{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9980787704130645,
  "eval_steps": 500,
  "global_step": 845,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 3.921568627450981e-07,
      "loss": 2.8563,
      "step": 1
    },
    {
      "epoch": 0.0,
      "learning_rate": 7.843137254901962e-07,
      "loss": 2.8115,
      "step": 2
    },
    {
      "epoch": 0.01,
      "learning_rate": 1.1764705882352942e-06,
      "loss": 3.2218,
      "step": 3
    },
    {
      "epoch": 0.01,
      "learning_rate": 1.5686274509803923e-06,
      "loss": 2.4117,
      "step": 4
    },
    {
      "epoch": 0.01,
      "learning_rate": 1.96078431372549e-06,
      "loss": 2.3211,
      "step": 5
    },
    {
      "epoch": 0.01,
      "learning_rate": 2.3529411764705885e-06,
      "loss": 4.132,
      "step": 6
    },
    {
      "epoch": 0.02,
      "learning_rate": 2.7450980392156867e-06,
      "loss": 1.6698,
      "step": 7
    },
    {
      "epoch": 0.02,
      "learning_rate": 3.1372549019607846e-06,
      "loss": 1.5128,
      "step": 8
    },
    {
      "epoch": 0.02,
      "learning_rate": 3.529411764705883e-06,
      "loss": 1.1126,
      "step": 9
    },
    {
      "epoch": 0.02,
      "learning_rate": 3.92156862745098e-06,
      "loss": 1.2495,
      "step": 10
    },
    {
      "epoch": 0.03,
      "learning_rate": 4.313725490196079e-06,
      "loss": 1.2488,
      "step": 11
    },
    {
      "epoch": 0.03,
      "learning_rate": 4.705882352941177e-06,
      "loss": 0.9805,
      "step": 12
    },
    {
      "epoch": 0.03,
      "learning_rate": 5.098039215686274e-06,
      "loss": 1.4122,
      "step": 13
    },
    {
      "epoch": 0.03,
      "learning_rate": 5.4901960784313735e-06,
      "loss": 1.2458,
      "step": 14
    },
    {
      "epoch": 0.04,
      "learning_rate": 5.882352941176471e-06,
      "loss": 1.1085,
      "step": 15
    },
    {
      "epoch": 0.04,
      "learning_rate": 6.274509803921569e-06,
      "loss": 1.2776,
      "step": 16
    },
    {
      "epoch": 0.04,
      "learning_rate": 6.666666666666667e-06,
      "loss": 1.1455,
      "step": 17
    },
    {
      "epoch": 0.04,
      "learning_rate": 7.058823529411766e-06,
      "loss": 2.7929,
      "step": 18
    },
    {
      "epoch": 0.04,
      "learning_rate": 7.450980392156863e-06,
      "loss": 1.0727,
      "step": 19
    },
    {
      "epoch": 0.05,
      "learning_rate": 7.84313725490196e-06,
      "loss": 1.0432,
      "step": 20
    },
    {
      "epoch": 0.05,
      "learning_rate": 8.23529411764706e-06,
      "loss": 1.3406,
      "step": 21
    },
    {
      "epoch": 0.05,
      "learning_rate": 8.627450980392157e-06,
      "loss": 1.2187,
      "step": 22
    },
    {
      "epoch": 0.05,
      "learning_rate": 9.019607843137256e-06,
      "loss": 1.2385,
      "step": 23
    },
    {
      "epoch": 0.06,
      "learning_rate": 9.411764705882354e-06,
      "loss": 1.0816,
      "step": 24
    },
    {
      "epoch": 0.06,
      "learning_rate": 9.803921568627451e-06,
      "loss": 1.2362,
      "step": 25
    },
    {
      "epoch": 0.06,
      "learning_rate": 1.0196078431372549e-05,
      "loss": 1.0961,
      "step": 26
    },
    {
      "epoch": 0.06,
      "learning_rate": 1.0588235294117648e-05,
      "loss": 3.3536,
      "step": 27
    },
    {
      "epoch": 0.07,
      "learning_rate": 1.0980392156862747e-05,
      "loss": 1.2555,
      "step": 28
    },
    {
      "epoch": 0.07,
      "learning_rate": 1.1372549019607844e-05,
      "loss": 1.2809,
      "step": 29
    },
    {
      "epoch": 0.07,
      "learning_rate": 1.1764705882352942e-05,
      "loss": 1.112,
      "step": 30
    },
    {
      "epoch": 0.07,
      "learning_rate": 1.215686274509804e-05,
      "loss": 0.9778,
      "step": 31
    },
    {
      "epoch": 0.08,
      "learning_rate": 1.2549019607843138e-05,
      "loss": 1.0019,
      "step": 32
    },
    {
      "epoch": 0.08,
      "learning_rate": 1.2941176470588238e-05,
      "loss": 1.111,
      "step": 33
    },
    {
      "epoch": 0.08,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 1.6903,
      "step": 34
    },
    {
      "epoch": 0.08,
      "learning_rate": 1.3725490196078432e-05,
      "loss": 1.3078,
      "step": 35
    },
    {
      "epoch": 0.09,
      "learning_rate": 1.4117647058823532e-05,
      "loss": 1.1624,
      "step": 36
    },
    {
      "epoch": 0.09,
      "learning_rate": 1.4509803921568629e-05,
      "loss": 1.0753,
      "step": 37
    },
    {
      "epoch": 0.09,
      "learning_rate": 1.4901960784313726e-05,
      "loss": 1.0034,
      "step": 38
    },
    {
      "epoch": 0.09,
      "learning_rate": 1.5294117647058822e-05,
      "loss": 0.9425,
      "step": 39
    },
    {
      "epoch": 0.09,
      "learning_rate": 1.568627450980392e-05,
      "loss": 1.1524,
      "step": 40
    },
    {
      "epoch": 0.1,
      "learning_rate": 1.607843137254902e-05,
      "loss": 1.4117,
      "step": 41
    },
    {
      "epoch": 0.1,
      "learning_rate": 1.647058823529412e-05,
      "loss": 0.912,
      "step": 42
    },
    {
      "epoch": 0.1,
      "learning_rate": 1.686274509803922e-05,
      "loss": 1.1033,
      "step": 43
    },
    {
      "epoch": 0.1,
      "learning_rate": 1.7254901960784314e-05,
      "loss": 0.8739,
      "step": 44
    },
    {
      "epoch": 0.11,
      "learning_rate": 1.7647058823529414e-05,
      "loss": 1.0451,
      "step": 45
    },
    {
      "epoch": 0.11,
      "learning_rate": 1.8039215686274513e-05,
      "loss": 0.9691,
      "step": 46
    },
    {
      "epoch": 0.11,
      "learning_rate": 1.843137254901961e-05,
      "loss": 1.0241,
      "step": 47
    },
    {
      "epoch": 0.11,
      "learning_rate": 1.8823529411764708e-05,
      "loss": 1.008,
      "step": 48
    },
    {
      "epoch": 0.12,
      "learning_rate": 1.9215686274509807e-05,
      "loss": 1.23,
      "step": 49
    },
    {
      "epoch": 0.12,
      "learning_rate": 1.9607843137254903e-05,
      "loss": 1.0657,
      "step": 50
    },
    {
      "epoch": 0.12,
      "learning_rate": 2e-05,
      "loss": 0.9921,
      "step": 51
    },
    {
      "epoch": 0.12,
      "learning_rate": 1.9987782529016497e-05,
      "loss": 1.2391,
      "step": 52
    },
    {
      "epoch": 0.13,
      "learning_rate": 1.997556505803299e-05,
      "loss": 1.0503,
      "step": 53
    },
    {
      "epoch": 0.13,
      "learning_rate": 1.9963347587049484e-05,
      "loss": 1.0485,
      "step": 54
    },
    {
      "epoch": 0.13,
      "learning_rate": 1.9951130116065975e-05,
      "loss": 1.015,
      "step": 55
    },
    {
      "epoch": 0.13,
      "learning_rate": 1.993891264508247e-05,
      "loss": 0.9249,
      "step": 56
    },
    {
      "epoch": 0.13,
      "learning_rate": 1.9926695174098962e-05,
      "loss": 0.9354,
      "step": 57
    },
    {
      "epoch": 0.14,
      "learning_rate": 1.9914477703115457e-05,
      "loss": 1.0506,
      "step": 58
    },
    {
      "epoch": 0.14,
      "learning_rate": 1.9902260232131952e-05,
      "loss": 1.1836,
      "step": 59
    },
    {
      "epoch": 0.14,
      "learning_rate": 1.9890042761148444e-05,
      "loss": 1.1941,
      "step": 60
    },
    {
      "epoch": 0.14,
      "learning_rate": 1.987782529016494e-05,
      "loss": 1.0519,
      "step": 61
    },
    {
      "epoch": 0.15,
      "learning_rate": 1.986560781918143e-05,
      "loss": 1.1899,
      "step": 62
    },
    {
      "epoch": 0.15,
      "learning_rate": 1.9853390348197926e-05,
      "loss": 1.0649,
      "step": 63
    },
    {
      "epoch": 0.15,
      "learning_rate": 1.9841172877214418e-05,
      "loss": 1.2611,
      "step": 64
    },
    {
      "epoch": 0.15,
      "learning_rate": 1.9828955406230913e-05,
      "loss": 1.352,
      "step": 65
    },
    {
      "epoch": 0.16,
      "learning_rate": 1.9816737935247404e-05,
      "loss": 1.0917,
      "step": 66
    },
    {
      "epoch": 0.16,
      "learning_rate": 1.98045204642639e-05,
      "loss": 1.0388,
      "step": 67
    },
    {
      "epoch": 0.16,
      "learning_rate": 1.979230299328039e-05,
      "loss": 1.0571,
      "step": 68
    },
    {
      "epoch": 0.16,
      "learning_rate": 1.9780085522296886e-05,
      "loss": 1.3069,
      "step": 69
    },
    {
      "epoch": 0.17,
      "learning_rate": 1.9767868051313378e-05,
      "loss": 1.8331,
      "step": 70
    },
    {
      "epoch": 0.17,
      "learning_rate": 1.9755650580329873e-05,
      "loss": 0.9785,
      "step": 71
    },
    {
      "epoch": 0.17,
      "learning_rate": 1.9743433109346365e-05,
      "loss": 1.1313,
      "step": 72
    },
    {
      "epoch": 0.17,
      "learning_rate": 1.973121563836286e-05,
      "loss": 0.9897,
      "step": 73
    },
    {
      "epoch": 0.17,
      "learning_rate": 1.9718998167379355e-05,
      "loss": 1.2494,
      "step": 74
    },
    {
      "epoch": 0.18,
      "learning_rate": 1.970678069639585e-05,
      "loss": 0.9476,
      "step": 75
    },
    {
      "epoch": 0.18,
      "learning_rate": 1.9694563225412342e-05,
      "loss": 1.2027,
      "step": 76
    },
    {
      "epoch": 0.18,
      "learning_rate": 1.9682345754428837e-05,
      "loss": 1.2106,
      "step": 77
    },
    {
      "epoch": 0.18,
      "learning_rate": 1.967012828344533e-05,
      "loss": 0.9073,
      "step": 78
    },
    {
      "epoch": 0.19,
      "learning_rate": 1.9657910812461824e-05,
      "loss": 1.1522,
      "step": 79
    },
    {
      "epoch": 0.19,
      "learning_rate": 1.9645693341478315e-05,
      "loss": 1.0476,
      "step": 80
    },
    {
      "epoch": 0.19,
      "learning_rate": 1.963347587049481e-05,
      "loss": 0.9493,
      "step": 81
    },
    {
      "epoch": 0.19,
      "learning_rate": 1.9621258399511302e-05,
      "loss": 0.9964,
      "step": 82
    },
    {
      "epoch": 0.2,
      "learning_rate": 1.9609040928527797e-05,
      "loss": 0.8192,
      "step": 83
    },
    {
      "epoch": 0.2,
      "learning_rate": 1.959682345754429e-05,
      "loss": 1.1224,
      "step": 84
    },
    {
      "epoch": 0.2,
      "learning_rate": 1.9584605986560784e-05,
      "loss": 1.0489,
      "step": 85
    },
    {
      "epoch": 0.2,
      "learning_rate": 1.9572388515577276e-05,
      "loss": 1.0041,
      "step": 86
    },
    {
      "epoch": 0.21,
      "learning_rate": 1.956017104459377e-05,
      "loss": 0.9065,
      "step": 87
    },
    {
      "epoch": 0.21,
      "learning_rate": 1.9547953573610263e-05,
      "loss": 1.7173,
      "step": 88
    },
    {
      "epoch": 0.21,
      "learning_rate": 1.9535736102626758e-05,
      "loss": 1.2975,
      "step": 89
    },
    {
      "epoch": 0.21,
      "learning_rate": 1.9523518631643253e-05,
      "loss": 0.8456,
      "step": 90
    },
    {
      "epoch": 0.22,
      "learning_rate": 1.9511301160659744e-05,
      "loss": 1.1463,
      "step": 91
    },
    {
      "epoch": 0.22,
      "learning_rate": 1.949908368967624e-05,
      "loss": 1.1324,
      "step": 92
    },
    {
      "epoch": 0.22,
      "learning_rate": 1.948686621869273e-05,
      "loss": 1.0128,
      "step": 93
    },
    {
      "epoch": 0.22,
      "learning_rate": 1.9474648747709226e-05,
      "loss": 1.0593,
      "step": 94
    },
    {
      "epoch": 0.22,
      "learning_rate": 1.9462431276725718e-05,
      "loss": 0.9871,
      "step": 95
    },
    {
      "epoch": 0.23,
      "learning_rate": 1.9450213805742213e-05,
      "loss": 1.3373,
      "step": 96
    },
    {
      "epoch": 0.23,
      "learning_rate": 1.9437996334758705e-05,
      "loss": 1.1009,
      "step": 97
    },
    {
      "epoch": 0.23,
      "learning_rate": 1.94257788637752e-05,
      "loss": 1.1652,
      "step": 98
    },
    {
      "epoch": 0.23,
      "learning_rate": 1.941356139279169e-05,
      "loss": 0.855,
      "step": 99
    },
    {
      "epoch": 0.24,
      "learning_rate": 1.9401343921808187e-05,
      "loss": 1.1619,
      "step": 100
    },
    {
      "epoch": 0.24,
      "learning_rate": 1.938912645082468e-05,
      "loss": 0.883,
      "step": 101
    },
    {
      "epoch": 0.24,
      "learning_rate": 1.9376908979841174e-05,
      "loss": 1.2267,
      "step": 102
    },
    {
      "epoch": 0.24,
      "learning_rate": 1.936469150885767e-05,
      "loss": 0.8898,
      "step": 103
    },
    {
      "epoch": 0.25,
      "learning_rate": 1.9352474037874164e-05,
      "loss": 0.9406,
      "step": 104
    },
    {
      "epoch": 0.25,
      "learning_rate": 1.9340256566890655e-05,
      "loss": 1.0517,
      "step": 105
    },
    {
      "epoch": 0.25,
      "learning_rate": 1.932803909590715e-05,
      "loss": 1.0964,
      "step": 106
    },
