Training in progress, step 100, 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 sha256:907ab6a0ae89db772ae066fb2052a5616b7b71563fa71453ebf2c8e2f0fcd5b2
|
| 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 sha256:74815f4c6901145dc008341c2caf282cceddf39d2dc5ef73927e806423c85a5e
|
| 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 sha256:2cf2b2d7ca16a98ed374235967d4dea68bc581943625ddd333bc11a0a503cc75
|
| 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 sha256:4e0244c146d76bf610ae39789eea36d0bff336b81d211db008e020e66921060c
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "miner_id_24/checkpoint-
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 50,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -373,6 +373,364 @@
|
|
| 373 |
"eval_samples_per_second": 25.782,
|
| 374 |
"eval_steps_per_second": 12.899,
|
| 375 |
"step": 50
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 376 |
}
|
| 377 |
],
|
| 378 |
"logging_steps": 1,
|
|
@@ -401,7 +759,7 @@
|
|
| 401 |
"attributes": {}
|
| 402 |
}
|
| 403 |
},
|
| 404 |
-
"total_flos":
|
| 405 |
"train_batch_size": 8,
|
| 406 |
"trial_name": null,
|
| 407 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.6289290189743042,
|
| 3 |
+
"best_model_checkpoint": "miner_id_24/checkpoint-100",
|
| 4 |
+
"epoch": 0.09854644000985464,
|
| 5 |
"eval_steps": 50,
|
| 6 |
+
"global_step": 100,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 373 |
"eval_samples_per_second": 25.782,
|
| 374 |
"eval_steps_per_second": 12.899,
|
| 375 |
"step": 50
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.05025868440502587,
|
| 379 |
+
"grad_norm": 0.12230159342288971,
|
| 380 |
+
"learning_rate": 9.285836503510562e-05,
|
| 381 |
+
"loss": 0.0571,
|
| 382 |
+
"step": 51
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 0.051244148805124415,
|
| 386 |
+
"grad_norm": 0.11277984827756882,
|
| 387 |
+
"learning_rate": 9.24024048078213e-05,
|
| 388 |
+
"loss": 0.1799,
|
| 389 |
+
"step": 52
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 0.052229613205222965,
|
| 393 |
+
"grad_norm": 0.1283257156610489,
|
| 394 |
+
"learning_rate": 9.193352839727121e-05,
|
| 395 |
+
"loss": 0.2379,
|
| 396 |
+
"step": 53
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 0.05321507760532151,
|
| 400 |
+
"grad_norm": 0.1701822280883789,
|
| 401 |
+
"learning_rate": 9.145187862775209e-05,
|
| 402 |
+
"loss": 0.3356,
|
| 403 |
+
"step": 54
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 0.05420054200542006,
|
| 407 |
+
"grad_norm": 0.20211534202098846,
|
| 408 |
+
"learning_rate": 9.09576022144496e-05,
|
| 409 |
+
"loss": 0.4511,
|
| 410 |
+
"step": 55
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.0551860064055186,
|
| 414 |
+
"grad_norm": 0.20168742537498474,
|
| 415 |
+
"learning_rate": 9.045084971874738e-05,
|
| 416 |
+
"loss": 0.558,
|
| 417 |
+
"step": 56
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 0.05617147080561715,
|
| 421 |
+
"grad_norm": 0.1796450912952423,
|
| 422 |
+
"learning_rate": 8.993177550236464e-05,
|
| 423 |
+
"loss": 0.5004,
|
| 424 |
+
"step": 57
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 0.057156935205715693,
|
| 428 |
+
"grad_norm": 0.21156959235668182,
|
| 429 |
+
"learning_rate": 8.940053768033609e-05,
|
| 430 |
+
"loss": 0.6535,
|
