{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.996749729144095,
  "eval_steps": 500,
  "global_step": 1383,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 16.709853484760444,
      "learning_rate": 2.3809523809523811e-07,
      "loss": 1.8301,
      "step": 1
    },
    {
      "epoch": 0.0,
      "grad_norm": 17.76074099398955,
      "learning_rate": 4.7619047619047623e-07,
      "loss": 1.8182,
      "step": 2
    },
    {
      "epoch": 0.01,
      "grad_norm": 12.12137117799721,
      "learning_rate": 7.142857142857143e-07,
      "loss": 1.5935,
      "step": 3
    },
    {
      "epoch": 0.01,
      "grad_norm": 18.819806190896873,
      "learning_rate": 9.523809523809525e-07,
      "loss": 1.922,
      "step": 4
    },
    {
      "epoch": 0.01,
      "grad_norm": 123.84026713377355,
      "learning_rate": 1.1904761904761906e-06,
      "loss": 2.2824,
      "step": 5
    },
    {
      "epoch": 0.01,
      "grad_norm": 24.027763659744394,
      "learning_rate": 1.4285714285714286e-06,
      "loss": 2.6041,
      "step": 6
    },
    {
      "epoch": 0.02,
      "grad_norm": 13.153268703370804,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 1.5593,
      "step": 7
    },
    {
      "epoch": 0.02,
      "grad_norm": 27.1264107206067,
      "learning_rate": 1.904761904761905e-06,
      "loss": 1.5941,
      "step": 8
    },
    {
      "epoch": 0.02,
      "grad_norm": 94.53444983832594,
      "learning_rate": 2.1428571428571427e-06,
      "loss": 1.6793,
      "step": 9
    },
    {
      "epoch": 0.02,
      "grad_norm": 20.694564890213154,
      "learning_rate": 2.380952380952381e-06,
      "loss": 2.1942,
      "step": 10
    },
    {
      "epoch": 0.02,
      "grad_norm": 13.512498569645478,
      "learning_rate": 2.6190476190476192e-06,
      "loss": 1.3492,
      "step": 11
    },
    {
      "epoch": 0.03,
      "grad_norm": 15.390180239158552,
      "learning_rate": 2.8571428571428573e-06,
      "loss": 1.417,
      "step": 12
    },
    {
      "epoch": 0.03,
      "grad_norm": 46.49955729937121,
      "learning_rate": 3.0952380952380957e-06,
      "loss": 2.2055,
      "step": 13
    },
    {
      "epoch": 0.03,
      "grad_norm": 12.788713482669845,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.472,
      "step": 14
    },
    {
      "epoch": 0.03,
      "grad_norm": 34.647064331379156,
      "learning_rate": 3.5714285714285718e-06,
      "loss": 1.894,
      "step": 15
    },
    {
      "epoch": 0.03,
      "grad_norm": 15.723778004088178,
      "learning_rate": 3.80952380952381e-06,
      "loss": 1.8272,
      "step": 16
    },
    {
      "epoch": 0.04,
      "grad_norm": 13.415386634497915,
      "learning_rate": 4.047619047619048e-06,
      "loss": 2.0885,
      "step": 17
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.183680122844404,
      "learning_rate": 4.2857142857142855e-06,
      "loss": 1.5831,
      "step": 18
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.561139019796537,
      "learning_rate": 4.523809523809524e-06,
      "loss": 1.4041,
      "step": 19
    },
    {
      "epoch": 0.04,
      "grad_norm": 12.114015847517962,
      "learning_rate": 4.761904761904762e-06,
      "loss": 2.019,
      "step": 20
    },
    {
      "epoch": 0.05,
      "grad_norm": 15.544777014735105,
      "learning_rate": 5e-06,
      "loss": 1.8317,
      "step": 21
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.248644488479986,
      "learning_rate": 5.2380952380952384e-06,
      "loss": 1.0241,
      "step": 22
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.794172629391284,
      "learning_rate": 5.476190476190477e-06,
      "loss": 1.0715,
      "step": 23
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.60123146591206,
      "learning_rate": 5.7142857142857145e-06,
      "loss": 0.8624,
      "step": 24
    },
    {
      "epoch": 0.05,
      "grad_norm": 6.876822151989595,
      "learning_rate": 5.9523809523809525e-06,
      "loss": 0.9872,
      "step": 25
    },
    {
      "epoch": 0.06,
      "grad_norm": 65.23342078707829,
      "learning_rate": 6.1904761904761914e-06,
      "loss": 1.5916,
      "step": 26
    },
    {
      "epoch": 0.06,
      "grad_norm": 19.716248969557167,
      "learning_rate": 6.4285714285714295e-06,
      "loss": 1.2537,
      "step": 27
    },
    {
      "epoch": 0.06,
      "grad_norm": 40.309027906143484,
      "learning_rate": 6.666666666666667e-06,
      "loss": 1.4646,
      "step": 28
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.23572597505871,
      "learning_rate": 6.9047619047619055e-06,
      "loss": 1.2538,
      "step": 29
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.78421674520985,
      "learning_rate": 7.1428571428571436e-06,
      "loss": 1.0255,
      "step": 30
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.54339528361063,
      "learning_rate": 7.380952380952382e-06,
      "loss": 1.5354,
      "step": 31
    },
    {
      "epoch": 0.07,
      "grad_norm": 10.933138768794096,
      "learning_rate": 7.61904761904762e-06,
      "loss": 1.1058,
      "step": 32
    },
    {
      "epoch": 0.07,
      "grad_norm": 11.4160352478607,
      "learning_rate": 7.857142857142858e-06,
      "loss": 1.0351,
      "step": 33
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.129012473979065,
      "learning_rate": 8.095238095238097e-06,
      "loss": 1.2369,
      "step": 34
    },
    {
      "epoch": 0.08,
      "grad_norm": 7.781606997214579,
      "learning_rate": 8.333333333333334e-06,
      "loss": 1.1036,
      "step": 35
    },
    {
      "epoch": 0.08,
      "grad_norm": 10.347149816110733,
      "learning_rate": 8.571428571428571e-06,
      "loss": 1.068,
      "step": 36
    },
    {
      "epoch": 0.08,
      "grad_norm": 20.6600053331111,
      "learning_rate": 8.80952380952381e-06,
      "loss": 1.0292,
      "step": 37
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.833730122259707,
      "learning_rate": 9.047619047619049e-06,
      "loss": 1.3301,
      "step": 38
    },
    {
      "epoch": 0.08,
      "grad_norm": 46.26871271599295,
      "learning_rate": 9.285714285714288e-06,
      "loss": 1.1459,
      "step": 39
    },
    {
      "epoch": 0.09,
      "grad_norm": 6.647375193132848,
      "learning_rate": 9.523809523809525e-06,
      "loss": 1.0104,
      "step": 40
    },
    {
      "epoch": 0.09,
      "grad_norm": 10.211938763321209,
      "learning_rate": 9.761904761904762e-06,
      "loss": 0.9854,
      "step": 41
    },
    {
      "epoch": 0.09,
      "grad_norm": 9.423016780892906,
      "learning_rate": 1e-05,
      "loss": 1.0041,
      "step": 42
    },
    {
      "epoch": 0.09,
      "grad_norm": 6.378281475980182,
      "learning_rate": 9.999986279118938e-06,
      "loss": 0.7482,
      "step": 43
    },
    {
      "epoch": 0.1,
      "grad_norm": 11.967973427715416,
      "learning_rate": 9.999945116551056e-06,
      "loss": 1.209,
      "step": 44
    },
    {
      "epoch": 0.1,
      "grad_norm": 15.013362990686305,
      "learning_rate": 9.999876512522269e-06,
      "loss": 0.9791,
      "step": 45
    },
    {
      "epoch": 0.1,
      "grad_norm": 22.346200515629217,
      "learning_rate": 9.9997804674091e-06,
      "loss": 1.1579,
      "step": 46
    },
    {
      "epoch": 0.1,
      "grad_norm": 15.473304342880718,
      "learning_rate": 9.999656981738679e-06,
      "loss": 1.1002,
      "step": 47
    },
    {
      "epoch": 0.1,
      "grad_norm": 13.637556586048346,
      "learning_rate": 9.999506056188736e-06,
      "loss": 1.028,
      "step": 48
    },
    {
      "epoch": 0.11,
      "grad_norm": 11.87068363113673,
      "learning_rate": 9.999327691587609e-06,
      "loss": 1.184,
      "step": 49
    },
    {
      "epoch": 0.11,
      "grad_norm": 12.962466012025889,
      "learning_rate": 9.99912188891422e-06,
      "loss": 0.7548,
      "step": 50
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.870472049099167,
      "learning_rate": 9.99888864929809e-06,
      "loss": 0.7095,
      "step": 51
    },
    {
      "epoch": 0.11,
      "grad_norm": 7.769393838862314,
      "learning_rate": 9.998627974019322e-06,
      "loss": 0.8385,
      "step": 52
    },
    {
      "epoch": 0.11,
      "grad_norm": 10.146788682294355,
      "learning_rate": 9.99833986450859e-06,
      "loss": 0.8668,
      "step": 53
    },
    {
      "epoch": 0.12,
      "grad_norm": 16.342968809803455,
      "learning_rate": 9.99802432234714e-06,
      "loss": 1.0532,
      "step": 54
    },
    {
      "epoch": 0.12,
      "grad_norm": 7.941859879040072,
      "learning_rate": 9.997681349266782e-06,
      "loss": 0.7458,
      "step": 55
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.082256291790866,
      "learning_rate": 9.997310947149872e-06,
      "loss": 1.1143,
      "step": 56
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.362992453741251,
      "learning_rate": 9.996913118029306e-06,
      "loss": 0.7627,
      "step": 57
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.127557425933743,
      "learning_rate": 9.996487864088512e-06,
      "loss": 0.8977,
      "step": 58
    },
    {
      "epoch": 0.13,
      "grad_norm": 13.41587985751095,
      "learning_rate": 9.996035187661433e-06,
      "loss": 0.7937,
      "step": 59
    },
    {
      "epoch": 0.13,
      "grad_norm": 7.587596574021975,
      "learning_rate": 9.995555091232516e-06,
      "loss": 1.144,
      "step": 60
    },
    {
      "epoch": 0.13,
      "grad_norm": 6.417800794002112,
      "learning_rate": 9.9950475774367e-06,
      "loss": 1.0541,
      "step": 61
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.372188268539198,
      "learning_rate": 9.994512649059401e-06,
      "loss": 0.9625,
      "step": 62
    },
    {
      "epoch": 0.14,
      "grad_norm": 15.336487087329994,
      "learning_rate": 9.99395030903649e-06,
      "loss": 1.2699,
      "step": 63
    },
    {
      "epoch": 0.14,
      "grad_norm": 7.636482601575623,
      "learning_rate": 9.993360560454293e-06,
      "loss": 0.9842,
      "step": 64
    },
    {
      "epoch": 0.14,
      "grad_norm": 19.589267733596557,
      "learning_rate": 9.992743406549556e-06,
      "loss": 0.7281,
      "step": 65
    },
    {
      "epoch": 0.14,
      "grad_norm": 16.420257619016123,
      "learning_rate": 9.992098850709434e-06,
      "loss": 0.9726,
      "step": 66
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.376109096365017,
      "learning_rate": 9.99142689647148e-06,
      "loss": 0.767,
      "step": 67
    },
    {
      "epoch": 0.15,
      "grad_norm": 6.933235299280333,
      "learning_rate": 9.990727547523616e-06,
      "loss": 0.8517,
      "step": 68
    },
    {
      "epoch": 0.15,
      "grad_norm": 11.73047596652497,
      "learning_rate": 9.990000807704114e-06,
      "loss": 1.0491,
      "step": 69
    },
    {
      "epoch": 0.15,
      "grad_norm": 7.905669593527127,
      "learning_rate": 9.989246681001577e-06,
      "loss": 0.6869,
      "step": 70
    },
    {
      "epoch": 0.15,
      "grad_norm": 8.555118643397824,
      "learning_rate": 9.988465171554921e-06,
      "loss": 1.0503,
      "step": 71
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.765801442864587,
      "learning_rate": 9.987656283653344e-06,
      "loss": 0.7513,
      "step": 72
    },
    {
      "epoch": 0.16,
      "grad_norm": 6.613609355206775,
      "learning_rate": 9.986820021736306e-06,
      "loss": 1.3825,
      "step": 73
    },
    {
      "epoch": 0.16,
      "grad_norm": 7.139915646581661,
      "learning_rate": 9.985956390393511e-06,
      "loss": 0.7673,
      "step": 74
    },
    {
      "epoch": 0.16,
      "grad_norm": 7.030138436848795,
      "learning_rate": 9.985065394364869e-06,
      "loss": 0.9911,
      "step": 75
    },
    {
      "epoch": 0.16,
      "grad_norm": 5.489349297862267,
      "learning_rate": 9.984147038540482e-06,
      "loss": 0.6724,
      "step": 76
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.991520253457708,
      "learning_rate": 9.983201327960607e-06,
      "loss": 0.9831,
      "step": 77
    },
    {
      "epoch": 0.17,
      "grad_norm": 10.021752398596858,
      "learning_rate": 9.982228267815644e-06,
      "loss": 1.0091,
      "step": 78
    },
    {
      "epoch": 0.17,
      "grad_norm": 9.222009010527715,
      "learning_rate": 9.981227863446082e-06,
      "loss": 0.766,
      "step": 79
    },
    {
      "epoch": 0.17,
      "grad_norm": 9.356349516249546,
      "learning_rate": 9.980200120342499e-06,
      "loss": 1.0214,
      "step": 80
    },
    {
      "epoch": 0.18,
      "grad_norm": 12.335354552339782,
      "learning_rate": 9.979145044145506e-06,
      "loss": 0.8724,
      "step": 81
    },
    {
      "epoch": 0.18,
      "grad_norm": 12.661542419985505,
      "learning_rate": 9.978062640645737e-06,
      "loss": 0.8929,
      "step": 82
    },
    {
      "epoch": 0.18,
      "grad_norm": 8.110473111865039,
      "learning_rate": 9.976952915783804e-06,
      "loss": 0.8315,
      "step": 83
    },
    {
      "epoch": 0.18,
      "grad_norm": 6.794559175947099,
      "learning_rate": 9.975815875650265e-06,
      "loss": 0.8811,
      "step": 84
    },
    {
      "epoch": 0.18,
      "grad_norm": 7.11765902706106,
      "learning_rate": 9.9746515264856e-06,
      "loss": 1.0472,
      "step": 85
    },
    {
      "epoch": 0.19,
      "grad_norm": 7.908404155306792,
      "learning_rate": 9.973459874680167e-06,
      "loss": 0.7223,
      "step": 86
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.174144453601226,
      "learning_rate": 9.972240926774167e-06,
      "loss": 0.9243,
      "step": 87
    },
    {
      "epoch": 0.19,
      "grad_norm": 7.375878883309969,
      "learning_rate": 9.970994689457623e-06,
      "loss": 0.5092,
      "step": 88
    },
    {
      "epoch": 0.19,
      "grad_norm": 12.430330322024414,
      "learning_rate": 9.969721169570319e-06,
      "loss": 0.9366,
      "step": 89
    },
    {
      "epoch": 0.2,
      "grad_norm": 7.302509671721146,
      "learning_rate": 9.968420374101782e-06,
      "loss": 1.0668,
      "step": 90
    },
    {
      "epoch": 0.2,
      "grad_norm": 7.968141735213008,
      "learning_rate": 9.967092310191237e-06,
      "loss": 0.6861,
      "step": 91
    },
    {
      "epoch": 0.2,
      "grad_norm": 5.538191349344046,
      "learning_rate": 9.965736985127568e-06,
      "loss": 0.5609,
      "step": 92
    },
    {
      "epoch": 0.2,
      "grad_norm": 4.8112824775455305,
      "learning_rate": 9.964354406349272e-06,
      "loss": 1.2046,
      "step": 93
    },
    {
      "epoch": 0.2,
      "grad_norm": 4.884064109722187,
      "learning_rate": 9.962944581444433e-06,
      "loss": 0.6504,
      "step": 94
    },
    {
      "epoch": 0.21,
      "grad_norm": 20.67672184687602,
      "learning_rate": 9.961507518150666e-06,
      "loss": 0.6738,
      "step": 95
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.688860261045402,
      "learning_rate": 9.960043224355081e-06,
      "loss": 1.2039,
      "step": 96
    },
    {
      "epoch": 0.21,
      "grad_norm": 7.126907539863814,
      "learning_rate": 9.958551708094237e-06,
      "loss": 0.784,
      "step": 97
    },
    {
      "epoch": 0.21,
      "grad_norm": 11.59800774389457,
      "learning_rate": 9.9570329775541e-06,
      "loss": 0.6421,
      "step": 98
    },
    {
      "epoch": 0.21,
      "grad_norm": 5.018966602413883,
      "learning_rate": 9.955487041070003e-06,
      "loss": 0.962,
      "step": 99
    },
    {
      "epoch": 0.22,
      "grad_norm": 7.097399514150202,
      "learning_rate": 9.953913907126584e-06,
      "loss": 0.7477,
      "step": 100
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.978733314858584,
      "learning_rate": 9.952313584357763e-06,
      "loss": 0.8233,
      "step": 101
    },
    {
      "epoch": 0.22,
      "grad_norm": 5.802927932563751,
      "learning_rate": 9.95068608154667e-06,
      "loss": 0.6661,
      "step": 102
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.444880658637484,
      "learning_rate": 9.949031407625616e-06,
      "loss": 1.0689,
      "step": 103
    },
    {
      "epoch": 0.23,
      "grad_norm": 6.56746988799901,
      "learning_rate": 9.947349571676037e-06,
      "loss": 1.0204,
      "step": 104
    },
    {
      "epoch": 0.23,
      "grad_norm": 5.20829934091558,
      "learning_rate": 9.945640582928438e-06,
      "loss": 1.265,
      "step": 105
    },
    {
      "epoch": 0.23,
      "grad_norm": 7.542342410431212,
      "learning_rate": 9.943904450762351e-06,
      "loss": 1.0481,
      "step": 106
    },
    {
      "epoch": 0.23,
      "grad_norm": 6.793272653060435,
      "learning_rate": 9.942141184706286e-06,
      "loss": 0.9086,
      "step": 107
    },
    {
      "epoch": 0.23,
      "grad_norm": 7.0840611126564115,
      "learning_rate": 9.940350794437663e-06,
      "loss": 0.7843,
      "step": 108
    },
    {
      "epoch": 0.24,
      "grad_norm": 13.848611795807416,
      "learning_rate": 9.938533289782778e-06,
      "loss": 0.5617,
      "step": 109
    },
    {
      "epoch": 0.24,
      "grad_norm": 12.486475663507044,
      "learning_rate": 9.936688680716737e-06,
      "loss": 1.1011,
      "step": 110
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.347823071118775,
      "learning_rate": 9.934816977363404e-06,
      "loss": 0.6698,
      "step": 111
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.154987399033947,
      "learning_rate": 9.932918189995345e-06,
      "loss": 0.8165,
      "step": 112
    },
    {
      "epoch": 0.24,
      "grad_norm": 6.058318739999231,
      "learning_rate": 9.930992329033777e-06,
      "loss": 0.6749,
      "step": 113
    },
    {
      "epoch": 0.25,
      "grad_norm": 21.22549560471258,
      "learning_rate": 9.929039405048502e-06,
      "loss": 0.9771,
      "step": 114
    },
    {
      "epoch": 0.25,
      "grad_norm": 5.983590522505441,
      "learning_rate": 9.927059428757857e-06,
      "loss": 0.7692,
      "step": 115
    },
    {
      "epoch": 0.25,
      "grad_norm": 7.3296331738631455,
      "learning_rate": 9.925052411028646e-06,
      "loss": 0.7862,
      "step": 116
    },
    {
      "epoch": 0.25,
      "grad_norm": 4.95611639385475,
      "learning_rate": 9.923018362876093e-06,
      "loss": 0.8135,
      "step": 117
    },
    {
      "epoch": 0.26,
      "grad_norm": 5.7796285184998055,
      "learning_rate": 9.920957295463772e-06,
      "loss": 1.0534,
      "step": 118
    },
    {
      "epoch": 0.26,
      "grad_norm": 6.023107868325316,
      "learning_rate": 9.918869220103542e-06,
      "loss": 1.1525,
      "step": 119
    },
    {
      "epoch": 0.26,
      "grad_norm": 5.368831486399966,
      "learning_rate": 9.916754148255501e-06,
      "loss": 0.8707,
      "step": 120
    },
    {
      "epoch": 0.26,
      "grad_norm": 12.221674004747765,
      "learning_rate": 9.914612091527908e-06,
      "loss": 0.9087,
      "step": 121
    },
    {
      "epoch": 0.26,
      "grad_norm": 10.199165857227971,
      "learning_rate": 9.912443061677125e-06,
      "loss": 0.7072,
      "step": 122
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.567216978746918,
      "learning_rate": 9.91024707060755e-06,
      "loss": 1.3963,
      "step": 123
    },
    {
      "epoch": 0.27,
      "grad_norm": 5.853800213263277,
      "learning_rate": 9.90802413037156e-06,
      "loss": 0.9134,
      "step": 124
    },
    {
      "epoch": 0.27,
      "grad_norm": 7.378918810527769,
      "learning_rate": 9.905774253169433e-06,
      "loss": 0.8595,
      "step": 125
    },
    {
      "epoch": 0.27,
      "grad_norm": 12.341225926537351,
      "learning_rate": 9.903497451349286e-06,
      "loss": 0.9243,
      "step": 126
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.248944845290005,
      "learning_rate": 9.901193737407011e-06,
      "loss": 0.8358,
      "step": 127
    },
    {
      "epoch": 0.28,
      "grad_norm": 4.258032575953594,
      "learning_rate": 9.898863123986203e-06,
      "loss": 0.5628,
      "step": 128
    },
    {
      "epoch": 0.28,
      "grad_norm": 7.21784397219793,
      "learning_rate": 9.896505623878088e-06,
      "loss": 1.3182,
      "step": 129
    },
    {
      "epoch": 0.28,
      "grad_norm": 6.329875700796257,
      "learning_rate": 9.89412125002146e-06,
      "loss": 0.8512,
      "step": 130
    },
    {
      "epoch": 0.28,
      "grad_norm": 10.581720625708526,
      "learning_rate": 9.8917100155026e-06,
      "loss": 1.0895,
      "step": 131
    },
    {
      "epoch": 0.29,
      "grad_norm": 6.137416355476592,
      "learning_rate": 9.889271933555214e-06,
      "loss": 0.8711,
      "step": 132
    },
    {
      "epoch": 0.29,
      "grad_norm": 5.0422201530433775,
      "learning_rate": 9.886807017560356e-06,
      "loss": 0.6656,
      "step": 133
    },
    {
      "epoch": 0.29,
      "grad_norm": 7.987457474658196,
      "learning_rate": 9.884315281046352e-06,
      "loss": 0.5781,
      "step": 134
    },
    {
      "epoch": 0.29,
      "grad_norm": 6.960528452954194,
      "learning_rate": 9.881796737688732e-06,
      "loss": 0.926,
      "step": 135
    },
    {
      "epoch": 0.29,
      "grad_norm": 11.201292617344443,
      "learning_rate": 9.879251401310148e-06,
      "loss": 0.6939,
      "step": 136
    },
    {
      "epoch": 0.3,
      "grad_norm": 5.1492310813022675,
      "learning_rate": 9.876679285880304e-06,
      "loss": 0.5426,
      "step": 137
    },
    {
      "epoch": 0.3,
      "grad_norm": 6.038882982259862,
      "learning_rate": 9.874080405515874e-06,
      "loss": 1.1484,
      "step": 138
    },
    {
      "epoch": 0.3,
      "grad_norm": 20.90304939482476,
      "learning_rate": 9.871454774480433e-06,
      "loss": 0.9477,
      "step": 139
    },
    {
      "epoch": 0.3,
      "grad_norm": 13.886241154862255,
      "learning_rate": 9.868802407184367e-06,
      "loss": 0.781,
      "step": 140
    },
    {
      "epoch": 0.31,
      "grad_norm": 7.07234206061503,
      "learning_rate": 9.866123318184803e-06,
      "loss": 0.9407,
      "step": 141
    },
    {
      "epoch": 0.31,
      "grad_norm": 6.094062520686632,
      "learning_rate": 9.863417522185525e-06,
      "loss": 0.5684,
      "step": 142
    },
    {
      "epoch": 0.31,
      "grad_norm": 7.455569887584047,
      "learning_rate": 9.860685034036897e-06,
      "loss": 1.1929,
      "step": 143
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.442237940285388,
      "learning_rate": 9.857925868735774e-06,
      "loss": 0.7326,
      "step": 144
    },
    {
      "epoch": 0.31,
      "grad_norm": 13.685907236075884,
      "learning_rate": 9.855140041425428e-06,
      "loss": 0.7155,
      "step": 145
    },
    {
      "epoch": 0.32,
      "grad_norm": 5.112555033407144,
      "learning_rate": 9.852327567395463e-06,
      "loss": 0.7351,
      "step": 146
    },
    {
      "epoch": 0.32,
      "grad_norm": 11.794842020685188,
      "learning_rate": 9.84948846208173e-06,
      "loss": 1.0658,
      "step": 147
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.531685603474545,
      "learning_rate": 9.846622741066232e-06,
      "loss": 0.7198,
      "step": 148
    },
    {
      "epoch": 0.32,
      "grad_norm": 7.666766380067266,
      "learning_rate": 9.843730420077061e-06,
      "loss": 0.6356,
      "step": 149
    },
    {
      "epoch": 0.33,
      "grad_norm": 5.876115120371116,
      "learning_rate": 9.840811514988294e-06,
      "loss": 0.6298,
      "step": 150
    },
    {
      "epoch": 0.33,
      "grad_norm": 7.6182072048956755,
      "learning_rate": 9.83786604181991e-06,
      "loss": 0.786,
      "step": 151
    },
    {
      "epoch": 0.33,
      "grad_norm": 7.866749635724763,
      "learning_rate": 9.834894016737705e-06,
      "loss": 1.1082,
      "step": 152
    },
    {
      "epoch": 0.33,
      "grad_norm": 8.526130765296307,
      "learning_rate": 9.831895456053197e-06,
      "loss": 0.8268,
      "step": 153
    },
    {
      "epoch": 0.33,
      "grad_norm": 8.261244862246361,
      "learning_rate": 9.828870376223546e-06,
      "loss": 0.8857,
      "step": 154
    },
    {
      "epoch": 0.34,
      "grad_norm": 6.735474990735441,
      "learning_rate": 9.825818793851456e-06,
      "loss": 0.7115,
      "step": 155
    },
    {
      "epoch": 0.34,
      "grad_norm": 8.768196144026826,
      "learning_rate": 9.822740725685087e-06,
      "loss": 0.9181,
      "step": 156
    },
    {
      "epoch": 0.34,
      "grad_norm": 6.3683851898909865,
      "learning_rate": 9.819636188617961e-06,
      "loss": 1.2787,
      "step": 157
    },
    {
      "epoch": 0.34,
      "grad_norm": 5.1773569212347414,
      "learning_rate": 9.81650519968887e-06,
      "loss": 0.8574,
      "step": 158
    },
    {
      "epoch": 0.34,
      "grad_norm": 6.310766361065571,
      "learning_rate": 9.81334777608179e-06,
      "loss": 1.4374,
      "step": 159
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.206477294414363,
      "learning_rate": 9.810163935125768e-06,
      "loss": 0.7034,
      "step": 160
    },
    {
      "epoch": 0.35,
      "grad_norm": 11.123431403479872,
      "learning_rate": 9.806953694294849e-06,
      "loss": 1.4288,
      "step": 161
    },
    {
      "epoch": 0.35,
      "grad_norm": 5.862609413415919,
      "learning_rate": 9.803717071207965e-06,
      "loss": 0.7613,
      "step": 162
    },
    {
      "epoch": 0.35,
      "grad_norm": 12.098717632708196,
      "learning_rate": 9.800454083628845e-06,
      "loss": 1.0396,
      "step": 163
    },
    {
      "epoch": 0.36,
      "grad_norm": 5.769556222192253,
      "learning_rate": 9.797164749465915e-06,
      "loss": 0.9211,
      "step": 164
    },
    {
      "epoch": 0.36,
      "grad_norm": 4.975746377742021,
      "learning_rate": 9.793849086772198e-06,
      "loss": 0.8666,
      "step": 165
    },
    {
      "epoch": 0.36,
      "grad_norm": 6.480223255517235,
      "learning_rate": 9.790507113745222e-06,
      "loss": 0.8274,
      "step": 166
    },
    {
      "epoch": 0.36,
      "grad_norm": 4.777496147161274,
      "learning_rate": 9.787138848726912e-06,
      "loss": 0.4873,
      "step": 167
    },
    {
      "epoch": 0.36,
      "grad_norm": 6.665695467411743,
      "learning_rate": 9.783744310203492e-06,
      "loss": 1.1125,
      "step": 168
    },
    {
      "epoch": 0.37,
      "grad_norm": 6.021810813078232,
      "learning_rate": 9.780323516805386e-06,
      "loss": 0.61,
      "step": 169
    },
    {
      "epoch": 0.37,
      "grad_norm": 5.177138524550743,
      "learning_rate": 9.776876487307115e-06,
      "loss": 0.7478,
      "step": 170
    },
    {
      "epoch": 0.37,
      "grad_norm": 6.168093726996426,
      "learning_rate": 9.77340324062719e-06,
      "loss": 0.8012,
      "step": 171
    },
    {
      "epoch": 0.37,
      "grad_norm": 5.6090394609599565,
      "learning_rate": 9.769903795828016e-06,
      "loss": 0.6071,
      "step": 172
    },
    {
      "epoch": 0.37,
      "grad_norm": 6.021333570468785,
      "learning_rate": 9.766378172115775e-06,
      "loss": 0.5914,
      "step": 173
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.465416271692636,
      "learning_rate": 9.76282638884034e-06,
      "loss": 0.9138,
      "step": 174
    },
    {
