End of training
Browse files- README.md +2 -2
- all_results.json +6 -6
- runs/Feb18_07-17-35_f3548d498004/events.out.tfevents.1708241409.f3548d498004.5521.4 +2 -2
- test_results.json +6 -6
- trainer_state.json +1545 -411
README.md
CHANGED
|
@@ -17,8 +17,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 17 |
|
| 18 |
This model is a fine-tuned version of [facebook/timesformer-base-finetuned-k400](https://huggingface.co/facebook/timesformer-base-finetuned-k400) on an unknown dataset.
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
-
- Loss: 1.
|
| 21 |
-
- Accuracy: 0.
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
|
|
|
| 17 |
|
| 18 |
This model is a fine-tuned version of [facebook/timesformer-base-finetuned-k400](https://huggingface.co/facebook/timesformer-base-finetuned-k400) on an unknown dataset.
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
+
- Loss: 1.6472
|
| 21 |
+
- Accuracy: 0.5994
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
all_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss":
|
| 5 |
-
"eval_runtime": 27.
|
| 6 |
-
"eval_samples_per_second": 12.
|
| 7 |
-
"eval_steps_per_second": 12.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 9.1,
|
| 3 |
+
"eval_accuracy": 0.5994236311239193,
|
| 4 |
+
"eval_loss": 1.647168755531311,
|
| 5 |
+
"eval_runtime": 27.7986,
|
| 6 |
+
"eval_samples_per_second": 12.483,
|
| 7 |
+
"eval_steps_per_second": 12.483
|
| 8 |
}
|
runs/Feb18_07-17-35_f3548d498004/events.out.tfevents.1708241409.f3548d498004.5521.4
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e99826f80870ce425c52b4dae60068e728d817e23c0800d7b7ee94e6aa06d7b0
|
| 3 |
+
size 734
|
test_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss":
|
| 5 |
-
"eval_runtime": 27.
|
| 6 |
-
"eval_samples_per_second": 12.
|
| 7 |
-
"eval_steps_per_second": 12.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 9.1,
|
| 3 |
+
"eval_accuracy": 0.5994236311239193,
|
| 4 |
+
"eval_loss": 1.647168755531311,
|
| 5 |
+
"eval_runtime": 27.7986,
|
| 6 |
+
"eval_samples_per_second": 12.483,
|
| 7 |
+
"eval_steps_per_second": 12.483
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,802 +1,1936 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "timesformer-base-finetuned-k400-finetuned-ucf101-subset/checkpoint-
|
| 4 |
-
"epoch":
|
| 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.
|
| 13 |
-
"learning_rate":
|
| 14 |
-
"loss":
|
| 15 |
"step": 10
|
| 16 |
},
|
| 17 |
{
|
| 18 |
-
"epoch": 0.
|
| 19 |
-
"learning_rate":
|
| 20 |
-
"loss":
|
| 21 |
"step": 20
|
| 22 |
},
|
| 23 |
{
|
| 24 |
-
"epoch": 0.
|
| 25 |
-
"learning_rate":
|
| 26 |
-
"loss":
|
| 27 |
"step": 30
|
| 28 |
},
|
| 29 |
{
|
| 30 |
-
"epoch": 0.
|
| 31 |
-
"learning_rate":
|
| 32 |
-
"loss":
|
| 33 |
"step": 40
|
| 34 |
},
|
| 35 |
{
|
| 36 |
-
"epoch": 0.
|
| 37 |
-
"learning_rate":
|
| 38 |
-
"loss":
|
| 39 |
"step": 50
|
| 40 |
},
|
| 41 |
{
|
| 42 |
-
"epoch": 0.
|
| 43 |
-
"learning_rate":
|
| 44 |
-
"loss":
|
| 45 |
"step": 60
|
| 46 |
},
|
| 47 |
{
|
| 48 |
-
"epoch": 0.
|
| 49 |
-
"learning_rate":
|
| 50 |
-
"loss":
|
| 51 |
"step": 70
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch": 0.
|
| 55 |
-
"learning_rate":
|
| 56 |
-
"loss":
|
| 57 |
"step": 80
|
| 58 |
},
|
| 59 |
{
|
| 60 |
-
"epoch": 0.
|
| 61 |
-
"learning_rate":
|
| 62 |
-
"loss":
|
| 63 |
"step": 90
|
| 64 |
},
|
| 65 |
{
|
| 66 |
-
"epoch": 0.
|
| 67 |
-
"learning_rate":
|
| 68 |
-
"loss":
|
| 69 |
"step": 100
|
| 70 |
},
|
| 71 |
{
|
| 72 |
-
"epoch": 0.
|
| 73 |
-
"learning_rate":
|
| 74 |
-
"loss":
|
| 75 |
"step": 110
|
| 76 |
},
|
| 77 |
{
|
| 78 |
-
"epoch": 0.
|
| 79 |
-
"learning_rate":
|
| 80 |
-
"loss":
|
| 81 |
"step": 120
|
| 82 |
},
|
| 83 |
{
|
| 84 |
-
"epoch": 0.
|
| 85 |
-
"learning_rate":
|
| 86 |
-
"loss":
|
| 87 |
"step": 130
|
| 88 |
},
|
| 89 |
{
|
| 90 |
-
"epoch": 0.
|
| 91 |
-
"learning_rate":
|
| 92 |
-
"loss": 2.
|
| 93 |
"step": 140
|
| 94 |
},
|
| 95 |
{
|
| 96 |
-
"epoch": 0.
|
| 97 |
-
"learning_rate":
|
| 98 |
-
"loss": 2.
|
| 99 |
"step": 150
|
| 100 |
},
|
| 101 |
{
|
| 102 |
-
"epoch": 0.
|
| 103 |
-
"learning_rate":
|
| 104 |
-
"loss":
|
| 105 |
"step": 160
|
| 106 |
},
|
| 107 |
{
|
| 108 |
-
"epoch": 0.
|
| 109 |
-
"learning_rate":
|
| 110 |
-
"loss":
|
| 111 |
"step": 170
|
| 112 |
},
|
| 113 |
{
|
| 114 |
-
"epoch": 0.
|
| 115 |
-
"learning_rate":
|
| 116 |
-
"loss":
|
| 117 |
"step": 180
|
| 118 |
},
|
| 119 |
{
|
| 120 |
-
"epoch": 0.
|
| 121 |
-
"learning_rate":
|
| 122 |
-
"loss": 2.
|
| 123 |
"step": 190
|
| 124 |
},
|
| 125 |
{
|
| 126 |
-
"epoch": 0.
|
| 127 |
-
"learning_rate":
|
| 128 |
-
"loss":
|
| 129 |
"step": 200
|
| 130 |
},
|
| 131 |
{
|
| 132 |
-
"epoch": 0.
|
| 133 |
-
"learning_rate":
|
| 134 |
-
"loss":
|
| 135 |
"step": 210
|
| 136 |
},
|
| 137 |
{
|
| 138 |
-
"epoch": 0.
|
| 139 |
-
"learning_rate":
|
| 140 |
-
"loss": 2.
|
| 141 |
"step": 220
|
| 142 |
},
|
| 143 |
{
|
| 144 |
-
"epoch": 0.
|
| 145 |
-
"learning_rate":
|
| 146 |
-
"loss":
|
| 147 |
"step": 230
|
| 148 |
},
|
| 149 |
{
|
| 150 |
-
"epoch": 0.
|
| 151 |
-
"learning_rate":
|
| 152 |
-
"loss":
|
| 153 |
"step": 240
|
| 154 |
},
|
| 155 |
{
|
| 156 |
-
"epoch": 0.
|
| 157 |
-
"learning_rate": 4.
|
| 158 |
-
"loss": 2.
|
| 159 |
"step": 250
|
| 160 |
},
|
| 161 |
{
|
| 162 |
-
"epoch": 0.
|
| 163 |
-
"learning_rate": 4.
|
| 164 |
-
"loss":
|
| 165 |
"step": 260
|
| 166 |
},
|
| 167 |
{
|
| 168 |
-
"epoch": 0.
|
| 169 |
-
"learning_rate": 4.
|
| 170 |
-
"loss":
|
| 171 |
"step": 270
|
| 172 |
},
|
| 173 |
{
|
| 174 |
-
"epoch": 0.
|
| 175 |
-
"learning_rate": 4.
|
| 176 |
-
"loss": 2.
|
| 177 |
"step": 280
|
| 178 |
},
|
| 179 |
{
|
| 180 |
-
"epoch": 0.
|
| 181 |
-
"learning_rate": 4.
|
| 182 |
-
"loss":
|
| 183 |
"step": 290
|
| 184 |
},
|
| 185 |
{
|
| 186 |
-
"epoch": 0.
|
| 187 |
-
"learning_rate":
|
| 188 |
-
"loss":
|
| 189 |
"step": 300
|
| 190 |
},
|
| 191 |
{
|
| 192 |
-
"epoch": 0.
|
| 193 |
-
"eval_accuracy": 0.
|
| 194 |
-
"eval_loss":
|
| 195 |
-
"eval_runtime": 11.
|
| 196 |
-
"eval_samples_per_second": 12.
|
| 197 |
-
"eval_steps_per_second": 12.
|
| 198 |
"step": 300
|
| 199 |
},
|
| 200 |
{
|
| 201 |
-
"epoch": 1.
|
| 202 |
-
"learning_rate": 4.
|
| 203 |
-
"loss":
|
| 204 |
"step": 310
|
| 205 |
},
|
| 206 |
{
|
| 207 |
-
"epoch": 1.
|
| 208 |
-
"learning_rate": 4.
|
| 209 |
-
"loss":
|
| 210 |
"step": 320
|
| 211 |
},
|
| 212 |
{
|
| 213 |
-
"epoch": 1.
|
| 214 |
-
"learning_rate": 4.
|
| 215 |
-
"loss": 2.
|
| 216 |
"step": 330
|
| 217 |
},
|
| 218 |
{
|
| 219 |
-
"epoch": 1.
|
| 220 |
-
"learning_rate":
|
| 221 |
-
"loss":
|
| 222 |
"step": 340
|
| 223 |
},
|
| 224 |
{
|
| 225 |
-
"epoch": 1.
|
| 226 |
-
"learning_rate":
|
| 227 |
-
"loss":
|
| 228 |
"step": 350
|
| 229 |
},
|
| 230 |
{
|
| 231 |
-
"epoch": 1.
|
| 232 |
-
"learning_rate":
|
| 233 |
-
"loss":
|
| 234 |
"step": 360
|
| 235 |
},
|
| 236 |
{
|
| 237 |
-
"epoch": 1.
|
| 238 |
-
"learning_rate":
|
| 239 |
-
"loss": 2.
|
| 240 |
"step": 370
|
| 241 |
},
|
| 242 |
{
|
| 243 |
-
"epoch": 1.
|
| 244 |
-
"learning_rate":
|
| 245 |
-
"loss": 2.
|
| 246 |
"step": 380
|
| 247 |
},
|
| 248 |
{
|
| 249 |
-
"epoch": 1.
|
| 250 |
-
"learning_rate":
|
| 251 |
-
"loss":
|
| 252 |
"step": 390
|
| 253 |
},
|
| 254 |
{
|
| 255 |
-
"epoch": 1.
|
| 256 |
-
"learning_rate":
|
| 257 |
-
"loss": 2.
|
| 258 |
"step": 400
|
| 259 |
},
|
| 260 |
{
|
| 261 |
-
"epoch": 1.
|
| 262 |
-
"learning_rate":
|
| 263 |
-
"loss":
|
| 264 |
"step": 410
|
| 265 |
},
|
| 266 |
{
|
| 267 |
-
"epoch": 1.
|
| 268 |
-
"learning_rate":
|
| 269 |
-
"loss":
|
| 270 |
"step": 420
|
| 271 |
},
|
| 272 |
{
|
| 273 |
-
"epoch": 1.
|
| 274 |
-
"learning_rate":
|
| 275 |
-
"loss":
|
| 276 |
"step": 430
|
| 277 |
},
|
| 278 |
{
|
| 279 |
-
"epoch": 1.
|
| 280 |
-
"learning_rate":
|
| 281 |
-
"loss":
|
| 282 |
"step": 440
|
| 283 |
},
|
| 284 |
{
|
| 285 |
-
"epoch": 1.
|
| 286 |
-
"learning_rate":
|
| 287 |
-
"loss":
|
| 288 |
"step": 450
|
| 289 |
},
|
| 290 |
{
|
| 291 |
-
"epoch": 1.
|
| 292 |
-
"learning_rate":
|
| 293 |
-
"loss":
|
| 294 |
"step": 460
|
| 295 |
},
|
| 296 |
{
|
| 297 |
-
"epoch": 1.
|
| 298 |
-
"learning_rate":
|
| 299 |
-
"loss":
|
| 300 |
"step": 470
|
| 301 |
},
|
| 302 |
{
|
| 303 |
-
"epoch": 1.
|
| 304 |
-
"learning_rate":
|
| 305 |
-
"loss": 1.
|
| 306 |
"step": 480
|
| 307 |
},
|
| 308 |
{
|
| 309 |
-
"epoch": 1.
|
| 310 |
-
"learning_rate":
|
| 311 |
-
"loss": 1.
|
| 312 |
"step": 490
|
| 313 |
},
|
| 314 |
{
|
| 315 |
-
"epoch": 1.
