Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +268 -457
trainer_state.json
CHANGED
|
@@ -1,667 +1,478 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
-
"epoch": 0.
|
| 13 |
-
"grad_norm": 5.
|
| 14 |
-
"learning_rate": 1.
|
| 15 |
-
"loss": 1.
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch": 0.
|
| 20 |
-
"grad_norm":
|
| 21 |
-
"learning_rate": 1.
|
| 22 |
-
"loss": 1.
|
| 23 |
"step": 10
|
| 24 |
},
|
| 25 |
{
|
| 26 |
-
"epoch": 0.
|
| 27 |
-
"grad_norm":
|
| 28 |
-
"learning_rate":
|
| 29 |
-
"loss": 0.
|
| 30 |
"step": 20
|
| 31 |
},
|
| 32 |
{
|
| 33 |
-
"epoch": 0.
|
| 34 |
-
"grad_norm": 0.
|
| 35 |
-
"learning_rate":
|
| 36 |
-
"loss": 0.
|
| 37 |
"step": 30
|
| 38 |
},
|
| 39 |
{
|
| 40 |
-
"epoch": 0.
|
| 41 |
-
"grad_norm": 0.
|
| 42 |
-
"learning_rate":
|
| 43 |
-
"loss": 0.
|
| 44 |
"step": 40
|
| 45 |
},
|
| 46 |
{
|
| 47 |
-
"epoch": 0.
|
| 48 |
-
"grad_norm": 0.
|
| 49 |
-
"learning_rate":
|
| 50 |
-
"loss": 0.
|
| 51 |
"step": 50
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch": 0.
|
| 55 |
-
"grad_norm": 0.
|
| 56 |
-
"learning_rate":
|
| 57 |
-
"loss": 0.
|
| 58 |
"step": 60
|
| 59 |
},
|
| 60 |
{
|
| 61 |
-
"epoch": 0.
|
| 62 |
-
"grad_norm": 0.
|
| 63 |
-
"learning_rate":
|
| 64 |
-
"loss": 0.
|
| 65 |
"step": 70
|
| 66 |
},
|
| 67 |
{
|
| 68 |
-
"epoch": 0.
|
| 69 |
-
"grad_norm": 0.
|
| 70 |
-
"learning_rate":
|
| 71 |
-
"loss": 0.
|
| 72 |
"step": 80
|
| 73 |
},
|
| 74 |
{
|
| 75 |
-
"epoch": 0.
|
| 76 |
-
"grad_norm": 0.
|
| 77 |
-
"learning_rate": 9.
|
| 78 |
-
"loss": 0.
|
| 79 |
"step": 90
|
| 80 |
},
|
| 81 |
{
|
| 82 |
-
"epoch": 0.
|
| 83 |
-
"grad_norm": 0.
|
| 84 |
-
"learning_rate": 9.
|
| 85 |
-
"loss": 0.
|
| 86 |
"step": 100
|
| 87 |
},
|
| 88 |
{
|
| 89 |
-
"epoch": 0.
|
| 90 |
-
"grad_norm": 0.
|
| 91 |
-
"learning_rate": 9.
|
| 92 |
-
"loss": 0.
|
| 93 |
"step": 110
|
| 94 |
},
|
| 95 |
{
|
| 96 |
-
"epoch": 0.
|
| 97 |
-
"grad_norm": 0.
|
| 98 |
-
"learning_rate": 9.
|
| 99 |
-
"loss": 0.
|
| 100 |
"step": 120
|
| 101 |
},
|
| 102 |
{
|
| 103 |
-
"epoch": 0.
|
| 104 |
-
"grad_norm": 0.
|
| 105 |
-
"learning_rate": 9.
|
| 106 |
-
"loss": 0.
|
| 107 |
"step": 130
|
| 108 |
},
|
| 109 |
{
|
| 110 |
-
"epoch": 0.
|
| 111 |
-
"grad_norm": 0.
|
| 112 |
-
"learning_rate": 9.
|
| 113 |
-
"loss": 0.
|
| 114 |
"step": 140
|
| 115 |
},
|
| 116 |
{
|
| 117 |
-
"epoch": 0.
|
| 118 |
-
"grad_norm": 0.
|
| 119 |
-
"learning_rate": 9.
|
| 120 |
-
"loss": 0.
|
| 121 |
"step": 150
|
| 122 |
},
|
| 123 |
{
|
| 124 |
-
"epoch": 0.
|
| 125 |
-
"grad_norm": 0.
|
| 126 |
-
"learning_rate": 9.
|
| 127 |
-
"loss": 0.
|
| 128 |
"step": 160
|
| 129 |
},
|
| 130 |
{
|
| 131 |
-
"epoch": 0.
|
| 132 |
-
"grad_norm": 0.
|
| 133 |
-
"learning_rate": 9.
|
| 134 |
-
"loss": 0.
|
| 135 |
"step": 170
|
| 136 |
},
|
| 137 |
{
|
| 138 |
-
"epoch": 0.
|
| 139 |
-
"grad_norm": 0.
|
| 140 |
-
"learning_rate": 9.
|
| 141 |
-
"loss": 0.
|
| 142 |
"step": 180
|
| 143 |
},
|
| 144 |
{
|
| 145 |
-
"epoch": 0.
|
| 146 |
-
"grad_norm": 0.
|
| 147 |
-
"learning_rate":
|
| 148 |
-
"loss": 0.
|
| 149 |
"step": 190
|
| 150 |
},
|
| 151 |
{
|
| 152 |
-
"epoch": 0.
|
| 153 |
-
"grad_norm": 0.
|
| 154 |
-
"learning_rate":
|
| 155 |
-
"loss": 0.
|
| 156 |
"step": 200
|
| 157 |
},
|
| 158 |
{
|
| 159 |
-
"epoch": 0.
|
| 160 |
-
"grad_norm": 0.
|
| 161 |
-
"learning_rate":
|
| 162 |
-
"loss": 0.
|
| 163 |
"step": 210
|
| 164 |
},
|
| 165 |
{
|
| 166 |
-
"epoch": 0.
|
| 167 |
-
"grad_norm": 0.
|
| 168 |
-
"learning_rate":
|
| 169 |
-
"loss": 0.
|
| 170 |
"step": 220
|
| 171 |
},
|
| 172 |
{
|
| 173 |
-
"epoch": 0.
|
| 174 |
-
"grad_norm": 0.
|
| 175 |
-
"learning_rate":
|
| 176 |
-
"loss": 0.
