Training in progress, step 1000, 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 295488936
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:17ba270b888a201fead48ad37c2c2e228e832cc5e2304c9d48ddcc2a4ab95b9d
|
| 3 |
size 295488936
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 591203178
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:94ad12c53cb7962a72e2d80a27249286394dc06a5b1f83bd4257087da8221ea0
|
| 3 |
size 591203178
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5edb34d031c0c2b447f3eaadb401a4c1e7e7e6d8c096e28b7092e01a8bd48c92
|
| 3 |
size 14244
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b7a13d6eba883692f8ed583bb8ce176c7e7a1118cd4c39d5498dbe981adfa197
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 100,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -397,6 +397,396 @@
|
|
| 397 |
"eval_news_finetune_val_samples_per_second": 1.396,
|
| 398 |
"eval_news_finetune_val_steps_per_second": 1.396,
|
| 399 |
"step": 500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 400 |
}
|
| 401 |
],
|
| 402 |
"logging_steps": 10,
|
|
@@ -416,7 +806,7 @@
|
|
| 416 |
"attributes": {}
|
| 417 |
}
|
| 418 |
},
|
| 419 |
-
"total_flos":
|
| 420 |
"train_batch_size": 1,
|
| 421 |
"trial_name": null,
|
| 422 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.7145409074669525,
|
| 5 |
"eval_steps": 100,
|
| 6 |
+
"global_step": 1000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 397 |
"eval_news_finetune_val_samples_per_second": 1.396,
|
| 398 |
"eval_news_finetune_val_steps_per_second": 1.396,
|
| 399 |
"step": 500
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 0.36441586280814575,
|
| 403 |
+
"grad_norm": 1.282714605331421,
|
| 404 |
+
"learning_rate": 9.985996777749747e-05,
|
| 405 |
+
"loss": 0.3636,
|
| 406 |
+
"step": 510
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 0.3715612718828153,
|
| 410 |
+
"grad_norm": 2.0360989570617676,
|
| 411 |
+
"learning_rate": 9.982713965133122e-05,
|
| 412 |
+
"loss": 0.4467,
|
| 413 |
+
"step": 520
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 0.37870668095748483,
|
| 417 |
+
"grad_norm": 1.7432626485824585,
|
| 418 |
+
"learning_rate": 9.979086430335417e-05,
|
| 419 |
+
"loss": 0.3875,
|
| 420 |
+
"step": 530
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 0.3858520900321543,
|
| 424 |
+
"grad_norm": 1.6053438186645508,
|
| 425 |
+
"learning_rate": 9.975114424322609e-05,
|
| 426 |
+
"loss": 0.3646,
|
| 427 |
+
"step": 540
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 0.39299749910682386,
|
| 431 |
+
"grad_norm": 1.2323070764541626,
|
| 432 |
+
"learning_rate": 9.970798221892452e-05,
|
| 433 |
+
"loss": 0.353,
|
| 434 |
+
"step": 550
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.4001429081814934,
|
| 438 |
+
"grad_norm": 1.16932213306427,
|
| 439 |
+
"learning_rate": 9.966138121655445e-05,
|
| 440 |
+
"loss": 0.331,
|
| 441 |
+
"step": 560
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.40728831725616294,
|
| 445 |
+
"grad_norm": 1.8134998083114624,
|
| 446 |
+
"learning_rate": 9.961134446014184e-05,
|
| 447 |
+
"loss": 0.3132,
|
| 448 |
+
"step": 570
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 0.4144337263308324,
|
| 452 |
+
"grad_norm": 1.4292124509811401,
|
| 453 |
+
"learning_rate": 9.955787541141055e-05,
|
| 454 |
+
"loss": 0.3017,
|
| 455 |
+
"step": 580
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 0.42157913540550196,
|
| 459 |
+
"grad_norm": 1.4605034589767456,
|
| 460 |
+
"learning_rate": 9.950097776954284e-05,
|
| 461 |
+
"loss": 0.3596,
|
| 462 |
+
"step": 590
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 0.4287245444801715,
|