      "epoch": 0.38,
      "grad_norm": 4.995763393453568,
      "learning_rate": 9.75924846549514e-06,
      "loss": 0.7426,
      "step": 175
    },
    {
      "epoch": 0.38,
      "grad_norm": 5.510409673936295,
      "learning_rate": 9.755644421717083e-06,
      "loss": 1.0583,
      "step": 176
    },
    {
      "epoch": 0.38,
      "grad_norm": 17.15352935488223,
      "learning_rate": 9.752014277286433e-06,
      "loss": 1.4155,
      "step": 177
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.550985130782966,
      "learning_rate": 9.7483580521267e-06,
      "loss": 1.0206,
      "step": 178
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.635851462278461,
      "learning_rate": 9.744675766304538e-06,
      "loss": 0.7578,
      "step": 179
    },
    {
      "epoch": 0.39,
      "grad_norm": 5.985148978001489,
      "learning_rate": 9.740967440029628e-06,
      "loss": 0.8446,
      "step": 180
    },
    {
      "epoch": 0.39,
      "grad_norm": 6.402241115833116,
      "learning_rate": 9.737233093654572e-06,
      "loss": 0.6192,
      "step": 181
    },
    {
      "epoch": 0.39,
      "grad_norm": 7.716272599859672,
      "learning_rate": 9.733472747674779e-06,
      "loss": 0.9123,
      "step": 182
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.379360411159151,
      "learning_rate": 9.729686422728353e-06,
      "loss": 1.1469,
      "step": 183
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.919448297223582,
      "learning_rate": 9.725874139595978e-06,
      "loss": 0.827,
      "step": 184
    },
    {
      "epoch": 0.4,
      "grad_norm": 5.878278608429362,
      "learning_rate": 9.722035919200812e-06,
      "loss": 0.5674,
      "step": 185
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.224053078854734,
      "learning_rate": 9.718171782608355e-06,
      "loss": 1.1627,
      "step": 186
    },
    {
      "epoch": 0.41,
      "grad_norm": 7.381664018700441,
      "learning_rate": 9.714281751026356e-06,
      "loss": 0.8423,
      "step": 187
    },
    {
      "epoch": 0.41,
      "grad_norm": 4.05746290605081,
      "learning_rate": 9.710365845804675e-06,
      "loss": 0.598,
      "step": 188
    },
    {
      "epoch": 0.41,
      "grad_norm": 5.333888866929324,
      "learning_rate": 9.706424088435183e-06,
      "loss": 1.0943,
      "step": 189
    },
    {
      "epoch": 0.41,
      "grad_norm": 13.711486027117282,
      "learning_rate": 9.702456500551632e-06,
      "loss": 0.7994,
      "step": 190
    },
    {
      "epoch": 0.41,
      "grad_norm": 5.911482750376984,
      "learning_rate": 9.698463103929542e-06,
      "loss": 0.8593,
      "step": 191
    },
    {
      "epoch": 0.42,
      "grad_norm": 7.220783493502536,
      "learning_rate": 9.694443920486083e-06,
      "loss": 0.7845,
      "step": 192
    },
    {
      "epoch": 0.42,
      "grad_norm": 7.721522076703853,
      "learning_rate": 9.690398972279949e-06,
      "loss": 1.0479,
      "step": 193
    },
    {
      "epoch": 0.42,
      "grad_norm": 6.672958973565715,
      "learning_rate": 9.686328281511241e-06,
      "loss": 1.219,
      "step": 194
    },
    {
      "epoch": 0.42,
      "grad_norm": 6.400563736354728,
      "learning_rate": 9.682231870521347e-06,
      "loss": 0.5703,
      "step": 195
    },
    {
      "epoch": 0.42,
      "grad_norm": 6.987796834138165,
      "learning_rate": 9.67810976179281e-06,
      "loss": 1.3393,
      "step": 196
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.308941072904819,
      "learning_rate": 9.673961977949219e-06,
      "loss": 0.9578,
      "step": 197
    },
    {
      "epoch": 0.43,
      "grad_norm": 16.97780721004731,
      "learning_rate": 9.669788541755072e-06,
      "loss": 0.8457,
      "step": 198
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.161572186161637,
      "learning_rate": 9.665589476115657e-06,
      "loss": 0.6846,
      "step": 199
    },
    {
      "epoch": 0.43,
      "grad_norm": 10.807655350864229,
      "learning_rate": 9.661364804076927e-06,
      "loss": 1.0723,
      "step": 200
    },
    {
      "epoch": 0.44,
      "grad_norm": 10.949138239136994,
      "learning_rate": 9.657114548825372e-06,
      "loss": 0.6734,
      "step": 201
    },
    {
      "epoch": 0.44,
      "grad_norm": 6.565171684759978,
      "learning_rate": 9.652838733687888e-06,
      "loss": 0.7715,
      "step": 202
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.257728318632259,
      "learning_rate": 9.648537382131659e-06,
      "loss": 0.6255,
      "step": 203
    },
    {
      "epoch": 0.44,
      "grad_norm": 6.324867963444289,
      "learning_rate": 9.644210517764014e-06,
      "loss": 0.7353,
      "step": 204
    },
    {
      "epoch": 0.44,
      "grad_norm": 11.450727057968573,
      "learning_rate": 9.639858164332314e-06,
      "loss": 1.1188,
      "step": 205
    },
    {
      "epoch": 0.45,
      "grad_norm": 22.83811208499536,
      "learning_rate": 9.635480345723805e-06,
      "loss": 0.8207,
      "step": 206
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.819748906007037,
      "learning_rate": 9.631077085965501e-06,
      "loss": 0.8453,
      "step": 207
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.109245952936767,
      "learning_rate": 9.626648409224041e-06,
      "loss": 0.5089,
      "step": 208
    },
    {
      "epoch": 0.45,
      "grad_norm": 216.41666635540267,
      "learning_rate": 9.622194339805565e-06,
      "loss": 0.8961,
      "step": 209
    },
    {
      "epoch": 0.46,
      "grad_norm": 12.353788519566036,
      "learning_rate": 9.617714902155576e-06,
      "loss": 0.7606,
      "step": 210
    },
    {
      "epoch": 0.46,
      "grad_norm": 5.767736785557985,
      "learning_rate": 9.613210120858805e-06,
      "loss": 0.6783,
      "step": 211
    },
    {
      "epoch": 0.46,
      "grad_norm": 7.280948638309139,
      "learning_rate": 9.608680020639081e-06,
      "loss": 0.8223,
      "step": 212
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.756037418636119,
      "learning_rate": 9.60412462635919e-06,
      "loss": 1.4771,
      "step": 213
    },
    {
      "epoch": 0.46,
      "grad_norm": 13.068271429666547,
      "learning_rate": 9.599543963020741e-06,
      "loss": 0.8577,
      "step": 214
    },
    {
      "epoch": 0.47,
      "grad_norm": 5.2856504108258076,
      "learning_rate": 9.594938055764029e-06,
      "loss": 0.9688,
      "step": 215
    },
    {
      "epoch": 0.47,
      "grad_norm": 6.371862947753952,
      "learning_rate": 9.590306929867896e-06,
      "loss": 0.8963,
      "step": 216
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.323360204864738,
      "learning_rate": 9.585650610749593e-06,
      "loss": 1.1522,
      "step": 217
    },
    {
      "epoch": 0.47,
      "grad_norm": 7.330538613693062,
      "learning_rate": 9.580969123964641e-06,
      "loss": 0.4873,
      "step": 218
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.142650410779465,
      "learning_rate": 9.576262495206689e-06,
      "loss": 0.8826,
      "step": 219
    },
    {
      "epoch": 0.48,
      "grad_norm": 10.131855404066036,
      "learning_rate": 9.571530750307374e-06,
      "loss": 0.7748,
      "step": 220
    },
    {
      "epoch": 0.48,
      "grad_norm": 10.46154601654344,
      "learning_rate": 9.56677391523618e-06,
      "loss": 0.8647,
      "step": 221
    },
    {
      "epoch": 0.48,
      "grad_norm": 6.7654652004305245,
      "learning_rate": 9.561992016100293e-06,
      "loss": 1.1686,
      "step": 222
    },
    {
      "epoch": 0.48,
      "grad_norm": 10.367330264867718,
      "learning_rate": 9.557185079144463e-06,
      "loss": 1.1446,
      "step": 223
    },
    {
      "epoch": 0.49,
      "grad_norm": 11.282509640425063,
      "learning_rate": 9.552353130750852e-06,
      "loss": 0.8017,
      "step": 224
    },
    {
      "epoch": 0.49,
      "grad_norm": 7.585893598182119,
      "learning_rate": 9.547496197438896e-06,
      "loss": 0.5994,
      "step": 225
    },
    {
      "epoch": 0.49,
      "grad_norm": 5.676001162036721,
      "learning_rate": 9.542614305865158e-06,
      "loss": 0.7552,
      "step": 226
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.663559991153523,
      "learning_rate": 9.53770748282318e-06,
      "loss": 0.6547,
      "step": 227
    },
    {
      "epoch": 0.49,
      "grad_norm": 10.21433519308261,
      "learning_rate": 9.532775755243334e-06,
      "loss": 0.9227,
      "step": 228
    },
    {
      "epoch": 0.5,
      "grad_norm": 11.46063556062438,
      "learning_rate": 9.527819150192681e-06,
      "loss": 0.5508,
      "step": 229
    },
    {
      "epoch": 0.5,
      "grad_norm": 5.911367022039519,
      "learning_rate": 9.522837694874814e-06,
      "loss": 0.8321,
      "step": 230
    },
    {
      "epoch": 0.5,
      "grad_norm": 11.759478808471373,
      "learning_rate": 9.517831416629717e-06,
      "loss": 0.7717,
      "step": 231
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.1967113444487,
      "learning_rate": 9.512800342933608e-06,
      "loss": 0.9552,
      "step": 232
    },
    {
      "epoch": 0.5,
      "grad_norm": 5.656713464341317,
      "learning_rate": 9.507744501398794e-06,
      "loss": 0.861,
      "step": 233
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.873311825449719,
      "learning_rate": 9.502663919773516e-06,
      "loss": 1.1296,
      "step": 234
    },
    {
      "epoch": 0.51,
      "grad_norm": 16.557151620940395,
      "learning_rate": 9.497558625941794e-06,
      "loss": 0.6016,
      "step": 235
    },
    {
      "epoch": 0.51,
      "grad_norm": 7.3372500236258595,
      "learning_rate": 9.492428647923281e-06,
      "loss": 0.7782,
      "step": 236
    },
    {
      "epoch": 0.51,
      "grad_norm": 7.269916724046493,
      "learning_rate": 9.487274013873104e-06,
      "loss": 0.9705,
      "step": 237
    },
    {
      "epoch": 0.52,
      "grad_norm": 6.849475264013678,
      "learning_rate": 9.482094752081711e-06,
      "loss": 0.9708,
      "step": 238
    },
    {
      "epoch": 0.52,
      "grad_norm": 7.004333218161702,
      "learning_rate": 9.47689089097472e-06,
      "loss": 1.3229,
      "step": 239
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.48043395832493,
      "learning_rate": 9.471662459112747e-06,
      "loss": 1.0887,
      "step": 240
    },
    {
      "epoch": 0.52,
      "grad_norm": 5.117144669269138,
      "learning_rate": 9.466409485191275e-06,
      "loss": 0.99,
      "step": 241
    },
    {
      "epoch": 0.52,
      "grad_norm": 11.922111013844766,
      "learning_rate": 9.461131998040473e-06,
      "loss": 0.9013,
      "step": 242
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.718938115583912,
      "learning_rate": 9.455830026625053e-06,
      "loss": 1.0077,
      "step": 243
    },
    {
      "epoch": 0.53,
      "grad_norm": 20.618351077842586,
      "learning_rate": 9.450503600044102e-06,
      "loss": 0.6754,
      "step": 244
    },
    {
      "epoch": 0.53,
      "grad_norm": 14.188665144782009,
      "learning_rate": 9.445152747530922e-06,
      "loss": 1.2685,
      "step": 245
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.716269578623072,
      "learning_rate": 9.439777498452883e-06,
      "loss": 0.987,
      "step": 246
    },
    {
      "epoch": 0.54,
      "grad_norm": 7.805450741086592,
      "learning_rate": 9.434377882311244e-06,
      "loss": 0.8349,
      "step": 247
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.597832513557751,
      "learning_rate": 9.428953928741002e-06,
      "loss": 0.5039,
      "step": 248
    },
    {
      "epoch": 0.54,
      "grad_norm": 8.784702802078924,
      "learning_rate": 9.423505667510724e-06,
      "loss": 0.6232,
      "step": 249
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.102567473308133,
      "learning_rate": 9.41803312852239e-06,
      "loss": 1.316,
      "step": 250
    },
    {
      "epoch": 0.54,
      "grad_norm": 5.920647506602279,
      "learning_rate": 9.41253634181122e-06,
      "loss": 0.6804,
      "step": 251
    },
    {
      "epoch": 0.55,
      "grad_norm": 7.340084576953514,
      "learning_rate": 9.40701533754552e-06,
      "loss": 0.965,
      "step": 252
    },
    {
      "epoch": 0.55,
      "grad_norm": 6.468531400202817,
      "learning_rate": 9.401470146026504e-06,
      "loss": 0.603,
      "step": 253
    },
    {
      "epoch": 0.55,
      "grad_norm": 6.426632098932166,
      "learning_rate": 9.39590079768814e-06,
      "loss": 0.956,
      "step": 254
    },
    {
      "epoch": 0.55,
      "grad_norm": 5.839573791420091,
      "learning_rate": 9.390307323096972e-06,
      "loss": 1.4254,
      "step": 255
    },
    {
      "epoch": 0.55,
      "grad_norm": 15.673775976510164,
      "learning_rate": 9.384689752951961e-06,
      "loss": 0.9265,
      "step": 256
    },
    {
      "epoch": 0.56,
      "grad_norm": 7.425153888514454,
      "learning_rate": 9.379048118084312e-06,
      "loss": 0.8622,
      "step": 257
    },
    {
      "epoch": 0.56,
      "grad_norm": 6.921227874785891,
      "learning_rate": 9.373382449457305e-06,
      "loss": 1.0439,
      "step": 258
    },
    {
      "epoch": 0.56,
      "grad_norm": 7.103842056856591,
      "learning_rate": 9.367692778166126e-06,
      "loss": 0.8448,
      "step": 259
    },
    {
      "epoch": 0.56,
      "grad_norm": 10.469204599848686,
      "learning_rate": 9.361979135437697e-06,
      "loss": 0.8551,
      "step": 260
    },
    {
      "epoch": 0.57,
      "grad_norm": 4.94006877290149,
      "learning_rate": 9.356241552630503e-06,
      "loss": 0.5999,
      "step": 261
    },
    {
      "epoch": 0.57,
      "grad_norm": 15.620442029317527,
      "learning_rate": 9.350480061234419e-06,
      "loss": 0.9909,
      "step": 262
    },
    {
      "epoch": 0.57,
      "grad_norm": 13.042210867358685,
      "learning_rate": 9.344694692870541e-06,
      "loss": 0.5547,
      "step": 263
    },
    {
      "epoch": 0.57,
      "grad_norm": 4.231985521531022,
      "learning_rate": 9.338885479291012e-06,
      "loss": 0.4277,
      "step": 264
    },
    {
      "epoch": 0.57,
      "grad_norm": 10.482731626851903,
      "learning_rate": 9.333052452378838e-06,
      "loss": 0.8931,
      "step": 265
    },
    {
      "epoch": 0.58,
      "grad_norm": 6.134860457945292,
      "learning_rate": 9.32719564414773e-06,
      "loss": 0.8861,
      "step": 266
    },
    {
      "epoch": 0.58,
      "grad_norm": 7.327440131549406,
      "learning_rate": 9.321315086741916e-06,
      "loss": 1.1791,
      "step": 267
    },
    {
      "epoch": 0.58,
      "grad_norm": 10.124316625706545,
      "learning_rate": 9.315410812435967e-06,
      "loss": 1.0424,
      "step": 268
    },
    {
      "epoch": 0.58,
      "grad_norm": 8.19815523229244,
      "learning_rate": 9.30948285363462e-06,
      "loss": 1.0448,
      "step": 269
    },
    {
      "epoch": 0.59,
      "grad_norm": 8.34298574007648,
      "learning_rate": 9.303531242872606e-06,
      "loss": 1.0967,
      "step": 270
    },
    {
      "epoch": 0.59,
      "grad_norm": 413.38378981076244,
      "learning_rate": 9.297556012814457e-06,
      "loss": 0.9364,
      "step": 271
    },
    {
      "epoch": 0.59,
      "grad_norm": 6.526039286115283,
      "learning_rate": 9.291557196254342e-06,
      "loss": 0.6435,
      "step": 272
    },
    {
      "epoch": 0.59,
      "grad_norm": 8.067471408168608,
      "learning_rate": 9.285534826115884e-06,
      "loss": 1.1205,
      "step": 273
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.688601246472631,
      "learning_rate": 9.279488935451971e-06,
      "loss": 1.1275,
      "step": 274
    },
    {
      "epoch": 0.6,
      "grad_norm": 7.788649449310776,
      "learning_rate": 9.27341955744458e-06,
      "loss": 0.8794,
      "step": 275
    },
    {
      "epoch": 0.6,
      "grad_norm": 10.127497182284744,
      "learning_rate": 9.2673267254046e-06,
      "loss": 0.9408,
      "step": 276
    },
    {
      "epoch": 0.6,
      "grad_norm": 18.775566960401378,
      "learning_rate": 9.261210472771637e-06,
      "loss": 1.0974,
      "step": 277
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.819611189100439,
      "learning_rate": 9.255070833113845e-06,
      "loss": 0.967,
      "step": 278
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.68116408443567,
      "learning_rate": 9.248907840127726e-06,
      "loss": 0.884,
      "step": 279
    },
    {
      "epoch": 0.61,
      "grad_norm": 6.365075617338463,
      "learning_rate": 9.24272152763796e-06,
      "loss": 0.3923,
      "step": 280
    },
    {
      "epoch": 0.61,
      "grad_norm": 7.411318519982267,
      "learning_rate": 9.236511929597206e-06,
      "loss": 0.6685,
      "step": 281
    },
    {
      "epoch": 0.61,
      "grad_norm": 7.452944974973433,
      "learning_rate": 9.230279080085933e-06,
      "loss": 0.9466,
      "step": 282
    },
    {
      "epoch": 0.61,
      "grad_norm": 10.286773276618867,
      "learning_rate": 9.224023013312212e-06,
      "loss": 1.1582,
      "step": 283
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.189669006492984,
      "learning_rate": 9.217743763611545e-06,
      "loss": 0.5304,
      "step": 284
    },
    {
      "epoch": 0.62,
      "grad_norm": 6.320587159563499,
      "learning_rate": 9.211441365446661e-06,
      "loss": 0.7365,
      "step": 285
    },
    {
      "epoch": 0.62,
      "grad_norm": 6.453292211028784,
      "learning_rate": 9.20511585340735e-06,
      "loss": 0.7461,
      "step": 286
    },
    {
      "epoch": 0.62,
      "grad_norm": 6.900310585348448,
      "learning_rate": 9.198767262210244e-06,
      "loss": 0.5479,
      "step": 287
    },
    {
      "epoch": 0.62,
      "grad_norm": 15.275980881581928,
      "learning_rate": 9.192395626698656e-06,
      "loss": 0.8681,
      "step": 288
    },
    {
      "epoch": 0.63,
      "grad_norm": 11.385500406400531,
      "learning_rate": 9.186000981842362e-06,
      "loss": 1.3033,
      "step": 289
    },
    {
      "epoch": 0.63,
      "grad_norm": 6.528106618508999,
      "learning_rate": 9.17958336273743e-06,
      "loss": 0.6513,
      "step": 290
    },
    {
      "epoch": 0.63,
      "grad_norm": 7.634484251525752,
      "learning_rate": 9.173142804606012e-06,
      "loss": 1.1624,
      "step": 291
    },
    {
      "epoch": 0.63,
      "grad_norm": 11.443618769301825,
      "learning_rate": 9.166679342796162e-06,
      "loss": 0.8847,
      "step": 292
    },
    {
      "epoch": 0.63,
      "grad_norm": 7.014853469005433,
      "learning_rate": 9.160193012781639e-06,
      "loss": 1.0795,
      "step": 293
    },
    {
      "epoch": 0.64,
      "grad_norm": 11.299733223069984,
      "learning_rate": 9.153683850161706e-06,
      "loss": 0.9537,
      "step": 294
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.546968230727561,
      "learning_rate": 9.147151890660942e-06,
      "loss": 0.7291,
      "step": 295
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.299221743536055,
      "learning_rate": 9.140597170129041e-06,
      "loss": 0.6803,
      "step": 296
    },
    {
      "epoch": 0.64,
      "grad_norm": 7.223935419120376,
      "learning_rate": 9.13401972454062e-06,
      "loss": 0.6249,
      "step": 297
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.383382221550068,
      "learning_rate": 9.12741958999502e-06,
      "loss": 0.6575,
      "step": 298
    },
    {
      "epoch": 0.65,
      "grad_norm": 11.472042897911813,
      "learning_rate": 9.120796802716104e-06,
      "loss": 0.7585,
      "step": 299
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.124718630371285,
      "learning_rate": 9.114151399052064e-06,
      "loss": 0.8384,
      "step": 300
    },
    {
      "epoch": 0.65,
      "grad_norm": 12.10589567665355,
      "learning_rate": 9.107483415475216e-06,
      "loss": 0.9969,
      "step": 301
    },
    {
      "epoch": 0.65,
      "grad_norm": 6.448167113986866,
      "learning_rate": 9.100792888581803e-06,
      "loss": 0.5601,
      "step": 302
    },
    {
      "epoch": 0.66,
      "grad_norm": 6.606265730034924,
      "learning_rate": 9.094079855091797e-06,
      "loss": 0.7746,
      "step": 303
    },
    {
      "epoch": 0.66,
      "grad_norm": 7.368010537358716,
      "learning_rate": 9.08734435184869e-06,
      "loss": 0.812,
      "step": 304
    },
    {
      "epoch": 0.66,
      "grad_norm": 22.03782984859312,
      "learning_rate": 9.080586415819296e-06,
      "loss": 1.3603,
      "step": 305
    },
    {
      "epoch": 0.66,
      "grad_norm": 11.127738700395815,
      "learning_rate": 9.073806084093556e-06,
      "loss": 1.2948,
      "step": 306
    },
    {
      "epoch": 0.67,
      "grad_norm": 8.096018452610016,
      "learning_rate": 9.067003393884313e-06,
      "loss": 0.8198,
      "step": 307
    },
    {
      "epoch": 0.67,
      "grad_norm": 8.13859614149178,
      "learning_rate": 9.06017838252713e-06,
      "loss": 0.6417,
      "step": 308
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.590306255914138,
      "learning_rate": 9.053331087480075e-06,
      "loss": 0.9276,
      "step": 309
    },
    {
      "epoch": 0.67,
      "grad_norm": 6.582364163209959,
      "learning_rate": 9.046461546323519e-06,
      "loss": 0.7733,
      "step": 310
    },
    {
      "epoch": 0.67,
      "grad_norm": 10.426941117046487,
      "learning_rate": 9.039569796759921e-06,
      "loss": 1.0951,
      "step": 311
    },
    {
      "epoch": 0.68,
      "grad_norm": 7.605901200835638,
      "learning_rate": 9.032655876613636e-06,
      "loss": 1.0577,
      "step": 312
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.867164233768532,
      "learning_rate": 9.02571982383069e-06,
      "loss": 1.1325,
      "step": 313
    },
    {
      "epoch": 0.68,
      "grad_norm": 11.830644660009698,
      "learning_rate": 9.018761676478585e-06,
      "loss": 0.9014,
      "step": 314
    },
    {
      "epoch": 0.68,
      "grad_norm": 7.810113685881364,
      "learning_rate": 9.01178147274609e-06,
      "loss": 0.5936,
      "step": 315
    },
    {
      "epoch": 0.68,
      "grad_norm": 16.179621903138063,
      "learning_rate": 9.00477925094302e-06,
      "loss": 0.7019,
      "step": 316
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.920766034543163,
      "learning_rate": 8.997755049500037e-06,
      "loss": 0.943,
      "step": 317
    },
    {
      "epoch": 0.69,
      "grad_norm": 14.114366213107283,
      "learning_rate": 8.990708906968431e-06,
      "loss": 0.8255,
      "step": 318
    },
    {
      "epoch": 0.69,
      "grad_norm": 8.120763677587066,
      "learning_rate": 8.98364086201992e-06,
      "loss": 0.8304,
      "step": 319
    },
    {
      "epoch": 0.69,
      "grad_norm": 6.08901272436489,
      "learning_rate": 8.976550953446426e-06,
      "loss": 0.9432,
      "step": 320
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.308842219969819,
      "learning_rate": 8.969439220159861e-06,
      "loss": 0.8121,
      "step": 321
    },
    {
      "epoch": 0.7,
      "grad_norm": 17.878200034172043,
      "learning_rate": 8.962305701191927e-06,
      "loss": 0.7953,
      "step": 322
    },
    {
      "epoch": 0.7,
      "grad_norm": 21.10206343331327,
      "learning_rate": 8.955150435693889e-06,
      "loss": 0.9148,
      "step": 323
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.524680666683667,
      "learning_rate": 8.947973462936366e-06,
      "loss": 0.8113,
      "step": 324
    },
    {
      "epoch": 0.7,
      "grad_norm": 76.70159355041507,
      "learning_rate": 8.940774822309116e-06,
      "loss": 0.9235,
      "step": 325
    },
    {
      "epoch": 0.71,
      "grad_norm": 7.213821466757655,
      "learning_rate": 8.933554553320813e-06,
      "loss": 0.7369,
      "step": 326
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.123907294112344,
      "learning_rate": 8.926312695598837e-06,
      "loss": 0.7895,
      "step": 327
    },
    {
      "epoch": 0.71,
      "grad_norm": 13.04638195649625,
      "learning_rate": 8.919049288889058e-06,
      "loss": 1.0482,
      "step": 328
    },
    {
      "epoch": 0.71,
      "grad_norm": 7.808002976953559,
      "learning_rate": 8.911764373055612e-06,
      "loss": 1.1717,
      "step": 329
    },
    {
      "epoch": 0.72,
      "grad_norm": 15.09363941341881,
      "learning_rate": 8.904457988080682e-06,
      "loss": 0.9796,
      "step": 330
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.911803545696616,
      "learning_rate": 8.897130174064285e-06,
      "loss": 0.4774,
      "step": 331
    },
    {
      "epoch": 0.72,
      "grad_norm": 12.499733990592587,
      "learning_rate": 8.889780971224047e-06,
      "loss": 0.8269,
      "step": 332
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.342667348077121,
      "learning_rate": 8.882410419894983e-06,
      "loss": 0.8154,
      "step": 333
    },
    {
      "epoch": 0.72,
      "grad_norm": 11.108624894210307,
      "learning_rate": 8.875018560529275e-06,
      "loss": 0.8289,
      "step": 334
    },
    {
      "epoch": 0.73,
      "grad_norm": 11.106091281600394,
      "learning_rate": 8.867605433696056e-06,
      "loss": 0.4796,
      "step": 335
    },
    {
      "epoch": 0.73,
      "grad_norm": 8.06640474401585,
      "learning_rate": 8.860171080081174e-06,
      "loss": 0.6012,
      "step": 336
    },
    {
      "epoch": 0.73,
      "grad_norm": 16.163041163425866,
      "learning_rate": 8.852715540486986e-06,
      "loss": 1.2754,
      "step": 337
    },
    {
      "epoch": 0.73,
      "grad_norm": 6.743624007382333,
      "learning_rate": 8.845238855832117e-06,
      "loss": 0.9331,
      "step": 338
    },
    {
      "epoch": 0.73,
      "grad_norm": 8.20473410893407,
      "learning_rate": 8.837741067151251e-06,
      "loss": 1.0095,
      "step": 339
    },
    {
      "epoch": 0.74,
      "grad_norm": 7.188332115359249,
      "learning_rate": 8.83022221559489e-06,
      "loss": 0.7177,
      "step": 340
    },
    {
      "epoch": 0.74,
      "grad_norm": 7.205581298250863,
      "learning_rate": 8.822682342429147e-06,
      "loss": 0.4657,
      "step": 341
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.126625989036473,
      "learning_rate": 8.8151214890355e-06,
      "loss": 0.6293,
      "step": 342
    },
    {
      "epoch": 0.74,
      "grad_norm": 6.55722860022889,
      "learning_rate": 8.807539696910574e-06,
      "loss": 0.8475,
      "step": 343
    },
    {
      "epoch": 0.75,
      "grad_norm": 7.419685394443201,
      "learning_rate": 8.79993700766592e-06,
      "loss": 0.9321,
      "step": 344
    },
    {
      "epoch": 0.75,
      "grad_norm": 6.0405535247974,
      "learning_rate": 8.792313463027777e-06,
      "loss": 0.9064,
      "step": 345
    },
    {
      "epoch": 0.75,
      "grad_norm": 6.291524102033553,
      "learning_rate": 8.784669104836842e-06,
      "loss": 0.8932,
      "step": 346
    },
    {
      "epoch": 0.75,
      "grad_norm": 23.969284831667945,
      "learning_rate": 8.777003975048048e-06,
      "loss": 0.8017,
      "step": 347
    },
    {
      "epoch": 0.75,
      "grad_norm": 6.988104236794416,
      "learning_rate": 8.76931811573033e-06,
      "loss": 0.6091,
      "step": 348
    },
    {
