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 319876032
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a007ef3819d0a180b2cc640168c0b397f14a7b2be24311a5b244e8f015424f6d
|
| 3 |
size 319876032
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 162933396
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:206d31fd0e94ad4b5ee8616fb918d46b4119f29e50d8445063f977fbd1c44d08
|
| 3 |
size 162933396
|
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 sha256:68fb56de6c6277dcb224030074007cc7a4f8ca144925b66a98fada92c578b58a
|
| 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 sha256:49d60a69e2379be2053e816cbaff31e6c931b5922dd86c71c9eaf473299cbf62
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 1.
|
| 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": 8.005,
|
| 374 |
"eval_steps_per_second": 2.003,
|
| 375 |
"step": 50
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 376 |
}
|
| 377 |
],
|
| 378 |
"logging_steps": 1,
|
|
@@ -396,12 +754,12 @@
|
|
| 396 |
"should_evaluate": false,
|
| 397 |
"should_log": false,
|
| 398 |
"should_save": true,
|
| 399 |
-
"should_training_stop":
|
| 400 |
},
|
| 401 |
"attributes": {}
|
| 402 |
}
|
| 403 |
},
|
| 404 |
-
"total_flos":
|
| 405 |
"train_batch_size": 8,
|
| 406 |
"trial_name": null,
|
| 407 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 1.5987274646759033,
|
| 3 |
+
"best_model_checkpoint": "miner_id_24/checkpoint-100",
|
| 4 |
+
"epoch": 0.03712641544458883,
|
| 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": 8.005,
|
| 374 |
"eval_steps_per_second": 2.003,
|
| 375 |
"step": 50
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.0189344718767403,
|
| 379 |
+
"grad_norm": 5.353964328765869,
|
| 380 |
+
"learning_rate": 5.695865504800327e-05,
|
| 381 |
+
"loss": 8.9486,
|
| 382 |
+
"step": 51
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 0.01930573603118619,
|
| 386 |
+
"grad_norm": 6.079212188720703,
|
| 387 |
+
"learning_rate": 5.522642316338268e-05,
|
| 388 |
+
"loss": 9.9081,
|
| 389 |
+
"step": 52
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 0.019677000185632076,
|
| 393 |
+
"grad_norm": 7.1975932121276855,
|
| 394 |
+
"learning_rate": 5.348782368720626e-05,
|
| 395 |
+
"loss": 8.2391,
|
| 396 |
+
"step": 53
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 0.020048264340077965,
|
| 400 |
+
"grad_norm": 6.638981342315674,
|
| 401 |
+
"learning_rate": 5.174497483512506e-05,
|
| 402 |
+
"loss": 8.5821,
|
| 403 |
+
"step": 54
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 0.020419528494523855,
|
| 407 |
+
"grad_norm": 9.038630485534668,
|
| 408 |
+
"learning_rate": 5e-05,
|
| 409 |
+
"loss": 9.6055,
|
| 410 |
+
"step": 55
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.02079079264896974,
|
| 414 |
+
"grad_norm": 9.090941429138184,
|
| 415 |
+
"learning_rate": 4.825502516487497e-05,
|
| 416 |
+
"loss": 9.724,
|
| 417 |
+
"step": 56
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 0.02116205680341563,
|
| 421 |
+
"grad_norm": 7.1579108238220215,
|
| 422 |
+
"learning_rate": 4.6512176312793736e-05,
|
| 423 |
+
"loss": 7.7207,
|
| 424 |
+
"step": 57
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 0.021533320957861517,
|
| 428 |
+
"grad_norm": 8.75041389465332,
|
| 429 |
+
"learning_rate": 4.477357683661734e-05,
|
| 430 |
+
"loss": 8.7275,
|
| 431 |
+
