Commit
·
67893cd
1
Parent(s):
1e5ab1d
Upload 8 files
Browse files- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +380 -932
- training_args.bin +1 -1
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 33661637
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c799f21594ab06d6b91129c311773ce342760d0c02360ffb01ea6208e4faace8
|
| 3 |
size 33661637
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14575
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f329b74b9ae80cb5e6b64ce551d508321a3549bc3d469325a37753fa7dd59f60
|
| 3 |
size 14575
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 627
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:57ad5ba60863d9ee4b493aebecbb69b4c26f9bad3b815096052e17626ba0bb35
|
| 3 |
size 627
|
trainer_state.json
CHANGED
|
@@ -1,1248 +1,696 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric":
|
| 3 |
-
"best_model_checkpoint": "experiments/checkpoint-
|
| 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":
|
| 13 |
-
"loss": 1.
|
| 14 |
"step": 5
|
| 15 |
},
|
| 16 |
{
|
| 17 |
-
"epoch": 0
|
| 18 |
-
"learning_rate":
|
| 19 |
-
"loss": 1.
|
| 20 |
"step": 10
|
| 21 |
},
|
| 22 |
{
|
| 23 |
-
"epoch": 0
|
| 24 |
-
"learning_rate":
|
| 25 |
-
"loss": 1.
|
| 26 |
"step": 15
|
| 27 |
},
|
| 28 |
{
|
| 29 |
-
"epoch": 0
|
| 30 |
-
"learning_rate": 1.
|
| 31 |
-
"loss": 1.
|
| 32 |
"step": 20
|
| 33 |
},
|
| 34 |
{
|
| 35 |
-
"epoch": 0
|
| 36 |
-
"learning_rate": 1.
|
| 37 |
-
"loss": 1.
|
| 38 |
"step": 25
|
| 39 |
},
|
| 40 |
{
|
| 41 |
-
"epoch": 0
|
| 42 |
-
"learning_rate": 1.
|
| 43 |
-
"loss": 1.
|
| 44 |
"step": 30
|
| 45 |
},
|
| 46 |
{
|
| 47 |
-
"epoch": 0
|
| 48 |
-
"learning_rate":
|
| 49 |
-
"loss": 1.
|
| 50 |
"step": 35
|
| 51 |
},
|
| 52 |
{
|
| 53 |
-
"epoch": 0
|
| 54 |
-
"learning_rate": 2.
|
| 55 |
-
"loss": 1.
|
| 56 |
"step": 40
|
| 57 |
},
|
| 58 |
{
|
| 59 |
-
"epoch": 0
|
| 60 |
-
"learning_rate": 2.
|
| 61 |
-
"loss": 1.
|
| 62 |
"step": 45
|
| 63 |
},
|
| 64 |
{
|
| 65 |
-
"epoch": 0
|
| 66 |
-
"learning_rate": 2.
|
| 67 |
-
"loss": 1.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 68 |
"step": 50
|
| 69 |
},
|
| 70 |
{
|
| 71 |
-
"epoch": 0
|
| 72 |
-
"learning_rate": 3.
|
| 73 |
-
"loss": 1.
|
| 74 |
"step": 55
|
| 75 |
},
|
| 76 |
{
|
| 77 |
-
"epoch": 0
|
| 78 |
-
"learning_rate": 3.
|
| 79 |
-
"loss": 1.
|
| 80 |
"step": 60
|
| 81 |
},
|
| 82 |
{
|
| 83 |
-
"epoch":
|
| 84 |
-
"learning_rate": 3.
|
| 85 |
-
"loss": 1.
|
| 86 |
"step": 65
|
| 87 |
},
|
| 88 |
{
|
| 89 |
-
"epoch":
|
| 90 |
-
"learning_rate": 4.
|
| 91 |
-
"loss": 1.
|
| 92 |
"step": 70
|
| 93 |
},
|
| 94 |
{
|
| 95 |
-
"epoch":
|
| 96 |
-
"learning_rate": 4.
|
| 97 |
-
"loss": 1.
|
| 98 |
"step": 75
|
| 99 |
},
|
| 100 |
{
|
| 101 |
-
"epoch":
|
| 102 |
-
"learning_rate": 4.
|
| 103 |
-
"loss":
|
| 104 |
"step": 80
|
| 105 |
},
|
| 106 |
{
|
| 107 |
-
"epoch":
|
| 108 |
-
"learning_rate":
|
| 109 |
-
"loss":
|
| 110 |
"step": 85
|
| 111 |
},
|
| 112 |
{
|
| 113 |
-
"epoch":
|
| 114 |
-
"learning_rate": 5.
|
| 115 |
-
"loss":
|
| 116 |
"step": 90
|
| 117 |
},
|
| 118 |
{
|
| 119 |
-
"epoch":
|
| 120 |
-
"learning_rate": 5.
|
| 121 |
-
"loss": 0.
|
| 122 |
"step": 95
|
| 123 |
},
|
| 124 |
{
|
| 125 |
-
"epoch":
|
| 126 |
-
"learning_rate": 5.
|
| 127 |
-
"loss": 0.
|
| 128 |
"step": 100
|
| 129 |
},
|
| 130 |
{
|
| 131 |
-
"epoch":
|
| 132 |
-
"
|
| 133 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 134 |
"step": 105
|
| 135 |
},
|
| 136 |
{
|
| 137 |
-
"epoch":
|
| 138 |
-
"learning_rate": 6.
|
| 139 |
-
"loss": 0.
|
| 140 |
"step": 110
|
| 141 |
},
|
| 142 |
{
|
| 143 |
-
"epoch":
|
| 144 |
-
"learning_rate": 6.
|
| 145 |
-
"loss": 0.
|
| 146 |
"step": 115
|
| 147 |
},
|
| 148 |
{
|
| 149 |
-
"epoch":
|
| 150 |
-
"learning_rate":
|
| 151 |
-
"loss": 0.
|
| 152 |
"step": 120
|
| 153 |
},
|
| 154 |
{
|
| 155 |
-
"epoch":
|
| 156 |
-
"learning_rate": 7.
|
| 157 |
-
"loss": 0.
|
| 158 |
"step": 125
|
| 159 |
},
|
| 160 |
{
|
| 161 |
-
"epoch":
|
| 162 |
-
"learning_rate": 7.
|
| 163 |
-
"loss": 0.
|
| 164 |
"step": 130
|
| 165 |
},
|
| 166 |
{
|
| 167 |
-
"epoch":
|
| 168 |
-
"learning_rate":
|
| 169 |
-
"loss": 0.
|
| 170 |
"step": 135
|
| 171 |
},
|
| 172 |
{
|
| 173 |
-
"epoch":
|
| 174 |
-
"learning_rate": 8.
|
| 175 |
-
"loss": 0.
|
| 176 |
"step": 140
|
| 177 |
},
|
| 178 |
{
|
| 179 |
-
"epoch":
|
| 180 |
-
"learning_rate": 8.
|
| 181 |
-
"loss": 0.
|
| 182 |
"step": 145
|
| 183 |
},
|
| 184 |
{
|
| 185 |
-
"epoch":
|
| 186 |
-
"learning_rate": 8.
|
| 187 |
-
"loss": 0.
|
| 188 |
"step": 150
|
| 189 |
},
|
| 190 |
{
|
| 191 |
-
"epoch":
|
| 192 |
-
"
|
| 193 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 194 |
"step": 155
|
| 195 |
},
|
| 196 |
{
|
| 197 |
-
"epoch":
|
| 198 |
-
"learning_rate": 9.
|
| 199 |
-
"loss": 0.
|
| 200 |
"step": 160
|
| 201 |
},
|
| 202 |
{
|
| 203 |
-
"epoch":
|
| 204 |
-
"learning_rate": 9.
|
| 205 |
-
"loss": 0.
|
| 206 |
"step": 165
|
| 207 |
},
|
| 208 |
{
|
| 209 |
-
"epoch":
|
| 210 |
-
"learning_rate":
|
| 211 |
-
"loss": 0.
|
| 212 |
"step": 170
|
| 213 |
},
|
| 214 |
{
|
| 215 |
-
"epoch":
|
| 216 |
-
"learning_rate": 0.
|
| 217 |
-
"loss": 0.
|
| 218 |
"step": 175
|
| 219 |
},
|
| 220 |
{
|
| 221 |
-
"epoch":
|
| 222 |
-
"learning_rate": 0.
|
| 223 |
-
"loss": 0.
|
| 224 |
"step": 180
|
| 225 |
},
|
| 226 |
{
|
| 227 |
-
"epoch":
|
| 228 |
-
"learning_rate": 0.
|
| 229 |
-
"loss": 0.
|
| 230 |
"step": 185
|
| 231 |
},
|
| 232 |
{
|
| 233 |
-
"epoch":
|
| 234 |
-
"learning_rate": 0.
|
| 235 |
-
"loss": 0.
|
| 236 |
"step": 190
|
| 237 |
},
|
| 238 |
{
|
| 239 |
-
"epoch":
|
| 240 |
-
"learning_rate": 0.
|
| 241 |
-
"loss": 0.
|
| 242 |
"step": 195
|
| 243 |
},
|
| 244 |
{
|
| 245 |
-
"epoch":
|
| 246 |
-
"learning_rate": 0.
|
| 247 |
-
"loss": 0.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 248 |
"step": 200
|
| 249 |
},
|
| 250 |
{
|
| 251 |
-
"epoch":
|
| 252 |
-
"learning_rate": 0.
|
| 253 |
-
"loss": 0.
|
| 254 |
"step": 205
|
| 255 |
},
|
| 256 |
{
|
| 257 |
-
"epoch":
|
| 258 |
-
"learning_rate": 0.
|
| 259 |
-
"loss": 0.
|
| 260 |
"step": 210
|
| 261 |
},
|
| 262 |
{
|
| 263 |
-
"epoch":
|
| 264 |
-
"learning_rate": 0.
|
| 265 |
-
"loss": 0.
|
| 266 |
"step": 215
|
| 267 |
},
|
| 268 |
{
|
| 269 |
-
"epoch":
|
| 270 |
-
"learning_rate": 0.
|
| 271 |
-
"loss": 0.
