Training in progress, step 3438, checkpoint
Browse files
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 516810008
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:88b916425e9c0934a0f7062407f1b73535e218ac98e877fcb8ee9b8017fd579d
|
| 3 |
size 516810008
|
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 sha256:9798aa39906f12cd89a0a56a12db91e16577b6948826e0e6863eded99960b381
|
| 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 sha256:805c0bccc0907c9d87d7cceb385924edbfe5fd82910b2d6ec679bbe7866a153f
|
| 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": 1.
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -23437,6 +23437,650 @@
|
|
| 23437 |
"learning_rate": 1.8731449773342625e-07,
|
| 23438 |
"loss": 0.0,
|
| 23439 |
"step": 3346
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 23440 |
}
|
| 23441 |
],
|
| 23442 |
"logging_steps": 1,
|
|
@@ -23451,12 +24095,12 @@
|
|
| 23451 |
"should_evaluate": false,
|
| 23452 |
"should_log": false,
|
| 23453 |
"should_save": true,
|
| 23454 |
-
"should_training_stop":
|
| 23455 |
},
|
| 23456 |
"attributes": {}
|
| 23457 |
}
|
| 23458 |
},
|
| 23459 |
-
"total_flos": 3.
|
| 23460 |
"train_batch_size": 2,
|
| 23461 |
"trial_name": null,
|
| 23462 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.9997091755125782,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 3438,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 23437 |
"learning_rate": 1.8731449773342625e-07,
|
| 23438 |
"loss": 0.0,
|
| 23439 |
"step": 3346
|
| 23440 |
+
},
|
| 23441 |
+
{
|
| 23442 |
+
"epoch": 1.946779118801803,
|
| 23443 |
+
"grad_norm": NaN,
|
| 23444 |
+
"learning_rate": 1.8326705005886624e-07,
|
| 23445 |
+
"loss": 0.0,
|
| 23446 |
+
"step": 3347
|
| 23447 |
+
},
|
| 23448 |
+
{
|
| 23449 |
+
"epoch": 1.947360767776647,
|
| 23450 |
+
"grad_norm": NaN,
|
| 23451 |
+
"learning_rate": 1.7926372916975965e-07,
|
| 23452 |
+
"loss": 0.0,
|
| 23453 |
+
"step": 3348
|
| 23454 |
+
},
|
| 23455 |
+
{
|
| 23456 |
+
"epoch": 1.9479424167514905,
|
| 23457 |
+
"grad_norm": NaN,
|
| 23458 |
+
"learning_rate": 1.7530453861218098e-07,
|
| 23459 |
+
"loss": 0.0,
|
| 23460 |
+
"step": 3349
|
| 23461 |
+
},
|
| 23462 |
+
{
|
| 23463 |
+
"epoch": 1.948524065726334,
|
| 23464 |
+
"grad_norm": NaN,
|
| 23465 |
+
"learning_rate": 1.7138948189311387e-07,
|
| 23466 |
+
"loss": 0.0,
|
| 23467 |
+
"step": 3350
|
| 23468 |
+
},
|
| 23469 |
+
{
|
| 23470 |
+
"epoch": 1.9491057147011779,
|
| 23471 |
+
"grad_norm": NaN,
|
| 23472 |
+
"learning_rate": 1.6751856248043985e-07,
|
| 23473 |
+
"loss": 0.0,
|
| 23474 |
+
"step": 3351
|
| 23475 |
+
},
|
| 23476 |
+
{
|
| 23477 |
+
"epoch": 1.9496873636760215,
|
| 23478 |
+
"grad_norm": NaN,
|
| 23479 |
+
