Training in progress, step 14000, checkpoint
Browse files- last-checkpoint/global_step14000/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- last-checkpoint/global_step14000/mp_rank_00_model_states.pt +3 -0
- last-checkpoint/latest +1 -1
- last-checkpoint/model.safetensors +1 -1
- last-checkpoint/rng_state.pth +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +585 -5
last-checkpoint/global_step14000/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:86a2f554ce1ec3d33d02f54c080f37b2dec7e2cc450cf17190781c61293c3d8e
|
| 3 |
+
size 761059696
|
last-checkpoint/global_step14000/mp_rank_00_model_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4d5093ee98781439808af99abf1b4a69c796044bd8d29fcf1271fdffc29bb205
|
| 3 |
+
size 129965712
|
last-checkpoint/latest
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
|
|
|
|
| 1 |
+
global_step14000
|
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 181508256
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3b168574268210f92c9e3c34fcaf08a2e263a9da68774e57a2622b5bb1eb076e
|
| 3 |
size 181508256
|
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 sha256:9ec63f9e79d0c1ca51e7a6a550657e26243aaf847e676e55fcadb7e9922d4621
|
| 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:a15f6bb9490f63e2fdfa164e81ea571822fffda0ba3283b70da7aa9b56f23b1b
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 96.
|
| 3 |
-
"best_model_checkpoint": "./iteboshi_temp/checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 1000,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -3487,6 +3487,586 @@
|
|
| 3487 |
"eval_steps_per_second": 3.181,
|
| 3488 |
"eval_wer": 96.82225365393681,
|
| 3489 |
"step": 12000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3490 |
}
|
| 3491 |
],
|
| 3492 |
"logging_steps": 25,
|
|
@@ -3506,7 +4086,7 @@
|
|
| 3506 |
"attributes": {}
|
| 3507 |
}
|
| 3508 |
},
|
| 3509 |
-
"total_flos": 2.
|
| 3510 |
"train_batch_size": 4,
|
| 3511 |
"trial_name": null,
|
| 3512 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 96.67138142385667,
|
| 3 |
+
"best_model_checkpoint": "./iteboshi_temp/checkpoint-14000",
|
| 4 |
+
"epoch": 15.418502202643172,
|
| 5 |
"eval_steps": 1000,
|
| 6 |
+
"global_step": 14000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 3487 |
"eval_steps_per_second": 3.181,
|
| 3488 |
"eval_wer": 96.82225365393681,
|
| 3489 |
"step": 12000
|
| 3490 |
+
},
|
| 3491 |
+
{
|
| 3492 |
+
"epoch": 13.243392070484582,
|
| 3493 |
+
"grad_norm": 0.23276664316654205,
|
| 3494 |
+
"learning_rate": 8.17948717948718e-06,
|
| 3495 |
+
"loss": 0.0127,
|
| 3496 |
+
"step": 12025
|
| 3497 |
+
},
|
| 3498 |
+
{
|
| 3499 |
+
"epoch": 13.270925110132158,
|
| 3500 |
+
