{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.097173828881146,
  "eval_steps": 500,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 6.020599913279623e-06,
      "loss": 1.2605,
      "step": 2
    },
    {
      "epoch": 0.01,
      "learning_rate": 1.2041199826559246e-05,
      "loss": 1.1266,
      "step": 4
    },
    {
      "epoch": 0.02,
      "learning_rate": 1.5563025007672873e-05,
      "loss": 0.9129,
      "step": 6
    },
    {
      "epoch": 0.02,
      "learning_rate": 1.806179973983887e-05,
      "loss": 0.7453,
      "step": 8
    },
    {
      "epoch": 0.03,
      "learning_rate": 1.9999999999999998e-05,
      "loss": 0.6517,
      "step": 10
    },
    {
      "epoch": 0.04,
      "learning_rate": 2e-05,
      "loss": 0.5941,
      "step": 12
    },
    {
      "epoch": 0.04,
      "learning_rate": 2e-05,
      "loss": 0.5568,
      "step": 14
    },
    {
      "epoch": 0.05,
      "learning_rate": 2e-05,
      "loss": 0.5374,
      "step": 16
    },
    {
      "epoch": 0.06,
      "learning_rate": 2e-05,
      "loss": 0.5121,
      "step": 18
    },
    {
      "epoch": 0.06,
      "learning_rate": 2e-05,
      "loss": 0.5023,
      "step": 20
    },
    {
      "epoch": 0.07,
      "learning_rate": 2e-05,
      "loss": 0.4889,
      "step": 22
    },
    {
      "epoch": 0.07,
      "learning_rate": 2e-05,
      "loss": 0.4846,
      "step": 24
    },
    {
      "epoch": 0.08,
      "learning_rate": 2e-05,
      "loss": 0.4739,
      "step": 26
    },
    {
      "epoch": 0.09,
      "learning_rate": 2e-05,
      "loss": 0.4667,
      "step": 28
    },
    {
      "epoch": 0.09,
      "learning_rate": 2e-05,
      "loss": 0.4566,
      "step": 30
    },
    {
      "epoch": 0.1,
      "learning_rate": 2e-05,
      "loss": 0.4544,
      "step": 32
    },
    {
      "epoch": 0.11,
      "learning_rate": 2e-05,
      "loss": 0.4409,
      "step": 34
    },
    {
      "epoch": 0.11,
      "learning_rate": 2e-05,
      "loss": 0.4397,
      "step": 36
    },
    {
      "epoch": 0.12,
      "learning_rate": 2e-05,
      "loss": 0.4408,
      "step": 38
    },
    {
      "epoch": 0.12,
      "learning_rate": 2e-05,
      "loss": 0.4376,
      "step": 40
    },
    {
      "epoch": 0.13,
      "learning_rate": 2e-05,
      "loss": 0.4279,
      "step": 42
    },
    {
      "epoch": 0.14,
      "learning_rate": 2e-05,
      "loss": 0.4267,
      "step": 44
    },
    {
      "epoch": 0.14,
      "learning_rate": 2e-05,
      "loss": 0.4209,
      "step": 46
    },
    {
      "epoch": 0.15,
      "learning_rate": 2e-05,
      "loss": 0.4169,
      "step": 48
    },
    {
      "epoch": 0.15,
      "learning_rate": 2e-05,
      "loss": 0.4171,
      "step": 50
    },
    {
      "epoch": 0.16,
      "learning_rate": 2e-05,
      "loss": 0.4162,
      "step": 52
    },
    {
      "epoch": 0.17,
      "learning_rate": 2e-05,
      "loss": 0.4187,
      "step": 54
    },
    {
      "epoch": 0.17,
      "learning_rate": 2e-05,
      "loss": 0.4138,
      "step": 56
    },
    {
      "epoch": 0.18,
      "learning_rate": 2e-05,
      "loss": 0.4184,
      "step": 58
    },
    {
      "epoch": 0.19,
      "learning_rate": 2e-05,
      "loss": 0.4118,
      "step": 60
    },
    {
      "epoch": 0.19,
      "learning_rate": 2e-05,
      "loss": 0.4039,
      "step": 62
    },
    {
      "epoch": 0.2,
      "learning_rate": 2e-05,
      "loss": 0.404,
      "step": 64
    },
    {
      "epoch": 0.2,
      "learning_rate": 2e-05,
      "loss": 0.4031,
      "step": 66
    },
    {
      "epoch": 0.21,
      "learning_rate": 2e-05,
      "loss": 0.3959,
      "step": 68
    },
    {
      "epoch": 0.22,
      "learning_rate": 2e-05,
      "loss": 0.4046,
      "step": 70
    },
    {
      "epoch": 0.22,
      "learning_rate": 2e-05,
      "loss": 0.3943,
      "step": 72
    },
    {
      "epoch": 0.23,
      "learning_rate": 2e-05,
      "loss": 0.3995,
      "step": 74
    },
    {
      "epoch": 0.24,
      "learning_rate": 2e-05,
      "loss": 0.3925,
      "step": 76
    },
    {
      "epoch": 0.24,
      "learning_rate": 2e-05,
      "loss": 0.3908,
      "step": 78
    },
    {
      "epoch": 0.25,
      "learning_rate": 2e-05,
      "loss": 0.3901,
      "step": 80
    },
    {
      "epoch": 0.25,
      "learning_rate": 2e-05,
      "loss": 0.3848,
      "step": 82
    },
    {
      "epoch": 0.26,
      "learning_rate": 2e-05,
      "loss": 0.3877,
      "step": 84
    },
    {
      "epoch": 0.27,
      "learning_rate": 2e-05,
      "loss": 0.3857,
      "step": 86
    },
    {
      "epoch": 0.27,
      "learning_rate": 2e-05,
      "loss": 0.381,
      "step": 88
    },
    {
      "epoch": 0.28,
      "learning_rate": 2e-05,
      "loss": 0.3841,
      "step": 90
    },
    {
      "epoch": 0.28,
      "learning_rate": 2e-05,
      "loss": 0.3895,
      "step": 92
    },
    {
      "epoch": 0.29,
      "learning_rate": 2e-05,
      "loss": 0.3774,
      "step": 94
    },
    {
      "epoch": 0.3,
      "learning_rate": 2e-05,
      "loss": 0.3794,
      "step": 96
    },
    {
      "epoch": 0.3,
      "learning_rate": 2e-05,
      "loss": 0.3761,
      "step": 98
    },
    {
      "epoch": 0.31,
      "learning_rate": 2e-05,
      "loss": 0.376,
      "step": 100
    },
    {
      "epoch": 0.32,
      "learning_rate": 2e-05,
      "loss": 0.3808,
      "step": 102
    },
    {
      "epoch": 0.32,
      "learning_rate": 2e-05,
      "loss": 0.3791,
      "step": 104
    },
    {
      "epoch": 0.33,
      "learning_rate": 2e-05,
      "loss": 0.3716,
      "step": 106
    },
    {
      "epoch": 0.33,
      "learning_rate": 2e-05,
      "loss": 0.377,
      "step": 108
    },
    {
      "epoch": 0.34,
      "learning_rate": 2e-05,
      "loss": 0.3684,
      "step": 110
    },
    {
      "epoch": 0.35,
      "learning_rate": 2e-05,
      "loss": 0.3665,
      "step": 112
    },
    {
      "epoch": 0.35,
      "learning_rate": 2e-05,
      "loss": 0.3705,
      "step": 114
    },
    {
      "epoch": 0.36,
      "learning_rate": 2e-05,
      "loss": 0.3719,
      "step": 116
    },
    {
      "epoch": 0.37,
      "learning_rate": 2e-05,
      "loss": 0.371,
      "step": 118
    },
    {
      "epoch": 0.37,
      "learning_rate": 2e-05,
      "loss": 0.3686,
      "step": 120
    },
    {
      "epoch": 0.38,
      "learning_rate": 2e-05,
      "loss": 0.3689,
      "step": 122
    },
    {
      "epoch": 0.38,
      "learning_rate": 2e-05,