| 431 |
+
"step": 58
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 0.05814239960581424,
|
| 435 |
+
"grad_norm": 0.19509942829608917,
|
| 436 |
+
"learning_rate": 8.885729807284856e-05,
|
| 437 |
+
"loss": 0.5512,
|
| 438 |
+
"step": 59
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 0.059127864005912786,
|
| 442 |
+
"grad_norm": 0.22828419506549835,
|
| 443 |
+
"learning_rate": 8.83022221559489e-05,
|
| 444 |
+
"loss": 0.8388,
|
| 445 |
+
"step": 60
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 0.060113328406011336,
|
| 449 |
+
"grad_norm": 0.20433418452739716,
|
| 450 |
+
"learning_rate": 8.773547901113862e-05,
|
| 451 |
+
"loss": 0.5995,
|
| 452 |
+
"step": 61
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.06109879280610988,
|
| 456 |
+
"grad_norm": 0.20512636005878448,
|
| 457 |
+
"learning_rate": 8.715724127386972e-05,
|
| 458 |
+
"loss": 0.6224,
|
| 459 |
+
"step": 62
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 0.06208425720620843,
|
| 463 |
+
"grad_norm": 0.21725022792816162,
|
| 464 |
+
"learning_rate": 8.656768508095853e-05,
|
| 465 |
+
"loss": 0.6128,
|
| 466 |
+
"step": 63
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 0.06306972160630697,
|
| 470 |
+
"grad_norm": 0.2174377739429474,
|
| 471 |
+
"learning_rate": 8.596699001693255e-05,
|
| 472 |
+
"loss": 0.6843,
|
| 473 |
+
"step": 64
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.06405518600640552,
|
| 477 |
+
"grad_norm": 0.2236977070569992,
|
| 478 |
+
"learning_rate": 8.535533905932738e-05,
|
| 479 |
+
"loss": 0.6858,
|
| 480 |
+
"step": 65
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 0.06504065040650407,
|
| 484 |
+
"grad_norm": 0.27690693736076355,
|
| 485 |
+
"learning_rate": 8.473291852294987e-05,
|
| 486 |
+
"loss": 0.9392,
|
| 487 |
+
"step": 66
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 0.06602611480660261,
|
| 491 |
+
"grad_norm": 0.22860707342624664,
|
| 492 |
+
"learning_rate": 8.409991800312493e-05,
|
| 493 |
+
"loss": 0.7385,
|
| 494 |
+
"step": 67
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.06701157920670116,
|
| 498 |
+
"grad_norm": 0.2298586070537567,
|
| 499 |
+
"learning_rate": 8.345653031794292e-05,
|
| 500 |
+
"loss": 0.7546,
|
| 501 |
+
"step": 68
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.0679970436067997,
|
| 505 |
+
"grad_norm": 0.2377099245786667,
|
| 506 |
+
"learning_rate": 8.280295144952536e-05,
|
| 507 |
+
"loss": 0.6743,
|
| 508 |
+
"step": 69
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 0.06898250800689826,
|
| 512 |
+
"grad_norm": 0.21426402032375336,
|
| 513 |
+
"learning_rate": 8.213938048432697e-05,
|
| 514 |
+
"loss": 0.5414,
|
| 515 |
+
"step": 70
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.0699679724069968,
|
| 519 |
+
"grad_norm": 0.23989328742027283,
|
| 520 |
+
"learning_rate": 8.146601955249188e-05,
|
| 521 |
+
"loss": 0.606,
|
| 522 |
+
"step": 71
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 0.07095343680709534,
|
| 526 |
+
"grad_norm": 0.2526426613330841,
|
| 527 |
+
"learning_rate": 8.07830737662829e-05,
|
| 528 |
+
"loss": 0.6877,
|
| 529 |
+
"step": 72
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 0.07193890120719389,
|
| 533 |
+
"grad_norm": 0.25667741894721985,
|
| 534 |
+
"learning_rate": 8.009075115760243e-05,
