Upload trainer_state.json
Browse files- trainer_state.json +288 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,288 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
+
"global_step": 2700,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.37,
|
| 12 |
+
"learning_rate": 4.814814814814815e-05,
|
| 13 |
+
"loss": 0.8478,
|
| 14 |
+
"step": 100
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.74,
|
| 18 |
+
"learning_rate": 4.62962962962963e-05,
|
| 19 |
+
"loss": 0.3793,
|
| 20 |
+
"step": 200
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 1.0,
|
| 24 |
+
"eval_bleu": 16.2484,
|
| 25 |
+
"eval_gen_len": 18.4167,
|
| 26 |
+
"eval_loss": 0.3601939380168915,
|
| 27 |
+
"eval_meteor": 0.4221,
|
| 28 |
+
"eval_runtime": 13.5749,
|
| 29 |
+
"eval_samples_per_second": 4.42,
|
| 30 |
+
"eval_steps_per_second": 2.21,
|
| 31 |
+
"step": 270
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 1.11,
|
| 35 |
+
"learning_rate": 4.4444444444444447e-05,
|
| 36 |
+
"loss": 0.3099,
|
| 37 |
+
"step": 300
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 1.48,
|
| 41 |
+
"learning_rate": 4.259259259259259e-05,
|
| 42 |
+
"loss": 0.274,
|
| 43 |
+
"step": 400
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"epoch": 1.85,
|
| 47 |
+
"learning_rate": 4.074074074074074e-05,
|
| 48 |
+
"loss": 0.2437,
|
| 49 |
+
"step": 500
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 2.0,
|
| 53 |
+
"eval_bleu": 17.2012,
|
| 54 |
+
"eval_gen_len": 18.45,
|
| 55 |
+
"eval_loss": 0.32714077830314636,
|
| 56 |
+
"eval_meteor": 0.4235,
|
| 57 |
+
"eval_runtime": 11.4881,
|
| 58 |
+
"eval_samples_per_second": 5.223,
|
| 59 |
+
"eval_steps_per_second": 2.611,
|
| 60 |
+
"step": 540
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 2.22,
|
| 64 |
+
"learning_rate": 3.888888888888889e-05,
|
| 65 |
+
"loss": 0.2115,
|
| 66 |
+
"step": 600
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 2.59,
|
| 70 |
+
"learning_rate": 3.7037037037037037e-05,
|
| 71 |
+
"loss": 0.1758,
|
| 72 |
+
"step": 700
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 2.96,
|
| 76 |
+
"learning_rate": 3.518518518518519e-05,
|
| 77 |
+
"loss": 0.1911,
|
| 78 |
+
"step": 800
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 3.0,
|
| 82 |
+
"eval_bleu": 17.6526,
|
| 83 |
+
"eval_gen_len": 18.5,
|
| 84 |
+
"eval_loss": 0.3079971969127655,
|
| 85 |
+
"eval_meteor": 0.4244,
|
| 86 |
+
"eval_runtime": 11.4927,
|
| 87 |
+
"eval_samples_per_second": 5.221,
|
| 88 |
+
"eval_steps_per_second": 2.61,
|
| 89 |
+
"step": 810
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 3.33,
|
| 93 |
+
"learning_rate": 3.3333333333333335e-05,
|
| 94 |
+
"loss": 0.136,
|
| 95 |
+
"step": 900
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 3.7,
|
| 99 |
+
"learning_rate": 3.148148148148148e-05,
|
| 100 |
+
"loss": 0.1508,
|
| 101 |
+
"step": 1000
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 4.0,
|
| 105 |
+
"eval_bleu": 17.6733,
|
| 106 |
+
"eval_gen_len": 18.5,
|
| 107 |
+
"eval_loss": 0.30199024081230164,
|
| 108 |
+
"eval_meteor": 0.4227,
|
| 109 |
+
"eval_runtime": 11.3901,
|
| 110 |
+
"eval_samples_per_second": 5.268,
|
| 111 |
+
"eval_steps_per_second": 2.634,
|
| 112 |
+
"step": 1080
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 4.07,
|
| 116 |
+
"learning_rate": 2.962962962962963e-05,
|
| 117 |
+
"loss": 0.1336,
|
| 118 |
+
"step": 1100
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 4.44,
|
| 122 |
+
"learning_rate": 2.777777777777778e-05,
|
| 123 |
+
"loss": 0.0853,
|
| 124 |
+
"step": 1200
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 4.81,
|
| 128 |
+
"learning_rate": 2.5925925925925925e-05,
|
| 129 |
+
"loss": 0.1042,
|
| 130 |
+
"step": 1300
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 5.0,
|
| 134 |
+
"eval_bleu": 17.1567,
|
| 135 |
+
"eval_gen_len": 18.45,
|
| 136 |
+
"eval_loss": 0.43005862832069397,
|
| 137 |
+
"eval_meteor": 0.421,
|
| 138 |
+
"eval_runtime": 11.4383,
|
| 139 |
+
"eval_samples_per_second": 5.246,
|
| 140 |
+
"eval_steps_per_second": 2.623,
|
| 141 |
+
"step": 1350
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 5.19,
|
| 145 |
+
"learning_rate": 2.4074074074074074e-05,
|
| 146 |
