Commit
•
20997d7
1
Parent(s):
8910a07
Training in progress, step 450000
Browse files- last-checkpoint/optimizer.pt +1 -1
- last-checkpoint/pytorch_model.bin +1 -1
- last-checkpoint/rng_state_0.pth +1 -1
- last-checkpoint/rng_state_1.pth +1 -1
- last-checkpoint/rng_state_2.pth +1 -1
- last-checkpoint/rng_state_3.pth +1 -1
- last-checkpoint/rng_state_4.pth +1 -1
- last-checkpoint/rng_state_5.pth +1 -1
- last-checkpoint/rng_state_6.pth +1 -1
- last-checkpoint/rng_state_7.pth +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +2162 -6
- last-checkpoint/training_args.bin +1 -1
- pytorch_model.bin +1 -1
- runs/Mar31_08-12-57_t1v-n-7a44a9fa-w-0/1680250902.9870226/events.out.tfevents.1680250902.t1v-n-7a44a9fa-w-0.3132829.1 +3 -0
- runs/Mar31_08-12-57_t1v-n-7a44a9fa-w-0/events.out.tfevents.1680250902.t1v-n-7a44a9fa-w-0.3132829.0 +3 -0
- training_args.bin +1 -1
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2693742553
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5245f434060d2339380f0d68095e220a13ebd01a40a31dfcdc5e08a32be82dbd
|
| 3 |
size 2693742553
|
last-checkpoint/pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1346893675
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b2688ccc6d66b88b6fcc4941eb91cb8f99c587c38ad79a52f32dec77cce83f0c
|
| 3 |
size 1346893675
|
last-checkpoint/rng_state_0.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 13611
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:84d44465521496476110b3bef1a6de56a0a7b1c72fceacbd25c97129c2c5aa00
|
| 3 |
size 13611
|
last-checkpoint/rng_state_1.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 13611
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:84d44465521496476110b3bef1a6de56a0a7b1c72fceacbd25c97129c2c5aa00
|
| 3 |
size 13611
|
last-checkpoint/rng_state_2.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 13611
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:84d44465521496476110b3bef1a6de56a0a7b1c72fceacbd25c97129c2c5aa00
|
| 3 |
size 13611
|
last-checkpoint/rng_state_3.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 13611
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:84d44465521496476110b3bef1a6de56a0a7b1c72fceacbd25c97129c2c5aa00
|
| 3 |
size 13611
|
last-checkpoint/rng_state_4.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 13611
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:84d44465521496476110b3bef1a6de56a0a7b1c72fceacbd25c97129c2c5aa00
|
| 3 |
size 13611
|
last-checkpoint/rng_state_5.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 13611
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:84d44465521496476110b3bef1a6de56a0a7b1c72fceacbd25c97129c2c5aa00
|
| 3 |
size 13611
|
last-checkpoint/rng_state_6.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 13611
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:84d44465521496476110b3bef1a6de56a0a7b1c72fceacbd25c97129c2c5aa00
|
| 3 |
size 13611
|
last-checkpoint/rng_state_7.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 13611
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:84d44465521496476110b3bef1a6de56a0a7b1c72fceacbd25c97129c2c5aa00
|
| 3 |
size 13611
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 623
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:78e735efa7e40e0dd22dcac5cb3724b0cbe120563d603ea4b62f22b0f40fc602
|
| 3 |
size 623
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
-
"global_step":
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
@@ -618,15 +618,2171 @@
|
|
| 618 |
{
|
| 619 |
"epoch": 0.1,
|
| 620 |
"eval_loss": 0.3987598419189453,
|
| 621 |
-
"eval_runtime": 34.
|
| 622 |
-
"eval_samples_per_second": 144.
|
| 623 |
-
"eval_steps_per_second": 2.