      "loss": 0.3519,
      "step": 124
    },
    {
      "epoch": 0.39,
      "learning_rate": 2e-05,
      "loss": 0.3601,
      "step": 126
    },
    {
      "epoch": 0.4,
      "learning_rate": 2e-05,
      "loss": 0.3615,
      "step": 128
    },
    {
      "epoch": 0.4,
      "learning_rate": 2e-05,
      "loss": 0.3602,
      "step": 130
    },
    {
      "epoch": 0.41,
      "learning_rate": 2e-05,
      "loss": 0.3618,
      "step": 132
    },
    {
      "epoch": 0.42,
      "learning_rate": 2e-05,
      "loss": 0.36,
      "step": 134
    },
    {
      "epoch": 0.42,
      "learning_rate": 2e-05,
      "loss": 0.3538,
      "step": 136
    },
    {
      "epoch": 0.43,
      "learning_rate": 2e-05,
      "loss": 0.3534,
      "step": 138
    },
    {
      "epoch": 0.43,
      "learning_rate": 2e-05,
      "loss": 0.3651,
      "step": 140
    },
    {
      "epoch": 0.44,
      "learning_rate": 2e-05,
      "loss": 0.3543,
      "step": 142
    },
    {
      "epoch": 0.45,
      "learning_rate": 2e-05,
      "loss": 0.3521,
      "step": 144
    },
    {
      "epoch": 0.45,
      "learning_rate": 2e-05,
      "loss": 0.3503,
      "step": 146
    },
    {
      "epoch": 0.46,
      "learning_rate": 2e-05,
      "loss": 0.3443,
      "step": 148
    },
    {
      "epoch": 0.46,
      "learning_rate": 2e-05,
      "loss": 0.3486,
      "step": 150
    },
    {
      "epoch": 0.47,
      "learning_rate": 2e-05,
      "loss": 0.348,
      "step": 152
    },
    {
      "epoch": 0.48,
      "learning_rate": 2e-05,
      "loss": 0.3509,
      "step": 154
    },
    {
      "epoch": 0.48,
      "learning_rate": 2e-05,
      "loss": 0.3406,
      "step": 156
    },
    {
      "epoch": 0.49,
      "learning_rate": 2e-05,
      "loss": 0.3476,
      "step": 158
    },
    {
      "epoch": 0.5,
      "learning_rate": 2e-05,
      "loss": 0.3385,
      "step": 160
    },
    {
      "epoch": 0.5,
      "learning_rate": 2e-05,
      "loss": 0.3419,
      "step": 162
    },
    {
      "epoch": 0.51,
      "learning_rate": 2e-05,
      "loss": 0.3449,
      "step": 164
    },
    {
      "epoch": 0.51,
      "learning_rate": 2e-05,
      "loss": 0.335,
      "step": 166
    },
    {
      "epoch": 0.52,
      "learning_rate": 2e-05,
      "loss": 0.3395,
      "step": 168
    },
    {
      "epoch": 0.53,
      "learning_rate": 2e-05,
      "loss": 0.3364,
      "step": 170
    },
    {
      "epoch": 0.53,
      "learning_rate": 2e-05,
      "loss": 0.3377,
      "step": 172
    },
    {
      "epoch": 0.54,
      "learning_rate": 2e-05,
      "loss": 0.339,
      "step": 174
    },
    {
      "epoch": 0.55,
      "learning_rate": 2e-05,
      "loss": 0.3358,
      "step": 176
    },
    {
      "epoch": 0.55,
      "learning_rate": 2e-05,
      "loss": 0.3402,
      "step": 178
    },
    {
      "epoch": 0.56,
      "learning_rate": 2e-05,
      "loss": 0.3324,
      "step": 180
    },
    {
      "epoch": 0.56,
      "learning_rate": 2e-05,
      "loss": 0.3363,
      "step": 182
    },
    {
      "epoch": 0.57,
      "learning_rate": 2e-05,
      "loss": 0.3329,
      "step": 184
    },
    {
      "epoch": 0.58,
      "learning_rate": 2e-05,
      "loss": 0.3282,
      "step": 186
    },
    {
      "epoch": 0.58,
      "learning_rate": 2e-05,
      "loss": 0.3363,
      "step": 188
    },
    {
      "epoch": 0.59,
      "learning_rate": 2e-05,
      "loss": 0.3334,
      "step": 190
    },
    {
      "epoch": 0.59,
      "learning_rate": 2e-05,
      "loss": 0.331,
      "step": 192
    },
    {
      "epoch": 0.6,
      "learning_rate": 2e-05,
      "loss": 0.3297,
      "step": 194
    },
    {
      "epoch": 0.61,
      "learning_rate": 2e-05,
      "loss": 0.331,
      "step": 196
    },
    {
      "epoch": 0.61,
      "learning_rate": 2e-05,
      "loss": 0.3277,
      "step": 198
    },
    {
      "epoch": 0.62,
      "learning_rate": 2e-05,
      "loss": 0.3246,
      "step": 200
    },
    {
      "epoch": 0.63,
      "learning_rate": 2e-05,
      "loss": 0.3269,
      "step": 202
    },
    {
      "epoch": 0.63,
      "learning_rate": 2e-05,
      "loss": 0.3269,
      "step": 204
    },
    {
      "epoch": 0.64,
      "learning_rate": 2e-05,
      "loss": 0.3261,
      "step": 206
    },
    {
      "epoch": 0.64,
      "learning_rate": 2e-05,
      "loss": 0.3288,
      "step": 208
    },
    {
      "epoch": 0.65,
      "learning_rate": 2e-05,
      "loss": 0.3266,
      "step": 210
    },
    {
      "epoch": 0.66,
      "learning_rate": 2e-05,
      "loss": 0.3239,
      "step": 212
    },
    {
      "epoch": 0.66,
      "learning_rate": 2e-05,
      "loss": 0.3247,
      "step": 214
    },
    {
      "epoch": 0.67,
      "learning_rate": 2e-05,
      "loss": 0.3215,
      "step": 216
    },
    {
      "epoch": 0.68,
      "learning_rate": 2e-05,
      "loss": 0.3212,
      "step": 218
    },
    {
      "epoch": 0.68,
      "learning_rate": 2e-05,
      "loss": 0.3195,
      "step": 220
    },
    {
      "epoch": 0.69,
      "learning_rate": 2e-05,
      "loss": 0.3187,
      "step": 222
    },
    {
      "epoch": 0.69,
      "learning_rate": 2e-05,
      "loss": 0.3179,
      "step": 224
    },
    {
      "epoch": 0.7,
      "learning_rate": 2e-05,
      "loss": 0.3159,
      "step": 226
    },
    {
      "epoch": 0.71,
      "learning_rate": 2e-05,
      "loss": 0.3193,
      "step": 228
    },
    {
      "epoch": 0.71,
      "learning_rate": 2e-05,
      "loss": 0.3168,
      "step": 230
    },
    {
      "epoch": 0.72,
      "learning_rate": 2e-05,
      "loss": 0.3164,
      "step": 232
    },
    {
      "epoch": 0.72,
      "learning_rate": 2e-05,
      "loss": 0.3174,
      "step": 234
    },
    {
      "epoch": 0.73,
      "learning_rate": 2e-05,
      "loss": 0.3185,
      "step": 236
    },
    {
      "epoch": 0.74,
      "learning_rate": 2e-05,
      "loss": 0.3173,
      "step": 238
    },
    {
      "epoch": 0.74,
      "learning_rate": 2e-05,
      "loss": 0.3132,
      "step": 240
    },
    {
      "epoch": 0.75,
      "learning_rate": 2e-05,
      "loss": 0.3116,
      "step": 242
    },
    {
      "epoch": 0.76,
      "learning_rate": 2e-05,
      "loss": 0.3142,
      "step": 244
    },
    {
      "epoch": 0.76,
      "learning_rate": 2e-05,
      "loss": 0.3086,
      "step": 246
    },
    {
      "epoch": 0.77,
      "learning_rate": 2e-05,
      "loss": 0.3154,
      "step": 248
    },
    {
      "epoch": 0.77,
      "learning_rate": 2e-05,