"grad_norm": 0.11635535210371017,
|
| 3501 |
+
"learning_rate": 8.153846153846154e-06,
|
| 3502 |
+
"loss": 0.0095,
|
| 3503 |
+
"step": 12050
|
| 3504 |
+
},
|
| 3505 |
+
{
|
| 3506 |
+
"epoch": 13.298458149779735,
|
| 3507 |
+
"grad_norm": 0.16364231705665588,
|
| 3508 |
+
"learning_rate": 8.12820512820513e-06,
|
| 3509 |
+
"loss": 0.0114,
|
| 3510 |
+
"step": 12075
|
| 3511 |
+
},
|
| 3512 |
+
{
|
| 3513 |
+
"epoch": 13.325991189427313,
|
| 3514 |
+
"grad_norm": 0.13409483432769775,
|
| 3515 |
+
"learning_rate": 8.102564102564103e-06,
|
| 3516 |
+
"loss": 0.0122,
|
| 3517 |
+
"step": 12100
|
| 3518 |
+
},
|
| 3519 |
+
{
|
| 3520 |
+
"epoch": 13.353524229074889,
|
| 3521 |
+
"grad_norm": 0.10241974890232086,
|
| 3522 |
+
"learning_rate": 8.076923076923077e-06,
|
| 3523 |
+
"loss": 0.0121,
|
| 3524 |
+
"step": 12125
|
| 3525 |
+
},
|
| 3526 |
+
{
|
| 3527 |
+
"epoch": 13.381057268722467,
|
| 3528 |
+
"grad_norm": 0.07747479528188705,
|
| 3529 |
+
"learning_rate": 8.051282051282052e-06,
|
| 3530 |
+
"loss": 0.0125,
|
| 3531 |
+
"step": 12150
|
| 3532 |
+
},
|
| 3533 |
+
{
|
| 3534 |
+
"epoch": 13.408590308370044,
|
| 3535 |
+
"grad_norm": 0.08573091775178909,
|
| 3536 |
+
"learning_rate": 8.025641025641026e-06,
|
| 3537 |
+
"loss": 0.0125,
|
| 3538 |
+
"step": 12175
|
| 3539 |
+
},
|
| 3540 |
+
{
|
| 3541 |
+
"epoch": 13.43612334801762,
|
| 3542 |
+
"grad_norm": 0.15942828357219696,
|
| 3543 |
+
"learning_rate": 8.000000000000001e-06,
|
| 3544 |
+
"loss": 0.0127,
|
| 3545 |
+
"step": 12200
|
| 3546 |
+
},
|
| 3547 |
+
{
|
| 3548 |
+
"epoch": 13.463656387665198,
|
| 3549 |
+
"grad_norm": 0.22725528478622437,
|
| 3550 |
+
"learning_rate": 7.974358974358975e-06,
|
| 3551 |
+
"loss": 0.0135,
|
| 3552 |
+
"step": 12225
|
| 3553 |
+
},
|
| 3554 |
+
{
|
| 3555 |
+
"epoch": 13.491189427312776,
|
| 3556 |
+
"grad_norm": 0.07710346579551697,
|
| 3557 |
+
"learning_rate": 7.948717948717949e-06,
|
| 3558 |
+
"loss": 0.01,
|
| 3559 |
+
"step": 12250
|
| 3560 |
+
},
|
| 3561 |
+
{
|
| 3562 |
+
"epoch": 13.518722466960352,
|
| 3563 |
+
"grad_norm": 0.06472910940647125,
|
| 3564 |
+
"learning_rate": 7.923076923076924e-06,
|
| 3565 |
+
"loss": 0.0126,
|
| 3566 |
+
"step": 12275
|
| 3567 |
+
},
|
| 3568 |
+
{
|
| 3569 |
+
"epoch": 13.54625550660793,
|
| 3570 |
+
"grad_norm": 0.20756784081459045,
|
| 3571 |
+
"learning_rate": 7.897435897435898e-06,
|
| 3572 |
+
"loss": 0.0135,
|
| 3573 |
+
"step": 12300
|
| 3574 |
+
},
|
| 3575 |
+
{
|
| 3576 |
+
"epoch": 13.573788546255507,
|
| 3577 |
+
"grad_norm": 0.1374279409646988,
|
| 3578 |
+
"learning_rate": 7.871794871794873e-06,
|
| 3579 |
+
"loss": 0.0128,
|
| 3580 |
+
"step": 12325
|
| 3581 |
+
},
|
| 3582 |
+
{
|
| 3583 |
+
"epoch": 13.601321585903083,
