Training in progress, step 12000
Browse files- optimizer.pt +1 -1
- rng_state.pth +1 -1
- runs/Jun07_12-33-16_DESKTOP-69FPKCK/events.out.tfevents.1717788805.DESKTOP-69FPKCK +2 -2
- scheduler.pt +1 -1
- trainer_state.json +712 -3
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 11230198
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bf5c3ae23e0061e18908dedb416fcb0d32e5ecadfc7f18abb693d4f1c6a53a96
|
| 3 |
size 11230198
|
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 sha256:69f50a692634404f2eebb2eab9f456865957578d752987bc52d843ac2a774366
|
| 3 |
size 14244
|
runs/Jun07_12-33-16_DESKTOP-69FPKCK/events.out.tfevents.1717788805.DESKTOP-69FPKCK
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:43356ad5026fe7bb50ebc6b79634c8aa625fabdcc290f5b9e335f623df1606f8
|
| 3 |
+
size 132761
|
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:af7225d5b0731cf57528f6961f709bb0e7ed929fa0d79711b3aa9685866e262d
|
| 3 |
size 1064
|
trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 2000,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -3552,6 +3552,715 @@
|
|
| 3552 |
"eval_samples_per_second": 2857.071,
|
| 3553 |
"eval_steps_per_second": 11.163,
|
| 3554 |
"step": 10000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3555 |
}
|
| 3556 |
],
|
| 3557 |
"logging_steps": 20,
|
|
@@ -3559,7 +4268,7 @@
|
|
| 3559 |
"num_input_tokens_seen": 0,
|
| 3560 |
"num_train_epochs": 3,
|
| 3561 |
"save_steps": 100,
|
| 3562 |
-
"total_flos":
|
| 3563 |
"train_batch_size": 256,
|
| 3564 |
"trial_name": null,
|
| 3565 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.05429397472615476,
|
| 5 |
"eval_steps": 2000,
|
| 6 |
+
"global_step": 12000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 3552 |
"eval_samples_per_second": 2857.071,
|
| 3553 |
"eval_steps_per_second": 11.163,
|
| 3554 |
"step": 10000
|
| 3555 |
+
},
|
| 3556 |
+
{
|
| 3557 |
+
"epoch": 0.04533546889633923,
|
| 3558 |
+
"grad_norm": 12.951713562011719,
|
| 3559 |
+
"learning_rate": 0.00013589720387295268,
|
| 3560 |
+
"loss": 9.154,
|
| 3561 |
+
"step": 10020
|
| 3562 |
+
},
|
| 3563 |
+
{
|
| 3564 |
+
"epoch": 0.04542595885421615,
|
| 3565 |
+
"grad_norm": 9.139362335205078,
|
| 3566 |
+
"learning_rate": 0.00013616867251832414,
|
| 3567 |
+
"loss": 9.154,
|
| 3568 |
+
"step": 10040
|
| 3569 |
+
},
|
| 3570 |
+
{
|
| 3571 |
+
"epoch": 0.04551644881209308,
|
| 3572 |
+
"grad_norm": 8.388337135314941,
|
| 3573 |
+
"learning_rate": 0.0001364401411636956,
|
| 3574 |
+
"loss": 9.1391,
|
| 3575 |
+
"step": 10060
|
| 3576 |
+
},
|
| 3577 |
+
{
|
| 3578 |
+
"epoch": 0.045606938769970004,
|
| 3579 |
+
"grad_norm": 10.0809326171875,
|
| 3580 |
+
"learning_rate": 0.00013671160980906704,
|
| 3581 |
+
"loss": 9.1417,
|
| 3582 |
+
"step": 10080
|
| 3583 |
+
},
|
| 3584 |
+
{
|
| 3585 |
+
"epoch": 0.04569742872784693,
|
| 3586 |
+
"grad_norm": 8.565701484680176,
|
| 3587 |
+
"learning_rate": 0.0001369830784544385,
|
| 3588 |
+
"loss": 9.1112,
|
| 3589 |
+
"step": 10100
|
| 3590 |
+
},
|
| 3591 |
+
{
|
| 3592 |
