Training in progress, epoch 2, checkpoint
Browse files
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 228140600
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ad737a62d5e6dd3601ece9ec89b866a23dce6e9660089db12fbc69ce938d925e
|
| 3 |
size 228140600
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 116484839
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0dfe45f7e8553ab326b74acce24981ad2310a19952e96422e10cdcd05d9f3261
|
| 3 |
size 116484839
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14709
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f4ff3bd83efcc74d45f6dc982dfad42de943c268219c0ad0ee388295c41e8e02
|
| 3 |
size 14709
|
last-checkpoint/scaler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1383
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4cb6f523fe7cbe7ec261f5a7daf8f68472cbad6a063d529646d1f827a9ef9fd3
|
| 3 |
size 1383
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:53f409af08acb24ba2f85422d6d830e93fdc97a01268b4582a53eec3cbfeb20a
|
| 3 |
size 1465
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": 438,
|
| 3 |
"best_metric": 1.2615772485733032,
|
| 4 |
"best_model_checkpoint": "./adapter-phase2/checkpoint-438",
|
| 5 |
-
"epoch":
|
| 6 |
"eval_steps": 500,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -450,6 +450,458 @@
|
|
| 450 |
"eval_samples_per_second": 3.459,
|
| 451 |
"eval_steps_per_second": 0.866,
|
| 452 |
"step": 438
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 453 |
}
|
| 454 |
],
|
| 455 |
"logging_steps": 10,
|
|
@@ -469,7 +921,7 @@
|
|
| 469 |
"attributes": {}
|
| 470 |
}
|
| 471 |
},
|
| 472 |
-
"total_flos":
|
| 473 |
"train_batch_size": 1,
|
| 474 |
"trial_name": null,
|
| 475 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": 438,
|
| 3 |
"best_metric": 1.2615772485733032,
|
| 4 |
"best_model_checkpoint": "./adapter-phase2/checkpoint-438",
|
| 5 |
+
"epoch": 2.0,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 876,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 450 |
"eval_samples_per_second": 3.459,
|
| 451 |
"eval_steps_per_second": 0.866,
|
| 452 |
"step": 438
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"entropy": 0.9707203358411789,
|
| 456 |
+
"epoch": 1.0045714285714287,
|
| 457 |
+
"grad_norm": 1.046062707901001,
|
| 458 |
+
"learning_rate": 7.99543378995434e-06,
|
| 459 |
+
"loss": 0.8608,
|
| 460 |
+
"mean_token_accuracy": 0.7782945515293824,
|
| 461 |
+
"num_tokens": 425275.0,
|
| 462 |
+
"step": 440
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"entropy": 0.7523622503504157,
|
| 466 |
+
"epoch": 1.0274285714285714,
|
| 467 |
+
"grad_norm": 1.2251920700073242,
|
| 468 |
+
"learning_rate": 7.949771689497718e-06,
|
| 469 |
+
"loss": 0.7473,
|
| 470 |
+
"mean_token_accuracy": 0.8092356324195862,
|
| 471 |
+
"num_tokens": 439850.0,
|
| 472 |
+
"step": 450
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"entropy": 0.84888547193259,
|
| 476 |
+
"epoch": 1.0502857142857143,
|
| 477 |
+
"grad_norm": 1.325343370437622,
|
| 478 |
+
"learning_rate": 7.904109589041097e-06,
|
| 479 |
+
"loss": 0.7907,
|
| 480 |
+
"mean_token_accuracy": 0.7977306388318539,
