Upload Step 100/22483
Browse files- adapter_model.safetensors +1 -1
- trainer_state.json +471 -3
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1047628488
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eef2a76f8c9c355a2831ca4dce9743f1165d7eec991513bcb577ef444d71f7a6
|
| 3 |
size 1047628488
|
trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.
|
| 6 |
"eval_steps": 10,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -328,6 +328,474 @@
|
|
| 328 |
"eval_samples_per_second": 0.273,
|
| 329 |
"eval_steps_per_second": 0.068,
|
| 330 |
"step": 40
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 331 |
}
|
| 332 |
],
|
| 333 |
"logging_steps": 1,
|
|
@@ -347,7 +815,7 @@
|
|
| 347 |
"attributes": {}
|
| 348 |
}
|
| 349 |
},
|
| 350 |
-
"total_flos":
|
| 351 |
"train_batch_size": 4,
|
| 352 |
"trial_name": null,
|
| 353 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.00444780500822844,
|
| 6 |
"eval_steps": 10,
|
| 7 |
+
"global_step": 100,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 328 |
"eval_samples_per_second": 0.273,
|
| 329 |
"eval_steps_per_second": 0.068,
|
| 330 |
"step": 40
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 0.0018236000533736602,
|
| 334 |
+
"grad_norm": 0.45847609639167786,
|
| 335 |
+
"learning_rate": 9.998264731957889e-07,
|
| 336 |
+
"loss": 1.8686,
|
| 337 |
+
"step": 41
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 0.0018680781034559446,
|
| 341 |
+
"grad_norm": 0.33932119607925415,
|
| 342 |
+
"learning_rate": 9.998222337069841e-07,
|
| 343 |
+
"loss": 2.2302,
|
| 344 |
+
"step": 42
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.001912556153538229,
|
| 348 |
+
"grad_norm": 1.3315902948379517,
|
| 349 |
+
"learning_rate": 9.998179938781784e-07,
|
| 350 |
+
"loss": 1.8721,
|
| 351 |
+
"step": 43
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 0.001957034203620513,
|
| 355 |
+
"grad_norm": 1.2163060903549194,
|
| 356 |
+
"learning_rate": 9.99813753709331e-07,
|
| 357 |
+
"loss": 2.1642,
|
| 358 |
+
"step": 44
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 0.0020015122537027975,
|
| 362 |
+
"grad_norm": 0.4775317907333374,
|
| 363 |
+
"learning_rate": 9.99809513200401e-07,
|
| 364 |
+
"loss": 2.4106,
|
| 365 |
+
"step": 45
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 0.002045990303785082,
|
| 369 |
+
"grad_norm": 0.28207871317863464,
|
| 370 |
+
"learning_rate": 9.998052723513476e-07,
|
| 371 |
+
"loss": 2.058,
|
| 372 |
+
"step": 46
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 0.0020904683538673662,
|
| 376 |
+
"grad_norm": 0.5639588236808777,
|
| 377 |
+
"learning_rate": 9.998010311621295e-07,
|
| 378 |
+
"loss": 2.2522,
|
| 379 |
+
"step": 47
|
| 380 |
+
},
|
| 381 |
+
{
|
| 382 |
+
"epoch": 0.002134946403949651,
|
| 383 |
+
"grad_norm": 0.45129290223121643,
|
| 384 |
+
"learning_rate": 9.997967896327061e-07,
|
| 385 |
+
"loss": 2.3281,
|
| 386 |
+
"step": 48
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.0021794244540319354,
|
| 390 |
+
"grad_norm": 0.4590243101119995,
|
| 391 |
+
"learning_rate": 9.997925477630364e-07,
|
| 392 |
+
"loss": 2.1991,
|
| 393 |
+
"step": 49
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 0.00222390250411422,
|
| 397 |
+
"grad_norm": 0.3881557881832123,
|
| 398 |
+