|
| 177 |
"step": 230
|
| 178 |
},
|
| 179 |
{
|
| 180 |
-
"epoch": 0.
|
| 181 |
-
"grad_norm": 0.
|
| 182 |
-
"learning_rate":
|
| 183 |
-
"loss": 0.
|
| 184 |
"step": 240
|
| 185 |
},
|
| 186 |
{
|
| 187 |
-
"epoch": 0.
|
| 188 |
-
"grad_norm": 0.
|
| 189 |
-
"learning_rate":
|
| 190 |
-
"loss": 0.
|
| 191 |
"step": 250
|
| 192 |
},
|
| 193 |
{
|
| 194 |
-
"epoch": 0.
|
| 195 |
-
"grad_norm": 0.
|
| 196 |
-
"learning_rate":
|
| 197 |
-
"loss": 0.
|
| 198 |
"step": 260
|
| 199 |
},
|
| 200 |
{
|
| 201 |
-
"epoch": 0.
|
| 202 |
-
"grad_norm": 0.
|
| 203 |
-
"learning_rate":
|
| 204 |
-
"loss": 0.
|
| 205 |
"step": 270
|
| 206 |
},
|
| 207 |
{
|
| 208 |
-
"epoch": 0.
|
| 209 |
-
"grad_norm": 0.
|
| 210 |
-
"learning_rate":
|
| 211 |
-
"loss": 0.
|
| 212 |
"step": 280
|
| 213 |
},
|
| 214 |
{
|
| 215 |
-
"epoch": 0.
|
| 216 |
-
"grad_norm": 0.
|
| 217 |
-
"learning_rate":
|
| 218 |
-
"loss": 0.
|
| 219 |
"step": 290
|
| 220 |
},
|
| 221 |
{
|
| 222 |
-
"epoch": 0.
|
| 223 |
-
"grad_norm": 0.
|
| 224 |
-
"learning_rate":
|
| 225 |
-
"loss": 0.
|
| 226 |
"step": 300
|
| 227 |
},
|
| 228 |
{
|
| 229 |
-
"epoch": 0.
|
| 230 |
-
"grad_norm": 0.
|
| 231 |
-
"learning_rate":
|
| 232 |
-
"loss": 0.
|
| 233 |
"step": 310
|
| 234 |
},
|
| 235 |
{
|
| 236 |
-
"epoch": 0.
|
| 237 |
-
"grad_norm": 0.
|
| 238 |
-
"learning_rate":
|
| 239 |
-
"loss": 0.
|
| 240 |
"step": 320
|
| 241 |
},
|
| 242 |
{
|
| 243 |
-
"epoch": 0.
|
| 244 |
-
"grad_norm": 0.
|
| 245 |
-
"learning_rate":
|
| 246 |
-
"loss": 0.
|
| 247 |
"step": 330
|
| 248 |
},
|
| 249 |
{
|
| 250 |
-
"epoch": 0.
|
| 251 |
-
"grad_norm": 0.
|
| 252 |
-
"learning_rate":
|
| 253 |
-
"loss": 0.
|
| 254 |
"step": 340
|
| 255 |
},
|
| 256 |
{
|
| 257 |
-
"epoch": 0.
|
| 258 |
-
"grad_norm": 0.
|
| 259 |
-
"learning_rate":
|
| 260 |
-
"loss": 0.
|
| 261 |
"step": 350
|
| 262 |
},
|
| 263 |
{
|
| 264 |
-
"epoch": 0.
|
| 265 |
-
"grad_norm": 0.
|
| 266 |
-
"learning_rate":
|
| 267 |
-
"loss": 0.
|
| 268 |
"step": 360
|
| 269 |
},
|
| 270 |
{
|
| 271 |
-
"epoch": 0.
|
| 272 |
-
"grad_norm": 0.
|
| 273 |
-
"learning_rate":
|
| 274 |
-
"loss": 0.
|
| 275 |
"step": 370
|
| 276 |
},
|
| 277 |
{
|
| 278 |
-
"epoch": 0.
|
| 279 |
-
"grad_norm": 0.
|
| 280 |
-
"learning_rate":
|
| 281 |
-
"loss": 0.
|
| 282 |
"step": 380
|
| 283 |
},
|
| 284 |
{
|
| 285 |
-
"epoch": 0.
|
| 286 |
-
"grad_norm": 0.
|
| 287 |
-
"learning_rate":
|
| 288 |
-
"loss": 0.
|
| 289 |
"step": 390
|
| 290 |
},
|
| 291 |
{
|
| 292 |
-
"epoch": 0.
|
| 293 |
-
"grad_norm": 0.
|
| 294 |
-
"learning_rate":
|
| 295 |
-
"loss": 0.
|
| 296 |
"step": 400
|
| 297 |
},
|
| 298 |
{
|
| 299 |
-
"epoch": 0.
|
| 300 |
-
"grad_norm": 0.
|
| 301 |
-
"learning_rate":
|
| 302 |
-
"loss": 0.
|
| 303 |
"step": 410
|
| 304 |
},
|
| 305 |
{
|
| 306 |
-
"epoch": 0.
|
| 307 |
-
"grad_norm": 0.
|
| 308 |
-
"learning_rate":
|
| 309 |
-
"loss": 0.
|
| 310 |
"step": 420
|
| 311 |
},
|
| 312 |
{
|
| 313 |
-
"epoch": 0.
|
| 314 |
-
"grad_norm": 0.
|
| 315 |
-
"learning_rate":
|
| 316 |
-
"loss": 0.
|
| 317 |
"step": 430
|
| 318 |
},
|
| 319 |
{
|
| 320 |
-
"epoch": 0.
|
| 321 |
-
"grad_norm": 0.
|
| 322 |
-
"learning_rate":
|
| 323 |
-
"loss": 0.
|
| 324 |
"step": 440
|
| 325 |
},
|
| 326 |
{
|
| 327 |
-
"epoch": 0.
|
| 328 |
-
"grad_norm": 0.
|
| 329 |
-
"learning_rate":
|
| 330 |
-
"loss": 0.
|
| 331 |
"step": 450
|
| 332 |
},
|
| 333 |
{
|
| 334 |
-
"epoch": 0.
|
| 335 |
-
"grad_norm": 0.
|
| 336 |
-
"learning_rate":
|
| 337 |
-
"loss": 0.
|
| 338 |
"step": 460
|
| 339 |
},
|
| 340 |
{
|
| 341 |
-
"epoch": 0.
