Commit
·
632b0a1
1
Parent(s):
38a1e05
End of training
Browse files- README.md +3 -0
- all_results.json +5 -5
- train_results.json +5 -5
- trainer_state.json +421 -1081
README.md
CHANGED
|
@@ -1,4 +1,7 @@
|
|
| 1 |
---
|
|
|
|
|
|
|
|
|
|
| 2 |
license: mit
|
| 3 |
base_model: facebook/mbart-large-50
|
| 4 |
tags:
|
|
|
|
| 1 |
---
|
| 2 |
+
language:
|
| 3 |
+
- zh
|
| 4 |
+
- yue
|
| 5 |
license: mit
|
| 6 |
base_model: facebook/mbart-large-50
|
| 7 |
tags:
|
all_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 5.0,
|
| 3 |
-
"train_loss": 0.
|
| 4 |
-
"train_runtime":
|
| 5 |
-
"train_samples":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 5.0,
|
| 3 |
+
"train_loss": 0.003351212132286996,
|
| 4 |
+
"train_runtime": 2726.0601,
|
| 5 |
+
"train_samples": 220940,
|
| 6 |
+
"train_samples_per_second": 405.237,
|
| 7 |
+
"train_steps_per_second": 25.328
|
| 8 |
}
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 5.0,
|
| 3 |
-
"train_loss": 0.
|
| 4 |
-
"train_runtime":
|
| 5 |
-
"train_samples":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 5.0,
|
| 3 |
+
"train_loss": 0.003351212132286996,
|
| 4 |
+
"train_runtime": 2726.0601,
|
| 5 |
+
"train_samples": 220940,
|
| 6 |
+
"train_samples_per_second": 405.237,
|
| 7 |
+
"train_steps_per_second": 25.328
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -3,1514 +3,854 @@
|
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
"epoch": 5.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.02,
|
| 13 |
-
"learning_rate": 4.
|
| 14 |
-
"loss": 1.
|
| 15 |
"step": 500
|
| 16 |
},
|
| 17 |
{
|
| 18 |
"epoch": 0.04,
|
| 19 |
-
"learning_rate": 4.
|
| 20 |
-
"loss": 0.
|
| 21 |
"step": 1000
|
| 22 |
},
|
| 23 |
{
|
| 24 |
-
"epoch": 0.
|
| 25 |
-
"learning_rate": 4.
|
| 26 |
-
"loss": 0.
|
| 27 |
"step": 1500
|
| 28 |
},
|
| 29 |
{
|
| 30 |
-
"epoch": 0.
|
| 31 |
-
"learning_rate": 4.
|
| 32 |
-
"loss": 0.
|
| 33 |
"step": 2000
|
| 34 |
},
|
| 35 |
{
|
| 36 |
-
"epoch": 0.
|
| 37 |
-
"learning_rate": 4.
|
| 38 |
-
"loss": 0.
|
| 39 |
"step": 2500
|
| 40 |
},
|
| 41 |
{
|
| 42 |
-
"epoch": 0.
|
| 43 |
-
"learning_rate": 4.
|
| 44 |
-
"loss": 0.
|
| 45 |
"step": 3000
|
| 46 |
},
|
| 47 |
{
|
| 48 |
-
"epoch": 0.
|
| 49 |
-
"learning_rate": 4.
|
| 50 |
-
"loss": 0.
|
| 51 |
"step": 3500
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch": 0.
|
| 55 |
-
"learning_rate": 4.
|
| 56 |
-
"loss": 0.
|
| 57 |
"step": 4000
|
| 58 |
},
|
| 59 |
{
|
| 60 |
-
"epoch": 0.
|
| 61 |
-
"learning_rate": 4.
|
| 62 |
-
"loss": 0.
|
| 63 |
"step": 4500
|
| 64 |
},
|
| 65 |
{
|
| 66 |
-
"epoch": 0.
|
| 67 |
-
"learning_rate": 4.
|
| 68 |
-
"loss": 0.
|
| 69 |
"step": 5000
|
| 70 |
},
|
| 71 |
{
|
| 72 |
-
"epoch": 0.
|
| 73 |
-
"learning_rate": 4.
|
| 74 |
-
"loss": 0.
|
| 75 |
"step": 5500
|
| 76 |
},
|
| 77 |
{
|
| 78 |
-
"epoch": 0.
|
| 79 |
-
"learning_rate": 4.
|
| 80 |
-
"loss": 0.
|
| 81 |
"step": 6000
|
| 82 |
},
|
| 83 |
{
|
| 84 |
-
"epoch": 0.
|
| 85 |
-
"learning_rate": 4.
|
| 86 |
-
"loss": 0.
|
| 87 |
"step": 6500
|
| 88 |
},
|
| 89 |
{
|
| 90 |
-
"epoch": 0.
|
| 91 |
-
"learning_rate": 4.
|
| 92 |
-
"loss": 0.
|
| 93 |
"step": 7000
|
| 94 |
},
|
| 95 |
{
|
| 96 |
-
"epoch": 0.
|
| 97 |
-
"learning_rate": 4.
|
| 98 |
-
"loss": 0.
|
| 99 |
"step": 7500
|
| 100 |
},
|
| 101 |
{
|
| 102 |
-
"epoch": 0.
|
| 103 |
-
"learning_rate": 4.
|
| 104 |
-
"loss": 0.
|
| 105 |
"step": 8000
|
| 106 |
},
|
| 107 |
{
|
| 108 |
-
"epoch": 0.
|
| 109 |
-
"learning_rate": 4.
|
| 110 |
-
"loss": 0.
|
| 111 |
"step": 8500
|
| 112 |
},
|
| 113 |
{
|
| 114 |
-
"epoch": 0.
|
| 115 |
-
"learning_rate": 4.
|
| 116 |
-
"loss": 0.
|
| 117 |
"step": 9000
|
| 118 |
},
|
| 119 |
{
|
| 120 |
-
"epoch": 0.
|
| 121 |
-
"learning_rate": 4.
|
| 122 |
-
"loss": 0.
|
| 123 |
"step": 9500
|
| 124 |
},
|
| 125 |
{
|
| 126 |
-
"epoch": 0.
|
| 127 |
-
"learning_rate": 4.
|
| 128 |
-
"loss": 0.
|
| 129 |
"step": 10000
|
| 130 |
},
|
| 131 |
{
|
| 132 |
-
"epoch": 0.
|
| 133 |
-
"learning_rate": 4.
|
| 134 |
-
"loss": 0.
|
| 135 |
"step": 10500
|
| 136 |
},
|
| 137 |
{
|
| 138 |
-
"epoch": 0.
|
| 139 |
-
"learning_rate": 4.
|
| 140 |
-
"loss": 0.
|
| 141 |
"step": 11000
|
| 142 |
},
|
| 143 |
{
|
| 144 |
-
"epoch": 0.
|
| 145 |
-
"learning_rate": 4.
|
| 146 |
-
"loss": 0.
|
| 147 |
"step": 11500
|
| 148 |
},
|
| 149 |
{
|
| 150 |
-
"epoch": 0.
|
| 151 |
-
"learning_rate": 4.
|
| 152 |
-
"loss": 0.
|
| 153 |
"step": 12000
|
| 154 |
},
|
| 155 |
{
|
| 156 |
-
"epoch": 0.
|
| 157 |
-
"learning_rate": 4.
|
| 158 |
-
"loss": 0.
|
| 159 |
"step": 12500
|
| 160 |
},
|
| 161 |
{
|
| 162 |
-
"epoch": 0.
|
| 163 |
-
"learning_rate": 4.
|
| 164 |
-
"loss": 0.
|
| 165 |
"step": 13000
|
| 166 |
},
|
| 167 |
{
|
| 168 |
-
"epoch": 0.
|
| 169 |
-
"learning_rate": 4.
|
| 170 |
-
"loss": 0.
|
| 171 |
"step": 13500
|
| 172 |
},
|
| 173 |
{
|
| 174 |
-
"epoch": 0.
|
| 175 |
-
"learning_rate": 4.
|
| 176 |
-
"loss": 0.
|
| 177 |
"step": 14000
|
| 178 |
},
|
| 179 |
{
|
| 180 |
-
"epoch": 0.
|
| 181 |
-
"learning_rate": 4.
|
| 182 |
-
"loss": 0.
|
| 183 |
"step": 14500
|
| 184 |
},
|
| 185 |
{
|
| 186 |
-
"epoch": 0.
|
| 187 |
-
"learning_rate": 4.
|
| 188 |
-
"loss": 0.
|
| 189 |
"step": 15000
|
| 190 |
},
|
| 191 |
{
|
| 192 |
-
"epoch": 0.
|
| 193 |
-
"learning_rate": 4.
|
| 194 |
-
"loss": 0.
|
| 195 |
"step": 15500
|
| 196 |
},
|
| 197 |
{
|
| 198 |
-
"epoch": 0.
|
| 199 |
-
"learning_rate": 4.
|
| 200 |
-
"loss": 0.
|
| 201 |
"step": 16000
|
| 202 |
},
|
| 203 |
{
|
| 204 |
-
"epoch": 0.
|
| 205 |
-
"learning_rate": 4.
|
| 206 |
-
"loss": 0.
|
| 207 |
"step": 16500
|
| 208 |
},
|
| 209 |
{
|
| 210 |
-
"epoch": 0.
|
| 211 |
-
"learning_rate": 4.
|
| 212 |
-
"loss": 0.
|
| 213 |
"step": 17000
|
| 214 |
},
|
| 215 |
{
|
| 216 |
-
"epoch": 0.
|
| 217 |
-
"learning_rate": 4.
|
| 218 |
-
"loss": 0.
|
| 219 |
"step": 17500
|
| 220 |
},
|
| 221 |
{
|
| 222 |
-
"epoch": 0.
|
| 223 |
-
"learning_rate": 4.
|
| 224 |
-
"loss": 0.
|
| 225 |
"step": 18000
|
| 226 |
},
|
| 227 |
{
|
| 228 |
-
"epoch": 0.
|
| 229 |
-
"learning_rate": 4.
|
| 230 |
-
"loss": 0.
|
| 231 |
"step": 18500
|
| 232 |
},
|
| 233 |
{
|
| 234 |
-
"epoch": 0.
|
| 235 |
-
"learning_rate": 4.
|
| 236 |
-
"loss": 0.
|
| 237 |
"step": 19000
|
| 238 |
},
|
| 239 |
{
|
| 240 |
-
"epoch": 0.
|
| 241 |
-
"learning_rate": 4.
|
| 242 |
-
"loss": 0.
|
| 243 |
"step": 19500
|
| 244 |
},
|
| 245 |
{
|
| 246 |
-
"epoch": 0.
|
| 247 |
-
"learning_rate": 4.
|
| 248 |
-
"loss": 0.
|
| 249 |
"step": 20000
|
| 250 |
},
|
| 251 |
{
|
| 252 |
-
"epoch": 0.
|
| 253 |
-
"learning_rate": 4.
|
| 254 |
-
"loss": 0.
|
| 255 |
"step": 20500
|
| 256 |
},
|
| 257 |
{
|
| 258 |
-
"epoch": 0.
|
| 259 |
-
"learning_rate": 4.
|
| 260 |
-
"loss": 0.
|
| 261 |
"step": 21000
|
| 262 |
},
|
| 263 |
{
|
| 264 |
-
"epoch": 0.
|
| 265 |
-
"learning_rate": 4.
|
| 266 |
-
"loss": 0.
|
| 267 |
"step": 21500
|
| 268 |
},
|
| 269 |
{
|
| 270 |
-
"epoch": 0.
|
| 271 |
-
"learning_rate": 4.
|
| 272 |
-
"loss": 0.
|
| 273 |
"step": 22000
|
| 274 |
},
|
| 275 |
{
|
| 276 |
-
"epoch": 0.
|
| 277 |
-
"learning_rate": 4.
|
| 278 |
-
"loss": 0.
|
| 279 |
"step": 22500
|
| 280 |
},
|
| 281 |
{
|
| 282 |
-
"epoch": 0.
|
| 283 |
-
"learning_rate": 4.
|
| 284 |
-
"loss": 0.
|
| 285 |
"step": 23000
|
| 286 |
},
|
| 287 |
{
|
| 288 |
-
"epoch": 0.
|
| 289 |
-
"learning_rate": 4.
|
| 290 |
-
"loss": 0.
|
| 291 |
"step": 23500
|
| 292 |
},
|
| 293 |
{
|
| 294 |
-
"epoch": 0.
|
| 295 |
-
"learning_rate": 4.
|
| 296 |
-
"loss": 0.
|
| 297 |
"step": 24000
|
| 298 |
},
|
| 299 |
{
|
| 300 |
-
"epoch": 0.
|
| 301 |
-
"learning_rate": 4.
|
| 302 |
-
"loss": 0.