+
"epoch": 0.04578791868572385,
|
| 3593 |
+
"grad_norm": 10.437520027160645,
|
| 3594 |
+
"learning_rate": 0.00013725454709980997,
|
| 3595 |
+
"loss": 9.1169,
|
| 3596 |
+
"step": 10120
|
| 3597 |
+
},
|
| 3598 |
+
{
|
| 3599 |
+
"epoch": 0.04587840864360078,
|
| 3600 |
+
"grad_norm": 8.615896224975586,
|
| 3601 |
+
"learning_rate": 0.00013752601574518143,
|
| 3602 |
+
"loss": 9.1003,
|
| 3603 |
+
"step": 10140
|
| 3604 |
+
},
|
| 3605 |
+
{
|
| 3606 |
+
"epoch": 0.0459688986014777,
|
| 3607 |
+
"grad_norm": 10.89583683013916,
|
| 3608 |
+
"learning_rate": 0.0001377974843905529,
|
| 3609 |
+
"loss": 9.101,
|
| 3610 |
+
"step": 10160
|
| 3611 |
+
},
|
| 3612 |
+
{
|
| 3613 |
+
"epoch": 0.046059388559354625,
|
| 3614 |
+
"grad_norm": 9.786931991577148,
|
| 3615 |
+
"learning_rate": 0.00013806895303592433,
|
| 3616 |
+
"loss": 9.0689,
|
| 3617 |
+
"step": 10180
|
| 3618 |
+
},
|
| 3619 |
+
{
|
| 3620 |
+
"epoch": 0.04614987851723155,
|
| 3621 |
+
"grad_norm": 9.010174751281738,
|
| 3622 |
+
"learning_rate": 0.0001383404216812958,
|
| 3623 |
+
"loss": 9.0579,
|
| 3624 |
+
"step": 10200
|
| 3625 |
+
},
|
| 3626 |
+
{
|
| 3627 |
+
"epoch": 0.04624036847510848,
|
| 3628 |
+
"grad_norm": 11.039669036865234,
|
| 3629 |
+
"learning_rate": 0.00013861189032666725,
|
| 3630 |
+
"loss": 9.0865,
|
| 3631 |
+
"step": 10220
|
| 3632 |
+
},
|
| 3633 |
+
{
|
| 3634 |
+
"epoch": 0.0463308584329854,
|
| 3635 |
+
"grad_norm": 12.055830001831055,
|
| 3636 |
+
"learning_rate": 0.00013888335897203872,
|
| 3637 |
+
"loss": 9.0955,
|
| 3638 |
+
"step": 10240
|
| 3639 |
+
},
|
| 3640 |
+
{
|
| 3641 |
+
"epoch": 0.04642134839086232,
|
| 3642 |
+
"grad_norm": 8.361885070800781,
|
| 3643 |
+
"learning_rate": 0.00013915482761741018,
|
| 3644 |
+
"loss": 9.07,
|
| 3645 |
+
"step": 10260
|
| 3646 |
+
},
|
| 3647 |
+
{
|
| 3648 |
+
"epoch": 0.046511838348739246,
|
| 3649 |
+
"grad_norm": 7.196146011352539,
|
| 3650 |
+
"learning_rate": 0.00013942629626278164,
|
| 3651 |
+
"loss": 9.0528,
|
| 3652 |
+
"step": 10280
|
| 3653 |
+
},
|
| 3654 |
+
{
|
| 3655 |
+
"epoch": 0.046602328306616175,
|
| 3656 |
+
"grad_norm": 9.67076587677002,
|
| 3657 |
+
"learning_rate": 0.0001396977649081531,
|
| 3658 |
+
"loss": 9.0546,
|
| 3659 |
+
"step": 10300
|
| 3660 |
+
},
|
| 3661 |
+
{
|
| 3662 |
+
"epoch": 0.0466928182644931,
|
| 3663 |
+
"grad_norm": 10.09327220916748,
|
| 3664 |
+
"learning_rate": 0.00013996923355352457,
|
| 3665 |
+
"loss": 9.0741,
|
| 3666 |
+
"step": 10320
|
| 3667 |
+
},
|
| 3668 |
+
{
|
| 3669 |
+
"epoch": 0.04678330822237002,
|
| 3670 |
+
"grad_norm": 9.639015197753906,
|
| 3671 |
+
"learning_rate": 0.00014024070219889603,
|
| 3672 |
+
"loss": 9.0633,
|
| 3673 |
+
"step": 10340
|
| 3674 |
+
},
|
| 3675 |
+
{
|
| 3676 |
+
"epoch": 0.04687379818024695,
|
| 3677 |
+
"grad_norm": 10.251932144165039,
|
| 3678 |
+
"learning_rate": 0.0001405121708442675,
|
| 3679 |
+
"loss": 9.0446,
|
| 3680 |
+
"step": 10360
|
| 3681 |
+
},
|
| 3682 |
+
{
|
| 3683 |
+
"epoch": 0.04696428813812387,
|
| 3684 |