|
| 342 |
-
"grad_norm": 0.
|
| 343 |
-
"learning_rate":
|
| 344 |
-
"loss": 0.
|
| 345 |
"step": 470
|
| 346 |
},
|
| 347 |
{
|
| 348 |
-
"epoch": 0.
|
| 349 |
-
"grad_norm": 0.
|
| 350 |
-
"learning_rate":
|
| 351 |
-
"loss": 0.
|
| 352 |
"step": 480
|
| 353 |
},
|
| 354 |
{
|
| 355 |
-
"epoch": 0.
|
| 356 |
-
"grad_norm": 0.
|
| 357 |
-
"learning_rate":
|
| 358 |
-
"loss": 0.
|
| 359 |
"step": 490
|
| 360 |
},
|
| 361 |
{
|
| 362 |
-
"epoch": 0.
|
| 363 |
-
"grad_norm": 0.
|
| 364 |
-
"learning_rate":
|
| 365 |
-
"loss": 0.
|
| 366 |
"step": 500
|
| 367 |
},
|
| 368 |
{
|
| 369 |
-
"epoch": 0.
|
| 370 |
-
"grad_norm": 0.
|
| 371 |
-
"learning_rate":
|
| 372 |
-
"loss": 0.
|
| 373 |
"step": 510
|
| 374 |
},
|
| 375 |
{
|
| 376 |
-
"epoch": 0.
|
| 377 |
-
"grad_norm": 0.
|
| 378 |
-
"learning_rate":
|
| 379 |
-
"loss": 0.
|
| 380 |
"step": 520
|
| 381 |
},
|
| 382 |
{
|
| 383 |
-
"epoch": 0.
|
| 384 |
-
"grad_norm": 0.
|
| 385 |
-
"learning_rate":
|
| 386 |
-
"loss": 0.
|
| 387 |
"step": 530
|
| 388 |
},
|
| 389 |
{
|
| 390 |
-
"epoch": 0.
|
| 391 |
-
"grad_norm": 0.
|
| 392 |
-
"learning_rate":
|
| 393 |
-
"loss": 0.
|
| 394 |
"step": 540
|
| 395 |
},
|
| 396 |
{
|
| 397 |
-
"epoch": 0.
|
| 398 |
-
"grad_norm": 0.
|
| 399 |
-
"learning_rate":
|
| 400 |
-
"loss": 0.
|
| 401 |
"step": 550
|
| 402 |
},
|
| 403 |
{
|
| 404 |
-
"epoch": 0.
|
| 405 |
-
"grad_norm": 0.
|
| 406 |
-
"learning_rate":
|
| 407 |
-
"loss": 0.
|
| 408 |
"step": 560
|
| 409 |
},
|
| 410 |
{
|
| 411 |
-
"epoch": 0.
|
| 412 |
-
"grad_norm": 0.
|
| 413 |
-
"learning_rate":
|
| 414 |
-
"loss": 0.
|
| 415 |
"step": 570
|
| 416 |
},
|
| 417 |
{
|
| 418 |
-
"epoch": 0.
|
| 419 |
-
"grad_norm": 0.
|
| 420 |
-
"learning_rate": 3.
|
| 421 |
-
"loss": 0.
|
| 422 |
"step": 580
|
| 423 |
},
|
| 424 |
{
|
| 425 |
-
"epoch": 0.
|
| 426 |
-
"grad_norm": 0.
|
| 427 |
-
"learning_rate":
|
| 428 |
-
"loss": 0.
|
| 429 |
"step": 590
|
| 430 |
},
|
| 431 |
{
|
| 432 |
-
"epoch": 0.
|
| 433 |
-
"grad_norm": 0.
|
| 434 |
-
"learning_rate":
|
| 435 |
-
"loss": 0.
|
| 436 |
"step": 600
|
| 437 |
},
|
| 438 |
{
|
| 439 |
-
"epoch": 0.
|
| 440 |
-
"grad_norm": 0.
|
| 441 |
-
"learning_rate":
|
| 442 |
-
"loss": 0.
|
| 443 |
"step": 610
|
| 444 |
},
|
| 445 |
{
|
| 446 |
-
"epoch": 0.
|
| 447 |
-
"grad_norm": 0.
|
| 448 |
-
"learning_rate":
|
| 449 |
-
"loss": 0.
|
| 450 |
"step": 620
|
| 451 |
},
|
| 452 |
{
|
| 453 |
-
"epoch": 0.
|
| 454 |
-
"grad_norm": 0.
|
| 455 |
-
"learning_rate":
|
| 456 |
-
"loss": 0.
|
| 457 |
"step": 630
|
| 458 |
},
|
| 459 |
{
|
| 460 |
-
"epoch": 0.
|
| 461 |
-
"grad_norm": 0.
|
| 462 |
-
"learning_rate":
|
| 463 |
-
"loss": 0.
|
| 464 |
"step": 640
|
| 465 |
},
|
| 466 |
{
|
| 467 |
-
"epoch": 0.
|
| 468 |
-
"grad_norm": 0.
|
| 469 |
-
"learning_rate":
|
| 470 |
-
"loss": 0.
