Training in progress, step 30000
Browse files- last-checkpoint/optimizer.pt +1 -1
- last-checkpoint/pytorch_model.bin +1 -1
- last-checkpoint/rng_state_0.pth +1 -1
- last-checkpoint/rng_state_1.pth +1 -1
- last-checkpoint/rng_state_2.pth +1 -1
- last-checkpoint/rng_state_3.pth +1 -1
- last-checkpoint/rng_state_4.pth +1 -1
- last-checkpoint/rng_state_5.pth +1 -1
- last-checkpoint/rng_state_6.pth +1 -1
- last-checkpoint/rng_state_7.pth +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +203 -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 50044241
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9a92335164a57171365050fd07a22cee199d87b5d4f31f9f266d934997806ef1
|
| 3 |
size 50044241
|
last-checkpoint/pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 25761253
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d17ced65414e5fa0fdf019e711ae42967ebbbc1772201d6c2fa0f444b22af2ac
|
| 3 |
size 25761253
|
last-checkpoint/rng_state_0.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14503
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0b1c87c4f861c3b9c96d42cb3e1fac9d804a68e14a3351872cbc3092e0046c79
|
| 3 |
size 14503
|
last-checkpoint/rng_state_1.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14503
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0b1c87c4f861c3b9c96d42cb3e1fac9d804a68e14a3351872cbc3092e0046c79
|
| 3 |
size 14503
|
last-checkpoint/rng_state_2.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14503
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0b1c87c4f861c3b9c96d42cb3e1fac9d804a68e14a3351872cbc3092e0046c79
|
| 3 |
size 14503
|
last-checkpoint/rng_state_3.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14503
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0b1c87c4f861c3b9c96d42cb3e1fac9d804a68e14a3351872cbc3092e0046c79
|
| 3 |
size 14503
|
last-checkpoint/rng_state_4.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14503
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0b1c87c4f861c3b9c96d42cb3e1fac9d804a68e14a3351872cbc3092e0046c79
|
| 3 |
size 14503
|
last-checkpoint/rng_state_5.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14503
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0b1c87c4f861c3b9c96d42cb3e1fac9d804a68e14a3351872cbc3092e0046c79
|
| 3 |
size 14503
|
last-checkpoint/rng_state_6.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14503
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0b1c87c4f861c3b9c96d42cb3e1fac9d804a68e14a3351872cbc3092e0046c79
|
| 3 |
size 14503
|
last-checkpoint/rng_state_7.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14503
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0b1c87c4f861c3b9c96d42cb3e1fac9d804a68e14a3351872cbc3092e0046c79
|
| 3 |
size 14503
|
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 sha256:285c70cc0b99f831fd558ed5749dae6f2d660f449e704cb70ab7d67faaef2748
|
| 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":
|
| 5 |
-
"global_step":
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
@@ -406,11 +406,211 @@
|
|
| 406 |
"eval_samples_per_second": 1070.586,
|
| 407 |
"eval_steps_per_second": 16.779,
|
| 408 |
"step": 20000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 409 |
}
|
| 410 |
],
|
| 411 |
"max_steps": 250000,
|
| 412 |
"num_train_epochs": 12,
|
| 413 |
-
"total_flos":
|
| 414 |
"trial_name": null,
|
| 415 |
"trial_params": null
|
| 416 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.3383297644539613,
|
| 5 |
+
"global_step": 30000,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
|
|
| 406 |
"eval_samples_per_second": 1070.586,
|
| 407 |
"eval_steps_per_second": 16.779,
|
| 408 |
"step": 20000
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.91,
|
| 412 |
+
"learning_rate": 0.000598349790079647,
|
| 413 |
+
"loss": 0.6773,
|
| 414 |
+
"step": 20500
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 0.94,
|
| 418 |
+
"learning_rate": 0.000598137291838376,
|
| 419 |
+
"loss": 0.6772,
|
| 420 |
+
"step": 21000
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 0.94,
|
| 424 |
+
"eval_loss": 0.6761835217475891,
|
| 425 |
+
"eval_runtime": 2.1367,
|
| 426 |
+
"eval_samples_per_second": 1075.027,
|
| 427 |
+
"eval_steps_per_second": 16.848,
|
| 428 |
+
"step": 21000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.96,
|
| 432 |
+
"learning_rate": 0.000597911970822327,
|
| 433 |
+
"loss": 0.6772,
|
| 434 |
+
"step": 21500
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.98,
|
| 438 |
+
"learning_rate": 0.000597673836887771,
|
| 439 |
+
"loss": 0.6773,
|
| 440 |
+
"step": 22000
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.98,
|
| 444 |
+
"eval_loss": 0.6774830222129822,
|
| 445 |
+
"eval_runtime": 2.1107,
|
| 446 |
+
"eval_samples_per_second": 1088.282,
|
| 447 |
+
"eval_steps_per_second": 17.056,
|
| 448 |
+
"step": 22000
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 1.0,
|
| 452 |
+
"learning_rate": 0.0005974229004514577,
|
| 453 |
+
"loss": 0.6773,
|
| 454 |
+
"step": 22500