"learning_rate": 1.6369178380294948e-07,
|
| 23480 |
+
"loss": 0.0,
|
| 23481 |
+
"step": 3352
|
| 23482 |
+
},
|
| 23483 |
+
{
|
| 23484 |
+
"epoch": 1.9502690126508653,
|
| 23485 |
+
"grad_norm": NaN,
|
| 23486 |
+
"learning_rate": 1.5990914925034794e-07,
|
| 23487 |
+
"loss": 0.0,
|
| 23488 |
+
"step": 3353
|
| 23489 |
+
},
|
| 23490 |
+
{
|
| 23491 |
+
"epoch": 1.9508506616257089,
|
| 23492 |
+
"grad_norm": NaN,
|
| 23493 |
+
"learning_rate": 1.561706621732162e-07,
|
| 23494 |
+
"loss": 0.0,
|
| 23495 |
+
"step": 3354
|
| 23496 |
+
},
|
| 23497 |
+
{
|
| 23498 |
+
"epoch": 1.9514323106005524,
|
| 23499 |
+
"grad_norm": NaN,
|
| 23500 |
+
"learning_rate": 1.5247632588304417e-07,
|
| 23501 |
+
"loss": 0.0,
|
| 23502 |
+
"step": 3355
|
| 23503 |
+
},
|
| 23504 |
+
{
|
| 23505 |
+
"epoch": 1.9520139595753963,
|
| 23506 |
+
"grad_norm": NaN,
|
| 23507 |
+
"learning_rate": 1.488261436522087e-07,
|
| 23508 |
+
"loss": 0.0,
|
| 23509 |
+
"step": 3356
|
| 23510 |
+
},
|
| 23511 |
+
{
|
| 23512 |
+
"epoch": 1.95259560855024,
|
| 23513 |
+
"grad_norm": NaN,
|
| 23514 |
+
"learning_rate": 1.45220118713979e-07,
|
| 23515 |
+
"loss": 0.0,
|
| 23516 |
+
"step": 3357
|
| 23517 |
+
},
|
| 23518 |
+
{
|
| 23519 |
+
"epoch": 1.9531772575250836,
|
| 23520 |
+
"grad_norm": NaN,
|
| 23521 |
+
"learning_rate": 1.4165825426250557e-07,
|
| 23522 |
+
"loss": 0.0,
|
| 23523 |
+
"step": 3358
|
| 23524 |
+
},
|
| 23525 |
+
{
|
| 23526 |
+
"epoch": 1.9537589064999272,
|
| 23527 |
+
"grad_norm": NaN,
|
| 23528 |
+
"learning_rate": 1.3814055345282573e-07,
|
| 23529 |
+
"loss": 0.0,
|
| 23530 |
+
"step": 3359
|
| 23531 |
+
},
|
| 23532 |
+
{
|
| 23533 |
+
"epoch": 1.9543405554747708,
|
| 23534 |
+
"grad_norm": NaN,
|
| 23535 |
+
"learning_rate": 1.3466701940085259e-07,
|
| 23536 |
+
"loss": 0.0,
|
| 23537 |
+
"step": 3360
|
| 23538 |
+
},
|
| 23539 |
+
{
|
| 23540 |
+
"epoch": 1.9549222044496146,
|
| 23541 |
+
"grad_norm": NaN,
|
| 23542 |
+
"learning_rate": 1.3123765518339716e-07,
|
| 23543 |
+
"loss": 0.0,
|
| 23544 |
+
"step": 3361
|
| 23545 |
+
},
|
| 23546 |
+
{
|
| 23547 |
+
"epoch": 1.9555038534244584,
|
| 23548 |
+
"grad_norm": NaN,
|
| 23549 |
+
"learning_rate": 1.2785246383811288e-07,
|
| 23550 |
+
"loss": 0.0,
|
| 23551 |
+
"step": 3362
|
| 23552 |
+
},
|
| 23553 |
+
{
|
| 23554 |
+
"epoch": 1.956085502399302,
|
| 23555 |
+
"grad_norm": NaN,
|
| 23556 |
+
"learning_rate": 1.2451144836355123e-07,
|
| 23557 |
+
"loss": 0.0,
|
| 23558 |
+
"step": 3363
|
| 23559 |
+
},
|
| 23560 |
+
{
|
| 23561 |
+
"epoch": 1.9566671513741456,
|
| 23562 |
+
"grad_norm": NaN,
|
| 23563 |
+
"learning_rate": 1.2121461171912262e-07,
|
| 23564 |
+
"loss": 0.0,
|
| 23565 |
+
"step": 3364
|
| 23566 |
+
},
|
| 23567 |
+
{
|
| 23568 |