|
| 471 |
"step": 650
|
| 472 |
-
},
|
| 473 |
-
{
|
| 474 |
-
"epoch": 0.7152533188837713,
|
| 475 |
-
"grad_norm": 0.1748624015228809,
|
| 476 |
-
"learning_rate": 2.2685995956697037e-06,
|
| 477 |
-
"loss": 0.1568,
|
| 478 |
-
"step": 660
|
| 479 |
-
},
|
| 480 |
-
{
|
| 481 |
-
"epoch": 0.7260904903820103,
|
| 482 |
-
"grad_norm": 0.1723078141044498,
|
| 483 |
-
"learning_rate": 2.1118889991365477e-06,
|
| 484 |
-
"loss": 0.1544,
|
| 485 |
-
"step": 670
|
| 486 |
-
},
|
| 487 |
-
{
|
| 488 |
-
"epoch": 0.7369276618802493,
|
| 489 |
-
"grad_norm": 0.14553092139119517,
|
| 490 |
-
"learning_rate": 1.9593255797775578e-06,
|
| 491 |
-
"loss": 0.1467,
|
| 492 |
-
"step": 680
|
| 493 |
-
},
|
| 494 |
-
{
|
| 495 |
-
"epoch": 0.7477648333784882,
|
| 496 |
-
"grad_norm": 0.16342951046374027,
|
| 497 |
-
"learning_rate": 1.811128410727454e-06,
|
| 498 |
-
"loss": 0.1565,
|
| 499 |
-
"step": 690
|
| 500 |
-
},
|
| 501 |
-
{
|
| 502 |
-
"epoch": 0.7586020048767271,
|
| 503 |
-
"grad_norm": 0.15898031074835672,
|
| 504 |
-
"learning_rate": 1.6675102954126203e-06,
|
| 505 |
-
"loss": 0.1672,
|
| 506 |
-
"step": 700
|
| 507 |
-
},
|
| 508 |
-
{
|
| 509 |
-
"epoch": 0.7694391763749662,
|
| 510 |
-
"grad_norm": 0.07132360800064767,
|
| 511 |
-
"learning_rate": 1.528677461976451e-06,
|
| 512 |
-
"loss": 0.1411,
|
| 513 |
-
"step": 710
|
| 514 |
-
},
|
| 515 |
-
{
|
| 516 |
-
"epoch": 0.7802763478732051,
|
| 517 |
-
"grad_norm": 0.14147704428723085,
|
| 518 |
-
"learning_rate": 1.3948292671464708e-06,
|
| 519 |
-
"loss": 0.1454,
|
| 520 |
-
"step": 720
|
| 521 |
-
},
|
| 522 |
-
{
|
| 523 |
-
"epoch": 0.791113519371444,
|
| 524 |
-
"grad_norm": 0.1217153272659515,
|
| 525 |
-
"learning_rate": 1.2661579099684345e-06,
|
| 526 |
-
"loss": 0.1367,
|
| 527 |
-
"step": 730
|
| 528 |
-
},
|
| 529 |
-
{
|
| 530 |
-
"epoch": 0.801950690869683,
|
| 531 |
-
"grad_norm": 0.14222802833323006,
|
| 532 |
-
"learning_rate": 1.1428481558184985e-06,
|
| 533 |
-
"loss": 0.1658,
|
| 534 |
-
"step": 740
|
| 535 |
-
},
|
| 536 |
-
{
|
| 537 |
-
"epoch": 0.812787862367922,
|
| 538 |
-
"grad_norm": 0.15290868142041023,
|
| 539 |
-
"learning_rate": 1.0250770710897513e-06,
|
| 540 |
-
"loss": 0.1837,
|
| 541 |
-
"step": 750
|
| 542 |
-
},
|
| 543 |
-
{
|
| 544 |
-
"epoch": 0.8236250338661609,
|
| 545 |
-
"grad_norm": 0.10220156710073305,
|
| 546 |
-
"learning_rate": 9.13013768934084e-07,
|
| 547 |
-
"loss": 0.1642,
|
| 548 |
-
"step": 760
|
| 549 |
-
},
|
| 550 |
-
{
|
| 551 |
-
"epoch": 0.8344622053643999,
|
| 552 |
-
"grad_norm": 0.13901575395063445,
|
| 553 |
-
"learning_rate": 8.068191664244945e-07,
|
| 554 |
-
"loss": 0.1637,
|
| 555 |
-
"step": 770
|
| 556 |
-
},
|
| 557 |
-
{
|
| 558 |
-
"epoch": 0.8452993768626389,
|
| 559 |
-
"grad_norm": 0.13154009484022863,
|
| 560 |
-
"learning_rate": 7.066457534865529e-07,
|
| 561 |
-
"loss": 0.1485,
|
| 562 |
-
"step": 780
|
| 563 |
-
},
|
| 564 |
-
{
|
| 565 |
-
"epoch": 0.8561365483608778,
|
| 566 |
-
"grad_norm": 0.15411518006395045,
|
| 567 |
-
"learning_rate": 6.126373739307856e-07,
|
| 568 |
-
"loss": 0.1351,
|
| 569 |
-
"step": 790
|
| 570 |
-
},
|
| 571 |
-
{
|
| 572 |
-
"epoch": 0.8669737198591168,
|
| 573 |
-
"grad_norm": 0.1465409663251719,
|
| 574 |
-
"learning_rate": 5.249290189004552e-07,
|
| 575 |
-
"loss": 0.1616,
|
| 576 |
-
"step": 800
|
| 577 |
-
},
|
| 578 |
-
{
|
| 579 |
-
"epoch": 0.8778108913573557,
|
| 580 |
-
"grad_norm": 0.151355342159477,
|
| 581 |
-
"learning_rate": 4.4364663303129664e-07,
|
| 582 |
-
"loss": 0.1338,
|
| 583 |
-
"step": 810
|
| 584 |
-
},
|
| 585 |
-
{
|
| 586 |
-
"epoch": 0.8886480628555947,
|
| 587 |
-
"grad_norm": 0.13076440882361037,
|
| 588 |
-
"learning_rate": 3.6890693360157104e-07,
|
| 589 |
-
"loss": 0.1415,
|
| 590 |
-
"step": 820
|
| 591 |
-
},
|
| 592 |
-
{
|
| 593 |
-
"epoch": 0.8994852343538337,
|
| 594 |
-
"grad_norm": 0.1112996143342554,
|
| 595 |
-
"learning_rate": 3.0081724293212653e-07,
|
| 596 |
-
"loss": 0.1552,
|
| 597 |
-
"step": 830
|
| 598 |
-
},
|
| 599 |
-
{
|
| 600 |
-
"epoch": 0.9103224058520726,
|
| 601 |
-
"grad_norm": 0.09563035386898121,
|
| 602 |
-
"learning_rate": 2.3947533427712667e-07,
|
| 603 |
-
"loss": 0.1501,