|
| 272 |
"step": 220
|
| 273 |
},
|
| 274 |
{
|
| 275 |
-
"epoch":
|
| 276 |
-
"learning_rate": 0.
|
| 277 |
-
"loss": 0.
|
| 278 |
"step": 225
|
| 279 |
},
|
| 280 |
{
|
| 281 |
-
"epoch":
|
| 282 |
-
"learning_rate": 0.
|
| 283 |
-
"loss": 0.
|
| 284 |
"step": 230
|
| 285 |
},
|
| 286 |
{
|
| 287 |
-
"epoch":
|
| 288 |
-
"learning_rate": 0.
|
| 289 |
-
"loss": 0.
|
| 290 |
"step": 235
|
| 291 |
},
|
| 292 |
{
|
| 293 |
-
"epoch":
|
| 294 |
-
"learning_rate": 0.
|
| 295 |
-
"loss": 0.
|
| 296 |
"step": 240
|
| 297 |
},
|
| 298 |
{
|
| 299 |
-
"epoch":
|
| 300 |
-
"learning_rate": 0.
|
| 301 |
-
"loss": 0.
|
| 302 |
"step": 245
|
| 303 |
},
|
| 304 |
{
|
| 305 |
-
"epoch":
|
| 306 |
-
"learning_rate": 0.
|
| 307 |
-
"loss": 0.
|
| 308 |
"step": 250
|
| 309 |
},
|
| 310 |
{
|
| 311 |
-
"epoch":
|
| 312 |
-
"eval_loss":
|
| 313 |
-
"eval_runtime":
|
| 314 |
-
"eval_samples_per_second": 7.
|
| 315 |
-
"eval_steps_per_second":
|
| 316 |
"step": 250
|
| 317 |
},
|
| 318 |
{
|
| 319 |
-
"epoch":
|
| 320 |
-
"learning_rate": 0.
|
| 321 |
-
"loss": 0.
|
| 322 |
"step": 255
|
| 323 |
},
|
| 324 |
{
|
| 325 |
-
"epoch":
|
| 326 |
-
"learning_rate": 0.
|
| 327 |
-
"loss": 0.
|
| 328 |
"step": 260
|
| 329 |
},
|
| 330 |
{
|
| 331 |
-
"epoch":
|
| 332 |
-
"learning_rate": 0.
|
| 333 |
-
"loss": 0.
|
| 334 |
"step": 265
|
| 335 |
},
|
| 336 |
{
|
| 337 |
-
"epoch":
|
| 338 |
-
"learning_rate": 0.
|
| 339 |
-
"loss": 0.
|
| 340 |
"step": 270
|
| 341 |
},
|
| 342 |
{
|
| 343 |
-
"epoch":
|
| 344 |
-
"learning_rate": 0.
|
| 345 |
-
"loss": 0.
|
| 346 |
"step": 275
|
| 347 |
},
|
| 348 |
{
|
| 349 |
-
"epoch":
|
| 350 |
-
"learning_rate": 0.
|
| 351 |
-
"loss": 0.
|
| 352 |
"step": 280
|
| 353 |
},
|
| 354 |
{
|
| 355 |
-
"epoch":
|
| 356 |
-
"learning_rate": 0.
|
| 357 |
-
"loss": 0.
|
| 358 |
"step": 285
|
| 359 |
},
|
| 360 |
{
|
| 361 |
-
"epoch":
|
| 362 |
-
"learning_rate": 0.
|
| 363 |
-
"loss": 0.
|
| 364 |
"step": 290
|
| 365 |
},
|
| 366 |
{
|
| 367 |
-
"epoch":
|
| 368 |
-
"learning_rate": 0.
|
| 369 |
-
"loss": 0.
|
| 370 |
"step": 295
|
| 371 |
},
|
| 372 |
{
|
| 373 |
-
"epoch":
|
| 374 |
-
"learning_rate": 0.
|
| 375 |
-
"loss": 0.
|
| 376 |
"step": 300
|
| 377 |
},
|
| 378 |
{
|
| 379 |
-
"epoch":
|
| 380 |
-
"
|
| 381 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 382 |
"step": 305
|
| 383 |
},
|
| 384 |
{
|
| 385 |
-
"epoch":
|
| 386 |
-
"learning_rate": 0.
|
| 387 |
-
"loss": 0.
|
| 388 |
"step": 310
|
| 389 |
},
|
| 390 |
{
|
| 391 |
-
"epoch":
|
| 392 |
-
"learning_rate": 0.
|
| 393 |
-
"loss": 0.
|
| 394 |
"step": 315
|
| 395 |
},
|
| 396 |
{
|
| 397 |
-
"epoch":
|
| 398 |
-
"learning_rate": 0.
|
| 399 |
-
"loss": 0.
|
| 400 |
"step": 320
|
| 401 |
},
|
| 402 |
{
|
| 403 |
-
"epoch":
|
| 404 |
-
"learning_rate": 0.
|
| 405 |
-
"loss": 0.
|
| 406 |
"step": 325
|
| 407 |
},
|
| 408 |
{
|
| 409 |
-
"epoch":
|
| 410 |
-
"learning_rate": 0.
|
| 411 |
-
"loss": 0.
|
| 412 |
"step": 330
|
| 413 |
},
|
| 414 |
{
|
| 415 |
-
"epoch":
|
| 416 |
-
"learning_rate": 0.
|
| 417 |
-
"loss": 0.
|
| 418 |
"step": 335
|
| 419 |
},
|
| 420 |
{
|
| 421 |
-
"epoch":
|
| 422 |
-
"learning_rate": 0.
|
| 423 |
-
"loss": 0.
|
| 424 |
"step": 340
|
| 425 |
},
|
| 426 |
{
|
| 427 |
-
"epoch":
|
| 428 |
-
"learning_rate": 0.
|
| 429 |
-
"loss": 0.
|
| 430 |
"step": 345
|
| 431 |
},
|
| 432 |
{
|
| 433 |
-
"epoch":
|
| 434 |
-
"learning_rate": 0.
|
| 435 |
-
"loss": 0.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 436 |
"step": 350
|
| 437 |
},
|
| 438 |
{
|
| 439 |
-
"epoch":
|
| 440 |
-
"learning_rate": 0.
|
| 441 |
-
"loss": 0.
|
| 442 |
"step": 355
|
| 443 |
},
|
| 444 |
{
|
| 445 |
-
"epoch":
|
| 446 |
-
"learning_rate": 0.
|
| 447 |
-
"loss": 0.
|
| 448 |
"step": 360
|
| 449 |
},
|
| 450 |
{
|
| 451 |
-
"epoch":
|
| 452 |
-
"learning_rate": 0.
|
| 453 |
-
"loss": 0.
|
| 454 |
"step": 365
|
| 455 |
},
|
| 456 |
{
|
| 457 |
-
"epoch":
|
| 458 |
-
"learning_rate": 0.
|
| 459 |
-
"loss": 0.
|
| 460 |
"step": 370
|
| 461 |
},
|
| 462 |
{
|
| 463 |
-
"epoch":
|
| 464 |
-
"learning_rate": 0.
|
| 465 |
-
"loss": 0.
|
| 466 |
"step": 375
|
| 467 |
},
|
| 468 |
{
|
| 469 |
-
"epoch":
|
| 470 |
-
"learning_rate": 0.
|
| 471 |
-
"loss": 0.
|
| 472 |
"step": 380
|
| 473 |
},
|
| 474 |
{
|
| 475 |
-
"epoch":
|
| 476 |
-
"learning_rate": 0.
|
| 477 |
-
"loss": 0.
|
| 478 |
"step": 385
|
| 479 |
},
|
| 480 |
{
|
| 481 |
-
"epoch":
|
| 482 |
-
"learning_rate": 0.
|
| 483 |
-
"loss": 0.
|
| 484 |
"step": 390
|
| 485 |
},
|
| 486 |
{
|
| 487 |
-
"epoch":
|
| 488 |
-
"learning_rate": 0.
|
| 489 |
-
"loss": 0.
|
| 490 |
"step": 395
|
| 491 |
},
|
| 492 |
{
|
| 493 |
-
"epoch":
|
| 494 |
-
"learning_rate": 0.
|
| 495 |
-
"loss": 0.
|
| 496 |
"step": 400
|
| 497 |
},
|
| 498 |
{
|
| 499 |
-
"epoch":
|
| 500 |
-
"
|
| 501 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 502 |
"step": 405
|
| 503 |
},
|
| 504 |
{
|
| 505 |
-
"epoch":
|
| 506 |
-
"learning_rate": 0.
|
| 507 |
-
"loss": 0.
|
| 508 |
"step": 410
|
| 509 |
},
|
| 510 |
{
|
| 511 |
-
"epoch":
|
| 512 |
-
"learning_rate": 0.
|
| 513 |
-
"loss": 0.
|
| 514 |
"step": 415
|
| 515 |
},
|
| 516 |
{
|
| 517 |
-
"epoch":
|
| 518 |
-
"learning_rate": 0.
|
| 519 |
-
"loss": 0.
|
| 520 |
"step": 420
|
| 521 |
},
|
| 522 |
{
|
| 523 |
-
"epoch":
|
| 524 |
-
"learning_rate": 0.
|
| 525 |
-
"loss": 0.
|
| 526 |
"step": 425
|
| 527 |
},
|
| 528 |
{
|
| 529 |
-
"epoch":
|
| 530 |
-
"learning_rate": 0.
|
| 531 |
-
"loss": 0.
|
| 532 |
"step": 430
|
| 533 |
},
|
| 534 |
{
|
| 535 |
-
"epoch":
|
| 536 |
-
"learning_rate": 0.
|
| 537 |
-
"loss": 0.
|
| 538 |
"step": 435
|
| 539 |
},
|
| 540 |
{
|
| 541 |
-
"epoch":
|
| 542 |
-
"learning_rate": 0.
|
| 543 |
-
"loss": 0.
|
| 544 |
"step": 440
|
| 545 |
},
|
| 546 |
{
|
| 547 |
-
"epoch":
|
| 548 |
-
"learning_rate": 0.
|
| 549 |
-
"loss": 0.
|
| 550 |
"step": 445
|
| 551 |
},
|
| 552 |
{
|
| 553 |
-
"epoch":
|
| 554 |
-
"learning_rate": 0.