|
| 303 |
"step": 24500
|
| 304 |
},
|
| 305 |
{
|
| 306 |
-
"epoch":
|
| 307 |
-
"learning_rate":
|
| 308 |
-
"loss": 0.
|
| 309 |
"step": 25000
|
| 310 |
},
|
| 311 |
{
|
| 312 |
-
"epoch":
|
| 313 |
-
"learning_rate":
|
| 314 |
-
"loss": 0.
|
| 315 |
"step": 25500
|
| 316 |
},
|
| 317 |
{
|
| 318 |
-
"epoch":
|
| 319 |
-
"learning_rate":
|
| 320 |
-
"loss": 0.
|
| 321 |
"step": 26000
|
| 322 |
},
|
| 323 |
{
|
| 324 |
-
"epoch":
|
| 325 |
-
"learning_rate":
|
| 326 |
-
"loss": 0.
|
| 327 |
"step": 26500
|
| 328 |
},
|
| 329 |
{
|
| 330 |
-
"epoch":
|
| 331 |
-
"learning_rate":
|
| 332 |
-
"loss": 0.
|
| 333 |
"step": 27000
|
| 334 |
},
|
| 335 |
{
|
| 336 |
-
"epoch": 1.
|
| 337 |
-
"learning_rate":
|
| 338 |
-
"loss": 0.
|
| 339 |
"step": 27500
|
| 340 |
},
|
| 341 |
{
|
| 342 |
-
"epoch": 1.
|
| 343 |
-
"learning_rate": 3.
|
| 344 |
-
"loss": 0.
|
| 345 |
"step": 28000
|
| 346 |
},
|
| 347 |
{
|
| 348 |
-
"epoch": 1.
|
| 349 |
-
"learning_rate": 3.
|
| 350 |
-
"loss": 0.
|
| 351 |
"step": 28500
|
| 352 |
},
|
| 353 |
{
|
| 354 |
-
"epoch": 1.
|
| 355 |
-
"learning_rate": 3.
|
| 356 |
-
"loss": 0.
|
| 357 |
"step": 29000
|
| 358 |
},
|
| 359 |
{
|
| 360 |
-
"epoch": 1.
|
| 361 |
-
"learning_rate": 3.
|
| 362 |
-
"loss": 0.
|
| 363 |
"step": 29500
|
| 364 |
},
|
| 365 |
{
|
| 366 |
-
"epoch": 1.
|
| 367 |
-
"learning_rate": 3.
|
| 368 |
-
"loss": 0.
|
| 369 |
"step": 30000
|
| 370 |
},
|
| 371 |
{
|
| 372 |
-
"epoch": 1.
|
| 373 |
-
"learning_rate": 3.
|
| 374 |
-
"loss": 0.
|
| 375 |
"step": 30500
|
| 376 |
},
|
| 377 |
{
|
| 378 |
-
"epoch": 1.
|
| 379 |
-
"learning_rate": 3.
|
| 380 |
-
"loss": 0.
|
| 381 |
"step": 31000
|
| 382 |
},
|
| 383 |
{
|
| 384 |
-
"epoch": 1.
|
| 385 |
-
"learning_rate": 3.
|
| 386 |
-
"loss": 0.
|
| 387 |
"step": 31500
|
| 388 |
},
|
| 389 |
{
|
| 390 |
-
"epoch": 1.
|
| 391 |
-
"learning_rate": 3.
|
| 392 |
-
"loss": 0.
|
| 393 |
"step": 32000
|
| 394 |
},
|
| 395 |
{
|
| 396 |
-
"epoch": 1.
|
| 397 |
-
"learning_rate": 3.
|
| 398 |
-
"loss": 0.
|
| 399 |
"step": 32500
|
| 400 |
},
|
| 401 |
{
|
| 402 |
-
"epoch": 1.
|
| 403 |
-
"learning_rate": 3.
|
| 404 |
-
"loss": 0.
|
| 405 |
"step": 33000
|
| 406 |
},
|
| 407 |
{
|
| 408 |
-
"epoch": 1.
|
| 409 |
-
"learning_rate": 3.
|
| 410 |
-
"loss": 0.
|
| 411 |
"step": 33500
|
| 412 |
},
|
| 413 |
{
|
| 414 |
-
"epoch": 1.
|
| 415 |
-
"learning_rate": 3.
|
| 416 |
-
"loss": 0.
|
| 417 |
"step": 34000
|
| 418 |
},
|
| 419 |
{
|
| 420 |
-
"epoch": 1.
|
| 421 |
-
"learning_rate": 3.
|
| 422 |
-
"loss": 0.
|
| 423 |
"step": 34500
|
| 424 |
},
|
| 425 |
{
|
| 426 |
-
"epoch": 1.
|
| 427 |
-
"learning_rate": 3.
|
| 428 |
-
"loss": 0.
|
| 429 |
"step": 35000
|
| 430 |
},
|
| 431 |
{
|
| 432 |
-
"epoch": 1.
|
| 433 |
-
"learning_rate": 3.
|
| 434 |
-
"loss": 0.
|
| 435 |
"step": 35500
|
| 436 |
},
|
| 437 |
{
|
| 438 |
-
"epoch": 1.
|
| 439 |
-
"learning_rate": 3.
|
| 440 |
-
"loss": 0.
|
| 441 |
"step": 36000
|
| 442 |
},
|
| 443 |
{
|
| 444 |
-
"epoch": 1.
|
| 445 |
-
"learning_rate": 3.
|
| 446 |
-
"loss": 0.
|
| 447 |
"step": 36500
|
| 448 |
},
|
| 449 |
{
|
| 450 |
-
"epoch": 1.
|
| 451 |
-
"learning_rate": 3.
|
| 452 |
-
"loss": 0.
|
| 453 |
"step": 37000
|
| 454 |
},
|
| 455 |
{
|
| 456 |
-
"epoch": 1.
|
| 457 |
-
"learning_rate": 3.
|
| 458 |
-
"loss": 0.
|
| 459 |
"step": 37500
|
| 460 |
},
|
| 461 |
{
|
| 462 |
-
"epoch": 1.
|
| 463 |
-
"learning_rate": 3.
|
| 464 |
-
"loss": 0.
|
| 465 |
"step": 38000
|
| 466 |
},
|
| 467 |
{
|
| 468 |
-
"epoch": 1.
|
| 469 |
-
"learning_rate": 3.
|
| 470 |
-
"loss": 0.
|
| 471 |
"step": 38500
|
| 472 |
},
|
| 473 |
{
|
| 474 |
-
"epoch": 1.
|
| 475 |
-
"learning_rate": 3.
|
| 476 |
-
"loss": 0.
|
| 477 |
"step": 39000
|
| 478 |
},
|
| 479 |
{
|
| 480 |
-
"epoch": 1.
|
| 481 |
-
"learning_rate": 3.
|
| 482 |
-
"loss": 0.
|
| 483 |
"step": 39500
|
| 484 |
},
|
| 485 |
{
|
| 486 |
-
"epoch": 1.
|
| 487 |
-
"learning_rate": 3.
|
| 488 |
-
"loss": 0.
|
| 489 |
"step": 40000
|
| 490 |
},
|
| 491 |
{
|
| 492 |
-
"epoch": 1.
|
| 493 |
-
"learning_rate": 3.
|
| 494 |
-
"loss": 0.
|
| 495 |
"step": 40500
|
| 496 |
},
|
| 497 |
{
|
| 498 |
-
"epoch": 1.
|
| 499 |
-
"learning_rate": 3.
|
| 500 |
-
"loss": 0.
|
| 501 |
"step": 41000
|
| 502 |
},
|
| 503 |
{
|
| 504 |
-
"epoch": 1.
|
| 505 |
-
"learning_rate": 3.
|
| 506 |
-
"loss": 0.
|
| 507 |
"step": 41500
|
| 508 |
},
|
| 509 |
{
|
| 510 |
-
"epoch": 1.
|
| 511 |
-
"learning_rate": 3.
|
| 512 |
-
"loss": 0.
|
| 513 |
"step": 42000
|
| 514 |
},
|
| 515 |
{
|
| 516 |
-
"epoch": 1.
|
| 517 |
-
"learning_rate": 3.
|
| 518 |
-
"loss": 0.
|
| 519 |
"step": 42500
|
| 520 |
},
|
| 521 |
{
|
| 522 |
-
"epoch": 1.
|
| 523 |
-
"learning_rate": 3.
|
| 524 |
-
"loss": 0.
|
| 525 |
"step": 43000
|
| 526 |
},
|
| 527 |
{
|
| 528 |
-
"epoch": 1.
|
| 529 |
-
"learning_rate": 3.
|
| 530 |
-
"loss": 0.
|
| 531 |
"step": 43500
|
| 532 |
},
|
| 533 |
{
|
| 534 |
-
"epoch": 1.
|
| 535 |
-
"learning_rate": 3.
|
| 536 |
-
"loss": 0.
|
| 537 |
"step": 44000
|
| 538 |
},
|
| 539 |
{
|
| 540 |
-
"epoch": 1.
|
| 541 |
-
"learning_rate": 3.
|
| 542 |
-
"loss": 0.
|
| 543 |
"step": 44500
|
| 544 |
},
|
| 545 |
{
|
| 546 |
-
"epoch": 1.
|
| 547 |
-
"learning_rate": 3.
|
| 548 |
-
"loss": 0.
|
| 549 |
"step": 45000
|
| 550 |
},
|
| 551 |
{
|
| 552 |
-
"epoch": 1.
|
| 553 |
-
"learning_rate": 3.
|
| 554 |
-
"loss": 0.
|
| 555 |
"step": 45500
|
| 556 |
},
|
| 557 |
{
|
| 558 |
-
"epoch": 1.
|
| 559 |
-
"learning_rate": 3.
|
| 560 |
-
"loss": 0.
|
| 561 |
"step": 46000
|
| 562 |
},
|
| 563 |
{
|
| 564 |
-
"epoch": 1.
|
| 565 |
-
"learning_rate": 3.
|
| 566 |
-
"loss": 0.
|
| 567 |
"step": 46500
|
| 568 |
},
|
| 569 |
{
|
| 570 |
-
"epoch": 1.
|
| 571 |
-
"learning_rate": 3.
|
| 572 |
-
"loss": 0.
|
| 573 |
"step": 47000
|
| 574 |
},
|
| 575 |
{
|
| 576 |
-
"epoch": 1.
|
| 577 |
-
"learning_rate": 3.
|
| 578 |
-
"loss": 0.
|
| 579 |
"step": 47500
|
| 580 |
},
|
| 581 |
{
|
| 582 |
-
"epoch": 1.
|
| 583 |
-
"learning_rate": 3.
|
| 584 |
-
"loss": 0.
|
| 585 |
"step": 48000
|
| 586 |
},
|
| 587 |
{
|
| 588 |
-
"epoch": 1.
|
| 589 |
-
"learning_rate": 3.
|
| 590 |
-
"loss": 0.
|
| 591 |
"step": 48500
|
| 592 |
},
|
| 593 |
{
|
| 594 |
-
"epoch": 1.
|
| 595 |
-
"learning_rate": 3.
|
| 596 |
-
"loss": 0.
|
| 597 |
"step": 49000
|
| 598 |
},
|
| 599 |
{
|
| 600 |
-
"epoch":
|
| 601 |
-
"learning_rate": 3.
|
| 602 |
-
"loss": 0.
|
| 603 |
"step": 49500
|
| 604 |
},
|
| 605 |
{
|
| 606 |
-
"epoch":
|
| 607 |
-
"learning_rate":
|
| 608 |
-
"loss": 0.
|
| 609 |
"step": 50000
|
| 610 |
},
|
| 611 |
{
|
| 612 |
-
"epoch":
|
| 613 |
-
"learning_rate":
|
| 614 |
-
"loss": 0.
|
| 615 |
"step": 50500
|
| 616 |
},
|
| 617 |
{
|
| 618 |
-
"epoch":
|
| 619 |
-
"learning_rate":
|
| 620 |
-
"loss": 0.
|
| 621 |
"step": 51000
|
| 622 |
},
|
| 623 |
{
|
| 624 |
-
"epoch":
|
| 625 |
-
"learning_rate":
|
| 626 |
-
"loss": 0.
|
| 627 |
"step": 51500
|
| 628 |
},
|
| 629 |
{
|
| 630 |
-
"epoch":
|
| 631 |
-
"learning_rate":
|
| 632 |
-
"loss": 0.
|
| 633 |
"step": 52000
|
| 634 |
},
|
| 635 |
{
|
| 636 |
-
"epoch":
|
| 637 |
-
"learning_rate":
|
| 638 |
-
"loss": 0.
|
| 639 |
"step": 52500
|
| 640 |
},
|
| 641 |
{
|
| 642 |
-
"epoch":
|
| 643 |
-
"learning_rate":
|
| 644 |
-
"loss": 0.
