longt5_xl_govreport_4096 / trainer_state.json
learn3r's picture
End of training
489785a
{
"best_metric": 1.3889119625091553,
"best_model_checkpoint": "/home/co-ou1/rds/hpc-work/models/longt5_xl_govreport_4096/checkpoint-272",
"epoch": 4.985108820160367,
"eval_steps": 500,
"global_step": 340,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.03,
"learning_rate": 0.001,
"loss": 2.6414,
"step": 2
},
{
"epoch": 0.06,
"learning_rate": 0.001,
"loss": 2.4747,
"step": 4
},
{
"epoch": 0.09,
"learning_rate": 0.001,
"loss": 1.8708,
"step": 6
},
{
"epoch": 0.12,
"learning_rate": 0.001,
"loss": 2.4373,
"step": 8
},
{
"epoch": 0.15,
"learning_rate": 0.001,
"loss": 2.2961,
"step": 10
},
{
"epoch": 0.18,
"learning_rate": 0.001,
"loss": 2.1959,
"step": 12
},
{
"epoch": 0.21,
"learning_rate": 0.001,
"loss": 1.9882,
"step": 14
},
{
"epoch": 0.23,
"learning_rate": 0.001,
"loss": 2.0164,
"step": 16
},
{
"epoch": 0.26,
"learning_rate": 0.001,
"loss": 1.8772,
"step": 18
},
{
"epoch": 0.29,
"learning_rate": 0.001,
"loss": 1.8516,
"step": 20
},
{
"epoch": 0.32,
"learning_rate": 0.001,
"loss": 1.8587,
"step": 22
},
{
"epoch": 0.35,
"learning_rate": 0.001,
"loss": 1.729,
"step": 24
},
{
"epoch": 0.38,
"learning_rate": 0.001,
"loss": 1.7265,
"step": 26
},
{
"epoch": 0.41,
"learning_rate": 0.001,
"loss": 1.7689,
"step": 28
},
{
"epoch": 0.44,
"learning_rate": 0.001,
"loss": 1.7899,
"step": 30
},
{
"epoch": 0.47,
"learning_rate": 0.001,
"loss": 1.6755,
"step": 32
},
{
"epoch": 0.5,
"learning_rate": 0.001,
"loss": 1.6701,
"step": 34
},
{
"epoch": 0.53,
"learning_rate": 0.001,
"loss": 1.6626,
"step": 36
},
{
"epoch": 0.56,
"learning_rate": 0.001,
"loss": 1.708,
"step": 38
},
{
"epoch": 0.59,
"learning_rate": 0.001,
"loss": 1.7019,
"step": 40
},
{
"epoch": 0.62,
"learning_rate": 0.001,
"loss": 1.6705,
"step": 42
},
{
"epoch": 0.65,
"learning_rate": 0.001,
"loss": 1.6039,
"step": 44
},
{
"epoch": 0.67,
"learning_rate": 0.001,
"loss": 1.6041,
"step": 46
},
{
"epoch": 0.7,
"learning_rate": 0.001,
"loss": 1.5829,
"step": 48
},
{
"epoch": 0.73,
"learning_rate": 0.001,
"loss": 1.6283,
"step": 50
},
{
"epoch": 0.76,
"learning_rate": 0.001,
"loss": 1.7403,
"step": 52
},
{
"epoch": 0.79,
"learning_rate": 0.001,
"loss": 1.6786,
"step": 54
},
{
"epoch": 0.82,
"learning_rate": 0.001,
"loss": 1.6499,
"step": 56
},
{
"epoch": 0.85,
"learning_rate": 0.001,
"loss": 1.5641,
"step": 58
},
{
"epoch": 0.88,
"learning_rate": 0.001,
"loss": 1.5824,
"step": 60
},
{
"epoch": 0.91,
"learning_rate": 0.001,
"loss": 1.6462,
"step": 62
},
{
"epoch": 0.94,
"learning_rate": 0.001,
"loss": 1.5787,
"step": 64
},
{
"epoch": 0.97,
"learning_rate": 0.001,
"loss": 1.6053,
"step": 66
},
{
"epoch": 1.0,
"learning_rate": 0.001,
"loss": 1.6043,