|
| 3584 |
+
"grad_norm": 0.08940647542476654,
|
| 3585 |
+
"learning_rate": 7.846153846153847e-06,
|
| 3586 |
+
"loss": 0.0122,
|
| 3587 |
+
"step": 12350
|
| 3588 |
+
},
|
| 3589 |
+
{
|
| 3590 |
+
"epoch": 13.62885462555066,
|
| 3591 |
+
"grad_norm": 0.174547016620636,
|
| 3592 |
+
"learning_rate": 7.820512820512822e-06,
|
| 3593 |
+
"loss": 0.0131,
|
| 3594 |
+
"step": 12375
|
| 3595 |
+
},
|
| 3596 |
+
{
|
| 3597 |
+
"epoch": 13.656387665198238,
|
| 3598 |
+
"grad_norm": 0.08119652420282364,
|
| 3599 |
+
"learning_rate": 7.794871794871796e-06,
|
| 3600 |
+
"loss": 0.0128,
|
| 3601 |
+
"step": 12400
|
| 3602 |
+
},
|
| 3603 |
+
{
|
| 3604 |
+
"epoch": 13.683920704845814,
|
| 3605 |
+
"grad_norm": 0.08605458587408066,
|
| 3606 |
+
"learning_rate": 7.76923076923077e-06,
|
| 3607 |
+
"loss": 0.0121,
|
| 3608 |
+
"step": 12425
|
| 3609 |
+
},
|
| 3610 |
+
{
|
| 3611 |
+
"epoch": 13.711453744493392,
|
| 3612 |
+
"grad_norm": 0.06772664934396744,
|
| 3613 |
+
"learning_rate": 7.743589743589745e-06,
|
| 3614 |
+
"loss": 0.0124,
|
| 3615 |
+
"step": 12450
|
| 3616 |
+
},
|
| 3617 |
+
{
|
| 3618 |
+
"epoch": 13.73898678414097,
|
| 3619 |
+
"grad_norm": 0.1438221037387848,
|
| 3620 |
+
"learning_rate": 7.717948717948718e-06,
|
| 3621 |
+
"loss": 0.0118,
|
| 3622 |
+
"step": 12475
|
| 3623 |
+
},
|
| 3624 |
+
{
|
| 3625 |
+
"epoch": 13.766519823788546,
|
| 3626 |
+
"grad_norm": 0.24703101813793182,
|
| 3627 |
+
"learning_rate": 7.692307692307694e-06,
|
| 3628 |
+
"loss": 0.013,
|
| 3629 |
+
"step": 12500
|
| 3630 |
+
},
|
| 3631 |
+
{
|
| 3632 |
+
"epoch": 13.794052863436123,
|
| 3633 |
+
"grad_norm": 0.06869100034236908,
|
| 3634 |
+
"learning_rate": 7.666666666666667e-06,
|
| 3635 |
+
"loss": 0.0105,
|
| 3636 |
+
"step": 12525
|
| 3637 |
+
},
|
| 3638 |
+
{
|
| 3639 |
+
"epoch": 13.821585903083701,
|
| 3640 |
+
"grad_norm": 0.2140737622976303,
|
| 3641 |
+
"learning_rate": 7.641025641025641e-06,
|
| 3642 |
+
"loss": 0.0153,
|
| 3643 |
+
"step": 12550
|
| 3644 |
+
},
|
| 3645 |
+
{
|
| 3646 |
+
"epoch": 13.849118942731277,
|
| 3647 |
+
"grad_norm": 0.14644251763820648,
|
| 3648 |
+
"learning_rate": 7.615384615384615e-06,
|
| 3649 |
+
"loss": 0.0113,
|
| 3650 |
+
"step": 12575
|
| 3651 |
+
},
|
| 3652 |
+
{
|
| 3653 |
+
"epoch": 13.876651982378855,
|
| 3654 |
+
"grad_norm": 0.0690101683139801,
|
| 3655 |
+
"learning_rate": 7.58974358974359e-06,
|
| 3656 |
+
"loss": 0.0114,
|
| 3657 |
+
"step": 12600
|
| 3658 |
+
},
|
| 3659 |
+
{
|
| 3660 |
+
"epoch": 13.904185022026432,
|
| 3661 |
+
"grad_norm": 0.07625989615917206,
|
| 3662 |
+
"learning_rate": 7.564102564102564e-06,
|
| 3663 |
+
"loss": 0.0117,
|
| 3664 |
+
"step": 12625
|
| 3665 |
+
},
|
| 3666 |
+