|
| 645 |
"step": 53000
|
| 646 |
},
|
| 647 |
{
|
| 648 |
-
"epoch":
|
| 649 |
-
"learning_rate":
|
| 650 |
-
"loss": 0.
|
| 651 |
"step": 53500
|
| 652 |
},
|
| 653 |
{
|
| 654 |
-
"epoch":
|
| 655 |
-
"learning_rate":
|
| 656 |
-
"loss": 0.
|
| 657 |
"step": 54000
|
| 658 |
},
|
| 659 |
{
|
| 660 |
-
"epoch":
|
| 661 |
-
"learning_rate":
|
| 662 |
-
"loss": 0.
|
| 663 |
"step": 54500
|
| 664 |
},
|
| 665 |
{
|
| 666 |
-
"epoch":
|
| 667 |
-
"learning_rate":
|
| 668 |
-
"loss": 0.
|
| 669 |
"step": 55000
|
| 670 |
},
|
| 671 |
{
|
| 672 |
-
"epoch": 2.
|
| 673 |
-
"learning_rate": 2.
|
| 674 |
-
"loss": 0.
|
| 675 |
"step": 55500
|
| 676 |
},
|
| 677 |
{
|
| 678 |
-
"epoch": 2.
|
| 679 |
-
"learning_rate": 2.
|
| 680 |
-
"loss": 0.
|
| 681 |
"step": 56000
|
| 682 |
},
|
| 683 |
{
|
| 684 |
-
"epoch": 2.
|
| 685 |
-
"learning_rate": 2.
|
| 686 |
-
"loss": 0.
|
| 687 |
"step": 56500
|
| 688 |
},
|
| 689 |
{
|
| 690 |
-
"epoch": 2.
|
| 691 |
-
"learning_rate": 2.
|
| 692 |
-
"loss": 0.
|
| 693 |
"step": 57000
|
| 694 |
},
|
| 695 |
{
|
| 696 |
-
"epoch": 2.
|
| 697 |
-
"learning_rate": 2.
|
| 698 |
-
"loss": 0.
|
| 699 |
"step": 57500
|
| 700 |
},
|
| 701 |
{
|
| 702 |
-
"epoch": 2.
|
| 703 |
-
"learning_rate": 2.
|
| 704 |
-
"loss": 0.
|
| 705 |
"step": 58000
|
| 706 |
},
|
| 707 |
{
|
| 708 |
-
"epoch": 2.
|
| 709 |
-
"learning_rate": 2.
|
| 710 |
-
"loss": 0.
|
| 711 |
"step": 58500
|
| 712 |
},
|
| 713 |
{
|
| 714 |
-
"epoch": 2.
|
| 715 |
-
"learning_rate": 2.
|
| 716 |
-
"loss": 0.
|
| 717 |
"step": 59000
|
| 718 |
},
|
| 719 |
{
|
| 720 |
-
"epoch": 2.
|
| 721 |
-
"learning_rate": 2.
|
| 722 |
-
"loss": 0.
|
| 723 |
"step": 59500
|
| 724 |
},
|
| 725 |
{
|
| 726 |
-
"epoch":
|
| 727 |
-
"learning_rate":
|
| 728 |
-
"loss": 0.
|
| 729 |
"step": 60000
|
| 730 |
},
|
| 731 |
{
|
| 732 |
-
"epoch":
|
| 733 |
-
"learning_rate":
|
| 734 |
-
"loss": 0.
|
| 735 |
"step": 60500
|
| 736 |
},
|
| 737 |
{
|
| 738 |
-
"epoch":
|
| 739 |
-
"learning_rate":
|
| 740 |
-
"loss": 0.
|
| 741 |
"step": 61000
|
| 742 |
},
|
| 743 |
{
|
| 744 |
-
"epoch":
|
| 745 |
-
"learning_rate":
|
| 746 |
-
"loss": 0.
|
| 747 |
"step": 61500
|
| 748 |
},
|
| 749 |
{
|
| 750 |
-
"epoch":
|
| 751 |
-
"learning_rate":
|
| 752 |
-
"loss": 0.
|
| 753 |
"step": 62000
|
| 754 |
},
|
| 755 |
{
|
| 756 |
-
"epoch":
|
| 757 |
-
"learning_rate":
|
| 758 |
-
"loss": 0.
|
| 759 |
"step": 62500
|
| 760 |
},
|
| 761 |
{
|
| 762 |
-
"epoch":
|
| 763 |
-
"learning_rate":
|
| 764 |
-
"loss": 0.
|
| 765 |
"step": 63000
|
| 766 |
},
|
| 767 |
{
|
| 768 |
-
"epoch":
|
| 769 |
-
"learning_rate":
|
| 770 |
-
"loss": 0.
|
| 771 |
"step": 63500
|
| 772 |
},
|
| 773 |
{
|
| 774 |
-
"epoch":
|
| 775 |
-
"learning_rate":
|
| 776 |
-
"loss": 0.
|
| 777 |
"step": 64000
|
| 778 |
},
|
| 779 |
{
|
| 780 |
-
"epoch":
|
| 781 |
-
"learning_rate":
|
| 782 |
-
"loss": 0.
|
| 783 |
"step": 64500
|
| 784 |
},
|
| 785 |
{
|
| 786 |
-
"epoch":
|
| 787 |
-
"learning_rate": 2.
|
| 788 |
-
"loss": 0.
|
| 789 |
"step": 65000
|
| 790 |
},
|
| 791 |
{
|
| 792 |
-
"epoch":
|
| 793 |
-
"learning_rate": 2.
|
| 794 |
-
"loss": 0.
|
| 795 |
"step": 65500
|
| 796 |
},
|
| 797 |
{
|
| 798 |
-
"epoch":
|
| 799 |
-
"learning_rate": 2.
|
| 800 |
-
"loss": 0.
|
| 801 |
"step": 66000
|
| 802 |
},
|
| 803 |
{
|
| 804 |
-
"epoch":
|
| 805 |
-
"learning_rate":
|
| 806 |
-
"loss": 0.
|
| 807 |
"step": 66500
|
| 808 |
},
|
| 809 |
{
|
| 810 |
-
"epoch":
|
| 811 |
-
"learning_rate":
|
| 812 |
-
"loss": 0.
|
| 813 |
"step": 67000
|
| 814 |
},
|
| 815 |
{
|
| 816 |
-
"epoch":
|
| 817 |
-
"learning_rate":
|
| 818 |
-
"loss": 0.
|
| 819 |
"step": 67500
|
| 820 |
},
|
| 821 |
-
{
|
| 822 |
-
"epoch": 2.74,
|
| 823 |
-
"learning_rate": 2.259775860678868e-05,
|
| 824 |
-
"loss": 0.0444,
|
| 825 |
-
"step": 68000
|
| 826 |
-
},
|
| 827 |
-
{
|
| 828 |
-
"epoch": 2.76,
|
| 829 |
-
"learning_rate": 2.239619446908006e-05,
|
| 830 |
-
"loss": 0.0452,
|
| 831 |
-
"step": 68500
|
| 832 |
-
},
|
| 833 |
-
{
|
| 834 |
-
"epoch": 2.78,
|
| 835 |
-
"learning_rate": 2.2194630331371444e-05,
|
| 836 |
-
"loss": 0.0424,
|
| 837 |
-
"step": 69000
|
| 838 |
-
},
|
| 839 |
-
{
|
| 840 |
-
"epoch": 2.8,
|
| 841 |
-
"learning_rate": 2.1993066193662826e-05,
|
| 842 |
-
"loss": 0.0442,
|
| 843 |
-
"step": 69500
|
| 844 |
-
},
|
| 845 |
-
{
|
| 846 |
-
"epoch": 2.82,
|
| 847 |
-
"learning_rate": 2.1791905184229623e-05,
|
| 848 |
-
"loss": 0.0383,
|
| 849 |
-
"step": 70000
|
| 850 |
-
},
|
| 851 |
-
{
|
| 852 |
-
"epoch": 2.84,
|
| 853 |
-
"learning_rate": 2.1590341046521002e-05,
|
| 854 |
-
"loss": 0.0446,
|
| 855 |
-
"step": 70500
|
| 856 |
-
},
|
| 857 |
-
{
|
| 858 |
-
"epoch": 2.86,
|
| 859 |
-
"learning_rate": 2.1388776908812385e-05,
|
| 860 |
-
"loss": 0.0419,
|
| 861 |
-
"step": 71000
|
| 862 |
-
},
|
| 863 |
-
{
|
| 864 |
-
"epoch": 2.88,
|
| 865 |
-
"learning_rate": 2.1187212771103764e-05,
|
| 866 |
-
"loss": 0.0418,
|
| 867 |
-
"step": 71500
|
| 868 |
-
},
|
| 869 |
-
{
|
| 870 |
-
"epoch": 2.9,
|
| 871 |
-
"learning_rate": 2.0986051761670564e-05,
|
| 872 |
-
"loss": 0.0409,
|
| 873 |
-
"step": 72000
|
| 874 |
-
},
|
| 875 |
-
{
|
| 876 |
-
"epoch": 2.92,
|
| 877 |
-
"learning_rate": 2.0784487623961947e-05,
|
| 878 |
-
"loss": 0.0378,
|
| 879 |
-
"step": 72500
|
| 880 |
-
},
|
| 881 |
-
{
|
| 882 |
-
"epoch": 2.94,
|
| 883 |
-
"learning_rate": 2.0582923486253326e-05,
|
| 884 |
-
"loss": 0.0363,
|
| 885 |
-
"step": 73000
|
| 886 |
-
},
|
| 887 |
-
{
|
| 888 |
-
"epoch": 2.96,
|
| 889 |
-
"learning_rate": 2.0381359348544705e-05,
|
| 890 |
-
"loss": 0.0363,
|
| 891 |
-
"step": 73500
|
| 892 |
-
},
|
| 893 |
-
{
|
| 894 |
-
"epoch": 2.98,
|
| 895 |
-
"learning_rate": 2.0179795210836088e-05,
|
| 896 |
-
"loss": 0.0366,
|
| 897 |
-
"step": 74000
|
| 898 |
-
},
|
| 899 |
-
{
|
| 900 |
-
"epoch": 3.0,
|
| 901 |
-
"learning_rate": 1.9978634201402888e-05,
|
| 902 |
-
"loss": 0.0342,
|
| 903 |
-
"step": 74500
|
| 904 |
-
},
|
| 905 |
-
{
|
| 906 |
-
"epoch": 3.02,
|
| 907 |
-
"learning_rate": 1.977707006369427e-05,
|
| 908 |
-
"loss": 0.0198,
|
| 909 |
-
"step": 75000
|
| 910 |
-
},
|
| 911 |
-
{
|
| 912 |
-
"epoch": 3.04,
|
| 913 |
-
"learning_rate": 1.957550592598565e-05,
|
| 914 |
-
"loss": 0.0209,
|
| 915 |
-
"step": 75500
|
| 916 |
-
},
|
| 917 |
-
{
|
| 918 |
-
"epoch": 3.06,
|
| 919 |
-
"learning_rate": 1.937394178827703e-05,
|
| 920 |
-
"loss": 0.0186,
|
| 921 |
-
"step": 76000
|
| 922 |
-
},
|
| 923 |
-
{
|
| 924 |
-
"epoch": 3.08,
|
| 925 |
-
"learning_rate": 1.9172377650568412e-05,
|
| 926 |
-
"loss": 0.0183,
|
| 927 |
-
"step": 76500
|
| 928 |
-
},
|
| 929 |
-
{
|
| 930 |
-
"epoch": 3.1,
|
| 931 |
-
"learning_rate": 1.897081351285979e-05,
|
| 932 |
-
"loss": 0.0187,
|
| 933 |
-
"step": 77000
|
| 934 |
-
},
|
| 935 |
-
{
|
| 936 |
-
"epoch": 3.12,
|
| 937 |
-
"learning_rate": 1.8769249375151174e-05,
|
| 938 |
-
"loss": 0.0192,
|
| 939 |
-
"step": 77500
|
| 940 |
-
},
|
| 941 |
-
{
|
| 942 |
-
"epoch": 3.14,
|
| 943 |
-
"learning_rate": 1.8567685237442557e-05,
|
| 944 |
-
"loss": 0.0163,
|
| 945 |
-
"step": 78000
|
| 946 |
-
},
|
| 947 |
-
{
|
| 948 |
-
"epoch": 3.16,
|
| 949 |
-
"learning_rate": 1.8366121099733936e-05,
|
| 950 |
-
"loss": 0.0218,
|
| 951 |
-
"step": 78500
|
| 952 |
-
},
|
| 953 |
-
{
|
| 954 |
-
"epoch": 3.18,
|
| 955 |
-
"learning_rate": 1.8164556962025315e-05,
|
| 956 |
-
"loss": 0.0179,
|
| 957 |
-
"step": 79000