+
"epoch": 1.9572488003489894,
|
| 23569 |
+
"grad_norm": NaN,
|
| 23570 |
+
"learning_rate": 1.179619568251078e-07,
|
| 23571 |
+
"loss": 0.0,
|
| 23572 |
+
"step": 3365
|
| 23573 |
+
},
|
| 23574 |
+
{
|
| 23575 |
+
"epoch": 1.9578304493238332,
|
| 23576 |
+
"grad_norm": NaN,
|
| 23577 |
+
"learning_rate": 1.147534865626465e-07,
|
| 23578 |
+
"loss": 0.0,
|
| 23579 |
+
"step": 3366
|
| 23580 |
+
},
|
| 23581 |
+
{
|
| 23582 |
+
"epoch": 1.9584120982986768,
|
| 23583 |
+
"grad_norm": NaN,
|
| 23584 |
+
"learning_rate": 1.1158920377375426e-07,
|
| 23585 |
+
"loss": 0.0,
|
| 23586 |
+
"step": 3367
|
| 23587 |
+
},
|
| 23588 |
+
{
|
| 23589 |
+
"epoch": 1.9589937472735204,
|
| 23590 |
+
"grad_norm": NaN,
|
| 23591 |
+
"learning_rate": 1.08469111261289e-07,
|
| 23592 |
+
"loss": 0.0,
|
| 23593 |
+
"step": 3368
|
| 23594 |
+
},
|
| 23595 |
+
{
|
| 23596 |
+
"epoch": 1.959575396248364,
|
| 23597 |
+
"grad_norm": NaN,
|
| 23598 |
+
"learning_rate": 1.0539321178897888e-07,
|
| 23599 |
+
"loss": 0.0,
|
| 23600 |
+
"step": 3369
|
| 23601 |
+
},
|
| 23602 |
+
{
|
| 23603 |
+
"epoch": 1.9601570452232078,
|
| 23604 |
+
"grad_norm": NaN,
|
| 23605 |
+
"learning_rate": 1.0236150808139999e-07,
|
| 23606 |
+
"loss": 0.0,
|
| 23607 |
+
"step": 3370
|
| 23608 |
+
},
|
| 23609 |
+
{
|
| 23610 |
+
"epoch": 1.9607386941980516,
|
| 23611 |
+
"grad_norm": NaN,
|
| 23612 |
+
"learning_rate": 9.937400282398201e-08,
|
| 23613 |
+
"loss": 0.0,
|
| 23614 |
+
"step": 3371
|
| 23615 |
+
},
|
| 23616 |
+
{
|
| 23617 |
+
"epoch": 1.9613203431728952,
|
| 23618 |
+
"grad_norm": NaN,
|
| 23619 |
+
"learning_rate": 9.643069866300259e-08,
|
| 23620 |
+
"loss": 0.0,
|
| 23621 |
+
"step": 3372
|
| 23622 |
+
},
|
| 23623 |
+
{
|
| 23624 |
+
"epoch": 1.9619019921477387,
|
| 23625 |
+
"grad_norm": NaN,
|
| 23626 |
+
"learning_rate": 9.353159820559287e-08,
|
| 23627 |
+
"loss": 0.0,
|
| 23628 |
+
"step": 3373
|
| 23629 |
+
},
|
| 23630 |
+
{
|
| 23631 |
+
"epoch": 1.9624836411225826,
|
| 23632 |
+
"grad_norm": NaN,
|
| 23633 |
+
"learning_rate": 9.067670401972095e-08,
|
| 23634 |
+
"loss": 0.0,
|
| 23635 |
+
"step": 3374
|
| 23636 |
+
},
|
| 23637 |
+
{
|
| 23638 |
+
"epoch": 1.9630652900974264,
|
| 23639 |
+
"grad_norm": NaN,
|
| 23640 |
+
"learning_rate": 8.786601863420286e-08,
|
| 23641 |
+
"loss": 0.0,
|
| 23642 |
+
"step": 3375
|
| 23643 |
+
},
|
| 23644 |
+
{
|
| 23645 |
+
"epoch": 1.96364693907227,
|
| 23646 |
+
"grad_norm": NaN,
|
| 23647 |
+
"learning_rate": 8.509954453869152e-08,
|
| 23648 |
+
"loss": 0.0,
|
| 23649 |
+
"step": 3376
|
| 23650 |
+
},
|
| 23651 |
+
{
|
| 23652 |
+
"epoch": 1.9642285880471135,
|
| 23653 |
+
"grad_norm": NaN,
|
| 23654 |
+
"learning_rate": 8.23772841836934e-08,
|
| 23655 |
+