{
|
| 3667 |
+
"epoch": 13.931718061674008,
|
| 3668 |
+
"grad_norm": 0.06748715043067932,
|
| 3669 |
+
"learning_rate": 7.538461538461539e-06,
|
| 3670 |
+
"loss": 0.0126,
|
| 3671 |
+
"step": 12650
|
| 3672 |
+
},
|
| 3673 |
+
{
|
| 3674 |
+
"epoch": 13.959251101321586,
|
| 3675 |
+
"grad_norm": 0.11845114827156067,
|
| 3676 |
+
"learning_rate": 7.512820512820513e-06,
|
| 3677 |
+
"loss": 0.0121,
|
| 3678 |
+
"step": 12675
|
| 3679 |
+
},
|
| 3680 |
+
{
|
| 3681 |
+
"epoch": 13.986784140969164,
|
| 3682 |
+
"grad_norm": 0.16305984556674957,
|
| 3683 |
+
"learning_rate": 7.487179487179488e-06,
|
| 3684 |
+
"loss": 0.0142,
|
| 3685 |
+
"step": 12700
|
| 3686 |
+
},
|
| 3687 |
+
{
|
| 3688 |
+
"epoch": 14.01431718061674,
|
| 3689 |
+
"grad_norm": 0.04868720471858978,
|
| 3690 |
+
"learning_rate": 7.461538461538462e-06,
|
| 3691 |
+
"loss": 0.0101,
|
| 3692 |
+
"step": 12725
|
| 3693 |
+
},
|
| 3694 |
+
{
|
| 3695 |
+
"epoch": 14.041850220264317,
|
| 3696 |
+
"grad_norm": 0.19972330331802368,
|
| 3697 |
+
"learning_rate": 7.435897435897437e-06,
|
| 3698 |
+
"loss": 0.0078,
|
| 3699 |
+
"step": 12750
|
| 3700 |
+
},
|
| 3701 |
+
{
|
| 3702 |
+
"epoch": 14.069383259911895,
|
| 3703 |
+
"grad_norm": 0.06579900532960892,
|
| 3704 |
+
"learning_rate": 7.410256410256411e-06,
|
| 3705 |
+
"loss": 0.0097,
|
| 3706 |
+
"step": 12775
|
| 3707 |
+
},
|
| 3708 |
+
{
|
| 3709 |
+
"epoch": 14.09691629955947,
|
| 3710 |
+
"grad_norm": 0.07141165435314178,
|
| 3711 |
+
"learning_rate": 7.384615384615386e-06,
|
| 3712 |
+
"loss": 0.0075,
|
| 3713 |
+
"step": 12800
|
| 3714 |
+
},
|
| 3715 |
+
{
|
| 3716 |
+
"epoch": 14.124449339207048,
|
| 3717 |
+
"grad_norm": 0.0851076990365982,
|
| 3718 |
+
"learning_rate": 7.35897435897436e-06,
|
| 3719 |
+
"loss": 0.0087,
|
| 3720 |
+
"step": 12825
|
| 3721 |
+
},
|
| 3722 |
+
{
|
| 3723 |
+
"epoch": 14.151982378854626,
|
| 3724 |
+
"grad_norm": 0.08377552032470703,
|
| 3725 |
+
"learning_rate": 7.333333333333333e-06,
|
| 3726 |
+
"loss": 0.0097,
|
| 3727 |
+
"step": 12850
|
| 3728 |
+
},
|
| 3729 |
+
{
|
| 3730 |
+
"epoch": 14.179515418502202,
|
| 3731 |
+
"grad_norm": 0.05744962766766548,
|
| 3732 |
+
"learning_rate": 7.307692307692308e-06,
|
| 3733 |
+
"loss": 0.007,
|
| 3734 |
+
"step": 12875
|
| 3735 |
+
},
|
| 3736 |
+
{
|
| 3737 |
+
"epoch": 14.20704845814978,
|
| 3738 |
+
"grad_norm": 0.04978534206748009,
|
| 3739 |
+
"learning_rate": 7.282051282051282e-06,
|
| 3740 |
+
"loss": 0.0079,
|
| 3741 |
+
"step": 12900
|
| 3742 |
+
},
|
| 3743 |
+
{
|
| 3744 |
+
"epoch": 14.234581497797357,
|
| 3745 |
+
"grad_norm": 0.07835716754198074,
|
| 3746 |
+
"learning_rate": 7.256410256410257e-06,
|
| 3747 |
+
"loss": 0.0082,
|
| 3748 |
+