+
"grad_norm": 11.07875919342041,
|
| 3685 |
+
"learning_rate": 0.00014078363948963896,
|
| 3686 |
+
"loss": 9.0418,
|
| 3687 |
+
"step": 10380
|
| 3688 |
+
},
|
| 3689 |
+
{
|
| 3690 |
+
"epoch": 0.047054778096000796,
|
| 3691 |
+
"grad_norm": 9.328507423400879,
|
| 3692 |
+
"learning_rate": 0.00014105510813501042,
|
| 3693 |
+
"loss": 9.0287,
|
| 3694 |
+
"step": 10400
|
| 3695 |
+
},
|
| 3696 |
+
{
|
| 3697 |
+
"epoch": 0.04714526805387772,
|
| 3698 |
+
"grad_norm": 7.056753635406494,
|
| 3699 |
+
"learning_rate": 0.00014132657678038186,
|
| 3700 |
+
"loss": 9.0362,
|
| 3701 |
+
"step": 10420
|
| 3702 |
+
},
|
| 3703 |
+
{
|
| 3704 |
+
"epoch": 0.04723575801175465,
|
| 3705 |
+
"grad_norm": 8.899680137634277,
|
| 3706 |
+
"learning_rate": 0.0001415980454257533,
|
| 3707 |
+
"loss": 9.036,
|
| 3708 |
+
"step": 10440
|
| 3709 |
+
},
|
| 3710 |
+
{
|
| 3711 |
+
"epoch": 0.04732624796963157,
|
| 3712 |
+
"grad_norm": 9.175132751464844,
|
| 3713 |
+
"learning_rate": 0.00014186951407112476,
|
| 3714 |
+
"loss": 9.0444,
|
| 3715 |
+
"step": 10460
|
| 3716 |
+
},
|
| 3717 |
+
{
|
| 3718 |
+
"epoch": 0.047416737927508494,
|
| 3719 |
+
"grad_norm": 9.374978065490723,
|
| 3720 |
+
"learning_rate": 0.00014214098271649622,
|
| 3721 |
+
"loss": 9.0372,
|
| 3722 |
+
"step": 10480
|
| 3723 |
+
},
|
| 3724 |
+
{
|
| 3725 |
+
"epoch": 0.04750722788538542,
|
| 3726 |
+
"grad_norm": 9.893750190734863,
|
| 3727 |
+
"learning_rate": 0.00014241245136186769,
|
| 3728 |
+
"loss": 9.0424,
|
| 3729 |
+
"step": 10500
|
| 3730 |
+
},
|
| 3731 |
+
{
|
| 3732 |
+
"epoch": 0.04759771784326235,
|
| 3733 |
+
"grad_norm": 7.787280082702637,
|
| 3734 |
+
"learning_rate": 0.00014265677314270202,
|
| 3735 |
+
"loss": 8.9691,
|
| 3736 |
+
"step": 10520
|
| 3737 |
+
},
|
| 3738 |
+
{
|
| 3739 |
+
"epoch": 0.04768820780113927,
|
| 3740 |
+
"grad_norm": 17.40734100341797,
|
| 3741 |
+
"learning_rate": 0.00014277893403311917,
|
| 3742 |
+
"loss": 8.2225,
|
| 3743 |
+
"step": 10540
|
| 3744 |
+
},
|
| 3745 |
+
{
|
| 3746 |
+
"epoch": 0.04777869775901619,
|
| 3747 |
+
"grad_norm": NaN,
|
| 3748 |
+
"learning_rate": 0.00014286037462673062,
|
| 3749 |
+
"loss": 6.6046,
|
| 3750 |
+
"step": 10560
|
| 3751 |
+
},
|
| 3752 |
+
{
|
| 3753 |
+
"epoch": 0.047869187716893115,
|
| 3754 |
+
"grad_norm": NaN,
|
| 3755 |
+
"learning_rate": 0.0001429146683558049,
|
| 3756 |
+
"loss": 3.0921,
|
| 3757 |
+
"step": 10580
|
| 3758 |
+
},
|
| 3759 |
+
{
|
| 3760 |
+
"epoch": 0.047959677674770045,
|
| 3761 |
+
"grad_norm": NaN,
|
| 3762 |
+
"learning_rate": 0.00014294181522034205,
|
| 3763 |
+
"loss": 3.9765,
|
| 3764 |
+
"step": 10600
|
| 3765 |
+
},
|
| 3766 |
+
{
|
| 3767 |
+
"epoch": 0.04805016763264697,
|
| 3768 |
+
"grad_norm": NaN,
|
| 3769 |
+
"learning_rate": 0.00014298253551714776,
|
| 3770 |
+
"loss": 6.9972,
|
| 3771 |
+
"step": 10620
|
| 3772 |
+
},
|
| 3773 |
+
{
|
| 3774 |
+
"epoch": 0.04814065759052389,
|
| 3775 |
+
"grad_norm": NaN,
|
| 3776 |
+
"learning_rate": 0.00014298253551714776,
|
| 3777 |
+
"loss": 0.0,