"learning_rate": 9.997883055530797e-07,
|
| 399 |
+
"loss": 2.2259,
|
| 400 |
+
"step": 50
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 0.00222390250411422,
|
| 404 |
+
"eval_loss": 1.9868642091751099,
|
| 405 |
+
"eval_runtime": 233.6732,
|
| 406 |
+
"eval_samples_per_second": 0.274,
|
| 407 |
+
"eval_steps_per_second": 0.068,
|
| 408 |
+
"step": 50
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.002268380554196504,
|
| 412 |
+
"grad_norm": 0.4352044463157654,
|
| 413 |
+
"learning_rate": 9.997840630027944e-07,
|
| 414 |
+
"loss": 2.2623,
|
| 415 |
+
"step": 51
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.0023128586042787886,
|
| 419 |
+
"grad_norm": 0.43397316336631775,
|
| 420 |
+
"learning_rate": 9.997798201121402e-07,
|
| 421 |
+
"loss": 1.8368,
|
| 422 |
+
"step": 52
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.002357336654361073,
|
| 426 |
+
"grad_norm": 0.38807961344718933,
|
| 427 |
+
"learning_rate": 9.99775576881076e-07,
|
| 428 |
+
"loss": 2.0101,
|
| 429 |
+
"step": 53
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.0024018147044433573,
|
| 433 |
+
"grad_norm": 0.6237647533416748,
|
| 434 |
+
"learning_rate": 9.997713333095603e-07,
|
| 435 |
+
"loss": 2.0238,
|
| 436 |
+
"step": 54
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.0024462927545256417,
|
| 440 |
+
"grad_norm": 0.34039369225502014,
|
| 441 |
+
"learning_rate": 9.997670893975529e-07,
|
| 442 |
+
"loss": 2.1,
|
| 443 |
+
"step": 55
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.002490770804607926,
|
| 447 |
+
"grad_norm": 0.4221118986606598,
|
| 448 |
+
"learning_rate": 9.997628451450122e-07,
|
| 449 |
+
"loss": 2.1649,
|
| 450 |
+
"step": 56
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.0025352488546902105,
|
| 454 |
+
"grad_norm": 0.38328638672828674,
|
| 455 |
+
"learning_rate": 9.997586005518976e-07,
|
| 456 |
+
"loss": 2.1151,
|
| 457 |
+
"step": 57
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.002579726904772495,
|
| 461 |
+
"grad_norm": 0.3308090567588806,
|
| 462 |
+
"learning_rate": 9.997543556181679e-07,
|
| 463 |
+
"loss": 2.1499,
|
| 464 |
+
"step": 58
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.0026242049548547792,
|
| 468 |
+
"grad_norm": 0.3598516881465912,
|
| 469 |
+
"learning_rate": 9.99750110343782e-07,
|
| 470 |
+
"loss": 2.1146,
|
| 471 |
+
"step": 59
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.0026686830049370636,
|
| 475 |
+
"grad_norm": 0.38582974672317505,
|
| 476 |
+
"learning_rate": 9.997458647286993e-07,
|
| 477 |
+
"loss": 1.8236,
|
| 478 |
+
"step": 60
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.0026686830049370636,
|
| 482 |
+
"eval_loss": 1.9794503450393677,
|
| 483 |
+
"eval_runtime": 312.5608,
|
| 484 |
+
"eval_samples_per_second": 0.205,
|
| 485 |
+
"eval_steps_per_second": 0.051,
|
| 486 |
+
"step": 60
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.002713161055019348,
|
| 490 |
+
"grad_norm": 0.5031485557556152,
|
| 491 |
+
"learning_rate": 9.997416187728787e-07,
|
| 492 |
+
"loss": 2.3825,
|
| 493 |
+
"step": 61
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.0027576391051016324,
|
| 497 |
+
"grad_norm": 0.40436115860939026,
|
| 498 |
+
"learning_rate": 9.997373724762788e-07,
|
| 499 |
+
"loss": 1.9051,
|
| 500 |
