Training in progress, step 40000, checkpoint
Browse files
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:214a1072eb6e2e4cc13be55f0e76f496dbf8421eeae52bb353439962f9793e2b
|
| 3 |
size 14645
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch":
|
| 6 |
"eval_steps": 10000,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -452,6 +452,154 @@
|
|
| 452 |
"eval_samples_per_second": 139.748,
|
| 453 |
"eval_steps_per_second": 4.367,
|
| 454 |
"step": 30000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 455 |
}
|
| 456 |
],
|
| 457 |
"logging_steps": 500,
|
|
@@ -471,7 +619,7 @@
|
|
| 471 |
"attributes": {}
|
| 472 |
}
|
| 473 |
},
|
| 474 |
-
"total_flos":
|
| 475 |
"train_batch_size": 32,
|
| 476 |
"trial_name": null,
|
| 477 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 3.3261267254282387,
|
| 6 |
"eval_steps": 10000,
|
| 7 |
+
"global_step": 40000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 452 |
"eval_samples_per_second": 139.748,
|
| 453 |
"eval_steps_per_second": 4.367,
|
| 454 |
"step": 30000
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 2.536171628139032,
|
| 458 |
+
"grad_norm": NaN,
|
| 459 |
+
"learning_rate": 0.0,
|
| 460 |
+
"loss": 0.0,
|
| 461 |
+
"step": 30500
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 2.577748212206885,
|
| 465 |
+
"grad_norm": NaN,
|
| 466 |
+
"learning_rate": 0.0,
|
| 467 |
+
"loss": 0.0,
|
| 468 |
+
"step": 31000
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 2.619324796274738,
|
| 472 |
+
"grad_norm": NaN,
|
| 473 |
+
"learning_rate": 0.0,
|
| 474 |
+
"loss": 0.0,
|
| 475 |
+
"step": 31500
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 2.660901380342591,
|
| 479 |
+
"grad_norm": NaN,
|
| 480 |
+
"learning_rate": 0.0,
|
| 481 |
+
"loss": 0.0,
|
| 482 |
+
"step": 32000
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 2.702477964410444,
|
| 486 |
+
"grad_norm": NaN,
|
| 487 |
+
"learning_rate": 0.0,
|
| 488 |
+
"loss": 0.0,
|
| 489 |
+
"step": 32500
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 2.7440545484782968,
|
| 493 |
+
"grad_norm": NaN,
|
| 494 |
+
"learning_rate": 0.0,
|
| 495 |
+
"loss": 0.0,
|
| 496 |
+
"step": 33000
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 2.78563113254615,
|
| 500 |
+
"grad_norm": NaN,
|
| 501 |
+
"learning_rate": 0.0,
|
| 502 |
+
"loss": 0.0,
|
| 503 |
+
"step": 33500
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 2.827207716614003,
|
| 507 |
+
"grad_norm": NaN,
|
| 508 |
+
"learning_rate": 0.0,
|
| 509 |
+
"loss": 0.0,
|
| 510 |
+
"step": 34000
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 2.868784300681856,
|
| 514 |
+
"grad_norm": NaN,
|
| 515 |
+
"learning_rate": 0.0,
|
| 516 |
+
"loss": 0.0,
|
| 517 |
+
"step": 34500
|
| 518 |
+
},
|
| 519 |
+
{
|
| 520 |
+
"epoch": 2.910360884749709,
|
| 521 |
+
"grad_norm": NaN,
|
| 522 |
+
"learning_rate": 0.0,
|
| 523 |
+
"loss": 0.0,
|
| 524 |
+
"step": 35000
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 2.9519374688175617,
|
| 528 |
+
"grad_norm": NaN,
|
| 529 |
+
"learning_rate": 0.0,
|
| 530 |
+
"loss": 0.0,
|
| 531 |
+
"step": 35500
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 2.993514052885415,
|
| 535 |
+
"grad_norm": NaN,
|
| 536 |
+
"learning_rate": 0.0,
|
| 537 |
+
"loss": 0.0,
|
| 538 |
+
"step": 36000
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 3.0350906369532678,
|
| 542 |
+
"grad_norm": NaN,
|
| 543 |
+
"learning_rate": 0.0,
|
| 544 |
+
"loss": 0.0,
|
| 545 |
+
"step": 36500
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 3.076667221021121,
|
| 549 |
+
"grad_norm": NaN,
|
| 550 |
+
"learning_rate": 0.0,
|
| 551 |
+
"loss": 0.0,
|
| 552 |
+
"step": 37000
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 3.118243805088974,
|
| 556 |
+
"grad_norm": NaN,
|
| 557 |
+
"learning_rate": 0.0,
|
| 558 |
+
"loss": 0.0,
|
| 559 |
+
"step": 37500
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 3.159820389156827,
|
| 563 |
+
"grad_norm": NaN,
|
| 564 |
+
"learning_rate": 0.0,
|
| 565 |
+
"loss": 0.0,
|
| 566 |
+
"step": 38000
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 3.20139697322468,
|
| 570 |
+
"grad_norm": NaN,
|
| 571 |
+
"learning_rate": 0.0,
|
| 572 |
+
"loss": 0.0,
|
| 573 |
+
"step": 38500
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 3.2429735572925327,
|
| 577 |
+
"grad_norm": NaN,
|
| 578 |
+
"learning_rate": 0.0,
|
| 579 |
+
"loss": 0.0,
|
| 580 |
+
"step": 39000
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 3.284550141360386,
|
| 584 |
+
"grad_norm": NaN,
|
| 585 |
+
"learning_rate": 0.0,
|
| 586 |
+
"loss": 0.0,
|
| 587 |
+
"step": 39500
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 3.3261267254282387,
|
| 591 |
+
"grad_norm": NaN,
|
| 592 |
+
"learning_rate": 0.0,
|
| 593 |
+
"loss": 0.0,
|
| 594 |
+
"step": 40000
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 3.3261267254282387,
|
| 598 |
+
"eval_loss": NaN,
|
| 599 |
+
"eval_runtime": 2753.2662,
|
| 600 |
+
"eval_samples_per_second": 139.771,
|
| 601 |
+
"eval_steps_per_second": 4.368,
|
| 602 |
+
"step": 40000
|
| 603 |
}
|
| 604 |
],
|
| 605 |
"logging_steps": 500,
|
|
|
|
| 619 |
"attributes": {}
|
| 620 |
}
|
| 621 |
},
|
| 622 |
+
"total_flos": 4.483885876471327e+17,
|
| 623 |
"train_batch_size": 32,
|
| 624 |
"trial_name": null,
|
| 625 |
"trial_params": null
|