    {
      "epoch": 0.25,
      "learning_rate": 1.9315821624923642e-05,
      "loss": 0.9225,
      "step": 107
    },
    {
      "epoch": 0.26,
      "learning_rate": 1.9303604153940137e-05,
      "loss": 1.0517,
      "step": 108
    },
    {
      "epoch": 0.26,
      "learning_rate": 1.929138668295663e-05,
      "loss": 1.0929,
      "step": 109
    },
    {
      "epoch": 0.26,
      "learning_rate": 1.9279169211973124e-05,
      "loss": 1.0169,
      "step": 110
    },
    {
      "epoch": 0.26,
      "learning_rate": 1.9266951740989616e-05,
      "loss": 0.9908,
      "step": 111
    },
    {
      "epoch": 0.26,
      "learning_rate": 1.925473427000611e-05,
      "loss": 0.9025,
      "step": 112
    },
    {
      "epoch": 0.27,
      "learning_rate": 1.9242516799022603e-05,
      "loss": 1.2115,
      "step": 113
    },
    {
      "epoch": 0.27,
      "learning_rate": 1.9230299328039098e-05,
      "loss": 1.0489,
      "step": 114
    },
    {
      "epoch": 0.27,
      "learning_rate": 1.921808185705559e-05,
      "loss": 0.9932,
      "step": 115
    },
    {
      "epoch": 0.27,
      "learning_rate": 1.9205864386072085e-05,
      "loss": 1.0111,
      "step": 116
    },
    {
      "epoch": 0.28,
      "learning_rate": 1.9193646915088576e-05,
      "loss": 1.4229,
      "step": 117
    },
    {
      "epoch": 0.28,
      "learning_rate": 1.918142944410507e-05,
      "loss": 1.1299,
      "step": 118
    },
    {
      "epoch": 0.28,
      "learning_rate": 1.9169211973121567e-05,
      "loss": 1.1633,
      "step": 119
    },
    {
      "epoch": 0.28,
      "learning_rate": 1.9156994502138058e-05,
      "loss": 0.9408,
      "step": 120
    },
    {
      "epoch": 0.29,
      "learning_rate": 1.9144777031154553e-05,
      "loss": 1.0087,
      "step": 121
    },
    {
      "epoch": 0.29,
      "learning_rate": 1.913255956017105e-05,
      "loss": 0.9289,
      "step": 122
    },
    {
      "epoch": 0.29,
      "learning_rate": 1.912034208918754e-05,
      "loss": 1.1067,
      "step": 123
    },
    {
      "epoch": 0.29,
      "learning_rate": 1.9108124618204035e-05,
      "loss": 1.2101,
      "step": 124
    },
    {
      "epoch": 0.3,
      "learning_rate": 1.9095907147220527e-05,
      "loss": 0.9399,
      "step": 125
    },
    {
      "epoch": 0.3,
      "learning_rate": 1.9083689676237022e-05,
      "loss": 1.0829,
      "step": 126
    },
    {
      "epoch": 0.3,
      "learning_rate": 1.9071472205253514e-05,
      "loss": 1.1841,
      "step": 127
    },
    {
      "epoch": 0.3,
      "learning_rate": 1.905925473427001e-05,
      "loss": 0.9907,
      "step": 128
    },
    {
      "epoch": 0.31,
      "learning_rate": 1.90470372632865e-05,
      "loss": 1.0977,
      "step": 129
    },
    {
      "epoch": 0.31,
      "learning_rate": 1.9034819792302996e-05,
      "loss": 1.0854,
      "step": 130
    },
    {
      "epoch": 0.31,
      "learning_rate": 1.9022602321319487e-05,
      "loss": 0.9287,
      "step": 131
    },
    {
      "epoch": 0.31,
      "learning_rate": 1.9010384850335982e-05,
      "loss": 0.7713,
      "step": 132
    },
    {
      "epoch": 0.31,
      "learning_rate": 1.8998167379352474e-05,
      "loss": 0.8631,
      "step": 133
    },
    {
      "epoch": 0.32,
      "learning_rate": 1.898594990836897e-05,
      "loss": 0.9045,
      "step": 134
    },
    {
      "epoch": 0.32,
      "learning_rate": 1.8973732437385464e-05,
      "loss": 0.7478,
      "step": 135
    },
    {
      "epoch": 0.32,
      "learning_rate": 1.8961514966401956e-05,
      "loss": 1.2063,
      "step": 136
    },
    {
      "epoch": 0.32,
      "learning_rate": 1.894929749541845e-05,
      "loss": 0.9594,
      "step": 137
    },
    {
      "epoch": 0.33,
      "learning_rate": 1.8937080024434943e-05,
      "loss": 1.0198,
      "step": 138
    },
    {
      "epoch": 0.33,
      "learning_rate": 1.8924862553451438e-05,
      "loss": 0.9273,
      "step": 139
    },
    {
      "epoch": 0.33,
      "learning_rate": 1.891264508246793e-05,
      "loss": 0.8572,
      "step": 140
    },
    {
      "epoch": 0.33,
      "learning_rate": 1.8900427611484425e-05,
      "loss": 1.0327,
      "step": 141
    },
    {
      "epoch": 0.34,
      "learning_rate": 1.8888210140500916e-05,
      "loss": 0.7909,
      "step": 142
    },
    {
      "epoch": 0.34,
      "learning_rate": 1.887599266951741e-05,
      "loss": 0.9771,
      "step": 143
    },
    {
      "epoch": 0.34,
      "learning_rate": 1.8863775198533903e-05,
      "loss": 1.7899,
      "step": 144
    },
    {
      "epoch": 0.34,
      "learning_rate": 1.88515577275504e-05,
      "loss": 0.8969,
      "step": 145
    },
    {
      "epoch": 0.35,
      "learning_rate": 1.883934025656689e-05,
      "loss": 0.9746,
      "step": 146
    },
    {
      "epoch": 0.35,
      "learning_rate": 1.8827122785583385e-05,
      "loss": 0.7487,
      "step": 147
    },
    {
      "epoch": 0.35,
      "learning_rate": 1.8814905314599877e-05,
      "loss": 0.9903,
      "step": 148
    },
    {
      "epoch": 0.35,
      "learning_rate": 1.8802687843616375e-05,
      "loss": 0.9306,
      "step": 149
    },
    {
      "epoch": 0.35,
      "learning_rate": 1.8790470372632867e-05,
      "loss": 0.8955,
      "step": 150
    },
    {
      "epoch": 0.36,
      "learning_rate": 1.8778252901649362e-05,
      "loss": 0.8673,
      "step": 151
    },
    {
      "epoch": 0.36,
      "learning_rate": 1.8766035430665854e-05,
      "loss": 1.12,
      "step": 152
    },
    {
      "epoch": 0.36,
      "learning_rate": 1.875381795968235e-05,
      "loss": 0.7883,
      "step": 153
    },
    {
      "epoch": 0.36,
      "learning_rate": 1.874160048869884e-05,
      "loss": 0.9563,
      "step": 154
    },
    {
      "epoch": 0.37,
      "learning_rate": 1.8729383017715336e-05,
      "loss": 1.0749,
      "step": 155
    },
    {
      "epoch": 0.37,
      "learning_rate": 1.8717165546731827e-05,
      "loss": 1.2149,
      "step": 156
    },
    {
      "epoch": 0.37,
      "learning_rate": 1.8704948075748323e-05,
      "loss": 1.4591,
      "step": 157
    },
    {
      "epoch": 0.37,
      "learning_rate": 1.8692730604764814e-05,
      "loss": 1.2051,
      "step": 158
    },
    {
      "epoch": 0.38,
      "learning_rate": 1.868051313378131e-05,
      "loss": 0.9812,
      "step": 159
    },
    {
      "epoch": 0.38,
      "learning_rate": 1.86682956627978e-05,
      "loss": 1.1543,
      "step": 160
    },
    {
      "epoch": 0.38,
      "learning_rate": 1.8656078191814296e-05,
      "loss": 1.1209,
      "step": 161
    },
    {
      "epoch": 0.38,
      "learning_rate": 1.8643860720830788e-05,
      "loss": 1.0923,
      "step": 162
    },
    {
      "epoch": 0.39,
      "learning_rate": 1.8631643249847283e-05,
      "loss": 1.0248,
      "step": 163
    },
    {
      "epoch": 0.39,
      "learning_rate": 1.8619425778863778e-05,
      "loss": 0.9989,
      "step": 164
    },
    {
      "epoch": 0.39,
      "learning_rate": 1.860720830788027e-05,
      "loss": 1.0276,
      "step": 165
    },
    {
      "epoch": 0.39,
      "learning_rate": 1.8594990836896765e-05,
      "loss": 1.0172,
      "step": 166
    },
    {
      "epoch": 0.39,
      "learning_rate": 1.8582773365913257e-05,
      "loss": 1.0659,
      "step": 167
    },
    {
      "epoch": 0.4,
      "learning_rate": 1.857055589492975e-05,
      "loss": 0.9678,
      "step": 168
    },
    {
      "epoch": 0.4,
      "learning_rate": 1.8558338423946243e-05,
      "loss": 1.0349,
      "step": 169
    },
    {
      "epoch": 0.4,
      "learning_rate": 1.854612095296274e-05,
      "loss": 1.1437,
      "step": 170
    },
    {
      "epoch": 0.4,
      "learning_rate": 1.853390348197923e-05,
      "loss": 0.9805,
      "step": 171
    },
    {
      "epoch": 0.41,
      "learning_rate": 1.8521686010995725e-05,
      "loss": 1.1063,
      "step": 172
    },
    {
      "epoch": 0.41,
      "learning_rate": 1.850946854001222e-05,
      "loss": 1.0671,
      "step": 173
    },
    {
      "epoch": 0.41,
      "learning_rate": 1.8497251069028712e-05,
      "loss": 1.0975,
      "step": 174
    },
    {
      "epoch": 0.41,
      "learning_rate": 1.8485033598045207e-05,
      "loss": 0.9989,
      "step": 175
    },
    {
      "epoch": 0.42,
      "learning_rate": 1.84728161270617e-05,
      "loss": 1.033,
      "step": 176
    },
    {
      "epoch": 0.42,
      "learning_rate": 1.8460598656078194e-05,
      "loss": 0.8797,
      "step": 177
    },
    {
      "epoch": 0.42,
      "learning_rate": 1.8448381185094686e-05,
      "loss": 1.0079,
      "step": 178
    },
    {
      "epoch": 0.42,
      "learning_rate": 1.843616371411118e-05,
      "loss": 1.0638,
      "step": 179
    },
    {
      "epoch": 0.43,
      "learning_rate": 1.8423946243127676e-05,
      "loss": 0.907,
      "step": 180
    },
    {
      "epoch": 0.43,
      "learning_rate": 1.8411728772144168e-05,
      "loss": 0.8491,
      "step": 181
    },
    {
      "epoch": 0.43,
      "learning_rate": 1.8399511301160663e-05,
      "loss": 1.0359,
      "step": 182
    },
    {
      "epoch": 0.43,
      "learning_rate": 1.8387293830177154e-05,
      "loss": 0.9798,
      "step": 183
    },
    {
      "epoch": 0.44,
      "learning_rate": 1.837507635919365e-05,
      "loss": 0.8019,
      "step": 184
    },
    {
      "epoch": 0.44,
      "learning_rate": 1.836285888821014e-05,
      "loss": 1.2839,
      "step": 185
    },
    {
      "epoch": 0.44,
      "learning_rate": 1.8350641417226636e-05,
      "loss": 1.007,
      "step": 186
    },
    {
      "epoch": 0.44,
      "learning_rate": 1.8338423946243128e-05,
      "loss": 0.8493,
      "step": 187
    },
    {
      "epoch": 0.44,
      "learning_rate": 1.8326206475259623e-05,
      "loss": 0.9633,
      "step": 188
    },
    {
      "epoch": 0.45,
      "learning_rate": 1.8313989004276115e-05,
      "loss": 0.7577,
      "step": 189
    },
    {
      "epoch": 0.45,
      "learning_rate": 1.830177153329261e-05,
      "loss": 2.1476,
      "step": 190
    },
    {
      "epoch": 0.45,
      "learning_rate": 1.82895540623091e-05,
      "loss": 0.7755,
      "step": 191
    },
    {
      "epoch": 0.45,
      "learning_rate": 1.8277336591325597e-05,
      "loss": 1.1204,
      "step": 192
    },
    {
      "epoch": 0.46,
      "learning_rate": 1.826511912034209e-05,
      "loss": 1.0931,
      "step": 193
    },
    {
      "epoch": 0.46,
      "learning_rate": 1.8252901649358587e-05,
      "loss": 0.8763,
      "step": 194
    },
    {
      "epoch": 0.46,
      "learning_rate": 1.824068417837508e-05,
      "loss": 0.9134,
      "step": 195
    },
    {
      "epoch": 0.46,
      "learning_rate": 1.8228466707391574e-05,
      "loss": 0.7959,
      "step": 196
    },
    {
      "epoch": 0.47,
      "learning_rate": 1.8216249236408065e-05,
      "loss": 0.7246,
      "step": 197
    },
    {
      "epoch": 0.47,
      "learning_rate": 1.820403176542456e-05,
      "loss": 1.1064,
      "step": 198
    },
    {
      "epoch": 0.47,
      "learning_rate": 1.8191814294441052e-05,
      "loss": 1.5641,
      "step": 199
    },
    {
      "epoch": 0.47,
      "learning_rate": 1.8179596823457547e-05,
      "loss": 1.2576,
      "step": 200
    },
    {
      "epoch": 0.48,
      "learning_rate": 1.816737935247404e-05,
      "loss": 1.1123,
      "step": 201
    },
    {
      "epoch": 0.48,
      "learning_rate": 1.8155161881490534e-05,
      "loss": 0.8261,
      "step": 202
    },
    {
      "epoch": 0.48,
      "learning_rate": 1.8142944410507026e-05,
      "loss": 1.1329,
      "step": 203
    },
    {
      "epoch": 0.48,
      "learning_rate": 1.813072693952352e-05,
      "loss": 0.7938,
      "step": 204
    },
    {
      "epoch": 0.48,
      "learning_rate": 1.8118509468540013e-05,
      "loss": 0.9276,
      "step": 205
    },
    {
      "epoch": 0.49,
      "learning_rate": 1.8106291997556508e-05,
      "loss": 0.8416,
      "step": 206
    },
    {
      "epoch": 0.49,
      "learning_rate": 1.8094074526573e-05,
      "loss": 1.0415,
      "step": 207
    },
    {
      "epoch": 0.49,
      "learning_rate": 1.8081857055589494e-05,
      "loss": 1.0821,
      "step": 208
    },
    {
      "epoch": 0.49,
      "learning_rate": 1.806963958460599e-05,
      "loss": 0.9006,
      "step": 209
    },
    {
      "epoch": 0.5,
      "learning_rate": 1.805742211362248e-05,
      "loss": 0.9101,
      "step": 210
    },
    {
      "epoch": 0.5,
      "learning_rate": 1.8045204642638976e-05,
      "loss": 1.0872,
      "step": 211
    },
    {
      "epoch": 0.5,
      "learning_rate": 1.8032987171655468e-05,
      "loss": 0.9797,
