Commit
·
4f68ff3
1
Parent(s):
7b1f997
Training in progress, step 12800
Browse files- last-checkpoint/optimizer.pt +2 -2
- last-checkpoint/pytorch_model.bin +1 -1
- last-checkpoint/rng_state.pth +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +396 -3
- pytorch_model.bin +1 -1
last-checkpoint/optimizer.pt
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:b3f77e87a0cf433d10838fb68e0ab0a000fe769bd6530dd014fc549fab4e8769
|
| 3 |
+
size 745634697
|
last-checkpoint/pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 372832803
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e369ec99ab61a02ee5a324cbae60bfddfa49210663a3dc92314e7fd8027b7e1c
|
| 3 |
size 372832803
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15523
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:545fcebd225c2fbcaaae084db32b315ff159bcb9f66f876ced049afa99cb2632
|
| 3 |
size 15523
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 623
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8a63c18679f872f561021a84d9bfcd3fad0c807bcef87d1a807b9818f9895c1f
|
| 3 |
size 623
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
-
"global_step":
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
@@ -399,11 +399,404 @@
|
|
| 399 |
"eval_samples_per_second": 22.53,
|
| 400 |
"eval_steps_per_second": 1.408,
|
| 401 |
"step": 6400
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 402 |
}
|
| 403 |
],
|
| 404 |
"max_steps": 64000,
|
| 405 |
"num_train_epochs": 9223372036854775807,
|
| 406 |
-
"total_flos":
|
| 407 |
"trial_name": null,
|
| 408 |
"trial_params": null
|
| 409 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.2,
|
| 5 |
+
"global_step": 12800,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
|
|
| 399 |
"eval_samples_per_second": 22.53,
|
| 400 |
"eval_steps_per_second": 1.408,
|
| 401 |
"step": 6400
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.1,
|
| 405 |
+
"learning_rate": 0.001,
|
| 406 |
+
"loss": 8.3018,
|
| 407 |
+
"step": 6500
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 0.1,
|
| 411 |
+
"learning_rate": 0.001,
|
| 412 |
+
"loss": 8.0897,
|
| 413 |
+
"step": 6600
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 0.1,
|
| 417 |
+
"learning_rate": 0.001,
|
| 418 |
+
"loss": 8.0611,
|
| 419 |
+
"step": 6700
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 0.11,
|
| 423 |
+
"learning_rate": 0.001,
|
| 424 |
+
"loss": 8.0538,
|
| 425 |
+
"step": 6800
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 0.11,
|
| 429 |
+
"learning_rate": 0.001,
|
| 430 |
+
"loss": 8.0351,
|
| 431 |
+
"step": 6900
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 0.11,
|
| 435 |
+
"learning_rate": 0.001,
|
| 436 |
+
"loss": 8.0364,
|
| 437 |
+
"step": 7000
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.11,
|
| 441 |
+
"learning_rate": 0.001,
|
| 442 |
+
"loss": 8.0244,
|
| 443 |
+
"step": 7100
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.11,
|
| 447 |
+
"learning_rate": 0.001,
|
| 448 |
+
"loss": 8.0338,
|
| 449 |
+
"step": 7200
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 0.11,
|
| 453 |
+
"learning_rate": 0.001,
|
| 454 |
+
"loss": 8.0224,
|
| 455 |
+
"step": 7300
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 0.12,
|
| 459 |
+
"learning_rate": 0.001,
|
| 460 |
+
"loss": 8.0215,
|
| 461 |
+
"step": 7400
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 0.12,
|
| 465 |
+
"learning_rate": 0.001,
|
| 466 |
+
"loss": 8.0214,
|
| 467 |
+
"step": 7500
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 0.12,
|
| 471 |
+
"learning_rate": 0.001,
|
| 472 |
+
"loss": 8.0183,
|
| 473 |
+
"step": 7600
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.12,
|
| 477 |
+
"learning_rate": 0.001,
|
| 478 |
+
"loss": 8.0246,
|
| 479 |
+
"step": 7700
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.12,
|
| 483 |
+
"learning_rate": 0.001,
|
| 484 |
+
"loss": 8.0169,
|
| 485 |
+
"step": 7800
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.12,
|
| 489 |
+
"learning_rate": 0.001,
|
| 490 |
+
"loss": 8.0206,