"step": 12925
|
| 3749 |
+
},
|
| 3750 |
+
{
|
| 3751 |
+
"epoch": 14.262114537444933,
|
| 3752 |
+
"grad_norm": 0.1253698170185089,
|
| 3753 |
+
"learning_rate": 7.230769230769231e-06,
|
| 3754 |
+
"loss": 0.0073,
|
| 3755 |
+
"step": 12950
|
| 3756 |
+
},
|
| 3757 |
+
{
|
| 3758 |
+
"epoch": 14.289647577092511,
|
| 3759 |
+
"grad_norm": 0.06475073099136353,
|
| 3760 |
+
"learning_rate": 7.205128205128206e-06,
|
| 3761 |
+
"loss": 0.0076,
|
| 3762 |
+
"step": 12975
|
| 3763 |
+
},
|
| 3764 |
+
{
|
| 3765 |
+
"epoch": 14.317180616740089,
|
| 3766 |
+
"grad_norm": 0.05178418755531311,
|
| 3767 |
+
"learning_rate": 7.17948717948718e-06,
|
| 3768 |
+
"loss": 0.008,
|
| 3769 |
+
"step": 13000
|
| 3770 |
+
},
|
| 3771 |
+
{
|
| 3772 |
+
"epoch": 14.317180616740089,
|
| 3773 |
+
"eval_cer": 57.12803640824,
|
| 3774 |
+
"eval_loss": 0.979995608329773,
|
| 3775 |
+
"eval_runtime": 852.1277,
|
| 3776 |
+
"eval_samples_per_second": 12.417,
|
| 3777 |
+
"eval_steps_per_second": 3.105,
|
| 3778 |
+
"eval_wer": 96.71852899575671,
|
| 3779 |
+
"step": 13000
|
| 3780 |
+
},
|
| 3781 |
+
{
|
| 3782 |
+
"epoch": 14.344713656387665,
|
| 3783 |
+
"grad_norm": 0.07015621662139893,
|
| 3784 |
+
"learning_rate": 7.153846153846155e-06,
|
| 3785 |
+
"loss": 0.0081,
|
| 3786 |
+
"step": 13025
|
| 3787 |
+
},
|
| 3788 |
+
{
|
| 3789 |
+
"epoch": 14.372246696035242,
|
| 3790 |
+
"grad_norm": 0.061675600707530975,
|
| 3791 |
+
"learning_rate": 7.128205128205129e-06,
|
| 3792 |
+
"loss": 0.0076,
|
| 3793 |
+
"step": 13050
|
| 3794 |
+
},
|
| 3795 |
+
{
|
| 3796 |
+
"epoch": 14.39977973568282,
|
| 3797 |
+
"grad_norm": 0.1807292252779007,
|
| 3798 |
+
"learning_rate": 7.102564102564104e-06,
|
| 3799 |
+
"loss": 0.009,
|
| 3800 |
+
"step": 13075
|
| 3801 |
+
},
|
| 3802 |
+
{
|
| 3803 |
+
"epoch": 14.427312775330396,
|
| 3804 |
+
"grad_norm": 0.05348524823784828,
|
| 3805 |
+
"learning_rate": 7.076923076923078e-06,
|
| 3806 |
+
"loss": 0.0085,
|
| 3807 |
+
"step": 13100
|
| 3808 |
+
},
|
| 3809 |
+
{
|
| 3810 |
+
"epoch": 14.454845814977974,
|
| 3811 |
+
"grad_norm": 0.040988489985466,
|
| 3812 |
+
"learning_rate": 7.051282051282053e-06,
|
| 3813 |
+
"loss": 0.0082,
|
| 3814 |
+
"step": 13125
|
| 3815 |
+
},
|
| 3816 |
+
{
|
| 3817 |
+
"epoch": 14.482378854625551,
|
| 3818 |
+
"grad_norm": 0.06997233629226685,
|
| 3819 |
+
"learning_rate": 7.025641025641025e-06,
|
| 3820 |
+
"loss": 0.0081,
|
| 3821 |
+
"step": 13150
|
| 3822 |
+
},
|
| 3823 |
+
{
|
| 3824 |
+
"epoch": 14.509911894273127,
|
| 3825 |
+
"grad_norm": 0.08544085174798965,
|
| 3826 |
+
"learning_rate": 7e-06,
|
| 3827 |
+
"loss": 0.0074,
|
| 3828 |
+
"step": 13175
|
| 3829 |
+
},
|
| 3830 |
+
{
|
| 3831 |
+
"epoch": 14.537444933920705,
