epoch": 0.620446818844099,
      "grad_norm": 0.3613235354423523,
      "learning_rate": 0.0001,
      "loss": 1.8425,
      "step": 2555
    },
    {
      "epoch": 0.6206896551724138,
      "grad_norm": 0.35716602206230164,
      "learning_rate": 0.0001,
      "loss": 1.7749,
      "step": 2556
    },
    {
      "epoch": 0.6209324915007285,
      "grad_norm": 0.3657638728618622,
      "learning_rate": 0.0001,
      "loss": 1.7389,
      "step": 2557
    },
    {
      "epoch": 0.6211753278290433,
      "grad_norm": 0.354874849319458,
      "learning_rate": 0.0001,
      "loss": 1.7393,
      "step": 2558
    },
    {
      "epoch": 0.621418164157358,
      "grad_norm": 0.3623853325843811,
      "learning_rate": 0.0001,
      "loss": 1.6127,
      "step": 2559
    },
    {
      "epoch": 0.6216610004856726,
      "grad_norm": 0.3865921199321747,
      "learning_rate": 0.0001,
      "loss": 1.7271,
      "step": 2560
    },
    {
      "epoch": 0.6219038368139874,
      "grad_norm": 0.38097888231277466,
      "learning_rate": 0.0001,
      "loss": 1.8403,
      "step": 2561
    },
    {
      "epoch": 0.6221466731423021,
      "grad_norm": 0.34053757786750793,
      "learning_rate": 0.0001,
      "loss": 1.4744,
      "step": 2562
    },
    {
      "epoch": 0.6223895094706168,
      "grad_norm": 0.35708364844322205,
      "learning_rate": 0.0001,
      "loss": 1.6543,
      "step": 2563
    },
    {
      "epoch": 0.6226323457989316,
      "grad_norm": 0.3617381155490875,
      "learning_rate": 0.0001,
      "loss": 1.7614,
      "step": 2564
    },
    {
      "epoch": 0.6228751821272462,
      "grad_norm": 0.38946405053138733,
      "learning_rate": 0.0001,
      "loss": 1.6281,
      "step": 2565
    },
    {
      "epoch": 0.6231180184555609,
      "grad_norm": 0.3813820481300354,
      "learning_rate": 0.0001,
      "loss": 1.7442,
      "step": 2566
    },
    {
      "epoch": 0.6233608547838757,
      "grad_norm": 0.3859807848930359,
      "learning_rate": 0.0001,
      "loss": 1.88,
      "step": 2567
    },
    {
      "epoch": 0.6236036911121904,
      "grad_norm": 0.3598704934120178,
      "learning_rate": 0.0001,
      "loss": 1.6405,
      "step": 2568
    },
    {
      "epoch": 0.6238465274405051,
      "grad_norm": 0.37942245602607727,
      "learning_rate": 0.0001,
      "loss": 1.782,
      "step": 2569
    },
    {
      "epoch": 0.6240893637688198,
      "grad_norm": 0.34600645303726196,
      "learning_rate": 0.0001,
      "loss": 1.591,
      "step": 2570
    },
    {
      "epoch": 0.6243322000971345,
      "grad_norm": 0.35516923666000366,
      "learning_rate": 0.0001,
      "loss": 1.6701,
      "step": 2571
    },
    {
      "epoch": 0.6245750364254492,
      "grad_norm": 0.365143746137619,
      "learning_rate": 0.0001,
      "loss": 1.7584,
      "step": 2572
    },
    {
      "epoch": 0.624817872753764,
      "grad_norm": 0.3468589782714844,
      "learning_rate": 0.0001,
      "loss": 1.5642,
      "step": 2573
    },
    {
      "epoch": 0.6250607090820787,
      "grad_norm": 0.3397524952888489,
      "learning_rate": 0.0001,
      "loss": 1.7321,
      "step": 2574
    },
    {
      "epoch": 0.6253035454103933,
      "grad_norm": 0.3729221224784851,
      "learning_rate": 0.0001,
      "loss": 1.7544,
      "step": 2575
    },
    {
      "epoch": 0.6255463817387081,
      "grad_norm": 0.3533952236175537,
      "learning_rate": 0.0001,
      "loss": 1.7231,
      "step": 2576
    },
    {
      "epoch": 0.6257892180670228,
      "grad_norm": 0.38188955187797546,
      "learning_rate": 0.0001,
      "loss": 1.6541,
      "step": 2577
    },
    {
      "epoch": 0.6260320543953375,
      "grad_norm": 0.38412585854530334,
      "learning_rate": 0.0001,
      "loss": 1.7092,
      "step": 2578
    },
    {
      "epoch": 0.6262748907236523,
      "grad_norm": 0.40415629744529724,
      "learning_rate": 0.0001,
      "loss": 1.8571,
      "step": 2579
    },
    {
      "epoch": 0.626517727051967,
      "grad_norm": 0.36079704761505127,
      "learning_rate": 0.0001,
      "loss": 1.6119,
      "step": 2580
    },
    {
      "epoch": 0.6267605633802817,
      "grad_norm": 0.3780444264411926,
      "learning_rate": 0.0001,
      "loss": 1.7924,
      "step": 2581
    },
    {
      "epoch": 0.6270033997085964,
      "grad_norm": 0.3634578287601471,
      "learning_rate": 0.0001,
      "loss": 1.7201,
      "step": 2582
    },
    {
      "epoch": 0.6272462360369111,
      "grad_norm": 0.3554176688194275,
      "learning_rate": 0.0001,
      "loss": 1.6298,
      "step": 2583
    },
    {
      "epoch": 0.6274890723652259,
      "grad_norm": 0.34220534563064575,
      "learning_rate": 0.0001,
      "loss": 1.6022,
      "step": 2584
    },
    {
      "epoch": 0.6277319086935406,
      "grad_norm": 0.38494858145713806,
      "learning_rate": 0.0001,
      "loss": 1.742,
      "step": 2585
    },
    {
      "epoch": 0.6279747450218552,
      "grad_norm": 0.3729349970817566,
      "learning_rate": 0.0001,
      "loss": 1.7127,
      "step": 2586
    },
    {
      "epoch": 0.62821758135017,
      "grad_norm": 0.35282763838768005,
      "learning_rate": 0.0001,
      "loss": 1.6586,
      "step": 2587
    },
    {
      "epoch": 0.6284604176784847,
      "grad_norm": 0.3638066053390503,
      "learning_rate": 0.0001,
      "loss": 1.8,
      "step": 2588
    },
    {
      "epoch": 0.6287032540067994,
      "grad_norm": 0.36281734704971313,
      "learning_rate": 0.0001,
      "loss": 1.7697,
      "step": 2589
    },
    {
      "epoch": 0.6289460903351142,
      "grad_norm": 0.38954979181289673,
      "learning_rate": 0.0001,
      "loss": 1.8074,
      "step": 2590
    },
    {
      "epoch": 0.6291889266634288,
      "grad_norm": 0.3460639715194702,
      "learning_rate": 0.0001,
      "loss": 1.7445,
      "step": 2591
    },
    {
      "epoch": 0.6294317629917435,
      "grad_norm": 0.3687112033367157,
      "learning_rate": 0.0001,
      "loss": 1.7311,
      "step": 2592
    },
    {
      "epoch": 0.6296745993200583,
      "grad_norm": 0.3805738091468811,
      "learning_rate": 0.0001,
      "loss": 1.8862,
      "step": 2593
    },
    {
      "epoch": 0.629917435648373,
      "grad_norm": 0.346627801656723,
      "learning_rate": 0.0001,
      "loss": 1.6794,
      "step": 2594
    },
    {
      "epoch": 0.6301602719766877,
      "grad_norm": 0.34023693203926086,
      "learning_rate": 0.0001,
      "loss": 1.5122,
      "step": 2595
    },
    {
      "epoch": 0.6304031083050025,
      "grad_norm": 0.373219758272171,
      "learning_rate": 0.0001,
      "loss": 1.6898,
      "step": 2596
    },
    {
      "epoch": 0.6306459446333171,
      "grad_norm": 0.37325000762939453,
      "learning_rate": 0.0001,
      "loss": 1.789,
      "step": 2597
    },
    {
      "epoch": 0.6308887809616318,
      "grad_norm": 0.38243454694747925,
      "learning_rate": 0.0001,
      "loss": 1.8489,
      "step": 2598
    },
    {
      "epoch": 0.6311316172899466,
      "grad_norm": 0.35410985350608826,
      "learning_rate": 0.0001,
      "loss": 1.6322,
      "step": 2599
    },
    {
      "epoch": 0.6313744536182613,
      "grad_norm": 0.3585367798805237,
      "learning_rate": 0.0001,
      "loss": 1.805,
      "step": 2600
    },
    {
      "epoch": 0.631617289946576,
      "grad_norm": 0.37655138969421387,
      "learning_rate": 0.0001,
      "loss": 1.7564,
      "step": 2601
    },
    {
      "epoch": 0.6318601262748907,
      "grad_norm": 0.3572215139865875,
      "learning_rate": 0.0001,
      "loss": 1.7444,
      "step": 2602
    },
    {
      "epoch": 0.6321029626032054,
      "grad_norm": 0.39637553691864014,
      "learning_rate": 0.0001,
      "loss": 1.8615,
      "step": 2603
    },
    {
      "epoch": 0.6323457989315202,
      "grad_norm": 0.35991159081459045,
      "learning_rate": 0.0001,
      "loss": 1.7543,
      "step": 2604
    },
    {
      "epoch": 0.6325886352598349,
      "grad_norm": 0.32327163219451904,
      "learning_rate": 0.0001,
      "loss": 1.3716,
      "step": 2605
    },
    {
      "epoch": 0.6328314715881496,
      "grad_norm": 0.3595682978630066,
      "learning_rate": 0.0001,
      "loss": 1.8769,
      "step": 2606
    },
    {
      "epoch": 0.6330743079164644,
      "grad_norm": 0.3780916929244995,
      "learning_rate": 0.0001,
      "loss": 1.7702,
      "step": 2607
    },
    {
      "epoch": 0.633317144244779,
      "grad_norm": 0.3648173213005066,
      "learning_rate": 0.0001,
      "loss": 1.774,
      "step": 2608
    },
    {
      "epoch": 0.6335599805730937,
      "grad_norm": 0.3711946904659271,
      "learning_rate": 0.0001,
      "loss": 1.7383,
      "step": 2609
    },
    {
      "epoch": 0.6338028169014085,
      "grad_norm": 0.3568876385688782,
      "learning_rate": 0.0001,
      "loss": 1.6766,
      "step": 2610
    },
    {
      "epoch": 0.6340456532297232,
      "grad_norm": 0.3626377284526825,
      "learning_rate": 0.0001,
      "loss": 1.6791,
      "step": 2611
    },
    {
      "epoch": 0.6342884895580378,
      "grad_norm": 0.340934157371521,
      "learning_rate": 0.0001,
      "loss": 1.6389,
      "step": 2612
    },
    {
      "epoch": 0.6345313258863526,
      "grad_norm": 0.3646719455718994,
      "learning_rate": 0.0001,
      "loss": 1.8103,
      "step": 2613
    },
    {
      "epoch": 0.6347741622146673,
      "grad_norm": 0.36216479539871216,
      "learning_rate": 0.0001,
      "loss": 1.8449,
      "step": 2614
    },
    {
      "epoch": 0.635016998542982,
      "grad_norm": 0.36999836564064026,
      "learning_rate": 0.0001,
      "loss": 1.8061,
      "step": 2615
    },
    {
      "epoch": 0.6352598348712968,
      "grad_norm": 0.37942442297935486,
      "learning_rate": 0.0001,
      "loss": 1.7946,
      "step": 2616
    },
    {
      "epoch": 0.6355026711996115,
      "grad_norm": 0.3583141565322876,
      "learning_rate": 0.0001,
      "loss": 1.6131,
      "step": 2617
    },
    {
      "epoch": 0.6357455075279261,
      "grad_norm": 0.35164546966552734,
      "learning_rate": 0.0001,
      "loss": 1.771,
      "step": 2618
    },
    {
      "epoch": 0.6359883438562409,
      "grad_norm": 0.35949042439460754,
      "learning_rate": 0.0001,
      "loss": 1.6526,
      "step": 2619
    },
    {
      "epoch": 0.6362311801845556,
      "grad_norm": 0.38381242752075195,
      "learning_rate": 0.0001,
      "loss": 1.6743,
      "step": 2620
    },
    {
      "epoch": 0.6364740165128703,
      "grad_norm": 0.3700384497642517,
      "learning_rate": 0.0001,
      "loss": 1.721,
      "step": 2621
    },
    {
      "epoch": 0.6367168528411851,
      "grad_norm": 0.3850778043270111,
      "learning_rate": 0.0001,
      "loss": 1.9642,
      "step": 2622
    },
    {
      "epoch": 0.6369596891694997,
      "grad_norm": 0.3443865180015564,
      "learning_rate": 0.0001,
      "loss": 1.6247,
      "step": 2623
    },
    {
      "epoch": 0.6372025254978144,
      "grad_norm": 0.39186325669288635,
      "learning_rate": 0.0001,
      "loss": 1.8336,
      "step": 2624
    },
    {
      "epoch": 0.6374453618261292,
      "grad_norm": 0.404778391122818,
      "learning_rate": 0.0001,
      "loss": 1.622,
      "step": 2625
    },
    {
      "epoch": 0.6376881981544439,
      "grad_norm": 0.35440367460250854,
      "learning_rate": 0.0001,
      "loss": 1.8359,
      "step": 2626
    },
    {
      "epoch": 0.6379310344827587,
      "grad_norm": 0.387531042098999,
      "learning_rate": 0.0001,
      "loss": 1.7942,
      "step": 2627
    },
    {
      "epoch": 0.6381738708110734,
      "grad_norm": 0.35026273131370544,
      "learning_rate": 0.0001,
      "loss": 1.8036,
      "step": 2628
    },
    {
      "epoch": 0.638416707139388,
      "grad_norm": 0.35991358757019043,
      "learning_rate": 0.0001,
      "loss": 1.6788,
      "step": 2629
    },
    {
      "epoch": 0.6386595434677028,
      "grad_norm": 0.3719328045845032,
      "learning_rate": 0.0001,
      "loss": 1.7535,
      "step": 2630
    },
    {
      "epoch": 0.6389023797960175,
      "grad_norm": 0.38722363114356995,
      "learning_rate": 0.0001,
      "loss": 1.741,
      "step": 2631
    },
    {
      "epoch": 0.6391452161243322,
      "grad_norm": 0.3769434690475464,
      "learning_rate": 0.0001,
      "loss": 1.7619,
      "step": 2632
    },
    {
      "epoch": 0.639388052452647,
      "grad_norm": 0.3432542085647583,
      "learning_rate": 0.0001,
      "loss": 1.7678,
      "step": 2633
    },
    {
      "epoch": 0.6396308887809616,
      "grad_norm": 0.36289897561073303,
      "learning_rate": 0.0001,
      "loss": 1.5774,
      "step": 2634
    },
    {
      "epoch": 0.6398737251092763,
      "grad_norm": 0.37560367584228516,
      "learning_rate": 0.0001,
      "loss": 1.8036,
      "step": 2635
    },
    {
      "epoch": 0.6401165614375911,
      "grad_norm": 0.33592575788497925,
      "learning_rate": 0.0001,
      "loss": 1.6657,
      "step": 2636
    },
    {
      "epoch": 0.6403593977659058,
      "grad_norm": 0.34603649377822876,
      "learning_rate": 0.0001,
      "loss": 1.6675,
      "step": 2637
    },
    {
      "epoch": 0.6406022340942205,
      "grad_norm": 0.35474568605422974,
      "learning_rate": 0.0001,
      "loss": 1.5955,
      "step": 2638
    },
    {
      "epoch": 0.6408450704225352,
      "grad_norm": 0.37561899423599243,
      "learning_rate": 0.0001,
      "loss": 1.79,
      "step": 2639
    },
    {
      "epoch": 0.6410879067508499,
      "grad_norm": 0.3733697533607483,
      "learning_rate": 0.0001,
      "loss": 1.6762,
      "step": 2640
    },
    {
      "epoch": 0.6413307430791646,
      "grad_norm": 0.3769213557243347,
      "learning_rate": 0.0001,
      "loss": 1.74,
      "step": 2641
    },
    {
      "epoch": 0.6415735794074794,
      "grad_norm": 0.3656337857246399,
      "learning_rate": 0.0001,
      "loss": 1.7458,
      "step": 2642
    },
    {
      "epoch": 0.6418164157357941,
      "grad_norm": 0.358331561088562,
      "learning_rate": 0.0001,
      "loss": 1.7657,
      "step": 2643
    },
    {
      "epoch": 0.6420592520641087,
      "grad_norm": 0.36741209030151367,
      "learning_rate": 0.0001,
      "loss": 1.676,
      "step": 2644
    },
    {
      "epoch": 0.6423020883924235,
      "grad_norm": 0.35470911860466003,
      "learning_rate": 0.0001,
      "loss": 1.7777,
      "step": 2645
    },
    {
      "epoch": 0.6425449247207382,
      "grad_norm": 0.3638315200805664,
      "learning_rate": 0.0001,
      "loss": 1.8138,
      "step": 2646
    },
    {
      "epoch": 0.6427877610490529,
      "grad_norm": 0.36279577016830444,
      "learning_rate": 0.0001,
      "loss": 1.6799,
      "step": 2647
    },
    {
      "epoch": 0.6430305973773677,
      "grad_norm": 0.38609495759010315,
      "learning_rate": 0.0001,
      "loss": 1.8977,
      "step": 2648
    },
    {
      "epoch": 0.6432734337056824,
      "grad_norm": 0.379506379365921,
      "learning_rate": 0.0001,
      "loss": 1.7512,
      "step": 2649
    },
    {
      "epoch": 0.6435162700339971,
      "grad_norm": 0.3665315508842468,
      "learning_rate": 0.0001,
      "loss": 1.8016,
      "step": 2650
    },
    {
      "epoch": 0.6437591063623118,
      "grad_norm": 0.36043789982795715,
      "learning_rate": 0.0001,
      "loss": 1.7292,
      "step": 2651
    },
    {
      "epoch": 0.6440019426906265,
      "grad_norm": 0.3489297330379486,
      "learning_rate": 0.0001,
      "loss": 1.6868,
      "step": 2652
    },
    {
      "epoch": 0.6442447790189413,
      "grad_norm": 0.37208572030067444,
      "learning_rate": 0.0001,
      "loss": 1.8481,
      "step": 2653
    },
    {
      "epoch": 0.644487615347256,
      "grad_norm": 0.35702675580978394,
      "learning_rate": 0.0001,
      "loss": 1.7971,
      "step": 2654
    },
    {
      "epoch": 0.6447304516755706,
      "grad_norm": 0.36701202392578125,
      "learning_rate": 0.0001,
      "loss": 1.7149,
      "step": 2655
    },
    {
      "epoch": 0.6449732880038854,
      "grad_norm": 0.35397055745124817,
      "learning_rate": 0.0001,
      "loss": 1.615,
      "step": 2656
    },
    {
      "epoch": 0.6452161243322001,
      "grad_norm": 0.3622033894062042,
      "learning_rate": 0.0001,
      "loss": 1.7409,
      "step": 2657
    },
    {
      "epoch": 0.6454589606605148,
      "grad_norm": 0.35813143849372864,
      "learning_rate": 0.0001,
      "loss": 1.5353,
      "step": 2658
    },
    {
      "epoch": 0.6457017969888296,
      "grad_norm": 0.352824330329895,
      "learning_rate": 0.0001,
      "loss": 1.6359,
      "step": 2659
    },
    {
      "epoch": 0.6459446333171442,
      "grad_norm": 0.37167853116989136,
      "learning_rate": 0.0001,
      "loss": 1.7035,
      "step": 2660
    },
    {
      "epoch": 0.6461874696454589,
      "grad_norm": 0.360668808221817,
      "learning_rate": 0.0001,
      "loss": 1.7314,
      "step": 2661
    },
    {
      "epoch": 0.6464303059737737,
      "grad_norm": 0.36218956112861633,
      "learning_rate": 0.0001,
      "loss": 1.7743,
      "step": 2662
    },
    {
      "epoch": 0.6466731423020884,
      "grad_norm": 0.3747027814388275,
      "learning_rate": 0.0001,
      "loss": 1.7106,
      "step": 2663
    },
    {
      "epoch": 0.6469159786304031,
      "grad_norm": 0.33726468682289124,
      "learning_rate": 0.0001,
      "loss": 1.5821,
      "step": 2664
    },
    {
      "epoch": 0.6471588149587179,
      "grad_norm": 0.3692800998687744,
      "learning_rate": 0.0001,
      "loss": 1.7861,
      "step": 2665
    },
    {
      "epoch": 0.6474016512870325,
      "grad_norm": 0.35162967443466187,
      "learning_rate": 0.0001,
      "loss": 1.7523,
      "step": 2666
    },
    {
      "epoch": 0.6476444876153472,
      "grad_norm": 0.3360744118690491,
      "learning_rate": 0.0001,
      "loss": 1.6948,
      "step": 2667
    },
    {
      "epoch": 0.647887323943662,
      "grad_norm": 0.3539160192012787,
      "learning_rate": 0.0001,
      "loss": 1.6048,
      "step": 2668
    },
    {
      "epoch": 0.6481301602719767,
      "grad_norm": 0.3462900221347809,
      "learning_rate": 0.0001,
      "loss": 1.7686,
      "step": 2669
    },
    {
      "epoch": 0.6483729966002914,
      "grad_norm": 0.35002318024635315,
      "learning_rate": 0.0001,
      "loss": 1.7655,
      "step": 2670
    },
    {
      "epoch": 0.6486158329286061,
      "grad_norm": 0.3510165810585022,
      "learning_rate": 0.0001,
      "loss": 1.6869,
      "step": 2671
    },
    {
      "epoch": 0.6488586692569208,
      "grad_norm": 0.3706774115562439,
      "learning_rate": 0.0001,
      "loss": 1.8192,
      "step": 2672
    },
    {
      "epoch": 0.6491015055852356,
      "grad_norm": 0.35293927788734436,
      "learning_rate": 0.0001,
      "loss": 1.7139,
      "step": 2673
    },
    {
      "epoch": 0.6493443419135503,
      "grad_norm": 0.35285595059394836,
      "learning_rate": 0.0001,
      "loss": 1.7176,
      "step": 2674
    },
    {
      "epoch": 0.649587178241865,
      "grad_norm": 0.3747918903827667,
      "learning_rate": 0.0001,
      "loss": 1.8401,
      "step": 2675
    },
    {
      "epoch": 0.6498300145701797,
      "grad_norm": 0.36907392740249634,
      "learning_rate": 0.0001,
      "loss": 1.8169,
      "step": 2676
    },
    {
      "epoch": 0.6500728508984944,
      "grad_norm": 0.37845489382743835,
      "learning_rate": 0.0001,
      "loss": 1.832,
      "step": 2677
    },
    {
      "epoch": 0.6503156872268091,
      "grad_norm": 0.3632284998893738,
      "learning_rate": 0.0001,
      "loss": 1.7413,
      "step": 2678
    },
    {
      "epoch": 0.6505585235551239,
      "grad_norm": 0.36341753602027893,
      "learning_rate": 0.0001,
      "loss": 1.7788,
      "step": 2679
    },
    {
      "epoch": 0.6508013598834386,
      "grad_norm": 0.3553708791732788,
      "learning_rate": 0.0001,
      "loss": 1.6001,
      "step": 2680
    },
    {
      "epoch": 0.6510441962117532,
      "grad_norm": 0.3506467640399933,
      "learning_rate": 0.0001,
      "loss": 1.656,
      "step": 2681
    },
    {
      "epoch": 0.651287032540068,
      "grad_norm": 0.3974582850933075,
      "learning_rate": 0.0001,
      "loss": 1.958,
      "step": 2682
    },
    {
      "epoch": 0.6515298688683827,
      "grad_norm": 0.3517482876777649,
      "learning_rate": 0.0001,
      "loss": 1.6594,
      "step": 2683
    },
    {
      "epoch": 0.6517727051966974,
      "grad_norm": 0.3600420653820038,
      "learning_rate": 0.0001,
      "loss": 1.4849,
      "step": 2684
    },
    {
      "epoch": 0.6520155415250122,
      "grad_norm": 0.36825674772262573,
      "learning_rate": 0.0001,
      "loss": 1.8789,
      "step": 2685
    },
    {
      "epoch": 0.6522583778533269,
      "grad_norm": 0.37107062339782715,
      "learning_rate": 0.0001,
      "loss": 1.7713,
      "step": 2686
    },
    {
      "epoch": 0.6525012141816415,
      "grad_norm": 0.3648766875267029,
      "learning_rate": 0.0001,
      "loss": 1.7974,
      "step": 2687
    },
    {
      "epoch": 0.6527440505099563,
      "grad_norm": 0.34810009598731995,
      "learning_rate": 0.0001,
      "loss": 1.693,
      "step": 2688
    },
    {
      "epoch": 0.652986886838271,
      "grad_norm": 0.3525703549385071,
      "learning_rate": 0.0001,
      "loss": 1.641,
      "step": 2689
    },
    {
      "epoch": 0.6532297231665857,
      "grad_norm": 0.39204278588294983,
      "learning_rate": 0.0001,
      "loss": 1.8355,
      "step": 2690
    },
    {
      "epoch": 0.6534725594949005,
      "grad_norm": 0.3839750289916992,
      "learning_rate": 0.0001,
      "loss": 1.7873,
      "step": 2691
    },
    {
      "epoch": 0.6537153958232151,
      "grad_norm": 0.3619327247142792,
      "learning_rate": 0.0001,
      "loss": 1.7816,
      "step": 2692
    },
    {
      "epoch": 0.6539582321515298,
      "grad_norm": 0.34967416524887085,
      "learning_rate": 0.0001,
      "loss": 1.5844,
      "step": 2693
    },
    {
      "epoch": 0.6542010684798446,
      "grad_norm": 0.35875624418258667,
      "learning_rate": 0.0001,
      "loss": 1.6636,
      "step": 2694
    },
    {
      "epoch": 0.6544439048081593,
      "grad_norm": 0.35244202613830566,
      "learning_rate": 0.0001,
      "loss": 1.5934,
      "step": 2695
    },
    {
      "epoch": 0.6546867411364741,
      "grad_norm": 0.37988707423210144,
      "learning_rate": 0.0001,
      "loss": 1.6672,
      "step": 2696
    },
    {
      "epoch": 0.6549295774647887,
      "grad_norm": 0.3807011544704437,
      "learning_rate": 0.0001,
      "loss": 1.7633,
      "step": 2697
    },
    {
      "epoch": 0.6551724137931034,
      "grad_norm": 0.3748406171798706,
      "learning_rate": 0.0001,
      "loss": 1.7146,
      "step": 2698
    },
    {
      "epoch": 0.6554152501214182,
      "grad_norm": 0.3780806362628937,
      "learning_rate": 0.0001,
      "loss": 1.786,
      "step": 2699
    },
    {
      "epoch": 0.6556580864497329,
      "grad_norm": 0.37719249725341797,
      "learning_rate": 0.0001,
      "loss": 1.7855,
      "step": 2700
    },
    {
      "epoch": 0.6559009227780476,
      "grad_norm": 0.36691877245903015,
      "learning_rate": 0.0001,
      "loss": 1.6269,
      "step": 2701
    },
    {
      "epoch": 0.6561437591063624,
      "grad_norm": 0.3625624477863312,
      "learning_rate": 0.0001,
      "loss": 1.5904,
      "step": 2702
    },
    {
      "epoch": 0.656386595434677,
      "grad_norm": 0.3556610345840454,
      "learning_rate": 0.0001,
      "loss": 1.9223,
      "step": 2703
    },
    {
      "epoch": 0.6566294317629917,
      "grad_norm": 0.3541751801967621,
      "learning_rate": 0.0001,
      "loss": 1.7873,
      "step": 2704
    },
    {
      "epoch": 0.6568722680913065,
      "grad_norm": 0.39495357871055603,
      "learning_rate": 0.0001,
      "loss": 1.8065,
      "step": 2705
    },
    {
      "epoch": 0.6571151044196212,
      "grad_norm": 0.362796425819397,
      "learning_rate": 0.0001,
      "loss": 1.6427,
      "step": 2706
    },
    {
      "epoch": 0.6573579407479359,
      "grad_norm": 0.35276418924331665,
      "learning_rate": 0.0001,
      "loss": 1.7092,
      "step": 2707
    },
    {
      "epoch": 0.6576007770762506,
      "grad_norm": 0.36206701397895813,
      "learning_rate": 0.0001,
      "loss": 1.6966,
      "step": 2708
    },
    {
      "epoch": 0.6578436134045653,
      "grad_norm": 0.36278581619262695,
      "learning_rate": 0.0001,
      "loss": 1.7378,
      "step": 2709
    },
    {
      "epoch": 0.65808644973288,
      "grad_norm": 0.38123491406440735,
      "learning_rate": 0.0001,
      "loss": 1.7312,
      "step": 2710
    },
    {
      "epoch": 0.6583292860611948,
      "grad_norm": 0.3612028658390045,
      "learning_rate": 0.0001,
      "loss": 1.8957,
      "step": 2711
    },
    {
      "epoch": 0.6585721223895095,
      "grad_norm": 0.3690524101257324,
      "learning_rate": 0.0001,
      "loss": 1.712,
      "step": 2712
    },
    {
      "epoch": 0.6588149587178241,
      "grad_norm": 0.35725656151771545,
      "learning_rate": 0.0001,
      "loss": 1.5862,
      "step": 2713
    },
    {
      "epoch": 0.6590577950461389,
      "grad_norm": 0.3594600260257721,
      "learning_rate": 0.0001,
      "loss": 1.6163,
      "step": 2714
    },
    {
      "epoch": 0.6593006313744536,
      "grad_norm": 0.34838706254959106,
      "learning_rate": 0.0001,
      "loss": 1.6886,
      "step": 2715
    },
    {
      "epoch": 0.6595434677027683,
      "grad_norm": 0.34556108713150024,
      "learning_rate": 0.0001,
      "loss": 1.6638,
      "step": 2716
    },
    {
      "epoch": 0.6597863040310831,
      "grad_norm": 0.34152528643608093,
      "learning_rate": 0.0001,
      "loss": 1.6916,
      "step": 2717
    },
    {
      "epoch": 0.6600291403593977,
      "grad_norm": 0.36784660816192627,
      "learning_rate": 0.0001,
      "loss": 1.8092,
      "step": 2718
    },
    {
      "epoch": 0.6602719766877125,
      "grad_norm": 0.35600847005844116,
      "learning_rate": 0.0001,
      "loss": 1.7819,
      "step": 2719
    },
    {
      "epoch": 0.6605148130160272,
      "grad_norm": 0.36031457781791687,
      "learning_rate": 0.0001,
      "loss": 1.6415,
      "step": 2720
    },
    {
      "epoch": 0.6607576493443419,
      "grad_norm": 0.3358955681324005,
      "learning_rate": 0.0001,
      "loss": 1.6377,
      "step": 2721
    },
    {
      "epoch": 0.6610004856726567,
      "grad_norm": 0.39089998602867126,
      "learning_rate": 0.0001,
      "loss": 1.8941,
      "step": 2722
    },
    {
      "epoch": 0.6612433220009714,
      "grad_norm": 0.3684552311897278,
      "learning_rate": 0.0001,
      "loss": 1.5994,
      "step": 2723
    },
    {
      "epoch": 0.661486158329286,
      "grad_norm": 0.38114041090011597,
      "learning_rate": 0.0001,
      "loss": 1.7248,
      "step": 2724
    },
    {
      "epoch": 0.6617289946576008,
      "grad_norm": 0.3674573600292206,
      "learning_rate": 0.0001,
      "loss": 1.653,
      "step": 2725
    },
    {
      "epoch": 0.6619718309859155,
      "grad_norm": 0.35975441336631775,
      "learning_rate": 0.0001,
      "loss": 1.6908,
      "step": 2726
    },
    {
      "epoch": 0.6622146673142302,
      "grad_norm": 0.37502220273017883,
      "learning_rate": 0.0001,
      "loss": 1.8135,
      "step": 2727
    },
    {
      "epoch": 0.662457503642545,
      "grad_norm": 0.3751458525657654,
      "learning_rate": 0.0001,
      "loss": 1.4792,
      "step": 2728
    },
    {
      "epoch": 0.6627003399708596,
      "grad_norm": 0.376924991607666,
      "learning_rate": 0.0001,
      "loss": 1.7133,
      "step": 2729
    },
    {
      "epoch": 0.6629431762991743,
      "grad_norm": 0.3626424968242645,
      "learning_rate": 0.0001,
      "loss": 1.7239,
      "step": 2730
    },
    {
      "epoch": 0.6631860126274891,
      "grad_norm": 0.4036501348018646,
      "learning_rate": 0.0001,
      "loss": 1.7198,
      "step": 2731
    },
    {
      "epoch": 0.6634288489558038,
      "grad_norm": 0.38245660066604614,
      "learning_rate": 0.0001,
      "loss": 1.6446,
      "step": 2732
    },
    {
      "epoch": 0.6636716852841185,
      "grad_norm": 0.3560940623283386,
      "learning_rate": 0.0001,
      "loss": 1.6699,
      "step": 2733
    },
    {
      "epoch": 0.6639145216124333,
      "grad_norm": 0.37661460041999817,
      "learning_rate": 0.0001,
      "loss": 1.8264,
      "step": 2734
    },
    {
      "epoch": 0.6641573579407479,
      "grad_norm": 0.38734447956085205,
      "learning_rate": 0.0001,
      "loss": 1.6237,
      "step": 2735
    },
    {
      "epoch": 0.6644001942690626,
      "grad_norm": 0.37723541259765625,
      "learning_rate": 0.0001,
      "loss": 1.625,
      "step": 2736
    },
    {
      "epoch": 0.6646430305973774,
      "grad_norm": 0.358579158782959,
      "learning_rate": 0.0001,
      "loss": 1.8073,
      "step": 2737
    },
    {
      "epoch": 0.6648858669256921,
      "grad_norm": 0.3974575400352478,
      "learning_rate": 0.0001,
      "loss": 1.8406,
      "step": 2738
    },
    {
      "epoch": 0.6651287032540067,
      "grad_norm": 0.39381760358810425,
      "learning_rate": 0.0001,
      "loss": 1.7404,
      "step": 2739
    },
    {
      "epoch": 0.6653715395823215,
      "grad_norm": 0.38782989978790283,
      "learning_rate": 0.0001,
      "loss": 1.6999,
      "step": 2740
    },
    {
      "epoch": 0.6656143759106362,
      "grad_norm": 0.38447389006614685,
      "learning_rate": 0.0001,
      "loss": 1.8801,
      "step": 2741
    },
    {
      "epoch": 0.665857212238951,
      "grad_norm": 0.3697817921638489,
      "learning_rate": 0.0001,
      "loss": 1.7613,
      "step": 2742
    },
    {
      "epoch": 0.6661000485672657,
      "grad_norm": 0.3970728814601898,
      "learning_rate": 0.0001,
      "loss": 1.6759,
      "step": 2743
    },
    {
      "epoch": 0.6663428848955804,
      "grad_norm": 0.3800773620605469,
      "learning_rate": 0.0001,
      "loss": 1.6984,
      "step": 2744
    },
    {
      "epoch": 0.6665857212238951,
      "grad_norm": 0.35148799419403076,
      "learning_rate": 0.0001,
      "loss": 1.6721,
      "step": 2745
    },
    {
      "epoch": 0.6668285575522098,
      "grad_norm": 0.3488072156906128,
      "learning_rate": 0.0001,
      "loss": 1.8215,
      "step": 2746
    },
    {
      "epoch": 0.6670713938805245,
      "grad_norm": 0.39117431640625,
      "learning_rate": 0.0001,
      "loss": 1.6453,
      "step": 2747
    },
    {
      "epoch": 0.6673142302088393,
      "grad_norm": 0.3691113591194153,
      "learning_rate": 0.0001,
      "loss": 1.709,
      "step": 2748
    },
    {
      "epoch": 0.667557066537154,
      "grad_norm": 0.3842622935771942,
      "learning_rate": 0.0001,
      "loss": 1.837,
      "step": 2749
    },
    {
      "epoch": 0.6677999028654686,
      "grad_norm": 0.36402004957199097,
      "learning_rate": 0.0001,
      "loss": 1.7675,
      "step": 2750
