End of training
Browse files- all_results.json +14 -14
- eval_results.json +9 -10
- runs/Jun01_07-22-16_7b8f3300e052/events.out.tfevents.1685607178.7b8f3300e052.6012.2 +3 -0
- train_results.json +6 -6
- trainer_state.json +367 -1218
all_results.json
CHANGED
|
@@ -1,17 +1,17 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_bleu":
|
| 4 |
-
"eval_em": 0.
|
| 5 |
-
"eval_gen_len":
|
| 6 |
-
"eval_loss":
|
| 7 |
"eval_rm": 0.22,
|
| 8 |
-
"eval_runtime":
|
| 9 |
-
"eval_samples":
|
| 10 |
-
"eval_samples_per_second":
|
| 11 |
-
"eval_steps_per_second": 0.
|
| 12 |
-
"train_loss":
|
| 13 |
-
"train_runtime":
|
| 14 |
-
"train_samples":
|
| 15 |
-
"train_samples_per_second":
|
| 16 |
-
"train_steps_per_second":
|
| 17 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 8.0,
|
| 3 |
+
"eval_bleu": 77.0649,
|
| 4 |
+
"eval_em": 0.0548,
|
| 5 |
+
"eval_gen_len": 47.473,
|
| 6 |
+
"eval_loss": 1.024766206741333,
|
| 7 |
"eval_rm": 0.22,
|
| 8 |
+
"eval_runtime": 396.6503,
|
| 9 |
+
"eval_samples": 1205,
|
| 10 |
+
"eval_samples_per_second": 3.038,
|
| 11 |
+
"eval_steps_per_second": 0.381,
|
| 12 |
+
"train_loss": 0.15130291744847968,
|
| 13 |
+
"train_runtime": 1637.6947,
|
| 14 |
+
"train_samples": 22890,
|
| 15 |
+
"train_samples_per_second": 111.816,
|
| 16 |
+
"train_steps_per_second": 6.99
|
| 17 |
}
|
eval_results.json
CHANGED
|
@@ -1,12 +1,11 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_bleu":
|
| 4 |
-
"eval_em": 0.
|
| 5 |
-
"eval_gen_len":
|
| 6 |
-
"eval_loss":
|
| 7 |
-
"
|
| 8 |
-
"
|
| 9 |
-
"
|
| 10 |
-
"
|
| 11 |
-
"eval_steps_per_second": 0.178
|
| 12 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 8.0,
|
| 3 |
+
"eval_bleu": 77.0649,
|
| 4 |
+
"eval_em": 0.0548,
|
| 5 |
+
"eval_gen_len": 47.473,
|
| 6 |
+
"eval_loss": 1.024766206741333,
|
| 7 |
+
"eval_runtime": 396.6503,
|
| 8 |
+
"eval_samples": 1205,
|
| 9 |
+
"eval_samples_per_second": 3.038,
|
| 10 |
+
"eval_steps_per_second": 0.381
|
|
|
|
| 11 |
}
|
runs/Jun01_07-22-16_7b8f3300e052/events.out.tfevents.1685607178.7b8f3300e052.6012.2
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:956aad7d033977572661fda753904a774c2cf2b4c7294bd0e2afd09cdfb5846e
|
| 3 |
+
size 504
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"train_loss":
|
| 4 |
-
"train_runtime":
|
| 5 |
-
"train_samples":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 8.0,
|
| 3 |
+
"train_loss": 0.15130291744847968,
|
| 4 |
+
"train_runtime": 1637.6947,
|
| 5 |
+
"train_samples": 22890,
|
| 6 |
+
"train_samples_per_second": 111.816,
|
| 7 |
+
"train_steps_per_second": 6.99
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,1615 +1,764 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
-
"global_step":
|
| 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.
|
| 12 |
"learning_rate": 2.0000000000000003e-06,
|
| 13 |
-
"loss":
|
| 14 |
"step": 100
|
| 15 |
},
|
| 16 |
{
|
| 17 |
-
"epoch": 0.
|
| 18 |
"learning_rate": 4.000000000000001e-06,
|
| 19 |
-
"loss":
|
| 20 |
"step": 200
|
| 21 |
},
|
| 22 |
{
|
| 23 |
-
"epoch": 0.
|
| 24 |
"learning_rate": 6e-06,
|
| 25 |
-
"loss": 41.
|
| 26 |
"step": 300
|
| 27 |
},
|
| 28 |
{
|
| 29 |
-
"epoch": 0.
|
| 30 |
"learning_rate": 8.000000000000001e-06,
|
| 31 |
-
"loss": 36.
|
| 32 |
"step": 400
|
| 33 |
},
|
| 34 |
{
|
| 35 |
-
"epoch": 0.
|
| 36 |
"learning_rate": 1e-05,
|
| 37 |
-
"loss": 33.
|
| 38 |
"step": 500
|
| 39 |
},
|
| 40 |
{
|
| 41 |
-
"epoch": 0.
|
| 42 |
"learning_rate": 1.2e-05,
|
| 43 |
-
"loss":
|
| 44 |
"step": 600
|
| 45 |
},
|
| 46 |
{
|
| 47 |
-
"epoch": 0.
|
| 48 |
"learning_rate": 1.4000000000000001e-05,
|
| 49 |
-
"loss":
|
| 50 |
"step": 700
|
| 51 |
},
|
| 52 |
{
|
| 53 |
-
"epoch": 0.
|
| 54 |
"learning_rate": 1.6000000000000003e-05,
|
| 55 |
-
"loss":
|
| 56 |
"step": 800
|
| 57 |
},
|
| 58 |
{
|
| 59 |
-
"epoch": 0.
|
| 60 |
"learning_rate": 1.8e-05,
|
| 61 |
-
"loss":
|
| 62 |
"step": 900
|
| 63 |
},
|
| 64 |
{
|
| 65 |
-
"epoch": 0.
|
| 66 |
"learning_rate": 2e-05,
|
| 67 |
-
"loss":
|
| 68 |
"step": 1000
|
| 69 |
},
|
| 70 |
{
|
| 71 |
-
"epoch": 0.
|
| 72 |
"learning_rate": 2.2000000000000003e-05,
|
| 73 |
-
"loss":
|
| 74 |
"step": 1100
|
| 75 |
},
|
| 76 |
{
|
| 77 |
-
"epoch": 0.
|
| 78 |
"learning_rate": 2.4e-05,
|
| 79 |
-
"loss":
|
| 80 |
"step": 1200
|
| 81 |
},
|
| 82 |
{
|
| 83 |
-
"epoch": 0.
|
| 84 |
"learning_rate": 2.6000000000000002e-05,
|
| 85 |
-
"loss":
|
| 86 |
"step": 1300
|
| 87 |
},
|
| 88 |
{
|
| 89 |
-
"epoch": 0.
|
| 90 |
"learning_rate": 2.8000000000000003e-05,
|
| 91 |
-
"loss":
|
| 92 |
"step": 1400
|
| 93 |
},
|
| 94 |
{
|
| 95 |
-
"epoch":
|
| 96 |
"learning_rate": 3e-05,
|
| 97 |
-
"loss":
|
| 98 |
"step": 1500
|
| 99 |
},
|
| 100 |
{
|
| 101 |
-
"epoch":
|
| 102 |
"learning_rate": 3.2000000000000005e-05,
|
| 103 |
-
"loss":
|
| 104 |
"step": 1600
|
| 105 |
},
|
| 106 |
{
|
| 107 |
-
"epoch":
|
| 108 |
"learning_rate": 3.4000000000000007e-05,
|
| 109 |
-
"loss":
|
| 110 |
"step": 1700
|
| 111 |
},
|
| 112 |
{
|
| 113 |
-
"epoch":
|
| 114 |
"learning_rate": 3.6e-05,
|
| 115 |
-
"loss":
|
| 116 |
"step": 1800
|
| 117 |
},
|
| 118 |
{
|
| 119 |
-
"epoch":
|
| 120 |
"learning_rate": 3.8e-05,
|
| 121 |
-
"loss":
|
| 122 |
"step": 1900
|
| 123 |
},
|
| 124 |
{
|
| 125 |
-
"epoch":
|
| 126 |
"learning_rate": 4e-05,
|
| 127 |
-
"loss":
|
| 128 |
"step": 2000
|
| 129 |
},
|
| 130 |
{
|
| 131 |
-
"epoch":
|
| 132 |
-
"eval_bleu":
|
| 133 |
"eval_em": 0.0,
|
| 134 |
-
"eval_gen_len":
|
| 135 |
-
"eval_loss": 6.
|
| 136 |
-
"
|
| 137 |
-
"
|
| 138 |
-
"
|
| 139 |
-
"eval_steps_per_second": 0.083,
|
| 140 |
"step": 2000
|
| 141 |
},
|
| 142 |
{
|
| 143 |
-
"epoch":
|
| 144 |
"learning_rate": 4.2e-05,
|
| 145 |
-
"loss":
|
| 146 |
"step": 2100
|
| 147 |
},
|
| 148 |
{
|
| 149 |
-
"epoch":
|
| 150 |
"learning_rate": 4.4000000000000006e-05,
|
| 151 |
-
"loss":
|
| 152 |
"step": 2200
|
| 153 |
},
|
| 154 |
{
|
| 155 |
-
"epoch":
|
| 156 |
"learning_rate": 4.600000000000001e-05,
|
| 157 |
-
"loss":
|
| 158 |
"step": 2300
|
| 159 |
},
|
| 160 |
{
|
| 161 |
-
"epoch": 1.
|
| 162 |
"learning_rate": 4.8e-05,
|
| 163 |
-
"loss":
|
| 164 |
"step": 2400
|
| 165 |
},
|
| 166 |
{
|
| 167 |
-
"epoch": 1.
|
| 168 |
"learning_rate": 5e-05,
|
| 169 |
-
"loss": 5.
|
| 170 |
"step": 2500
|
| 171 |
},
|
| 172 |
{
|
| 173 |
-
"epoch": 1.
|
| 174 |
-
"learning_rate": 4.
|
| 175 |
-
"loss":
|
| 176 |
"step": 2600
|
| 177 |
},
|
| 178 |
{
|
| 179 |
-
"epoch": 1.
|
| 180 |
-
"learning_rate": 4.
|
| 181 |
-
"loss":
|
| 182 |
"step": 2700
|
| 183 |
},
|
| 184 |
{
|
| 185 |
-
"epoch": 1.
|
| 186 |
-
"learning_rate": 4.
|
| 187 |
-
"loss": 4.
|
| 188 |
"step": 2800
|
| 189 |
},
|
| 190 |
{
|
| 191 |
-
"epoch":
|
| 192 |
-
"learning_rate": 4.
|
| 193 |
-
"loss":
|
| 194 |
"step": 2900
|
| 195 |
},
|
| 196 |
{
|
| 197 |
-
"epoch": 1
|
| 198 |
-
"learning_rate": 4.
|
| 199 |
-
"loss": 4.
|
| 200 |
"step": 3000
|
| 201 |
},
|
| 202 |
{
|
| 203 |
-
"epoch":
|
| 204 |
-
"learning_rate": 4.
|
| 205 |
-
"loss": 4.
|
| 206 |
"step": 3100
|
| 207 |
},
|
| 208 |
{
|
| 209 |
-
"epoch":
|
| 210 |
-
"learning_rate": 4.
|
| 211 |
-
"loss":
|
| 212 |
"step": 3200
|
| 213 |
},
|
| 214 |
{
|
| 215 |
-
"epoch":
|
| 216 |
-
"learning_rate": 4.
|
| 217 |
-
"loss":
|
| 218 |
"step": 3300
|
| 219 |
},
|
| 220 |
{
|
| 221 |
-
"epoch":
|
| 222 |
-
"learning_rate": 4.
|
| 223 |
-
"loss":
|
| 224 |
"step": 3400
|
| 225 |
},
|
| 226 |
{
|
| 227 |
-
"epoch":
|
| 228 |
-
"learning_rate": 4.
|
| 229 |
-
"loss":
|
| 230 |
"step": 3500
|
| 231 |
},
|
| 232 |
{
|
| 233 |
-
"epoch":
|
| 234 |
-
"learning_rate": 4.
|
| 235 |
-
"loss": 3.
|
| 236 |
"step": 3600
|
| 237 |
},
|
| 238 |
{
|
| 239 |
-
"epoch":
|
| 240 |
-
"learning_rate": 4.
|
| 241 |
-
"loss": 3.
|
| 242 |
"step": 3700
|
| 243 |
},
|
| 244 |
{
|
| 245 |
-
"epoch":
|
| 246 |
-
"learning_rate": 4.
|
| 247 |
-
"loss": 3.
|
| 248 |
"step": 3800
|
| 249 |
},
|
| 250 |
{
|
| 251 |
-
"epoch":
|
| 252 |
-
"learning_rate": 4.
|
| 253 |
-
"loss": 3.
|
| 254 |
"step": 3900
|
| 255 |
},
|
| 256 |
{
|
| 257 |
-
"epoch":
|
| 258 |
-
"learning_rate": 4.
|
| 259 |
-
"loss": 3.