      "loss": 0.3053,
      "step": 250
    },
    {
      "epoch": 0.78,
      "learning_rate": 2e-05,
      "loss": 0.3108,
      "step": 252
    },
    {
      "epoch": 0.79,
      "learning_rate": 2e-05,
      "loss": 0.3081,
      "step": 254
    },
    {
      "epoch": 0.79,
      "learning_rate": 2e-05,
      "loss": 0.3041,
      "step": 256
    },
    {
      "epoch": 0.8,
      "learning_rate": 2e-05,
      "loss": 0.3112,
      "step": 258
    },
    {
      "epoch": 0.81,
      "learning_rate": 2e-05,
      "loss": 0.3077,
      "step": 260
    },
    {
      "epoch": 0.81,
      "learning_rate": 2e-05,
      "loss": 0.3032,
      "step": 262
    },
    {
      "epoch": 0.82,
      "learning_rate": 2e-05,
      "loss": 0.3068,
      "step": 264
    },
    {
      "epoch": 0.82,
      "learning_rate": 2e-05,
      "loss": 0.3111,
      "step": 266
    },
    {
      "epoch": 0.83,
      "learning_rate": 2e-05,
      "loss": 0.3061,
      "step": 268
    },
    {
      "epoch": 0.84,
      "learning_rate": 2e-05,
      "loss": 0.3015,
      "step": 270
    },
    {
      "epoch": 0.84,
      "learning_rate": 2e-05,
      "loss": 0.3043,
      "step": 272
    },
    {
      "epoch": 0.85,
      "learning_rate": 2e-05,
      "loss": 0.2996,
      "step": 274
    },
    {
      "epoch": 0.85,
      "learning_rate": 2e-05,
      "loss": 0.3047,
      "step": 276
    },
    {
      "epoch": 0.86,
      "learning_rate": 2e-05,
      "loss": 0.2996,
      "step": 278
    },
    {
      "epoch": 0.87,
      "learning_rate": 2e-05,
      "loss": 0.3038,
      "step": 280
    },
    {
      "epoch": 0.87,
      "learning_rate": 2e-05,
      "loss": 0.2977,
      "step": 282
    },
    {
      "epoch": 0.88,
      "learning_rate": 2e-05,
      "loss": 0.2981,
      "step": 284
    },
    {
      "epoch": 0.89,
      "learning_rate": 2e-05,
      "loss": 0.3054,
      "step": 286
    },
    {
      "epoch": 0.89,
      "learning_rate": 2e-05,
      "loss": 0.3002,
      "step": 288
    },
    {
      "epoch": 0.9,
      "learning_rate": 2e-05,
      "loss": 0.3028,
      "step": 290
    },
    {
      "epoch": 0.9,
      "learning_rate": 2e-05,
      "loss": 0.2942,
      "step": 292
    },
    {
      "epoch": 0.91,
      "learning_rate": 2e-05,
      "loss": 0.2966,
      "step": 294
    },
    {
      "epoch": 0.92,
      "learning_rate": 2e-05,
      "loss": 0.2955,
      "step": 296
    },
    {
      "epoch": 0.92,
      "learning_rate": 2e-05,
      "loss": 0.3015,
      "step": 298
    },
    {
      "epoch": 0.93,
      "learning_rate": 2e-05,
      "loss": 0.2933,
      "step": 300
    },
    {
      "epoch": 0.94,
      "learning_rate": 2e-05,
      "loss": 0.2935,
      "step": 302
    },
    {
      "epoch": 0.94,
      "learning_rate": 2e-05,
      "loss": 0.2979,
      "step": 304
    },
    {
      "epoch": 0.95,
      "learning_rate": 2e-05,
      "loss": 0.2951,
      "step": 306
    },
    {
      "epoch": 0.95,
      "learning_rate": 2e-05,
      "loss": 0.2901,
      "step": 308
    },
    {
      "epoch": 0.96,
      "learning_rate": 2e-05,
      "loss": 0.2885,
      "step": 310
    },
    {
      "epoch": 0.97,
      "learning_rate": 2e-05,
      "loss": 0.2897,
      "step": 312
    },
    {
      "epoch": 0.97,
      "learning_rate": 2e-05,
      "loss": 0.291,
      "step": 314
    },
    {
      "epoch": 0.98,
      "learning_rate": 2e-05,
      "loss": 0.2919,
      "step": 316
    },
    {
      "epoch": 0.98,
      "learning_rate": 2e-05,
      "loss": 0.2869,
      "step": 318
    },
    {
      "epoch": 0.99,
      "learning_rate": 2e-05,
      "loss": 0.2873,
      "step": 320
    },
    {
      "epoch": 1.0,
      "learning_rate": 2e-05,
      "loss": 0.2913,
      "step": 322
    },
    {
      "epoch": 1.0,
      "learning_rate": 2e-05,
      "loss": 0.2778,
      "step": 324
    },
    {
      "epoch": 1.01,
      "learning_rate": 2e-05,
      "loss": 0.2778,
      "step": 326
    },
    {
      "epoch": 1.02,
      "learning_rate": 2e-05,
      "loss": 0.2808,
      "step": 328
    },
    {
      "epoch": 1.02,
      "learning_rate": 2e-05,
      "loss": 0.2713,
      "step": 330
    },
    {
      "epoch": 1.03,
      "learning_rate": 2e-05,
      "loss": 0.2719,
      "step": 332
    },
    {
      "epoch": 1.03,
      "learning_rate": 2e-05,
      "loss": 0.2749,
      "step": 334
    },
    {
      "epoch": 1.04,
      "learning_rate": 2e-05,
      "loss": 0.269,
      "step": 336
    },
    {
      "epoch": 1.05,
      "learning_rate": 2e-05,
      "loss": 0.272,
      "step": 338
    },
    {
      "epoch": 1.05,
      "learning_rate": 2e-05,
      "loss": 0.2662,
      "step": 340
    },
    {
      "epoch": 1.06,
      "learning_rate": 2e-05,
      "loss": 0.2745,
      "step": 342
    },
    {
      "epoch": 1.07,
      "learning_rate": 2e-05,
      "loss": 0.2697,
      "step": 344
    },
    {
      "epoch": 1.07,
      "learning_rate": 2e-05,
      "loss": 0.2725,
      "step": 346
    },
    {
      "epoch": 1.08,
      "learning_rate": 2e-05,
      "loss": 0.2697,
      "step": 348
    },
    {
      "epoch": 1.08,
      "learning_rate": 2e-05,
      "loss": 0.2783,
      "step": 350
    },
    {
      "epoch": 1.09,
      "learning_rate": 2e-05,
      "loss": 0.2699,
      "step": 352
    },
    {
      "epoch": 1.1,
      "learning_rate": 2e-05,
      "loss": 0.2677,
      "step": 354
    },
    {
      "epoch": 1.1,
      "learning_rate": 2e-05,
      "loss": 0.2667,
      "step": 356
    },
    {
      "epoch": 1.11,
      "learning_rate": 2e-05,
      "loss": 0.2746,
      "step": 358
    },
    {
      "epoch": 1.11,
      "learning_rate": 2e-05,
      "loss": 0.2714,
      "step": 360
    },
    {
      "epoch": 1.12,
      "learning_rate": 2e-05,
      "loss": 0.2677,
      "step": 362
    },
    {
      "epoch": 1.13,
      "learning_rate": 2e-05,
      "loss": 0.2656,
      "step": 364
    },
    {
      "epoch": 1.13,
      "learning_rate": 2e-05,
      "loss": 0.2667,
      "step": 366
    },
    {
      "epoch": 1.14,
      "learning_rate": 2e-05,
      "loss": 0.266,
      "step": 368
    },
    {
      "epoch": 1.15,
      "learning_rate": 2e-05,
      "loss": 0.2691,
      "step": 370
    },
    {
      "epoch": 1.15,
      "learning_rate": 2e-05,
      "loss": 0.2671,
      "step": 372
    },
    {
      "epoch": 1.16,
      "learning_rate": 2e-05,
      "loss": 0.2679,
      "step": 374
    },
    {
      "epoch": 1.16,