|
| 958 |
-
},
|
| 959 |
-
{
|
| 960 |
-
"epoch": 3.2,
|
| 961 |
-
"learning_rate": 1.7962992824316698e-05,
|
| 962 |
-
"loss": 0.0201,
|
| 963 |
-
"step": 79500
|
| 964 |
-
},
|
| 965 |
-
{
|
| 966 |
-
"epoch": 3.23,
|
| 967 |
-
"learning_rate": 1.776142868660808e-05,
|
| 968 |
-
"loss": 0.0181,
|
| 969 |
-
"step": 80000
|
| 970 |
-
},
|
| 971 |
-
{
|
| 972 |
-
"epoch": 3.25,
|
| 973 |
-
"learning_rate": 1.755986454889946e-05,
|
| 974 |
-
"loss": 0.0211,
|
| 975 |
-
"step": 80500
|
| 976 |
-
},
|
| 977 |
-
{
|
| 978 |
-
"epoch": 3.27,
|
| 979 |
-
"learning_rate": 1.735870353946626e-05,
|
| 980 |
-
"loss": 0.0182,
|
| 981 |
-
"step": 81000
|
| 982 |
-
},
|
| 983 |
-
{
|
| 984 |
-
"epoch": 3.29,
|
| 985 |
-
"learning_rate": 1.7157542530033057e-05,
|
| 986 |
-
"loss": 0.0199,
|
| 987 |
-
"step": 81500
|
| 988 |
-
},
|
| 989 |
-
{
|
| 990 |
-
"epoch": 3.31,
|
| 991 |
-
"learning_rate": 1.695597839232444e-05,
|
| 992 |
-
"loss": 0.0173,
|
| 993 |
-
"step": 82000
|
| 994 |
-
},
|
| 995 |
-
{
|
| 996 |
-
"epoch": 3.33,
|
| 997 |
-
"learning_rate": 1.675441425461582e-05,
|
| 998 |
-
"loss": 0.0195,
|
| 999 |
-
"step": 82500
|
| 1000 |
-
},
|
| 1001 |
-
{
|
| 1002 |
-
"epoch": 3.35,
|
| 1003 |
-
"learning_rate": 1.65528501169072e-05,
|
| 1004 |
-
"loss": 0.0212,
|
| 1005 |
-
"step": 83000
|
| 1006 |
-
},
|
| 1007 |
-
{
|
| 1008 |
-
"epoch": 3.37,
|
| 1009 |
-
"learning_rate": 1.6351285979198584e-05,
|
| 1010 |
-
"loss": 0.0186,
|
| 1011 |
-
"step": 83500
|
| 1012 |
-
},
|
| 1013 |
-
{
|
| 1014 |
-
"epoch": 3.39,
|
| 1015 |
-
"learning_rate": 1.6150124969765377e-05,
|
| 1016 |
-
"loss": 0.0188,
|
| 1017 |
-
"step": 84000
|
| 1018 |
-
},
|
| 1019 |
-
{
|
| 1020 |
-
"epoch": 3.41,
|
| 1021 |
-
"learning_rate": 1.5948963960332177e-05,
|
| 1022 |
-
"loss": 0.018,
|
| 1023 |
-
"step": 84500
|
| 1024 |
-
},
|
| 1025 |
-
{
|
| 1026 |
-
"epoch": 3.43,
|
| 1027 |
-
"learning_rate": 1.5747802950898977e-05,
|
| 1028 |
-
"loss": 0.0202,
|
| 1029 |
-
"step": 85000
|
| 1030 |
-
},
|
| 1031 |
-
{
|
| 1032 |
-
"epoch": 3.45,
|
| 1033 |
-
"learning_rate": 1.554623881319036e-05,
|
| 1034 |
-
"loss": 0.0186,
|
| 1035 |
-
"step": 85500
|
| 1036 |
-
},
|
| 1037 |
-
{
|
| 1038 |
-
"epoch": 3.47,
|
| 1039 |
-
"learning_rate": 1.534467467548174e-05,
|
| 1040 |
-
"loss": 0.0182,
|
| 1041 |
-
"step": 86000
|
| 1042 |
-
},
|
| 1043 |
-
{
|
| 1044 |
-
"epoch": 3.49,
|
| 1045 |
-
"learning_rate": 1.514311053777312e-05,
|
| 1046 |
-
"loss": 0.0172,
|
| 1047 |
-
"step": 86500
|
| 1048 |
-
},
|
| 1049 |
-
{
|
| 1050 |
-
"epoch": 3.51,
|
| 1051 |
-
"learning_rate": 1.4941546400064501e-05,
|
| 1052 |
-
"loss": 0.0164,
|
| 1053 |
-
"step": 87000
|
| 1054 |
-
},
|
| 1055 |
-
{
|
| 1056 |
-
"epoch": 3.53,
|
| 1057 |
-
"learning_rate": 1.4739982262355884e-05,
|
| 1058 |
-
"loss": 0.015,
|
| 1059 |
-
"step": 87500
|
| 1060 |
-
},
|
| 1061 |
-
{
|
| 1062 |
-
"epoch": 3.55,
|
| 1063 |
-
"learning_rate": 1.4538418124647263e-05,
|
| 1064 |
-
"loss": 0.0171,
|
| 1065 |
-
"step": 88000
|
| 1066 |
-
},
|
| 1067 |
-
{
|
| 1068 |
-
"epoch": 3.57,
|
| 1069 |
-
"learning_rate": 1.4336853986938644e-05,
|
| 1070 |
-
"loss": 0.0148,
|
| 1071 |
-
"step": 88500
|
| 1072 |
-
},
|
| 1073 |
-
{
|
| 1074 |
-
"epoch": 3.59,
|
| 1075 |
-
"learning_rate": 1.4135289849230027e-05,
|
| 1076 |
-
"loss": 0.0168,
|
| 1077 |
-
"step": 89000
|
| 1078 |
-
},
|
| 1079 |
-
{
|
| 1080 |
-
"epoch": 3.61,
|
| 1081 |
-
"learning_rate": 1.3933725711521406e-05,
|
| 1082 |
-
"loss": 0.0161,
|
| 1083 |
-
"step": 89500
|
| 1084 |
-
},
|
| 1085 |
-
{
|
| 1086 |
-
"epoch": 3.63,
|
| 1087 |
-
"learning_rate": 1.3732564702088205e-05,
|
| 1088 |
-
"loss": 0.0152,
|
| 1089 |
-
"step": 90000
|
| 1090 |
-
},
|
| 1091 |
-
{
|
| 1092 |
-
"epoch": 3.65,
|
| 1093 |
-
"learning_rate": 1.3531000564379587e-05,
|
| 1094 |
-
"loss": 0.0155,
|
| 1095 |
-
"step": 90500
|
| 1096 |
-
},
|
| 1097 |
-
{
|
| 1098 |
-
"epoch": 3.67,
|
| 1099 |
-
"learning_rate": 1.3329436426670968e-05,
|
| 1100 |
-
"loss": 0.0155,
|
| 1101 |
-
"step": 91000
|
| 1102 |
-
},
|
| 1103 |
-
{
|
| 1104 |
-
"epoch": 3.69,
|
| 1105 |
-
"learning_rate": 1.3127872288962348e-05,
|
| 1106 |
-
"loss": 0.0137,
|
| 1107 |
-
"step": 91500
|
| 1108 |
-
},
|
| 1109 |
-
{
|
| 1110 |
-
"epoch": 3.71,
|
| 1111 |
-
"learning_rate": 1.2926711279529146e-05,
|
| 1112 |
-
"loss": 0.0154,
|
| 1113 |
-
"step": 92000
|
| 1114 |
-
},
|
| 1115 |
-
{
|
| 1116 |
-
"epoch": 3.73,
|
| 1117 |
-
"learning_rate": 1.2725147141820529e-05,
|
| 1118 |
-
"loss": 0.0152,
|
| 1119 |
-
"step": 92500
|
| 1120 |
-
},
|
| 1121 |
-
{
|
| 1122 |
-
"epoch": 3.75,
|
| 1123 |
-
"learning_rate": 1.2523583004111908e-05,
|
| 1124 |
-
"loss": 0.0156,
|
| 1125 |
-
"step": 93000
|
| 1126 |
-
},
|
| 1127 |
-
{
|
| 1128 |
-
"epoch": 3.77,
|
| 1129 |
-
"learning_rate": 1.232201886640329e-05,
|
| 1130 |
-
"loss": 0.0147,
|
| 1131 |
-
"step": 93500
|
| 1132 |
-
},
|
| 1133 |
-
{
|
| 1134 |
-
"epoch": 3.79,
|
| 1135 |
-
"learning_rate": 1.2120454728694672e-05,
|
| 1136 |
-
"loss": 0.0158,
|
| 1137 |
-
"step": 94000
|
| 1138 |
-
},
|
| 1139 |
-
{
|
| 1140 |
-
"epoch": 3.81,
|
| 1141 |
-
"learning_rate": 1.1918890590986053e-05,
|
| 1142 |
-
"loss": 0.015,
|
| 1143 |
-
"step": 94500
|
| 1144 |
-
},
|
| 1145 |
-
{
|
| 1146 |
-
"epoch": 3.83,
|
| 1147 |
-
"learning_rate": 1.1717326453277434e-05,
|
| 1148 |
-
"loss": 0.0131,
|
| 1149 |
-
"step": 95000
|
| 1150 |
-
},
|
| 1151 |
-
{
|
| 1152 |
-
"epoch": 3.85,
|
| 1153 |
-
"learning_rate": 1.1515762315568815e-05,
|
| 1154 |
-
"loss": 0.0145,
|
| 1155 |
-
"step": 95500
|
| 1156 |
-
},
|
| 1157 |
-
{
|
| 1158 |
-
"epoch": 3.87,
|
| 1159 |
-
"learning_rate": 1.1314198177860195e-05,
|
| 1160 |
-
"loss": 0.0136,
|
| 1161 |
-
"step": 96000
|
| 1162 |
-
},
|
| 1163 |
-
{
|
| 1164 |
-
"epoch": 3.89,
|
| 1165 |
-
"learning_rate": 1.1112634040151578e-05,
|
| 1166 |
-
"loss": 0.0121,
|
| 1167 |
-
"step": 96500
|
| 1168 |
-
},
|
| 1169 |
-
{
|
| 1170 |
-
"epoch": 3.91,
|
| 1171 |
-
"learning_rate": 1.0911069902442957e-05,
|
| 1172 |
-
"loss": 0.0125,
|
| 1173 |
-
"step": 97000
|
| 1174 |
-
},
|
| 1175 |
-
{
|
| 1176 |
-
"epoch": 3.93,
|
| 1177 |
-
"learning_rate": 1.0709505764734338e-05,
|
| 1178 |
-
"loss": 0.0116,
|
| 1179 |
-
"step": 97500
|
| 1180 |
-
},
|
| 1181 |
-
{
|
| 1182 |
-
"epoch": 3.95,
|
| 1183 |
-
"learning_rate": 1.0508344755301137e-05,
|
| 1184 |
-
"loss": 0.0105,
|
| 1185 |
-
"step": 98000
|
| 1186 |
-
},
|
| 1187 |
-
{
|
| 1188 |
-
"epoch": 3.97,
|
| 1189 |
-
"learning_rate": 1.0306780617592518e-05,
|
| 1190 |
-
"loss": 0.0128,
|
| 1191 |
-
"step": 98500
|
| 1192 |
-
},
|
| 1193 |
-
{
|
| 1194 |
-
"epoch": 3.99,
|
| 1195 |
-
"learning_rate": 1.01052164798839e-05,
|
| 1196 |
-
"loss": 0.0115,
|
| 1197 |
-
"step": 99000
|
| 1198 |
-
},
|
| 1199 |
-
{
|
| 1200 |
-
"epoch": 4.01,
|
| 1201 |
-
"learning_rate": 9.90365234217528e-06,
|
| 1202 |
-
"loss": 0.009,
|
| 1203 |
-
"step": 99500
|
| 1204 |
-
},
|
| 1205 |
-
{
|
| 1206 |
-
"epoch": 4.03,
|
| 1207 |
-
"learning_rate": 9.70208820446666e-06,
|
| 1208 |
-
"loss": 0.0071,
|
| 1209 |
-
"step": 100000
|
| 1210 |
-
},
|
| 1211 |
-
{
|
| 1212 |
-
"epoch": 4.05,
|
| 1213 |
-
"learning_rate": 9.500524066758043e-06,
|
| 1214 |
-
"loss": 0.006,
|
| 1215 |
-
"step": 100500
|
| 1216 |
-
},
|
| 1217 |
-
{
|
| 1218 |
-
"epoch": 4.07,
|
| 1219 |
-
"learning_rate": 9.298959929049424e-06,
|
| 1220 |
-
"loss": 0.0063,
|
| 1221 |