|
| 260 |
"step": 4000
|
| 261 |
},
|
| 262 |
{
|
| 263 |
-
"epoch":
|
| 264 |
-
"eval_bleu":
|
| 265 |
"eval_em": 0.0,
|
| 266 |
-
"eval_gen_len":
|
| 267 |
-
"eval_loss":
|
| 268 |
-
"
|
| 269 |
-
"
|
| 270 |
-
"
|
| 271 |
-
"eval_steps_per_second": 0.231,
|
| 272 |
"step": 4000
|
| 273 |
},
|
| 274 |
{
|
| 275 |
-
"epoch":
|
| 276 |
-
"learning_rate": 4.
|
| 277 |
-
"loss":
|
| 278 |
"step": 4100
|
| 279 |
},
|
| 280 |
{
|
| 281 |
-
"epoch":
|
| 282 |
-
"learning_rate": 4.
|
| 283 |
-
"loss":
|
| 284 |
"step": 4200
|
| 285 |
},
|
| 286 |
{
|
| 287 |
-
"epoch":
|
| 288 |
-
"learning_rate":
|
| 289 |
-
"loss":
|
| 290 |
"step": 4300
|
| 291 |
},
|
| 292 |
{
|
| 293 |
-
"epoch":
|
| 294 |
-
"learning_rate":
|
| 295 |
-
"loss":
|
| 296 |
"step": 4400
|
| 297 |
},
|
| 298 |
{
|
| 299 |
-
"epoch":
|
| 300 |
-
"learning_rate":
|
| 301 |
-
"loss":
|
| 302 |
"step": 4500
|
| 303 |
},
|
| 304 |
{
|
| 305 |
-
"epoch":
|
| 306 |
-
"learning_rate":
|
| 307 |
-
"loss":
|
| 308 |
"step": 4600
|
| 309 |
},
|
| 310 |
{
|
| 311 |
-
"epoch":
|
| 312 |
-
"learning_rate":
|
| 313 |
-
"loss":
|
| 314 |
"step": 4700
|
| 315 |
},
|
| 316 |
{
|
| 317 |
-
"epoch":
|
| 318 |
-
"learning_rate":
|
| 319 |
-
"loss":
|
| 320 |
"step": 4800
|
| 321 |
},
|
| 322 |
{
|
| 323 |
-
"epoch":
|
| 324 |
-
"learning_rate":
|
| 325 |
-
"loss": 2.
|
| 326 |
"step": 4900
|
| 327 |
},
|
| 328 |
{
|
| 329 |
-
"epoch":
|
| 330 |
-
"learning_rate":
|
| 331 |
-
"loss": 2.
|
| 332 |
"step": 5000
|
| 333 |
},
|
| 334 |
{
|
| 335 |
-
"epoch":
|
| 336 |
-
"learning_rate":
|
| 337 |
-
"loss": 2.
|
| 338 |
"step": 5100
|
| 339 |
},
|
| 340 |
{
|
| 341 |
-
"epoch":
|
| 342 |
-
"learning_rate":
|
| 343 |
-
"loss": 2.
|
| 344 |
"step": 5200
|
| 345 |
},
|
| 346 |
{
|
| 347 |
-
"epoch":
|
| 348 |
-
"learning_rate":
|
| 349 |
-
"loss": 2.
|
| 350 |
"step": 5300
|
| 351 |
},
|
| 352 |
{
|
| 353 |
-
"epoch":
|
| 354 |
-
"learning_rate":
|
| 355 |
-
"loss":
|
| 356 |
"step": 5400
|
| 357 |
},
|
| 358 |
{
|
| 359 |
-
"epoch":
|
| 360 |
-
"learning_rate":
|
| 361 |
-
"loss":
|
| 362 |
"step": 5500
|
| 363 |
},
|
| 364 |
{
|
| 365 |
-
"epoch":
|
| 366 |
-
"learning_rate":
|
| 367 |
-
"loss":
|
| 368 |
"step": 5600
|
| 369 |
},
|
| 370 |
{
|
| 371 |
-
"epoch":
|
| 372 |
-
"learning_rate":
|
| 373 |
-
"loss":
|
| 374 |
"step": 5700
|
| 375 |
},
|
| 376 |
{
|
| 377 |
-
"epoch":
|
| 378 |
-
"learning_rate":
|
| 379 |
-
"loss":
|
| 380 |
"step": 5800
|
| 381 |
},
|
| 382 |
{
|
| 383 |
-
"epoch":
|
| 384 |
-
"learning_rate":
|
| 385 |
-
"loss":
|
| 386 |
"step": 5900
|
| 387 |
},
|
| 388 |
{
|
| 389 |
-
"epoch":
|
| 390 |
-
"learning_rate":
|
| 391 |
-
"loss":
|
| 392 |
"step": 6000
|
| 393 |
},
|
| 394 |
{
|
| 395 |
-
"epoch":
|
| 396 |
-
"eval_bleu":
|
| 397 |
-
"eval_em": 0.
|
| 398 |
-
"eval_gen_len": 44.
|
| 399 |
-
"eval_loss":
|
| 400 |
-
"
|
| 401 |
-
"
|
| 402 |
-
"
|
| 403 |
-
"eval_steps_per_second": 0.149,
|
| 404 |
"step": 6000
|
| 405 |
},
|
| 406 |
{
|
| 407 |
-
"epoch":
|
| 408 |
-
"learning_rate":
|
| 409 |
-
"loss":
|
| 410 |
"step": 6100
|
| 411 |
},
|
| 412 |
{
|
| 413 |
-
"epoch":
|
| 414 |
-
"learning_rate":
|
| 415 |
-
"loss":
|
| 416 |
"step": 6200
|
| 417 |
},
|
| 418 |
{
|
| 419 |
-
"epoch":
|
| 420 |
-
"learning_rate":
|
| 421 |
-
"loss":
|
| 422 |
"step": 6300
|
| 423 |
},
|
| 424 |
{
|
| 425 |
-
"epoch":
|
| 426 |
-
"learning_rate":
|
| 427 |
-
"loss":
|
| 428 |
"step": 6400
|
| 429 |
},
|
| 430 |
{
|
| 431 |
-
"epoch":
|
| 432 |
-
"learning_rate":
|
| 433 |
-
"loss":
|
| 434 |
"step": 6500
|
| 435 |
},
|
| 436 |
{
|
| 437 |
-
"epoch":
|
| 438 |
-
"learning_rate":
|
| 439 |
-
"loss":
|
| 440 |
"step": 6600
|
| 441 |
},
|
| 442 |
{
|
| 443 |
-
"epoch":
|
| 444 |
-
"learning_rate":
|
| 445 |
-
"loss":
|
| 446 |
"step": 6700
|
| 447 |
},
|
| 448 |
{
|
| 449 |
-
"epoch":
|
| 450 |
-
"learning_rate":
|
| 451 |
-
"loss":
|
| 452 |
"step": 6800
|
| 453 |
},
|
| 454 |
{
|
| 455 |
-
"epoch":
|
| 456 |
-
"learning_rate":
|
| 457 |
-
"loss":
|
| 458 |
"step": 6900
|
| 459 |
},
|
| 460 |
{
|
| 461 |
-
"epoch":
|
| 462 |
-
"learning_rate":
|
| 463 |
-
"loss":
|
| 464 |
"step": 7000
|
| 465 |
},
|
| 466 |
{
|
| 467 |
-
"epoch":
|
| 468 |
-
"learning_rate":
|
| 469 |
-
"loss": 1.
|
| 470 |
"step": 7100
|
| 471 |
},
|
| 472 |
{
|
| 473 |
-
"epoch":
|
| 474 |
-
"learning_rate":
|
| 475 |
-
"loss": 1.
|
| 476 |
"step": 7200
|
| 477 |
},
|
| 478 |
{
|
| 479 |
-
"epoch":
|
| 480 |
-
"learning_rate":
|
| 481 |
-
"loss": 1.
|
| 482 |
"step": 7300
|
| 483 |
},
|
| 484 |
{
|
| 485 |
-
"epoch":
|
| 486 |
-
"learning_rate":
|
| 487 |
-
"loss": 1.
|
| 488 |
"step": 7400
|
| 489 |
},
|
| 490 |
{
|
| 491 |
-
"epoch":
|
| 492 |
-
"learning_rate":
|
| 493 |
-
"loss": 1.
|
| 494 |
"step": 7500
|
| 495 |
},
|
| 496 |
{
|
| 497 |
-
"epoch":
|
| 498 |
-
"learning_rate":
|
| 499 |
-
"loss": 1.
|
| 500 |
"step": 7600
|
| 501 |
},
|
| 502 |
{
|
| 503 |
-
"epoch":
|
| 504 |
-
"learning_rate":
|
| 505 |
-
"loss": 1.
|
| 506 |
"step": 7700
|
| 507 |
},
|
| 508 |
{
|
| 509 |
-
"epoch":
|
| 510 |
-
"learning_rate":
|
| 511 |
-
"loss": 1.
|
| 512 |
"step": 7800
|
| 513 |
},
|
| 514 |
{
|
| 515 |
-
"epoch":
|
| 516 |
-
"learning_rate":
|
| 517 |
-
"loss": 1.
|
| 518 |
"step": 7900
|
| 519 |
},
|
| 520 |
{
|
| 521 |
-
"epoch":
|
| 522 |
-
"learning_rate":
|
| 523 |
-
"loss": 1.
|
| 524 |
"step": 8000
|
| 525 |
},
|
| 526 |
{
|
| 527 |
-
"epoch":
|
| 528 |
-
"eval_bleu":
|
| 529 |
-
"eval_em": 0.
|
| 530 |
-
"eval_gen_len": 46.
|
| 531 |
-
"eval_loss": 1.
|
| 532 |
-
"
|
| 533 |
-
"
|
| 534 |
-
"
|
| 535 |
-
"eval_steps_per_second": 0.158,
|
| 536 |
"step": 8000
|
| 537 |
},
|
| 538 |
{
|
| 539 |
-
"epoch":
|
| 540 |
-
"learning_rate":
|
| 541 |
-
"loss": 1.
|
| 542 |
"step": 8100
|
| 543 |
},
|
| 544 |
{
|
| 545 |
-
"epoch":
|
| 546 |
-
"learning_rate":
|
| 547 |
-
"loss": 1.
|
| 548 |
"step": 8200
|
| 549 |
},
|
| 550 |
{
|
| 551 |
-
"epoch":
|
| 552 |
-
"learning_rate":
|
| 553 |
-
"loss":
|
| 554 |
"step": 8300
|
| 555 |
},
|
| 556 |
{
|
| 557 |
-
"epoch":
|
| 558 |
-
"learning_rate":
|
| 559 |
-
"loss": 1.
|
| 560 |
"step": 8400
|
| 561 |
},
|
| 562 |
{
|
| 563 |
-
"epoch":
|
| 564 |
-
"learning_rate":
|
| 565 |
-
"loss":
|
| 566 |
"step": 8500
|
| 567 |
},
|
| 568 |
{
|
| 569 |
-
"epoch":
|
| 570 |
-
"learning_rate":
|
| 571 |
-
"loss":
|
| 572 |
"step": 8600
|
| 573 |
},
|
| 574 |
{
|
| 575 |
-
"epoch":
|
| 576 |
-
"learning_rate":
|
| 577 |
-
"loss":
|
| 578 |
"step": 8700
|
| 579 |
},
|
| 580 |
{
|
| 581 |
-
"epoch":
|
| 582 |
-
"learning_rate":
|
| 583 |
-
"loss":
|
| 584 |
"step": 8800
|
| 585 |
},
|
| 586 |
{
|
| 587 |
-
"epoch":
|
| 588 |
-
"learning_rate":
|
| 589 |
-
"loss":
|
| 590 |
"step": 8900
|
| 591 |
},
|
| 592 |
{
|
| 593 |
-
"epoch":
|
| 594 |
-
"learning_rate":
|
| 595 |
-
"loss":
|
| 596 |
"step": 9000
|
| 597 |
},
|
| 598 |
{
|
| 599 |
-
"epoch":
|
| 600 |
-
"learning_rate":
|
| 601 |
-
"loss":
|
| 602 |
"step": 9100
|
| 603 |
},
|
| 604 |
{
|
| 605 |
-
"epoch":
|
| 606 |
-
"learning_rate":
|
| 607 |
-
"loss":
|
| 608 |
"step": 9200
|
| 609 |
},
|
| 610 |
{
|
| 611 |
-
"epoch":
|
| 612 |
-
"learning_rate":
|
| 613 |
-
"loss":
|
| 614 |
"step": 9300
|
| 615 |
},
|
| 616 |
{
|
| 617 |
-
"epoch":
|
| 618 |
-
"learning_rate":
|
| 619 |
-
"loss":
|
| 620 |
"step": 9400
|
| 621 |
},
|
| 622 |
{
|
| 623 |
-
"epoch":
|
| 624 |
-
"learning_rate":
|
| 625 |
-
"loss":
|
| 626 |
"step": 9500
|
| 627 |
},
|
| 628 |
{
|
| 629 |
-
"epoch":
|
| 630 |
-
"learning_rate":
|
| 631 |
-
"loss":
|
| 632 |
"step": 9600
|
| 633 |
},
|
| 634 |
{
|
| 635 |
-
"epoch":
|
| 636 |
-
"learning_rate":
|
| 637 |
-
"loss":
|
| 638 |
"step": 9700
|
| 639 |
},
|
| 640 |
{
|
| 641 |
-
"epoch":
|
| 642 |
-
"learning_rate":
|
| 643 |
-
"loss":
|
| 644 |
"step": 9800
|
| 645 |
},
|
| 646 |
{
|
| 647 |
-
"epoch":
|
| 648 |
-
"learning_rate":
|
| 649 |
-
"loss":
|
| 650 |
"step": 9900
|
| 651 |
},
|
| 652 |
{
|
| 653 |
-
"epoch":
|
| 654 |
-
"learning_rate":
|
| 655 |
-
"loss":
|
| 656 |
"step": 10000
|
| 657 |
},
|
| 658 |
{
|
| 659 |
-
"epoch":
|
| 660 |
-
"eval_bleu":
|
| 661 |
-
"eval_em": 0.