"loss": 0.0,
|
| 23656 |
+
"step": 3377
|
| 23657 |
+
},
|
| 23658 |
+
{
|
| 23659 |
+
"epoch": 1.9648102370219571,
|
| 23660 |
+
"grad_norm": NaN,
|
| 23661 |
+
"learning_rate": 7.96992399805241e-08,
|
| 23662 |
+
"loss": 0.0,
|
| 23663 |
+
"step": 3378
|
| 23664 |
+
},
|
| 23665 |
+
{
|
| 23666 |
+
"epoch": 1.965391885996801,
|
| 23667 |
+
"grad_norm": NaN,
|
| 23668 |
+
"learning_rate": 7.706541430135273e-08,
|
| 23669 |
+
"loss": 0.0,
|
| 23670 |
+
"step": 3379
|
| 23671 |
+
},
|
| 23672 |
+
{
|
| 23673 |
+
"epoch": 1.9659735349716447,
|
| 23674 |
+
"grad_norm": NaN,
|
| 23675 |
+
"learning_rate": 7.447580947917975e-08,
|
| 23676 |
+
"loss": 0.0,
|
| 23677 |
+
"step": 3380
|
| 23678 |
+
},
|
| 23679 |
+
{
|
| 23680 |
+
"epoch": 1.9665551839464883,
|
| 23681 |
+
"grad_norm": NaN,
|
| 23682 |
+
"learning_rate": 7.193042780782588e-08,
|
| 23683 |
+
"loss": 0.0,
|
| 23684 |
+
"step": 3381
|
| 23685 |
+
},
|
| 23686 |
+
{
|
| 23687 |
+
"epoch": 1.967136832921332,
|
| 23688 |
+
"grad_norm": NaN,
|
| 23689 |
+
"learning_rate": 6.942927154194867e-08,
|
| 23690 |
+
"loss": 0.0,
|
| 23691 |
+
"step": 3382
|
| 23692 |
+
},
|
| 23693 |
+
{
|
| 23694 |
+
"epoch": 1.9677184818961757,
|
| 23695 |
+
"grad_norm": NaN,
|
| 23696 |
+
"learning_rate": 6.697234289703147e-08,
|
| 23697 |
+
"loss": 0.0,
|
| 23698 |
+
"step": 3383
|
| 23699 |
+
},
|
| 23700 |
+
{
|
| 23701 |
+
"epoch": 1.9683001308710193,
|
| 23702 |
+
"grad_norm": NaN,
|
| 23703 |
+
"learning_rate": 6.455964404937232e-08,
|
| 23704 |
+
"loss": 0.0,
|
| 23705 |
+
"step": 3384
|
| 23706 |
+
},
|
| 23707 |
+
{
|
| 23708 |
+
"epoch": 1.968881779845863,
|
| 23709 |
+
"grad_norm": NaN,
|
| 23710 |
+
"learning_rate": 6.219117713610056e-08,
|
| 23711 |
+
"loss": 0.0,
|
| 23712 |
+
"step": 3385
|
| 23713 |
+
},
|
| 23714 |
+
{
|
| 23715 |
+
"epoch": 1.9694634288207067,
|
| 23716 |
+
"grad_norm": NaN,
|
| 23717 |
+
"learning_rate": 5.986694425516026e-08,
|
| 23718 |
+
"loss": 0.0,
|
| 23719 |
+
"step": 3386
|
| 23720 |
+
},
|
| 23721 |
+
{
|
| 23722 |
+
"epoch": 1.9700450777955503,
|
| 23723 |
+
"grad_norm": NaN,
|
| 23724 |
+
"learning_rate": 5.7586947465315675e-08,
|
| 23725 |
+
"loss": 0.0,
|
| 23726 |
+
"step": 3387
|
| 23727 |
+
},
|
| 23728 |
+
{
|
| 23729 |
+
"epoch": 1.970626726770394,
|
| 23730 |
+
"grad_norm": NaN,
|
| 23731 |
+
"learning_rate": 5.535118878615131e-08,
|
| 23732 |
+
"loss": 0.0,
|
| 23733 |
+
"step": 3388
|
| 23734 |
+
},
|
| 23735 |
+
{
|
| 23736 |
+
"epoch": 1.9712083757452379,
|
| 23737 |
+
"grad_norm": NaN,
|
| 23738 |
+
"learning_rate": 5.315967019806078e-08,
|
| 23739 |
+
"loss": 0.0,
|
| 23740 |
+
"step": 3389
|
| 23741 |
+
},
|
| 23742 |
+
{
|
| 23743 |
+
"epoch": 1.9717900247200815,
|
| 23744 |
+
"grad_norm": NaN,