      "learning_rate": 2e-05,
      "loss": 0.263,
      "step": 376
    },
    {
      "epoch": 1.17,
      "learning_rate": 2e-05,
      "loss": 0.262,
      "step": 378
    },
    {
      "epoch": 1.18,
      "learning_rate": 2e-05,
      "loss": 0.2668,
      "step": 380
    },
    {
      "epoch": 1.18,
      "learning_rate": 2e-05,
      "loss": 0.265,
      "step": 382
    },
    {
      "epoch": 1.19,
      "learning_rate": 2e-05,
      "loss": 0.2674,
      "step": 384
    },
    {
      "epoch": 1.2,
      "learning_rate": 2e-05,
      "loss": 0.2614,
      "step": 386
    },
    {
      "epoch": 1.2,
      "learning_rate": 2e-05,
      "loss": 0.2612,
      "step": 388
    },
    {
      "epoch": 1.21,
      "learning_rate": 2e-05,
      "loss": 0.2582,
      "step": 390
    },
    {
      "epoch": 1.21,
      "learning_rate": 2e-05,
      "loss": 0.2662,
      "step": 392
    },
    {
      "epoch": 1.22,
      "learning_rate": 2e-05,
      "loss": 0.2652,
      "step": 394
    },
    {
      "epoch": 1.23,
      "learning_rate": 2e-05,
      "loss": 0.2618,
      "step": 396
    },
    {
      "epoch": 1.23,
      "learning_rate": 2e-05,
      "loss": 0.2647,
      "step": 398
    },
    {
      "epoch": 1.24,
      "learning_rate": 2e-05,
      "loss": 0.2592,
      "step": 400
    },
    {
      "epoch": 1.25,
      "learning_rate": 2e-05,
      "loss": 0.2633,
      "step": 402
    },
    {
      "epoch": 1.25,
      "learning_rate": 2e-05,
      "loss": 0.2651,
      "step": 404
    },
    {
      "epoch": 1.26,
      "learning_rate": 2e-05,
      "loss": 0.2573,
      "step": 406
    },
    {
      "epoch": 1.26,
      "learning_rate": 2e-05,
      "loss": 0.2633,
      "step": 408
    },
    {
      "epoch": 1.27,
      "learning_rate": 2e-05,
      "loss": 0.2537,
      "step": 410
    },
    {
      "epoch": 1.28,
      "learning_rate": 2e-05,
      "loss": 0.2625,
      "step": 412
    },
    {
      "epoch": 1.28,
      "learning_rate": 2e-05,
      "loss": 0.2636,
      "step": 414
    },
    {
      "epoch": 1.29,
      "learning_rate": 2e-05,
      "loss": 0.2549,
      "step": 416
    },
    {
      "epoch": 1.29,
      "learning_rate": 2e-05,
      "loss": 0.2596,
      "step": 418
    },
    {
      "epoch": 1.3,
      "learning_rate": 2e-05,
      "loss": 0.257,
      "step": 420
    },
    {
      "epoch": 1.31,
      "learning_rate": 2e-05,
      "loss": 0.2605,
      "step": 422
    },
    {
      "epoch": 1.31,
      "learning_rate": 2e-05,
      "loss": 0.2531,
      "step": 424
    },
    {
      "epoch": 1.32,
      "learning_rate": 2e-05,
      "loss": 0.2547,
      "step": 426
    },
    {
      "epoch": 1.33,
      "learning_rate": 2e-05,
      "loss": 0.255,
      "step": 428
    },
    {
      "epoch": 1.33,
      "learning_rate": 2e-05,
      "loss": 0.2518,
      "step": 430
    },
    {
      "epoch": 1.34,
      "learning_rate": 2e-05,
      "loss": 0.25,
      "step": 432
    },
    {
      "epoch": 1.34,
      "learning_rate": 2e-05,
      "loss": 0.2552,
      "step": 434
    },
    {
      "epoch": 1.35,
      "learning_rate": 2e-05,
      "loss": 0.2532,
      "step": 436
    },
    {
      "epoch": 1.36,
      "learning_rate": 2e-05,
      "loss": 0.2533,
      "step": 438
    },
    {
      "epoch": 1.36,
      "learning_rate": 2e-05,
      "loss": 0.2543,
      "step": 440
    },
    {
      "epoch": 1.37,
      "learning_rate": 2e-05,
      "loss": 0.2508,
      "step": 442
    },
    {
      "epoch": 1.38,
      "learning_rate": 2e-05,
      "loss": 0.2504,
      "step": 444
    },
    {
      "epoch": 1.38,
      "learning_rate": 2e-05,
      "loss": 0.2477,
      "step": 446
    },
    {
      "epoch": 1.39,
      "learning_rate": 2e-05,
      "loss": 0.2535,
      "step": 448
    },
    {
      "epoch": 1.39,
      "learning_rate": 2e-05,
      "loss": 0.2458,
      "step": 450
    },
    {
      "epoch": 1.4,
      "learning_rate": 2e-05,
      "loss": 0.2494,
      "step": 452
    },
    {
      "epoch": 1.41,
      "learning_rate": 2e-05,
      "loss": 0.2466,
      "step": 454
    },
    {
      "epoch": 1.41,
      "learning_rate": 2e-05,
      "loss": 0.2472,
      "step": 456
    },
    {
      "epoch": 1.42,
      "learning_rate": 2e-05,
      "loss": 0.2477,
      "step": 458
    },
    {
      "epoch": 1.42,
      "learning_rate": 2e-05,
      "loss": 0.2501,
      "step": 460
    },
    {
      "epoch": 1.43,
      "learning_rate": 2e-05,
      "loss": 0.2534,
      "step": 462
    },
    {
      "epoch": 1.44,
      "learning_rate": 2e-05,
      "loss": 0.2443,
      "step": 464
    },
    {
      "epoch": 1.44,
      "learning_rate": 2e-05,
      "loss": 0.2463,
      "step": 466
    },
    {
      "epoch": 1.45,
      "learning_rate": 2e-05,
      "loss": 0.2426,
      "step": 468
    },
    {
      "epoch": 1.46,
      "learning_rate": 2e-05,
      "loss": 0.246,
      "step": 470
    },
    {
      "epoch": 1.46,
      "learning_rate": 2e-05,
      "loss": 0.2474,
      "step": 472
    },
    {
      "epoch": 1.47,
      "learning_rate": 2e-05,
      "loss": 0.247,
      "step": 474
    },
    {
      "epoch": 1.47,
      "learning_rate": 2e-05,
      "loss": 0.241,
      "step": 476
    },
    {
      "epoch": 1.48,
      "learning_rate": 2e-05,
      "loss": 0.2498,
      "step": 478
    },
    {
      "epoch": 1.49,
      "learning_rate": 2e-05,
      "loss": 0.2443,
      "step": 480
    },
    {
      "epoch": 1.49,
      "learning_rate": 2e-05,
      "loss": 0.2503,
      "step": 482
    },
    {
      "epoch": 1.5,
      "learning_rate": 2e-05,
      "loss": 0.2483,
      "step": 484
    },
    {
      "epoch": 1.51,
      "learning_rate": 2e-05,
      "loss": 0.2467,
      "step": 486
    },
    {
      "epoch": 1.51,
      "learning_rate": 2e-05,
      "loss": 0.2473,
      "step": 488
    },
    {
      "epoch": 1.52,
      "learning_rate": 2e-05,
      "loss": 0.2412,
      "step": 490
    },
    {
      "epoch": 1.52,
      "learning_rate": 2e-05,
      "loss": 0.2401,
      "step": 492
    },
    {
      "epoch": 1.53,
      "learning_rate": 2e-05,
      "loss": 0.2448,
      "step": 494
    },
    {
      "epoch": 1.54,
      "learning_rate": 2e-05,
      "loss": 0.2373,
      "step": 496
    },
    {
      "epoch": 1.54,
      "learning_rate": 2e-05,
      "loss": 0.2425,
      "step": 498
    },
    {
      "epoch": 1.55,
      "learning_rate": 2e-05,
      "loss": 0.2375,
      "step": 500
    },
    {
      "epoch": 1.55,