    },
    {
      "epoch": 0.6680427391937834,
      "grad_norm": 0.35241395235061646,
      "learning_rate": 0.0001,
      "loss": 1.71,
      "step": 2751
    },
    {
      "epoch": 0.6682855755220981,
      "grad_norm": 0.35116317868232727,
      "learning_rate": 0.0001,
      "loss": 1.6839,
      "step": 2752
    },
    {
      "epoch": 0.6685284118504128,
      "grad_norm": 0.3499755561351776,
      "learning_rate": 0.0001,
      "loss": 1.761,
      "step": 2753
    },
    {
      "epoch": 0.6687712481787276,
      "grad_norm": 0.3414444625377655,
      "learning_rate": 0.0001,
      "loss": 1.6769,
      "step": 2754
    },
    {
      "epoch": 0.6690140845070423,
      "grad_norm": 0.3383598029613495,
      "learning_rate": 0.0001,
      "loss": 1.3583,
      "step": 2755
    },
    {
      "epoch": 0.6692569208353569,
      "grad_norm": 0.3479803204536438,
      "learning_rate": 0.0001,
      "loss": 1.778,
      "step": 2756
    },
    {
      "epoch": 0.6694997571636717,
      "grad_norm": 0.3396441340446472,
      "learning_rate": 0.0001,
      "loss": 1.6624,
      "step": 2757
    },
    {
      "epoch": 0.6697425934919864,
      "grad_norm": 0.36957108974456787,
      "learning_rate": 0.0001,
      "loss": 1.7515,
      "step": 2758
    },
    {
      "epoch": 0.6699854298203011,
      "grad_norm": 0.37735646963119507,
      "learning_rate": 0.0001,
      "loss": 1.8101,
      "step": 2759
    },
    {
      "epoch": 0.6702282661486159,
      "grad_norm": 0.37875017523765564,
      "learning_rate": 0.0001,
      "loss": 1.9074,
      "step": 2760
    },
    {
      "epoch": 0.6704711024769305,
      "grad_norm": 0.37791571021080017,
      "learning_rate": 0.0001,
      "loss": 1.9334,
      "step": 2761
    },
    {
      "epoch": 0.6707139388052452,
      "grad_norm": 0.36337870359420776,
      "learning_rate": 0.0001,
      "loss": 1.7473,
      "step": 2762
    },
    {
      "epoch": 0.67095677513356,
      "grad_norm": 0.4029197096824646,
      "learning_rate": 0.0001,
      "loss": 1.8366,
      "step": 2763
    },
    {
      "epoch": 0.6711996114618747,
      "grad_norm": 0.3762977421283722,
      "learning_rate": 0.0001,
      "loss": 1.6142,
      "step": 2764
    },
    {
      "epoch": 0.6714424477901895,
      "grad_norm": 0.3758258819580078,
      "learning_rate": 0.0001,
      "loss": 1.8017,
      "step": 2765
    },
    {
      "epoch": 0.6716852841185041,
      "grad_norm": 0.3770975172519684,
      "learning_rate": 0.0001,
      "loss": 1.8988,
      "step": 2766
    },
    {
      "epoch": 0.6719281204468188,
      "grad_norm": 0.35840946435928345,
      "learning_rate": 0.0001,
      "loss": 1.6796,
      "step": 2767
    },
    {
      "epoch": 0.6721709567751336,
      "grad_norm": 0.3695116937160492,
      "learning_rate": 0.0001,
      "loss": 1.6958,
      "step": 2768
    },
    {
      "epoch": 0.6724137931034483,
      "grad_norm": 0.37600207328796387,
      "learning_rate": 0.0001,
      "loss": 1.7066,
      "step": 2769
    },
    {
      "epoch": 0.672656629431763,
      "grad_norm": 0.3682120740413666,
      "learning_rate": 0.0001,
      "loss": 1.8241,
      "step": 2770
    },
    {
      "epoch": 0.6728994657600778,
      "grad_norm": 0.3645362854003906,
      "learning_rate": 0.0001,
      "loss": 1.5875,
      "step": 2771
    },
    {
      "epoch": 0.6731423020883924,
      "grad_norm": 0.39225202798843384,
      "learning_rate": 0.0001,
      "loss": 1.7588,
      "step": 2772
    },
    {
      "epoch": 0.6733851384167071,
      "grad_norm": 0.3571353852748871,
      "learning_rate": 0.0001,
      "loss": 1.6266,
      "step": 2773
    },
    {
      "epoch": 0.6736279747450219,
      "grad_norm": 0.3667353689670563,
      "learning_rate": 0.0001,
      "loss": 1.7951,
      "step": 2774
    },
    {
      "epoch": 0.6738708110733366,
      "grad_norm": 0.3817640542984009,
      "learning_rate": 0.0001,
      "loss": 1.8936,
      "step": 2775
    },
    {
      "epoch": 0.6741136474016513,
      "grad_norm": 0.35165515542030334,
      "learning_rate": 0.0001,
      "loss": 1.8528,
      "step": 2776
    },
    {
      "epoch": 0.674356483729966,
      "grad_norm": 0.38937145471572876,
      "learning_rate": 0.0001,
      "loss": 1.7106,
      "step": 2777
    },
    {
      "epoch": 0.6745993200582807,
      "grad_norm": 0.3663376271724701,
      "learning_rate": 0.0001,
      "loss": 1.7573,
      "step": 2778
    },
    {
      "epoch": 0.6748421563865954,
      "grad_norm": 0.3608282506465912,
      "learning_rate": 0.0001,
      "loss": 1.6956,
      "step": 2779
    },
    {
      "epoch": 0.6750849927149102,
      "grad_norm": 0.3559147119522095,
      "learning_rate": 0.0001,
      "loss": 1.6093,
      "step": 2780
    },
    {
      "epoch": 0.6753278290432249,
      "grad_norm": 0.3778630197048187,
      "learning_rate": 0.0001,
      "loss": 1.8678,
      "step": 2781
    },
    {
      "epoch": 0.6755706653715395,
      "grad_norm": 0.35783466696739197,
      "learning_rate": 0.0001,
      "loss": 1.6642,
      "step": 2782
    },
    {
      "epoch": 0.6758135016998543,
      "grad_norm": 0.3654789924621582,
      "learning_rate": 0.0001,
      "loss": 1.7328,
      "step": 2783
    },
    {
      "epoch": 0.676056338028169,
      "grad_norm": 0.35034000873565674,
      "learning_rate": 0.0001,
      "loss": 1.4726,
      "step": 2784
    },
    {
      "epoch": 0.6762991743564837,
      "grad_norm": 0.35254645347595215,
      "learning_rate": 0.0001,
      "loss": 1.6496,
      "step": 2785
    },
    {
      "epoch": 0.6765420106847985,
      "grad_norm": 0.35973024368286133,
      "learning_rate": 0.0001,
      "loss": 1.6284,
      "step": 2786
    },
    {
      "epoch": 0.6767848470131131,
      "grad_norm": 0.35698050260543823,
      "learning_rate": 0.0001,
      "loss": 1.7358,
      "step": 2787
    },
    {
      "epoch": 0.6770276833414279,
      "grad_norm": 0.38141682744026184,
      "learning_rate": 0.0001,
      "loss": 1.8136,
      "step": 2788
    },
    {
      "epoch": 0.6772705196697426,
      "grad_norm": 0.33951547741889954,
      "learning_rate": 0.0001,
      "loss": 1.6219,
      "step": 2789
    },
    {
      "epoch": 0.6775133559980573,
      "grad_norm": 0.3280647397041321,
      "learning_rate": 0.0001,
      "loss": 1.4926,
      "step": 2790
    },
    {
      "epoch": 0.6777561923263721,
      "grad_norm": 0.38123634457588196,
      "learning_rate": 0.0001,
      "loss": 1.7184,
      "step": 2791
    },
    {
      "epoch": 0.6779990286546868,
      "grad_norm": 0.4019968807697296,
      "learning_rate": 0.0001,
      "loss": 1.9837,
      "step": 2792
    },
    {
      "epoch": 0.6782418649830014,
      "grad_norm": 0.3574753701686859,
      "learning_rate": 0.0001,
      "loss": 1.584,
      "step": 2793
    },
    {
      "epoch": 0.6784847013113162,
      "grad_norm": 0.36737924814224243,
      "learning_rate": 0.0001,
      "loss": 1.5873,
      "step": 2794
    },
    {
      "epoch": 0.6787275376396309,
      "grad_norm": 0.3804500102996826,
      "learning_rate": 0.0001,
      "loss": 1.7311,
      "step": 2795
    },
    {
      "epoch": 0.6789703739679456,
      "grad_norm": 0.34975457191467285,
      "learning_rate": 0.0001,
      "loss": 1.5167,
      "step": 2796
    },
    {
      "epoch": 0.6792132102962604,
      "grad_norm": 0.3589155673980713,
      "learning_rate": 0.0001,
      "loss": 1.756,
      "step": 2797
    },
    {
      "epoch": 0.679456046624575,
      "grad_norm": 0.3515421152114868,
      "learning_rate": 0.0001,
      "loss": 1.7368,
      "step": 2798
    },
    {
      "epoch": 0.6796988829528897,
      "grad_norm": 0.3487219512462616,
      "learning_rate": 0.0001,
      "loss": 1.588,
      "step": 2799
    },
    {
      "epoch": 0.6799417192812045,
      "grad_norm": 0.3846356272697449,
      "learning_rate": 0.0001,
      "loss": 1.6109,
      "step": 2800
    },
    {
      "epoch": 0.6801845556095192,
      "grad_norm": 0.3570033609867096,
      "learning_rate": 0.0001,
      "loss": 1.6234,
      "step": 2801
    },
    {
      "epoch": 0.6804273919378339,
      "grad_norm": 0.3869325518608093,
      "learning_rate": 0.0001,
      "loss": 1.7455,
      "step": 2802
    },
    {
      "epoch": 0.6806702282661486,
      "grad_norm": 0.35647138953208923,
      "learning_rate": 0.0001,
      "loss": 1.6681,
      "step": 2803
    },
    {
      "epoch": 0.6809130645944633,
      "grad_norm": 0.3512788414955139,
      "learning_rate": 0.0001,
      "loss": 1.5214,
      "step": 2804
    },
    {
      "epoch": 0.681155900922778,
      "grad_norm": 0.38840922713279724,
      "learning_rate": 0.0001,
      "loss": 1.8565,
      "step": 2805
    },
    {
      "epoch": 0.6813987372510928,
      "grad_norm": 0.3611179292201996,
      "learning_rate": 0.0001,
      "loss": 1.6661,
      "step": 2806
    },
    {
      "epoch": 0.6816415735794075,
      "grad_norm": 0.3591524660587311,
      "learning_rate": 0.0001,
      "loss": 1.791,
      "step": 2807
    },
    {
      "epoch": 0.6818844099077221,
      "grad_norm": 0.38657450675964355,
      "learning_rate": 0.0001,
      "loss": 1.7421,
      "step": 2808
    },
    {
      "epoch": 0.6821272462360369,
      "grad_norm": 0.36568185687065125,
      "learning_rate": 0.0001,
      "loss": 1.732,
      "step": 2809
    },
    {
      "epoch": 0.6823700825643516,
      "grad_norm": 0.36642563343048096,
      "learning_rate": 0.0001,
      "loss": 1.7422,
      "step": 2810
    },
    {
      "epoch": 0.6826129188926664,
      "grad_norm": 0.39892640709877014,
      "learning_rate": 0.0001,
      "loss": 1.965,
      "step": 2811
    },
    {
      "epoch": 0.6828557552209811,
      "grad_norm": 0.39376381039619446,
      "learning_rate": 0.0001,
      "loss": 1.8844,
      "step": 2812
    },
    {
      "epoch": 0.6830985915492958,
      "grad_norm": 0.35928022861480713,
      "learning_rate": 0.0001,
      "loss": 1.6422,
      "step": 2813
    },
    {
      "epoch": 0.6833414278776105,
      "grad_norm": 0.3709716796875,
      "learning_rate": 0.0001,
      "loss": 1.721,
      "step": 2814
    },
    {
      "epoch": 0.6835842642059252,
      "grad_norm": 0.3744778037071228,
      "learning_rate": 0.0001,
      "loss": 1.8316,
      "step": 2815
    },
    {
      "epoch": 0.6838271005342399,
      "grad_norm": 0.3626949191093445,
      "learning_rate": 0.0001,
      "loss": 1.608,
      "step": 2816
    },
    {
      "epoch": 0.6840699368625547,
      "grad_norm": 0.37098953127861023,
      "learning_rate": 0.0001,
      "loss": 1.8472,
      "step": 2817
    },
    {
      "epoch": 0.6843127731908694,
      "grad_norm": 0.35896000266075134,
      "learning_rate": 0.0001,
      "loss": 1.8092,
      "step": 2818
    },
    {
      "epoch": 0.684555609519184,
      "grad_norm": 0.3691065013408661,
      "learning_rate": 0.0001,
      "loss": 1.792,
      "step": 2819
    },
    {
      "epoch": 0.6847984458474988,
      "grad_norm": 0.3581167161464691,
      "learning_rate": 0.0001,
      "loss": 1.6491,
      "step": 2820
    },
    {
      "epoch": 0.6850412821758135,
      "grad_norm": 0.3448159098625183,
      "learning_rate": 0.0001,
      "loss": 1.7149,
      "step": 2821
    },
    {
      "epoch": 0.6852841185041282,
      "grad_norm": 0.3719812035560608,
      "learning_rate": 0.0001,
      "loss": 1.7454,
      "step": 2822
    },
    {
      "epoch": 0.685526954832443,
      "grad_norm": 0.35453951358795166,
      "learning_rate": 0.0001,
      "loss": 1.7427,
      "step": 2823
    },
    {
      "epoch": 0.6857697911607576,
      "grad_norm": 0.3828970193862915,
      "learning_rate": 0.0001,
      "loss": 1.7638,
      "step": 2824
    },
    {
      "epoch": 0.6860126274890723,
      "grad_norm": 0.34769728779792786,
      "learning_rate": 0.0001,
      "loss": 1.6054,
      "step": 2825
    },
    {
      "epoch": 0.6862554638173871,
      "grad_norm": 0.3649182915687561,
      "learning_rate": 0.0001,
      "loss": 1.8765,
      "step": 2826
    },
    {
      "epoch": 0.6864983001457018,
      "grad_norm": 0.3718123137950897,
      "learning_rate": 0.0001,
      "loss": 1.8563,
      "step": 2827
    },
    {
      "epoch": 0.6867411364740165,
      "grad_norm": 0.38165369629859924,
      "learning_rate": 0.0001,
      "loss": 1.7033,
      "step": 2828
    },
    {
      "epoch": 0.6869839728023313,
      "grad_norm": 0.3523254990577698,
      "learning_rate": 0.0001,
      "loss": 1.6182,
      "step": 2829
    },
    {
      "epoch": 0.6872268091306459,
      "grad_norm": 0.3618701994419098,
      "learning_rate": 0.0001,
      "loss": 1.6489,
      "step": 2830
    },
    {
      "epoch": 0.6874696454589606,
      "grad_norm": 0.3542143404483795,
      "learning_rate": 0.0001,
      "loss": 1.7151,
      "step": 2831
    },
    {
      "epoch": 0.6877124817872754,
      "grad_norm": 0.3631667494773865,
      "learning_rate": 0.0001,
      "loss": 1.7198,
      "step": 2832
    },
    {
      "epoch": 0.6879553181155901,
      "grad_norm": 0.35690179467201233,
      "learning_rate": 0.0001,
      "loss": 1.7414,
      "step": 2833
    },
    {
      "epoch": 0.6881981544439048,
      "grad_norm": 0.3542751371860504,
      "learning_rate": 0.0001,
      "loss": 1.7052,
      "step": 2834
    },
    {
      "epoch": 0.6884409907722195,
      "grad_norm": 0.36255085468292236,
      "learning_rate": 0.0001,
      "loss": 1.7469,
      "step": 2835
    },
    {
      "epoch": 0.6886838271005342,
      "grad_norm": 0.3458736836910248,
      "learning_rate": 0.0001,
      "loss": 1.6074,
      "step": 2836
    },
    {
      "epoch": 0.688926663428849,
      "grad_norm": 0.35049203038215637,
      "learning_rate": 0.0001,
      "loss": 1.6416,
      "step": 2837
    },
    {
      "epoch": 0.6891694997571637,
      "grad_norm": 0.3697221875190735,
      "learning_rate": 0.0001,
      "loss": 1.7879,
      "step": 2838
    },
    {
      "epoch": 0.6894123360854784,
      "grad_norm": 0.34680137038230896,
      "learning_rate": 0.0001,
      "loss": 1.6623,
      "step": 2839
    },
    {
      "epoch": 0.6896551724137931,
      "grad_norm": 0.4107730984687805,
      "learning_rate": 0.0001,
      "loss": 1.7127,
      "step": 2840
    },
    {
      "epoch": 0.6898980087421078,
      "grad_norm": 0.36184608936309814,
      "learning_rate": 0.0001,
      "loss": 1.7176,
      "step": 2841
    },
    {
      "epoch": 0.6901408450704225,
      "grad_norm": 0.3733268082141876,
      "learning_rate": 0.0001,
      "loss": 1.7568,
      "step": 2842
    },
    {
      "epoch": 0.6903836813987373,
      "grad_norm": 0.3708178997039795,
      "learning_rate": 0.0001,
      "loss": 1.8397,
      "step": 2843
    },
    {
      "epoch": 0.690626517727052,
      "grad_norm": 0.34265971183776855,
      "learning_rate": 0.0001,
      "loss": 1.7555,
      "step": 2844
    },
    {
      "epoch": 0.6908693540553666,
      "grad_norm": 0.3513777256011963,
      "learning_rate": 0.0001,
      "loss": 1.6347,
      "step": 2845
    },
    {
      "epoch": 0.6911121903836814,
      "grad_norm": 0.3604748249053955,
      "learning_rate": 0.0001,
      "loss": 1.8085,
      "step": 2846
    },
    {
      "epoch": 0.6913550267119961,
      "grad_norm": 0.36842814087867737,
      "learning_rate": 0.0001,
      "loss": 1.7507,
      "step": 2847
    },
    {
      "epoch": 0.6915978630403108,
      "grad_norm": 0.3745664656162262,
      "learning_rate": 0.0001,
      "loss": 1.6002,
      "step": 2848
    },
    {
      "epoch": 0.6918406993686256,
      "grad_norm": 0.3670363128185272,
      "learning_rate": 0.0001,
      "loss": 1.714,
      "step": 2849
    },
    {
      "epoch": 0.6920835356969403,
      "grad_norm": 0.3517747223377228,
      "learning_rate": 0.0001,
      "loss": 1.6155,
      "step": 2850
    },
    {
      "epoch": 0.6923263720252549,
      "grad_norm": 0.3622724115848541,
      "learning_rate": 0.0001,
      "loss": 1.7528,
      "step": 2851
    },
    {
      "epoch": 0.6925692083535697,
      "grad_norm": 0.36922353506088257,
      "learning_rate": 0.0001,
      "loss": 1.6394,
      "step": 2852
    },
    {
      "epoch": 0.6928120446818844,
      "grad_norm": 0.3616962432861328,
      "learning_rate": 0.0001,
      "loss": 1.6272,
      "step": 2853
    },
    {
      "epoch": 0.6930548810101991,
      "grad_norm": 0.36178451776504517,
      "learning_rate": 0.0001,
      "loss": 1.6249,
      "step": 2854
    },
    {
      "epoch": 0.6932977173385139,
      "grad_norm": 0.3681275248527527,
      "learning_rate": 0.0001,
      "loss": 1.7381,
      "step": 2855
    },
    {
      "epoch": 0.6935405536668285,
      "grad_norm": 0.37139177322387695,
      "learning_rate": 0.0001,
      "loss": 1.6641,
      "step": 2856
    },
    {
      "epoch": 0.6937833899951432,
      "grad_norm": 0.4304312765598297,
      "learning_rate": 0.0001,
      "loss": 1.6573,
      "step": 2857
    },
    {
      "epoch": 0.694026226323458,
      "grad_norm": 0.35460364818573,
      "learning_rate": 0.0001,
      "loss": 1.6168,
      "step": 2858
    },
    {
      "epoch": 0.6942690626517727,
      "grad_norm": 0.35733842849731445,
      "learning_rate": 0.0001,
      "loss": 1.6385,
      "step": 2859
    },
    {
      "epoch": 0.6945118989800875,
      "grad_norm": 0.3486827611923218,
      "learning_rate": 0.0001,
      "loss": 1.6365,
      "step": 2860
    },
    {
      "epoch": 0.6947547353084021,
      "grad_norm": 0.3677719235420227,
      "learning_rate": 0.0001,
      "loss": 1.7202,
      "step": 2861
    },
    {
      "epoch": 0.6949975716367168,
      "grad_norm": 0.4203099310398102,
      "learning_rate": 0.0001,
      "loss": 1.806,
      "step": 2862
    },
    {
      "epoch": 0.6952404079650316,
      "grad_norm": 0.3546834886074066,
      "learning_rate": 0.0001,
      "loss": 1.6461,
      "step": 2863
    },
    {
      "epoch": 0.6954832442933463,
      "grad_norm": 0.34874799847602844,
      "learning_rate": 0.0001,
      "loss": 1.6211,
      "step": 2864
    },
    {
      "epoch": 0.695726080621661,
      "grad_norm": 0.3491641581058502,
      "learning_rate": 0.0001,
      "loss": 1.6389,
      "step": 2865
    },
    {
      "epoch": 0.6959689169499758,
      "grad_norm": 0.3777684271335602,
      "learning_rate": 0.0001,
      "loss": 1.6018,
      "step": 2866
    },
    {
      "epoch": 0.6962117532782904,
      "grad_norm": 0.35895413160324097,
      "learning_rate": 0.0001,
      "loss": 1.5895,
      "step": 2867
    },
    {
      "epoch": 0.6964545896066051,
      "grad_norm": 0.3357354402542114,
      "learning_rate": 0.0001,
      "loss": 1.6225,
      "step": 2868
    },
    {
      "epoch": 0.6966974259349199,
      "grad_norm": 0.36405596137046814,
      "learning_rate": 0.0001,
      "loss": 1.6561,
      "step": 2869
    },
    {
      "epoch": 0.6969402622632346,
      "grad_norm": 0.3359208405017853,
      "learning_rate": 0.0001,
      "loss": 1.6273,
      "step": 2870
    },
    {
      "epoch": 0.6971830985915493,
      "grad_norm": 0.333678662776947,
      "learning_rate": 0.0001,
      "loss": 1.544,
      "step": 2871
    },
    {
      "epoch": 0.697425934919864,
      "grad_norm": 0.35640749335289,
      "learning_rate": 0.0001,
      "loss": 1.6572,
      "step": 2872
    },
    {
      "epoch": 0.6976687712481787,
      "grad_norm": 0.34949982166290283,
      "learning_rate": 0.0001,
      "loss": 1.7163,
      "step": 2873
    },
    {
      "epoch": 0.6979116075764934,
      "grad_norm": 0.3794506788253784,
      "learning_rate": 0.0001,
      "loss": 1.8469,
      "step": 2874
    },
    {
      "epoch": 0.6981544439048082,
      "grad_norm": 0.36223942041397095,
      "learning_rate": 0.0001,
      "loss": 1.802,
      "step": 2875
    },
    {
      "epoch": 0.6983972802331229,
      "grad_norm": 0.3552629351615906,
      "learning_rate": 0.0001,
      "loss": 1.6599,
      "step": 2876
    },
    {
      "epoch": 0.6986401165614375,
      "grad_norm": 0.35770854353904724,
      "learning_rate": 0.0001,
      "loss": 1.7617,
      "step": 2877
    },
    {
      "epoch": 0.6988829528897523,
      "grad_norm": 0.35151875019073486,
      "learning_rate": 0.0001,
      "loss": 1.6911,
      "step": 2878
    },
    {
      "epoch": 0.699125789218067,
      "grad_norm": 0.3744480013847351,
      "learning_rate": 0.0001,
      "loss": 1.6648,
      "step": 2879
    },
    {
      "epoch": 0.6993686255463817,
      "grad_norm": 0.3655317723751068,
      "learning_rate": 0.0001,
      "loss": 1.7782,
      "step": 2880
    },
    {
      "epoch": 0.6996114618746965,
      "grad_norm": 0.3632819950580597,
      "learning_rate": 0.0001,
      "loss": 1.6503,
      "step": 2881
    },
    {
      "epoch": 0.6998542982030111,
      "grad_norm": 0.365756094455719,
      "learning_rate": 0.0001,
      "loss": 1.6464,
      "step": 2882
    },
    {
      "epoch": 0.7000971345313259,
      "grad_norm": 0.34989890456199646,
      "learning_rate": 0.0001,
      "loss": 1.6492,
      "step": 2883
    },
    {
      "epoch": 0.7003399708596406,
      "grad_norm": 0.35626035928726196,
      "learning_rate": 0.0001,
      "loss": 1.4924,
      "step": 2884
    },
    {
      "epoch": 0.7005828071879553,
      "grad_norm": 0.38774046301841736,
      "learning_rate": 0.0001,
      "loss": 1.7844,
      "step": 2885
    },
    {
      "epoch": 0.7008256435162701,
      "grad_norm": 0.3692142963409424,
      "learning_rate": 0.0001,
      "loss": 1.7175,
      "step": 2886
    },
    {
      "epoch": 0.7010684798445848,
      "grad_norm": 0.33797192573547363,
      "learning_rate": 0.0001,
      "loss": 1.6566,
      "step": 2887
    },
    {
      "epoch": 0.7013113161728994,
      "grad_norm": 0.3465418219566345,
      "learning_rate": 0.0001,
      "loss": 1.595,
      "step": 2888
    },
    {
      "epoch": 0.7015541525012142,
      "grad_norm": 0.353791743516922,
      "learning_rate": 0.0001,
      "loss": 1.8116,
      "step": 2889
    },
    {
      "epoch": 0.7017969888295289,
      "grad_norm": 0.3460966944694519,
      "learning_rate": 0.0001,
      "loss": 1.6999,
      "step": 2890
    },
    {
      "epoch": 0.7020398251578436,
      "grad_norm": 0.35177546739578247,
      "learning_rate": 0.0001,
      "loss": 1.7354,
      "step": 2891
    },
    {
      "epoch": 0.7022826614861584,
      "grad_norm": 0.3689141869544983,
      "learning_rate": 0.0001,
      "loss": 1.704,
      "step": 2892
    },
    {
      "epoch": 0.702525497814473,
      "grad_norm": 0.35670557618141174,
      "learning_rate": 0.0001,
      "loss": 1.6332,
      "step": 2893
    },
    {
      "epoch": 0.7027683341427877,
      "grad_norm": 0.32746395468711853,
      "learning_rate": 0.0001,
      "loss": 1.3454,
      "step": 2894
    },
    {
      "epoch": 0.7030111704711025,
      "grad_norm": 0.351516455411911,
      "learning_rate": 0.0001,
      "loss": 1.7503,
      "step": 2895
    },
    {
      "epoch": 0.7032540067994172,
      "grad_norm": 0.3373326361179352,
      "learning_rate": 0.0001,
      "loss": 1.5746,
      "step": 2896
    },
    {
      "epoch": 0.7034968431277319,
      "grad_norm": 0.36539536714553833,
      "learning_rate": 0.0001,
      "loss": 1.6384,
      "step": 2897
    },
    {
      "epoch": 0.7037396794560467,
      "grad_norm": 0.6192057728767395,
      "learning_rate": 0.0001,
      "loss": 1.7346,
      "step": 2898
    },
    {
      "epoch": 0.7039825157843613,
      "grad_norm": 0.3435933291912079,
      "learning_rate": 0.0001,
      "loss": 1.6926,
      "step": 2899
    },
    {
      "epoch": 0.704225352112676,
      "grad_norm": 0.3592207729816437,
      "learning_rate": 0.0001,
      "loss": 1.686,
      "step": 2900
    },
    {
      "epoch": 0.7044681884409908,
      "grad_norm": 0.33544716238975525,
      "learning_rate": 0.0001,
      "loss": 1.6719,
      "step": 2901
    },
    {
      "epoch": 0.7047110247693055,
      "grad_norm": 0.379833847284317,
      "learning_rate": 0.0001,
      "loss": 1.7507,
      "step": 2902
    },
    {
      "epoch": 0.7049538610976201,
      "grad_norm": 0.3567056655883789,
      "learning_rate": 0.0001,
      "loss": 1.5885,
      "step": 2903
    },
    {
      "epoch": 0.7051966974259349,
      "grad_norm": 0.36313581466674805,
      "learning_rate": 0.0001,
      "loss": 1.773,
      "step": 2904
    },
    {
      "epoch": 0.7054395337542496,
      "grad_norm": 0.35985293984413147,
      "learning_rate": 0.0001,
      "loss": 1.8007,
      "step": 2905
    },
    {
      "epoch": 0.7056823700825644,
      "grad_norm": 0.3519170582294464,
      "learning_rate": 0.0001,
      "loss": 1.7713,
      "step": 2906
    },
    {
      "epoch": 0.7059252064108791,
      "grad_norm": 0.35734522342681885,
      "learning_rate": 0.0001,
      "loss": 1.5381,
      "step": 2907
    },
    {
      "epoch": 0.7061680427391938,
      "grad_norm": 0.38351139426231384,
      "learning_rate": 0.0001,
      "loss": 1.621,
      "step": 2908
    },
    {
      "epoch": 0.7064108790675085,
      "grad_norm": 0.39112213253974915,
      "learning_rate": 0.0001,
      "loss": 1.8291,
      "step": 2909
    },
    {
      "epoch": 0.7066537153958232,
      "grad_norm": 0.3656308948993683,
      "learning_rate": 0.0001,
      "loss": 1.8392,
      "step": 2910
    },
    {
      "epoch": 0.7068965517241379,
      "grad_norm": 0.3724428713321686,
      "learning_rate": 0.0001,
      "loss": 1.8628,
      "step": 2911
    },
    {
      "epoch": 0.7071393880524527,
      "grad_norm": 0.38988369703292847,
      "learning_rate": 0.0001,
      "loss": 1.7934,
      "step": 2912
    },
    {
      "epoch": 0.7073822243807674,
      "grad_norm": 0.34413278102874756,
      "learning_rate": 0.0001,
      "loss": 1.6022,
      "step": 2913
    },
    {
      "epoch": 0.707625060709082,
      "grad_norm": 0.3609641492366791,
      "learning_rate": 0.0001,
      "loss": 1.7286,
      "step": 2914
    },
    {
      "epoch": 0.7078678970373968,
      "grad_norm": 0.3715507984161377,
      "learning_rate": 0.0001,
      "loss": 1.8692,
      "step": 2915
    },
    {
      "epoch": 0.7081107333657115,
      "grad_norm": 0.3761596083641052,
      "learning_rate": 0.0001,
      "loss": 1.7286,
      "step": 2916
    },
    {
      "epoch": 0.7083535696940262,
      "grad_norm": 0.35243546962738037,
      "learning_rate": 0.0001,
      "loss": 1.7567,
      "step": 2917
    },
    {
      "epoch": 0.708596406022341,
      "grad_norm": 0.36848515272140503,
      "learning_rate": 0.0001,
      "loss": 1.7247,
      "step": 2918
    },
    {
      "epoch": 0.7088392423506557,
      "grad_norm": 0.36660486459732056,
      "learning_rate": 0.0001,
      "loss": 1.9038,
      "step": 2919
    },
    {
      "epoch": 0.7090820786789703,
      "grad_norm": 0.3525727093219757,
      "learning_rate": 0.0001,
      "loss": 1.7261,
      "step": 2920
    },
    {
      "epoch": 0.7093249150072851,
      "grad_norm": 0.402280330657959,
      "learning_rate": 0.0001,
      "loss": 1.8771,
      "step": 2921
    },
    {
      "epoch": 0.7095677513355998,
      "grad_norm": 0.3789054751396179,
      "learning_rate": 0.0001,
      "loss": 1.7986,
      "step": 2922
    },
    {
      "epoch": 0.7098105876639145,
      "grad_norm": 0.36707112193107605,
      "learning_rate": 0.0001,
      "loss": 1.7941,
      "step": 2923
    },
    {
      "epoch": 0.7100534239922293,
      "grad_norm": 0.35683396458625793,
      "learning_rate": 0.0001,
      "loss": 1.7371,
      "step": 2924
    },
    {
      "epoch": 0.7102962603205439,
      "grad_norm": 0.37030884623527527,
      "learning_rate": 0.0001,
      "loss": 1.6644,
      "step": 2925
    },
    {
      "epoch": 0.7105390966488586,
      "grad_norm": 0.38591259717941284,
      "learning_rate": 0.0001,
      "loss": 1.7018,
      "step": 2926
    },
    {
      "epoch": 0.7107819329771734,
      "grad_norm": 0.3896026611328125,
      "learning_rate": 0.0001,
      "loss": 1.7778,
      "step": 2927
    },
    {
      "epoch": 0.7110247693054881,
      "grad_norm": 0.3603595793247223,
      "learning_rate": 0.0001,
      "loss": 1.5979,
      "step": 2928
    },
    {
      "epoch": 0.7112676056338029,
      "grad_norm": 0.3804043233394623,
      "learning_rate": 0.0001,
      "loss": 1.7772,
      "step": 2929
    },
    {
      "epoch": 0.7115104419621175,
      "grad_norm": 0.3650084435939789,
      "learning_rate": 0.0001,
      "loss": 1.7025,
      "step": 2930
    },
    {
      "epoch": 0.7117532782904322,
      "grad_norm": 0.37033146619796753,
      "learning_rate": 0.0001,
      "loss": 1.8184,
      "step": 2931
    },
    {
      "epoch": 0.711996114618747,
      "grad_norm": 0.3535173237323761,
      "learning_rate": 0.0001,
      "loss": 1.6509,
      "step": 2932
    },
    {
      "epoch": 0.7122389509470617,
      "grad_norm": 0.3423483967781067,
      "learning_rate": 0.0001,
      "loss": 1.5214,
      "step": 2933
    },
    {
      "epoch": 0.7124817872753764,
      "grad_norm": 0.36774322390556335,
      "learning_rate": 0.0001,
      "loss": 1.8374,
      "step": 2934
    },
    {
      "epoch": 0.7127246236036912,
      "grad_norm": 0.3929206430912018,
      "learning_rate": 0.0001,
      "loss": 1.8696,
      "step": 2935
    },
    {
      "epoch": 0.7129674599320058,
      "grad_norm": 0.36784839630126953,
      "learning_rate": 0.0001,
      "loss": 1.6886,
      "step": 2936
    },
    {
      "epoch": 0.7132102962603205,
      "grad_norm": 0.39813923835754395,
      "learning_rate": 0.0001,
      "loss": 1.8575,
      "step": 2937
    },
    {
      "epoch": 0.7134531325886353,
      "grad_norm": 0.3539840877056122,
      "learning_rate": 0.0001,
      "loss": 1.7424,
      "step": 2938
    },
    {
      "epoch": 0.71369596891695,
      "grad_norm": 0.35620298981666565,
      "learning_rate": 0.0001,
      "loss": 1.6124,
      "step": 2939
    },
    {
      "epoch": 0.7139388052452647,
      "grad_norm": 0.35802677273750305,
      "learning_rate": 0.0001,
      "loss": 1.6363,
      "step": 2940
    },
    {
      "epoch": 0.7141816415735794,
      "grad_norm": 0.3788433074951172,
      "learning_rate": 0.0001,
      "loss": 1.8535,
      "step": 2941
    },
    {
      "epoch": 0.7144244779018941,
      "grad_norm": 0.3388618230819702,
      "learning_rate": 0.0001,
      "loss": 1.5753,
      "step": 2942
    },
    {
      "epoch": 0.7146673142302088,
      "grad_norm": 0.37521353363990784,
      "learning_rate": 0.0001,
      "loss": 1.7129,
      "step": 2943
    },
    {
      "epoch": 0.7149101505585236,
      "grad_norm": 0.3805568218231201,
      "learning_rate": 0.0001,
      "loss": 1.815,
      "step": 2944
    },
    {
      "epoch": 0.7151529868868383,
      "grad_norm": 0.3507835865020752,
      "learning_rate": 0.0001,
      "loss": 1.6339,
      "step": 2945
    },
    {
      "epoch": 0.7153958232151529,
      "grad_norm": 0.36764758825302124,
      "learning_rate": 0.0001,