|
| 316 |
-
"learning_rate":
|
| 317 |
-
"loss":
|
| 318 |
"step": 500
|
| 319 |
},
|
| 320 |
{
|
| 321 |
-
"epoch": 1.
|
| 322 |
-
"learning_rate":
|
| 323 |
-
"loss":
|
| 324 |
"step": 510
|
| 325 |
},
|
| 326 |
{
|
| 327 |
-
"epoch": 1.
|
| 328 |
-
"learning_rate":
|
| 329 |
-
"loss": 2.
|
| 330 |
"step": 520
|
| 331 |
},
|
| 332 |
{
|
| 333 |
-
"epoch": 1.
|
| 334 |
-
"learning_rate":
|
| 335 |
-
"loss":
|
| 336 |
"step": 530
|
| 337 |
},
|
| 338 |
{
|
| 339 |
-
"epoch": 1.
|
| 340 |
-
"learning_rate":
|
| 341 |
-
"loss":
|
| 342 |
"step": 540
|
| 343 |
},
|
| 344 |
{
|
| 345 |
-
"epoch": 1.
|
| 346 |
-
"learning_rate":
|
| 347 |
-
"loss": 2.
|
| 348 |
"step": 550
|
| 349 |
},
|
| 350 |
{
|
| 351 |
-
"epoch": 1.
|
| 352 |
-
"learning_rate":
|
| 353 |
-
"loss": 1.
|
| 354 |
"step": 560
|
| 355 |
},
|
| 356 |
{
|
| 357 |
-
"epoch": 1.
|
| 358 |
-
"learning_rate":
|
| 359 |
-
"loss": 2.
|
| 360 |
"step": 570
|
| 361 |
},
|
| 362 |
{
|
| 363 |
-
"epoch": 1.
|
| 364 |
-
"learning_rate":
|
| 365 |
-
"loss": 1.
|
| 366 |
"step": 580
|
| 367 |
},
|
| 368 |
{
|
| 369 |
-
"epoch": 1.
|
| 370 |
-
"learning_rate":
|
| 371 |
-
"loss":
|
| 372 |
"step": 590
|
| 373 |
},
|
| 374 |
{
|
| 375 |
-
"epoch": 1.
|
| 376 |
-
"learning_rate":
|
| 377 |
-
"loss": 2.
|
| 378 |
"step": 600
|
| 379 |
},
|
| 380 |
{
|
| 381 |
-
"epoch": 1.
|
| 382 |
-
"eval_accuracy": 0.
|
| 383 |
-
"eval_loss": 2.
|
| 384 |
-
"eval_runtime": 11.
|
| 385 |
-
"eval_samples_per_second": 12.
|
| 386 |
-
"eval_steps_per_second": 12.
|
| 387 |
"step": 600
|
| 388 |
},
|
| 389 |
{
|
| 390 |
-
"epoch": 2.
|
| 391 |
-
"learning_rate":
|
| 392 |
-
"loss": 2.
|
| 393 |
"step": 610
|
| 394 |
},
|
| 395 |
{
|
| 396 |
-
"epoch": 2.
|
| 397 |
-
"learning_rate":
|
| 398 |
-
"loss":
|
| 399 |
"step": 620
|
| 400 |
},
|
| 401 |
{
|
| 402 |
-
"epoch": 2.
|
| 403 |
-
"learning_rate":
|
| 404 |
-
"loss": 2.
|
| 405 |
"step": 630
|
| 406 |
},
|
| 407 |
{
|
| 408 |
-
"epoch": 2.
|
| 409 |
-
"learning_rate":
|
| 410 |
-
"loss":
|
| 411 |
"step": 640
|
| 412 |
},
|
| 413 |
{
|
| 414 |
-
"epoch": 2.
|
| 415 |
-
"learning_rate":
|
| 416 |
-
"loss":
|
| 417 |
"step": 650
|
| 418 |
},
|
| 419 |
{
|
| 420 |
-
"epoch": 2.
|
| 421 |
-
"learning_rate":
|
| 422 |
-
"loss": 1.
|
| 423 |
"step": 660
|
| 424 |
},
|
| 425 |
{
|
| 426 |
-
"epoch": 2.
|
| 427 |
-
"learning_rate":
|
| 428 |
-
"loss":
|
| 429 |
"step": 670
|
| 430 |
},
|
| 431 |
{
|
| 432 |
-
"epoch": 2.
|
| 433 |
-
"learning_rate":
|
| 434 |
-
"loss":
|
| 435 |
"step": 680
|
| 436 |
},
|
| 437 |
{
|
| 438 |
-
"epoch": 2.
|
| 439 |
-
"learning_rate":
|
| 440 |
-
"loss": 1.
|
| 441 |
"step": 690
|
| 442 |
},
|
| 443 |
{
|
| 444 |
-
"epoch": 2.
|
| 445 |
-
"learning_rate":
|
| 446 |
-
"loss":
|
| 447 |
"step": 700
|
| 448 |
},
|
| 449 |
{
|
| 450 |
-
"epoch": 2.
|
| 451 |
-
"learning_rate":
|
| 452 |
-
"loss":
|
| 453 |
"step": 710
|
| 454 |
},
|
| 455 |
{
|
| 456 |
-
"epoch": 2.
|
| 457 |
-
"learning_rate":
|
| 458 |
-
"loss":
|
| 459 |
"step": 720
|
| 460 |
},
|
| 461 |
{
|
| 462 |
-
"epoch": 2.
|
| 463 |
-
"learning_rate":
|
| 464 |
-
"loss": 2.
|
| 465 |
"step": 730
|
| 466 |
},
|
| 467 |
{
|
| 468 |
-
"epoch": 2.
|
| 469 |
-
"learning_rate":
|
| 470 |
-
"loss": 1.
|
| 471 |
"step": 740
|
| 472 |
},
|
| 473 |
{
|
| 474 |
-
"epoch": 2.
|
| 475 |
-
"learning_rate":
|
| 476 |
-
"loss": 1.
|
| 477 |
"step": 750
|
| 478 |
},
|
| 479 |
{
|
| 480 |
-
"epoch": 2.
|
| 481 |
-
"learning_rate":
|
| 482 |
-
"loss": 2.
|
| 483 |
"step": 760
|
| 484 |
},
|
| 485 |
{
|
| 486 |
-
"epoch": 2.
|
| 487 |
-
"learning_rate":
|
| 488 |
-
"loss": 1.
|
| 489 |
"step": 770
|
| 490 |
},
|
| 491 |
{
|
| 492 |
-
"epoch": 2.
|
| 493 |
-
"learning_rate":
|
| 494 |
-
"loss":
|
| 495 |
"step": 780
|
| 496 |
},
|
| 497 |
{
|
| 498 |
-
"epoch": 2.
|
| 499 |
-
"learning_rate":
|
| 500 |
-
"loss":
|
| 501 |
"step": 790
|
| 502 |
},
|
| 503 |
{
|
| 504 |
-
"epoch": 2.
|
| 505 |
-
"learning_rate":
|
| 506 |
-
"loss": 1.
|
| 507 |
"step": 800
|
| 508 |
},
|
| 509 |
{
|
| 510 |
-
"epoch": 2.
|
| 511 |
-
"learning_rate":
|
| 512 |
-
"loss":
|
| 513 |
"step": 810
|
| 514 |
},
|
| 515 |
{
|
| 516 |
-
"epoch": 2.
|
| 517 |
-
"learning_rate":
|
| 518 |
-
"loss": 2.
|
| 519 |
"step": 820
|
| 520 |
},
|
| 521 |
{
|
| 522 |
-
"epoch": 2.
|
| 523 |
-
"learning_rate":
|
| 524 |
-
"loss":
|
| 525 |
"step": 830
|
| 526 |
},
|
| 527 |
{
|
| 528 |
-
"epoch": 2.
|
| 529 |
-
"learning_rate":
|
| 530 |
-
"loss": 2.
|
| 531 |
"step": 840
|
| 532 |
},
|
| 533 |
{
|
| 534 |
-
"epoch": 2.
|
| 535 |
-
"learning_rate":
|
| 536 |
-
"loss":
|
| 537 |
"step": 850
|
| 538 |
},
|
| 539 |
{
|
| 540 |
-
"epoch": 2.
|
| 541 |
-
"learning_rate":
|
| 542 |
-
"loss":
|
| 543 |
"step": 860
|
| 544 |
},
|
| 545 |
{
|
| 546 |
-
"epoch": 2.
|
| 547 |
-
"learning_rate":
|
| 548 |
-
"loss":
|
| 549 |
"step": 870
|
| 550 |
},
|
| 551 |
{
|
| 552 |
-
"epoch": 2.
|
| 553 |
-
"learning_rate":
|
| 554 |
-
"loss":
|
| 555 |
"step": 880
|
| 556 |
},
|
| 557 |
{
|
| 558 |
-
"epoch": 2.
|
| 559 |
-
"learning_rate":
|
| 560 |
-
"loss": 1.
|
| 561 |
"step": 890
|
| 562 |
},
|
| 563 |
{
|
| 564 |
-
"epoch": 2.
|
| 565 |
-
"learning_rate":
|
| 566 |
-
"loss":
|
| 567 |
"step": 900
|
| 568 |
},
|
| 569 |
{
|
| 570 |
-
"epoch": 2.
|
| 571 |
-
"eval_accuracy": 0.
|
| 572 |
-
"eval_loss":
|
| 573 |
-
"eval_runtime": 11.
|
| 574 |
-
"eval_samples_per_second": 12.
|
| 575 |
-
"eval_steps_per_second": 12.
|
| 576 |
"step": 900
|
| 577 |
},
|
| 578 |
{
|
| 579 |
-
"epoch": 3.
|
| 580 |
-
"learning_rate":
|
| 581 |
-
"loss": 1.
|
| 582 |
"step": 910
|
| 583 |
},
|
| 584 |
{
|
| 585 |
-
"epoch": 3.
|
| 586 |
-
"learning_rate":
|
| 587 |
-
"loss": 1.
|
| 588 |
"step": 920
|
| 589 |
},
|
| 590 |
{
|
| 591 |
-
"epoch": 3.
|
| 592 |
-
"learning_rate":
|
| 593 |
-
"loss":
|
| 594 |
"step": 930
|
| 595 |
},
|
| 596 |
{
|
| 597 |
-
"epoch": 3.
|
| 598 |
-
"learning_rate":
|
| 599 |
-
"loss": 1.
|
| 600 |
"step": 940
|
| 601 |
},
|
| 602 |
{
|
| 603 |
-
"epoch": 3.
|
| 604 |
-
"learning_rate":
|
| 605 |
-
"loss":
|
| 606 |
"step": 950
|
| 607 |
},
|
| 608 |
{
|
| 609 |
-
"epoch": 3.
|
| 610 |
-
"learning_rate":
|
| 611 |
-
"loss": 1.
|
| 612 |
"step": 960
|
| 613 |
},
|
| 614 |
{
|
| 615 |
-
"epoch": 3.
|
| 616 |
-
"learning_rate":
|
| 617 |
-
"loss": 1.
|
| 618 |
"step": 970
|
| 619 |
},
|
| 620 |
{
|
| 621 |
-
"epoch": 3.
|
| 622 |
-
"learning_rate":
|
| 623 |
-
"loss":
|
| 624 |
"step": 980
|
| 625 |
},
|
| 626 |
{
|
| 627 |
-
"epoch": 3.
|
| 628 |
-
"learning_rate":
|
| 629 |
-
"loss":
|
| 630 |
"step": 990
|
| 631 |
},
|
| 632 |
{
|
| 633 |
-
"epoch": 3.
|
| 634 |
-
"learning_rate":
|
| 635 |
-
"loss":
|
| 636 |
"step": 1000
|
| 637 |
},
|
| 638 |
{
|
| 639 |
-
"epoch": 3.
|
| 640 |
-
"learning_rate":
|
| 641 |
-
"loss": 1.
|
| 642 |
"step": 1010
|
| 643 |
},
|
| 644 |
{
|
| 645 |
-
"epoch": 3.
|
| 646 |
-
"learning_rate":
|
| 647 |
-
"loss": 2.
|
| 648 |
"step": 1020
|
| 649 |
},
|
| 650 |
{
|
| 651 |
-
"epoch": 3.
|
| 652 |
-
"learning_rate":
|
| 653 |
-
"loss":
|
| 654 |
"step": 1030
|
| 655 |
},
|
| 656 |
{
|
| 657 |
-
"epoch": 3.
|
| 658 |
-
"learning_rate":
|
| 659 |
-
"loss":
|
| 660 |
"step": 1040
|
| 661 |
},
|
| 662 |
{
|
| 663 |
-
"epoch": 3.
|
| 664 |
-
"learning_rate":
|
| 665 |
-
"loss":
|
| 666 |
"step": 1050
|
| 667 |
},
|
| 668 |
{
|
| 669 |
-
"epoch": 3.
|
| 670 |
-
"learning_rate":
|
| 671 |
-
"loss": 1.
|
| 672 |
"step": 1060
|
| 673 |
},
|
| 674 |
{
|
| 675 |
-
"epoch": 3.
|
| 676 |
-
"learning_rate":
|
| 677 |
-
"loss": 1.
|
| 678 |
"step": 1070
|
| 679 |
},
|
| 680 |
{
|
| 681 |
-
"epoch": 3.
|
| 682 |
-
"learning_rate":
|
| 683 |
-
"loss":
|
| 684 |
"step": 1080
|
| 685 |
},
|
| 686 |
{
|
| 687 |
-
"epoch": 3.