-
"step": 101000
|
| 1222 |
-
},
|
| 1223 |
-
{
|
| 1224 |
-
"epoch": 4.09,
|
| 1225 |
-
"learning_rate": 9.097395791340805e-06,
|
| 1226 |
-
"loss": 0.006,
|
| 1227 |
-
"step": 101500
|
| 1228 |
-
},
|
| 1229 |
-
{
|
| 1230 |
-
"epoch": 4.11,
|
| 1231 |
-
"learning_rate": 8.895831653632186e-06,
|
| 1232 |
-
"loss": 0.0065,
|
| 1233 |
-
"step": 102000
|
| 1234 |
-
},
|
| 1235 |
-
{
|
| 1236 |
-
"epoch": 4.13,
|
| 1237 |
-
"learning_rate": 8.694267515923567e-06,
|
| 1238 |
-
"loss": 0.0059,
|
| 1239 |
-
"step": 102500
|
| 1240 |
-
},
|
| 1241 |
-
{
|
| 1242 |
-
"epoch": 4.15,
|
| 1243 |
-
"learning_rate": 8.492703378214948e-06,
|
| 1244 |
-
"loss": 0.0063,
|
| 1245 |
-
"step": 103000
|
| 1246 |
-
},
|
| 1247 |
-
{
|
| 1248 |
-
"epoch": 4.17,
|
| 1249 |
-
"learning_rate": 8.291542368781747e-06,
|
| 1250 |
-
"loss": 0.007,
|
| 1251 |
-
"step": 103500
|
| 1252 |
-
},
|
| 1253 |
-
{
|
| 1254 |
-
"epoch": 4.19,
|
| 1255 |
-
"learning_rate": 8.089978231073128e-06,
|
| 1256 |
-
"loss": 0.0059,
|
| 1257 |
-
"step": 104000
|
| 1258 |
-
},
|
| 1259 |
-
{
|
| 1260 |
-
"epoch": 4.21,
|
| 1261 |
-
"learning_rate": 7.888414093364509e-06,
|
| 1262 |
-
"loss": 0.0062,
|
| 1263 |
-
"step": 104500
|
| 1264 |
-
},
|
| 1265 |
-
{
|
| 1266 |
-
"epoch": 4.23,
|
| 1267 |
-
"learning_rate": 7.686849955655891e-06,
|
| 1268 |
-
"loss": 0.0055,
|
| 1269 |
-
"step": 105000
|
| 1270 |
-
},
|
| 1271 |
-
{
|
| 1272 |
-
"epoch": 4.25,
|
| 1273 |
-
"learning_rate": 7.4852858179472715e-06,
|
| 1274 |
-
"loss": 0.0055,
|
| 1275 |
-
"step": 105500
|
| 1276 |
-
},
|
| 1277 |
-
{
|
| 1278 |
-
"epoch": 4.27,
|
| 1279 |
-
"learning_rate": 7.283721680238652e-06,
|
| 1280 |
-
"loss": 0.0048,
|
| 1281 |
-
"step": 106000
|
| 1282 |
-
},
|
| 1283 |
-
{
|
| 1284 |
-
"epoch": 4.29,
|
| 1285 |
-
"learning_rate": 7.0821575425300335e-06,
|
| 1286 |
-
"loss": 0.0063,
|
| 1287 |
-
"step": 106500
|
| 1288 |
-
},
|
| 1289 |
-
{
|
| 1290 |
-
"epoch": 4.31,
|
| 1291 |
-
"learning_rate": 6.880996533096831e-06,
|
| 1292 |
-
"loss": 0.0059,
|
| 1293 |
-
"step": 107000
|
| 1294 |
-
},
|
| 1295 |
-
{
|
| 1296 |
-
"epoch": 4.33,
|
| 1297 |
-
"learning_rate": 6.679432395388213e-06,
|
| 1298 |
-
"loss": 0.0052,
|
| 1299 |
-
"step": 107500
|
| 1300 |
-
},
|
| 1301 |
-
{
|
| 1302 |
-
"epoch": 4.35,
|
| 1303 |
-
"learning_rate": 6.477868257679594e-06,
|
| 1304 |
-
"loss": 0.005,
|
| 1305 |
-
"step": 108000
|
| 1306 |
-
},
|
| 1307 |
-
{
|
| 1308 |
-
"epoch": 4.37,
|
| 1309 |
-
"learning_rate": 6.276304119970976e-06,
|
| 1310 |
-
"loss": 0.0043,
|
| 1311 |
-
"step": 108500
|
| 1312 |
-
},
|
| 1313 |
-
{
|
| 1314 |
-
"epoch": 4.39,
|
| 1315 |
-
"learning_rate": 6.074739982262357e-06,
|
| 1316 |
-
"loss": 0.0056,
|
| 1317 |
-
"step": 109000
|
| 1318 |
-
},
|
| 1319 |
-
{
|
| 1320 |
-
"epoch": 4.41,
|
| 1321 |
-
"learning_rate": 5.873175844553738e-06,
|
| 1322 |
-
"loss": 0.0046,
|
| 1323 |
-
"step": 109500
|
| 1324 |
-
},
|
| 1325 |
-
{
|
| 1326 |
-
"epoch": 4.43,
|
| 1327 |
-
"learning_rate": 5.672014835120536e-06,
|
| 1328 |
-
"loss": 0.0056,
|
| 1329 |
-
"step": 110000
|
| 1330 |
-
},
|
| 1331 |
-
{
|
| 1332 |
-
"epoch": 4.45,
|
| 1333 |
-
"learning_rate": 5.470450697411917e-06,
|
| 1334 |
-
"loss": 0.0049,
|
| 1335 |
-
"step": 110500
|
| 1336 |
-
},
|
| 1337 |
-
{
|
| 1338 |
-
"epoch": 4.47,
|
| 1339 |
-
"learning_rate": 5.269289687978715e-06,
|
| 1340 |
-
"loss": 0.006,
|
| 1341 |
-
"step": 111000
|
| 1342 |
-
},
|
| 1343 |
-
{
|
| 1344 |
-
"epoch": 4.49,
|
| 1345 |
-
"learning_rate": 5.067725550270096e-06,
|
| 1346 |
-
"loss": 0.0042,
|
| 1347 |
-
"step": 111500
|
| 1348 |
-
},
|
| 1349 |
-
{
|
| 1350 |
-
"epoch": 4.52,
|
| 1351 |
-
"learning_rate": 4.866161412561477e-06,
|
| 1352 |
-
"loss": 0.0048,
|
| 1353 |
-
"step": 112000
|
| 1354 |
-
},
|
| 1355 |
-
{
|
| 1356 |
-
"epoch": 4.54,
|
| 1357 |
-
"learning_rate": 4.664597274852858e-06,
|
| 1358 |
-
"loss": 0.0065,
|
| 1359 |
-
"step": 112500
|
| 1360 |
-
},
|
| 1361 |
-
{
|
| 1362 |
-
"epoch": 4.56,
|
| 1363 |
-
"learning_rate": 4.463033137144239e-06,
|
| 1364 |
-
"loss": 0.0052,
|
| 1365 |
-
"step": 113000
|
| 1366 |
-
},
|
| 1367 |
-
{
|
| 1368 |
-
"epoch": 4.58,
|
| 1369 |
-
"learning_rate": 4.261468999435621e-06,
|
| 1370 |
-
"loss": 0.0051,
|
| 1371 |
-
"step": 113500
|
| 1372 |
-
},
|
| 1373 |
-
{
|
| 1374 |
-
"epoch": 4.6,
|
| 1375 |
-
"learning_rate": 4.059904861727002e-06,
|
| 1376 |
-
"loss": 0.0037,
|
| 1377 |
-
"step": 114000
|
| 1378 |
-
},
|
| 1379 |
-
{
|
| 1380 |
-
"epoch": 4.62,
|
| 1381 |
-
"learning_rate": 3.858340724018382e-06,
|
| 1382 |
-
"loss": 0.0038,
|
| 1383 |
-
"step": 114500
|
| 1384 |
-
},
|
| 1385 |
-
{
|
| 1386 |
-
"epoch": 4.64,
|
| 1387 |
-
"learning_rate": 3.656776586309764e-06,
|
| 1388 |
-
"loss": 0.0044,
|
| 1389 |
-
"step": 115000
|
| 1390 |
-
},
|
| 1391 |
-
{
|
| 1392 |
-
"epoch": 4.66,
|
| 1393 |
-
"learning_rate": 3.455615576876562e-06,
|
| 1394 |
-
"loss": 0.0038,
|
| 1395 |
-
"step": 115500
|
| 1396 |
-
},
|
| 1397 |
-
{
|
| 1398 |
-
"epoch": 4.68,
|
| 1399 |
-
"learning_rate": 3.2544545674433607e-06,
|
| 1400 |
-
"loss": 0.0046,
|
| 1401 |
-
"step": 116000
|
| 1402 |
-
},
|
| 1403 |
-
{
|
| 1404 |
-
"epoch": 4.7,
|
| 1405 |
-
"learning_rate": 3.0528904297347417e-06,
|
| 1406 |
-
"loss": 0.0044,
|
| 1407 |
-
"step": 116500
|
| 1408 |
-
},
|
| 1409 |
-
{
|
| 1410 |
-
"epoch": 4.72,
|
| 1411 |
-
"learning_rate": 2.8513262920261227e-06,
|
| 1412 |
-
"loss": 0.0038,
|
| 1413 |
-
"step": 117000
|
| 1414 |
-
},
|
| 1415 |
-
{
|
| 1416 |
-
"epoch": 4.74,
|
| 1417 |
-
"learning_rate": 2.649762154317504e-06,
|
| 1418 |
-
"loss": 0.0041,
|
| 1419 |
-
"step": 117500
|
| 1420 |
-
},
|
| 1421 |
-
{
|
| 1422 |
-
"epoch": 4.76,
|
| 1423 |
-
"learning_rate": 2.448198016608885e-06,
|
| 1424 |
-
"loss": 0.0035,
|
| 1425 |
-
"step": 118000
|
| 1426 |
-
},
|
| 1427 |
-
{
|
| 1428 |
-
"epoch": 4.78,
|
| 1429 |
-
"learning_rate": 2.246633878900266e-06,
|
| 1430 |
-
"loss": 0.0038,
|
| 1431 |
-
"step": 118500
|
| 1432 |
-
},
|
| 1433 |
-
{
|
| 1434 |
-
"epoch": 4.8,
|
| 1435 |
-
"learning_rate": 2.0450697411916475e-06,
|
| 1436 |
-
"loss": 0.0042,
|
| 1437 |
-
"step": 119000
|
| 1438 |
-
},
|
| 1439 |
-
{
|
| 1440 |
-
"epoch": 4.82,
|
| 1441 |
-
"learning_rate": 1.8435056034830283e-06,
|
| 1442 |
-
"loss": 0.0036,
|
| 1443 |
-
"step": 119500
|
| 1444 |
-
},
|
| 1445 |
-
{
|
| 1446 |
-
"epoch": 4.84,
|
| 1447 |
-
"learning_rate": 1.6419414657744097e-06,
|
| 1448 |
-
"loss": 0.0028,
|
| 1449 |
-
"step": 120000
|
| 1450 |
-
},
|
| 1451 |
-
{
|
| 1452 |
-
"epoch": 4.86,
|
| 1453 |
-
"learning_rate": 1.4407804563412078e-06,
|
| 1454 |
-
"loss": 0.0031,
|
| 1455 |
-
"step": 120500
|
| 1456 |
-
},
|
| 1457 |
-
{
|
| 1458 |
-
"epoch": 4.88,
|
| 1459 |
-
"learning_rate": 1.2392163186325888e-06,
|
| 1460 |
-
"loss": 0.0024,
|
| 1461 |
-
"step": 121000
|
| 1462 |
-
},
|
| 1463 |
-
{
|
| 1464 |
-
"epoch": 4.9,
|
| 1465 |
-
"learning_rate": 1.03765218092397e-06,
|
| 1466 |
-
"loss": 0.0037,
|
| 1467 |
-
"step": 121500
|
| 1468 |
-
},
|
| 1469 |
{
|
| 1470 |
"epoch": 4.92,
|
| 1471 |
-
"learning_rate":
|
| 1472 |
-
"loss": 0.
|
| 1473 |
-
"step":
|
| 1474 |
-
},
|
| 1475 |
-
{
|
| 1476 |
-
"epoch": 4.94,
|
| 1477 |
-
"learning_rate": 6.349270337821494e-07,
|
| 1478 |
-
"loss": 0.0022,
|
| 1479 |
-
"step": 122500
|
| 1480 |
},
|
| 1481 |
{
|
| 1482 |
"epoch": 4.96,
|
| 1483 |
-
"learning_rate": 4.
|
| 1484 |
-
"loss": 0.