      "step": 212
    },
    {
      "epoch": 0.5,
      "learning_rate": 1.8020769700671963e-05,
      "loss": 1.0248,
      "step": 213
    },
    {
      "epoch": 0.51,
      "learning_rate": 1.8008552229688455e-05,
      "loss": 1.1993,
      "step": 214
    },
    {
      "epoch": 0.51,
      "learning_rate": 1.799633475870495e-05,
      "loss": 0.8838,
      "step": 215
    },
    {
      "epoch": 0.51,
      "learning_rate": 1.7984117287721442e-05,
      "loss": 0.8972,
      "step": 216
    },
    {
      "epoch": 0.51,
      "learning_rate": 1.7971899816737937e-05,
      "loss": 1.1412,
      "step": 217
    },
    {
      "epoch": 0.52,
      "learning_rate": 1.795968234575443e-05,
      "loss": 1.0733,
      "step": 218
    },
    {
      "epoch": 0.52,
      "learning_rate": 1.7947464874770924e-05,
      "loss": 1.0435,
      "step": 219
    },
    {
      "epoch": 0.52,
      "learning_rate": 1.7935247403787415e-05,
      "loss": 1.0862,
      "step": 220
    },
    {
      "epoch": 0.52,
      "learning_rate": 1.792302993280391e-05,
      "loss": 1.0815,
      "step": 221
    },
    {
      "epoch": 0.52,
      "learning_rate": 1.7910812461820402e-05,
      "loss": 1.1491,
      "step": 222
    },
    {
      "epoch": 0.53,
      "learning_rate": 1.7898594990836897e-05,
      "loss": 1.0378,
      "step": 223
    },
    {
      "epoch": 0.53,
      "learning_rate": 1.7886377519853392e-05,
      "loss": 1.0162,
      "step": 224
    },
    {
      "epoch": 0.53,
      "learning_rate": 1.7874160048869887e-05,
      "loss": 0.8259,
      "step": 225
    },
    {
      "epoch": 0.53,
      "learning_rate": 1.786194257788638e-05,
      "loss": 0.8914,
      "step": 226
    },
    {
      "epoch": 0.54,
      "learning_rate": 1.7849725106902874e-05,
      "loss": 0.8749,
      "step": 227
    },
    {
      "epoch": 0.54,
      "learning_rate": 1.7837507635919366e-05,
      "loss": 1.0432,
      "step": 228
    },
    {
      "epoch": 0.54,
      "learning_rate": 1.782529016493586e-05,
      "loss": 1.0119,
      "step": 229
    },
    {
      "epoch": 0.54,
      "learning_rate": 1.7813072693952353e-05,
      "loss": 1.1508,
      "step": 230
    },
    {
      "epoch": 0.55,
      "learning_rate": 1.7800855222968848e-05,
      "loss": 1.1533,
      "step": 231
    },
    {
      "epoch": 0.55,
      "learning_rate": 1.778863775198534e-05,
      "loss": 1.2029,
      "step": 232
    },
    {
      "epoch": 0.55,
      "learning_rate": 1.7776420281001835e-05,
      "loss": 1.1147,
      "step": 233
    },
    {
      "epoch": 0.55,
      "learning_rate": 1.7764202810018326e-05,
      "loss": 0.9685,
      "step": 234
    },
    {
      "epoch": 0.56,
      "learning_rate": 1.775198533903482e-05,
      "loss": 0.8694,
      "step": 235
    },
    {
      "epoch": 0.56,
      "learning_rate": 1.7739767868051313e-05,
      "loss": 0.9998,
      "step": 236
    },
    {
      "epoch": 0.56,
      "learning_rate": 1.7727550397067808e-05,
      "loss": 0.9252,
      "step": 237
    },
    {
      "epoch": 0.56,
      "learning_rate": 1.77153329260843e-05,
      "loss": 0.9679,
      "step": 238
    },
    {
      "epoch": 0.57,
      "learning_rate": 1.7703115455100795e-05,
      "loss": 1.1989,
      "step": 239
    },
    {
      "epoch": 0.57,
      "learning_rate": 1.769089798411729e-05,
      "loss": 1.0249,
      "step": 240
    },
    {
      "epoch": 0.57,
      "learning_rate": 1.7678680513133785e-05,
      "loss": 0.9713,
      "step": 241
    },
    {
      "epoch": 0.57,
      "learning_rate": 1.7666463042150277e-05,
      "loss": 1.0119,
      "step": 242
    },
    {
      "epoch": 0.57,
      "learning_rate": 1.7654245571166772e-05,
      "loss": 1.0165,
      "step": 243
    },
    {
      "epoch": 0.58,
      "learning_rate": 1.7642028100183264e-05,
      "loss": 0.848,
      "step": 244
    },
    {
      "epoch": 0.58,
      "learning_rate": 1.762981062919976e-05,
      "loss": 0.9093,
      "step": 245
    },
    {
      "epoch": 0.58,
      "learning_rate": 1.761759315821625e-05,
      "loss": 0.929,
      "step": 246
    },
    {
      "epoch": 0.58,
      "learning_rate": 1.7605375687232746e-05,
      "loss": 1.1163,
      "step": 247
    },
    {
      "epoch": 0.59,
      "learning_rate": 1.7593158216249237e-05,
      "loss": 1.4804,
      "step": 248
    },
    {
      "epoch": 0.59,
      "learning_rate": 1.7580940745265732e-05,
      "loss": 1.04,
      "step": 249
    },
    {
      "epoch": 0.59,
      "learning_rate": 1.7568723274282224e-05,
      "loss": 0.915,
      "step": 250
    },
    {
      "epoch": 0.59,
      "learning_rate": 1.755650580329872e-05,
      "loss": 1.1308,
      "step": 251
    },
    {
      "epoch": 0.6,
      "learning_rate": 1.754428833231521e-05,
      "loss": 1.029,
      "step": 252
    },
    {
      "epoch": 0.6,
      "learning_rate": 1.7532070861331706e-05,
      "loss": 1.1648,
      "step": 253
    },
    {
      "epoch": 0.6,
      "learning_rate": 1.75198533903482e-05,
      "loss": 1.0718,
      "step": 254
    },
    {
      "epoch": 0.6,
      "learning_rate": 1.7507635919364693e-05,
      "loss": 0.7947,
      "step": 255
    },
    {
      "epoch": 0.61,
      "learning_rate": 1.7495418448381188e-05,
      "loss": 1.0397,
      "step": 256
    },
    {
      "epoch": 0.61,
      "learning_rate": 1.748320097739768e-05,
      "loss": 0.7579,
      "step": 257
    },
    {
      "epoch": 0.61,
      "learning_rate": 1.7470983506414175e-05,
      "loss": 0.8241,
      "step": 258
    },
    {
      "epoch": 0.61,
      "learning_rate": 1.7458766035430666e-05,
      "loss": 1.7378,
      "step": 259
    },
    {
      "epoch": 0.61,
      "learning_rate": 1.744654856444716e-05,
      "loss": 0.9632,
      "step": 260
    },
    {
      "epoch": 0.62,
      "learning_rate": 1.7434331093463653e-05,
      "loss": 0.9171,
      "step": 261
    },
    {
      "epoch": 0.62,
      "learning_rate": 1.742211362248015e-05,
      "loss": 0.887,
      "step": 262
    },
    {
      "epoch": 0.62,
      "learning_rate": 1.740989615149664e-05,
      "loss": 1.0072,
      "step": 263
    },
    {
      "epoch": 0.62,
      "learning_rate": 1.7397678680513135e-05,
      "loss": 1.051,
      "step": 264
    },
    {
      "epoch": 0.63,
      "learning_rate": 1.7385461209529627e-05,
      "loss": 0.95,
      "step": 265
    },
    {
      "epoch": 0.63,
      "learning_rate": 1.7373243738546122e-05,
      "loss": 0.8019,
      "step": 266
    },
    {
      "epoch": 0.63,
      "learning_rate": 1.7361026267562614e-05,
      "loss": 1.014,
      "step": 267
    },
    {
      "epoch": 0.63,
      "learning_rate": 1.734880879657911e-05,
      "loss": 0.8493,
      "step": 268
    },
    {
      "epoch": 0.64,
      "learning_rate": 1.7336591325595604e-05,
      "loss": 1.2377,
      "step": 269
    },
    {
      "epoch": 0.64,
      "learning_rate": 1.73243738546121e-05,
      "loss": 0.9735,
      "step": 270
    },
    {
      "epoch": 0.64,
      "learning_rate": 1.731215638362859e-05,
      "loss": 1.1114,
      "step": 271
    },
    {
      "epoch": 0.64,
      "learning_rate": 1.7299938912645086e-05,
      "loss": 1.1334,
      "step": 272
    },
    {
      "epoch": 0.65,
      "learning_rate": 1.7287721441661577e-05,
      "loss": 0.8667,
      "step": 273
    },
    {
      "epoch": 0.65,
      "learning_rate": 1.7275503970678073e-05,
      "loss": 0.9092,
      "step": 274
    },
    {
      "epoch": 0.65,
      "learning_rate": 1.7263286499694564e-05,
      "loss": 1.004,
      "step": 275
    },
    {
      "epoch": 0.65,
      "learning_rate": 1.725106902871106e-05,
      "loss": 1.027,
      "step": 276
    },
    {
      "epoch": 0.65,
      "learning_rate": 1.723885155772755e-05,
      "loss": 0.9408,
      "step": 277
    },
    {
      "epoch": 0.66,
      "learning_rate": 1.7226634086744046e-05,
      "loss": 0.8829,
      "step": 278
    },
    {
      "epoch": 0.66,
      "learning_rate": 1.7214416615760538e-05,
      "loss": 1.3557,
      "step": 279
    },
    {
      "epoch": 0.66,
      "learning_rate": 1.7202199144777033e-05,
      "loss": 0.9898,
      "step": 280
    },
    {
      "epoch": 0.66,
      "learning_rate": 1.7189981673793525e-05,
      "loss": 0.9547,
      "step": 281
    },
    {
      "epoch": 0.67,
      "learning_rate": 1.717776420281002e-05,
      "loss": 0.9013,
      "step": 282
    },
    {
      "epoch": 0.67,
      "learning_rate": 1.716554673182651e-05,
      "loss": 0.9221,
      "step": 283
    },
    {
      "epoch": 0.67,
      "learning_rate": 1.7153329260843007e-05,
      "loss": 0.9773,
      "step": 284
    },
    {
      "epoch": 0.67,
      "learning_rate": 1.71411117898595e-05,
      "loss": 0.9181,
      "step": 285
    },
    {
      "epoch": 0.68,
      "learning_rate": 1.7128894318875993e-05,
      "loss": 1.0452,
      "step": 286
    },
    {
      "epoch": 0.68,
      "learning_rate": 1.711667684789249e-05,
      "loss": 1.0132,
      "step": 287
    },
    {
      "epoch": 0.68,
      "learning_rate": 1.710445937690898e-05,
      "loss": 0.8378,
      "step": 288
    },
    {
      "epoch": 0.68,
      "learning_rate": 1.7092241905925475e-05,
      "loss": 0.9312,
      "step": 289
    },
    {
      "epoch": 0.69,
      "learning_rate": 1.7080024434941967e-05,
      "loss": 0.863,
      "step": 290
    },
    {
      "epoch": 0.69,
      "learning_rate": 1.7067806963958462e-05,
      "loss": 1.1051,
      "step": 291
    },
    {
      "epoch": 0.69,
      "learning_rate": 1.7055589492974954e-05,
      "loss": 0.8351,
      "step": 292
    },
    {
      "epoch": 0.69,
      "learning_rate": 1.704337202199145e-05,
      "loss": 1.1086,
      "step": 293
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.7031154551007944e-05,
      "loss": 0.9148,
      "step": 294
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.7018937080024436e-05,
      "loss": 1.0763,
      "step": 295
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.700671960904093e-05,
      "loss": 0.9651,
      "step": 296
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.6994502138057422e-05,
      "loss": 0.7762,
      "step": 297
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.6982284667073918e-05,
      "loss": 0.8788,
      "step": 298
    },
    {
      "epoch": 0.71,
      "learning_rate": 1.6970067196090413e-05,
      "loss": 0.9629,
      "step": 299
    },
    {
      "epoch": 0.71,
      "learning_rate": 1.6957849725106904e-05,
      "loss": 0.9064,
      "step": 300
    },
    {
      "epoch": 0.71,
      "learning_rate": 1.69456322541234e-05,
      "loss": 0.9137,
      "step": 301
    },
    {
      "epoch": 0.71,
      "learning_rate": 1.693341478313989e-05,
      "loss": 1.0518,
      "step": 302
    },
    {
      "epoch": 0.72,
      "learning_rate": 1.6921197312156386e-05,
      "loss": 1.1436,
      "step": 303
    },
    {
      "epoch": 0.72,
      "learning_rate": 1.6908979841172878e-05,
      "loss": 1.063,
      "step": 304
    },
    {
      "epoch": 0.72,
      "learning_rate": 1.6896762370189373e-05,
      "loss": 0.8848,
      "step": 305
    },
    {
      "epoch": 0.72,
      "learning_rate": 1.6884544899205865e-05,
      "loss": 0.7027,
      "step": 306
    },
    {
      "epoch": 0.73,
      "learning_rate": 1.687232742822236e-05,
      "loss": 0.929,
      "step": 307
    },
    {
      "epoch": 0.73,
      "learning_rate": 1.686010995723885e-05,
      "loss": 0.9704,
      "step": 308
    },
    {
      "epoch": 0.73,
      "learning_rate": 1.6847892486255347e-05,
      "loss": 0.8492,
      "step": 309
    },
    {
      "epoch": 0.73,
      "learning_rate": 1.683567501527184e-05,
      "loss": 0.8724,
      "step": 310
    },
    {
      "epoch": 0.74,
      "learning_rate": 1.6823457544288334e-05,
      "loss": 0.8877,
      "step": 311
    },
    {
      "epoch": 0.74,
      "learning_rate": 1.6811240073304825e-05,
      "loss": 0.9331,
      "step": 312
    },
    {
      "epoch": 0.74,
      "learning_rate": 1.679902260232132e-05,
      "loss": 0.9574,
      "step": 313
    },
    {
      "epoch": 0.74,
      "learning_rate": 1.6786805131337815e-05,
      "loss": 1.0783,
      "step": 314
    },
    {
      "epoch": 0.74,
      "learning_rate": 1.677458766035431e-05,
      "loss": 0.9318,
      "step": 315
    },
    {
      "epoch": 0.75,
      "learning_rate": 1.6762370189370802e-05,
      "loss": 0.783,
      "step": 316
    },
    {
      "epoch": 0.75,
      "learning_rate": 1.6750152718387297e-05,
      "loss": 0.8419,
      "step": 317
    },
    {
      "epoch": 0.75,
      "learning_rate": 1.673793524740379e-05,
      "loss": 1.1953,