|
| 604 |
-
"step": 840
|
| 605 |
-
},
|
| 606 |
-
{
|
| 607 |
-
"epoch": 0.9211595773503116,
|
| 608 |
-
"grad_norm": 0.15792781910845813,
|
| 609 |
-
"learning_rate": 1.8496929142674424e-07,
|
| 610 |
-
"loss": 0.1491,
|
| 611 |
-
"step": 850
|
| 612 |
-
},
|
| 613 |
-
{
|
| 614 |
-
"epoch": 0.9319967488485505,
|
| 615 |
-
"grad_norm": 0.16164911200152776,
|
| 616 |
-
"learning_rate": 1.3737738222341813e-07,
|
| 617 |
-
"loss": 0.1552,
|
| 618 |
-
"step": 860
|
| 619 |
-
},
|
| 620 |
-
{
|
| 621 |
-
"epoch": 0.9428339203467895,
|
| 622 |
-
"grad_norm": 0.15162312410480808,
|
| 623 |
-
"learning_rate": 9.67679461733051e-08,
|
| 624 |
-
"loss": 0.1379,
|
| 625 |
-
"step": 870
|
| 626 |
-
},
|
| 627 |
-
{
|
| 628 |
-
"epoch": 0.9536710918450284,
|
| 629 |
-
"grad_norm": 0.11915620491115217,
|
| 630 |
-
"learning_rate": 6.319929631430077e-08,
|
| 631 |
-
"loss": 0.1203,
|
| 632 |
-
"step": 880
|
| 633 |
-
},
|
| 634 |
-
{
|
| 635 |
-
"epoch": 0.9645082633432674,
|
| 636 |
-
"grad_norm": 0.16926863067932774,
|
| 637 |
-
"learning_rate": 3.671963548155244e-08,
|
| 638 |
-
"loss": 0.1523,
|
| 639 |
-
"step": 890
|
| 640 |
-
},
|
| 641 |
-
{
|
| 642 |
-
"epoch": 0.9753454348415064,
|
| 643 |
-
"grad_norm": 0.14805382379357954,
|
| 644 |
-
"learning_rate": 1.736698709069673e-08,
|
| 645 |
-
"loss": 0.1425,
|
| 646 |
-
"step": 900
|
| 647 |
-
},
|
| 648 |
-
{
|
| 649 |
-
"epoch": 0.9861826063397453,
|
| 650 |
-
"grad_norm": 0.17214363284720982,
|
| 651 |
-
"learning_rate": 5.169140538207051e-09,
|
| 652 |
-
"loss": 0.1496,
|
| 653 |
-
"step": 910
|
| 654 |
-
},
|
| 655 |
-
{
|
| 656 |
-
"epoch": 0.9970197778379842,
|
| 657 |
-
"grad_norm": 0.15327122063501583,
|
| 658 |
-
"learning_rate": 1.4361129727025193e-10,
|
| 659 |
-
"loss": 0.1459,
|
| 660 |
-
"step": 920
|
| 661 |
}
|
| 662 |
],
|
| 663 |
"logging_steps": 10,
|
| 664 |
-
"max_steps":
|
| 665 |
"num_input_tokens_seen": 0,
|
| 666 |
"num_train_epochs": 1,
|
| 667 |
"save_steps": 500,
|
|
@@ -677,7 +488,7 @@
|
|
| 677 |
"attributes": {}
|
| 678 |
}
|
| 679 |
},
|
| 680 |
-
"total_flos":
|
| 681 |
"train_batch_size": 1,
|
| 682 |
"trial_name": null,
|
| 683 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.9988545246277205,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 654,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.0015273004963726614,
|
| 13 |
+
"grad_norm": 5.274756775371289,
|
| 14 |
+
"learning_rate": 1.5151515151515152e-07,
|
| 15 |
+
"loss": 1.3724,
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.015273004963726614,
|
| 20 |
+
"grad_norm": 2.4356823026963044,
|
| 21 |
+
"learning_rate": 1.5151515151515152e-06,
|
| 22 |
+
"loss": 1.3215,
|
| 23 |
"step": 10
|
| 24 |
},
|
| 25 |
{
|
| 26 |
+
"epoch": 0.030546009927453228,
|
| 27 |
+
"grad_norm": 0.8415959934340891,
|
| 28 |
+
"learning_rate": 3.0303030303030305e-06,
|
| 29 |
+
"loss": 0.7041,
|
| 30 |
"step": 20
|
| 31 |
},
|
| 32 |
{
|
| 33 |
+
"epoch": 0.045819014891179836,
|
| 34 |
+
"grad_norm": 0.3951472895674848,
|
| 35 |
+
"learning_rate": 4.5454545454545455e-06,
|
| 36 |
+
"loss": 0.3621,
|
| 37 |
"step": 30
|
| 38 |
},
|
| 39 |
{
|
| 40 |
+
"epoch": 0.061092019854906456,
|
| 41 |
+
"grad_norm": 0.3302170549897179,
|
| 42 |
+
"learning_rate": 6.060606060606061e-06,
|
| 43 |
+
"loss": 0.2555,
|
| 44 |
"step": 40
|
| 45 |
},
|
| 46 |
{
|
| 47 |
+
"epoch": 0.07636502481863307,
|
| 48 |
+
"grad_norm": 0.3855993238810659,
|
| 49 |
+
"learning_rate": 7.5757575757575764e-06,
|
| 50 |
+
"loss": 0.21,
|
| 51 |
"step": 50
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.09163802978235967,
|
| 55 |
+
"grad_norm": 0.3643941392356843,
|
| 56 |
+
"learning_rate": 9.090909090909091e-06,
|
| 57 |
+
"loss": 0.1707,
|
| 58 |
"step": 60
|
| 59 |
},
|
| 60 |
{
|
| 61 |
+
"epoch": 0.10691103474608629,
|
| 62 |
+
"grad_norm": 0.24356795945348528,
|
| 63 |
+
"learning_rate": 9.99885820390154e-06,
|
| 64 |
+
"loss": 0.1628,
|
| 65 |
"step": 70
|
| 66 |
},
|
| 67 |
{
|
| 68 |
+
"epoch": 0.12218403970981291,
|
| 69 |
+
"grad_norm": 0.18282398542383,
|
| 70 |
+
"learning_rate": 9.986018985905901e-06,
|
| 71 |
+
"loss": 0.1537,
|
| 72 |
"step": 80
|
| 73 |
},
|
| 74 |
{
|
| 75 |
+