|
| 481 |
+
"num_tokens": 451136.0,
|
| 482 |
+
"step": 460
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"entropy": 0.9735806178301573,
|
| 486 |
+
"epoch": 1.0731428571428572,
|
| 487 |
+
"grad_norm": 1.6439032554626465,
|
| 488 |
+
"learning_rate": 7.858447488584475e-06,
|
| 489 |
+
"loss": 0.9513,
|
| 490 |
+
"mean_token_accuracy": 0.7617030199617147,
|
| 491 |
+
"num_tokens": 459486.0,
|
| 492 |
+
"step": 470
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"entropy": 1.0277788739651441,
|
| 496 |
+
"epoch": 1.096,
|
| 497 |
+
"grad_norm": 1.8182581663131714,
|
| 498 |
+
"learning_rate": 7.812785388127855e-06,
|
| 499 |
+
"loss": 0.9526,
|
| 500 |
+
"mean_token_accuracy": 0.7630892738699913,
|
| 501 |
+
"num_tokens": 465975.0,
|
| 502 |
+
"step": 480
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"entropy": 0.943339848332107,
|
| 506 |
+
"epoch": 1.1188571428571428,
|
| 507 |
+
"grad_norm": 1.1697943210601807,
|
| 508 |
+
"learning_rate": 7.767123287671234e-06,
|
| 509 |
+
"loss": 0.851,
|
| 510 |
+
"mean_token_accuracy": 0.7827658370137215,
|
| 511 |
+
"num_tokens": 473929.0,
|
| 512 |
+
"step": 490
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"entropy": 0.7733788685873151,
|
| 516 |
+
"epoch": 1.1417142857142857,
|
| 517 |
+
"grad_norm": 1.2411632537841797,
|
| 518 |
+
"learning_rate": 7.721461187214612e-06,
|
| 519 |
+
"loss": 0.7691,
|
| 520 |
+
"mean_token_accuracy": 0.8111804500222206,
|
| 521 |
+
"num_tokens": 488306.0,
|
| 522 |
+
"step": 500
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"entropy": 0.824394048191607,
|
| 526 |
+
"epoch": 1.1645714285714286,
|
| 527 |
+
"grad_norm": 1.3971821069717407,
|
| 528 |
+
"learning_rate": 7.675799086757991e-06,
|
| 529 |
+
"loss": 0.7429,
|
| 530 |
+
"mean_token_accuracy": 0.8088308341801167,
|
| 531 |
+
"num_tokens": 499208.0,
|
| 532 |
+
"step": 510
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"entropy": 0.9718045836314559,
|
| 536 |
+
"epoch": 1.1874285714285715,
|
| 537 |
+
"grad_norm": 1.7808269262313843,
|
| 538 |
+
"learning_rate": 7.630136986301371e-06,
|
| 539 |
+
"loss": 0.9365,
|
| 540 |
+
"mean_token_accuracy": 0.762304800376296,
|
| 541 |
+
"num_tokens": 507299.0,
|
| 542 |
+
"step": 520
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"entropy": 0.9984221205115318,
|
| 546 |
+
"epoch": 1.2102857142857144,
|
| 547 |
+
"grad_norm": 2.0445668697357178,
|
| 548 |
+
"learning_rate": 7.58447488584475e-06,
|
| 549 |
+
"loss": 0.9299,
|
| 550 |
+
"mean_token_accuracy": 0.7667763099074364,
|
| 551 |
+
"num_tokens": 513558.0,
|
| 552 |
+
"step": 530
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"entropy": 0.9143548993393779,
|
| 556 |
+
"epoch": 1.233142857142857,
|
| 557 |
+
"grad_norm": 1.4540475606918335,
|
| 558 |
+
"learning_rate": 7.538812785388129e-06,
|
| 559 |
+
"loss": 0.7977,
|
| 560 |
+
"mean_token_accuracy": 0.7940818261355161,
|
| 561 |
+
"num_tokens": 521477.0,
|
| 562 |
+
"step": 540
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"entropy": 0.7225218357518315,
|
| 566 |
+
"epoch": 1.256,
|
| 567 |
+
"grad_norm": 1.4530831575393677,