|
| 23745 |
+
"learning_rate": 5.101239364225796e-08,
|
| 23746 |
+
"loss": 0.0,
|
| 23747 |
+
"step": 3390
|
| 23748 |
+
},
|
| 23749 |
+
{
|
| 23750 |
+
"epoch": 1.972371673694925,
|
| 23751 |
+
"grad_norm": NaN,
|
| 23752 |
+
"learning_rate": 4.890936102075472e-08,
|
| 23753 |
+
"loss": 0.0,
|
| 23754 |
+
"step": 3391
|
| 23755 |
+
},
|
| 23756 |
+
{
|
| 23757 |
+
"epoch": 1.9729533226697686,
|
| 23758 |
+
"grad_norm": NaN,
|
| 23759 |
+
"learning_rate": 4.685057419638317e-08,
|
| 23760 |
+
"loss": 0.0,
|
| 23761 |
+
"step": 3392
|
| 23762 |
+
},
|
| 23763 |
+
{
|
| 23764 |
+
"epoch": 1.9735349716446124,
|
| 23765 |
+
"grad_norm": NaN,
|
| 23766 |
+
"learning_rate": 4.4836034992779e-08,
|
| 23767 |
+
"loss": 0.0,
|
| 23768 |
+
"step": 3393
|
| 23769 |
+
},
|
| 23770 |
+
{
|
| 23771 |
+
"epoch": 1.9741166206194563,
|
| 23772 |
+
"grad_norm": NaN,
|
| 23773 |
+
"learning_rate": 4.286574519438702e-08,
|
| 23774 |
+
"loss": 0.0,
|
| 23775 |
+
"step": 3394
|
| 23776 |
+
},
|
| 23777 |
+
{
|
| 23778 |
+
"epoch": 1.9746982695942998,
|
| 23779 |
+
"grad_norm": NaN,
|
| 23780 |
+
"learning_rate": 4.0939706546461175e-08,
|
| 23781 |
+
"loss": 0.0,
|
| 23782 |
+
"step": 3395
|
| 23783 |
+
},
|
| 23784 |
+
{
|
| 23785 |
+
"epoch": 1.9752799185691434,
|
| 23786 |
+
"grad_norm": NaN,
|
| 23787 |
+
"learning_rate": 3.905792075504233e-08,
|
| 23788 |
+
"loss": 0.0,
|
| 23789 |
+
"step": 3396
|
| 23790 |
+
},
|
| 23791 |
+
{
|
| 23792 |
+
"epoch": 1.9758615675439872,
|
| 23793 |
+
"grad_norm": NaN,
|
| 23794 |
+
"learning_rate": 3.722038948698603e-08,
|
| 23795 |
+
"loss": 0.0,
|
| 23796 |
+
"step": 3397
|
| 23797 |
+
},
|
| 23798 |
+
{
|
| 23799 |
+
"epoch": 1.976443216518831,
|
| 23800 |
+
"grad_norm": NaN,
|
| 23801 |
+
"learning_rate": 3.542711436995139e-08,
|
| 23802 |
+
"loss": 0.0,
|
| 23803 |
+
"step": 3398
|
| 23804 |
+
},
|
| 23805 |
+
{
|
| 23806 |
+
"epoch": 1.9770248654936746,
|
| 23807 |
+
"grad_norm": NaN,
|
| 23808 |
+
"learning_rate": 3.3678096992384446e-08,
|
| 23809 |
+
"loss": 0.0,
|
| 23810 |
+
"step": 3399
|
| 23811 |
+
},
|
| 23812 |
+
{
|
| 23813 |
+
"epoch": 1.9776065144685182,
|
| 23814 |
+
"grad_norm": NaN,
|
| 23815 |
+
"learning_rate": 3.197333890353482e-08,
|
| 23816 |
+
"loss": 0.0,
|
| 23817 |
+
"step": 3400
|
| 23818 |
+
},
|
| 23819 |
+
{
|
| 23820 |
+
"epoch": 1.9781881634433618,
|
| 23821 |
+
"grad_norm": NaN,
|
| 23822 |
+
"learning_rate": 3.031284161344461e-08,
|
| 23823 |
+
"loss": 0.0,
|
| 23824 |
+
"step": 3401
|
| 23825 |
+
},
|
| 23826 |
+
{
|
| 23827 |
+
"epoch": 1.9787698124182056,
|
| 23828 |
+
"grad_norm": NaN,
|
| 23829 |
+
"learning_rate": 2.8696606592959475e-08,
|
| 23830 |
+
"loss": 0.0,
|
| 23831 |
+
"step": 3402
|
| 23832 |
+
},
|