|
| 555 |
-
"loss": 0.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 556 |
"step": 450
|
| 557 |
},
|
| 558 |
{
|
| 559 |
-
"epoch":
|
| 560 |
-
"learning_rate": 0.
|
| 561 |
-
"loss": 0.
|
| 562 |
"step": 455
|
| 563 |
},
|
| 564 |
{
|
| 565 |
-
"epoch":
|
| 566 |
-
"learning_rate": 0.
|
| 567 |
-
"loss": 0.
|
| 568 |
"step": 460
|
| 569 |
},
|
| 570 |
{
|
| 571 |
-
"epoch":
|
| 572 |
-
"learning_rate": 0.
|
| 573 |
-
"loss": 0.
|
| 574 |
"step": 465
|
| 575 |
},
|
| 576 |
{
|
| 577 |
-
"epoch":
|
| 578 |
-
"learning_rate": 0.
|
| 579 |
-
"loss": 0.
|
| 580 |
"step": 470
|
| 581 |
},
|
| 582 |
{
|
| 583 |
-
"epoch":
|
| 584 |
-
"learning_rate": 0.
|
| 585 |
-
"loss": 0.
|
| 586 |
"step": 475
|
| 587 |
},
|
| 588 |
{
|
| 589 |
-
"epoch":
|
| 590 |
-
"learning_rate": 0.
|
| 591 |
-
"loss": 0.
|
| 592 |
"step": 480
|
| 593 |
},
|
| 594 |
{
|
| 595 |
-
"epoch":
|
| 596 |
-
"learning_rate": 0.
|
| 597 |
-
"loss": 0.
|
| 598 |
"step": 485
|
| 599 |
},
|
| 600 |
{
|
| 601 |
-
"epoch":
|
| 602 |
-
"learning_rate": 0.
|
| 603 |
-
"loss": 0.
|
| 604 |
"step": 490
|
| 605 |
},
|
| 606 |
{
|
| 607 |
-
"epoch":
|
| 608 |
-
"learning_rate": 0.
|
| 609 |
-
"loss": 0.
|
| 610 |
"step": 495
|
| 611 |
},
|
| 612 |
{
|
| 613 |
-
"epoch":
|
| 614 |
-
"learning_rate": 0.
|
| 615 |
-
"loss": 0.
|
| 616 |
"step": 500
|
| 617 |
},
|
| 618 |
{
|
| 619 |
-
"epoch":
|
| 620 |
-
"eval_loss":
|
| 621 |
-
"eval_runtime":
|
| 622 |
-
"eval_samples_per_second": 7.
|
| 623 |
-
"eval_steps_per_second":
|
| 624 |
"step": 500
|
| 625 |
-
},
|
| 626 |
-
{
|
| 627 |
-
"epoch": 8.08,
|
| 628 |
-
"learning_rate": 0.00029984210526315787,
|
| 629 |
-
"loss": 0.441,
|
| 630 |
-
"step": 505
|
| 631 |
-
},
|
| 632 |
-
{
|
| 633 |
-
"epoch": 8.16,
|
| 634 |
-
"learning_rate": 0.00029968421052631577,
|
| 635 |
-
"loss": 0.4404,
|
| 636 |
-
"step": 510
|
| 637 |
-
},
|
| 638 |
-
{
|
| 639 |
-
"epoch": 8.24,
|
| 640 |
-
"learning_rate": 0.00029952631578947366,
|
| 641 |
-
"loss": 0.4452,
|
| 642 |
-
"step": 515
|
| 643 |
-
},
|
| 644 |
-
{
|
| 645 |
-
"epoch": 8.32,
|
| 646 |
-
"learning_rate": 0.00029936842105263156,
|
| 647 |
-
"loss": 0.4385,
|
| 648 |
-
"step": 520
|
| 649 |
-
},
|
| 650 |
-
{
|
| 651 |
-
"epoch": 8.4,
|
| 652 |
-
"learning_rate": 0.00029921052631578946,
|
| 653 |
-
"loss": 0.4457,
|
| 654 |
-
"step": 525
|
| 655 |
-
},
|
| 656 |
-
{
|
| 657 |
-
"epoch": 8.48,
|
| 658 |
-
"learning_rate": 0.00029905263157894735,
|
| 659 |
-
"loss": 0.4407,
|
| 660 |
-
"step": 530
|
| 661 |
-
},
|
| 662 |
-
{
|
| 663 |
-
"epoch": 8.56,
|
| 664 |
-
"learning_rate": 0.00029889473684210525,
|
| 665 |
-
"loss": 0.4432,
|
| 666 |
-
"step": 535
|
| 667 |
-
},
|
| 668 |
-
{
|
| 669 |
-
"epoch": 8.64,
|
| 670 |
-
"learning_rate": 0.00029873684210526315,
|
| 671 |
-
"loss": 0.4408,
|
| 672 |
-
"step": 540
|
| 673 |
-
},
|
| 674 |
-
{
|
| 675 |
-
"epoch": 8.72,
|
| 676 |
-
"learning_rate": 0.00029857894736842104,
|
| 677 |
-
"loss": 0.4431,
|
| 678 |
-
"step": 545
|
| 679 |
-
},
|
| 680 |
-
{
|
| 681 |
-
"epoch": 8.8,
|
| 682 |
-
"learning_rate": 0.00029842105263157894,
|
| 683 |
-
"loss": 0.4403,
|
| 684 |
-
"step": 550
|
| 685 |
-
},
|
| 686 |
-
{
|
| 687 |
-
"epoch": 8.88,
|
| 688 |
-
"learning_rate": 0.0002982631578947368,
|
| 689 |
-
"loss": 0.4357,
|
| 690 |
-
"step": 555
|
| 691 |
-
},
|
| 692 |
-
{
|
| 693 |
-
"epoch": 8.96,
|
| 694 |
-
"learning_rate": 0.00029810526315789473,
|
| 695 |
-
"loss": 0.4411,
|
| 696 |
-
"step": 560
|
| 697 |
-
},
|
| 698 |
-
{
|
| 699 |
-
"epoch": 9.04,
|
| 700 |
-
"learning_rate": 0.00029794736842105263,
|
| 701 |
-
"loss": 0.4347,
|
| 702 |
-
"step": 565
|
| 703 |
-
},
|
| 704 |
-
{
|
| 705 |
-
"epoch": 9.12,
|
| 706 |
-
"learning_rate": 0.0002977894736842105,
|
| 707 |
-
"loss": 0.4317,
|
| 708 |
-
"step": 570
|
| 709 |
-
},
|
| 710 |
-
{
|
| 711 |
-
"epoch": 9.2,
|
| 712 |
-
"learning_rate": 0.00029763157894736837,
|
| 713 |
-
"loss": 0.4332,
|
| 714 |
-
"step": 575
|
| 715 |
-
},
|
| 716 |
-
{
|
| 717 |
-
"epoch": 9.28,
|
| 718 |
-
"learning_rate": 0.00029747368421052627,
|
| 719 |
-
"loss": 0.4261,
|
| 720 |
-
"step": 580
|
| 721 |
-
},
|
| 722 |
-
{
|
| 723 |
-
"epoch": 9.36,
|
| 724 |
-
"learning_rate": 0.0002973157894736842,
|
| 725 |
-
"loss": 0.4348,
|
| 726 |
-
"step": 585
|
| 727 |
-
},
|
| 728 |
-
{
|
| 729 |
-
"epoch": 9.44,
|
| 730 |
-
"learning_rate": 0.00029715789473684206,
|
| 731 |
-
"loss": 0.434,
|
| 732 |
-
"step": 590
|
| 733 |
-
},
|
| 734 |
-
{
|
| 735 |
-
"epoch": 9.52,
|
| 736 |
-
"learning_rate": 0.00029699999999999996,
|
| 737 |
-
"loss": 0.4334,
|
| 738 |
-
"step": 595
|
| 739 |
-
},
|
| 740 |
-
{
|
| 741 |
-
"epoch": 9.6,
|
| 742 |
-
"learning_rate": 0.00029684210526315785,
|
| 743 |
-
"loss": 0.4348,
|
| 744 |
-
"step": 600
|
| 745 |
-
},
|
| 746 |
-
{
|
| 747 |
-
"epoch": 9.68,
|
| 748 |
-
"learning_rate": 0.0002966842105263158,
|
| 749 |
-
"loss": 0.4292,
|
| 750 |
-
"step": 605
|
| 751 |
-
},
|
| 752 |
-
{
|
| 753 |
-
"epoch": 9.76,
|
| 754 |
-
"learning_rate": 0.00029652631578947364,
|
| 755 |
-
"loss": 0.4332,
|
| 756 |
-
"step": 610
|
| 757 |
-
},
|
| 758 |
-
{
|
| 759 |
-
"epoch": 9.84,
|
| 760 |
-
"learning_rate": 0.00029636842105263154,
|
| 761 |
-
"loss": 0.4302,
|
| 762 |
-
"step": 615
|
| 763 |
-
},
|
| 764 |
-
{
|
| 765 |
-
"epoch": 9.92,
|
| 766 |
-
"learning_rate": 0.00029621052631578944,
|
| 767 |
-
"loss": 0.4303,
|
| 768 |
-
"step": 620
|
| 769 |
-
},
|
| 770 |
-
{
|
| 771 |
-
"epoch": 10.0,
|
| 772 |
-
"learning_rate": 0.00029605263157894733,
|
| 773 |
-
"loss": 0.427,
|
| 774 |
-
"step": 625
|
| 775 |
-
},
|
| 776 |
-
{
|
| 777 |
-
"epoch": 10.08,
|
| 778 |
-
"learning_rate": 0.00029589473684210523,
|
| 779 |
-
"loss": 0.4234,
|
| 780 |
-
"step": 630
|
| 781 |
-
},
|
| 782 |
-
{
|
| 783 |
-
"epoch": 10.16,
|
| 784 |
-
"learning_rate": 0.00029573684210526313,
|
| 785 |
-
"loss": 0.4251,
|
| 786 |
-
"step": 635
|
| 787 |
-
},
|
| 788 |
-
{
|
| 789 |
-
"epoch": 10.24,
|
| 790 |
-
"learning_rate": 0.000295578947368421,
|
| 791 |
-
"loss": 0.4208,
|
| 792 |
-
"step": 640
|
| 793 |
-
},
|
| 794 |
-
{
|
| 795 |
-
"epoch": 10.32,
|