|
| 3832 |
+
"grad_norm": 0.16296857595443726,
|
| 3833 |
+
"learning_rate": 6.974358974358974e-06,
|
| 3834 |
+
"loss": 0.0086,
|
| 3835 |
+
"step": 13200
|
| 3836 |
+
},
|
| 3837 |
+
{
|
| 3838 |
+
"epoch": 14.564977973568283,
|
| 3839 |
+
"grad_norm": 0.04533977061510086,
|
| 3840 |
+
"learning_rate": 6.948717948717949e-06,
|
| 3841 |
+
"loss": 0.0074,
|
| 3842 |
+
"step": 13225
|
| 3843 |
+
},
|
| 3844 |
+
{
|
| 3845 |
+
"epoch": 14.592511013215859,
|
| 3846 |
+
"grad_norm": 0.11487758159637451,
|
| 3847 |
+
"learning_rate": 6.923076923076923e-06,
|
| 3848 |
+
"loss": 0.0088,
|
| 3849 |
+
"step": 13250
|
| 3850 |
+
},
|
| 3851 |
+
{
|
| 3852 |
+
"epoch": 14.620044052863436,
|
| 3853 |
+
"grad_norm": 0.05602938309311867,
|
| 3854 |
+
"learning_rate": 6.897435897435898e-06,
|
| 3855 |
+
"loss": 0.0078,
|
| 3856 |
+
"step": 13275
|
| 3857 |
+
},
|
| 3858 |
+
{
|
| 3859 |
+
"epoch": 14.647577092511014,
|
| 3860 |
+
"grad_norm": 0.06876658648252487,
|
| 3861 |
+
"learning_rate": 6.871794871794872e-06,
|
| 3862 |
+
"loss": 0.009,
|
| 3863 |
+
"step": 13300
|
| 3864 |
+
},
|
| 3865 |
+
{
|
| 3866 |
+
"epoch": 14.67511013215859,
|
| 3867 |
+
"grad_norm": 0.21141541004180908,
|
| 3868 |
+
"learning_rate": 6.846153846153847e-06,
|
| 3869 |
+
"loss": 0.008,
|
| 3870 |
+
"step": 13325
|
| 3871 |
+
},
|
| 3872 |
+
{
|
| 3873 |
+
"epoch": 14.702643171806168,
|
| 3874 |
+
"grad_norm": 0.25995275378227234,
|
| 3875 |
+
"learning_rate": 6.820512820512821e-06,
|
| 3876 |
+
"loss": 0.0083,
|
| 3877 |
+
"step": 13350
|
| 3878 |
+
},
|
| 3879 |
+
{
|
| 3880 |
+
"epoch": 14.730176211453745,
|
| 3881 |
+
"grad_norm": 0.07600809633731842,
|
| 3882 |
+
"learning_rate": 6.794871794871796e-06,
|
| 3883 |
+
"loss": 0.0088,
|
| 3884 |
+
"step": 13375
|
| 3885 |
+
},
|
| 3886 |
+
{
|
| 3887 |
+
"epoch": 14.757709251101321,
|
| 3888 |
+
"grad_norm": 0.08052767813205719,
|
| 3889 |
+
"learning_rate": 6.76923076923077e-06,
|
| 3890 |
+
"loss": 0.0086,
|
| 3891 |
+
"step": 13400
|
| 3892 |
+
},
|
| 3893 |
+
{
|
| 3894 |
+
"epoch": 14.785242290748899,
|
| 3895 |
+
"grad_norm": 0.0512065626680851,
|
| 3896 |
+
"learning_rate": 6.743589743589745e-06,
|
| 3897 |
+
"loss": 0.0077,
|
| 3898 |
+
"step": 13425
|
| 3899 |
+
},
|
| 3900 |
+
{
|
| 3901 |
+
"epoch": 14.812775330396477,
|
| 3902 |
+
"grad_norm": 0.06173788756132126,
|
| 3903 |
+
"learning_rate": 6.717948717948718e-06,
|
| 3904 |
+
"loss": 0.0082,
|
| 3905 |
+
"step": 13450
|
| 3906 |
+
},
|
| 3907 |
+
{
|
| 3908 |
+
"epoch": 14.840308370044053,
|
| 3909 |
+
"grad_norm": 0.05340331420302391,
|
| 3910 |
+
"learning_rate": 6.692307692307692e-06,
|
| 3911 |
+
"loss": 0.0082,
|
| 3912 |
+
"step": 13475
|
| 3913 |
+
},
|
| 3914 |