|
| 3778 |
+
"step": 10640
|
| 3779 |
+
},
|
| 3780 |
+
{
|
| 3781 |
+
"epoch": 0.04823114754840081,
|
| 3782 |
+
"grad_norm": NaN,
|
| 3783 |
+
"learning_rate": 0.00014298253551714776,
|
| 3784 |
+
"loss": 0.0,
|
| 3785 |
+
"step": 10660
|
| 3786 |
+
},
|
| 3787 |
+
{
|
| 3788 |
+
"epoch": 0.04832163750627774,
|
| 3789 |
+
"grad_norm": NaN,
|
| 3790 |
+
"learning_rate": 0.00014298253551714776,
|
| 3791 |
+
"loss": 0.0,
|
| 3792 |
+
"step": 10680
|
| 3793 |
+
},
|
| 3794 |
+
{
|
| 3795 |
+
"epoch": 0.048412127464154665,
|
| 3796 |
+
"grad_norm": NaN,
|
| 3797 |
+
"learning_rate": 0.00014298253551714776,
|
| 3798 |
+
"loss": 0.0,
|
| 3799 |
+
"step": 10700
|
| 3800 |
+
},
|
| 3801 |
+
{
|
| 3802 |
+
"epoch": 0.04850261742203159,
|
| 3803 |
+
"grad_norm": NaN,
|
| 3804 |
+
"learning_rate": 0.00014298253551714776,
|
| 3805 |
+
"loss": 0.0,
|
| 3806 |
+
"step": 10720
|
| 3807 |
+
},
|
| 3808 |
+
{
|
| 3809 |
+
"epoch": 0.04859310737990852,
|
| 3810 |
+
"grad_norm": NaN,
|
| 3811 |
+
"learning_rate": 0.00014298253551714776,
|
| 3812 |
+
"loss": 0.0,
|
| 3813 |
+
"step": 10740
|
| 3814 |
+
},
|
| 3815 |
+
{
|
| 3816 |
+
"epoch": 0.04868359733778544,
|
| 3817 |
+
"grad_norm": NaN,
|
| 3818 |
+
"learning_rate": 0.00014298253551714776,
|
| 3819 |
+
"loss": 0.0,
|
| 3820 |
+
"step": 10760
|
| 3821 |
+
},
|
| 3822 |
+
{
|
| 3823 |
+
"epoch": 0.04877408729566236,
|
| 3824 |
+
"grad_norm": NaN,
|
| 3825 |
+
"learning_rate": 0.00014298253551714776,
|
| 3826 |
+
"loss": 0.0,
|
| 3827 |
+
"step": 10780
|
| 3828 |
+
},
|
| 3829 |
+
{
|
| 3830 |
+
"epoch": 0.048864577253539286,
|
| 3831 |
+
"grad_norm": NaN,
|
| 3832 |
+
"learning_rate": 0.00014298253551714776,
|
| 3833 |
+
"loss": 0.0,
|
| 3834 |
+
"step": 10800
|
| 3835 |
+
},
|
| 3836 |
+
{
|
| 3837 |
+
"epoch": 0.048955067211416216,
|
| 3838 |
+
"grad_norm": NaN,
|
| 3839 |
+
"learning_rate": 0.00014298253551714776,
|
| 3840 |
+
"loss": 0.0,
|
| 3841 |
+
"step": 10820
|
| 3842 |
+
},
|
| 3843 |
+
{
|
| 3844 |
+
"epoch": 0.04904555716929314,
|
| 3845 |
+
"grad_norm": NaN,
|
| 3846 |
+
"learning_rate": 0.00014298253551714776,
|
| 3847 |
+
"loss": 0.0,
|
| 3848 |
+
"step": 10840
|
| 3849 |
+
},
|
| 3850 |
+
{
|
| 3851 |
+
"epoch": 0.04913604712717006,
|
| 3852 |
+
"grad_norm": NaN,
|
| 3853 |
+
"learning_rate": 0.00014298253551714776,
|
| 3854 |
+
"loss": 0.0,
|
| 3855 |
+
"step": 10860
|
| 3856 |
+
},
|
| 3857 |
+
{
|
| 3858 |
+
"epoch": 0.049226537085046984,
|
| 3859 |
+
"grad_norm": NaN,
|
| 3860 |
+
"learning_rate": 0.00014298253551714776,
|
| 3861 |
+
"loss": 0.0,
|
| 3862 |
+
"step": 10880
|
| 3863 |
+
},
|
| 3864 |
+
{
|
| 3865 |
+
"epoch": 0.049317027042923914,
|
| 3866 |
+
"grad_norm": NaN,
|
| 3867 |
+
"learning_rate": 0.00014298253551714776,
|
| 3868 |
+
"loss": 0.0,
|
| 3869 |
+
"step": 10900
|
| 3870 |
+
},
|
| 3871 |
+
{
|
| 3872 |
+
"epoch": 0.04940751700080084,
|
| 3873 |
+
"grad_norm": NaN,
|
| 3874 |
+
"learning_rate": 0.00014298253551714776,
|
| 3875 |
+
"loss": 0.0,
|
| 3876 |
+
"step": 10920
|
| 3877 |
+
},