"epoch": 0.13745704467353953,
|
| 76 |
+
"grad_norm": 0.17824704057472698,
|
| 77 |
+
"learning_rate": 9.95895006911623e-06,
|
| 78 |
+
"loss": 0.1472,
|
| 79 |
"step": 90
|
| 80 |
},
|
| 81 |
{
|
| 82 |
+
"epoch": 0.15273004963726614,
|
| 83 |
+
"grad_norm": 0.20039388213383333,
|
| 84 |
+
"learning_rate": 9.917728706052765e-06,
|
| 85 |
+
"loss": 0.1512,
|
| 86 |
"step": 100
|
| 87 |
},
|
| 88 |
{
|
| 89 |
+
"epoch": 0.16800305460099274,
|
| 90 |
+
"grad_norm": 0.21394666608149668,
|
| 91 |
+
"learning_rate": 9.862472539183757e-06,
|
| 92 |
+
"loss": 0.1446,
|
| 93 |
"step": 110
|
| 94 |
},
|
| 95 |
{
|
| 96 |
+
"epoch": 0.18327605956471935,
|
| 97 |
+
"grad_norm": 0.1399694568156125,
|
| 98 |
+
"learning_rate": 9.793339265183303e-06,
|
| 99 |
+
"loss": 0.1379,
|
| 100 |
"step": 120
|
| 101 |
},
|
| 102 |
{
|
| 103 |
+
"epoch": 0.19854906452844598,
|
| 104 |
+
"grad_norm": 0.17756723955927178,
|
| 105 |
+
"learning_rate": 9.710526184877667e-06,
|
| 106 |
+
"loss": 0.1446,
|
| 107 |
"step": 130
|
| 108 |
},
|
| 109 |
{
|
| 110 |
+
"epoch": 0.21382206949217258,
|
| 111 |
+
"grad_norm": 0.1256183224074801,
|
| 112 |
+
"learning_rate": 9.61426964016452e-06,
|
| 113 |
+
"loss": 0.1598,
|
| 114 |
"step": 140
|
| 115 |
},
|
| 116 |
{
|
| 117 |
+
"epoch": 0.2290950744558992,
|
| 118 |
+
"grad_norm": 0.1925308832459223,
|
| 119 |
+
"learning_rate": 9.504844339512096e-06,
|
| 120 |
+
"loss": 0.1703,
|
| 121 |
"step": 150
|
| 122 |
},
|
| 123 |
{
|
| 124 |
+
"epoch": 0.24436807941962582,
|
| 125 |
+
"grad_norm": 0.18425814277453983,
|
| 126 |
+
"learning_rate": 9.382562573963238e-06,
|
| 127 |
+
"loss": 0.1347,
|
| 128 |
"step": 160
|
| 129 |
},
|
| 130 |
{
|
| 131 |
+
"epoch": 0.2596410843833524,
|
| 132 |
+
"grad_norm": 0.16364274896349304,
|
| 133 |
+
"learning_rate": 9.24777332588177e-06,
|
| 134 |
+
"loss": 0.1291,
|
| 135 |
"step": 170
|
| 136 |
},
|
| 137 |
{
|
| 138 |
+
"epoch": 0.27491408934707906,
|
| 139 |
+
"grad_norm": 0.1575283767974318,
|
| 140 |
+
"learning_rate": 9.10086127298478e-06,
|
| 141 |
+
"loss": 0.1455,
|
| 142 |
"step": 180
|
| 143 |
},
|
| 144 |
{
|
| 145 |
+
"epoch": 0.29018709431080564,
|
| 146 |
+
"grad_norm": 0.16299547711079815,
|
| 147 |
+
"learning_rate": 8.94224569050324e-06,
|
| 148 |
+
"loss": 0.1448,
|
| 149 |
"step": 190
|
| 150 |
},
|
| 151 |
{
|
| 152 |
+
"epoch": 0.30546009927453227,
|
| 153 |
+
"grad_norm": 0.1678304349758231,
|
| 154 |
+
"learning_rate": 8.772379254604074e-06,
|
| 155 |
+
"loss": 0.1329,
|
| 156 |
"step": 200
|
| 157 |
},
|
| 158 |
{
|
| 159 |
+
"epoch": 0.3207331042382589,
|
| 160 |
+
"grad_norm": 0.18133792526251943,
|
| 161 |
+
"learning_rate": 8.591746750488639e-06,
|
| 162 |
+
"loss": 0.1523,
|
| 163 |
"step": 210
|
| 164 |
},
|
| 165 |
{
|
| 166 |
+
"epoch": 0.3360061092019855,
|
| 167 |
+
"grad_norm": 0.138988177087972,
|
| 168 |
+
"learning_rate": 8.400863688854598e-06,
|
| 169 |
+
"loss": 0.1487,
|
| 170 |
"step": 220
|
| 171 |
},
|
| 172 |
{
|
| 173 |
+
"epoch": 0.3512791141657121,
|
| 174 |
+
"grad_norm": 0.10975239156978413,
|
| 175 |
+
"learning_rate": 8.200274834669675e-06,
|
| 176 |
+
"loss": 0.1454,
|
| 177 |
"step": 230
|
| 178 |
},
|
| 179 |
{
|
| 180 |
+
"epoch": 0.3665521191294387,
|
| 181 |
+
"grad_norm": 0.13022143377921225,
|
| 182 |
+
"learning_rate": 7.99055265245608e-06,
|
| 183 |
+
"loss": 0.1414,
|
| 184 |
"step": 240
|
| 185 |
},
|
| 186 |
{
|
| 187 |
+
"epoch": 0.3818251240931653,
|
| 188 |
+
"grad_norm": 0.09325598021130543,
|
| 189 |
+
"learning_rate": 7.772295672522615e-06,
|
| 190 |
+
"loss": 0.1406,
|
| 191 |
"step": 250
|
| 192 |
},
|
| 193 |
{
|
| 194 |
+
"epoch": 0.39709812905689196,
|
| 195 |
+
"grad_norm": 0.10192417440664955,
|
| 196 |
+
"learning_rate": 7.546126782807117e-06,
|
| 197 |
+
"loss": 0.1243,
|
| 198 |
"step": 260
|
| 199 |
},
|
| 200 |
{
|
| 201 |
+
"epoch": 0.41237113402061853,
|
| 202 |
+
"grad_norm": 0.10214456159318729,
|
| 203 |
+
"learning_rate": 7.312691451204178e-06,
|
| 204 |
+
"loss": 0.1381,
|
| 205 |
"step": 270
|
| 206 |
},
|
| 207 |
{
|
| 208 |
+
"epoch": 0.42764413898434517,
|
| 209 |
+
"grad_norm": 0.11650402115334278,
|
| 210 |
+
"learning_rate": 7.072655883451478e-06,
|
| 211 |
+
"loss": 0.1155,