+
{
|
| 3915 |
+
"epoch": 14.86784140969163,
|
| 3916 |
+
"grad_norm": 0.05996181070804596,
|
| 3917 |
+
"learning_rate": 6.666666666666667e-06,
|
| 3918 |
+
"loss": 0.0087,
|
| 3919 |
+
"step": 13500
|
| 3920 |
+
},
|
| 3921 |
+
{
|
| 3922 |
+
"epoch": 14.895374449339208,
|
| 3923 |
+
"grad_norm": 0.05534656345844269,
|
| 3924 |
+
"learning_rate": 6.641025641025641e-06,
|
| 3925 |
+
"loss": 0.0092,
|
| 3926 |
+
"step": 13525
|
| 3927 |
+
},
|
| 3928 |
+
{
|
| 3929 |
+
"epoch": 14.922907488986784,
|
| 3930 |
+
"grad_norm": 0.0450417622923851,
|
| 3931 |
+
"learning_rate": 6.615384615384616e-06,
|
| 3932 |
+
"loss": 0.0072,
|
| 3933 |
+
"step": 13550
|
| 3934 |
+
},
|
| 3935 |
+
{
|
| 3936 |
+
"epoch": 14.950440528634362,
|
| 3937 |
+
"grad_norm": 0.07858394831418991,
|
| 3938 |
+
"learning_rate": 6.58974358974359e-06,
|
| 3939 |
+
"loss": 0.0081,
|
| 3940 |
+
"step": 13575
|
| 3941 |
+
},
|
| 3942 |
+
{
|
| 3943 |
+
"epoch": 14.97797356828194,
|
| 3944 |
+
"grad_norm": 0.12585800886154175,
|
| 3945 |
+
"learning_rate": 6.564102564102565e-06,
|
| 3946 |
+
"loss": 0.0097,
|
| 3947 |
+
"step": 13600
|
| 3948 |
+
},
|
| 3949 |
+
{
|
| 3950 |
+
"epoch": 15.005506607929515,
|
| 3951 |
+
"grad_norm": 0.08572439104318619,
|
| 3952 |
+
"learning_rate": 6.538461538461539e-06,
|
| 3953 |
+
"loss": 0.0096,
|
| 3954 |
+
"step": 13625
|
| 3955 |
+
},
|
| 3956 |
+
{
|
| 3957 |
+
"epoch": 15.033039647577093,
|
| 3958 |
+
"grad_norm": 0.051955390721559525,
|
| 3959 |
+
"learning_rate": 6.512820512820514e-06,
|
| 3960 |
+
"loss": 0.0057,
|
| 3961 |
+
"step": 13650
|
| 3962 |
+
},
|
| 3963 |
+
{
|
| 3964 |
+
"epoch": 15.060572687224669,
|
| 3965 |
+
"grad_norm": 0.04480992630124092,
|
| 3966 |
+
"learning_rate": 6.487179487179488e-06,
|
| 3967 |
+
"loss": 0.0052,
|
| 3968 |
+
"step": 13675
|
| 3969 |
+
},
|
| 3970 |
+
{
|
| 3971 |
+
"epoch": 15.088105726872246,
|
| 3972 |
+
"grad_norm": 0.03293057531118393,
|
| 3973 |
+
"learning_rate": 6.461538461538463e-06,
|
| 3974 |
+
"loss": 0.0057,
|
| 3975 |
+
"step": 13700
|
| 3976 |
+
},
|
| 3977 |
+
{
|
| 3978 |
+
"epoch": 15.115638766519824,
|
| 3979 |
+
"grad_norm": 0.0531000941991806,
|
| 3980 |
+
"learning_rate": 6.435897435897437e-06,
|
| 3981 |
+
"loss": 0.0059,
|
| 3982 |
+
"step": 13725
|
| 3983 |
+
},
|
| 3984 |
+
{
|
| 3985 |
+
"epoch": 15.1431718061674,
|
| 3986 |
+
"grad_norm": 0.030901802703738213,
|
| 3987 |
+
"learning_rate": 6.410256410256412e-06,
|
| 3988 |
+
"loss": 0.0058,
|
| 3989 |
+
"step": 13750
|
| 3990 |
+
},
|
| 3991 |
+
{
|
| 3992 |
+
"epoch": 15.170704845814978,
|
| 3993 |
+
"grad_norm": 0.04868703335523605,
|
| 3994 |
+
"learning_rate": 6.384615384615384e-06,
|
| 3995 |
+
"loss": 0.0056,
|
| 3996 |
+