      "loss": 1.5961,
      "step": 2946
    },
    {
      "epoch": 0.7156386595434677,
      "grad_norm": 0.3647833466529846,
      "learning_rate": 0.0001,
      "loss": 1.6382,
      "step": 2947
    },
    {
      "epoch": 0.7158814958717824,
      "grad_norm": 0.38218680024147034,
      "learning_rate": 0.0001,
      "loss": 1.9101,
      "step": 2948
    },
    {
      "epoch": 0.7161243322000971,
      "grad_norm": 0.34345221519470215,
      "learning_rate": 0.0001,
      "loss": 1.7023,
      "step": 2949
    },
    {
      "epoch": 0.7163671685284119,
      "grad_norm": 0.3746163249015808,
      "learning_rate": 0.0001,
      "loss": 1.7223,
      "step": 2950
    },
    {
      "epoch": 0.7166100048567265,
      "grad_norm": 0.3593830168247223,
      "learning_rate": 0.0001,
      "loss": 1.5907,
      "step": 2951
    },
    {
      "epoch": 0.7168528411850413,
      "grad_norm": 0.35187071561813354,
      "learning_rate": 0.0001,
      "loss": 1.6425,
      "step": 2952
    },
    {
      "epoch": 0.717095677513356,
      "grad_norm": 0.36968597769737244,
      "learning_rate": 0.0001,
      "loss": 1.5398,
      "step": 2953
    },
    {
      "epoch": 0.7173385138416707,
      "grad_norm": 0.35344114899635315,
      "learning_rate": 0.0001,
      "loss": 1.5868,
      "step": 2954
    },
    {
      "epoch": 0.7175813501699855,
      "grad_norm": 0.3591856360435486,
      "learning_rate": 0.0001,
      "loss": 1.5745,
      "step": 2955
    },
    {
      "epoch": 0.7178241864983002,
      "grad_norm": 0.3720721900463104,
      "learning_rate": 0.0001,
      "loss": 1.776,
      "step": 2956
    },
    {
      "epoch": 0.7180670228266148,
      "grad_norm": 0.36205706000328064,
      "learning_rate": 0.0001,
      "loss": 1.7799,
      "step": 2957
    },
    {
      "epoch": 0.7183098591549296,
      "grad_norm": 0.3556406497955322,
      "learning_rate": 0.0001,
      "loss": 1.7076,
      "step": 2958
    },
    {
      "epoch": 0.7185526954832443,
      "grad_norm": 0.3606123924255371,
      "learning_rate": 0.0001,
      "loss": 1.7962,
      "step": 2959
    },
    {
      "epoch": 0.718795531811559,
      "grad_norm": 0.35451430082321167,
      "learning_rate": 0.0001,
      "loss": 1.6689,
      "step": 2960
    },
    {
      "epoch": 0.7190383681398738,
      "grad_norm": 0.3367184102535248,
      "learning_rate": 0.0001,
      "loss": 1.6781,
      "step": 2961
    },
    {
      "epoch": 0.7192812044681884,
      "grad_norm": 0.356923907995224,
      "learning_rate": 0.0001,
      "loss": 1.6426,
      "step": 2962
    },
    {
      "epoch": 0.7195240407965031,
      "grad_norm": 0.38560399413108826,
      "learning_rate": 0.0001,
      "loss": 2.0482,
      "step": 2963
    },
    {
      "epoch": 0.7197668771248179,
      "grad_norm": 0.357217401266098,
      "learning_rate": 0.0001,
      "loss": 1.8023,
      "step": 2964
    },
    {
      "epoch": 0.7200097134531326,
      "grad_norm": 0.36859866976737976,
      "learning_rate": 0.0001,
      "loss": 1.6244,
      "step": 2965
    },
    {
      "epoch": 0.7202525497814473,
      "grad_norm": 0.3331631124019623,
      "learning_rate": 0.0001,
      "loss": 1.6152,
      "step": 2966
    },
    {
      "epoch": 0.720495386109762,
      "grad_norm": 0.3537011444568634,
      "learning_rate": 0.0001,
      "loss": 1.8049,
      "step": 2967
    },
    {
      "epoch": 0.7207382224380767,
      "grad_norm": 0.38022685050964355,
      "learning_rate": 0.0001,
      "loss": 1.8788,
      "step": 2968
    },
    {
      "epoch": 0.7209810587663914,
      "grad_norm": 0.3764863610267639,
      "learning_rate": 0.0001,
      "loss": 1.7169,
      "step": 2969
    },
    {
      "epoch": 0.7212238950947062,
      "grad_norm": 0.36463648080825806,
      "learning_rate": 0.0001,
      "loss": 1.8591,
      "step": 2970
    },
    {
      "epoch": 0.7214667314230209,
      "grad_norm": 0.34684088826179504,
      "learning_rate": 0.0001,
      "loss": 1.6458,
      "step": 2971
    },
    {
      "epoch": 0.7217095677513355,
      "grad_norm": 0.3718288838863373,
      "learning_rate": 0.0001,
      "loss": 1.7119,
      "step": 2972
    },
    {
      "epoch": 0.7219524040796503,
      "grad_norm": 0.3367420732975006,
      "learning_rate": 0.0001,
      "loss": 1.6449,
      "step": 2973
    },
    {
      "epoch": 0.722195240407965,
      "grad_norm": 0.3648276627063751,
      "learning_rate": 0.0001,
      "loss": 1.6718,
      "step": 2974
    },
    {
      "epoch": 0.7224380767362798,
      "grad_norm": 0.36360329389572144,
      "learning_rate": 0.0001,
      "loss": 1.8129,
      "step": 2975
    },
    {
      "epoch": 0.7226809130645945,
      "grad_norm": 0.38630610704421997,
      "learning_rate": 0.0001,
      "loss": 1.6333,
      "step": 2976
    },
    {
      "epoch": 0.7229237493929092,
      "grad_norm": 0.36072230339050293,
      "learning_rate": 0.0001,
      "loss": 1.7083,
      "step": 2977
    },
    {
      "epoch": 0.7231665857212239,
      "grad_norm": 0.3993285000324249,
      "learning_rate": 0.0001,
      "loss": 1.7232,
      "step": 2978
    },
    {
      "epoch": 0.7234094220495386,
      "grad_norm": 0.36294326186180115,
      "learning_rate": 0.0001,
      "loss": 1.7431,
      "step": 2979
    },
    {
      "epoch": 0.7236522583778533,
      "grad_norm": 0.35069945454597473,
      "learning_rate": 0.0001,
      "loss": 1.6702,
      "step": 2980
    },
    {
      "epoch": 0.7238950947061681,
      "grad_norm": 0.35238704085350037,
      "learning_rate": 0.0001,
      "loss": 1.6071,
      "step": 2981
    },
    {
      "epoch": 0.7241379310344828,
      "grad_norm": 0.3727080225944519,
      "learning_rate": 0.0001,
      "loss": 1.8298,
      "step": 2982
    },
    {
      "epoch": 0.7243807673627974,
      "grad_norm": 0.3415742516517639,
      "learning_rate": 0.0001,
      "loss": 1.6771,
      "step": 2983
    },
    {
      "epoch": 0.7246236036911122,
      "grad_norm": 0.3733143210411072,
      "learning_rate": 0.0001,
      "loss": 1.7415,
      "step": 2984
    },
    {
      "epoch": 0.7248664400194269,
      "grad_norm": 0.4065549671649933,
      "learning_rate": 0.0001,
      "loss": 1.6586,
      "step": 2985
    },
    {
      "epoch": 0.7251092763477416,
      "grad_norm": 0.3903333246707916,
      "learning_rate": 0.0001,
      "loss": 1.8753,
      "step": 2986
    },
    {
      "epoch": 0.7253521126760564,
      "grad_norm": 0.36670219898223877,
      "learning_rate": 0.0001,
      "loss": 1.7198,
      "step": 2987
    },
    {
      "epoch": 0.725594949004371,
      "grad_norm": 0.372465580701828,
      "learning_rate": 0.0001,
      "loss": 1.6701,
      "step": 2988
    },
    {
      "epoch": 0.7258377853326857,
      "grad_norm": 0.3704904317855835,
      "learning_rate": 0.0001,
      "loss": 1.7167,
      "step": 2989
    },
    {
      "epoch": 0.7260806216610005,
      "grad_norm": 0.36772340536117554,
      "learning_rate": 0.0001,
      "loss": 1.7857,
      "step": 2990
    },
    {
      "epoch": 0.7263234579893152,
      "grad_norm": 0.3715022802352905,
      "learning_rate": 0.0001,
      "loss": 1.7589,
      "step": 2991
    },
    {
      "epoch": 0.7265662943176299,
      "grad_norm": 0.35818248987197876,
      "learning_rate": 0.0001,
      "loss": 1.646,
      "step": 2992
    },
    {
      "epoch": 0.7268091306459447,
      "grad_norm": 0.3635038435459137,
      "learning_rate": 0.0001,
      "loss": 1.6426,
      "step": 2993
    },
    {
      "epoch": 0.7270519669742593,
      "grad_norm": 0.3737863004207611,
      "learning_rate": 0.0001,
      "loss": 1.7217,
      "step": 2994
    },
    {
      "epoch": 0.727294803302574,
      "grad_norm": 0.37764856219291687,
      "learning_rate": 0.0001,
      "loss": 1.6232,
      "step": 2995
    },
    {
      "epoch": 0.7275376396308888,
      "grad_norm": 0.34850314259529114,
      "learning_rate": 0.0001,
      "loss": 1.5159,
      "step": 2996
    },
    {
      "epoch": 0.7277804759592035,
      "grad_norm": 0.36448803544044495,
      "learning_rate": 0.0001,
      "loss": 1.6992,
      "step": 2997
    },
    {
      "epoch": 0.7280233122875183,
      "grad_norm": 0.36929845809936523,
      "learning_rate": 0.0001,
      "loss": 1.7365,
      "step": 2998
    },
    {
      "epoch": 0.7282661486158329,
      "grad_norm": 0.3795127868652344,
      "learning_rate": 0.0001,
      "loss": 1.7614,
      "step": 2999
    },
    {
      "epoch": 0.7285089849441476,
      "grad_norm": 0.3693547248840332,
      "learning_rate": 0.0001,
      "loss": 1.8719,
      "step": 3000
    },
    {
      "epoch": 0.7287518212724624,
      "grad_norm": 0.33187270164489746,
      "learning_rate": 0.0001,
      "loss": 1.6259,
      "step": 3001
    },
    {
      "epoch": 0.7289946576007771,
      "grad_norm": 0.3986161947250366,
      "learning_rate": 0.0001,
      "loss": 1.7832,
      "step": 3002
    },
    {
      "epoch": 0.7292374939290918,
      "grad_norm": 0.36065754294395447,
      "learning_rate": 0.0001,
      "loss": 1.8036,
      "step": 3003
    },
    {
      "epoch": 0.7294803302574066,
      "grad_norm": 0.35556501150131226,
      "learning_rate": 0.0001,
      "loss": 1.5954,
      "step": 3004
    },
    {
      "epoch": 0.7297231665857212,
      "grad_norm": 0.3530411720275879,
      "learning_rate": 0.0001,
      "loss": 1.7182,
      "step": 3005
    },
    {
      "epoch": 0.7299660029140359,
      "grad_norm": 0.40550586581230164,
      "learning_rate": 0.0001,
      "loss": 1.6723,
      "step": 3006
    },
    {
      "epoch": 0.7302088392423507,
      "grad_norm": 0.37494075298309326,
      "learning_rate": 0.0001,
      "loss": 1.7971,
      "step": 3007
    },
    {
      "epoch": 0.7304516755706654,
      "grad_norm": 0.38023585081100464,
      "learning_rate": 0.0001,
      "loss": 1.6531,
      "step": 3008
    },
    {
      "epoch": 0.73069451189898,
      "grad_norm": 0.3713429868221283,
      "learning_rate": 0.0001,
      "loss": 1.9022,
      "step": 3009
    },
    {
      "epoch": 0.7309373482272948,
      "grad_norm": 0.41121378540992737,
      "learning_rate": 0.0001,
      "loss": 1.7952,
      "step": 3010
    },
    {
      "epoch": 0.7311801845556095,
      "grad_norm": 0.3646053373813629,
      "learning_rate": 0.0001,
      "loss": 1.8038,
      "step": 3011
    },
    {
      "epoch": 0.7314230208839242,
      "grad_norm": 0.34566327929496765,
      "learning_rate": 0.0001,
      "loss": 1.4985,
      "step": 3012
    },
    {
      "epoch": 0.731665857212239,
      "grad_norm": 0.3599144220352173,
      "learning_rate": 0.0001,
      "loss": 1.6797,
      "step": 3013
    },
    {
      "epoch": 0.7319086935405537,
      "grad_norm": 0.34851646423339844,
      "learning_rate": 0.0001,
      "loss": 1.5837,
      "step": 3014
    },
    {
      "epoch": 0.7321515298688683,
      "grad_norm": 0.3771142065525055,
      "learning_rate": 0.0001,
      "loss": 1.6967,
      "step": 3015
    },
    {
      "epoch": 0.7323943661971831,
      "grad_norm": 0.34787657856941223,
      "learning_rate": 0.0001,
      "loss": 1.6806,
      "step": 3016
    },
    {
      "epoch": 0.7326372025254978,
      "grad_norm": 0.3611025810241699,
      "learning_rate": 0.0001,
      "loss": 1.5241,
      "step": 3017
    },
    {
      "epoch": 0.7328800388538125,
      "grad_norm": 0.3520788848400116,
      "learning_rate": 0.0001,
      "loss": 1.7057,
      "step": 3018
    },
    {
      "epoch": 0.7331228751821273,
      "grad_norm": 0.3628937900066376,
      "learning_rate": 0.0001,
      "loss": 1.5687,
      "step": 3019
    },
    {
      "epoch": 0.7333657115104419,
      "grad_norm": 0.35097360610961914,
      "learning_rate": 0.0001,
      "loss": 1.61,
      "step": 3020
    },
    {
      "epoch": 0.7336085478387567,
      "grad_norm": 0.37622305750846863,
      "learning_rate": 0.0001,
      "loss": 1.7781,
      "step": 3021
    },
    {
      "epoch": 0.7338513841670714,
      "grad_norm": 0.39038702845573425,
      "learning_rate": 0.0001,
      "loss": 1.8756,
      "step": 3022
    },
    {
      "epoch": 0.7340942204953861,
      "grad_norm": 0.381448358297348,
      "learning_rate": 0.0001,
      "loss": 1.6518,
      "step": 3023
    },
    {
      "epoch": 0.7343370568237009,
      "grad_norm": 0.3688746988773346,
      "learning_rate": 0.0001,
      "loss": 1.761,
      "step": 3024
    },
    {
      "epoch": 0.7345798931520156,
      "grad_norm": 0.37213847041130066,
      "learning_rate": 0.0001,
      "loss": 1.7584,
      "step": 3025
    },
    {
      "epoch": 0.7348227294803302,
      "grad_norm": 0.37292101979255676,
      "learning_rate": 0.0001,
      "loss": 1.686,
      "step": 3026
    },
    {
      "epoch": 0.735065565808645,
      "grad_norm": 0.3280561566352844,
      "learning_rate": 0.0001,
      "loss": 1.4754,
      "step": 3027
    },
    {
      "epoch": 0.7353084021369597,
      "grad_norm": 0.3627604842185974,
      "learning_rate": 0.0001,
      "loss": 1.6521,
      "step": 3028
    },
    {
      "epoch": 0.7355512384652744,
      "grad_norm": 0.3830265402793884,
      "learning_rate": 0.0001,
      "loss": 1.8071,
      "step": 3029
    },
    {
      "epoch": 0.7357940747935892,
      "grad_norm": 0.3494594991207123,
      "learning_rate": 0.0001,
      "loss": 1.5432,
      "step": 3030
    },
    {
      "epoch": 0.7360369111219038,
      "grad_norm": 0.3553903102874756,
      "learning_rate": 0.0001,
      "loss": 1.7695,
      "step": 3031
    },
    {
      "epoch": 0.7362797474502185,
      "grad_norm": 0.3740635812282562,
      "learning_rate": 0.0001,
      "loss": 1.7267,
      "step": 3032
    },
    {
      "epoch": 0.7365225837785333,
      "grad_norm": 0.34639549255371094,
      "learning_rate": 0.0001,
      "loss": 1.6156,
      "step": 3033
    },
    {
      "epoch": 0.736765420106848,
      "grad_norm": 0.4009366035461426,
      "learning_rate": 0.0001,
      "loss": 1.6878,
      "step": 3034
    },
    {
      "epoch": 0.7370082564351627,
      "grad_norm": 0.3689112067222595,
      "learning_rate": 0.0001,
      "loss": 1.6398,
      "step": 3035
    },
    {
      "epoch": 0.7372510927634774,
      "grad_norm": 0.3872334063053131,
      "learning_rate": 0.0001,
      "loss": 1.769,
      "step": 3036
    },
    {
      "epoch": 0.7374939290917921,
      "grad_norm": 0.3638404607772827,
      "learning_rate": 0.0001,
      "loss": 1.6397,
      "step": 3037
    },
    {
      "epoch": 0.7377367654201068,
      "grad_norm": 0.3534509241580963,
      "learning_rate": 0.0001,
      "loss": 1.6901,
      "step": 3038
    },
    {
      "epoch": 0.7379796017484216,
      "grad_norm": 0.36130160093307495,
      "learning_rate": 0.0001,
      "loss": 1.628,
      "step": 3039
    },
    {
      "epoch": 0.7382224380767363,
      "grad_norm": 0.35487523674964905,
      "learning_rate": 0.0001,
      "loss": 1.6811,
      "step": 3040
    },
    {
      "epoch": 0.7384652744050509,
      "grad_norm": 0.3612043857574463,
      "learning_rate": 0.0001,
      "loss": 1.7166,
      "step": 3041
    },
    {
      "epoch": 0.7387081107333657,
      "grad_norm": 0.37539663910865784,
      "learning_rate": 0.0001,
      "loss": 1.8484,
      "step": 3042
    },
    {
      "epoch": 0.7389509470616804,
      "grad_norm": 0.35037562251091003,
      "learning_rate": 0.0001,
      "loss": 1.6824,
      "step": 3043
    },
    {
      "epoch": 0.7391937833899952,
      "grad_norm": 0.36021849513053894,
      "learning_rate": 0.0001,
      "loss": 1.7272,
      "step": 3044
    },
    {
      "epoch": 0.7394366197183099,
      "grad_norm": 0.38384193181991577,
      "learning_rate": 0.0001,
      "loss": 1.7511,
      "step": 3045
    },
    {
      "epoch": 0.7396794560466246,
      "grad_norm": 0.3991917669773102,
      "learning_rate": 0.0001,
      "loss": 1.6954,
      "step": 3046
    },
    {
      "epoch": 0.7399222923749393,
      "grad_norm": 0.3840140402317047,
      "learning_rate": 0.0001,
      "loss": 1.6328,
      "step": 3047
    },
    {
      "epoch": 0.740165128703254,
      "grad_norm": 0.36042118072509766,
      "learning_rate": 0.0001,
      "loss": 1.6662,
      "step": 3048
    },
    {
      "epoch": 0.7404079650315687,
      "grad_norm": 0.3783377707004547,
      "learning_rate": 0.0001,
      "loss": 1.6943,
      "step": 3049
    },
    {
      "epoch": 0.7406508013598835,
      "grad_norm": 0.3565843403339386,
      "learning_rate": 0.0001,
      "loss": 1.7268,
      "step": 3050
    },
    {
      "epoch": 0.7408936376881982,
      "grad_norm": 0.34464597702026367,
      "learning_rate": 0.0001,
      "loss": 1.7426,
      "step": 3051
    },
    {
      "epoch": 0.7411364740165128,
      "grad_norm": 0.36159613728523254,
      "learning_rate": 0.0001,
      "loss": 1.7378,
      "step": 3052
    },
    {
      "epoch": 0.7413793103448276,
      "grad_norm": 0.3757307231426239,
      "learning_rate": 0.0001,
      "loss": 1.7074,
      "step": 3053
    },
    {
      "epoch": 0.7416221466731423,
      "grad_norm": 0.3887554705142975,
      "learning_rate": 0.0001,
      "loss": 1.8408,
      "step": 3054
    },
    {
      "epoch": 0.741864983001457,
      "grad_norm": 0.38285601139068604,
      "learning_rate": 0.0001,
      "loss": 1.7693,
      "step": 3055
    },
    {
      "epoch": 0.7421078193297718,
      "grad_norm": 0.3463350236415863,
      "learning_rate": 0.0001,
      "loss": 1.6086,
      "step": 3056
    },
    {
      "epoch": 0.7423506556580864,
      "grad_norm": 0.3648509979248047,
      "learning_rate": 0.0001,
      "loss": 1.6091,
      "step": 3057
    },
    {
      "epoch": 0.7425934919864011,
      "grad_norm": 0.34089362621307373,
      "learning_rate": 0.0001,
      "loss": 1.5889,
      "step": 3058
    },
    {
      "epoch": 0.7428363283147159,
      "grad_norm": 0.38336682319641113,
      "learning_rate": 0.0001,
      "loss": 1.6736,
      "step": 3059
    },
    {
      "epoch": 0.7430791646430306,
      "grad_norm": 0.37535834312438965,
      "learning_rate": 0.0001,
      "loss": 1.8989,
      "step": 3060
    },
    {
      "epoch": 0.7433220009713453,
      "grad_norm": 0.3540712594985962,
      "learning_rate": 0.0001,
      "loss": 1.7103,
      "step": 3061
    },
    {
      "epoch": 0.74356483729966,
      "grad_norm": 0.3841947019100189,
      "learning_rate": 0.0001,
      "loss": 1.7098,
      "step": 3062
    },
    {
      "epoch": 0.7438076736279747,
      "grad_norm": 0.38840192556381226,
      "learning_rate": 0.0001,
      "loss": 1.5798,
      "step": 3063
    },
    {
      "epoch": 0.7440505099562894,
      "grad_norm": 0.3693578243255615,
      "learning_rate": 0.0001,
      "loss": 1.6931,
      "step": 3064
    },
    {
      "epoch": 0.7442933462846042,
      "grad_norm": 0.3905344605445862,
      "learning_rate": 0.0001,
      "loss": 1.6618,
      "step": 3065
    },
    {
      "epoch": 0.7445361826129189,
      "grad_norm": 0.37690243124961853,
      "learning_rate": 0.0001,
      "loss": 1.7186,
      "step": 3066
    },
    {
      "epoch": 0.7447790189412337,
      "grad_norm": 0.37685608863830566,
      "learning_rate": 0.0001,
      "loss": 1.6061,
      "step": 3067
    },
    {
      "epoch": 0.7450218552695483,
      "grad_norm": 0.33402764797210693,
      "learning_rate": 0.0001,
      "loss": 1.581,
      "step": 3068
    },
    {
      "epoch": 0.745264691597863,
      "grad_norm": 0.3556559085845947,
      "learning_rate": 0.0001,
      "loss": 1.6959,
      "step": 3069
    },
    {
      "epoch": 0.7455075279261778,
      "grad_norm": 0.3709651231765747,
      "learning_rate": 0.0001,
      "loss": 1.6965,
      "step": 3070
    },
    {
      "epoch": 0.7457503642544925,
      "grad_norm": 0.37813064455986023,
      "learning_rate": 0.0001,
      "loss": 1.521,
      "step": 3071
    },
    {
      "epoch": 0.7459932005828072,
      "grad_norm": 0.3927704691886902,
      "learning_rate": 0.0001,
      "loss": 1.7058,
      "step": 3072
    },
    {
      "epoch": 0.746236036911122,
      "grad_norm": 0.3572740852832794,
      "learning_rate": 0.0001,
      "loss": 1.8172,
      "step": 3073
    },
    {
      "epoch": 0.7464788732394366,
      "grad_norm": 0.39202433824539185,
      "learning_rate": 0.0001,
      "loss": 1.6642,
      "step": 3074
    },
    {
      "epoch": 0.7467217095677513,
      "grad_norm": 0.37363842129707336,
      "learning_rate": 0.0001,
      "loss": 1.6773,
      "step": 3075
    },
    {
      "epoch": 0.7469645458960661,
      "grad_norm": 0.375654399394989,
      "learning_rate": 0.0001,
      "loss": 1.7064,
      "step": 3076
    },
    {
      "epoch": 0.7472073822243808,
      "grad_norm": 0.359342485666275,
      "learning_rate": 0.0001,
      "loss": 1.6948,
      "step": 3077
    },
    {
      "epoch": 0.7474502185526954,
      "grad_norm": 0.3476933538913727,
      "learning_rate": 0.0001,
      "loss": 1.6175,
      "step": 3078
    },
    {
      "epoch": 0.7476930548810102,
      "grad_norm": 0.38797739148139954,
      "learning_rate": 0.0001,
      "loss": 1.741,
      "step": 3079
    },
    {
      "epoch": 0.7479358912093249,
      "grad_norm": 0.3711453676223755,
      "learning_rate": 0.0001,
      "loss": 1.6954,
      "step": 3080
    },
    {
      "epoch": 0.7481787275376396,
      "grad_norm": 0.37687355279922485,
      "learning_rate": 0.0001,
      "loss": 1.7159,
      "step": 3081
    },
    {
      "epoch": 0.7484215638659544,
      "grad_norm": 0.36160585284233093,
      "learning_rate": 0.0001,
      "loss": 1.7158,
      "step": 3082
    },
    {
      "epoch": 0.748664400194269,
      "grad_norm": 0.3428528606891632,
      "learning_rate": 0.0001,
      "loss": 1.5954,
      "step": 3083
    },
    {
      "epoch": 0.7489072365225837,
      "grad_norm": 0.38035425543785095,
      "learning_rate": 0.0001,
      "loss": 1.7305,
      "step": 3084
    },
    {
      "epoch": 0.7491500728508985,
      "grad_norm": 0.37780630588531494,
      "learning_rate": 0.0001,
      "loss": 1.6278,
      "step": 3085
    },
    {
      "epoch": 0.7493929091792132,
      "grad_norm": 0.3517446219921112,
      "learning_rate": 0.0001,
      "loss": 1.7097,
      "step": 3086
    },
    {
      "epoch": 0.7496357455075279,
      "grad_norm": 0.384857177734375,
      "learning_rate": 0.0001,
      "loss": 1.7569,
      "step": 3087
    },
    {
      "epoch": 0.7498785818358427,
      "grad_norm": 0.35901719331741333,
      "learning_rate": 0.0001,
      "loss": 1.6987,
      "step": 3088
    },
    {
      "epoch": 0.7501214181641573,
      "grad_norm": 0.3643746078014374,
      "learning_rate": 0.0001,
      "loss": 1.8,
      "step": 3089
    },
    {
      "epoch": 0.7503642544924721,
      "grad_norm": 0.34800952672958374,
      "learning_rate": 0.0001,
      "loss": 1.802,
      "step": 3090
    },
    {
      "epoch": 0.7506070908207868,
      "grad_norm": 0.3544056713581085,
      "learning_rate": 0.0001,
      "loss": 1.7547,
      "step": 3091
    },
    {
      "epoch": 0.7508499271491015,
      "grad_norm": 0.32631686329841614,
      "learning_rate": 0.0001,
      "loss": 1.6443,
      "step": 3092
    },
    {
      "epoch": 0.7510927634774163,
      "grad_norm": 0.368476927280426,
      "learning_rate": 0.0001,
      "loss": 1.6593,
      "step": 3093
    },
    {
      "epoch": 0.751335599805731,
      "grad_norm": 0.3730914890766144,
      "learning_rate": 0.0001,
      "loss": 1.7031,
      "step": 3094
    },
    {
      "epoch": 0.7515784361340456,
      "grad_norm": 0.3924466669559479,
      "learning_rate": 0.0001,
      "loss": 1.708,
      "step": 3095
    },
    {
      "epoch": 0.7518212724623604,
      "grad_norm": 0.3600999414920807,
      "learning_rate": 0.0001,
      "loss": 1.5777,
      "step": 3096
    },
    {
      "epoch": 0.7520641087906751,
      "grad_norm": 0.3866841197013855,
      "learning_rate": 0.0001,
      "loss": 1.6916,
      "step": 3097
    },
    {
      "epoch": 0.7523069451189898,
      "grad_norm": 0.39760568737983704,
      "learning_rate": 0.0001,
      "loss": 1.7249,
      "step": 3098
    },
    {
      "epoch": 0.7525497814473046,
      "grad_norm": 0.3764370083808899,
      "learning_rate": 0.0001,
      "loss": 1.8291,
      "step": 3099
    },
    {
      "epoch": 0.7527926177756192,
      "grad_norm": 0.3639552593231201,
      "learning_rate": 0.0001,
      "loss": 1.8444,
      "step": 3100
    },
    {
      "epoch": 0.7530354541039339,
      "grad_norm": 0.34328848123550415,
      "learning_rate": 0.0001,
      "loss": 1.5405,
      "step": 3101
    },
    {
      "epoch": 0.7532782904322487,
      "grad_norm": 0.36111384630203247,
      "learning_rate": 0.0001,
      "loss": 1.7172,
      "step": 3102
    },
    {
      "epoch": 0.7535211267605634,
      "grad_norm": 0.3573237955570221,
      "learning_rate": 0.0001,
      "loss": 1.6974,
      "step": 3103
    },
    {
      "epoch": 0.753763963088878,
      "grad_norm": 0.36783355474472046,
      "learning_rate": 0.0001,
      "loss": 1.7718,
      "step": 3104
    },
    {
      "epoch": 0.7540067994171928,
      "grad_norm": 0.3471871614456177,
      "learning_rate": 0.0001,
      "loss": 1.7798,
      "step": 3105
    },
    {
      "epoch": 0.7542496357455075,
      "grad_norm": 0.37970730662345886,
      "learning_rate": 0.0001,
      "loss": 1.8948,
      "step": 3106
    },
    {
      "epoch": 0.7544924720738222,
      "grad_norm": 0.37085381150245667,
      "learning_rate": 0.0001,
      "loss": 1.6477,
      "step": 3107
    },
    {
      "epoch": 0.754735308402137,
      "grad_norm": 0.39417344331741333,
      "learning_rate": 0.0001,
      "loss": 1.8159,
      "step": 3108
    },
    {
      "epoch": 0.7549781447304517,
      "grad_norm": 0.3625888526439667,
      "learning_rate": 0.0001,
      "loss": 1.6386,
      "step": 3109
    },
    {
      "epoch": 0.7552209810587663,
      "grad_norm": 0.3569849729537964,
      "learning_rate": 0.0001,
      "loss": 1.644,
      "step": 3110
    },
    {
      "epoch": 0.7554638173870811,
      "grad_norm": 0.3573492169380188,
      "learning_rate": 0.0001,
      "loss": 1.7321,
      "step": 3111
    },
    {
      "epoch": 0.7557066537153958,
      "grad_norm": 0.3579084873199463,
      "learning_rate": 0.0001,
      "loss": 1.6692,
      "step": 3112
    },
    {
      "epoch": 0.7559494900437106,
      "grad_norm": 0.3422337770462036,
      "learning_rate": 0.0001,
      "loss": 1.7031,
      "step": 3113
    },
    {
      "epoch": 0.7561923263720253,
      "grad_norm": 0.3654865026473999,
      "learning_rate": 0.0001,
      "loss": 1.806,
      "step": 3114
    },
    {
      "epoch": 0.75643516270034,
      "grad_norm": 0.34083136916160583,
      "learning_rate": 0.0001,
      "loss": 1.5453,
      "step": 3115
    },
    {
      "epoch": 0.7566779990286547,
      "grad_norm": 0.3878861665725708,
      "learning_rate": 0.0001,
      "loss": 1.8776,
      "step": 3116
    },
    {
      "epoch": 0.7569208353569694,
      "grad_norm": 0.3917778730392456,
      "learning_rate": 0.0001,
      "loss": 1.7033,
      "step": 3117
    },
    {
      "epoch": 0.7571636716852841,
      "grad_norm": 0.3512471914291382,
      "learning_rate": 0.0001,
      "loss": 1.6972,
      "step": 3118
    },
    {
      "epoch": 0.7574065080135989,
      "grad_norm": 0.3681905269622803,
      "learning_rate": 0.0001,
      "loss": 1.6938,
      "step": 3119
    },
    {
      "epoch": 0.7576493443419136,
      "grad_norm": 0.373844712972641,
      "learning_rate": 0.0001,
      "loss": 1.6651,
      "step": 3120
    },
    {
      "epoch": 0.7578921806702282,
      "grad_norm": 0.36612239480018616,
      "learning_rate": 0.0001,
      "loss": 1.6204,
      "step": 3121
    },
    {
      "epoch": 0.758135016998543,
      "grad_norm": 0.35830995440483093,
      "learning_rate": 0.0001,
      "loss": 1.6255,
      "step": 3122
    },
    {
      "epoch": 0.7583778533268577,
      "grad_norm": 0.3689367473125458,
      "learning_rate": 0.0001,
      "loss": 1.6545,
      "step": 3123
    },
    {
      "epoch": 0.7586206896551724,
      "grad_norm": 0.36791351437568665,
      "learning_rate": 0.0001,
      "loss": 1.702,
      "step": 3124
    },
    {
      "epoch": 0.7588635259834872,
      "grad_norm": 0.37115055322647095,
      "learning_rate": 0.0001,
      "loss": 1.7153,
      "step": 3125
    },
    {
      "epoch": 0.7591063623118018,
      "grad_norm": 0.3720249831676483,
      "learning_rate": 0.0001,
      "loss": 1.7402,
      "step": 3126
    },
    {
      "epoch": 0.7593491986401165,
      "grad_norm": 0.3667154014110565,
      "learning_rate": 0.0001,
      "loss": 1.8012,
      "step": 3127
    },
    {
      "epoch": 0.7595920349684313,
      "grad_norm": 0.369220495223999,
      "learning_rate": 0.0001,
      "loss": 1.8846,
      "step": 3128
    },
    {
      "epoch": 0.759834871296746,
      "grad_norm": 0.37342166900634766,
      "learning_rate": 0.0001,
      "loss": 1.7446,
      "step": 3129
    },
    {
      "epoch": 0.7600777076250607,
      "grad_norm": 0.34881383180618286,
      "learning_rate": 0.0001,
      "loss": 1.6816,
      "step": 3130
    },
    {
      "epoch": 0.7603205439533754,
      "grad_norm": 0.34620028734207153,
      "learning_rate": 0.0001,
      "loss": 1.6644,
      "step": 3131
    },
    {
      "epoch": 0.7605633802816901,
      "grad_norm": 0.3626459538936615,
      "learning_rate": 0.0001,
      "loss": 1.7328,
      "step": 3132
    },
    {
      "epoch": 0.7608062166100048,
      "grad_norm": 0.3597855567932129,
      "learning_rate": 0.0001,
      "loss": 1.7039,
      "step": 3133
    },
    {
      "epoch": 0.7610490529383196,
      "grad_norm": 0.37836751341819763,
      "learning_rate": 0.0001,
      "loss": 1.916,
      "step": 3134
    },
    {
      "epoch": 0.7612918892666343,
      "grad_norm": 0.3582778871059418,
      "learning_rate": 0.0001,
      "loss": 1.6265,
      "step": 3135
    },
    {
      "epoch": 0.7615347255949491,
      "grad_norm": 0.34617099165916443,
      "learning_rate": 0.0001,
      "loss": 1.6059,
      "step": 3136
    },
    {
      "epoch": 0.7617775619232637,
      "grad_norm": 0.3793935775756836,
      "learning_rate": 0.0001,
      "loss": 1.7159,
      "step": 3137
    },
    {
      "epoch": 0.7620203982515784,
      "grad_norm": 0.3514363765716553,
      "learning_rate": 0.0001,
      "loss": 1.7584,
      "step": 3138
    },
    {
      "epoch": 0.7622632345798932,
      "grad_norm": 0.36340010166168213,
      "learning_rate": 0.0001,
      "loss": 1.7109,
      "step": 3139
    },
    {
      "epoch": 0.7625060709082079,
      "grad_norm": 0.3772117793560028,
      "learning_rate": 0.0001,
      "loss": 1.8007,
      "step": 3140
    },
    {
      "epoch": 0.7627489072365226,
      "grad_norm": 0.3619104027748108,
      "learning_rate": 0.0001,
      "loss": 1.7076,
      "step": 3141
    },
    {
      "epoch": 0.7629917435648373,
      "grad_norm": 0.366484671831131,