"step": 68
},
{
"epoch": 1.0,
"eval_loss": 1.4757200479507446,
"eval_runtime": 333.5276,
"eval_samples_per_second": 2.914,
"eval_steps_per_second": 0.729,
"step": 68
},
{
"epoch": 1.03,
"learning_rate": 0.001,
"loss": 1.549,
"step": 70
},
{
"epoch": 1.06,
"learning_rate": 0.001,
"loss": 1.4379,
"step": 72
},
{
"epoch": 1.08,
"learning_rate": 0.001,
"loss": 1.5224,
"step": 74
},
{
"epoch": 1.11,
"learning_rate": 0.001,
"loss": 1.5106,
"step": 76
},
{
"epoch": 1.14,
"learning_rate": 0.001,
"loss": 1.4786,
"step": 78
},
{
"epoch": 1.17,
"learning_rate": 0.001,
"loss": 1.5101,
"step": 80
},
{
"epoch": 1.2,
"learning_rate": 0.001,
"loss": 1.4835,
"step": 82
},
{
"epoch": 1.23,
"learning_rate": 0.001,
"loss": 1.4749,
"step": 84
},
{
"epoch": 1.26,
"learning_rate": 0.001,
"loss": 1.488,
"step": 86
},
{
"epoch": 1.29,
"learning_rate": 0.001,
"loss": 1.4604,
"step": 88
},
{
"epoch": 1.32,
"learning_rate": 0.001,
"loss": 1.5358,
"step": 90
},
{
"epoch": 1.35,
"learning_rate": 0.001,
"loss": 1.5396,
"step": 92
},
{
"epoch": 1.38,
"learning_rate": 0.001,
"loss": 1.5071,
"step": 94
},
{
"epoch": 1.41,
"learning_rate": 0.001,
"loss": 1.5219,
"step": 96
},
{
"epoch": 1.44,
"learning_rate": 0.001,
"loss": 1.4589,
"step": 98
},
{
"epoch": 1.47,
"learning_rate": 0.001,
"loss": 1.4392,
"step": 100
},
{
"epoch": 1.5,
"learning_rate": 0.001,
"loss": 1.485,
"step": 102
},
{
"epoch": 1.52,
"learning_rate": 0.001,
"loss": 1.4617,
"step": 104
},
{
"epoch": 1.55,
"learning_rate": 0.001,
"loss": 1.4659,
"step": 106
},
{
"epoch": 1.58,
"learning_rate": 0.001,
"loss": 1.4908,
"step": 108
},
{
"epoch": 1.61,
"learning_rate": 0.001,
"loss": 1.4929,
"step": 110
},
{
"epoch": 1.64,
"learning_rate": 0.001,
"loss": 1.4538,
"step": 112
},
{
"epoch": 1.67,
"learning_rate": 0.001,
"loss": 1.4815,
"step": 114
},
{
"epoch": 1.7,
"learning_rate": 0.001,
"loss": 1.4525,
"step": 116
},
{
"epoch": 1.73,
"learning_rate": 0.001,
"loss": 1.4664,
"step": 118
},
{
"epoch": 1.76,
"learning_rate": 0.001,
"loss": 1.4498,
"step": 120
},
{
"epoch": 1.79,
"learning_rate": 0.001,
"loss": 1.5247,
"step": 122
},
{
"epoch": 1.82,
"learning_rate": 0.001,
"loss": 1.5033,
"step": 124
},
{
"epoch": 1.85,
"learning_rate": 0.001,
"loss": 1.4434,
"step": 126
},
{
"epoch": 1.88,
"learning_rate": 0.001,
"loss": 1.4806,
"step": 128
},
{
"epoch": 1.91,
"learning_rate": 0.001,
"loss": 1.4551,
"step": 130
},
{
"epoch": 1.94,
"learning_rate": 0.001,
"loss": 1.4359,
"step": 132
},
{
"epoch": 1.96,
"learning_rate": 0.001,
"loss": 1.4424,
"step": 134
},
{
"epoch": 1.99,
"learning_rate": 0.001,
"loss": 1.4471,
"step": 136
},
{
"epoch": 1.99,
"eval_loss": 1.401979923248291,
"eval_runtime": 333.5413,
"eval_samples_per_second": 2.914,
"eval_steps_per_second": 0.729,