|
| 688 |
-
"learning_rate":
|
| 689 |
-
"loss": 1.
|
| 690 |
"step": 1090
|
| 691 |
},
|
| 692 |
{
|
| 693 |
-
"epoch": 3.
|
| 694 |
-
"learning_rate":
|
| 695 |
-
"loss":
|
| 696 |
"step": 1100
|
| 697 |
},
|
| 698 |
{
|
| 699 |
-
"epoch": 3.
|
| 700 |
-
"learning_rate":
|
| 701 |
-
"loss": 2.
|
| 702 |
"step": 1110
|
| 703 |
},
|
| 704 |
{
|
| 705 |
-
"epoch": 3.
|
| 706 |
-
"learning_rate": 3.
|
| 707 |
-
"loss": 1.
|
| 708 |
"step": 1120
|
| 709 |
},
|
| 710 |
{
|
| 711 |
-
"epoch": 3.
|
| 712 |
-
"learning_rate": 3.
|
| 713 |
-
"loss": 1.
|
| 714 |
"step": 1130
|
| 715 |
},
|
| 716 |
{
|
| 717 |
-
"epoch": 3.
|
| 718 |
-
"learning_rate":
|
| 719 |
-
"loss":
|
| 720 |
"step": 1140
|
| 721 |
},
|
| 722 |
{
|
| 723 |
-
"epoch": 3.
|
| 724 |
-
"learning_rate":
|
| 725 |
-
"loss":
|
| 726 |
"step": 1150
|
| 727 |
},
|
| 728 |
{
|
| 729 |
-
"epoch": 3.
|
| 730 |
-
"learning_rate":
|
| 731 |
-
"loss": 1.
|
| 732 |
"step": 1160
|
| 733 |
},
|
| 734 |
{
|
| 735 |
-
"epoch": 3.
|
| 736 |
-
"learning_rate":
|
| 737 |
-
"loss":
|
| 738 |
"step": 1170
|
| 739 |
},
|
| 740 |
{
|
| 741 |
-
"epoch": 3.
|
| 742 |
-
"learning_rate":
|
| 743 |
-
"loss":
|
| 744 |
"step": 1180
|
| 745 |
},
|
| 746 |
{
|
| 747 |
-
"epoch": 3.
|
| 748 |
-
"learning_rate":
|
| 749 |
-
"loss": 1.
|
| 750 |
"step": 1190
|
| 751 |
},
|
| 752 |
{
|
| 753 |
-
"epoch": 3.
|
| 754 |
-
"learning_rate":
|
| 755 |
-
"loss":
|
| 756 |
"step": 1200
|
| 757 |
},
|
| 758 |
{
|
| 759 |
-
"epoch": 3.
|
| 760 |
-
"eval_accuracy": 0.
|
| 761 |
-
"eval_loss": 2.
|
| 762 |
-
"eval_runtime": 11.
|
| 763 |
-
"eval_samples_per_second": 12.
|
| 764 |
-
"eval_steps_per_second": 12.
|
| 765 |
"step": 1200
|
| 766 |
},
|
| 767 |
{
|
| 768 |
-
"epoch":
|
| 769 |
-
"
|
| 770 |
-
"
|
| 771 |
-
"
|
| 772 |
-
"train_runtime": 280.7931,
|
| 773 |
-
"train_samples_per_second": 4.274,
|
| 774 |
-
"train_steps_per_second": 4.274
|
| 775 |
},
|
| 776 |
{
|
| 777 |
-
"epoch":
|
| 778 |
-
"
|
| 779 |
-
"
|
| 780 |
-
"
|
| 781 |
-
"eval_samples_per_second": 12.817,
|
| 782 |
-
"eval_steps_per_second": 12.817,
|
| 783 |
-
"step": 1200
|
| 784 |
},
|
| 785 |
{
|
| 786 |
-
"epoch":
|
| 787 |
-
"
|
| 788 |
-
"
|
| 789 |
-
"
|
| 790 |
-
|
| 791 |
-
|
| 792 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 793 |
}
|
| 794 |
],
|
| 795 |
"logging_steps": 10,
|
| 796 |
-
"max_steps":
|
| 797 |
"num_train_epochs": 9223372036854775807,
|
| 798 |
"save_steps": 500,
|
| 799 |
-
"total_flos":
|
| 800 |
"trial_name": null,
|
| 801 |
"trial_params": null
|
| 802 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.436241610738255,
|
| 3 |
+
"best_model_checkpoint": "timesformer-base-finetuned-k400-finetuned-ucf101-subset/checkpoint-2700",
|
| 4 |
+
"epoch": 9.1,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 3000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.0,
|
| 13 |
+
"learning_rate": 1.6666666666666667e-06,
|
| 14 |
+
"loss": 1.3924,
|
| 15 |
"step": 10
|
| 16 |
},
|
| 17 |
{
|
| 18 |
+
"epoch": 0.01,
|
| 19 |
+
"learning_rate": 3.3333333333333333e-06,
|
| 20 |
+
"loss": 1.9129,
|
| 21 |
"step": 20
|
| 22 |
},
|
| 23 |
{
|
| 24 |
+
"epoch": 0.01,
|
| 25 |
+
"learning_rate": 5e-06,
|
| 26 |
+
"loss": 1.3532,
|
| 27 |
"step": 30
|
| 28 |
},
|
| 29 |
{
|
| 30 |
+
"epoch": 0.01,
|
| 31 |
+
"learning_rate": 6.666666666666667e-06,
|
| 32 |
+
"loss": 1.12,
|
| 33 |
"step": 40
|
| 34 |
},
|
| 35 |
{
|
| 36 |
+
"epoch": 0.02,
|
| 37 |
+
"learning_rate": 8.333333333333334e-06,
|
| 38 |
+
"loss": 1.14,
|
| 39 |
"step": 50
|
| 40 |
},
|
| 41 |
{
|
| 42 |
+
"epoch": 0.02,
|
| 43 |
+
"learning_rate": 1e-05,
|
| 44 |
+
"loss": 1.9799,
|
| 45 |
"step": 60
|
| 46 |
},
|
| 47 |
{
|
| 48 |
+
"epoch": 0.02,
|
| 49 |
+
"learning_rate": 1.1666666666666668e-05,
|
| 50 |
+
"loss": 1.6834,
|
| 51 |
"step": 70
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.03,
|
| 55 |
+
"learning_rate": 1.3333333333333333e-05,
|
| 56 |
+
"loss": 1.7657,
|
| 57 |
"step": 80
|
| 58 |
},
|
| 59 |
{
|
| 60 |
+
"epoch": 0.03,
|
| 61 |
+
"learning_rate": 1.5e-05,
|
| 62 |
+
"loss": 0.8917,
|
| 63 |
"step": 90
|
| 64 |
},
|
| 65 |
{
|
| 66 |
+
"epoch": 0.03,
|
| 67 |
+
"learning_rate": 1.6666666666666667e-05,
|
| 68 |
+
"loss": 1.8642,
|
| 69 |
"step": 100
|
| 70 |
},
|
| 71 |
{
|
| 72 |
+
"epoch": 0.04,
|
| 73 |
+
"learning_rate": 1.8333333333333333e-05,
|
| 74 |
+
"loss": 2.9366,
|
| 75 |
"step": 110
|
| 76 |
},
|
| 77 |
{
|
| 78 |
+
"epoch": 0.04,
|
| 79 |
+
"learning_rate": 2e-05,
|
| 80 |
+
"loss": 1.7314,
|
| 81 |
"step": 120
|
| 82 |
},
|
| 83 |
{
|
| 84 |
+
"epoch": 0.04,
|
| 85 |
+
"learning_rate": 2.1666666666666667e-05,
|
| 86 |
+
"loss": 1.2063,
|
| 87 |
"step": 130
|
| 88 |
},
|
| 89 |
{
|
| 90 |
+
"epoch": 0.05,
|
| 91 |
+
"learning_rate": 2.3333333333333336e-05,
|
| 92 |
+
"loss": 2.1836,
|
| 93 |
"step": 140
|
| 94 |
},
|
| 95 |
{
|
| 96 |
+
"epoch": 0.05,
|
| 97 |
+
"learning_rate": 2.5e-05,
|
| 98 |
+
"loss": 2.3437,
|
| 99 |
"step": 150
|
| 100 |
},
|
| 101 |
{
|
| 102 |
+
"epoch": 0.05,
|
| 103 |
+
"learning_rate": 2.6666666666666667e-05,
|
| 104 |
+
"loss": 3.0346,
|
| 105 |
"step": 160
|
| 106 |
},
|
| 107 |
{
|
| 108 |
+
"epoch": 0.06,
|
| 109 |
+
"learning_rate": 2.8333333333333335e-05,
|
| 110 |
+
"loss": 1.5671,
|
| 111 |
"step": 170
|
| 112 |
},
|
| 113 |
{
|
| 114 |
+
"epoch": 0.06,
|
| 115 |
+
"learning_rate": 3e-05,
|
| 116 |
+
"loss": 0.9069,
|
| 117 |
"step": 180
|
| 118 |
},
|
| 119 |
{
|
| 120 |
+
"epoch": 0.06,
|
| 121 |
+
"learning_rate": 3.1666666666666666e-05,
|
| 122 |
+
"loss": 2.7771,
|
| 123 |
"step": 190
|
| 124 |
},
|
| 125 |
{
|
| 126 |
+
"epoch": 0.07,
|
| 127 |
+
"learning_rate": 3.3333333333333335e-05,
|
| 128 |
+
"loss": 1.5606,
|
| 129 |
"step": 200
|
| 130 |
},
|
| 131 |
{
|
| 132 |
+
"epoch": 0.07,
|
| 133 |
+
"learning_rate": 3.5e-05,
|
| 134 |
+
"loss": 2.1344,
|
| 135 |
"step": 210
|
| 136 |
},
|
| 137 |
{
|
| 138 |
+
"epoch": 0.07,
|
| 139 |
+
"learning_rate": 3.6666666666666666e-05,
|
| 140 |
+
"loss": 2.5326,
|
| 141 |
"step": 220
|
| 142 |
},
|
| 143 |
{
|
| 144 |
+
"epoch": 0.08,
|
| 145 |
+
"learning_rate": 3.8333333333333334e-05,
|
| 146 |
+
"loss": 3.2706,
|
| 147 |
"step": 230
|
| 148 |
},
|
| 149 |
{
|
| 150 |
+
"epoch": 0.08,
|
| 151 |
+
"learning_rate": 4e-05,
|
| 152 |
+
"loss": 1.9432,
|
| 153 |
"step": 240
|
| 154 |
},
|
| 155 |
{
|
| 156 |
+
"epoch": 0.08,
|
| 157 |
+
"learning_rate": 4.166666666666667e-05,
|
| 158 |
+
"loss": 2.5327,
|
| 159 |
"step": 250
|
| 160 |
},
|
| 161 |
{
|
| 162 |
+
"epoch": 0.09,
|
| 163 |
+
"learning_rate": 4.3333333333333334e-05,
|
| 164 |
+
"loss": 2.7838,
|
| 165 |
"step": 260
|
| 166 |
},
|
| 167 |
{
|
| 168 |
+
"epoch": 0.09,
|
| 169 |
+
"learning_rate": 4.5e-05,
|
| 170 |
+
"loss": 1.2561,
|
| 171 |
"step": 270
|
| 172 |
},
|
| 173 |
{
|
| 174 |
+
"epoch": 0.09,
|
| 175 |
+
"learning_rate": 4.666666666666667e-05,
|
| 176 |
+
"loss": 2.3613,
|
| 177 |
"step": 280
|
| 178 |
},
|
| 179 |
{
|
| 180 |
+
"epoch": 0.1,
|
| 181 |
+
"learning_rate": 4.8333333333333334e-05,
|
| 182 |
+
"loss": 4.2268,
|
| 183 |
"step": 290
|
| 184 |
},
|
| 185 |
{
|
| 186 |
+
"epoch": 0.1,
|
| 187 |
+
"learning_rate": 5e-05,
|
| 188 |
+
"loss": 1.4112,
|
| 189 |
"step": 300
|
| 190 |
},
|
| 191 |
{
|
| 192 |
+
"epoch": 0.1,
|
| 193 |
+
"eval_accuracy": 0.11409395973154363,
|
| 194 |
+
"eval_loss": 4.316457748413086,
|
| 195 |
+
"eval_runtime": 11.8355,
|
| 196 |
+
"eval_samples_per_second": 12.589,
|
| 197 |
+
"eval_steps_per_second": 12.589,
|
| 198 |
"step": 300
|
| 199 |
},
|
| 200 |
{
|
| 201 |
+
"epoch": 1.0,
|
| 202 |
+
"learning_rate": 4.981481481481482e-05,
|
| 203 |
+
"loss": 3.9062,
|
| 204 |
"step": 310
|
| 205 |
},
|
| 206 |
{
|
| 207 |
+
"epoch": 1.01,
|
| 208 |
+
"learning_rate": 4.962962962962963e-05,
|
| 209 |
+
"loss": 1.7452,
|
| 210 |
"step": 320
|
| 211 |
},
|
| 212 |
{
|
| 213 |
+
"epoch": 1.01,
|
| 214 |
+
"learning_rate": 4.9444444444444446e-05,
|
| 215 |
+
"loss": 2.8437,
|
| 216 |
"step": 330
|
| 217 |
},
|
| 218 |
{
|
| 219 |
+
"epoch": 1.01,
|
| 220 |
+
"learning_rate": 4.925925925925926e-05,
|
| 221 |
+
"loss": 2.381,
|
| 222 |
"step": 340
|
| 223 |
},
|
| 224 |
{
|
| 225 |
+
"epoch": 1.02,
|
| 226 |
+
"learning_rate": 4.9074074074074075e-05,
|
| 227 |
+
"loss": 2.7739,
|
| 228 |
"step": 350
|
| 229 |
},
|
| 230 |
{
|
| 231 |
+
"epoch": 1.02,
|
| 232 |
+
"learning_rate": 4.888888888888889e-05,
|
| 233 |
+
"loss": 1.9838,
|
| 234 |
"step": 360
|
| 235 |
},
|
| 236 |
{
|
| 237 |
+
"epoch": 1.02,
|
| 238 |
+
"learning_rate": 4.8703703703703704e-05,
|
| 239 |
+
"loss": 2.6442,
|
| 240 |
"step": 370
|
| 241 |
},
|
| 242 |
{
|
| 243 |
+
"epoch": 1.03,
|
| 244 |
+
"learning_rate": 4.851851851851852e-05,
|
| 245 |
+
"loss": 2.4327,
|
| 246 |
"step": 380
|
| 247 |
},
|
| 248 |
{
|
| 249 |
+
"epoch": 1.03,
|
| 250 |
+
"learning_rate": 4.8333333333333334e-05,
|
| 251 |
+
"loss": 1.7824,
|
| 252 |
"step": 390
|
| 253 |
},
|
| 254 |
{
|
| 255 |
+
"epoch": 1.03,
|
| 256 |
+
"learning_rate": 4.814814814814815e-05,
|
| 257 |
+
"loss": 2.5717,
|
| 258 |
"step": 400
|
| 259 |
},
|
| 260 |
{
|
| 261 |
+
"epoch": 1.04,
|
| 262 |
+
"learning_rate": 4.796296296296296e-05,
|
| 263 |
+
"loss": 2.0453,
|
| 264 |
"step": 410
|
| 265 |
},
|
| 266 |
{
|
| 267 |
+
"epoch": 1.04,
|
| 268 |
+