      "learning_rate": 0.0001,
      "loss": 1.749,
      "step": 3142
    },
    {
      "epoch": 0.763234579893152,
      "grad_norm": 0.37760522961616516,
      "learning_rate": 0.0001,
      "loss": 1.6653,
      "step": 3143
    },
    {
      "epoch": 0.7634774162214667,
      "grad_norm": 0.35971128940582275,
      "learning_rate": 0.0001,
      "loss": 1.7426,
      "step": 3144
    },
    {
      "epoch": 0.7637202525497815,
      "grad_norm": 0.3709493577480316,
      "learning_rate": 0.0001,
      "loss": 1.8147,
      "step": 3145
    },
    {
      "epoch": 0.7639630888780962,
      "grad_norm": 0.3713069558143616,
      "learning_rate": 0.0001,
      "loss": 1.6445,
      "step": 3146
    },
    {
      "epoch": 0.7642059252064108,
      "grad_norm": 0.3586064577102661,
      "learning_rate": 0.0001,
      "loss": 1.6478,
      "step": 3147
    },
    {
      "epoch": 0.7644487615347256,
      "grad_norm": 0.34258317947387695,
      "learning_rate": 0.0001,
      "loss": 1.696,
      "step": 3148
    },
    {
      "epoch": 0.7646915978630403,
      "grad_norm": 0.3723694682121277,
      "learning_rate": 0.0001,
      "loss": 1.7666,
      "step": 3149
    },
    {
      "epoch": 0.764934434191355,
      "grad_norm": 0.41363194584846497,
      "learning_rate": 0.0001,
      "loss": 1.7405,
      "step": 3150
    },
    {
      "epoch": 0.7651772705196698,
      "grad_norm": 0.36498725414276123,
      "learning_rate": 0.0001,
      "loss": 1.7503,
      "step": 3151
    },
    {
      "epoch": 0.7654201068479844,
      "grad_norm": 0.3648782968521118,
      "learning_rate": 0.0001,
      "loss": 1.7038,
      "step": 3152
    },
    {
      "epoch": 0.7656629431762991,
      "grad_norm": 0.3406696617603302,
      "learning_rate": 0.0001,
      "loss": 1.6182,
      "step": 3153
    },
    {
      "epoch": 0.7659057795046139,
      "grad_norm": 0.34064993262290955,
      "learning_rate": 0.0001,
      "loss": 1.6491,
      "step": 3154
    },
    {
      "epoch": 0.7661486158329286,
      "grad_norm": 0.3697326183319092,
      "learning_rate": 0.0001,
      "loss": 1.7287,
      "step": 3155
    },
    {
      "epoch": 0.7663914521612433,
      "grad_norm": 0.3929752707481384,
      "learning_rate": 0.0001,
      "loss": 1.7711,
      "step": 3156
    },
    {
      "epoch": 0.7666342884895581,
      "grad_norm": 0.3791561424732208,
      "learning_rate": 0.0001,
      "loss": 1.5454,
      "step": 3157
    },
    {
      "epoch": 0.7668771248178727,
      "grad_norm": 0.35193535685539246,
      "learning_rate": 0.0001,
      "loss": 1.6301,
      "step": 3158
    },
    {
      "epoch": 0.7671199611461875,
      "grad_norm": 0.3666219711303711,
      "learning_rate": 0.0001,
      "loss": 1.7659,
      "step": 3159
    },
    {
      "epoch": 0.7673627974745022,
      "grad_norm": 0.3909856379032135,
      "learning_rate": 0.0001,
      "loss": 1.6792,
      "step": 3160
    },
    {
      "epoch": 0.7676056338028169,
      "grad_norm": 0.38402125239372253,
      "learning_rate": 0.0001,
      "loss": 1.6724,
      "step": 3161
    },
    {
      "epoch": 0.7678484701311317,
      "grad_norm": 0.3770609498023987,
      "learning_rate": 0.0001,
      "loss": 1.6962,
      "step": 3162
    },
    {
      "epoch": 0.7680913064594463,
      "grad_norm": 0.3623834550380707,
      "learning_rate": 0.0001,
      "loss": 1.6792,
      "step": 3163
    },
    {
      "epoch": 0.768334142787761,
      "grad_norm": 0.38644856214523315,
      "learning_rate": 0.0001,
      "loss": 1.862,
      "step": 3164
    },
    {
      "epoch": 0.7685769791160758,
      "grad_norm": 0.36994674801826477,
      "learning_rate": 0.0001,
      "loss": 1.8121,
      "step": 3165
    },
    {
      "epoch": 0.7688198154443905,
      "grad_norm": 0.3379240334033966,
      "learning_rate": 0.0001,
      "loss": 1.6578,
      "step": 3166
    },
    {
      "epoch": 0.7690626517727052,
      "grad_norm": 0.3766805827617645,
      "learning_rate": 0.0001,
      "loss": 1.7598,
      "step": 3167
    },
    {
      "epoch": 0.76930548810102,
      "grad_norm": 0.3783973455429077,
      "learning_rate": 0.0001,
      "loss": 1.7333,
      "step": 3168
    },
    {
      "epoch": 0.7695483244293346,
      "grad_norm": 0.354950487613678,
      "learning_rate": 0.0001,
      "loss": 1.6588,
      "step": 3169
    },
    {
      "epoch": 0.7697911607576493,
      "grad_norm": 0.37209805846214294,
      "learning_rate": 0.0001,
      "loss": 1.6948,
      "step": 3170
    },
    {
      "epoch": 0.7700339970859641,
      "grad_norm": 0.38874351978302,
      "learning_rate": 0.0001,
      "loss": 1.7529,
      "step": 3171
    },
    {
      "epoch": 0.7702768334142788,
      "grad_norm": 0.3658055365085602,
      "learning_rate": 0.0001,
      "loss": 1.7363,
      "step": 3172
    },
    {
      "epoch": 0.7705196697425934,
      "grad_norm": 0.37785252928733826,
      "learning_rate": 0.0001,
      "loss": 1.6342,
      "step": 3173
    },
    {
      "epoch": 0.7707625060709082,
      "grad_norm": 0.35146087408065796,
      "learning_rate": 0.0001,
      "loss": 1.6762,
      "step": 3174
    },
    {
      "epoch": 0.7710053423992229,
      "grad_norm": 0.3499802350997925,
      "learning_rate": 0.0001,
      "loss": 1.7154,
      "step": 3175
    },
    {
      "epoch": 0.7712481787275376,
      "grad_norm": 0.3618769943714142,
      "learning_rate": 0.0001,
      "loss": 1.7542,
      "step": 3176
    },
    {
      "epoch": 0.7714910150558524,
      "grad_norm": 0.37386730313301086,
      "learning_rate": 0.0001,
      "loss": 1.7876,
      "step": 3177
    },
    {
      "epoch": 0.7717338513841671,
      "grad_norm": 0.3643173575401306,
      "learning_rate": 0.0001,
      "loss": 1.6575,
      "step": 3178
    },
    {
      "epoch": 0.7719766877124817,
      "grad_norm": 0.3616684377193451,
      "learning_rate": 0.0001,
      "loss": 1.6947,
      "step": 3179
    },
    {
      "epoch": 0.7722195240407965,
      "grad_norm": 0.3882141709327698,
      "learning_rate": 0.0001,
      "loss": 1.5895,
      "step": 3180
    },
    {
      "epoch": 0.7724623603691112,
      "grad_norm": 0.35174623131752014,
      "learning_rate": 0.0001,
      "loss": 1.5863,
      "step": 3181
    },
    {
      "epoch": 0.772705196697426,
      "grad_norm": 0.3746509552001953,
      "learning_rate": 0.0001,
      "loss": 1.7051,
      "step": 3182
    },
    {
      "epoch": 0.7729480330257407,
      "grad_norm": 0.3511878252029419,
      "learning_rate": 0.0001,
      "loss": 1.6308,
      "step": 3183
    },
    {
      "epoch": 0.7731908693540553,
      "grad_norm": 0.36943355202674866,
      "learning_rate": 0.0001,
      "loss": 1.8696,
      "step": 3184
    },
    {
      "epoch": 0.7734337056823701,
      "grad_norm": 0.3677257299423218,
      "learning_rate": 0.0001,
      "loss": 1.6531,
      "step": 3185
    },
    {
      "epoch": 0.7736765420106848,
      "grad_norm": 0.37608298659324646,
      "learning_rate": 0.0001,
      "loss": 1.7762,
      "step": 3186
    },
    {
      "epoch": 0.7739193783389995,
      "grad_norm": 0.35966455936431885,
      "learning_rate": 0.0001,
      "loss": 1.6512,
      "step": 3187
    },
    {
      "epoch": 0.7741622146673143,
      "grad_norm": 0.35690024495124817,
      "learning_rate": 0.0001,
      "loss": 1.5475,
      "step": 3188
    },
    {
      "epoch": 0.774405050995629,
      "grad_norm": 0.3662291169166565,
      "learning_rate": 0.0001,
      "loss": 1.7397,
      "step": 3189
    },
    {
      "epoch": 0.7746478873239436,
      "grad_norm": 0.3676881492137909,
      "learning_rate": 0.0001,
      "loss": 1.7465,
      "step": 3190
    },
    {
      "epoch": 0.7748907236522584,
      "grad_norm": 0.36643221974372864,
      "learning_rate": 0.0001,
      "loss": 1.6811,
      "step": 3191
    },
    {
      "epoch": 0.7751335599805731,
      "grad_norm": 0.3677716553211212,
      "learning_rate": 0.0001,
      "loss": 1.7394,
      "step": 3192
    },
    {
      "epoch": 0.7753763963088878,
      "grad_norm": 0.3882474899291992,
      "learning_rate": 0.0001,
      "loss": 1.9068,
      "step": 3193
    },
    {
      "epoch": 0.7756192326372026,
      "grad_norm": 0.36003318428993225,
      "learning_rate": 0.0001,
      "loss": 1.5764,
      "step": 3194
    },
    {
      "epoch": 0.7758620689655172,
      "grad_norm": 0.3616494834423065,
      "learning_rate": 0.0001,
      "loss": 1.7874,
      "step": 3195
    },
    {
      "epoch": 0.7761049052938319,
      "grad_norm": 0.36378318071365356,
      "learning_rate": 0.0001,
      "loss": 1.8228,
      "step": 3196
    },
    {
      "epoch": 0.7763477416221467,
      "grad_norm": 0.3754560053348541,
      "learning_rate": 0.0001,
      "loss": 1.8119,
      "step": 3197
    },
    {
      "epoch": 0.7765905779504614,
      "grad_norm": 0.38387665152549744,
      "learning_rate": 0.0001,
      "loss": 1.7857,
      "step": 3198
    },
    {
      "epoch": 0.7768334142787761,
      "grad_norm": 0.3557393550872803,
      "learning_rate": 0.0001,
      "loss": 1.704,
      "step": 3199
    },
    {
      "epoch": 0.7770762506070908,
      "grad_norm": 0.36891213059425354,
      "learning_rate": 0.0001,
      "loss": 1.6919,
      "step": 3200
    },
    {
      "epoch": 0.7773190869354055,
      "grad_norm": 0.38402387499809265,
      "learning_rate": 0.0001,
      "loss": 1.7416,
      "step": 3201
    },
    {
      "epoch": 0.7775619232637202,
      "grad_norm": 0.37884989380836487,
      "learning_rate": 0.0001,
      "loss": 1.8708,
      "step": 3202
    },
    {
      "epoch": 0.777804759592035,
      "grad_norm": 0.37382736802101135,
      "learning_rate": 0.0001,
      "loss": 1.8483,
      "step": 3203
    },
    {
      "epoch": 0.7780475959203497,
      "grad_norm": 0.38139647245407104,
      "learning_rate": 0.0001,
      "loss": 1.8322,
      "step": 3204
    },
    {
      "epoch": 0.7782904322486645,
      "grad_norm": 0.3646901547908783,
      "learning_rate": 0.0001,
      "loss": 1.6103,
      "step": 3205
    },
    {
      "epoch": 0.7785332685769791,
      "grad_norm": 0.3592281937599182,
      "learning_rate": 0.0001,
      "loss": 1.7267,
      "step": 3206
    },
    {
      "epoch": 0.7787761049052938,
      "grad_norm": 0.40910977125167847,
      "learning_rate": 0.0001,
      "loss": 1.7943,
      "step": 3207
    },
    {
      "epoch": 0.7790189412336086,
      "grad_norm": 0.3776848316192627,
      "learning_rate": 0.0001,
      "loss": 1.7475,
      "step": 3208
    },
    {
      "epoch": 0.7792617775619233,
      "grad_norm": 0.38275858759880066,
      "learning_rate": 0.0001,
      "loss": 1.7987,
      "step": 3209
    },
    {
      "epoch": 0.779504613890238,
      "grad_norm": 0.38461169600486755,
      "learning_rate": 0.0001,
      "loss": 1.7708,
      "step": 3210
    },
    {
      "epoch": 0.7797474502185527,
      "grad_norm": 0.36892619729042053,
      "learning_rate": 0.0001,
      "loss": 1.6829,
      "step": 3211
    },
    {
      "epoch": 0.7799902865468674,
      "grad_norm": 0.3585686981678009,
      "learning_rate": 0.0001,
      "loss": 1.6911,
      "step": 3212
    },
    {
      "epoch": 0.7802331228751821,
      "grad_norm": 0.358227014541626,
      "learning_rate": 0.0001,
      "loss": 1.5892,
      "step": 3213
    },
    {
      "epoch": 0.7804759592034969,
      "grad_norm": 0.37613505125045776,
      "learning_rate": 0.0001,
      "loss": 1.7145,
      "step": 3214
    },
    {
      "epoch": 0.7807187955318116,
      "grad_norm": 0.3856728672981262,
      "learning_rate": 0.0001,
      "loss": 1.7167,
      "step": 3215
    },
    {
      "epoch": 0.7809616318601262,
      "grad_norm": 0.3552917540073395,
      "learning_rate": 0.0001,
      "loss": 1.6204,
      "step": 3216
    },
    {
      "epoch": 0.781204468188441,
      "grad_norm": 0.3964640498161316,
      "learning_rate": 0.0001,
      "loss": 1.9078,
      "step": 3217
    },
    {
      "epoch": 0.7814473045167557,
      "grad_norm": 0.3828042447566986,
      "learning_rate": 0.0001,
      "loss": 1.7612,
      "step": 3218
    },
    {
      "epoch": 0.7816901408450704,
      "grad_norm": 0.37503907084465027,
      "learning_rate": 0.0001,
      "loss": 1.5186,
      "step": 3219
    },
    {
      "epoch": 0.7819329771733852,
      "grad_norm": 0.395419716835022,
      "learning_rate": 0.0001,
      "loss": 1.5914,
      "step": 3220
    },
    {
      "epoch": 0.7821758135016998,
      "grad_norm": 0.34283438324928284,
      "learning_rate": 0.0001,
      "loss": 1.6393,
      "step": 3221
    },
    {
      "epoch": 0.7824186498300145,
      "grad_norm": 0.3886133134365082,
      "learning_rate": 0.0001,
      "loss": 1.7853,
      "step": 3222
    },
    {
      "epoch": 0.7826614861583293,
      "grad_norm": 0.3807658851146698,
      "learning_rate": 0.0001,
      "loss": 1.6544,
      "step": 3223
    },
    {
      "epoch": 0.782904322486644,
      "grad_norm": 0.3506130874156952,
      "learning_rate": 0.0001,
      "loss": 1.6089,
      "step": 3224
    },
    {
      "epoch": 0.7831471588149587,
      "grad_norm": 0.3771856129169464,
      "learning_rate": 0.0001,
      "loss": 1.7467,
      "step": 3225
    },
    {
      "epoch": 0.7833899951432735,
      "grad_norm": 0.37501615285873413,
      "learning_rate": 0.0001,
      "loss": 1.6253,
      "step": 3226
    },
    {
      "epoch": 0.7836328314715881,
      "grad_norm": 0.3541441261768341,
      "learning_rate": 0.0001,
      "loss": 1.6923,
      "step": 3227
    },
    {
      "epoch": 0.7838756677999029,
      "grad_norm": 0.3698452413082123,
      "learning_rate": 0.0001,
      "loss": 1.7794,
      "step": 3228
    },
    {
      "epoch": 0.7841185041282176,
      "grad_norm": 0.3618011176586151,
      "learning_rate": 0.0001,
      "loss": 1.6233,
      "step": 3229
    },
    {
      "epoch": 0.7843613404565323,
      "grad_norm": 0.3783111870288849,
      "learning_rate": 0.0001,
      "loss": 1.7096,
      "step": 3230
    },
    {
      "epoch": 0.7846041767848471,
      "grad_norm": 0.3707350790500641,
      "learning_rate": 0.0001,
      "loss": 1.7606,
      "step": 3231
    },
    {
      "epoch": 0.7848470131131617,
      "grad_norm": 0.34714633226394653,
      "learning_rate": 0.0001,
      "loss": 1.6162,
      "step": 3232
    },
    {
      "epoch": 0.7850898494414764,
      "grad_norm": 0.3586050868034363,
      "learning_rate": 0.0001,
      "loss": 1.6275,
      "step": 3233
    },
    {
      "epoch": 0.7853326857697912,
      "grad_norm": 0.3658125400543213,
      "learning_rate": 0.0001,
      "loss": 1.7027,
      "step": 3234
    },
    {
      "epoch": 0.7855755220981059,
      "grad_norm": 0.35601937770843506,
      "learning_rate": 0.0001,
      "loss": 1.6929,
      "step": 3235
    },
    {
      "epoch": 0.7858183584264206,
      "grad_norm": 0.377439945936203,
      "learning_rate": 0.0001,
      "loss": 1.7305,
      "step": 3236
    },
    {
      "epoch": 0.7860611947547353,
      "grad_norm": 0.3666037619113922,
      "learning_rate": 0.0001,
      "loss": 1.7382,
      "step": 3237
    },
    {
      "epoch": 0.78630403108305,
      "grad_norm": 0.3553290367126465,
      "learning_rate": 0.0001,
      "loss": 1.6883,
      "step": 3238
    },
    {
      "epoch": 0.7865468674113647,
      "grad_norm": 0.36804431676864624,
      "learning_rate": 0.0001,
      "loss": 1.7385,
      "step": 3239
    },
    {
      "epoch": 0.7867897037396795,
      "grad_norm": 0.3820963501930237,
      "learning_rate": 0.0001,
      "loss": 1.831,
      "step": 3240
    },
    {
      "epoch": 0.7870325400679942,
      "grad_norm": 0.3654751181602478,
      "learning_rate": 0.0001,
      "loss": 1.6394,
      "step": 3241
    },
    {
      "epoch": 0.7872753763963088,
      "grad_norm": 0.37659505009651184,
      "learning_rate": 0.0001,
      "loss": 1.7862,
      "step": 3242
    },
    {
      "epoch": 0.7875182127246236,
      "grad_norm": 0.3688657283782959,
      "learning_rate": 0.0001,
      "loss": 1.6812,
      "step": 3243
    },
    {
      "epoch": 0.7877610490529383,
      "grad_norm": 0.37824395298957825,
      "learning_rate": 0.0001,
      "loss": 1.6927,
      "step": 3244
    },
    {
      "epoch": 0.788003885381253,
      "grad_norm": 0.36304405331611633,
      "learning_rate": 0.0001,
      "loss": 1.5247,
      "step": 3245
    },
    {
      "epoch": 0.7882467217095678,
      "grad_norm": 0.36583948135375977,
      "learning_rate": 0.0001,
      "loss": 1.7472,
      "step": 3246
    },
    {
      "epoch": 0.7884895580378825,
      "grad_norm": 0.3592684864997864,
      "learning_rate": 0.0001,
      "loss": 1.7444,
      "step": 3247
    },
    {
      "epoch": 0.7887323943661971,
      "grad_norm": 0.3583742380142212,
      "learning_rate": 0.0001,
      "loss": 1.6766,
      "step": 3248
    },
    {
      "epoch": 0.7889752306945119,
      "grad_norm": 0.3744942247867584,
      "learning_rate": 0.0001,
      "loss": 1.6817,
      "step": 3249
    },
    {
      "epoch": 0.7892180670228266,
      "grad_norm": 0.38749590516090393,
      "learning_rate": 0.0001,
      "loss": 1.5955,
      "step": 3250
    },
    {
      "epoch": 0.7894609033511414,
      "grad_norm": 0.35330215096473694,
      "learning_rate": 0.0001,
      "loss": 1.755,
      "step": 3251
    },
    {
      "epoch": 0.7897037396794561,
      "grad_norm": 0.36863937973976135,
      "learning_rate": 0.0001,
      "loss": 1.7673,
      "step": 3252
    },
    {
      "epoch": 0.7899465760077707,
      "grad_norm": 0.4081970751285553,
      "learning_rate": 0.0001,
      "loss": 1.7699,
      "step": 3253
    },
    {
      "epoch": 0.7901894123360855,
      "grad_norm": 0.3634713888168335,
      "learning_rate": 0.0001,
      "loss": 1.7146,
      "step": 3254
    },
    {
      "epoch": 0.7904322486644002,
      "grad_norm": 0.35160213708877563,
      "learning_rate": 0.0001,
      "loss": 1.7863,
      "step": 3255
    },
    {
      "epoch": 0.7906750849927149,
      "grad_norm": 0.3982776403427124,
      "learning_rate": 0.0001,
      "loss": 1.9224,
      "step": 3256
    },
    {
      "epoch": 0.7909179213210297,
      "grad_norm": 0.381833016872406,
      "learning_rate": 0.0001,
      "loss": 1.6776,
      "step": 3257
    },
    {
      "epoch": 0.7911607576493443,
      "grad_norm": 0.3521765172481537,
      "learning_rate": 0.0001,
      "loss": 1.7072,
      "step": 3258
    },
    {
      "epoch": 0.791403593977659,
      "grad_norm": 0.37415578961372375,
      "learning_rate": 0.0001,
      "loss": 1.815,
      "step": 3259
    },
    {
      "epoch": 0.7916464303059738,
      "grad_norm": 0.3757515549659729,
      "learning_rate": 0.0001,
      "loss": 1.7893,
      "step": 3260
    },
    {
      "epoch": 0.7918892666342885,
      "grad_norm": 0.3607475459575653,
      "learning_rate": 0.0001,
      "loss": 1.8022,
      "step": 3261
    },
    {
      "epoch": 0.7921321029626032,
      "grad_norm": 0.3828256130218506,
      "learning_rate": 0.0001,
      "loss": 1.7075,
      "step": 3262
    },
    {
      "epoch": 0.792374939290918,
      "grad_norm": 0.3982337415218353,
      "learning_rate": 0.0001,
      "loss": 1.9973,
      "step": 3263
    },
    {
      "epoch": 0.7926177756192326,
      "grad_norm": 0.35930705070495605,
      "learning_rate": 0.0001,
      "loss": 1.6939,
      "step": 3264
    },
    {
      "epoch": 0.7928606119475473,
      "grad_norm": 0.3679361641407013,
      "learning_rate": 0.0001,
      "loss": 1.7,
      "step": 3265
    },
    {
      "epoch": 0.7931034482758621,
      "grad_norm": 0.3649376928806305,
      "learning_rate": 0.0001,
      "loss": 1.6656,
      "step": 3266
    },
    {
      "epoch": 0.7933462846041768,
      "grad_norm": 0.3413490951061249,
      "learning_rate": 0.0001,
      "loss": 1.5184,
      "step": 3267
    },
    {
      "epoch": 0.7935891209324915,
      "grad_norm": 0.3625178039073944,
      "learning_rate": 0.0001,
      "loss": 1.6144,
      "step": 3268
    },
    {
      "epoch": 0.7938319572608062,
      "grad_norm": 0.36175259947776794,
      "learning_rate": 0.0001,
      "loss": 1.7998,
      "step": 3269
    },
    {
      "epoch": 0.7940747935891209,
      "grad_norm": 0.35525405406951904,
      "learning_rate": 0.0001,
      "loss": 1.831,
      "step": 3270
    },
    {
      "epoch": 0.7943176299174356,
      "grad_norm": 0.3651837706565857,
      "learning_rate": 0.0001,
      "loss": 1.7108,
      "step": 3271
    },
    {
      "epoch": 0.7945604662457504,
      "grad_norm": 0.34924477338790894,
      "learning_rate": 0.0001,
      "loss": 1.6447,
      "step": 3272
    },
    {
      "epoch": 0.7948033025740651,
      "grad_norm": 0.37113112211227417,
      "learning_rate": 0.0001,
      "loss": 1.7026,
      "step": 3273
    },
    {
      "epoch": 0.7950461389023799,
      "grad_norm": 0.3404080867767334,
      "learning_rate": 0.0001,
      "loss": 1.613,
      "step": 3274
    },
    {
      "epoch": 0.7952889752306945,
      "grad_norm": 0.38144057989120483,
      "learning_rate": 0.0001,
      "loss": 1.9026,
      "step": 3275
    },
    {
      "epoch": 0.7955318115590092,
      "grad_norm": 0.3685757517814636,
      "learning_rate": 0.0001,
      "loss": 1.7348,
      "step": 3276
    },
    {
      "epoch": 0.795774647887324,
      "grad_norm": 0.3650652766227722,
      "learning_rate": 0.0001,
      "loss": 1.7319,
      "step": 3277
    },
    {
      "epoch": 0.7960174842156387,
      "grad_norm": 0.3495384752750397,
      "learning_rate": 0.0001,
      "loss": 1.5908,
      "step": 3278
    },
    {
      "epoch": 0.7962603205439533,
      "grad_norm": 0.4011358618736267,
      "learning_rate": 0.0001,
      "loss": 1.6868,
      "step": 3279
    },
    {
      "epoch": 0.7965031568722681,
      "grad_norm": 0.37123218178749084,
      "learning_rate": 0.0001,
      "loss": 1.7107,
      "step": 3280
    },
    {
      "epoch": 0.7967459932005828,
      "grad_norm": 0.3909901976585388,
      "learning_rate": 0.0001,
      "loss": 1.7587,
      "step": 3281
    },
    {
      "epoch": 0.7969888295288975,
      "grad_norm": 0.39892497658729553,
      "learning_rate": 0.0001,
      "loss": 1.8257,
      "step": 3282
    },
    {
      "epoch": 0.7972316658572123,
      "grad_norm": 0.3590630888938904,
      "learning_rate": 0.0001,
      "loss": 1.6685,
      "step": 3283
    },
    {
      "epoch": 0.797474502185527,
      "grad_norm": 0.3754388093948364,
      "learning_rate": 0.0001,
      "loss": 1.8419,
      "step": 3284
    },
    {
      "epoch": 0.7977173385138416,
      "grad_norm": 0.39368849992752075,
      "learning_rate": 0.0001,
      "loss": 1.6508,
      "step": 3285
    },
    {
      "epoch": 0.7979601748421564,
      "grad_norm": 0.3552359938621521,
      "learning_rate": 0.0001,
      "loss": 1.7739,
      "step": 3286
    },
    {
      "epoch": 0.7982030111704711,
      "grad_norm": 0.3980265259742737,
      "learning_rate": 0.0001,
      "loss": 1.7555,
      "step": 3287
    },
    {
      "epoch": 0.7984458474987858,
      "grad_norm": 0.36231398582458496,
      "learning_rate": 0.0001,
      "loss": 1.5782,
      "step": 3288
    },
    {
      "epoch": 0.7986886838271006,
      "grad_norm": 0.3526536822319031,
      "learning_rate": 0.0001,
      "loss": 1.6628,
      "step": 3289
    },
    {
      "epoch": 0.7989315201554152,
      "grad_norm": 0.401947557926178,
      "learning_rate": 0.0001,
      "loss": 1.7001,
      "step": 3290
    },
    {
      "epoch": 0.7991743564837299,
      "grad_norm": 0.3727927505970001,
      "learning_rate": 0.0001,
      "loss": 1.6996,
      "step": 3291
    },
    {
      "epoch": 0.7994171928120447,
      "grad_norm": 0.37071672081947327,
      "learning_rate": 0.0001,
      "loss": 1.6681,
      "step": 3292
    },
    {
      "epoch": 0.7996600291403594,
      "grad_norm": 0.38639959692955017,
      "learning_rate": 0.0001,
      "loss": 1.7947,
      "step": 3293
    },
    {
      "epoch": 0.7999028654686741,
      "grad_norm": 0.35984519124031067,
      "learning_rate": 0.0001,
      "loss": 1.7382,
      "step": 3294
    },
    {
      "epoch": 0.8001457017969889,
      "grad_norm": 0.3773345947265625,
      "learning_rate": 0.0001,
      "loss": 1.8011,
      "step": 3295
    },
    {
      "epoch": 0.8003885381253035,
      "grad_norm": 0.3689408004283905,
      "learning_rate": 0.0001,
      "loss": 1.6109,
      "step": 3296
    },
    {
      "epoch": 0.8006313744536183,
      "grad_norm": 0.36196985840797424,
      "learning_rate": 0.0001,
      "loss": 1.6933,
      "step": 3297
    },
    {
      "epoch": 0.800874210781933,
      "grad_norm": 0.3734223544597626,
      "learning_rate": 0.0001,
      "loss": 1.6886,
      "step": 3298
    },
    {
      "epoch": 0.8011170471102477,
      "grad_norm": 0.38002875447273254,
      "learning_rate": 0.0001,
      "loss": 1.7362,
      "step": 3299
    },
    {
      "epoch": 0.8013598834385625,
      "grad_norm": 0.3728928864002228,
      "learning_rate": 0.0001,
      "loss": 1.8203,
      "step": 3300
    },
    {
      "epoch": 0.8016027197668771,
      "grad_norm": 0.3470155894756317,
      "learning_rate": 0.0001,
      "loss": 1.6447,
      "step": 3301
    },
    {
      "epoch": 0.8018455560951918,
      "grad_norm": 0.36024266481399536,
      "learning_rate": 0.0001,
      "loss": 1.8473,
      "step": 3302
    },
    {
      "epoch": 0.8020883924235066,
      "grad_norm": 0.3593677282333374,
      "learning_rate": 0.0001,
      "loss": 1.7207,
      "step": 3303
    },
    {
      "epoch": 0.8023312287518213,
      "grad_norm": 0.37037888169288635,
      "learning_rate": 0.0001,
      "loss": 1.7462,
      "step": 3304
    },
    {
      "epoch": 0.802574065080136,
      "grad_norm": 0.37236881256103516,
      "learning_rate": 0.0001,
      "loss": 1.7995,
      "step": 3305
    },
    {
      "epoch": 0.8028169014084507,
      "grad_norm": 0.3446558117866516,
      "learning_rate": 0.0001,
      "loss": 1.6447,
      "step": 3306
    },
    {
      "epoch": 0.8030597377367654,
      "grad_norm": 0.3672121465206146,
      "learning_rate": 0.0001,
      "loss": 1.7503,
      "step": 3307
    },
    {
      "epoch": 0.8033025740650801,
      "grad_norm": 0.3923403322696686,
      "learning_rate": 0.0001,
      "loss": 1.717,
      "step": 3308
    },
    {
      "epoch": 0.8035454103933949,
      "grad_norm": 0.39351925253868103,
      "learning_rate": 0.0001,
      "loss": 1.6714,
      "step": 3309
    },
    {
      "epoch": 0.8037882467217096,
      "grad_norm": 0.37669646739959717,
      "learning_rate": 0.0001,
      "loss": 1.7114,
      "step": 3310
    },
    {
      "epoch": 0.8040310830500242,
      "grad_norm": 0.3658166825771332,
      "learning_rate": 0.0001,
      "loss": 1.7235,
      "step": 3311
    },
    {
      "epoch": 0.804273919378339,
      "grad_norm": 0.3610089421272278,
      "learning_rate": 0.0001,
      "loss": 1.6012,
      "step": 3312
    },
    {
      "epoch": 0.8045167557066537,
      "grad_norm": 0.3617137670516968,
      "learning_rate": 0.0001,
      "loss": 1.6845,
      "step": 3313
    },
    {
      "epoch": 0.8047595920349684,
      "grad_norm": 0.37453964352607727,
      "learning_rate": 0.0001,
      "loss": 1.7513,
      "step": 3314
    },
    {
      "epoch": 0.8050024283632832,
      "grad_norm": 0.38044503331184387,
      "learning_rate": 0.0001,
      "loss": 1.7897,
      "step": 3315
    },
    {
      "epoch": 0.8052452646915979,
      "grad_norm": 0.3552364408969879,
      "learning_rate": 0.0001,
      "loss": 1.545,
      "step": 3316
    },
    {
      "epoch": 0.8054881010199125,
      "grad_norm": 0.38444262742996216,
      "learning_rate": 0.0001,
      "loss": 1.7455,
      "step": 3317
    },
    {
      "epoch": 0.8057309373482273,
      "grad_norm": 0.36202672123908997,
      "learning_rate": 0.0001,
      "loss": 1.6165,
      "step": 3318
    },
    {
      "epoch": 0.805973773676542,
      "grad_norm": 0.38152796030044556,
      "learning_rate": 0.0001,
      "loss": 1.8369,
      "step": 3319
    },
    {
      "epoch": 0.8062166100048568,
      "grad_norm": 0.3856159448623657,
      "learning_rate": 0.0001,
      "loss": 1.7539,
      "step": 3320
    },
    {
      "epoch": 0.8064594463331715,
      "grad_norm": 0.37758535146713257,
      "learning_rate": 0.0001,
      "loss": 1.7429,
      "step": 3321
    },
    {
      "epoch": 0.8067022826614861,
      "grad_norm": 0.4032396376132965,
      "learning_rate": 0.0001,
      "loss": 1.7092,
      "step": 3322
    },
    {
      "epoch": 0.8069451189898009,
      "grad_norm": 0.3742205798625946,
      "learning_rate": 0.0001,
      "loss": 1.6522,
      "step": 3323
    },
    {
      "epoch": 0.8071879553181156,
      "grad_norm": 0.3685240149497986,
      "learning_rate": 0.0001,
      "loss": 1.818,
      "step": 3324
    },
    {
      "epoch": 0.8074307916464303,
      "grad_norm": 0.39305341243743896,
      "learning_rate": 0.0001,
      "loss": 1.9695,
      "step": 3325
    },
    {
      "epoch": 0.8076736279747451,
      "grad_norm": 0.38635092973709106,
      "learning_rate": 0.0001,
      "loss": 1.6922,
      "step": 3326
    },
    {
      "epoch": 0.8079164643030597,
      "grad_norm": 0.3558487594127655,
      "learning_rate": 0.0001,
      "loss": 1.588,
      "step": 3327
    },
    {
      "epoch": 0.8081593006313744,
      "grad_norm": 0.37422701716423035,
      "learning_rate": 0.0001,
      "loss": 1.6272,
      "step": 3328
    },
    {
      "epoch": 0.8084021369596892,
      "grad_norm": 0.3510863184928894,
      "learning_rate": 0.0001,
      "loss": 1.7223,
      "step": 3329
    },
    {
      "epoch": 0.8086449732880039,
      "grad_norm": 0.38053059577941895,
      "learning_rate": 0.0001,
      "loss": 1.7258,
      "step": 3330
    },
    {
      "epoch": 0.8088878096163186,
      "grad_norm": 0.39362215995788574,
      "learning_rate": 0.0001,
      "loss": 1.6726,
      "step": 3331
    },
    {
      "epoch": 0.8091306459446334,
      "grad_norm": 0.3837401866912842,
      "learning_rate": 0.0001,
      "loss": 1.8808,
      "step": 3332
    },
    {
      "epoch": 0.809373482272948,
      "grad_norm": 0.35187309980392456,
      "learning_rate": 0.0001,
      "loss": 1.6576,
      "step": 3333
    },
    {
      "epoch": 0.8096163186012627,
      "grad_norm": 0.4003773331642151,
      "learning_rate": 0.0001,
      "loss": 1.8689,
      "step": 3334
    },
    {
      "epoch": 0.8098591549295775,
      "grad_norm": 0.3761930465698242,
      "learning_rate": 0.0001,
      "loss": 1.6872,
      "step": 3335
    },
    {
      "epoch": 0.8101019912578922,
      "grad_norm": 0.34285593032836914,
      "learning_rate": 0.0001,
      "loss": 1.471,
      "step": 3336
    },
    {
      "epoch": 0.8103448275862069,
      "grad_norm": 0.36871927976608276,
      "learning_rate": 0.0001,
      "loss": 1.7828,
      "step": 3337
    },
    {
      "epoch": 0.8105876639145216,