+
"step": 62
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.0028021171551839167,
|
| 504 |
+
"grad_norm": 0.3216610252857208,
|
| 505 |
+
"learning_rate": 9.997331258388588e-07,
|
| 506 |
+
"loss": 1.7655,
|
| 507 |
+
"step": 63
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.002846595205266201,
|
| 511 |
+
"grad_norm": 0.30226317048072815,
|
| 512 |
+
"learning_rate": 9.997288788605777e-07,
|
| 513 |
+
"loss": 2.215,
|
| 514 |
+
"step": 64
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.0028910732553484855,
|
| 518 |
+
"grad_norm": 0.34857413172721863,
|
| 519 |
+
"learning_rate": 9.997246315413945e-07,
|
| 520 |
+
"loss": 2.1704,
|
| 521 |
+
"step": 65
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.00293555130543077,
|
| 525 |
+
"grad_norm": 0.4939625859260559,
|
| 526 |
+
"learning_rate": 9.99720383881268e-07,
|
| 527 |
+
"loss": 2.1591,
|
| 528 |
+
"step": 66
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.0029800293555130542,
|
| 532 |
+
"grad_norm": 0.6396478414535522,
|
| 533 |
+
"learning_rate": 9.997161358801571e-07,
|
| 534 |
+
"loss": 2.4183,
|
| 535 |
+
"step": 67
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.0030245074055953386,
|
| 539 |
+
"grad_norm": 0.3547438681125641,
|
| 540 |
+
"learning_rate": 9.99711887538021e-07,
|
| 541 |
+
"loss": 2.3338,
|
| 542 |
+
"step": 68
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.003068985455677623,
|
| 546 |
+
"grad_norm": 0.455522745847702,
|
| 547 |
+
"learning_rate": 9.997076388548186e-07,
|
| 548 |
+
"loss": 2.2559,
|
| 549 |
+
"step": 69
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.0031134635057599074,
|
| 553 |
+
"grad_norm": 0.5139729976654053,
|
| 554 |
+
"learning_rate": 9.997033898305084e-07,
|
| 555 |
+
"loss": 2.4271,
|
| 556 |
+
"step": 70
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.0031134635057599074,
|
| 560 |
+
"eval_loss": 1.974100947380066,
|
| 561 |
+
"eval_runtime": 231.1208,
|
| 562 |
+
"eval_samples_per_second": 0.277,
|
| 563 |
+
"eval_steps_per_second": 0.069,
|
| 564 |
+
"step": 70
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 0.0031579415558421918,
|
| 568 |
+
"grad_norm": 0.3868389427661896,
|
| 569 |
+
"learning_rate": 9.996991404650499e-07,
|
| 570 |
+
"loss": 1.8754,
|
| 571 |
+
"step": 71
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 0.003202419605924476,
|
| 575 |
+
"grad_norm": 0.4664241373538971,
|
| 576 |
+
"learning_rate": 9.996948907584016e-07,
|
| 577 |
+
"loss": 1.9934,
|
| 578 |
+
"step": 72
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.0032468976560067605,
|
| 582 |
+
"grad_norm": 0.3952767848968506,
|
| 583 |
+
"learning_rate": 9.996906407105226e-07,
|
| 584 |
+
"loss": 2.0883,
|
| 585 |
+
"step": 73
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.003291375706089045,
|
| 589 |
+
"grad_norm": 0.4785691797733307,
|
| 590 |
+
"learning_rate": 9.996863903213718e-07,
|
| 591 |
+
"loss": 2.3203,
|
| 592 |
+
"step": 74
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 0.0033358537561713293,
|
| 596 |
+
"grad_norm": 0.4103385806083679,
|
| 597 |
+
"learning_rate": 9.99682139590908e-07,
|
| 598 |
+
"loss": 2.0406,
|
| 599 |
+
"step": 75
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 0.0033803318062536137,
|
| 603 |
+