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 1.03,
|
| 458 |
+
"learning_rate": 0.0005971591724901598,
|
| 459 |
+
"loss": 0.6773,
|
| 460 |
+
"step": 23000
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 1.03,
|
| 464 |
+
"eval_loss": 0.6764113903045654,
|
| 465 |
+
"eval_runtime": 2.206,
|
| 466 |
+
"eval_samples_per_second": 1041.236,
|
| 467 |
+
"eval_steps_per_second": 16.319,
|
| 468 |
+
"step": 23000
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 1.05,
|
| 472 |
+
"learning_rate": 0.0005968826645401927,
|
| 473 |
+
"loss": 0.6772,
|
| 474 |
+
"step": 23500
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 1.07,
|
| 478 |
+
"learning_rate": 0.00059659338869691,
|
| 479 |
+
"loss": 0.6772,
|
| 480 |
+
"step": 24000
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 1.07,
|
| 484 |
+
"eval_loss": 0.6768463253974915,
|
| 485 |
+
"eval_runtime": 2.2173,
|
| 486 |
+
"eval_samples_per_second": 1035.939,
|
| 487 |
+
"eval_steps_per_second": 16.236,
|
| 488 |
+
"step": 24000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 1.09,
|
| 492 |
+
"learning_rate": 0.0005962913576141742,
|
| 493 |
+
"loss": 0.6773,
|
| 494 |
+
"step": 24500
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.12,
|
| 498 |
+
"learning_rate": 0.000595976584503803,
|
| 499 |
+
"loss": 0.6772,
|
| 500 |
+
"step": 25000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.12,
|
| 504 |
+
"eval_loss": 0.6769193410873413,
|
| 505 |
+
"eval_runtime": 2.2077,
|
| 506 |
+
"eval_samples_per_second": 1040.465,
|
| 507 |
+
"eval_steps_per_second": 16.307,
|
| 508 |
+
"step": 25000
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 1.14,
|
| 512 |
+
"learning_rate": 0.0005956490831349923,
|
| 513 |
+
"loss": 0.6773,
|
| 514 |
+
"step": 25500
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.16,
|
| 518 |
+
"learning_rate": 0.0005953088678337129,
|
| 519 |
+
"loss": 0.6772,
|
| 520 |
+
"step": 26000
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.16,
|
| 524 |
+
"eval_loss": 0.6775402426719666,
|
| 525 |
+
"eval_runtime": 2.2094,
|
| 526 |
+
"eval_samples_per_second": 1039.671,
|
| 527 |
+
"eval_steps_per_second": 16.294,
|
| 528 |
+
"step": 26000
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.18,
|
| 532 |
+
"learning_rate": 0.0005949559534820841,
|
| 533 |
+
"loss": 0.6772,
|
| 534 |
+
"step": 26500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 1.2,
|
| 538 |
+
"learning_rate": 0.0005945903555177229,
|
| 539 |
+
"loss": 0.6772,
|
| 540 |
+
"step": 27000
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 1.2,
|
| 544 |
+
"eval_loss": 0.6775718331336975,
|
| 545 |
+
"eval_runtime": 2.1372,
|
| 546 |
+
"eval_samples_per_second": 1074.76,
|
| 547 |
+
"eval_steps_per_second": 16.844,
|
| 548 |
+
"step": 27000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.23,
|
| 552 |
+
"learning_rate": 0.0005942120899330687,
|
| 553 |
+
"loss": 0.6772,
|
| 554 |
+
"step": 27500
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 1.25,
|
| 558 |
+
"learning_rate": 0.0005938211732746836,
|
| 559 |
+
"loss": 0.6772,
|
| 560 |
+
"step": 28000
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 1.25,
|
| 564 |
+
"eval_loss": 0.6771659255027771,
|
| 565 |
+
"eval_runtime": 2.191,
|
| 566 |
+
"eval_samples_per_second": 1048.378,
|
| 567 |
+
"eval_steps_per_second": 16.431,
|
| 568 |
+
"step": 28000
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 1.27,
|
| 572 |
+
"learning_rate": 0.0005934176226425286,
|
| 573 |
+
"loss": 0.6772,
|
| 574 |
+
"step": 28500
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 1.29,
|
| 578 |
+
"learning_rate": 0.0005930014556892158,
|
| 579 |
+
"loss": 0.6772,
|
| 580 |
+
"step": 29000
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 1.29,
|
| 584 |
+
"eval_loss": 0.6768732666969299,
|
| 585 |
+
"eval_runtime": 2.2912,
|
| 586 |
+
"eval_samples_per_second": 1002.553,
|
| 587 |
+
"eval_steps_per_second": 15.713,
|
| 588 |
+
"step": 29000
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 1.32,
|
| 592 |
+
"learning_rate": 0.0005925726906192357,
|
| 593 |
+
"loss": 0.6772,
|
| 594 |
+
"step": 29500
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 1.34,
|
| 598 |
+
"learning_rate": 0.0005921313461881617,
|
| 599 |
+
"loss": 0.6773,
|
| 600 |
+
"step": 30000
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 1.34,
|
| 604 |
+
"eval_loss": 0.6772350072860718,
|
| 605 |
+
"eval_runtime": 2.2465,
|
| 606 |
+
"eval_samples_per_second": 1022.498,
|
| 607 |
+
"eval_steps_per_second": 16.025,
|
| 608 |
+
"step": 30000
|
| 609 |
}
|
| 610 |
],
|
| 611 |
"max_steps": 250000,
|
| 612 |
"num_train_epochs": 12,
|
| 613 |
+
"total_flos": 4.804968766152387e+20,
|
| 614 |
"trial_name": null,
|
| 615 |
"trial_params": null
|
| 616 |
}
|
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 25761253
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d17ced65414e5fa0fdf019e711ae42967ebbbc1772201d6c2fa0f444b22af2ac
|
| 3 |
size 25761253
|