      "epoch": 0.76,
      "grad_norm": 6.315506929807876,
      "learning_rate": 8.761611569066388e-06,
      "loss": 0.7777,
      "step": 349
    },
    {
      "epoch": 0.76,
      "grad_norm": 12.56290699774491,
      "learning_rate": 8.753884377352472e-06,
      "loss": 0.5312,
      "step": 350
    },
    {
      "epoch": 0.76,
      "grad_norm": 8.842349481163227,
      "learning_rate": 8.74613658299813e-06,
      "loss": 1.0589,
      "step": 351
    },
    {
      "epoch": 0.76,
      "grad_norm": 16.052496292077887,
      "learning_rate": 8.738368228525988e-06,
      "loss": 0.5814,
      "step": 352
    },
    {
      "epoch": 0.76,
      "grad_norm": 13.05604938880339,
      "learning_rate": 8.730579356571514e-06,
      "loss": 0.7891,
      "step": 353
    },
    {
      "epoch": 0.77,
      "grad_norm": 7.96275463605765,
      "learning_rate": 8.72277000988278e-06,
      "loss": 0.6414,
      "step": 354
    },
    {
      "epoch": 0.77,
      "grad_norm": 6.754877917391566,
      "learning_rate": 8.714940231320237e-06,
      "loss": 0.8882,
      "step": 355
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.975706573231873,
      "learning_rate": 8.707090063856466e-06,
      "loss": 0.7497,
      "step": 356
    },
    {
      "epoch": 0.77,
      "grad_norm": 12.78951987846906,
      "learning_rate": 8.699219550575954e-06,
      "loss": 0.7604,
      "step": 357
    },
    {
      "epoch": 0.78,
      "grad_norm": 5.688345595628288,
      "learning_rate": 8.691328734674851e-06,
      "loss": 1.0541,
      "step": 358
    },
    {
      "epoch": 0.78,
      "grad_norm": 5.1367711226593,
      "learning_rate": 8.683417659460735e-06,
      "loss": 0.4163,
      "step": 359
    },
    {
      "epoch": 0.78,
      "grad_norm": 13.284019140486247,
      "learning_rate": 8.675486368352376e-06,
      "loss": 0.6912,
      "step": 360
    },
    {
      "epoch": 0.78,
      "grad_norm": 14.944446951037683,
      "learning_rate": 8.667534904879495e-06,
      "loss": 0.9886,
      "step": 361
    },
    {
      "epoch": 0.78,
      "grad_norm": 7.706659997012531,
      "learning_rate": 8.659563312682524e-06,
      "loss": 0.8324,
      "step": 362
    },
    {
      "epoch": 0.79,
      "grad_norm": 8.575454387871599,
      "learning_rate": 8.651571635512372e-06,
      "loss": 1.3185,
      "step": 363
    },
    {
      "epoch": 0.79,
      "grad_norm": 7.78644278384784,
      "learning_rate": 8.64355991723018e-06,
      "loss": 0.7319,
      "step": 364
    },
    {
      "epoch": 0.79,
      "grad_norm": 8.336459375323221,
      "learning_rate": 8.635528201807079e-06,
      "loss": 0.7317,
      "step": 365
    },
    {
      "epoch": 0.79,
      "grad_norm": 7.779957664972544,
      "learning_rate": 8.627476533323957e-06,
      "loss": 0.8981,
      "step": 366
    },
    {
      "epoch": 0.8,
      "grad_norm": 6.2789072313027505,
      "learning_rate": 8.619404955971208e-06,
      "loss": 0.7661,
      "step": 367
    },
    {
      "epoch": 0.8,
      "grad_norm": 11.100271108867858,
      "learning_rate": 8.61131351404849e-06,
      "loss": 0.6775,
      "step": 368
    },
    {
      "epoch": 0.8,
      "grad_norm": 14.389819996594321,
      "learning_rate": 8.603202251964492e-06,
      "loss": 0.5948,
      "step": 369
    },
    {
      "epoch": 0.8,
      "grad_norm": 6.032078382746744,
      "learning_rate": 8.595071214236675e-06,
      "loss": 0.9381,
      "step": 370
    },
    {
      "epoch": 0.8,
      "grad_norm": 6.8853518453201925,
      "learning_rate": 8.586920445491043e-06,
      "loss": 1.0514,
      "step": 371
    },
    {
      "epoch": 0.81,
      "grad_norm": 5.9323272695627,
      "learning_rate": 8.578749990461884e-06,
      "loss": 0.6577,
      "step": 372
    },
    {
      "epoch": 0.81,
      "grad_norm": 5.862051940987071,
      "learning_rate": 8.570559893991537e-06,
      "loss": 0.9836,
      "step": 373
    },
    {
      "epoch": 0.81,
      "grad_norm": 7.954312208610544,
      "learning_rate": 8.562350201030139e-06,
      "loss": 0.8495,
      "step": 374
    },
    {
      "epoch": 0.81,
      "grad_norm": 5.995698712841757,
      "learning_rate": 8.554120956635375e-06,
      "loss": 1.032,
      "step": 375
    },
    {
      "epoch": 0.81,
      "grad_norm": 12.075228175931256,
      "learning_rate": 8.54587220597224e-06,
      "loss": 1.3547,
      "step": 376
    },
    {
      "epoch": 0.82,
      "grad_norm": 6.19245486688663,
      "learning_rate": 8.537603994312786e-06,
      "loss": 0.7102,
      "step": 377
    },
    {
      "epoch": 0.82,
      "grad_norm": 8.525497239832244,
      "learning_rate": 8.52931636703587e-06,
      "loss": 0.8742,
      "step": 378
    },
    {
      "epoch": 0.82,
      "grad_norm": 9.394821060377648,
      "learning_rate": 8.521009369626914e-06,
      "loss": 1.3331,
      "step": 379
    },
    {
      "epoch": 0.82,
      "grad_norm": 6.669952244150261,
      "learning_rate": 8.512683047677644e-06,
      "loss": 0.4839,
      "step": 380
    },
    {
      "epoch": 0.83,
      "grad_norm": 5.112511421085029,
      "learning_rate": 8.504337446885854e-06,
      "loss": 0.9033,
      "step": 381
    },
    {
      "epoch": 0.83,
      "grad_norm": 6.617206159661717,
      "learning_rate": 8.495972613055137e-06,
      "loss": 0.6206,
      "step": 382
    },
    {
      "epoch": 0.83,
      "grad_norm": 9.835064668153016,
      "learning_rate": 8.487588592094652e-06,
      "loss": 0.9622,
      "step": 383
    },
    {
      "epoch": 0.83,
      "grad_norm": 8.78077016064967,
      "learning_rate": 8.47918543001886e-06,
      "loss": 0.7874,
      "step": 384
    },
    {
      "epoch": 0.83,
      "grad_norm": 4.772988044852947,
      "learning_rate": 8.470763172947276e-06,
      "loss": 0.8958,
      "step": 385
    },
    {
      "epoch": 0.84,
      "grad_norm": 18.75541773349608,
      "learning_rate": 8.462321867104217e-06,
      "loss": 1.0347,
      "step": 386
    },
    {
      "epoch": 0.84,
      "grad_norm": 7.249716140887136,
      "learning_rate": 8.453861558818542e-06,
      "loss": 0.6069,
      "step": 387
    },
    {
      "epoch": 0.84,
      "grad_norm": 6.008406999352871,
      "learning_rate": 8.445382294523406e-06,
      "loss": 0.8651,
      "step": 388
    },
    {
      "epoch": 0.84,
      "grad_norm": 8.00693319313131,
      "learning_rate": 8.436884120755997e-06,
      "loss": 0.9206,
      "step": 389
    },
    {
      "epoch": 0.85,
      "grad_norm": 16.79743099139263,
      "learning_rate": 8.428367084157292e-06,
      "loss": 1.2097,
      "step": 390
    },
    {
      "epoch": 0.85,
      "grad_norm": 7.399956557005277,
      "learning_rate": 8.419831231471785e-06,
      "loss": 0.9206,
      "step": 391
    },
    {
      "epoch": 0.85,
      "grad_norm": 24.09824617567442,
      "learning_rate": 8.411276609547246e-06,
      "loss": 0.6516,
      "step": 392
    },
    {
      "epoch": 0.85,
      "grad_norm": 11.563367705378283,
      "learning_rate": 8.402703265334455e-06,
      "loss": 1.1148,
      "step": 393
    },
    {
      "epoch": 0.85,
      "grad_norm": 5.714990277953213,
      "learning_rate": 8.394111245886948e-06,
      "loss": 1.0459,
      "step": 394
    },
    {
      "epoch": 0.86,
      "grad_norm": 16.75851353717472,
      "learning_rate": 8.385500598360752e-06,
      "loss": 0.9293,
      "step": 395
    },
    {
      "epoch": 0.86,
      "grad_norm": 10.829310711134095,
      "learning_rate": 8.376871370014139e-06,
      "loss": 0.9969,
      "step": 396
    },
    {
      "epoch": 0.86,
      "grad_norm": 5.943808728967222,
      "learning_rate": 8.368223608207351e-06,
      "loss": 0.8583,
      "step": 397
    },
    {
      "epoch": 0.86,
      "grad_norm": 7.599733481311216,
      "learning_rate": 8.359557360402357e-06,
      "loss": 0.8544,
      "step": 398
    },
    {
      "epoch": 0.86,
      "grad_norm": 44.08862196639092,
      "learning_rate": 8.350872674162578e-06,
      "loss": 0.8271,
      "step": 399
    },
    {
      "epoch": 0.87,
      "grad_norm": 7.229328310770392,
      "learning_rate": 8.34216959715263e-06,
      "loss": 0.9676,
      "step": 400
    },
    {
      "epoch": 0.87,
      "grad_norm": 6.9873344736457685,
      "learning_rate": 8.333448177138071e-06,
      "loss": 1.2777,
      "step": 401
    },
    {
      "epoch": 0.87,
      "grad_norm": 7.435846012381489,
      "learning_rate": 8.324708461985124e-06,
      "loss": 0.8482,
      "step": 402
    },
    {
      "epoch": 0.87,
      "grad_norm": 5.223728740321737,
      "learning_rate": 8.315950499660427e-06,
      "loss": 0.5472,
      "step": 403
    },
    {
      "epoch": 0.88,
      "grad_norm": 6.914373371781798,
      "learning_rate": 8.307174338230765e-06,
      "loss": 0.7123,
      "step": 404
    },
    {
      "epoch": 0.88,
      "grad_norm": 10.972346600485958,
      "learning_rate": 8.298380025862805e-06,
      "loss": 1.0286,
      "step": 405
    },
    {
      "epoch": 0.88,
      "grad_norm": 7.280269101738496,
      "learning_rate": 8.28956761082283e-06,
      "loss": 0.8731,
      "step": 406
    },
    {
      "epoch": 0.88,
      "grad_norm": 6.709624034213731,
      "learning_rate": 8.280737141476482e-06,
      "loss": 0.788,
      "step": 407
    },
    {
      "epoch": 0.88,
      "grad_norm": 5.871202043420699,
      "learning_rate": 8.271888666288488e-06,
      "loss": 0.9205,
      "step": 408
    },
    {
      "epoch": 0.89,
      "grad_norm": 8.420859396418175,
      "learning_rate": 8.263022233822397e-06,
      "loss": 0.887,
      "step": 409
    },
    {
      "epoch": 0.89,
      "grad_norm": 5.727170902405111,
      "learning_rate": 8.254137892740318e-06,
      "loss": 0.7956,
      "step": 410
    },
    {
      "epoch": 0.89,
      "grad_norm": 6.996836658105899,
      "learning_rate": 8.245235691802644e-06,
      "loss": 0.6177,
      "step": 411
    },
    {
      "epoch": 0.89,
      "grad_norm": 7.944032306080663,
      "learning_rate": 8.23631567986779e-06,
      "loss": 1.0121,
      "step": 412
    },
    {
      "epoch": 0.89,
      "grad_norm": 8.001921721080434,
      "learning_rate": 8.227377905891927e-06,
      "loss": 0.8717,
      "step": 413
    },
    {
      "epoch": 0.9,
      "grad_norm": 5.8047443729770425,
      "learning_rate": 8.218422418928709e-06,
      "loss": 0.6057,
      "step": 414
    },
    {
      "epoch": 0.9,
      "grad_norm": 6.909522698352482,
      "learning_rate": 8.209449268129003e-06,
      "loss": 0.6102,
      "step": 415
    },
    {
      "epoch": 0.9,
      "grad_norm": 7.627924032642646,
      "learning_rate": 8.200458502740623e-06,
      "loss": 0.7516,
      "step": 416
    },
    {
      "epoch": 0.9,
      "grad_norm": 7.234104159211019,
      "learning_rate": 8.191450172108058e-06,
      "loss": 1.0543,
      "step": 417
    },
    {
      "epoch": 0.91,
      "grad_norm": 5.74890216207692,
      "learning_rate": 8.182424325672203e-06,
      "loss": 0.7833,
      "step": 418
    },
    {
      "epoch": 0.91,
      "grad_norm": 5.6667347726502495,
      "learning_rate": 8.173381012970084e-06,
      "loss": 0.8892,
      "step": 419
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.465391227446869,
      "learning_rate": 8.164320283634585e-06,
      "loss": 1.0338,
      "step": 420
    },
    {
      "epoch": 0.91,
      "grad_norm": 7.519857385413438,
      "learning_rate": 8.155242187394184e-06,
      "loss": 0.693,
      "step": 421
    },
    {
      "epoch": 0.91,
      "grad_norm": 6.5631576153404145,
      "learning_rate": 8.146146774072674e-06,
      "loss": 0.7426,
      "step": 422
    },
    {
      "epoch": 0.92,
      "grad_norm": 5.761851605401293,
      "learning_rate": 8.137034093588885e-06,
      "loss": 1.3586,
      "step": 423
    },
    {
      "epoch": 0.92,
      "grad_norm": 7.522057539557152,
      "learning_rate": 8.127904195956424e-06,
      "loss": 0.7734,
      "step": 424
    },
    {
      "epoch": 0.92,
      "grad_norm": 7.980723302846689,
      "learning_rate": 8.118757131283383e-06,
      "loss": 0.7056,
      "step": 425
    },
    {
      "epoch": 0.92,
      "grad_norm": 5.719299592554976,
      "learning_rate": 8.109592949772076e-06,
      "loss": 0.8606,
      "step": 426
    },
    {
      "epoch": 0.93,
      "grad_norm": 3.8514900814203483,
      "learning_rate": 8.100411701718765e-06,
      "loss": 0.812,
      "step": 427
    },
    {
      "epoch": 0.93,
      "grad_norm": 6.844609481945989,
      "learning_rate": 8.091213437513371e-06,
      "loss": 0.565,
      "step": 428
    },
    {
      "epoch": 0.93,
      "grad_norm": 10.690868382905714,
      "learning_rate": 8.081998207639212e-06,
      "loss": 1.0165,
      "step": 429
    },
    {
      "epoch": 0.93,
      "grad_norm": 6.625562355238185,
      "learning_rate": 8.072766062672717e-06,
      "loss": 0.7614,
      "step": 430
    },
    {
      "epoch": 0.93,
      "grad_norm": 4.482398510932476,
      "learning_rate": 8.06351705328315e-06,
      "loss": 0.6091,
      "step": 431
    },
    {
      "epoch": 0.94,
      "grad_norm": 8.990874693953495,
      "learning_rate": 8.054251230232333e-06,
      "loss": 0.8274,
      "step": 432
    },
    {
      "epoch": 0.94,
      "grad_norm": 6.812605911529919,
      "learning_rate": 8.044968644374373e-06,
      "loss": 0.5275,
      "step": 433
    },
    {
      "epoch": 0.94,
      "grad_norm": 4.395743645944749,
      "learning_rate": 8.035669346655368e-06,
      "loss": 0.4632,
      "step": 434
    },
    {
      "epoch": 0.94,
      "grad_norm": 4.951494414948849,
      "learning_rate": 8.026353388113142e-06,
      "loss": 0.6928,
      "step": 435
    },
    {
      "epoch": 0.94,
      "grad_norm": 4.991251125151261,
      "learning_rate": 8.017020819876962e-06,
      "loss": 0.9211,
      "step": 436
    },
    {
      "epoch": 0.95,
      "grad_norm": 5.245943404860766,
      "learning_rate": 8.007671693167248e-06,
      "loss": 0.5762,
      "step": 437
    },
    {
      "epoch": 0.95,
      "grad_norm": 5.902748704231252,
      "learning_rate": 7.998306059295302e-06,
      "loss": 1.0082,
      "step": 438
    },
    {
      "epoch": 0.95,
      "grad_norm": 5.20681935611651,
      "learning_rate": 7.988923969663027e-06,
      "loss": 0.7212,
      "step": 439
    },
    {
      "epoch": 0.95,
      "grad_norm": 6.708135448207899,
      "learning_rate": 7.979525475762634e-06,
      "loss": 0.8072,
      "step": 440
    },
    {
      "epoch": 0.96,
      "grad_norm": 16.08304977604758,
      "learning_rate": 7.97011062917637e-06,
      "loss": 1.0204,
      "step": 441
    },
    {
      "epoch": 0.96,
      "grad_norm": 5.00424730893097,
      "learning_rate": 7.960679481576233e-06,
      "loss": 0.3967,
      "step": 442
    },
    {
      "epoch": 0.96,
      "grad_norm": 4.302678134652087,
      "learning_rate": 7.951232084723685e-06,
      "loss": 0.8321,
      "step": 443
    },
    {
      "epoch": 0.96,
      "grad_norm": 8.21007014790003,
      "learning_rate": 7.941768490469368e-06,
      "loss": 0.7974,
      "step": 444
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.022463974705133,
      "learning_rate": 7.932288750752819e-06,
      "loss": 0.9476,
      "step": 445
    },
    {
      "epoch": 0.97,
      "grad_norm": 6.840491476409611,
      "learning_rate": 7.922792917602197e-06,
      "loss": 0.6898,
      "step": 446
    },
    {
      "epoch": 0.97,
      "grad_norm": 5.595399845237821,
      "learning_rate": 7.913281043133978e-06,
      "loss": 0.8424,
      "step": 447
    },
    {
      "epoch": 0.97,
      "grad_norm": 4.852853791832801,
      "learning_rate": 7.903753179552682e-06,
      "loss": 0.8927,
      "step": 448
    },
    {
      "epoch": 0.97,
      "grad_norm": 5.641108040840753,
      "learning_rate": 7.89420937915058e-06,
      "loss": 0.6338,
      "step": 449
    },
    {
      "epoch": 0.98,
      "grad_norm": 7.50426166681464,
      "learning_rate": 7.884649694307413e-06,
      "loss": 0.9013,
      "step": 450
    },
    {
      "epoch": 0.98,
      "grad_norm": 4.761508690988941,
      "learning_rate": 7.875074177490103e-06,
      "loss": 1.1413,
      "step": 451
    },
    {
      "epoch": 0.98,
      "grad_norm": 6.359928059115919,
      "learning_rate": 7.86548288125246e-06,
      "loss": 0.9018,
      "step": 452
    },
    {
      "epoch": 0.98,
      "grad_norm": 6.283695371962136,
      "learning_rate": 7.855875858234894e-06,
      "loss": 0.8888,
      "step": 453
    },
    {
      "epoch": 0.98,
      "grad_norm": 11.179687942698198,
      "learning_rate": 7.846253161164138e-06,
      "loss": 0.7023,
      "step": 454
    },
    {
      "epoch": 0.99,
      "grad_norm": 5.392544646151634,
      "learning_rate": 7.836614842852942e-06,
      "loss": 0.8921,
      "step": 455
    },
    {
      "epoch": 0.99,
      "grad_norm": 5.79092613241167,
      "learning_rate": 7.826960956199796e-06,
      "loss": 0.4287,
      "step": 456
    },
    {
      "epoch": 0.99,
      "grad_norm": 14.228006552395009,
      "learning_rate": 7.817291554188628e-06,
      "loss": 1.0252,
      "step": 457
    },
    {
      "epoch": 0.99,
      "grad_norm": 8.132685804097838,
      "learning_rate": 7.80760668988853e-06,
      "loss": 0.6667,
      "step": 458
    },
    {
      "epoch": 0.99,
      "grad_norm": 3.605469023915293,
      "learning_rate": 7.797906416453445e-06,
      "loss": 0.5812,
      "step": 459
    },
    {
      "epoch": 1.0,
      "grad_norm": 6.611834868021434,
      "learning_rate": 7.788190787121896e-06,
      "loss": 1.0812,
      "step": 460
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.748807627498234,
      "learning_rate": 7.778459855216678e-06,
      "loss": 0.6416,
      "step": 461
    },
    {
      "epoch": 1.0,
      "grad_norm": 8.71663158186708,
      "learning_rate": 7.768713674144578e-06,
      "loss": 1.1982,
      "step": 462
    },
    {
      "epoch": 1.0,
      "grad_norm": 5.082596557974946,
      "learning_rate": 7.758952297396068e-06,
      "loss": 0.7038,
      "step": 463
    },
    {
      "epoch": 1.01,
      "grad_norm": 5.403842287463601,
      "learning_rate": 7.749175778545026e-06,
      "loss": 0.4491,
      "step": 464
    },
    {
      "epoch": 1.01,
      "grad_norm": 7.934590453833129,
      "learning_rate": 7.739384171248436e-06,
      "loss": 0.993,
      "step": 465
    },
    {
      "epoch": 1.01,
      "grad_norm": 8.490913153253894,
      "learning_rate": 7.729577529246084e-06,
      "loss": 0.8978,
      "step": 466
    },
    {
      "epoch": 1.01,
      "grad_norm": 3.884865911481896,
      "learning_rate": 7.719755906360282e-06,
      "loss": 0.6971,
      "step": 467
    },
    {
      "epoch": 1.01,
      "grad_norm": 4.622054258747089,
      "learning_rate": 7.709919356495555e-06,
      "loss": 0.7152,
      "step": 468
    },
    {
      "epoch": 1.02,
      "grad_norm": 5.510801486476476,
      "learning_rate": 7.700067933638357e-06,
      "loss": 1.0191,
      "step": 469
    },
    {
      "epoch": 1.02,
      "grad_norm": 5.94393183217501,
      "learning_rate": 7.690201691856768e-06,
      "loss": 0.739,
      "step": 470
    },
    {
      "epoch": 1.02,
      "grad_norm": 4.873265886826739,
      "learning_rate": 7.6803206853002e-06,
      "loss": 0.411,
      "step": 471
    },
    {
      "epoch": 1.02,
      "grad_norm": 4.65437550393593,
      "learning_rate": 7.670424968199099e-06,
      "loss": 0.754,
      "step": 472
    },
    {
      "epoch": 1.02,
      "grad_norm": 4.215632778543796,
      "learning_rate": 7.660514594864648e-06,
      "loss": 0.8239,
      "step": 473
    },
    {
      "epoch": 1.03,
      "grad_norm": 3.780992039892973,
      "learning_rate": 7.650589619688468e-06,
      "loss": 0.5503,
      "step": 474
    },
    {
      "epoch": 1.03,
      "grad_norm": 5.6058544068026315,
      "learning_rate": 7.640650097142322e-06,
      "loss": 0.8284,
      "step": 475
    },
    {
      "epoch": 1.03,
      "grad_norm": 5.781756060053529,
      "learning_rate": 7.630696081777813e-06,
      "loss": 1.4773,
      "step": 476
    },
    {
      "epoch": 1.03,
      "grad_norm": 4.660894527377529,
      "learning_rate": 7.620727628226081e-06,
      "loss": 0.7705,
      "step": 477
    },
    {
      "epoch": 1.04,
      "grad_norm": 4.920659077506768,
      "learning_rate": 7.610744791197518e-06,
      "loss": 1.0657,
      "step": 478
    },
    {
      "epoch": 1.04,
      "grad_norm": 4.480697880095817,
      "learning_rate": 7.6007476254814495e-06,
      "loss": 0.7933,
      "step": 479
    },
    {
      "epoch": 1.04,
      "grad_norm": 15.91453917863499,
      "learning_rate": 7.590736185945843e-06,
      "loss": 0.8726,
      "step": 480
    },
    {
      "epoch": 1.04,
      "grad_norm": 4.968656543049703,
      "learning_rate": 7.580710527537008e-06,
      "loss": 0.6285,
      "step": 481
    },
    {
      "epoch": 1.04,
      "grad_norm": 4.028478484782784,
      "learning_rate": 7.570670705279291e-06,
      "loss": 0.6504,
      "step": 482
    },
    {
      "epoch": 1.05,
      "grad_norm": 4.824118211800543,
      "learning_rate": 7.560616774274775e-06,
      "loss": 0.6768,
      "step": 483
    },
    {
      "epoch": 1.05,
      "grad_norm": 4.180129952629621,
      "learning_rate": 7.550548789702979e-06,
      "loss": 0.8852,
      "step": 484
    },
    {
      "epoch": 1.05,
      "grad_norm": 5.0351350082494495,
      "learning_rate": 7.540466806820545e-06,
      "loss": 1.1165,
      "step": 485
    },
    {
      "epoch": 1.05,
      "grad_norm": 3.4368496578527776,
      "learning_rate": 7.5303708809609514e-06,
      "loss": 0.4866,
      "step": 486
    },
    {
      "epoch": 1.06,
      "grad_norm": 6.516962778169572,
      "learning_rate": 7.520261067534198e-06,
      "loss": 1.3718,
      "step": 487
    },
    {
      "epoch": 1.06,
      "grad_norm": 11.624220983367362,
      "learning_rate": 7.510137422026502e-06,
      "loss": 0.8864,
      "step": 488
    },
    {
      "epoch": 1.06,
      "grad_norm": 5.092522208844558,
      "learning_rate": 7.500000000000001e-06,
      "loss": 1.2468,
      "step": 489
    },
    {
      "epoch": 1.06,
      "grad_norm": 6.383241927395272,
      "learning_rate": 7.489848857092436e-06,
      "loss": 0.5698,
      "step": 490
    },
    {
      "epoch": 1.06,
      "grad_norm": 12.095506327024182,
      "learning_rate": 7.479684049016859e-06,
      "loss": 0.7068,
      "step": 491
    },
    {
      "epoch": 1.07,
      "grad_norm": 3.662070391715651,
      "learning_rate": 7.469505631561318e-06,
      "loss": 0.4461,
      "step": 492
    },
    {
      "epoch": 1.07,
      "grad_norm": 4.63200066859428,
      "learning_rate": 7.459313660588557e-06,
      "loss": 0.7645,
      "step": 493
    },
    {
      "epoch": 1.07,
      "grad_norm": 3.0344717928553098,
      "learning_rate": 7.449108192035701e-06,
      "loss": 0.9319,
      "step": 494
    },
    {
      "epoch": 1.07,
      "grad_norm": 3.937233258865784,
      "learning_rate": 7.4388892819139625e-06,
      "loss": 0.7877,
      "step": 495
    },
    {
      "epoch": 1.07,
      "grad_norm": 3.9766267535245468,
      "learning_rate": 7.428656986308318e-06,
      "loss": 0.5804,
      "step": 496
    },
    {
      "epoch": 1.08,
      "grad_norm": 7.429389518289982,
      "learning_rate": 7.4184113613772134e-06,
      "loss": 1.4882,
      "step": 497
    },
    {
      "epoch": 1.08,
      "grad_norm": 8.803915013316674,
      "learning_rate": 7.408152463352249e-06,
      "loss": 0.8911,
      "step": 498
    },
    {
      "epoch": 1.08,
      "grad_norm": 4.932660499571029,
      "learning_rate": 7.397880348537873e-06,
      "loss": 0.4862,
      "step": 499
    },
    {
      "epoch": 1.08,
      "grad_norm": 4.688953879099508,
      "learning_rate": 7.387595073311072e-06,
      "loss": 0.7006,
      "step": 500
    },
    {
      "epoch": 1.09,
      "grad_norm": 4.2805593552725485,
      "learning_rate": 7.3772966941210585e-06,
      "loss": 0.9436,
      "step": 501
    },
    {
      "epoch": 1.09,
      "grad_norm": 4.752371651596414,
      "learning_rate": 7.366985267488971e-06,
      "loss": 0.8117,
      "step": 502
    },
    {
      "epoch": 1.09,
      "grad_norm": 3.7191175842560518,
      "learning_rate": 7.356660850007551e-06,
      "loss": 0.6944,
      "step": 503
    },
    {
      "epoch": 1.09,
      "grad_norm": 13.272982073458952,
      "learning_rate": 7.346323498340839e-06,
      "loss": 1.2487,
      "step": 504
    },
    {
      "epoch": 1.09,
      "grad_norm": 5.02570922956184,
      "learning_rate": 7.335973269223865e-06,
      "loss": 0.7171,
      "step": 505
    },
    {
      "epoch": 1.1,
      "grad_norm": 3.6956174766095082,
      "learning_rate": 7.325610219462336e-06,
      "loss": 0.8076,
      "step": 506
    },
    {
      "epoch": 1.1,
      "grad_norm": 4.886080994573779,
      "learning_rate": 7.3152344059323165e-06,
      "loss": 1.1337,
      "step": 507
    },
    {
      "epoch": 1.1,
      "grad_norm": 7.550958577975679,
      "learning_rate": 7.304845885579933e-06,
      "loss": 0.6753,
      "step": 508
    },
    {
      "epoch": 1.1,
      "grad_norm": 5.392339863989488,
      "learning_rate": 7.294444715421043e-06,
      "loss": 0.707,
      "step": 509
    },
    {
      "epoch": 1.11,
      "grad_norm": 4.737058245977314,
      "learning_rate": 7.284030952540937e-06,
      "loss": 0.6724,
      "step": 510
    },
    {
      "epoch": 1.11,
      "grad_norm": 4.9761773797491,
      "learning_rate": 7.273604654094012e-06,
      "loss": 1.1948,
      "step": 511
    },
    {
      "epoch": 1.11,
      "grad_norm": 6.336422460667102,
      "learning_rate": 7.2631658773034715e-06,
      "loss": 0.6149,
      "step": 512
    },
    {
      "epoch": 1.11,
      "grad_norm": 5.461667584851034,
      "learning_rate": 7.252714679461001e-06,
      "loss": 1.0776,
      "step": 513
    },
    {
      "epoch": 1.11,
      "grad_norm": 3.8209830405316394,
      "learning_rate": 7.2422511179264555e-06,
      "loss": 0.735,
      "step": 514
    },
    {
      "epoch": 1.12,
      "grad_norm": 17.063776373405748,
      "learning_rate": 7.231775250127551e-06,
      "loss": 0.9933,
      "step": 515
    },
    {
      "epoch": 1.12,
      "grad_norm": 3.9408242290403863,
      "learning_rate": 7.221287133559537e-06,
      "loss": 0.9487,
      "step": 516
    },
    {
      "epoch": 1.12,
      "grad_norm": 4.715665961164967,
      "learning_rate": 7.2107868257849e-06,
      "loss": 0.6645,
      "step": 517
    },
    {
      "epoch": 1.12,
      "grad_norm": 4.887012293181657,
      "learning_rate": 7.200274384433026e-06,
      "loss": 0.8304,
      "step": 518
    },
    {
      "epoch": 1.12,
      "grad_norm": 6.120583880764877,
      "learning_rate": 7.189749867199899e-06,
      "loss": 1.2551,
      "step": 519
    },
    {
      "epoch": 1.13,
      "grad_norm": 2.8841236376632926,
      "learning_rate": 7.1792133318477775e-06,
      "loss": 0.4285,
      "step": 520
    },
    {
      "epoch": 1.13,
      "grad_norm": 3.6648188998725875,
      "learning_rate": 7.1686648362048824e-06,
      "loss": 0.5227,
      "step": 521
    },
    {
      "epoch": 1.13,
      "grad_norm": 3.46842115966258,
      "learning_rate": 7.1581044381650735e-06,
      "loss": 0.708,
      "step": 522