| 466 |
+
"grad_norm": 1.2365972995758057,
|
| 467 |
+
"learning_rate": 9.944065547092345e-05,
|
| 468 |
+
"loss": 0.3399,
|
| 469 |
+
"step": 600
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 0.4287245444801715,
|
| 473 |
+
"eval_news_finetune_val_loss": 0.36549311876296997,
|
| 474 |
+
"eval_news_finetune_val_runtime": 1002.8044,
|
| 475 |
+
"eval_news_finetune_val_samples_per_second": 1.396,
|
| 476 |
+
"eval_news_finetune_val_steps_per_second": 1.396,
|
| 477 |
+
"step": 600
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 0.43586995355484104,
|
| 481 |
+
"grad_norm": 1.0590678453445435,
|
| 482 |
+
"learning_rate": 9.937691268886725e-05,
|
| 483 |
+
"loss": 0.3747,
|
| 484 |
+
"step": 610
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 0.4430153626295105,
|
| 488 |
+
"grad_norm": 0.9111473560333252,
|
| 489 |
+
"learning_rate": 9.930975383333056e-05,
|
| 490 |
+
"loss": 0.2868,
|
| 491 |
+
"step": 620
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 0.45016077170418006,
|
| 495 |
+
"grad_norm": 2.0456018447875977,
|
| 496 |
+
"learning_rate": 9.923918355060599e-05,
|
| 497 |
+
"loss": 0.3289,
|
| 498 |
+
"step": 630
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 0.4573061807788496,
|
| 502 |
+
"grad_norm": 1.5998501777648926,
|
| 503 |
+
"learning_rate": 9.916520672300107e-05,
|
| 504 |
+
"loss": 0.3664,
|
| 505 |
+
"step": 640
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 0.4644515898535191,
|
| 509 |
+
"grad_norm": 1.0773181915283203,
|
| 510 |
+
"learning_rate": 9.908782846850037e-05,
|
| 511 |
+
"loss": 0.3432,
|
| 512 |
+
"step": 650
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.4715969989281886,
|
| 516 |
+
"grad_norm": 1.244042158126831,
|
| 517 |
+
"learning_rate": 9.900705414041154e-05,
|
| 518 |
+
"loss": 0.3242,
|
| 519 |
+
"step": 660
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 0.47874240800285817,
|
| 523 |
+
"grad_norm": 1.8120310306549072,
|
| 524 |
+
"learning_rate": 9.892288932699484e-05,
|
| 525 |
+
"loss": 0.317,
|
| 526 |
+
"step": 670
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 0.4858878170775277,
|
| 530 |
+
"grad_norm": 0.7863224148750305,
|
| 531 |
+
"learning_rate": 9.883533985107663e-05,
|
| 532 |
+
"loss": 0.322,
|
| 533 |
+
"step": 680
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 0.4930332261521972,
|
| 537 |
+
"grad_norm": 1.223832130432129,
|
| 538 |
+
"learning_rate": 9.874441176964642e-05,
|
| 539 |
+
"loss": 0.343,
|
| 540 |
+
"step": 690
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 0.5001786352268668,
|
| 544 |
+
"grad_norm": 0.9870743155479431,
|
| 545 |
+
"learning_rate": 9.865011137343787e-05,
|
| 546 |
+
"loss": 0.3278,
|
| 547 |
+
"step": 700
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 0.5001786352268668,
|
| 551 |
+
"eval_news_finetune_val_loss": 0.35386842489242554,
|
| 552 |
+
"eval_news_finetune_val_runtime": 1003.4109,
|
| 553 |
+
"eval_news_finetune_val_samples_per_second": 1.395,
|
| 554 |
+
"eval_news_finetune_val_steps_per_second": 1.395,
|
| 555 |
+
"step": 700
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.5073240443015362,
|
| 559 |
+
"grad_norm": 1.3699963092803955,
|
| 560 |
+
"learning_rate": 9.85524451864936e-05,
|
| 561 |
+
"loss": 0.3902,
|
| 562 |
+
"step": 710
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.5144694533762058,
|
| 566 |
+
"grad_norm": 1.7188071012496948,
|
| 567 |
+
"learning_rate": 9.845141996571384e-05,
|
| 568 |
+
"loss": 0.369,
|
| 569 |
+
"step": 720
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.5216148624508753,
|
| 573 |
+
"grad_norm": 0.4889034628868103,
|
| 574 |
+
"learning_rate": 9.834704270038888e-05,