"step": 136
},
{
"epoch": 2.02,
"learning_rate": 0.001,
"loss": 1.3704,
"step": 138
},
{
"epoch": 2.05,
"learning_rate": 0.001,
"loss": 1.3246,
"step": 140
},
{
"epoch": 2.08,
"learning_rate": 0.001,
"loss": 1.3714,
"step": 142
},
{
"epoch": 2.11,
"learning_rate": 0.001,
"loss": 1.3134,
"step": 144
},
{
"epoch": 2.14,
"learning_rate": 0.001,
"loss": 1.2998,
"step": 146
},
{
"epoch": 2.17,
"learning_rate": 0.001,
"loss": 1.3643,
"step": 148
},
{
"epoch": 2.2,
"learning_rate": 0.001,
"loss": 1.306,
"step": 150
},
{
"epoch": 2.23,
"learning_rate": 0.001,
"loss": 1.3152,
"step": 152
},
{
"epoch": 2.26,
"learning_rate": 0.001,
"loss": 1.3086,
"step": 154
},
{
"epoch": 2.29,
"learning_rate": 0.001,
"loss": 1.3033,
"step": 156
},
{
"epoch": 2.32,
"learning_rate": 0.001,
"loss": 1.343,
"step": 158
},
{
"epoch": 2.35,
"learning_rate": 0.001,
"loss": 1.3399,
"step": 160
},
{
"epoch": 2.38,
"learning_rate": 0.001,
"loss": 1.3134,
"step": 162
},
{
"epoch": 2.4,
"learning_rate": 0.001,
"loss": 1.3472,
"step": 164
},
{
"epoch": 2.43,
"learning_rate": 0.001,
"loss": 1.3261,
"step": 166
},
{
"epoch": 2.46,
"learning_rate": 0.001,
"loss": 1.3162,
"step": 168
},
{
"epoch": 2.49,
"learning_rate": 0.001,
"loss": 1.3194,
"step": 170
},
{
"epoch": 2.52,
"learning_rate": 0.001,
"loss": 1.3236,
"step": 172
},
{
"epoch": 2.55,
"learning_rate": 0.001,
"loss": 1.3033,
"step": 174
},
{
"epoch": 2.58,
"learning_rate": 0.001,
"loss": 1.3212,
"step": 176
},
{
"epoch": 2.61,
"learning_rate": 0.001,
"loss": 1.3433,
"step": 178
},
{
"epoch": 2.64,
"learning_rate": 0.001,
"loss": 1.34,
"step": 180
},
{
"epoch": 2.67,
"learning_rate": 0.001,
"loss": 1.3306,
"step": 182
},
{
"epoch": 2.7,
"learning_rate": 0.001,
"loss": 1.3377,
"step": 184
},
{
"epoch": 2.73,
"learning_rate": 0.001,
"loss": 1.3048,
"step": 186
},
{
"epoch": 2.76,
"learning_rate": 0.001,
"loss": 1.3186,
"step": 188
},
{
"epoch": 2.79,
"learning_rate": 0.001,
"loss": 1.3405,
"step": 190
},
{
"epoch": 2.82,
"learning_rate": 0.001,
"loss": 1.3266,
"step": 192
},
{
"epoch": 2.84,
"learning_rate": 0.001,
"loss": 1.3182,
"step": 194
},
{
"epoch": 2.87,
"learning_rate": 0.001,
"loss": 1.3755,
"step": 196
},
{
"epoch": 2.9,
"learning_rate": 0.001,
"loss": 1.3422,
"step": 198
},
{
"epoch": 2.93,
"learning_rate": 0.001,
"loss": 1.3247,
"step": 200
},
{
"epoch": 2.96,
"learning_rate": 0.001,
"loss": 1.312,
"step": 202
},
{
"epoch": 2.99,
"learning_rate": 0.001,
"loss": 1.315,
"step": 204
},
{
"epoch": 2.99,
"eval_loss": 1.3935041427612305,
"eval_runtime": 333.777,
"eval_samples_per_second": 2.912,
"eval_steps_per_second": 0.728,
"step": 204
},
{
"epoch": 3.02,
"learning_rate": 0.001,
"loss": 1.2264,
"step": 206
},
{
"epoch": 3.05,
"learning_rate": 0.001,