      "learning_rate": 2e-05,
      "loss": 0.2441,
      "step": 502
    },
    {
      "epoch": 1.56,
      "learning_rate": 2e-05,
      "loss": 0.2383,
      "step": 504
    },
    {
      "epoch": 1.57,
      "learning_rate": 2e-05,
      "loss": 0.2471,
      "step": 506
    },
    {
      "epoch": 1.57,
      "learning_rate": 2e-05,
      "loss": 0.2385,
      "step": 508
    },
    {
      "epoch": 1.58,
      "learning_rate": 2e-05,
      "loss": 0.2385,
      "step": 510
    },
    {
      "epoch": 1.59,
      "learning_rate": 2e-05,
      "loss": 0.2411,
      "step": 512
    },
    {
      "epoch": 1.59,
      "learning_rate": 2e-05,
      "loss": 0.2346,
      "step": 514
    },
    {
      "epoch": 1.6,
      "learning_rate": 2e-05,
      "loss": 0.2373,
      "step": 516
    },
    {
      "epoch": 1.6,
      "learning_rate": 2e-05,
      "loss": 0.2404,
      "step": 518
    },
    {
      "epoch": 1.61,
      "learning_rate": 2e-05,
      "loss": 0.2403,
      "step": 520
    },
    {
      "epoch": 1.62,
      "learning_rate": 2e-05,
      "loss": 0.2413,
      "step": 522
    },
    {
      "epoch": 1.62,
      "learning_rate": 2e-05,
      "loss": 0.2323,
      "step": 524
    },
    {
      "epoch": 1.63,
      "learning_rate": 2e-05,
      "loss": 0.2376,
      "step": 526
    },
    {
      "epoch": 1.64,
      "learning_rate": 2e-05,
      "loss": 0.2362,
      "step": 528
    },
    {
      "epoch": 1.64,
      "learning_rate": 2e-05,
      "loss": 0.2359,
      "step": 530
    },
    {
      "epoch": 1.65,
      "learning_rate": 2e-05,
      "loss": 0.2376,
      "step": 532
    },
    {
      "epoch": 1.65,
      "learning_rate": 2e-05,
      "loss": 0.2351,
      "step": 534
    },
    {
      "epoch": 1.66,
      "learning_rate": 2e-05,
      "loss": 0.2341,
      "step": 536
    },
    {
      "epoch": 1.67,
      "learning_rate": 2e-05,
      "loss": 0.2345,
      "step": 538
    },
    {
      "epoch": 1.67,
      "learning_rate": 2e-05,
      "loss": 0.2354,
      "step": 540
    },
    {
      "epoch": 1.68,
      "learning_rate": 2e-05,
      "loss": 0.2365,
      "step": 542
    },
    {
      "epoch": 1.68,
      "learning_rate": 2e-05,
      "loss": 0.2335,
      "step": 544
    },
    {
      "epoch": 1.69,
      "learning_rate": 2e-05,
      "loss": 0.2342,
      "step": 546
    },
    {
      "epoch": 1.7,
      "learning_rate": 2e-05,
      "loss": 0.2319,
      "step": 548
    },
    {
      "epoch": 1.7,
      "learning_rate": 2e-05,
      "loss": 0.2388,
      "step": 550
    },
    {
      "epoch": 1.71,
      "learning_rate": 2e-05,
      "loss": 0.2362,
      "step": 552
    },
    {
      "epoch": 1.72,
      "learning_rate": 2e-05,
      "loss": 0.2342,
      "step": 554
    },
    {
      "epoch": 1.72,
      "learning_rate": 2e-05,
      "loss": 0.2282,
      "step": 556
    },
    {
      "epoch": 1.73,
      "learning_rate": 2e-05,
      "loss": 0.2354,
      "step": 558
    },
    {
      "epoch": 1.73,
      "learning_rate": 2e-05,
      "loss": 0.2337,
      "step": 560
    },
    {
      "epoch": 1.74,
      "learning_rate": 2e-05,
      "loss": 0.2286,
      "step": 562
    },
    {
      "epoch": 1.75,
      "learning_rate": 2e-05,
      "loss": 0.2323,
      "step": 564
    },
    {
      "epoch": 1.75,
      "learning_rate": 2e-05,
      "loss": 0.2298,
      "step": 566
    },
    {
      "epoch": 1.76,
      "learning_rate": 2e-05,
      "loss": 0.2302,
      "step": 568
    },
    {
      "epoch": 1.77,
      "learning_rate": 2e-05,
      "loss": 0.2296,
      "step": 570
    },
    {
      "epoch": 1.77,
      "learning_rate": 2e-05,
      "loss": 0.2329,
      "step": 572
    },
    {
      "epoch": 1.78,
      "learning_rate": 2e-05,
      "loss": 0.2298,
      "step": 574
    },
    {
      "epoch": 1.78,
      "learning_rate": 2e-05,
      "loss": 0.228,
      "step": 576
    },
    {
      "epoch": 1.79,
      "learning_rate": 2e-05,
      "loss": 0.2262,
      "step": 578
    },
    {
      "epoch": 1.8,
      "learning_rate": 2e-05,
      "loss": 0.2296,
      "step": 580
    },
    {
      "epoch": 1.8,
      "learning_rate": 2e-05,
      "loss": 0.2284,
      "step": 582
    },
    {
      "epoch": 1.81,
      "learning_rate": 2e-05,
      "loss": 0.2299,
      "step": 584
    },
    {
      "epoch": 1.81,
      "learning_rate": 2e-05,
      "loss": 0.2276,
      "step": 586
    },
    {
      "epoch": 1.82,
      "learning_rate": 2e-05,
      "loss": 0.2315,
      "step": 588
    },
    {
      "epoch": 1.83,
      "learning_rate": 2e-05,
      "loss": 0.2304,
      "step": 590
    },
    {
      "epoch": 1.83,
      "learning_rate": 2e-05,
      "loss": 0.2253,
      "step": 592
    },
    {
      "epoch": 1.84,
      "learning_rate": 2e-05,
      "loss": 0.2243,
      "step": 594
    },
    {
      "epoch": 1.85,
      "learning_rate": 2e-05,
      "loss": 0.2282,
      "step": 596
    },
    {
      "epoch": 1.85,
      "learning_rate": 2e-05,
      "loss": 0.2277,
      "step": 598
    },
    {
      "epoch": 1.86,
      "learning_rate": 2e-05,
      "loss": 0.2306,
      "step": 600
    },
    {
      "epoch": 1.86,
      "learning_rate": 2e-05,
      "loss": 0.23,
      "step": 602
    },
    {
      "epoch": 1.87,
      "learning_rate": 2e-05,
      "loss": 0.2295,
      "step": 604
    },
    {
      "epoch": 1.88,
      "learning_rate": 2e-05,
      "loss": 0.2266,
      "step": 606
    },
    {
      "epoch": 1.88,
      "learning_rate": 2e-05,
      "loss": 0.2191,
      "step": 608
    },
    {
      "epoch": 1.89,
      "learning_rate": 2e-05,
      "loss": 0.2219,
      "step": 610
    },
    {
      "epoch": 1.9,
      "learning_rate": 2e-05,
      "loss": 0.2277,
      "step": 612
    },
    {
      "epoch": 1.9,
      "learning_rate": 2e-05,
      "loss": 0.2237,
      "step": 614
    },
    {
      "epoch": 1.91,
      "learning_rate": 2e-05,
      "loss": 0.2229,
      "step": 616
    },
    {
      "epoch": 1.91,
      "learning_rate": 2e-05,
      "loss": 0.2209,
      "step": 618
    },
    {
      "epoch": 1.92,
      "learning_rate": 2e-05,
      "loss": 0.2229,
      "step": 620
    },
    {
      "epoch": 1.93,
      "learning_rate": 2e-05,
      "loss": 0.2208,
      "step": 622
    },
    {
      "epoch": 1.93,
      "learning_rate": 2e-05,
      "loss": 0.2245,
      "step": 624
    },
    {
      "epoch": 1.94,
      "learning_rate": 2e-05,
      "loss": 0.2245,
      "step": 626
    },
    {