"step": 58
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 0.021904585112307407,
|
| 435 |
+
"grad_norm": 6.073078155517578,
|
| 436 |
+
"learning_rate": 4.3041344951996746e-05,
|
| 437 |
+
"loss": 7.8166,
|
| 438 |
+
"step": 59
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 0.022275849266753293,
|
| 442 |
+
"grad_norm": 6.027640342712402,
|
| 443 |
+
"learning_rate": 4.131759111665349e-05,
|
| 444 |
+
"loss": 6.9793,
|
| 445 |
+
"step": 60
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 0.022647113421199183,
|
| 449 |
+
"grad_norm": 6.702777862548828,
|
| 450 |
+
"learning_rate": 3.960441545911204e-05,
|
| 451 |
+
"loss": 7.5993,
|
| 452 |
+
"step": 61
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.023018377575645073,
|
| 456 |
+
"grad_norm": 6.559276103973389,
|
| 457 |
+
"learning_rate": 3.790390522001662e-05,
|
| 458 |
+
"loss": 6.5231,
|
| 459 |
+
"step": 62
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 0.02338964173009096,
|
| 463 |
+
"grad_norm": 6.18765926361084,
|
| 464 |
+
"learning_rate": 3.6218132209150045e-05,
|
| 465 |
+
"loss": 6.9596,
|
| 466 |
+
"step": 63
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 0.02376090588453685,
|
| 470 |
+
"grad_norm": 5.145318508148193,
|
| 471 |
+
"learning_rate": 3.4549150281252636e-05,
|
| 472 |
+
"loss": 6.4935,
|
| 473 |
+
"step": 64
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.024132170038982735,
|
| 477 |
+
"grad_norm": 4.60317850112915,
|
| 478 |
+
"learning_rate": 3.289899283371657e-05,
|
| 479 |
+
"loss": 6.1899,
|
| 480 |
+
"step": 65
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 0.024503434193428625,
|
| 484 |
+
"grad_norm": 4.280679225921631,
|
| 485 |
+
"learning_rate": 3.12696703292044e-05,
|
| 486 |
+
"loss": 6.1858,
|
| 487 |
+
"step": 66
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 0.02487469834787451,
|
| 491 |
+
"grad_norm": 4.138360023498535,
|
| 492 |
+
"learning_rate": 2.9663167846209998e-05,
|
| 493 |
+
"loss": 6.1128,
|
| 494 |
+
"step": 67
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.0252459625023204,
|
| 498 |
+
"grad_norm": 4.611503601074219,
|
| 499 |
+
"learning_rate": 2.8081442660546125e-05,
|
| 500 |
+
"loss": 6.3028,
|
| 501 |
+
"step": 68
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.02561722665676629,
|
| 505 |
+
"grad_norm": 4.154057502746582,
|
| 506 |
+
"learning_rate": 2.6526421860705473e-05,
|
| 507 |
+
"loss": 5.5962,
|
| 508 |
+
"step": 69
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 0.025988490811212177,
|
| 512 |
+
"grad_norm": 3.594250202178955,
|
| 513 |
+
"learning_rate": 2.500000000000001e-05,
|
| 514 |
+
"loss": 5.6007,
|
| 515 |
+
"step": 70
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.026359754965658067,
|
| 519 |
+
"grad_norm": 4.3070454597473145,
|
| 520 |
+
"learning_rate": 2.350403678833976e-05,
|
| 521 |
+
"loss": 6.1225,
|
| 522 |
+
"step": 71
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 0.026731019120103953,
|
| 526 |
+
"grad_norm": 4.306888580322266,
|
| 527 |
+
"learning_rate": 2.2040354826462668e-05,
|
| 528 |
+
"loss": 6.8154,
|
| 529 |
+
"step": 72
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 0.027102283274549843,
|
| 533 |
+
"grad_norm": 3.854240655899048,
|
| 534 |
+
"learning_rate": 2.061073738537635e-05,
|
| 535 |
+
"loss": 5.8671,
|
| 536 |
+
"step": 73