      "grad_norm": 0.4032946825027466,
      "learning_rate": 0.0001,
      "loss": 1.9575,
      "step": 3338
    },
    {
      "epoch": 0.8108305002428363,
      "grad_norm": 0.3758159875869751,
      "learning_rate": 0.0001,
      "loss": 1.702,
      "step": 3339
    },
    {
      "epoch": 0.811073336571151,
      "grad_norm": 0.396798312664032,
      "learning_rate": 0.0001,
      "loss": 1.9003,
      "step": 3340
    },
    {
      "epoch": 0.8113161728994658,
      "grad_norm": 0.39947810769081116,
      "learning_rate": 0.0001,
      "loss": 1.8968,
      "step": 3341
    },
    {
      "epoch": 0.8115590092277805,
      "grad_norm": 0.36287468671798706,
      "learning_rate": 0.0001,
      "loss": 1.6774,
      "step": 3342
    },
    {
      "epoch": 0.8118018455560952,
      "grad_norm": 0.37061214447021484,
      "learning_rate": 0.0001,
      "loss": 1.5444,
      "step": 3343
    },
    {
      "epoch": 0.8120446818844099,
      "grad_norm": 0.3543872535228729,
      "learning_rate": 0.0001,
      "loss": 1.7137,
      "step": 3344
    },
    {
      "epoch": 0.8122875182127246,
      "grad_norm": 0.40300488471984863,
      "learning_rate": 0.0001,
      "loss": 1.764,
      "step": 3345
    },
    {
      "epoch": 0.8125303545410394,
      "grad_norm": 0.34426894783973694,
      "learning_rate": 0.0001,
      "loss": 1.5996,
      "step": 3346
    },
    {
      "epoch": 0.8127731908693541,
      "grad_norm": 0.3738694489002228,
      "learning_rate": 0.0001,
      "loss": 1.6846,
      "step": 3347
    },
    {
      "epoch": 0.8130160271976687,
      "grad_norm": 0.38625332713127136,
      "learning_rate": 0.0001,
      "loss": 1.7664,
      "step": 3348
    },
    {
      "epoch": 0.8132588635259835,
      "grad_norm": 0.38722801208496094,
      "learning_rate": 0.0001,
      "loss": 1.922,
      "step": 3349
    },
    {
      "epoch": 0.8135016998542982,
      "grad_norm": 0.38053563237190247,
      "learning_rate": 0.0001,
      "loss": 1.836,
      "step": 3350
    },
    {
      "epoch": 0.8137445361826129,
      "grad_norm": 0.3337772786617279,
      "learning_rate": 0.0001,
      "loss": 1.4316,
      "step": 3351
    },
    {
      "epoch": 0.8139873725109277,
      "grad_norm": 0.3726194202899933,
      "learning_rate": 0.0001,
      "loss": 1.6717,
      "step": 3352
    },
    {
      "epoch": 0.8142302088392424,
      "grad_norm": 0.36288002133369446,
      "learning_rate": 0.0001,
      "loss": 1.8114,
      "step": 3353
    },
    {
      "epoch": 0.814473045167557,
      "grad_norm": 0.3562511205673218,
      "learning_rate": 0.0001,
      "loss": 1.6636,
      "step": 3354
    },
    {
      "epoch": 0.8147158814958718,
      "grad_norm": 0.360158234834671,
      "learning_rate": 0.0001,
      "loss": 1.6963,
      "step": 3355
    },
    {
      "epoch": 0.8149587178241865,
      "grad_norm": 0.36954307556152344,
      "learning_rate": 0.0001,
      "loss": 1.5675,
      "step": 3356
    },
    {
      "epoch": 0.8152015541525012,
      "grad_norm": 0.3674972653388977,
      "learning_rate": 0.0001,
      "loss": 1.7673,
      "step": 3357
    },
    {
      "epoch": 0.815444390480816,
      "grad_norm": 0.3780413866043091,
      "learning_rate": 0.0001,
      "loss": 1.68,
      "step": 3358
    },
    {
      "epoch": 0.8156872268091306,
      "grad_norm": 0.3669142723083496,
      "learning_rate": 0.0001,
      "loss": 1.7466,
      "step": 3359
    },
    {
      "epoch": 0.8159300631374453,
      "grad_norm": 0.34627634286880493,
      "learning_rate": 0.0001,
      "loss": 1.5616,
      "step": 3360
    },
    {
      "epoch": 0.8161728994657601,
      "grad_norm": 0.3495223820209503,
      "learning_rate": 0.0001,
      "loss": 1.7176,
      "step": 3361
    },
    {
      "epoch": 0.8164157357940748,
      "grad_norm": 0.3799058198928833,
      "learning_rate": 0.0001,
      "loss": 1.8741,
      "step": 3362
    },
    {
      "epoch": 0.8166585721223895,
      "grad_norm": 0.3761565685272217,
      "learning_rate": 0.0001,
      "loss": 1.7821,
      "step": 3363
    },
    {
      "epoch": 0.8169014084507042,
      "grad_norm": 0.39104485511779785,
      "learning_rate": 0.0001,
      "loss": 1.8327,
      "step": 3364
    },
    {
      "epoch": 0.8171442447790189,
      "grad_norm": 0.4119609296321869,
      "learning_rate": 0.0001,
      "loss": 1.8113,
      "step": 3365
    },
    {
      "epoch": 0.8173870811073336,
      "grad_norm": 0.37373781204223633,
      "learning_rate": 0.0001,
      "loss": 1.7457,
      "step": 3366
    },
    {
      "epoch": 0.8176299174356484,
      "grad_norm": 0.38254791498184204,
      "learning_rate": 0.0001,
      "loss": 1.7419,
      "step": 3367
    },
    {
      "epoch": 0.8178727537639631,
      "grad_norm": 0.36546850204467773,
      "learning_rate": 0.0001,
      "loss": 1.8308,
      "step": 3368
    },
    {
      "epoch": 0.8181155900922779,
      "grad_norm": 0.39306992292404175,
      "learning_rate": 0.0001,
      "loss": 1.8687,
      "step": 3369
    },
    {
      "epoch": 0.8183584264205925,
      "grad_norm": 0.3737972378730774,
      "learning_rate": 0.0001,
      "loss": 1.9084,
      "step": 3370
    },
    {
      "epoch": 0.8186012627489072,
      "grad_norm": 0.3632451891899109,
      "learning_rate": 0.0001,
      "loss": 1.7257,
      "step": 3371
    },
    {
      "epoch": 0.818844099077222,
      "grad_norm": 0.3642801344394684,
      "learning_rate": 0.0001,
      "loss": 1.7191,
      "step": 3372
    },
    {
      "epoch": 0.8190869354055367,
      "grad_norm": 0.3375829756259918,
      "learning_rate": 0.0001,
      "loss": 1.4421,
      "step": 3373
    },
    {
      "epoch": 0.8193297717338514,
      "grad_norm": 0.3556838035583496,
      "learning_rate": 0.0001,
      "loss": 1.5868,
      "step": 3374
    },
    {
      "epoch": 0.8195726080621661,
      "grad_norm": 0.348859041929245,
      "learning_rate": 0.0001,
      "loss": 1.5974,
      "step": 3375
    },
    {
      "epoch": 0.8198154443904808,
      "grad_norm": 0.3708842098712921,
      "learning_rate": 0.0001,
      "loss": 1.743,
      "step": 3376
    },
    {
      "epoch": 0.8200582807187955,
      "grad_norm": 0.36357876658439636,
      "learning_rate": 0.0001,
      "loss": 1.7435,
      "step": 3377
    },
    {
      "epoch": 0.8203011170471103,
      "grad_norm": 0.3622495234012604,
      "learning_rate": 0.0001,
      "loss": 1.7109,
      "step": 3378
    },
    {
      "epoch": 0.820543953375425,
      "grad_norm": 0.35432299971580505,
      "learning_rate": 0.0001,
      "loss": 1.7539,
      "step": 3379
    },
    {
      "epoch": 0.8207867897037396,
      "grad_norm": 0.35265427827835083,
      "learning_rate": 0.0001,
      "loss": 1.5913,
      "step": 3380
    },
    {
      "epoch": 0.8210296260320544,
      "grad_norm": 0.37112993001937866,
      "learning_rate": 0.0001,
      "loss": 1.5671,
      "step": 3381
    },
    {
      "epoch": 0.8212724623603691,
      "grad_norm": 0.35254931449890137,
      "learning_rate": 0.0001,
      "loss": 1.7341,
      "step": 3382
    },
    {
      "epoch": 0.8215152986886838,
      "grad_norm": 0.38896042108535767,
      "learning_rate": 0.0001,
      "loss": 1.8627,
      "step": 3383
    },
    {
      "epoch": 0.8217581350169986,
      "grad_norm": 0.37757647037506104,
      "learning_rate": 0.0001,
      "loss": 1.8308,
      "step": 3384
    },
    {
      "epoch": 0.8220009713453132,
      "grad_norm": 0.3572361171245575,
      "learning_rate": 0.0001,
      "loss": 1.6857,
      "step": 3385
    },
    {
      "epoch": 0.8222438076736279,
      "grad_norm": 0.35898885130882263,
      "learning_rate": 0.0001,
      "loss": 1.8175,
      "step": 3386
    },
    {
      "epoch": 0.8224866440019427,
      "grad_norm": 0.36892169713974,
      "learning_rate": 0.0001,
      "loss": 1.675,
      "step": 3387
    },
    {
      "epoch": 0.8227294803302574,
      "grad_norm": 0.3749498426914215,
      "learning_rate": 0.0001,
      "loss": 1.6009,
      "step": 3388
    },
    {
      "epoch": 0.8229723166585721,
      "grad_norm": 0.3872259557247162,
      "learning_rate": 0.0001,
      "loss": 1.8164,
      "step": 3389
    },
    {
      "epoch": 0.8232151529868869,
      "grad_norm": 0.36481043696403503,
      "learning_rate": 0.0001,
      "loss": 1.7736,
      "step": 3390
    },
    {
      "epoch": 0.8234579893152015,
      "grad_norm": 0.3672960698604584,
      "learning_rate": 0.0001,
      "loss": 1.8317,
      "step": 3391
    },
    {
      "epoch": 0.8237008256435163,
      "grad_norm": 0.5431075692176819,
      "learning_rate": 0.0001,
      "loss": 1.7295,
      "step": 3392
    },
    {
      "epoch": 0.823943661971831,
      "grad_norm": 0.36839866638183594,
      "learning_rate": 0.0001,
      "loss": 1.7219,
      "step": 3393
    },
    {
      "epoch": 0.8241864983001457,
      "grad_norm": 0.36228734254837036,
      "learning_rate": 0.0001,
      "loss": 1.7996,
      "step": 3394
    },
    {
      "epoch": 0.8244293346284605,
      "grad_norm": 0.3687099516391754,
      "learning_rate": 0.0001,
      "loss": 1.8457,
      "step": 3395
    },
    {
      "epoch": 0.8246721709567751,
      "grad_norm": 0.3856877088546753,
      "learning_rate": 0.0001,
      "loss": 1.5591,
      "step": 3396
    },
    {
      "epoch": 0.8249150072850898,
      "grad_norm": 0.3498803675174713,
      "learning_rate": 0.0001,
      "loss": 1.6165,
      "step": 3397
    },
    {
      "epoch": 0.8251578436134046,
      "grad_norm": 0.3575108051300049,
      "learning_rate": 0.0001,
      "loss": 1.6455,
      "step": 3398
    },
    {
      "epoch": 0.8254006799417193,
      "grad_norm": 0.3399507403373718,
      "learning_rate": 0.0001,
      "loss": 1.5308,
      "step": 3399
    },
    {
      "epoch": 0.825643516270034,
      "grad_norm": 0.39992624521255493,
      "learning_rate": 0.0001,
      "loss": 1.8403,
      "step": 3400
    },
    {
      "epoch": 0.8258863525983487,
      "grad_norm": 0.34879302978515625,
      "learning_rate": 0.0001,
      "loss": 1.6182,
      "step": 3401
    },
    {
      "epoch": 0.8261291889266634,
      "grad_norm": 0.3787181079387665,
      "learning_rate": 0.0001,
      "loss": 1.8406,
      "step": 3402
    },
    {
      "epoch": 0.8263720252549781,
      "grad_norm": 0.366046667098999,
      "learning_rate": 0.0001,
      "loss": 1.6661,
      "step": 3403
    },
    {
      "epoch": 0.8266148615832929,
      "grad_norm": 0.38305869698524475,
      "learning_rate": 0.0001,
      "loss": 1.6747,
      "step": 3404
    },
    {
      "epoch": 0.8268576979116076,
      "grad_norm": 0.3716723322868347,
      "learning_rate": 0.0001,
      "loss": 1.8145,
      "step": 3405
    },
    {
      "epoch": 0.8271005342399222,
      "grad_norm": 0.3804088830947876,
      "learning_rate": 0.0001,
      "loss": 1.6595,
      "step": 3406
    },
    {
      "epoch": 0.827343370568237,
      "grad_norm": 0.36902347207069397,
      "learning_rate": 0.0001,
      "loss": 1.7445,
      "step": 3407
    },
    {
      "epoch": 0.8275862068965517,
      "grad_norm": 0.3657947778701782,
      "learning_rate": 0.0001,
      "loss": 1.719,
      "step": 3408
    },
    {
      "epoch": 0.8278290432248664,
      "grad_norm": 0.3669492304325104,
      "learning_rate": 0.0001,
      "loss": 1.7128,
      "step": 3409
    },
    {
      "epoch": 0.8280718795531812,
      "grad_norm": 0.34419408440589905,
      "learning_rate": 0.0001,
      "loss": 1.6568,
      "step": 3410
    },
    {
      "epoch": 0.8283147158814959,
      "grad_norm": 0.39030420780181885,
      "learning_rate": 0.0001,
      "loss": 1.899,
      "step": 3411
    },
    {
      "epoch": 0.8285575522098105,
      "grad_norm": 0.3708794116973877,
      "learning_rate": 0.0001,
      "loss": 1.5666,
      "step": 3412
    },
    {
      "epoch": 0.8288003885381253,
      "grad_norm": 0.3762510120868683,
      "learning_rate": 0.0001,
      "loss": 1.6719,
      "step": 3413
    },
    {
      "epoch": 0.82904322486644,
      "grad_norm": 0.36242151260375977,
      "learning_rate": 0.0001,
      "loss": 1.713,
      "step": 3414
    },
    {
      "epoch": 0.8292860611947548,
      "grad_norm": 0.35218676924705505,
      "learning_rate": 0.0001,
      "loss": 1.6207,
      "step": 3415
    },
    {
      "epoch": 0.8295288975230695,
      "grad_norm": 0.3742830157279968,
      "learning_rate": 0.0001,
      "loss": 1.8128,
      "step": 3416
    },
    {
      "epoch": 0.8297717338513841,
      "grad_norm": 0.360321044921875,
      "learning_rate": 0.0001,
      "loss": 1.5888,
      "step": 3417
    },
    {
      "epoch": 0.8300145701796989,
      "grad_norm": 0.3791470527648926,
      "learning_rate": 0.0001,
      "loss": 1.8753,
      "step": 3418
    },
    {
      "epoch": 0.8302574065080136,
      "grad_norm": 0.3563266396522522,
      "learning_rate": 0.0001,
      "loss": 1.6355,
      "step": 3419
    },
    {
      "epoch": 0.8305002428363283,
      "grad_norm": 0.3572905361652374,
      "learning_rate": 0.0001,
      "loss": 1.678,
      "step": 3420
    },
    {
      "epoch": 0.8307430791646431,
      "grad_norm": 0.3519369065761566,
      "learning_rate": 0.0001,
      "loss": 1.6316,
      "step": 3421
    },
    {
      "epoch": 0.8309859154929577,
      "grad_norm": 0.3865967392921448,
      "learning_rate": 0.0001,
      "loss": 1.6014,
      "step": 3422
    },
    {
      "epoch": 0.8312287518212724,
      "grad_norm": 0.36829259991645813,
      "learning_rate": 0.0001,
      "loss": 1.5808,
      "step": 3423
    },
    {
      "epoch": 0.8314715881495872,
      "grad_norm": 0.5048966407775879,
      "learning_rate": 0.0001,
      "loss": 1.783,
      "step": 3424
    },
    {
      "epoch": 0.8317144244779019,
      "grad_norm": 0.3445897400379181,
      "learning_rate": 0.0001,
      "loss": 1.5677,
      "step": 3425
    },
    {
      "epoch": 0.8319572608062166,
      "grad_norm": 0.4075135588645935,
      "learning_rate": 0.0001,
      "loss": 1.8443,
      "step": 3426
    },
    {
      "epoch": 0.8322000971345314,
      "grad_norm": 0.3757880926132202,
      "learning_rate": 0.0001,
      "loss": 1.8096,
      "step": 3427
    },
    {
      "epoch": 0.832442933462846,
      "grad_norm": 0.4158633351325989,
      "learning_rate": 0.0001,
      "loss": 1.722,
      "step": 3428
    },
    {
      "epoch": 0.8326857697911607,
      "grad_norm": 0.33717289566993713,
      "learning_rate": 0.0001,
      "loss": 1.6177,
      "step": 3429
    },
    {
      "epoch": 0.8329286061194755,
      "grad_norm": 0.3926994204521179,
      "learning_rate": 0.0001,
      "loss": 1.7179,
      "step": 3430
    },
    {
      "epoch": 0.8331714424477902,
      "grad_norm": 0.3628498911857605,
      "learning_rate": 0.0001,
      "loss": 1.7354,
      "step": 3431
    },
    {
      "epoch": 0.8334142787761049,
      "grad_norm": 0.3821134567260742,
      "learning_rate": 0.0001,
      "loss": 1.6441,
      "step": 3432
    },
    {
      "epoch": 0.8336571151044196,
      "grad_norm": 0.3743181526660919,
      "learning_rate": 0.0001,
      "loss": 1.7191,
      "step": 3433
    },
    {
      "epoch": 0.8338999514327343,
      "grad_norm": 0.38392940163612366,
      "learning_rate": 0.0001,
      "loss": 1.8581,
      "step": 3434
    },
    {
      "epoch": 0.834142787761049,
      "grad_norm": 0.3931395411491394,
      "learning_rate": 0.0001,
      "loss": 1.7641,
      "step": 3435
    },
    {
      "epoch": 0.8343856240893638,
      "grad_norm": 0.3985886871814728,
      "learning_rate": 0.0001,
      "loss": 1.7006,
      "step": 3436
    },
    {
      "epoch": 0.8346284604176785,
      "grad_norm": 0.3533487617969513,
      "learning_rate": 0.0001,
      "loss": 1.6744,
      "step": 3437
    },
    {
      "epoch": 0.8348712967459933,
      "grad_norm": 0.3607969582080841,
      "learning_rate": 0.0001,
      "loss": 1.7189,
      "step": 3438
    },
    {
      "epoch": 0.8351141330743079,
      "grad_norm": 0.35621362924575806,
      "learning_rate": 0.0001,
      "loss": 1.5766,
      "step": 3439
    },
    {
      "epoch": 0.8353569694026226,
      "grad_norm": 0.3671231269836426,
      "learning_rate": 0.0001,
      "loss": 1.7925,
      "step": 3440
    },
    {
      "epoch": 0.8355998057309374,
      "grad_norm": 0.3626062572002411,
      "learning_rate": 0.0001,
      "loss": 1.5959,
      "step": 3441
    },
    {
      "epoch": 0.8358426420592521,
      "grad_norm": 0.3930570185184479,
      "learning_rate": 0.0001,
      "loss": 1.676,
      "step": 3442
    },
    {
      "epoch": 0.8360854783875667,
      "grad_norm": 0.3632255792617798,
      "learning_rate": 0.0001,
      "loss": 1.8025,
      "step": 3443
    },
    {
      "epoch": 0.8363283147158815,
      "grad_norm": 0.3855805993080139,
      "learning_rate": 0.0001,
      "loss": 1.6734,
      "step": 3444
    },
    {
      "epoch": 0.8365711510441962,
      "grad_norm": 0.36899393796920776,
      "learning_rate": 0.0001,
      "loss": 1.707,
      "step": 3445
    },
    {
      "epoch": 0.8368139873725109,
      "grad_norm": 0.39814889430999756,
      "learning_rate": 0.0001,
      "loss": 1.7721,
      "step": 3446
    },
    {
      "epoch": 0.8370568237008257,
      "grad_norm": 0.34774938225746155,
      "learning_rate": 0.0001,
      "loss": 1.581,
      "step": 3447
    },
    {
      "epoch": 0.8372996600291404,
      "grad_norm": 0.39525139331817627,
      "learning_rate": 0.0001,
      "loss": 1.8362,
      "step": 3448
    },
    {
      "epoch": 0.837542496357455,
      "grad_norm": 0.37141239643096924,
      "learning_rate": 0.0001,
      "loss": 1.7525,
      "step": 3449
    },
    {
      "epoch": 0.8377853326857698,
      "grad_norm": 0.38118383288383484,
      "learning_rate": 0.0001,
      "loss": 1.6791,
      "step": 3450
    },
    {
      "epoch": 0.8380281690140845,
      "grad_norm": 0.3800981640815735,
      "learning_rate": 0.0001,
      "loss": 1.6531,
      "step": 3451
    },
    {
      "epoch": 0.8382710053423992,
      "grad_norm": 0.38465872406959534,
      "learning_rate": 0.0001,
      "loss": 1.8196,
      "step": 3452
    },
    {
      "epoch": 0.838513841670714,
      "grad_norm": 0.33680155873298645,
      "learning_rate": 0.0001,
      "loss": 1.5088,
      "step": 3453
    },
    {
      "epoch": 0.8387566779990286,
      "grad_norm": 0.3530154228210449,
      "learning_rate": 0.0001,
      "loss": 1.7184,
      "step": 3454
    },
    {
      "epoch": 0.8389995143273433,
      "grad_norm": 0.3624134659767151,
      "learning_rate": 0.0001,
      "loss": 1.6819,
      "step": 3455
    },
    {
      "epoch": 0.8392423506556581,
      "grad_norm": 0.3938109874725342,
      "learning_rate": 0.0001,
      "loss": 1.625,
      "step": 3456
    },
    {
      "epoch": 0.8394851869839728,
      "grad_norm": 0.39321231842041016,
      "learning_rate": 0.0001,
      "loss": 1.9044,
      "step": 3457
    },
    {
      "epoch": 0.8397280233122875,
      "grad_norm": 0.36550790071487427,
      "learning_rate": 0.0001,
      "loss": 1.691,
      "step": 3458
    },
    {
      "epoch": 0.8399708596406023,
      "grad_norm": 0.3689191937446594,
      "learning_rate": 0.0001,
      "loss": 1.6638,
      "step": 3459
    },
    {
      "epoch": 0.8402136959689169,
      "grad_norm": 0.36174771189689636,
      "learning_rate": 0.0001,
      "loss": 1.7413,
      "step": 3460
    },
    {
      "epoch": 0.8404565322972317,
      "grad_norm": 0.37011727690696716,
      "learning_rate": 0.0001,
      "loss": 1.794,
      "step": 3461
    },
    {
      "epoch": 0.8406993686255464,
      "grad_norm": 0.37291502952575684,
      "learning_rate": 0.0001,
      "loss": 1.7135,
      "step": 3462
    },
    {
      "epoch": 0.8409422049538611,
      "grad_norm": 0.40054547786712646,
      "learning_rate": 0.0001,
      "loss": 1.651,
      "step": 3463
    },
    {
      "epoch": 0.8411850412821759,
      "grad_norm": 0.3493271768093109,
      "learning_rate": 0.0001,
      "loss": 1.702,
      "step": 3464
    },
    {
      "epoch": 0.8414278776104905,
      "grad_norm": 0.36861565709114075,
      "learning_rate": 0.0001,
      "loss": 1.5792,
      "step": 3465
    },
    {
      "epoch": 0.8416707139388052,
      "grad_norm": 0.37903013825416565,
      "learning_rate": 0.0001,
      "loss": 1.8688,
      "step": 3466
    },
    {
      "epoch": 0.84191355026712,
      "grad_norm": 0.36558017134666443,
      "learning_rate": 0.0001,
      "loss": 1.6547,
      "step": 3467
    },
    {
      "epoch": 0.8421563865954347,
      "grad_norm": 0.3424556255340576,
      "learning_rate": 0.0001,
      "loss": 1.528,
      "step": 3468
    },
    {
      "epoch": 0.8423992229237494,
      "grad_norm": 0.3787643611431122,
      "learning_rate": 0.0001,
      "loss": 1.7236,
      "step": 3469
    },
    {
      "epoch": 0.8426420592520641,
      "grad_norm": 0.3981590270996094,
      "learning_rate": 0.0001,
      "loss": 1.8502,
      "step": 3470
    },
    {
      "epoch": 0.8428848955803788,
      "grad_norm": 0.35316038131713867,
      "learning_rate": 0.0001,
      "loss": 1.5544,
      "step": 3471
    },
    {
      "epoch": 0.8431277319086935,
      "grad_norm": 0.36104702949523926,
      "learning_rate": 0.0001,
      "loss": 1.6113,
      "step": 3472
    },
    {
      "epoch": 0.8433705682370083,
      "grad_norm": 0.3761642277240753,
      "learning_rate": 0.0001,
      "loss": 1.7298,
      "step": 3473
    },
    {
      "epoch": 0.843613404565323,
      "grad_norm": 0.3798404932022095,
      "learning_rate": 0.0001,
      "loss": 1.7458,
      "step": 3474
    },
    {
      "epoch": 0.8438562408936376,
      "grad_norm": 0.36524298787117004,
      "learning_rate": 0.0001,
      "loss": 1.724,
      "step": 3475
    },
    {
      "epoch": 0.8440990772219524,
      "grad_norm": 0.3702562749385834,
      "learning_rate": 0.0001,
      "loss": 1.8075,
      "step": 3476
    },
    {
      "epoch": 0.8443419135502671,
      "grad_norm": 0.369831383228302,
      "learning_rate": 0.0001,
      "loss": 1.6562,
      "step": 3477
    },
    {
      "epoch": 0.8445847498785818,
      "grad_norm": 0.35366007685661316,
      "learning_rate": 0.0001,
      "loss": 1.6473,
      "step": 3478
    },
    {
      "epoch": 0.8448275862068966,
      "grad_norm": 0.3676561117172241,
      "learning_rate": 0.0001,
      "loss": 1.7891,
      "step": 3479
    },
    {
      "epoch": 0.8450704225352113,
      "grad_norm": 0.36195895075798035,
      "learning_rate": 0.0001,
      "loss": 1.6914,
      "step": 3480
    },
    {
      "epoch": 0.8453132588635259,
      "grad_norm": 0.36972975730895996,
      "learning_rate": 0.0001,
      "loss": 1.7172,
      "step": 3481
    },
    {
      "epoch": 0.8455560951918407,
      "grad_norm": 0.37382155656814575,
      "learning_rate": 0.0001,
      "loss": 1.6309,
      "step": 3482
    },
    {
      "epoch": 0.8457989315201554,
      "grad_norm": 0.3648379445075989,
      "learning_rate": 0.0001,
      "loss": 1.7619,
      "step": 3483
    },
    {
      "epoch": 0.8460417678484702,
      "grad_norm": 0.36794230341911316,
      "learning_rate": 0.0001,
      "loss": 1.6669,
      "step": 3484
    },
    {
      "epoch": 0.8462846041767849,
      "grad_norm": 0.3614242970943451,
      "learning_rate": 0.0001,
      "loss": 1.6528,
      "step": 3485
    },
    {
      "epoch": 0.8465274405050995,
      "grad_norm": 0.3677575886249542,
      "learning_rate": 0.0001,
      "loss": 1.7447,
      "step": 3486
    },
    {
      "epoch": 0.8467702768334143,
      "grad_norm": 0.371786504983902,
      "learning_rate": 0.0001,
      "loss": 1.5776,
      "step": 3487
    },
    {
      "epoch": 0.847013113161729,
      "grad_norm": 0.36296510696411133,
      "learning_rate": 0.0001,
      "loss": 1.6704,
      "step": 3488
    },
    {
      "epoch": 0.8472559494900437,
      "grad_norm": 0.3505253791809082,
      "learning_rate": 0.0001,
      "loss": 1.7259,
      "step": 3489
    },
    {
      "epoch": 0.8474987858183585,
      "grad_norm": 0.34446287155151367,
      "learning_rate": 0.0001,
      "loss": 1.504,
      "step": 3490
    },
    {
      "epoch": 0.8477416221466731,
      "grad_norm": 0.362392783164978,
      "learning_rate": 0.0001,
      "loss": 1.5495,
      "step": 3491
    },
    {
      "epoch": 0.8479844584749878,
      "grad_norm": 0.3405704200267792,
      "learning_rate": 0.0001,
      "loss": 1.5627,
      "step": 3492
    },
    {
      "epoch": 0.8482272948033026,
      "grad_norm": 0.36630943417549133,
      "learning_rate": 0.0001,
      "loss": 1.6703,
      "step": 3493
    },
    {
      "epoch": 0.8484701311316173,
      "grad_norm": 0.35960787534713745,
      "learning_rate": 0.0001,
      "loss": 1.7689,
      "step": 3494
    },
    {
      "epoch": 0.848712967459932,
      "grad_norm": 0.35360845923423767,
      "learning_rate": 0.0001,
      "loss": 1.6148,
      "step": 3495
    },
    {
      "epoch": 0.8489558037882468,
      "grad_norm": 0.3669142723083496,
      "learning_rate": 0.0001,
      "loss": 1.683,
      "step": 3496
    },
    {
      "epoch": 0.8491986401165614,
      "grad_norm": 0.36652398109436035,
      "learning_rate": 0.0001,
      "loss": 1.6641,
      "step": 3497
    },
    {
      "epoch": 0.8494414764448761,
      "grad_norm": 0.4173559844493866,
      "learning_rate": 0.0001,
      "loss": 1.846,
      "step": 3498
    },
    {
      "epoch": 0.8496843127731909,
      "grad_norm": 0.37224480509757996,
      "learning_rate": 0.0001,
      "loss": 1.7263,
      "step": 3499
    },
    {
      "epoch": 0.8499271491015056,
      "grad_norm": 0.35871750116348267,
      "learning_rate": 0.0001,
      "loss": 1.5461,
      "step": 3500
    },
    {
      "epoch": 0.8501699854298203,
      "grad_norm": 0.37326905131340027,
      "learning_rate": 0.0001,
      "loss": 1.7384,
      "step": 3501
    },
    {
      "epoch": 0.850412821758135,
      "grad_norm": 0.4032769799232483,
      "learning_rate": 0.0001,
      "loss": 1.7194,
      "step": 3502
    },
    {
      "epoch": 0.8506556580864497,
      "grad_norm": 0.354282945394516,
      "learning_rate": 0.0001,
      "loss": 1.6118,
      "step": 3503
    },
    {
      "epoch": 0.8508984944147644,
      "grad_norm": 0.3966940939426422,
      "learning_rate": 0.0001,
      "loss": 1.7711,
      "step": 3504
    },
    {
      "epoch": 0.8511413307430792,
      "grad_norm": 0.37445423007011414,
      "learning_rate": 0.0001,
      "loss": 1.6806,
      "step": 3505
    },
    {
      "epoch": 0.8513841670713939,
      "grad_norm": 0.35755404829978943,
      "learning_rate": 0.0001,
      "loss": 1.5763,
      "step": 3506
    },
    {
      "epoch": 0.8516270033997086,
      "grad_norm": 0.407652348279953,
      "learning_rate": 0.0001,
      "loss": 1.6067,
      "step": 3507
    },
    {
      "epoch": 0.8518698397280233,
      "grad_norm": 0.37608498334884644,
      "learning_rate": 0.0001,
      "loss": 1.8628,
      "step": 3508
    },
    {
      "epoch": 0.852112676056338,
      "grad_norm": 0.37260276079177856,
      "learning_rate": 0.0001,
      "loss": 1.6454,
      "step": 3509
    },
    {
      "epoch": 0.8523555123846528,
      "grad_norm": 0.38507553935050964,
      "learning_rate": 0.0001,
      "loss": 1.7672,
      "step": 3510
    },
    {
      "epoch": 0.8525983487129675,
      "grad_norm": 0.3733923137187958,
      "learning_rate": 0.0001,
      "loss": 1.6671,
      "step": 3511
    },
    {
      "epoch": 0.8528411850412821,
      "grad_norm": 0.3613528609275818,
      "learning_rate": 0.0001,
      "loss": 1.6734,
      "step": 3512
    },
    {
      "epoch": 0.8530840213695969,
      "grad_norm": 0.3519417345523834,
      "learning_rate": 0.0001,
      "loss": 1.7748,
      "step": 3513
    },
    {
      "epoch": 0.8533268576979116,
      "grad_norm": 0.37620311975479126,
      "learning_rate": 0.0001,
      "loss": 1.6402,
      "step": 3514
    },
    {
      "epoch": 0.8535696940262263,
      "grad_norm": 0.32235080003738403,
      "learning_rate": 0.0001,
      "loss": 1.5878,
      "step": 3515
    },
    {
      "epoch": 0.8538125303545411,
      "grad_norm": 0.3643632233142853,
      "learning_rate": 0.0001,
      "loss": 1.5403,
      "step": 3516
    },
    {
      "epoch": 0.8540553666828558,
      "grad_norm": 0.37543153762817383,
      "learning_rate": 0.0001,
      "loss": 1.8312,
      "step": 3517
    },
    {
      "epoch": 0.8542982030111704,
      "grad_norm": 0.36344969272613525,
      "learning_rate": 0.0001,
      "loss": 1.8156,
      "step": 3518
    },
    {
      "epoch": 0.8545410393394852,
      "grad_norm": 0.3584924340248108,
      "learning_rate": 0.0001,
      "loss": 1.631,
      "step": 3519
    },
    {
      "epoch": 0.8547838756677999,
      "grad_norm": 0.3910501301288605,
      "learning_rate": 0.0001,
      "loss": 1.6077,
      "step": 3520
    },
    {
      "epoch": 0.8550267119961146,
      "grad_norm": 0.3873438239097595,
      "learning_rate": 0.0001,
      "loss": 1.7625,
      "step": 3521
    },
    {
      "epoch": 0.8552695483244294,
      "grad_norm": 0.37016668915748596,
      "learning_rate": 0.0001,
      "loss": 1.6542,
      "step": 3522
    },
    {
      "epoch": 0.855512384652744,
      "grad_norm": 0.34312549233436584,
      "learning_rate": 0.0001,
      "loss": 1.5476,
      "step": 3523
    },
    {
      "epoch": 0.8557552209810587,
      "grad_norm": 0.3576250672340393,
      "learning_rate": 0.0001,
      "loss": 1.7494,
      "step": 3524
    },
    {
      "epoch": 0.8559980573093735,
      "grad_norm": 0.39972227811813354,
      "learning_rate": 0.0001,
      "loss": 1.7725,
      "step": 3525
    },
    {
      "epoch": 0.8562408936376882,
      "grad_norm": 0.41491860151290894,
      "learning_rate": 0.0001,
      "loss": 1.9628,
      "step": 3526
    },
    {
      "epoch": 0.8564837299660029,
      "grad_norm": 0.35856276750564575,
      "learning_rate": 0.0001,
      "loss": 1.6327,
      "step": 3527
    },
    {
      "epoch": 0.8567265662943176,
      "grad_norm": 0.3800105154514313,
      "learning_rate": 0.0001,
      "loss": 1.6284,
      "step": 3528
    },
    {
      "epoch": 0.8569694026226323,
      "grad_norm": 0.36823752522468567,
      "learning_rate": 0.0001,
      "loss": 1.7279,
      "step": 3529
    },
    {
      "epoch": 0.8572122389509471,
      "grad_norm": 0.40262046456336975,
      "learning_rate": 0.0001,
      "loss": 1.8158,
      "step": 3530
    },
    {
      "epoch": 0.8574550752792618,
      "grad_norm": 0.35609057545661926,
      "learning_rate": 0.0001,
      "loss": 1.7015,
      "step": 3531
    },
    {
      "epoch": 0.8576979116075765,
      "grad_norm": 0.35543057322502136,
      "learning_rate": 0.0001,
      "loss": 1.5971,
      "step": 3532
    },
    {
      "epoch": 0.8579407479358913,
      "grad_norm": 0.3998347520828247,
      "learning_rate": 0.0001,
      "loss": 1.755,
      "step": 3533
    },
    {