      "epoch": 1.95,
      "learning_rate": 2e-05,
      "loss": 0.2219,
      "step": 628
    },
    {
      "epoch": 1.95,
      "learning_rate": 2e-05,
      "loss": 0.2218,
      "step": 630
    },
    {
      "epoch": 1.96,
      "learning_rate": 2e-05,
      "loss": 0.218,
      "step": 632
    },
    {
      "epoch": 1.96,
      "learning_rate": 2e-05,
      "loss": 0.2238,
      "step": 634
    },
    {
      "epoch": 1.97,
      "learning_rate": 2e-05,
      "loss": 0.2176,
      "step": 636
    },
    {
      "epoch": 1.98,
      "learning_rate": 2e-05,
      "loss": 0.2167,
      "step": 638
    },
    {
      "epoch": 1.98,
      "learning_rate": 2e-05,
      "loss": 0.2226,
      "step": 640
    },
    {
      "epoch": 1.99,
      "learning_rate": 2e-05,
      "loss": 0.2182,
      "step": 642
    },
    {
      "epoch": 1.99,
      "learning_rate": 2e-05,
      "loss": 0.2144,
      "step": 644
    },
    {
      "epoch": 2.0,
      "learning_rate": 2e-05,
      "loss": 0.2122,
      "step": 646
    },
    {
      "epoch": 2.01,
      "learning_rate": 2e-05,
      "loss": 0.201,
      "step": 648
    },
    {
      "epoch": 2.01,
      "learning_rate": 2e-05,
      "loss": 0.2033,
      "step": 650
    },
    {
      "epoch": 2.02,
      "learning_rate": 2e-05,
      "loss": 0.1941,
      "step": 652
    },
    {
      "epoch": 2.03,
      "learning_rate": 2e-05,
      "loss": 0.2006,
      "step": 654
    },
    {
      "epoch": 2.03,
      "learning_rate": 2e-05,
      "loss": 0.1975,
      "step": 656
    },
    {
      "epoch": 2.04,
      "learning_rate": 2e-05,
      "loss": 0.2006,
      "step": 658
    },
    {
      "epoch": 2.04,
      "learning_rate": 2e-05,
      "loss": 0.1975,
      "step": 660
    },
    {
      "epoch": 2.05,
      "learning_rate": 2e-05,
      "loss": 0.1962,
      "step": 662
    },
    {
      "epoch": 2.06,
      "learning_rate": 2e-05,
      "loss": 0.195,
      "step": 664
    },
    {
      "epoch": 2.06,
      "learning_rate": 2e-05,
      "loss": 0.1967,
      "step": 666
    },
    {
      "epoch": 2.07,
      "learning_rate": 2e-05,
      "loss": 0.1918,
      "step": 668
    },
    {
      "epoch": 2.08,
      "learning_rate": 2e-05,
      "loss": 0.1927,
      "step": 670
    },
    {
      "epoch": 2.08,
      "learning_rate": 2e-05,
      "loss": 0.1934,
      "step": 672
    },
    {
      "epoch": 2.09,
      "learning_rate": 2e-05,
      "loss": 0.1963,
      "step": 674
    },
    {
      "epoch": 2.09,
      "learning_rate": 2e-05,
      "loss": 0.1951,
      "step": 676
    },
    {
      "epoch": 2.1,
      "learning_rate": 2e-05,
      "loss": 0.1886,
      "step": 678
    },
    {
      "epoch": 2.11,
      "learning_rate": 2e-05,
      "loss": 0.1939,
      "step": 680
    },
    {
      "epoch": 2.11,
      "learning_rate": 2e-05,
      "loss": 0.1978,
      "step": 682
    },
    {
      "epoch": 2.12,
      "learning_rate": 2e-05,
      "loss": 0.1941,
      "step": 684
    },
    {
      "epoch": 2.12,
      "learning_rate": 2e-05,
      "loss": 0.1955,
      "step": 686
    },
    {
      "epoch": 2.13,
      "learning_rate": 2e-05,
      "loss": 0.194,
      "step": 688
    },
    {
      "epoch": 2.14,
      "learning_rate": 2e-05,
      "loss": 0.1971,
      "step": 690
    },
    {
      "epoch": 2.14,
      "learning_rate": 2e-05,
      "loss": 0.19,
      "step": 692
    },
    {
      "epoch": 2.15,
      "learning_rate": 2e-05,
      "loss": 0.1887,
      "step": 694
    },
    {
      "epoch": 2.16,
      "learning_rate": 2e-05,
      "loss": 0.1923,
      "step": 696
    },
    {
      "epoch": 2.16,
      "learning_rate": 2e-05,
      "loss": 0.1909,
      "step": 698
    },
    {
      "epoch": 2.17,
      "learning_rate": 2e-05,
      "loss": 0.1932,
      "step": 700
    },
    {
      "epoch": 2.17,
      "learning_rate": 2e-05,
      "loss": 0.1929,
      "step": 702
    },
    {
      "epoch": 2.18,
      "learning_rate": 2e-05,
      "loss": 0.1981,
      "step": 704
    },
    {
      "epoch": 2.19,
      "learning_rate": 2e-05,
      "loss": 0.1923,
      "step": 706
    },
    {
      "epoch": 2.19,
      "learning_rate": 2e-05,
      "loss": 0.1896,
      "step": 708
    },
    {
      "epoch": 2.2,
      "learning_rate": 2e-05,
      "loss": 0.1946,
      "step": 710
    },
    {
      "epoch": 2.21,
      "learning_rate": 2e-05,
      "loss": 0.1913,
      "step": 712
    },
    {
      "epoch": 2.21,
      "learning_rate": 2e-05,
      "loss": 0.1938,
      "step": 714
    },
    {
      "epoch": 2.22,
      "learning_rate": 2e-05,
      "loss": 0.1949,
      "step": 716
    },
    {
      "epoch": 2.22,
      "learning_rate": 2e-05,
      "loss": 0.1935,
      "step": 718
    },
    {
      "epoch": 2.23,
      "learning_rate": 2e-05,
      "loss": 0.1912,
      "step": 720
    },
    {
      "epoch": 2.24,
      "learning_rate": 2e-05,
      "loss": 0.1934,
      "step": 722
    },
    {
      "epoch": 2.24,
      "learning_rate": 2e-05,
      "loss": 0.1953,
      "step": 724
    },
    {
      "epoch": 2.25,
      "learning_rate": 2e-05,
      "loss": 0.1892,
      "step": 726
    },
    {
      "epoch": 2.25,
      "learning_rate": 2e-05,
      "loss": 0.1918,
      "step": 728
    },
    {
      "epoch": 2.26,
      "learning_rate": 2e-05,
      "loss": 0.1923,
      "step": 730
    },
    {
      "epoch": 2.27,
      "learning_rate": 2e-05,
      "loss": 0.1818,
      "step": 732
    },
    {
      "epoch": 2.27,
      "learning_rate": 2e-05,
      "loss": 0.1873,
      "step": 734
    },
    {
      "epoch": 2.28,
      "learning_rate": 2e-05,
      "loss": 0.1898,
      "step": 736
    },
    {
      "epoch": 2.29,
      "learning_rate": 2e-05,
      "loss": 0.1882,
      "step": 738
    },
    {
      "epoch": 2.29,
      "learning_rate": 2e-05,
      "loss": 0.1854,
      "step": 740
    },
    {
      "epoch": 2.3,
      "learning_rate": 2e-05,
      "loss": 0.189,
      "step": 742
    },
    {
      "epoch": 2.3,
      "learning_rate": 2e-05,
      "loss": 0.1919,
      "step": 744
    },
    {
      "epoch": 2.31,
      "learning_rate": 2e-05,
      "loss": 0.1902,
      "step": 746
    },
    {
      "epoch": 2.32,
      "learning_rate": 2e-05,
      "loss": 0.1865,
      "step": 748
    },
    {
      "epoch": 2.32,
      "learning_rate": 2e-05,
      "loss": 0.1858,
      "step": 750
    },
    {
      "epoch": 2.33,
      "learning_rate": 2e-05,
      "loss": 0.1833,
      "step": 752