|
| 491 |
+
"step": 7900
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 0.12,
|
| 495 |
+
"learning_rate": 0.001,
|
| 496 |
+
"loss": 8.012,
|
| 497 |
+
"step": 8000
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 0.13,
|
| 501 |
+
"learning_rate": 0.001,
|
| 502 |
+
"loss": 8.0098,
|
| 503 |
+
"step": 8100
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 0.13,
|
| 507 |
+
"learning_rate": 0.001,
|
| 508 |
+
"loss": 8.0122,
|
| 509 |
+
"step": 8200
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 0.13,
|
| 513 |
+
"learning_rate": 0.001,
|
| 514 |
+
"loss": 7.997,
|
| 515 |
+
"step": 8300
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.13,
|
| 519 |
+
"learning_rate": 0.001,
|
| 520 |
+
"loss": 7.9958,
|
| 521 |
+
"step": 8400
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.13,
|
| 525 |
+
"learning_rate": 0.001,
|
| 526 |
+
"loss": 8.0232,
|
| 527 |
+
"step": 8500
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.13,
|
| 531 |
+
"learning_rate": 0.001,
|
| 532 |
+
"loss": 7.9921,
|
| 533 |
+
"step": 8600
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 0.14,
|
| 537 |
+
"learning_rate": 0.001,
|
| 538 |
+
"loss": 7.9862,
|
| 539 |
+
"step": 8700
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 0.14,
|
| 543 |
+
"learning_rate": 0.001,
|
| 544 |
+
"loss": 8.0073,
|
| 545 |
+
"step": 8800
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 0.14,
|
| 549 |
+
"learning_rate": 0.001,
|
| 550 |
+
"loss": 7.9879,
|
| 551 |
+
"step": 8900
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 0.14,
|
| 555 |
+
"learning_rate": 0.001,
|
| 556 |
+
"loss": 8.0064,
|
| 557 |
+
"step": 9000
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 0.14,
|
| 561 |
+
"learning_rate": 0.001,
|
| 562 |
+
"loss": 8.0153,
|
| 563 |
+
"step": 9100
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.14,
|
| 567 |
+
"learning_rate": 0.001,
|
| 568 |
+
"loss": 8.0111,
|
| 569 |
+
"step": 9200
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.15,
|
| 573 |
+
"learning_rate": 0.001,
|
| 574 |
+
"loss": 7.986,
|
| 575 |
+
"step": 9300
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 0.15,
|
| 579 |
+
"learning_rate": 0.001,
|
| 580 |
+
"loss": 8.0042,
|
| 581 |
+
"step": 9400
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 0.15,
|
| 585 |
+
"learning_rate": 0.001,
|
| 586 |
+
"loss": 8.0028,
|
| 587 |
+
"step": 9500
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 0.15,
|
| 591 |
+
"learning_rate": 0.001,
|
| 592 |
+
"loss": 7.9986,
|
| 593 |
+
"step": 9600
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 0.15,
|
| 597 |
+
"learning_rate": 0.001,
|
| 598 |
+
"loss": 8.0031,
|
| 599 |
+
"step": 9700
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 0.15,
|
| 603 |
+
"learning_rate": 0.001,
|
| 604 |
+
"loss": 8.0015,
|
| 605 |
+
"step": 9800
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.15,
|
| 609 |
+
"learning_rate": 0.001,
|
| 610 |
+
"loss": 8.015,
|
| 611 |
+
"step": 9900
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.16,
|
| 615 |
+
"learning_rate": 0.001,
|
| 616 |
+
"loss": 8.0086,
|
| 617 |
+
"step": 10000
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 0.16,
|
| 621 |
+
"learning_rate": 0.001,
|
| 622 |
+
"loss": 7.9976,
|
| 623 |
+
"step": 10100
|
| 624 |
+
},
|
| 625 |
+
{
|
| 626 |
+
"epoch": 0.16,
|
| 627 |
+
"learning_rate": 0.001,
|
| 628 |
+
"loss": 8.0013,
|
| 629 |
+
"step": 10200
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 0.16,
|
| 633 |
+
"learning_rate": 0.001,
|
| 634 |
+
"loss": 8.0002,
|
| 635 |
+
"step": 10300
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 0.16,
|
| 639 |
+
"learning_rate": 0.001,
|
| 640 |
+
"loss": 8.0142,
|
| 641 |
+
"step": 10400
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 0.16,
|
| 645 |
+
"learning_rate": 0.001,
|
| 646 |
+
"loss": 8.0136,
|
| 647 |
+
"step": 10500
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.17,
|
| 651 |
+
"learning_rate": 0.001,
|
| 652 |
+
"loss": 7.9931,
|
| 653 |
+
"step": 10600