      "step": 318
    },
    {
      "epoch": 0.75,
      "learning_rate": 1.6725717776420284e-05,
      "loss": 0.9503,
      "step": 319
    },
    {
      "epoch": 0.76,
      "learning_rate": 1.6713500305436776e-05,
      "loss": 0.7846,
      "step": 320
    },
    {
      "epoch": 0.76,
      "learning_rate": 1.670128283445327e-05,
      "loss": 0.949,
      "step": 321
    },
    {
      "epoch": 0.76,
      "learning_rate": 1.6689065363469763e-05,
      "loss": 1.0251,
      "step": 322
    },
    {
      "epoch": 0.76,
      "learning_rate": 1.6676847892486258e-05,
      "loss": 1.1247,
      "step": 323
    },
    {
      "epoch": 0.77,
      "learning_rate": 1.666463042150275e-05,
      "loss": 0.9591,
      "step": 324
    },
    {
      "epoch": 0.77,
      "learning_rate": 1.6652412950519245e-05,
      "loss": 0.9174,
      "step": 325
    },
    {
      "epoch": 0.77,
      "learning_rate": 1.6640195479535736e-05,
      "loss": 0.8984,
      "step": 326
    },
    {
      "epoch": 0.77,
      "learning_rate": 1.662797800855223e-05,
      "loss": 0.8926,
      "step": 327
    },
    {
      "epoch": 0.78,
      "learning_rate": 1.6615760537568723e-05,
      "loss": 0.9084,
      "step": 328
    },
    {
      "epoch": 0.78,
      "learning_rate": 1.6603543066585218e-05,
      "loss": 0.9717,
      "step": 329
    },
    {
      "epoch": 0.78,
      "learning_rate": 1.6591325595601713e-05,
      "loss": 1.2356,
      "step": 330
    },
    {
      "epoch": 0.78,
      "learning_rate": 1.6579108124618205e-05,
      "loss": 0.8197,
      "step": 331
    },
    {
      "epoch": 0.79,
      "learning_rate": 1.65668906536347e-05,
      "loss": 1.1189,
      "step": 332
    },
    {
      "epoch": 0.79,
      "learning_rate": 1.6554673182651192e-05,
      "loss": 1.0264,
      "step": 333
    },
    {
      "epoch": 0.79,
      "learning_rate": 1.6542455711667687e-05,
      "loss": 1.0084,
      "step": 334
    },
    {
      "epoch": 0.79,
      "learning_rate": 1.653023824068418e-05,
      "loss": 1.0175,
      "step": 335
    },
    {
      "epoch": 0.79,
      "learning_rate": 1.6518020769700674e-05,
      "loss": 0.8441,
      "step": 336
    },
    {
      "epoch": 0.8,
      "learning_rate": 1.6505803298717165e-05,
      "loss": 1.2329,
      "step": 337
    },
    {
      "epoch": 0.8,
      "learning_rate": 1.649358582773366e-05,
      "loss": 0.994,
      "step": 338
    },
    {
      "epoch": 0.8,
      "learning_rate": 1.6481368356750152e-05,
      "loss": 1.2737,
      "step": 339
    },
    {
      "epoch": 0.8,
      "learning_rate": 1.6469150885766647e-05,
      "loss": 1.2804,
      "step": 340
    },
    {
      "epoch": 0.81,
      "learning_rate": 1.645693341478314e-05,
      "loss": 0.7583,
      "step": 341
    },
    {
      "epoch": 0.81,
      "learning_rate": 1.6444715943799634e-05,
      "loss": 1.0893,
      "step": 342
    },
    {
      "epoch": 0.81,
      "learning_rate": 1.6432498472816126e-05,
      "loss": 1.1396,
      "step": 343
    },
    {
      "epoch": 0.81,
      "learning_rate": 1.6420281001832624e-05,
      "loss": 0.9067,
      "step": 344
    },
    {
      "epoch": 0.82,
      "learning_rate": 1.6408063530849116e-05,
      "loss": 1.6975,
      "step": 345
    },
    {
      "epoch": 0.82,
      "learning_rate": 1.639584605986561e-05,
      "loss": 0.7273,
      "step": 346
    },
    {
      "epoch": 0.82,
      "learning_rate": 1.6383628588882103e-05,
      "loss": 0.9293,
      "step": 347
    },
    {
      "epoch": 0.82,
      "learning_rate": 1.6371411117898598e-05,
      "loss": 0.7945,
      "step": 348
    },
    {
      "epoch": 0.83,
      "learning_rate": 1.635919364691509e-05,
      "loss": 0.8376,
      "step": 349
    },
    {
      "epoch": 0.83,
      "learning_rate": 1.6346976175931585e-05,
      "loss": 0.7766,
      "step": 350
    },
    {
      "epoch": 0.83,
      "learning_rate": 1.6334758704948076e-05,
      "loss": 1.1176,
      "step": 351
    },
    {
      "epoch": 0.83,
      "learning_rate": 1.632254123396457e-05,
      "loss": 0.9348,
      "step": 352
    },
    {
      "epoch": 0.83,
      "learning_rate": 1.6310323762981063e-05,
      "loss": 1.0182,
      "step": 353
    },
    {
      "epoch": 0.84,
      "learning_rate": 1.6298106291997558e-05,
      "loss": 1.0181,
      "step": 354
    },
    {
      "epoch": 0.84,
      "learning_rate": 1.628588882101405e-05,
      "loss": 0.7864,
      "step": 355
    },
    {
      "epoch": 0.84,
      "learning_rate": 1.6273671350030545e-05,
      "loss": 0.9777,
      "step": 356
    },
    {
      "epoch": 0.84,
      "learning_rate": 1.6261453879047037e-05,
      "loss": 0.7503,
      "step": 357
    },
    {
      "epoch": 0.85,
      "learning_rate": 1.6249236408063532e-05,
      "loss": 0.8894,
      "step": 358
    },
    {
      "epoch": 0.85,
      "learning_rate": 1.6237018937080027e-05,
      "loss": 0.8838,
      "step": 359
    },
    {
      "epoch": 0.85,
      "learning_rate": 1.622480146609652e-05,
      "loss": 0.874,
      "step": 360
    },
    {
      "epoch": 0.85,
      "learning_rate": 1.6212583995113014e-05,
      "loss": 0.9243,
      "step": 361
    },
    {
      "epoch": 0.86,
      "learning_rate": 1.620036652412951e-05,
      "loss": 0.9355,
      "step": 362
    },
    {
      "epoch": 0.86,
      "learning_rate": 1.6188149053146e-05,
      "loss": 0.9969,
      "step": 363
    },
    {
      "epoch": 0.86,
      "learning_rate": 1.6175931582162496e-05,
      "loss": 0.7542,
      "step": 364
    },
    {
      "epoch": 0.86,
      "learning_rate": 1.6163714111178987e-05,
      "loss": 1.7259,
      "step": 365
    },
    {
      "epoch": 0.87,
      "learning_rate": 1.6151496640195482e-05,
      "loss": 1.0189,
      "step": 366
    },
    {
      "epoch": 0.87,
      "learning_rate": 1.6139279169211974e-05,
      "loss": 0.8796,
      "step": 367
    },
    {
      "epoch": 0.87,
      "learning_rate": 1.612706169822847e-05,
      "loss": 1.0183,
      "step": 368
    },
    {
      "epoch": 0.87,
      "learning_rate": 1.611484422724496e-05,
      "loss": 0.9069,
      "step": 369
    },
    {
      "epoch": 0.87,
      "learning_rate": 1.6102626756261456e-05,
      "loss": 0.7681,
      "step": 370
    },
    {
      "epoch": 0.88,
      "learning_rate": 1.6090409285277948e-05,
      "loss": 0.8676,
      "step": 371
    },
    {
      "epoch": 0.88,
      "learning_rate": 1.6078191814294443e-05,
      "loss": 1.0638,
      "step": 372
    },
    {
      "epoch": 0.88,
      "learning_rate": 1.6065974343310935e-05,
      "loss": 0.8624,
      "step": 373
    },
    {
      "epoch": 0.88,
      "learning_rate": 1.605375687232743e-05,
      "loss": 0.8557,
      "step": 374
    },
    {
      "epoch": 0.89,
      "learning_rate": 1.6041539401343925e-05,
      "loss": 0.9156,
      "step": 375
    },
    {
      "epoch": 0.89,
      "learning_rate": 1.6029321930360416e-05,
      "loss": 0.9213,
      "step": 376
    },
    {
      "epoch": 0.89,
      "learning_rate": 1.601710445937691e-05,
      "loss": 0.8925,
      "step": 377
    },
    {
      "epoch": 0.89,
      "learning_rate": 1.6004886988393403e-05,
      "loss": 1.0731,
      "step": 378
    },
    {
      "epoch": 0.9,
      "learning_rate": 1.59926695174099e-05,
      "loss": 0.9418,
      "step": 379
    },
    {
      "epoch": 0.9,
      "learning_rate": 1.598045204642639e-05,
      "loss": 1.1517,
      "step": 380
    },
    {
      "epoch": 0.9,
      "learning_rate": 1.5968234575442885e-05,
      "loss": 1.0124,
      "step": 381
    },
    {
      "epoch": 0.9,
      "learning_rate": 1.5956017104459377e-05,
      "loss": 1.0369,
      "step": 382
    },
    {
      "epoch": 0.91,
      "learning_rate": 1.5943799633475872e-05,
      "loss": 0.9406,
      "step": 383
    },
    {
      "epoch": 0.91,
      "learning_rate": 1.5931582162492364e-05,
      "loss": 1.1757,
      "step": 384
    },
    {
      "epoch": 0.91,
      "learning_rate": 1.591936469150886e-05,
      "loss": 1.0063,
      "step": 385
    },
    {
      "epoch": 0.91,
      "learning_rate": 1.590714722052535e-05,
      "loss": 1.6491,
      "step": 386
    },
    {
      "epoch": 0.92,
      "learning_rate": 1.5894929749541846e-05,
      "loss": 1.0071,
      "step": 387
    },
    {
      "epoch": 0.92,
      "learning_rate": 1.5882712278558337e-05,
      "loss": 0.926,
      "step": 388
    },
    {
      "epoch": 0.92,
      "learning_rate": 1.5870494807574836e-05,
      "loss": 0.91,
      "step": 389
    },
    {
      "epoch": 0.92,
      "learning_rate": 1.5858277336591327e-05,
      "loss": 0.9837,
      "step": 390
    },
    {
      "epoch": 0.92,
      "learning_rate": 1.5846059865607823e-05,
      "loss": 0.787,
      "step": 391
    },
    {
      "epoch": 0.93,
      "learning_rate": 1.5833842394624314e-05,
      "loss": 1.0002,
      "step": 392
    },
    {
      "epoch": 0.93,
      "learning_rate": 1.582162492364081e-05,
      "loss": 1.0767,
      "step": 393
    },
    {
      "epoch": 0.93,
      "learning_rate": 1.58094074526573e-05,
      "loss": 0.9251,
      "step": 394
    },
    {
      "epoch": 0.93,
      "learning_rate": 1.5797189981673796e-05,
      "loss": 1.2487,
      "step": 395
    },
    {
      "epoch": 0.94,
      "learning_rate": 1.5784972510690288e-05,
      "loss": 0.9963,
      "step": 396
    },
    {
      "epoch": 0.94,
      "learning_rate": 1.5772755039706783e-05,
      "loss": 0.6831,
      "step": 397
    },
    {
      "epoch": 0.94,
      "learning_rate": 1.5760537568723275e-05,
      "loss": 0.9607,
      "step": 398
    },
    {
      "epoch": 0.94,
      "learning_rate": 1.574832009773977e-05,
      "loss": 0.8419,
      "step": 399
    },
    {
      "epoch": 0.95,
      "learning_rate": 1.573610262675626e-05,
      "loss": 1.1296,
      "step": 400
    },
    {
      "epoch": 0.95,
      "learning_rate": 1.5723885155772757e-05,
      "loss": 1.1621,
      "step": 401
    },
    {
      "epoch": 0.95,
      "learning_rate": 1.5711667684789248e-05,
      "loss": 0.7839,
      "step": 402
    },
    {
      "epoch": 0.95,
      "learning_rate": 1.5699450213805743e-05,
      "loss": 0.9428,
      "step": 403
    },
    {
      "epoch": 0.96,
      "learning_rate": 1.568723274282224e-05,
      "loss": 0.8488,
      "step": 404
    },
    {
      "epoch": 0.96,
      "learning_rate": 1.567501527183873e-05,
      "loss": 0.8306,
      "step": 405
    },
    {
      "epoch": 0.96,
      "learning_rate": 1.5662797800855225e-05,
      "loss": 1.0083,
      "step": 406
    },
    {
      "epoch": 0.96,
      "learning_rate": 1.5650580329871717e-05,
      "loss": 0.9262,
      "step": 407
    },
    {
      "epoch": 0.96,
      "learning_rate": 1.5638362858888212e-05,
      "loss": 0.769,
      "step": 408
    },
    {
      "epoch": 0.97,
      "learning_rate": 1.5626145387904704e-05,
      "loss": 0.9818,
      "step": 409
    },
    {
      "epoch": 0.97,
      "learning_rate": 1.56139279169212e-05,
      "loss": 0.8393,
      "step": 410
    },
    {
      "epoch": 0.97,
      "learning_rate": 1.560171044593769e-05,
      "loss": 0.953,
      "step": 411
    },
    {
      "epoch": 0.97,
      "learning_rate": 1.5589492974954186e-05,
      "loss": 0.9442,
      "step": 412
    },
    {
      "epoch": 0.98,
      "learning_rate": 1.557727550397068e-05,
      "loss": 0.9067,
      "step": 413
    },
    {
      "epoch": 0.98,
      "learning_rate": 1.5565058032987173e-05,
      "loss": 1.0057,
      "step": 414
    },
    {
      "epoch": 0.98,
      "learning_rate": 1.5552840562003668e-05,
      "loss": 0.8088,
      "step": 415
    },
    {
      "epoch": 0.98,
      "learning_rate": 1.554062309102016e-05,
      "loss": 0.7332,
      "step": 416
    },
    {
      "epoch": 0.99,
      "learning_rate": 1.5528405620036654e-05,
      "loss": 1.084,
      "step": 417
    },
    {
      "epoch": 0.99,
      "learning_rate": 1.5516188149053146e-05,
      "loss": 0.8262,
      "step": 418
    },
    {
      "epoch": 0.99,
      "learning_rate": 1.550397067806964e-05,
      "loss": 0.6978,
      "step": 419
    },
    {
      "epoch": 0.99,
      "learning_rate": 1.5491753207086136e-05,
      "loss": 0.8885,
      "step": 420
    },
    {
      "epoch": 1.0,
      "learning_rate": 1.5479535736102628e-05,
      "loss": 0.9892,
      "step": 421
    },
    {
      "epoch": 1.0,
      "learning_rate": 1.5467318265119123e-05,
      "loss": 1.0634,
      "step": 422
    },
    {
      "epoch": 1.0,
      "learning_rate": 1.5455100794135615e-05,
      "loss": 1.2635,
      "step": 423
    },
    {
      "epoch": 1.0,
      "learning_rate": 1.544288332315211e-05,