| 23833 |
+
{
|
| 23834 |
+
"epoch": 1.9793514613930494,
|
| 23835 |
+
"grad_norm": NaN,
|
| 23836 |
+
"learning_rate": 2.7124635273712006e-08,
|
| 23837 |
+
"loss": 0.0,
|
| 23838 |
+
"step": 3403
|
| 23839 |
+
},
|
| 23840 |
+
{
|
| 23841 |
+
"epoch": 1.979933110367893,
|
| 23842 |
+
"grad_norm": NaN,
|
| 23843 |
+
"learning_rate": 2.5596929048116168e-08,
|
| 23844 |
+
"loss": 0.0,
|
| 23845 |
+
"step": 3404
|
| 23846 |
+
},
|
| 23847 |
+
{
|
| 23848 |
+
"epoch": 1.9805147593427366,
|
| 23849 |
+
"grad_norm": NaN,
|
| 23850 |
+
"learning_rate": 2.41134892694006e-08,
|
| 23851 |
+
"loss": 0.0,
|
| 23852 |
+
"step": 3405
|
| 23853 |
+
},
|
| 23854 |
+
{
|
| 23855 |
+
"epoch": 1.9810964083175804,
|
| 23856 |
+
"grad_norm": NaN,
|
| 23857 |
+
"learning_rate": 2.2674317251558664e-08,
|
| 23858 |
+
"loss": 0.0,
|
| 23859 |
+
"step": 3406
|
| 23860 |
+
},
|
| 23861 |
+
{
|
| 23862 |
+
"epoch": 1.981678057292424,
|
| 23863 |
+
"grad_norm": NaN,
|
| 23864 |
+
"learning_rate": 2.127941426938729e-08,
|
| 23865 |
+
"loss": 0.0,
|
| 23866 |
+
"step": 3407
|
| 23867 |
+
},
|
| 23868 |
+
{
|
| 23869 |
+
"epoch": 1.9822597062672678,
|
| 23870 |
+
"grad_norm": NaN,
|
| 23871 |
+
"learning_rate": 1.9928781558475883e-08,
|
| 23872 |
+
"loss": 0.0,
|
| 23873 |
+
"step": 3408
|
| 23874 |
+
},
|
| 23875 |
+
{
|
| 23876 |
+
"epoch": 1.9828413552421114,
|
| 23877 |
+
"grad_norm": NaN,
|
| 23878 |
+
"learning_rate": 1.862242031517858e-08,
|
| 23879 |
+
"loss": 0.0,
|
| 23880 |
+
"step": 3409
|
| 23881 |
+
},
|
| 23882 |
+
{
|
| 23883 |
+
"epoch": 1.983423004216955,
|
| 23884 |
+
"grad_norm": NaN,
|
| 23885 |
+
"learning_rate": 1.7360331696653075e-08,
|
| 23886 |
+
"loss": 0.0,
|
| 23887 |
+
"step": 3410
|
| 23888 |
+
},
|
| 23889 |
+
{
|
| 23890 |
+
"epoch": 1.9840046531917988,
|
| 23891 |
+
"grad_norm": NaN,
|
| 23892 |
+
"learning_rate": 1.614251682083845e-08,
|
| 23893 |
+
"loss": 0.0,
|
| 23894 |
+
"step": 3411
|
| 23895 |
+
},
|
| 23896 |
+
{
|
| 23897 |
+
"epoch": 1.9845863021666426,
|
| 23898 |
+
"grad_norm": NaN,
|
| 23899 |
+
"learning_rate": 1.496897676644404e-08,
|
| 23900 |
+
"loss": 0.0,
|
| 23901 |
+
"step": 3412
|
| 23902 |
+
},
|
| 23903 |
+
{
|
| 23904 |
+
"epoch": 1.9851679511414861,
|
| 23905 |
+
"grad_norm": NaN,
|
| 23906 |
+
"learning_rate": 1.3839712572977227e-08,
|
| 23907 |
+
"loss": 0.0,
|
| 23908 |
+
"step": 3413
|
| 23909 |
+
},
|
| 23910 |
+
{
|
| 23911 |
+
"epoch": 1.9857496001163297,
|
| 23912 |
+
"grad_norm": NaN,
|
| 23913 |
+
"learning_rate": 1.275472524072674e-08,
|
| 23914 |
+
"loss": 0.0,
|
| 23915 |
+
"step": 3414
|
| 23916 |
+
},
|
| 23917 |
+
{
|
| 23918 |
+
"epoch": 1.9863312490911733,
|
| 23919 |
+
"grad_norm": NaN,
|
| 23920 |
+
"learning_rate": 1.1714015730740492e-08,