|
| 624 |
"step": 100000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 625 |
}
|
| 626 |
],
|
| 627 |
"max_steps": 1000000,
|
| 628 |
"num_train_epochs": 9223372036854775807,
|
| 629 |
-
"total_flos":
|
| 630 |
"trial_name": null,
|
| 631 |
"trial_params": null
|
| 632 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.017126,
|
| 5 |
+
"global_step": 450000,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
|
|
| 618 |
{
|
| 619 |
"epoch": 0.1,
|
| 620 |
"eval_loss": 0.3987598419189453,
|
| 621 |
+
"eval_runtime": 34.6601,
|
| 622 |
+
"eval_samples_per_second": 144.258,
|
| 623 |
+
"eval_steps_per_second": 2.279,
|
| 624 |
"step": 100000
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 0.1,
|
| 628 |
+
"learning_rate": 9.929058033379181e-05,
|
| 629 |
+
"loss": 0.7347,
|
| 630 |
+
"step": 101000
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 0.1,
|
| 634 |
+
"learning_rate": 9.926255646355804e-05,
|
| 635 |
+
"loss": 0.7343,
|
| 636 |
+
"step": 102000
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 0.1,
|
| 640 |
+
"learning_rate": 9.923399386589933e-05,
|
| 641 |
+
"loss": 0.731,
|
| 642 |
+
"step": 103000
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 0.1,
|
| 646 |
+
"learning_rate": 9.92048928531717e-05,
|
| 647 |
+
"loss": 0.7306,
|
| 648 |
+
"step": 104000
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 0.1,
|
| 652 |
+
"learning_rate": 9.917525374361912e-05,
|
| 653 |
+
"loss": 0.7338,
|
| 654 |
+
"step": 105000
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.11,
|
| 658 |
+
"learning_rate": 9.914507686137019e-05,
|
| 659 |
+
"loss": 0.7435,
|
| 660 |
+
"step": 106000
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.11,
|
| 664 |
+
"learning_rate": 9.911436253643445e-05,
|
| 665 |
+
"loss": 0.7392,
|
| 666 |
+
"step": 107000
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 0.11,
|
| 670 |
+
"learning_rate": 9.90831111046988e-05,
|
| 671 |
+
"loss": 0.7355,
|
| 672 |
+
"step": 108000
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 0.11,
|
| 676 |
+
"learning_rate": 9.905132290792394e-05,
|
| 677 |
+
"loss": 0.7403,
|
| 678 |
+
"step": 109000
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 0.11,
|
| 682 |
+
"learning_rate": 9.901899829374047e-05,
|
| 683 |
+
"loss": 0.7348,
|
| 684 |
+
"step": 110000
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 0.11,
|
| 688 |
+
"learning_rate": 9.89861376156452e-05,
|
| 689 |
+
"loss": 0.7248,
|
| 690 |
+
"step": 111000
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 0.11,
|
| 694 |
+
"learning_rate": 9.895274123299723e-05,
|
| 695 |
+
"loss": 0.7224,
|
| 696 |
+
"step": 112000
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.11,
|
| 700 |
+
"learning_rate": 9.891880951101407e-05,
|
| 701 |
+
"loss": 0.7168,
|
| 702 |
+
"step": 113000
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.11,
|
| 706 |
+
"learning_rate": 9.888434282076758e-05,
|
| 707 |
+
"loss": 0.7186,
|
| 708 |
+
"step": 114000
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 0.12,
|
| 712 |
+
"learning_rate": 9.884934153917997e-05,
|
| 713 |
+
"loss": 0.7245,
|
| 714 |
+
"step": 115000
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 0.12,
|
| 718 |
+
"learning_rate": 9.881380604901964e-05,
|
| 719 |
+
"loss": 0.7274,
|
| 720 |
+
"step": 116000
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 0.12,
|
| 724 |
+
"learning_rate": 9.877773673889701e-05,
|
| 725 |
+
"loss": 0.7287,
|
| 726 |
+
"step": 117000
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 0.12,
|
| 730 |
+
"learning_rate": 9.87411340032603e-05,
|
| 731 |
+
"loss": 0.7225,
|
| 732 |
+
"step": 118000
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 0.12,
|
| 736 |
+
"learning_rate": 9.870399824239117e-05,
|
| 737 |
+
"loss": 0.7337,
|
| 738 |
+
"step": 119000
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.12,
|
| 742 |
+
"learning_rate": 9.86663298624003e-05,
|
| 743 |
+
"loss": 0.7229,
|
| 744 |
+
"step": 120000
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.12,
|
| 748 |
+
"learning_rate": 9.862812927522309e-05,
|
| 749 |
+
"loss": 0.7127,
|
| 750 |
+
"step": 121000
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 0.12,
|
| 754 |
+
"learning_rate": 9.858939689861506e-05,
|
| 755 |
+
"loss": 0.7089,
|
| 756 |
+
"step": 122000
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 0.12,
|
| 760 |
+
"learning_rate": 9.855013315614725e-05,
|
| 761 |
+
"loss": 0.7057,
|
| 762 |
+
"step": 123000
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 0.12,