      "loss": 0.8863,
      "step": 424
    },
    {
      "epoch": 1.0,
      "learning_rate": 1.54306658521686e-05,
      "loss": 0.8724,
      "step": 425
    },
    {
      "epoch": 1.01,
      "learning_rate": 1.5418448381185097e-05,
      "loss": 0.8895,
      "step": 426
    },
    {
      "epoch": 1.01,
      "learning_rate": 1.540623091020159e-05,
      "loss": 0.7665,
      "step": 427
    },
    {
      "epoch": 1.01,
      "learning_rate": 1.5394013439218084e-05,
      "loss": 0.7186,
      "step": 428
    },
    {
      "epoch": 1.01,
      "learning_rate": 1.5381795968234575e-05,
      "loss": 0.7365,
      "step": 429
    },
    {
      "epoch": 1.02,
      "learning_rate": 1.536957849725107e-05,
      "loss": 0.9939,
      "step": 430
    },
    {
      "epoch": 1.02,
      "learning_rate": 1.5357361026267562e-05,
      "loss": 0.9525,
      "step": 431
    },
    {
      "epoch": 1.02,
      "learning_rate": 1.5345143555284057e-05,
      "loss": 0.7501,
      "step": 432
    },
    {
      "epoch": 1.02,
      "learning_rate": 1.533292608430055e-05,
      "loss": 1.0303,
      "step": 433
    },
    {
      "epoch": 1.03,
      "learning_rate": 1.5320708613317047e-05,
      "loss": 1.0044,
      "step": 434
    },
    {
      "epoch": 1.03,
      "learning_rate": 1.530849114233354e-05,
      "loss": 1.0009,
      "step": 435
    },
    {
      "epoch": 1.03,
      "learning_rate": 1.5296273671350034e-05,
      "loss": 1.0215,
      "step": 436
    },
    {
      "epoch": 1.03,
      "learning_rate": 1.5284056200366526e-05,
      "loss": 1.0151,
      "step": 437
    },
    {
      "epoch": 1.04,
      "learning_rate": 1.527183872938302e-05,
      "loss": 0.9823,
      "step": 438
    },
    {
      "epoch": 1.04,
      "learning_rate": 1.5259621258399513e-05,
      "loss": 0.862,
      "step": 439
    },
    {
      "epoch": 1.04,
      "learning_rate": 1.5247403787416006e-05,
      "loss": 0.8483,
      "step": 440
    },
    {
      "epoch": 1.04,
      "learning_rate": 1.52351863164325e-05,
      "loss": 0.947,
      "step": 441
    },
    {
      "epoch": 1.05,
      "learning_rate": 1.5222968845448993e-05,
      "loss": 0.8914,
      "step": 442
    },
    {
      "epoch": 1.05,
      "learning_rate": 1.5210751374465486e-05,
      "loss": 1.0391,
      "step": 443
    },
    {
      "epoch": 1.05,
      "learning_rate": 1.519853390348198e-05,
      "loss": 0.9706,
      "step": 444
    },
    {
      "epoch": 1.05,
      "learning_rate": 1.5186316432498473e-05,
      "loss": 0.6686,
      "step": 445
    },
    {
      "epoch": 1.05,
      "learning_rate": 1.5174098961514966e-05,
      "loss": 1.177,
      "step": 446
    },
    {
      "epoch": 1.06,
      "learning_rate": 1.516188149053146e-05,
      "loss": 0.9745,
      "step": 447
    },
    {
      "epoch": 1.06,
      "learning_rate": 1.5149664019547953e-05,
      "loss": 0.8828,
      "step": 448
    },
    {
      "epoch": 1.06,
      "learning_rate": 1.513744654856445e-05,
      "loss": 0.5288,
      "step": 449
    },
    {
      "epoch": 1.06,
      "learning_rate": 1.5125229077580943e-05,
      "loss": 1.6787,
      "step": 450
    },
    {
      "epoch": 1.07,
      "learning_rate": 1.5113011606597437e-05,
      "loss": 0.8921,
      "step": 451
    },
    {
      "epoch": 1.07,
      "learning_rate": 1.510079413561393e-05,
      "loss": 1.1841,
      "step": 452
    },
    {
      "epoch": 1.07,
      "learning_rate": 1.5088576664630424e-05,
      "loss": 0.9546,
      "step": 453
    },
    {
      "epoch": 1.07,
      "learning_rate": 1.5076359193646917e-05,
      "loss": 0.786,
      "step": 454
    },
    {
      "epoch": 1.08,
      "learning_rate": 1.506414172266341e-05,
      "loss": 0.9826,
      "step": 455
    },
    {
      "epoch": 1.08,
      "learning_rate": 1.5051924251679904e-05,
      "loss": 0.9842,
      "step": 456
    },
    {
      "epoch": 1.08,
      "learning_rate": 1.5039706780696397e-05,
      "loss": 0.9016,
      "step": 457
    },
    {
      "epoch": 1.08,
      "learning_rate": 1.502748930971289e-05,
      "loss": 0.7724,
      "step": 458
    },
    {
      "epoch": 1.09,
      "learning_rate": 1.5015271838729384e-05,
      "loss": 0.6341,
      "step": 459
    },
    {
      "epoch": 1.09,
      "learning_rate": 1.5003054367745877e-05,
      "loss": 0.8749,
      "step": 460
    },
    {
      "epoch": 1.09,
      "learning_rate": 1.4990836896762371e-05,
      "loss": 0.6795,
      "step": 461
    },
    {
      "epoch": 1.09,
      "learning_rate": 1.4978619425778864e-05,
      "loss": 1.0035,
      "step": 462
    },
    {
      "epoch": 1.09,
      "learning_rate": 1.4966401954795358e-05,
      "loss": 0.6718,
      "step": 463
    },
    {
      "epoch": 1.1,
      "learning_rate": 1.4954184483811851e-05,
      "loss": 0.853,
      "step": 464
    },
    {
      "epoch": 1.1,
      "learning_rate": 1.4941967012828346e-05,
      "loss": 0.8484,
      "step": 465
    },
    {
      "epoch": 1.1,
      "learning_rate": 1.492974954184484e-05,
      "loss": 0.8504,
      "step": 466
    },
    {
      "epoch": 1.1,
      "learning_rate": 1.4917532070861333e-05,
      "loss": 1.0019,
      "step": 467
    },
    {
      "epoch": 1.11,
      "learning_rate": 1.4905314599877826e-05,
      "loss": 0.7926,
      "step": 468
    },
    {
      "epoch": 1.11,
      "learning_rate": 1.489309712889432e-05,
      "loss": 0.951,
      "step": 469
    },
    {
      "epoch": 1.11,
      "learning_rate": 1.4880879657910813e-05,
      "loss": 0.9752,
      "step": 470
    },
    {
      "epoch": 1.11,
      "learning_rate": 1.4868662186927307e-05,
      "loss": 0.7827,
      "step": 471
    },
    {
      "epoch": 1.12,
      "learning_rate": 1.48564447159438e-05,
      "loss": 1.0879,
      "step": 472
    },
    {
      "epoch": 1.12,
      "learning_rate": 1.4844227244960293e-05,
      "loss": 0.985,
      "step": 473
    },
    {
      "epoch": 1.12,
      "learning_rate": 1.4832009773976788e-05,
      "loss": 0.662,
      "step": 474
    },
    {
      "epoch": 1.12,
      "learning_rate": 1.4819792302993282e-05,
      "loss": 0.7518,
      "step": 475
    },
    {
      "epoch": 1.13,
      "learning_rate": 1.4807574832009775e-05,
      "loss": 0.6804,
      "step": 476
    },
    {
      "epoch": 1.13,
      "learning_rate": 1.4795357361026269e-05,
      "loss": 0.7282,
      "step": 477
    },
    {
      "epoch": 1.13,
      "learning_rate": 1.4783139890042762e-05,
      "loss": 0.8529,
      "step": 478
    },
    {
      "epoch": 1.13,
      "learning_rate": 1.4770922419059255e-05,
      "loss": 1.0995,
      "step": 479
    },
    {
      "epoch": 1.14,
      "learning_rate": 1.475870494807575e-05,
      "loss": 0.9315,
      "step": 480
    },
    {
      "epoch": 1.14,
      "learning_rate": 1.4746487477092244e-05,
      "loss": 0.9777,
      "step": 481
    },
    {
      "epoch": 1.14,
      "learning_rate": 1.4734270006108737e-05,
      "loss": 0.8812,
      "step": 482
    },
    {
      "epoch": 1.14,
      "learning_rate": 1.472205253512523e-05,
      "loss": 0.7139,
      "step": 483
    },
    {
      "epoch": 1.14,
      "learning_rate": 1.4709835064141724e-05,
      "loss": 0.871,
      "step": 484
    },
    {
      "epoch": 1.15,
      "learning_rate": 1.4697617593158218e-05,
      "loss": 0.9647,
      "step": 485
    },
    {
      "epoch": 1.15,
      "learning_rate": 1.4685400122174711e-05,
      "loss": 0.9651,
      "step": 486
    },
    {
      "epoch": 1.15,
      "learning_rate": 1.4673182651191204e-05,
      "loss": 0.821,
      "step": 487
    },
    {
      "epoch": 1.15,
      "learning_rate": 1.4660965180207698e-05,
      "loss": 0.6677,
      "step": 488
    },
    {
      "epoch": 1.16,
      "learning_rate": 1.4648747709224191e-05,
      "loss": 0.8127,
      "step": 489
    },
    {
      "epoch": 1.16,
      "learning_rate": 1.4636530238240685e-05,
      "loss": 0.9843,
      "step": 490
    },
    {
      "epoch": 1.16,
      "learning_rate": 1.4624312767257178e-05,
      "loss": 0.958,
      "step": 491
    },
    {
      "epoch": 1.16,
      "learning_rate": 1.4612095296273671e-05,
      "loss": 0.9279,
      "step": 492
    },
    {
      "epoch": 1.17,
      "learning_rate": 1.4599877825290165e-05,
      "loss": 0.9219,
      "step": 493
    },
    {
      "epoch": 1.17,
      "learning_rate": 1.4587660354306658e-05,
      "loss": 0.7247,
      "step": 494
    },
    {
      "epoch": 1.17,
      "learning_rate": 1.4575442883323155e-05,
      "loss": 0.7789,
      "step": 495
    },
    {
      "epoch": 1.17,
      "learning_rate": 1.4563225412339648e-05,
      "loss": 0.8834,
      "step": 496
    },
    {
      "epoch": 1.18,
      "learning_rate": 1.4551007941356142e-05,
      "loss": 1.0376,
      "step": 497
    },
    {
      "epoch": 1.18,
      "learning_rate": 1.4538790470372635e-05,
      "loss": 0.6958,
      "step": 498
    },
    {
      "epoch": 1.18,
      "learning_rate": 1.4526572999389129e-05,
      "loss": 0.7884,
      "step": 499
    },
    {
      "epoch": 1.18,
      "learning_rate": 1.4514355528405622e-05,
      "loss": 0.7412,
      "step": 500
    },
    {
      "epoch": 1.18,
      "learning_rate": 1.4502138057422115e-05,
      "loss": 1.04,
      "step": 501
    },
    {
      "epoch": 1.19,
      "learning_rate": 1.4489920586438609e-05,
      "loss": 0.9107,
      "step": 502
    },
    {
      "epoch": 1.19,
      "learning_rate": 1.4477703115455102e-05,
      "loss": 0.8633,
      "step": 503
    },
    {
      "epoch": 1.19,
      "learning_rate": 1.4465485644471596e-05,
      "loss": 0.97,
      "step": 504
    },
    {
      "epoch": 1.19,
      "learning_rate": 1.4453268173488089e-05,
      "loss": 0.8615,
      "step": 505
    },
    {
      "epoch": 1.2,
      "learning_rate": 1.4441050702504582e-05,
      "loss": 0.846,
      "step": 506
    },
    {
      "epoch": 1.2,
      "learning_rate": 1.4428833231521076e-05,
      "loss": 0.7451,
      "step": 507
    },
    {
      "epoch": 1.2,
      "learning_rate": 1.441661576053757e-05,
      "loss": 0.6941,
      "step": 508
    },
    {
      "epoch": 1.2,
      "learning_rate": 1.4404398289554063e-05,
      "loss": 0.8699,
      "step": 509
    },
    {
      "epoch": 1.21,
      "learning_rate": 1.4392180818570558e-05,
      "loss": 0.926,
      "step": 510
    },
    {
      "epoch": 1.21,
      "learning_rate": 1.4379963347587051e-05,
      "loss": 0.6859,
      "step": 511
    },
    {
      "epoch": 1.21,
      "learning_rate": 1.4367745876603545e-05,
      "loss": 1.0583,
      "step": 512
    },
    {
      "epoch": 1.21,
      "learning_rate": 1.4355528405620038e-05,
      "loss": 0.888,
      "step": 513
    },
    {
      "epoch": 1.22,
      "learning_rate": 1.4343310934636531e-05,
      "loss": 1.9128,
      "step": 514
    },
    {
      "epoch": 1.22,
      "learning_rate": 1.4331093463653025e-05,
      "loss": 0.9135,
      "step": 515
    },
    {
      "epoch": 1.22,
      "learning_rate": 1.4318875992669518e-05,
      "loss": 0.9358,
      "step": 516
    },
    {
      "epoch": 1.22,
      "learning_rate": 1.4306658521686012e-05,
      "loss": 1.0008,
      "step": 517
    },
    {
      "epoch": 1.22,
      "learning_rate": 1.4294441050702505e-05,
      "loss": 1.059,
      "step": 518
    },
    {
      "epoch": 1.23,
      "learning_rate": 1.4282223579718998e-05,
      "loss": 0.8735,
      "step": 519
    },
    {
      "epoch": 1.23,
      "learning_rate": 1.4270006108735492e-05,
      "loss": 0.9229,
      "step": 520
    },
    {
      "epoch": 1.23,
      "learning_rate": 1.4257788637751985e-05,
      "loss": 0.8591,
      "step": 521
    },
    {
      "epoch": 1.23,
      "learning_rate": 1.4245571166768479e-05,
      "loss": 0.6778,
      "step": 522
    },
    {
      "epoch": 1.24,
      "learning_rate": 1.4233353695784972e-05,
      "loss": 0.9968,
      "step": 523
    },
    {
      "epoch": 1.24,
      "learning_rate": 1.4221136224801465e-05,
      "loss": 0.932,
      "step": 524
    },
    {
      "epoch": 1.24,
      "learning_rate": 1.4208918753817962e-05,
      "loss": 0.8142,
      "step": 525
    },
    {
      "epoch": 1.24,
      "learning_rate": 1.4196701282834456e-05,
      "loss": 0.8707,
      "step": 526
    },
    {
      "epoch": 1.25,
      "learning_rate": 1.4184483811850949e-05,
      "loss": 1.3432,
      "step": 527
    },
    {
      "epoch": 1.25,
      "learning_rate": 1.4172266340867442e-05,
      "loss": 0.9905,
      "step": 528
    },
    {
      "epoch": 1.25,
      "learning_rate": 1.4160048869883936e-05,
      "loss": 0.9144,
      "step": 529
    },
    {
      "epoch": 1.25,