"step": 13775
|
| 3997 |
+
},
|
| 3998 |
+
{
|
| 3999 |
+
"epoch": 15.198237885462555,
|
| 4000 |
+
"grad_norm": 0.045157793909311295,
|
| 4001 |
+
"learning_rate": 6.358974358974359e-06,
|
| 4002 |
+
"loss": 0.0054,
|
| 4003 |
+
"step": 13800
|
| 4004 |
+
},
|
| 4005 |
+
{
|
| 4006 |
+
"epoch": 15.225770925110131,
|
| 4007 |
+
"grad_norm": 0.07468298077583313,
|
| 4008 |
+
"learning_rate": 6.333333333333333e-06,
|
| 4009 |
+
"loss": 0.0071,
|
| 4010 |
+
"step": 13825
|
| 4011 |
+
},
|
| 4012 |
+
{
|
| 4013 |
+
"epoch": 15.253303964757709,
|
| 4014 |
+
"grad_norm": 0.06386591494083405,
|
| 4015 |
+
"learning_rate": 6.307692307692308e-06,
|
| 4016 |
+
"loss": 0.006,
|
| 4017 |
+
"step": 13850
|
| 4018 |
+
},
|
| 4019 |
+
{
|
| 4020 |
+
"epoch": 15.280837004405287,
|
| 4021 |
+
"grad_norm": 0.04140784963965416,
|
| 4022 |
+
"learning_rate": 6.282051282051282e-06,
|
| 4023 |
+
"loss": 0.006,
|
| 4024 |
+
"step": 13875
|
| 4025 |
+
},
|
| 4026 |
+
{
|
| 4027 |
+
"epoch": 15.308370044052863,
|
| 4028 |
+
"grad_norm": 0.09238462150096893,
|
| 4029 |
+
"learning_rate": 6.256410256410257e-06,
|
| 4030 |
+
"loss": 0.0061,
|
| 4031 |
+
"step": 13900
|
| 4032 |
+
},
|
| 4033 |
+
{
|
| 4034 |
+
"epoch": 15.33590308370044,
|
| 4035 |
+
"grad_norm": 0.2956899106502533,
|
| 4036 |
+
"learning_rate": 6.230769230769231e-06,
|
| 4037 |
+
"loss": 0.006,
|
| 4038 |
+
"step": 13925
|
| 4039 |
+
},
|
| 4040 |
+
{
|
| 4041 |
+
"epoch": 15.363436123348018,
|
| 4042 |
+
"grad_norm": 0.048573561012744904,
|
| 4043 |
+
"learning_rate": 6.205128205128206e-06,
|
| 4044 |
+
"loss": 0.0069,
|
| 4045 |
+
"step": 13950
|
| 4046 |
+
},
|
| 4047 |
+
{
|
| 4048 |
+
"epoch": 15.390969162995594,
|
| 4049 |
+
"grad_norm": 0.17084334790706635,
|
| 4050 |
+
"learning_rate": 6.17948717948718e-06,
|
| 4051 |
+
"loss": 0.0068,
|
| 4052 |
+
"step": 13975
|
| 4053 |
+
},
|
| 4054 |
+
{
|
| 4055 |
+
"epoch": 15.418502202643172,
|
| 4056 |
+
"grad_norm": 0.06327039748430252,
|
| 4057 |
+
"learning_rate": 6.153846153846155e-06,
|
| 4058 |
+
"loss": 0.0062,
|
| 4059 |
+
"step": 14000
|
| 4060 |
+
},
|
| 4061 |
+
{
|
| 4062 |
+
"epoch": 15.418502202643172,
|
| 4063 |
+
"eval_cer": 55.30074627721161,
|
| 4064 |
+
"eval_loss": 0.9948021769523621,
|
| 4065 |
+
"eval_runtime": 827.2524,
|
| 4066 |
+
"eval_samples_per_second": 12.791,
|
| 4067 |
+
"eval_steps_per_second": 3.199,
|
| 4068 |
+
"eval_wer": 96.67138142385667,
|
| 4069 |
+
"step": 14000
|
| 4070 |
}
|
| 4071 |
],
|
| 4072 |
"logging_steps": 25,
|
|
|
|
| 4086 |
"attributes": {}
|
| 4087 |
}
|
| 4088 |
},
|
| 4089 |
+
"total_flos": 2.3634390360491295e+19,
|
| 4090 |
"train_batch_size": 4,
|
| 4091 |
"trial_name": null,
|
| 4092 |
"trial_params": null
|