|
| 662 |
-
"eval_gen_len":
|
| 663 |
-
"eval_loss": 1.
|
| 664 |
-
"
|
| 665 |
-
"
|
| 666 |
-
"
|
| 667 |
-
"eval_steps_per_second": 0.174,
|
| 668 |
"step": 10000
|
| 669 |
},
|
| 670 |
{
|
| 671 |
-
"epoch":
|
| 672 |
-
"learning_rate":
|
| 673 |
-
"loss":
|
| 674 |
"step": 10100
|
| 675 |
},
|
| 676 |
{
|
| 677 |
-
"epoch":
|
| 678 |
-
"learning_rate":
|
| 679 |
-
"loss":
|
| 680 |
"step": 10200
|
| 681 |
},
|
| 682 |
{
|
| 683 |
-
"epoch":
|
| 684 |
-
"learning_rate":
|
| 685 |
-
"loss":
|
| 686 |
"step": 10300
|
| 687 |
},
|
| 688 |
{
|
| 689 |
-
"epoch":
|
| 690 |
-
"learning_rate":
|
| 691 |
-
"loss":
|
| 692 |
"step": 10400
|
| 693 |
},
|
| 694 |
{
|
| 695 |
-
"epoch":
|
| 696 |
-
"learning_rate":
|
| 697 |
-
"loss":
|
| 698 |
"step": 10500
|
| 699 |
},
|
| 700 |
{
|
| 701 |
-
"epoch":
|
| 702 |
-
"learning_rate":
|
| 703 |
-
"loss": 0.
|
| 704 |
"step": 10600
|
| 705 |
},
|
| 706 |
{
|
| 707 |
-
"epoch":
|
| 708 |
-
"learning_rate":
|
| 709 |
-
"loss":
|
| 710 |
"step": 10700
|
| 711 |
},
|
| 712 |
{
|
| 713 |
-
"epoch":
|
| 714 |
-
"learning_rate": 3.
|
| 715 |
-
"loss":
|
| 716 |
"step": 10800
|
| 717 |
},
|
| 718 |
{
|
| 719 |
-
"epoch":
|
| 720 |
-
"learning_rate": 3.
|
| 721 |
-
"loss":
|
| 722 |
"step": 10900
|
| 723 |
},
|
| 724 |
{
|
| 725 |
-
"epoch":
|
| 726 |
-
"learning_rate":
|
| 727 |
-
"loss":
|
| 728 |
"step": 11000
|
| 729 |
},
|
| 730 |
{
|
| 731 |
-
"epoch":
|
| 732 |
-
"learning_rate":
|
| 733 |
-
"loss":
|
| 734 |
"step": 11100
|
| 735 |
},
|
| 736 |
{
|
| 737 |
-
"epoch":
|
| 738 |
-
"learning_rate":
|
| 739 |
-
"loss": 0.
|
| 740 |
"step": 11200
|
| 741 |
},
|
| 742 |
{
|
| 743 |
-
"epoch":
|
| 744 |
-
"learning_rate":
|
| 745 |
-
"loss":
|
| 746 |
"step": 11300
|
| 747 |
},
|
| 748 |
-
{
|
| 749 |
-
"epoch": 4.73,
|
| 750 |
-
"learning_rate": 2.9398148148148146e-05,
|
| 751 |
-
"loss": 1.0798,
|
| 752 |
-
"step": 11400
|
| 753 |
-
},
|
| 754 |
-
{
|
| 755 |
-
"epoch": 4.77,
|
| 756 |
-
"learning_rate": 2.916666666666667e-05,
|
| 757 |
-
"loss": 1.0331,
|
| 758 |
-
"step": 11500
|
| 759 |
-
},
|
| 760 |
-
{
|
| 761 |
-
"epoch": 4.81,
|
| 762 |
-
"learning_rate": 2.8935185185185186e-05,
|
| 763 |
-
"loss": 0.9571,
|
| 764 |
-
"step": 11600
|
| 765 |
-
},
|
| 766 |
-
{
|
| 767 |
-
"epoch": 4.85,
|
| 768 |
-
"learning_rate": 2.8703703703703706e-05,
|
| 769 |
-
"loss": 0.9468,
|
| 770 |
-
"step": 11700
|
| 771 |
-
},
|
| 772 |
-
{
|
| 773 |
-
"epoch": 4.9,
|
| 774 |
-
"learning_rate": 2.8472222222222223e-05,
|
| 775 |
-
"loss": 0.9652,
|
| 776 |
-
"step": 11800
|
| 777 |
-
},
|
| 778 |
-
{
|
| 779 |
-
"epoch": 4.94,
|
| 780 |
-
"learning_rate": 2.824074074074074e-05,
|
| 781 |
-
"loss": 0.968,
|
| 782 |
-
"step": 11900
|
| 783 |
-
},
|
| 784 |
-
{
|
| 785 |
-
"epoch": 4.98,
|
| 786 |
-
"learning_rate": 2.8009259259259263e-05,
|
| 787 |
-
"loss": 0.9597,
|
| 788 |
-
"step": 12000
|
| 789 |
-
},
|
| 790 |
-
{
|
| 791 |
-
"epoch": 4.98,
|
| 792 |
-
"eval_bleu": 74.2575,
|
| 793 |
-
"eval_em": 0.04,
|
| 794 |
-
"eval_gen_len": 47.52,
|
| 795 |
-
"eval_loss": 0.9973701238632202,
|
| 796 |
-
"eval_rm": 0.16,
|
| 797 |
-
"eval_runtime": 41.5642,
|
| 798 |
-
"eval_samples_per_second": 1.203,
|
| 799 |
-
"eval_steps_per_second": 0.168,
|
| 800 |
-
"step": 12000
|
| 801 |
-
},
|
| 802 |
-
{
|
| 803 |
-
"epoch": 5.02,
|
| 804 |
-
"learning_rate": 2.777777777777778e-05,
|
| 805 |
-
"loss": 0.8246,
|
| 806 |
-
"step": 12100
|
| 807 |
-
},
|
| 808 |
-
{
|
| 809 |
-
"epoch": 5.06,
|
| 810 |
-
"learning_rate": 2.75462962962963e-05,
|
| 811 |
-
"loss": 0.7769,
|
| 812 |
-
"step": 12200
|
| 813 |
-
},
|
| 814 |
-
{
|
| 815 |
-
"epoch": 5.1,
|
| 816 |
-
"learning_rate": 2.7314814814814816e-05,
|
| 817 |
-
"loss": 0.7765,
|
| 818 |
-
"step": 12300
|
| 819 |
-
},
|
| 820 |
-
{
|
| 821 |
-
"epoch": 5.15,
|
| 822 |
-
"learning_rate": 2.7083333333333332e-05,
|
| 823 |
-
"loss": 0.7842,
|
| 824 |
-
"step": 12400
|
| 825 |
-
},
|
| 826 |
-
{
|
| 827 |
-
"epoch": 5.19,
|
| 828 |
-
"learning_rate": 2.6851851851851855e-05,
|
| 829 |
-
"loss": 0.7961,
|
| 830 |
-
"step": 12500
|
| 831 |
-
},
|
| 832 |
-
{
|
| 833 |
-
"epoch": 5.23,
|
| 834 |
-
"learning_rate": 2.6620370370370372e-05,
|
| 835 |
-
"loss": 0.7886,
|
| 836 |
-
"step": 12600
|
| 837 |
-
},
|
| 838 |
-
{
|
| 839 |
-
"epoch": 5.27,
|
| 840 |
-
"learning_rate": 2.6388888888888892e-05,
|
| 841 |
-
"loss": 0.7946,
|
| 842 |
-
"step": 12700
|
| 843 |
-
},
|
| 844 |
-
{
|
| 845 |
-
"epoch": 5.31,
|
| 846 |
-
"learning_rate": 2.615740740740741e-05,
|
| 847 |
-
"loss": 0.7629,
|
| 848 |
-
"step": 12800
|
| 849 |
-
},
|
| 850 |
-
{
|
| 851 |
-
"epoch": 5.35,
|
| 852 |
-
"learning_rate": 2.5925925925925925e-05,
|
| 853 |
-
"loss": 0.7594,
|
| 854 |
-
"step": 12900
|
| 855 |
-
},
|
| 856 |
-
{
|
| 857 |
-
"epoch": 5.39,
|
| 858 |
-
"learning_rate": 2.5694444444444445e-05,
|
| 859 |
-
"loss": 0.7429,
|
| 860 |
-
"step": 13000
|
| 861 |
-
},
|
| 862 |
-
{
|
| 863 |
-
"epoch": 5.44,
|
| 864 |
-
"learning_rate": 2.5462962962962965e-05,
|
| 865 |
-
"loss": 0.7468,
|
| 866 |
-
"step": 13100
|
| 867 |
-
},
|
| 868 |
-
{
|
| 869 |
-
"epoch": 5.48,
|
| 870 |
-
"learning_rate": 2.5231481481481485e-05,
|
| 871 |
-
"loss": 0.7168,
|
| 872 |
-
"step": 13200
|
| 873 |
-
},
|
| 874 |
-
{
|
| 875 |
-
"epoch": 5.52,
|
| 876 |
-
"learning_rate": 2.5e-05,
|
| 877 |
-
"loss": 0.7351,
|
| 878 |
-
"step": 13300
|
| 879 |
-
},
|
| 880 |
-
{
|
| 881 |
-
"epoch": 5.56,
|
| 882 |
-
"learning_rate": 2.4768518518518518e-05,
|
| 883 |
-
"loss": 0.7345,
|
| 884 |
-
"step": 13400
|
| 885 |
-
},
|
| 886 |
-
{
|
| 887 |
-
"epoch": 5.6,
|
| 888 |
-
"learning_rate": 2.4537037037037038e-05,
|
| 889 |
-
"loss": 0.7124,
|
| 890 |
-
"step": 13500
|
| 891 |
-
},
|
| 892 |
-
{
|
| 893 |
-
"epoch": 5.64,
|
| 894 |
-
"learning_rate": 2.4305555555555558e-05,
|
| 895 |
-
"loss": 0.7162,
|
| 896 |
-
"step": 13600
|
| 897 |
-
},
|
| 898 |
-
{
|
| 899 |
-
"epoch": 5.68,
|
| 900 |
-
"learning_rate": 2.4074074074074074e-05,
|
| 901 |
-
"loss": 0.7175,
|
| 902 |
-
"step": 13700
|
| 903 |
-
},
|
| 904 |
-
{
|
| 905 |
-
"epoch": 5.73,
|
| 906 |
-
"learning_rate": 2.3842592592592594e-05,
|
| 907 |
-
"loss": 0.7096,
|
| 908 |
-
"step": 13800
|
| 909 |
-
},
|
| 910 |
-
{
|
| 911 |
-
"epoch": 5.77,
|
| 912 |
-
"learning_rate": 2.361111111111111e-05,
|
| 913 |
-
"loss": 0.7534,
|
| 914 |
-
"step": 13900
|
| 915 |
-
},
|
| 916 |
-
{
|
| 917 |
-
"epoch": 5.81,
|
| 918 |
-
"learning_rate": 2.337962962962963e-05,
|
| 919 |
-
"loss": 0.7033,
|
| 920 |
-
"step": 14000
|
| 921 |
-
},
|
| 922 |
-
{
|
| 923 |
-
"epoch": 5.81,
|
| 924 |
-
"eval_bleu": 74.9164,
|
| 925 |
-
"eval_em": 0.08,
|
| 926 |
-
"eval_gen_len": 45.64,
|
| 927 |
-
"eval_loss": 0.9389866590499878,
|
| 928 |
-
"eval_rm": 0.2,
|
| 929 |
-
"eval_runtime": 43.8848,
|
| 930 |
-
"eval_samples_per_second": 1.139,
|
| 931 |
-
"eval_steps_per_second": 0.16,
|
| 932 |
-
"step": 14000
|
| 933 |
-
},
|
| 934 |
-
{
|
| 935 |
-
"epoch": 5.85,
|
| 936 |
-
"learning_rate": 2.314814814814815e-05,
|
| 937 |
-
"loss": 0.7113,
|
| 938 |
-
"step": 14100
|
| 939 |
-
},
|
| 940 |
-
{
|
| 941 |
-
"epoch": 5.89,
|
| 942 |
-
"learning_rate": 2.2916666666666667e-05,
|
| 943 |
-
"loss": 0.6708,
|
| 944 |
-
"step": 14200
|
| 945 |
-
},
|
| 946 |
-
{
|
| 947 |
-
"epoch": 5.93,
|
| 948 |
-
"learning_rate": 2.2685185185185187e-05,
|
| 949 |
-
"loss": 0.7416,
|
| 950 |
-
"step": 14300
|
| 951 |
-
},
|
| 952 |
-
{
|
| 953 |
-
"epoch": 5.98,
|
| 954 |
-
"learning_rate": 2.2453703703703703e-05,
|
| 955 |
-
"loss": 0.7072,
|
| 956 |
-
"step": 14400
|
| 957 |
-
},
|
| 958 |
-
{
|
| 959 |
-
"epoch": 6.02,
|
| 960 |
-
"learning_rate": 2.2222222222222223e-05,
|
| 961 |
-
"loss": 0.635,
|
| 962 |
-
"step": 14500
|
| 963 |
-
},
|
| 964 |
-
{
|
| 965 |
-
"epoch": 6.06,
|
| 966 |
-
"learning_rate": 2.1990740740740743e-05,
|
| 967 |
-
"loss": 0.5427,
|
| 968 |
-
"step": 14600
|
| 969 |
-
},
|
| 970 |
-
{
|
| 971 |
-
"epoch": 6.1,
|
| 972 |
-
"learning_rate": 2.175925925925926e-05,
|
| 973 |
-
"loss": 0.5473,
|
| 974 |
-
"step": 14700
|
| 975 |
-
},
|
| 976 |
-
{
|
| 977 |
-
"epoch": 6.14,
|
| 978 |
-
"learning_rate": 2.152777777777778e-05,
|
| 979 |
-
"loss": 0.5598,
|