|
| 212 |
"step": 280
|
| 213 |
},
|
| 214 |
{
|
| 215 |
+
"epoch": 0.4429171439480718,
|
| 216 |
+
"grad_norm": 0.09899147898887324,
|
| 217 |
+
"learning_rate": 6.8267051218319766e-06,
|
| 218 |
+
"loss": 0.1372,
|
| 219 |
"step": 290
|
| 220 |
},
|
| 221 |
{
|
| 222 |
+
"epoch": 0.4581901489117984,
|
| 223 |
+
"grad_norm": 0.11909991495605356,
|
| 224 |
+
"learning_rate": 6.575541090118105e-06,
|
| 225 |
+
"loss": 0.1449,
|
| 226 |
"step": 300
|
| 227 |
},
|
| 228 |
{
|
| 229 |
+
"epoch": 0.473463153875525,
|
| 230 |
+
"grad_norm": 0.11324258672549159,
|
| 231 |
+
"learning_rate": 6.319880590337549e-06,
|
| 232 |
+
"loss": 0.1355,
|
| 233 |
"step": 310
|
| 234 |
},
|
| 235 |
{
|
| 236 |
+
"epoch": 0.48873615883925164,
|
| 237 |
+
"grad_norm": 0.15684721556128514,
|
| 238 |
+
"learning_rate": 6.060453257077686e-06,
|
| 239 |
+
"loss": 0.1448,
|
| 240 |
"step": 320
|
| 241 |
},
|
| 242 |
{
|
| 243 |
+
"epoch": 0.5040091638029782,
|
| 244 |
+
"grad_norm": 0.13428699506671887,
|
| 245 |
+
"learning_rate": 5.797999475166897e-06,
|
| 246 |
+
"loss": 0.161,
|
| 247 |
"step": 330
|
| 248 |
},
|
| 249 |
{
|
| 250 |
+
"epoch": 0.5192821687667049,
|
| 251 |
+
"grad_norm": 0.16260953903247685,
|
| 252 |
+
"learning_rate": 5.533268266675601e-06,
|
| 253 |
+
"loss": 0.1301,
|
| 254 |
"step": 340
|
| 255 |
},
|
| 256 |
{
|
| 257 |
+
"epoch": 0.5345551737304315,
|
| 258 |
+
"grad_norm": 0.138117775450493,
|
| 259 |
+
"learning_rate": 5.267015153267246e-06,
|
| 260 |
+
"loss": 0.1153,
|
| 261 |
"step": 350
|
| 262 |
},
|
| 263 |
{
|
| 264 |
+
"epoch": 0.5498281786941581,
|
| 265 |
+
"grad_norm": 0.09342193422842268,
|
| 266 |
+
"learning_rate": 5e-06,
|
| 267 |
+
"loss": 0.1388,
|
| 268 |
"step": 360
|
| 269 |
},
|
| 270 |
{
|
| 271 |
+
"epoch": 0.5651011836578846,
|
| 272 |
+
"grad_norm": 0.12521354455054481,
|
| 273 |
+
"learning_rate": 4.732984846732755e-06,
|
| 274 |
+
"loss": 0.1469,
|
| 275 |
"step": 370
|
| 276 |
},
|
| 277 |
{
|
| 278 |
+
"epoch": 0.5803741886216113,
|
| 279 |
+
"grad_norm": 0.12074643796303608,
|
| 280 |
+
"learning_rate": 4.466731733324399e-06,
|
| 281 |
+
"loss": 0.1336,
|
| 282 |
"step": 380
|
| 283 |
},
|
| 284 |
{
|
| 285 |
+
"epoch": 0.5956471935853379,
|
| 286 |
+
"grad_norm": 0.10811612216811932,
|
| 287 |
+
"learning_rate": 4.2020005248331056e-06,
|
| 288 |
+
"loss": 0.1444,
|
| 289 |
"step": 390
|
| 290 |
},
|
| 291 |
{
|
| 292 |
+
"epoch": 0.6109201985490645,
|
| 293 |
+
"grad_norm": 0.07932149519427452,
|
| 294 |
+
"learning_rate": 3.939546742922318e-06,
|
| 295 |
+
"loss": 0.1217,
|
| 296 |
"step": 400
|
| 297 |
},
|
| 298 |
{
|
| 299 |
+
"epoch": 0.6261932035127912,
|
| 300 |
+
"grad_norm": 0.13495076484320434,
|
| 301 |
+
"learning_rate": 3.6801194096624515e-06,
|
| 302 |
+
"loss": 0.1486,
|
| 303 |
"step": 410
|
| 304 |
},
|
| 305 |
{
|
| 306 |
+
"epoch": 0.6414662084765178,
|
| 307 |
+
"grad_norm": 0.09577670394462665,
|
| 308 |
+
"learning_rate": 3.424458909881897e-06,
|
| 309 |
+
"loss": 0.1351,
|
| 310 |
"step": 420
|
| 311 |
},
|
| 312 |
{
|
| 313 |
+
"epoch": 0.6567392134402443,
|
| 314 |
+
"grad_norm": 0.13052717933517727,
|
| 315 |
+
"learning_rate": 3.173294878168025e-06,
|
| 316 |
+
"loss": 0.1115,
|
| 317 |
"step": 430
|
| 318 |
},
|
| 319 |
{
|
| 320 |
+
"epoch": 0.672012218403971,
|
| 321 |
+
"grad_norm": 0.10868560247173807,
|
| 322 |
+
"learning_rate": 2.9273441165485227e-06,
|
| 323 |
+
"loss": 0.1429,
|
| 324 |
"step": 440
|
| 325 |
},
|
| 326 |
{
|
| 327 |
+
"epoch": 0.6872852233676976,
|
| 328 |
+
"grad_norm": 0.08492718279444092,
|
| 329 |
+
"learning_rate": 2.687308548795825e-06,
|
| 330 |
+
"loss": 0.1006,
|
| 331 |
"step": 450
|
| 332 |
},
|
| 333 |
{
|
| 334 |
+
"epoch": 0.7025582283314242,
|
| 335 |
+
"grad_norm": 0.11879878396233616,
|
| 336 |
+
"learning_rate": 2.4538732171928847e-06,
|
| 337 |
+
"loss": 0.1441,
|
| 338 |
"step": 460
|
| 339 |
},
|
| 340 |
{
|
| 341 |
+
"epoch": 0.7178312332951509,
|
| 342 |
+
"grad_norm": 0.08627366089386783,
|
| 343 |
+
"learning_rate": 2.2277043274773856e-06,
|
| 344 |
+
"loss": 0.121,
|
| 345 |
"step": 470
|
| 346 |
},
|
| 347 |
{
|
| 348 |
+
"epoch": 0.7331042382588774,
|
| 349 |
+
"grad_norm": 0.11860493468588708,