+
"loss": 0.0795,
|
| 147 |
+
"step": 1400
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 5.56,
|
| 151 |
+
"learning_rate": 2.2222222222222223e-05,
|
| 152 |
+
"loss": 0.0776,
|
| 153 |
+
"step": 1500
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 5.93,
|
| 157 |
+
"learning_rate": 2.037037037037037e-05,
|
| 158 |
+
"loss": 0.0712,
|
| 159 |
+
"step": 1600
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 6.0,
|
| 163 |
+
"eval_bleu": 17.5338,
|
| 164 |
+
"eval_gen_len": 18.5,
|
| 165 |
+
"eval_loss": 0.39506202936172485,
|
| 166 |
+
"eval_meteor": 0.4211,
|
| 167 |
+
"eval_runtime": 11.4467,
|
| 168 |
+
"eval_samples_per_second": 5.242,
|
| 169 |
+
"eval_steps_per_second": 2.621,
|
| 170 |
+
"step": 1620
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 6.3,
|
| 174 |
+
"learning_rate": 1.8518518518518518e-05,
|
| 175 |
+
"loss": 0.0548,
|
| 176 |
+
"step": 1700
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 6.67,
|
| 180 |
+
"learning_rate": 1.6666666666666667e-05,
|
| 181 |
+
"loss": 0.0516,
|
| 182 |
+
"step": 1800
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 7.0,
|
| 186 |
+
"eval_bleu": 17.1366,
|
| 187 |
+
"eval_gen_len": 18.4667,
|
| 188 |
+
"eval_loss": 0.44624677300453186,
|
| 189 |
+
"eval_meteor": 0.4185,
|
| 190 |
+
"eval_runtime": 11.4875,
|
| 191 |
+
"eval_samples_per_second": 5.223,
|
| 192 |
+
"eval_steps_per_second": 2.612,
|
| 193 |
+
"step": 1890
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 7.04,
|
| 197 |
+
"learning_rate": 1.4814814814814815e-05,
|
| 198 |
+
"loss": 0.0557,
|
| 199 |
+
"step": 1900
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 7.41,
|
| 203 |
+
"learning_rate": 1.2962962962962962e-05,
|
| 204 |
+
"loss": 0.0485,
|
| 205 |
+
"step": 2000
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 7.78,
|
| 209 |
+
"learning_rate": 1.1111111111111112e-05,
|
| 210 |
+
"loss": 0.043,
|
| 211 |
+
"step": 2100
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 8.0,
|
| 215 |
+
"eval_bleu": 17.6692,
|
| 216 |
+
"eval_gen_len": 18.5,
|
| 217 |
+
"eval_loss": 0.4671081006526947,
|
| 218 |
+
"eval_meteor": 0.4217,
|
| 219 |
+
"eval_runtime": 11.4117,
|
| 220 |
+
"eval_samples_per_second": 5.258,
|
| 221 |
+
"eval_steps_per_second": 2.629,
|
| 222 |
+
"step": 2160
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 8.15,
|
| 226 |
+
"learning_rate": 9.259259259259259e-06,
|
| 227 |
+
"loss": 0.0412,
|
| 228 |
+
"step": 2200
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 8.52,
|
| 232 |
+
"learning_rate": 7.4074074074074075e-06,
|
| 233 |
+
"loss": 0.0402,
|
| 234 |
+
"step": 2300
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 8.89,
|
| 238 |
+
"learning_rate": 5.555555555555556e-06,
|
| 239 |
+
"loss": 0.0243,
|
| 240 |
+
"step": 2400
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 9.0,
|
| 244 |
+
"eval_bleu": 17.5591,
|
| 245 |
+
"eval_gen_len": 18.5,
|
| 246 |
+
"eval_loss": 0.5049145221710205,
|
| 247 |
+
"eval_meteor": 0.4214,
|
| 248 |
+
"eval_runtime": 11.458,
|
| 249 |
+
"eval_samples_per_second": 5.237,
|
| 250 |
+
"eval_steps_per_second": 2.618,
|
| 251 |
+
"step": 2430
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 9.26,
|
| 255 |
+
"learning_rate": 3.7037037037037037e-06,
|
| 256 |
+
"loss": 0.0475,
|
| 257 |
+
"step": 2500
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 9.63,
|
| 261 |
+
"learning_rate": 1.8518518518518519e-06,
|
| 262 |
+
"loss": 0.0324,
|
| 263 |
+
"step": 2600
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 10.0,
|
| 267 |
+
"learning_rate": 0.0,
|
| 268 |
+
"loss": 0.0313,
|
| 269 |
+
"step": 2700
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 10.0,
|
| 273 |
+
"eval_bleu": 17.4599,
|
| 274 |
+
"eval_gen_len": 18.5,
|
| 275 |
+
"eval_loss": 0.5202592015266418,
|
| 276 |
+
"eval_meteor": 0.4218,
|
| 277 |
+
"eval_runtime": 11.5675,
|
| 278 |
+
"eval_samples_per_second": 5.187,
|
| 279 |
+
"eval_steps_per_second": 2.593,
|
| 280 |
+
"step": 2700
|
| 281 |
+
}
|
| 282 |
+
],
|
| 283 |
+
"max_steps": 2700,
|
| 284 |
+
"num_train_epochs": 10,
|
| 285 |
+
"total_flos": 163370071388160.0,
|
| 286 |
+
"trial_name": null,
|
| 287 |
+
"trial_params": null
|
| 288 |
+
}
|