      "learning_rate": 1.4147831398900429e-05,
      "loss": 0.9362,
      "step": 530
    },
    {
      "epoch": 1.26,
      "learning_rate": 1.4135613927916923e-05,
      "loss": 0.8545,
      "step": 531
    },
    {
      "epoch": 1.26,
      "learning_rate": 1.4123396456933416e-05,
      "loss": 0.9268,
      "step": 532
    },
    {
      "epoch": 1.26,
      "learning_rate": 1.411117898594991e-05,
      "loss": 0.8372,
      "step": 533
    },
    {
      "epoch": 1.26,
      "learning_rate": 1.4098961514966403e-05,
      "loss": 0.7659,
      "step": 534
    },
    {
      "epoch": 1.27,
      "learning_rate": 1.4086744043982896e-05,
      "loss": 0.7959,
      "step": 535
    },
    {
      "epoch": 1.27,
      "learning_rate": 1.407452657299939e-05,
      "loss": 0.7957,
      "step": 536
    },
    {
      "epoch": 1.27,
      "learning_rate": 1.4062309102015883e-05,
      "loss": 0.8758,
      "step": 537
    },
    {
      "epoch": 1.27,
      "learning_rate": 1.4050091631032376e-05,
      "loss": 0.8478,
      "step": 538
    },
    {
      "epoch": 1.27,
      "learning_rate": 1.403787416004887e-05,
      "loss": 0.9166,
      "step": 539
    },
    {
      "epoch": 1.28,
      "learning_rate": 1.4025656689065365e-05,
      "loss": 0.8067,
      "step": 540
    },
    {
      "epoch": 1.28,
      "learning_rate": 1.4013439218081858e-05,
      "loss": 0.7642,
      "step": 541
    },
    {
      "epoch": 1.28,
      "learning_rate": 1.4001221747098352e-05,
      "loss": 0.925,
      "step": 542
    },
    {
      "epoch": 1.28,
      "learning_rate": 1.3989004276114847e-05,
      "loss": 0.9245,
      "step": 543
    },
    {
      "epoch": 1.29,
      "learning_rate": 1.397678680513134e-05,
      "loss": 1.3363,
      "step": 544
    },
    {
      "epoch": 1.29,
      "learning_rate": 1.3964569334147834e-05,
      "loss": 0.703,
      "step": 545
    },
    {
      "epoch": 1.29,
      "learning_rate": 1.3952351863164327e-05,
      "loss": 0.8265,
      "step": 546
    },
    {
      "epoch": 1.29,
      "learning_rate": 1.394013439218082e-05,
      "loss": 0.8719,
      "step": 547
    },
    {
      "epoch": 1.3,
      "learning_rate": 1.3927916921197314e-05,
      "loss": 0.9639,
      "step": 548
    },
    {
      "epoch": 1.3,
      "learning_rate": 1.3915699450213807e-05,
      "loss": 0.9368,
      "step": 549
    },
    {
      "epoch": 1.3,
      "learning_rate": 1.39034819792303e-05,
      "loss": 0.8262,
      "step": 550
    },
    {
      "epoch": 1.3,
      "learning_rate": 1.3891264508246794e-05,
      "loss": 0.8692,
      "step": 551
    },
    {
      "epoch": 1.31,
      "learning_rate": 1.3879047037263287e-05,
      "loss": 0.8015,
      "step": 552
    },
    {
      "epoch": 1.31,
      "learning_rate": 1.386682956627978e-05,
      "loss": 0.789,
      "step": 553
    },
    {
      "epoch": 1.31,
      "learning_rate": 1.3854612095296274e-05,
      "loss": 0.7983,
      "step": 554
    },
    {
      "epoch": 1.31,
      "learning_rate": 1.384239462431277e-05,
      "loss": 0.9553,
      "step": 555
    },
    {
      "epoch": 1.31,
      "learning_rate": 1.3830177153329263e-05,
      "loss": 0.9573,
      "step": 556
    },
    {
      "epoch": 1.32,
      "learning_rate": 1.3817959682345756e-05,
      "loss": 0.8161,
      "step": 557
    },
    {
      "epoch": 1.32,
      "learning_rate": 1.380574221136225e-05,
      "loss": 0.7292,
      "step": 558
    },
    {
      "epoch": 1.32,
      "learning_rate": 1.3793524740378743e-05,
      "loss": 0.9365,
      "step": 559
    },
    {
      "epoch": 1.32,
      "learning_rate": 1.3781307269395236e-05,
      "loss": 0.9629,
      "step": 560
    },
    {
      "epoch": 1.33,
      "learning_rate": 1.376908979841173e-05,
      "loss": 0.823,
      "step": 561
    },
    {
      "epoch": 1.33,
      "learning_rate": 1.3756872327428223e-05,
      "loss": 0.8126,
      "step": 562
    },
    {
      "epoch": 1.33,
      "learning_rate": 1.3744654856444716e-05,
      "loss": 0.8997,
      "step": 563
    },
    {
      "epoch": 1.33,
      "learning_rate": 1.373243738546121e-05,
      "loss": 0.7755,
      "step": 564
    },
    {
      "epoch": 1.34,
      "learning_rate": 1.3720219914477703e-05,
      "loss": 0.8907,
      "step": 565
    },
    {
      "epoch": 1.34,
      "learning_rate": 1.3708002443494197e-05,
      "loss": 0.9902,
      "step": 566
    },
    {
      "epoch": 1.34,
      "learning_rate": 1.369578497251069e-05,
      "loss": 0.7797,
      "step": 567
    },
    {
      "epoch": 1.34,
      "learning_rate": 1.3683567501527183e-05,
      "loss": 0.8291,
      "step": 568
    },
    {
      "epoch": 1.35,
      "learning_rate": 1.3671350030543677e-05,
      "loss": 0.7132,
      "step": 569
    },
    {
      "epoch": 1.35,
      "learning_rate": 1.3659132559560174e-05,
      "loss": 0.8713,
      "step": 570
    },
    {
      "epoch": 1.35,
      "learning_rate": 1.3646915088576667e-05,
      "loss": 0.8353,
      "step": 571
    },
    {
      "epoch": 1.35,
      "learning_rate": 1.363469761759316e-05,
      "loss": 1.1653,
      "step": 572
    },
    {
      "epoch": 1.35,
      "learning_rate": 1.3622480146609654e-05,
      "loss": 0.8543,
      "step": 573
    },
    {
      "epoch": 1.36,
      "learning_rate": 1.3610262675626147e-05,
      "loss": 0.7127,
      "step": 574
    },
    {
      "epoch": 1.36,
      "learning_rate": 1.359804520464264e-05,
      "loss": 0.7931,
      "step": 575
    },
    {
      "epoch": 1.36,
      "learning_rate": 1.3585827733659134e-05,
      "loss": 1.0101,
      "step": 576
    },
    {
      "epoch": 1.36,
      "learning_rate": 1.3573610262675627e-05,
      "loss": 0.7975,
      "step": 577
    },
    {
      "epoch": 1.37,
      "learning_rate": 1.3561392791692121e-05,
      "loss": 0.8408,
      "step": 578
    },
    {
      "epoch": 1.37,
      "learning_rate": 1.3549175320708614e-05,
      "loss": 0.8974,
      "step": 579
    },
    {
      "epoch": 1.37,
      "learning_rate": 1.3536957849725108e-05,
      "loss": 0.8791,
      "step": 580
    },
    {
      "epoch": 1.37,
      "learning_rate": 1.3524740378741601e-05,
      "loss": 0.654,
      "step": 581
    },
    {
      "epoch": 1.38,
      "learning_rate": 1.3512522907758094e-05,
      "loss": 0.9567,
      "step": 582
    },
    {
      "epoch": 1.38,
      "learning_rate": 1.3500305436774588e-05,
      "loss": 0.9604,
      "step": 583
    },
    {
      "epoch": 1.38,
      "learning_rate": 1.3488087965791081e-05,
      "loss": 0.8445,
      "step": 584
    },
    {
      "epoch": 1.38,
      "learning_rate": 1.3475870494807576e-05,
      "loss": 0.9758,
      "step": 585
    },
    {
      "epoch": 1.39,
      "learning_rate": 1.346365302382407e-05,
      "loss": 0.8508,
      "step": 586
    },
    {
      "epoch": 1.39,
      "learning_rate": 1.3451435552840563e-05,
      "loss": 0.789,
      "step": 587
    },
    {
      "epoch": 1.39,
      "learning_rate": 1.3439218081857057e-05,
      "loss": 0.946,
      "step": 588
    },
    {
      "epoch": 1.39,
      "learning_rate": 1.342700061087355e-05,
      "loss": 0.7667,
      "step": 589
    },
    {
      "epoch": 1.4,
      "learning_rate": 1.3414783139890043e-05,
      "loss": 1.0173,
      "step": 590
    },
    {
      "epoch": 1.4,
      "learning_rate": 1.3402565668906537e-05,
      "loss": 0.781,
      "step": 591
    },
    {
      "epoch": 1.4,
      "learning_rate": 1.339034819792303e-05,
      "loss": 0.8253,
      "step": 592
    },
    {
      "epoch": 1.4,
      "learning_rate": 1.3378130726939524e-05,
      "loss": 0.985,
      "step": 593
    },
    {
      "epoch": 1.4,
      "learning_rate": 1.3365913255956019e-05,
      "loss": 1.0426,
      "step": 594
    },
    {
      "epoch": 1.41,
      "learning_rate": 1.3353695784972512e-05,
      "loss": 0.7838,
      "step": 595
    },
    {
      "epoch": 1.41,
      "learning_rate": 1.3341478313989005e-05,
      "loss": 0.8737,
      "step": 596
    },
    {
      "epoch": 1.41,
      "learning_rate": 1.3329260843005499e-05,
      "loss": 0.9306,
      "step": 597
    },
    {
      "epoch": 1.41,
      "learning_rate": 1.3317043372021992e-05,
      "loss": 0.7315,
      "step": 598
    },
    {
      "epoch": 1.42,
      "learning_rate": 1.3304825901038486e-05,
      "loss": 1.3077,
      "step": 599
    },
    {
      "epoch": 1.42,
      "learning_rate": 1.329260843005498e-05,
      "loss": 0.7888,
      "step": 600
    },
    {
      "epoch": 1.42,
      "learning_rate": 1.3280390959071474e-05,
      "loss": 0.9452,
      "step": 601
    },
    {
      "epoch": 1.42,
      "learning_rate": 1.3268173488087968e-05,
      "loss": 1.0367,
      "step": 602
    },
    {
      "epoch": 1.43,
      "learning_rate": 1.3255956017104461e-05,
      "loss": 1.4717,
      "step": 603
    },
    {
      "epoch": 1.43,
      "learning_rate": 1.3243738546120954e-05,
      "loss": 1.4204,
      "step": 604
    },
    {
      "epoch": 1.43,
      "learning_rate": 1.3231521075137448e-05,
      "loss": 0.7385,
      "step": 605
    },
    {
      "epoch": 1.43,
      "learning_rate": 1.3219303604153941e-05,
      "loss": 0.7257,
      "step": 606
    },
    {
      "epoch": 1.44,
      "learning_rate": 1.3207086133170435e-05,
      "loss": 0.8938,
      "step": 607
    },
    {
      "epoch": 1.44,
      "learning_rate": 1.3194868662186928e-05,
      "loss": 0.6473,
      "step": 608
    },
    {
      "epoch": 1.44,
      "learning_rate": 1.3182651191203421e-05,
      "loss": 0.8329,
      "step": 609
    },
    {
      "epoch": 1.44,
      "learning_rate": 1.3170433720219915e-05,
      "loss": 0.6972,
      "step": 610
    },
    {
      "epoch": 1.44,
      "learning_rate": 1.3158216249236408e-05,
      "loss": 0.894,
      "step": 611
    },
    {
      "epoch": 1.45,
      "learning_rate": 1.3145998778252902e-05,
      "loss": 0.751,
      "step": 612
    },
    {
      "epoch": 1.45,
      "learning_rate": 1.3133781307269395e-05,
      "loss": 0.9107,
      "step": 613
    },
    {
      "epoch": 1.45,
      "learning_rate": 1.3121563836285888e-05,
      "loss": 1.0217,
      "step": 614
    },
    {
      "epoch": 1.45,
      "learning_rate": 1.3109346365302385e-05,
      "loss": 0.8694,
      "step": 615
    },
    {
      "epoch": 1.46,
      "learning_rate": 1.3097128894318879e-05,
      "loss": 0.7461,
      "step": 616
    },
    {
      "epoch": 1.46,
      "learning_rate": 1.3084911423335372e-05,
      "loss": 0.916,
      "step": 617
    },
    {
      "epoch": 1.46,
      "learning_rate": 1.3072693952351865e-05,
      "loss": 0.8143,
      "step": 618
    },
    {
      "epoch": 1.46,
      "learning_rate": 1.3060476481368359e-05,
      "loss": 0.8181,
      "step": 619
    },
    {
      "epoch": 1.47,
      "learning_rate": 1.3048259010384852e-05,
      "loss": 0.9167,
      "step": 620
    },
    {
      "epoch": 1.47,
      "learning_rate": 1.3036041539401346e-05,
      "loss": 0.8222,
      "step": 621
    },
    {
      "epoch": 1.47,
      "learning_rate": 1.3023824068417839e-05,
      "loss": 0.907,
      "step": 622
    },
    {
      "epoch": 1.47,
      "learning_rate": 1.3011606597434332e-05,
      "loss": 1.0519,
      "step": 623
    },
    {
      "epoch": 1.48,
      "learning_rate": 1.2999389126450826e-05,
      "loss": 0.8482,
      "step": 624
    },
    {
      "epoch": 1.48,
      "learning_rate": 1.298717165546732e-05,
      "loss": 0.9613,
      "step": 625
    },
    {
      "epoch": 1.48,
      "learning_rate": 1.2974954184483813e-05,
      "loss": 0.749,
      "step": 626
    },
    {
      "epoch": 1.48,
      "learning_rate": 1.2962736713500306e-05,
      "loss": 0.8903,
      "step": 627
    },
    {
      "epoch": 1.48,
      "learning_rate": 1.29505192425168e-05,
      "loss": 0.7888,
      "step": 628
    },
    {
      "epoch": 1.49,
      "learning_rate": 1.2938301771533293e-05,
      "loss": 0.7947,
      "step": 629
    },
    {
      "epoch": 1.49,
      "learning_rate": 1.2926084300549788e-05,
      "loss": 0.7678,
      "step": 630
    },
    {
      "epoch": 1.49,
      "learning_rate": 1.2913866829566281e-05,
      "loss": 1.0733,
      "step": 631
    },
    {
      "epoch": 1.49,
      "learning_rate": 1.2901649358582775e-05,
      "loss": 1.0331,
      "step": 632
    },
    {
      "epoch": 1.5,
      "learning_rate": 1.2889431887599268e-05,
      "loss": 0.7706,
      "step": 633
    },
    {
      "epoch": 1.5,
      "learning_rate": 1.2877214416615762e-05,
      "loss": 0.9589,
      "step": 634
    },
    {
      "epoch": 1.5,
      "learning_rate": 1.2864996945632255e-05,
      "loss": 0.7827,
      "step": 635
    },
    {