|
| 1485 |
-
"step":
|
| 1486 |
-
},
|
| 1487 |
-
{
|
| 1488 |
-
"epoch": 4.98,
|
| 1489 |
-
"learning_rate": 2.317987583649117e-07,
|
| 1490 |
-
"loss": 0.0029,
|
| 1491 |
-
"step": 123500
|
| 1492 |
},
|
| 1493 |
{
|
| 1494 |
"epoch": 5.0,
|
| 1495 |
-
"learning_rate":
|
| 1496 |
-
"loss": 0.
|
| 1497 |
-
"step":
|
| 1498 |
},
|
| 1499 |
{
|
| 1500 |
"epoch": 5.0,
|
| 1501 |
-
"step":
|
| 1502 |
-
"total_flos":
|
| 1503 |
-
"train_loss": 0.
|
| 1504 |
-
"train_runtime":
|
| 1505 |
-
"train_samples_per_second":
|
| 1506 |
-
"train_steps_per_second":
|
| 1507 |
}
|
| 1508 |
],
|
| 1509 |
"logging_steps": 500,
|
| 1510 |
-
"max_steps":
|
| 1511 |
"num_train_epochs": 5,
|
| 1512 |
"save_steps": 500,
|
| 1513 |
-
"total_flos":
|
| 1514 |
"trial_name": null,
|
| 1515 |
"trial_params": null
|
| 1516 |
}
|
|
|
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
"epoch": 5.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 69045,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.02,
|
| 13 |
+
"learning_rate": 4.9820769063654144e-05,
|
| 14 |
+
"loss": 1.7116,
|
| 15 |
"step": 500
|
| 16 |
},
|
| 17 |
{
|
| 18 |
"epoch": 0.04,
|
| 19 |
+
"learning_rate": 4.9639727713809836e-05,
|
| 20 |
+
"loss": 0.8933,
|
| 21 |
"step": 1000
|
| 22 |
},
|
| 23 |
{
|
| 24 |
+
"epoch": 0.05,
|
| 25 |
+
"learning_rate": 4.9458686363965534e-05,
|
| 26 |
+
"loss": 0.7493,
|
| 27 |
"step": 1500
|
| 28 |
},
|
| 29 |
{
|
| 30 |
+
"epoch": 0.07,
|
| 31 |
+
"learning_rate": 4.927800709682092e-05,
|
| 32 |
+
"loss": 0.718,
|
| 33 |
"step": 2000
|
| 34 |
},
|
| 35 |
{
|
| 36 |
+
"epoch": 0.09,
|
| 37 |
+
"learning_rate": 4.9096965746976617e-05,
|
| 38 |
+
"loss": 0.6011,
|
| 39 |
"step": 2500
|
| 40 |
},
|
| 41 |
{
|
| 42 |
+
"epoch": 0.11,
|
| 43 |
+
"learning_rate": 4.891592439713231e-05,
|
| 44 |
+
"loss": 0.5649,
|
| 45 |
"step": 3000
|
| 46 |
},
|
| 47 |
{
|
| 48 |
+
"epoch": 0.13,
|
| 49 |
+
"learning_rate": 4.8734883047288006e-05,
|
| 50 |
+
"loss": 0.5612,
|
| 51 |
"step": 3500
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.14,
|
| 55 |
+
"learning_rate": 4.85538416974437e-05,
|
| 56 |
+
"loss": 0.5292,
|
| 57 |
"step": 4000
|
| 58 |
},
|
| 59 |
{
|
| 60 |
+
"epoch": 0.16,
|
| 61 |
+
"learning_rate": 4.8372800347599396e-05,
|
| 62 |
+
"loss": 0.4878,
|
| 63 |
"step": 4500
|
| 64 |
},
|
| 65 |
{
|
| 66 |
+
"epoch": 0.18,
|
| 67 |
+
"learning_rate": 4.819175899775509e-05,
|
| 68 |
+
"loss": 0.5319,
|
| 69 |
"step": 5000
|
| 70 |
},
|
| 71 |
{
|
| 72 |
+
"epoch": 0.2,
|
| 73 |
+
"learning_rate": 4.8010717647910786e-05,
|
| 74 |
+
"loss": 0.4588,
|
| 75 |
"step": 5500
|
| 76 |
},
|
| 77 |
{
|
| 78 |
+
"epoch": 0.22,
|
| 79 |
+
"learning_rate": 4.782967629806648e-05,
|
| 80 |
+
"loss": 0.4249,
|
| 81 |
"step": 6000
|
| 82 |
},
|
| 83 |
{
|
| 84 |
+
"epoch": 0.24,
|
| 85 |
+
"learning_rate": 4.7648634948222176e-05,
|
| 86 |
+
"loss": 0.4334,
|
| 87 |
"step": 6500
|
| 88 |
},
|
| 89 |
{
|
| 90 |
+
"epoch": 0.25,
|
| 91 |
+
"learning_rate": 4.7467593598377874e-05,
|
| 92 |
+
"loss": 0.4356,
|
| 93 |
"step": 7000
|
| 94 |
},
|
| 95 |
{
|
| 96 |
+
"epoch": 0.27,
|
| 97 |
+
"learning_rate": 4.7286552248533566e-05,
|
| 98 |
+
"loss": 0.4183,
|
| 99 |
"step": 7500
|
| 100 |
},
|
| 101 |
{
|
| 102 |
+
"epoch": 0.29,
|
| 103 |
+
"learning_rate": 4.7105510898689264e-05,
|
| 104 |
+
"loss": 0.3894,
|
| 105 |
"step": 8000
|
| 106 |
},
|
| 107 |
{
|
| 108 |
+
"epoch": 0.31,
|
| 109 |
+
"learning_rate": 4.6924469548844955e-05,
|
| 110 |
+
"loss": 0.3652,
|
| 111 |
"step": 8500
|
| 112 |
},
|
| 113 |
{
|
| 114 |
+
"epoch": 0.33,
|
| 115 |
+
"learning_rate": 4.6743428199000654e-05,
|
| 116 |
+
"loss": 0.3811,
|
| 117 |
"step": 9000
|
| 118 |
},
|
| 119 |
{
|
| 120 |
+
"epoch": 0.34,
|
| 121 |
+
"learning_rate": 4.6562386849156345e-05,
|
| 122 |
+
"loss": 0.3613,
|
| 123 |
"step": 9500
|
| 124 |
},
|
| 125 |
{
|
| 126 |
+
"epoch": 0.36,
|
| 127 |
+
"learning_rate": 4.6381707582011736e-05,
|
| 128 |
+
"loss": 0.3654,
|
| 129 |
"step": 10000
|
| 130 |
},
|
| 131 |
{
|
| 132 |
+
"epoch": 0.38,
|
| 133 |
+
"learning_rate": 4.620066623216743e-05,
|
| 134 |
+
"loss": 0.3379,
|
| 135 |
"step": 10500
|
| 136 |
},
|
| 137 |
{
|
| 138 |
+
"epoch": 0.4,
|
| 139 |
+
"learning_rate": 4.6019624882323126e-05,
|
| 140 |
+
"loss": 0.3438,
|
| 141 |
"step": 11000
|
| 142 |
},
|
| 143 |
{
|
| 144 |
+
"epoch": 0.42,
|
| 145 |
+
"learning_rate": 4.583858353247882e-05,
|
| 146 |
+
"loss": 0.3448,
|
| 147 |
"step": 11500
|
| 148 |
},
|
| 149 |
{
|
| 150 |
+
"epoch": 0.43,
|
| 151 |
+
"learning_rate": 4.5657542182634516e-05,
|
| 152 |
+
"loss": 0.3283,
|
| 153 |
"step": 12000
|
| 154 |
},
|
| 155 |
{
|
| 156 |
+
"epoch": 0.45,
|
| 157 |
+
"learning_rate": 4.5476500832790214e-05,
|
| 158 |
+
"loss": 0.3169,
|
| 159 |
"step": 12500
|
| 160 |
},
|
| 161 |
{
|
| 162 |
+
"epoch": 0.47,
|
| 163 |
+
"learning_rate": 4.5295459482945906e-05,
|
| 164 |
+
"loss": 0.3153,
|
| 165 |
"step": 13000
|
| 166 |
},
|
| 167 |
{
|
| 168 |
+
"epoch": 0.49,
|
| 169 |
+
"learning_rate": 4.5114418133101604e-05,
|
| 170 |
+
"loss": 0.3148,
|
| 171 |
"step": 13500
|
| 172 |
},
|
| 173 |
{
|
| 174 |
+
"epoch": 0.51,
|
| 175 |
+
"learning_rate": 4.4933376783257296e-05,
|
| 176 |
+
"loss": 0.3144,
|
| 177 |
"step": 14000
|
| 178 |
},
|
| 179 |
{
|
| 180 |
+
"epoch": 0.53,
|
| 181 |
+
"learning_rate": 4.4752335433412994e-05,
|
| 182 |
+
"loss": 0.3019,
|
| 183 |
"step": 14500
|
| 184 |
},
|
| 185 |
{
|
| 186 |
+
"epoch": 0.54,
|
| 187 |
+
"learning_rate": 4.4571294083568686e-05,
|
| 188 |
+
"loss": 0.3035,
|
| 189 |
"step": 15000
|
| 190 |
},
|
| 191 |
{
|
| 192 |
+
"epoch": 0.56,
|
| 193 |
+
"learning_rate": 4.4390252733724384e-05,
|
| 194 |
+
"loss": 0.2955,
|
| 195 |
"step": 15500
|
| 196 |
},
|
| 197 |
{
|
| 198 |
+
"epoch": 0.58,
|
| 199 |
+
"learning_rate": 4.4209211383880075e-05,
|
| 200 |
+
"loss": 0.2821,
|
| 201 |
"step": 16000
|
| 202 |
},
|
| 203 |
{
|
| 204 |
+
"epoch": 0.6,
|
| 205 |
+
"learning_rate": 4.4028170034035774e-05,
|
| 206 |
+
"loss": 0.2722,
|
| 207 |
"step": 16500
|
| 208 |
},
|
| 209 |
{
|
| 210 |
+
"epoch": 0.62,
|
| 211 |
+
"learning_rate": 4.384712868419147e-05,
|
| 212 |
+
"loss": 0.2771,
|
| 213 |
"step": 17000
|
| 214 |
},
|
| 215 |
{
|
| 216 |
+
"epoch": 0.63,
|
| 217 |
+
"learning_rate": 4.3666449417046856e-05,
|
| 218 |
+
"loss": 0.2767,
|
| 219 |
"step": 17500
|
| 220 |
},
|
| 221 |
{
|
| 222 |
+
"epoch": 0.65,
|
| 223 |
+
"learning_rate": 4.348540806720255e-05,
|
| 224 |
+
"loss": 0.2724,
|
| 225 |
"step": 18000
|
| 226 |
},
|
| 227 |
{
|
| 228 |
+
"epoch": 0.67,
|
| 229 |
+
"learning_rate": 4.3304366717358246e-05,
|
| 230 |
+
"loss": 0.2537,
|
| 231 |
"step": 18500
|
| 232 |
},
|
| 233 |
{
|
| 234 |
+
"epoch": 0.69,
|
| 235 |
+
"learning_rate": 4.3123325367513944e-05,
|
| 236 |
+
"loss": 0.2591,
|
| 237 |
"step": 19000
|
| 238 |
},
|
| 239 |
{
|
| 240 |
+
"epoch": 0.71,
|
| 241 |
+
"learning_rate": 4.2942284017669636e-05,
|
| 242 |
+
"loss": 0.2573,
|
| 243 |
"step": 19500
|
| 244 |
},
|
| 245 |
{
|
| 246 |
+
"epoch": 0.72,
|
| 247 |
+
"learning_rate": 4.2761242667825334e-05,
|
| 248 |
+
"loss": 0.2586,
|
| 249 |
"step": 20000
|
| 250 |
},
|
| 251 |
{
|
| 252 |
+
"epoch": 0.74,
|
| 253 |
+
"learning_rate": 4.2580201317981026e-05,
|
| 254 |
+
"loss": 0.2514,
|
| 255 |
"step": 20500
|
| 256 |
},
|
| 257 |
{
|
| 258 |
+
"epoch": 0.76,
|
| 259 |
+
"learning_rate": 4.2399159968136724e-05,
|
| 260 |
+
"loss": 0.2403,
|
| 261 |
"step": 21000
|
| 262 |
},
|
| 263 |
{
|
| 264 |
+
"epoch": 0.78,
|
| 265 |
+
"learning_rate": 4.2218118618292416e-05,
|
| 266 |
+
"loss": 0.2443,
|
| 267 |
"step": 21500
|
| 268 |
},
|
| 269 |
{
|
| 270 |
+
"epoch": 0.8,
|
| 271 |
+
"learning_rate": 4.203780143384749e-05,
|
| 272 |
+
"loss": 0.2393,