|
| 535 |
+
"loss": 0.8371,
|
| 536 |
+
"step": 73
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 0.07292436560729244,
|
| 540 |
+
"grad_norm": 0.2683584690093994,
|
| 541 |
+
"learning_rate": 7.938926261462366e-05,
|
| 542 |
+
"loss": 0.9071,
|
| 543 |
+
"step": 74
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 0.07390983000739099,
|
| 547 |
+
"grad_norm": 0.2747768461704254,
|
| 548 |
+
"learning_rate": 7.86788218175523e-05,
|
| 549 |
+
"loss": 0.7603,
|
| 550 |
+
"step": 75
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 0.07489529440748953,
|
| 554 |
+
"grad_norm": 0.2705404758453369,
|
| 555 |
+
"learning_rate": 7.795964517353735e-05,
|
| 556 |
+
"loss": 0.7255,
|
| 557 |
+
"step": 76
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 0.07588075880758807,
|
| 561 |
+
"grad_norm": 0.2925010323524475,
|
| 562 |
+
"learning_rate": 7.723195175075136e-05,
|
| 563 |
+
"loss": 0.7684,
|
| 564 |
+
"step": 77
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 0.07686622320768663,
|
| 568 |
+
"grad_norm": 0.2544151544570923,
|
| 569 |
+
"learning_rate": 7.649596321166024e-05,
|
| 570 |
+
"loss": 0.6322,
|
| 571 |
+
"step": 78
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 0.07785168760778517,
|
| 575 |
+
"grad_norm": 0.27934426069259644,
|
| 576 |
+
"learning_rate": 7.575190374550272e-05,
|
| 577 |
+
"loss": 0.7391,
|
| 578 |
+
"step": 79
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.07883715200788372,
|
| 582 |
+
"grad_norm": 0.3051762878894806,
|
| 583 |
+
"learning_rate": 7.500000000000001e-05,
|
| 584 |
+
"loss": 0.7494,
|
| 585 |
+
"step": 80
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.07982261640798226,
|
| 589 |
+
"grad_norm": 0.28656357526779175,
|
| 590 |
+
"learning_rate": 7.424048101231686e-05,
|
| 591 |
+
"loss": 0.7278,
|
| 592 |
+
"step": 81
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 0.08080808080808081,
|
| 596 |
+
"grad_norm": 0.3144802153110504,
|
| 597 |
+
"learning_rate": 7.347357813929454e-05,
|
| 598 |
+
"loss": 0.719,
|
| 599 |
+
"step": 82
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 0.08179354520817936,
|
| 603 |
+
"grad_norm": 0.3164990246295929,
|
| 604 |
+
"learning_rate": 7.269952498697734e-05,
|
| 605 |
+
"loss": 0.7447,
|
| 606 |
+
"step": 83
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 0.0827790096082779,
|
| 610 |
+
"grad_norm": 0.28684890270233154,
|
| 611 |
+
"learning_rate": 7.191855733945387e-05,
|
| 612 |
+
"loss": 0.7701,
|
| 613 |
+
"step": 84
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 0.08376447400837644,
|
| 617 |
+
"grad_norm": 0.328673779964447,
|
| 618 |
+
"learning_rate": 7.113091308703498e-05,
|
| 619 |
+
"loss": 0.6504,
|
| 620 |
+
"step": 85
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 0.084749938408475,
|
| 624 |
+
"grad_norm": 0.3479191064834595,
|
| 625 |
+
"learning_rate": 7.033683215379002e-05,
|
| 626 |
+
"loss": 0.816,
|
| 627 |
+
"step": 86
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 0.08573540280857354,
|
| 631 |
+
"grad_norm": 0.30887213349342346,
|
| 632 |
+
"learning_rate": 6.953655642446368e-05,
|
| 633 |
+
"loss": 0.6048,
|
| 634 |
+
"step": 87
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 0.08672086720867209,
|
| 638 |
+
"grad_norm": 0.3754432797431946,
|
| 639 |