|
| 3878 |
+
{
|
| 3879 |
+
"epoch": 0.04949800695867776,
|
| 3880 |
+
"grad_norm": NaN,
|
| 3881 |
+
"learning_rate": 0.00014298253551714776,
|
| 3882 |
+
"loss": 0.0,
|
| 3883 |
+
"step": 10940
|
| 3884 |
+
},
|
| 3885 |
+
{
|
| 3886 |
+
"epoch": 0.04958849691655468,
|
| 3887 |
+
"grad_norm": NaN,
|
| 3888 |
+
"learning_rate": 0.00014298253551714776,
|
| 3889 |
+
"loss": 0.0,
|
| 3890 |
+
"step": 10960
|
| 3891 |
+
},
|
| 3892 |
+
{
|
| 3893 |
+
"epoch": 0.04967898687443161,
|
| 3894 |
+
"grad_norm": NaN,
|
| 3895 |
+
"learning_rate": 0.00014298253551714776,
|
| 3896 |
+
"loss": 0.0,
|
| 3897 |
+
"step": 10980
|
| 3898 |
+
},
|
| 3899 |
+
{
|
| 3900 |
+
"epoch": 0.049769476832308535,
|
| 3901 |
+
"grad_norm": NaN,
|
| 3902 |
+
"learning_rate": 0.00014298253551714776,
|
| 3903 |
+
"loss": 0.0,
|
| 3904 |
+
"step": 11000
|
| 3905 |
+
},
|
| 3906 |
+
{
|
| 3907 |
+
"epoch": 0.04985996679018546,
|
| 3908 |
+
"grad_norm": NaN,
|
| 3909 |
+
"learning_rate": 0.00014298253551714776,
|
| 3910 |
+
"loss": 0.0,
|
| 3911 |
+
"step": 11020
|
| 3912 |
+
},
|
| 3913 |
+
{
|
| 3914 |
+
"epoch": 0.04995045674806239,
|
| 3915 |
+
"grad_norm": NaN,
|
| 3916 |
+
"learning_rate": 0.00014298253551714776,
|
| 3917 |
+
"loss": 0.0,
|
| 3918 |
+
"step": 11040
|
| 3919 |
+
},
|
| 3920 |
+
{
|
| 3921 |
+
"epoch": 0.05004094670593931,
|
| 3922 |
+
"grad_norm": NaN,
|
| 3923 |
+
"learning_rate": 0.00014298253551714776,
|
| 3924 |
+
"loss": 0.0,
|
| 3925 |
+
"step": 11060
|
| 3926 |
+
},
|
| 3927 |
+
{
|
| 3928 |
+
"epoch": 0.05013143666381623,
|
| 3929 |
+
"grad_norm": NaN,
|
| 3930 |
+
"learning_rate": 0.00014298253551714776,
|
| 3931 |
+
"loss": 0.0,
|
| 3932 |
+
"step": 11080
|
| 3933 |
+
},
|
| 3934 |
+
{
|
| 3935 |
+
"epoch": 0.050221926621693155,
|
| 3936 |
+
"grad_norm": NaN,
|
| 3937 |
+
"learning_rate": 0.00014298253551714776,
|
| 3938 |
+
"loss": 0.0,
|
| 3939 |
+
"step": 11100
|
| 3940 |
+
},
|
| 3941 |
+
{
|
| 3942 |
+
"epoch": 0.050312416579570085,
|
| 3943 |
+
"grad_norm": NaN,
|
| 3944 |
+
"learning_rate": 0.00014298253551714776,
|
| 3945 |
+
"loss": 0.0,
|
| 3946 |
+
"step": 11120
|
| 3947 |
+
},
|
| 3948 |
+
{
|
| 3949 |
+
"epoch": 0.05040290653744701,
|
| 3950 |
+
"grad_norm": NaN,
|
| 3951 |
+
"learning_rate": 0.00014298253551714776,
|
| 3952 |
+
"loss": 0.0,
|
| 3953 |
+
"step": 11140
|
| 3954 |
+
},
|
| 3955 |
+
{
|
| 3956 |
+
"epoch": 0.05049339649532393,
|
| 3957 |
+
"grad_norm": NaN,
|
| 3958 |
+
"learning_rate": 0.00014298253551714776,
|
| 3959 |
+
"loss": 0.0,
|
| 3960 |
+
"step": 11160
|
| 3961 |
+
},
|
| 3962 |
+
{
|
| 3963 |
+
"epoch": 0.05058388645320085,
|
| 3964 |
+
"grad_norm": NaN,
|
| 3965 |
+
"learning_rate": 0.00014298253551714776,
|
| 3966 |
+
"loss": 0.0,
|
| 3967 |
+
"step": 11180
|
| 3968 |
+
},
|
| 3969 |
+
{
|
| 3970 |
+
"epoch": 0.05067437641107778,
|
| 3971 |
+
"grad_norm": NaN,
|
| 3972 |
+
"learning_rate": 0.00014298253551714776,
|
| 3973 |
+
"loss": 0.0,
|
| 3974 |
+
"step": 11200
|
| 3975 |
+
},
|
| 3976 |
+
{
|
| 3977 |
+
"epoch": 0.050764866368954706,