      "epoch": 0.8581835842642059,
      "grad_norm": 0.35681387782096863,
      "learning_rate": 0.0001,
      "loss": 1.639,
      "step": 3534
    },
    {
      "epoch": 0.8584264205925206,
      "grad_norm": 0.35703030228614807,
      "learning_rate": 0.0001,
      "loss": 1.6203,
      "step": 3535
    },
    {
      "epoch": 0.8586692569208354,
      "grad_norm": 0.39074602723121643,
      "learning_rate": 0.0001,
      "loss": 1.7393,
      "step": 3536
    },
    {
      "epoch": 0.8589120932491501,
      "grad_norm": 0.37359052896499634,
      "learning_rate": 0.0001,
      "loss": 1.7177,
      "step": 3537
    },
    {
      "epoch": 0.8591549295774648,
      "grad_norm": 0.3567638397216797,
      "learning_rate": 0.0001,
      "loss": 1.6187,
      "step": 3538
    },
    {
      "epoch": 0.8593977659057795,
      "grad_norm": 0.3632083535194397,
      "learning_rate": 0.0001,
      "loss": 1.8805,
      "step": 3539
    },
    {
      "epoch": 0.8596406022340942,
      "grad_norm": 0.36028143763542175,
      "learning_rate": 0.0001,
      "loss": 1.7016,
      "step": 3540
    },
    {
      "epoch": 0.8598834385624089,
      "grad_norm": 0.3658236563205719,
      "learning_rate": 0.0001,
      "loss": 1.8488,
      "step": 3541
    },
    {
      "epoch": 0.8601262748907237,
      "grad_norm": 0.3709207773208618,
      "learning_rate": 0.0001,
      "loss": 1.553,
      "step": 3542
    },
    {
      "epoch": 0.8603691112190384,
      "grad_norm": 0.3640122711658478,
      "learning_rate": 0.0001,
      "loss": 1.6917,
      "step": 3543
    },
    {
      "epoch": 0.860611947547353,
      "grad_norm": 0.36741796135902405,
      "learning_rate": 0.0001,
      "loss": 1.7421,
      "step": 3544
    },
    {
      "epoch": 0.8608547838756678,
      "grad_norm": 0.36788538098335266,
      "learning_rate": 0.0001,
      "loss": 1.5018,
      "step": 3545
    },
    {
      "epoch": 0.8610976202039825,
      "grad_norm": 0.3530402183532715,
      "learning_rate": 0.0001,
      "loss": 1.6676,
      "step": 3546
    },
    {
      "epoch": 0.8613404565322972,
      "grad_norm": 0.39383524656295776,
      "learning_rate": 0.0001,
      "loss": 1.7636,
      "step": 3547
    },
    {
      "epoch": 0.861583292860612,
      "grad_norm": 0.3982522189617157,
      "learning_rate": 0.0001,
      "loss": 1.7767,
      "step": 3548
    },
    {
      "epoch": 0.8618261291889266,
      "grad_norm": 0.3580620586872101,
      "learning_rate": 0.0001,
      "loss": 1.597,
      "step": 3549
    },
    {
      "epoch": 0.8620689655172413,
      "grad_norm": 0.35779640078544617,
      "learning_rate": 0.0001,
      "loss": 1.6438,
      "step": 3550
    },
    {
      "epoch": 0.8623118018455561,
      "grad_norm": 0.3500531017780304,
      "learning_rate": 0.0001,
      "loss": 1.6659,
      "step": 3551
    },
    {
      "epoch": 0.8625546381738708,
      "grad_norm": 0.34910285472869873,
      "learning_rate": 0.0001,
      "loss": 1.658,
      "step": 3552
    },
    {
      "epoch": 0.8627974745021856,
      "grad_norm": 0.35888442397117615,
      "learning_rate": 0.0001,
      "loss": 1.5498,
      "step": 3553
    },
    {
      "epoch": 0.8630403108305003,
      "grad_norm": 0.38242191076278687,
      "learning_rate": 0.0001,
      "loss": 1.6856,
      "step": 3554
    },
    {
      "epoch": 0.8632831471588149,
      "grad_norm": 0.3862399160861969,
      "learning_rate": 0.0001,
      "loss": 1.7477,
      "step": 3555
    },
    {
      "epoch": 0.8635259834871297,
      "grad_norm": 0.3789171576499939,
      "learning_rate": 0.0001,
      "loss": 1.6583,
      "step": 3556
    },
    {
      "epoch": 0.8637688198154444,
      "grad_norm": 0.3593288064002991,
      "learning_rate": 0.0001,
      "loss": 1.5699,
      "step": 3557
    },
    {
      "epoch": 0.8640116561437591,
      "grad_norm": 0.3673674762248993,
      "learning_rate": 0.0001,
      "loss": 1.7593,
      "step": 3558
    },
    {
      "epoch": 0.8642544924720739,
      "grad_norm": 0.38266125321388245,
      "learning_rate": 0.0001,
      "loss": 1.6827,
      "step": 3559
    },
    {
      "epoch": 0.8644973288003885,
      "grad_norm": 0.39288368821144104,
      "learning_rate": 0.0001,
      "loss": 1.8141,
      "step": 3560
    },
    {
      "epoch": 0.8647401651287032,
      "grad_norm": 0.3890032172203064,
      "learning_rate": 0.0001,
      "loss": 1.5849,
      "step": 3561
    },
    {
      "epoch": 0.864983001457018,
      "grad_norm": 0.36325764656066895,
      "learning_rate": 0.0001,
      "loss": 1.6591,
      "step": 3562
    },
    {
      "epoch": 0.8652258377853327,
      "grad_norm": 0.37350988388061523,
      "learning_rate": 0.0001,
      "loss": 1.7085,
      "step": 3563
    },
    {
      "epoch": 0.8654686741136474,
      "grad_norm": 0.3759698271751404,
      "learning_rate": 0.0001,
      "loss": 1.6911,
      "step": 3564
    },
    {
      "epoch": 0.8657115104419622,
      "grad_norm": 0.39357560873031616,
      "learning_rate": 0.0001,
      "loss": 1.7312,
      "step": 3565
    },
    {
      "epoch": 0.8659543467702768,
      "grad_norm": 0.37965479493141174,
      "learning_rate": 0.0001,
      "loss": 1.808,
      "step": 3566
    },
    {
      "epoch": 0.8661971830985915,
      "grad_norm": 0.3746955096721649,
      "learning_rate": 0.0001,
      "loss": 1.7676,
      "step": 3567
    },
    {
      "epoch": 0.8664400194269063,
      "grad_norm": 0.38889482617378235,
      "learning_rate": 0.0001,
      "loss": 1.7351,
      "step": 3568
    },
    {
      "epoch": 0.866682855755221,
      "grad_norm": 0.3991025388240814,
      "learning_rate": 0.0001,
      "loss": 1.8694,
      "step": 3569
    },
    {
      "epoch": 0.8669256920835356,
      "grad_norm": 0.3674953579902649,
      "learning_rate": 0.0001,
      "loss": 1.6615,
      "step": 3570
    },
    {
      "epoch": 0.8671685284118504,
      "grad_norm": 0.3669956922531128,
      "learning_rate": 0.0001,
      "loss": 1.8463,
      "step": 3571
    },
    {
      "epoch": 0.8674113647401651,
      "grad_norm": 0.3681383728981018,
      "learning_rate": 0.0001,
      "loss": 1.6519,
      "step": 3572
    },
    {
      "epoch": 0.8676542010684798,
      "grad_norm": 0.37041378021240234,
      "learning_rate": 0.0001,
      "loss": 1.8967,
      "step": 3573
    },
    {
      "epoch": 0.8678970373967946,
      "grad_norm": 0.34929654002189636,
      "learning_rate": 0.0001,
      "loss": 1.4997,
      "step": 3574
    },
    {
      "epoch": 0.8681398737251093,
      "grad_norm": 0.3579554855823517,
      "learning_rate": 0.0001,
      "loss": 1.6965,
      "step": 3575
    },
    {
      "epoch": 0.868382710053424,
      "grad_norm": 0.3582157492637634,
      "learning_rate": 0.0001,
      "loss": 1.6105,
      "step": 3576
    },
    {
      "epoch": 0.8686255463817387,
      "grad_norm": 0.37660810351371765,
      "learning_rate": 0.0001,
      "loss": 1.8288,
      "step": 3577
    },
    {
      "epoch": 0.8688683827100534,
      "grad_norm": 0.36132362484931946,
      "learning_rate": 0.0001,
      "loss": 1.5915,
      "step": 3578
    },
    {
      "epoch": 0.8691112190383682,
      "grad_norm": 0.37348681688308716,
      "learning_rate": 0.0001,
      "loss": 1.6533,
      "step": 3579
    },
    {
      "epoch": 0.8693540553666829,
      "grad_norm": 0.38948729634284973,
      "learning_rate": 0.0001,
      "loss": 1.7371,
      "step": 3580
    },
    {
      "epoch": 0.8695968916949975,
      "grad_norm": 0.35966864228248596,
      "learning_rate": 0.0001,
      "loss": 1.4848,
      "step": 3581
    },
    {
      "epoch": 0.8698397280233123,
      "grad_norm": 0.37547367811203003,
      "learning_rate": 0.0001,
      "loss": 1.8124,
      "step": 3582
    },
    {
      "epoch": 0.870082564351627,
      "grad_norm": 0.35957401990890503,
      "learning_rate": 0.0001,
      "loss": 1.609,
      "step": 3583
    },
    {
      "epoch": 0.8703254006799417,
      "grad_norm": 0.36093026399612427,
      "learning_rate": 0.0001,
      "loss": 1.7151,
      "step": 3584
    },
    {
      "epoch": 0.8705682370082565,
      "grad_norm": 0.3634076714515686,
      "learning_rate": 0.0001,
      "loss": 1.652,
      "step": 3585
    },
    {
      "epoch": 0.8708110733365712,
      "grad_norm": 0.3747778534889221,
      "learning_rate": 0.0001,
      "loss": 1.7686,
      "step": 3586
    },
    {
      "epoch": 0.8710539096648858,
      "grad_norm": 0.3573356568813324,
      "learning_rate": 0.0001,
      "loss": 1.7262,
      "step": 3587
    },
    {
      "epoch": 0.8712967459932006,
      "grad_norm": 0.3708506226539612,
      "learning_rate": 0.0001,
      "loss": 1.7808,
      "step": 3588
    },
    {
      "epoch": 0.8715395823215153,
      "grad_norm": 0.3369008004665375,
      "learning_rate": 0.0001,
      "loss": 1.603,
      "step": 3589
    },
    {
      "epoch": 0.87178241864983,
      "grad_norm": 0.37004029750823975,
      "learning_rate": 0.0001,
      "loss": 1.6321,
      "step": 3590
    },
    {
      "epoch": 0.8720252549781448,
      "grad_norm": 0.3614467680454254,
      "learning_rate": 0.0001,
      "loss": 1.5722,
      "step": 3591
    },
    {
      "epoch": 0.8722680913064594,
      "grad_norm": 0.3672022223472595,
      "learning_rate": 0.0001,
      "loss": 1.7806,
      "step": 3592
    },
    {
      "epoch": 0.8725109276347741,
      "grad_norm": 0.37763553857803345,
      "learning_rate": 0.0001,
      "loss": 1.6137,
      "step": 3593
    },
    {
      "epoch": 0.8727537639630889,
      "grad_norm": 0.3805335760116577,
      "learning_rate": 0.0001,
      "loss": 1.7193,
      "step": 3594
    },
    {
      "epoch": 0.8729966002914036,
      "grad_norm": 0.3677334785461426,
      "learning_rate": 0.0001,
      "loss": 1.7612,
      "step": 3595
    },
    {
      "epoch": 0.8732394366197183,
      "grad_norm": 0.36495909094810486,
      "learning_rate": 0.0001,
      "loss": 1.6714,
      "step": 3596
    },
    {
      "epoch": 0.873482272948033,
      "grad_norm": 0.37761765718460083,
      "learning_rate": 0.0001,
      "loss": 1.7882,
      "step": 3597
    },
    {
      "epoch": 0.8737251092763477,
      "grad_norm": 0.3680744767189026,
      "learning_rate": 0.0001,
      "loss": 1.6574,
      "step": 3598
    },
    {
      "epoch": 0.8739679456046625,
      "grad_norm": 0.35450413823127747,
      "learning_rate": 0.0001,
      "loss": 1.6302,
      "step": 3599
    },
    {
      "epoch": 0.8742107819329772,
      "grad_norm": 0.36951565742492676,
      "learning_rate": 0.0001,
      "loss": 1.7608,
      "step": 3600
    },
    {
      "epoch": 0.8744536182612919,
      "grad_norm": 0.34576210379600525,
      "learning_rate": 0.0001,
      "loss": 1.5916,
      "step": 3601
    },
    {
      "epoch": 0.8746964545896067,
      "grad_norm": 0.34308797121047974,
      "learning_rate": 0.0001,
      "loss": 1.5128,
      "step": 3602
    },
    {
      "epoch": 0.8749392909179213,
      "grad_norm": 0.3728722333908081,
      "learning_rate": 0.0001,
      "loss": 1.6882,
      "step": 3603
    },
    {
      "epoch": 0.875182127246236,
      "grad_norm": 0.3887479603290558,
      "learning_rate": 0.0001,
      "loss": 1.6687,
      "step": 3604
    },
    {
      "epoch": 0.8754249635745508,
      "grad_norm": 0.37138471007347107,
      "learning_rate": 0.0001,
      "loss": 1.6479,
      "step": 3605
    },
    {
      "epoch": 0.8756677999028655,
      "grad_norm": 0.3760085701942444,
      "learning_rate": 0.0001,
      "loss": 1.7381,
      "step": 3606
    },
    {
      "epoch": 0.8759106362311802,
      "grad_norm": 0.3597986698150635,
      "learning_rate": 0.0001,
      "loss": 1.5603,
      "step": 3607
    },
    {
      "epoch": 0.8761534725594949,
      "grad_norm": 0.3942025899887085,
      "learning_rate": 0.0001,
      "loss": 1.8538,
      "step": 3608
    },
    {
      "epoch": 0.8763963088878096,
      "grad_norm": 0.3929956257343292,
      "learning_rate": 0.0001,
      "loss": 1.6928,
      "step": 3609
    },
    {
      "epoch": 0.8766391452161243,
      "grad_norm": 0.415682315826416,
      "learning_rate": 0.0001,
      "loss": 1.584,
      "step": 3610
    },
    {
      "epoch": 0.8768819815444391,
      "grad_norm": 0.3906037211418152,
      "learning_rate": 0.0001,
      "loss": 1.8822,
      "step": 3611
    },
    {
      "epoch": 0.8771248178727538,
      "grad_norm": 0.3817678391933441,
      "learning_rate": 0.0001,
      "loss": 1.6585,
      "step": 3612
    },
    {
      "epoch": 0.8773676542010684,
      "grad_norm": 0.37527331709861755,
      "learning_rate": 0.0001,
      "loss": 1.8332,
      "step": 3613
    },
    {
      "epoch": 0.8776104905293832,
      "grad_norm": 0.34259480237960815,
      "learning_rate": 0.0001,
      "loss": 1.6764,
      "step": 3614
    },
    {
      "epoch": 0.8778533268576979,
      "grad_norm": 0.37580329179763794,
      "learning_rate": 0.0001,
      "loss": 1.5702,
      "step": 3615
    },
    {
      "epoch": 0.8780961631860126,
      "grad_norm": 0.38391757011413574,
      "learning_rate": 0.0001,
      "loss": 1.7783,
      "step": 3616
    },
    {
      "epoch": 0.8783389995143274,
      "grad_norm": 0.36990684270858765,
      "learning_rate": 0.0001,
      "loss": 1.7667,
      "step": 3617
    },
    {
      "epoch": 0.878581835842642,
      "grad_norm": 0.37066319584846497,
      "learning_rate": 0.0001,
      "loss": 1.452,
      "step": 3618
    },
    {
      "epoch": 0.8788246721709567,
      "grad_norm": 0.3481326401233673,
      "learning_rate": 0.0001,
      "loss": 1.6822,
      "step": 3619
    },
    {
      "epoch": 0.8790675084992715,
      "grad_norm": 0.40429386496543884,
      "learning_rate": 0.0001,
      "loss": 1.7546,
      "step": 3620
    },
    {
      "epoch": 0.8793103448275862,
      "grad_norm": 0.40886256098747253,
      "learning_rate": 0.0001,
      "loss": 1.7198,
      "step": 3621
    },
    {
      "epoch": 0.879553181155901,
      "grad_norm": 0.38799768686294556,
      "learning_rate": 0.0001,
      "loss": 1.9331,
      "step": 3622
    },
    {
      "epoch": 0.8797960174842157,
      "grad_norm": 0.37042146921157837,
      "learning_rate": 0.0001,
      "loss": 1.6779,
      "step": 3623
    },
    {
      "epoch": 0.8800388538125303,
      "grad_norm": 0.3842093348503113,
      "learning_rate": 0.0001,
      "loss": 1.6768,
      "step": 3624
    },
    {
      "epoch": 0.8802816901408451,
      "grad_norm": 0.39390814304351807,
      "learning_rate": 0.0001,
      "loss": 1.7565,
      "step": 3625
    },
    {
      "epoch": 0.8805245264691598,
      "grad_norm": 0.3362022340297699,
      "learning_rate": 0.0001,
      "loss": 1.4937,
      "step": 3626
    },
    {
      "epoch": 0.8807673627974745,
      "grad_norm": 0.36625936627388,
      "learning_rate": 0.0001,
      "loss": 1.7513,
      "step": 3627
    },
    {
      "epoch": 0.8810101991257893,
      "grad_norm": 0.3637999892234802,
      "learning_rate": 0.0001,
      "loss": 1.5824,
      "step": 3628
    },
    {
      "epoch": 0.8812530354541039,
      "grad_norm": 0.3684556484222412,
      "learning_rate": 0.0001,
      "loss": 1.742,
      "step": 3629
    },
    {
      "epoch": 0.8814958717824186,
      "grad_norm": 0.3652805685997009,
      "learning_rate": 0.0001,
      "loss": 1.4867,
      "step": 3630
    },
    {
      "epoch": 0.8817387081107334,
      "grad_norm": 0.3626108765602112,
      "learning_rate": 0.0001,
      "loss": 1.5675,
      "step": 3631
    },
    {
      "epoch": 0.8819815444390481,
      "grad_norm": 0.40082186460494995,
      "learning_rate": 0.0001,
      "loss": 1.6533,
      "step": 3632
    },
    {
      "epoch": 0.8822243807673628,
      "grad_norm": 0.34865519404411316,
      "learning_rate": 0.0001,
      "loss": 1.5657,
      "step": 3633
    },
    {
      "epoch": 0.8824672170956775,
      "grad_norm": 0.3757842779159546,
      "learning_rate": 0.0001,
      "loss": 1.7309,
      "step": 3634
    },
    {
      "epoch": 0.8827100534239922,
      "grad_norm": 0.36966630816459656,
      "learning_rate": 0.0001,
      "loss": 1.7184,
      "step": 3635
    },
    {
      "epoch": 0.8829528897523069,
      "grad_norm": 0.409877210855484,
      "learning_rate": 0.0001,
      "loss": 1.8053,
      "step": 3636
    },
    {
      "epoch": 0.8831957260806217,
      "grad_norm": 0.3520601689815521,
      "learning_rate": 0.0001,
      "loss": 1.4993,
      "step": 3637
    },
    {
      "epoch": 0.8834385624089364,
      "grad_norm": 0.3983246088027954,
      "learning_rate": 0.0001,
      "loss": 1.7949,
      "step": 3638
    },
    {
      "epoch": 0.883681398737251,
      "grad_norm": 0.3668357729911804,
      "learning_rate": 0.0001,
      "loss": 1.7983,
      "step": 3639
    },
    {
      "epoch": 0.8839242350655658,
      "grad_norm": 0.3689560294151306,
      "learning_rate": 0.0001,
      "loss": 1.806,
      "step": 3640
    },
    {
      "epoch": 0.8841670713938805,
      "grad_norm": 0.37663447856903076,
      "learning_rate": 0.0001,
      "loss": 1.7754,
      "step": 3641
    },
    {
      "epoch": 0.8844099077221952,
      "grad_norm": 0.3643248677253723,
      "learning_rate": 0.0001,
      "loss": 1.7468,
      "step": 3642
    },
    {
      "epoch": 0.88465274405051,
      "grad_norm": 0.36683642864227295,
      "learning_rate": 0.0001,
      "loss": 1.815,
      "step": 3643
    },
    {
      "epoch": 0.8848955803788247,
      "grad_norm": 0.3876796364784241,
      "learning_rate": 0.0001,
      "loss": 1.7493,
      "step": 3644
    },
    {
      "epoch": 0.8851384167071394,
      "grad_norm": 0.3821657598018646,
      "learning_rate": 0.0001,
      "loss": 1.6686,
      "step": 3645
    },
    {
      "epoch": 0.8853812530354541,
      "grad_norm": 0.3620165288448334,
      "learning_rate": 0.0001,
      "loss": 1.57,
      "step": 3646
    },
    {
      "epoch": 0.8856240893637688,
      "grad_norm": 0.3420584499835968,
      "learning_rate": 0.0001,
      "loss": 1.5326,
      "step": 3647
    },
    {
      "epoch": 0.8858669256920836,
      "grad_norm": 0.37859800457954407,
      "learning_rate": 0.0001,
      "loss": 1.7683,
      "step": 3648
    },
    {
      "epoch": 0.8861097620203983,
      "grad_norm": 0.369691401720047,
      "learning_rate": 0.0001,
      "loss": 1.7959,
      "step": 3649
    },
    {
      "epoch": 0.8863525983487129,
      "grad_norm": 0.37752020359039307,
      "learning_rate": 0.0001,
      "loss": 1.6003,
      "step": 3650
    },
    {
      "epoch": 0.8865954346770277,
      "grad_norm": 0.3806999921798706,
      "learning_rate": 0.0001,
      "loss": 1.8619,
      "step": 3651
    },
    {
      "epoch": 0.8868382710053424,
      "grad_norm": 0.3760840594768524,
      "learning_rate": 0.0001,
      "loss": 1.7124,
      "step": 3652
    },
    {
      "epoch": 0.8870811073336571,
      "grad_norm": 0.3627361059188843,
      "learning_rate": 0.0001,
      "loss": 1.7258,
      "step": 3653
    },
    {
      "epoch": 0.8873239436619719,
      "grad_norm": 0.3710281252861023,
      "learning_rate": 0.0001,
      "loss": 1.6779,
      "step": 3654
    },
    {
      "epoch": 0.8875667799902865,
      "grad_norm": 0.36439019441604614,
      "learning_rate": 0.0001,
      "loss": 1.8435,
      "step": 3655
    },
    {
      "epoch": 0.8878096163186012,
      "grad_norm": 0.385211318731308,
      "learning_rate": 0.0001,
      "loss": 1.8635,
      "step": 3656
    },
    {
      "epoch": 0.888052452646916,
      "grad_norm": 0.3514195382595062,
      "learning_rate": 0.0001,
      "loss": 1.5961,
      "step": 3657
    },
    {
      "epoch": 0.8882952889752307,
      "grad_norm": 0.3596414625644684,
      "learning_rate": 0.0001,
      "loss": 1.6581,
      "step": 3658
    },
    {
      "epoch": 0.8885381253035454,
      "grad_norm": 0.37253424525260925,
      "learning_rate": 0.0001,
      "loss": 1.7471,
      "step": 3659
    },
    {
      "epoch": 0.8887809616318602,
      "grad_norm": 0.3801403343677521,
      "learning_rate": 0.0001,
      "loss": 1.7299,
      "step": 3660
    },
    {
      "epoch": 0.8890237979601748,
      "grad_norm": 0.3856620192527771,
      "learning_rate": 0.0001,
      "loss": 1.7593,
      "step": 3661
    },
    {
      "epoch": 0.8892666342884895,
      "grad_norm": 0.3800942301750183,
      "learning_rate": 0.0001,
      "loss": 1.7336,
      "step": 3662
    },
    {
      "epoch": 0.8895094706168043,
      "grad_norm": 0.3820795714855194,
      "learning_rate": 0.0001,
      "loss": 1.8671,
      "step": 3663
    },
    {
      "epoch": 0.889752306945119,
      "grad_norm": 0.3742009103298187,
      "learning_rate": 0.0001,
      "loss": 1.7299,
      "step": 3664
    },
    {
      "epoch": 0.8899951432734337,
      "grad_norm": 0.33877500891685486,
      "learning_rate": 0.0001,
      "loss": 1.6424,
      "step": 3665
    },
    {
      "epoch": 0.8902379796017484,
      "grad_norm": 0.35122042894363403,
      "learning_rate": 0.0001,
      "loss": 1.5245,
      "step": 3666
    },
    {
      "epoch": 0.8904808159300631,
      "grad_norm": 0.345304936170578,
      "learning_rate": 0.0001,
      "loss": 1.5726,
      "step": 3667
    },
    {
      "epoch": 0.8907236522583779,
      "grad_norm": 0.3724261522293091,
      "learning_rate": 0.0001,
      "loss": 1.6823,
      "step": 3668
    },
    {
      "epoch": 0.8909664885866926,
      "grad_norm": 0.3617289066314697,
      "learning_rate": 0.0001,
      "loss": 1.6174,
      "step": 3669
    },
    {
      "epoch": 0.8912093249150073,
      "grad_norm": 0.36587101221084595,
      "learning_rate": 0.0001,
      "loss": 1.6748,
      "step": 3670
    },
    {
      "epoch": 0.891452161243322,
      "grad_norm": 0.3875347673892975,
      "learning_rate": 0.0001,
      "loss": 1.8046,
      "step": 3671
    },
    {
      "epoch": 0.8916949975716367,
      "grad_norm": 0.3689577579498291,
      "learning_rate": 0.0001,
      "loss": 1.6712,
      "step": 3672
    },
    {
      "epoch": 0.8919378338999514,
      "grad_norm": 0.3996396064758301,
      "learning_rate": 0.0001,
      "loss": 1.7927,
      "step": 3673
    },
    {
      "epoch": 0.8921806702282662,
      "grad_norm": 0.3754492700099945,
      "learning_rate": 0.0001,
      "loss": 1.6346,
      "step": 3674
    },
    {
      "epoch": 0.8924235065565809,
      "grad_norm": 0.37411370873451233,
      "learning_rate": 0.0001,
      "loss": 1.72,
      "step": 3675
    },
    {
      "epoch": 0.8926663428848955,
      "grad_norm": 0.35851696133613586,
      "learning_rate": 0.0001,
      "loss": 1.7167,
      "step": 3676
    },
    {
      "epoch": 0.8929091792132103,
      "grad_norm": 0.4156700372695923,
      "learning_rate": 0.0001,
      "loss": 1.7594,
      "step": 3677
    },
    {
      "epoch": 0.893152015541525,
      "grad_norm": 0.3887973129749298,
      "learning_rate": 0.0001,
      "loss": 1.8126,
      "step": 3678
    },
    {
      "epoch": 0.8933948518698397,
      "grad_norm": 0.38091030716896057,
      "learning_rate": 0.0001,
      "loss": 1.8359,
      "step": 3679
    },
    {
      "epoch": 0.8936376881981545,
      "grad_norm": 0.40002167224884033,
      "learning_rate": 0.0001,
      "loss": 1.5869,
      "step": 3680
    },
    {
      "epoch": 0.8938805245264692,
      "grad_norm": 0.3787648379802704,
      "learning_rate": 0.0001,
      "loss": 1.6256,
      "step": 3681
    },
    {
      "epoch": 0.8941233608547838,
      "grad_norm": 0.3811265230178833,
      "learning_rate": 0.0001,
      "loss": 1.6876,
      "step": 3682
    },
    {
      "epoch": 0.8943661971830986,
      "grad_norm": 0.3632388114929199,
      "learning_rate": 0.0001,
      "loss": 1.7248,
      "step": 3683
    },
    {
      "epoch": 0.8946090335114133,
      "grad_norm": 0.3976041376590729,
      "learning_rate": 0.0001,
      "loss": 1.667,
      "step": 3684
    },
    {
      "epoch": 0.894851869839728,
      "grad_norm": 0.37803617119789124,
      "learning_rate": 0.0001,
      "loss": 1.6494,
      "step": 3685
    },
    {
      "epoch": 0.8950947061680428,
      "grad_norm": 0.3664271831512451,
      "learning_rate": 0.0001,
      "loss": 1.7021,
      "step": 3686
    },
    {
      "epoch": 0.8953375424963574,
      "grad_norm": 0.35976269841194153,
      "learning_rate": 0.0001,
      "loss": 1.6618,
      "step": 3687
    },
    {
      "epoch": 0.8955803788246721,
      "grad_norm": 0.35800105333328247,
      "learning_rate": 0.0001,
      "loss": 1.6671,
      "step": 3688
    },
    {
      "epoch": 0.8958232151529869,
      "grad_norm": 0.342416375875473,
      "learning_rate": 0.0001,
      "loss": 1.6609,
      "step": 3689
    },
    {
      "epoch": 0.8960660514813016,
      "grad_norm": 0.3788902759552002,
      "learning_rate": 0.0001,
      "loss": 1.6947,
      "step": 3690
    },
    {
      "epoch": 0.8963088878096164,
      "grad_norm": 0.4403901696205139,
      "learning_rate": 0.0001,
      "loss": 1.6916,
      "step": 3691
    },
    {
      "epoch": 0.896551724137931,
      "grad_norm": 0.3746020197868347,
      "learning_rate": 0.0001,
      "loss": 1.7877,
      "step": 3692
    },
    {
      "epoch": 0.8967945604662457,
      "grad_norm": 0.3692503273487091,
      "learning_rate": 0.0001,
      "loss": 1.6245,
      "step": 3693
    },
    {
      "epoch": 0.8970373967945605,
      "grad_norm": 0.38501957058906555,
      "learning_rate": 0.0001,
      "loss": 1.8146,
      "step": 3694
    },
    {
      "epoch": 0.8972802331228752,
      "grad_norm": 0.3959503769874573,
      "learning_rate": 0.0001,
      "loss": 1.7034,
      "step": 3695
    },
    {
      "epoch": 0.8975230694511899,
      "grad_norm": 0.35865819454193115,
      "learning_rate": 0.0001,
      "loss": 1.7013,
      "step": 3696
    },
    {
      "epoch": 0.8977659057795047,
      "grad_norm": 0.3601124584674835,
      "learning_rate": 0.0001,
      "loss": 1.6855,
      "step": 3697
    },
    {
      "epoch": 0.8980087421078193,
      "grad_norm": 0.40033718943595886,
      "learning_rate": 0.0001,
      "loss": 1.8622,
      "step": 3698
    },
    {
      "epoch": 0.898251578436134,
      "grad_norm": 0.35714730620384216,
      "learning_rate": 0.0001,
      "loss": 1.718,
      "step": 3699
    },
    {
      "epoch": 0.8984944147644488,
      "grad_norm": 0.37696969509124756,
      "learning_rate": 0.0001,
      "loss": 1.6804,
      "step": 3700
    },
    {
      "epoch": 0.8987372510927635,
      "grad_norm": 0.3685283660888672,
      "learning_rate": 0.0001,
      "loss": 1.6738,
      "step": 3701
    },
    {
      "epoch": 0.8989800874210782,
      "grad_norm": 0.3915679156780243,
      "learning_rate": 0.0001,
      "loss": 1.7391,
      "step": 3702
    },
    {
      "epoch": 0.8992229237493929,
      "grad_norm": 0.37142446637153625,
      "learning_rate": 0.0001,
      "loss": 1.7583,
      "step": 3703
    },
    {
      "epoch": 0.8994657600777076,
      "grad_norm": 0.3911544978618622,
      "learning_rate": 0.0001,
      "loss": 1.8277,
      "step": 3704
    },
    {
      "epoch": 0.8997085964060223,
      "grad_norm": 0.3824489116668701,
      "learning_rate": 0.0001,
      "loss": 1.7645,
      "step": 3705
    },
    {
      "epoch": 0.8999514327343371,
      "grad_norm": 0.36203762888908386,
      "learning_rate": 0.0001,
      "loss": 1.7458,
      "step": 3706
    },
    {
      "epoch": 0.9001942690626518,
      "grad_norm": 0.3630374073982239,
      "learning_rate": 0.0001,
      "loss": 1.6611,
      "step": 3707
    },
    {
      "epoch": 0.9004371053909664,
      "grad_norm": 0.3603218197822571,
      "learning_rate": 0.0001,
      "loss": 1.5932,
      "step": 3708
    },
    {
      "epoch": 0.9006799417192812,
      "grad_norm": 0.34822002053260803,
      "learning_rate": 0.0001,
      "loss": 1.5359,
      "step": 3709
    },
    {
      "epoch": 0.9009227780475959,
      "grad_norm": 0.36869946122169495,
      "learning_rate": 0.0001,
      "loss": 1.6381,
      "step": 3710
    },
    {
      "epoch": 0.9011656143759106,
      "grad_norm": 0.37184393405914307,
      "learning_rate": 0.0001,
      "loss": 1.7159,
      "step": 3711
    },
    {
      "epoch": 0.9014084507042254,
      "grad_norm": 0.3584497272968292,
      "learning_rate": 0.0001,
      "loss": 1.7466,
      "step": 3712
    },
    {
      "epoch": 0.90165128703254,
      "grad_norm": 0.3819580674171448,
      "learning_rate": 0.0001,
      "loss": 1.7638,
      "step": 3713
    },
    {
      "epoch": 0.9018941233608548,
      "grad_norm": 0.3784763514995575,
      "learning_rate": 0.0001,
      "loss": 1.7203,
      "step": 3714
    },
    {
      "epoch": 0.9021369596891695,
      "grad_norm": 0.3671453297138214,
      "learning_rate": 0.0001,
      "loss": 1.7007,
      "step": 3715
    },
    {
      "epoch": 0.9023797960174842,
      "grad_norm": 0.3914904296398163,
      "learning_rate": 0.0001,
      "loss": 1.8468,
      "step": 3716
    },
    {
      "epoch": 0.902622632345799,
      "grad_norm": 0.3483647406101227,
      "learning_rate": 0.0001,
      "loss": 1.5126,
      "step": 3717
    },
    {
      "epoch": 0.9028654686741137,
      "grad_norm": 0.3823911249637604,
      "learning_rate": 0.0001,
      "loss": 1.6252,
      "step": 3718
    },
    {
      "epoch": 0.9031083050024283,
      "grad_norm": 0.3677743673324585,
      "learning_rate": 0.0001,
      "loss": 1.7183,
      "step": 3719
    },
    {
      "epoch": 0.9033511413307431,
      "grad_norm": 0.38830864429473877,
      "learning_rate": 0.0001,
      "loss": 1.757,
      "step": 3720
    },
    {
      "epoch": 0.9035939776590578,
      "grad_norm": 0.3823908865451813,
      "learning_rate": 0.0001,
      "loss": 1.6822,
      "step": 3721
    },
    {
      "epoch": 0.9038368139873725,
      "grad_norm": 0.3889397382736206,
      "learning_rate": 0.0001,
      "loss": 1.8468,
      "step": 3722
    },
    {
      "epoch": 0.9040796503156873,
      "grad_norm": 0.3682743012905121,
      "learning_rate": 0.0001,
      "loss": 1.6469,
      "step": 3723
    },
    {
      "epoch": 0.9043224866440019,
      "grad_norm": 0.41113588213920593,
      "learning_rate": 0.0001,
      "loss": 1.9098,
      "step": 3724
    },
    {
      "epoch": 0.9045653229723166,
      "grad_norm": 0.3677564859390259,
      "learning_rate": 0.0001,
      "loss": 1.6514,
      "step": 3725
    },
    {
      "epoch": 0.9048081593006314,
      "grad_norm": 0.3636486232280731,
      "learning_rate": 0.0001,
      "loss": 1.527,
      "step": 3726
    },
    {
      "epoch": 0.9050509956289461,
      "grad_norm": 0.4054773151874542,
      "learning_rate": 0.0001,
      "loss": 1.6989,
      "step": 3727
    },
    {
      "epoch": 0.9052938319572608,
      "grad_norm": 0.3783499598503113,
      "learning_rate": 0.0001,
      "loss": 1.7008,
      "step": 3728
    },
    {
      "epoch": 0.9055366682855756,
      "grad_norm": 0.3834938406944275,
      "learning_rate": 0.0001,
      "loss": 1.7659,
      "step": 3729