| 796 |
-
"learning_rate": 0.0002954210526315789,
|
| 797 |
-
"loss": 0.4252,
|
| 798 |
-
"step": 645
|
| 799 |
-
},
|
| 800 |
-
{
|
| 801 |
-
"epoch": 10.4,
|
| 802 |
-
"learning_rate": 0.0002952631578947368,
|
| 803 |
-
"loss": 0.4263,
|
| 804 |
-
"step": 650
|
| 805 |
-
},
|
| 806 |
-
{
|
| 807 |
-
"epoch": 10.48,
|
| 808 |
-
"learning_rate": 0.0002951052631578947,
|
| 809 |
-
"loss": 0.4221,
|
| 810 |
-
"step": 655
|
| 811 |
-
},
|
| 812 |
-
{
|
| 813 |
-
"epoch": 10.56,
|
| 814 |
-
"learning_rate": 0.0002949473684210526,
|
| 815 |
-
"loss": 0.4169,
|
| 816 |
-
"step": 660
|
| 817 |
-
},
|
| 818 |
-
{
|
| 819 |
-
"epoch": 10.64,
|
| 820 |
-
"learning_rate": 0.0002947894736842105,
|
| 821 |
-
"loss": 0.4282,
|
| 822 |
-
"step": 665
|
| 823 |
-
},
|
| 824 |
-
{
|
| 825 |
-
"epoch": 10.72,
|
| 826 |
-
"learning_rate": 0.0002946315789473684,
|
| 827 |
-
"loss": 0.4207,
|
| 828 |
-
"step": 670
|
| 829 |
-
},
|
| 830 |
-
{
|
| 831 |
-
"epoch": 10.8,
|
| 832 |
-
"learning_rate": 0.0002944736842105263,
|
| 833 |
-
"loss": 0.4257,
|
| 834 |
-
"step": 675
|
| 835 |
-
},
|
| 836 |
-
{
|
| 837 |
-
"epoch": 10.88,
|
| 838 |
-
"learning_rate": 0.0002943157894736842,
|
| 839 |
-
"loss": 0.4207,
|
| 840 |
-
"step": 680
|
| 841 |
-
},
|
| 842 |
-
{
|
| 843 |
-
"epoch": 10.96,
|
| 844 |
-
"learning_rate": 0.0002941578947368421,
|
| 845 |
-
"loss": 0.4171,
|
| 846 |
-
"step": 685
|
| 847 |
-
},
|
| 848 |
-
{
|
| 849 |
-
"epoch": 11.04,
|
| 850 |
-
"learning_rate": 0.000294,
|
| 851 |
-
"loss": 0.4215,
|
| 852 |
-
"step": 690
|
| 853 |
-
},
|
| 854 |
-
{
|
| 855 |
-
"epoch": 11.12,
|
| 856 |
-
"learning_rate": 0.00029384210526315783,
|
| 857 |
-
"loss": 0.4112,
|
| 858 |
-
"step": 695
|
| 859 |
-
},
|
| 860 |
-
{
|
| 861 |
-
"epoch": 11.2,
|
| 862 |
-
"learning_rate": 0.0002936842105263158,
|
| 863 |
-
"loss": 0.4127,
|
| 864 |
-
"step": 700
|
| 865 |
-
},
|
| 866 |
-
{
|
| 867 |
-
"epoch": 11.28,
|
| 868 |
-
"learning_rate": 0.0002935263157894737,
|
| 869 |
-
"loss": 0.4125,
|
| 870 |
-
"step": 705
|
| 871 |
-
},
|
| 872 |
-
{
|
| 873 |
-
"epoch": 11.36,
|
| 874 |
-
"learning_rate": 0.0002933684210526316,
|
| 875 |
-
"loss": 0.4149,
|
| 876 |
-
"step": 710
|
| 877 |
-
},
|
| 878 |
-
{
|
| 879 |
-
"epoch": 11.44,
|
| 880 |
-
"learning_rate": 0.0002932105263157894,
|
| 881 |
-
"loss": 0.4144,
|
| 882 |
-
"step": 715
|
| 883 |
-
},
|
| 884 |
-
{
|
| 885 |
-
"epoch": 11.52,
|
| 886 |
-
"learning_rate": 0.0002930526315789473,
|
| 887 |
-
"loss": 0.4157,
|
| 888 |
-
"step": 720
|
| 889 |
-
},
|
| 890 |
-
{
|
| 891 |
-
"epoch": 11.6,
|
| 892 |
-
"learning_rate": 0.00029289473684210527,
|
| 893 |
-
"loss": 0.4087,
|
| 894 |
-
"step": 725
|
| 895 |
-
},
|
| 896 |
-
{
|
| 897 |
-
"epoch": 11.68,
|
| 898 |
-
"learning_rate": 0.00029273684210526316,
|
| 899 |
-
"loss": 0.4162,
|
| 900 |
-
"step": 730
|
| 901 |
-
},
|
| 902 |
-
{
|
| 903 |
-
"epoch": 11.76,
|
| 904 |
-
"learning_rate": 0.000292578947368421,
|
| 905 |
-
"loss": 0.4104,
|
| 906 |
-
"step": 735
|
| 907 |
-
},
|
| 908 |
-
{
|
| 909 |
-
"epoch": 11.84,
|
| 910 |
-
"learning_rate": 0.0002924210526315789,
|
| 911 |
-
"loss": 0.414,
|
| 912 |
-
"step": 740
|
| 913 |
-
},
|
| 914 |
-
{
|
| 915 |
-
"epoch": 11.92,
|
| 916 |
-
"learning_rate": 0.00029226315789473685,
|
| 917 |
-
"loss": 0.4132,
|
| 918 |
-
"step": 745
|
| 919 |
-
},
|
| 920 |
-
{
|
| 921 |
-
"epoch": 12.0,
|
| 922 |
-
"learning_rate": 0.0002921052631578947,
|
| 923 |
-
"loss": 0.4223,
|
| 924 |
-
"step": 750
|
| 925 |
-
},
|
| 926 |
-
{
|
| 927 |
-
"epoch": 12.0,
|
| 928 |
-
"eval_loss": 0.44360578060150146,
|
| 929 |
-
"eval_runtime": 265.1184,
|
| 930 |
-
"eval_samples_per_second": 7.548,
|
| 931 |
-
"eval_steps_per_second": 0.947,
|
| 932 |
-
"step": 750
|
| 933 |
-
},
|
| 934 |
-
{
|
| 935 |
-
"epoch": 12.08,
|
| 936 |
-
"learning_rate": 0.0002919473684210526,
|
| 937 |
-
"loss": 0.4084,
|
| 938 |
-
"step": 755
|
| 939 |
-
},
|
| 940 |
-
{
|
| 941 |
-
"epoch": 12.16,
|
| 942 |
-
"learning_rate": 0.0002917894736842105,
|
| 943 |
-
"loss": 0.4045,
|
| 944 |
-
"step": 760
|
| 945 |
-
},
|
| 946 |
-
{
|
| 947 |
-
"epoch": 12.24,
|
| 948 |
-
"learning_rate": 0.0002916315789473684,
|
| 949 |
-
"loss": 0.4084,
|
| 950 |
-
"step": 765
|
| 951 |
-
},
|
| 952 |
-
{
|
| 953 |
-
"epoch": 12.32,
|
| 954 |
-
"learning_rate": 0.0002914736842105263,
|
| 955 |
-
"loss": 0.4053,
|
| 956 |
-
"step": 770
|
| 957 |
-
},
|
| 958 |
-
{
|
| 959 |
-
"epoch": 12.4,
|
| 960 |
-
"learning_rate": 0.0002913157894736842,
|
| 961 |
-
"loss": 0.4079,
|
| 962 |
-
"step": 775
|
| 963 |
-
},
|
| 964 |
-
{
|
| 965 |
-
"epoch": 12.48,
|
| 966 |
-
"learning_rate": 0.0002911578947368421,
|
| 967 |
-
"loss": 0.4045,
|
| 968 |
-
"step": 780
|
| 969 |
-
},
|
| 970 |
-
{
|
| 971 |
-
"epoch": 12.56,
|
| 972 |
-
"learning_rate": 0.00029099999999999997,
|
| 973 |
-
"loss": 0.4009,
|
| 974 |
-
"step": 785
|
| 975 |
-
},
|
| 976 |
-
{
|
| 977 |
-
"epoch": 12.64,
|
| 978 |
-
"learning_rate": 0.00029084210526315787,
|
| 979 |
-
"loss": 0.4064,
|
| 980 |
-
"step": 790
|
| 981 |
-
},
|
| 982 |
-
{
|
| 983 |
-
"epoch": 12.72,
|
| 984 |
-
"learning_rate": 0.00029068421052631577,
|
| 985 |
-
"loss": 0.4104,
|
| 986 |
-
"step": 795
|
| 987 |
-
},
|
| 988 |
-
{
|
| 989 |
-
"epoch": 12.8,
|
| 990 |
-
"learning_rate": 0.00029052631578947366,
|
| 991 |
-
"loss": 0.4121,
|
| 992 |
-
"step": 800
|
| 993 |
-
},
|
| 994 |
-
{
|
| 995 |
-
"epoch": 12.88,
|
| 996 |
-
"learning_rate": 0.00029036842105263156,
|
| 997 |
-
"loss": 0.4064,
|
| 998 |
-
"step": 805
|
| 999 |
-
},
|
| 1000 |
-
{
|
| 1001 |
-
"epoch": 12.96,
|
| 1002 |
-
"learning_rate": 0.00029021052631578945,
|
| 1003 |
-
"loss": 0.4126,
|
| 1004 |
-
"step": 810
|
| 1005 |
-
},
|
| 1006 |
-
{
|
| 1007 |
-
"epoch": 13.04,
|
| 1008 |
-
"learning_rate": 0.00029005263157894735,
|
| 1009 |
-
"loss": 0.4028,
|
| 1010 |
-
"step": 815
|
| 1011 |
-
},
|
| 1012 |
-
{
|
| 1013 |
-
"epoch": 13.12,
|
| 1014 |
-
"learning_rate": 0.00028989473684210525,
|
| 1015 |
-
"loss": 0.3973,
|
| 1016 |
-
"step": 820
|
| 1017 |
-
},
|
| 1018 |
-
{
|
| 1019 |
-
"epoch": 13.2,
|
| 1020 |
-