|
| 766 |
+
"learning_rate": 9.851033847720166e-05,
|
| 767 |
+
"loss": 0.7075,
|
| 768 |
+
"step": 124000
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 0.12,
|
| 772 |
+
"learning_rate": 9.847001329696653e-05,
|
| 773 |
+
"loss": 0.716,
|
| 774 |
+
"step": 125000
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 0.13,
|
| 778 |
+
"learning_rate": 9.842915805643155e-05,
|
| 779 |
+
"loss": 0.7194,
|
| 780 |
+
"step": 126000
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.13,
|
| 784 |
+
"learning_rate": 9.838777320238312e-05,
|
| 785 |
+
"loss": 0.7189,
|
| 786 |
+
"step": 127000
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.13,
|
| 790 |
+
"learning_rate": 9.834585918739936e-05,
|
| 791 |
+
"loss": 0.7149,
|
| 792 |
+
"step": 128000
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 0.13,
|
| 796 |
+
"learning_rate": 9.830341646984521e-05,
|
| 797 |
+
"loss": 0.7232,
|
| 798 |
+
"step": 129000
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 0.13,
|
| 802 |
+
"learning_rate": 9.826044551386744e-05,
|
| 803 |
+
"loss": 0.7177,
|
| 804 |
+
"step": 130000
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 0.13,
|
| 808 |
+
"learning_rate": 9.821694678938953e-05,
|
| 809 |
+
"loss": 0.7018,
|
| 810 |
+
"step": 131000
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 0.13,
|
| 814 |
+
"learning_rate": 9.817292077210659e-05,
|
| 815 |
+
"loss": 0.7007,
|
| 816 |
+
"step": 132000
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 0.13,
|
| 820 |
+
"learning_rate": 9.812836794348004e-05,
|
| 821 |
+
"loss": 0.6968,
|
| 822 |
+
"step": 133000
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.13,
|
| 826 |
+
"learning_rate": 9.808328879073251e-05,
|
| 827 |
+
"loss": 0.702,
|
| 828 |
+
"step": 134000
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.14,
|
| 832 |
+
"learning_rate": 9.803768380684242e-05,
|
| 833 |
+
"loss": 0.7089,
|
| 834 |
+
"step": 135000
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 0.14,
|
| 838 |
+
"learning_rate": 9.799155349053851e-05,
|
| 839 |
+
"loss": 0.7109,
|
| 840 |
+
"step": 136000
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 0.14,
|
| 844 |
+
"learning_rate": 9.794489834629455e-05,
|
| 845 |
+
"loss": 0.7107,
|
| 846 |
+
"step": 137000
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 0.14,
|
| 850 |
+
"learning_rate": 9.789771888432375e-05,
|
| 851 |
+
"loss": 0.7074,
|
| 852 |
+
"step": 138000
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 0.14,
|
| 856 |
+
"learning_rate": 9.785001562057309e-05,
|
| 857 |
+
"loss": 0.7147,
|
| 858 |
+
"step": 139000
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 0.14,
|
| 862 |
+
"learning_rate": 9.780178907671789e-05,
|
| 863 |
+
"loss": 0.7074,
|
| 864 |
+
"step": 140000
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.14,
|
| 868 |
+
"learning_rate": 9.775303978015585e-05,
|
| 869 |
+
"loss": 0.6941,
|
| 870 |
+
"step": 141000
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.14,
|
| 874 |
+
"learning_rate": 9.77037682640015e-05,
|
| 875 |
+
"loss": 0.6937,
|
| 876 |
+
"step": 142000
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 0.14,
|
| 880 |
+
"learning_rate": 9.765397506708023e-05,
|
| 881 |
+
"loss": 0.6884,
|
| 882 |
+
"step": 143000
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 0.14,
|
| 886 |
+
"learning_rate": 9.760366073392246e-05,
|
| 887 |
+
"loss": 0.6907,
|
| 888 |
+
"step": 144000
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 1.0,
|
| 892 |
+
"learning_rate": 9.755282581475769e-05,
|
| 893 |
+
"loss": 0.6948,
|
| 894 |
+
"step": 145000
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 1.0,
|
| 898 |
+
"learning_rate": 9.750147086550844e-05,
|
| 899 |
+
"loss": 0.6918,
|
| 900 |
+
"step": 146000
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 1.0,
|
| 904 |
+
"learning_rate": 9.744959644778422e-05,
|
| 905 |
+
"loss": 0.699,
|
| 906 |
+
"step": 147000
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 1.0,
|
| 910 |
+
"learning_rate": 9.739720312887535e-05,
|
| 911 |
+
"loss": 0.699,
|
| 912 |
+
"step": 148000
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 1.0,
|
| 916 |
+
"learning_rate": 9.734429148174675e-05,
|
| 917 |
+
"loss": 0.7016,
|
| 918 |
+
"step": 149000
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 1.01,
|
| 922 |
+
"learning_rate": 9.729086208503174e-05,
|
| 923 |
+
"loss": 0.701,
|
| 924 |
+
"step": 150000
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 1.01,
|
| 928 |
+
"eval_loss": 0.3547865152359009,