      "epoch": 1.5,
      "learning_rate": 1.2852779474648748e-05,
      "loss": 0.86,
      "step": 636
    },
    {
      "epoch": 1.51,
      "learning_rate": 1.2840562003665242e-05,
      "loss": 0.8908,
      "step": 637
    },
    {
      "epoch": 1.51,
      "learning_rate": 1.2828344532681735e-05,
      "loss": 0.9065,
      "step": 638
    },
    {
      "epoch": 1.51,
      "learning_rate": 1.2816127061698229e-05,
      "loss": 0.847,
      "step": 639
    },
    {
      "epoch": 1.51,
      "learning_rate": 1.2803909590714722e-05,
      "loss": 1.051,
      "step": 640
    },
    {
      "epoch": 1.52,
      "learning_rate": 1.2791692119731215e-05,
      "loss": 1.0868,
      "step": 641
    },
    {
      "epoch": 1.52,
      "learning_rate": 1.2779474648747709e-05,
      "loss": 1.0014,
      "step": 642
    },
    {
      "epoch": 1.52,
      "learning_rate": 1.2767257177764202e-05,
      "loss": 0.8227,
      "step": 643
    },
    {
      "epoch": 1.52,
      "learning_rate": 1.2755039706780696e-05,
      "loss": 0.88,
      "step": 644
    },
    {
      "epoch": 1.53,
      "learning_rate": 1.2742822235797192e-05,
      "loss": 0.7747,
      "step": 645
    },
    {
      "epoch": 1.53,
      "learning_rate": 1.2730604764813686e-05,
      "loss": 0.7716,
      "step": 646
    },
    {
      "epoch": 1.53,
      "learning_rate": 1.2718387293830179e-05,
      "loss": 0.8671,
      "step": 647
    },
    {
      "epoch": 1.53,
      "learning_rate": 1.2706169822846673e-05,
      "loss": 0.7988,
      "step": 648
    },
    {
      "epoch": 1.53,
      "learning_rate": 1.2693952351863166e-05,
      "loss": 0.7907,
      "step": 649
    },
    {
      "epoch": 1.54,
      "learning_rate": 1.268173488087966e-05,
      "loss": 0.7658,
      "step": 650
    },
    {
      "epoch": 1.54,
      "learning_rate": 1.2669517409896153e-05,
      "loss": 0.9084,
      "step": 651
    },
    {
      "epoch": 1.54,
      "learning_rate": 1.2657299938912646e-05,
      "loss": 0.7117,
      "step": 652
    },
    {
      "epoch": 1.54,
      "learning_rate": 1.264508246792914e-05,
      "loss": 0.7998,
      "step": 653
    },
    {
      "epoch": 1.55,
      "learning_rate": 1.2632864996945633e-05,
      "loss": 0.7648,
      "step": 654
    },
    {
      "epoch": 1.55,
      "learning_rate": 1.2620647525962126e-05,
      "loss": 0.7891,
      "step": 655
    },
    {
      "epoch": 1.55,
      "learning_rate": 1.260843005497862e-05,
      "loss": 0.7822,
      "step": 656
    },
    {
      "epoch": 1.55,
      "learning_rate": 1.2596212583995113e-05,
      "loss": 0.9386,
      "step": 657
    },
    {
      "epoch": 1.56,
      "learning_rate": 1.2583995113011607e-05,
      "loss": 1.0717,
      "step": 658
    },
    {
      "epoch": 1.56,
      "learning_rate": 1.25717776420281e-05,
      "loss": 0.883,
      "step": 659
    },
    {
      "epoch": 1.56,
      "learning_rate": 1.2559560171044595e-05,
      "loss": 0.8948,
      "step": 660
    },
    {
      "epoch": 1.56,
      "learning_rate": 1.2547342700061088e-05,
      "loss": 0.8952,
      "step": 661
    },
    {
      "epoch": 1.57,
      "learning_rate": 1.2535125229077582e-05,
      "loss": 0.8745,
      "step": 662
    },
    {
      "epoch": 1.57,
      "learning_rate": 1.2522907758094075e-05,
      "loss": 1.0238,
      "step": 663
    },
    {
      "epoch": 1.57,
      "learning_rate": 1.251069028711057e-05,
      "loss": 0.835,
      "step": 664
    },
    {
      "epoch": 1.57,
      "learning_rate": 1.2498472816127064e-05,
      "loss": 0.9643,
      "step": 665
    },
    {
      "epoch": 1.57,
      "learning_rate": 1.2486255345143557e-05,
      "loss": 0.8413,
      "step": 666
    },
    {
      "epoch": 1.58,
      "learning_rate": 1.247403787416005e-05,
      "loss": 0.8589,
      "step": 667
    },
    {
      "epoch": 1.58,
      "learning_rate": 1.2461820403176544e-05,
      "loss": 0.9187,
      "step": 668
    },
    {
      "epoch": 1.58,
      "learning_rate": 1.2449602932193037e-05,
      "loss": 0.6343,
      "step": 669
    },
    {
      "epoch": 1.58,
      "learning_rate": 1.243738546120953e-05,
      "loss": 0.7666,
      "step": 670
    },
    {
      "epoch": 1.59,
      "learning_rate": 1.2425167990226024e-05,
      "loss": 0.7723,
      "step": 671
    },
    {
      "epoch": 1.59,
      "learning_rate": 1.2412950519242518e-05,
      "loss": 0.7481,
      "step": 672
    },
    {
      "epoch": 1.59,
      "learning_rate": 1.2400733048259011e-05,
      "loss": 1.0699,
      "step": 673
    },
    {
      "epoch": 1.59,
      "learning_rate": 1.2388515577275504e-05,
      "loss": 0.9871,
      "step": 674
    },
    {
      "epoch": 1.6,
      "learning_rate": 1.2376298106292e-05,
      "loss": 0.7944,
      "step": 675
    },
    {
      "epoch": 1.6,
      "learning_rate": 1.2364080635308493e-05,
      "loss": 0.8405,
      "step": 676
    },
    {
      "epoch": 1.6,
      "learning_rate": 1.2351863164324986e-05,
      "loss": 1.097,
      "step": 677
    },
    {
      "epoch": 1.6,
      "learning_rate": 1.233964569334148e-05,
      "loss": 0.8661,
      "step": 678
    },
    {
      "epoch": 1.61,
      "learning_rate": 1.2327428222357973e-05,
      "loss": 0.9293,
      "step": 679
    },
    {
      "epoch": 1.61,
      "learning_rate": 1.2315210751374466e-05,
      "loss": 0.8302,
      "step": 680
    },
    {
      "epoch": 1.61,
      "learning_rate": 1.230299328039096e-05,
      "loss": 0.8256,
      "step": 681
    },
    {
      "epoch": 1.61,
      "learning_rate": 1.2290775809407453e-05,
      "loss": 0.9635,
      "step": 682
    },
    {
      "epoch": 1.62,
      "learning_rate": 1.2278558338423947e-05,
      "loss": 0.9612,
      "step": 683
    },
    {
      "epoch": 1.62,
      "learning_rate": 1.226634086744044e-05,
      "loss": 0.6661,
      "step": 684
    },
    {
      "epoch": 1.62,
      "learning_rate": 1.2254123396456933e-05,
      "loss": 0.8132,
      "step": 685
    },
    {
      "epoch": 1.62,
      "learning_rate": 1.2241905925473427e-05,
      "loss": 0.8554,
      "step": 686
    },
    {
      "epoch": 1.62,
      "learning_rate": 1.222968845448992e-05,
      "loss": 0.9417,
      "step": 687
    },
    {
      "epoch": 1.63,
      "learning_rate": 1.2217470983506414e-05,
      "loss": 0.7557,
      "step": 688
    },
    {
      "epoch": 1.63,
      "learning_rate": 1.2205253512522907e-05,
      "loss": 0.7313,
      "step": 689
    },
    {
      "epoch": 1.63,
      "learning_rate": 1.2193036041539404e-05,
      "loss": 1.0282,
      "step": 690
    },
    {
      "epoch": 1.63,
      "learning_rate": 1.2180818570555897e-05,
      "loss": 0.8616,
      "step": 691
    },
    {
      "epoch": 1.64,
      "learning_rate": 1.216860109957239e-05,
      "loss": 1.0365,
      "step": 692
    },
    {
      "epoch": 1.64,
      "learning_rate": 1.2156383628588884e-05,
      "loss": 0.6458,
      "step": 693
    },
    {
      "epoch": 1.64,
      "learning_rate": 1.2144166157605377e-05,
      "loss": 0.8812,
      "step": 694
    },
    {
      "epoch": 1.64,
      "learning_rate": 1.2131948686621871e-05,
      "loss": 0.9863,
      "step": 695
    },
    {
      "epoch": 1.65,
      "learning_rate": 1.2119731215638364e-05,
      "loss": 0.9672,
      "step": 696
    },
    {
      "epoch": 1.65,
      "learning_rate": 1.2107513744654858e-05,
      "loss": 0.7714,
      "step": 697
    },
    {
      "epoch": 1.65,
      "learning_rate": 1.2095296273671351e-05,
      "loss": 1.8343,
      "step": 698
    },
    {
      "epoch": 1.65,
      "learning_rate": 1.2083078802687844e-05,
      "loss": 1.1402,
      "step": 699
    },
    {
      "epoch": 1.66,
      "learning_rate": 1.2070861331704338e-05,
      "loss": 0.7542,
      "step": 700
    },
    {
      "epoch": 1.66,
      "learning_rate": 1.2058643860720831e-05,
      "loss": 0.9874,
      "step": 701
    },
    {
      "epoch": 1.66,
      "learning_rate": 1.2046426389737325e-05,
      "loss": 0.9331,
      "step": 702
    },
    {
      "epoch": 1.66,
      "learning_rate": 1.2034208918753818e-05,
      "loss": 0.9548,
      "step": 703
    },
    {
      "epoch": 1.66,
      "learning_rate": 1.2021991447770312e-05,
      "loss": 0.7545,
      "step": 704
    },
    {
      "epoch": 1.67,
      "learning_rate": 1.2009773976786807e-05,
      "loss": 1.3738,
      "step": 705
    },
    {
      "epoch": 1.67,
      "learning_rate": 1.19975565058033e-05,
      "loss": 0.9293,
      "step": 706
    },
    {
      "epoch": 1.67,
      "learning_rate": 1.1985339034819793e-05,
      "loss": 0.7363,
      "step": 707
    },
    {
      "epoch": 1.67,
      "learning_rate": 1.1973121563836287e-05,
      "loss": 0.9742,
      "step": 708
    },
    {
      "epoch": 1.68,
      "learning_rate": 1.196090409285278e-05,
      "loss": 0.9045,
      "step": 709
    },
    {
      "epoch": 1.68,
      "learning_rate": 1.1948686621869274e-05,
      "loss": 1.3182,
      "step": 710
    },
    {
      "epoch": 1.68,
      "learning_rate": 1.1936469150885767e-05,
      "loss": 0.9714,
      "step": 711
    },
    {
      "epoch": 1.68,
      "learning_rate": 1.192425167990226e-05,
      "loss": 0.9094,
      "step": 712
    },
    {
      "epoch": 1.69,
      "learning_rate": 1.1912034208918754e-05,
      "loss": 1.4162,
      "step": 713
    },
    {
      "epoch": 1.69,
      "learning_rate": 1.1899816737935247e-05,
      "loss": 0.8216,
      "step": 714
    },
    {
      "epoch": 1.69,
      "learning_rate": 1.1887599266951742e-05,
      "loss": 0.8592,
      "step": 715
    },
    {
      "epoch": 1.69,
      "learning_rate": 1.1875381795968236e-05,
      "loss": 0.8486,
      "step": 716
    },
    {
      "epoch": 1.7,
      "learning_rate": 1.1863164324984729e-05,
      "loss": 0.7479,
      "step": 717
    },
    {
      "epoch": 1.7,
      "learning_rate": 1.1850946854001223e-05,
      "loss": 0.9526,
      "step": 718
    },
    {
      "epoch": 1.7,
      "learning_rate": 1.1838729383017716e-05,
      "loss": 1.3393,
      "step": 719
    },
    {
      "epoch": 1.7,
      "learning_rate": 1.1826511912034211e-05,
      "loss": 0.7805,
      "step": 720
    },
    {
      "epoch": 1.7,
      "learning_rate": 1.1814294441050704e-05,
      "loss": 0.8528,
      "step": 721
    },
    {
      "epoch": 1.71,
      "learning_rate": 1.1802076970067198e-05,
      "loss": 1.4923,
      "step": 722
    },
    {
      "epoch": 1.71,
      "learning_rate": 1.1789859499083691e-05,
      "loss": 0.9028,
      "step": 723
    },
    {
      "epoch": 1.71,
      "learning_rate": 1.1777642028100185e-05,
      "loss": 0.8582,
      "step": 724
    },
    {
      "epoch": 1.71,
      "learning_rate": 1.1765424557116678e-05,
      "loss": 0.7968,
      "step": 725
    },
    {
      "epoch": 1.72,
      "learning_rate": 1.1753207086133171e-05,
      "loss": 0.8106,
      "step": 726
    },
    {
      "epoch": 1.72,
      "learning_rate": 1.1740989615149665e-05,
      "loss": 1.3575,
      "step": 727
    },
    {
      "epoch": 1.72,
      "learning_rate": 1.1728772144166158e-05,
      "loss": 0.7235,
      "step": 728
    },
    {
      "epoch": 1.72,
      "learning_rate": 1.1716554673182652e-05,
      "loss": 0.8314,
      "step": 729
    },
    {
      "epoch": 1.73,
      "learning_rate": 1.1704337202199145e-05,
      "loss": 1.0177,
      "step": 730
    },
    {
      "epoch": 1.73,
      "learning_rate": 1.1692119731215638e-05,
      "loss": 0.8382,
      "step": 731
    },
    {
      "epoch": 1.73,
      "learning_rate": 1.1679902260232132e-05,
      "loss": 0.95,
      "step": 732
    },
    {
      "epoch": 1.73,
      "learning_rate": 1.1667684789248625e-05,
      "loss": 0.6696,
      "step": 733
    },
    {
      "epoch": 1.74,
      "learning_rate": 1.1655467318265119e-05,
      "loss": 1.0264,
      "step": 734
    },
    {
      "epoch": 1.74,
      "learning_rate": 1.1643249847281615e-05,
      "loss": 1.1022,
      "step": 735
    },
    {
      "epoch": 1.74,
      "learning_rate": 1.1631032376298109e-05,
      "loss": 0.9275,
      "step": 736
    },
    {
      "epoch": 1.74,
      "learning_rate": 1.1618814905314602e-05,
      "loss": 0.7903,
      "step": 737
    },
    {
      "epoch": 1.75,
      "learning_rate": 1.1606597434331096e-05,
      "loss": 0.9102,
      "step": 738
    },
    {
      "epoch": 1.75,
      "learning_rate": 1.1594379963347589e-05,
      "loss": 0.8238,
      "step": 739
    },
    {
      "epoch": 1.75,
      "learning_rate": 1.1582162492364082e-05,
      "loss": 0.8609,
      "step": 740
    },
    {
      "epoch": 1.75,
      "learning_rate": 1.1569945021380576e-05,
      "loss": 0.8274,