"grad_norm": 0.45813262462615967,
|
| 604 |
+
"learning_rate": 9.996778885190904e-07,
|
| 605 |
+
"loss": 2.1745,
|
| 606 |
+
"step": 76
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 0.003424809856335898,
|
| 610 |
+
"grad_norm": 0.34197866916656494,
|
| 611 |
+
"learning_rate": 9.996736371058771e-07,
|
| 612 |
+
"loss": 2.0839,
|
| 613 |
+
"step": 77
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 0.0034692879064181824,
|
| 617 |
+
"grad_norm": 0.4917107820510864,
|
| 618 |
+
"learning_rate": 9.996693853512279e-07,
|
| 619 |
+
"loss": 1.9646,
|
| 620 |
+
"step": 78
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 0.0035137659565004672,
|
| 624 |
+
"grad_norm": 0.570755124092102,
|
| 625 |
+
"learning_rate": 9.99665133255101e-07,
|
| 626 |
+
"loss": 2.4128,
|
| 627 |
+
"step": 79
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 0.0035582440065827516,
|
| 631 |
+
"grad_norm": 0.6334550380706787,
|
| 632 |
+
"learning_rate": 9.996608808174557e-07,
|
| 633 |
+
"loss": 2.3972,
|
| 634 |
+
"step": 80
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 0.0035582440065827516,
|
| 638 |
+
"eval_loss": 1.9703996181488037,
|
| 639 |
+
"eval_runtime": 236.153,
|
| 640 |
+
"eval_samples_per_second": 0.271,
|
| 641 |
+
"eval_steps_per_second": 0.068,
|
| 642 |
+
"step": 80
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 0.003602722056665036,
|
| 646 |
+
"grad_norm": 0.44049328565597534,
|
| 647 |
+
"learning_rate": 9.996566280382507e-07,
|
| 648 |
+
"loss": 2.389,
|
| 649 |
+
"step": 81
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 0.0036472001067473204,
|
| 653 |
+
"grad_norm": 0.5198694467544556,
|
| 654 |
+
"learning_rate": 9.996523749174444e-07,
|
| 655 |
+
"loss": 1.8092,
|
| 656 |
+
"step": 82
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 0.0036916781568296047,
|
| 660 |
+
"grad_norm": 0.4297351837158203,
|
| 661 |
+
"learning_rate": 9.996481214549966e-07,
|
| 662 |
+
"loss": 1.9158,
|
| 663 |
+
"step": 83
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 0.003736156206911889,
|
| 667 |
+
"grad_norm": 0.5207564234733582,
|
| 668 |
+
"learning_rate": 9.996438676508653e-07,
|
| 669 |
+
"loss": 2.3368,
|
| 670 |
+
"step": 84
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 0.0037806342569941735,
|
| 674 |
+
"grad_norm": 3.4639275074005127,
|
| 675 |
+
"learning_rate": 9.996396135050097e-07,
|
| 676 |
+
"loss": 2.0157,
|
| 677 |
+
"step": 85
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 0.003825112307076458,
|
| 681 |
+
"grad_norm": 0.5132240056991577,
|
| 682 |
+
"learning_rate": 9.996353590173885e-07,
|
| 683 |
+
"loss": 2.1738,
|
| 684 |
+
"step": 86
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 0.0038695903571587423,
|
| 688 |
+
"grad_norm": 0.559355616569519,
|
| 689 |
+
"learning_rate": 9.996311041879605e-07,
|
| 690 |
+
"loss": 2.2668,
|
| 691 |
+
"step": 87
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 0.003914068407241026,
|
| 695 |
+
"grad_norm": 0.40078872442245483,
|
| 696 |
+
"learning_rate": 9.996268490166847e-07,
|
| 697 |
+
"loss": 2.0339,
|
| 698 |
+
"step": 88
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 0.003958546457323311,
|
| 702 |
+
"grad_norm": 0.43362948298454285,
|
| 703 |
+
"learning_rate": 9.996225935035196e-07,
|
| 704 |
+
"loss": 2.3476,