|
| 575 |
+
"loss": 0.3174,
|
| 576 |
+
"step": 730
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.5287602715255448,
|
| 580 |
+
"grad_norm": 0.8782143592834473,
|
| 581 |
+
"learning_rate": 9.823932061171561e-05,
|
| 582 |
+
"loss": 0.3501,
|
| 583 |
+
"step": 740
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.5359056806002144,
|
| 587 |
+
"grad_norm": 2.4089126586914062,
|
| 588 |
+
"learning_rate": 9.812826115229789e-05,
|
| 589 |
+
"loss": 0.3292,
|
| 590 |
+
"step": 750
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.5430510896748839,
|
| 594 |
+
"grad_norm": 1.6382787227630615,
|
| 595 |
+
"learning_rate": 9.801387200563096e-05,
|
| 596 |
+
"loss": 0.459,
|
| 597 |
+
"step": 760
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.5501964987495535,
|
| 601 |
+
"grad_norm": 1.443916916847229,
|
| 602 |
+
"learning_rate": 9.789616108556992e-05,
|
| 603 |
+
"loss": 0.3409,
|
| 604 |
+
"step": 770
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.5573419078242229,
|
| 608 |
+
"grad_norm": 1.632278323173523,
|
| 609 |
+
"learning_rate": 9.77751365357821e-05,
|
| 610 |
+
"loss": 0.281,
|
| 611 |
+
"step": 780
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.5644873168988924,
|
| 615 |
+
"grad_norm": 2.1452109813690186,
|
| 616 |
+
"learning_rate": 9.765080672918374e-05,
|
| 617 |
+
"loss": 0.3511,
|
| 618 |
+
"step": 790
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.571632725973562,
|
| 622 |
+
"grad_norm": 1.2721842527389526,
|
| 623 |
+
"learning_rate": 9.752318026736078e-05,
|
| 624 |
+
"loss": 0.2298,
|
| 625 |
+
"step": 800
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.571632725973562,
|
| 629 |
+
"eval_news_finetune_val_loss": 0.34554028511047363,
|
| 630 |
+
"eval_news_finetune_val_runtime": 1003.3342,
|
| 631 |
+
"eval_news_finetune_val_samples_per_second": 1.395,
|
| 632 |
+
"eval_news_finetune_val_steps_per_second": 1.395,
|
| 633 |
+
"step": 800
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.5787781350482315,
|
| 637 |
+
"grad_norm": 2.5264174938201904,
|
| 638 |
+
"learning_rate": 9.739226597997359e-05,
|
| 639 |
+
"loss": 0.3214,
|
| 640 |
+
"step": 810
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.585923544122901,
|
| 644 |
+
"grad_norm": 1.4553183317184448,
|
| 645 |
+
"learning_rate": 9.725807292414629e-05,
|
| 646 |
+
"loss": 0.2697,
|
| 647 |
+
"step": 820
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.5930689531975706,
|
| 651 |
+
"grad_norm": 2.2111873626708984,
|
| 652 |
+
"learning_rate": 9.712061038384002e-05,
|
| 653 |
+
"loss": 0.3315,
|
| 654 |
+
"step": 830
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.6002143622722401,
|
| 658 |
+
"grad_norm": 1.4308302402496338,
|
| 659 |
+
"learning_rate": 9.697988786921071e-05,
|
| 660 |
+
"loss": 0.4036,
|
| 661 |
+
"step": 840
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.6073597713469097,
|
| 665 |
+
"grad_norm": 1.8136054277420044,
|
| 666 |
+
"learning_rate": 9.683591511595107e-05,
|
| 667 |
+
"loss": 0.2946,
|
| 668 |
+
"step": 850
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.6145051804215791,
|
| 672 |
+
"grad_norm": 1.8586084842681885,
|
| 673 |
+
"learning_rate": 9.668870208461713e-05,
|
| 674 |
+
"loss": 0.2259,
|
| 675 |
+
"step": 860
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.6216505894962486,
|
| 679 |
+
"grad_norm": 1.1640444993972778,
|
| 680 |
+
"learning_rate": 9.653825895993908e-05,
|
| 681 |
+
"loss": 0.4,
|
| 682 |
+
"step": 870
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.6287959985709182,
|
| 686 |
+
"grad_norm": 1.386013388633728,
|
| 687 |
+
"learning_rate": 9.63845961501166e-05,
|