    },
    {
      "epoch": 1.13,
      "grad_norm": 12.793489093636781,
      "learning_rate": 7.14753219568754e-06,
      "loss": 0.7177,
      "step": 523
    },
    {
      "epoch": 1.14,
      "grad_norm": 6.046364871079263,
      "learning_rate": 7.136948166796472e-06,
      "loss": 0.7475,
      "step": 524
    },
    {
      "epoch": 1.14,
      "grad_norm": 4.1565675816256515,
      "learning_rate": 7.126352409580749e-06,
      "loss": 0.9295,
      "step": 525
    },
    {
      "epoch": 1.14,
      "grad_norm": 5.958603322195073,
      "learning_rate": 7.115744982193624e-06,
      "loss": 0.9321,
      "step": 526
    },
    {
      "epoch": 1.14,
      "grad_norm": 4.168498774538549,
      "learning_rate": 7.105125942852396e-06,
      "loss": 0.9775,
      "step": 527
    },
    {
      "epoch": 1.14,
      "grad_norm": 4.623755754097411,
      "learning_rate": 7.094495349838093e-06,
      "loss": 1.1091,
      "step": 528
    },
    {
      "epoch": 1.15,
      "grad_norm": 4.879823949399993,
      "learning_rate": 7.083853261495159e-06,
      "loss": 1.1664,
      "step": 529
    },
    {
      "epoch": 1.15,
      "grad_norm": 4.806129301641277,
      "learning_rate": 7.073199736231123e-06,
      "loss": 0.7362,
      "step": 530
    },
    {
      "epoch": 1.15,
      "grad_norm": 4.293114279485207,
      "learning_rate": 7.062534832516288e-06,
      "loss": 0.9095,
      "step": 531
    },
    {
      "epoch": 1.15,
      "grad_norm": 9.30363068252378,
      "learning_rate": 7.051858608883404e-06,
      "loss": 1.0258,
      "step": 532
    },
    {
      "epoch": 1.15,
      "grad_norm": 4.764239368556652,
      "learning_rate": 7.041171123927347e-06,
      "loss": 0.5543,
      "step": 533
    },
    {
      "epoch": 1.16,
      "grad_norm": 6.273590648797182,
      "learning_rate": 7.0304724363048025e-06,
      "loss": 0.7014,
      "step": 534
    },
    {
      "epoch": 1.16,
      "grad_norm": 4.336978580036685,
      "learning_rate": 7.019762604733939e-06,
      "loss": 0.6642,
      "step": 535
    },
    {
      "epoch": 1.16,
      "grad_norm": 4.474360896575037,
      "learning_rate": 7.009041687994085e-06,
      "loss": 0.5569,
      "step": 536
    },
    {
      "epoch": 1.16,
      "grad_norm": 4.221112064073276,
      "learning_rate": 6.998309744925411e-06,
      "loss": 1.1484,
      "step": 537
    },
    {
      "epoch": 1.17,
      "grad_norm": 8.44513903277495,
      "learning_rate": 6.987566834428605e-06,
      "loss": 0.7669,
      "step": 538
    },
    {
      "epoch": 1.17,
      "grad_norm": 4.48773002400597,
      "learning_rate": 6.97681301546454e-06,
      "loss": 0.6679,
      "step": 539
    },
    {
      "epoch": 1.17,
      "grad_norm": 4.421708178079644,
      "learning_rate": 6.9660483470539704e-06,
      "loss": 0.7137,
      "step": 540
    },
    {
      "epoch": 1.17,
      "grad_norm": 9.278354208580838,
      "learning_rate": 6.955272888277188e-06,
      "loss": 1.0701,
      "step": 541
    },
    {
      "epoch": 1.17,
      "grad_norm": 6.140314428944263,
      "learning_rate": 6.944486698273704e-06,
      "loss": 0.8633,
      "step": 542
    },
    {
      "epoch": 1.18,
      "grad_norm": 4.968022964706049,
      "learning_rate": 6.933689836241939e-06,
      "loss": 0.9712,
      "step": 543
    },
    {
      "epoch": 1.18,
      "grad_norm": 5.577098343126258,
      "learning_rate": 6.92288236143887e-06,
      "loss": 0.6306,
      "step": 544
    },
    {
      "epoch": 1.18,
      "grad_norm": 7.2812646295214645,
      "learning_rate": 6.912064333179729e-06,
      "loss": 0.5139,
      "step": 545
    },
    {
      "epoch": 1.18,
      "grad_norm": 4.190003760068685,
      "learning_rate": 6.901235810837668e-06,
      "loss": 0.475,
      "step": 546
    },
    {
      "epoch": 1.19,
      "grad_norm": 3.315129006420571,
      "learning_rate": 6.890396853843436e-06,
      "loss": 0.4961,
      "step": 547
    },
    {
      "epoch": 1.19,
      "grad_norm": 3.2273916354830576,
      "learning_rate": 6.879547521685046e-06,
      "loss": 0.5158,
      "step": 548
    },
    {
      "epoch": 1.19,
      "grad_norm": 3.846406102548112,
      "learning_rate": 6.868687873907458e-06,
      "loss": 0.6876,
      "step": 549
    },
    {
      "epoch": 1.19,
      "grad_norm": 3.4009413970175566,
      "learning_rate": 6.857817970112246e-06,
      "loss": 0.6599,
      "step": 550
    },
    {
      "epoch": 1.19,
      "grad_norm": 3.7071152914824803,
      "learning_rate": 6.846937869957272e-06,
      "loss": 0.6584,
      "step": 551
    },
    {
      "epoch": 1.2,
      "grad_norm": 4.514893875440874,
      "learning_rate": 6.836047633156361e-06,
      "loss": 1.055,
      "step": 552
    },
    {
      "epoch": 1.2,
      "grad_norm": 4.613262630800641,
      "learning_rate": 6.8251473194789695e-06,
      "loss": 0.7814,
      "step": 553
    },
    {
      "epoch": 1.2,
      "grad_norm": 4.178023119488768,
      "learning_rate": 6.814236988749863e-06,
      "loss": 0.7497,
      "step": 554
    },
    {
      "epoch": 1.2,
      "grad_norm": 5.399813203820247,
      "learning_rate": 6.8033167008487784e-06,
      "loss": 0.7111,
      "step": 555
    },
    {
      "epoch": 1.2,
      "grad_norm": 5.923197620910671,
      "learning_rate": 6.792386515710106e-06,
      "loss": 0.6366,
      "step": 556
    },
    {
      "epoch": 1.21,
      "grad_norm": 4.746346535814018,
      "learning_rate": 6.7814464933225535e-06,
      "loss": 1.2167,
      "step": 557
    },
    {
      "epoch": 1.21,
      "grad_norm": 4.028439115247876,
      "learning_rate": 6.77049669372882e-06,
      "loss": 0.5582,
      "step": 558
    },
    {
      "epoch": 1.21,
      "grad_norm": 3.5041563276504553,
      "learning_rate": 6.759537177025263e-06,
      "loss": 0.7432,
      "step": 559
    },
    {
      "epoch": 1.21,
      "grad_norm": 6.5679975816289256,
      "learning_rate": 6.748568003361576e-06,
      "loss": 0.5429,
      "step": 560
    },
    {
      "epoch": 1.22,
      "grad_norm": 3.83897683304579,
      "learning_rate": 6.737589232940445e-06,
      "loss": 0.6961,
      "step": 561
    },
    {
      "epoch": 1.22,
      "grad_norm": 5.0520725262276525,
      "learning_rate": 6.726600926017234e-06,
      "loss": 0.4801,
      "step": 562
    },
    {
      "epoch": 1.22,
      "grad_norm": 3.900451446996258,
      "learning_rate": 6.715603142899645e-06,
      "loss": 0.6139,
      "step": 563
    },
    {
      "epoch": 1.22,
      "grad_norm": 5.183678015498263,
      "learning_rate": 6.704595943947385e-06,
      "loss": 1.1473,
      "step": 564
    },
    {
      "epoch": 1.22,
      "grad_norm": 3.2156011176538377,
      "learning_rate": 6.693579389571844e-06,
      "loss": 0.6277,
      "step": 565
    },
    {
      "epoch": 1.23,
      "grad_norm": 4.771821614828381,
      "learning_rate": 6.682553540235754e-06,
      "loss": 0.916,
      "step": 566
    },
    {
      "epoch": 1.23,
      "grad_norm": 3.7310312029700152,
      "learning_rate": 6.671518456452859e-06,
      "loss": 0.7245,
      "step": 567
    },
    {
      "epoch": 1.23,
      "grad_norm": 4.236929851154136,
      "learning_rate": 6.6604741987875905e-06,
      "loss": 0.8908,
      "step": 568
    },
    {
      "epoch": 1.23,
      "grad_norm": 5.918040187647108,
      "learning_rate": 6.649420827854729e-06,
      "loss": 0.4952,
      "step": 569
    },
    {
      "epoch": 1.24,
      "grad_norm": 4.026444456642722,
      "learning_rate": 6.638358404319064e-06,
      "loss": 0.5557,
      "step": 570
    },
    {
      "epoch": 1.24,
      "grad_norm": 3.7726770863791916,
      "learning_rate": 6.62728698889508e-06,
      "loss": 0.7233,
      "step": 571
    },
    {
      "epoch": 1.24,
      "grad_norm": 5.588043463477253,
      "learning_rate": 6.616206642346603e-06,
      "loss": 0.7511,
      "step": 572
    },
    {
      "epoch": 1.24,
      "grad_norm": 4.096165156912804,
      "learning_rate": 6.605117425486483e-06,
      "loss": 0.9166,
      "step": 573
    },
    {
      "epoch": 1.24,
      "grad_norm": 5.053499557257062,
      "learning_rate": 6.594019399176246e-06,
      "loss": 0.6312,
      "step": 574
    },
    {
      "epoch": 1.25,
      "grad_norm": 4.819663139282108,
      "learning_rate": 6.582912624325777e-06,
      "loss": 1.0514,
      "step": 575
    },
    {
      "epoch": 1.25,
      "grad_norm": 8.765617347316468,
      "learning_rate": 6.571797161892965e-06,
      "loss": 0.6527,
      "step": 576
    },
    {
      "epoch": 1.25,
      "grad_norm": 4.16056048833256,
      "learning_rate": 6.5606730728833904e-06,
      "loss": 1.1848,
      "step": 577
    },
    {
      "epoch": 1.25,
      "grad_norm": 5.232763900496471,
      "learning_rate": 6.549540418349969e-06,
      "loss": 1.1141,
      "step": 578
    },
    {
      "epoch": 1.25,
      "grad_norm": 3.5342310207226517,
      "learning_rate": 6.538399259392637e-06,
      "loss": 0.7612,
      "step": 579
    },
    {
      "epoch": 1.26,
      "grad_norm": 7.098473409366254,
      "learning_rate": 6.527249657157998e-06,
      "loss": 0.9515,
      "step": 580
    },
    {
      "epoch": 1.26,
      "grad_norm": 5.315601893630016,
      "learning_rate": 6.516091672839e-06,
      "loss": 0.9006,
      "step": 581
    },
    {
      "epoch": 1.26,
      "grad_norm": 4.707098814291459,
      "learning_rate": 6.504925367674595e-06,
      "loss": 0.8706,
      "step": 582
    },
    {
      "epoch": 1.26,
      "grad_norm": 3.8000836101963626,
      "learning_rate": 6.4937508029493965e-06,
      "loss": 0.7443,
      "step": 583
    },
    {
      "epoch": 1.27,
      "grad_norm": 4.351853235391886,
      "learning_rate": 6.482568039993356e-06,
      "loss": 0.5871,
      "step": 584
    },
    {
      "epoch": 1.27,
      "grad_norm": 7.313439106072162,
      "learning_rate": 6.471377140181419e-06,
      "loss": 1.0306,
      "step": 585
    },
    {
      "epoch": 1.27,
      "grad_norm": 4.985523293750241,
      "learning_rate": 6.4601781649331885e-06,
      "loss": 0.9134,
      "step": 586
    },
    {
      "epoch": 1.27,
      "grad_norm": 6.242346127512318,
      "learning_rate": 6.4489711757125814e-06,
      "loss": 1.0284,
      "step": 587
    },
    {
      "epoch": 1.27,
      "grad_norm": 4.191394201465743,
      "learning_rate": 6.437756234027512e-06,
      "loss": 0.9135,
      "step": 588
    },
    {
      "epoch": 1.28,
      "grad_norm": 3.7001840248830526,
      "learning_rate": 6.4265334014295284e-06,
      "loss": 0.5841,
      "step": 589
    },
    {
      "epoch": 1.28,
      "grad_norm": 4.832495905956136,
      "learning_rate": 6.415302739513492e-06,
      "loss": 0.819,
      "step": 590
    },
    {
      "epoch": 1.28,
      "grad_norm": 4.285002468017606,
      "learning_rate": 6.40406430991723e-06,
      "loss": 0.6539,
      "step": 591
    },
    {
      "epoch": 1.28,
      "grad_norm": 7.547589049140377,
      "learning_rate": 6.392818174321213e-06,
      "loss": 0.6665,
      "step": 592
    },
    {
      "epoch": 1.28,
      "grad_norm": 6.854518599895364,
      "learning_rate": 6.3815643944481866e-06,
      "loss": 0.9152,
      "step": 593
    },
    {
      "epoch": 1.29,
      "grad_norm": 4.398005246216444,
      "learning_rate": 6.370303032062869e-06,
      "loss": 0.6803,
      "step": 594
    },
    {
      "epoch": 1.29,
      "grad_norm": 6.845640196241702,
      "learning_rate": 6.359034148971581e-06,
      "loss": 1.1216,
      "step": 595
    },
    {
      "epoch": 1.29,
      "grad_norm": 4.685009566091401,
      "learning_rate": 6.347757807021926e-06,
      "loss": 0.744,
      "step": 596
    },
    {
      "epoch": 1.29,
      "grad_norm": 4.890597056272389,
      "learning_rate": 6.336474068102444e-06,
      "loss": 0.7789,
      "step": 597
    },
    {
      "epoch": 1.3,
      "grad_norm": 7.294863791115077,
      "learning_rate": 6.325182994142267e-06,
      "loss": 1.0683,
      "step": 598
    },
    {
      "epoch": 1.3,
      "grad_norm": 7.726982584797627,
      "learning_rate": 6.3138846471107925e-06,
      "loss": 0.9563,
      "step": 599
    },
    {
      "epoch": 1.3,
      "grad_norm": 4.746835361270983,
      "learning_rate": 6.302579089017328e-06,
      "loss": 0.5483,
      "step": 600
    },
    {
      "epoch": 1.3,
      "grad_norm": 3.3693051399473886,
      "learning_rate": 6.291266381910761e-06,
      "loss": 0.871,
      "step": 601
    },
    {
      "epoch": 1.3,
      "grad_norm": 3.370358413889282,
      "learning_rate": 6.279946587879216e-06,
      "loss": 0.5022,
      "step": 602
    },
    {
      "epoch": 1.31,
      "grad_norm": 4.405289510530603,
      "learning_rate": 6.268619769049713e-06,
      "loss": 0.8934,
      "step": 603
    },
    {
      "epoch": 1.31,
      "grad_norm": 7.312615270365992,
      "learning_rate": 6.2572859875878225e-06,
      "loss": 1.436,
      "step": 604
    },
    {
      "epoch": 1.31,
      "grad_norm": 4.726057261419138,
      "learning_rate": 6.245945305697335e-06,
      "loss": 1.0793,
      "step": 605
    },
    {
      "epoch": 1.31,
      "grad_norm": 8.031269538478732,
      "learning_rate": 6.234597785619906e-06,
      "loss": 1.2222,
      "step": 606
    },
    {
      "epoch": 1.32,
      "grad_norm": 5.026622143556818,
      "learning_rate": 6.223243489634727e-06,
      "loss": 0.6623,
      "step": 607
    },
    {
      "epoch": 1.32,
      "grad_norm": 3.7840932939307383,
      "learning_rate": 6.211882480058175e-06,
      "loss": 1.0045,
      "step": 608
    },
    {
      "epoch": 1.32,
      "grad_norm": 5.943589600924059,
      "learning_rate": 6.200514819243476e-06,
      "loss": 0.5478,
      "step": 609
    },
    {
      "epoch": 1.32,
      "grad_norm": 3.812090291975411,
      "learning_rate": 6.189140569580356e-06,
      "loss": 0.8765,
      "step": 610
    },
    {
      "epoch": 1.32,
      "grad_norm": 3.8390497819223697,
      "learning_rate": 6.1777597934947084e-06,
      "loss": 0.7506,
      "step": 611
    },
    {
      "epoch": 1.33,
      "grad_norm": 4.1795489387778195,
      "learning_rate": 6.166372553448241e-06,
      "loss": 0.9004,
      "step": 612
    },
    {
      "epoch": 1.33,
      "grad_norm": 3.012761715714778,
      "learning_rate": 6.154978911938143e-06,
      "loss": 0.6815,
      "step": 613
    },
    {
      "epoch": 1.33,
      "grad_norm": 3.086963231140802,
      "learning_rate": 6.143578931496732e-06,
      "loss": 0.5118,
      "step": 614
    },
    {
      "epoch": 1.33,
      "grad_norm": 4.5763051909063535,
      "learning_rate": 6.132172674691119e-06,
      "loss": 1.0439,
      "step": 615
    },
    {
      "epoch": 1.33,
      "grad_norm": 5.572026592855522,
      "learning_rate": 6.120760204122862e-06,
      "loss": 0.7312,
      "step": 616
    },
    {
      "epoch": 1.34,
      "grad_norm": 5.256436954727932,
      "learning_rate": 6.109341582427621e-06,
      "loss": 1.0365,
      "step": 617
    },
    {
      "epoch": 1.34,
      "grad_norm": 3.506773959393946,
      "learning_rate": 6.097916872274815e-06,
      "loss": 0.6215,
      "step": 618
    },
    {
      "epoch": 1.34,
      "grad_norm": 4.831802197030479,
      "learning_rate": 6.086486136367281e-06,
      "loss": 0.778,
      "step": 619
    },
    {
      "epoch": 1.34,
      "grad_norm": 6.190794369942719,
      "learning_rate": 6.075049437440927e-06,
      "loss": 0.8434,
      "step": 620
    },
    {
      "epoch": 1.35,
      "grad_norm": 4.891451244272097,
      "learning_rate": 6.063606838264384e-06,
      "loss": 0.9422,
      "step": 621
    },
    {
      "epoch": 1.35,
      "grad_norm": 7.278346471027911,
      "learning_rate": 6.0521584016386735e-06,
      "loss": 0.7854,
      "step": 622
    },
    {
      "epoch": 1.35,
      "grad_norm": 4.492232814776734,
      "learning_rate": 6.040704190396847e-06,
      "loss": 0.7011,
      "step": 623
    },
    {
      "epoch": 1.35,
      "grad_norm": 4.685211357904295,
      "learning_rate": 6.029244267403652e-06,
      "loss": 0.9886,
      "step": 624
    },
    {
      "epoch": 1.35,
      "grad_norm": 4.283060495046248,
      "learning_rate": 6.0177786955551874e-06,
      "loss": 1.2869,
      "step": 625
    },
    {
      "epoch": 1.36,
      "grad_norm": 4.129998668654544,
      "learning_rate": 6.006307537778552e-06,
      "loss": 0.5871,
      "step": 626
    },
    {
      "epoch": 1.36,
      "grad_norm": 4.938376916804373,
      "learning_rate": 5.9948308570315e-06,
      "loss": 0.8201,
      "step": 627
    },
    {
      "epoch": 1.36,
      "grad_norm": 3.710371101001817,
      "learning_rate": 5.983348716302101e-06,
      "loss": 0.7914,
      "step": 628
    },
    {
      "epoch": 1.36,
      "grad_norm": 4.33068147708345,
      "learning_rate": 5.97186117860839e-06,
      "loss": 1.598,
      "step": 629
    },
    {
      "epoch": 1.37,
      "grad_norm": 5.0834374798650295,
      "learning_rate": 5.960368306998023e-06,
      "loss": 1.0107,
      "step": 630
    },
    {
      "epoch": 1.37,
      "grad_norm": 4.543910516801835,
      "learning_rate": 5.948870164547932e-06,
      "loss": 0.9229,
      "step": 631
    },
    {
      "epoch": 1.37,
      "grad_norm": 4.085463158903914,
      "learning_rate": 5.9373668143639694e-06,
      "loss": 0.9099,
      "step": 632
    },
    {
      "epoch": 1.37,
      "grad_norm": 8.278059620563297,
      "learning_rate": 5.92585831958058e-06,
      "loss": 0.3782,
      "step": 633
    },
    {
      "epoch": 1.37,
      "grad_norm": 3.367935418838566,
      "learning_rate": 5.914344743360435e-06,
      "loss": 0.8209,
      "step": 634
    },
    {
      "epoch": 1.38,
      "grad_norm": 3.702344211547487,
      "learning_rate": 5.902826148894102e-06,
      "loss": 0.7207,
      "step": 635
    },
    {
      "epoch": 1.38,
      "grad_norm": 3.5254705320240913,
      "learning_rate": 5.891302599399686e-06,
      "loss": 1.0203,
      "step": 636
    },
    {
      "epoch": 1.38,
      "grad_norm": 4.933676158072319,
      "learning_rate": 5.8797741581224866e-06,
      "loss": 1.1941,
      "step": 637
    },
    {
      "epoch": 1.38,
      "grad_norm": 4.188637928188261,
      "learning_rate": 5.8682408883346535e-06,
      "loss": 0.8622,
      "step": 638
    },
    {
      "epoch": 1.38,
      "grad_norm": 4.243287145852907,
      "learning_rate": 5.856702853334833e-06,
      "loss": 0.9248,
      "step": 639
    },
    {
      "epoch": 1.39,
      "grad_norm": 16.4313378114192,
      "learning_rate": 5.845160116447833e-06,
      "loss": 0.6026,
      "step": 640
    },
    {
      "epoch": 1.39,
      "grad_norm": 3.8248044358462567,
      "learning_rate": 5.833612741024256e-06,
      "loss": 0.8453,
      "step": 641
    },
    {
      "epoch": 1.39,
      "grad_norm": 3.5670671121728383,
      "learning_rate": 5.8220607904401725e-06,
      "loss": 0.5977,
      "step": 642
    },
    {
      "epoch": 1.39,
      "grad_norm": 5.611520944736675,
      "learning_rate": 5.810504328096756e-06,
      "loss": 0.9563,
      "step": 643
    },
    {
      "epoch": 1.4,
      "grad_norm": 6.357203921917924,
      "learning_rate": 5.798943417419944e-06,
      "loss": 1.1505,
      "step": 644
    },
    {
      "epoch": 1.4,
      "grad_norm": 9.210773250786598,
      "learning_rate": 5.78737812186009e-06,
      "loss": 0.6838,
      "step": 645
    },
    {
      "epoch": 1.4,
      "grad_norm": 5.225137934649469,
      "learning_rate": 5.775808504891612e-06,
      "loss": 0.7771,
      "step": 646
    },
    {
      "epoch": 1.4,
      "grad_norm": 9.201838597339028,
      "learning_rate": 5.764234630012643e-06,
      "loss": 0.7846,
      "step": 647
    },
    {
      "epoch": 1.4,
      "grad_norm": 5.358226984333985,
      "learning_rate": 5.752656560744692e-06,
      "loss": 0.9535,
      "step": 648
    },
    {
      "epoch": 1.41,
      "grad_norm": 6.3326056143169,
      "learning_rate": 5.741074360632278e-06,
      "loss": 0.8374,
      "step": 649
    },
    {
      "epoch": 1.41,
      "grad_norm": 3.899065421948101,
      "learning_rate": 5.729488093242601e-06,
      "loss": 0.865,
      "step": 650
    },
    {
      "epoch": 1.41,
      "grad_norm": 3.8766465588406374,
      "learning_rate": 5.717897822165179e-06,
      "loss": 0.6312,
      "step": 651
    },
    {
      "epoch": 1.41,
      "grad_norm": 4.004354935718821,
      "learning_rate": 5.706303611011502e-06,
      "loss": 0.6085,
      "step": 652
    },
    {
      "epoch": 1.41,
      "grad_norm": 3.6167722873745642,
      "learning_rate": 5.694705523414691e-06,
      "loss": 0.5921,
      "step": 653
    },
    {
      "epoch": 1.42,
      "grad_norm": 11.54820270884967,
      "learning_rate": 5.6831036230291345e-06,
      "loss": 0.9268,
      "step": 654
    },
    {
      "epoch": 1.42,
      "grad_norm": 3.978205239439661,
      "learning_rate": 5.671497973530152e-06,
      "loss": 0.6033,
      "step": 655
    },
    {
      "epoch": 1.42,
      "grad_norm": 5.47494789876935,
      "learning_rate": 5.659888638613638e-06,
      "loss": 0.7947,
      "step": 656
    },
    {
      "epoch": 1.42,
      "grad_norm": 7.523404347952476,
      "learning_rate": 5.648275681995716e-06,
      "loss": 0.8731,
      "step": 657
    },
    {
      "epoch": 1.43,
      "grad_norm": 8.509525466610395,
      "learning_rate": 5.636659167412381e-06,
      "loss": 0.6351,
      "step": 658
    },
    {
      "epoch": 1.43,
      "grad_norm": 5.358684440617374,
      "learning_rate": 5.625039158619161e-06,
      "loss": 0.8919,
      "step": 659
    },
    {
      "epoch": 1.43,
      "grad_norm": 4.153602392716318,
      "learning_rate": 5.613415719390759e-06,
      "loss": 0.7977,
      "step": 660
    },
    {
      "epoch": 1.43,
      "grad_norm": 4.583696678657548,
      "learning_rate": 5.601788913520706e-06,
      "loss": 0.5462,
      "step": 661
    },
    {
      "epoch": 1.43,
      "grad_norm": 4.350413106197617,
      "learning_rate": 5.590158804821011e-06,
      "loss": 0.5968,
      "step": 662
    },
    {
      "epoch": 1.44,
      "grad_norm": 6.341699181909662,
      "learning_rate": 5.578525457121807e-06,
      "loss": 0.9821,
      "step": 663
    },
    {
      "epoch": 1.44,
      "grad_norm": 4.0474611893396775,
      "learning_rate": 5.566888934271007e-06,
      "loss": 0.5021,
      "step": 664
    },
    {
      "epoch": 1.44,
      "grad_norm": 4.9412944576207485,
      "learning_rate": 5.5552493001339535e-06,
      "loss": 0.6499,
      "step": 665
    },
    {
      "epoch": 1.44,
      "grad_norm": 4.046643035538699,
      "learning_rate": 5.543606618593053e-06,
      "loss": 0.7663,
      "step": 666
    },
    {
      "epoch": 1.45,
      "grad_norm": 4.5511750984030765,
      "learning_rate": 5.531960953547452e-06,
      "loss": 0.756,
      "step": 667
    },
    {
      "epoch": 1.45,
      "grad_norm": 4.812669094190228,
      "learning_rate": 5.520312368912661e-06,
      "loss": 0.508,
      "step": 668
    },
    {
      "epoch": 1.45,
      "grad_norm": 4.158377195033854,
      "learning_rate": 5.508660928620216e-06,
      "loss": 0.8016,
      "step": 669
    },
    {
      "epoch": 1.45,
      "grad_norm": 5.282502994512615,
      "learning_rate": 5.497006696617333e-06,
      "loss": 1.0109,
      "step": 670
    },
    {
      "epoch": 1.45,
      "grad_norm": 3.745634514837313,
      "learning_rate": 5.485349736866541e-06,
      "loss": 0.6266,
      "step": 671
    },
    {
      "epoch": 1.46,
      "grad_norm": 3.6831496390833034,
      "learning_rate": 5.473690113345343e-06,
      "loss": 0.8167,
      "step": 672
    },
    {
      "epoch": 1.46,
      "grad_norm": 5.911708962760637,
      "learning_rate": 5.462027890045862e-06,
      "loss": 1.3342,
      "step": 673
    },
    {
      "epoch": 1.46,
      "grad_norm": 5.768437390344328,
      "learning_rate": 5.450363130974492e-06,
      "loss": 0.8137,
      "step": 674
    },
    {
      "epoch": 1.46,
      "grad_norm": 3.5018277853479356,
      "learning_rate": 5.438695900151537e-06,
      "loss": 0.6084,
      "step": 675
    },
    {
      "epoch": 1.46,
      "grad_norm": 3.861207188385205,
      "learning_rate": 5.427026261610877e-06,
      "loss": 0.8033,
      "step": 676
    },
    {
      "epoch": 1.47,
      "grad_norm": 5.622487493364923,
      "learning_rate": 5.4153542793995985e-06,
      "loss": 1.1146,
      "step": 677
    },
    {
      "epoch": 1.47,
      "grad_norm": 5.195838368832894,
      "learning_rate": 5.403680017577653e-06,
      "loss": 0.8582,
      "step": 678
    },
    {
      "epoch": 1.47,
      "grad_norm": 3.788380054670546,
      "learning_rate": 5.392003540217505e-06,
      "loss": 1.3347,
      "step": 679
    },
    {
      "epoch": 1.47,
      "grad_norm": 3.980479710568579,
      "learning_rate": 5.380324911403776e-06,
      "loss": 1.2648,
      "step": 680
    },
    {
      "epoch": 1.48,
      "grad_norm": 2.8696529482551454,
      "learning_rate": 5.368644195232896e-06,
      "loss": 0.5512,
      "step": 681
    },
    {
      "epoch": 1.48,
      "grad_norm": 4.532676231340781,
      "learning_rate": 5.356961455812754e-06,
      "loss": 0.8363,
      "step": 682
    },
    {
      "epoch": 1.48,
      "grad_norm": 4.126021049654493,
      "learning_rate": 5.34527675726234e-06,
      "loss": 0.6017,
      "step": 683
    },
    {
      "epoch": 1.48,
      "grad_norm": 4.522780874819132,
      "learning_rate": 5.3335901637113985e-06,
      "loss": 0.7216,
      "step": 684
    },
    {
      "epoch": 1.48,
      "grad_norm": 5.617814121129761,
      "learning_rate": 5.321901739300074e-06,
      "loss": 0.8176,
      "step": 685
    },
    {
      "epoch": 1.49,
      "grad_norm": 8.11473871816298,
      "learning_rate": 5.310211548178556e-06,
      "loss": 1.0311,
      "step": 686
    },
    {
      "epoch": 1.49,
      "grad_norm": 4.17991505820069,
      "learning_rate": 5.298519654506736e-06,
      "loss": 0.5411,
      "step": 687
    },
    {
      "epoch": 1.49,
      "grad_norm": 5.454035373368481,
      "learning_rate": 5.286826122453847e-06,
      "loss": 1.1865,
      "step": 688
    },
    {
      "epoch": 1.49,
      "grad_norm": 4.217368877185839,
      "learning_rate": 5.275131016198112e-06,
      "loss": 0.8641,
      "step": 689
    },
    {
      "epoch": 1.5,
      "grad_norm": 3.981680730499904,
      "learning_rate": 5.2634343999263985e-06,
      "loss": 0.5225,
      "step": 690
    },
    {
      "epoch": 1.5,
      "grad_norm": 6.732351400589309,
      "learning_rate": 5.251736337833857e-06,
      "loss": 1.1268,
      "step": 691
    },
    {
      "epoch": 1.5,
      "grad_norm": 4.508824123570986,
      "learning_rate": 5.2400368941235745e-06,
      "loss": 1.0734,
      "step": 692
    },
    {
      "epoch": 1.5,
      "grad_norm": 3.8887452547112376,
      "learning_rate": 5.228336133006223e-06,
      "loss": 0.6999,
      "step": 693
    },
    {
      "epoch": 1.5,
      "grad_norm": 5.967305981759056,
      "learning_rate": 5.216634118699701e-06,
      "loss": 0.9105,
      "step": 694
    },
    {
      "epoch": 1.51,
      "grad_norm": 5.067040638746576,
      "learning_rate": 5.20493091542879e-06,
      "loss": 0.9205,
      "step": 695
    },
    {
      "epoch": 1.51,
      "grad_norm": 14.123435475367321,