| 980 |
-
"step": 14800
|
| 981 |
-
},
|
| 982 |
-
{
|
| 983 |
-
"epoch": 6.18,
|
| 984 |
-
"learning_rate": 2.1296296296296296e-05,
|
| 985 |
-
"loss": 0.5542,
|
| 986 |
-
"step": 14900
|
| 987 |
-
},
|
| 988 |
-
{
|
| 989 |
-
"epoch": 6.22,
|
| 990 |
-
"learning_rate": 2.1064814814814816e-05,
|
| 991 |
-
"loss": 0.5598,
|
| 992 |
-
"step": 15000
|
| 993 |
-
},
|
| 994 |
-
{
|
| 995 |
-
"epoch": 6.27,
|
| 996 |
-
"learning_rate": 2.0833333333333336e-05,
|
| 997 |
-
"loss": 0.5746,
|
| 998 |
-
"step": 15100
|
| 999 |
-
},
|
| 1000 |
-
{
|
| 1001 |
-
"epoch": 6.31,
|
| 1002 |
-
"learning_rate": 2.0601851851851853e-05,
|
| 1003 |
-
"loss": 0.5545,
|
| 1004 |
-
"step": 15200
|
| 1005 |
-
},
|
| 1006 |
-
{
|
| 1007 |
-
"epoch": 6.35,
|
| 1008 |
-
"learning_rate": 2.037037037037037e-05,
|
| 1009 |
-
"loss": 0.5896,
|
| 1010 |
-
"step": 15300
|
| 1011 |
-
},
|
| 1012 |
-
{
|
| 1013 |
-
"epoch": 6.39,
|
| 1014 |
-
"learning_rate": 2.013888888888889e-05,
|
| 1015 |
-
"loss": 0.5367,
|
| 1016 |
-
"step": 15400
|
| 1017 |
-
},
|
| 1018 |
-
{
|
| 1019 |
-
"epoch": 6.43,
|
| 1020 |
-
"learning_rate": 1.990740740740741e-05,
|
| 1021 |
-
"loss": 0.5577,
|
| 1022 |
-
"step": 15500
|
| 1023 |
-
},
|
| 1024 |
-
{
|
| 1025 |
-
"epoch": 6.47,
|
| 1026 |
-
"learning_rate": 1.967592592592593e-05,
|
| 1027 |
-
"loss": 0.5199,
|
| 1028 |
-
"step": 15600
|
| 1029 |
-
},
|
| 1030 |
-
{
|
| 1031 |
-
"epoch": 6.51,
|
| 1032 |
-
"learning_rate": 1.9444444444444445e-05,
|
| 1033 |
-
"loss": 0.5139,
|
| 1034 |
-
"step": 15700
|
| 1035 |
-
},
|
| 1036 |
-
{
|
| 1037 |
-
"epoch": 6.56,
|
| 1038 |
-
"learning_rate": 1.9212962962962962e-05,
|
| 1039 |
-
"loss": 0.5623,
|
| 1040 |
-
"step": 15800
|
| 1041 |
-
},
|
| 1042 |
-
{
|
| 1043 |
-
"epoch": 6.6,
|
| 1044 |
-
"learning_rate": 1.8981481481481482e-05,
|
| 1045 |
-
"loss": 0.5468,
|
| 1046 |
-
"step": 15900
|
| 1047 |
-
},
|
| 1048 |
-
{
|
| 1049 |
-
"epoch": 6.64,
|
| 1050 |
-
"learning_rate": 1.8750000000000002e-05,
|
| 1051 |
-
"loss": 0.5446,
|
| 1052 |
-
"step": 16000
|
| 1053 |
-
},
|
| 1054 |
-
{
|
| 1055 |
-
"epoch": 6.64,
|
| 1056 |
-
"eval_bleu": 75.9224,
|
| 1057 |
-
"eval_em": 0.06,
|
| 1058 |
-
"eval_gen_len": 47.34,
|
| 1059 |
-
"eval_loss": 0.9507510662078857,
|
| 1060 |
-
"eval_rm": 0.14,
|
| 1061 |
-
"eval_runtime": 42.686,
|
| 1062 |
-
"eval_samples_per_second": 1.171,
|
| 1063 |
-
"eval_steps_per_second": 0.164,
|
| 1064 |
-
"step": 16000
|
| 1065 |
-
},
|
| 1066 |
-
{
|
| 1067 |
-
"epoch": 6.68,
|
| 1068 |
-
"learning_rate": 1.8518518518518518e-05,
|
| 1069 |
-
"loss": 0.5024,
|
| 1070 |
-
"step": 16100
|
| 1071 |
-
},
|
| 1072 |
-
{
|
| 1073 |
-
"epoch": 6.72,
|
| 1074 |
-
"learning_rate": 1.8287037037037038e-05,
|
| 1075 |
-
"loss": 0.585,
|
| 1076 |
-
"step": 16200
|
| 1077 |
-
},
|
| 1078 |
-
{
|
| 1079 |
-
"epoch": 6.76,
|
| 1080 |
-
"learning_rate": 1.8055555555555555e-05,
|
| 1081 |
-
"loss": 0.5118,
|
| 1082 |
-
"step": 16300
|
| 1083 |
-
},
|
| 1084 |
-
{
|
| 1085 |
-
"epoch": 6.8,
|
| 1086 |
-
"learning_rate": 1.7824074074074075e-05,
|
| 1087 |
-
"loss": 0.5362,
|
| 1088 |
-
"step": 16400
|
| 1089 |
-
},
|
| 1090 |
-
{
|
| 1091 |
-
"epoch": 6.85,
|
| 1092 |
-
"learning_rate": 1.7592592592592595e-05,
|
| 1093 |
-
"loss": 0.5509,
|
| 1094 |
-
"step": 16500
|
| 1095 |
-
},
|
| 1096 |
-
{
|
| 1097 |
-
"epoch": 6.89,
|
| 1098 |
-
"learning_rate": 1.736111111111111e-05,
|
| 1099 |
-
"loss": 0.5336,
|
| 1100 |
-
"step": 16600
|
| 1101 |
-
},
|
| 1102 |
-
{
|
| 1103 |
-
"epoch": 6.93,
|
| 1104 |
-
"learning_rate": 1.712962962962963e-05,
|
| 1105 |
-
"loss": 0.5307,
|
| 1106 |
-
"step": 16700
|
| 1107 |
-
},
|
| 1108 |
-
{
|
| 1109 |
-
"epoch": 6.97,
|
| 1110 |
-
"learning_rate": 1.6898148148148148e-05,
|
| 1111 |
-
"loss": 0.5012,
|
| 1112 |
-
"step": 16800
|
| 1113 |
-
},
|
| 1114 |
-
{
|
| 1115 |
-
"epoch": 7.01,
|
| 1116 |
-
"learning_rate": 1.6666666666666667e-05,
|
| 1117 |
-
"loss": 0.5114,
|
| 1118 |
-
"step": 16900
|
| 1119 |
-
},
|
| 1120 |
-
{
|
| 1121 |
-
"epoch": 7.05,
|
| 1122 |
-
"learning_rate": 1.6435185185185187e-05,
|
| 1123 |
-
"loss": 0.4098,
|
| 1124 |
-
"step": 17000
|
| 1125 |
-
},
|
| 1126 |
-
{
|
| 1127 |
-
"epoch": 7.1,
|
| 1128 |
-
"learning_rate": 1.6203703703703704e-05,
|
| 1129 |
-
"loss": 0.3945,
|
| 1130 |
-
"step": 17100
|
| 1131 |
-
},
|
| 1132 |
-
{
|
| 1133 |
-
"epoch": 7.14,
|
| 1134 |
-
"learning_rate": 1.597222222222222e-05,
|
| 1135 |
-
"loss": 0.4309,
|
| 1136 |
-
"step": 17200
|
| 1137 |
-
},
|
| 1138 |
-
{
|
| 1139 |
-
"epoch": 7.18,
|
| 1140 |
-
"learning_rate": 1.574074074074074e-05,
|
| 1141 |
-
"loss": 0.446,
|
| 1142 |
-
"step": 17300
|
| 1143 |
-
},
|
| 1144 |
-
{
|
| 1145 |
-
"epoch": 7.22,
|
| 1146 |
-
"learning_rate": 1.550925925925926e-05,
|
| 1147 |
-
"loss": 0.3874,
|
| 1148 |
-
"step": 17400
|
| 1149 |
-
},
|
| 1150 |
-
{
|
| 1151 |
-
"epoch": 7.26,
|
| 1152 |
-
"learning_rate": 1.527777777777778e-05,
|
| 1153 |
-
"loss": 0.4103,
|
| 1154 |
-
"step": 17500
|
| 1155 |
-
},
|
| 1156 |
-
{
|
| 1157 |
-
"epoch": 7.3,
|
| 1158 |
-
"learning_rate": 1.5046296296296297e-05,
|
| 1159 |
-
"loss": 0.4097,
|
| 1160 |
-
"step": 17600
|
| 1161 |
-
},
|
| 1162 |
-
{
|
| 1163 |
-
"epoch": 7.34,
|
| 1164 |
-
"learning_rate": 1.4814814814814815e-05,
|
| 1165 |
-
"loss": 0.3853,
|
| 1166 |
-
"step": 17700
|
| 1167 |
-
},
|
| 1168 |
-
{
|
| 1169 |
-
"epoch": 7.39,
|
| 1170 |
-
"learning_rate": 1.4583333333333335e-05,
|
| 1171 |
-
"loss": 0.3967,
|
| 1172 |
-
"step": 17800
|
| 1173 |
-
},
|
| 1174 |
-
{
|
| 1175 |
-
"epoch": 7.43,
|
| 1176 |
-
"learning_rate": 1.4351851851851853e-05,
|
| 1177 |
-
"loss": 0.4024,
|
| 1178 |
-
"step": 17900
|
| 1179 |
-
},
|
| 1180 |
-
{
|
| 1181 |
-
"epoch": 7.47,
|
| 1182 |
-
"learning_rate": 1.412037037037037e-05,
|
| 1183 |
-
"loss": 0.4362,
|
| 1184 |
-
"step": 18000
|
| 1185 |
-
},
|
| 1186 |
-
{
|
| 1187 |
-
"epoch": 7.47,
|
| 1188 |
-
"eval_bleu": 77.2209,
|
| 1189 |
-
"eval_em": 0.1,
|
| 1190 |
-
"eval_gen_len": 44.36,
|
| 1191 |
-
"eval_loss": 0.8897470235824585,
|
| 1192 |
-
"eval_rm": 0.22,
|
| 1193 |
-
"eval_runtime": 38.4221,
|
| 1194 |
-
"eval_samples_per_second": 1.301,
|
| 1195 |
-
"eval_steps_per_second": 0.182,
|
| 1196 |
-
"step": 18000
|
| 1197 |
-
},
|
| 1198 |
-
{
|
| 1199 |
-
"epoch": 7.51,
|
| 1200 |
-
"learning_rate": 1.388888888888889e-05,
|
| 1201 |
-
"loss": 0.4094,
|
| 1202 |
-
"step": 18100
|
| 1203 |
-
},
|
| 1204 |
-
{
|
| 1205 |
-
"epoch": 7.55,
|
| 1206 |
-
"learning_rate": 1.3657407407407408e-05,
|
| 1207 |
-
"loss": 0.4048,
|
| 1208 |
-
"step": 18200
|
| 1209 |
-
},
|
| 1210 |
-
{
|
| 1211 |
-
"epoch": 7.59,
|
| 1212 |
-
"learning_rate": 1.3425925925925928e-05,
|
| 1213 |
-
"loss": 0.4047,
|
| 1214 |
-
"step": 18300
|
| 1215 |
-
},
|
| 1216 |
-
{
|
| 1217 |
-
"epoch": 7.63,
|
| 1218 |
-
"learning_rate": 1.3194444444444446e-05,
|
| 1219 |
-
"loss": 0.4203,
|
| 1220 |
-
"step": 18400
|
| 1221 |
-
},
|
| 1222 |
-
{
|
| 1223 |
-
"epoch": 7.68,
|
| 1224 |
-
"learning_rate": 1.2962962962962962e-05,
|
| 1225 |
-
"loss": 0.4176,
|
| 1226 |
-
"step": 18500
|
| 1227 |
-
},
|
| 1228 |
-
{
|
| 1229 |
-
"epoch": 7.72,
|
| 1230 |
-
"learning_rate": 1.2731481481481482e-05,
|
| 1231 |
-
"loss": 0.4351,
|
| 1232 |
-
"step": 18600
|
| 1233 |
-
},
|
| 1234 |
-
{
|
| 1235 |
-
"epoch": 7.76,
|
| 1236 |
-
"learning_rate": 1.25e-05,
|
| 1237 |
-
"loss": 0.3931,
|
| 1238 |
-
"step": 18700
|
| 1239 |
-
},
|
| 1240 |
-
{
|
| 1241 |
-
"epoch": 7.8,
|
| 1242 |
-
"learning_rate": 1.2268518518518519e-05,
|
| 1243 |
-
"loss": 0.3883,
|
| 1244 |
-
"step": 18800
|
| 1245 |
-
},
|
| 1246 |
-
{
|
| 1247 |
-
"epoch": 7.84,
|
| 1248 |
-
"learning_rate": 1.2037037037037037e-05,
|
| 1249 |
-
"loss": 0.3879,
|
| 1250 |
-
"step": 18900
|
| 1251 |
-
},
|
| 1252 |
-
{
|
| 1253 |
-
"epoch": 7.88,
|
| 1254 |
-
"learning_rate": 1.1805555555555555e-05,
|
| 1255 |
-
"loss": 0.3854,
|
| 1256 |
-
"step": 19000
|
| 1257 |
-
},
|
| 1258 |
-
{
|
| 1259 |
-
"epoch": 7.93,
|
| 1260 |
-
"learning_rate": 1.1574074074074075e-05,
|
| 1261 |
-
"loss": 0.3694,
|
| 1262 |
-
"step": 19100
|
| 1263 |
-
},
|
| 1264 |
{
|
| 1265 |
"epoch": 7.97,
|
| 1266 |
-
"learning_rate":
|
| 1267 |
-
"loss": 0.