    },
    {
      "epoch": 2.34,
      "learning_rate": 2e-05,
      "loss": 0.1829,
      "step": 754
    },
    {
      "epoch": 2.34,
      "learning_rate": 2e-05,
      "loss": 0.1845,
      "step": 756
    },
    {
      "epoch": 2.35,
      "learning_rate": 2e-05,
      "loss": 0.182,
      "step": 758
    },
    {
      "epoch": 2.35,
      "learning_rate": 2e-05,
      "loss": 0.1854,
      "step": 760
    },
    {
      "epoch": 2.36,
      "learning_rate": 2e-05,
      "loss": 0.1876,
      "step": 762
    },
    {
      "epoch": 2.37,
      "learning_rate": 2e-05,
      "loss": 0.1868,
      "step": 764
    },
    {
      "epoch": 2.37,
      "learning_rate": 2e-05,
      "loss": 0.1911,
      "step": 766
    },
    {
      "epoch": 2.38,
      "learning_rate": 2e-05,
      "loss": 0.1804,
      "step": 768
    },
    {
      "epoch": 2.38,
      "learning_rate": 2e-05,
      "loss": 0.1791,
      "step": 770
    },
    {
      "epoch": 2.39,
      "learning_rate": 2e-05,
      "loss": 0.1889,
      "step": 772
    },
    {
      "epoch": 2.4,
      "learning_rate": 2e-05,
      "loss": 0.1827,
      "step": 774
    },
    {
      "epoch": 2.4,
      "learning_rate": 2e-05,
      "loss": 0.1854,
      "step": 776
    },
    {
      "epoch": 2.41,
      "learning_rate": 2e-05,
      "loss": 0.1826,
      "step": 778
    },
    {
      "epoch": 2.42,
      "learning_rate": 2e-05,
      "loss": 0.1793,
      "step": 780
    },
    {
      "epoch": 2.42,
      "learning_rate": 2e-05,
      "loss": 0.1859,
      "step": 782
    },
    {
      "epoch": 2.43,
      "learning_rate": 2e-05,
      "loss": 0.1827,
      "step": 784
    },
    {
      "epoch": 2.43,
      "learning_rate": 2e-05,
      "loss": 0.1894,
      "step": 786
    },
    {
      "epoch": 2.44,
      "learning_rate": 2e-05,
      "loss": 0.1878,
      "step": 788
    },
    {
      "epoch": 2.45,
      "learning_rate": 2e-05,
      "loss": 0.1858,
      "step": 790
    },
    {
      "epoch": 2.45,
      "learning_rate": 2e-05,
      "loss": 0.1836,
      "step": 792
    },
    {
      "epoch": 2.46,
      "learning_rate": 2e-05,
      "loss": 0.184,
      "step": 794
    },
    {
      "epoch": 2.47,
      "learning_rate": 2e-05,
      "loss": 0.1828,
      "step": 796
    },
    {
      "epoch": 2.47,
      "learning_rate": 2e-05,
      "loss": 0.1804,
      "step": 798
    },
    {
      "epoch": 2.48,
      "learning_rate": 2e-05,
      "loss": 0.1764,
      "step": 800
    },
    {
      "epoch": 2.48,
      "learning_rate": 2e-05,
      "loss": 0.181,
      "step": 802
    },
    {
      "epoch": 2.49,
      "learning_rate": 2e-05,
      "loss": 0.1881,
      "step": 804
    },
    {
      "epoch": 2.5,
      "learning_rate": 2e-05,
      "loss": 0.1848,
      "step": 806
    },
    {
      "epoch": 2.5,
      "learning_rate": 2e-05,
      "loss": 0.1799,
      "step": 808
    },
    {
      "epoch": 2.51,
      "learning_rate": 2e-05,
      "loss": 0.1846,
      "step": 810
    },
    {
      "epoch": 2.51,
      "learning_rate": 2e-05,
      "loss": 0.1799,
      "step": 812
    },
    {
      "epoch": 2.52,
      "learning_rate": 2e-05,
      "loss": 0.1805,
      "step": 814
    },
    {
      "epoch": 2.53,
      "learning_rate": 2e-05,
      "loss": 0.1825,
      "step": 816
    },
    {
      "epoch": 2.53,
      "learning_rate": 2e-05,
      "loss": 0.1804,
      "step": 818
    },
    {
      "epoch": 2.54,
      "learning_rate": 2e-05,
      "loss": 0.1805,
      "step": 820
    },
    {
      "epoch": 2.55,
      "learning_rate": 2e-05,
      "loss": 0.1799,
      "step": 822
    },
    {
      "epoch": 2.55,
      "learning_rate": 2e-05,
      "loss": 0.1797,
      "step": 824
    },
    {
      "epoch": 2.56,
      "learning_rate": 2e-05,
      "loss": 0.1778,
      "step": 826
    },
    {
      "epoch": 2.56,
      "learning_rate": 2e-05,
      "loss": 0.1851,
      "step": 828
    },
    {
      "epoch": 2.57,
      "learning_rate": 2e-05,
      "loss": 0.1743,
      "step": 830
    },
    {
      "epoch": 2.58,
      "learning_rate": 2e-05,
      "loss": 0.1766,
      "step": 832
    },
    {
      "epoch": 2.58,
      "learning_rate": 2e-05,
      "loss": 0.1852,
      "step": 834
    },
    {
      "epoch": 2.59,
      "learning_rate": 2e-05,
      "loss": 0.1737,
      "step": 836
    },
    {
      "epoch": 2.6,
      "learning_rate": 2e-05,
      "loss": 0.1758,
      "step": 838
    },
    {
      "epoch": 2.6,
      "learning_rate": 2e-05,
      "loss": 0.1754,
      "step": 840
    },
    {
      "epoch": 2.61,
      "learning_rate": 2e-05,
      "loss": 0.1754,
      "step": 842
    },
    {
      "epoch": 2.61,
      "learning_rate": 2e-05,
      "loss": 0.1757,
      "step": 844
    },
    {
      "epoch": 2.62,
      "learning_rate": 2e-05,
      "loss": 0.1755,
      "step": 846
    },
    {
      "epoch": 2.63,
      "learning_rate": 2e-05,
      "loss": 0.1755,
      "step": 848
    },
    {
      "epoch": 2.63,
      "learning_rate": 2e-05,
      "loss": 0.1754,
      "step": 850
    },
    {
      "epoch": 2.64,
      "learning_rate": 2e-05,
      "loss": 0.1751,
      "step": 852
    },
    {
      "epoch": 2.64,
      "learning_rate": 2e-05,
      "loss": 0.1726,
      "step": 854
    },
    {
      "epoch": 2.65,
      "learning_rate": 2e-05,
      "loss": 0.1727,
      "step": 856
    },
    {
      "epoch": 2.66,
      "learning_rate": 2e-05,
      "loss": 0.1743,
      "step": 858
    },
    {
      "epoch": 2.66,
      "learning_rate": 2e-05,
      "loss": 0.1769,
      "step": 860
    },
    {
      "epoch": 2.67,
      "learning_rate": 2e-05,
      "loss": 0.1833,
      "step": 862
    },
    {
      "epoch": 2.68,
      "learning_rate": 2e-05,
      "loss": 0.1737,
      "step": 864
    },
    {
      "epoch": 2.68,
      "learning_rate": 2e-05,
      "loss": 0.1762,
      "step": 866
    },
    {
      "epoch": 2.69,
      "learning_rate": 2e-05,
      "loss": 0.172,
      "step": 868
    },
    {
      "epoch": 2.69,
      "learning_rate": 2e-05,
      "loss": 0.1753,
      "step": 870
    },
    {
      "epoch": 2.7,
      "learning_rate": 2e-05,
      "loss": 0.1739,
      "step": 872
    },
    {
      "epoch": 2.71,
      "learning_rate": 2e-05,
      "loss": 0.1752,
      "step": 874
    },
    {
      "epoch": 2.71,
      "learning_rate": 2e-05,
      "loss": 0.1748,
      "step": 876
    },
    {
      "epoch": 2.72,
      "learning_rate": 2e-05,
      "loss": 0.1813,