|
| 568 |
+
"learning_rate": 7.4931506849315075e-06,
|
| 569 |
+
"loss": 0.7282,
|
| 570 |
+
"mean_token_accuracy": 0.8143456902354955,
|
| 571 |
+
"num_tokens": 536156.0,
|
| 572 |
+
"step": 550
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"entropy": 0.826616644486785,
|
| 576 |
+
"epoch": 1.278857142857143,
|
| 577 |
+
"grad_norm": 1.307198166847229,
|
| 578 |
+
"learning_rate": 7.447488584474887e-06,
|
| 579 |
+
"loss": 0.7509,
|
| 580 |
+
"mean_token_accuracy": 0.8082947298884392,
|
| 581 |
+
"num_tokens": 547250.0,
|
| 582 |
+
"step": 560
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"entropy": 0.9597857438027859,
|
| 586 |
+
"epoch": 1.3017142857142856,
|
| 587 |
+
"grad_norm": 2.1991994380950928,
|
| 588 |
+
"learning_rate": 7.401826484018265e-06,
|
| 589 |
+
"loss": 0.9294,
|
| 590 |
+
"mean_token_accuracy": 0.7654231000691653,
|
| 591 |
+
"num_tokens": 555523.0,
|
| 592 |
+
"step": 570
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"entropy": 1.0113731533288957,
|
| 596 |
+
"epoch": 1.3245714285714285,
|
| 597 |
+
"grad_norm": 2.2134881019592285,
|
| 598 |
+
"learning_rate": 7.356164383561645e-06,
|
| 599 |
+
"loss": 0.9149,
|
| 600 |
+
"mean_token_accuracy": 0.7716960549354553,
|
| 601 |
+
"num_tokens": 561790.0,
|
| 602 |
+
"step": 580
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"entropy": 0.89712286721915,
|
| 606 |
+
"epoch": 1.3474285714285714,
|
| 607 |
+
"grad_norm": 1.2084845304489136,
|
| 608 |
+
"learning_rate": 7.310502283105023e-06,
|
| 609 |
+
"loss": 0.7891,
|
| 610 |
+
"mean_token_accuracy": 0.7911178763955832,
|
| 611 |
+
"num_tokens": 569844.0,
|
| 612 |
+
"step": 590
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"entropy": 0.7331022916361689,
|
| 616 |
+
"epoch": 1.3702857142857143,
|
| 617 |
+
"grad_norm": 1.3023542165756226,
|
| 618 |
+
"learning_rate": 7.269406392694065e-06,
|
| 619 |
+
"loss": 0.7457,
|
| 620 |
+
"mean_token_accuracy": 0.8113049529492855,
|
| 621 |
+
"num_tokens": 584459.0,
|
| 622 |
+
"step": 600
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"entropy": 0.7879349924623966,
|
| 626 |
+
"epoch": 1.3931428571428572,
|
| 627 |
+
"grad_norm": 1.555379867553711,
|
| 628 |
+
"learning_rate": 7.223744292237444e-06,
|
| 629 |
+
"loss": 0.7306,
|
| 630 |
+
"mean_token_accuracy": 0.8145640216767788,
|
| 631 |
+
"num_tokens": 595804.0,
|
| 632 |
+
"step": 610
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"entropy": 0.9201723251491785,
|
| 636 |
+
"epoch": 1.416,
|
| 637 |
+
"grad_norm": 2.0131261348724365,
|
| 638 |
+
"learning_rate": 7.178082191780823e-06,
|
| 639 |
+
"loss": 0.881,
|
| 640 |
+
"mean_token_accuracy": 0.7761596899479628,
|
| 641 |
+
"num_tokens": 604098.0,
|
| 642 |
+
"step": 620
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"entropy": 1.0043058268725873,
|
| 646 |
+
"epoch": 1.4388571428571428,
|
| 647 |
+
"grad_norm": 1.952837586402893,
|
| 648 |
+
"learning_rate": 7.132420091324202e-06,
|
| 649 |
+
"loss": 0.9229,
|
| 650 |
+
"mean_token_accuracy": 0.7723097205162048,
|
| 651 |
+
"num_tokens": 610481.0,
|
| 652 |
+
"step": 630