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.17,
|
| 657 |
+
"learning_rate": 0.001,
|
| 658 |
+
"loss": 8.0043,
|
| 659 |
+
"step": 10700
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 0.17,
|
| 663 |
+
"learning_rate": 0.001,
|
| 664 |
+
"loss": 8.0092,
|
| 665 |
+
"step": 10800
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 0.17,
|
| 669 |
+
"learning_rate": 0.001,
|
| 670 |
+
"loss": 7.9803,
|
| 671 |
+
"step": 10900
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 0.17,
|
| 675 |
+
"learning_rate": 0.001,
|
| 676 |
+
"loss": 8.011,
|
| 677 |
+
"step": 11000
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 0.17,
|
| 681 |
+
"learning_rate": 0.001,
|
| 682 |
+
"loss": 7.9888,
|
| 683 |
+
"step": 11100
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 0.17,
|
| 687 |
+
"learning_rate": 0.001,
|
| 688 |
+
"loss": 8.0009,
|
| 689 |
+
"step": 11200
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.18,
|
| 693 |
+
"learning_rate": 0.001,
|
| 694 |
+
"loss": 8.0118,
|
| 695 |
+
"step": 11300
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.18,
|
| 699 |
+
"learning_rate": 0.001,
|
| 700 |
+
"loss": 7.9928,
|
| 701 |
+
"step": 11400
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 0.18,
|
| 705 |
+
"learning_rate": 0.001,
|
| 706 |
+
"loss": 8.0043,
|
| 707 |
+
"step": 11500
|
| 708 |
+
},
|
| 709 |
+
{
|
| 710 |
+
"epoch": 0.18,
|
| 711 |
+
"learning_rate": 0.001,
|
| 712 |
+
"loss": 7.9995,
|
| 713 |
+
"step": 11600
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.18,
|
| 717 |
+
"learning_rate": 0.001,
|
| 718 |
+
"loss": 8.0006,
|
| 719 |
+
"step": 11700
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 0.18,
|
| 723 |
+
"learning_rate": 0.001,
|
| 724 |
+
"loss": 8.0018,
|
| 725 |
+
"step": 11800
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 0.19,
|
| 729 |
+
"learning_rate": 0.001,
|
| 730 |
+
"loss": 7.9999,
|
| 731 |
+
"step": 11900
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.19,
|
| 735 |
+
"learning_rate": 0.001,
|
| 736 |
+
"loss": 8.0084,
|
| 737 |
+
"step": 12000
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.19,
|
| 741 |
+
"learning_rate": 0.001,
|
| 742 |
+
"loss": 7.9975,
|
| 743 |
+
"step": 12100
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 0.19,
|
| 747 |
+
"learning_rate": 0.001,
|
| 748 |
+
"loss": 7.9834,
|
| 749 |
+
"step": 12200
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 0.19,
|
| 753 |
+
"learning_rate": 0.001,
|
| 754 |
+
"loss": 8.0164,
|
| 755 |
+
"step": 12300
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 0.19,
|
| 759 |
+
"learning_rate": 0.001,
|
| 760 |
+
"loss": 8.0209,
|
| 761 |
+
"step": 12400
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 0.2,
|
| 765 |
+
"learning_rate": 0.001,
|
| 766 |
+
"loss": 8.0044,
|
| 767 |
+
"step": 12500
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 0.2,
|
| 771 |
+
"learning_rate": 0.001,
|
| 772 |
+
"loss": 8.0032,
|
| 773 |
+
"step": 12600
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.2,
|
| 777 |
+
"learning_rate": 0.001,
|
| 778 |
+
"loss": 8.0112,
|
| 779 |
+
"step": 12700
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.2,
|
| 783 |
+
"learning_rate": 0.001,
|
| 784 |
+
"loss": 8.0055,
|
| 785 |
+
"step": 12800
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 0.2,
|
| 789 |
+
"eval_accuracy": 0.034224256973106555,
|
| 790 |
+
"eval_loss": 7.999759674072266,
|
| 791 |
+
"eval_runtime": 14321.7563,
|
| 792 |
+
"eval_samples_per_second": 22.892,
|
| 793 |
+
"eval_steps_per_second": 1.431,
|
| 794 |
+
"step": 12800
|
| 795 |
}
|
| 796 |
],
|
| 797 |
"max_steps": 64000,
|
| 798 |
"num_train_epochs": 9223372036854775807,
|
| 799 |
+
"total_flos": 1.353967057502208e+17,
|
| 800 |
"trial_name": null,
|
| 801 |
"trial_params": null
|
| 802 |
}
|
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 372832803
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e369ec99ab61a02ee5a324cbae60bfddfa49210663a3dc92314e7fd8027b7e1c
|
| 3 |
size 372832803
|