+
"learning_rate": 6.873032967079561e-05,
|
| 640 |
+
"loss": 0.728,
|
| 641 |
+
"step": 88
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 0.08770633160877063,
|
| 645 |
+
"grad_norm": 0.3510769009590149,
|
| 646 |
+
"learning_rate": 6.7918397477265e-05,
|
| 647 |
+
"loss": 0.7624,
|
| 648 |
+
"step": 89
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 0.08869179600886919,
|
| 652 |
+
"grad_norm": 0.37762296199798584,
|
| 653 |
+
"learning_rate": 6.710100716628344e-05,
|
| 654 |
+
"loss": 0.6676,
|
| 655 |
+
"step": 90
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 0.08967726040896773,
|
| 659 |
+
"grad_norm": 0.34411317110061646,
|
| 660 |
+
"learning_rate": 6.627840772285784e-05,
|
| 661 |
+
"loss": 0.6249,
|
| 662 |
+
"step": 91
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 0.09066272480906627,
|
| 666 |
+
"grad_norm": 0.3781382739543915,
|
| 667 |
+
"learning_rate": 6.545084971874738e-05,
|
| 668 |
+
"loss": 0.5981,
|
| 669 |
+
"step": 92
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 0.09164818920916482,
|
| 673 |
+
"grad_norm": 0.4274924099445343,
|
| 674 |
+
"learning_rate": 6.461858523613684e-05,
|
| 675 |
+
"loss": 0.8801,
|
| 676 |
+
"step": 93
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 0.09263365360926337,
|
| 680 |
+
"grad_norm": 0.4174174964427948,
|
| 681 |
+
"learning_rate": 6.378186779084995e-05,
|
| 682 |
+
"loss": 0.6756,
|
| 683 |
+
"step": 94
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 0.09361911800936192,
|
| 687 |
+
"grad_norm": 0.47355586290359497,
|
| 688 |
+
"learning_rate": 6.294095225512603e-05,
|
| 689 |
+
"loss": 0.8179,
|
| 690 |
+
"step": 95
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 0.09460458240946046,
|
| 694 |
+
"grad_norm": 0.4370453953742981,
|
| 695 |
+
"learning_rate": 6.209609477998338e-05,
|
| 696 |
+
"loss": 0.7869,
|
| 697 |
+
"step": 96
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 0.095590046809559,
|
| 701 |
+
"grad_norm": 0.4838176667690277,
|
| 702 |
+
"learning_rate": 6.124755271719325e-05,
|
| 703 |
+
"loss": 0.8112,
|
| 704 |
+
"step": 97
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 0.09657551120965756,
|
| 708 |
+
"grad_norm": 0.5100703835487366,
|
| 709 |
+
"learning_rate": 6.0395584540887963e-05,
|
| 710 |
+
"loss": 0.7746,
|
| 711 |
+
"step": 98
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 0.0975609756097561,
|
| 715 |
+
"grad_norm": 0.6169785261154175,
|
| 716 |
+
"learning_rate": 5.9540449768827246e-05,
|
| 717 |
+
"loss": 0.8988,
|
| 718 |
+
"step": 99
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 0.09854644000985464,
|
| 722 |
+
"grad_norm": 0.8071909546852112,
|
| 723 |
+
"learning_rate": 5.868240888334653e-05,
|
| 724 |
+
"loss": 0.6597,
|
| 725 |
+
"step": 100
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 0.09854644000985464,
|
| 729 |
+
"eval_loss": 0.6289290189743042,
|
| 730 |
+
"eval_runtime": 66.4133,
|
| 731 |
+
"eval_samples_per_second": 25.733,
|
| 732 |
+
"eval_steps_per_second": 12.874,
|
| 733 |
+
"step": 100
|
| 734 |
}
|
| 735 |
],
|
| 736 |
"logging_steps": 1,
|
|
|
|
| 759 |
"attributes": {}
|
| 760 |
}
|
| 761 |
},
|
| 762 |
+
"total_flos": 3.09564631154688e+16,
|
| 763 |
"train_batch_size": 8,
|
| 764 |
"trial_name": null,
|
| 765 |
"trial_params": null
|