"learning_rate": 4.7777777777777784e-05,
|
| 269 |
+
"loss": 0.9527,
|
| 270 |
"step": 420
|
| 271 |
},
|
| 272 |
{
|
| 273 |
+
"epoch": 1.04,
|
| 274 |
+
"learning_rate": 4.759259259259259e-05,
|
| 275 |
+
"loss": 1.9192,
|
| 276 |
"step": 430
|
| 277 |
},
|
| 278 |
{
|
| 279 |
+
"epoch": 1.05,
|
| 280 |
+
"learning_rate": 4.740740740740741e-05,
|
| 281 |
+
"loss": 1.8401,
|
| 282 |
"step": 440
|
| 283 |
},
|
| 284 |
{
|
| 285 |
+
"epoch": 1.05,
|
| 286 |
+
"learning_rate": 4.722222222222222e-05,
|
| 287 |
+
"loss": 1.7712,
|
| 288 |
"step": 450
|
| 289 |
},
|
| 290 |
{
|
| 291 |
+
"epoch": 1.05,
|
| 292 |
+
"learning_rate": 4.703703703703704e-05,
|
| 293 |
+
"loss": 1.7476,
|
| 294 |
"step": 460
|
| 295 |
},
|
| 296 |
{
|
| 297 |
+
"epoch": 1.06,
|
| 298 |
+
"learning_rate": 4.685185185185185e-05,
|
| 299 |
+
"loss": 3.8317,
|
| 300 |
"step": 470
|
| 301 |
},
|
| 302 |
{
|
| 303 |
+
"epoch": 1.06,
|
| 304 |
+
"learning_rate": 4.666666666666667e-05,
|
| 305 |
+
"loss": 1.8526,
|
| 306 |
"step": 480
|
| 307 |
},
|
| 308 |
{
|
| 309 |
+
"epoch": 1.06,
|
| 310 |
+
"learning_rate": 4.648148148148148e-05,
|
| 311 |
+
"loss": 1.9277,
|
| 312 |
"step": 490
|
| 313 |
},
|
| 314 |
{
|
| 315 |
+
"epoch": 1.07,
|
| 316 |
+
"learning_rate": 4.62962962962963e-05,
|
| 317 |
+
"loss": 1.5954,
|
| 318 |
"step": 500
|
| 319 |
},
|
| 320 |
{
|
| 321 |
+
"epoch": 1.07,
|
| 322 |
+
"learning_rate": 4.6111111111111115e-05,
|
| 323 |
+
"loss": 1.4111,
|
| 324 |
"step": 510
|
| 325 |
},
|
| 326 |
{
|
| 327 |
+
"epoch": 1.07,
|
| 328 |
+
"learning_rate": 4.592592592592593e-05,
|
| 329 |
+
"loss": 2.5728,
|
| 330 |
"step": 520
|
| 331 |
},
|
| 332 |
{
|
| 333 |
+
"epoch": 1.08,
|
| 334 |
+
"learning_rate": 4.5740740740740745e-05,
|
| 335 |
+
"loss": 2.1665,
|
| 336 |
"step": 530
|
| 337 |
},
|
| 338 |
{
|
| 339 |
+
"epoch": 1.08,
|
| 340 |
+
"learning_rate": 4.555555555555556e-05,
|
| 341 |
+
"loss": 1.7534,
|
| 342 |
"step": 540
|
| 343 |
},
|
| 344 |
{
|
| 345 |
+
"epoch": 1.08,
|
| 346 |
+
"learning_rate": 4.5370370370370374e-05,
|
| 347 |
+
"loss": 2.4386,
|
| 348 |
"step": 550
|
| 349 |
},
|
| 350 |
{
|
| 351 |
+
"epoch": 1.09,
|
| 352 |
+
"learning_rate": 4.518518518518519e-05,
|
| 353 |
+
"loss": 1.9791,
|
| 354 |
"step": 560
|
| 355 |
},
|
| 356 |
{
|
| 357 |
+
"epoch": 1.09,
|
| 358 |
+
"learning_rate": 4.5e-05,
|
| 359 |
+
"loss": 2.5686,
|
| 360 |
"step": 570
|
| 361 |
},
|
| 362 |
{
|
| 363 |
+
"epoch": 1.09,
|
| 364 |
+
"learning_rate": 4.481481481481482e-05,
|
| 365 |
+
"loss": 1.7416,
|
| 366 |
"step": 580
|
| 367 |
},
|
| 368 |
{
|
| 369 |
+
"epoch": 1.1,
|
| 370 |
+
"learning_rate": 4.462962962962963e-05,
|
| 371 |
+
"loss": 2.3083,
|
| 372 |
"step": 590
|
| 373 |
},
|
| 374 |
{
|
| 375 |
+
"epoch": 1.1,
|
| 376 |
+
"learning_rate": 4.4444444444444447e-05,
|
| 377 |
+
"loss": 2.1452,
|
| 378 |
"step": 600
|
| 379 |
},
|
| 380 |
{
|
| 381 |
+
"epoch": 1.1,
|
| 382 |
+
"eval_accuracy": 0.37583892617449666,
|
| 383 |
+
"eval_loss": 2.0348801612854004,
|
| 384 |
+
"eval_runtime": 11.9169,
|
| 385 |
+
"eval_samples_per_second": 12.503,
|
| 386 |
+
"eval_steps_per_second": 12.503,
|
| 387 |
"step": 600
|
| 388 |
},
|
| 389 |
{
|
| 390 |
+
"epoch": 2.0,
|
| 391 |
+
"learning_rate": 4.425925925925926e-05,
|
| 392 |
+
"loss": 2.1612,
|
| 393 |
"step": 610
|
| 394 |
},
|
| 395 |
{
|
| 396 |
+
"epoch": 2.01,
|
| 397 |
+
"learning_rate": 4.4074074074074076e-05,
|
| 398 |
+
"loss": 2.1526,
|
| 399 |
"step": 620
|
| 400 |
},
|
| 401 |
{
|
| 402 |
+
"epoch": 2.01,
|
| 403 |
+
"learning_rate": 4.388888888888889e-05,
|
| 404 |
+
"loss": 2.0271,
|
| 405 |
"step": 630
|
| 406 |
},
|
| 407 |
{
|
| 408 |
+
"epoch": 2.01,
|
| 409 |
+
"learning_rate": 4.3703703703703705e-05,
|
| 410 |
+
"loss": 0.9482,
|
| 411 |
"step": 640
|
| 412 |
},
|
| 413 |
{
|
| 414 |
+
"epoch": 2.02,
|
| 415 |
+
"learning_rate": 4.351851851851852e-05,
|
| 416 |
+
"loss": 2.8068,
|
| 417 |
"step": 650
|
| 418 |
},
|
| 419 |
{
|
| 420 |
+
"epoch": 2.02,
|
| 421 |
+
"learning_rate": 4.3333333333333334e-05,
|
| 422 |
+
"loss": 1.6808,
|
| 423 |
"step": 660
|
| 424 |
},
|
| 425 |
{
|
| 426 |
+
"epoch": 2.02,
|
| 427 |
+
"learning_rate": 4.314814814814815e-05,
|
| 428 |
+
"loss": 3.8616,
|
| 429 |
"step": 670
|
| 430 |
},
|
| 431 |
{
|
| 432 |
+
"epoch": 2.03,
|
| 433 |
+
"learning_rate": 4.296296296296296e-05,
|
| 434 |
+
"loss": 0.9547,
|
| 435 |
"step": 680
|
| 436 |
},
|
| 437 |
{
|
| 438 |
+
"epoch": 2.03,
|
| 439 |
+
"learning_rate": 4.277777777777778e-05,
|
| 440 |
+
"loss": 1.786,
|
| 441 |
"step": 690
|
| 442 |
},
|
| 443 |
{
|
| 444 |
+
"epoch": 2.03,
|
| 445 |
+
"learning_rate": 4.259259259259259e-05,
|
| 446 |
+
"loss": 2.885,
|
| 447 |
"step": 700
|
| 448 |
},
|
| 449 |
{
|
| 450 |
+
"epoch": 2.04,
|
| 451 |
+
"learning_rate": 4.240740740740741e-05,
|
| 452 |
+
"loss": 1.6305,
|
| 453 |
"step": 710
|
| 454 |
},
|
| 455 |
{
|
| 456 |
+
"epoch": 2.04,
|
| 457 |
+
"learning_rate": 4.222222222222222e-05,
|
| 458 |
+
"loss": 2.182,
|
| 459 |
"step": 720
|
| 460 |
},
|
| 461 |
{
|
| 462 |
+
"epoch": 2.04,
|
| 463 |
+
"learning_rate": 4.203703703703704e-05,
|
| 464 |
+
"loss": 2.5439,
|
| 465 |
"step": 730
|
| 466 |
},
|
| 467 |
{
|
| 468 |
+
"epoch": 2.05,
|
| 469 |
+
"learning_rate": 4.185185185185185e-05,
|
| 470 |
+
"loss": 1.4649,
|
| 471 |
"step": 740
|
| 472 |
},
|
| 473 |
{
|
| 474 |
+
"epoch": 2.05,
|
| 475 |
+
"learning_rate": 4.166666666666667e-05,
|
| 476 |
+
"loss": 1.6346,
|
| 477 |
"step": 750
|
| 478 |
},
|
| 479 |
{
|
| 480 |
+
"epoch": 2.05,
|
| 481 |
+
"learning_rate": 4.148148148148148e-05,
|
| 482 |
+
"loss": 2.7063,
|
| 483 |
"step": 760
|
| 484 |
},
|
| 485 |
{
|
| 486 |
+
"epoch": 2.06,
|
| 487 |
+
"learning_rate": 4.12962962962963e-05,
|
| 488 |
+
"loss": 1.6567,
|
| 489 |
"step": 770
|
| 490 |
},
|
| 491 |
{
|
| 492 |
+
"epoch": 2.06,
|
| 493 |
+
"learning_rate": 4.111111111111111e-05,
|
| 494 |
+
"loss": 2.4469,
|
| 495 |
"step": 780
|
| 496 |
},
|
| 497 |
{
|
| 498 |
+
"epoch": 2.06,
|
| 499 |
+
"learning_rate": 4.092592592592593e-05,
|
| 500 |
+
"loss": 0.9031,
|
| 501 |
"step": 790
|
| 502 |
},
|
| 503 |
{
|
| 504 |
+
"epoch": 2.07,
|
| 505 |
+
"learning_rate": 4.074074074074074e-05,
|
| 506 |
+
"loss": 1.6897,
|
| 507 |
"step": 800
|
| 508 |
},
|
| 509 |
{
|
| 510 |
+
"epoch": 2.07,
|
| 511 |
+
"learning_rate": 4.055555555555556e-05,
|
| 512 |
+
"loss": 2.0307,
|
| 513 |
"step": 810
|
| 514 |
},
|
| 515 |
{
|
| 516 |
+
"epoch": 2.07,
|
| 517 |
+
"learning_rate": 4.0370370370370374e-05,
|
| 518 |
+
"loss": 2.1667,
|
| 519 |
"step": 820
|
| 520 |
},
|
| 521 |
{
|
| 522 |
+
"epoch": 2.08,
|
| 523 |
+
"learning_rate": 4.018518518518519e-05,
|
| 524 |
+
"loss": 0.8811,
|
| 525 |
"step": 830
|
| 526 |
},
|
| 527 |
{
|
| 528 |
+
"epoch": 2.08,
|
| 529 |
+
"learning_rate": 4e-05,
|
| 530 |
+
"loss": 2.7105,
|
| 531 |
"step": 840
|
| 532 |
},
|
| 533 |
{
|
| 534 |
+
"epoch": 2.08,
|
| 535 |
+
"learning_rate": 3.981481481481482e-05,
|
| 536 |
+
"loss": 2.8756,
|
| 537 |
"step": 850
|
| 538 |
},
|
| 539 |
{
|
| 540 |
+
"epoch": 2.09,
|
| 541 |
+
"learning_rate": 3.962962962962963e-05,
|
| 542 |
+
"loss": 1.3152,
|
| 543 |
"step": 860
|
| 544 |
},
|
| 545 |
{
|
| 546 |
+
"epoch": 2.09,
|
| 547 |
+
"learning_rate": 3.944444444444445e-05,
|
| 548 |
+
"loss": 1.4837,
|
| 549 |
"step": 870
|
| 550 |
},
|
| 551 |
{
|
| 552 |
+
"epoch": 2.09,
|
| 553 |
+
"learning_rate": 3.925925925925926e-05,
|
| 554 |
+
"loss": 2.2555,
|
| 555 |
"step": 880
|
| 556 |
},
|
| 557 |
{
|
| 558 |
+
"epoch": 2.1,
|
| 559 |
+
"learning_rate": 3.9074074074074076e-05,
|
| 560 |
+
"loss": 1.9877,
|
| 561 |
"step": 890
|
| 562 |
},
|
| 563 |
{
|
| 564 |
+
"epoch": 2.1,
|
| 565 |
+
"learning_rate": 3.888888888888889e-05,
|
| 566 |
+
"loss": 2.156,
|
| 567 |
"step": 900
|
| 568 |
},
|
| 569 |
{
|
| 570 |
+
"epoch": 2.1,
|
| 571 |
+
"eval_accuracy": 0.2751677852348993,
|
| 572 |
+
"eval_loss": 1.9145618677139282,
|
| 573 |
+
"eval_runtime": 11.9012,
|
| 574 |
+
"eval_samples_per_second": 12.52,
|
| 575 |
+
"eval_steps_per_second": 12.52,
|
| 576 |
"step": 900
|
| 577 |
},
|
| 578 |
{
|
| 579 |
+
"epoch": 3.0,
|
| 580 |
+
"learning_rate": 3.8703703703703705e-05,
|
| 581 |
+
"loss": 1.5921,
|
| 582 |
"step": 910
|
| 583 |
},
|
| 584 |
{
|
| 585 |
+
"epoch": 3.01,
|
| 586 |
+
"learning_rate": 3.851851851851852e-05,
|
| 587 |
+
"loss": 1.7994,
|
| 588 |
"step": 920
|
| 589 |
},
|
| 590 |
{
|
| 591 |
+
"epoch": 3.01,
|
| 592 |
+
"learning_rate": 3.8333333333333334e-05,
|
| 593 |
+
"loss": 2.6093,
|
| 594 |
"step": 930
|
| 595 |
},
|
| 596 |
{
|
| 597 |
+
"epoch": 3.01,
|
| 598 |
+
"learning_rate": 3.814814814814815e-05,
|
| 599 |
+
"loss": 1.5036,
|
| 600 |
"step": 940
|
| 601 |
},
|
| 602 |
{
|
| 603 |
+
"epoch": 3.02,
|
| 604 |
+
"learning_rate": 3.7962962962962964e-05,
|
| 605 |
+
"loss": 1.6625,
|
| 606 |
"step": 950
|
| 607 |
},
|
| 608 |
{
|
| 609 |
+
"epoch": 3.02,
|
| 610 |
+
"learning_rate": 3.777777777777778e-05,
|
| 611 |
+
"loss": 1.7184,
|
| 612 |
"step": 960
|
| 613 |
},
|
| 614 |
{
|
| 615 |
+
"epoch": 3.02,
|
| 616 |
+
"learning_rate": 3.759259259259259e-05,
|
| 617 |
+
"loss": 1.3917,
|
| 618 |
"step": 970
|
| 619 |
},
|
| 620 |
{
|
| 621 |
+
"epoch": 3.03,
|
| 622 |
+
"learning_rate": 3.740740740740741e-05,
|
| 623 |
+
"loss": 1.2039,
|