|
| 23921 |
+
"loss": 0.0,
|
| 23922 |
+
"step": 3415
|
| 23923 |
+
},
|
| 23924 |
+
{
|
| 23925 |
+
"epoch": 1.9869128980660171,
|
| 23926 |
+
"grad_norm": NaN,
|
| 23927 |
+
"learning_rate": 1.0717584964869964e-08,
|
| 23928 |
+
"loss": 0.0,
|
| 23929 |
+
"step": 3416
|
| 23930 |
+
},
|
| 23931 |
+
{
|
| 23932 |
+
"epoch": 1.987494547040861,
|
| 23933 |
+
"grad_norm": NaN,
|
| 23934 |
+
"learning_rate": 9.765433825736914e-09,
|
| 23935 |
+
"loss": 0.0,
|
| 23936 |
+
"step": 3417
|
| 23937 |
+
},
|
| 23938 |
+
{
|
| 23939 |
+
"epoch": 1.9880761960157045,
|
| 23940 |
+
"grad_norm": NaN,
|
| 23941 |
+
"learning_rate": 8.857563156738913e-09,
|
| 23942 |
+
"loss": 0.0,
|
| 23943 |
+
"step": 3418
|
| 23944 |
+
},
|
| 23945 |
+
{
|
| 23946 |
+
"epoch": 1.988657844990548,
|
| 23947 |
+
"grad_norm": NaN,
|
| 23948 |
+
"learning_rate": 7.993973762049356e-09,
|
| 23949 |
+
"loss": 0.0,
|
| 23950 |
+
"step": 3419
|
| 23951 |
+
},
|
| 23952 |
+
{
|
| 23953 |
+
"epoch": 1.989239493965392,
|
| 23954 |
+
"grad_norm": NaN,
|
| 23955 |
+
"learning_rate": 7.1746664066230094e-09,
|
| 23956 |
+
"loss": 0.0,
|
| 23957 |
+
"step": 3420
|
| 23958 |
+
},
|
| 23959 |
+
{
|
| 23960 |
+
"epoch": 1.9898211429402357,
|
| 23961 |
+
"grad_norm": NaN,
|
| 23962 |
+
"learning_rate": 6.399641816184909e-09,
|
| 23963 |
+
"loss": 0.0,
|
| 23964 |
+
"step": 3421
|
| 23965 |
+
},
|
| 23966 |
+
{
|
| 23967 |
+
"epoch": 1.9904027919150793,
|
| 23968 |
+
"grad_norm": NaN,
|
| 23969 |
+
"learning_rate": 5.668900677235911e-09,
|
| 23970 |
+
"loss": 0.0,
|
| 23971 |
+
"step": 3422
|
| 23972 |
+
},
|
| 23973 |
+
{
|
| 23974 |
+
"epoch": 1.9909844408899229,
|
| 23975 |
+
"grad_norm": NaN,
|
| 23976 |
+
"learning_rate": 4.982443637063794e-09,
|
| 23977 |
+
"loss": 0.0,
|
| 23978 |
+
"step": 3423
|
| 23979 |
+
},
|
| 23980 |
+
{
|
| 23981 |
+
"epoch": 1.9915660898647665,
|
| 23982 |
+
"grad_norm": NaN,
|
| 23983 |
+
"learning_rate": 4.340271303715504e-09,
|
| 23984 |
+
"loss": 0.0,
|
| 23985 |
+
"step": 3424
|
| 23986 |
+
},
|
| 23987 |
+
{
|
| 23988 |
+
"epoch": 1.9921477388396103,
|
| 23989 |
+
"grad_norm": NaN,
|
| 23990 |
+
"learning_rate": 3.742384246008257e-09,
|
| 23991 |
+
"loss": 0.0,
|
| 23992 |
+
"step": 3425
|
| 23993 |
+
},
|
| 23994 |
+
{
|
| 23995 |
+
"epoch": 1.992729387814454,
|
| 23996 |
+
"grad_norm": NaN,
|
| 23997 |
+
"learning_rate": 3.188782993551742e-09,
|
| 23998 |
+
"loss": 0.0,
|
| 23999 |
+
"step": 3426
|
| 24000 |
+
},
|
| 24001 |
+
{
|
| 24002 |
+
"epoch": 1.9933110367892977,
|
| 24003 |
+
"grad_norm": NaN,
|
| 24004 |
+
"learning_rate": 2.679468036709265e-09,
|
| 24005 |
+
"loss": 0.0,
|
| 24006 |
+
"step": 3427
|
| 24007 |
+
},
|
| 24008 |
+
{
|
| 24009 |
+
"epoch": 1.9938926857641412,