"learning_rate": 0.00028973684210526314,
|
| 1021 |
-
"loss": 0.3977,
|
| 1022 |
-
"step": 825
|
| 1023 |
-
},
|
| 1024 |
-
{
|
| 1025 |
-
"epoch": 13.28,
|
| 1026 |
-
"learning_rate": 0.00028957894736842104,
|
| 1027 |
-
"loss": 0.3989,
|
| 1028 |
-
"step": 830
|
| 1029 |
-
},
|
| 1030 |
-
{
|
| 1031 |
-
"epoch": 13.36,
|
| 1032 |
-
"learning_rate": 0.0002894210526315789,
|
| 1033 |
-
"loss": 0.4025,
|
| 1034 |
-
"step": 835
|
| 1035 |
-
},
|
| 1036 |
-
{
|
| 1037 |
-
"epoch": 13.44,
|
| 1038 |
-
"learning_rate": 0.00028926315789473683,
|
| 1039 |
-
"loss": 0.4013,
|
| 1040 |
-
"step": 840
|
| 1041 |
-
},
|
| 1042 |
-
{
|
| 1043 |
-
"epoch": 13.52,
|
| 1044 |
-
"learning_rate": 0.00028910526315789473,
|
| 1045 |
-
"loss": 0.4044,
|
| 1046 |
-
"step": 845
|
| 1047 |
-
},
|
| 1048 |
-
{
|
| 1049 |
-
"epoch": 13.6,
|
| 1050 |
-
"learning_rate": 0.00028894736842105263,
|
| 1051 |
-
"loss": 0.4037,
|
| 1052 |
-
"step": 850
|
| 1053 |
-
},
|
| 1054 |
-
{
|
| 1055 |
-
"epoch": 13.68,
|
| 1056 |
-
"learning_rate": 0.00028878947368421047,
|
| 1057 |
-
"loss": 0.4,
|
| 1058 |
-
"step": 855
|
| 1059 |
-
},
|
| 1060 |
-
{
|
| 1061 |
-
"epoch": 13.76,
|
| 1062 |
-
"learning_rate": 0.00028863157894736837,
|
| 1063 |
-
"loss": 0.4023,
|
| 1064 |
-
"step": 860
|
| 1065 |
-
},
|
| 1066 |
-
{
|
| 1067 |
-
"epoch": 13.84,
|
| 1068 |
-
"learning_rate": 0.0002884736842105263,
|
| 1069 |
-
"loss": 0.402,
|
| 1070 |
-
"step": 865
|
| 1071 |
-
},
|
| 1072 |
-
{
|
| 1073 |
-
"epoch": 13.92,
|
| 1074 |
-
"learning_rate": 0.0002883157894736842,
|
| 1075 |
-
"loss": 0.4033,
|
| 1076 |
-
"step": 870
|
| 1077 |
-
},
|
| 1078 |
-
{
|
| 1079 |
-
"epoch": 14.0,
|
| 1080 |
-
"learning_rate": 0.00028815789473684206,
|
| 1081 |
-
"loss": 0.4009,
|
| 1082 |
-
"step": 875
|
| 1083 |
-
},
|
| 1084 |
-
{
|
| 1085 |
-
"epoch": 14.08,
|
| 1086 |
-
"learning_rate": 0.00028799999999999995,
|
| 1087 |
-
"loss": 0.3952,
|
| 1088 |
-
"step": 880
|
| 1089 |
-
},
|
| 1090 |
-
{
|
| 1091 |
-
"epoch": 14.16,
|
| 1092 |
-
"learning_rate": 0.0002878421052631579,
|
| 1093 |
-
"loss": 0.3911,
|
| 1094 |
-
"step": 885
|
| 1095 |
-
},
|
| 1096 |
-
{
|
| 1097 |
-
"epoch": 14.24,
|
| 1098 |
-
"learning_rate": 0.00028768421052631575,
|
| 1099 |
-
"loss": 0.392,
|
| 1100 |
-
"step": 890
|
| 1101 |
-
},
|
| 1102 |
-
{
|
| 1103 |
-
"epoch": 14.32,
|
| 1104 |
-
"learning_rate": 0.00028752631578947364,
|
| 1105 |
-
"loss": 0.3912,
|
| 1106 |
-
"step": 895
|
| 1107 |
-
},
|
| 1108 |
-
{
|
| 1109 |
-
"epoch": 14.4,
|
| 1110 |
-
"learning_rate": 0.00028736842105263154,
|
| 1111 |
-
"loss": 0.3918,
|
| 1112 |
-
"step": 900
|
| 1113 |
-
},
|
| 1114 |
-
{
|
| 1115 |
-
"epoch": 14.48,
|
| 1116 |
-
"learning_rate": 0.00028721052631578944,
|
| 1117 |
-
"loss": 0.3955,
|
| 1118 |
-
"step": 905
|
| 1119 |
-
},
|
| 1120 |
-
{
|
| 1121 |
-
"epoch": 14.56,
|
| 1122 |
-
"learning_rate": 0.00028705263157894733,
|
| 1123 |
-
"loss": 0.3979,
|
| 1124 |
-
"step": 910
|
| 1125 |
-
},
|
| 1126 |
-
{
|
| 1127 |
-
"epoch": 14.64,
|
| 1128 |
-
"learning_rate": 0.00028689473684210523,
|
| 1129 |
-
"loss": 0.396,
|
| 1130 |
-
"step": 915
|
| 1131 |
-
},
|
| 1132 |
-
{
|
| 1133 |
-
"epoch": 14.72,
|
| 1134 |
-
"learning_rate": 0.0002867368421052631,
|
| 1135 |
-
"loss": 0.3957,
|
| 1136 |
-
"step": 920
|
| 1137 |
-
},
|
| 1138 |
-
{
|
| 1139 |
-
"epoch": 14.8,
|
| 1140 |
-
"learning_rate": 0.000286578947368421,
|
| 1141 |
-
"loss": 0.3975,
|
| 1142 |
-
"step": 925
|
| 1143 |
-
},
|
| 1144 |
-
{
|
| 1145 |
-
"epoch": 14.88,
|
| 1146 |
-
"learning_rate": 0.0002864210526315789,
|
| 1147 |
-
"loss": 0.398,
|
| 1148 |
-
"step": 930
|
| 1149 |
-
},
|
| 1150 |
-
{
|
| 1151 |
-
"epoch": 14.96,
|
| 1152 |
-
"learning_rate": 0.0002862631578947368,
|
| 1153 |
-
"loss": 0.3965,
|
| 1154 |
-
"step": 935
|
| 1155 |
-
},
|
| 1156 |
-
{
|
| 1157 |
-
"epoch": 15.04,
|
| 1158 |
-
"learning_rate": 0.0002861052631578947,
|
| 1159 |
-
"loss": 0.3906,
|
| 1160 |
-
"step": 940
|
| 1161 |
-
},
|
| 1162 |
-
{
|
| 1163 |
-
"epoch": 15.12,
|
| 1164 |
-
"learning_rate": 0.0002859473684210526,
|
| 1165 |
-
"loss": 0.3852,
|
| 1166 |
-
"step": 945
|
| 1167 |
-
},
|
| 1168 |
-
{
|
| 1169 |
-
"epoch": 15.2,
|
| 1170 |
-
"learning_rate": 0.0002857894736842105,
|
| 1171 |
-
"loss": 0.3933,
|
| 1172 |
-
"step": 950
|
| 1173 |
-
},
|
| 1174 |
-
{
|
| 1175 |
-
"epoch": 15.28,
|
| 1176 |
-
"learning_rate": 0.0002856315789473684,
|
| 1177 |
-
"loss": 0.3857,
|
| 1178 |
-
"step": 955
|
| 1179 |
-
},
|
| 1180 |
-
{
|
| 1181 |
-
"epoch": 15.36,
|
| 1182 |
-
"learning_rate": 0.0002854736842105263,
|
| 1183 |
-
"loss": 0.3897,
|
| 1184 |
-
"step": 960
|
| 1185 |
-
},
|
| 1186 |
-
{
|
| 1187 |
-
"epoch": 15.44,
|
| 1188 |
-
"learning_rate": 0.0002853157894736842,
|
| 1189 |
-
"loss": 0.3885,
|
| 1190 |
-
"step": 965
|
| 1191 |
-
},
|
| 1192 |
-
{
|
| 1193 |
-
"epoch": 15.52,
|
| 1194 |
-
"learning_rate": 0.0002851578947368421,
|
| 1195 |
-
"loss": 0.3882,
|
| 1196 |
-
"step": 970
|
| 1197 |
-
},
|
| 1198 |
-
{
|
| 1199 |
-
"epoch": 15.6,
|
| 1200 |
-
"learning_rate": 0.000285,
|
| 1201 |
-
"loss": 0.3961,
|
| 1202 |
-
"step": 975
|
| 1203 |
-
},
|
| 1204 |
-
{
|
| 1205 |
-
"epoch": 15.68,
|
| 1206 |
-
"learning_rate": 0.0002848421052631579,
|
| 1207 |
-
"loss": 0.3886,
|
| 1208 |
-
"step": 980
|
| 1209 |
-
},
|
| 1210 |
-
{
|
| 1211 |
-
"epoch": 15.76,
|
| 1212 |
-
"learning_rate": 0.0002846842105263158,
|
| 1213 |
-
"loss": 0.391,
|
| 1214 |
-
"step": 985
|
| 1215 |
-
},
|
| 1216 |
-
{
|
| 1217 |
-
"epoch": 15.84,
|
| 1218 |
-
"learning_rate": 0.0002845263157894737,
|
| 1219 |
-
"loss": 0.391,
|
| 1220 |
-
"step": 990
|
| 1221 |
-
},
|
| 1222 |
-
{
|
| 1223 |
-
"epoch": 15.92,
|
| 1224 |
-
"learning_rate": 0.0002843684210526315,
|
| 1225 |
-
"loss": 0.3901,
|
| 1226 |
-
"step": 995
|
| 1227 |
-
},
|
| 1228 |
-
{
|
| 1229 |
-
"epoch": 16.0,
|
| 1230 |
-
"learning_rate": 0.0002842105263157894,
|
| 1231 |
-
"loss": 0.3869,
|
| 1232 |
-
"step": 1000
|
| 1233 |
-
},
|
| 1234 |
-
{
|
| 1235 |
-
"epoch": 16.0,
|
| 1236 |
-
"eval_loss": 0.43608614802360535,
|
| 1237 |
-
"eval_runtime": 265.3444,
|
| 1238 |
-