      "learning_rate": 5.193226587424793e-06,
      "loss": 0.68,
      "step": 696
    },
    {
      "epoch": 1.51,
      "grad_norm": 3.588883243120534,
      "learning_rate": 5.181521198925183e-06,
      "loss": 0.5447,
      "step": 697
    },
    {
      "epoch": 1.51,
      "grad_norm": 4.272806408820429,
      "learning_rate": 5.169814814173263e-06,
      "loss": 0.805,
      "step": 698
    },
    {
      "epoch": 1.51,
      "grad_norm": 10.856741311019459,
      "learning_rate": 5.158107497417795e-06,
      "loss": 1.6276,
      "step": 699
    },
    {
      "epoch": 1.52,
      "grad_norm": 4.6213509709197425,
      "learning_rate": 5.14639931291266e-06,
      "loss": 0.5754,
      "step": 700
    },
    {
      "epoch": 1.52,
      "grad_norm": 4.3376588182364735,
      "learning_rate": 5.134690324916502e-06,
      "loss": 0.7027,
      "step": 701
    },
    {
      "epoch": 1.52,
      "grad_norm": 4.473414310469182,
      "learning_rate": 5.122980597692372e-06,
      "loss": 0.8444,
      "step": 702
    },
    {
      "epoch": 1.52,
      "grad_norm": 5.043458006574416,
      "learning_rate": 5.11127019550738e-06,
      "loss": 0.9398,
      "step": 703
    },
    {
      "epoch": 1.53,
      "grad_norm": 5.171554900425086,
      "learning_rate": 5.099559182632342e-06,
      "loss": 0.793,
      "step": 704
    },
    {
      "epoch": 1.53,
      "grad_norm": 3.0656759902516413,
      "learning_rate": 5.087847623341421e-06,
      "loss": 0.8292,
      "step": 705
    },
    {
      "epoch": 1.53,
      "grad_norm": 3.684208014048517,
      "learning_rate": 5.076135581911784e-06,
      "loss": 0.7525,
      "step": 706
    },
    {
      "epoch": 1.53,
      "grad_norm": 5.476531438726525,
      "learning_rate": 5.0644231226232434e-06,
      "loss": 0.518,
      "step": 707
    },
    {
      "epoch": 1.53,
      "grad_norm": 2.5888884739665667,
      "learning_rate": 5.052710309757899e-06,
      "loss": 0.4806,
      "step": 708
    },
    {
      "epoch": 1.54,
      "grad_norm": 4.5148049894393205,
      "learning_rate": 5.040997207599798e-06,
      "loss": 0.6303,
      "step": 709
    },
    {
      "epoch": 1.54,
      "grad_norm": 4.696669970337422,
      "learning_rate": 5.029283880434575e-06,
      "loss": 0.8405,
      "step": 710
    },
    {
      "epoch": 1.54,
      "grad_norm": 4.105444460901763,
      "learning_rate": 5.0175703925490936e-06,
      "loss": 0.9452,
      "step": 711
    },
    {
      "epoch": 1.54,
      "grad_norm": 3.7111636187580013,
      "learning_rate": 5.005856808231108e-06,
      "loss": 0.6518,
      "step": 712
    },
    {
      "epoch": 1.54,
      "grad_norm": 3.5656324134372595,
      "learning_rate": 4.994143191768893e-06,
      "loss": 0.6231,
      "step": 713
    },
    {
      "epoch": 1.55,
      "grad_norm": 3.628453449182133,
      "learning_rate": 4.982429607450907e-06,
      "loss": 0.8894,
      "step": 714
    },
    {
      "epoch": 1.55,
      "grad_norm": 3.8249129287281125,
      "learning_rate": 4.970716119565427e-06,
      "loss": 0.5449,
      "step": 715
    },
    {
      "epoch": 1.55,
      "grad_norm": 3.4639102794783585,
      "learning_rate": 4.959002792400205e-06,
      "loss": 1.0635,
      "step": 716
    },
    {
      "epoch": 1.55,
      "grad_norm": 2.841679684437223,
      "learning_rate": 4.947289690242103e-06,
      "loss": 0.5156,
      "step": 717
    },
    {
      "epoch": 1.56,
      "grad_norm": 4.0728175754161,
      "learning_rate": 4.935576877376759e-06,
      "loss": 0.7773,
      "step": 718
    },
    {
      "epoch": 1.56,
      "grad_norm": 2.888494502202359,
      "learning_rate": 4.9238644180882175e-06,
      "loss": 0.8587,
      "step": 719
    },
    {
      "epoch": 1.56,
      "grad_norm": 3.7974346245928463,
      "learning_rate": 4.91215237665858e-06,
      "loss": 0.6479,
      "step": 720
    },
    {
      "epoch": 1.56,
      "grad_norm": 5.045516935360225,
      "learning_rate": 4.900440817367661e-06,
      "loss": 0.5905,
      "step": 721
    },
    {
      "epoch": 1.56,
      "grad_norm": 7.910118355713333,
      "learning_rate": 4.88872980449262e-06,
      "loss": 0.6173,
      "step": 722
    },
    {
      "epoch": 1.57,
      "grad_norm": 3.6008491832675267,
      "learning_rate": 4.877019402307629e-06,
      "loss": 0.6901,
      "step": 723
    },
    {
      "epoch": 1.57,
      "grad_norm": 3.247362927972009,
      "learning_rate": 4.8653096750835e-06,
      "loss": 0.8205,
      "step": 724
    },
    {
      "epoch": 1.57,
      "grad_norm": 3.741012616491536,
      "learning_rate": 4.853600687087342e-06,
      "loss": 1.1564,
      "step": 725
    },
    {
      "epoch": 1.57,
      "grad_norm": 3.445143220604005,
      "learning_rate": 4.841892502582206e-06,
      "loss": 0.6222,
      "step": 726
    },
    {
      "epoch": 1.58,
      "grad_norm": 3.095861011332695,
      "learning_rate": 4.830185185826739e-06,
      "loss": 0.7213,
      "step": 727
    },
    {
      "epoch": 1.58,
      "grad_norm": 9.699084314826322,
      "learning_rate": 4.818478801074818e-06,
      "loss": 0.8693,
      "step": 728
    },
    {
      "epoch": 1.58,
      "grad_norm": 6.023226690844008,
      "learning_rate": 4.806773412575211e-06,
      "loss": 0.6292,
      "step": 729
    },
    {
      "epoch": 1.58,
      "grad_norm": 3.5515098977690887,
      "learning_rate": 4.795069084571211e-06,
      "loss": 0.6723,
      "step": 730
    },
    {
      "epoch": 1.58,
      "grad_norm": 3.7006074616388904,
      "learning_rate": 4.7833658813002995e-06,
      "loss": 0.7684,
      "step": 731
    },
    {
      "epoch": 1.59,
      "grad_norm": 3.413485495131064,
      "learning_rate": 4.7716638669937784e-06,
      "loss": 0.8193,
      "step": 732
    },
    {
      "epoch": 1.59,
      "grad_norm": 3.4756569770754346,
      "learning_rate": 4.759963105876428e-06,
      "loss": 1.0946,
      "step": 733
    },
    {
      "epoch": 1.59,
      "grad_norm": 4.870729758495444,
      "learning_rate": 4.748263662166145e-06,
      "loss": 0.611,
      "step": 734
    },
    {
      "epoch": 1.59,
      "grad_norm": 4.7372571923083,
      "learning_rate": 4.736565600073602e-06,
      "loss": 0.878,
      "step": 735
    },
    {
      "epoch": 1.59,
      "grad_norm": 5.0881982744001455,
      "learning_rate": 4.724868983801889e-06,
      "loss": 1.1393,
      "step": 736
    },
    {
      "epoch": 1.6,
      "grad_norm": 5.3254718357922295,
      "learning_rate": 4.713173877546155e-06,
      "loss": 0.6504,
      "step": 737
    },
    {
      "epoch": 1.6,
      "grad_norm": 5.172499918688033,
      "learning_rate": 4.701480345493266e-06,
      "loss": 1.2642,
      "step": 738
    },
    {
      "epoch": 1.6,
      "grad_norm": 3.5689986829390334,
      "learning_rate": 4.689788451821445e-06,
      "loss": 1.0345,
      "step": 739
    },
    {
      "epoch": 1.6,
      "grad_norm": 3.705432695773935,
      "learning_rate": 4.678098260699928e-06,
      "loss": 0.7087,
      "step": 740
    },
    {
      "epoch": 1.61,
      "grad_norm": 3.219179549547455,
      "learning_rate": 4.666409836288603e-06,
      "loss": 0.6367,
      "step": 741
    },
    {
      "epoch": 1.61,
      "grad_norm": 5.255732078999917,
      "learning_rate": 4.654723242737661e-06,
      "loss": 0.7115,
      "step": 742
    },
    {
      "epoch": 1.61,
      "grad_norm": 4.149858290734801,
      "learning_rate": 4.643038544187246e-06,
      "loss": 0.9283,
      "step": 743
    },
    {
      "epoch": 1.61,
      "grad_norm": 3.1276721307868645,
      "learning_rate": 4.631355804767106e-06,
      "loss": 0.6885,
      "step": 744
    },
    {
      "epoch": 1.61,
      "grad_norm": 4.99286365373704,
      "learning_rate": 4.619675088596226e-06,
      "loss": 0.5559,
      "step": 745
    },
    {
      "epoch": 1.62,
      "grad_norm": 4.159295260704925,
      "learning_rate": 4.607996459782498e-06,
      "loss": 0.9156,
      "step": 746
    },
    {
      "epoch": 1.62,
      "grad_norm": 4.278878515177881,
      "learning_rate": 4.596319982422348e-06,
      "loss": 0.8749,
      "step": 747
    },
    {
      "epoch": 1.62,
      "grad_norm": 3.3018026506666307,
      "learning_rate": 4.584645720600403e-06,
      "loss": 0.572,
      "step": 748
    },
    {
      "epoch": 1.62,
      "grad_norm": 2.756721452694666,
      "learning_rate": 4.572973738389124e-06,
      "loss": 0.9567,
      "step": 749
    },
    {
      "epoch": 1.63,
      "grad_norm": 7.1749632535732815,
      "learning_rate": 4.561304099848464e-06,
      "loss": 0.7456,
      "step": 750
    },
    {
      "epoch": 1.63,
      "grad_norm": 3.815152076975536,
      "learning_rate": 4.549636869025511e-06,
      "loss": 0.9782,
      "step": 751
    },
    {
      "epoch": 1.63,
      "grad_norm": 3.150240679362041,
      "learning_rate": 4.5379721099541385e-06,
      "loss": 1.0206,
      "step": 752
    },
    {
      "epoch": 1.63,
      "grad_norm": 2.8397660912234395,
      "learning_rate": 4.526309886654659e-06,
      "loss": 0.5227,
      "step": 753
    },
    {
      "epoch": 1.63,
      "grad_norm": 6.620867432088285,
      "learning_rate": 4.514650263133461e-06,
      "loss": 0.9101,
      "step": 754
    },
    {
      "epoch": 1.64,
      "grad_norm": 5.289513984757383,
      "learning_rate": 4.502993303382669e-06,
      "loss": 0.5222,
      "step": 755
    },
    {
      "epoch": 1.64,
      "grad_norm": 3.603217855651493,
      "learning_rate": 4.491339071379783e-06,
      "loss": 0.6508,
      "step": 756
    },
    {
      "epoch": 1.64,
      "grad_norm": 3.1462942318180485,
      "learning_rate": 4.47968763108734e-06,
      "loss": 0.6058,
      "step": 757
    },
    {
      "epoch": 1.64,
      "grad_norm": 6.6357774648749395,
      "learning_rate": 4.46803904645255e-06,
      "loss": 0.8383,
      "step": 758
    },
    {
      "epoch": 1.64,
      "grad_norm": 3.471641388495333,
      "learning_rate": 4.4563933814069475e-06,
      "loss": 0.5906,
      "step": 759
    },
    {
      "epoch": 1.65,
      "grad_norm": 5.403850980249393,
      "learning_rate": 4.444750699866047e-06,
      "loss": 0.8193,
      "step": 760
    },
    {
      "epoch": 1.65,
      "grad_norm": 9.275967904210173,
      "learning_rate": 4.433111065728992e-06,
      "loss": 0.6853,
      "step": 761
    },
    {
      "epoch": 1.65,
      "grad_norm": 3.4792714696266387,
      "learning_rate": 4.4214745428781946e-06,
      "loss": 0.5158,
      "step": 762
    },
    {
      "epoch": 1.65,
      "grad_norm": 3.757595573689217,
      "learning_rate": 4.409841195178991e-06,
      "loss": 0.8048,
      "step": 763
    },
    {
      "epoch": 1.66,
      "grad_norm": 4.537297139241674,
      "learning_rate": 4.3982110864792956e-06,
      "loss": 0.5375,
      "step": 764
    },
    {
      "epoch": 1.66,
      "grad_norm": 3.981207693136639,
      "learning_rate": 4.386584280609242e-06,
      "loss": 0.9523,
      "step": 765
    },
    {
      "epoch": 1.66,
      "grad_norm": 3.6323230981396373,
      "learning_rate": 4.37496084138084e-06,
      "loss": 1.1545,
      "step": 766
    },
    {
      "epoch": 1.66,
      "grad_norm": 3.2913471490659716,
      "learning_rate": 4.363340832587621e-06,
      "loss": 1.1955,
      "step": 767
    },
    {
      "epoch": 1.66,
      "grad_norm": 4.236475031441623,
      "learning_rate": 4.351724318004286e-06,
      "loss": 0.8675,
      "step": 768
    },
    {
      "epoch": 1.67,
      "grad_norm": 3.4983772075862323,
      "learning_rate": 4.340111361386361e-06,
      "loss": 0.8009,
      "step": 769
    },
    {
      "epoch": 1.67,
      "grad_norm": 3.292318693894747,
      "learning_rate": 4.328502026469849e-06,
      "loss": 0.8934,
      "step": 770
    },
    {
      "epoch": 1.67,
      "grad_norm": 3.4845101924924564,
      "learning_rate": 4.316896376970866e-06,
      "loss": 0.9893,
      "step": 771
    },
    {
      "epoch": 1.67,
      "grad_norm": 7.380461100126218,
      "learning_rate": 4.305294476585312e-06,
      "loss": 0.8119,
      "step": 772
    },
    {
      "epoch": 1.67,
      "grad_norm": 3.368804382836309,
      "learning_rate": 4.293696388988498e-06,
      "loss": 0.6158,
      "step": 773
    },
    {
      "epoch": 1.68,
      "grad_norm": 7.120007802273314,
      "learning_rate": 4.282102177834822e-06,
      "loss": 0.6013,
      "step": 774
    },
    {
      "epoch": 1.68,
      "grad_norm": 3.8660709943595273,
      "learning_rate": 4.2705119067574006e-06,
      "loss": 0.8132,
      "step": 775
    },
    {
      "epoch": 1.68,
      "grad_norm": 4.547279992795479,
      "learning_rate": 4.258925639367723e-06,
      "loss": 0.6929,
      "step": 776
    },
    {
      "epoch": 1.68,
      "grad_norm": 3.588721538636798,
      "learning_rate": 4.2473434392553115e-06,
      "loss": 0.6578,
      "step": 777
    },
    {
      "epoch": 1.69,
      "grad_norm": 4.054166652252499,
      "learning_rate": 4.235765369987358e-06,
      "loss": 0.7853,
      "step": 778
    },
    {
      "epoch": 1.69,
      "grad_norm": 3.4196806828943505,
      "learning_rate": 4.224191495108391e-06,
      "loss": 0.737,
      "step": 779
    },
    {
      "epoch": 1.69,
      "grad_norm": 4.0922277734657895,
      "learning_rate": 4.212621878139912e-06,
      "loss": 0.5667,
      "step": 780
    },
    {
      "epoch": 1.69,
      "grad_norm": 3.8095466792086126,
      "learning_rate": 4.201056582580059e-06,
      "loss": 0.6303,
      "step": 781
    },
    {
      "epoch": 1.69,
      "grad_norm": 3.18086598969423,
      "learning_rate": 4.189495671903246e-06,
      "loss": 0.652,
      "step": 782
    },
    {
      "epoch": 1.7,
      "grad_norm": 2.551781196269752,
      "learning_rate": 4.177939209559828e-06,
      "loss": 0.4617,
      "step": 783
    },
    {
      "epoch": 1.7,
      "grad_norm": 3.812000989565637,
      "learning_rate": 4.1663872589757445e-06,
      "loss": 0.7181,
      "step": 784
    },
    {
      "epoch": 1.7,
      "grad_norm": 3.5749453407426754,
      "learning_rate": 4.154839883552169e-06,
      "loss": 0.6589,
      "step": 785
    },
    {
      "epoch": 1.7,
      "grad_norm": 3.0896004504679513,
      "learning_rate": 4.143297146665167e-06,
      "loss": 0.6696,
      "step": 786
    },
    {
      "epoch": 1.71,
      "grad_norm": 3.4340907306165236,
      "learning_rate": 4.131759111665349e-06,
      "loss": 0.7689,
      "step": 787
    },
    {
      "epoch": 1.71,
      "grad_norm": 2.666118112935884,
      "learning_rate": 4.120225841877515e-06,
      "loss": 0.4608,
      "step": 788
    },
    {
      "epoch": 1.71,
      "grad_norm": 4.5072197764582675,
      "learning_rate": 4.108697400600316e-06,
      "loss": 0.7534,
      "step": 789
    },
    {
      "epoch": 1.71,
      "grad_norm": 3.511509633699172,
      "learning_rate": 4.0971738511059e-06,
      "loss": 1.3334,
      "step": 790
    },
    {
      "epoch": 1.71,
      "grad_norm": 4.929478043534673,
      "learning_rate": 4.085655256639565e-06,
      "loss": 0.8954,
      "step": 791
    },
    {
      "epoch": 1.72,
      "grad_norm": 3.518543261966062,
      "learning_rate": 4.074141680419422e-06,
      "loss": 0.5551,
      "step": 792
    },
    {
      "epoch": 1.72,
      "grad_norm": 3.6310867582410298,
      "learning_rate": 4.062633185636031e-06,
      "loss": 0.5767,
      "step": 793
    },
    {
      "epoch": 1.72,
      "grad_norm": 3.7125482879008938,
      "learning_rate": 4.051129835452071e-06,
      "loss": 0.4141,
      "step": 794
    },
    {
      "epoch": 1.72,
      "grad_norm": 4.44966422686601,
      "learning_rate": 4.039631693001976e-06,
      "loss": 0.8329,
      "step": 795
    },
    {
      "epoch": 1.72,
      "grad_norm": 3.5906673919049545,
      "learning_rate": 4.028138821391611e-06,
      "loss": 0.939,
      "step": 796
    },
    {
      "epoch": 1.73,
      "grad_norm": 4.868486296437826,
      "learning_rate": 4.016651283697901e-06,
      "loss": 0.9175,
      "step": 797
    },
    {
      "epoch": 1.73,
      "grad_norm": 3.9797569136130453,
      "learning_rate": 4.005169142968503e-06,
      "loss": 0.7102,
      "step": 798
    },
    {
      "epoch": 1.73,
      "grad_norm": 5.3132735917584055,
      "learning_rate": 3.99369246222145e-06,
      "loss": 0.9272,
      "step": 799
    },
    {
      "epoch": 1.73,
      "grad_norm": 3.689024991339006,
      "learning_rate": 3.982221304444813e-06,
      "loss": 1.122,
      "step": 800
    },
    {
      "epoch": 1.74,
      "grad_norm": 3.189839935211865,
      "learning_rate": 3.970755732596349e-06,
      "loss": 0.4422,
      "step": 801
    },
    {
      "epoch": 1.74,
      "grad_norm": 5.769722968566529,
      "learning_rate": 3.959295809603155e-06,
      "loss": 0.8105,
      "step": 802
    },
    {
      "epoch": 1.74,
      "grad_norm": 5.553634743316624,
      "learning_rate": 3.947841598361329e-06,
      "loss": 0.5204,
      "step": 803
    },
    {
      "epoch": 1.74,
      "grad_norm": 4.1210930560206345,
      "learning_rate": 3.936393161735616e-06,
      "loss": 0.5955,
      "step": 804
    },
    {
      "epoch": 1.74,
      "grad_norm": 5.3814012948576995,
      "learning_rate": 3.924950562559074e-06,
      "loss": 0.7806,
      "step": 805
    },
    {
      "epoch": 1.75,
      "grad_norm": 5.53486596810532,
      "learning_rate": 3.91351386363272e-06,
      "loss": 0.647,
      "step": 806
    },
    {
      "epoch": 1.75,
      "grad_norm": 3.253424021566948,
      "learning_rate": 3.902083127725186e-06,
      "loss": 0.941,
      "step": 807
    },
    {
      "epoch": 1.75,
      "grad_norm": 4.1606421405576866,
      "learning_rate": 3.890658417572379e-06,
      "loss": 0.7955,
      "step": 808
    },
    {
      "epoch": 1.75,
      "grad_norm": 5.374397002104239,
      "learning_rate": 3.879239795877139e-06,
      "loss": 0.675,
      "step": 809
    },
    {
      "epoch": 1.76,
      "grad_norm": 3.1427336826137084,
      "learning_rate": 3.867827325308882e-06,
      "loss": 0.4083,
      "step": 810
    },
    {
      "epoch": 1.76,
      "grad_norm": 10.11396066678661,
      "learning_rate": 3.8564210685032695e-06,
      "loss": 0.8555,
      "step": 811
    },
    {
      "epoch": 1.76,
      "grad_norm": 4.515795368719006,
      "learning_rate": 3.845021088061858e-06,
      "loss": 0.9614,
      "step": 812
    },
    {
      "epoch": 1.76,
      "grad_norm": 6.306834158852401,
      "learning_rate": 3.83362744655176e-06,
      "loss": 0.7661,
      "step": 813
    },
    {
      "epoch": 1.76,
      "grad_norm": 4.794094388964206,
      "learning_rate": 3.822240206505293e-06,
      "loss": 0.9652,
      "step": 814
    },
    {
      "epoch": 1.77,
      "grad_norm": 2.872331494679321,
      "learning_rate": 3.810859430419646e-06,
      "loss": 0.3409,
      "step": 815
    },
    {
      "epoch": 1.77,
      "grad_norm": 4.329510973826048,
      "learning_rate": 3.799485180756526e-06,
      "loss": 0.8902,
      "step": 816
    },
    {
      "epoch": 1.77,
      "grad_norm": 5.082847057607486,
      "learning_rate": 3.788117519941825e-06,
      "loss": 0.7542,
      "step": 817
    },
    {
      "epoch": 1.77,
      "grad_norm": 3.061121717249633,
      "learning_rate": 3.776756510365275e-06,
      "loss": 0.9546,
      "step": 818
    },
    {
      "epoch": 1.77,
      "grad_norm": 4.151491106179622,
      "learning_rate": 3.765402214380095e-06,
      "loss": 0.7729,
      "step": 819
    },
    {
      "epoch": 1.78,
      "grad_norm": 3.685369999134383,
      "learning_rate": 3.7540546943026677e-06,
      "loss": 0.7164,
      "step": 820
    },
    {
      "epoch": 1.78,
      "grad_norm": 4.101567012278009,
      "learning_rate": 3.7427140124121774e-06,
      "loss": 0.6588,
      "step": 821
    },
    {
      "epoch": 1.78,
      "grad_norm": 2.9497252224567743,
      "learning_rate": 3.731380230950288e-06,
      "loss": 0.6358,
      "step": 822
    },
    {
      "epoch": 1.78,
      "grad_norm": 12.568225027937032,
      "learning_rate": 3.720053412120784e-06,
      "loss": 0.8321,
      "step": 823
    },
    {
      "epoch": 1.79,
      "grad_norm": 3.759246624017155,
      "learning_rate": 3.7087336180892395e-06,
      "loss": 0.7644,
      "step": 824
    },
    {
      "epoch": 1.79,
      "grad_norm": 3.62345304109213,
      "learning_rate": 3.6974209109826724e-06,
      "loss": 0.6483,
      "step": 825
    },
    {
      "epoch": 1.79,
      "grad_norm": 8.52075171643868,
      "learning_rate": 3.686115352889209e-06,
      "loss": 0.7505,
      "step": 826
    },
    {
      "epoch": 1.79,
      "grad_norm": 3.2323516380477835,
      "learning_rate": 3.674817005857735e-06,
      "loss": 0.6018,
      "step": 827
    },
    {
      "epoch": 1.79,
      "grad_norm": 4.211372771264943,
      "learning_rate": 3.663525931897559e-06,
      "loss": 1.0932,
      "step": 828
    },
    {
      "epoch": 1.8,
      "grad_norm": 4.8878414004606405,
      "learning_rate": 3.6522421929780746e-06,
      "loss": 0.8801,
      "step": 829
    },
    {
      "epoch": 1.8,
      "grad_norm": 3.8082900685871035,
      "learning_rate": 3.6409658510284208e-06,
      "loss": 0.6203,
      "step": 830
    },
    {
      "epoch": 1.8,
      "grad_norm": 3.6495767535006958,
      "learning_rate": 3.6296969679371325e-06,
      "loss": 0.7162,
      "step": 831
    },
    {
      "epoch": 1.8,
      "grad_norm": 3.5310708308115455,
      "learning_rate": 3.6184356055518143e-06,
      "loss": 0.5489,
      "step": 832
    },
    {
      "epoch": 1.8,
      "grad_norm": 4.519488845602426,
      "learning_rate": 3.6071818256787906e-06,
      "loss": 0.8231,
      "step": 833
    },
    {
      "epoch": 1.81,
      "grad_norm": 3.680805542278986,
      "learning_rate": 3.595935690082769e-06,
      "loss": 0.5912,
      "step": 834
    },
    {
      "epoch": 1.81,
      "grad_norm": 4.8389558092932115,
      "learning_rate": 3.5846972604865103e-06,
      "loss": 0.6906,
      "step": 835
    },
    {
      "epoch": 1.81,
      "grad_norm": 4.654908465364193,
      "learning_rate": 3.5734665985704732e-06,
      "loss": 0.7744,
      "step": 836
    },
    {
      "epoch": 1.81,
      "grad_norm": 5.334453080170469,
      "learning_rate": 3.56224376597249e-06,
      "loss": 0.8189,
      "step": 837
    },
    {
      "epoch": 1.82,
      "grad_norm": 3.6600207975950636,
      "learning_rate": 3.551028824287418e-06,
      "loss": 0.6004,
      "step": 838
    },
    {
      "epoch": 1.82,
      "grad_norm": 4.258554408115963,
      "learning_rate": 3.5398218350668136e-06,
      "loss": 0.8109,
      "step": 839
    },
    {
      "epoch": 1.82,
      "grad_norm": 6.693380933513454,
      "learning_rate": 3.528622859818582e-06,
      "loss": 0.7812,
      "step": 840
    },
    {
      "epoch": 1.82,
      "grad_norm": 5.539828940170528,
      "learning_rate": 3.517431960006645e-06,
      "loss": 0.9981,
      "step": 841
    },
    {
      "epoch": 1.82,
      "grad_norm": 3.1560155284777145,
      "learning_rate": 3.506249197050604e-06,
      "loss": 0.6148,
      "step": 842
    },
    {
      "epoch": 1.83,
      "grad_norm": 4.067935719607778,
      "learning_rate": 3.495074632325407e-06,
      "loss": 0.6073,
      "step": 843
    },
    {
      "epoch": 1.83,
      "grad_norm": 5.3924145674828114,
      "learning_rate": 3.4839083271610007e-06,
      "loss": 0.7521,
      "step": 844
    },
    {
      "epoch": 1.83,
      "grad_norm": 4.098029673550654,
      "learning_rate": 3.472750342842003e-06,
      "loss": 0.8263,
      "step": 845
    },
    {
      "epoch": 1.83,
      "grad_norm": 4.581742837012406,
      "learning_rate": 3.461600740607366e-06,
      "loss": 1.0475,
      "step": 846
    },
    {
      "epoch": 1.84,
      "grad_norm": 3.2624616055869273,
      "learning_rate": 3.4504595816500318e-06,
      "loss": 0.5945,
      "step": 847
    },
    {
      "epoch": 1.84,
      "grad_norm": 3.882864173941595,
      "learning_rate": 3.4393269271166117e-06,
      "loss": 1.2057,
      "step": 848
    },
    {
      "epoch": 1.84,
      "grad_norm": 2.6098567153049883,
      "learning_rate": 3.4282028381070366e-06,
      "loss": 0.4498,
      "step": 849
    },
    {
      "epoch": 1.84,
      "grad_norm": 3.8898265178629163,
      "learning_rate": 3.4170873756742263e-06,
      "loss": 0.9574,
      "step": 850
    },
    {
      "epoch": 1.84,
      "grad_norm": 3.2575391506729066,
      "learning_rate": 3.405980600823754e-06,
      "loss": 0.6424,
      "step": 851
    },
    {
      "epoch": 1.85,
      "grad_norm": 3.904966626495903,
      "learning_rate": 3.3948825745135196e-06,
      "loss": 0.9632,
      "step": 852
    },
    {
      "epoch": 1.85,
      "grad_norm": 3.1680824181919536,
      "learning_rate": 3.383793357653398e-06,
      "loss": 0.5572,
      "step": 853
    },
    {
      "epoch": 1.85,
      "grad_norm": 3.2729061138403215,
      "learning_rate": 3.372713011104922e-06,
      "loss": 0.494,
      "step": 854
    },
    {
      "epoch": 1.85,
      "grad_norm": 3.469414094962235,
      "learning_rate": 3.361641595680937e-06,
      "loss": 0.5889,
      "step": 855
    },
    {
      "epoch": 1.85,
      "grad_norm": 4.940486687229903,
      "learning_rate": 3.350579172145273e-06,
      "loss": 1.0901,
      "step": 856
    },
    {
      "epoch": 1.86,
      "grad_norm": 4.292363357859409,
      "learning_rate": 3.3395258012124103e-06,
      "loss": 0.8299,
      "step": 857
    },
    {
      "epoch": 1.86,
      "grad_norm": 3.1533992835877807,
      "learning_rate": 3.3284815435471423e-06,
      "loss": 0.8111,
      "step": 858
    },
    {
      "epoch": 1.86,
      "grad_norm": 3.8814189330333293,
      "learning_rate": 3.3174464597642497e-06,
      "loss": 0.7773,
      "step": 859
    },
    {
      "epoch": 1.86,
      "grad_norm": 3.7381050483686904,
      "learning_rate": 3.306420610428157e-06,
      "loss": 0.7228,
      "step": 860
    },
    {
      "epoch": 1.87,
      "grad_norm": 3.8952439069588345,
      "learning_rate": 3.295404056052616e-06,
      "loss": 0.6839,
      "step": 861
    },
    {
      "epoch": 1.87,
      "grad_norm": 5.693583999884172,
      "learning_rate": 3.284396857100357e-06,
      "loss": 1.0729,
      "step": 862
    },
    {
      "epoch": 1.87,
      "grad_norm": 3.4565350102390147,
      "learning_rate": 3.273399073982768e-06,
      "loss": 0.5441,
      "step": 863
    },
    {
      "epoch": 1.87,
      "grad_norm": 3.52468061427538,
      "learning_rate": 3.2624107670595567e-06,
      "loss": 0.8231,
      "step": 864
    },
    {
      "epoch": 1.87,
      "grad_norm": 4.332974061838427,
      "learning_rate": 3.251431996638427e-06,
      "loss": 0.7278,
      "step": 865
    },
    {
      "epoch": 1.88,
      "grad_norm": 4.834362669335944,
      "learning_rate": 3.2404628229747386e-06,
      "loss": 1.2918,
      "step": 866
    },
    {
      "epoch": 1.88,
      "grad_norm": 4.080335073868095,
      "learning_rate": 3.2295033062711823e-06,
      "loss": 0.7472,
      "step": 867
    },
    {
      "epoch": 1.88,
      "grad_norm": 2.6096209194849838,
      "learning_rate": 3.2185535066774477e-06,
      "loss": 0.3796,
      "step": 868
    },
    {
      "epoch": 1.88,
      "grad_norm": 3.517487407873049,
      "learning_rate": 3.2076134842898955e-06,
      "loss": 0.5771,