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"entropy": 0.8940085913985968,
|
| 656 |
+
"epoch": 1.4617142857142857,
|
| 657 |
+
"grad_norm": 1.2801399230957031,
|
| 658 |
+
"learning_rate": 7.086757990867581e-06,
|
| 659 |
+
"loss": 0.8006,
|
| 660 |
+
"mean_token_accuracy": 0.7930479496717453,
|
| 661 |
+
"num_tokens": 618699.0,
|
| 662 |
+
"step": 640
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"entropy": 0.6966889450326562,
|
| 666 |
+
"epoch": 1.4845714285714287,
|
| 667 |
+
"grad_norm": 1.557562232017517,
|
| 668 |
+
"learning_rate": 7.0410958904109596e-06,
|
| 669 |
+
"loss": 0.665,
|
| 670 |
+
"mean_token_accuracy": 0.8264754865318537,
|
| 671 |
+
"num_tokens": 632856.0,
|
| 672 |
+
"step": 650
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"entropy": 0.8100471086800098,
|
| 676 |
+
"epoch": 1.5074285714285716,
|
| 677 |
+
"grad_norm": 1.7616751194000244,
|
| 678 |
+
"learning_rate": 6.995433789954339e-06,
|
| 679 |
+
"loss": 0.7669,
|
| 680 |
+
"mean_token_accuracy": 0.8096333492547274,
|
| 681 |
+
"num_tokens": 643712.0,
|
| 682 |
+
"step": 660
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"entropy": 0.9476521443575621,
|
| 686 |
+
"epoch": 1.5302857142857142,
|
| 687 |
+
"grad_norm": 1.97320556640625,
|
| 688 |
+
"learning_rate": 6.9497716894977175e-06,
|
| 689 |
+
"loss": 0.8769,
|
| 690 |
+
"mean_token_accuracy": 0.7822451706975698,
|
| 691 |
+
"num_tokens": 651732.0,
|
| 692 |
+
"step": 670
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"entropy": 0.9541807420551777,
|
| 696 |
+
"epoch": 1.5531428571428572,
|
| 697 |
+
"grad_norm": 2.2813711166381836,
|
| 698 |
+
"learning_rate": 6.904109589041097e-06,
|
| 699 |
+
"loss": 0.8731,
|
| 700 |
+
"mean_token_accuracy": 0.7764547783881426,
|
| 701 |
+
"num_tokens": 658104.0,
|
| 702 |
+
"step": 680
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"entropy": 0.8891686601564288,
|
| 706 |
+
"epoch": 1.576,
|
| 707 |
+
"grad_norm": 1.2347137928009033,
|
| 708 |
+
"learning_rate": 6.858447488584475e-06,
|
| 709 |
+
"loss": 0.8099,
|
| 710 |
+
"mean_token_accuracy": 0.795854776352644,
|
| 711 |
+
"num_tokens": 666681.0,
|
| 712 |
+
"step": 690
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"entropy": 0.7062053712084889,
|
| 716 |
+
"epoch": 1.5988571428571428,
|
| 717 |
+
"grad_norm": 1.505817174911499,
|
| 718 |
+
"learning_rate": 6.812785388127855e-06,
|
| 719 |
+
"loss": 0.6689,
|
| 720 |
+
"mean_token_accuracy": 0.8225430808961391,
|
| 721 |
+
"num_tokens": 681161.0,
|
| 722 |
+
"step": 700
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"entropy": 0.7627910353243351,
|
| 726 |
+
"epoch": 1.6217142857142857,
|
| 727 |
+
"grad_norm": 1.7354750633239746,
|
| 728 |
+
"learning_rate": 6.767123287671233e-06,
|
| 729 |
+
"loss": 0.7217,
|
| 730 |
+
"mean_token_accuracy": 0.8088484812527895,
|
| 731 |
+
"num_tokens": 692262.0,
|
| 732 |
+
"step": 710
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"entropy": 0.9181301448494196,
|
| 736 |
+
"epoch": 1.6445714285714286,
|
| 737 |
+
"grad_norm": 1.9427331686019897,
|
| 738 |
+
"learning_rate": 6.721461187214613e-06,
|
| 739 |
+
"loss": 0.8664,