|
| 1268 |
-
"step":
|
| 1269 |
-
},
|
| 1270 |
-
{
|
| 1271 |
-
"epoch": 8.01,
|
| 1272 |
-
"learning_rate": 1.1111111111111112e-05,
|
| 1273 |
-
"loss": 0.3859,
|
| 1274 |
-
"step": 19300
|
| 1275 |
-
},
|
| 1276 |
-
{
|
| 1277 |
-
"epoch": 8.05,
|
| 1278 |
-
"learning_rate": 1.087962962962963e-05,
|
| 1279 |
-
"loss": 0.2946,
|
| 1280 |
-
"step": 19400
|
| 1281 |
-
},
|
| 1282 |
-
{
|
| 1283 |
-
"epoch": 8.09,
|
| 1284 |
-
"learning_rate": 1.0648148148148148e-05,
|
| 1285 |
-
"loss": 0.2993,
|
| 1286 |
-
"step": 19500
|
| 1287 |
-
},
|
| 1288 |
-
{
|
| 1289 |
-
"epoch": 8.13,
|
| 1290 |
-
"learning_rate": 1.0416666666666668e-05,
|
| 1291 |
-
"loss": 0.3085,
|
| 1292 |
-
"step": 19600
|
| 1293 |
-
},
|
| 1294 |
-
{
|
| 1295 |
-
"epoch": 8.17,
|
| 1296 |
-
"learning_rate": 1.0185185185185185e-05,
|
| 1297 |
-
"loss": 0.3064,
|
| 1298 |
-
"step": 19700
|
| 1299 |
-
},
|
| 1300 |
-
{
|
| 1301 |
-
"epoch": 8.22,
|
| 1302 |
-
"learning_rate": 9.953703703703704e-06,
|
| 1303 |
-
"loss": 0.2933,
|
| 1304 |
-
"step": 19800
|
| 1305 |
-
},
|
| 1306 |
-
{
|
| 1307 |
-
"epoch": 8.26,
|
| 1308 |
-
"learning_rate": 9.722222222222223e-06,
|
| 1309 |
-
"loss": 0.3256,
|
| 1310 |
-
"step": 19900
|
| 1311 |
-
},
|
| 1312 |
-
{
|
| 1313 |
-
"epoch": 8.3,
|
| 1314 |
-
"learning_rate": 9.490740740740741e-06,
|
| 1315 |
-
"loss": 0.3231,
|
| 1316 |
-
"step": 20000
|
| 1317 |
-
},
|
| 1318 |
-
{
|
| 1319 |
-
"epoch": 8.3,
|
| 1320 |
-
"eval_bleu": 78.5526,
|
| 1321 |
-
"eval_em": 0.16,
|
| 1322 |
-
"eval_gen_len": 46.66,
|
| 1323 |
-
"eval_loss": 0.923231840133667,
|
| 1324 |
-
"eval_rm": 0.26,
|
| 1325 |
-
"eval_runtime": 38.1607,
|
| 1326 |
-
"eval_samples_per_second": 1.31,
|
| 1327 |
-
"eval_steps_per_second": 0.183,
|
| 1328 |
-
"step": 20000
|
| 1329 |
-
},
|
| 1330 |
-
{
|
| 1331 |
-
"epoch": 8.34,
|
| 1332 |
-
"learning_rate": 9.259259259259259e-06,
|
| 1333 |
-
"loss": 0.3179,
|
| 1334 |
-
"step": 20100
|
| 1335 |
-
},
|
| 1336 |
-
{
|
| 1337 |
-
"epoch": 8.38,
|
| 1338 |
-
"learning_rate": 9.027777777777777e-06,
|
| 1339 |
-
"loss": 0.318,
|
| 1340 |
-
"step": 20200
|
| 1341 |
-
},
|
| 1342 |
-
{
|
| 1343 |
-
"epoch": 8.42,
|
| 1344 |
-
"learning_rate": 8.796296296296297e-06,
|
| 1345 |
-
"loss": 0.3244,
|
| 1346 |
-
"step": 20300
|
| 1347 |
-
},
|
| 1348 |
-
{
|
| 1349 |
-
"epoch": 8.46,
|
| 1350 |
-
"learning_rate": 8.564814814814816e-06,
|
| 1351 |
-
"loss": 0.2916,
|
| 1352 |
-
"step": 20400
|
| 1353 |
-
},
|
| 1354 |
-
{
|
| 1355 |
-
"epoch": 8.51,
|
| 1356 |
-
"learning_rate": 8.333333333333334e-06,
|
| 1357 |
-
"loss": 0.2984,
|
| 1358 |
-
"step": 20500
|
| 1359 |
-
},
|
| 1360 |
-
{
|
| 1361 |
-
"epoch": 8.55,
|
| 1362 |
-
"learning_rate": 8.101851851851852e-06,
|
| 1363 |
-
"loss": 0.2965,
|
| 1364 |
-
"step": 20600
|
| 1365 |
-
},
|
| 1366 |
-
{
|
| 1367 |
-
"epoch": 8.59,
|
| 1368 |
-
"learning_rate": 7.87037037037037e-06,
|
| 1369 |
-
"loss": 0.3038,
|
| 1370 |
-
"step": 20700
|
| 1371 |
-
},
|
| 1372 |
-
{
|
| 1373 |
-
"epoch": 8.63,
|
| 1374 |
-
"learning_rate": 7.63888888888889e-06,
|
| 1375 |
-
"loss": 0.3222,
|
| 1376 |
-
"step": 20800
|
| 1377 |
-
},
|
| 1378 |
-
{
|
| 1379 |
-
"epoch": 8.67,
|
| 1380 |
-
"learning_rate": 7.4074074074074075e-06,
|
| 1381 |
-
"loss": 0.3272,
|
| 1382 |
-
"step": 20900
|
| 1383 |
-
},
|
| 1384 |
-
{
|
| 1385 |
-
"epoch": 8.71,
|
| 1386 |
-
"learning_rate": 7.1759259259259266e-06,
|
| 1387 |
-
"loss": 0.3165,
|
| 1388 |
-
"step": 21000
|
| 1389 |
-
},
|
| 1390 |
-
{
|
| 1391 |
-
"epoch": 8.76,
|
| 1392 |
-
"learning_rate": 6.944444444444445e-06,
|
| 1393 |
-
"loss": 0.329,
|
| 1394 |
-
"step": 21100
|
| 1395 |
-
},
|
| 1396 |
-
{
|
| 1397 |
-
"epoch": 8.8,
|
| 1398 |
-
"learning_rate": 6.712962962962964e-06,
|
| 1399 |
-
"loss": 0.2912,
|
| 1400 |
-
"step": 21200
|
| 1401 |
-
},
|
| 1402 |
-
{
|
| 1403 |
-
"epoch": 8.84,
|
| 1404 |
-
"learning_rate": 6.481481481481481e-06,
|
| 1405 |
-
"loss": 0.3109,
|
| 1406 |
-
"step": 21300
|
| 1407 |
-
},
|
| 1408 |
-
{
|
| 1409 |
-
"epoch": 8.88,
|
| 1410 |
-
"learning_rate": 6.25e-06,
|
| 1411 |
-
"loss": 0.2998,
|
| 1412 |
-
"step": 21400
|
| 1413 |
-
},
|
| 1414 |
-
{
|
| 1415 |
-
"epoch": 8.92,
|
| 1416 |
-
"learning_rate": 6.0185185185185185e-06,
|
| 1417 |
-
"loss": 0.3262,
|
| 1418 |
-
"step": 21500
|
| 1419 |
-
},
|
| 1420 |
-
{
|
| 1421 |
-
"epoch": 8.96,
|
| 1422 |
-
"learning_rate": 5.787037037037038e-06,
|
| 1423 |
-
"loss": 0.2877,
|
| 1424 |
-
"step": 21600
|
| 1425 |
-
},
|
| 1426 |
-
{
|
| 1427 |
-
"epoch": 9.0,
|
| 1428 |
-
"learning_rate": 5.555555555555556e-06,
|
| 1429 |
-
"loss": 0.2983,
|
| 1430 |
-
"step": 21700
|
| 1431 |
-
},
|
| 1432 |
-
{
|
| 1433 |
-
"epoch": 9.05,
|
| 1434 |
-
"learning_rate": 5.324074074074074e-06,
|
| 1435 |
-
"loss": 0.2471,
|
| 1436 |
-
"step": 21800
|
| 1437 |
-
},
|
| 1438 |
-
{
|
| 1439 |
-
"epoch": 9.09,
|
| 1440 |
-
"learning_rate": 5.092592592592592e-06,
|
| 1441 |
-
"loss": 0.2438,
|
| 1442 |
-
"step": 21900
|
| 1443 |
-
},
|
| 1444 |
-
{
|
| 1445 |
-
"epoch": 9.13,
|
| 1446 |
-
"learning_rate": 4.861111111111111e-06,
|
| 1447 |
-
"loss": 0.2584,
|
| 1448 |
-
"step": 22000
|
| 1449 |
-
},
|
| 1450 |
-
{
|
| 1451 |
-
"epoch": 9.13,
|
| 1452 |
-
"eval_bleu": 79.9787,
|
| 1453 |
-
"eval_em": 0.1,
|
| 1454 |
-
"eval_gen_len": 47.52,
|
| 1455 |
-
"eval_loss": 0.9404464960098267,
|
| 1456 |
-
"eval_rm": 0.2,
|
| 1457 |
-
"eval_runtime": 40.3319,
|
| 1458 |
-
"eval_samples_per_second": 1.24,
|
| 1459 |
-
"eval_steps_per_second": 0.174,
|
| 1460 |
-
"step": 22000
|
| 1461 |
-
},
|
| 1462 |
-
{
|
| 1463 |
-
"epoch": 9.17,
|
| 1464 |
-
"learning_rate": 4.6296296296296296e-06,
|
| 1465 |
-
"loss": 0.2477,
|
| 1466 |
-
"step": 22100
|
| 1467 |
-
},
|
| 1468 |
-
{
|
| 1469 |
-
"epoch": 9.21,
|
| 1470 |
-
"learning_rate": 4.398148148148149e-06,
|
| 1471 |
-
"loss": 0.2323,
|
| 1472 |
-
"step": 22200
|
| 1473 |
-
},
|
| 1474 |
-
{
|
| 1475 |
-
"epoch": 9.25,
|
| 1476 |
-
"learning_rate": 4.166666666666667e-06,
|
| 1477 |
-
"loss": 0.2489,
|
| 1478 |
-
"step": 22300
|
| 1479 |
-
},
|
| 1480 |
-
{
|
| 1481 |
-
"epoch": 9.29,
|
| 1482 |
-
"learning_rate": 3.935185185185185e-06,
|
| 1483 |
-
"loss": 0.265,
|
| 1484 |
-
"step": 22400
|
| 1485 |
-
},
|
| 1486 |
-
{
|
| 1487 |
-
"epoch": 9.34,
|
| 1488 |
-
"learning_rate": 3.7037037037037037e-06,
|
| 1489 |
-
"loss": 0.2298,
|
| 1490 |
-
"step": 22500
|
| 1491 |
-
},
|
| 1492 |
-
{
|
| 1493 |
-
"epoch": 9.38,
|
| 1494 |
-
"learning_rate": 3.4722222222222224e-06,
|
| 1495 |
-
"loss": 0.2694,
|
| 1496 |
-
"step": 22600
|
| 1497 |
-
},
|
| 1498 |
-
{
|
| 1499 |
-
"epoch": 9.42,
|
| 1500 |
-
"learning_rate": 3.2407407407407406e-06,
|
| 1501 |
-
"loss": 0.2439,
|
| 1502 |
-
"step": 22700
|
| 1503 |
-
},
|
| 1504 |
-
{
|
| 1505 |
-
"epoch": 9.46,
|
| 1506 |
-
"learning_rate": 3.0092592592592593e-06,
|
| 1507 |
-
"loss": 0.2415,
|
| 1508 |
-
"step": 22800
|
| 1509 |
-
},
|
| 1510 |
-
{
|
| 1511 |
-
"epoch": 9.5,
|
| 1512 |
-
"learning_rate": 2.777777777777778e-06,
|
| 1513 |
-
"loss": 0.2612,
|
| 1514 |
-
"step": 22900
|
| 1515 |
-
},
|
| 1516 |
-
{
|
| 1517 |
-
"epoch": 9.54,
|
| 1518 |
-
"learning_rate": 2.546296296296296e-06,
|
| 1519 |
-
"loss": 0.2541,
|
| 1520 |
-
"step": 23000
|
| 1521 |
-
},
|
| 1522 |
-
{
|
| 1523 |
-
"epoch": 9.59,
|
| 1524 |
-
"learning_rate": 2.3148148148148148e-06,
|
| 1525 |
-
"loss": 0.2387,
|
| 1526 |
-
"step": 23100
|
| 1527 |
-
},
|
| 1528 |
-
{
|
| 1529 |
-
"epoch": 9.63,
|
| 1530 |
-