      "step": 869
    },
    {
      "epoch": 1.89,
      "grad_norm": 3.426730924330036,
      "learning_rate": 3.1966832991512232e-06,
      "loss": 0.8781,
      "step": 870
    },
    {
      "epoch": 1.89,
      "grad_norm": 10.664511888686405,
      "learning_rate": 3.1857630112501397e-06,
      "loss": 0.7429,
      "step": 871
    },
    {
      "epoch": 1.89,
      "grad_norm": 3.4963520697702903,
      "learning_rate": 3.174852680521032e-06,
      "loss": 0.5781,
      "step": 872
    },
    {
      "epoch": 1.89,
      "grad_norm": 4.240686229058446,
      "learning_rate": 3.16395236684364e-06,
      "loss": 0.9904,
      "step": 873
    },
    {
      "epoch": 1.89,
      "grad_norm": 3.7385937879676963,
      "learning_rate": 3.1530621300427294e-06,
      "loss": 0.6248,
      "step": 874
    },
    {
      "epoch": 1.9,
      "grad_norm": 3.2563955445344686,
      "learning_rate": 3.1421820298877554e-06,
      "loss": 0.6759,
      "step": 875
    },
    {
      "epoch": 1.9,
      "grad_norm": 3.6600694716292286,
      "learning_rate": 3.131312126092544e-06,
      "loss": 0.5502,
      "step": 876
    },
    {
      "epoch": 1.9,
      "grad_norm": 7.2538600057634035,
      "learning_rate": 3.1204524783149546e-06,
      "loss": 0.7111,
      "step": 877
    },
    {
      "epoch": 1.9,
      "grad_norm": 3.4777684384331216,
      "learning_rate": 3.1096031461565656e-06,
      "loss": 0.5092,
      "step": 878
    },
    {
      "epoch": 1.9,
      "grad_norm": 2.8812735478965306,
      "learning_rate": 3.098764189162332e-06,
      "loss": 0.7524,
      "step": 879
    },
    {
      "epoch": 1.91,
      "grad_norm": 5.364182283202637,
      "learning_rate": 3.087935666820273e-06,
      "loss": 0.831,
      "step": 880
    },
    {
      "epoch": 1.91,
      "grad_norm": 3.3953503297462335,
      "learning_rate": 3.0771176385611318e-06,
      "loss": 0.9355,
      "step": 881
    },
    {
      "epoch": 1.91,
      "grad_norm": 7.482988017394584,
      "learning_rate": 3.0663101637580626e-06,
      "loss": 1.0444,
      "step": 882
    },
    {
      "epoch": 1.91,
      "grad_norm": 24.07170157003501,
      "learning_rate": 3.055513301726296e-06,
      "loss": 1.1011,
      "step": 883
    },
    {
      "epoch": 1.92,
      "grad_norm": 3.4316528976323317,
      "learning_rate": 3.044727111722815e-06,
      "loss": 0.6053,
      "step": 884
    },
    {
      "epoch": 1.92,
      "grad_norm": 3.38557833515691,
      "learning_rate": 3.03395165294603e-06,
      "loss": 0.5275,
      "step": 885
    },
    {
      "epoch": 1.92,
      "grad_norm": 3.4763149274770804,
      "learning_rate": 3.02318698453546e-06,
      "loss": 0.6878,
      "step": 886
    },
    {
      "epoch": 1.92,
      "grad_norm": 3.674740882218427,
      "learning_rate": 3.0124331655713966e-06,
      "loss": 0.5598,
      "step": 887
    },
    {
      "epoch": 1.92,
      "grad_norm": 2.793725780843057,
      "learning_rate": 3.0016902550745896e-06,
      "loss": 0.6041,
      "step": 888
    },
    {
      "epoch": 1.93,
      "grad_norm": 6.120276736431503,
      "learning_rate": 2.990958312005916e-06,
      "loss": 1.1304,
      "step": 889
    },
    {
      "epoch": 1.93,
      "grad_norm": 3.989489627388308,
      "learning_rate": 2.980237395266061e-06,
      "loss": 0.7831,
      "step": 890
    },
    {
      "epoch": 1.93,
      "grad_norm": 4.144666659943107,
      "learning_rate": 2.9695275636951983e-06,
      "loss": 0.7328,
      "step": 891
    },
    {
      "epoch": 1.93,
      "grad_norm": 4.27318475572958,
      "learning_rate": 2.958828876072654e-06,
      "loss": 0.7659,
      "step": 892
    },
    {
      "epoch": 1.93,
      "grad_norm": 2.6610556959682508,
      "learning_rate": 2.9481413911165984e-06,
      "loss": 0.5647,
      "step": 893
    },
    {
      "epoch": 1.94,
      "grad_norm": 4.380486687678224,
      "learning_rate": 2.9374651674837128e-06,
      "loss": 0.7599,
      "step": 894
    },
    {
      "epoch": 1.94,
      "grad_norm": 5.172560737903105,
      "learning_rate": 2.9268002637688788e-06,
      "loss": 0.7527,
      "step": 895
    },
    {
      "epoch": 1.94,
      "grad_norm": 3.5581352925036547,
      "learning_rate": 2.9161467385048425e-06,
      "loss": 0.6063,
      "step": 896
    },
    {
      "epoch": 1.94,
      "grad_norm": 3.267990667488901,
      "learning_rate": 2.9055046501619088e-06,
      "loss": 0.651,
      "step": 897
    },
    {
      "epoch": 1.95,
      "grad_norm": 3.6831917996386423,
      "learning_rate": 2.894874057147606e-06,
      "loss": 1.0493,
      "step": 898
    },
    {
      "epoch": 1.95,
      "grad_norm": 3.3699298807546394,
      "learning_rate": 2.8842550178063777e-06,
      "loss": 1.0716,
      "step": 899
    },
    {
      "epoch": 1.95,
      "grad_norm": 3.1048940500456603,
      "learning_rate": 2.8736475904192516e-06,
      "loss": 0.6683,
      "step": 900
    },
    {
      "epoch": 1.95,
      "grad_norm": 3.781736611900924,
      "learning_rate": 2.863051833203531e-06,
      "loss": 0.7018,
      "step": 901
    },
    {
      "epoch": 1.95,
      "grad_norm": 3.440534255835195,
      "learning_rate": 2.852467804312463e-06,
      "loss": 0.8496,
      "step": 902
    },
    {
      "epoch": 1.96,
      "grad_norm": 3.8786677230510858,
      "learning_rate": 2.841895561834927e-06,
      "loss": 0.8237,
      "step": 903
    },
    {
      "epoch": 1.96,
      "grad_norm": 7.09470769395165,
      "learning_rate": 2.8313351637951196e-06,
      "loss": 0.4336,
      "step": 904
    },
    {
      "epoch": 1.96,
      "grad_norm": 3.6170075361977427,
      "learning_rate": 2.8207866681522233e-06,
      "loss": 0.5937,
      "step": 905
    },
    {
      "epoch": 1.96,
      "grad_norm": 3.3725634329628402,
      "learning_rate": 2.810250132800103e-06,
      "loss": 0.6167,
      "step": 906
    },
    {
      "epoch": 1.97,
      "grad_norm": 4.603029972519325,
      "learning_rate": 2.7997256155669737e-06,
      "loss": 1.0976,
      "step": 907
    },
    {
      "epoch": 1.97,
      "grad_norm": 3.2563443697114023,
      "learning_rate": 2.7892131742151007e-06,
      "loss": 0.9893,
      "step": 908
    },
    {
      "epoch": 1.97,
      "grad_norm": 3.3478433109908567,
      "learning_rate": 2.778712866440464e-06,
      "loss": 1.0481,
      "step": 909
    },
    {
      "epoch": 1.97,
      "grad_norm": 3.8754323715796746,
      "learning_rate": 2.7682247498724536e-06,
      "loss": 0.6663,
      "step": 910
    },
    {
      "epoch": 1.97,
      "grad_norm": 3.728280220198159,
      "learning_rate": 2.7577488820735465e-06,
      "loss": 0.7614,
      "step": 911
    },
    {
      "epoch": 1.98,
      "grad_norm": 3.24699808100551,
      "learning_rate": 2.7472853205389997e-06,
      "loss": 0.737,
      "step": 912
    },
    {
      "epoch": 1.98,
      "grad_norm": 3.3445131281866023,
      "learning_rate": 2.736834122696529e-06,
      "loss": 0.9125,
      "step": 913
    },
    {
      "epoch": 1.98,
      "grad_norm": 5.574569643603356,
      "learning_rate": 2.7263953459059888e-06,
      "loss": 0.8048,
      "step": 914
    },
    {
      "epoch": 1.98,
      "grad_norm": 4.522234392161867,
      "learning_rate": 2.715969047459066e-06,
      "loss": 0.9942,
      "step": 915
    },
    {
      "epoch": 1.98,
      "grad_norm": 3.1427723242778804,
      "learning_rate": 2.705555284578958e-06,
      "loss": 0.716,
      "step": 916
    },
    {
      "epoch": 1.99,
      "grad_norm": 4.050379978712173,
      "learning_rate": 2.6951541144200676e-06,
      "loss": 0.8721,
      "step": 917
    },
    {
      "epoch": 1.99,
      "grad_norm": 3.1242821454387326,
      "learning_rate": 2.6847655940676843e-06,
      "loss": 1.1733,
      "step": 918
    },
    {
      "epoch": 1.99,
      "grad_norm": 3.4250876504642624,
      "learning_rate": 2.6743897805376672e-06,
      "loss": 0.7195,
      "step": 919
    },
    {
      "epoch": 1.99,
      "grad_norm": 3.7217477129056338,
      "learning_rate": 2.664026730776136e-06,
      "loss": 0.8821,
      "step": 920
    },
    {
      "epoch": 2.0,
      "grad_norm": 5.523733880102304,
      "learning_rate": 2.6536765016591626e-06,
      "loss": 0.9302,
      "step": 921
    },
    {
      "epoch": 2.0,
      "grad_norm": 16.94363299298687,
      "learning_rate": 2.64333914999245e-06,
      "loss": 0.8664,
      "step": 922
    },
    {
      "epoch": 2.0,
      "grad_norm": 3.953063093268987,
      "learning_rate": 2.63301473251103e-06,
      "loss": 1.006,
      "step": 923
    },
    {
      "epoch": 2.0,
      "grad_norm": 7.871036341712749,
      "learning_rate": 2.622703305878941e-06,
      "loss": 0.9341,
      "step": 924
    },
    {
      "epoch": 2.0,
      "grad_norm": 3.919307324227037,
      "learning_rate": 2.6124049266889296e-06,
      "loss": 0.6258,
      "step": 925
    },
    {
      "epoch": 2.01,
      "grad_norm": 4.243198650022596,
      "learning_rate": 2.6021196514621283e-06,
      "loss": 0.4821,
      "step": 926
    },
    {
      "epoch": 2.01,
      "grad_norm": 6.607896692734867,
      "learning_rate": 2.5918475366477536e-06,
      "loss": 0.8829,
      "step": 927
    },
    {
      "epoch": 2.01,
      "grad_norm": 3.3957216771032,
      "learning_rate": 2.5815886386227882e-06,
      "loss": 0.7431,
      "step": 928
    },
    {
      "epoch": 2.01,
      "grad_norm": 3.5228994644419007,
      "learning_rate": 2.5713430136916828e-06,
      "loss": 1.0325,
      "step": 929
    },
    {
      "epoch": 2.02,
      "grad_norm": 4.213582019224428,
      "learning_rate": 2.5611107180860395e-06,
      "loss": 1.1798,
      "step": 930
    },
    {
      "epoch": 2.02,
      "grad_norm": 3.5071868277546927,
      "learning_rate": 2.5508918079643e-06,
      "loss": 0.6067,
      "step": 931
    },
    {
      "epoch": 2.02,
      "grad_norm": 3.7222323750058215,
      "learning_rate": 2.540686339411446e-06,
      "loss": 0.8965,
      "step": 932
    },
    {
      "epoch": 2.02,
      "grad_norm": 4.124876309815822,
      "learning_rate": 2.530494368438683e-06,
      "loss": 0.5735,
      "step": 933
    },
    {
      "epoch": 2.02,
      "grad_norm": 5.0735620158038985,
      "learning_rate": 2.520315950983141e-06,
      "loss": 0.9851,
      "step": 934
    },
    {
      "epoch": 2.03,
      "grad_norm": 2.890135544902828,
      "learning_rate": 2.5101511429075654e-06,
      "loss": 0.5405,
      "step": 935
    },
    {
      "epoch": 2.03,
      "grad_norm": 3.9999528539240594,
      "learning_rate": 2.5000000000000015e-06,
      "loss": 0.7653,
      "step": 936
    },
    {
      "epoch": 2.03,
      "grad_norm": 3.0256756135079503,
      "learning_rate": 2.489862577973498e-06,
      "loss": 0.4422,
      "step": 937
    },
    {
      "epoch": 2.03,
      "grad_norm": 2.8720942464825185,
      "learning_rate": 2.4797389324658037e-06,
      "loss": 0.8503,
      "step": 938
    },
    {
      "epoch": 2.03,
      "grad_norm": 3.7514143847372616,
      "learning_rate": 2.4696291190390494e-06,
      "loss": 0.8386,
      "step": 939
    },
    {
      "epoch": 2.04,
      "grad_norm": 3.7574921460015793,
      "learning_rate": 2.459533193179457e-06,
      "loss": 1.057,
      "step": 940
    },
    {
      "epoch": 2.04,
      "grad_norm": 4.858917782664826,
      "learning_rate": 2.4494512102970247e-06,
      "loss": 0.7795,
      "step": 941
    },
    {
      "epoch": 2.04,
      "grad_norm": 5.900070950856582,
      "learning_rate": 2.4393832257252253e-06,
      "loss": 1.1636,
      "step": 942
    },
    {
      "epoch": 2.04,
      "grad_norm": 3.4182890619855524,
      "learning_rate": 2.42932929472071e-06,
      "loss": 0.7024,
      "step": 943
    },
    {
      "epoch": 2.05,
      "grad_norm": 2.6272711502194275,
      "learning_rate": 2.4192894724629943e-06,
      "loss": 0.5219,
      "step": 944
    },
    {
      "epoch": 2.05,
      "grad_norm": 3.2584081721395615,
      "learning_rate": 2.4092638140541586e-06,
      "loss": 0.7251,
      "step": 945
    },
    {
      "epoch": 2.05,
      "grad_norm": 3.3671627955726313,
      "learning_rate": 2.399252374518551e-06,
      "loss": 1.0014,
      "step": 946
    },
    {
      "epoch": 2.05,
      "grad_norm": 4.810795614836694,
      "learning_rate": 2.3892552088024827e-06,
      "loss": 0.4545,
      "step": 947
    },
    {
      "epoch": 2.05,
      "grad_norm": 8.673623733437953,
      "learning_rate": 2.3792723717739197e-06,
      "loss": 0.88,
      "step": 948
    },
    {
      "epoch": 2.06,
      "grad_norm": 4.365382226431478,
      "learning_rate": 2.3693039182221907e-06,
      "loss": 0.6554,
      "step": 949
    },
    {
      "epoch": 2.06,
      "grad_norm": 2.8149835592552086,
      "learning_rate": 2.3593499028576793e-06,
      "loss": 0.4599,
      "step": 950
    },
    {
      "epoch": 2.06,
      "grad_norm": 3.5192624846846763,
      "learning_rate": 2.349410380311532e-06,
      "loss": 0.7618,
      "step": 951
    },
    {
      "epoch": 2.06,
      "grad_norm": 3.1595511543307264,
      "learning_rate": 2.3394854051353534e-06,
      "loss": 0.7603,
      "step": 952
    },
    {
      "epoch": 2.07,
      "grad_norm": 4.659661635705528,
      "learning_rate": 2.329575031800903e-06,
      "loss": 1.1096,
      "step": 953
    },
    {
      "epoch": 2.07,
      "grad_norm": 4.715158995230258,
      "learning_rate": 2.319679314699801e-06,
      "loss": 0.8287,
      "step": 954
    },
    {
      "epoch": 2.07,
      "grad_norm": 4.013905706270423,
      "learning_rate": 2.3097983081432334e-06,
      "loss": 0.9549,
      "step": 955
    },
    {
      "epoch": 2.07,
      "grad_norm": 7.846039846655144,
      "learning_rate": 2.299932066361643e-06,
      "loss": 0.8468,
      "step": 956
    },
    {
      "epoch": 2.07,
      "grad_norm": 4.924257238623335,
      "learning_rate": 2.290080643504446e-06,
      "loss": 0.8954,
      "step": 957
    },
    {
      "epoch": 2.08,
      "grad_norm": 3.436664408871294,
      "learning_rate": 2.2802440936397203e-06,
      "loss": 0.9313,
      "step": 958
    },
    {
      "epoch": 2.08,
      "grad_norm": 6.079806645842031,
      "learning_rate": 2.2704224707539164e-06,
      "loss": 1.0962,
      "step": 959
    },
    {
      "epoch": 2.08,
      "grad_norm": 3.7675731638306615,
      "learning_rate": 2.2606158287515662e-06,
      "loss": 0.8263,
      "step": 960
    },
    {
      "epoch": 2.08,
      "grad_norm": 3.6487997998290176,
      "learning_rate": 2.250824221454976e-06,
      "loss": 0.5801,
      "step": 961
    },
    {
      "epoch": 2.08,
      "grad_norm": 2.932946857849537,
      "learning_rate": 2.2410477026039335e-06,
      "loss": 0.7822,
      "step": 962
    },
    {
      "epoch": 2.09,
      "grad_norm": 3.507348086445544,
      "learning_rate": 2.2312863258554236e-06,
      "loss": 0.6426,
      "step": 963
    },
    {
      "epoch": 2.09,
      "grad_norm": 3.0329849860510727,
      "learning_rate": 2.221540144783323e-06,
      "loss": 0.7527,
      "step": 964
    },
    {
      "epoch": 2.09,
      "grad_norm": 4.329423058411885,
      "learning_rate": 2.211809212878106e-06,
      "loss": 0.5473,
      "step": 965
    },
    {
      "epoch": 2.09,
      "grad_norm": 4.179974193985937,
      "learning_rate": 2.2020935835465567e-06,
      "loss": 1.4559,
      "step": 966
    },
    {
      "epoch": 2.1,
      "grad_norm": 3.95760142756448,
      "learning_rate": 2.1923933101114713e-06,
      "loss": 0.9247,
      "step": 967
    },
    {
      "epoch": 2.1,
      "grad_norm": 9.798576340687958,
      "learning_rate": 2.182708445811371e-06,
      "loss": 1.0239,
      "step": 968
    },
    {
      "epoch": 2.1,
      "grad_norm": 6.260950317599809,
      "learning_rate": 2.1730390438002056e-06,
      "loss": 0.8968,
      "step": 969
    },
    {
      "epoch": 2.1,
      "grad_norm": 2.599448056163962,
      "learning_rate": 2.1633851571470595e-06,
      "loss": 0.4632,
      "step": 970
    },
    {
      "epoch": 2.1,
      "grad_norm": 3.468840384975282,
      "learning_rate": 2.1537468388358645e-06,
      "loss": 0.4913,
      "step": 971
    },
    {
      "epoch": 2.11,
      "grad_norm": 4.276722054645036,
      "learning_rate": 2.1441241417651072e-06,
      "loss": 0.8904,
      "step": 972
    },
    {
      "epoch": 2.11,
      "grad_norm": 4.254793819495432,
      "learning_rate": 2.134517118747541e-06,
      "loss": 0.5505,
      "step": 973
    },
    {
      "epoch": 2.11,
      "grad_norm": 4.965552073937336,
      "learning_rate": 2.1249258225098974e-06,
      "loss": 0.8492,
      "step": 974
    },
    {
      "epoch": 2.11,
      "grad_norm": 5.911978814751599,
      "learning_rate": 2.1153503056925872e-06,
      "loss": 0.8407,
      "step": 975
    },
    {
      "epoch": 2.11,
      "grad_norm": 4.14946384334899,
      "learning_rate": 2.1057906208494204e-06,
      "loss": 0.7988,
      "step": 976
    },
    {
      "epoch": 2.12,
      "grad_norm": 4.164867881029668,
      "learning_rate": 2.09624682044732e-06,
      "loss": 1.0988,
      "step": 977
    },
    {
      "epoch": 2.12,
      "grad_norm": 2.658068852930561,
      "learning_rate": 2.086718956866024e-06,
      "loss": 0.3764,
      "step": 978
    },
    {
      "epoch": 2.12,
      "grad_norm": 3.7340885241651143,
      "learning_rate": 2.0772070823978034e-06,
      "loss": 0.9143,
      "step": 979
    },
    {
      "epoch": 2.12,
      "grad_norm": 3.1120335582369956,
      "learning_rate": 2.06771124924718e-06,
      "loss": 0.4599,
      "step": 980
    },
    {
      "epoch": 2.13,
      "grad_norm": 3.0870228581822516,
      "learning_rate": 2.0582315095306343e-06,
      "loss": 0.6423,
      "step": 981
    },
    {
      "epoch": 2.13,
      "grad_norm": 15.809260480096588,
      "learning_rate": 2.0487679152763173e-06,
      "loss": 0.8614,
      "step": 982
    },
    {
      "epoch": 2.13,
      "grad_norm": 4.351250464612998,
      "learning_rate": 2.0393205184237687e-06,
      "loss": 0.7455,
      "step": 983
    },
    {
      "epoch": 2.13,
      "grad_norm": 7.3428911053029156,
      "learning_rate": 2.0298893708236307e-06,
      "loss": 1.1378,
      "step": 984
    },
    {
      "epoch": 2.13,
      "grad_norm": 5.240730097513499,
      "learning_rate": 2.0204745242373665e-06,
      "loss": 0.8594,
      "step": 985
    },
    {
      "epoch": 2.14,
      "grad_norm": 3.96916840566228,
      "learning_rate": 2.011076030336974e-06,
      "loss": 1.2131,
      "step": 986
    },
    {
      "epoch": 2.14,
      "grad_norm": 4.025453747280727,
      "learning_rate": 2.0016939407046987e-06,
      "loss": 1.0998,
      "step": 987
    },
    {
      "epoch": 2.14,
      "grad_norm": 5.113978492467684,
      "learning_rate": 1.992328306832755e-06,
      "loss": 0.7903,
      "step": 988
    },
    {
      "epoch": 2.14,
      "grad_norm": 3.696288906346202,
      "learning_rate": 1.9829791801230398e-06,
      "loss": 0.9494,
      "step": 989
    },
    {
      "epoch": 2.15,
      "grad_norm": 3.0821256284898313,
      "learning_rate": 1.9736466118868573e-06,
      "loss": 1.2399,
      "step": 990
    },
    {
      "epoch": 2.15,
      "grad_norm": 2.2184154982856725,
      "learning_rate": 1.9643306533446332e-06,
      "loss": 0.4059,
      "step": 991
    },
    {
      "epoch": 2.15,
      "grad_norm": 3.4414336554393636,
      "learning_rate": 1.9550313556256294e-06,
      "loss": 0.5361,
      "step": 992
    },
    {
      "epoch": 2.15,
      "grad_norm": 4.229002393039922,
      "learning_rate": 1.945748769767667e-06,
      "loss": 0.9436,
      "step": 993
    },
    {
      "epoch": 2.15,
      "grad_norm": 4.127647528620698,
      "learning_rate": 1.9364829467168522e-06,
      "loss": 1.2559,
      "step": 994
    },
    {
      "epoch": 2.16,
      "grad_norm": 4.856341113330408,
      "learning_rate": 1.927233937327285e-06,
      "loss": 0.6056,
      "step": 995
    },
    {
      "epoch": 2.16,
      "grad_norm": 2.9476485058670283,
      "learning_rate": 1.9180017923607884e-06,
      "loss": 1.0021,
      "step": 996
    },
    {
      "epoch": 2.16,
      "grad_norm": 6.209826720555463,
      "learning_rate": 1.9087865624866297e-06,
      "loss": 0.5974,
      "step": 997
    },
    {
      "epoch": 2.16,
      "grad_norm": 4.329974615002508,
      "learning_rate": 1.8995882982812352e-06,
      "loss": 0.8706,
      "step": 998
    },
    {
      "epoch": 2.16,
      "grad_norm": 4.59809899848369,
      "learning_rate": 1.8904070502279242e-06,
      "loss": 0.957,
      "step": 999
    },
    {
      "epoch": 2.17,
      "grad_norm": 3.7836682815944545,
      "learning_rate": 1.8812428687166195e-06,
      "loss": 0.6117,
      "step": 1000
    },
    {
      "epoch": 2.17,
      "grad_norm": 8.527775027230831,
      "learning_rate": 1.8720958040435772e-06,
      "loss": 0.6178,
      "step": 1001
    },
    {
      "epoch": 2.17,
      "grad_norm": 3.6933136458280904,
      "learning_rate": 1.8629659064111138e-06,
      "loss": 1.0519,
      "step": 1002
    },
    {
      "epoch": 2.17,
      "grad_norm": 3.9200237600199053,
      "learning_rate": 1.8538532259273272e-06,
      "loss": 0.534,
      "step": 1003
    },
    {
      "epoch": 2.18,
      "grad_norm": 4.503483019983542,
      "learning_rate": 1.844757812605817e-06,
      "loss": 0.5481,
      "step": 1004
    },
    {
      "epoch": 2.18,
      "grad_norm": 3.3158209383420285,
      "learning_rate": 1.8356797163654172e-06,
      "loss": 0.5771,
      "step": 1005
    },
    {
      "epoch": 2.18,
      "grad_norm": 4.03673203989928,
      "learning_rate": 1.8266189870299184e-06,
      "loss": 0.8207,
      "step": 1006
    },
    {
      "epoch": 2.18,
      "grad_norm": 4.362575812767571,
      "learning_rate": 1.8175756743277967e-06,
      "loss": 0.6848,
      "step": 1007
    },
    {
      "epoch": 2.18,
      "grad_norm": 3.529596802268623,
      "learning_rate": 1.8085498278919421e-06,
      "loss": 0.763,
      "step": 1008
    },
    {
      "epoch": 2.19,
      "grad_norm": 4.202720698552692,
      "learning_rate": 1.7995414972593784e-06,
      "loss": 0.9208,
      "step": 1009
    },
    {
      "epoch": 2.19,
      "grad_norm": 4.4534429324019715,
      "learning_rate": 1.7905507318709997e-06,
      "loss": 1.0568,
      "step": 1010
    },
    {
      "epoch": 2.19,
      "grad_norm": 5.657703962496298,
      "learning_rate": 1.7815775810712921e-06,
      "loss": 0.8386,
      "step": 1011
    },
    {
      "epoch": 2.19,
      "grad_norm": 7.15645645265648,
      "learning_rate": 1.772622094108074e-06,
      "loss": 0.7625,
      "step": 1012
    },
    {
      "epoch": 2.2,
      "grad_norm": 4.313831039213513,
      "learning_rate": 1.7636843201322106e-06,
      "loss": 1.1595,
      "step": 1013
    },
    {
      "epoch": 2.2,
      "grad_norm": 4.0729744730552575,
      "learning_rate": 1.754764308197358e-06,
      "loss": 0.7605,
      "step": 1014
    },
    {
      "epoch": 2.2,
      "grad_norm": 3.7874753718870817,
      "learning_rate": 1.7458621072596827e-06,
      "loss": 0.8222,
      "step": 1015
    },
    {
      "epoch": 2.2,
      "grad_norm": 4.690874094650301,
      "learning_rate": 1.7369777661776032e-06,
      "loss": 0.8995,
      "step": 1016
    },
    {
      "epoch": 2.2,
      "grad_norm": 2.9860114540198484,
      "learning_rate": 1.728111333711514e-06,
      "loss": 0.7115,
      "step": 1017
    },
    {
      "epoch": 2.21,
      "grad_norm": 4.851971909779037,
      "learning_rate": 1.7192628585235188e-06,
      "loss": 0.9985,
      "step": 1018
    },
    {
      "epoch": 2.21,
      "grad_norm": 3.696323069097701,
      "learning_rate": 1.7104323891771697e-06,
      "loss": 0.5773,
      "step": 1019
    },
    {
      "epoch": 2.21,
      "grad_norm": 3.95403964150847,
      "learning_rate": 1.7016199741371958e-06,
      "loss": 0.5239,
      "step": 1020
    },
    {
      "epoch": 2.21,
      "grad_norm": 4.185290949308418,
      "learning_rate": 1.6928256617692357e-06,
      "loss": 0.6401,
      "step": 1021
    },
    {
      "epoch": 2.21,
      "grad_norm": 8.323315625289952,
      "learning_rate": 1.6840495003395741e-06,
      "loss": 0.671,
      "step": 1022
    },
    {
      "epoch": 2.22,
      "grad_norm": 3.6173400100771023,
      "learning_rate": 1.6752915380148772e-06,
      "loss": 1.0549,
      "step": 1023
    },
    {
      "epoch": 2.22,
      "grad_norm": 3.1209639402858014,
      "learning_rate": 1.6665518228619316e-06,
      "loss": 0.7976,
      "step": 1024
    },
    {
      "epoch": 2.22,
      "grad_norm": 4.3112526635176085,
      "learning_rate": 1.6578304028473703e-06,
      "loss": 0.6233,
      "step": 1025
    },
    {
      "epoch": 2.22,
      "grad_norm": 4.277484137748271,
      "learning_rate": 1.6491273258374241e-06,
      "loss": 0.7358,
      "step": 1026
    },
    {
      "epoch": 2.23,
      "grad_norm": 3.4175476694064164,
      "learning_rate": 1.6404426395976446e-06,
      "loss": 0.9737,
      "step": 1027
    },
    {
      "epoch": 2.23,
      "grad_norm": 3.6676738722666706,
      "learning_rate": 1.6317763917926494e-06,
      "loss": 0.7246,
      "step": 1028
    },
    {
      "epoch": 2.23,
      "grad_norm": 4.231275335588269,
      "learning_rate": 1.6231286299858635e-06,
      "loss": 0.8862,
      "step": 1029
    },
    {
      "epoch": 2.23,
      "grad_norm": 3.5208637129710056,
      "learning_rate": 1.6144994016392484e-06,
      "loss": 0.6832,
      "step": 1030
    },
    {
      "epoch": 2.23,
      "grad_norm": 4.830508142961262,
      "learning_rate": 1.6058887541130541e-06,
      "loss": 0.7411,
      "step": 1031
    },
    {
      "epoch": 2.24,
      "grad_norm": 3.3142066079793637,
      "learning_rate": 1.5972967346655449e-06,
      "loss": 0.6873,
      "step": 1032
    },
    {
      "epoch": 2.24,
      "grad_norm": 2.708956333932293,
      "learning_rate": 1.5887233904527548e-06,
      "loss": 1.0716,
      "step": 1033
    },
    {
      "epoch": 2.24,
      "grad_norm": 3.819871948715081,
      "learning_rate": 1.5801687685282169e-06,
      "loss": 0.7969,
      "step": 1034
    },
    {
      "epoch": 2.24,
      "grad_norm": 3.2574927732780843,
      "learning_rate": 1.5716329158427097e-06,
      "loss": 0.6045,
      "step": 1035
    },
    {
      "epoch": 2.24,
      "grad_norm": 3.978283974709393,
      "learning_rate": 1.5631158792440027e-06,
      "loss": 0.4923,
      "step": 1036
    },
    {
      "epoch": 2.25,
      "grad_norm": 4.786076311615979,
      "learning_rate": 1.5546177054765954e-06,
      "loss": 0.6364,
      "step": 1037
    },
    {
      "epoch": 2.25,
      "grad_norm": 3.6988477514629587,
      "learning_rate": 1.546138441181459e-06,
      "loss": 0.4704,
      "step": 1038
    },
    {
      "epoch": 2.25,
      "grad_norm": 4.813474512012489,
      "learning_rate": 1.5376781328957846e-06,
      "loss": 1.0928,
      "step": 1039
    },
    {
      "epoch": 2.25,
      "grad_norm": 4.099393158541576,
      "learning_rate": 1.5292368270527259e-06,
      "loss": 0.4666,
      "step": 1040
    },
    {
      "epoch": 2.26,
      "grad_norm": 4.729208297173249,
      "learning_rate": 1.5208145699811417e-06,
      "loss": 0.6468,
      "step": 1041
    },
    {
      "epoch": 2.26,
      "grad_norm": 4.2952031961692585,