|
| 740 |
+
"mean_token_accuracy": 0.7764203164726495,
|
| 741 |
+
"num_tokens": 700252.0,
|
| 742 |
+
"step": 720
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"entropy": 0.970825233310461,
|
| 746 |
+
"epoch": 1.6674285714285715,
|
| 747 |
+
"grad_norm": 2.231489419937134,
|
| 748 |
+
"learning_rate": 6.675799086757991e-06,
|
| 749 |
+
"loss": 0.8727,
|
| 750 |
+
"mean_token_accuracy": 0.77991351634264,
|
| 751 |
+
"num_tokens": 706466.0,
|
| 752 |
+
"step": 730
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"entropy": 0.8769128978252411,
|
| 756 |
+
"epoch": 1.6902857142857144,
|
| 757 |
+
"grad_norm": 1.3580577373504639,
|
| 758 |
+
"learning_rate": 6.630136986301371e-06,
|
| 759 |
+
"loss": 0.7826,
|
| 760 |
+
"mean_token_accuracy": 0.7997685220092535,
|
| 761 |
+
"num_tokens": 714701.0,
|
| 762 |
+
"step": 740
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"entropy": 0.6923451218754053,
|
| 766 |
+
"epoch": 1.713142857142857,
|
| 767 |
+
"grad_norm": 1.4095361232757568,
|
| 768 |
+
"learning_rate": 6.584474885844749e-06,
|
| 769 |
+
"loss": 0.6984,
|
| 770 |
+
"mean_token_accuracy": 0.8204937841743231,
|
| 771 |
+
"num_tokens": 729622.0,
|
| 772 |
+
"step": 750
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"entropy": 0.7426450593397022,
|
| 776 |
+
"epoch": 1.736,
|
| 777 |
+
"grad_norm": 1.5736570358276367,
|
| 778 |
+
"learning_rate": 6.538812785388129e-06,
|
| 779 |
+
"loss": 0.667,
|
| 780 |
+
"mean_token_accuracy": 0.8291565012186766,
|
| 781 |
+
"num_tokens": 740772.0,
|
| 782 |
+
"step": 760
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"entropy": 0.910079357214272,
|
| 786 |
+
"epoch": 1.758857142857143,
|
| 787 |
+
"grad_norm": 2.1047656536102295,
|
| 788 |
+
"learning_rate": 6.493150684931508e-06,
|
| 789 |
+
"loss": 0.875,
|
| 790 |
+
"mean_token_accuracy": 0.7781037461012602,
|
| 791 |
+
"num_tokens": 748857.0,
|
| 792 |
+
"step": 770
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"entropy": 0.9749910116195679,
|
| 796 |
+
"epoch": 1.7817142857142856,
|
| 797 |
+
"grad_norm": 2.2609705924987793,
|
| 798 |
+
"learning_rate": 6.447488584474887e-06,
|
| 799 |
+
"loss": 0.9058,
|
| 800 |
+
"mean_token_accuracy": 0.7749961122870446,
|
| 801 |
+
"num_tokens": 755273.0,
|
| 802 |
+
"step": 780
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"entropy": 0.8688624935224653,
|
| 806 |
+
"epoch": 1.8045714285714287,
|
| 807 |
+
"grad_norm": 2.156954765319824,
|
| 808 |
+
"learning_rate": 6.401826484018266e-06,
|
| 809 |
+
"loss": 0.7568,
|
| 810 |
+
"mean_token_accuracy": 0.8001648161560297,
|
| 811 |
+
"num_tokens": 763404.0,
|
| 812 |
+
"step": 790
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"entropy": 0.6553533479571343,
|
| 816 |
+
"epoch": 1.8274285714285714,
|
| 817 |
+
"grad_norm": 1.5286246538162231,
|
| 818 |
+
"learning_rate": 6.356164383561645e-06,
|
| 819 |
+
"loss": 0.6357,
|
| 820 |
+
"mean_token_accuracy": 0.8322514686733484,
|
| 821 |
+
"num_tokens": 777652.0,
|
| 822 |
+
"step": 800
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"entropy": 0.7381465582177043,
|
| 826 |
+
"epoch": 1.8502857142857143,