|
| 3978 |
+
"grad_norm": NaN,
|
| 3979 |
+
"learning_rate": 0.00014298253551714776,
|
| 3980 |
+
"loss": 0.0,
|
| 3981 |
+
"step": 11220
|
| 3982 |
+
},
|
| 3983 |
+
{
|
| 3984 |
+
"epoch": 0.05085535632683163,
|
| 3985 |
+
"grad_norm": NaN,
|
| 3986 |
+
"learning_rate": 0.00014298253551714776,
|
| 3987 |
+
"loss": 0.0,
|
| 3988 |
+
"step": 11240
|
| 3989 |
+
},
|
| 3990 |
+
{
|
| 3991 |
+
"epoch": 0.05094584628470855,
|
| 3992 |
+
"grad_norm": NaN,
|
| 3993 |
+
"learning_rate": 0.00014298253551714776,
|
| 3994 |
+
"loss": 0.0,
|
| 3995 |
+
"step": 11260
|
| 3996 |
+
},
|
| 3997 |
+
{
|
| 3998 |
+
"epoch": 0.05103633624258548,
|
| 3999 |
+
"grad_norm": NaN,
|
| 4000 |
+
"learning_rate": 0.00014298253551714776,
|
| 4001 |
+
"loss": 0.0,
|
| 4002 |
+
"step": 11280
|
| 4003 |
+
},
|
| 4004 |
+
{
|
| 4005 |
+
"epoch": 0.051126826200462404,
|
| 4006 |
+
"grad_norm": NaN,
|
| 4007 |
+
"learning_rate": 0.00014298253551714776,
|
| 4008 |
+
"loss": 0.0,
|
| 4009 |
+
"step": 11300
|
| 4010 |
+
},
|
| 4011 |
+
{
|
| 4012 |
+
"epoch": 0.05121731615833933,
|
| 4013 |
+
"grad_norm": NaN,
|
| 4014 |
+
"learning_rate": 0.00014298253551714776,
|
| 4015 |
+
"loss": 0.0,
|
| 4016 |
+
"step": 11320
|
| 4017 |
+
},
|
| 4018 |
+
{
|
| 4019 |
+
"epoch": 0.051307806116216256,
|
| 4020 |
+
"grad_norm": NaN,
|
| 4021 |
+
"learning_rate": 0.00014298253551714776,
|
| 4022 |
+
"loss": 0.0,
|
| 4023 |
+
"step": 11340
|
| 4024 |
+
},
|
| 4025 |
+
{
|
| 4026 |
+
"epoch": 0.05139829607409318,
|
| 4027 |
+
"grad_norm": NaN,
|
| 4028 |
+
"learning_rate": 0.00014298253551714776,
|
| 4029 |
+
"loss": 0.0,
|
| 4030 |
+
"step": 11360
|
| 4031 |
+
},
|
| 4032 |
+
{
|
| 4033 |
+
"epoch": 0.0514887860319701,
|
| 4034 |
+
"grad_norm": NaN,
|
| 4035 |
+
"learning_rate": 0.00014298253551714776,
|
| 4036 |
+
"loss": 0.0,
|
| 4037 |
+
"step": 11380
|
| 4038 |
+
},
|
| 4039 |
+
{
|
| 4040 |
+
"epoch": 0.051579275989847025,
|
| 4041 |
+
"grad_norm": NaN,
|
| 4042 |
+
"learning_rate": 0.00014298253551714776,
|
| 4043 |
+
"loss": 0.0,
|
| 4044 |
+
"step": 11400
|
| 4045 |
+
},
|
| 4046 |
+
{
|
| 4047 |
+
"epoch": 0.051669765947723954,
|
| 4048 |
+
"grad_norm": NaN,
|
| 4049 |
+
"learning_rate": 0.00014298253551714776,
|
| 4050 |
+
"loss": 0.0,
|
| 4051 |
+
"step": 11420
|
| 4052 |
+
},
|
| 4053 |
+
{
|
| 4054 |
+
"epoch": 0.05176025590560088,
|
| 4055 |
+
"grad_norm": NaN,
|
| 4056 |
+
"learning_rate": 0.00014298253551714776,
|
| 4057 |
+
"loss": 0.0,
|
| 4058 |
+
"step": 11440
|
| 4059 |
+
},
|
| 4060 |
+
{
|
| 4061 |
+
"epoch": 0.0518507458634778,
|
| 4062 |
+
"grad_norm": NaN,
|
| 4063 |
+
"learning_rate": 0.00014298253551714776,
|
| 4064 |
+
"loss": 0.0,
|
| 4065 |
+
"step": 11460
|
| 4066 |
+
},
|
| 4067 |
+
{
|
| 4068 |
+
"epoch": 0.05194123582135472,
|
| 4069 |
+
"grad_norm": NaN,
|
| 4070 |
+
"learning_rate": 0.00014298253551714776,
|
| 4071 |
+
"loss": 0.0,
|
| 4072 |
+
"step": 11480
|
| 4073 |
+
},
|
| 4074 |
+
{
|
| 4075 |
+
"epoch": 0.05203172577923165,
|
| 4076 |
+
"grad_norm": NaN,
|
| 4077 |