    },
    {
      "epoch": 0.9057795046138902,
      "grad_norm": 0.3626881241798401,
      "learning_rate": 0.0001,
      "loss": 1.7242,
      "step": 3730
    },
    {
      "epoch": 0.9060223409422049,
      "grad_norm": 0.37479889392852783,
      "learning_rate": 0.0001,
      "loss": 1.6904,
      "step": 3731
    },
    {
      "epoch": 0.9062651772705197,
      "grad_norm": 0.38191792368888855,
      "learning_rate": 0.0001,
      "loss": 1.7189,
      "step": 3732
    },
    {
      "epoch": 0.9065080135988344,
      "grad_norm": 0.3727664649486542,
      "learning_rate": 0.0001,
      "loss": 1.5817,
      "step": 3733
    },
    {
      "epoch": 0.906750849927149,
      "grad_norm": 0.3694825768470764,
      "learning_rate": 0.0001,
      "loss": 1.6841,
      "step": 3734
    },
    {
      "epoch": 0.9069936862554638,
      "grad_norm": 0.36148035526275635,
      "learning_rate": 0.0001,
      "loss": 1.7172,
      "step": 3735
    },
    {
      "epoch": 0.9072365225837785,
      "grad_norm": 0.42565613985061646,
      "learning_rate": 0.0001,
      "loss": 1.7159,
      "step": 3736
    },
    {
      "epoch": 0.9074793589120933,
      "grad_norm": 0.4023953080177307,
      "learning_rate": 0.0001,
      "loss": 1.7227,
      "step": 3737
    },
    {
      "epoch": 0.907722195240408,
      "grad_norm": 0.35882890224456787,
      "learning_rate": 0.0001,
      "loss": 1.6104,
      "step": 3738
    },
    {
      "epoch": 0.9079650315687227,
      "grad_norm": 0.4042493999004364,
      "learning_rate": 0.0001,
      "loss": 1.6337,
      "step": 3739
    },
    {
      "epoch": 0.9082078678970374,
      "grad_norm": 0.3844760060310364,
      "learning_rate": 0.0001,
      "loss": 1.7242,
      "step": 3740
    },
    {
      "epoch": 0.9084507042253521,
      "grad_norm": 0.3652195930480957,
      "learning_rate": 0.0001,
      "loss": 1.785,
      "step": 3741
    },
    {
      "epoch": 0.9086935405536668,
      "grad_norm": 0.38959187269210815,
      "learning_rate": 0.0001,
      "loss": 1.8141,
      "step": 3742
    },
    {
      "epoch": 0.9089363768819816,
      "grad_norm": 0.37486904859542847,
      "learning_rate": 0.0001,
      "loss": 1.7031,
      "step": 3743
    },
    {
      "epoch": 0.9091792132102963,
      "grad_norm": 0.360111266374588,
      "learning_rate": 0.0001,
      "loss": 1.4131,
      "step": 3744
    },
    {
      "epoch": 0.9094220495386109,
      "grad_norm": 0.37102243304252625,
      "learning_rate": 0.0001,
      "loss": 1.8195,
      "step": 3745
    },
    {
      "epoch": 0.9096648858669257,
      "grad_norm": 0.36768385767936707,
      "learning_rate": 0.0001,
      "loss": 1.6769,
      "step": 3746
    },
    {
      "epoch": 0.9099077221952404,
      "grad_norm": 0.37154293060302734,
      "learning_rate": 0.0001,
      "loss": 1.7091,
      "step": 3747
    },
    {
      "epoch": 0.9101505585235551,
      "grad_norm": 0.3653973639011383,
      "learning_rate": 0.0001,
      "loss": 1.6808,
      "step": 3748
    },
    {
      "epoch": 0.9103933948518699,
      "grad_norm": 0.3743472695350647,
      "learning_rate": 0.0001,
      "loss": 1.7694,
      "step": 3749
    },
    {
      "epoch": 0.9106362311801846,
      "grad_norm": 0.3726973831653595,
      "learning_rate": 0.0001,
      "loss": 1.7917,
      "step": 3750
    },
    {
      "epoch": 0.9108790675084992,
      "grad_norm": 0.36880922317504883,
      "learning_rate": 0.0001,
      "loss": 1.703,
      "step": 3751
    },
    {
      "epoch": 0.911121903836814,
      "grad_norm": 0.3703029751777649,
      "learning_rate": 0.0001,
      "loss": 1.7295,
      "step": 3752
    },
    {
      "epoch": 0.9113647401651287,
      "grad_norm": 0.3539191484451294,
      "learning_rate": 0.0001,
      "loss": 1.7341,
      "step": 3753
    },
    {
      "epoch": 0.9116075764934434,
      "grad_norm": 0.3663886785507202,
      "learning_rate": 0.0001,
      "loss": 1.6153,
      "step": 3754
    },
    {
      "epoch": 0.9118504128217582,
      "grad_norm": 0.3673847019672394,
      "learning_rate": 0.0001,
      "loss": 1.7479,
      "step": 3755
    },
    {
      "epoch": 0.9120932491500728,
      "grad_norm": 0.3632906973361969,
      "learning_rate": 0.0001,
      "loss": 1.7295,
      "step": 3756
    },
    {
      "epoch": 0.9123360854783875,
      "grad_norm": 0.37695321440696716,
      "learning_rate": 0.0001,
      "loss": 1.6302,
      "step": 3757
    },
    {
      "epoch": 0.9125789218067023,
      "grad_norm": 0.3578214645385742,
      "learning_rate": 0.0001,
      "loss": 1.6948,
      "step": 3758
    },
    {
      "epoch": 0.912821758135017,
      "grad_norm": 0.4027178883552551,
      "learning_rate": 0.0001,
      "loss": 1.7601,
      "step": 3759
    },
    {
      "epoch": 0.9130645944633318,
      "grad_norm": 0.38166484236717224,
      "learning_rate": 0.0001,
      "loss": 1.7154,
      "step": 3760
    },
    {
      "epoch": 0.9133074307916464,
      "grad_norm": 0.3768790364265442,
      "learning_rate": 0.0001,
      "loss": 1.7746,
      "step": 3761
    },
    {
      "epoch": 0.9135502671199611,
      "grad_norm": 0.3585771322250366,
      "learning_rate": 0.0001,
      "loss": 1.5268,
      "step": 3762
    },
    {
      "epoch": 0.9137931034482759,
      "grad_norm": 0.35531941056251526,
      "learning_rate": 0.0001,
      "loss": 1.6352,
      "step": 3763
    },
    {
      "epoch": 0.9140359397765906,
      "grad_norm": 0.37948161363601685,
      "learning_rate": 0.0001,
      "loss": 1.7789,
      "step": 3764
    },
    {
      "epoch": 0.9142787761049053,
      "grad_norm": 0.3917665183544159,
      "learning_rate": 0.0001,
      "loss": 1.8557,
      "step": 3765
    },
    {
      "epoch": 0.91452161243322,
      "grad_norm": 0.3560101091861725,
      "learning_rate": 0.0001,
      "loss": 1.6469,
      "step": 3766
    },
    {
      "epoch": 0.9147644487615347,
      "grad_norm": 0.37302523851394653,
      "learning_rate": 0.0001,
      "loss": 1.6948,
      "step": 3767
    },
    {
      "epoch": 0.9150072850898494,
      "grad_norm": 0.3936711251735687,
      "learning_rate": 0.0001,
      "loss": 1.7581,
      "step": 3768
    },
    {
      "epoch": 0.9152501214181642,
      "grad_norm": 0.3600723445415497,
      "learning_rate": 0.0001,
      "loss": 1.6131,
      "step": 3769
    },
    {
      "epoch": 0.9154929577464789,
      "grad_norm": 0.3593932092189789,
      "learning_rate": 0.0001,
      "loss": 1.5959,
      "step": 3770
    },
    {
      "epoch": 0.9157357940747936,
      "grad_norm": 0.3671623468399048,
      "learning_rate": 0.0001,
      "loss": 1.6203,
      "step": 3771
    },
    {
      "epoch": 0.9159786304031083,
      "grad_norm": 0.3588806986808777,
      "learning_rate": 0.0001,
      "loss": 1.616,
      "step": 3772
    },
    {
      "epoch": 0.916221466731423,
      "grad_norm": 0.3804694712162018,
      "learning_rate": 0.0001,
      "loss": 1.7997,
      "step": 3773
    },
    {
      "epoch": 0.9164643030597377,
      "grad_norm": 0.36302250623703003,
      "learning_rate": 0.0001,
      "loss": 1.7443,
      "step": 3774
    },
    {
      "epoch": 0.9167071393880525,
      "grad_norm": 0.34705448150634766,
      "learning_rate": 0.0001,
      "loss": 1.578,
      "step": 3775
    },
    {
      "epoch": 0.9169499757163672,
      "grad_norm": 0.359637975692749,
      "learning_rate": 0.0001,
      "loss": 1.6751,
      "step": 3776
    },
    {
      "epoch": 0.9171928120446818,
      "grad_norm": 0.3869219422340393,
      "learning_rate": 0.0001,
      "loss": 1.7907,
      "step": 3777
    },
    {
      "epoch": 0.9174356483729966,
      "grad_norm": 0.3761953115463257,
      "learning_rate": 0.0001,
      "loss": 1.8541,
      "step": 3778
    },
    {
      "epoch": 0.9176784847013113,
      "grad_norm": 0.3853410482406616,
      "learning_rate": 0.0001,
      "loss": 1.7271,
      "step": 3779
    },
    {
      "epoch": 0.917921321029626,
      "grad_norm": 0.37119677662849426,
      "learning_rate": 0.0001,
      "loss": 1.7134,
      "step": 3780
    },
    {
      "epoch": 0.9181641573579408,
      "grad_norm": 0.369057834148407,
      "learning_rate": 0.0001,
      "loss": 1.7957,
      "step": 3781
    },
    {
      "epoch": 0.9184069936862554,
      "grad_norm": 0.37399792671203613,
      "learning_rate": 0.0001,
      "loss": 1.7382,
      "step": 3782
    },
    {
      "epoch": 0.9186498300145702,
      "grad_norm": 0.38667723536491394,
      "learning_rate": 0.0001,
      "loss": 1.7569,
      "step": 3783
    },
    {
      "epoch": 0.9188926663428849,
      "grad_norm": 0.39617523550987244,
      "learning_rate": 0.0001,
      "loss": 1.6252,
      "step": 3784
    },
    {
      "epoch": 0.9191355026711996,
      "grad_norm": 0.38363882899284363,
      "learning_rate": 0.0001,
      "loss": 1.8444,
      "step": 3785
    },
    {
      "epoch": 0.9193783389995144,
      "grad_norm": 0.4173693358898163,
      "learning_rate": 0.0001,
      "loss": 1.9011,
      "step": 3786
    },
    {
      "epoch": 0.919621175327829,
      "grad_norm": 0.4081033766269684,
      "learning_rate": 0.0001,
      "loss": 1.772,
      "step": 3787
    },
    {
      "epoch": 0.9198640116561437,
      "grad_norm": 0.3917405903339386,
      "learning_rate": 0.0001,
      "loss": 1.8649,
      "step": 3788
    },
    {
      "epoch": 0.9201068479844585,
      "grad_norm": 0.36920738220214844,
      "learning_rate": 0.0001,
      "loss": 1.6866,
      "step": 3789
    },
    {
      "epoch": 0.9203496843127732,
      "grad_norm": 0.359141081571579,
      "learning_rate": 0.0001,
      "loss": 1.8148,
      "step": 3790
    },
    {
      "epoch": 0.9205925206410879,
      "grad_norm": 0.35884061455726624,
      "learning_rate": 0.0001,
      "loss": 1.5744,
      "step": 3791
    },
    {
      "epoch": 0.9208353569694027,
      "grad_norm": 0.38075488805770874,
      "learning_rate": 0.0001,
      "loss": 1.7728,
      "step": 3792
    },
    {
      "epoch": 0.9210781932977173,
      "grad_norm": 0.36001816391944885,
      "learning_rate": 0.0001,
      "loss": 1.6481,
      "step": 3793
    },
    {
      "epoch": 0.921321029626032,
      "grad_norm": 0.37972912192344666,
      "learning_rate": 0.0001,
      "loss": 1.601,
      "step": 3794
    },
    {
      "epoch": 0.9215638659543468,
      "grad_norm": 0.36784154176712036,
      "learning_rate": 0.0001,
      "loss": 1.679,
      "step": 3795
    },
    {
      "epoch": 0.9218067022826615,
      "grad_norm": 0.36483627557754517,
      "learning_rate": 0.0001,
      "loss": 1.5875,
      "step": 3796
    },
    {
      "epoch": 0.9220495386109762,
      "grad_norm": 0.3781699538230896,
      "learning_rate": 0.0001,
      "loss": 1.8234,
      "step": 3797
    },
    {
      "epoch": 0.922292374939291,
      "grad_norm": 0.3724982440471649,
      "learning_rate": 0.0001,
      "loss": 1.6933,
      "step": 3798
    },
    {
      "epoch": 0.9225352112676056,
      "grad_norm": 0.39953988790512085,
      "learning_rate": 0.0001,
      "loss": 1.7848,
      "step": 3799
    },
    {
      "epoch": 0.9227780475959203,
      "grad_norm": 0.3567099869251251,
      "learning_rate": 0.0001,
      "loss": 1.7083,
      "step": 3800
    },
    {
      "epoch": 0.9230208839242351,
      "grad_norm": 0.38667944073677063,
      "learning_rate": 0.0001,
      "loss": 1.7049,
      "step": 3801
    },
    {
      "epoch": 0.9232637202525498,
      "grad_norm": 0.4046122133731842,
      "learning_rate": 0.0001,
      "loss": 1.7332,
      "step": 3802
    },
    {
      "epoch": 0.9235065565808644,
      "grad_norm": 0.3688383400440216,
      "learning_rate": 0.0001,
      "loss": 1.5917,
      "step": 3803
    },
    {
      "epoch": 0.9237493929091792,
      "grad_norm": 0.3840380609035492,
      "learning_rate": 0.0001,
      "loss": 1.555,
      "step": 3804
    },
    {
      "epoch": 0.9239922292374939,
      "grad_norm": 0.36803650856018066,
      "learning_rate": 0.0001,
      "loss": 1.8166,
      "step": 3805
    },
    {
      "epoch": 0.9242350655658087,
      "grad_norm": 0.3958253860473633,
      "learning_rate": 0.0001,
      "loss": 1.8586,
      "step": 3806
    },
    {
      "epoch": 0.9244779018941234,
      "grad_norm": 0.3751855492591858,
      "learning_rate": 0.0001,
      "loss": 1.7138,
      "step": 3807
    },
    {
      "epoch": 0.924720738222438,
      "grad_norm": 0.3692433834075928,
      "learning_rate": 0.0001,
      "loss": 1.6246,
      "step": 3808
    },
    {
      "epoch": 0.9249635745507528,
      "grad_norm": 0.3688707947731018,
      "learning_rate": 0.0001,
      "loss": 1.8355,
      "step": 3809
    },
    {
      "epoch": 0.9252064108790675,
      "grad_norm": 0.39333146810531616,
      "learning_rate": 0.0001,
      "loss": 1.7563,
      "step": 3810
    },
    {
      "epoch": 0.9254492472073822,
      "grad_norm": 0.3726106584072113,
      "learning_rate": 0.0001,
      "loss": 1.6823,
      "step": 3811
    },
    {
      "epoch": 0.925692083535697,
      "grad_norm": 0.36215999722480774,
      "learning_rate": 0.0001,
      "loss": 1.6521,
      "step": 3812
    },
    {
      "epoch": 0.9259349198640117,
      "grad_norm": 0.38629502058029175,
      "learning_rate": 0.0001,
      "loss": 1.7459,
      "step": 3813
    },
    {
      "epoch": 0.9261777561923263,
      "grad_norm": 0.3587053418159485,
      "learning_rate": 0.0001,
      "loss": 1.6508,
      "step": 3814
    },
    {
      "epoch": 0.9264205925206411,
      "grad_norm": 0.3717343509197235,
      "learning_rate": 0.0001,
      "loss": 1.7678,
      "step": 3815
    },
    {
      "epoch": 0.9266634288489558,
      "grad_norm": 0.37315699458122253,
      "learning_rate": 0.0001,
      "loss": 1.6838,
      "step": 3816
    },
    {
      "epoch": 0.9269062651772705,
      "grad_norm": 0.3515320122241974,
      "learning_rate": 0.0001,
      "loss": 1.524,
      "step": 3817
    },
    {
      "epoch": 0.9271491015055853,
      "grad_norm": 0.36067575216293335,
      "learning_rate": 0.0001,
      "loss": 1.7694,
      "step": 3818
    },
    {
      "epoch": 0.9273919378339,
      "grad_norm": 0.37640073895454407,
      "learning_rate": 0.0001,
      "loss": 1.7702,
      "step": 3819
    },
    {
      "epoch": 0.9276347741622146,
      "grad_norm": 0.3961258828639984,
      "learning_rate": 0.0001,
      "loss": 1.8072,
      "step": 3820
    },
    {
      "epoch": 0.9278776104905294,
      "grad_norm": 0.373757004737854,
      "learning_rate": 0.0001,
      "loss": 1.6027,
      "step": 3821
    },
    {
      "epoch": 0.9281204468188441,
      "grad_norm": 0.36781755089759827,
      "learning_rate": 0.0001,
      "loss": 1.6513,
      "step": 3822
    },
    {
      "epoch": 0.9283632831471588,
      "grad_norm": 0.3814017176628113,
      "learning_rate": 0.0001,
      "loss": 1.7724,
      "step": 3823
    },
    {
      "epoch": 0.9286061194754736,
      "grad_norm": 0.40861770510673523,
      "learning_rate": 0.0001,
      "loss": 1.6388,
      "step": 3824
    },
    {
      "epoch": 0.9288489558037882,
      "grad_norm": 0.3678155839443207,
      "learning_rate": 0.0001,
      "loss": 1.5734,
      "step": 3825
    },
    {
      "epoch": 0.9290917921321029,
      "grad_norm": 0.3855389654636383,
      "learning_rate": 0.0001,
      "loss": 1.7397,
      "step": 3826
    },
    {
      "epoch": 0.9293346284604177,
      "grad_norm": 0.3705948293209076,
      "learning_rate": 0.0001,
      "loss": 1.6456,
      "step": 3827
    },
    {
      "epoch": 0.9295774647887324,
      "grad_norm": 0.3740312159061432,
      "learning_rate": 0.0001,
      "loss": 1.7897,
      "step": 3828
    },
    {
      "epoch": 0.9298203011170472,
      "grad_norm": 0.38665881752967834,
      "learning_rate": 0.0001,
      "loss": 1.8266,
      "step": 3829
    },
    {
      "epoch": 0.9300631374453618,
      "grad_norm": 0.36732590198516846,
      "learning_rate": 0.0001,
      "loss": 1.7069,
      "step": 3830
    },
    {
      "epoch": 0.9303059737736765,
      "grad_norm": 0.37280961871147156,
      "learning_rate": 0.0001,
      "loss": 1.8124,
      "step": 3831
    },
    {
      "epoch": 0.9305488101019913,
      "grad_norm": 0.4017069637775421,
      "learning_rate": 0.0001,
      "loss": 1.8979,
      "step": 3832
    },
    {
      "epoch": 0.930791646430306,
      "grad_norm": 0.3593195378780365,
      "learning_rate": 0.0001,
      "loss": 1.7046,
      "step": 3833
    },
    {
      "epoch": 0.9310344827586207,
      "grad_norm": 0.3705736994743347,
      "learning_rate": 0.0001,
      "loss": 1.6845,
      "step": 3834
    },
    {
      "epoch": 0.9312773190869355,
      "grad_norm": 0.38185861706733704,
      "learning_rate": 0.0001,
      "loss": 1.6715,
      "step": 3835
    },
    {
      "epoch": 0.9315201554152501,
      "grad_norm": 0.37733927369117737,
      "learning_rate": 0.0001,
      "loss": 1.7452,
      "step": 3836
    },
    {
      "epoch": 0.9317629917435648,
      "grad_norm": 0.4084572196006775,
      "learning_rate": 0.0001,
      "loss": 1.8061,
      "step": 3837
    },
    {
      "epoch": 0.9320058280718796,
      "grad_norm": 0.3832477927207947,
      "learning_rate": 0.0001,
      "loss": 1.7575,
      "step": 3838
    },
    {
      "epoch": 0.9322486644001943,
      "grad_norm": 0.4194474518299103,
      "learning_rate": 0.0001,
      "loss": 1.8246,
      "step": 3839
    },
    {
      "epoch": 0.932491500728509,
      "grad_norm": 0.36019614338874817,
      "learning_rate": 0.0001,
      "loss": 1.7305,
      "step": 3840
    },
    {
      "epoch": 0.9327343370568237,
      "grad_norm": 0.3596409857273102,
      "learning_rate": 0.0001,
      "loss": 1.6803,
      "step": 3841
    },
    {
      "epoch": 0.9329771733851384,
      "grad_norm": 0.3968151807785034,
      "learning_rate": 0.0001,
      "loss": 1.8563,
      "step": 3842
    },
    {
      "epoch": 0.9332200097134531,
      "grad_norm": 0.33716362714767456,
      "learning_rate": 0.0001,
      "loss": 1.6342,
      "step": 3843
    },
    {
      "epoch": 0.9334628460417679,
      "grad_norm": 0.3743828535079956,
      "learning_rate": 0.0001,
      "loss": 1.7532,
      "step": 3844
    },
    {
      "epoch": 0.9337056823700826,
      "grad_norm": 0.355603963136673,
      "learning_rate": 0.0001,
      "loss": 1.5373,
      "step": 3845
    },
    {
      "epoch": 0.9339485186983972,
      "grad_norm": 0.36174502968788147,
      "learning_rate": 0.0001,
      "loss": 1.6813,
      "step": 3846
    },
    {
      "epoch": 0.934191355026712,
      "grad_norm": 0.355954110622406,
      "learning_rate": 0.0001,
      "loss": 1.5443,
      "step": 3847
    },
    {
      "epoch": 0.9344341913550267,
      "grad_norm": 0.3647748529911041,
      "learning_rate": 0.0001,
      "loss": 1.4645,
      "step": 3848
    },
    {
      "epoch": 0.9346770276833414,
      "grad_norm": 0.3363376259803772,
      "learning_rate": 0.0001,
      "loss": 1.52,
      "step": 3849
    },
    {
      "epoch": 0.9349198640116562,
      "grad_norm": 0.3913493752479553,
      "learning_rate": 0.0001,
      "loss": 1.7376,
      "step": 3850
    },
    {
      "epoch": 0.9351627003399708,
      "grad_norm": 0.4021153748035431,
      "learning_rate": 0.0001,
      "loss": 1.9212,
      "step": 3851
    },
    {
      "epoch": 0.9354055366682856,
      "grad_norm": 0.3638807237148285,
      "learning_rate": 0.0001,
      "loss": 1.7574,
      "step": 3852
    },
    {
      "epoch": 0.9356483729966003,
      "grad_norm": 0.36912328004837036,
      "learning_rate": 0.0001,
      "loss": 1.8145,
      "step": 3853
    },
    {
      "epoch": 0.935891209324915,
      "grad_norm": 0.3873084485530853,
      "learning_rate": 0.0001,
      "loss": 1.7894,
      "step": 3854
    },
    {
      "epoch": 0.9361340456532298,
      "grad_norm": 0.3698597848415375,
      "learning_rate": 0.0001,
      "loss": 1.76,
      "step": 3855
    },
    {
      "epoch": 0.9363768819815445,
      "grad_norm": 0.37964126467704773,
      "learning_rate": 0.0001,
      "loss": 1.8156,
      "step": 3856
    },
    {
      "epoch": 0.9366197183098591,
      "grad_norm": 0.3631061911582947,
      "learning_rate": 0.0001,
      "loss": 1.6366,
      "step": 3857
    },
    {
      "epoch": 0.9368625546381739,
      "grad_norm": 0.3854769766330719,
      "learning_rate": 0.0001,
      "loss": 1.6486,
      "step": 3858
    },
    {
      "epoch": 0.9371053909664886,
      "grad_norm": 0.3673323094844818,
      "learning_rate": 0.0001,
      "loss": 1.741,
      "step": 3859
    },
    {
      "epoch": 0.9373482272948033,
      "grad_norm": 0.3874039947986603,
      "learning_rate": 0.0001,
      "loss": 1.8782,
      "step": 3860
    },
    {
      "epoch": 0.9375910636231181,
      "grad_norm": 0.3830685615539551,
      "learning_rate": 0.0001,
      "loss": 1.7935,
      "step": 3861
    },
    {
      "epoch": 0.9378338999514327,
      "grad_norm": 0.3593977689743042,
      "learning_rate": 0.0001,
      "loss": 1.6822,
      "step": 3862
    },
    {
      "epoch": 0.9380767362797474,
      "grad_norm": 0.3805672228336334,
      "learning_rate": 0.0001,
      "loss": 1.8244,
      "step": 3863
    },
    {
      "epoch": 0.9383195726080622,
      "grad_norm": 0.35292261838912964,
      "learning_rate": 0.0001,
      "loss": 1.6097,
      "step": 3864
    },
    {
      "epoch": 0.9385624089363769,
      "grad_norm": 0.3597135543823242,
      "learning_rate": 0.0001,
      "loss": 1.5708,
      "step": 3865
    },
    {
      "epoch": 0.9388052452646916,
      "grad_norm": 0.3945610523223877,
      "learning_rate": 0.0001,
      "loss": 1.8461,
      "step": 3866
    },
    {
      "epoch": 0.9390480815930063,
      "grad_norm": 0.36307209730148315,
      "learning_rate": 0.0001,
      "loss": 1.655,
      "step": 3867
    },
    {
      "epoch": 0.939290917921321,
      "grad_norm": 0.3550439178943634,
      "learning_rate": 0.0001,
      "loss": 1.616,
      "step": 3868
    },
    {
      "epoch": 0.9395337542496357,
      "grad_norm": 0.35855671763420105,
      "learning_rate": 0.0001,
      "loss": 1.6995,
      "step": 3869
    },
    {
      "epoch": 0.9397765905779505,
      "grad_norm": 0.3556819260120392,
      "learning_rate": 0.0001,
      "loss": 1.486,
      "step": 3870
    },
    {
      "epoch": 0.9400194269062652,
      "grad_norm": 0.38059186935424805,
      "learning_rate": 0.0001,
      "loss": 1.7648,
      "step": 3871
    },
    {
      "epoch": 0.9402622632345798,
      "grad_norm": 0.35753384232521057,
      "learning_rate": 0.0001,
      "loss": 1.5893,
      "step": 3872
    },
    {
      "epoch": 0.9405050995628946,
      "grad_norm": 0.37142637372016907,
      "learning_rate": 0.0001,
      "loss": 1.7485,
      "step": 3873
    },
    {
      "epoch": 0.9407479358912093,
      "grad_norm": 0.3409467041492462,
      "learning_rate": 0.0001,
      "loss": 1.5089,
      "step": 3874
    },
    {
      "epoch": 0.940990772219524,
      "grad_norm": 0.35744860768318176,
      "learning_rate": 0.0001,
      "loss": 1.7779,
      "step": 3875
    },
    {
      "epoch": 0.9412336085478388,
      "grad_norm": 0.3719422221183777,
      "learning_rate": 0.0001,
      "loss": 1.5868,
      "step": 3876
    },
    {
      "epoch": 0.9414764448761535,
      "grad_norm": 0.3818061649799347,
      "learning_rate": 0.0001,
      "loss": 1.7106,
      "step": 3877
    },
    {
      "epoch": 0.9417192812044682,
      "grad_norm": 0.3610047399997711,
      "learning_rate": 0.0001,
      "loss": 1.5557,
      "step": 3878
    },
    {
      "epoch": 0.9419621175327829,
      "grad_norm": 0.36827659606933594,
      "learning_rate": 0.0001,
      "loss": 1.7735,
      "step": 3879
    },
    {
      "epoch": 0.9422049538610976,
      "grad_norm": 0.3562065660953522,
      "learning_rate": 0.0001,
      "loss": 1.6174,
      "step": 3880
    },
    {
      "epoch": 0.9424477901894124,
      "grad_norm": 0.3670153021812439,
      "learning_rate": 0.0001,
      "loss": 1.7448,
      "step": 3881
    },
    {
      "epoch": 0.9426906265177271,
      "grad_norm": 0.37725064158439636,
      "learning_rate": 0.0001,
      "loss": 1.6493,
      "step": 3882
    },
    {
      "epoch": 0.9429334628460417,
      "grad_norm": 0.3852989375591278,
      "learning_rate": 0.0001,
      "loss": 1.8689,
      "step": 3883
    },
    {
      "epoch": 0.9431762991743565,
      "grad_norm": 0.35966503620147705,
      "learning_rate": 0.0001,
      "loss": 1.5603,
      "step": 3884
    },
    {
      "epoch": 0.9434191355026712,
      "grad_norm": 0.36065688729286194,
      "learning_rate": 0.0001,
      "loss": 1.6563,
      "step": 3885
    },
    {
      "epoch": 0.9436619718309859,
      "grad_norm": 0.37432312965393066,
      "learning_rate": 0.0001,
      "loss": 1.7617,
      "step": 3886
    },
    {
      "epoch": 0.9439048081593007,
      "grad_norm": 0.372538298368454,
      "learning_rate": 0.0001,
      "loss": 1.8255,
      "step": 3887
    },
    {
      "epoch": 0.9441476444876153,
      "grad_norm": 0.3747842013835907,
      "learning_rate": 0.0001,
      "loss": 1.8332,
      "step": 3888
    },
    {
      "epoch": 0.94439048081593,
      "grad_norm": 0.367948055267334,
      "learning_rate": 0.0001,
      "loss": 1.7293,
      "step": 3889
    },
    {
      "epoch": 0.9446333171442448,
      "grad_norm": 0.3862752914428711,
      "learning_rate": 0.0001,
      "loss": 1.8169,
      "step": 3890
    },
    {
      "epoch": 0.9448761534725595,
      "grad_norm": 0.38457968831062317,
      "learning_rate": 0.0001,
      "loss": 1.654,
      "step": 3891
    },
    {
      "epoch": 0.9451189898008742,
      "grad_norm": 0.37712398171424866,
      "learning_rate": 0.0001,
      "loss": 1.8021,
      "step": 3892
    },
    {
      "epoch": 0.945361826129189,
      "grad_norm": 0.36352089047431946,
      "learning_rate": 0.0001,
      "loss": 1.6227,
      "step": 3893
    },
    {
      "epoch": 0.9456046624575036,
      "grad_norm": 0.37528321146965027,
      "learning_rate": 0.0001,
      "loss": 1.629,
      "step": 3894
    },
    {
      "epoch": 0.9458474987858183,
      "grad_norm": 0.3859131634235382,
      "learning_rate": 0.0001,
      "loss": 1.7157,
      "step": 3895
    },
    {
      "epoch": 0.9460903351141331,
      "grad_norm": 0.367913156747818,
      "learning_rate": 0.0001,
      "loss": 1.773,
      "step": 3896
    },
    {
      "epoch": 0.9463331714424478,
      "grad_norm": 0.3919924199581146,
      "learning_rate": 0.0001,
      "loss": 1.7986,
      "step": 3897
    },
    {
      "epoch": 0.9465760077707625,
      "grad_norm": 0.3621267080307007,
      "learning_rate": 0.0001,
      "loss": 1.766,
      "step": 3898
    },
    {
      "epoch": 0.9468188440990772,
      "grad_norm": 0.375805526971817,
      "learning_rate": 0.0001,
      "loss": 1.7918,
      "step": 3899
    },
    {
      "epoch": 0.9470616804273919,
      "grad_norm": 0.38171714544296265,
      "learning_rate": 0.0001,
      "loss": 1.779,
      "step": 3900
    },
    {
      "epoch": 0.9473045167557067,
      "grad_norm": 0.3639489710330963,
      "learning_rate": 0.0001,
      "loss": 1.7265,
      "step": 3901
    },
    {
      "epoch": 0.9475473530840214,
      "grad_norm": 0.3477586805820465,
      "learning_rate": 0.0001,
      "loss": 1.6296,
      "step": 3902
    },
    {
      "epoch": 0.9477901894123361,
      "grad_norm": 0.37721481919288635,
      "learning_rate": 0.0001,
      "loss": 1.6934,
      "step": 3903
    },
    {
      "epoch": 0.9480330257406508,
      "grad_norm": 0.3980064392089844,
      "learning_rate": 0.0001,
      "loss": 1.736,
      "step": 3904
    },
    {
      "epoch": 0.9482758620689655,
      "grad_norm": 0.3610898554325104,
      "learning_rate": 0.0001,
      "loss": 1.6393,
      "step": 3905
    },
    {
      "epoch": 0.9485186983972802,
      "grad_norm": 0.3711906671524048,
      "learning_rate": 0.0001,
      "loss": 1.7033,
      "step": 3906
    },
    {
      "epoch": 0.948761534725595,
      "grad_norm": 0.38572028279304504,
      "learning_rate": 0.0001,
      "loss": 1.754,
      "step": 3907
    },
    {
      "epoch": 0.9490043710539097,
      "grad_norm": 0.36621928215026855,
      "learning_rate": 0.0001,
      "loss": 1.6322,
      "step": 3908
    },
    {
      "epoch": 0.9492472073822243,
      "grad_norm": 0.3782571256160736,
      "learning_rate": 0.0001,
      "loss": 1.8551,
      "step": 3909
    },
    {
      "epoch": 0.9494900437105391,
      "grad_norm": 0.36709949374198914,
      "learning_rate": 0.0001,
      "loss": 1.5676,
      "step": 3910
    },
    {
      "epoch": 0.9497328800388538,
      "grad_norm": 0.38570162653923035,
      "learning_rate": 0.0001,
      "loss": 1.7557,
      "step": 3911
    },
    {
      "epoch": 0.9499757163671685,
      "grad_norm": 0.3895321488380432,
      "learning_rate": 0.0001,
      "loss": 1.7039,
      "step": 3912
    },
    {
      "epoch": 0.9502185526954833,
      "grad_norm": 0.35144907236099243,
      "learning_rate": 0.0001,
      "loss": 1.5781,
      "step": 3913
    },
    {
      "epoch": 0.950461389023798,
      "grad_norm": 0.38092508912086487,
      "learning_rate": 0.0001,
      "loss": 1.6042,
      "step": 3914
    },
    {
      "epoch": 0.9507042253521126,
      "grad_norm": 0.3619476854801178,
      "learning_rate": 0.0001,
      "loss": 1.5631,
      "step": 3915
    },
    {
      "epoch": 0.9509470616804274,
      "grad_norm": 0.40158382058143616,
      "learning_rate": 0.0001,
      "loss": 1.7766,
      "step": 3916
    },
    {
      "epoch": 0.9511898980087421,
      "grad_norm": 0.3472714424133301,
      "learning_rate": 0.0001,
      "loss": 1.5479,
      "step": 3917
    },
    {
      "epoch": 0.9514327343370568,
      "grad_norm": 0.36904680728912354,
      "learning_rate": 0.0001,
      "loss": 1.7577,
      "step": 3918
    },
    {
      "epoch": 0.9516755706653716,
      "grad_norm": 0.3941371440887451,
      "learning_rate": 0.0001,
      "loss": 1.8633,
      "step": 3919
    },
    {
      "epoch": 0.9519184069936862,
      "grad_norm": 0.3843080401420593,
      "learning_rate": 0.0001,
      "loss": 1.7322,
      "step": 3920
    },
    {
      "epoch": 0.9521612433220009,
      "grad_norm": 0.3651866912841797,
      "learning_rate": 0.0001,
      "loss": 1.6752,
      "step": 3921
    },
    {
      "epoch": 0.9524040796503157,
      "grad_norm": 0.38083866238594055,
      "learning_rate": 0.0001,
      "loss": 1.6955,
      "step": 3922
    },
    {
      "epoch": 0.9526469159786304,
      "grad_norm": 0.3733740746974945,
      "learning_rate": 0.0001,
      "loss": 1.5993,
      "step": 3923
    },
    {
      "epoch": 0.9528897523069452,
      "grad_norm": 0.380954384803772,
      "learning_rate": 0.0001,
      "loss": 1.7203,
      "step": 3924
    },
    {
      "epoch": 0.9531325886352598,
      "grad_norm": 0.3862300515174866,
      "learning_rate": 0.0001,
      "loss": 1.6818,