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 0.02747354742899573,
|
| 540 |
+
"grad_norm": 3.8052146434783936,
|
| 541 |
+
"learning_rate": 1.9216926233717085e-05,
|
| 542 |
+
"loss": 6.1667,
|
| 543 |
+
"step": 74
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 0.02784481158344162,
|
| 547 |
+
"grad_norm": 3.809180974960327,
|
| 548 |
+
"learning_rate": 1.7860619515673033e-05,
|
| 549 |
+
"loss": 5.8298,
|
| 550 |
+
"step": 75
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 0.028216075737887508,
|
| 554 |
+
"grad_norm": 3.352525234222412,
|
| 555 |
+
"learning_rate": 1.6543469682057106e-05,
|
| 556 |
+
"loss": 5.3726,
|
| 557 |
+
"step": 76
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 0.028587339892333394,
|
| 561 |
+
"grad_norm": 3.5558321475982666,
|
| 562 |
+
"learning_rate": 1.526708147705013e-05,
|
| 563 |
+
"loss": 6.2531,
|
| 564 |
+
"step": 77
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 0.028958604046779284,
|
| 568 |
+
"grad_norm": 3.92938494682312,
|
| 569 |
+
"learning_rate": 1.4033009983067452e-05,
|
| 570 |
+
"loss": 5.5628,
|
| 571 |
+
"step": 78
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 0.02932986820122517,
|
| 575 |
+
"grad_norm": 3.702071189880371,
|
| 576 |
+
"learning_rate": 1.2842758726130283e-05,
|
| 577 |
+
"loss": 5.8864,
|
| 578 |
+
"step": 79
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.02970113235567106,
|
| 582 |
+
"grad_norm": 4.398648262023926,
|
| 583 |
+
"learning_rate": 1.1697777844051105e-05,
|
| 584 |
+
"loss": 6.06,
|
| 585 |
+
"step": 80
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.03007239651011695,
|
| 589 |
+
"grad_norm": 3.4324252605438232,
|
| 590 |
+
"learning_rate": 1.0599462319663905e-05,
|
| 591 |
+
"loss": 5.1271,
|
| 592 |
+
"step": 81
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 0.030443660664562836,
|
| 596 |
+
"grad_norm": 3.377044439315796,
|
| 597 |
+
"learning_rate": 9.549150281252633e-06,
|
| 598 |
+
"loss": 5.4419,
|
| 599 |
+
"step": 82
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 0.030814924819008726,
|
| 603 |
+
"grad_norm": 3.836601734161377,
|
| 604 |
+
"learning_rate": 8.548121372247918e-06,
|
| 605 |
+
"loss": 5.5232,
|
| 606 |
+
"step": 83
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 0.031186188973454612,
|
| 610 |
+
"grad_norm": 4.520117282867432,
|
| 611 |
+
"learning_rate": 7.597595192178702e-06,
|
| 612 |
+
"loss": 6.1833,
|
| 613 |
+
"step": 84
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 0.0315574531279005,
|
| 617 |
+
"grad_norm": 3.46451473236084,
|
| 618 |
+
"learning_rate": 6.698729810778065e-06,
|
| 619 |
+
"loss": 5.3309,
|
| 620 |
+
"step": 85
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 0.03192871728234639,
|
| 624 |
+
"grad_norm": 4.2184157371521,
|
| 625 |
+
"learning_rate": 5.852620357053651e-06,
|
| 626 |
+
"loss": 5.6528,
|
| 627 |
+
"step": 86
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 0.032299981436792274,
|
| 631 |
+
"grad_norm": 4.261045455932617,
|
| 632 |
+
"learning_rate": 5.060297685041659e-06,
|
| 633 |
+
"loss": 6.3858,
|
| 634 |
+
"step": 87
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 0.032671245591238164,
|
| 638 |
+
"grad_norm": 3.678689479827881,
|
| 639 |
+
"learning_rate": 4.322727117869951e-06,
|
| 640 |
+
"loss": 5.5889,
|
| 641 |
+
"step": 88