"eval_samples_per_second": 7.541,
|
| 1239 |
-
"eval_steps_per_second": 0.946,
|
| 1240 |
-
"step": 1000
|
| 1241 |
}
|
| 1242 |
],
|
| 1243 |
"max_steps": 10000,
|
| 1244 |
-
"num_train_epochs":
|
| 1245 |
-
"total_flos":
|
| 1246 |
"trial_name": null,
|
| 1247 |
"trial_params": null
|
| 1248 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 1.9192386865615845,
|
| 3 |
+
"best_model_checkpoint": "experiments/checkpoint-500",
|
| 4 |
+
"epoch": 100.0,
|
| 5 |
+
"global_step": 500,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
| 9 |
"log_history": [
|
| 10 |
{
|
| 11 |
+
"epoch": 1.0,
|
| 12 |
+
"learning_rate": 1.8e-06,
|
| 13 |
+
"loss": 1.7688,
|
| 14 |
"step": 5
|
| 15 |
},
|
| 16 |
{
|
| 17 |
+
"epoch": 2.0,
|
| 18 |
+
"learning_rate": 4.8e-06,
|
| 19 |
+
"loss": 1.7684,
|
| 20 |
"step": 10
|
| 21 |
},
|
| 22 |
{
|
| 23 |
+
"epoch": 3.0,
|
| 24 |
+
"learning_rate": 7.799999999999998e-06,
|
| 25 |
+
"loss": 1.7564,
|
| 26 |
"step": 15
|
| 27 |
},
|
| 28 |
{
|
| 29 |
+
"epoch": 4.0,
|
| 30 |
+
"learning_rate": 1.0799999999999998e-05,
|
| 31 |
+
"loss": 1.7424,
|
| 32 |
"step": 20
|
| 33 |
},
|
| 34 |
{
|
| 35 |
+
"epoch": 5.0,
|
| 36 |
+
"learning_rate": 1.3799999999999998e-05,
|
| 37 |
+
"loss": 1.727,
|
| 38 |
"step": 25
|
| 39 |
},
|
| 40 |
{
|
| 41 |
+
"epoch": 6.0,
|
| 42 |
+
"learning_rate": 1.68e-05,
|
| 43 |
+
"loss": 1.7135,
|
| 44 |
"step": 30
|
| 45 |
},
|
| 46 |
{
|
| 47 |
+
"epoch": 7.0,
|
| 48 |
+
"learning_rate": 1.98e-05,
|
| 49 |
+
"loss": 1.701,
|
| 50 |
"step": 35
|
| 51 |
},
|
| 52 |
{
|
| 53 |
+
"epoch": 8.0,
|
| 54 |
+
"learning_rate": 2.28e-05,
|
| 55 |
+
"loss": 1.6797,
|
| 56 |
"step": 40
|
| 57 |
},
|
| 58 |
{
|
| 59 |
+
"epoch": 9.0,
|
| 60 |
+
"learning_rate": 2.5799999999999997e-05,
|
| 61 |
+
"loss": 1.6547,
|
| 62 |
"step": 45
|
| 63 |
},
|
| 64 |
{
|
| 65 |
+
"epoch": 10.0,
|
| 66 |
+
"learning_rate": 2.88e-05,
|
| 67 |
+
"loss": 1.6245,
|
| 68 |
+
"step": 50
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 10.0,
|
| 72 |
+
"eval_loss": 1.5684312582015991,
|
| 73 |
+
"eval_runtime": 1.4166,
|
| 74 |
+
"eval_samples_per_second": 7.059,
|
| 75 |
+
"eval_steps_per_second": 1.412,
|
| 76 |
"step": 50
|
| 77 |
},
|
| 78 |
{
|
| 79 |
+
"epoch": 11.0,
|
| 80 |
+
"learning_rate": 3.1799999999999994e-05,
|
| 81 |
+
"loss": 1.5841,
|
| 82 |
"step": 55
|
| 83 |
},
|
| 84 |
{
|
| 85 |
+
"epoch": 12.0,
|
| 86 |
+
"learning_rate": 3.48e-05,
|
| 87 |
+
"loss": 1.5316,
|
| 88 |
"step": 60
|
| 89 |
},
|
| 90 |
{
|
| 91 |
+
"epoch": 13.0,
|
| 92 |
+
"learning_rate": 3.78e-05,
|
| 93 |
+
"loss": 1.4644,
|
| 94 |
"step": 65
|
| 95 |
},
|
| 96 |
{
|
| 97 |
+
"epoch": 14.0,
|
| 98 |
+
"learning_rate": 4.08e-05,
|
| 99 |
+
"loss": 1.3728,
|
| 100 |
"step": 70
|
| 101 |
},
|
| 102 |
{
|
| 103 |
+
"epoch": 15.0,
|
| 104 |
+
"learning_rate": 4.3799999999999994e-05,
|
| 105 |
+
"loss": 1.2692,
|
| 106 |
"step": 75
|
| 107 |
},
|
| 108 |
{
|
| 109 |
+
"epoch": 16.0,
|
| 110 |
+
"learning_rate": 4.56e-05,
|
| 111 |
+
"loss": 1.1998,
|
| 112 |
"step": 80
|
| 113 |
},
|
| 114 |
{
|
| 115 |
+
"epoch": 17.0,
|
| 116 |
+
"learning_rate": 4.8599999999999995e-05,
|
| 117 |
+
"loss": 1.1159,
|
| 118 |
"step": 85
|
| 119 |
},
|
| 120 |
{
|
| 121 |
+
"epoch": 18.0,
|
| 122 |
+
"learning_rate": 5.1599999999999994e-05,
|
| 123 |
+
"loss": 1.0442,
|
| 124 |
"step": 90
|
| 125 |
},
|
| 126 |
{
|
| 127 |
+
"epoch": 19.0,
|
| 128 |
+
"learning_rate": 5.459999999999999e-05,
|
| 129 |
+
"loss": 0.9944,
|
| 130 |
"step": 95
|
| 131 |
},
|
| 132 |
{
|
| 133 |
+
"epoch": 20.0,
|
| 134 |
+
"learning_rate": 5.76e-05,
|
| 135 |
+
"loss": 0.9518,
|
| 136 |
"step": 100
|
| 137 |
},
|
| 138 |
{
|
| 139 |
+
"epoch": 20.0,
|
| 140 |
+
"eval_loss": 0.9818406105041504,
|
| 141 |
+
"eval_runtime": 1.3835,
|
| 142 |
+
"eval_samples_per_second": 7.228,
|
| 143 |
+
"eval_steps_per_second": 1.446,
|
| 144 |
+
"step": 100
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 21.0,
|
| 148 |
+
"learning_rate": 6.0599999999999996e-05,
|
| 149 |
+
"loss": 0.908,
|
| 150 |
"step": 105
|
| 151 |
},
|
| 152 |
{
|
| 153 |
+
"epoch": 22.0,
|
| 154 |
+
"learning_rate": 6.359999999999999e-05,
|
| 155 |
+
"loss": 0.8678,
|
| 156 |
"step": 110
|
| 157 |
},
|
| 158 |
{
|
| 159 |
+
"epoch": 23.0,
|
| 160 |
+
"learning_rate": 6.659999999999999e-05,
|
| 161 |
+
"loss": 0.8303,
|
| 162 |
"step": 115
|
| 163 |
},
|
| 164 |
{
|
| 165 |
+
"epoch": 24.0,
|
| 166 |
+
"learning_rate": 6.96e-05,
|
| 167 |
+
"loss": 0.7928,
|
| 168 |
"step": 120
|
| 169 |
},
|
| 170 |
{
|
| 171 |
+
"epoch": 25.0,
|
| 172 |
+
"learning_rate": 7.259999999999999e-05,
|
| 173 |
+
"loss": 0.7594,
|
| 174 |
"step": 125
|
| 175 |
},
|
| 176 |
{
|
| 177 |
+
"epoch": 26.0,
|
| 178 |
+
"learning_rate": 7.56e-05,
|
| 179 |
+
"loss": 0.73,
|
| 180 |
"step": 130
|
| 181 |
},
|
| 182 |
{
|
| 183 |
+
"epoch": 27.0,
|
| 184 |
+
"learning_rate": 7.86e-05,
|
| 185 |
+
"loss": 0.7034,
|
| 186 |
"step": 135
|
| 187 |
},
|
| 188 |
{
|
| 189 |
+
"epoch": 28.0,
|
| 190 |
+
"learning_rate": 8.16e-05,
|
| 191 |
+
"loss": 0.6777,
|
| 192 |
"step": 140
|
| 193 |
},
|
| 194 |
{
|
| 195 |
+
"epoch": 29.0,
|
| 196 |
+
"learning_rate": 8.459999999999998e-05,
|
| 197 |
+
"loss": 0.6493,
|
| 198 |
"step": 145
|
| 199 |
},
|
| 200 |
{
|
| 201 |
+
"epoch": 30.0,
|
| 202 |
+
"learning_rate": 8.759999999999999e-05,
|
| 203 |
+
"loss": 0.6249,
|
| 204 |
"step": 150
|
| 205 |
},
|
| 206 |
{
|
| 207 |
+
"epoch": 30.0,
|
| 208 |
+
"eval_loss": 0.8599645495414734,
|
| 209 |
+
"eval_runtime": 1.4099,
|
| 210 |
+
"eval_samples_per_second": 7.093,
|
| 211 |
+
"eval_steps_per_second": 1.419,
|
| 212 |
+
"step": 150
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 31.0,
|
| 216 |
+
"learning_rate": 9.059999999999999e-05,
|
| 217 |
+
"loss": 0.6007,
|
| 218 |
"step": 155
|
| 219 |
},
|
| 220 |
{
|
| 221 |
+
"epoch": 32.0,
|
| 222 |
+
"learning_rate": 9.36e-05,
|
| 223 |
+
"loss": 0.5716,
|
| 224 |
"step": 160
|
| 225 |
},
|
| 226 |
{
|
| 227 |
+
"epoch": 33.0,
|
| 228 |
+
"learning_rate": 9.659999999999999e-05,
|
| 229 |
+
"loss": 0.5465,
|
| 230 |
"step": 165
|
| 231 |
},
|