|
| 350 |
+
"learning_rate": 2.00944734754392e-06,
|
| 351 |
+
"loss": 0.104,
|
| 352 |
"step": 480
|
| 353 |
},
|
| 354 |
{
|
| 355 |
+
"epoch": 0.748377243222604,
|
| 356 |
+
"grad_norm": 0.13566719962532361,
|
| 357 |
+
"learning_rate": 1.7997251653303249e-06,
|
| 358 |
+
"loss": 0.1223,
|
| 359 |
"step": 490
|
| 360 |
},
|
| 361 |
{
|
| 362 |
+
"epoch": 0.7636502481863306,
|
| 363 |
+
"grad_norm": 0.11713757986892694,
|
| 364 |
+
"learning_rate": 1.5991363111454023e-06,
|
| 365 |
+
"loss": 0.1297,
|
| 366 |
"step": 500
|
| 367 |
},
|
| 368 |
{
|
| 369 |
+
"epoch": 0.7789232531500573,
|
| 370 |
+
"grad_norm": 0.0932474027124396,
|
| 371 |
+
"learning_rate": 1.4082532495113627e-06,
|
| 372 |
+
"loss": 0.129,
|
| 373 |
"step": 510
|
| 374 |
},
|
| 375 |
{
|
| 376 |
+
"epoch": 0.7941962581137839,
|
| 377 |
+
"grad_norm": 0.06960734581371894,
|
| 378 |
+
"learning_rate": 1.2276207453959283e-06,
|
| 379 |
+
"loss": 0.102,
|
| 380 |
"step": 520
|
| 381 |
},
|
| 382 |
{
|
| 383 |
+
"epoch": 0.8094692630775105,
|
| 384 |
+
"grad_norm": 0.1054188314922148,
|
| 385 |
+
"learning_rate": 1.0577543094967613e-06,
|
| 386 |
+
"loss": 0.1182,
|
| 387 |
"step": 530
|
| 388 |
},
|
| 389 |
{
|
| 390 |
+
"epoch": 0.8247422680412371,
|
| 391 |
+
"grad_norm": 0.15433498716395788,
|
| 392 |
+
"learning_rate": 8.991387270152202e-07,
|
| 393 |
+
"loss": 0.1306,
|
| 394 |
"step": 540
|
| 395 |
},
|
| 396 |
{
|
| 397 |
+
"epoch": 0.8400152730049637,
|
| 398 |
+
"grad_norm": 0.10854241303868371,
|
| 399 |
+
"learning_rate": 7.522266741182305e-07,
|
| 400 |
+
"loss": 0.11,
|
| 401 |
"step": 550
|
| 402 |
},
|
| 403 |
{
|
| 404 |
+
"epoch": 0.8552882779686903,
|
| 405 |
+
"grad_norm": 0.13631634648737045,
|
| 406 |
+
"learning_rate": 6.174374260367611e-07,
|
| 407 |
+
"loss": 0.1162,
|
| 408 |
"step": 560
|
| 409 |
},
|
| 410 |
{
|
| 411 |
+
"epoch": 0.870561282932417,
|
| 412 |
+
"grad_norm": 0.11447961702982919,
|
| 413 |
+
"learning_rate": 4.951556604879049e-07,
|
| 414 |
+
"loss": 0.1396,
|
| 415 |
"step": 570
|
| 416 |
},
|
| 417 |
{
|
| 418 |
+
"epoch": 0.8858342878961436,
|
| 419 |
+
"grad_norm": 0.1527666931393435,
|
| 420 |
+
"learning_rate": 3.8573035983548167e-07,
|
| 421 |
+
"loss": 0.1353,
|
| 422 |
"step": 580
|
| 423 |
},
|
| 424 |
{
|
| 425 |
+
"epoch": 0.9011072928598702,
|
| 426 |
+
"grad_norm": 0.13839722389025982,
|
| 427 |
+
"learning_rate": 2.894738151223331e-07,
|
| 428 |
+
"loss": 0.1193,
|
| 429 |
"step": 590
|
| 430 |
},
|
| 431 |
{
|
| 432 |
+
"epoch": 0.9163802978235968,
|
| 433 |
+
"grad_norm": 0.13196785663691574,
|
| 434 |
+
"learning_rate": 2.0666073481669714e-07,
|
| 435 |
+
"loss": 0.1147,
|
| 436 |
"step": 600
|
| 437 |
},
|
| 438 |
{
|
| 439 |
+
"epoch": 0.9316533027873234,
|
| 440 |
+
"grad_norm": 0.09878735062205103,
|
| 441 |
+
"learning_rate": 1.375274608162447e-07,
|
| 442 |
+
"loss": 0.1261,
|
| 443 |
"step": 610
|
| 444 |
},
|
| 445 |
{
|
| 446 |
+
"epoch": 0.94692630775105,
|
| 447 |
+
"grad_norm": 0.07952980044315368,
|
| 448 |
+
"learning_rate": 8.227129394723643e-08,
|
| 449 |
+
"loss": 0.0992,
|
| 450 |
"step": 620
|
| 451 |
},
|
| 452 |
{
|
| 453 |
+
"epoch": 0.9621993127147767,
|
| 454 |
+
"grad_norm": 0.09015417303190752,
|
| 455 |
+
"learning_rate": 4.104993088376974e-08,
|
| 456 |
+
"loss": 0.1212,
|
| 457 |
"step": 630
|
| 458 |
},
|
| 459 |
{
|
| 460 |
+
"epoch": 0.9774723176785033,
|
| 461 |
+
"grad_norm": 0.12899056246523352,
|
| 462 |
+
"learning_rate": 1.3981014094099354e-08,
|
| 463 |
+
"loss": 0.131,
|
| 464 |
"step": 640
|
| 465 |
},
|
| 466 |
{
|
| 467 |
+
"epoch": 0.9927453226422298,
|
| 468 |
+
"grad_norm": 0.08443201140015646,
|
| 469 |
+
"learning_rate": 1.1417960984605459e-09,
|
| 470 |
+
"loss": 0.1302,
|
| 471 |
"step": 650
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 472 |
}
|
| 473 |
],
|
| 474 |
"logging_steps": 10,
|
| 475 |
+
"max_steps": 654,
|
| 476 |
"num_input_tokens_seen": 0,
|
| 477 |
"num_train_epochs": 1,
|
| 478 |
"save_steps": 500,
|
|
|
|
| 488 |
"attributes": {}
|
| 489 |
}
|
| 490 |
},
|
| 491 |
+
"total_flos": 32812144386048.0,
|
| 492 |
"train_batch_size": 1,
|
| 493 |
"trial_name": null,
|
| 494 |
"trial_params": null
|