|
| 705 |
+
"step": 89
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 0.004003024507405595,
|
| 709 |
+
"grad_norm": 0.6087605953216553,
|
| 710 |
+
"learning_rate": 9.99618337648424e-07,
|
| 711 |
+
"loss": 2.3268,
|
| 712 |
+
"step": 90
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 0.004003024507405595,
|
| 716 |
+
"eval_loss": 1.9674957990646362,
|
| 717 |
+
"eval_runtime": 241.976,
|
| 718 |
+
"eval_samples_per_second": 0.264,
|
| 719 |
+
"eval_steps_per_second": 0.066,
|
| 720 |
+
"step": 90
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 0.004047502557487879,
|
| 724 |
+
"grad_norm": 3.6922800540924072,
|
| 725 |
+
"learning_rate": 9.996140814513573e-07,
|
| 726 |
+
"loss": 2.2244,
|
| 727 |
+
"step": 91
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 0.004091980607570164,
|
| 731 |
+
"grad_norm": 0.6901090145111084,
|
| 732 |
+
"learning_rate": 9.996098249122776e-07,
|
| 733 |
+
"loss": 2.127,
|
| 734 |
+
"step": 92
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 0.004136458657652448,
|
| 738 |
+
"grad_norm": 0.42567893862724304,
|
| 739 |
+
"learning_rate": 9.99605568031144e-07,
|
| 740 |
+
"loss": 1.99,
|
| 741 |
+
"step": 93
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 0.0041809367077347325,
|
| 745 |
+
"grad_norm": 0.37529805302619934,
|
| 746 |
+
"learning_rate": 9.996013108079149e-07,
|
| 747 |
+
"loss": 2.1369,
|
| 748 |
+
"step": 94
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 0.004225414757817018,
|
| 752 |
+
"grad_norm": 0.5616635084152222,
|
| 753 |
+
"learning_rate": 9.995970532425493e-07,
|
| 754 |
+
"loss": 1.8421,
|
| 755 |
+
"step": 95
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 0.004269892807899302,
|
| 759 |
+
"grad_norm": 0.3917858898639679,
|
| 760 |
+
"learning_rate": 9.995927953350061e-07,
|
| 761 |
+
"loss": 2.0905,
|
| 762 |
+
"step": 96
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 0.0043143708579815865,
|
| 766 |
+
"grad_norm": 0.3693113923072815,
|
| 767 |
+
"learning_rate": 9.99588537085244e-07,
|
| 768 |
+
"loss": 2.326,
|
| 769 |
+
"step": 97
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 0.004358848908063871,
|
| 773 |
+
"grad_norm": 0.4595705270767212,
|
| 774 |
+
"learning_rate": 9.995842784932216e-07,
|
| 775 |
+
"loss": 1.8433,
|
| 776 |
+
"step": 98
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 0.004403326958146155,
|
| 780 |
+
"grad_norm": 0.46681633591651917,
|
| 781 |
+
"learning_rate": 9.995800195588977e-07,
|
| 782 |
+
"loss": 2.21,
|
| 783 |
+
"step": 99
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 0.00444780500822844,
|
| 787 |
+
"grad_norm": 0.5430231690406799,
|
| 788 |
+
"learning_rate": 9.99575760282231e-07,
|
| 789 |
+
"loss": 2.4849,
|
| 790 |
+
"step": 100
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 0.00444780500822844,
|
| 794 |
+
"eval_loss": 1.964964509010315,
|
| 795 |
+
"eval_runtime": 231.4209,
|
| 796 |
+
"eval_samples_per_second": 0.277,
|
| 797 |
+
"eval_steps_per_second": 0.069,
|
| 798 |
+
"step": 100
|
| 799 |
}
|
| 800 |
],
|
| 801 |
"logging_steps": 1,
|
|
|
|
| 815 |
"attributes": {}
|
| 816 |
}
|
| 817 |
},
|
| 818 |
+
"total_flos": 2.248725545091072e+17,
|
| 819 |
"train_batch_size": 4,
|
| 820 |
"trial_name": null,
|
| 821 |
"trial_params": null
|