      "step": 741
    },
    {
      "epoch": 1.75,
      "learning_rate": 1.155772755039707e-05,
      "loss": 0.9712,
      "step": 742
    },
    {
      "epoch": 1.76,
      "learning_rate": 1.1545510079413563e-05,
      "loss": 0.9417,
      "step": 743
    },
    {
      "epoch": 1.76,
      "learning_rate": 1.1533292608430056e-05,
      "loss": 0.8531,
      "step": 744
    },
    {
      "epoch": 1.76,
      "learning_rate": 1.152107513744655e-05,
      "loss": 0.8248,
      "step": 745
    },
    {
      "epoch": 1.76,
      "learning_rate": 1.1508857666463043e-05,
      "loss": 0.9558,
      "step": 746
    },
    {
      "epoch": 1.77,
      "learning_rate": 1.1496640195479536e-05,
      "loss": 1.0065,
      "step": 747
    },
    {
      "epoch": 1.77,
      "learning_rate": 1.148442272449603e-05,
      "loss": 1.0631,
      "step": 748
    },
    {
      "epoch": 1.77,
      "learning_rate": 1.1472205253512523e-05,
      "loss": 0.8718,
      "step": 749
    },
    {
      "epoch": 1.77,
      "learning_rate": 1.1459987782529018e-05,
      "loss": 0.9382,
      "step": 750
    },
    {
      "epoch": 1.78,
      "learning_rate": 1.1447770311545512e-05,
      "loss": 0.9873,
      "step": 751
    },
    {
      "epoch": 1.78,
      "learning_rate": 1.1435552840562005e-05,
      "loss": 0.7277,
      "step": 752
    },
    {
      "epoch": 1.78,
      "learning_rate": 1.1423335369578498e-05,
      "loss": 0.7512,
      "step": 753
    },
    {
      "epoch": 1.78,
      "learning_rate": 1.1411117898594992e-05,
      "loss": 0.8206,
      "step": 754
    },
    {
      "epoch": 1.79,
      "learning_rate": 1.1398900427611485e-05,
      "loss": 0.8375,
      "step": 755
    },
    {
      "epoch": 1.79,
      "learning_rate": 1.1386682956627979e-05,
      "loss": 0.9206,
      "step": 756
    },
    {
      "epoch": 1.79,
      "learning_rate": 1.1374465485644472e-05,
      "loss": 0.869,
      "step": 757
    },
    {
      "epoch": 1.79,
      "learning_rate": 1.1362248014660965e-05,
      "loss": 0.8234,
      "step": 758
    },
    {
      "epoch": 1.79,
      "learning_rate": 1.1350030543677459e-05,
      "loss": 0.7956,
      "step": 759
    },
    {
      "epoch": 1.8,
      "learning_rate": 1.1337813072693952e-05,
      "loss": 0.7696,
      "step": 760
    },
    {
      "epoch": 1.8,
      "learning_rate": 1.1325595601710446e-05,
      "loss": 0.9689,
      "step": 761
    },
    {
      "epoch": 1.8,
      "learning_rate": 1.1313378130726939e-05,
      "loss": 0.8306,
      "step": 762
    },
    {
      "epoch": 1.8,
      "learning_rate": 1.1301160659743432e-05,
      "loss": 0.7371,
      "step": 763
    },
    {
      "epoch": 1.81,
      "learning_rate": 1.1288943188759926e-05,
      "loss": 0.8887,
      "step": 764
    },
    {
      "epoch": 1.81,
      "learning_rate": 1.1276725717776423e-05,
      "loss": 0.7669,
      "step": 765
    },
    {
      "epoch": 1.81,
      "learning_rate": 1.1264508246792916e-05,
      "loss": 1.0305,
      "step": 766
    },
    {
      "epoch": 1.81,
      "learning_rate": 1.125229077580941e-05,
      "loss": 0.5774,
      "step": 767
    },
    {
      "epoch": 1.82,
      "learning_rate": 1.1240073304825903e-05,
      "loss": 0.9091,
      "step": 768
    },
    {
      "epoch": 1.82,
      "learning_rate": 1.1227855833842396e-05,
      "loss": 1.001,
      "step": 769
    },
    {
      "epoch": 1.82,
      "learning_rate": 1.121563836285889e-05,
      "loss": 0.9184,
      "step": 770
    },
    {
      "epoch": 1.82,
      "learning_rate": 1.1203420891875383e-05,
      "loss": 0.8154,
      "step": 771
    },
    {
      "epoch": 1.83,
      "learning_rate": 1.1191203420891876e-05,
      "loss": 0.9085,
      "step": 772
    },
    {
      "epoch": 1.83,
      "learning_rate": 1.117898594990837e-05,
      "loss": 0.892,
      "step": 773
    },
    {
      "epoch": 1.83,
      "learning_rate": 1.1166768478924863e-05,
      "loss": 0.7947,
      "step": 774
    },
    {
      "epoch": 1.83,
      "learning_rate": 1.1154551007941357e-05,
      "loss": 0.8033,
      "step": 775
    },
    {
      "epoch": 1.83,
      "learning_rate": 1.114233353695785e-05,
      "loss": 1.4454,
      "step": 776
    },
    {
      "epoch": 1.84,
      "learning_rate": 1.1130116065974343e-05,
      "loss": 1.1096,
      "step": 777
    },
    {
      "epoch": 1.84,
      "learning_rate": 1.1117898594990837e-05,
      "loss": 0.8875,
      "step": 778
    },
    {
      "epoch": 1.84,
      "learning_rate": 1.110568112400733e-05,
      "loss": 0.7709,
      "step": 779
    },
    {
      "epoch": 1.84,
      "learning_rate": 1.1093463653023825e-05,
      "loss": 0.8432,
      "step": 780
    },
    {
      "epoch": 1.85,
      "learning_rate": 1.1081246182040319e-05,
      "loss": 0.887,
      "step": 781
    },
    {
      "epoch": 1.85,
      "learning_rate": 1.1069028711056812e-05,
      "loss": 0.9143,
      "step": 782
    },
    {
      "epoch": 1.85,
      "learning_rate": 1.1056811240073305e-05,
      "loss": 0.8532,
      "step": 783
    },
    {
      "epoch": 1.85,
      "learning_rate": 1.10445937690898e-05,
      "loss": 0.7214,
      "step": 784
    },
    {
      "epoch": 1.86,
      "learning_rate": 1.1032376298106294e-05,
      "loss": 0.8941,
      "step": 785
    },
    {
      "epoch": 1.86,
      "learning_rate": 1.1020158827122787e-05,
      "loss": 0.6898,
      "step": 786
    },
    {
      "epoch": 1.86,
      "learning_rate": 1.100794135613928e-05,
      "loss": 0.7767,
      "step": 787
    },
    {
      "epoch": 1.86,
      "learning_rate": 1.0995723885155774e-05,
      "loss": 1.0031,
      "step": 788
    },
    {
      "epoch": 1.87,
      "learning_rate": 1.0983506414172268e-05,
      "loss": 0.7288,
      "step": 789
    },
    {
      "epoch": 1.87,
      "learning_rate": 1.0971288943188761e-05,
      "loss": 0.902,
      "step": 790
    },
    {
      "epoch": 1.87,
      "learning_rate": 1.0959071472205254e-05,
      "loss": 1.6741,
      "step": 791
    },
    {
      "epoch": 1.87,
      "learning_rate": 1.0946854001221748e-05,
      "loss": 0.8963,
      "step": 792
    },
    {
      "epoch": 1.88,
      "learning_rate": 1.0934636530238241e-05,
      "loss": 0.8454,
      "step": 793
    },
    {
      "epoch": 1.88,
      "learning_rate": 1.0922419059254735e-05,
      "loss": 0.7937,
      "step": 794
    },
    {
      "epoch": 1.88,
      "learning_rate": 1.091020158827123e-05,
      "loss": 0.9366,
      "step": 795
    },
    {
      "epoch": 1.88,
      "learning_rate": 1.0897984117287723e-05,
      "loss": 0.7642,
      "step": 796
    },
    {
      "epoch": 1.88,
      "learning_rate": 1.0885766646304216e-05,
      "loss": 1.0383,
      "step": 797
    },
    {
      "epoch": 1.89,
      "learning_rate": 1.087354917532071e-05,
      "loss": 0.8115,
      "step": 798
    },
    {
      "epoch": 1.89,
      "learning_rate": 1.0861331704337203e-05,
      "loss": 0.7814,
      "step": 799
    },
    {
      "epoch": 1.89,
      "learning_rate": 1.0849114233353697e-05,
      "loss": 0.7422,
      "step": 800
    },
    {
      "epoch": 1.89,
      "learning_rate": 1.083689676237019e-05,
      "loss": 0.9207,
      "step": 801
    },
    {
      "epoch": 1.9,
      "learning_rate": 1.0824679291386684e-05,
      "loss": 0.978,
      "step": 802
    },
    {
      "epoch": 1.9,
      "learning_rate": 1.0812461820403177e-05,
      "loss": 0.8223,
      "step": 803
    },
    {
      "epoch": 1.9,
      "learning_rate": 1.080024434941967e-05,
      "loss": 0.7823,
      "step": 804
    },
    {
      "epoch": 1.9,
      "learning_rate": 1.0788026878436164e-05,
      "loss": 1.227,
      "step": 805
    },
    {
      "epoch": 1.91,
      "learning_rate": 1.0775809407452657e-05,
      "loss": 0.7274,
      "step": 806
    },
    {
      "epoch": 1.91,
      "learning_rate": 1.076359193646915e-05,
      "loss": 1.0557,
      "step": 807
    },
    {
      "epoch": 1.91,
      "learning_rate": 1.0751374465485644e-05,
      "loss": 0.6012,
      "step": 808
    },
    {
      "epoch": 1.91,
      "learning_rate": 1.0739156994502137e-05,
      "loss": 0.7529,
      "step": 809
    },
    {
      "epoch": 1.92,
      "learning_rate": 1.0726939523518634e-05,
      "loss": 0.697,
      "step": 810
    },
    {
      "epoch": 1.92,
      "learning_rate": 1.0714722052535128e-05,
      "loss": 0.7619,
      "step": 811
    },
    {
      "epoch": 1.92,
      "learning_rate": 1.0702504581551621e-05,
      "loss": 0.6546,
      "step": 812
    },
    {
      "epoch": 1.92,
      "learning_rate": 1.0690287110568114e-05,
      "loss": 1.0423,
      "step": 813
    },
    {
      "epoch": 1.92,
      "learning_rate": 1.0678069639584608e-05,
      "loss": 0.8163,
      "step": 814
    },
    {
      "epoch": 1.93,
      "learning_rate": 1.0665852168601101e-05,
      "loss": 0.8361,
      "step": 815
    },
    {
      "epoch": 1.93,
      "learning_rate": 1.0653634697617595e-05,
      "loss": 1.0374,
      "step": 816
    },
    {
      "epoch": 1.93,
      "learning_rate": 1.0641417226634088e-05,
      "loss": 0.6715,
      "step": 817
    },
    {
      "epoch": 1.93,
      "learning_rate": 1.0629199755650581e-05,
      "loss": 0.8869,
      "step": 818
    },
    {
      "epoch": 1.94,
      "learning_rate": 1.0616982284667075e-05,
      "loss": 0.7648,
      "step": 819
    },
    {
      "epoch": 1.94,
      "learning_rate": 1.0604764813683568e-05,
      "loss": 0.617,
      "step": 820
    },
    {
      "epoch": 1.94,
      "learning_rate": 1.0592547342700062e-05,
      "loss": 0.8328,
      "step": 821
    },
    {
      "epoch": 1.94,
      "learning_rate": 1.0580329871716555e-05,
      "loss": 0.6538,
      "step": 822
    },
    {
      "epoch": 1.95,
      "learning_rate": 1.0568112400733048e-05,
      "loss": 0.7676,
      "step": 823
    },
    {
      "epoch": 1.95,
      "learning_rate": 1.0555894929749542e-05,
      "loss": 0.8372,
      "step": 824
    },
    {
      "epoch": 1.95,
      "learning_rate": 1.0543677458766037e-05,
      "loss": 1.4951,
      "step": 825
    },
    {
      "epoch": 1.95,
      "learning_rate": 1.053145998778253e-05,
      "loss": 0.8325,
      "step": 826
    },
    {
      "epoch": 1.96,
      "learning_rate": 1.0519242516799024e-05,
      "loss": 0.7662,
      "step": 827
    },
    {
      "epoch": 1.96,
      "learning_rate": 1.0507025045815517e-05,
      "loss": 0.7526,
      "step": 828
    },
    {
      "epoch": 1.96,
      "learning_rate": 1.049480757483201e-05,
      "loss": 0.9022,
      "step": 829
    },
    {
      "epoch": 1.96,
      "learning_rate": 1.0482590103848504e-05,
      "loss": 0.7836,
      "step": 830
    },
    {
      "epoch": 1.96,
      "learning_rate": 1.0470372632864997e-05,
      "loss": 0.8463,
      "step": 831
    },
    {
      "epoch": 1.97,
      "learning_rate": 1.045815516188149e-05,
      "loss": 1.0518,
      "step": 832
    },
    {
      "epoch": 1.97,
      "learning_rate": 1.0445937690897984e-05,
      "loss": 0.6923,
      "step": 833
    },
    {
      "epoch": 1.97,
      "learning_rate": 1.0433720219914477e-05,
      "loss": 1.0485,
      "step": 834
    },
    {
      "epoch": 1.97,
      "learning_rate": 1.0421502748930973e-05,
      "loss": 0.9231,
      "step": 835
    },
    {
      "epoch": 1.98,
      "learning_rate": 1.0409285277947466e-05,
      "loss": 0.8881,
      "step": 836
    },
    {
      "epoch": 1.98,
      "learning_rate": 1.039706780696396e-05,
      "loss": 0.9175,
      "step": 837
    },
    {
      "epoch": 1.98,
      "learning_rate": 1.0384850335980453e-05,
      "loss": 0.8728,
      "step": 838
    },
    {
      "epoch": 1.98,
      "learning_rate": 1.0372632864996946e-05,
      "loss": 0.6343,
      "step": 839
    },
    {
      "epoch": 1.99,
      "learning_rate": 1.0360415394013441e-05,
      "loss": 0.7374,
      "step": 840
    },
    {
      "epoch": 1.99,
      "learning_rate": 1.0348197923029935e-05,
      "loss": 1.014,
      "step": 841
    },
    {
      "epoch": 1.99,
      "learning_rate": 1.0335980452046428e-05,
      "loss": 0.7877,
      "step": 842
    },
    {
      "epoch": 1.99,
      "learning_rate": 1.0323762981062921e-05,
      "loss": 0.8771,
      "step": 843
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.0311545510079415e-05,
      "loss": 0.7503,
      "step": 844
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.0299328039095908e-05,
      "loss": 0.9533,
      "step": 845
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 1688,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 4.061471742005084e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}