| 624 |
"step": 980
|
| 625 |
},
|
| 626 |
{
|
| 627 |
+
"epoch": 3.03,
|
| 628 |
+
"learning_rate": 3.722222222222222e-05,
|
| 629 |
+
"loss": 0.7643,
|
| 630 |
"step": 990
|
| 631 |
},
|
| 632 |
{
|
| 633 |
+
"epoch": 3.03,
|
| 634 |
+
"learning_rate": 3.7037037037037037e-05,
|
| 635 |
+
"loss": 2.1679,
|
| 636 |
"step": 1000
|
| 637 |
},
|
| 638 |
{
|
| 639 |
+
"epoch": 3.04,
|
| 640 |
+
"learning_rate": 3.685185185185185e-05,
|
| 641 |
+
"loss": 1.7749,
|
| 642 |
"step": 1010
|
| 643 |
},
|
| 644 |
{
|
| 645 |
+
"epoch": 3.04,
|
| 646 |
+
"learning_rate": 3.6666666666666666e-05,
|
| 647 |
+
"loss": 2.1065,
|
| 648 |
"step": 1020
|
| 649 |
},
|
| 650 |
{
|
| 651 |
+
"epoch": 3.04,
|
| 652 |
+
"learning_rate": 3.648148148148148e-05,
|
| 653 |
+
"loss": 1.6983,
|
| 654 |
"step": 1030
|
| 655 |
},
|
| 656 |
{
|
| 657 |
+
"epoch": 3.05,
|
| 658 |
+
"learning_rate": 3.62962962962963e-05,
|
| 659 |
+
"loss": 1.1671,
|
| 660 |
"step": 1040
|
| 661 |
},
|
| 662 |
{
|
| 663 |
+
"epoch": 3.05,
|
| 664 |
+
"learning_rate": 3.611111111111111e-05,
|
| 665 |
+
"loss": 3.9637,
|
| 666 |
"step": 1050
|
| 667 |
},
|
| 668 |
{
|
| 669 |
+
"epoch": 3.05,
|
| 670 |
+
"learning_rate": 3.592592592592593e-05,
|
| 671 |
+
"loss": 1.8483,
|
| 672 |
"step": 1060
|
| 673 |
},
|
| 674 |
{
|
| 675 |
+
"epoch": 3.06,
|
| 676 |
+
"learning_rate": 3.574074074074074e-05,
|
| 677 |
+
"loss": 1.6249,
|
| 678 |
"step": 1070
|
| 679 |
},
|
| 680 |
{
|
| 681 |
+
"epoch": 3.06,
|
| 682 |
+
"learning_rate": 3.555555555555556e-05,
|
| 683 |
+
"loss": 4.3664,
|
| 684 |
"step": 1080
|
| 685 |
},
|
| 686 |
{
|
| 687 |
+
"epoch": 3.06,
|
| 688 |
+
"learning_rate": 3.537037037037037e-05,
|
| 689 |
+
"loss": 1.3076,
|
| 690 |
"step": 1090
|
| 691 |
},
|
| 692 |
{
|
| 693 |
+
"epoch": 3.07,
|
| 694 |
+
"learning_rate": 3.518518518518519e-05,
|
| 695 |
+
"loss": 1.6146,
|
| 696 |
"step": 1100
|
| 697 |
},
|
| 698 |
{
|
| 699 |
+
"epoch": 3.07,
|
| 700 |
+
"learning_rate": 3.5e-05,
|
| 701 |
+
"loss": 2.7497,
|
| 702 |
"step": 1110
|
| 703 |
},
|
| 704 |
{
|
| 705 |
+
"epoch": 3.07,
|
| 706 |
+
"learning_rate": 3.481481481481482e-05,
|
| 707 |
+
"loss": 1.2038,
|
| 708 |
"step": 1120
|
| 709 |
},
|
| 710 |
{
|
| 711 |
+
"epoch": 3.08,
|
| 712 |
+
"learning_rate": 3.4629629629629626e-05,
|
| 713 |
+
"loss": 1.6241,
|
| 714 |
"step": 1130
|
| 715 |
},
|
| 716 |
{
|
| 717 |
+
"epoch": 3.08,
|
| 718 |
+
"learning_rate": 3.444444444444445e-05,
|
| 719 |
+
"loss": 1.3569,
|
| 720 |
"step": 1140
|
| 721 |
},
|
| 722 |
{
|
| 723 |
+
"epoch": 3.08,
|
| 724 |
+
"learning_rate": 3.425925925925926e-05,
|
| 725 |
+
"loss": 1.6633,
|
| 726 |
"step": 1150
|
| 727 |
},
|
| 728 |
{
|
| 729 |
+
"epoch": 3.09,
|
| 730 |
+
"learning_rate": 3.4074074074074077e-05,
|
| 731 |
+
"loss": 1.2731,
|
| 732 |
"step": 1160
|
| 733 |
},
|
| 734 |
{
|
| 735 |
+
"epoch": 3.09,
|
| 736 |
+
"learning_rate": 3.388888888888889e-05,
|
| 737 |
+
"loss": 0.6022,
|
| 738 |
"step": 1170
|
| 739 |
},
|
| 740 |
{
|
| 741 |
+
"epoch": 3.09,
|
| 742 |
+
"learning_rate": 3.3703703703703706e-05,
|
| 743 |
+
"loss": 0.8256,
|
| 744 |
"step": 1180
|
| 745 |
},
|
| 746 |
{
|
| 747 |
+
"epoch": 3.1,
|
| 748 |
+
"learning_rate": 3.351851851851852e-05,
|
| 749 |
+
"loss": 1.4921,
|
| 750 |
"step": 1190
|
| 751 |
},
|
| 752 |
{
|
| 753 |
+
"epoch": 3.1,
|
| 754 |
+
"learning_rate": 3.3333333333333335e-05,
|
| 755 |
+
"loss": 0.7848,
|
| 756 |
"step": 1200
|
| 757 |
},
|
| 758 |
{
|
| 759 |
+
"epoch": 3.1,
|
| 760 |
+
"eval_accuracy": 0.2550335570469799,
|
| 761 |
+
"eval_loss": 2.933370590209961,
|
| 762 |
+
"eval_runtime": 11.7529,
|
| 763 |
+
"eval_samples_per_second": 12.678,
|
| 764 |
+
"eval_steps_per_second": 12.678,
|
| 765 |
"step": 1200
|
| 766 |
},
|
| 767 |
{
|
| 768 |
+
"epoch": 4.0,
|
| 769 |
+
"learning_rate": 3.314814814814815e-05,
|
| 770 |
+
"loss": 2.8528,
|
| 771 |
+
"step": 1210
|
|
|
|
|
|
|
|
|
|
| 772 |
},
|
| 773 |
{
|
| 774 |
+
"epoch": 4.01,
|
| 775 |
+
"learning_rate": 3.2962962962962964e-05,
|
| 776 |
+
"loss": 2.5306,
|
| 777 |
+
"step": 1220
|
|
|
|
|
|
|
|
|
|
| 778 |
},
|
| 779 |
{
|
| 780 |
+
"epoch": 4.01,
|
| 781 |
+
"learning_rate": 3.277777777777778e-05,
|
| 782 |
+
"loss": 1.8899,
|
| 783 |
+
"step": 1230
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 4.01,
|
| 787 |
+
"learning_rate": 3.25925925925926e-05,
|
| 788 |
+
"loss": 0.9383,
|
| 789 |
+
"step": 1240
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 4.02,
|
| 793 |
+
"learning_rate": 3.240740740740741e-05,
|
| 794 |
+
"loss": 1.5755,
|
| 795 |
+
"step": 1250
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 4.02,
|
| 799 |
+
"learning_rate": 3.222222222222223e-05,
|
| 800 |
+
"loss": 1.2614,
|
| 801 |
+
"step": 1260
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 4.02,
|
| 805 |
+
"learning_rate": 3.203703703703704e-05,
|
| 806 |
+
"loss": 2.0738,
|
| 807 |
+
"step": 1270
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 4.03,
|
| 811 |
+
"learning_rate": 3.185185185185185e-05,
|
| 812 |
+
"loss": 1.7824,
|
| 813 |
+
"step": 1280
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 4.03,
|
| 817 |
+
"learning_rate": 3.1666666666666666e-05,
|
| 818 |
+
"loss": 0.779,
|
| 819 |
+
"step": 1290
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 4.03,
|
| 823 |
+
"learning_rate": 3.148148148148148e-05,
|
| 824 |
+
"loss": 4.432,
|
| 825 |
+
"step": 1300
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 4.04,
|
| 829 |
+
"learning_rate": 3.1296296296296295e-05,
|
| 830 |
+
"loss": 1.934,
|
| 831 |
+
"step": 1310
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 4.04,
|
| 835 |
+
"learning_rate": 3.111111111111111e-05,
|
| 836 |
+
"loss": 2.0968,
|
| 837 |
+
"step": 1320
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 4.04,
|
| 841 |
+
"learning_rate": 3.0925925925925924e-05,
|
| 842 |
+
"loss": 1.7849,
|
| 843 |
+
"step": 1330
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 4.05,
|
| 847 |
+
"learning_rate": 3.074074074074074e-05,
|
| 848 |
+
"loss": 1.3378,
|
| 849 |
+
"step": 1340
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 4.05,
|
| 853 |
+
"learning_rate": 3.055555555555556e-05,
|
| 854 |
+
"loss": 0.8933,
|
| 855 |
+
"step": 1350
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 4.05,
|
| 859 |
+
"learning_rate": 3.037037037037037e-05,
|
| 860 |
+
"loss": 0.8768,
|
| 861 |
+
"step": 1360
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 4.06,
|
| 865 |
+
"learning_rate": 3.018518518518519e-05,
|
| 866 |
+
"loss": 1.1746,
|
| 867 |
+
"step": 1370
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 4.06,
|
| 871 |
+
"learning_rate": 3e-05,
|
| 872 |
+
"loss": 2.3593,
|
| 873 |
+
"step": 1380
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 4.06,
|
| 877 |
+
"learning_rate": 2.981481481481482e-05,
|
| 878 |
+
"loss": 2.9821,
|
| 879 |
+
"step": 1390
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 4.07,
|
| 883 |
+
"learning_rate": 2.962962962962963e-05,
|
| 884 |
+
"loss": 2.0901,
|
| 885 |
+
"step": 1400
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 4.07,
|
| 889 |
+
"learning_rate": 2.9444444444444448e-05,
|
| 890 |
+
"loss": 1.8877,
|
| 891 |
+
"step": 1410
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 4.07,
|
| 895 |
+
"learning_rate": 2.925925925925926e-05,
|
| 896 |
+
"loss": 1.3303,
|
| 897 |
+
"step": 1420
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 4.08,
|
| 901 |
+
"learning_rate": 2.9074074074074077e-05,
|
| 902 |
+
"loss": 1.0231,
|
| 903 |
+
"step": 1430
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 4.08,
|
| 907 |
+
"learning_rate": 2.8888888888888888e-05,
|
| 908 |
+
"loss": 1.271,
|
| 909 |
+
"step": 1440
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 4.08,
|
| 913 |
+
"learning_rate": 2.8703703703703706e-05,
|
| 914 |
+
"loss": 1.4303,
|
| 915 |
+
"step": 1450
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 4.09,
|
| 919 |
+
"learning_rate": 2.851851851851852e-05,
|
| 920 |
+
"loss": 1.4579,
|
| 921 |
+
"step": 1460
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 4.09,
|
| 925 |
+
"learning_rate": 2.8333333333333335e-05,
|
| 926 |
+
"loss": 1.1311,
|
| 927 |
+
"step": 1470
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 4.09,
|
| 931 |
+
"learning_rate": 2.814814814814815e-05,
|
| 932 |
+
"loss": 1.2384,
|
| 933 |
+
"step": 1480
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 4.1,
|
| 937 |
+
"learning_rate": 2.7962962962962965e-05,
|
| 938 |
+
"loss": 1.3813,
|
| 939 |
+
"step": 1490
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 4.1,
|
| 943 |
+
"learning_rate": 2.777777777777778e-05,
|
| 944 |
+
"loss": 3.2405,
|
| 945 |
+
"step": 1500
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 4.1,
|
| 949 |
+
"eval_accuracy": 0.35570469798657717,
|
| 950 |
+
"eval_loss": 2.689954996109009,
|
| 951 |
+
"eval_runtime": 11.9635,
|
| 952 |
+
"eval_samples_per_second": 12.455,
|
| 953 |
+
"eval_steps_per_second": 12.455,
|
| 954 |
+
"step": 1500
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 5.0,
|
| 958 |
+
"learning_rate": 2.7592592592592594e-05,
|
| 959 |
+
"loss": 1.1833,
|
| 960 |
+
"step": 1510
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 5.01,
|
| 964 |
+
"learning_rate": 2.7407407407407408e-05,
|