| 232 |
{
|
| 233 |
+
"epoch": 34.0,
|
| 234 |
+
"learning_rate": 9.96e-05,
|
| 235 |
+
"loss": 0.5191,
|
| 236 |
"step": 170
|
| 237 |
},
|
| 238 |
{
|
| 239 |
+
"epoch": 35.0,
|
| 240 |
+
"learning_rate": 0.0001026,
|
| 241 |
+
"loss": 0.4947,
|
| 242 |
"step": 175
|
| 243 |
},
|
| 244 |
{
|
| 245 |
+
"epoch": 36.0,
|
| 246 |
+
"learning_rate": 0.00010559999999999998,
|
| 247 |
+
"loss": 0.4681,
|
| 248 |
"step": 180
|
| 249 |
},
|
| 250 |
{
|
| 251 |
+
"epoch": 37.0,
|
| 252 |
+
"learning_rate": 0.00010859999999999998,
|
| 253 |
+
"loss": 0.4417,
|
| 254 |
"step": 185
|
| 255 |
},
|
| 256 |
{
|
| 257 |
+
"epoch": 38.0,
|
| 258 |
+
"learning_rate": 0.00011159999999999999,
|
| 259 |
+
"loss": 0.4116,
|
| 260 |
"step": 190
|
| 261 |
},
|
| 262 |
{
|
| 263 |
+
"epoch": 39.0,
|
| 264 |
+
"learning_rate": 0.0001146,
|
| 265 |
+
"loss": 0.3804,
|
| 266 |
"step": 195
|
| 267 |
},
|
| 268 |
{
|
| 269 |
+
"epoch": 40.0,
|
| 270 |
+
"learning_rate": 0.0001176,
|
| 271 |
+
"loss": 0.3544,
|
| 272 |
+
"step": 200
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 40.0,
|
| 276 |
+
"eval_loss": 1.1081293821334839,
|
| 277 |
+
"eval_runtime": 1.3764,
|
| 278 |
+
"eval_samples_per_second": 7.265,
|
| 279 |
+
"eval_steps_per_second": 1.453,
|
| 280 |
"step": 200
|
| 281 |
},
|
| 282 |
{
|
| 283 |
+
"epoch": 41.0,
|
| 284 |
+
"learning_rate": 0.00012059999999999999,
|
| 285 |
+
"loss": 0.3248,
|
| 286 |
"step": 205
|
| 287 |
},
|
| 288 |
{
|
| 289 |
+
"epoch": 42.0,
|
| 290 |
+
"learning_rate": 0.0001236,
|
| 291 |
+
"loss": 0.2931,
|
| 292 |
"step": 210
|
| 293 |
},
|
| 294 |
{
|
| 295 |
+
"epoch": 43.0,
|
| 296 |
+
"learning_rate": 0.0001266,
|
| 297 |
+
"loss": 0.2677,
|
| 298 |
"step": 215
|
| 299 |
},
|
| 300 |
{
|
| 301 |
+
"epoch": 44.0,
|
| 302 |
+
"learning_rate": 0.00012959999999999998,
|
| 303 |
+
"loss": 0.2386,
|
| 304 |
"step": 220
|
| 305 |
},
|
| 306 |
{
|
| 307 |
+
"epoch": 45.0,
|
| 308 |
+
"learning_rate": 0.0001326,
|
| 309 |
+
"loss": 0.2142,
|
| 310 |
"step": 225
|
| 311 |
},
|
| 312 |
{
|
| 313 |
+
"epoch": 46.0,
|
| 314 |
+
"learning_rate": 0.0001356,
|
| 315 |
+
"loss": 0.1932,
|
| 316 |
"step": 230
|
| 317 |
},
|
| 318 |
{
|
| 319 |
+
"epoch": 47.0,
|
| 320 |
+
"learning_rate": 0.0001386,
|
| 321 |
+
"loss": 0.1709,
|
| 322 |
"step": 235
|
| 323 |
},
|
| 324 |
{
|
| 325 |
+
"epoch": 48.0,
|
| 326 |
+
"learning_rate": 0.00014159999999999997,
|
| 327 |
+
"loss": 0.1571,
|
| 328 |
"step": 240
|
| 329 |
},
|
| 330 |
{
|
| 331 |
+
"epoch": 49.0,
|
| 332 |
+
"learning_rate": 0.0001446,
|
| 333 |
+
"loss": 0.1417,
|
| 334 |
"step": 245
|
| 335 |
},
|
| 336 |
{
|
| 337 |
+
"epoch": 50.0,
|
| 338 |
+
"learning_rate": 0.00014759999999999998,
|
| 339 |
+
"loss": 0.1184,
|
| 340 |
"step": 250
|
| 341 |
},
|
| 342 |
{
|
| 343 |
+
"epoch": 50.0,
|
| 344 |
+
"eval_loss": 1.5212451219558716,
|
| 345 |
+
"eval_runtime": 1.3827,
|
| 346 |
+
"eval_samples_per_second": 7.232,
|
| 347 |
+
"eval_steps_per_second": 1.446,
|
| 348 |
"step": 250
|
| 349 |
},
|
| 350 |
{
|
| 351 |
+
"epoch": 51.0,
|
| 352 |
+
"learning_rate": 0.00015059999999999997,
|
| 353 |
+
"loss": 0.1096,
|
| 354 |
"step": 255
|
| 355 |
},
|
| 356 |
{
|
| 357 |
+
"epoch": 52.0,
|
| 358 |
+
"learning_rate": 0.0001536,
|
| 359 |
+
"loss": 0.1037,
|
| 360 |
"step": 260
|
| 361 |
},
|
| 362 |
{
|
| 363 |
+
"epoch": 53.0,
|
| 364 |
+
"learning_rate": 0.00015659999999999998,
|
| 365 |
+
"loss": 0.095,
|
| 366 |
"step": 265
|
| 367 |
},
|
| 368 |
{
|
| 369 |
+
"epoch": 54.0,
|
| 370 |
+
"learning_rate": 0.0001596,
|
| 371 |
+
"loss": 0.0865,
|
| 372 |
"step": 270
|
| 373 |
},
|
| 374 |
{
|
| 375 |
+
"epoch": 55.0,
|
| 376 |
+
"learning_rate": 0.0001626,
|
| 377 |
+
"loss": 0.0808,
|
| 378 |
"step": 275
|
| 379 |
},
|
| 380 |
{
|
| 381 |
+
"epoch": 56.0,
|
| 382 |
+
"learning_rate": 0.0001656,
|
| 383 |
+
"loss": 0.0794,
|
| 384 |
"step": 280
|
| 385 |
},
|
| 386 |
{
|
| 387 |
+
"epoch": 57.0,
|
| 388 |
+
"learning_rate": 0.0001686,
|
| 389 |
+
"loss": 0.075,
|
| 390 |
"step": 285
|
| 391 |
},
|
| 392 |
{
|
| 393 |
+
"epoch": 58.0,
|
| 394 |
+
"learning_rate": 0.00017159999999999997,
|
| 395 |
+
"loss": 0.0726,
|
| 396 |
"step": 290
|
| 397 |
},
|
| 398 |
{
|
| 399 |
+
"epoch": 59.0,
|
| 400 |
+
"learning_rate": 0.00017459999999999996,
|
| 401 |
+
"loss": 0.0696,
|
| 402 |
"step": 295
|
| 403 |
},
|
| 404 |
{
|
| 405 |
+
"epoch": 60.0,
|
| 406 |
+
"learning_rate": 0.00017759999999999998,
|
| 407 |
+
"loss": 0.0665,
|
| 408 |
"step": 300
|
| 409 |
},
|
| 410 |
{
|
| 411 |
+
"epoch": 60.0,
|
| 412 |
+
"eval_loss": 1.7048699855804443,
|
| 413 |
+
"eval_runtime": 1.3753,
|
| 414 |
+
"eval_samples_per_second": 7.271,
|
| 415 |
+
"eval_steps_per_second": 1.454,
|
| 416 |
+
"step": 300
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 61.0,
|
| 420 |
+
"learning_rate": 0.00018059999999999997,
|
| 421 |
+
"loss": 0.065,
|
| 422 |
"step": 305
|
| 423 |
},
|
| 424 |
{
|
| 425 |
+
"epoch": 62.0,
|
| 426 |
+
"learning_rate": 0.0001836,
|
| 427 |
+
"loss": 0.0623,
|
| 428 |
"step": 310
|
| 429 |
},
|
| 430 |
{
|
| 431 |
+
"epoch": 63.0,
|
| 432 |
+
"learning_rate": 0.00018659999999999998,
|
| 433 |
+
"loss": 0.0574,
|
| 434 |
"step": 315
|
| 435 |
},
|
| 436 |
{
|
| 437 |
+
"epoch": 64.0,
|
| 438 |
+
"learning_rate": 0.00018959999999999997,
|
| 439 |
+
"loss": 0.0577,
|
| 440 |
"step": 320
|
| 441 |
},
|
| 442 |
{
|
| 443 |
+
"epoch": 65.0,
|
| 444 |
+
"learning_rate": 0.0001926,
|
| 445 |
+
"loss": 0.0597,
|
| 446 |
"step": 325
|
| 447 |
},
|
| 448 |
{
|
| 449 |
+
"epoch": 66.0,
|
| 450 |
+
"learning_rate": 0.00019559999999999998,
|
| 451 |
+
"loss": 0.0546,
|
| 452 |
"step": 330
|
| 453 |
},
|
| 454 |
{
|
| 455 |
+
"epoch": 67.0,
|
| 456 |
+
"learning_rate": 0.0001986,
|
| 457 |
+
"loss": 0.0603,
|
| 458 |
"step": 335
|
| 459 |
},
|
| 460 |
{
|
| 461 |
+
"epoch": 68.0,
|
| 462 |
+
"learning_rate": 0.0002016,
|
| 463 |
+
"loss": 0.0555,
|
| 464 |
"step": 340
|
| 465 |
},
|
| 466 |
{
|
| 467 |
+
"epoch": 69.0,
|
| 468 |
+
"learning_rate": 0.00020459999999999999,
|
| 469 |
+
"loss": 0.0551,
|
| 470 |
"step": 345
|
| 471 |
},