|
| 24010 |
+
"grad_norm": NaN,
|
| 24011 |
+
"learning_rate": 2.2144398266199518e-09,
|
| 24012 |
+
"loss": 0.0,
|
| 24013 |
+
"step": 3428
|
| 24014 |
+
},
|
| 24015 |
+
{
|
| 24016 |
+
"epoch": 1.994474334738985,
|
| 24017 |
+
"grad_norm": NaN,
|
| 24018 |
+
"learning_rate": 1.7936987752098511e-09,
|
| 24019 |
+
"loss": 0.0,
|
| 24020 |
+
"step": 3429
|
| 24021 |
+
},
|
| 24022 |
+
{
|
| 24023 |
+
"epoch": 1.9950559837138289,
|
| 24024 |
+
"grad_norm": NaN,
|
| 24025 |
+
"learning_rate": 1.417245255153077e-09,
|
| 24026 |
+
"loss": 0.0,
|
| 24027 |
+
"step": 3430
|
| 24028 |
+
},
|
| 24029 |
+
{
|
| 24030 |
+
"epoch": 1.9956376326886724,
|
| 24031 |
+
"grad_norm": NaN,
|
| 24032 |
+
"learning_rate": 1.0850795999051143e-09,
|
| 24033 |
+
"loss": 0.0,
|
| 24034 |
+
"step": 3431
|
| 24035 |
+
},
|
| 24036 |
+
{
|
| 24037 |
+
"epoch": 1.996219281663516,
|
| 24038 |
+
"grad_norm": NaN,
|
| 24039 |
+
"learning_rate": 7.972021036972699e-10,
|
| 24040 |
+
"loss": 0.0,
|
| 24041 |
+
"step": 3432
|
| 24042 |
+
},
|
| 24043 |
+
{
|
| 24044 |
+
"epoch": 1.9968009306383596,
|
| 24045 |
+
"grad_norm": NaN,
|
| 24046 |
+
"learning_rate": 5.536130215311186e-10,
|
| 24047 |
+
"loss": 0.0,
|
| 24048 |
+
"step": 3433
|
| 24049 |
+
},
|
| 24050 |
+
{
|
| 24051 |
+
"epoch": 1.9973825796132034,
|
| 24052 |
+
"grad_norm": NaN,
|
| 24053 |
+
"learning_rate": 3.5431256916185207e-10,
|
| 24054 |
+
"loss": 0.0,
|
| 24055 |
+
"step": 3434
|
| 24056 |
+
},
|
| 24057 |
+
{
|
| 24058 |
+
"epoch": 1.9979642285880472,
|
| 24059 |
+
"grad_norm": NaN,
|
| 24060 |
+
"learning_rate": 1.9930092313158455e-10,
|
| 24061 |
+
"loss": 0.0,
|
| 24062 |
+
"step": 3435
|
| 24063 |
+
},
|
| 24064 |
+
{
|
| 24065 |
+
"epoch": 1.9985458775628908,
|
| 24066 |
+
"grad_norm": NaN,
|
| 24067 |
+
"learning_rate": 8.857822075269973e-11,
|
| 24068 |
+
"loss": 0.0,
|
| 24069 |
+
"step": 3436
|
| 24070 |
+
},
|
| 24071 |
+
{
|
| 24072 |
+
"epoch": 1.9991275265377344,
|
| 24073 |
+
"grad_norm": NaN,
|
| 24074 |
+
"learning_rate": 2.2144560091197363e-11,
|
| 24075 |
+
"loss": 0.0,
|
| 24076 |
+
"step": 3437
|
| 24077 |
+
},
|
| 24078 |
+
{
|
| 24079 |
+
"epoch": 1.9997091755125782,
|
| 24080 |
+
"grad_norm": NaN,
|
| 24081 |
+
"learning_rate": 0.0,
|
| 24082 |
+
"loss": 0.0,
|
| 24083 |
+
"step": 3438
|
| 24084 |
}
|
| 24085 |
],
|
| 24086 |
"logging_steps": 1,
|
|
|
|
| 24095 |
"should_evaluate": false,
|
| 24096 |
"should_log": false,
|
| 24097 |
"should_save": true,
|
| 24098 |
+
"should_training_stop": true
|
| 24099 |
},
|
| 24100 |
"attributes": {}
|
| 24101 |
}
|
| 24102 |
},
|
| 24103 |
+
"total_flos": 3.203855633350656e+16,
|
| 24104 |
"train_batch_size": 2,
|
| 24105 |
"trial_name": null,
|
| 24106 |
"trial_params": null
|