| 965 |
+
"loss": 0.7545,
|
| 966 |
+
"step": 1520
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 5.01,
|
| 970 |
+
"learning_rate": 2.7222222222222223e-05,
|
| 971 |
+
"loss": 0.6151,
|
| 972 |
+
"step": 1530
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 5.01,
|
| 976 |
+
"learning_rate": 2.7037037037037037e-05,
|
| 977 |
+
"loss": 1.5773,
|
| 978 |
+
"step": 1540
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 5.02,
|
| 982 |
+
"learning_rate": 2.6851851851851855e-05,
|
| 983 |
+
"loss": 0.6904,
|
| 984 |
+
"step": 1550
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 5.02,
|
| 988 |
+
"learning_rate": 2.6666666666666667e-05,
|
| 989 |
+
"loss": 1.0375,
|
| 990 |
+
"step": 1560
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 5.02,
|
| 994 |
+
"learning_rate": 2.6481481481481485e-05,
|
| 995 |
+
"loss": 1.0612,
|
| 996 |
+
"step": 1570
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 5.03,
|
| 1000 |
+
"learning_rate": 2.6296296296296296e-05,
|
| 1001 |
+
"loss": 1.7255,
|
| 1002 |
+
"step": 1580
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 5.03,
|
| 1006 |
+
"learning_rate": 2.6111111111111114e-05,
|
| 1007 |
+
"loss": 1.8048,
|
| 1008 |
+
"step": 1590
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 5.03,
|
| 1012 |
+
"learning_rate": 2.5925925925925925e-05,
|
| 1013 |
+
"loss": 0.8129,
|
| 1014 |
+
"step": 1600
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 5.04,
|
| 1018 |
+
"learning_rate": 2.5740740740740743e-05,
|
| 1019 |
+
"loss": 2.9754,
|
| 1020 |
+
"step": 1610
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 5.04,
|
| 1024 |
+
"learning_rate": 2.5555555555555554e-05,
|
| 1025 |
+
"loss": 1.661,
|
| 1026 |
+
"step": 1620
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 5.04,
|
| 1030 |
+
"learning_rate": 2.5370370370370372e-05,
|
| 1031 |
+
"loss": 1.0065,
|
| 1032 |
+
"step": 1630
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 5.05,
|
| 1036 |
+
"learning_rate": 2.5185185185185183e-05,
|
| 1037 |
+
"loss": 1.6616,
|
| 1038 |
+
"step": 1640
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 5.05,
|
| 1042 |
+
"learning_rate": 2.5e-05,
|
| 1043 |
+
"loss": 1.8583,
|
| 1044 |
+
"step": 1650
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 5.05,
|
| 1048 |
+
"learning_rate": 2.4814814814814816e-05,
|
| 1049 |
+
"loss": 0.9394,
|
| 1050 |
+
"step": 1660
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"epoch": 5.06,
|
| 1054 |
+
"learning_rate": 2.462962962962963e-05,
|
| 1055 |
+
"loss": 0.7979,
|
| 1056 |
+
"step": 1670
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 5.06,
|
| 1060 |
+
"learning_rate": 2.4444444444444445e-05,
|
| 1061 |
+
"loss": 2.1306,
|
| 1062 |
+
"step": 1680
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 5.06,
|
| 1066 |
+
"learning_rate": 2.425925925925926e-05,
|
| 1067 |
+
"loss": 1.2191,
|
| 1068 |
+
"step": 1690
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 5.07,
|
| 1072 |
+
"learning_rate": 2.4074074074074074e-05,
|
| 1073 |
+
"loss": 0.7189,
|
| 1074 |
+
"step": 1700
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 5.07,
|
| 1078 |
+
"learning_rate": 2.3888888888888892e-05,
|
| 1079 |
+
"loss": 2.9497,
|
| 1080 |
+
"step": 1710
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 5.07,
|
| 1084 |
+
"learning_rate": 2.3703703703703707e-05,
|
| 1085 |
+
"loss": 1.1491,
|
| 1086 |
+
"step": 1720
|
| 1087 |
+
},
|
| 1088 |
+
{
|
| 1089 |
+
"epoch": 5.08,
|
| 1090 |
+
"learning_rate": 2.351851851851852e-05,
|
| 1091 |
+
"loss": 3.3826,
|
| 1092 |
+
"step": 1730
|
| 1093 |
+
},
|
| 1094 |
+
{
|
| 1095 |
+
"epoch": 5.08,
|
| 1096 |
+
"learning_rate": 2.3333333333333336e-05,
|
| 1097 |
+
"loss": 1.846,
|
| 1098 |
+
"step": 1740
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 5.08,
|
| 1102 |
+
"learning_rate": 2.314814814814815e-05,
|
| 1103 |
+
"loss": 1.1924,
|
| 1104 |
+
"step": 1750
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 5.09,
|
| 1108 |
+
"learning_rate": 2.2962962962962965e-05,
|
| 1109 |
+
"loss": 1.1331,
|
| 1110 |
+
"step": 1760
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 5.09,
|
| 1114 |
+
"learning_rate": 2.277777777777778e-05,
|
| 1115 |
+
"loss": 1.4474,
|
| 1116 |
+
"step": 1770
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 5.09,
|
| 1120 |
+
"learning_rate": 2.2592592592592594e-05,
|
| 1121 |
+
"loss": 1.565,
|
| 1122 |
+
"step": 1780
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 5.1,
|
| 1126 |
+
"learning_rate": 2.240740740740741e-05,
|
| 1127 |
+
"loss": 2.0442,
|
| 1128 |
+
"step": 1790
|
| 1129 |
+
},
|
| 1130 |
+
{
|
| 1131 |
+
"epoch": 5.1,
|
| 1132 |
+
"learning_rate": 2.2222222222222223e-05,
|
| 1133 |
+
"loss": 2.6131,
|
| 1134 |
+
"step": 1800
|
| 1135 |
+
},
|
| 1136 |
+
{
|
| 1137 |
+
"epoch": 5.1,
|
| 1138 |
+
"eval_accuracy": 0.3691275167785235,
|
| 1139 |
+
"eval_loss": 2.5973401069641113,
|
| 1140 |
+
"eval_runtime": 11.8379,
|
| 1141 |
+
"eval_samples_per_second": 12.587,
|
| 1142 |
+
"eval_steps_per_second": 12.587,
|
| 1143 |
+
"step": 1800
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 6.0,
|
| 1147 |
+
"learning_rate": 2.2037037037037038e-05,
|
| 1148 |
+
"loss": 1.1066,
|
| 1149 |
+
"step": 1810
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"epoch": 6.01,
|
| 1153 |
+
"learning_rate": 2.1851851851851852e-05,
|
| 1154 |
+
"loss": 1.4162,
|
| 1155 |
+
"step": 1820
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"epoch": 6.01,
|
| 1159 |
+
"learning_rate": 2.1666666666666667e-05,
|
| 1160 |
+
"loss": 1.5759,
|
| 1161 |
+
"step": 1830
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 6.01,
|
| 1165 |
+
"learning_rate": 2.148148148148148e-05,
|
| 1166 |
+
"loss": 2.3782,
|
| 1167 |
+
"step": 1840
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 6.02,
|
| 1171 |
+
"learning_rate": 2.1296296296296296e-05,
|
| 1172 |
+
"loss": 0.8711,
|
| 1173 |
+
"step": 1850
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 6.02,
|
| 1177 |
+
"learning_rate": 2.111111111111111e-05,
|
| 1178 |
+
"loss": 0.997,
|
| 1179 |
+
"step": 1860
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 6.02,
|
| 1183 |
+
"learning_rate": 2.0925925925925925e-05,
|
| 1184 |
+
"loss": 1.8742,
|
| 1185 |
+
"step": 1870
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 6.03,
|
| 1189 |
+
"learning_rate": 2.074074074074074e-05,
|
| 1190 |
+
"loss": 1.3593,
|
| 1191 |
+
"step": 1880
|
| 1192 |
+
},
|
| 1193 |
+
{
|
| 1194 |
+
"epoch": 6.03,
|
| 1195 |
+
"learning_rate": 2.0555555555555555e-05,
|
| 1196 |
+
"loss": 1.1478,
|
| 1197 |
+
"step": 1890
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 6.03,
|
| 1201 |
+
"learning_rate": 2.037037037037037e-05,
|
| 1202 |
+
"loss": 1.1945,
|
| 1203 |
+
"step": 1900
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 6.04,
|
| 1207 |
+
"learning_rate": 2.0185185185185187e-05,
|
| 1208 |
+
"loss": 1.4263,
|
| 1209 |
+
"step": 1910
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 6.04,
|
| 1213 |
+
"learning_rate": 2e-05,
|
| 1214 |
+
"loss": 0.8273,
|
| 1215 |
+
"step": 1920
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 6.04,
|
| 1219 |
+
"learning_rate": 1.9814814814814816e-05,
|
| 1220 |
+
"loss": 1.0158,
|
| 1221 |
+
"step": 1930
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 6.05,
|
| 1225 |
+
"learning_rate": 1.962962962962963e-05,
|
| 1226 |
+
"loss": 0.6081,
|
| 1227 |
+
"step": 1940
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 6.05,
|
| 1231 |
+
"learning_rate": 1.9444444444444445e-05,
|
| 1232 |
+
"loss": 2.2851,
|
| 1233 |
+
"step": 1950
|
| 1234 |
+
},
|
| 1235 |
+
{
|
| 1236 |
+
"epoch": 6.05,
|
| 1237 |
+
"learning_rate": 1.925925925925926e-05,
|
| 1238 |
+
"loss": 1.5421,
|
| 1239 |
+
"step": 1960
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"epoch": 6.06,
|
| 1243 |
+
"learning_rate": 1.9074074074074075e-05,
|
| 1244 |
+
"loss": 1.5585,
|
| 1245 |
+
"step": 1970
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 6.06,
|
| 1249 |
+
"learning_rate": 1.888888888888889e-05,
|
| 1250 |
+
"loss": 1.721,
|
| 1251 |
+
"step": 1980
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 6.06,
|
| 1255 |
+
"learning_rate": 1.8703703703703704e-05,
|
| 1256 |
+
"loss": 0.8218,
|
| 1257 |
+
"step": 1990
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 6.07,
|
| 1261 |
+
"learning_rate": 1.8518518518518518e-05,
|
| 1262 |
+
"loss": 0.8604,
|
| 1263 |
+
"step": 2000
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 6.07,
|
| 1267 |
+
"learning_rate": 1.8333333333333333e-05,
|
| 1268 |
+
"loss": 1.4615,
|
| 1269 |
+
"step": 2010
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 6.07,
|
| 1273 |
+
"learning_rate": 1.814814814814815e-05,
|
| 1274 |
+
"loss": 0.4951,
|
| 1275 |
+
"step": 2020
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"epoch": 6.08,
|
| 1279 |
+
"learning_rate": 1.7962962962962965e-05,
|
| 1280 |
+
"loss": 1.5857,
|
| 1281 |
+
"step": 2030
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"epoch": 6.08,
|
| 1285 |
+
"learning_rate": 1.777777777777778e-05,
|
| 1286 |
+
"loss": 0.1666,
|
| 1287 |
+
"step": 2040
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 6.08,
|
| 1291 |
+
"learning_rate": 1.7592592592592595e-05,
|
| 1292 |