"learning_rate": 2.0833333333333334e-06,
|
| 1531 |
-
"loss": 0.2461,
|
| 1532 |
-
"step": 23200
|
| 1533 |
-
},
|
| 1534 |
-
{
|
| 1535 |
-
"epoch": 9.67,
|
| 1536 |
-
"learning_rate": 1.8518518518518519e-06,
|
| 1537 |
-
"loss": 0.2495,
|
| 1538 |
-
"step": 23300
|
| 1539 |
-
},
|
| 1540 |
-
{
|
| 1541 |
-
"epoch": 9.71,
|
| 1542 |
-
"learning_rate": 1.6203703703703703e-06,
|
| 1543 |
-
"loss": 0.239,
|
| 1544 |
-
"step": 23400
|
| 1545 |
-
},
|
| 1546 |
-
{
|
| 1547 |
-
"epoch": 9.75,
|
| 1548 |
-
"learning_rate": 1.388888888888889e-06,
|
| 1549 |
-
"loss": 0.245,
|
| 1550 |
-
"step": 23500
|
| 1551 |
-
},
|
| 1552 |
-
{
|
| 1553 |
-
"epoch": 9.79,
|
| 1554 |
-
"learning_rate": 1.1574074074074074e-06,
|
| 1555 |
-
"loss": 0.225,
|
| 1556 |
-
"step": 23600
|
| 1557 |
-
},
|
| 1558 |
-
{
|
| 1559 |
-
"epoch": 9.83,
|
| 1560 |
-
"learning_rate": 9.259259259259259e-07,
|
| 1561 |
-
"loss": 0.2644,
|
| 1562 |
-
"step": 23700
|
| 1563 |
-
},
|
| 1564 |
-
{
|
| 1565 |
-
"epoch": 9.88,
|
| 1566 |
-
"learning_rate": 6.944444444444445e-07,
|
| 1567 |
-
"loss": 0.2264,
|
| 1568 |
-
"step": 23800
|
| 1569 |
-
},
|
| 1570 |
-
{
|
| 1571 |
-
"epoch": 9.92,
|
| 1572 |
-
"learning_rate": 4.6296296296296297e-07,
|
| 1573 |
-
"loss": 0.2532,
|
| 1574 |
-
"step": 23900
|
| 1575 |
-
},
|
| 1576 |
-
{
|
| 1577 |
-
"epoch": 9.96,
|
| 1578 |
-
"learning_rate": 2.3148148148148148e-07,
|
| 1579 |
-
"loss": 0.2387,
|
| 1580 |
-
"step": 24000
|
| 1581 |
-
},
|
| 1582 |
-
{
|
| 1583 |
-
"epoch": 9.96,
|
| 1584 |
-
"eval_bleu": 81.2301,
|
| 1585 |
-
"eval_em": 0.1,
|
| 1586 |
-
"eval_gen_len": 46.9,
|
| 1587 |
-
"eval_loss": 0.9337316155433655,
|
| 1588 |
-
"eval_rm": 0.22,
|
| 1589 |
-
"eval_runtime": 38.8763,
|
| 1590 |
-
"eval_samples_per_second": 1.286,
|
| 1591 |
-
"eval_steps_per_second": 0.18,
|
| 1592 |
-
"step": 24000
|
| 1593 |
-
},
|
| 1594 |
-
{
|
| 1595 |
-
"epoch": 10.0,
|
| 1596 |
-
"learning_rate": 0.0,
|
| 1597 |
-
"loss": 0.2259,
|
| 1598 |
-
"step": 24100
|
| 1599 |
},
|
| 1600 |
{
|
| 1601 |
-
"epoch":
|
| 1602 |
-
"step":
|
| 1603 |
-
"total_flos":
|
| 1604 |
-
"train_loss":
|
| 1605 |
-
"train_runtime":
|
| 1606 |
-
"train_samples_per_second":
|
| 1607 |
-
"train_steps_per_second":
|
| 1608 |
}
|
| 1609 |
],
|
| 1610 |
-
"max_steps":
|
| 1611 |
-
"num_train_epochs":
|
| 1612 |
-
"total_flos":
|
| 1613 |
"trial_name": null,
|
| 1614 |
"trial_params": null
|
| 1615 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 8.0,
|
| 5 |
+
"global_step": 11448,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
| 9 |
"log_history": [
|
| 10 |
{
|
| 11 |
+
"epoch": 0.07,
|
| 12 |
"learning_rate": 2.0000000000000003e-06,
|
| 13 |
+
"loss": 86.0421,
|
| 14 |
"step": 100
|
| 15 |
},
|
| 16 |
{
|
| 17 |
+
"epoch": 0.14,
|
| 18 |
"learning_rate": 4.000000000000001e-06,
|
| 19 |
+
"loss": 51.6706,
|
| 20 |
"step": 200
|
| 21 |
},
|
| 22 |
{
|
| 23 |
+
"epoch": 0.21,
|
| 24 |
"learning_rate": 6e-06,
|
| 25 |
+
"loss": 41.1349,
|
| 26 |
"step": 300
|
| 27 |
},
|
| 28 |
{
|
| 29 |
+
"epoch": 0.28,
|
| 30 |
"learning_rate": 8.000000000000001e-06,
|
| 31 |
+
"loss": 36.5061,
|
| 32 |
"step": 400
|
| 33 |
},
|
| 34 |
{
|
| 35 |
+
"epoch": 0.35,
|
| 36 |
"learning_rate": 1e-05,
|
| 37 |
+
"loss": 33.1858,
|
| 38 |
"step": 500
|
| 39 |
},
|
| 40 |
{
|
| 41 |
+
"epoch": 0.42,
|
| 42 |
"learning_rate": 1.2e-05,
|
| 43 |
+
"loss": 30.5206,
|
| 44 |
"step": 600
|
| 45 |
},
|
| 46 |
{
|
| 47 |
+
"epoch": 0.49,
|
| 48 |
"learning_rate": 1.4000000000000001e-05,
|
| 49 |
+
"loss": 28.0073,
|
| 50 |
"step": 700
|
| 51 |
},
|
| 52 |
{
|
| 53 |
+
"epoch": 0.56,
|
| 54 |
"learning_rate": 1.6000000000000003e-05,
|
| 55 |
+
"loss": 26.1939,
|
| 56 |
"step": 800
|
| 57 |
},
|
| 58 |
{
|
| 59 |
+
"epoch": 0.63,
|
| 60 |
"learning_rate": 1.8e-05,
|
| 61 |
+
"loss": 24.3465,
|
| 62 |
"step": 900
|
| 63 |
},
|
| 64 |
{
|
| 65 |
+
"epoch": 0.7,
|
| 66 |
"learning_rate": 2e-05,
|
| 67 |
+
"loss": 22.5759,
|
| 68 |
"step": 1000
|
| 69 |
},
|
| 70 |
{
|
| 71 |
+
"epoch": 0.77,
|
| 72 |
"learning_rate": 2.2000000000000003e-05,
|
| 73 |
+
"loss": 20.9294,
|
| 74 |
"step": 1100
|
| 75 |
},
|
| 76 |
{
|
| 77 |
+
"epoch": 0.84,
|
| 78 |
"learning_rate": 2.4e-05,
|
| 79 |
+
"loss": 19.3762,
|
| 80 |
"step": 1200
|
| 81 |
},
|
| 82 |
{
|
| 83 |
+
"epoch": 0.91,
|
| 84 |
"learning_rate": 2.6000000000000002e-05,
|
| 85 |
+
"loss": 17.72,
|
| 86 |
"step": 1300
|
| 87 |
},
|
| 88 |
{
|
| 89 |
+
"epoch": 0.98,
|
| 90 |
"learning_rate": 2.8000000000000003e-05,
|
| 91 |
+
"loss": 15.7901,
|
| 92 |
"step": 1400
|
| 93 |
},
|
| 94 |
{
|
| 95 |
+
"epoch": 1.05,
|
| 96 |
"learning_rate": 3e-05,
|
| 97 |
+
"loss": 14.0008,
|
| 98 |
"step": 1500
|
| 99 |
},
|
| 100 |
{
|
| 101 |
+
"epoch": 1.12,
|
| 102 |
"learning_rate": 3.2000000000000005e-05,
|
| 103 |
+
"loss": 12.3777,
|
| 104 |
"step": 1600
|
| 105 |
},
|
| 106 |
{
|
| 107 |
+
"epoch": 1.19,
|
| 108 |
"learning_rate": 3.4000000000000007e-05,
|
| 109 |
+
"loss": 10.7261,
|
| 110 |
"step": 1700
|
| 111 |
},
|
| 112 |
{
|
| 113 |
+
"epoch": 1.26,
|
| 114 |
"learning_rate": 3.6e-05,
|
| 115 |
+
"loss": 9.1024,
|
| 116 |
"step": 1800
|
| 117 |
},
|
| 118 |
{
|
| 119 |
+
"epoch": 1.33,
|
| 120 |
"learning_rate": 3.8e-05,
|
| 121 |
+
"loss": 7.4676,
|
| 122 |
"step": 1900
|
| 123 |
},
|
| 124 |
{
|
| 125 |
+
"epoch": 1.4,
|
| 126 |
"learning_rate": 4e-05,
|
| 127 |
+
"loss": 6.6044,
|
| 128 |
"step": 2000
|
| 129 |
},
|
| 130 |
{
|
| 131 |
+
"epoch": 1.4,
|
| 132 |
+
"eval_bleu": 3.8045,
|
| 133 |
"eval_em": 0.0,
|
| 134 |
+
"eval_gen_len": 158.8473,
|
| 135 |
+
"eval_loss": 6.171305179595947,
|
| 136 |
+
"eval_runtime": 1556.9786,
|
| 137 |
+
"eval_samples_per_second": 0.774,
|
| 138 |
+
"eval_steps_per_second": 0.097,
|
|
|
|
| 139 |
"step": 2000
|
| 140 |
},
|
| 141 |
{
|
| 142 |
+
"epoch": 1.47,
|
| 143 |
"learning_rate": 4.2e-05,
|
| 144 |
+
"loss": 6.0941,
|
| 145 |
"step": 2100
|
| 146 |
},
|
| 147 |
{
|
| 148 |
+
"epoch": 1.54,
|
| 149 |
"learning_rate": 4.4000000000000006e-05,
|
| 150 |
+
"loss": 5.6741,
|
| 151 |
"step": 2200
|
| 152 |
},
|
| 153 |
{
|
| 154 |
+
"epoch": 1.61,
|
| 155 |
"learning_rate": 4.600000000000001e-05,
|
| 156 |
+
"loss": 5.4757,
|
| 157 |
"step": 2300
|
| 158 |
},
|
| 159 |
{
|
| 160 |
+
"epoch": 1.68,
|
| 161 |
"learning_rate": 4.8e-05,
|
| 162 |
+
"loss": 5.242,
|
| 163 |
"step": 2400
|
| 164 |
},
|
| 165 |
{
|
| 166 |
+
"epoch": 1.75,
|
| 167 |
"learning_rate": 5e-05,
|
| 168 |
+
"loss": 5.0108,
|
| 169 |
"step": 2500
|
| 170 |
},
|
| 171 |
{
|
| 172 |
+
"epoch": 1.82,
|
| 173 |
+
"learning_rate": 4.944121591417077e-05,
|
| 174 |
+
"loss": 4.8595,
|
| 175 |
"step": 2600
|
| 176 |
},
|
| 177 |
{
|
| 178 |
+
"epoch": 1.89,
|
| 179 |
+
"learning_rate": 4.888243182834153e-05,
|
| 180 |
+
"loss": 4.695,
|
| 181 |
"step": 2700
|
| 182 |
},
|
| 183 |
{
|
| 184 |
+
"epoch": 1.96,
|
| 185 |
+
"learning_rate": 4.8323647742512295e-05,
|
| 186 |
+
"loss": 4.5706,
|
| 187 |
"step": 2800
|
| 188 |
},
|
| 189 |
{
|
| 190 |
+
"epoch": 2.03,
|
| 191 |
+
"learning_rate": 4.776486365668306e-05,
|
| 192 |
+
"loss": 4.2498,
|
| 193 |
"step": 2900
|
| 194 |
},
|
| 195 |
{
|
| 196 |
+
"epoch": 2.1,
|
| 197 |
+
"learning_rate": 4.720607957085382e-05,
|
| 198 |
+
"loss": 4.1223,
|
| 199 |
"step": 3000
|
| 200 |
},
|
| 201 |
{
|
| 202 |
+
"epoch": 2.17,
|
| 203 |
+
"learning_rate": 4.664729548502459e-05,
|
| 204 |
+
"loss": 4.0181,
|
| 205 |
"step": 3100
|
| 206 |
},
|
| 207 |
{
|
| 208 |
+
"epoch": 2.24,
|