"loss": 1.1606,
"step": 208
},
{
"epoch": 3.08,
"learning_rate": 0.001,
"loss": 1.1569,
"step": 210
},
{
"epoch": 3.11,
"learning_rate": 0.001,
"loss": 1.1434,
"step": 212
},
{
"epoch": 3.14,
"learning_rate": 0.001,
"loss": 1.1736,
"step": 214
},
{
"epoch": 3.17,
"learning_rate": 0.001,
"loss": 1.17,
"step": 216
},
{
"epoch": 3.2,
"learning_rate": 0.001,
"loss": 1.1802,
"step": 218
},
{
"epoch": 3.23,
"learning_rate": 0.001,
"loss": 1.1856,
"step": 220
},
{
"epoch": 3.25,
"learning_rate": 0.001,
"loss": 1.1918,
"step": 222
},
{
"epoch": 3.28,
"learning_rate": 0.001,
"loss": 1.1751,
"step": 224
},
{
"epoch": 3.31,
"learning_rate": 0.001,
"loss": 1.1896,
"step": 226
},
{
"epoch": 3.34,
"learning_rate": 0.001,
"loss": 1.1837,
"step": 228
},
{
"epoch": 3.37,
"learning_rate": 0.001,
"loss": 1.2053,
"step": 230
},
{
"epoch": 3.4,
"learning_rate": 0.001,
"loss": 1.2031,
"step": 232
},
{
"epoch": 3.43,
"learning_rate": 0.001,
"loss": 1.1886,
"step": 234
},
{
"epoch": 3.46,
"learning_rate": 0.001,
"loss": 1.1769,
"step": 236
},
{
"epoch": 3.49,
"learning_rate": 0.001,
"loss": 1.1951,
"step": 238
},
{
"epoch": 3.52,
"learning_rate": 0.001,
"loss": 1.1943,
"step": 240
},
{
"epoch": 3.55,
"learning_rate": 0.001,
"loss": 1.1816,
"step": 242
},
{
"epoch": 3.58,
"learning_rate": 0.001,
"loss": 1.1763,
"step": 244
},
{
"epoch": 3.61,
"learning_rate": 0.001,
"loss": 1.2028,
"step": 246
},
{
"epoch": 3.64,
"learning_rate": 0.001,
"loss": 1.2199,
"step": 248
},
{
"epoch": 3.67,
"learning_rate": 0.001,
"loss": 1.1948,
"step": 250
},
{
"epoch": 3.69,
"learning_rate": 0.001,
"loss": 1.1859,
"step": 252
},
{
"epoch": 3.72,
"learning_rate": 0.001,
"loss": 1.1864,
"step": 254
},
{
"epoch": 3.75,
"learning_rate": 0.001,
"loss": 1.2012,
"step": 256
},
{
"epoch": 3.78,
"learning_rate": 0.001,
"loss": 1.2177,
"step": 258
},
{
"epoch": 3.81,
"learning_rate": 0.001,
"loss": 1.2156,
"step": 260
},
{
"epoch": 3.84,
"learning_rate": 0.001,
"loss": 1.204,
"step": 262
},
{
"epoch": 3.87,
"learning_rate": 0.001,
"loss": 1.2053,
"step": 264
},
{
"epoch": 3.9,
"learning_rate": 0.001,
"loss": 1.2138,
"step": 266
},
{
"epoch": 3.93,
"learning_rate": 0.001,
"loss": 1.1992,
"step": 268
},
{
"epoch": 3.96,
"learning_rate": 0.001,
"loss": 1.1903,
"step": 270
},
{
"epoch": 3.99,
"learning_rate": 0.001,
"loss": 1.2098,
"step": 272
},
{
"epoch": 3.99,
"eval_loss": 1.3889119625091553,
"eval_runtime": 333.5624,
"eval_samples_per_second": 2.914,
"eval_steps_per_second": 0.728,
"step": 272
},
{
"epoch": 4.02,
"learning_rate": 0.001,
"loss": 1.1135,
"step": 274
},
{
"epoch": 4.05,
"learning_rate": 0.001,
"loss": 1.0318,
"step": 276
},
{
"epoch": 4.08,
"learning_rate": 0.001,
"loss": 1.068,
"step": 278
},
{
"epoch": 4.11,