+
"loss": 0.3136,
|
| 1293 |
+
"step": 2050
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 6.09,
|
| 1297 |
+
"learning_rate": 1.740740740740741e-05,
|
| 1298 |
+
"loss": 0.5031,
|
| 1299 |
+
"step": 2060
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 6.09,
|
| 1303 |
+
"learning_rate": 1.7222222222222224e-05,
|
| 1304 |
+
"loss": 1.0772,
|
| 1305 |
+
"step": 2070
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 6.09,
|
| 1309 |
+
"learning_rate": 1.7037037037037038e-05,
|
| 1310 |
+
"loss": 1.581,
|
| 1311 |
+
"step": 2080
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 6.1,
|
| 1315 |
+
"learning_rate": 1.6851851851851853e-05,
|
| 1316 |
+
"loss": 1.8657,
|
| 1317 |
+
"step": 2090
|
| 1318 |
+
},
|
| 1319 |
+
{
|
| 1320 |
+
"epoch": 6.1,
|
| 1321 |
+
"learning_rate": 1.6666666666666667e-05,
|
| 1322 |
+
"loss": 1.5998,
|
| 1323 |
+
"step": 2100
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"epoch": 6.1,
|
| 1327 |
+
"eval_accuracy": 0.42953020134228187,
|
| 1328 |
+
"eval_loss": 3.1545114517211914,
|
| 1329 |
+
"eval_runtime": 11.9443,
|
| 1330 |
+
"eval_samples_per_second": 12.475,
|
| 1331 |
+
"eval_steps_per_second": 12.475,
|
| 1332 |
+
"step": 2100
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 7.0,
|
| 1336 |
+
"learning_rate": 1.6481481481481482e-05,
|
| 1337 |
+
"loss": 1.1743,
|
| 1338 |
+
"step": 2110
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 7.01,
|
| 1342 |
+
"learning_rate": 1.62962962962963e-05,
|
| 1343 |
+
"loss": 0.6246,
|
| 1344 |
+
"step": 2120
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 7.01,
|
| 1348 |
+
"learning_rate": 1.6111111111111115e-05,
|
| 1349 |
+
"loss": 1.1438,
|
| 1350 |
+
"step": 2130
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 7.01,
|
| 1354 |
+
"learning_rate": 1.5925925925925926e-05,
|
| 1355 |
+
"loss": 1.7971,
|
| 1356 |
+
"step": 2140
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 7.02,
|
| 1360 |
+
"learning_rate": 1.574074074074074e-05,
|
| 1361 |
+
"loss": 2.8378,
|
| 1362 |
+
"step": 2150
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 7.02,
|
| 1366 |
+
"learning_rate": 1.5555555555555555e-05,
|
| 1367 |
+
"loss": 0.6635,
|
| 1368 |
+
"step": 2160
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 7.02,
|
| 1372 |
+
"learning_rate": 1.537037037037037e-05,
|
| 1373 |
+
"loss": 0.7129,
|
| 1374 |
+
"step": 2170
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 7.03,
|
| 1378 |
+
"learning_rate": 1.5185185185185186e-05,
|
| 1379 |
+
"loss": 1.2923,
|
| 1380 |
+
"step": 2180
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"epoch": 7.03,
|
| 1384 |
+
"learning_rate": 1.5e-05,
|
| 1385 |
+
"loss": 0.9154,
|
| 1386 |
+
"step": 2190
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 7.03,
|
| 1390 |
+
"learning_rate": 1.4814814814814815e-05,
|
| 1391 |
+
"loss": 1.2448,
|
| 1392 |
+
"step": 2200
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 7.04,
|
| 1396 |
+
"learning_rate": 1.462962962962963e-05,
|
| 1397 |
+
"loss": 1.1172,
|
| 1398 |
+
"step": 2210
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 7.04,
|
| 1402 |
+
"learning_rate": 1.4444444444444444e-05,
|
| 1403 |
+
"loss": 2.2937,
|
| 1404 |
+
"step": 2220
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 7.04,
|
| 1408 |
+
"learning_rate": 1.425925925925926e-05,
|
| 1409 |
+
"loss": 0.7835,
|
| 1410 |
+
"step": 2230
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 7.05,
|
| 1414 |
+
"learning_rate": 1.4074074074074075e-05,
|
| 1415 |
+
"loss": 0.0882,
|
| 1416 |
+
"step": 2240
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 7.05,
|
| 1420 |
+
"learning_rate": 1.388888888888889e-05,
|
| 1421 |
+
"loss": 0.2712,
|
| 1422 |
+
"step": 2250
|
| 1423 |
+
},
|
| 1424 |
+
{
|
| 1425 |
+
"epoch": 7.05,
|
| 1426 |
+
"learning_rate": 1.3703703703703704e-05,
|
| 1427 |
+
"loss": 2.0246,
|
| 1428 |
+
"step": 2260
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 7.06,
|
| 1432 |
+
"learning_rate": 1.3518518518518519e-05,
|
| 1433 |
+
"loss": 0.2939,
|
| 1434 |
+
"step": 2270
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 7.06,
|
| 1438 |
+
"learning_rate": 1.3333333333333333e-05,
|
| 1439 |
+
"loss": 0.2676,
|
| 1440 |
+
"step": 2280
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 7.06,
|
| 1444 |
+
"learning_rate": 1.3148148148148148e-05,
|
| 1445 |
+
"loss": 2.1027,
|
| 1446 |
+
"step": 2290
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 7.07,
|
| 1450 |
+
"learning_rate": 1.2962962962962962e-05,
|
| 1451 |
+
"loss": 1.2775,
|
| 1452 |
+
"step": 2300
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 7.07,
|
| 1456 |
+
"learning_rate": 1.2777777777777777e-05,
|
| 1457 |
+
"loss": 2.663,
|
| 1458 |
+
"step": 2310
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 7.07,
|
| 1462 |
+
"learning_rate": 1.2592592592592592e-05,
|
| 1463 |
+
"loss": 0.8683,
|
| 1464 |
+
"step": 2320
|
| 1465 |
+
},
|
| 1466 |
+
{
|
| 1467 |
+
"epoch": 7.08,
|
| 1468 |
+
"learning_rate": 1.2407407407407408e-05,
|
| 1469 |
+
"loss": 1.361,
|
| 1470 |
+
"step": 2330
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 7.08,
|
| 1474 |
+
"learning_rate": 1.2222222222222222e-05,
|
| 1475 |
+
"loss": 2.4195,
|
| 1476 |
+
"step": 2340
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 7.08,
|
| 1480 |
+
"learning_rate": 1.2037037037037037e-05,
|
| 1481 |
+
"loss": 1.2075,
|
| 1482 |
+
"step": 2350
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 7.09,
|
| 1486 |
+
"learning_rate": 1.1851851851851853e-05,
|
| 1487 |
+
"loss": 2.6057,
|
| 1488 |
+
"step": 2360
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 7.09,
|
| 1492 |
+
"learning_rate": 1.1666666666666668e-05,
|
| 1493 |
+
"loss": 1.0114,
|
| 1494 |
+
"step": 2370
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 7.09,
|
| 1498 |
+
"learning_rate": 1.1481481481481482e-05,
|
| 1499 |
+
"loss": 1.1405,
|
| 1500 |
+
"step": 2380
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 7.1,
|
| 1504 |
+
"learning_rate": 1.1296296296296297e-05,
|
| 1505 |
+
"loss": 0.8811,
|
| 1506 |
+
"step": 2390
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 7.1,
|
| 1510 |
+
"learning_rate": 1.1111111111111112e-05,
|
| 1511 |
+
"loss": 1.1689,
|
| 1512 |
+
"step": 2400
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 7.1,
|
| 1516 |
+
"eval_accuracy": 0.2953020134228188,
|
| 1517 |
+
"eval_loss": 3.3012382984161377,
|
| 1518 |
+
"eval_runtime": 11.9208,
|
| 1519 |
+
"eval_samples_per_second": 12.499,
|
| 1520 |
+
"eval_steps_per_second": 12.499,
|
| 1521 |
+
"step": 2400
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 8.0,
|
| 1525 |
+
"learning_rate": 1.0925925925925926e-05,
|
| 1526 |
+
"loss": 1.3559,
|
| 1527 |
+
"step": 2410
|
| 1528 |
+
},
|
| 1529 |
+
{
|
| 1530 |
+
"epoch": 8.01,
|
| 1531 |
+
"learning_rate": 1.074074074074074e-05,
|
| 1532 |
+
"loss": 0.6597,
|
| 1533 |
+
"step": 2420
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 8.01,
|
| 1537 |
+
"learning_rate": 1.0555555555555555e-05,
|
| 1538 |
+
"loss": 1.0085,
|
| 1539 |
+
"step": 2430
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 8.01,
|
| 1543 |
+
"learning_rate": 1.037037037037037e-05,
|
| 1544 |
+
"loss": 0.5666,
|
| 1545 |
+
"step": 2440
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 8.02,
|
| 1549 |
+
"learning_rate": 1.0185185185185185e-05,
|
| 1550 |
+
"loss": 0.1961,
|
| 1551 |
+
"step": 2450
|
| 1552 |
+
},
|
| 1553 |
+
{
|
| 1554 |
+
"epoch": 8.02,
|
| 1555 |
+
"learning_rate": 1e-05,
|
| 1556 |
+
"loss": 1.4929,
|
| 1557 |
+
"step": 2460
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 8.02,
|
| 1561 |
+
"learning_rate": 9.814814814814815e-06,
|
| 1562 |
+
"loss": 1.2987,
|
| 1563 |
+
"step": 2470
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 8.03,
|
| 1567 |
+
"learning_rate": 9.62962962962963e-06,
|
| 1568 |
+
"loss": 1.113,
|
| 1569 |
+
"step": 2480
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 8.03,
|
| 1573 |
+
"learning_rate": 9.444444444444445e-06,
|
| 1574 |
+
"loss": 2.0659,
|
| 1575 |
+
"step": 2490
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"epoch": 8.03,
|
| 1579 |
+
"learning_rate": 9.259259259259259e-06,
|
| 1580 |
+
"loss": 1.2915,
|
| 1581 |
+
"step": 2500
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"epoch": 8.04,
|
| 1585 |
+
"learning_rate": 9.074074074074075e-06,
|
| 1586 |
+
"loss": 1.2701,
|
| 1587 |
+
"step": 2510
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"epoch": 8.04,
|
| 1591 |
+
"learning_rate": 8.88888888888889e-06,
|
| 1592 |
+
"loss": 1.2432,
|
| 1593 |
+
"step": 2520
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 8.04,
|
| 1597 |
+
"learning_rate": 8.703703703703705e-06,
|
| 1598 |
+
"loss": 0.7502,
|
| 1599 |
+
"step": 2530
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 8.05,
|
| 1603 |
+
"learning_rate": 8.518518518518519e-06,
|
| 1604 |
+
"loss": 0.6665,
|
| 1605 |
+
"step": 2540
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 8.05,
|
| 1609 |
+
"learning_rate": 8.333333333333334e-06,
|
| 1610 |
+
"loss": 1.2312,
|
| 1611 |
+
"step": 2550
|
| 1612 |
+
},
|
| 1613 |
+
{
|
| 1614 |
+
"epoch": 8.05,
|
| 1615 |
+
"learning_rate": 8.14814814814815e-06,
|
| 1616 |
+
"loss": 0.4243,
|
| 1617 |
+