| 209 |
+
"learning_rate": 4.6088511399195353e-05,
|
| 210 |
+
"loss": 3.8722,
|
| 211 |
"step": 3200
|
| 212 |
},
|
| 213 |
{
|
| 214 |
+
"epoch": 2.31,
|
| 215 |
+
"learning_rate": 4.552972731336611e-05,
|
| 216 |
+
"loss": 3.7786,
|
| 217 |
"step": 3300
|
| 218 |
},
|
| 219 |
{
|
| 220 |
+
"epoch": 2.38,
|
| 221 |
+
"learning_rate": 4.497094322753688e-05,
|
| 222 |
+
"loss": 3.6403,
|
| 223 |
"step": 3400
|
| 224 |
},
|
| 225 |
{
|
| 226 |
+
"epoch": 2.45,
|
| 227 |
+
"learning_rate": 4.4412159141707646e-05,
|
| 228 |
+
"loss": 3.5437,
|
| 229 |
"step": 3500
|
| 230 |
},
|
| 231 |
{
|
| 232 |
+
"epoch": 2.52,
|
| 233 |
+
"learning_rate": 4.385337505587841e-05,
|
| 234 |
+
"loss": 3.389,
|
| 235 |
"step": 3600
|
| 236 |
},
|
| 237 |
{
|
| 238 |
+
"epoch": 2.59,
|
| 239 |
+
"learning_rate": 4.329459097004918e-05,
|
| 240 |
+
"loss": 3.2395,
|
| 241 |
"step": 3700
|
| 242 |
},
|
| 243 |
{
|
| 244 |
+
"epoch": 2.66,
|
| 245 |
+
"learning_rate": 4.2735806884219945e-05,
|
| 246 |
+
"loss": 3.1786,
|
| 247 |
"step": 3800
|
| 248 |
},
|
| 249 |
{
|
| 250 |
+
"epoch": 2.73,
|
| 251 |
+
"learning_rate": 4.2177022798390704e-05,
|
| 252 |
+
"loss": 3.0657,
|
| 253 |
"step": 3900
|
| 254 |
},
|
| 255 |
{
|
| 256 |
+
"epoch": 2.8,
|
| 257 |
+
"learning_rate": 4.161823871256147e-05,
|
| 258 |
+
"loss": 3.032,
|
| 259 |
"step": 4000
|
| 260 |
},
|
| 261 |
{
|
| 262 |
+
"epoch": 2.8,
|
| 263 |
+
"eval_bleu": 27.701,
|
| 264 |
"eval_em": 0.0,
|
| 265 |
+
"eval_gen_len": 33.9568,
|
| 266 |
+
"eval_loss": 2.904534339904785,
|
| 267 |
+
"eval_runtime": 297.5953,
|
| 268 |
+
"eval_samples_per_second": 4.049,
|
| 269 |
+
"eval_steps_per_second": 0.507,
|
|
|
|
| 270 |
"step": 4000
|
| 271 |
},
|
| 272 |
{
|
| 273 |
+
"epoch": 2.87,
|
| 274 |
+
"learning_rate": 4.105945462673223e-05,
|
| 275 |
+
"loss": 2.8755,
|
| 276 |
"step": 4100
|
| 277 |
},
|
| 278 |
{
|
| 279 |
+
"epoch": 2.94,
|
| 280 |
+
"learning_rate": 4.0500670540903e-05,
|
| 281 |
+
"loss": 2.8396,
|
| 282 |
"step": 4200
|
| 283 |
},
|
| 284 |
{
|
| 285 |
+
"epoch": 3.0,
|
| 286 |
+
"learning_rate": 3.994188645507376e-05,
|
| 287 |
+
"loss": 2.7454,
|
| 288 |
"step": 4300
|
| 289 |
},
|
| 290 |
{
|
| 291 |
+
"epoch": 3.07,
|
| 292 |
+
"learning_rate": 3.938310236924452e-05,
|
| 293 |
+
"loss": 2.5218,
|
| 294 |
"step": 4400
|
| 295 |
},
|
| 296 |
{
|
| 297 |
+
"epoch": 3.14,
|
| 298 |
+
"learning_rate": 3.882431828341529e-05,
|
| 299 |
+
"loss": 2.4895,
|
| 300 |
"step": 4500
|
| 301 |
},
|
| 302 |
{
|
| 303 |
+
"epoch": 3.21,
|
| 304 |
+
"learning_rate": 3.8265534197586055e-05,
|
| 305 |
+
"loss": 2.4554,
|
| 306 |
"step": 4600
|
| 307 |
},
|
| 308 |
{
|
| 309 |
+
"epoch": 3.28,
|
| 310 |
+
"learning_rate": 3.7706750111756815e-05,
|
| 311 |
+
"loss": 2.3573,
|
| 312 |
"step": 4700
|
| 313 |
},
|
| 314 |
{
|
| 315 |
+
"epoch": 3.35,
|
| 316 |
+
"learning_rate": 3.714796602592758e-05,
|
| 317 |
+
"loss": 2.2979,
|
| 318 |
"step": 4800
|
| 319 |
},
|
| 320 |
{
|
| 321 |
+
"epoch": 3.42,
|
| 322 |
+
"learning_rate": 3.658918194009835e-05,
|
| 323 |
+
"loss": 2.1874,
|
| 324 |
"step": 4900
|
| 325 |
},
|
| 326 |
{
|
| 327 |
+
"epoch": 3.49,
|
| 328 |
+
"learning_rate": 3.603039785426911e-05,
|
| 329 |
+
"loss": 2.1803,
|
| 330 |
"step": 5000
|
| 331 |
},
|
| 332 |
{
|
| 333 |
+
"epoch": 3.56,
|
| 334 |
+
"learning_rate": 3.5471613768439874e-05,
|
| 335 |
+
"loss": 2.1553,
|
| 336 |
"step": 5100
|
| 337 |
},
|
| 338 |
{
|
| 339 |
+
"epoch": 3.63,
|
| 340 |
+
"learning_rate": 3.491282968261064e-05,
|
| 341 |
+
"loss": 2.0567,
|
| 342 |
"step": 5200
|
| 343 |
},
|
| 344 |
{
|
| 345 |
+
"epoch": 3.7,
|
| 346 |
+
"learning_rate": 3.4354045596781406e-05,
|
| 347 |
+
"loss": 2.0147,
|
| 348 |
"step": 5300
|
| 349 |
},
|
| 350 |
{
|
| 351 |
+
"epoch": 3.77,
|
| 352 |
+
"learning_rate": 3.379526151095217e-05,
|
| 353 |
+
"loss": 1.9817,
|
| 354 |
"step": 5400
|
| 355 |
},
|
| 356 |
{
|
| 357 |
+
"epoch": 3.84,
|
| 358 |
+
"learning_rate": 3.323647742512294e-05,
|
| 359 |
+
"loss": 1.8843,
|
| 360 |
"step": 5500
|
| 361 |
},
|
| 362 |
{
|
| 363 |
+
"epoch": 3.91,
|
| 364 |
+
"learning_rate": 3.26776933392937e-05,
|
| 365 |
+
"loss": 1.8849,
|
| 366 |
"step": 5600
|
| 367 |
},
|
| 368 |
{
|
| 369 |
+
"epoch": 3.98,
|
| 370 |
+
"learning_rate": 3.2118909253464465e-05,
|
| 371 |
+
"loss": 1.8232,
|
| 372 |
"step": 5700
|
| 373 |
},
|
| 374 |
{
|
| 375 |
+
"epoch": 4.05,
|
| 376 |
+
"learning_rate": 3.156012516763523e-05,
|
| 377 |
+
"loss": 1.6714,
|
| 378 |
"step": 5800
|
| 379 |
},
|
| 380 |
{
|
| 381 |
+
"epoch": 4.12,
|
| 382 |
+
"learning_rate": 3.100134108180599e-05,
|
| 383 |
+
"loss": 1.6047,
|
| 384 |
"step": 5900
|
| 385 |
},
|
| 386 |
{
|
| 387 |
+
"epoch": 4.19,
|
| 388 |
+
"learning_rate": 3.0442556995976757e-05,
|
| 389 |
+
"loss": 1.5893,
|
| 390 |
"step": 6000
|
| 391 |
},
|
| 392 |
{
|
| 393 |
+
"epoch": 4.19,
|
| 394 |
+
"eval_bleu": 57.7509,
|
| 395 |
+
"eval_em": 0.0033,
|
| 396 |
+
"eval_gen_len": 44.132,
|
| 397 |
+
"eval_loss": 1.7326730489730835,
|
| 398 |
+
"eval_runtime": 426.2734,
|
| 399 |
+
"eval_samples_per_second": 2.827,
|
| 400 |
+
"eval_steps_per_second": 0.354,
|
|
|
|
| 401 |
"step": 6000
|
| 402 |
},
|
| 403 |
{
|
| 404 |
+
"epoch": 4.26,
|
| 405 |
+
"learning_rate": 2.9883772910147524e-05,
|
| 406 |
+
"loss": 1.5864,
|
| 407 |
"step": 6100
|
| 408 |
},
|
| 409 |
{
|
| 410 |
+
"epoch": 4.33,
|
| 411 |
+
"learning_rate": 2.9324988824318283e-05,
|
| 412 |
+
"loss": 1.5608,
|
| 413 |
"step": 6200
|
| 414 |
},
|
| 415 |
{
|
| 416 |
+
"epoch": 4.4,
|
| 417 |
+
"learning_rate": 2.876620473848905e-05,
|
| 418 |
+
"loss": 1.5144,
|
| 419 |
"step": 6300
|
| 420 |
},
|
| 421 |
{
|
| 422 |
+
"epoch": 4.47,
|
| 423 |
+
"learning_rate": 2.8207420652659816e-05,
|
| 424 |
+
"loss": 1.4582,
|
| 425 |
"step": 6400
|
| 426 |
},
|
| 427 |
{
|
| 428 |
+
"epoch": 4.54,
|
| 429 |
+
"learning_rate": 2.7648636566830576e-05,
|
| 430 |
+
"loss": 1.4793,
|
| 431 |
"step": 6500
|
| 432 |
},
|
| 433 |
{
|
| 434 |
+
"epoch": 4.61,
|
| 435 |
+
"learning_rate": 2.7089852481001342e-05,
|
| 436 |
+
"loss": 1.472,
|
| 437 |
"step": 6600
|
| 438 |
},
|
| 439 |
{
|
| 440 |
+
"epoch": 4.68,
|
| 441 |
+
"learning_rate": 2.653106839517211e-05,
|
| 442 |
+
"loss": 1.4424,
|
| 443 |
"step": 6700
|
| 444 |
},
|
| 445 |
{
|
| 446 |
+
"epoch": 4.75,
|
| 447 |
+
"learning_rate": 2.597228430934287e-05,
|
| 448 |
+
"loss": 1.3779,
|
| 449 |
"step": 6800
|
| 450 |
},
|
| 451 |
{
|
| 452 |
+
"epoch": 4.82,
|
| 453 |
+
"learning_rate": 2.5413500223513638e-05,
|
| 454 |
+
"loss": 1.3611,
|
| 455 |
"step": 6900
|
| 456 |
},
|
| 457 |
{
|
| 458 |
+
"epoch": 4.89,
|
| 459 |
+
"learning_rate": 2.48547161376844e-05,
|
| 460 |
+
"loss": 1.3311,
|
| 461 |
"step": 7000
|
| 462 |
},
|
| 463 |
{
|
| 464 |
+
"epoch": 4.96,
|
| 465 |
+
"learning_rate": 2.4295932051855164e-05,
|
| 466 |
+
"loss": 1.3164,
|
| 467 |
"step": 7100
|
| 468 |
},
|
| 469 |
{
|
| 470 |
+
"epoch": 5.03,
|
| 471 |
+
"learning_rate": 2.373714796602593e-05,
|
| 472 |
+
"loss": 1.2119,
|
| 473 |
"step": 7200
|
| 474 |
},
|
| 475 |
{
|
| 476 |
+
"epoch": 5.1,
|
| 477 |
+
"learning_rate": 2.3178363880196693e-05,
|
| 478 |
+
"loss": 1.1122,
|
| 479 |
"step": 7300
|
| 480 |
},
|
| 481 |
{
|
| 482 |
+
"epoch": 5.17,
|
| 483 |
+
"learning_rate": 2.2619579794367456e-05,
|
| 484 |
+
"loss": 1.1198,
|
| 485 |
"step": 7400
|
| 486 |
},
|
| 487 |
{
|
| 488 |
+
"epoch": 5.24,
|
| 489 |
+
"learning_rate": 2.206079570853822e-05,
|
| 490 |