|
| 472 |
{
|
| 473 |
+
"epoch": 70.0,
|
| 474 |
+
"learning_rate": 0.00020759999999999998,
|
| 475 |
+
"loss": 0.0529,
|
| 476 |
+
"step": 350
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 70.0,
|
| 480 |
+
"eval_loss": 1.901442527770996,
|
| 481 |
+
"eval_runtime": 1.3785,
|
| 482 |
+
"eval_samples_per_second": 7.254,
|
| 483 |
+
"eval_steps_per_second": 1.451,
|
| 484 |
"step": 350
|
| 485 |
},
|
| 486 |
{
|
| 487 |
+
"epoch": 71.0,
|
| 488 |
+
"learning_rate": 0.00021059999999999997,
|
| 489 |
+
"loss": 0.051,
|
| 490 |
"step": 355
|
| 491 |
},
|
| 492 |
{
|
| 493 |
+
"epoch": 72.0,
|
| 494 |
+
"learning_rate": 0.00021359999999999996,
|
| 495 |
+
"loss": 0.05,
|
| 496 |
"step": 360
|
| 497 |
},
|
| 498 |
{
|
| 499 |
+
"epoch": 73.0,
|
| 500 |
+
"learning_rate": 0.00021659999999999998,
|
| 501 |
+
"loss": 0.0493,
|
| 502 |
"step": 365
|
| 503 |
},
|
| 504 |
{
|
| 505 |
+
"epoch": 74.0,
|
| 506 |
+
"learning_rate": 0.00021959999999999997,
|
| 507 |
+
"loss": 0.0465,
|
| 508 |
"step": 370
|
| 509 |
},
|
| 510 |
{
|
| 511 |
+
"epoch": 75.0,
|
| 512 |
+
"learning_rate": 0.0002226,
|
| 513 |
+
"loss": 0.0504,
|
| 514 |
"step": 375
|
| 515 |
},
|
| 516 |
{
|
| 517 |
+
"epoch": 76.0,
|
| 518 |
+
"learning_rate": 0.00022559999999999998,
|
| 519 |
+
"loss": 0.0491,
|
| 520 |
"step": 380
|
| 521 |
},
|
| 522 |
{
|
| 523 |
+
"epoch": 77.0,
|
| 524 |
+
"learning_rate": 0.00022859999999999997,
|
| 525 |
+
"loss": 0.0485,
|
| 526 |
"step": 385
|
| 527 |
},
|
| 528 |
{
|
| 529 |
+
"epoch": 78.0,
|
| 530 |
+
"learning_rate": 0.0002316,
|
| 531 |
+
"loss": 0.0451,
|
| 532 |
"step": 390
|
| 533 |
},
|
| 534 |
{
|
| 535 |
+
"epoch": 79.0,
|
| 536 |
+
"learning_rate": 0.00023459999999999998,
|
| 537 |
+
"loss": 0.0478,
|
| 538 |
"step": 395
|
| 539 |
},
|
| 540 |
{
|
| 541 |
+
"epoch": 80.0,
|
| 542 |
+
"learning_rate": 0.0002376,
|
| 543 |
+
"loss": 0.0435,
|
| 544 |
"step": 400
|
| 545 |
},
|
| 546 |
{
|
| 547 |
+
"epoch": 80.0,
|
| 548 |
+
"eval_loss": 1.9840269088745117,
|
| 549 |
+
"eval_runtime": 1.3787,
|
| 550 |
+
"eval_samples_per_second": 7.253,
|
| 551 |
+
"eval_steps_per_second": 1.451,
|
| 552 |
+
"step": 400
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 81.0,
|
| 556 |
+
"learning_rate": 0.0002406,
|
| 557 |
+
"loss": 0.0429,
|
| 558 |
"step": 405
|
| 559 |
},
|
| 560 |
{
|
| 561 |
+
"epoch": 82.0,
|
| 562 |
+
"learning_rate": 0.00024359999999999999,
|
| 563 |
+
"loss": 0.0464,
|
| 564 |
"step": 410
|
| 565 |
},
|
| 566 |
{
|
| 567 |
+
"epoch": 83.0,
|
| 568 |
+
"learning_rate": 0.0002466,
|
| 569 |
+
"loss": 0.0458,
|
| 570 |
"step": 415
|
| 571 |
},
|
| 572 |
{
|
| 573 |
+
"epoch": 84.0,
|
| 574 |
+
"learning_rate": 0.00024959999999999994,
|
| 575 |
+
"loss": 0.0441,
|
| 576 |
"step": 420
|
| 577 |
},
|
| 578 |
{
|
| 579 |
+
"epoch": 85.0,
|
| 580 |
+
"learning_rate": 0.00025259999999999996,
|
| 581 |
+
"loss": 0.0421,
|
| 582 |
"step": 425
|
| 583 |
},
|
| 584 |
{
|
| 585 |
+
"epoch": 86.0,
|
| 586 |
+
"learning_rate": 0.0002556,
|
| 587 |
+
"loss": 0.0433,
|
| 588 |
"step": 430
|
| 589 |
},
|
| 590 |
{
|
| 591 |
+
"epoch": 87.0,
|
| 592 |
+
"learning_rate": 0.0002586,
|
| 593 |
+
"loss": 0.0444,
|
| 594 |
"step": 435
|
| 595 |
},
|
| 596 |
{
|
| 597 |
+
"epoch": 88.0,
|
| 598 |
+
"learning_rate": 0.00026159999999999996,
|
| 599 |
+
"loss": 0.0472,
|
| 600 |
"step": 440
|
| 601 |
},
|
| 602 |
{
|
| 603 |
+
"epoch": 89.0,
|
| 604 |
+
"learning_rate": 0.0002646,
|
| 605 |
+
"loss": 0.0442,
|
| 606 |
"step": 445
|
| 607 |
},
|
| 608 |
{
|
| 609 |
+
"epoch": 90.0,
|
| 610 |
+
"learning_rate": 0.0002676,
|
| 611 |
+
"loss": 0.0431,
|
| 612 |
+
"step": 450
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 90.0,
|
| 616 |
+
"eval_loss": 1.839582085609436,
|
| 617 |
+
"eval_runtime": 1.3848,
|
| 618 |
+
"eval_samples_per_second": 7.221,
|
| 619 |
+
"eval_steps_per_second": 1.444,
|
| 620 |
"step": 450
|
| 621 |
},
|
| 622 |
{
|
| 623 |
+
"epoch": 91.0,
|
| 624 |
+
"learning_rate": 0.00027059999999999996,
|
| 625 |
+
"loss": 0.0431,
|
| 626 |
"step": 455
|
| 627 |
},
|
| 628 |
{
|
| 629 |
+
"epoch": 92.0,
|
| 630 |
+
"learning_rate": 0.0002736,
|
| 631 |
+
"loss": 0.044,
|
| 632 |
"step": 460
|
| 633 |
},
|
| 634 |
{
|
| 635 |
+
"epoch": 93.0,
|
| 636 |
+
"learning_rate": 0.0002766,
|
| 637 |
+
"loss": 0.0429,
|
| 638 |
"step": 465
|
| 639 |
},
|
| 640 |
{
|
| 641 |
+
"epoch": 94.0,
|
| 642 |
+
"learning_rate": 0.00027959999999999997,
|
| 643 |
+
"loss": 0.042,
|
| 644 |
"step": 470
|
| 645 |
},
|
| 646 |
{
|
| 647 |
+
"epoch": 95.0,
|
| 648 |
+
"learning_rate": 0.0002826,
|
| 649 |
+
"loss": 0.0415,
|
| 650 |
"step": 475
|
| 651 |
},
|
| 652 |
{
|
| 653 |
+
"epoch": 96.0,
|
| 654 |
+
"learning_rate": 0.00028559999999999995,
|
| 655 |
+
"loss": 0.0433,
|
| 656 |
"step": 480
|
| 657 |
},
|
| 658 |
{
|
| 659 |
+
"epoch": 97.0,
|
| 660 |
+
"learning_rate": 0.00028859999999999997,
|
| 661 |
+
"loss": 0.0405,
|
| 662 |
"step": 485
|
| 663 |
},
|
| 664 |
{
|
| 665 |
+
"epoch": 98.0,
|
| 666 |
+
"learning_rate": 0.0002916,
|
| 667 |
+
"loss": 0.0416,
|
| 668 |
"step": 490
|
| 669 |
},
|
| 670 |
{
|
| 671 |
+
"epoch": 99.0,
|
| 672 |
+
"learning_rate": 0.00029459999999999995,
|
| 673 |
+
"loss": 0.0378,
|
| 674 |
"step": 495
|
| 675 |
},
|
| 676 |
{
|
| 677 |
+
"epoch": 100.0,
|
| 678 |
+
"learning_rate": 0.00029759999999999997,
|
| 679 |
+
"loss": 0.0379,
|
| 680 |
"step": 500
|
| 681 |
},
|
| 682 |
{
|
| 683 |
+
"epoch": 100.0,
|
| 684 |
+
"eval_loss": 1.9192386865615845,
|
| 685 |
+
"eval_runtime": 1.3756,
|
| 686 |
+
"eval_samples_per_second": 7.269,
|
| 687 |
+
"eval_steps_per_second": 1.454,
|
| 688 |
"step": 500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 689 |
}
|
| 690 |
],
|
| 691 |
"max_steps": 10000,
|
| 692 |
+
"num_train_epochs": 2000,
|
| 693 |
+
"total_flos": 8.0956959522816e+16,
|
| 694 |
"trial_name": null,
|
| 695 |
"trial_params": null
|
| 696 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3963
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ccf3cbed56dd6bc8e51a2fe8aa833170620c235dac2584c4369f03afd0d2c443
|
| 3 |
size 3963
|