      "learning_rate": 1.5124114079053492e-06,
      "loss": 0.5907,
      "step": 1042
    },
    {
      "epoch": 2.26,
      "grad_norm": 3.1541819309379755,
      "learning_rate": 1.5040273869448652e-06,
      "loss": 0.7636,
      "step": 1043
    },
    {
      "epoch": 2.26,
      "grad_norm": 3.143275669523735,
      "learning_rate": 1.4956625531141495e-06,
      "loss": 0.6156,
      "step": 1044
    },
    {
      "epoch": 2.26,
      "grad_norm": 4.639342564354085,
      "learning_rate": 1.4873169523223568e-06,
      "loss": 0.7773,
      "step": 1045
    },
    {
      "epoch": 2.27,
      "grad_norm": 4.040050143234123,
      "learning_rate": 1.4789906303730888e-06,
      "loss": 0.8212,
      "step": 1046
    },
    {
      "epoch": 2.27,
      "grad_norm": 3.3094417174689545,
      "learning_rate": 1.470683632964131e-06,
      "loss": 0.8892,
      "step": 1047
    },
    {
      "epoch": 2.27,
      "grad_norm": 3.2342692030216655,
      "learning_rate": 1.462396005687216e-06,
      "loss": 0.7523,
      "step": 1048
    },
    {
      "epoch": 2.27,
      "grad_norm": 3.174814822095073,
      "learning_rate": 1.4541277940277604e-06,
      "loss": 0.889,
      "step": 1049
    },
    {
      "epoch": 2.28,
      "grad_norm": 2.71305711785336,
      "learning_rate": 1.4458790433646264e-06,
      "loss": 0.3376,
      "step": 1050
    },
    {
      "epoch": 2.28,
      "grad_norm": 3.774461354461016,
      "learning_rate": 1.4376497989698635e-06,
      "loss": 0.4734,
      "step": 1051
    },
    {
      "epoch": 2.28,
      "grad_norm": 4.457397536855854,
      "learning_rate": 1.4294401060084634e-06,
      "loss": 0.6474,
      "step": 1052
    },
    {
      "epoch": 2.28,
      "grad_norm": 3.986523258550281,
      "learning_rate": 1.4212500095381176e-06,
      "loss": 0.56,
      "step": 1053
    },
    {
      "epoch": 2.28,
      "grad_norm": 3.3879283010301573,
      "learning_rate": 1.4130795545089588e-06,
      "loss": 0.8175,
      "step": 1054
    },
    {
      "epoch": 2.29,
      "grad_norm": 3.95245530956708,
      "learning_rate": 1.4049287857633264e-06,
      "loss": 0.882,
      "step": 1055
    },
    {
      "epoch": 2.29,
      "grad_norm": 3.696074115237081,
      "learning_rate": 1.3967977480355106e-06,
      "loss": 1.2675,
      "step": 1056
    },
    {
      "epoch": 2.29,
      "grad_norm": 4.009028397302704,
      "learning_rate": 1.388686485951512e-06,
      "loss": 0.7589,
      "step": 1057
    },
    {
      "epoch": 2.29,
      "grad_norm": 3.647218184915129,
      "learning_rate": 1.3805950440287936e-06,
      "loss": 0.7788,
      "step": 1058
    },
    {
      "epoch": 2.29,
      "grad_norm": 5.825470000314362,
      "learning_rate": 1.3725234666760428e-06,
      "loss": 1.0217,
      "step": 1059
    },
    {
      "epoch": 2.3,
      "grad_norm": 4.217192230007281,
      "learning_rate": 1.3644717981929213e-06,
      "loss": 1.3266,
      "step": 1060
    },
    {
      "epoch": 2.3,
      "grad_norm": 4.082055508837191,
      "learning_rate": 1.356440082769822e-06,
      "loss": 0.9868,
      "step": 1061
    },
    {
      "epoch": 2.3,
      "grad_norm": 4.127748136188224,
      "learning_rate": 1.3484283644876289e-06,
      "loss": 0.5976,
      "step": 1062
    },
    {
      "epoch": 2.3,
      "grad_norm": 3.7668032403101246,
      "learning_rate": 1.3404366873174778e-06,
      "loss": 0.6069,
      "step": 1063
    },
    {
      "epoch": 2.31,
      "grad_norm": 2.275613626492213,
      "learning_rate": 1.3324650951205064e-06,
      "loss": 0.3493,
      "step": 1064
    },
    {
      "epoch": 2.31,
      "grad_norm": 3.789295303795199,
      "learning_rate": 1.3245136316476253e-06,
      "loss": 0.5435,
      "step": 1065
    },
    {
      "epoch": 2.31,
      "grad_norm": 2.8378046655277824,
      "learning_rate": 1.3165823405392668e-06,
      "loss": 0.5181,
      "step": 1066
    },
    {
      "epoch": 2.31,
      "grad_norm": 3.9590247696842145,
      "learning_rate": 1.3086712653251504e-06,
      "loss": 0.5756,
      "step": 1067
    },
    {
      "epoch": 2.31,
      "grad_norm": 3.655691645699385,
      "learning_rate": 1.3007804494240478e-06,
      "loss": 0.8166,
      "step": 1068
    },
    {
      "epoch": 2.32,
      "grad_norm": 4.17394710746782,
      "learning_rate": 1.2929099361435348e-06,
      "loss": 1.0573,
      "step": 1069
    },
    {
      "epoch": 2.32,
      "grad_norm": 4.532621297993638,
      "learning_rate": 1.2850597686797644e-06,
      "loss": 0.8463,
      "step": 1070
    },
    {
      "epoch": 2.32,
      "grad_norm": 4.9056975904849205,
      "learning_rate": 1.2772299901172198e-06,
      "loss": 0.5864,
      "step": 1071
    },
    {
      "epoch": 2.32,
      "grad_norm": 3.8851952743872853,
      "learning_rate": 1.2694206434284878e-06,
      "loss": 0.9052,
      "step": 1072
    },
    {
      "epoch": 2.33,
      "grad_norm": 4.914410716342356,
      "learning_rate": 1.261631771474014e-06,
      "loss": 0.9956,
      "step": 1073
    },
    {
      "epoch": 2.33,
      "grad_norm": 3.3267859695488604,
      "learning_rate": 1.2538634170018727e-06,
      "loss": 0.4481,
      "step": 1074
    },
    {
      "epoch": 2.33,
      "grad_norm": 4.554248216193431,
      "learning_rate": 1.246115622647529e-06,
      "loss": 0.9084,
      "step": 1075
    },
    {
      "epoch": 2.33,
      "grad_norm": 4.4835978158366565,
      "learning_rate": 1.2383884309336114e-06,
      "loss": 0.5021,
      "step": 1076
    },
    {
      "epoch": 2.33,
      "grad_norm": 4.3494161616916465,
      "learning_rate": 1.2306818842696716e-06,
      "loss": 0.6337,
      "step": 1077
    },
    {
      "epoch": 2.34,
      "grad_norm": 3.3851782056081823,
      "learning_rate": 1.222996024951953e-06,
      "loss": 0.612,
      "step": 1078
    },
    {
      "epoch": 2.34,
      "grad_norm": 3.412479003746657,
      "learning_rate": 1.21533089516316e-06,
      "loss": 1.061,
      "step": 1079
    },
    {
      "epoch": 2.34,
      "grad_norm": 4.063745332718891,
      "learning_rate": 1.2076865369722246e-06,
      "loss": 0.9612,
      "step": 1080
    },
    {
      "epoch": 2.34,
      "grad_norm": 3.2420984782356124,
      "learning_rate": 1.2000629923340801e-06,
      "loss": 0.5312,
      "step": 1081
    },
    {
      "epoch": 2.34,
      "grad_norm": 3.991103369920871,
      "learning_rate": 1.1924603030894277e-06,
      "loss": 0.8386,
      "step": 1082
    },
    {
      "epoch": 2.35,
      "grad_norm": 3.684291825678522,
      "learning_rate": 1.184878510964504e-06,
      "loss": 0.5188,
      "step": 1083
    },
    {
      "epoch": 2.35,
      "grad_norm": 3.6534169814147446,
      "learning_rate": 1.1773176575708544e-06,
      "loss": 0.5278,
      "step": 1084
    },
    {
      "epoch": 2.35,
      "grad_norm": 2.9910215049979145,
      "learning_rate": 1.1697777844051105e-06,
      "loss": 0.5045,
      "step": 1085
    },
    {
      "epoch": 2.35,
      "grad_norm": 11.794204812101183,
      "learning_rate": 1.1622589328487505e-06,
      "loss": 0.9754,
      "step": 1086
    },
    {
      "epoch": 2.36,
      "grad_norm": 6.462389051326817,
      "learning_rate": 1.1547611441678836e-06,
      "loss": 0.9763,
      "step": 1087
    },
    {
      "epoch": 2.36,
      "grad_norm": 3.019231159524319,
      "learning_rate": 1.1472844595130145e-06,
      "loss": 0.9418,
      "step": 1088
    },
    {
      "epoch": 2.36,
      "grad_norm": 4.770771756595,
      "learning_rate": 1.1398289199188262e-06,
      "loss": 0.8821,
      "step": 1089
    },
    {
      "epoch": 2.36,
      "grad_norm": 5.572981453232036,
      "learning_rate": 1.132394566303946e-06,
      "loss": 1.1711,
      "step": 1090
    },
    {
      "epoch": 2.36,
      "grad_norm": 3.101407332624737,
      "learning_rate": 1.124981439470726e-06,
      "loss": 0.6692,
      "step": 1091
    },
    {
      "epoch": 2.37,
      "grad_norm": 2.8468680677286913,
      "learning_rate": 1.1175895801050185e-06,
      "loss": 0.602,
      "step": 1092
    },
    {
      "epoch": 2.37,
      "grad_norm": 2.8067660056383987,
      "learning_rate": 1.110219028775954e-06,
      "loss": 0.4187,
      "step": 1093
    },
    {
      "epoch": 2.37,
      "grad_norm": 5.27580433519872,
      "learning_rate": 1.1028698259357162e-06,
      "loss": 0.6387,
      "step": 1094
    },
    {
      "epoch": 2.37,
      "grad_norm": 4.525564639738765,
      "learning_rate": 1.09554201191932e-06,
      "loss": 0.9667,
      "step": 1095
    },
    {
      "epoch": 2.37,
      "grad_norm": 3.079793740756796,
      "learning_rate": 1.0882356269443912e-06,
      "loss": 0.778,
      "step": 1096
    },
    {
      "epoch": 2.38,
      "grad_norm": 2.967562992309466,
      "learning_rate": 1.080950711110943e-06,
      "loss": 0.5263,
      "step": 1097
    },
    {
      "epoch": 2.38,
      "grad_norm": 4.483310781831877,
      "learning_rate": 1.0736873044011632e-06,
      "loss": 1.0524,
      "step": 1098
    },
    {
      "epoch": 2.38,
      "grad_norm": 5.177188556242962,
      "learning_rate": 1.066445446679189e-06,
      "loss": 0.7285,
      "step": 1099
    },
    {
      "epoch": 2.38,
      "grad_norm": 4.9343650462783035,
      "learning_rate": 1.0592251776908857e-06,
      "loss": 0.6376,
      "step": 1100
    },
    {
      "epoch": 2.39,
      "grad_norm": 4.00343539787233,
      "learning_rate": 1.052026537063634e-06,
      "loss": 0.4683,
      "step": 1101
    },
    {
      "epoch": 2.39,
      "grad_norm": 2.6424773216618664,
      "learning_rate": 1.044849564306112e-06,
      "loss": 0.316,
      "step": 1102
    },
    {
      "epoch": 2.39,
      "grad_norm": 4.071097113586533,
      "learning_rate": 1.037694298808074e-06,
      "loss": 1.0268,
      "step": 1103
    },
    {
      "epoch": 2.39,
      "grad_norm": 4.268111514665191,
      "learning_rate": 1.03056077984014e-06,
      "loss": 0.7985,
      "step": 1104
    },
    {
      "epoch": 2.39,
      "grad_norm": 4.291604686869361,
      "learning_rate": 1.023449046553575e-06,
      "loss": 0.6545,
      "step": 1105
    },
    {
      "epoch": 2.4,
      "grad_norm": 4.89982858592844,
      "learning_rate": 1.0163591379800796e-06,
      "loss": 0.7632,
      "step": 1106
    },
    {
      "epoch": 2.4,
      "grad_norm": 3.94165080114321,
      "learning_rate": 1.0092910930315698e-06,
      "loss": 0.7757,
      "step": 1107
    },
    {
      "epoch": 2.4,
      "grad_norm": 5.867206971543332,
      "learning_rate": 1.002244950499966e-06,
      "loss": 0.772,
      "step": 1108
    },
    {
      "epoch": 2.4,
      "grad_norm": 3.9290296914963436,
      "learning_rate": 9.952207490569816e-07,
      "loss": 0.7619,
      "step": 1109
    },
    {
      "epoch": 2.41,
      "grad_norm": 3.5316728546575415,
      "learning_rate": 9.882185272539107e-07,
      "loss": 1.1061,
      "step": 1110
    },
    {
      "epoch": 2.41,
      "grad_norm": 2.89889015079531,
      "learning_rate": 9.81238323521415e-07,
      "loss": 0.7458,
      "step": 1111
    },
    {
      "epoch": 2.41,
      "grad_norm": 4.6376447162565935,
      "learning_rate": 9.742801761693122e-07,
      "loss": 0.8282,
      "step": 1112
    },
    {
      "epoch": 2.41,
      "grad_norm": 3.7097684434343976,
      "learning_rate": 9.673441233863661e-07,
      "loss": 0.7635,
      "step": 1113
    },
    {
      "epoch": 2.41,
      "grad_norm": 5.030603587536502,
      "learning_rate": 9.604302032400787e-07,
      "loss": 0.8801,
      "step": 1114
    },
    {
      "epoch": 2.42,
      "grad_norm": 6.233851850396977,
      "learning_rate": 9.535384536764807e-07,
      "loss": 0.9537,
      "step": 1115
    },
    {
      "epoch": 2.42,
      "grad_norm": 2.873362541343195,
      "learning_rate": 9.466689125199247e-07,
      "loss": 0.4023,
      "step": 1116
    },
    {
      "epoch": 2.42,
      "grad_norm": 3.074271690810135,
      "learning_rate": 9.39821617472872e-07,
      "loss": 0.8144,
      "step": 1117
    },
    {
      "epoch": 2.42,
      "grad_norm": 4.266692475146756,
      "learning_rate": 9.329966061156887e-07,
      "loss": 0.8052,
      "step": 1118
    },
    {
      "epoch": 2.42,
      "grad_norm": 2.6861629287027604,
      "learning_rate": 9.261939159064465e-07,
      "loss": 0.6277,
      "step": 1119
    },
    {
      "epoch": 2.43,
      "grad_norm": 2.899276590307398,
      "learning_rate": 9.194135841807028e-07,
      "loss": 0.7832,
      "step": 1120
    },
    {
      "epoch": 2.43,
      "grad_norm": 4.569950086021629,
      "learning_rate": 9.12655648151311e-07,
      "loss": 0.8706,
      "step": 1121
    },
    {
      "epoch": 2.43,
      "grad_norm": 2.5465472002878804,
      "learning_rate": 9.059201449082045e-07,
      "loss": 0.3566,
      "step": 1122
    },
    {
      "epoch": 2.43,
      "grad_norm": 2.8934558719412773,
      "learning_rate": 8.992071114181977e-07,
      "loss": 0.9903,
      "step": 1123
    },
    {
      "epoch": 2.44,
      "grad_norm": 35.29958917611168,
      "learning_rate": 8.925165845247858e-07,
      "loss": 0.8789,
      "step": 1124
    },
    {
      "epoch": 2.44,
      "grad_norm": 2.731422129853721,
      "learning_rate": 8.858486009479384e-07,
      "loss": 0.5998,
      "step": 1125
    },
    {
      "epoch": 2.44,
      "grad_norm": 4.14346734161125,
      "learning_rate": 8.792031972838966e-07,
      "loss": 1.1381,
      "step": 1126
    },
    {
      "epoch": 2.44,
      "grad_norm": 4.1962665176007015,
      "learning_rate": 8.7258041000498e-07,
      "loss": 0.4847,
      "step": 1127
    },
    {
      "epoch": 2.44,
      "grad_norm": 3.325885692663973,
      "learning_rate": 8.659802754593805e-07,
      "loss": 0.7411,
      "step": 1128
    },
    {
      "epoch": 2.45,
      "grad_norm": 3.708308565163734,
      "learning_rate": 8.594028298709605e-07,
      "loss": 0.915,
      "step": 1129
    },
    {
      "epoch": 2.45,
      "grad_norm": 2.4554229593221493,
      "learning_rate": 8.528481093390606e-07,
      "loss": 0.4377,
      "step": 1130
    },
    {
      "epoch": 2.45,
      "grad_norm": 3.9956844140018455,
      "learning_rate": 8.463161498382949e-07,
      "loss": 0.7974,
      "step": 1131
    },
    {
      "epoch": 2.45,
      "grad_norm": 5.485343881744705,
      "learning_rate": 8.398069872183607e-07,
      "loss": 0.8849,
      "step": 1132
    },
    {
      "epoch": 2.46,
      "grad_norm": 8.539989248571814,
      "learning_rate": 8.333206572038377e-07,
      "loss": 0.7398,
      "step": 1133
    },
    {
      "epoch": 2.46,
      "grad_norm": 4.155275844761884,
      "learning_rate": 8.268571953939897e-07,
      "loss": 0.7859,
      "step": 1134
    },
    {
      "epoch": 2.46,
      "grad_norm": 3.9925386614383274,
      "learning_rate": 8.204166372625727e-07,
      "loss": 0.5918,
      "step": 1135
    },
    {
      "epoch": 2.46,
      "grad_norm": 8.358829288204815,
      "learning_rate": 8.139990181576391e-07,
      "loss": 0.7719,
      "step": 1136
    },
    {
      "epoch": 2.46,
      "grad_norm": 3.9028579453611307,
      "learning_rate": 8.07604373301345e-07,
      "loss": 0.8872,
      "step": 1137
    },
    {
      "epoch": 2.47,
      "grad_norm": 3.211083871727704,
      "learning_rate": 8.012327377897561e-07,
      "loss": 0.541,
      "step": 1138
    },
    {
      "epoch": 2.47,
      "grad_norm": 4.872510272084048,
      "learning_rate": 7.948841465926533e-07,
      "loss": 0.648,
      "step": 1139
    },
    {
      "epoch": 2.47,
      "grad_norm": 3.278241443618759,
      "learning_rate": 7.885586345533397e-07,
      "loss": 0.7341,
      "step": 1140
    },
    {
      "epoch": 2.47,
      "grad_norm": 11.969777347156834,
      "learning_rate": 7.822562363884584e-07,
      "loss": 0.7185,
      "step": 1141
    },
    {
      "epoch": 2.47,
      "grad_norm": 2.6956087870026746,
      "learning_rate": 7.759769866877892e-07,
      "loss": 0.744,
      "step": 1142
    },
    {
      "epoch": 2.48,
      "grad_norm": 3.548278233008131,
      "learning_rate": 7.697209199140676e-07,
      "loss": 0.4974,
      "step": 1143
    },
    {
      "epoch": 2.48,
      "grad_norm": 3.3936725555394585,
      "learning_rate": 7.634880704027936e-07,
      "loss": 0.4715,
      "step": 1144
    },
    {
      "epoch": 2.48,
      "grad_norm": 3.94805057993403,
      "learning_rate": 7.572784723620424e-07,
      "loss": 1.2849,
      "step": 1145
    },
    {
      "epoch": 2.48,
      "grad_norm": 3.529310190303956,
      "learning_rate": 7.510921598722765e-07,
      "loss": 0.8144,
      "step": 1146
    },
    {
      "epoch": 2.49,
      "grad_norm": 3.9268920471102753,
      "learning_rate": 7.449291668861575e-07,
      "loss": 0.7064,
      "step": 1147
    },
    {
      "epoch": 2.49,
      "grad_norm": 7.010528181385532,
      "learning_rate": 7.387895272283635e-07,
      "loss": 1.1697,
      "step": 1148
    },
    {
      "epoch": 2.49,
      "grad_norm": 2.967293089185817,
      "learning_rate": 7.326732745954001e-07,
      "loss": 0.5221,
      "step": 1149
    },
    {
      "epoch": 2.49,
      "grad_norm": 3.325446213030333,
      "learning_rate": 7.265804425554202e-07,
      "loss": 0.5931,
      "step": 1150
    },
    {
      "epoch": 2.49,
      "grad_norm": 4.873159247947369,
      "learning_rate": 7.205110645480307e-07,
      "loss": 0.8414,
      "step": 1151
    },
    {
      "epoch": 2.5,
      "grad_norm": 5.213847899451591,
      "learning_rate": 7.144651738841174e-07,
      "loss": 0.8516,
      "step": 1152
    },
    {
      "epoch": 2.5,
      "grad_norm": 4.291838382178979,
      "learning_rate": 7.084428037456587e-07,
      "loss": 1.0744,
      "step": 1153
    },
    {
      "epoch": 2.5,
      "grad_norm": 4.073917885715483,
      "learning_rate": 7.024439871855448e-07,
      "loss": 1.0687,
      "step": 1154
    },
    {
      "epoch": 2.5,
      "grad_norm": 3.504272595685579,
      "learning_rate": 6.96468757127396e-07,
      "loss": 0.6589,
      "step": 1155
    },
    {
      "epoch": 2.5,
      "grad_norm": 4.374418457104451,
      "learning_rate": 6.905171463653798e-07,
      "loss": 0.8327,
      "step": 1156
    },
    {
      "epoch": 2.51,
      "grad_norm": 3.8177081032571496,
      "learning_rate": 6.845891875640331e-07,
      "loss": 1.0254,
      "step": 1157
    },
    {
      "epoch": 2.51,
      "grad_norm": 3.816156550512672,
      "learning_rate": 6.786849132580841e-07,
      "loss": 0.8273,
      "step": 1158
    },
    {
      "epoch": 2.51,
      "grad_norm": 2.874239498677427,
      "learning_rate": 6.728043558522706e-07,
      "loss": 1.0256,
      "step": 1159
    },
    {
      "epoch": 2.51,
      "grad_norm": 3.567049089717947,
      "learning_rate": 6.669475476211628e-07,
      "loss": 1.13,
      "step": 1160
    },
    {
      "epoch": 2.52,
      "grad_norm": 4.319667655697863,
      "learning_rate": 6.611145207089897e-07,
      "loss": 0.768,
      "step": 1161
    },
    {
      "epoch": 2.52,
      "grad_norm": 3.2393455444532404,
      "learning_rate": 6.55305307129459e-07,
      "loss": 0.6062,
      "step": 1162
    },
    {
      "epoch": 2.52,
      "grad_norm": 4.279559230703671,
      "learning_rate": 6.49519938765582e-07,
      "loss": 0.469,
      "step": 1163
    },
    {
      "epoch": 2.52,
      "grad_norm": 3.1092108285442204,
      "learning_rate": 6.437584473694991e-07,
      "loss": 0.7332,
      "step": 1164
    },
    {
      "epoch": 2.52,
      "grad_norm": 4.398874032328195,
      "learning_rate": 6.380208645623037e-07,
      "loss": 0.9302,
      "step": 1165
    },
    {
      "epoch": 2.53,
      "grad_norm": 5.485611308918295,
      "learning_rate": 6.323072218338739e-07,
      "loss": 0.451,
      "step": 1166
    },
    {
      "epoch": 2.53,
      "grad_norm": 4.366185278845239,
      "learning_rate": 6.266175505426958e-07,
      "loss": 0.7235,
      "step": 1167
    },
    {
      "epoch": 2.53,
      "grad_norm": 3.7238873272816817,
      "learning_rate": 6.209518819156895e-07,
      "loss": 0.6972,
      "step": 1168
    },
    {
      "epoch": 2.53,
      "grad_norm": 3.459934304625689,
      "learning_rate": 6.15310247048041e-07,
      "loss": 0.9183,
      "step": 1169
    },
    {
      "epoch": 2.54,
      "grad_norm": 4.479765701739508,
      "learning_rate": 6.096926769030298e-07,
      "loss": 1.1313,
      "step": 1170
    },
    {
      "epoch": 2.54,
      "grad_norm": 5.703513837671345,
      "learning_rate": 6.040992023118624e-07,
      "loss": 1.1461,
      "step": 1171
    },
    {
      "epoch": 2.54,
      "grad_norm": 3.871240039774051,
      "learning_rate": 5.985298539734973e-07,
      "loss": 0.7328,
      "step": 1172
    },
    {
      "epoch": 2.54,
      "grad_norm": 2.9900545689536164,
      "learning_rate": 5.929846624544821e-07,
      "loss": 0.754,
      "step": 1173
    },
    {
      "epoch": 2.54,
      "grad_norm": 3.5638730318870557,
      "learning_rate": 5.874636581887804e-07,
      "loss": 0.8462,
      "step": 1174
    },
    {
      "epoch": 2.55,
      "grad_norm": 4.983501212877505,
      "learning_rate": 5.819668714776122e-07,
      "loss": 0.4921,
      "step": 1175
    },
    {
      "epoch": 2.55,
      "grad_norm": 3.5878540138317154,
      "learning_rate": 5.76494332489278e-07,
      "loss": 0.9391,
      "step": 1176
    },
    {
      "epoch": 2.55,
      "grad_norm": 3.615693425408774,
      "learning_rate": 5.710460712589993e-07,
      "loss": 0.4171,
      "step": 1177
    },
    {
      "epoch": 2.55,
      "grad_norm": 5.349884610186352,
      "learning_rate": 5.656221176887572e-07,
      "loss": 0.9766,
      "step": 1178
    },
    {
      "epoch": 2.55,
      "grad_norm": 3.6022714720477156,
      "learning_rate": 5.602225015471175e-07,
      "loss": 0.945,
      "step": 1179
    },
    {
      "epoch": 2.56,
      "grad_norm": 3.5904749873324735,
      "learning_rate": 5.548472524690784e-07,
      "loss": 0.7807,
      "step": 1180
    },
    {
      "epoch": 2.56,
      "grad_norm": 6.2779712703383925,
      "learning_rate": 5.494963999559011e-07,
      "loss": 0.7399,
      "step": 1181
    },
    {
      "epoch": 2.56,
      "grad_norm": 3.6998393538915852,
      "learning_rate": 5.441699733749479e-07,
      "loss": 0.5907,
      "step": 1182
    },
    {
      "epoch": 2.56,
      "grad_norm": 4.550505894047273,
      "learning_rate": 5.388680019595266e-07,
      "loss": 0.6824,
      "step": 1183
    },
    {
      "epoch": 2.57,
      "grad_norm": 4.994550541279491,
      "learning_rate": 5.335905148087256e-07,
      "loss": 0.9723,
      "step": 1184
    },
    {
      "epoch": 2.57,
      "grad_norm": 3.7759282364766045,
      "learning_rate": 5.283375408872538e-07,
      "loss": 0.7567,
      "step": 1185
    },
    {
      "epoch": 2.57,
      "grad_norm": 3.1544311287632523,
      "learning_rate": 5.231091090252832e-07,
      "loss": 0.5102,
      "step": 1186
    },
    {
      "epoch": 2.57,
      "grad_norm": 3.999942909152997,
      "learning_rate": 5.179052479182889e-07,
      "loss": 0.775,
      "step": 1187
    },
    {
      "epoch": 2.57,
      "grad_norm": 3.592053792754821,
      "learning_rate": 5.127259861268974e-07,
      "loss": 0.7948,
      "step": 1188
    },
    {
      "epoch": 2.58,
      "grad_norm": 4.159246526852843,
      "learning_rate": 5.075713520767201e-07,
      "loss": 0.6045,
      "step": 1189
    },
    {
      "epoch": 2.58,
      "grad_norm": 4.254150523122296,
      "learning_rate": 5.024413740582074e-07,
      "loss": 0.8733,
      "step": 1190
    },
    {
      "epoch": 2.58,
      "grad_norm": 4.65456139029262,
      "learning_rate": 4.973360802264859e-07,
      "loss": 0.718,
      "step": 1191
    },
    {
      "epoch": 2.58,
      "grad_norm": 4.917495614013312,
      "learning_rate": 4.922554986012068e-07,
      "loss": 0.7542,
      "step": 1192
    },
    {
      "epoch": 2.59,
      "grad_norm": 3.2998623377448952,
      "learning_rate": 4.871996570663934e-07,
      "loss": 0.7912,
      "step": 1193
    },
    {
      "epoch": 2.59,
      "grad_norm": 7.063984229403393,
      "learning_rate": 4.82168583370285e-07,
      "loss": 0.8302,
      "step": 1194
    },
    {
      "epoch": 2.59,
      "grad_norm": 4.006968298605131,
      "learning_rate": 4.771623051251878e-07,
      "loss": 1.0133,
      "step": 1195
    },
    {
      "epoch": 2.59,
      "grad_norm": 2.585942645521526,
      "learning_rate": 4.721808498073205e-07,
      "loss": 0.8539,
      "step": 1196
    },
    {
      "epoch": 2.59,
      "grad_norm": 4.292635394607348,
      "learning_rate": 4.6722424475666715e-07,
      "loss": 0.5966,
      "step": 1197
    },
    {
      "epoch": 2.6,
      "grad_norm": 2.829951568975223,
      "learning_rate": 4.622925171768211e-07,
      "loss": 0.5376,
      "step": 1198
    },
    {
      "epoch": 2.6,
      "grad_norm": 3.340291358675051,
      "learning_rate": 4.57385694134842e-07,
      "loss": 0.6756,
      "step": 1199
    },
    {
      "epoch": 2.6,
      "grad_norm": 3.566500784328458,
      "learning_rate": 4.5250380256110335e-07,
      "loss": 0.6052,
      "step": 1200
    },
    {
      "epoch": 2.6,
      "grad_norm": 3.1032865466394703,
      "learning_rate": 4.476468692491487e-07,
      "loss": 0.4768,
      "step": 1201
    },
    {
      "epoch": 2.6,
      "grad_norm": 2.8412283649010512,
      "learning_rate": 4.428149208555388e-07,
      "loss": 0.4835,
      "step": 1202
    },
    {
      "epoch": 2.61,
      "grad_norm": 3.1603339449986114,
      "learning_rate": 4.380079838997087e-07,
      "loss": 0.3973,
      "step": 1203
    },
    {
      "epoch": 2.61,
      "grad_norm": 3.3202778182791284,
      "learning_rate": 4.3322608476382255e-07,
      "loss": 0.562,
      "step": 1204
    },
    {
      "epoch": 2.61,
      "grad_norm": 7.850824996314681,
      "learning_rate": 4.2846924969262736e-07,
      "loss": 0.684,
      "step": 1205
    },
    {
      "epoch": 2.61,
      "grad_norm": 3.7284696082195463,
      "learning_rate": 4.237375047933118e-07,
      "loss": 0.9255,
      "step": 1206
    },
    {
      "epoch": 2.62,
      "grad_norm": 4.055332292091045,
      "learning_rate": 4.190308760353595e-07,
      "loss": 1.0129,
      "step": 1207
    },
    {
      "epoch": 2.62,
      "grad_norm": 4.843747650963575,
      "learning_rate": 4.1434938925040804e-07,
      "loss": 0.9376,
      "step": 1208
    },
    {
      "epoch": 2.62,
      "grad_norm": 4.044308082624069,
      "learning_rate": 4.0969307013210445e-07,
      "loss": 0.6334,
      "step": 1209
    },
    {
      "epoch": 2.62,
      "grad_norm": 3.916953995416515,
      "learning_rate": 4.050619442359721e-07,
      "loss": 0.7055,
      "step": 1210
    },
    {
      "epoch": 2.62,
      "grad_norm": 3.65659512352173,
      "learning_rate": 4.004560369792593e-07,
      "loss": 0.8702,
      "step": 1211
    },
    {
      "epoch": 2.63,
      "grad_norm": 4.502079949870695,
      "learning_rate": 3.958753736408105e-07,
      "loss": 1.1807,
      "step": 1212
    },
    {
      "epoch": 2.63,
      "grad_norm": 4.36684085789138,
      "learning_rate": 3.91319979360919e-07,
      "loss": 0.6011,
      "step": 1213
    },
    {
      "epoch": 2.63,
      "grad_norm": 4.026283960595477,
      "learning_rate": 3.867898791411956e-07,