|
| 273 |
"step": 22000
|
| 274 |
},
|
| 275 |
{
|
| 276 |
+
"epoch": 0.81,
|
| 277 |
+
"learning_rate": 4.185676008400319e-05,
|
| 278 |
+
"loss": 0.2301,
|
| 279 |
"step": 22500
|
| 280 |
},
|
| 281 |
{
|
| 282 |
+
"epoch": 0.83,
|
| 283 |
+
"learning_rate": 4.167571873415889e-05,
|
| 284 |
+
"loss": 0.2324,
|
| 285 |
"step": 23000
|
| 286 |
},
|
| 287 |
{
|
| 288 |
+
"epoch": 0.85,
|
| 289 |
+
"learning_rate": 4.149467738431458e-05,
|
| 290 |
+
"loss": 0.2353,
|
| 291 |
"step": 23500
|
| 292 |
},
|
| 293 |
{
|
| 294 |
+
"epoch": 0.87,
|
| 295 |
+
"learning_rate": 4.131363603447028e-05,
|
| 296 |
+
"loss": 0.2248,
|
| 297 |
"step": 24000
|
| 298 |
},
|
| 299 |
{
|
| 300 |
+
"epoch": 0.89,
|
| 301 |
+
"learning_rate": 4.113259468462597e-05,
|
| 302 |
+
"loss": 0.2207,
|
| 303 |
"step": 24500
|
| 304 |
},
|
| 305 |
{
|
| 306 |
+
"epoch": 0.91,
|
| 307 |
+
"learning_rate": 4.095155333478167e-05,
|
| 308 |
+
"loss": 0.2178,
|
| 309 |
"step": 25000
|
| 310 |
},
|
| 311 |
{
|
| 312 |
+
"epoch": 0.92,
|
| 313 |
+
"learning_rate": 4.077051198493736e-05,
|
| 314 |
+
"loss": 0.2048,
|
| 315 |
"step": 25500
|
| 316 |
},
|
| 317 |
{
|
| 318 |
+
"epoch": 0.94,
|
| 319 |
+
"learning_rate": 4.0589832717792745e-05,
|
| 320 |
+
"loss": 0.2163,
|
| 321 |
"step": 26000
|
| 322 |
},
|
| 323 |
{
|
| 324 |
+
"epoch": 0.96,
|
| 325 |
+
"learning_rate": 4.040879136794844e-05,
|
| 326 |
+
"loss": 0.2139,
|
| 327 |
"step": 26500
|
| 328 |
},
|
| 329 |
{
|
| 330 |
+
"epoch": 0.98,
|
| 331 |
+
"learning_rate": 4.0227750018104134e-05,
|
| 332 |
+
"loss": 0.2135,
|
| 333 |
"step": 27000
|
| 334 |
},
|
| 335 |
{
|
| 336 |
+
"epoch": 1.0,
|
| 337 |
+
"learning_rate": 4.004670866825983e-05,
|
| 338 |
+
"loss": 0.1986,
|
| 339 |
"step": 27500
|
| 340 |
},
|
| 341 |
{
|
| 342 |
+
"epoch": 1.01,
|
| 343 |
+
"learning_rate": 3.986566731841553e-05,
|
| 344 |
+
"loss": 0.1597,
|
| 345 |
"step": 28000
|
| 346 |
},
|
| 347 |
{
|
| 348 |
+
"epoch": 1.03,
|
| 349 |
+
"learning_rate": 3.968462596857123e-05,
|
| 350 |
+
"loss": 0.154,
|
| 351 |
"step": 28500
|
| 352 |
},
|
| 353 |
{
|
| 354 |
+
"epoch": 1.05,
|
| 355 |
+
"learning_rate": 3.950358461872692e-05,
|
| 356 |
+
"loss": 0.1381,
|
| 357 |
"step": 29000
|
| 358 |
},
|
| 359 |
{
|
| 360 |
+
"epoch": 1.07,
|
| 361 |
+
"learning_rate": 3.932326743428199e-05,
|
| 362 |
+
"loss": 0.1525,
|
| 363 |
"step": 29500
|
| 364 |
},
|
| 365 |
{
|
| 366 |
+
"epoch": 1.09,
|
| 367 |
+
"learning_rate": 3.914222608443769e-05,
|
| 368 |
+
"loss": 0.143,
|
| 369 |
"step": 30000
|
| 370 |
},
|
| 371 |
{
|
| 372 |
+
"epoch": 1.1,
|
| 373 |
+
"learning_rate": 3.896118473459338e-05,
|
| 374 |
+
"loss": 0.1422,
|
| 375 |
"step": 30500
|
| 376 |
},
|
| 377 |
{
|
| 378 |
+
"epoch": 1.12,
|
| 379 |
+
"learning_rate": 3.878014338474908e-05,
|
| 380 |
+
"loss": 0.1348,
|
| 381 |
"step": 31000
|
| 382 |
},
|
| 383 |
{
|
| 384 |
+
"epoch": 1.14,
|
| 385 |
+
"learning_rate": 3.859910203490477e-05,
|
| 386 |
+
"loss": 0.1421,
|
| 387 |
"step": 31500
|
| 388 |
},
|
| 389 |
{
|
| 390 |
+
"epoch": 1.16,
|
| 391 |
+
"learning_rate": 3.841806068506047e-05,
|
| 392 |
+
"loss": 0.1438,
|
| 393 |
"step": 32000
|
| 394 |
},
|
| 395 |
{
|
| 396 |
+
"epoch": 1.18,
|
| 397 |
+
"learning_rate": 3.823701933521617e-05,
|
| 398 |
+
"loss": 0.1375,
|
| 399 |
"step": 32500
|
| 400 |
},
|
| 401 |
{
|
| 402 |
+
"epoch": 1.19,
|
| 403 |
+
"learning_rate": 3.805597798537186e-05,
|
| 404 |
+
"loss": 0.1418,
|
| 405 |
"step": 33000
|
| 406 |
},
|
| 407 |
{
|
| 408 |
+
"epoch": 1.21,
|
| 409 |
+
"learning_rate": 3.787493663552756e-05,
|
| 410 |
+
"loss": 0.1336,
|
| 411 |
"step": 33500
|
| 412 |
},
|
| 413 |
{
|
| 414 |
+
"epoch": 1.23,
|
| 415 |
+
"learning_rate": 3.769425736838294e-05,
|
| 416 |
+
"loss": 0.1339,
|
| 417 |
"step": 34000
|
| 418 |
},
|
| 419 |
{
|
| 420 |
+
"epoch": 1.25,
|
| 421 |
+
"learning_rate": 3.751321601853864e-05,
|
| 422 |
+
"loss": 0.1383,
|
| 423 |
"step": 34500
|
| 424 |
},
|
| 425 |
{
|
| 426 |
+
"epoch": 1.27,
|
| 427 |
+
"learning_rate": 3.733217466869433e-05,
|
| 428 |
+
"loss": 0.1298,
|
| 429 |
"step": 35000
|
| 430 |
},
|
| 431 |
{
|
| 432 |
+
"epoch": 1.29,
|
| 433 |
+
"learning_rate": 3.715113331885003e-05,
|
| 434 |
+
"loss": 0.1314,
|
| 435 |
"step": 35500
|
| 436 |
},
|
| 437 |
{
|
| 438 |
+
"epoch": 1.3,
|
| 439 |
+
"learning_rate": 3.697009196900572e-05,
|
| 440 |
+
"loss": 0.1249,
|
| 441 |
"step": 36000
|
| 442 |
},
|
| 443 |
{
|
| 444 |
+
"epoch": 1.32,
|
| 445 |
+
"learning_rate": 3.678905061916142e-05,
|
| 446 |
+
"loss": 0.1289,
|
| 447 |
"step": 36500
|
| 448 |
},
|
| 449 |
{
|
| 450 |
+
"epoch": 1.34,
|
| 451 |
+
"learning_rate": 3.660800926931711e-05,
|
| 452 |
+
"loss": 0.1289,
|
| 453 |
"step": 37000
|
| 454 |
},
|
| 455 |
{
|
| 456 |
+
"epoch": 1.36,
|
| 457 |
+
"learning_rate": 3.642696791947281e-05,
|
| 458 |
+
"loss": 0.1266,
|
| 459 |
"step": 37500
|
| 460 |
},
|
| 461 |
{
|
| 462 |
+
"epoch": 1.38,
|
| 463 |
+
"learning_rate": 3.62459265696285e-05,
|
| 464 |
+
"loss": 0.1201,
|
| 465 |
"step": 38000
|
| 466 |
},
|
| 467 |
{
|
| 468 |
+
"epoch": 1.39,
|
| 469 |
+
"learning_rate": 3.606524730248389e-05,
|
| 470 |
+
"loss": 0.1224,
|
| 471 |
"step": 38500
|
| 472 |
},
|
| 473 |
{
|
| 474 |
+
"epoch": 1.41,
|
| 475 |
+
"learning_rate": 3.5884205952639583e-05,
|
| 476 |
+
"loss": 0.1191,
|
| 477 |
"step": 39000
|
| 478 |
},
|
| 479 |
{
|
| 480 |
+
"epoch": 1.43,
|
| 481 |
+
"learning_rate": 3.570316460279528e-05,
|
| 482 |
+
"loss": 0.1157,
|
| 483 |
"step": 39500
|
| 484 |
},
|
| 485 |
{
|
| 486 |
+
"epoch": 1.45,
|
| 487 |
+
"learning_rate": 3.552212325295098e-05,
|
| 488 |
+
"loss": 0.1167,
|
| 489 |
"step": 40000
|
| 490 |
},
|
| 491 |
{
|
| 492 |
+
"epoch": 1.47,
|
| 493 |
+
"learning_rate": 3.534144398580636e-05,
|
| 494 |
+
"loss": 0.119,
|
| 495 |
"step": 40500
|
| 496 |
},
|
| 497 |
{
|
| 498 |
+
"epoch": 1.48,
|
| 499 |
+
"learning_rate": 3.5160402635962056e-05,
|
| 500 |
+
"loss": 0.1092,
|
| 501 |
"step": 41000
|
| 502 |
},
|
| 503 |
{
|
| 504 |
+
"epoch": 1.5,
|
| 505 |
+
"learning_rate": 3.4979361286117754e-05,
|
| 506 |
+
"loss": 0.1142,
|
| 507 |
"step": 41500
|
| 508 |
},
|
| 509 |
{
|
| 510 |
+
"epoch": 1.52,
|
| 511 |
+
"learning_rate": 3.479831993627345e-05,
|
| 512 |
+
"loss": 0.1156,
|
| 513 |
"step": 42000
|
| 514 |
},
|
| 515 |
{
|
| 516 |
+
"epoch": 1.54,
|
| 517 |
+
"learning_rate": 3.4617278586429144e-05,
|
| 518 |
+
"loss": 0.1122,
|
| 519 |
"step": 42500
|
| 520 |
},
|
| 521 |
{
|
| 522 |
+
"epoch": 1.56,
|
| 523 |
+
"learning_rate": 3.443623723658484e-05,
|
| 524 |
+
"loss": 0.1104,
|
| 525 |
"step": 43000
|
| 526 |
},
|
| 527 |
{
|
| 528 |
+
"epoch": 1.58,
|
| 529 |
+
"learning_rate": 3.4255195886740534e-05,
|
| 530 |
+
"loss": 0.1076,
|
| 531 |
"step": 43500
|
| 532 |
},
|
| 533 |
{
|
| 534 |
+
"epoch": 1.59,
|
| 535 |
+
"learning_rate": 3.407415453689623e-05,
|
| 536 |
+
"loss": 0.1096,
|
| 537 |
"step": 44000
|
| 538 |
},
|
| 539 |
{
|
| 540 |
+
"epoch": 1.61,
|
| 541 |
+
"learning_rate": 3.3893113187051924e-05,
|
| 542 |
+
"loss": 0.1094,
|
| 543 |
"step": 44500
|
| 544 |
},
|
| 545 |
{
|
| 546 |
+
"epoch": 1.63,
|
| 547 |
+
"learning_rate": 3.371243391990731e-05,
|
| 548 |
+
"loss": 0.105,
|
| 549 |
"step": 45000
|
| 550 |
},
|
| 551 |
{
|
| 552 |
+
"epoch": 1.65,
|
| 553 |
+
"learning_rate": 3.3531392570063e-05,
|
| 554 |
+
"loss": 0.1059,
|
| 555 |
"step": 45500
|
| 556 |
},
|
| 557 |
{
|
| 558 |
+
"epoch": 1.67,
|
| 559 |
+
"learning_rate": 3.33503512202187e-05,
|
| 560 |
+
"loss": 0.1012,
|
| 561 |
"step": 46000
|
| 562 |
},
|
| 563 |
{
|
| 564 |
+
"epoch": 1.68,
|
| 565 |
+
"learning_rate": 3.3169309870374396e-05,
|
| 566 |
+
"loss": 0.0991,
|
| 567 |
"step": 46500
|
| 568 |
},
|
| 569 |
{
|
| 570 |
+