+
"learning_rate": 0.00014298253551714776,
|
| 4078 |
+
"loss": 0.0,
|
| 4079 |
+
"step": 11500
|
| 4080 |
+
},
|
| 4081 |
+
{
|
| 4082 |
+
"epoch": 0.052122215737108575,
|
| 4083 |
+
"grad_norm": NaN,
|
| 4084 |
+
"learning_rate": 0.00014298253551714776,
|
| 4085 |
+
"loss": 0.0,
|
| 4086 |
+
"step": 11520
|
| 4087 |
+
},
|
| 4088 |
+
{
|
| 4089 |
+
"epoch": 0.0522127056949855,
|
| 4090 |
+
"grad_norm": NaN,
|
| 4091 |
+
"learning_rate": 0.00014298253551714776,
|
| 4092 |
+
"loss": 0.0,
|
| 4093 |
+
"step": 11540
|
| 4094 |
+
},
|
| 4095 |
+
{
|
| 4096 |
+
"epoch": 0.05230319565286242,
|
| 4097 |
+
"grad_norm": NaN,
|
| 4098 |
+
"learning_rate": 0.00014298253551714776,
|
| 4099 |
+
"loss": 0.0,
|
| 4100 |
+
"step": 11560
|
| 4101 |
+
},
|
| 4102 |
+
{
|
| 4103 |
+
"epoch": 0.05239368561073935,
|
| 4104 |
+
"grad_norm": NaN,
|
| 4105 |
+
"learning_rate": 0.00014298253551714776,
|
| 4106 |
+
"loss": 0.0,
|
| 4107 |
+
"step": 11580
|
| 4108 |
+
},
|
| 4109 |
+
{
|
| 4110 |
+
"epoch": 0.05248417556861627,
|
| 4111 |
+
"grad_norm": NaN,
|
| 4112 |
+
"learning_rate": 0.00014298253551714776,
|
| 4113 |
+
"loss": 0.0,
|
| 4114 |
+
"step": 11600
|
| 4115 |
+
},
|
| 4116 |
+
{
|
| 4117 |
+
"epoch": 0.052574665526493196,
|
| 4118 |
+
"grad_norm": NaN,
|
| 4119 |
+
"learning_rate": 0.00014298253551714776,
|
| 4120 |
+
"loss": 0.0,
|
| 4121 |
+
"step": 11620
|
| 4122 |
+
},
|
| 4123 |
+
{
|
| 4124 |
+
"epoch": 0.05266515548437012,
|
| 4125 |
+
"grad_norm": NaN,
|
| 4126 |
+
"learning_rate": 0.00014298253551714776,
|
| 4127 |
+
"loss": 0.0,
|
| 4128 |
+
"step": 11640
|
| 4129 |
+
},
|
| 4130 |
+
{
|
| 4131 |
+
"epoch": 0.05275564544224705,
|
| 4132 |
+
"grad_norm": NaN,
|
| 4133 |
+
"learning_rate": 0.00014298253551714776,
|
| 4134 |
+
"loss": 0.0,
|
| 4135 |
+
"step": 11660
|
| 4136 |
+
},
|
| 4137 |
+
{
|
| 4138 |
+
"epoch": 0.05284613540012397,
|
| 4139 |
+
"grad_norm": NaN,
|
| 4140 |
+
"learning_rate": 0.00014298253551714776,
|
| 4141 |
+
"loss": 0.0,
|
| 4142 |
+
"step": 11680
|
| 4143 |
+
},
|
| 4144 |
+
{
|
| 4145 |
+
"epoch": 0.052936625358000894,
|
| 4146 |
+
"grad_norm": NaN,
|
| 4147 |
+
"learning_rate": 0.00014298253551714776,
|
| 4148 |
+
"loss": 0.0,
|
| 4149 |
+
"step": 11700
|
| 4150 |
+
},
|
| 4151 |
+
{
|
| 4152 |
+
"epoch": 0.053027115315877824,
|
| 4153 |
+
"grad_norm": NaN,
|
| 4154 |
+
"learning_rate": 0.00014298253551714776,
|
| 4155 |
+
"loss": 0.0,
|
| 4156 |
+
"step": 11720
|
| 4157 |
+
},
|
| 4158 |
+
{
|
| 4159 |
+
"epoch": 0.053117605273754746,
|
| 4160 |
+
"grad_norm": NaN,
|
| 4161 |
+
"learning_rate": 0.00014298253551714776,
|
| 4162 |
+
"loss": 0.0,
|
| 4163 |
+
"step": 11740
|
| 4164 |
+
},
|
| 4165 |
+
{
|
| 4166 |
+
"epoch": 0.05320809523163167,
|
| 4167 |
+
"grad_norm": NaN,
|
| 4168 |
+
"learning_rate": 0.00014298253551714776,
|
| 4169 |
+
"loss": 0.0,
|
| 4170 |
+
"step": 11760
|
| 4171 |
+
},
|
| 4172 |
+
{
|
| 4173 |
+
"epoch": 0.05329858518950859,
|
| 4174 |
+
"grad_norm": NaN,
|
| 4175 |
+
"learning_rate": 0.00014298253551714776,
|