|
| 827 |
+
"grad_norm": 1.889930248260498,
|
| 828 |
+
"learning_rate": 6.3105022831050235e-06,
|
| 829 |
+
"loss": 0.6995,
|
| 830 |
+
"mean_token_accuracy": 0.8194405883550644,
|
| 831 |
+
"num_tokens": 788541.0,
|
| 832 |
+
"step": 810
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"entropy": 0.9207667458802462,
|
| 836 |
+
"epoch": 1.8731428571428572,
|
| 837 |
+
"grad_norm": 2.3677663803100586,
|
| 838 |
+
"learning_rate": 6.264840182648403e-06,
|
| 839 |
+
"loss": 0.876,
|
| 840 |
+
"mean_token_accuracy": 0.7714111492037773,
|
| 841 |
+
"num_tokens": 796574.0,
|
| 842 |
+
"step": 820
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"entropy": 0.9494761880487204,
|
| 846 |
+
"epoch": 1.896,
|
| 847 |
+
"grad_norm": 2.424638032913208,
|
| 848 |
+
"learning_rate": 6.219178082191781e-06,
|
| 849 |
+
"loss": 0.8548,
|
| 850 |
+
"mean_token_accuracy": 0.7811690699309111,
|
| 851 |
+
"num_tokens": 802836.0,
|
| 852 |
+
"step": 830
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"entropy": 0.8909835416823626,
|
| 856 |
+
"epoch": 1.9188571428571428,
|
| 857 |
+
"grad_norm": 1.3449039459228516,
|
| 858 |
+
"learning_rate": 6.173515981735161e-06,
|
| 859 |
+
"loss": 0.7825,
|
| 860 |
+
"mean_token_accuracy": 0.7954777158796787,
|
| 861 |
+
"num_tokens": 810726.0,
|
| 862 |
+
"step": 840
|
| 863 |
+
},
|
| 864 |
+
{
|
| 865 |
+
"entropy": 0.6921561988070607,
|
| 866 |
+
"epoch": 1.9417142857142857,
|
| 867 |
+
"grad_norm": 1.490689992904663,
|
| 868 |
+
"learning_rate": 6.127853881278539e-06,
|
| 869 |
+
"loss": 0.6554,
|
| 870 |
+
"mean_token_accuracy": 0.8262197155505419,
|
| 871 |
+
"num_tokens": 824145.0,
|
| 872 |
+
"step": 850
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"entropy": 0.8137379666790366,
|
| 876 |
+
"epoch": 1.9645714285714284,
|
| 877 |
+
"grad_norm": 2.0120434761047363,
|
| 878 |
+
"learning_rate": 6.082191780821919e-06,
|
| 879 |
+
"loss": 0.8024,
|
| 880 |
+
"mean_token_accuracy": 0.7950452182441949,
|
| 881 |
+
"num_tokens": 833220.0,
|
| 882 |
+
"step": 860
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"entropy": 0.9502449594438076,
|
| 886 |
+
"epoch": 1.9874285714285715,
|
| 887 |
+
"grad_norm": 2.679570198059082,
|
| 888 |
+
"learning_rate": 6.036529680365297e-06,
|
| 889 |
+
"loss": 0.8545,
|
| 890 |
+
"mean_token_accuracy": 0.781839894503355,
|
| 891 |
+
"num_tokens": 839758.0,
|
| 892 |
+
"step": 870
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 2.0,
|
| 896 |
+
"eval_accuracy": 0.00894328845369237,
|
| 897 |
+
"eval_entropy": 0.9275054344799528,
|
| 898 |
+
"eval_loss": 1.3659894466400146,
|
| 899 |
+
"eval_mean_token_accuracy": 0.7276899333626147,
|
| 900 |
+
"eval_num_tokens": 842388.0,
|
| 901 |
+
"eval_runtime": 299.6651,
|
| 902 |
+
"eval_samples_per_second": 3.451,
|
| 903 |
+
"eval_steps_per_second": 0.864,
|
| 904 |
+
"step": 876
|
| 905 |
}
|
| 906 |
],
|
| 907 |
"logging_steps": 10,
|
|
|
|
| 921 |
"attributes": {}
|
| 922 |
}
|
| 923 |
},
|
| 924 |
+
"total_flos": 5.880058660122624e+16,
|
| 925 |
"train_batch_size": 1,
|
| 926 |
"trial_name": null,
|
| 927 |
"trial_params": null
|