"epoch": 1.7,
|
| 571 |
+
"learning_rate": 3.2988268520530094e-05,
|
| 572 |
+
"loss": 0.0978,
|
| 573 |
"step": 47000
|
| 574 |
},
|
| 575 |
{
|
| 576 |
+
"epoch": 1.72,
|
| 577 |
+
"learning_rate": 3.2807227170685786e-05,
|
| 578 |
+
"loss": 0.1003,
|
| 579 |
"step": 47500
|
| 580 |
},
|
| 581 |
{
|
| 582 |
+
"epoch": 1.74,
|
| 583 |
+
"learning_rate": 3.2626185820841484e-05,
|
| 584 |
+
"loss": 0.1,
|
| 585 |
"step": 48000
|
| 586 |
},
|
| 587 |
{
|
| 588 |
+
"epoch": 1.76,
|
| 589 |
+
"learning_rate": 3.244550655369686e-05,
|
| 590 |
+
"loss": 0.0976,
|
| 591 |
"step": 48500
|
| 592 |
},
|
| 593 |
{
|
| 594 |
+
"epoch": 1.77,
|
| 595 |
+
"learning_rate": 3.226446520385256e-05,
|
| 596 |
+
"loss": 0.0961,
|
| 597 |
"step": 49000
|
| 598 |
},
|
| 599 |
{
|
| 600 |
+
"epoch": 1.79,
|
| 601 |
+
"learning_rate": 3.208342385400825e-05,
|
| 602 |
+
"loss": 0.1011,
|
| 603 |
"step": 49500
|
| 604 |
},
|
| 605 |
{
|
| 606 |
+
"epoch": 1.81,
|
| 607 |
+
"learning_rate": 3.190238250416395e-05,
|
| 608 |
+
"loss": 0.0902,
|
| 609 |
"step": 50000
|
| 610 |
},
|
| 611 |
{
|
| 612 |
+
"epoch": 1.83,
|
| 613 |
+
"learning_rate": 3.1721703237019334e-05,
|
| 614 |
+
"loss": 0.0936,
|
| 615 |
"step": 50500
|
| 616 |
},
|
| 617 |
{
|
| 618 |
+
"epoch": 1.85,
|
| 619 |
+
"learning_rate": 3.154066188717503e-05,
|
| 620 |
+
"loss": 0.0903,
|
| 621 |
"step": 51000
|
| 622 |
},
|
| 623 |
{
|
| 624 |
+
"epoch": 1.86,
|
| 625 |
+
"learning_rate": 3.135962053733073e-05,
|
| 626 |
+
"loss": 0.082,
|
| 627 |
"step": 51500
|
| 628 |
},
|
| 629 |
{
|
| 630 |
+
"epoch": 1.88,
|
| 631 |
+
"learning_rate": 3.1178941270186115e-05,
|
| 632 |
+
"loss": 0.0841,
|
| 633 |
"step": 52000
|
| 634 |
},
|
| 635 |
{
|
| 636 |
+
"epoch": 1.9,
|
| 637 |
+
"learning_rate": 3.0997899920341806e-05,
|
| 638 |
+
"loss": 0.0898,
|
| 639 |
"step": 52500
|
| 640 |
},
|
| 641 |
{
|
| 642 |
+
"epoch": 1.92,
|
| 643 |
+
"learning_rate": 3.0816858570497505e-05,
|
| 644 |
+
"loss": 0.085,
|
| 645 |
"step": 53000
|
| 646 |
},
|
| 647 |
{
|
| 648 |
+
"epoch": 1.94,
|
| 649 |
+
"learning_rate": 3.06358172206532e-05,
|
| 650 |
+
"loss": 0.0807,
|
| 651 |
"step": 53500
|
| 652 |
},
|
| 653 |
{
|
| 654 |
+
"epoch": 1.96,
|
| 655 |
+
"learning_rate": 3.0454775870808895e-05,
|
| 656 |
+
"loss": 0.0886,
|
| 657 |
"step": 54000
|
| 658 |
},
|
| 659 |
{
|
| 660 |
+
"epoch": 1.97,
|
| 661 |
+
"learning_rate": 3.0273734520964593e-05,
|
| 662 |
+
"loss": 0.0818,
|
| 663 |
"step": 54500
|
| 664 |
},
|
| 665 |
{
|
| 666 |
+
"epoch": 1.99,
|
| 667 |
+
"learning_rate": 3.0092693171120284e-05,
|
| 668 |
+
"loss": 0.0839,
|
| 669 |
"step": 55000
|
| 670 |
},
|
| 671 |
{
|
| 672 |
+
"epoch": 2.01,
|
| 673 |
+
"learning_rate": 2.9911651821275983e-05,
|
| 674 |
+
"loss": 0.0581,
|
| 675 |
"step": 55500
|
| 676 |
},
|
| 677 |
{
|
| 678 |
+
"epoch": 2.03,
|
| 679 |
+
"learning_rate": 2.9730610471431674e-05,
|
| 680 |
+
"loss": 0.0504,
|
| 681 |
"step": 56000
|
| 682 |
},
|
| 683 |
{
|
| 684 |
+
"epoch": 2.05,
|
| 685 |
+
"learning_rate": 2.9549569121587373e-05,
|
| 686 |
+
"loss": 0.0486,
|
| 687 |
"step": 56500
|
| 688 |
},
|
| 689 |
{
|
| 690 |
+
"epoch": 2.06,
|
| 691 |
+
"learning_rate": 2.9368527771743064e-05,
|
| 692 |
+
"loss": 0.0463,
|
| 693 |
"step": 57000
|
| 694 |
},
|
| 695 |
{
|
| 696 |
+
"epoch": 2.08,
|
| 697 |
+
"learning_rate": 2.9187486421898762e-05,
|
| 698 |
+
"loss": 0.0545,
|
| 699 |
"step": 57500
|
| 700 |
},
|
| 701 |
{
|
| 702 |
+
"epoch": 2.1,
|
| 703 |
+
"learning_rate": 2.900644507205446e-05,
|
| 704 |
+
"loss": 0.0482,
|
| 705 |
"step": 58000
|
| 706 |
},
|
| 707 |
{
|
| 708 |
+
"epoch": 2.12,
|
| 709 |
+
"learning_rate": 2.8825403722210152e-05,
|
| 710 |
+
"loss": 0.0431,
|
| 711 |
"step": 58500
|
| 712 |
},
|
| 713 |
{
|
| 714 |
+
"epoch": 2.14,
|
| 715 |
+
"learning_rate": 2.864436237236585e-05,
|
| 716 |
+
"loss": 0.0494,
|
| 717 |
"step": 59000
|
| 718 |
},
|
| 719 |
{
|
| 720 |
+
"epoch": 2.15,
|
| 721 |
+
"learning_rate": 2.8463683105221235e-05,
|
| 722 |
+
"loss": 0.0522,
|
| 723 |
"step": 59500
|
| 724 |
},
|
| 725 |
{
|
| 726 |
+
"epoch": 4.34,
|
| 727 |
+
"learning_rate": 6.565283510753857e-06,
|
| 728 |
+
"loss": 0.0392,
|
| 729 |
"step": 60000
|
| 730 |
},
|
| 731 |
{
|
| 732 |
+
"epoch": 4.38,
|
| 733 |
+
"learning_rate": 6.203200811065248e-06,
|
| 734 |
+
"loss": 0.0329,
|
| 735 |
"step": 60500
|
| 736 |
},
|
| 737 |
{
|
| 738 |
+
"epoch": 4.42,
|
| 739 |
+
"learning_rate": 5.841118111376639e-06,
|
| 740 |
+
"loss": 0.0306,
|
| 741 |
"step": 61000
|
| 742 |
},
|
| 743 |
{
|
| 744 |
+
"epoch": 4.45,
|
| 745 |
+
"learning_rate": 5.479759577087408e-06,
|
| 746 |
+
"loss": 0.0281,
|
| 747 |
"step": 61500
|
| 748 |
},
|
| 749 |
{
|
| 750 |
+
"epoch": 4.49,
|
| 751 |
+
"learning_rate": 5.117676877398798e-06,
|
| 752 |
+
"loss": 0.0275,
|
| 753 |
"step": 62000
|
| 754 |
},
|
| 755 |
{
|
| 756 |
+
"epoch": 4.53,
|
| 757 |
+
"learning_rate": 4.755594177710189e-06,
|
| 758 |
+
"loss": 0.0238,
|
| 759 |
"step": 62500
|
| 760 |
},
|
| 761 |
{
|
| 762 |
+
"epoch": 4.56,
|
| 763 |
+
"learning_rate": 4.394235643420958e-06,
|
| 764 |
+
"loss": 0.0253,
|
| 765 |
"step": 63000
|
| 766 |
},
|
| 767 |
{
|
| 768 |
+
"epoch": 4.6,
|
| 769 |
+
"learning_rate": 4.032152943732348e-06,
|
| 770 |
+
"loss": 0.0222,
|
| 771 |
"step": 63500
|
| 772 |
},
|
| 773 |
{
|
| 774 |
+
"epoch": 4.63,
|
| 775 |
+
"learning_rate": 3.67007024404374e-06,
|
| 776 |
+
"loss": 0.0234,
|
| 777 |
"step": 64000
|
| 778 |
},
|
| 779 |
{
|
| 780 |
+
"epoch": 4.67,
|
| 781 |
+
"learning_rate": 3.307987544355131e-06,
|
| 782 |
+
"loss": 0.0214,
|
| 783 |
"step": 64500
|
| 784 |
},
|
| 785 |
{
|
| 786 |
+
"epoch": 4.71,
|
| 787 |
+
"learning_rate": 2.945904844666522e-06,
|
| 788 |
+
"loss": 0.0217,
|
| 789 |
"step": 65000
|
| 790 |
},
|
| 791 |
{
|
| 792 |
+
"epoch": 4.74,
|
| 793 |
+
"learning_rate": 2.583822144977913e-06,
|
| 794 |
+
"loss": 0.022,
|
| 795 |
"step": 65500
|
| 796 |
},
|
| 797 |
{
|
| 798 |
+
"epoch": 4.78,
|
| 799 |
+
"learning_rate": 2.221739445289304e-06,
|
| 800 |
+
"loss": 0.0213,
|
| 801 |
"step": 66000
|
| 802 |
},
|
| 803 |
{
|
| 804 |
+
"epoch": 4.82,
|
| 805 |
+
"learning_rate": 1.8596567456006953e-06,
|
| 806 |
+
"loss": 0.0216,
|
| 807 |
"step": 66500
|
| 808 |
},
|
| 809 |
{
|
| 810 |
+
"epoch": 4.85,
|
| 811 |
+
"learning_rate": 1.4975740459120864e-06,
|
| 812 |
+
"loss": 0.0209,
|
| 813 |
"step": 67000
|
| 814 |
},
|
| 815 |
{
|
| 816 |
+
"epoch": 4.89,
|
| 817 |
+
"learning_rate": 1.1362155116228547e-06,
|
| 818 |
+
"loss": 0.0193,
|
| 819 |
"step": 67500
|
| 820 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 821 |
{
|
| 822 |
"epoch": 4.92,
|
| 823 |
+
"learning_rate": 7.748569773336231e-07,
|
| 824 |
+
"loss": 0.0199,
|
| 825 |
+
"step": 68000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 826 |
},
|
| 827 |
{
|
| 828 |
"epoch": 4.96,
|
| 829 |
+
"learning_rate": 4.1349844304439136e-07,
|
| 830 |
+
"loss": 0.0183,
|
| 831 |
+
"step": 68500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 832 |
},
|
| 833 |
{
|
| 834 |
"epoch": 5.0,
|
| 835 |
+
"learning_rate": 5.141574335578247e-08,
|
| 836 |
+
"loss": 0.0212,
|
| 837 |
+
"step": 69000
|
| 838 |
},
|
| 839 |
{
|
| 840 |
"epoch": 5.0,
|
| 841 |
+
"step": 69045,
|
| 842 |
+
"total_flos": 8.175381920769638e+16,
|
| 843 |
+
"train_loss": 0.003351212132286996,
|
| 844 |
+
"train_runtime": 2726.0601,
|
| 845 |
+
"train_samples_per_second": 405.237,
|
| 846 |
+
"train_steps_per_second": 25.328
|
| 847 |
}
|
| 848 |
],
|
| 849 |
"logging_steps": 500,
|
| 850 |
+
"max_steps": 69045,
|
| 851 |
"num_train_epochs": 5,
|
| 852 |
"save_steps": 500,
|
| 853 |
+
"total_flos": 8.175381920769638e+16,
|
| 854 |
"trial_name": null,
|
| 855 |
"trial_params": null
|
| 856 |
}
|