| 4176 |
+
"loss": 0.0,
|
| 4177 |
+
"step": 11780
|
| 4178 |
+
},
|
| 4179 |
+
{
|
| 4180 |
+
"epoch": 0.05338907514738552,
|
| 4181 |
+
"grad_norm": NaN,
|
| 4182 |
+
"learning_rate": 0.00014298253551714776,
|
| 4183 |
+
"loss": 0.0,
|
| 4184 |
+
"step": 11800
|
| 4185 |
+
},
|
| 4186 |
+
{
|
| 4187 |
+
"epoch": 0.053479565105262444,
|
| 4188 |
+
"grad_norm": NaN,
|
| 4189 |
+
"learning_rate": 0.00014298253551714776,
|
| 4190 |
+
"loss": 0.0,
|
| 4191 |
+
"step": 11820
|
| 4192 |
+
},
|
| 4193 |
+
{
|
| 4194 |
+
"epoch": 0.05357005506313937,
|
| 4195 |
+
"grad_norm": NaN,
|
| 4196 |
+
"learning_rate": 0.00014298253551714776,
|
| 4197 |
+
"loss": 0.0,
|
| 4198 |
+
"step": 11840
|
| 4199 |
+
},
|
| 4200 |
+
{
|
| 4201 |
+
"epoch": 0.05366054502101629,
|
| 4202 |
+
"grad_norm": NaN,
|
| 4203 |
+
"learning_rate": 0.00014298253551714776,
|
| 4204 |
+
"loss": 0.0,
|
| 4205 |
+
"step": 11860
|
| 4206 |
+
},
|
| 4207 |
+
{
|
| 4208 |
+
"epoch": 0.05375103497889322,
|
| 4209 |
+
"grad_norm": NaN,
|
| 4210 |
+
"learning_rate": 0.00014298253551714776,
|
| 4211 |
+
"loss": 0.0,
|
| 4212 |
+
"step": 11880
|
| 4213 |
+
},
|
| 4214 |
+
{
|
| 4215 |
+
"epoch": 0.05384152493677014,
|
| 4216 |
+
"grad_norm": NaN,
|
| 4217 |
+
"learning_rate": 0.00014298253551714776,
|
| 4218 |
+
"loss": 0.0,
|
| 4219 |
+
"step": 11900
|
| 4220 |
+
},
|
| 4221 |
+
{
|
| 4222 |
+
"epoch": 0.053932014894647065,
|
| 4223 |
+
"grad_norm": NaN,
|
| 4224 |
+
"learning_rate": 0.00014298253551714776,
|
| 4225 |
+
"loss": 0.0,
|
| 4226 |
+
"step": 11920
|
| 4227 |
+
},
|
| 4228 |
+
{
|
| 4229 |
+
"epoch": 0.05402250485252399,
|
| 4230 |
+
"grad_norm": NaN,
|
| 4231 |
+
"learning_rate": 0.00014298253551714776,
|
| 4232 |
+
"loss": 0.0,
|
| 4233 |
+
"step": 11940
|
| 4234 |
+
},
|
| 4235 |
+
{
|
| 4236 |
+
"epoch": 0.05411299481040092,
|
| 4237 |
+
"grad_norm": NaN,
|
| 4238 |
+
"learning_rate": 0.00014298253551714776,
|
| 4239 |
+
"loss": 0.0,
|
| 4240 |
+
"step": 11960
|
| 4241 |
+
},
|
| 4242 |
+
{
|
| 4243 |
+
"epoch": 0.05420348476827784,
|
| 4244 |
+
"grad_norm": NaN,
|
| 4245 |
+
"learning_rate": 0.00014298253551714776,
|
| 4246 |
+
"loss": 0.0,
|
| 4247 |
+
"step": 11980
|
| 4248 |
+
},
|
| 4249 |
+
{
|
| 4250 |
+
"epoch": 0.05429397472615476,
|
| 4251 |
+
"grad_norm": NaN,
|
| 4252 |
+
"learning_rate": 0.00014298253551714776,
|
| 4253 |
+
"loss": 0.0,
|
| 4254 |
+
"step": 12000
|
| 4255 |
+
},
|
| 4256 |
+
{
|
| 4257 |
+
"epoch": 0.05429397472615476,
|
| 4258 |
+
"eval_accuracy": 0.021626624590642192,
|
| 4259 |
+
"eval_loss": NaN,
|
| 4260 |
+
"eval_runtime": 218.9297,
|
| 4261 |
+
"eval_samples_per_second": 2776.417,
|
| 4262 |
+
"eval_steps_per_second": 10.848,
|
| 4263 |
+
"step": 12000
|
| 4264 |
}
|
| 4265 |
],
|
| 4266 |
"logging_steps": 20,
|
|
|
|
| 4268 |
"num_input_tokens_seen": 0,
|
| 4269 |
"num_train_epochs": 3,
|
| 4270 |
"save_steps": 100,
|
| 4271 |
+
"total_flos": 4315086323712000.0,
|
| 4272 |
"train_batch_size": 256,
|
| 4273 |
"trial_name": null,
|
| 4274 |
"trial_params": null
|