      "step": 878
    },
    {
      "epoch": 2.73,
      "learning_rate": 2e-05,
      "loss": 0.1745,
      "step": 880
    },
    {
      "epoch": 2.73,
      "learning_rate": 2e-05,
      "loss": 0.1737,
      "step": 882
    },
    {
      "epoch": 2.74,
      "learning_rate": 2e-05,
      "loss": 0.1769,
      "step": 884
    },
    {
      "epoch": 2.74,
      "learning_rate": 2e-05,
      "loss": 0.176,
      "step": 886
    },
    {
      "epoch": 2.75,
      "learning_rate": 2e-05,
      "loss": 0.1721,
      "step": 888
    },
    {
      "epoch": 2.76,
      "learning_rate": 2e-05,
      "loss": 0.174,
      "step": 890
    },
    {
      "epoch": 2.76,
      "learning_rate": 2e-05,
      "loss": 0.1713,
      "step": 892
    },
    {
      "epoch": 2.77,
      "learning_rate": 2e-05,
      "loss": 0.1702,
      "step": 894
    },
    {
      "epoch": 2.78,
      "learning_rate": 2e-05,
      "loss": 0.1749,
      "step": 896
    },
    {
      "epoch": 2.78,
      "learning_rate": 2e-05,
      "loss": 0.1737,
      "step": 898
    },
    {
      "epoch": 2.79,
      "learning_rate": 2e-05,
      "loss": 0.175,
      "step": 900
    },
    {
      "epoch": 2.79,
      "learning_rate": 2e-05,
      "loss": 0.1686,
      "step": 902
    },
    {
      "epoch": 2.8,
      "learning_rate": 2e-05,
      "loss": 0.1714,
      "step": 904
    },
    {
      "epoch": 2.81,
      "learning_rate": 2e-05,
      "loss": 0.1728,
      "step": 906
    },
    {
      "epoch": 2.81,
      "learning_rate": 2e-05,
      "loss": 0.1686,
      "step": 908
    },
    {
      "epoch": 2.82,
      "learning_rate": 2e-05,
      "loss": 0.1734,
      "step": 910
    },
    {
      "epoch": 2.82,
      "learning_rate": 2e-05,
      "loss": 0.1727,
      "step": 912
    },
    {
      "epoch": 2.83,
      "learning_rate": 2e-05,
      "loss": 0.1728,
      "step": 914
    },
    {
      "epoch": 2.84,
      "learning_rate": 2e-05,
      "loss": 0.1684,
      "step": 916
    },
    {
      "epoch": 2.84,
      "learning_rate": 2e-05,
      "loss": 0.17,
      "step": 918
    },
    {
      "epoch": 2.85,
      "learning_rate": 2e-05,
      "loss": 0.1664,
      "step": 920
    },
    {
      "epoch": 2.86,
      "learning_rate": 2e-05,
      "loss": 0.1657,
      "step": 922
    },
    {
      "epoch": 2.86,
      "learning_rate": 2e-05,
      "loss": 0.1695,
      "step": 924
    },
    {
      "epoch": 2.87,
      "learning_rate": 2e-05,
      "loss": 0.1731,
      "step": 926
    },
    {
      "epoch": 2.87,
      "learning_rate": 2e-05,
      "loss": 0.1681,
      "step": 928
    },
    {
      "epoch": 2.88,
      "learning_rate": 2e-05,
      "loss": 0.1683,
      "step": 930
    },
    {
      "epoch": 2.89,
      "learning_rate": 2e-05,
      "loss": 0.1685,
      "step": 932
    },
    {
      "epoch": 2.89,
      "learning_rate": 2e-05,
      "loss": 0.168,
      "step": 934
    },
    {
      "epoch": 2.9,
      "learning_rate": 2e-05,
      "loss": 0.1702,
      "step": 936
    },
    {
      "epoch": 2.91,
      "learning_rate": 2e-05,
      "loss": 0.1693,
      "step": 938
    },
    {
      "epoch": 2.91,
      "learning_rate": 2e-05,
      "loss": 0.1717,
      "step": 940
    },
    {
      "epoch": 2.92,
      "learning_rate": 2e-05,
      "loss": 0.1701,
      "step": 942
    },
    {
      "epoch": 2.92,
      "learning_rate": 2e-05,
      "loss": 0.1641,
      "step": 944
    },
    {
      "epoch": 2.93,
      "learning_rate": 2e-05,
      "loss": 0.1703,
      "step": 946
    },
    {
      "epoch": 2.94,
      "learning_rate": 2e-05,
      "loss": 0.1676,
      "step": 948
    },
    {
      "epoch": 2.94,
      "learning_rate": 2e-05,
      "loss": 0.1709,
      "step": 950
    },
    {
      "epoch": 2.95,
      "learning_rate": 2e-05,
      "loss": 0.1755,
      "step": 952
    },
    {
      "epoch": 2.95,
      "learning_rate": 2e-05,
      "loss": 0.1705,
      "step": 954
    },
    {
      "epoch": 2.96,
      "learning_rate": 2e-05,
      "loss": 0.1643,
      "step": 956
    },
    {
      "epoch": 2.97,
      "learning_rate": 2e-05,
      "loss": 0.1663,
      "step": 958
    },
    {
      "epoch": 2.97,
      "learning_rate": 2e-05,
      "loss": 0.1647,
      "step": 960
    },
    {
      "epoch": 2.98,
      "learning_rate": 2e-05,
      "loss": 0.1619,
      "step": 962
    },
    {
      "epoch": 2.99,
      "learning_rate": 2e-05,
      "loss": 0.17,
      "step": 964
    },
    {
      "epoch": 2.99,
      "learning_rate": 2e-05,
      "loss": 0.165,
      "step": 966
    },
    {
      "epoch": 3.0,
      "learning_rate": 2e-05,
      "loss": 0.1603,
      "step": 968
    },
    {
      "epoch": 3.0,
      "learning_rate": 2e-05,
      "loss": 0.1437,
      "step": 970
    },
    {
      "epoch": 3.01,
      "learning_rate": 2e-05,
      "loss": 0.1377,
      "step": 972
    },
    {
      "epoch": 3.02,
      "learning_rate": 2e-05,
      "loss": 0.1427,
      "step": 974
    },
    {
      "epoch": 3.02,
      "learning_rate": 2e-05,
      "loss": 0.148,
      "step": 976
    },
    {
      "epoch": 3.03,
      "learning_rate": 2e-05,
      "loss": 0.1383,
      "step": 978
    },
    {
      "epoch": 3.04,
      "learning_rate": 2e-05,
      "loss": 0.1441,
      "step": 980
    },
    {
      "epoch": 3.04,
      "learning_rate": 2e-05,
      "loss": 0.1416,
      "step": 982
    },
    {
      "epoch": 3.05,
      "learning_rate": 2e-05,
      "loss": 0.143,
      "step": 984
    },
    {
      "epoch": 3.05,
      "learning_rate": 2e-05,
      "loss": 0.1388,
      "step": 986
    },
    {
      "epoch": 3.06,
      "learning_rate": 2e-05,
      "loss": 0.1423,
      "step": 988
    },
    {
      "epoch": 3.07,
      "learning_rate": 2e-05,
      "loss": 0.1396,
      "step": 990
    },
    {
      "epoch": 3.07,
      "learning_rate": 2e-05,
      "loss": 0.1447,
      "step": 992
    },
    {
      "epoch": 3.08,
      "learning_rate": 2e-05,
      "loss": 0.1381,
      "step": 994
    },
    {
      "epoch": 3.08,
      "learning_rate": 2e-05,
      "loss": 0.1402,
      "step": 996
    },
    {
      "epoch": 3.09,
      "learning_rate": 2e-05,
      "loss": 0.1392,
      "step": 998
    },
    {
      "epoch": 3.1,
      "learning_rate": 2e-05,
      "loss": 0.1361,
      "step": 1000
    }
  ],
  "logging_steps": 2,
  "max_steps": 1288,
  "num_train_epochs": 4,
  "save_steps": 200,
  "total_flos": 2.2885242503168e+16,
  "trial_name": null,
  "trial_params": null
}