| 688 |
+
"loss": 0.2804,
|
| 689 |
+
"step": 880
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.6359414076455877,
|
| 693 |
+
"grad_norm": 2.1413650512695312,
|
| 694 |
+
"learning_rate": 9.622772428609887e-05,
|
| 695 |
+
"loss": 0.3593,
|
| 696 |
+
"step": 890
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.6430868167202572,
|
| 700 |
+
"grad_norm": 1.5462217330932617,
|
| 701 |
+
"learning_rate": 9.606765422084908e-05,
|
| 702 |
+
"loss": 0.3058,
|
| 703 |
+
"step": 900
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.6430868167202572,
|
| 707 |
+
"eval_news_finetune_val_loss": 0.3292103707790375,
|
| 708 |
+
"eval_news_finetune_val_runtime": 1003.4558,
|
| 709 |
+
"eval_news_finetune_val_samples_per_second": 1.395,
|
| 710 |
+
"eval_news_finetune_val_steps_per_second": 1.395,
|
| 711 |
+
"step": 900
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 0.6502322257949268,
|
| 715 |
+
"grad_norm": 1.0373942852020264,
|
| 716 |
+
"learning_rate": 9.590439702859351e-05,
|
| 717 |
+
"loss": 0.3318,
|
| 718 |
+
"step": 910
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 0.6573776348695963,
|
| 722 |
+
"grad_norm": 1.2724213600158691,
|
| 723 |
+
"learning_rate": 9.573796400405544e-05,
|
| 724 |
+
"loss": 0.3328,
|
| 725 |
+
"step": 920
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 0.6645230439442658,
|
| 729 |
+
"grad_norm": 0.8528966903686523,
|
| 730 |
+
"learning_rate": 9.55683666616737e-05,
|
| 731 |
+
"loss": 0.2673,
|
| 732 |
+
"step": 930
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 0.6716684530189353,
|
| 736 |
+
"grad_norm": 1.65499746799469,
|
| 737 |
+
"learning_rate": 9.539561673480612e-05,
|
| 738 |
+
"loss": 0.3538,
|
| 739 |
+
"step": 940
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 0.6788138620936048,
|
| 743 |
+
"grad_norm": 2.341379404067993,
|
| 744 |
+
"learning_rate": 9.521972617491767e-05,
|
| 745 |
+
"loss": 0.3228,
|
| 746 |
+
"step": 950
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 0.6859592711682744,
|
| 750 |
+
"grad_norm": 1.4938244819641113,
|
| 751 |
+
"learning_rate": 9.504070715075372e-05,
|
| 752 |
+
"loss": 0.3974,
|
| 753 |
+
"step": 960
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 0.6931046802429439,
|
| 757 |
+
"grad_norm": 1.0390361547470093,
|
| 758 |
+
"learning_rate": 9.485857204749811e-05,
|
| 759 |
+
"loss": 0.3236,
|
| 760 |
+
"step": 970
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 0.7002500893176135,
|
| 764 |
+
"grad_norm": 3.8845393657684326,
|
| 765 |
+
"learning_rate": 9.467333346591632e-05,
|
| 766 |
+
"loss": 0.3027,
|
| 767 |
+
"step": 980
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 0.707395498392283,
|
| 771 |
+
"grad_norm": 1.3295674324035645,
|
| 772 |
+
"learning_rate": 9.448500422148364e-05,
|
| 773 |
+
"loss": 0.3005,
|
| 774 |
+
"step": 990
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 0.7145409074669525,
|
| 778 |
+
"grad_norm": 1.0146369934082031,
|
| 779 |
+
"learning_rate": 9.429359734349863e-05,
|
| 780 |
+
"loss": 0.294,
|
| 781 |
+
"step": 1000
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 0.7145409074669525,
|
| 785 |
+
"eval_news_finetune_val_loss": 0.3208242654800415,
|
| 786 |
+
"eval_news_finetune_val_runtime": 1003.2491,
|
| 787 |
+
"eval_news_finetune_val_samples_per_second": 1.395,
|
| 788 |
+
"eval_news_finetune_val_steps_per_second": 1.395,
|
| 789 |
+
"step": 1000
|
| 790 |
}
|
| 791 |
],
|
| 792 |
"logging_steps": 10,
|
|
|
|
| 806 |
"attributes": {}
|
| 807 |
}
|
| 808 |
},
|
| 809 |
+
"total_flos": 2.760906963972096e+16,
|
| 810 |
"train_batch_size": 1,
|
| 811 |
"trial_name": null,
|
| 812 |
"trial_params": null
|