"learning_rate": 0.001,
"loss": 1.0386,
"step": 280
},
{
"epoch": 4.13,
"learning_rate": 0.001,
"loss": 1.0727,
"step": 282
},
{
"epoch": 4.16,
"learning_rate": 0.001,
"loss": 1.0592,
"step": 284
},
{
"epoch": 4.19,
"learning_rate": 0.001,
"loss": 1.0348,
"step": 286
},
{
"epoch": 4.22,
"learning_rate": 0.001,
"loss": 1.045,
"step": 288
},
{
"epoch": 4.25,
"learning_rate": 0.001,
"loss": 1.0535,
"step": 290
},
{
"epoch": 4.28,
"learning_rate": 0.001,
"loss": 1.0628,
"step": 292
},
{
"epoch": 4.31,
"learning_rate": 0.001,
"loss": 1.0699,
"step": 294
},
{
"epoch": 4.34,
"learning_rate": 0.001,
"loss": 1.0482,
"step": 296
},
{
"epoch": 4.37,
"learning_rate": 0.001,
"loss": 1.0703,
"step": 298
},
{
"epoch": 4.4,
"learning_rate": 0.001,
"loss": 1.0498,
"step": 300
},
{
"epoch": 4.43,
"learning_rate": 0.001,
"loss": 1.0775,
"step": 302
},
{
"epoch": 4.46,
"learning_rate": 0.001,
"loss": 1.0609,
"step": 304
},
{
"epoch": 4.49,
"learning_rate": 0.001,
"loss": 1.0985,
"step": 306
},
{
"epoch": 4.52,
"learning_rate": 0.001,
"loss": 1.081,
"step": 308
},
{
"epoch": 4.55,
"learning_rate": 0.001,
"loss": 1.0942,
"step": 310
},
{
"epoch": 4.57,
"learning_rate": 0.001,
"loss": 1.0789,
"step": 312
},
{
"epoch": 4.6,
"learning_rate": 0.001,
"loss": 1.0897,
"step": 314
},
{
"epoch": 4.63,
"learning_rate": 0.001,
"loss": 1.0869,
"step": 316
},
{
"epoch": 4.66,
"learning_rate": 0.001,
"loss": 1.1108,
"step": 318
},
{
"epoch": 4.69,
"learning_rate": 0.001,
"loss": 1.0687,
"step": 320
},
{
"epoch": 4.72,
"learning_rate": 0.001,
"loss": 1.1109,
"step": 322
},
{
"epoch": 4.75,
"learning_rate": 0.001,
"loss": 1.0838,
"step": 324
},
{
"epoch": 4.78,
"learning_rate": 0.001,
"loss": 1.0932,
"step": 326
},
{
"epoch": 4.81,
"learning_rate": 0.001,
"loss": 1.0923,
"step": 328
},
{
"epoch": 4.84,
"learning_rate": 0.001,
"loss": 1.0662,
"step": 330
},
{
"epoch": 4.87,
"learning_rate": 0.001,
"loss": 1.0727,
"step": 332
},
{
"epoch": 4.9,
"learning_rate": 0.001,
"loss": 1.0809,
"step": 334
},
{
"epoch": 4.93,
"learning_rate": 0.001,
"loss": 1.1087,
"step": 336
},
{
"epoch": 4.96,
"learning_rate": 0.001,
"loss": 1.0912,
"step": 338
},
{
"epoch": 4.99,
"learning_rate": 0.001,
"loss": 1.0888,
"step": 340
},
{
"epoch": 4.99,
"eval_loss": 1.4388501644134521,
"eval_runtime": 333.5483,
"eval_samples_per_second": 2.914,
"eval_steps_per_second": 0.729,
"step": 340
},
{
"epoch": 4.99,
"step": 340,
"total_flos": 5.95419516573578e+18,
"train_loss": 1.3779389661901138,
"train_runtime": 112099.6799,
"train_samples_per_second": 0.779,
"train_steps_per_second": 0.003
}
],
"logging_steps": 2,
"max_steps": 340,
"num_train_epochs": 5,
"save_steps": 500,
"total_flos": 5.95419516573578e+18,
"trial_name": null,
"trial_params": null
}