+
"loss": 1.0416,
|
| 491 |
"step": 7500
|
| 492 |
},
|
| 493 |
{
|
| 494 |
+
"epoch": 5.31,
|
| 495 |
+
"learning_rate": 2.1502011622708985e-05,
|
| 496 |
+
"loss": 1.1042,
|
| 497 |
"step": 7600
|
| 498 |
},
|
| 499 |
{
|
| 500 |
+
"epoch": 5.38,
|
| 501 |
+
"learning_rate": 2.0943227536879752e-05,
|
| 502 |
+
"loss": 1.0715,
|
| 503 |
"step": 7700
|
| 504 |
},
|
| 505 |
{
|
| 506 |
+
"epoch": 5.45,
|
| 507 |
+
"learning_rate": 2.0384443451050515e-05,
|
| 508 |
+
"loss": 1.0815,
|
| 509 |
"step": 7800
|
| 510 |
},
|
| 511 |
{
|
| 512 |
+
"epoch": 5.52,
|
| 513 |
+
"learning_rate": 1.982565936522128e-05,
|
| 514 |
+
"loss": 1.0445,
|
| 515 |
"step": 7900
|
| 516 |
},
|
| 517 |
{
|
| 518 |
+
"epoch": 5.59,
|
| 519 |
+
"learning_rate": 1.9266875279392044e-05,
|
| 520 |
+
"loss": 1.0512,
|
| 521 |
"step": 8000
|
| 522 |
},
|
| 523 |
{
|
| 524 |
+
"epoch": 5.59,
|
| 525 |
+
"eval_bleu": 71.1272,
|
| 526 |
+
"eval_em": 0.0241,
|
| 527 |
+
"eval_gen_len": 46.0672,
|
| 528 |
+
"eval_loss": 1.2382431030273438,
|
| 529 |
+
"eval_runtime": 358.1089,
|
| 530 |
+
"eval_samples_per_second": 3.365,
|
| 531 |
+
"eval_steps_per_second": 0.422,
|
|
|
|
| 532 |
"step": 8000
|
| 533 |
},
|
| 534 |
{
|
| 535 |
+
"epoch": 5.66,
|
| 536 |
+
"learning_rate": 1.8708091193562807e-05,
|
| 537 |
+
"loss": 1.0201,
|
| 538 |
"step": 8100
|
| 539 |
},
|
| 540 |
{
|
| 541 |
+
"epoch": 5.73,
|
| 542 |
+
"learning_rate": 1.8149307107733573e-05,
|
| 543 |
+
"loss": 1.0377,
|
| 544 |
"step": 8200
|
| 545 |
},
|
| 546 |
{
|
| 547 |
+
"epoch": 5.8,
|
| 548 |
+
"learning_rate": 1.7590523021904336e-05,
|
| 549 |
+
"loss": 0.986,
|
| 550 |
"step": 8300
|
| 551 |
},
|
| 552 |
{
|
| 553 |
+
"epoch": 5.87,
|
| 554 |
+
"learning_rate": 1.70317389360751e-05,
|
| 555 |
+
"loss": 1.0244,
|
| 556 |
"step": 8400
|
| 557 |
},
|
| 558 |
{
|
| 559 |
+
"epoch": 5.94,
|
| 560 |
+
"learning_rate": 1.6472954850245866e-05,
|
| 561 |
+
"loss": 0.9654,
|
| 562 |
"step": 8500
|
| 563 |
},
|
| 564 |
{
|
| 565 |
+
"epoch": 6.01,
|
| 566 |
+
"learning_rate": 1.5914170764416632e-05,
|
| 567 |
+
"loss": 0.9454,
|
| 568 |
"step": 8600
|
| 569 |
},
|
| 570 |
{
|
| 571 |
+
"epoch": 6.08,
|
| 572 |
+
"learning_rate": 1.5355386678587395e-05,
|
| 573 |
+
"loss": 0.8179,
|
| 574 |
"step": 8700
|
| 575 |
},
|
| 576 |
{
|
| 577 |
+
"epoch": 6.15,
|
| 578 |
+
"learning_rate": 1.479660259275816e-05,
|
| 579 |
+
"loss": 0.8433,
|
| 580 |
"step": 8800
|
| 581 |
},
|
| 582 |
{
|
| 583 |
+
"epoch": 6.22,
|
| 584 |
+
"learning_rate": 1.4237818506928924e-05,
|
| 585 |
+
"loss": 0.8235,
|
| 586 |
"step": 8900
|
| 587 |
},
|
| 588 |
{
|
| 589 |
+
"epoch": 6.29,
|
| 590 |
+
"learning_rate": 1.3679034421099687e-05,
|
| 591 |
+
"loss": 0.832,
|
| 592 |
"step": 9000
|
| 593 |
},
|
| 594 |
{
|
| 595 |
+
"epoch": 6.36,
|
| 596 |
+
"learning_rate": 1.312025033527045e-05,
|
| 597 |
+
"loss": 0.8019,
|
| 598 |
"step": 9100
|
| 599 |
},
|
| 600 |
{
|
| 601 |
+
"epoch": 6.43,
|
| 602 |
+
"learning_rate": 1.2561466249441217e-05,
|
| 603 |
+
"loss": 0.806,
|
| 604 |
"step": 9200
|
| 605 |
},
|
| 606 |
{
|
| 607 |
+
"epoch": 6.5,
|
| 608 |
+
"learning_rate": 1.2002682163611981e-05,
|
| 609 |
+
"loss": 0.7985,
|
| 610 |
"step": 9300
|
| 611 |
},
|
| 612 |
{
|
| 613 |
+
"epoch": 6.57,
|
| 614 |
+
"learning_rate": 1.1443898077782746e-05,
|
| 615 |
+
"loss": 0.801,
|
| 616 |
"step": 9400
|
| 617 |
},
|
| 618 |
{
|
| 619 |
+
"epoch": 6.64,
|
| 620 |
+
"learning_rate": 1.0885113991953509e-05,
|
| 621 |
+
"loss": 0.7758,
|
| 622 |
"step": 9500
|
| 623 |
},
|
| 624 |
{
|
| 625 |
+
"epoch": 6.71,
|
| 626 |
+
"learning_rate": 1.0326329906124274e-05,
|
| 627 |
+
"loss": 0.804,
|
| 628 |
"step": 9600
|
| 629 |
},
|
| 630 |
{
|
| 631 |
+
"epoch": 6.78,
|
| 632 |
+
"learning_rate": 9.767545820295038e-06,
|
| 633 |
+
"loss": 0.7846,
|
| 634 |
"step": 9700
|
| 635 |
},
|
| 636 |
{
|
| 637 |
+
"epoch": 6.85,
|
| 638 |
+
"learning_rate": 9.208761734465803e-06,
|
| 639 |
+
"loss": 0.7736,
|
| 640 |
"step": 9800
|
| 641 |
},
|
| 642 |
{
|
| 643 |
+
"epoch": 6.92,
|
| 644 |
+
"learning_rate": 8.649977648636568e-06,
|
| 645 |
+
"loss": 0.7625,
|
| 646 |
"step": 9900
|
| 647 |
},
|
| 648 |
{
|
| 649 |
+
"epoch": 6.99,
|
| 650 |
+
"learning_rate": 8.09119356280733e-06,
|
| 651 |
+
"loss": 0.7515,
|
| 652 |
"step": 10000
|
| 653 |
},
|
| 654 |
{
|
| 655 |
+
"epoch": 6.99,
|
| 656 |
+
"eval_bleu": 75.7356,
|
| 657 |
+
"eval_em": 0.0432,
|
| 658 |
+
"eval_gen_len": 47.2896,
|
| 659 |
+
"eval_loss": 1.057088017463684,
|
| 660 |
+
"eval_runtime": 378.6301,
|
| 661 |
+
"eval_samples_per_second": 3.183,
|
| 662 |
+
"eval_steps_per_second": 0.399,
|
|
|
|
| 663 |
"step": 10000
|
| 664 |
},
|
| 665 |
{
|
| 666 |
+
"epoch": 7.06,
|
| 667 |
+
"learning_rate": 7.532409476978096e-06,
|
| 668 |
+
"loss": 0.6703,
|
| 669 |
"step": 10100
|
| 670 |
},
|
| 671 |
{
|
| 672 |
+
"epoch": 7.13,
|
| 673 |
+
"learning_rate": 6.973625391148861e-06,
|
| 674 |
+
"loss": 0.6731,
|
| 675 |
"step": 10200
|
| 676 |
},
|
| 677 |
{
|
| 678 |
+
"epoch": 7.2,
|
| 679 |
+
"learning_rate": 6.414841305319625e-06,
|
| 680 |
+
"loss": 0.6584,
|
| 681 |
"step": 10300
|
| 682 |
},
|
| 683 |
{
|
| 684 |
+
"epoch": 7.27,
|
| 685 |
+
"learning_rate": 5.856057219490389e-06,
|
| 686 |
+
"loss": 0.6758,
|
| 687 |
"step": 10400
|
| 688 |
},
|
| 689 |
{
|
| 690 |
+
"epoch": 7.34,
|
| 691 |
+
"learning_rate": 5.297273133661153e-06,
|
| 692 |
+
"loss": 0.6801,
|
| 693 |
"step": 10500
|
| 694 |
},
|
| 695 |
{
|
| 696 |
+
"epoch": 7.41,
|
| 697 |
+
"learning_rate": 4.738489047831918e-06,
|
| 698 |
+
"loss": 0.6556,
|
| 699 |
"step": 10600
|
| 700 |
},
|
| 701 |
{
|
| 702 |
+
"epoch": 7.48,
|
| 703 |
+
"learning_rate": 4.1797049620026825e-06,
|
| 704 |
+
"loss": 0.6413,
|
| 705 |
"step": 10700
|
| 706 |
},
|
| 707 |
{
|
| 708 |
+
"epoch": 7.55,
|
| 709 |
+
"learning_rate": 3.6209208761734468e-06,
|
| 710 |
+
"loss": 0.6548,
|
| 711 |
"step": 10800
|
| 712 |
},
|
| 713 |
{
|
| 714 |
+
"epoch": 7.62,
|
| 715 |
+
"learning_rate": 3.062136790344211e-06,
|
| 716 |
+
"loss": 0.6631,
|
| 717 |
"step": 10900
|
| 718 |
},
|
| 719 |
{
|
| 720 |
+
"epoch": 7.69,
|
| 721 |
+
"learning_rate": 2.5033527045149757e-06,
|
| 722 |
+
"loss": 0.6458,
|
| 723 |
"step": 11000
|
| 724 |
},
|
| 725 |
{
|
| 726 |
+
"epoch": 7.76,
|
| 727 |
+
"learning_rate": 1.94456861868574e-06,
|
| 728 |
+
"loss": 0.629,
|
| 729 |
"step": 11100
|
| 730 |
},
|
| 731 |
{
|
| 732 |
+
"epoch": 7.83,
|
| 733 |
+
"learning_rate": 1.3857845328565042e-06,
|
| 734 |
+
"loss": 0.643,
|
| 735 |
"step": 11200
|
| 736 |
},
|
| 737 |
{
|
| 738 |
+
"epoch": 7.9,
|
| 739 |
+
"learning_rate": 8.270004470272687e-07,
|
| 740 |
+
"loss": 0.6475,
|
| 741 |
"step": 11300
|
| 742 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 743 |
{
|
| 744 |
"epoch": 7.97,
|
| 745 |
+
"learning_rate": 2.682163611980331e-07,
|
| 746 |
+
"loss": 0.633,
|
| 747 |
+
"step": 11400
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 748 |
},
|
| 749 |
{
|
| 750 |
+
"epoch": 8.0,
|
| 751 |
+
"step": 11448,
|
| 752 |
+
"total_flos": 7528840539566280.0,
|
| 753 |
+
"train_loss": 0.15130291744847968,
|
| 754 |
+
"train_runtime": 1637.6947,
|
| 755 |
+
"train_samples_per_second": 111.816,
|
| 756 |
+
"train_steps_per_second": 6.99
|
| 757 |
}
|
| 758 |
],
|
| 759 |
+
"max_steps": 11448,
|
| 760 |
+
"num_train_epochs": 8,
|
| 761 |
+
"total_flos": 7528840539566280.0,
|
| 762 |
"trial_name": null,
|
| 763 |
"trial_params": null
|
| 764 |
}
|