"step": 2560
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 8.06,
|
| 1621 |
+
"learning_rate": 7.962962962962963e-06,
|
| 1622 |
+
"loss": 0.8543,
|
| 1623 |
+
"step": 2570
|
| 1624 |
+
},
|
| 1625 |
+
{
|
| 1626 |
+
"epoch": 8.06,
|
| 1627 |
+
"learning_rate": 7.777777777777777e-06,
|
| 1628 |
+
"loss": 2.307,
|
| 1629 |
+
"step": 2580
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 8.06,
|
| 1633 |
+
"learning_rate": 7.592592592592593e-06,
|
| 1634 |
+
"loss": 1.0244,
|
| 1635 |
+
"step": 2590
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 8.07,
|
| 1639 |
+
"learning_rate": 7.4074074074074075e-06,
|
| 1640 |
+
"loss": 0.9076,
|
| 1641 |
+
"step": 2600
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 8.07,
|
| 1645 |
+
"learning_rate": 7.222222222222222e-06,
|
| 1646 |
+
"loss": 0.6468,
|
| 1647 |
+
"step": 2610
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 8.07,
|
| 1651 |
+
"learning_rate": 7.0370370370370375e-06,
|
| 1652 |
+
"loss": 0.8508,
|
| 1653 |
+
"step": 2620
|
| 1654 |
+
},
|
| 1655 |
+
{
|
| 1656 |
+
"epoch": 8.08,
|
| 1657 |
+
"learning_rate": 6.851851851851852e-06,
|
| 1658 |
+
"loss": 0.118,
|
| 1659 |
+
"step": 2630
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"epoch": 8.08,
|
| 1663 |
+
"learning_rate": 6.666666666666667e-06,
|
| 1664 |
+
"loss": 1.6374,
|
| 1665 |
+
"step": 2640
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 8.08,
|
| 1669 |
+
"learning_rate": 6.481481481481481e-06,
|
| 1670 |
+
"loss": 0.6333,
|
| 1671 |
+
"step": 2650
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 8.09,
|
| 1675 |
+
"learning_rate": 6.296296296296296e-06,
|
| 1676 |
+
"loss": 2.5795,
|
| 1677 |
+
"step": 2660
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"epoch": 8.09,
|
| 1681 |
+
"learning_rate": 6.111111111111111e-06,
|
| 1682 |
+
"loss": 2.5085,
|
| 1683 |
+
"step": 2670
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 8.09,
|
| 1687 |
+
"learning_rate": 5.925925925925927e-06,
|
| 1688 |
+
"loss": 1.2708,
|
| 1689 |
+
"step": 2680
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 8.1,
|
| 1693 |
+
"learning_rate": 5.740740740740741e-06,
|
| 1694 |
+
"loss": 0.1959,
|
| 1695 |
+
"step": 2690
|
| 1696 |
+
},
|
| 1697 |
+
{
|
| 1698 |
+
"epoch": 8.1,
|
| 1699 |
+
"learning_rate": 5.555555555555556e-06,
|
| 1700 |
+
"loss": 0.2398,
|
| 1701 |
+
"step": 2700
|
| 1702 |
+
},
|
| 1703 |
+
{
|
| 1704 |
+
"epoch": 8.1,
|
| 1705 |
+
"eval_accuracy": 0.436241610738255,
|
| 1706 |
+
"eval_loss": 2.723142385482788,
|
| 1707 |
+
"eval_runtime": 11.9115,
|
| 1708 |
+
"eval_samples_per_second": 12.509,
|
| 1709 |
+
"eval_steps_per_second": 12.509,
|
| 1710 |
+
"step": 2700
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 9.0,
|
| 1714 |
+
"learning_rate": 5.37037037037037e-06,
|
| 1715 |
+
"loss": 0.6692,
|
| 1716 |
+
"step": 2710
|
| 1717 |
+
},
|
| 1718 |
+
{
|
| 1719 |
+
"epoch": 9.01,
|
| 1720 |
+
"learning_rate": 5.185185185185185e-06,
|
| 1721 |
+
"loss": 0.4314,
|
| 1722 |
+
"step": 2720
|
| 1723 |
+
},
|
| 1724 |
+
{
|
| 1725 |
+
"epoch": 9.01,
|
| 1726 |
+
"learning_rate": 5e-06,
|
| 1727 |
+
"loss": 1.6368,
|
| 1728 |
+
"step": 2730
|
| 1729 |
+
},
|
| 1730 |
+
{
|
| 1731 |
+
"epoch": 9.01,
|
| 1732 |
+
"learning_rate": 4.814814814814815e-06,
|
| 1733 |
+
"loss": 0.4331,
|
| 1734 |
+
"step": 2740
|
| 1735 |
+
},
|
| 1736 |
+
{
|
| 1737 |
+
"epoch": 9.02,
|
| 1738 |
+
"learning_rate": 4.6296296296296296e-06,
|
| 1739 |
+
"loss": 0.5237,
|
| 1740 |
+
"step": 2750
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"epoch": 9.02,
|
| 1744 |
+
"learning_rate": 4.444444444444445e-06,
|
| 1745 |
+
"loss": 1.5999,
|
| 1746 |
+
"step": 2760
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 9.02,
|
| 1750 |
+
"learning_rate": 4.2592592592592596e-06,
|
| 1751 |
+
"loss": 1.7222,
|
| 1752 |
+
"step": 2770
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 9.03,
|
| 1756 |
+
"learning_rate": 4.074074074074075e-06,
|
| 1757 |
+
"loss": 0.5853,
|
| 1758 |
+
"step": 2780
|
| 1759 |
+
},
|
| 1760 |
+
{
|
| 1761 |
+
"epoch": 9.03,
|
| 1762 |
+
"learning_rate": 3.888888888888889e-06,
|
| 1763 |
+
"loss": 1.5566,
|
| 1764 |
+
"step": 2790
|
| 1765 |
+
},
|
| 1766 |
+
{
|
| 1767 |
+
"epoch": 9.03,
|
| 1768 |
+
"learning_rate": 3.7037037037037037e-06,
|
| 1769 |
+
"loss": 0.5217,
|
| 1770 |
+
"step": 2800
|
| 1771 |
+
},
|
| 1772 |
+
{
|
| 1773 |
+
"epoch": 9.04,
|
| 1774 |
+
"learning_rate": 3.5185185185185187e-06,
|
| 1775 |
+
"loss": 0.373,
|
| 1776 |
+
"step": 2810
|
| 1777 |
+
},
|
| 1778 |
+
{
|
| 1779 |
+
"epoch": 9.04,
|
| 1780 |
+
"learning_rate": 3.3333333333333333e-06,
|
| 1781 |
+
"loss": 0.9443,
|
| 1782 |
+
"step": 2820
|
| 1783 |
+
},
|
| 1784 |
+
{
|
| 1785 |
+
"epoch": 9.04,
|
| 1786 |
+
"learning_rate": 3.148148148148148e-06,
|
| 1787 |
+
"loss": 0.2011,
|
| 1788 |
+
"step": 2830
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 9.05,
|
| 1792 |
+
"learning_rate": 2.9629629629629633e-06,
|
| 1793 |
+
"loss": 0.6479,
|
| 1794 |
+
"step": 2840
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 9.05,
|
| 1798 |
+
"learning_rate": 2.777777777777778e-06,
|
| 1799 |
+
"loss": 1.0128,
|
| 1800 |
+
"step": 2850
|
| 1801 |
+
},
|
| 1802 |
+
{
|
| 1803 |
+
"epoch": 9.05,
|
| 1804 |
+
"learning_rate": 2.5925925925925925e-06,
|
| 1805 |
+
"loss": 1.3453,
|
| 1806 |
+
"step": 2860
|
| 1807 |
+
},
|
| 1808 |
+
{
|
| 1809 |
+
"epoch": 9.06,
|
| 1810 |
+
"learning_rate": 2.4074074074074075e-06,
|
| 1811 |
+
"loss": 1.9262,
|
| 1812 |
+
"step": 2870
|
| 1813 |
+
},
|
| 1814 |
+
{
|
| 1815 |
+
"epoch": 9.06,
|
| 1816 |
+
"learning_rate": 2.2222222222222225e-06,
|
| 1817 |
+
"loss": 0.9785,
|
| 1818 |
+
"step": 2880
|
| 1819 |
+
},
|
| 1820 |
+
{
|
| 1821 |
+
"epoch": 9.06,
|
| 1822 |
+
"learning_rate": 2.0370370370370375e-06,
|
| 1823 |
+
"loss": 0.6525,
|
| 1824 |
+
"step": 2890
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"epoch": 9.07,
|
| 1828 |
+
"learning_rate": 1.8518518518518519e-06,
|
| 1829 |
+
"loss": 0.6206,
|
| 1830 |
+
"step": 2900
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 9.07,
|
| 1834 |
+
"learning_rate": 1.6666666666666667e-06,
|
| 1835 |
+
"loss": 0.7492,
|
| 1836 |
+
"step": 2910
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 9.07,
|
| 1840 |
+
"learning_rate": 1.4814814814814817e-06,
|
| 1841 |
+
"loss": 1.004,
|
| 1842 |
+
"step": 2920
|
| 1843 |
+
},
|
| 1844 |
+
{
|
| 1845 |
+
"epoch": 9.08,
|
| 1846 |
+
"learning_rate": 1.2962962962962962e-06,
|
| 1847 |
+
"loss": 2.0529,
|
| 1848 |
+
"step": 2930
|
| 1849 |
+
},
|
| 1850 |
+
{
|
| 1851 |
+
"epoch": 9.08,
|
| 1852 |
+
"learning_rate": 1.1111111111111112e-06,
|
| 1853 |
+
"loss": 1.3948,
|
| 1854 |
+
"step": 2940
|
| 1855 |
+
},
|
| 1856 |
+
{
|
| 1857 |
+
"epoch": 9.08,
|
| 1858 |
+
"learning_rate": 9.259259259259259e-07,
|
| 1859 |
+
"loss": 1.0218,
|
| 1860 |
+
"step": 2950
|
| 1861 |
+
},
|
| 1862 |
+
{
|
| 1863 |
+
"epoch": 9.09,
|
| 1864 |
+
"learning_rate": 7.407407407407408e-07,
|
| 1865 |
+
"loss": 1.5455,
|
| 1866 |
+
"step": 2960
|
| 1867 |
+
},
|
| 1868 |
+
{
|
| 1869 |
+
"epoch": 9.09,
|
| 1870 |
+
"learning_rate": 5.555555555555556e-07,
|
| 1871 |
+
"loss": 0.1965,
|
| 1872 |
+
"step": 2970
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 9.09,
|
| 1876 |
+
"learning_rate": 3.703703703703704e-07,
|
| 1877 |
+
"loss": 0.3648,
|
| 1878 |
+
"step": 2980
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 9.1,
|
| 1882 |
+
"learning_rate": 1.851851851851852e-07,
|
| 1883 |
+
"loss": 0.4616,
|
| 1884 |
+
"step": 2990
|
| 1885 |
+
},
|
| 1886 |
+
{
|
| 1887 |
+
"epoch": 9.1,
|
| 1888 |
+
"learning_rate": 0.0,
|
| 1889 |
+
"loss": 1.2448,
|
| 1890 |
+
"step": 3000
|
| 1891 |
+
},
|
| 1892 |
+
{
|
| 1893 |
+
"epoch": 9.1,
|
| 1894 |
+
"eval_accuracy": 0.42953020134228187,
|
| 1895 |
+
"eval_loss": 2.694873809814453,
|
| 1896 |
+
"eval_runtime": 11.9402,
|
| 1897 |
+
"eval_samples_per_second": 12.479,
|
| 1898 |
+
"eval_steps_per_second": 12.479,
|
| 1899 |
+
"step": 3000
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 9.1,
|
| 1903 |
+
"step": 3000,
|
| 1904 |
+
"total_flos": 2.635076136075264e+18,
|
| 1905 |
+
"train_loss": 1.5726345704197884,
|
| 1906 |
+
"train_runtime": 705.2236,
|
| 1907 |
+
"train_samples_per_second": 4.254,
|
| 1908 |
+
"train_steps_per_second": 4.254
|
| 1909 |
+
},
|
| 1910 |
+
{
|
| 1911 |
+
"epoch": 9.1,
|
| 1912 |
+
"eval_accuracy": 0.6005747126436781,
|
| 1913 |
+
"eval_loss": 1.64243745803833,
|
| 1914 |
+
"eval_runtime": 28.1079,
|
| 1915 |
+
"eval_samples_per_second": 12.381,
|
| 1916 |
+
"eval_steps_per_second": 12.381,
|
| 1917 |
+
"step": 3000
|
| 1918 |
+
},
|
| 1919 |
+
{
|
| 1920 |
+
"epoch": 9.1,
|
| 1921 |
+
"eval_accuracy": 0.5994236311239193,
|
| 1922 |
+
"eval_loss": 1.647168755531311,
|
| 1923 |
+
"eval_runtime": 27.7986,
|
| 1924 |
+
"eval_samples_per_second": 12.483,
|
| 1925 |
+
"eval_steps_per_second": 12.483,
|
| 1926 |
+
"step": 3000
|
| 1927 |
}
|
| 1928 |
],
|
| 1929 |
"logging_steps": 10,
|
| 1930 |
+
"max_steps": 3000,
|
| 1931 |
"num_train_epochs": 9223372036854775807,
|
| 1932 |
"save_steps": 500,
|
| 1933 |
+
"total_flos": 2.635076136075264e+18,
|
| 1934 |
"trial_name": null,
|
| 1935 |
"trial_params": null
|
| 1936 |
}
|