      "step": 3925
    },
    {
      "epoch": 0.9533754249635745,
      "grad_norm": 0.40166255831718445,
      "learning_rate": 0.0001,
      "loss": 1.7067,
      "step": 3926
    },
    {
      "epoch": 0.9536182612918893,
      "grad_norm": 0.3671046495437622,
      "learning_rate": 0.0001,
      "loss": 1.6741,
      "step": 3927
    },
    {
      "epoch": 0.953861097620204,
      "grad_norm": 0.3554973602294922,
      "learning_rate": 0.0001,
      "loss": 1.6256,
      "step": 3928
    },
    {
      "epoch": 0.9541039339485187,
      "grad_norm": 0.3730063736438751,
      "learning_rate": 0.0001,
      "loss": 1.7237,
      "step": 3929
    },
    {
      "epoch": 0.9543467702768335,
      "grad_norm": 0.36285051703453064,
      "learning_rate": 0.0001,
      "loss": 1.6649,
      "step": 3930
    },
    {
      "epoch": 0.9545896066051481,
      "grad_norm": 0.36590075492858887,
      "learning_rate": 0.0001,
      "loss": 1.4982,
      "step": 3931
    },
    {
      "epoch": 0.9548324429334628,
      "grad_norm": 0.38101863861083984,
      "learning_rate": 0.0001,
      "loss": 1.7006,
      "step": 3932
    },
    {
      "epoch": 0.9550752792617776,
      "grad_norm": 0.38884878158569336,
      "learning_rate": 0.0001,
      "loss": 1.7229,
      "step": 3933
    },
    {
      "epoch": 0.9553181155900923,
      "grad_norm": 0.40515512228012085,
      "learning_rate": 0.0001,
      "loss": 1.7678,
      "step": 3934
    },
    {
      "epoch": 0.955560951918407,
      "grad_norm": 0.4170911908149719,
      "learning_rate": 0.0001,
      "loss": 1.7991,
      "step": 3935
    },
    {
      "epoch": 0.9558037882467217,
      "grad_norm": 0.36002835631370544,
      "learning_rate": 0.0001,
      "loss": 1.6004,
      "step": 3936
    },
    {
      "epoch": 0.9560466245750364,
      "grad_norm": 0.38889098167419434,
      "learning_rate": 0.0001,
      "loss": 1.51,
      "step": 3937
    },
    {
      "epoch": 0.9562894609033511,
      "grad_norm": 0.393414169549942,
      "learning_rate": 0.0001,
      "loss": 1.6472,
      "step": 3938
    },
    {
      "epoch": 0.9565322972316659,
      "grad_norm": 0.38613879680633545,
      "learning_rate": 0.0001,
      "loss": 1.7916,
      "step": 3939
    },
    {
      "epoch": 0.9567751335599806,
      "grad_norm": 0.384308785200119,
      "learning_rate": 0.0001,
      "loss": 1.8509,
      "step": 3940
    },
    {
      "epoch": 0.9570179698882952,
      "grad_norm": 0.39423057436943054,
      "learning_rate": 0.0001,
      "loss": 1.7192,
      "step": 3941
    },
    {
      "epoch": 0.95726080621661,
      "grad_norm": 0.3493192195892334,
      "learning_rate": 0.0001,
      "loss": 1.4989,
      "step": 3942
    },
    {
      "epoch": 0.9575036425449247,
      "grad_norm": 0.3633043169975281,
      "learning_rate": 0.0001,
      "loss": 1.6156,
      "step": 3943
    },
    {
      "epoch": 0.9577464788732394,
      "grad_norm": 0.3607250154018402,
      "learning_rate": 0.0001,
      "loss": 1.6399,
      "step": 3944
    },
    {
      "epoch": 0.9579893152015542,
      "grad_norm": 0.3949401080608368,
      "learning_rate": 0.0001,
      "loss": 1.7105,
      "step": 3945
    },
    {
      "epoch": 0.9582321515298688,
      "grad_norm": 0.4055007994174957,
      "learning_rate": 0.0001,
      "loss": 1.8245,
      "step": 3946
    },
    {
      "epoch": 0.9584749878581836,
      "grad_norm": 0.368541955947876,
      "learning_rate": 0.0001,
      "loss": 1.7052,
      "step": 3947
    },
    {
      "epoch": 0.9587178241864983,
      "grad_norm": 0.39234501123428345,
      "learning_rate": 0.0001,
      "loss": 1.7882,
      "step": 3948
    },
    {
      "epoch": 0.958960660514813,
      "grad_norm": 0.37539973855018616,
      "learning_rate": 0.0001,
      "loss": 1.768,
      "step": 3949
    },
    {
      "epoch": 0.9592034968431278,
      "grad_norm": 0.3946327269077301,
      "learning_rate": 0.0001,
      "loss": 1.6774,
      "step": 3950
    },
    {
      "epoch": 0.9594463331714425,
      "grad_norm": 0.399637371301651,
      "learning_rate": 0.0001,
      "loss": 1.9732,
      "step": 3951
    },
    {
      "epoch": 0.9596891694997571,
      "grad_norm": 0.4541732668876648,
      "learning_rate": 0.0001,
      "loss": 1.6445,
      "step": 3952
    },
    {
      "epoch": 0.9599320058280719,
      "grad_norm": 0.3677199184894562,
      "learning_rate": 0.0001,
      "loss": 1.6625,
      "step": 3953
    },
    {
      "epoch": 0.9601748421563866,
      "grad_norm": 0.3860390782356262,
      "learning_rate": 0.0001,
      "loss": 1.8102,
      "step": 3954
    },
    {
      "epoch": 0.9604176784847013,
      "grad_norm": 0.36609208583831787,
      "learning_rate": 0.0001,
      "loss": 1.6788,
      "step": 3955
    },
    {
      "epoch": 0.9606605148130161,
      "grad_norm": 0.39333397150039673,
      "learning_rate": 0.0001,
      "loss": 1.6569,
      "step": 3956
    },
    {
      "epoch": 0.9609033511413307,
      "grad_norm": 0.39077579975128174,
      "learning_rate": 0.0001,
      "loss": 1.826,
      "step": 3957
    },
    {
      "epoch": 0.9611461874696454,
      "grad_norm": 0.37753769755363464,
      "learning_rate": 0.0001,
      "loss": 1.6429,
      "step": 3958
    },
    {
      "epoch": 0.9613890237979602,
      "grad_norm": 0.36394500732421875,
      "learning_rate": 0.0001,
      "loss": 1.6673,
      "step": 3959
    },
    {
      "epoch": 0.9616318601262749,
      "grad_norm": 0.35576146841049194,
      "learning_rate": 0.0001,
      "loss": 1.7007,
      "step": 3960
    },
    {
      "epoch": 0.9618746964545896,
      "grad_norm": 0.35573631525039673,
      "learning_rate": 0.0001,
      "loss": 1.6759,
      "step": 3961
    },
    {
      "epoch": 0.9621175327829043,
      "grad_norm": 0.40379825234413147,
      "learning_rate": 0.0001,
      "loss": 1.8536,
      "step": 3962
    },
    {
      "epoch": 0.962360369111219,
      "grad_norm": 0.35767439007759094,
      "learning_rate": 0.0001,
      "loss": 1.6918,
      "step": 3963
    },
    {
      "epoch": 0.9626032054395337,
      "grad_norm": 0.4079418182373047,
      "learning_rate": 0.0001,
      "loss": 1.9354,
      "step": 3964
    },
    {
      "epoch": 0.9628460417678485,
      "grad_norm": 0.37000563740730286,
      "learning_rate": 0.0001,
      "loss": 1.6034,
      "step": 3965
    },
    {
      "epoch": 0.9630888780961632,
      "grad_norm": 0.36689653992652893,
      "learning_rate": 0.0001,
      "loss": 1.6183,
      "step": 3966
    },
    {
      "epoch": 0.9633317144244778,
      "grad_norm": 0.3575601577758789,
      "learning_rate": 0.0001,
      "loss": 1.7348,
      "step": 3967
    },
    {
      "epoch": 0.9635745507527926,
      "grad_norm": 0.3624182641506195,
      "learning_rate": 0.0001,
      "loss": 1.7706,
      "step": 3968
    },
    {
      "epoch": 0.9638173870811073,
      "grad_norm": 0.3750225603580475,
      "learning_rate": 0.0001,
      "loss": 1.8725,
      "step": 3969
    },
    {
      "epoch": 0.9640602234094221,
      "grad_norm": 0.4041099548339844,
      "learning_rate": 0.0001,
      "loss": 1.9005,
      "step": 3970
    },
    {
      "epoch": 0.9643030597377368,
      "grad_norm": 0.38125571608543396,
      "learning_rate": 0.0001,
      "loss": 1.7553,
      "step": 3971
    },
    {
      "epoch": 0.9645458960660515,
      "grad_norm": 0.3875647783279419,
      "learning_rate": 0.0001,
      "loss": 1.8006,
      "step": 3972
    },
    {
      "epoch": 0.9647887323943662,
      "grad_norm": 0.3712661564350128,
      "learning_rate": 0.0001,
      "loss": 1.6987,
      "step": 3973
    },
    {
      "epoch": 0.9650315687226809,
      "grad_norm": 0.35640186071395874,
      "learning_rate": 0.0001,
      "loss": 1.7312,
      "step": 3974
    },
    {
      "epoch": 0.9652744050509956,
      "grad_norm": 0.3601911962032318,
      "learning_rate": 0.0001,
      "loss": 1.74,
      "step": 3975
    },
    {
      "epoch": 0.9655172413793104,
      "grad_norm": 0.3869849443435669,
      "learning_rate": 0.0001,
      "loss": 1.9955,
      "step": 3976
    },
    {
      "epoch": 0.9657600777076251,
      "grad_norm": 0.38196122646331787,
      "learning_rate": 0.0001,
      "loss": 1.8113,
      "step": 3977
    },
    {
      "epoch": 0.9660029140359397,
      "grad_norm": 0.3775624930858612,
      "learning_rate": 0.0001,
      "loss": 1.664,
      "step": 3978
    },
    {
      "epoch": 0.9662457503642545,
      "grad_norm": 0.3636208176612854,
      "learning_rate": 0.0001,
      "loss": 1.7108,
      "step": 3979
    },
    {
      "epoch": 0.9664885866925692,
      "grad_norm": 0.37071430683135986,
      "learning_rate": 0.0001,
      "loss": 1.7082,
      "step": 3980
    },
    {
      "epoch": 0.9667314230208839,
      "grad_norm": 0.37515515089035034,
      "learning_rate": 0.0001,
      "loss": 1.6976,
      "step": 3981
    },
    {
      "epoch": 0.9669742593491987,
      "grad_norm": 0.3826182186603546,
      "learning_rate": 0.0001,
      "loss": 1.6529,
      "step": 3982
    },
    {
      "epoch": 0.9672170956775133,
      "grad_norm": 0.3860194683074951,
      "learning_rate": 0.0001,
      "loss": 1.8768,
      "step": 3983
    },
    {
      "epoch": 0.967459932005828,
      "grad_norm": 0.4301397502422333,
      "learning_rate": 0.0001,
      "loss": 1.8274,
      "step": 3984
    },
    {
      "epoch": 0.9677027683341428,
      "grad_norm": 0.37735873460769653,
      "learning_rate": 0.0001,
      "loss": 1.7217,
      "step": 3985
    },
    {
      "epoch": 0.9679456046624575,
      "grad_norm": 0.40235722064971924,
      "learning_rate": 0.0001,
      "loss": 1.8216,
      "step": 3986
    },
    {
      "epoch": 0.9681884409907722,
      "grad_norm": 0.38523074984550476,
      "learning_rate": 0.0001,
      "loss": 1.7234,
      "step": 3987
    },
    {
      "epoch": 0.968431277319087,
      "grad_norm": 0.4025649428367615,
      "learning_rate": 0.0001,
      "loss": 1.7751,
      "step": 3988
    },
    {
      "epoch": 0.9686741136474016,
      "grad_norm": 0.3688255250453949,
      "learning_rate": 0.0001,
      "loss": 1.7218,
      "step": 3989
    },
    {
      "epoch": 0.9689169499757163,
      "grad_norm": 0.4351263642311096,
      "learning_rate": 0.0001,
      "loss": 1.8507,
      "step": 3990
    },
    {
      "epoch": 0.9691597863040311,
      "grad_norm": 0.37375950813293457,
      "learning_rate": 0.0001,
      "loss": 1.5947,
      "step": 3991
    },
    {
      "epoch": 0.9694026226323458,
      "grad_norm": 0.39403340220451355,
      "learning_rate": 0.0001,
      "loss": 1.7429,
      "step": 3992
    },
    {
      "epoch": 0.9696454589606606,
      "grad_norm": 0.38074350357055664,
      "learning_rate": 0.0001,
      "loss": 1.6262,
      "step": 3993
    },
    {
      "epoch": 0.9698882952889752,
      "grad_norm": 0.3697483241558075,
      "learning_rate": 0.0001,
      "loss": 1.5779,
      "step": 3994
    },
    {
      "epoch": 0.9701311316172899,
      "grad_norm": 0.3529646694660187,
      "learning_rate": 0.0001,
      "loss": 1.5475,
      "step": 3995
    },
    {
      "epoch": 0.9703739679456047,
      "grad_norm": 0.3757368326187134,
      "learning_rate": 0.0001,
      "loss": 1.6681,
      "step": 3996
    },
    {
      "epoch": 0.9706168042739194,
      "grad_norm": 0.41512370109558105,
      "learning_rate": 0.0001,
      "loss": 1.803,
      "step": 3997
    },
    {
      "epoch": 0.9708596406022341,
      "grad_norm": 0.3747655153274536,
      "learning_rate": 0.0001,
      "loss": 1.7778,
      "step": 3998
    },
    {
      "epoch": 0.9711024769305489,
      "grad_norm": 0.3599449396133423,
      "learning_rate": 0.0001,
      "loss": 1.6616,
      "step": 3999
    },
    {
      "epoch": 0.9713453132588635,
      "grad_norm": 0.39672431349754333,
      "learning_rate": 0.0001,
      "loss": 1.8923,
      "step": 4000
    },
    {
      "epoch": 0.9715881495871782,
      "grad_norm": 0.36832189559936523,
      "learning_rate": 0.0001,
      "loss": 1.5115,
      "step": 4001
    },
    {
      "epoch": 0.971830985915493,
      "grad_norm": 0.37256690859794617,
      "learning_rate": 0.0001,
      "loss": 1.8418,
      "step": 4002
    },
    {
      "epoch": 0.9720738222438077,
      "grad_norm": 0.39560651779174805,
      "learning_rate": 0.0001,
      "loss": 1.6514,
      "step": 4003
    },
    {
      "epoch": 0.9723166585721223,
      "grad_norm": 0.377238005399704,
      "learning_rate": 0.0001,
      "loss": 1.7049,
      "step": 4004
    },
    {
      "epoch": 0.9725594949004371,
      "grad_norm": 0.354621559381485,
      "learning_rate": 0.0001,
      "loss": 1.6693,
      "step": 4005
    },
    {
      "epoch": 0.9728023312287518,
      "grad_norm": 0.368179589509964,
      "learning_rate": 0.0001,
      "loss": 1.7526,
      "step": 4006
    },
    {
      "epoch": 0.9730451675570665,
      "grad_norm": 0.4011801481246948,
      "learning_rate": 0.0001,
      "loss": 1.705,
      "step": 4007
    },
    {
      "epoch": 0.9732880038853813,
      "grad_norm": 0.3765765130519867,
      "learning_rate": 0.0001,
      "loss": 1.6606,
      "step": 4008
    },
    {
      "epoch": 0.973530840213696,
      "grad_norm": 0.38201904296875,
      "learning_rate": 0.0001,
      "loss": 1.7826,
      "step": 4009
    },
    {
      "epoch": 0.9737736765420106,
      "grad_norm": 0.376761257648468,
      "learning_rate": 0.0001,
      "loss": 1.5094,
      "step": 4010
    },
    {
      "epoch": 0.9740165128703254,
      "grad_norm": 0.36620327830314636,
      "learning_rate": 0.0001,
      "loss": 1.6652,
      "step": 4011
    },
    {
      "epoch": 0.9742593491986401,
      "grad_norm": 0.39449259638786316,
      "learning_rate": 0.0001,
      "loss": 1.8538,
      "step": 4012
    },
    {
      "epoch": 0.9745021855269548,
      "grad_norm": 0.4007440507411957,
      "learning_rate": 0.0001,
      "loss": 1.7131,
      "step": 4013
    },
    {
      "epoch": 0.9747450218552696,
      "grad_norm": 0.36265507340431213,
      "learning_rate": 0.0001,
      "loss": 1.5114,
      "step": 4014
    },
    {
      "epoch": 0.9749878581835842,
      "grad_norm": 0.35982275009155273,
      "learning_rate": 0.0001,
      "loss": 1.647,
      "step": 4015
    },
    {
      "epoch": 0.975230694511899,
      "grad_norm": 0.4066230058670044,
      "learning_rate": 0.0001,
      "loss": 1.8601,
      "step": 4016
    },
    {
      "epoch": 0.9754735308402137,
      "grad_norm": 0.3756217658519745,
      "learning_rate": 0.0001,
      "loss": 1.6316,
      "step": 4017
    },
    {
      "epoch": 0.9757163671685284,
      "grad_norm": 0.39345574378967285,
      "learning_rate": 0.0001,
      "loss": 1.7711,
      "step": 4018
    },
    {
      "epoch": 0.9759592034968432,
      "grad_norm": 0.37413719296455383,
      "learning_rate": 0.0001,
      "loss": 1.7526,
      "step": 4019
    },
    {
      "epoch": 0.9762020398251579,
      "grad_norm": 0.37252193689346313,
      "learning_rate": 0.0001,
      "loss": 1.6338,
      "step": 4020
    },
    {
      "epoch": 0.9764448761534725,
      "grad_norm": 0.37571969628334045,
      "learning_rate": 0.0001,
      "loss": 1.6935,
      "step": 4021
    },
    {
      "epoch": 0.9766877124817873,
      "grad_norm": 0.3627742826938629,
      "learning_rate": 0.0001,
      "loss": 1.6191,
      "step": 4022
    },
    {
      "epoch": 0.976930548810102,
      "grad_norm": 0.4085633158683777,
      "learning_rate": 0.0001,
      "loss": 1.7949,
      "step": 4023
    },
    {
      "epoch": 0.9771733851384167,
      "grad_norm": 0.3718832731246948,
      "learning_rate": 0.0001,
      "loss": 1.7685,
      "step": 4024
    },
    {
      "epoch": 0.9774162214667315,
      "grad_norm": 0.3533092439174652,
      "learning_rate": 0.0001,
      "loss": 1.5106,
      "step": 4025
    },
    {
      "epoch": 0.9776590577950461,
      "grad_norm": 0.37770017981529236,
      "learning_rate": 0.0001,
      "loss": 1.702,
      "step": 4026
    },
    {
      "epoch": 0.9779018941233608,
      "grad_norm": 0.37683406472206116,
      "learning_rate": 0.0001,
      "loss": 1.7316,
      "step": 4027
    },
    {
      "epoch": 0.9781447304516756,
      "grad_norm": 0.3591940999031067,
      "learning_rate": 0.0001,
      "loss": 1.6739,
      "step": 4028
    },
    {
      "epoch": 0.9783875667799903,
      "grad_norm": 0.36840593814849854,
      "learning_rate": 0.0001,
      "loss": 1.679,
      "step": 4029
    },
    {
      "epoch": 0.978630403108305,
      "grad_norm": 0.35593312978744507,
      "learning_rate": 0.0001,
      "loss": 1.5049,
      "step": 4030
    },
    {
      "epoch": 0.9788732394366197,
      "grad_norm": 0.3793998062610626,
      "learning_rate": 0.0001,
      "loss": 1.6473,
      "step": 4031
    },
    {
      "epoch": 0.9791160757649344,
      "grad_norm": 0.3793684244155884,
      "learning_rate": 0.0001,
      "loss": 1.7173,
      "step": 4032
    },
    {
      "epoch": 0.9793589120932491,
      "grad_norm": 0.3691100776195526,
      "learning_rate": 0.0001,
      "loss": 1.6269,
      "step": 4033
    },
    {
      "epoch": 0.9796017484215639,
      "grad_norm": 0.3731459379196167,
      "learning_rate": 0.0001,
      "loss": 1.6124,
      "step": 4034
    },
    {
      "epoch": 0.9798445847498786,
      "grad_norm": 0.3664308786392212,
      "learning_rate": 0.0001,
      "loss": 1.6795,
      "step": 4035
    },
    {
      "epoch": 0.9800874210781932,
      "grad_norm": 0.3941652774810791,
      "learning_rate": 0.0001,
      "loss": 1.6496,
      "step": 4036
    },
    {
      "epoch": 0.980330257406508,
      "grad_norm": 0.3622600734233856,
      "learning_rate": 0.0001,
      "loss": 1.6641,
      "step": 4037
    },
    {
      "epoch": 0.9805730937348227,
      "grad_norm": 0.38590314984321594,
      "learning_rate": 0.0001,
      "loss": 1.8057,
      "step": 4038
    },
    {
      "epoch": 0.9808159300631375,
      "grad_norm": 0.4126585125923157,
      "learning_rate": 0.0001,
      "loss": 1.9473,
      "step": 4039
    },
    {
      "epoch": 0.9810587663914522,
      "grad_norm": 0.37464776635169983,
      "learning_rate": 0.0001,
      "loss": 1.7544,
      "step": 4040
    },
    {
      "epoch": 0.9813016027197669,
      "grad_norm": 0.3711651861667633,
      "learning_rate": 0.0001,
      "loss": 1.7174,
      "step": 4041
    },
    {
      "epoch": 0.9815444390480816,
      "grad_norm": 0.36861082911491394,
      "learning_rate": 0.0001,
      "loss": 1.6576,
      "step": 4042
    },
    {
      "epoch": 0.9817872753763963,
      "grad_norm": 0.3674752414226532,
      "learning_rate": 0.0001,
      "loss": 1.8249,
      "step": 4043
    },
    {
      "epoch": 0.982030111704711,
      "grad_norm": 0.3985970616340637,
      "learning_rate": 0.0001,
      "loss": 1.7037,
      "step": 4044
    },
    {
      "epoch": 0.9822729480330258,
      "grad_norm": 0.4111131727695465,
      "learning_rate": 0.0001,
      "loss": 1.8352,
      "step": 4045
    },
    {
      "epoch": 0.9825157843613405,
      "grad_norm": 0.39453092217445374,
      "learning_rate": 0.0001,
      "loss": 1.7642,
      "step": 4046
    },
    {
      "epoch": 0.9827586206896551,
      "grad_norm": 0.37974950671195984,
      "learning_rate": 0.0001,
      "loss": 1.6592,
      "step": 4047
    },
    {
      "epoch": 0.9830014570179699,
      "grad_norm": 0.3616590201854706,
      "learning_rate": 0.0001,
      "loss": 1.7328,
      "step": 4048
    },
    {
      "epoch": 0.9832442933462846,
      "grad_norm": 0.34769493341445923,
      "learning_rate": 0.0001,
      "loss": 1.6637,
      "step": 4049
    },
    {
      "epoch": 0.9834871296745993,
      "grad_norm": 0.36645328998565674,
      "learning_rate": 0.0001,
      "loss": 1.7429,
      "step": 4050
    },
    {
      "epoch": 0.9837299660029141,
      "grad_norm": 0.3879911005496979,
      "learning_rate": 0.0001,
      "loss": 1.689,
      "step": 4051
    },
    {
      "epoch": 0.9839728023312287,
      "grad_norm": 0.3560008108615875,
      "learning_rate": 0.0001,
      "loss": 1.55,
      "step": 4052
    },
    {
      "epoch": 0.9842156386595434,
      "grad_norm": 0.35079431533813477,
      "learning_rate": 0.0001,
      "loss": 1.6343,
      "step": 4053
    },
    {
      "epoch": 0.9844584749878582,
      "grad_norm": 0.3875662684440613,
      "learning_rate": 0.0001,
      "loss": 1.6194,
      "step": 4054
    },
    {
      "epoch": 0.9847013113161729,
      "grad_norm": 0.39650678634643555,
      "learning_rate": 0.0001,
      "loss": 1.8524,
      "step": 4055
    },
    {
      "epoch": 0.9849441476444876,
      "grad_norm": 0.34503695368766785,
      "learning_rate": 0.0001,
      "loss": 1.6514,
      "step": 4056
    },
    {
      "epoch": 0.9851869839728024,
      "grad_norm": 0.3781222403049469,
      "learning_rate": 0.0001,
      "loss": 1.8418,
      "step": 4057
    },
    {
      "epoch": 0.985429820301117,
      "grad_norm": 0.3813125789165497,
      "learning_rate": 0.0001,
      "loss": 1.6916,
      "step": 4058
    },
    {
      "epoch": 0.9856726566294317,
      "grad_norm": 0.3896786570549011,
      "learning_rate": 0.0001,
      "loss": 1.7672,
      "step": 4059
    },
    {
      "epoch": 0.9859154929577465,
      "grad_norm": 0.3770349323749542,
      "learning_rate": 0.0001,
      "loss": 1.6771,
      "step": 4060
    },
    {
      "epoch": 0.9861583292860612,
      "grad_norm": 0.3951159417629242,
      "learning_rate": 0.0001,
      "loss": 1.7921,
      "step": 4061
    },
    {
      "epoch": 0.986401165614376,
      "grad_norm": 0.3489895462989807,
      "learning_rate": 0.0001,
      "loss": 1.5675,
      "step": 4062
    },
    {
      "epoch": 0.9866440019426906,
      "grad_norm": 0.3497614860534668,
      "learning_rate": 0.0001,
      "loss": 1.6035,
      "step": 4063
    },
    {
      "epoch": 0.9868868382710053,
      "grad_norm": 0.3808273375034332,
      "learning_rate": 0.0001,
      "loss": 1.7012,
      "step": 4064
    },
    {
      "epoch": 0.9871296745993201,
      "grad_norm": 0.38426393270492554,
      "learning_rate": 0.0001,
      "loss": 1.6891,
      "step": 4065
    },
    {
      "epoch": 0.9873725109276348,
      "grad_norm": 0.38831883668899536,
      "learning_rate": 0.0001,
      "loss": 1.8116,
      "step": 4066
    },
    {
      "epoch": 0.9876153472559495,
      "grad_norm": 0.3684384524822235,
      "learning_rate": 0.0001,
      "loss": 1.6775,
      "step": 4067
    },
    {
      "epoch": 0.9878581835842642,
      "grad_norm": 0.3812960088253021,
      "learning_rate": 0.0001,
      "loss": 1.729,
      "step": 4068
    },
    {
      "epoch": 0.9881010199125789,
      "grad_norm": 0.35406169295310974,
      "learning_rate": 0.0001,
      "loss": 1.4515,
      "step": 4069
    },
    {
      "epoch": 0.9883438562408936,
      "grad_norm": 0.3798271715641022,
      "learning_rate": 0.0001,
      "loss": 1.7221,
      "step": 4070
    },
    {
      "epoch": 0.9885866925692084,
      "grad_norm": 0.3925783038139343,
      "learning_rate": 0.0001,
      "loss": 1.6974,
      "step": 4071
    },
    {
      "epoch": 0.9888295288975231,
      "grad_norm": 0.3802926540374756,
      "learning_rate": 0.0001,
      "loss": 1.5563,
      "step": 4072
    },
    {
      "epoch": 0.9890723652258377,
      "grad_norm": 0.3390859365463257,
      "learning_rate": 0.0001,
      "loss": 1.5779,
      "step": 4073
    },
    {
      "epoch": 0.9893152015541525,
      "grad_norm": 0.44702109694480896,
      "learning_rate": 0.0001,
      "loss": 1.8813,
      "step": 4074
    },
    {
      "epoch": 0.9895580378824672,
      "grad_norm": 0.37999674677848816,
      "learning_rate": 0.0001,
      "loss": 1.6977,
      "step": 4075
    },
    {
      "epoch": 0.9898008742107819,
      "grad_norm": 0.3615684509277344,
      "learning_rate": 0.0001,
      "loss": 1.6233,
      "step": 4076
    },
    {
      "epoch": 0.9900437105390967,
      "grad_norm": 0.3654703199863434,
      "learning_rate": 0.0001,
      "loss": 1.7714,
      "step": 4077
    },
    {
      "epoch": 0.9902865468674114,
      "grad_norm": 0.3864661157131195,
      "learning_rate": 0.0001,
      "loss": 1.7514,
      "step": 4078
    },
    {
      "epoch": 0.990529383195726,
      "grad_norm": 0.3851757347583771,
      "learning_rate": 0.0001,
      "loss": 1.642,
      "step": 4079
    },
    {
      "epoch": 0.9907722195240408,
      "grad_norm": 0.35797446966171265,
      "learning_rate": 0.0001,
      "loss": 1.7346,
      "step": 4080
    },
    {
      "epoch": 0.9910150558523555,
      "grad_norm": 0.3770246207714081,
      "learning_rate": 0.0001,
      "loss": 1.5539,
      "step": 4081
    },
    {
      "epoch": 0.9912578921806702,
      "grad_norm": 0.3706028461456299,
      "learning_rate": 0.0001,
      "loss": 1.6336,
      "step": 4082
    },
    {
      "epoch": 0.991500728508985,
      "grad_norm": 0.362161248922348,
      "learning_rate": 0.0001,
      "loss": 1.6249,
      "step": 4083
    },
    {
      "epoch": 0.9917435648372996,
      "grad_norm": 0.34968650341033936,
      "learning_rate": 0.0001,
      "loss": 1.5998,
      "step": 4084
    },
    {
      "epoch": 0.9919864011656144,
      "grad_norm": 0.38522493839263916,
      "learning_rate": 0.0001,
      "loss": 1.7859,
      "step": 4085
    },
    {
      "epoch": 0.9922292374939291,
      "grad_norm": 0.3981729745864868,
      "learning_rate": 0.0001,
      "loss": 1.6727,
      "step": 4086
    },
    {
      "epoch": 0.9924720738222438,
      "grad_norm": 0.3760201632976532,
      "learning_rate": 0.0001,
      "loss": 1.6928,
      "step": 4087
    },
    {
      "epoch": 0.9927149101505586,
      "grad_norm": 0.3839087188243866,
      "learning_rate": 0.0001,
      "loss": 1.8002,
      "step": 4088
    },
    {
      "epoch": 0.9929577464788732,
      "grad_norm": 0.4088483452796936,
      "learning_rate": 0.0001,
      "loss": 1.8349,
      "step": 4089
    },
    {
      "epoch": 0.9932005828071879,
      "grad_norm": 0.37059393525123596,
      "learning_rate": 0.0001,
      "loss": 1.7424,
      "step": 4090
    },
    {
      "epoch": 0.9934434191355027,
      "grad_norm": 0.4004857838153839,
      "learning_rate": 0.0001,
      "loss": 1.7749,
      "step": 4091
    },
    {
      "epoch": 0.9936862554638174,
      "grad_norm": 0.37603604793548584,
      "learning_rate": 0.0001,
      "loss": 1.624,
      "step": 4092
    },
    {
      "epoch": 0.9939290917921321,
      "grad_norm": 0.38639020919799805,
      "learning_rate": 0.0001,
      "loss": 1.5217,
      "step": 4093
    },
    {
      "epoch": 0.9941719281204469,
      "grad_norm": 0.3561013340950012,
      "learning_rate": 0.0001,
      "loss": 1.6206,
      "step": 4094
    },
    {
      "epoch": 0.9944147644487615,
      "grad_norm": 0.38830888271331787,
      "learning_rate": 0.0001,
      "loss": 1.7821,
      "step": 4095
    },
    {
      "epoch": 0.9946576007770762,
      "grad_norm": 0.37778744101524353,
      "learning_rate": 0.0001,
      "loss": 1.6911,
      "step": 4096
    },
    {
      "epoch": 0.994900437105391,
      "grad_norm": 0.3863472044467926,
      "learning_rate": 0.0001,
      "loss": 1.7793,
      "step": 4097
    },
    {
      "epoch": 0.9951432734337057,
      "grad_norm": 0.39536169171333313,
      "learning_rate": 0.0001,
      "loss": 1.8105,
      "step": 4098
    },
    {
      "epoch": 0.9953861097620204,
      "grad_norm": 0.38470953702926636,
      "learning_rate": 0.0001,
      "loss": 1.6353,
      "step": 4099
    },
    {
      "epoch": 0.9956289460903351,
      "grad_norm": 0.36688435077667236,
      "learning_rate": 0.0001,
      "loss": 1.8089,
      "step": 4100
    },
    {
      "epoch": 0.9958717824186498,
      "grad_norm": 0.3831002116203308,
      "learning_rate": 0.0001,
      "loss": 1.8249,
      "step": 4101
    },
    {
      "epoch": 0.9961146187469645,
      "grad_norm": 0.3933352530002594,
      "learning_rate": 0.0001,
      "loss": 1.7054,
      "step": 4102
    },
    {
      "epoch": 0.9963574550752793,
      "grad_norm": 0.36921703815460205,
      "learning_rate": 0.0001,
      "loss": 1.6879,
      "step": 4103
    },
    {
      "epoch": 0.996600291403594,
      "grad_norm": 0.39410069584846497,
      "learning_rate": 0.0001,
      "loss": 1.7148,
      "step": 4104
    },
    {
      "epoch": 0.9968431277319086,
      "grad_norm": 0.3676906228065491,
      "learning_rate": 0.0001,
      "loss": 1.6285,
      "step": 4105
    },
    {
      "epoch": 0.9970859640602234,
      "grad_norm": 0.3701580762863159,
      "learning_rate": 0.0001,
      "loss": 1.6518,
      "step": 4106
    },
    {
      "epoch": 0.9973288003885381,
      "grad_norm": 0.36121150851249695,
      "learning_rate": 0.0001,
      "loss": 1.6662,
      "step": 4107
    },
    {
      "epoch": 0.9975716367168529,
      "grad_norm": 0.36982738971710205,
      "learning_rate": 0.0001,
      "loss": 1.6054,
      "step": 4108
    },
    {
      "epoch": 0.9978144730451676,
      "grad_norm": 0.37975645065307617,
      "learning_rate": 0.0001,
      "loss": 1.7856,
      "step": 4109
    },
    {
      "epoch": 0.9980573093734822,
      "grad_norm": 0.3816273808479309,
      "learning_rate": 0.0001,
      "loss": 1.5978,
      "step": 4110
    },
    {
      "epoch": 0.998300145701797,
      "grad_norm": 0.3661021292209625,
      "learning_rate": 0.0001,
      "loss": 1.6231,
      "step": 4111
    },
    {
      "epoch": 0.9985429820301117,
      "grad_norm": 0.3899565041065216,
      "learning_rate": 0.0001,
      "loss": 1.7643,
      "step": 4112
    },
    {
      "epoch": 0.9987858183584264,
      "grad_norm": 0.39457520842552185,
      "learning_rate": 0.0001,
      "loss": 1.8735,
      "step": 4113
    },
    {
      "epoch": 0.9990286546867412,
      "grad_norm": 0.3817576766014099,
      "learning_rate": 0.0001,
      "loss": 1.7596,
      "step": 4114
    },
    {
      "epoch": 0.9992714910150559,
      "grad_norm": 0.4239327609539032,
      "learning_rate": 0.0001,
      "loss": 1.9292,
      "step": 4115
    },
    {
      "epoch": 0.9995143273433705,
      "grad_norm": 0.4147722125053406,
      "learning_rate": 0.0001,
      "loss": 1.7747,
      "step": 4116
    },
    {
      "epoch": 0.9997571636716853,
      "grad_norm": 0.3927810788154602,
      "learning_rate": 0.0001,
      "loss": 1.6095,
      "step": 4117
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.3737819790840149,
      "learning_rate": 0.0001,
      "loss": 1.7487,
      "step": 4118
    },
    {
      "epoch": 1.0,
      "step": 4118,
      "total_flos": 6.212058917275435e+18,
      "train_loss": 1.7339236946173118,
      "train_runtime": 80202.1439,
      "train_samples_per_second": 0.205,
      "train_steps_per_second": 0.051
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 4118,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 6.212058917275435e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}