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 0.033042509745684054,
|
| 645 |
+
"grad_norm": 4.6800665855407715,
|
| 646 |
+
"learning_rate": 3.6408072716606346e-06,
|
| 647 |
+
"loss": 6.5155,
|
| 648 |
+
"step": 89
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 0.033413773900129944,
|
| 652 |
+
"grad_norm": 4.469486713409424,
|
| 653 |
+
"learning_rate": 3.0153689607045845e-06,
|
| 654 |
+
"loss": 6.2689,
|
| 655 |
+
"step": 90
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 0.03378503805457583,
|
| 659 |
+
"grad_norm": 3.791959762573242,
|
| 660 |
+
"learning_rate": 2.4471741852423237e-06,
|
| 661 |
+
"loss": 6.4441,
|
| 662 |
+
"step": 91
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 0.034156302209021716,
|
| 666 |
+
"grad_norm": 3.8623321056365967,
|
| 667 |
+
"learning_rate": 1.9369152030840556e-06,
|
| 668 |
+
"loss": 6.4281,
|
| 669 |
+
"step": 92
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 0.034527566363467606,
|
| 673 |
+
"grad_norm": 6.107036113739014,
|
| 674 |
+
"learning_rate": 1.4852136862001764e-06,
|
| 675 |
+
"loss": 6.3856,
|
| 676 |
+
"step": 93
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 0.034898830517913496,
|
| 680 |
+
"grad_norm": 3.526797294616699,
|
| 681 |
+
"learning_rate": 1.0926199633097157e-06,
|
| 682 |
+
"loss": 5.6764,
|
| 683 |
+
"step": 94
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 0.035270094672359385,
|
| 687 |
+
"grad_norm": 4.093867301940918,
|
| 688 |
+
"learning_rate": 7.596123493895991e-07,
|
| 689 |
+
"loss": 5.6925,
|
| 690 |
+
"step": 95
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 0.035641358826805275,
|
| 694 |
+
"grad_norm": 3.6119606494903564,
|
| 695 |
+
"learning_rate": 4.865965629214819e-07,
|
| 696 |
+
"loss": 5.3956,
|
| 697 |
+
"step": 96
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 0.03601262298125116,
|
| 701 |
+
"grad_norm": 4.349273681640625,
|
| 702 |
+
"learning_rate": 2.7390523158633554e-07,
|
| 703 |
+
"loss": 6.1551,
|
| 704 |
+
"step": 97
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 0.03638388713569705,
|
| 708 |
+
"grad_norm": 3.9628746509552,
|
| 709 |
+
"learning_rate": 1.2179748700879012e-07,
|
| 710 |
+
"loss": 5.6607,
|
| 711 |
+
"step": 98
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 0.03675515129014294,
|
| 715 |
+
"grad_norm": 4.282620906829834,
|
| 716 |
+
"learning_rate": 3.04586490452119e-08,
|
| 717 |
+
"loss": 6.0429,
|
| 718 |
+
"step": 99
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 0.03712641544458883,
|
| 722 |
+
"grad_norm": 4.604611873626709,
|
| 723 |
+
"learning_rate": 0.0,
|
| 724 |
+
"loss": 6.4466,
|
| 725 |
+
"step": 100
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 0.03712641544458883,
|
| 729 |
+
"eval_loss": 1.5987274646759033,
|
| 730 |
+
"eval_runtime": 565.1895,
|
| 731 |
+
"eval_samples_per_second": 8.027,
|
| 732 |
+
"eval_steps_per_second": 2.008,
|
| 733 |
+
"step": 100
|
| 734 |
}
|
| 735 |
],
|
| 736 |
"logging_steps": 1,
|
|
|
|
| 754 |
"should_evaluate": false,
|
| 755 |
"should_log": false,
|
| 756 |
"should_save": true,
|
| 757 |
+
"should_training_stop": true
|
| 758 |
},
|
| 759 |
"attributes": {}
|
| 760 |
}
|
| 761 |
},
|
| 762 |
+
"total_flos": 1.5777468946317312e+17,
|
| 763 |
"train_batch_size": 8,
|
| 764 |
"trial_name": null,
|
| 765 |
"trial_params": null
|