      "loss": 0.4615,
      "step": 1214
    },
    {
      "epoch": 2.63,
      "grad_norm": 3.853542159816083,
      "learning_rate": 3.822850978444254e-07,
      "loss": 1.3125,
      "step": 1215
    },
    {
      "epoch": 2.63,
      "grad_norm": 2.8269703586336767,
      "learning_rate": 3.778056601944358e-07,
      "loss": 0.6861,
      "step": 1216
    },
    {
      "epoch": 2.64,
      "grad_norm": 3.5814274883776083,
      "learning_rate": 3.733515907759594e-07,
      "loss": 1.0328,
      "step": 1217
    },
    {
      "epoch": 2.64,
      "grad_norm": 9.808015972735548,
      "learning_rate": 3.6892291403449963e-07,
      "loss": 0.6525,
      "step": 1218
    },
    {
      "epoch": 2.64,
      "grad_norm": 3.256726763031335,
      "learning_rate": 3.645196542761953e-07,
      "loss": 0.7215,
      "step": 1219
    },
    {
      "epoch": 2.64,
      "grad_norm": 3.459534626129964,
      "learning_rate": 3.6014183566768725e-07,
      "loss": 0.66,
      "step": 1220
    },
    {
      "epoch": 2.65,
      "grad_norm": 3.2570390358673986,
      "learning_rate": 3.557894822359864e-07,
      "loss": 0.9572,
      "step": 1221
    },
    {
      "epoch": 2.65,
      "grad_norm": 5.2803976399265595,
      "learning_rate": 3.5146261786834225e-07,
      "loss": 0.5007,
      "step": 1222
    },
    {
      "epoch": 2.65,
      "grad_norm": 4.034277456748412,
      "learning_rate": 3.471612663121121e-07,
      "loss": 0.4671,
      "step": 1223
    },
    {
      "epoch": 2.65,
      "grad_norm": 4.403721059495357,
      "learning_rate": 3.428854511746293e-07,
      "loss": 0.7494,
      "step": 1224
    },
    {
      "epoch": 2.65,
      "grad_norm": 3.255957136774676,
      "learning_rate": 3.386351959230738e-07,
      "loss": 0.7854,
      "step": 1225
    },
    {
      "epoch": 2.66,
      "grad_norm": 4.863528743432865,
      "learning_rate": 3.344105238843437e-07,
      "loss": 0.6159,
      "step": 1226
    },
    {
      "epoch": 2.66,
      "grad_norm": 4.160154096984732,
      "learning_rate": 3.302114582449295e-07,
      "loss": 1.1903,
      "step": 1227
    },
    {
      "epoch": 2.66,
      "grad_norm": 3.713268698495127,
      "learning_rate": 3.2603802205078195e-07,
      "loss": 0.7883,
      "step": 1228
    },
    {
      "epoch": 2.66,
      "grad_norm": 4.012520227502371,
      "learning_rate": 3.2189023820719034e-07,
      "loss": 0.709,
      "step": 1229
    },
    {
      "epoch": 2.67,
      "grad_norm": 4.167277337464063,
      "learning_rate": 3.177681294786539e-07,
      "loss": 0.9512,
      "step": 1230
    },
    {
      "epoch": 2.67,
      "grad_norm": 3.8617658304838134,
      "learning_rate": 3.136717184887589e-07,
      "loss": 0.6393,
      "step": 1231
    },
    {
      "epoch": 2.67,
      "grad_norm": 3.4497551575318464,
      "learning_rate": 3.0960102772005174e-07,
      "loss": 0.6874,
      "step": 1232
    },
    {
      "epoch": 2.67,
      "grad_norm": 3.1704999209237332,
      "learning_rate": 3.055560795139173e-07,
      "loss": 1.0812,
      "step": 1233
    },
    {
      "epoch": 2.67,
      "grad_norm": 3.536455446086755,
      "learning_rate": 3.015368960704584e-07,
      "loss": 0.8365,
      "step": 1234
    },
    {
      "epoch": 2.68,
      "grad_norm": 3.153091330095435,
      "learning_rate": 2.975434994483689e-07,
      "loss": 0.6632,
      "step": 1235
    },
    {
      "epoch": 2.68,
      "grad_norm": 2.4879590486011107,
      "learning_rate": 2.9357591156481793e-07,
      "loss": 0.4945,
      "step": 1236
    },
    {
      "epoch": 2.68,
      "grad_norm": 5.819571960910157,
      "learning_rate": 2.896341541953257e-07,
      "loss": 0.5708,
      "step": 1237
    },
    {
      "epoch": 2.68,
      "grad_norm": 5.8405115380708175,
      "learning_rate": 2.85718248973646e-07,
      "loss": 1.0022,
      "step": 1238
    },
    {
      "epoch": 2.68,
      "grad_norm": 2.7558082281327767,
      "learning_rate": 2.8182821739164534e-07,
      "loss": 0.5431,
      "step": 1239
    },
    {
      "epoch": 2.69,
      "grad_norm": 3.533321536980292,
      "learning_rate": 2.779640807991896e-07,
      "loss": 0.6858,
      "step": 1240
    },
    {
      "epoch": 2.69,
      "grad_norm": 3.6273475245248608,
      "learning_rate": 2.74125860404022e-07,
      "loss": 0.7401,
      "step": 1241
    },
    {
      "epoch": 2.69,
      "grad_norm": 4.456877900764199,
      "learning_rate": 2.7031357727164865e-07,
      "loss": 0.7417,
      "step": 1242
    },
    {
      "epoch": 2.69,
      "grad_norm": 3.970722527489924,
      "learning_rate": 2.665272523252216e-07,
      "loss": 0.7894,
      "step": 1243
    },
    {
      "epoch": 2.7,
      "grad_norm": 4.6539778520342,
      "learning_rate": 2.627669063454291e-07,
      "loss": 0.8177,
      "step": 1244
    },
    {
      "epoch": 2.7,
      "grad_norm": 3.2279372141250104,
      "learning_rate": 2.5903255997037246e-07,
      "loss": 0.8495,
      "step": 1245
    },
    {
      "epoch": 2.7,
      "grad_norm": 4.419333105016233,
      "learning_rate": 2.553242336954631e-07,
      "loss": 0.857,
      "step": 1246
    },
    {
      "epoch": 2.7,
      "grad_norm": 3.219582760231803,
      "learning_rate": 2.516419478733012e-07,
      "loss": 0.7043,
      "step": 1247
    },
    {
      "epoch": 2.7,
      "grad_norm": 3.156484238843353,
      "learning_rate": 2.479857227135685e-07,
      "loss": 0.7978,
      "step": 1248
    },
    {
      "epoch": 2.71,
      "grad_norm": 3.6625739013638547,
      "learning_rate": 2.443555782829188e-07,
      "loss": 1.1694,
      "step": 1249
    },
    {
      "epoch": 2.71,
      "grad_norm": 3.942333672766869,
      "learning_rate": 2.407515345048622e-07,
      "loss": 0.8283,
      "step": 1250
    },
    {
      "epoch": 2.71,
      "grad_norm": 4.797621882321987,
      "learning_rate": 2.3717361115966343e-07,
      "loss": 0.8573,
      "step": 1251
    },
    {
      "epoch": 2.71,
      "grad_norm": 4.232824254283906,
      "learning_rate": 2.3362182788422395e-07,
      "loss": 0.3124,
      "step": 1252
    },
    {
      "epoch": 2.72,
      "grad_norm": 3.0663476229509206,
      "learning_rate": 2.300962041719851e-07,
      "loss": 0.612,
      "step": 1253
    },
    {
      "epoch": 2.72,
      "grad_norm": 3.204055460164806,
      "learning_rate": 2.2659675937281078e-07,
      "loss": 0.7835,
      "step": 1254
    },
    {
      "epoch": 2.72,
      "grad_norm": 4.800383192813089,
      "learning_rate": 2.2312351269288712e-07,
      "loss": 0.5857,
      "step": 1255
    },
    {
      "epoch": 2.72,
      "grad_norm": 3.85990701972133,
      "learning_rate": 2.1967648319461577e-07,
      "loss": 0.3568,
      "step": 1256
    },
    {
      "epoch": 2.72,
      "grad_norm": 3.945022255034745,
      "learning_rate": 2.1625568979651012e-07,
      "loss": 0.582,
      "step": 1257
    },
    {
      "epoch": 2.73,
      "grad_norm": 4.251083546100536,
      "learning_rate": 2.1286115127308992e-07,
      "loss": 1.1017,
      "step": 1258
    },
    {
      "epoch": 2.73,
      "grad_norm": 10.263320490997698,
      "learning_rate": 2.0949288625477903e-07,
      "loss": 1.0074,
      "step": 1259
    },
    {
      "epoch": 2.73,
      "grad_norm": 4.670488597843965,
      "learning_rate": 2.061509132278028e-07,
      "loss": 0.8381,
      "step": 1260
    },
    {
      "epoch": 2.73,
      "grad_norm": 4.200300501452165,
      "learning_rate": 2.028352505340858e-07,
      "loss": 0.8736,
      "step": 1261
    },
    {
      "epoch": 2.73,
      "grad_norm": 4.578494878279605,
      "learning_rate": 1.9954591637115495e-07,
      "loss": 0.8713,
      "step": 1262
    },
    {
      "epoch": 2.74,
      "grad_norm": 4.009155204411217,
      "learning_rate": 1.9628292879203482e-07,
      "loss": 0.7406,
      "step": 1263
    },
    {
      "epoch": 2.74,
      "grad_norm": 4.023489782187626,
      "learning_rate": 1.9304630570515182e-07,
      "loss": 1.3087,
      "step": 1264
    },
    {
      "epoch": 2.74,
      "grad_norm": 9.332074482562328,
      "learning_rate": 1.8983606487423255e-07,
      "loss": 0.8784,
      "step": 1265
    },
    {
      "epoch": 2.74,
      "grad_norm": 3.7423454410691224,
      "learning_rate": 1.866522239182117e-07,
      "loss": 1.0179,
      "step": 1266
    },
    {
      "epoch": 2.75,
      "grad_norm": 3.47932716996629,
      "learning_rate": 1.8349480031112977e-07,
      "loss": 0.5978,
      "step": 1267
    },
    {
      "epoch": 2.75,
      "grad_norm": 4.719788409401892,
      "learning_rate": 1.8036381138204051e-07,
      "loss": 0.9179,
      "step": 1268
    },
    {
      "epoch": 2.75,
      "grad_norm": 4.84067509863434,
      "learning_rate": 1.7725927431491375e-07,
      "loss": 0.983,
      "step": 1269
    },
    {
      "epoch": 2.75,
      "grad_norm": 6.489623076690749,
      "learning_rate": 1.7418120614854427e-07,
      "loss": 0.6269,
      "step": 1270
    },
    {
      "epoch": 2.75,
      "grad_norm": 3.6623968802091005,
      "learning_rate": 1.711296237764548e-07,
      "loss": 0.654,
      "step": 1271
    },
    {
      "epoch": 2.76,
      "grad_norm": 3.5963754361301996,
      "learning_rate": 1.6810454394680431e-07,
      "loss": 0.8288,
      "step": 1272
    },
    {
      "epoch": 2.76,
      "grad_norm": 3.9158182301230084,
      "learning_rate": 1.6510598326229645e-07,
      "loss": 0.7982,
      "step": 1273
    },
    {
      "epoch": 2.76,
      "grad_norm": 4.733369410831597,
      "learning_rate": 1.6213395818009016e-07,
      "loss": 0.711,
      "step": 1274
    },
    {
      "epoch": 2.76,
      "grad_norm": 3.5399742141853707,
      "learning_rate": 1.5918848501170647e-07,
      "loss": 0.9369,
      "step": 1275
    },
    {
      "epoch": 2.76,
      "grad_norm": 3.7854491084742747,
      "learning_rate": 1.5626957992293966e-07,
      "loss": 0.6426,
      "step": 1276
    },
    {
      "epoch": 2.77,
      "grad_norm": 3.8357469418482455,
      "learning_rate": 1.5337725893376954e-07,
      "loss": 1.2189,
      "step": 1277
    },
    {
      "epoch": 2.77,
      "grad_norm": 3.0642565296545343,
      "learning_rate": 1.505115379182731e-07,
      "loss": 0.4267,
      "step": 1278
    },
    {
      "epoch": 2.77,
      "grad_norm": 4.922477437289356,
      "learning_rate": 1.47672432604537e-07,
      "loss": 1.1891,
      "step": 1279
    },
    {
      "epoch": 2.77,
      "grad_norm": 5.303758551591483,
      "learning_rate": 1.4485995857457246e-07,
      "loss": 0.9994,
      "step": 1280
    },
    {
      "epoch": 2.78,
      "grad_norm": 3.8089212575591618,
      "learning_rate": 1.420741312642282e-07,
      "loss": 0.8304,
      "step": 1281
    },
    {
      "epoch": 2.78,
      "grad_norm": 3.6408858341116432,
      "learning_rate": 1.3931496596310545e-07,
      "loss": 0.6992,
      "step": 1282
    },
    {
      "epoch": 2.78,
      "grad_norm": 3.097620868700892,
      "learning_rate": 1.3658247781447642e-07,
      "loss": 0.7126,
      "step": 1283
    },
    {
      "epoch": 2.78,
      "grad_norm": 5.466524772583007,
      "learning_rate": 1.338766818151982e-07,
      "loss": 1.0262,
      "step": 1284
    },
    {
      "epoch": 2.78,
      "grad_norm": 4.561900455757345,
      "learning_rate": 1.3119759281563392e-07,
      "loss": 0.8549,
      "step": 1285
    },
    {
      "epoch": 2.79,
      "grad_norm": 2.995616566062134,
      "learning_rate": 1.2854522551956738e-07,
      "loss": 0.7228,
      "step": 1286
    },
    {
      "epoch": 2.79,
      "grad_norm": 6.792315169637671,
      "learning_rate": 1.2591959448412628e-07,
      "loss": 0.8102,
      "step": 1287
    },
    {
      "epoch": 2.79,
      "grad_norm": 6.256449466929809,
      "learning_rate": 1.2332071411969792e-07,
      "loss": 0.6893,
      "step": 1288
    },
    {
      "epoch": 2.79,
      "grad_norm": 3.217750710603355,
      "learning_rate": 1.2074859868985377e-07,
      "loss": 0.9917,
      "step": 1289
    },
    {
      "epoch": 2.8,
      "grad_norm": 4.37944179686217,
      "learning_rate": 1.1820326231126944e-07,
      "loss": 0.6519,
      "step": 1290
    },
    {
      "epoch": 2.8,
      "grad_norm": 3.749122017248098,
      "learning_rate": 1.1568471895364863e-07,
      "loss": 0.8021,
      "step": 1291
    },
    {
      "epoch": 2.8,
      "grad_norm": 6.671608818439088,
      "learning_rate": 1.1319298243964549e-07,
      "loss": 0.7168,
      "step": 1292
    },
    {
      "epoch": 2.8,
      "grad_norm": 3.2193367735145215,
      "learning_rate": 1.107280664447874e-07,
      "loss": 0.5189,
      "step": 1293
    },
    {
      "epoch": 2.8,
      "grad_norm": 3.4775535672194033,
      "learning_rate": 1.082899844974017e-07,
      "loss": 0.4383,
      "step": 1294
    },
    {
      "epoch": 2.81,
      "grad_norm": 3.278461425789548,
      "learning_rate": 1.0587874997854186e-07,
      "loss": 0.496,
      "step": 1295
    },
    {
      "epoch": 2.81,
      "grad_norm": 3.572993178585463,
      "learning_rate": 1.0349437612191259e-07,
      "loss": 0.7011,
      "step": 1296
    },
    {
      "epoch": 2.81,
      "grad_norm": 5.664653907384666,
      "learning_rate": 1.0113687601379818e-07,
      "loss": 0.8093,
      "step": 1297
    },
    {
      "epoch": 2.81,
      "grad_norm": 5.2159070149965245,
      "learning_rate": 9.880626259298976e-08,
      "loss": 0.6524,
      "step": 1298
    },
    {
      "epoch": 2.81,
      "grad_norm": 3.10016254287963,
      "learning_rate": 9.650254865071428e-08,
      "loss": 0.6014,
      "step": 1299
    },
    {
      "epoch": 2.82,
      "grad_norm": 3.5295205642288185,
      "learning_rate": 9.422574683056795e-08,
      "loss": 0.8091,
      "step": 1300
    },
    {
      "epoch": 2.82,
      "grad_norm": 4.262415545877463,
      "learning_rate": 9.197586962843952e-08,
      "loss": 0.8392,
      "step": 1301
    },
    {
      "epoch": 2.82,
      "grad_norm": 2.5800622199414134,
      "learning_rate": 8.975292939244928e-08,
      "loss": 0.7041,
      "step": 1302
    },
    {
      "epoch": 2.82,
      "grad_norm": 3.9053847053981148,
      "learning_rate": 8.755693832287581e-08,
      "loss": 0.9696,
      "step": 1303
    },
    {
      "epoch": 2.83,
      "grad_norm": 3.9298599077517826,
      "learning_rate": 8.538790847209211e-08,
      "loss": 0.904,
      "step": 1304
    },
    {
      "epoch": 2.83,
      "grad_norm": 3.072538195183696,
      "learning_rate": 8.324585174449895e-08,
      "loss": 0.6277,
      "step": 1305
    },
    {
      "epoch": 2.83,
      "grad_norm": 4.585423590823913,
      "learning_rate": 8.11307798964589e-08,
      "loss": 0.6718,
      "step": 1306
    },
    {
      "epoch": 2.83,
      "grad_norm": 4.321505012873704,
      "learning_rate": 7.90427045362302e-08,
      "loss": 1.0418,
      "step": 1307
    },
    {
      "epoch": 2.83,
      "grad_norm": 3.0705219085812265,
      "learning_rate": 7.698163712390683e-08,
      "loss": 0.7948,
      "step": 1308
    },
    {
      "epoch": 2.84,
      "grad_norm": 3.4524691676666817,
      "learning_rate": 7.494758897135412e-08,
      "loss": 0.9096,
      "step": 1309
    },
    {
      "epoch": 2.84,
      "grad_norm": 3.522705420157837,
      "learning_rate": 7.294057124214438e-08,
      "loss": 0.8142,
      "step": 1310
    },
    {
      "epoch": 2.84,
      "grad_norm": 3.0803851855618705,
      "learning_rate": 7.096059495149855e-08,
      "loss": 0.6888,
      "step": 1311
    },
    {
      "epoch": 2.84,
      "grad_norm": 4.024131639554142,
      "learning_rate": 6.900767096622352e-08,
      "loss": 1.2485,
      "step": 1312
    },
    {
      "epoch": 2.85,
      "grad_norm": 3.9363741417911178,
      "learning_rate": 6.708181000465552e-08,
      "loss": 0.6075,
      "step": 1313
    },
    {
      "epoch": 2.85,
      "grad_norm": 3.9240914110106178,
      "learning_rate": 6.518302263659737e-08,
      "loss": 1.2087,
      "step": 1314
    },
    {
      "epoch": 2.85,
      "grad_norm": 2.9278715467335825,
      "learning_rate": 6.331131928326407e-08,
      "loss": 0.4839,
      "step": 1315
    },
    {
      "epoch": 2.85,
      "grad_norm": 4.816353315198928,
      "learning_rate": 6.146671021722284e-08,
      "loss": 1.3021,
      "step": 1316
    },
    {
      "epoch": 2.85,
      "grad_norm": 6.177182623144272,
      "learning_rate": 5.964920556233767e-08,
      "loss": 1.2263,
      "step": 1317
    },
    {
      "epoch": 2.86,
      "grad_norm": 1.9137056853114345,
      "learning_rate": 5.785881529371595e-08,
      "loss": 0.2995,
      "step": 1318
    },
    {
      "epoch": 2.86,
      "grad_norm": 5.548693942447032,
      "learning_rate": 5.609554923764915e-08,
      "loss": 0.7514,
      "step": 1319
    },
    {
      "epoch": 2.86,
      "grad_norm": 3.9389860455667853,
      "learning_rate": 5.435941707156389e-08,
      "loss": 0.8312,
      "step": 1320
    },
    {
      "epoch": 2.86,
      "grad_norm": 4.34630801963709,
      "learning_rate": 5.265042832396428e-08,
      "loss": 1.064,
      "step": 1321
    },
    {
      "epoch": 2.86,
      "grad_norm": 4.119427347321792,
      "learning_rate": 5.0968592374384116e-08,
      "loss": 1.1361,
      "step": 1322
    },
    {
      "epoch": 2.87,
      "grad_norm": 4.8740791376287484,
      "learning_rate": 4.931391845333089e-08,
      "loss": 0.5424,
      "step": 1323
    },
    {
      "epoch": 2.87,
      "grad_norm": 3.6454829156327113,
      "learning_rate": 4.768641564223852e-08,
      "loss": 0.9653,
      "step": 1324
    },
    {
      "epoch": 2.87,
      "grad_norm": 4.529939847842828,
      "learning_rate": 4.608609287341581e-08,
      "loss": 0.8173,
      "step": 1325
    },
    {
      "epoch": 2.87,
      "grad_norm": 4.104442798370329,
      "learning_rate": 4.451295892999863e-08,
      "loss": 0.8131,
      "step": 1326
    },
    {
      "epoch": 2.88,
      "grad_norm": 5.309105217036068,
      "learning_rate": 4.296702244590056e-08,
      "loss": 1.0391,
      "step": 1327
    },
    {
      "epoch": 2.88,
      "grad_norm": 3.4755981904081357,
      "learning_rate": 4.144829190576516e-08,
      "loss": 0.5436,
      "step": 1328
    },
    {
      "epoch": 2.88,
      "grad_norm": 4.871697098270554,
      "learning_rate": 3.99567756449204e-08,
      "loss": 0.5537,
      "step": 1329
    },
    {
      "epoch": 2.88,
      "grad_norm": 4.773660781790573,
      "learning_rate": 3.84924818493343e-08,
      "loss": 0.7863,
      "step": 1330
    },
    {
      "epoch": 2.88,
      "grad_norm": 5.312901342161548,
      "learning_rate": 3.705541855556716e-08,
      "loss": 0.6786,
      "step": 1331
    },
    {
      "epoch": 2.89,
      "grad_norm": 3.105647346956529,
      "learning_rate": 3.5645593650728284e-08,
      "loss": 0.5748,
      "step": 1332
    },
    {
      "epoch": 2.89,
      "grad_norm": 3.6237297262895427,
      "learning_rate": 3.426301487243433e-08,
      "loss": 0.9587,
      "step": 1333
    },
    {
      "epoch": 2.89,
      "grad_norm": 3.1071903628455835,
      "learning_rate": 3.290768980876324e-08,
      "loss": 0.8411,
      "step": 1334
    },
    {
      "epoch": 2.89,
      "grad_norm": 9.627139007602706,
      "learning_rate": 3.157962589821872e-08,
      "loss": 1.324,
      "step": 1335
    },
    {
      "epoch": 2.89,
      "grad_norm": 3.378481281392402,
      "learning_rate": 3.027883042968249e-08,
      "loss": 0.9775,
      "step": 1336
    },
    {
      "epoch": 2.9,
      "grad_norm": 3.160057721844543,
      "learning_rate": 2.9005310542378205e-08,
      "loss": 1.1101,
      "step": 1337
    },
    {
      "epoch": 2.9,
      "grad_norm": 3.562908982460253,
      "learning_rate": 2.77590732258326e-08,
      "loss": 0.826,
      "step": 1338
    },
    {
      "epoch": 2.9,
      "grad_norm": 5.3118860189745325,
      "learning_rate": 2.6540125319834964e-08,
      "loss": 1.0001,
      "step": 1339
    },
    {
      "epoch": 2.9,
      "grad_norm": 9.910506633676052,
      "learning_rate": 2.5348473514400507e-08,
      "loss": 0.6877,
      "step": 1340
    },
    {
      "epoch": 2.91,
      "grad_norm": 4.5475897635023115,
      "learning_rate": 2.4184124349734828e-08,
      "loss": 0.6111,
      "step": 1341
    },
    {
      "epoch": 2.91,
      "grad_norm": 8.133959889050892,
      "learning_rate": 2.3047084216196724e-08,
      "loss": 1.013,
      "step": 1342
    },
    {
      "epoch": 2.91,
      "grad_norm": 5.255799915405694,
      "learning_rate": 2.1937359354262665e-08,
      "loss": 0.9179,
      "step": 1343
    },
    {
      "epoch": 2.91,
      "grad_norm": 2.8603131551449446,
      "learning_rate": 2.085495585449404e-08,
      "loss": 0.8517,
      "step": 1344
    },
    {
      "epoch": 2.91,
      "grad_norm": 3.863589349327183,
      "learning_rate": 1.979987965750274e-08,
      "loss": 0.8371,
      "step": 1345
    },
    {
      "epoch": 2.92,
      "grad_norm": 3.4328310044251378,
      "learning_rate": 1.8772136553918408e-08,
      "loss": 0.9482,
      "step": 1346
    },
    {
      "epoch": 2.92,
      "grad_norm": 3.1944498547873947,
      "learning_rate": 1.7771732184357905e-08,
      "loss": 0.7028,
      "step": 1347
    },
    {
      "epoch": 2.92,
      "grad_norm": 4.8725547663292925,
      "learning_rate": 1.679867203939256e-08,
      "loss": 0.5892,
      "step": 1348
    },
    {
      "epoch": 2.92,
      "grad_norm": 3.030282750566577,
      "learning_rate": 1.5852961459519868e-08,
      "loss": 0.9101,
      "step": 1349
    },
    {
      "epoch": 2.93,
      "grad_norm": 2.792208664769457,
      "learning_rate": 1.4934605635132383e-08,
      "loss": 0.6938,
      "step": 1350
    },
    {
      "epoch": 2.93,
      "grad_norm": 4.7800584743152355,
      "learning_rate": 1.4043609606489983e-08,
      "loss": 0.4413,
      "step": 1351
    },
    {
      "epoch": 2.93,
      "grad_norm": 3.738022927358067,
      "learning_rate": 1.3179978263694326e-08,
      "loss": 0.9008,
      "step": 1352
    },
    {
      "epoch": 2.93,
      "grad_norm": 4.348293735000973,
      "learning_rate": 1.2343716346657209e-08,
      "loss": 1.1252,
      "step": 1353
    },
    {
      "epoch": 2.93,
      "grad_norm": 5.216821000218361,
      "learning_rate": 1.1534828445080027e-08,
      "loss": 0.9662,
      "step": 1354
    },
    {
      "epoch": 2.94,
      "grad_norm": 9.109515011024174,
      "learning_rate": 1.0753318998423246e-08,
      "loss": 0.833,
      "step": 1355
    },
    {
      "epoch": 2.94,
      "grad_norm": 4.9869023870412965,
      "learning_rate": 9.999192295886973e-09,
      "loss": 0.9768,
      "step": 1356
    },
    {
      "epoch": 2.94,
      "grad_norm": 3.366432836294807,
      "learning_rate": 9.272452476384308e-09,
      "loss": 0.8287,
      "step": 1357
    },
    {
      "epoch": 2.94,
      "grad_norm": 4.365739748824055,
      "learning_rate": 8.5731035285197e-09,
      "loss": 0.5413,
      "step": 1358
    },
    {
      "epoch": 2.94,
      "grad_norm": 3.7821274979507513,
      "learning_rate": 7.90114929056618e-09,
      "loss": 0.8889,
      "step": 1359
    },
    {
      "epoch": 2.95,
      "grad_norm": 6.352496070789191,
      "learning_rate": 7.256593450444827e-09,
      "loss": 0.6783,
      "step": 1360
    },
    {
      "epoch": 2.95,
      "grad_norm": 3.9962959235334927,
      "learning_rate": 6.639439545707005e-09,
      "loss": 0.5531,
      "step": 1361
    },
    {
      "epoch": 2.95,
      "grad_norm": 3.4430221911229806,
      "learning_rate": 6.04969096350938e-09,
      "loss": 0.7127,
      "step": 1362
    },
    {
      "epoch": 2.95,
      "grad_norm": 3.7550755744435125,
      "learning_rate": 5.487350940600044e-09,
      "loss": 0.6897,
      "step": 1363
    },
    {
      "epoch": 2.96,
      "grad_norm": 4.985906783417026,
      "learning_rate": 4.952422563300197e-09,
      "loss": 0.5017,
      "step": 1364
    },
    {
      "epoch": 2.96,
      "grad_norm": 3.4487418907979275,
      "learning_rate": 4.444908767484712e-09,
      "loss": 0.5373,
      "step": 1365
    },
    {
      "epoch": 2.96,
      "grad_norm": 2.7522394305722924,
      "learning_rate": 3.964812338567714e-09,
      "loss": 0.3756,
      "step": 1366
    },
    {
      "epoch": 2.96,
      "grad_norm": 3.6938438989662785,
      "learning_rate": 3.5121359114886898e-09,
      "loss": 0.851,
      "step": 1367
    },
    {
      "epoch": 2.96,
      "grad_norm": 6.916989315725603,
      "learning_rate": 3.0868819706947327e-09,
      "loss": 1.0307,
      "step": 1368
    },
    {
      "epoch": 2.97,
      "grad_norm": 2.2889284338065385,
      "learning_rate": 2.6890528501288814e-09,
      "loss": 0.6268,
      "step": 1369
    },
    {
      "epoch": 2.97,
      "grad_norm": 4.16103858757101,
      "learning_rate": 2.3186507332184636e-09,
      "loss": 0.6373,
      "step": 1370
    },
    {
      "epoch": 2.97,
      "grad_norm": 3.0926001894778175,
      "learning_rate": 1.9756776528601085e-09,
      "loss": 0.4641,
      "step": 1371
    },
    {
      "epoch": 2.97,
      "grad_norm": 4.685297870340439,
      "learning_rate": 1.660135491411974e-09,
      "loss": 0.8284,
      "step": 1372
    },
    {
      "epoch": 2.98,
      "grad_norm": 5.181241960207382,
      "learning_rate": 1.3720259806793146e-09,
      "loss": 0.804,
      "step": 1373
    },
    {
      "epoch": 2.98,
      "grad_norm": 5.571173852939345,
      "learning_rate": 1.111350701909486e-09,
      "loss": 0.7632,
      "step": 1374
    },
    {
      "epoch": 2.98,
      "grad_norm": 4.327512731624748,
      "learning_rate": 8.781110857802866e-10,
      "loss": 1.2718,
      "step": 1375
    },
    {
      "epoch": 2.98,
      "grad_norm": 4.80884721638967,
      "learning_rate": 6.723084123921864e-10,
      "loss": 0.6652,
      "step": 1376
    },
    {
      "epoch": 2.98,
      "grad_norm": 4.894496313112796,
      "learning_rate": 4.939438112638861e-10,
      "loss": 1.0874,
      "step": 1377
    },
    {
      "epoch": 2.99,
      "grad_norm": 4.712465819209399,
      "learning_rate": 3.430182613223254e-10,
      "loss": 1.0413,
      "step": 1378
    },
    {
      "epoch": 2.99,
      "grad_norm": 2.9681069243272327,
      "learning_rate": 2.1953259090101708e-10,
      "loss": 0.5211,
      "step": 1379
    },
    {
      "epoch": 2.99,
      "grad_norm": 2.4429471644862324,
      "learning_rate": 1.2348747773172075e-10,
      "loss": 0.5446,
      "step": 1380
    },
    {
      "epoch": 2.99,
      "grad_norm": 4.147691975121293,
      "learning_rate": 5.488344894444275e-11,
      "loss": 0.7822,
      "step": 1381
    },
    {
      "epoch": 2.99,
      "grad_norm": 4.1122106360432875,
      "learning_rate": 1.3720881062440073e-11,
      "loss": 1.2627,
      "step": 1382
    },
    {
      "epoch": 3.0,
      "grad_norm": 10.15022130866667,
      "learning_rate": 0.0,
      "loss": 0.472,
      "step": 1383
    },
    {
      "epoch": 3.0,
      "step": 1383,
      "total_flos": 588032156729344.0,
      "train_loss": 0.8329939518342049,
      "train_runtime": 22422.0656,
      "train_samples_per_second": 3.954,
      "train_steps_per_second": 0.062
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 1383,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 200,
  "total_flos": 588032156729344.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}