|
| 929 |
+
"eval_runtime": 24.2232,
|
| 930 |
+
"eval_samples_per_second": 206.414,
|
| 931 |
+
"eval_steps_per_second": 3.261,
|
| 932 |
+
"step": 150000
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 1.01,
|
| 936 |
+
"learning_rate": 9.723691552302562e-05,
|
| 937 |
+
"loss": 0.7063,
|
| 938 |
+
"step": 151000
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 1.01,
|
| 942 |
+
"learning_rate": 9.718245238567939e-05,
|
| 943 |
+
"loss": 0.6952,
|
| 944 |
+
"step": 152000
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 1.01,
|
| 948 |
+
"learning_rate": 9.712747326859315e-05,
|
| 949 |
+
"loss": 0.6859,
|
| 950 |
+
"step": 153000
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 1.01,
|
| 954 |
+
"learning_rate": 9.707197877300974e-05,
|
| 955 |
+
"loss": 0.6848,
|
| 956 |
+
"step": 154000
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 1.01,
|
| 960 |
+
"learning_rate": 9.701596950580806e-05,
|
| 961 |
+
"loss": 0.684,
|
| 962 |
+
"step": 155000
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 1.01,
|
| 966 |
+
"learning_rate": 9.695944607949649e-05,
|
| 967 |
+
"loss": 0.6841,
|
| 968 |
+
"step": 156000
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 1.01,
|
| 972 |
+
"learning_rate": 9.690240911220618e-05,
|
| 973 |
+
"loss": 0.6924,
|
| 974 |
+
"step": 157000
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 1.01,
|
| 978 |
+
"learning_rate": 9.684485922768422e-05,
|
| 979 |
+
"loss": 0.6949,
|
| 980 |
+
"step": 158000
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 1.01,
|
| 984 |
+
"learning_rate": 9.6786797055287e-05,
|
| 985 |
+
"loss": 0.6914,
|
| 986 |
+
"step": 159000
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 1.02,
|
| 990 |
+
"learning_rate": 9.672822322997305e-05,
|
| 991 |
+
"loss": 0.6928,
|
| 992 |
+
"step": 160000
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 1.02,
|
| 996 |
+
"learning_rate": 9.66691383922964e-05,
|
| 997 |
+
"loss": 0.6963,
|
| 998 |
+
"step": 161000
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 1.02,
|
| 1002 |
+
"learning_rate": 9.660954318839933e-05,
|
| 1003 |
+
"loss": 0.6901,
|
| 1004 |
+
"step": 162000
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 1.02,
|
| 1008 |
+
"learning_rate": 9.654943827000548e-05,
|
| 1009 |
+
"loss": 0.6782,
|
| 1010 |
+
"step": 163000
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 1.02,
|
| 1014 |
+
"learning_rate": 9.648882429441257e-05,
|
| 1015 |
+
"loss": 0.6799,
|
| 1016 |
+
"step": 164000
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 1.02,
|
| 1020 |
+
"learning_rate": 9.642770192448536e-05,
|
| 1021 |
+
"loss": 0.6753,
|
| 1022 |
+
"step": 165000
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 1.02,
|
| 1026 |
+
"learning_rate": 9.636607182864827e-05,
|
| 1027 |
+
"loss": 0.6747,
|
| 1028 |
+
"step": 166000
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 1.02,
|
| 1032 |
+
"learning_rate": 9.630393468087818e-05,
|
| 1033 |
+
"loss": 0.689,
|
| 1034 |
+
"step": 167000
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 1.02,
|
| 1038 |
+
"learning_rate": 9.624129116069694e-05,
|
| 1039 |
+
"loss": 0.6869,
|
| 1040 |
+
"step": 168000
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 1.02,
|
| 1044 |
+
"learning_rate": 9.617814195316411e-05,
|
| 1045 |
+
"loss": 0.6863,
|
| 1046 |
+
"step": 169000
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 1.03,
|
| 1050 |
+
"learning_rate": 9.611448774886924e-05,
|
| 1051 |
+
"loss": 0.6848,
|
| 1052 |
+
"step": 170000
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 1.03,
|
| 1056 |
+
"learning_rate": 9.605032924392457e-05,
|
| 1057 |
+
"loss": 0.6935,
|
| 1058 |
+
"step": 171000
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 1.03,
|
| 1062 |
+
"learning_rate": 9.598566713995718e-05,
|
| 1063 |
+
"loss": 0.6813,
|
| 1064 |
+
"step": 172000
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 1.03,
|
| 1068 |
+
"learning_rate": 9.59205021441015e-05,
|
| 1069 |
+
"loss": 0.6707,
|
| 1070 |
+
"step": 173000
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 1.03,
|
| 1074 |
+
"learning_rate": 9.58548349689915e-05,
|
| 1075 |
+
"loss": 0.6704,
|
| 1076 |
+
"step": 174000
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 1.03,
|
| 1080 |
+
"learning_rate": 9.578866633275288e-05,
|
| 1081 |
+
"loss": 0.6673,
|
| 1082 |
+
"step": 175000
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 1.03,
|
| 1086 |
+
"learning_rate": 9.572199695899522e-05,
|
| 1087 |
+
"loss": 0.6675,
|
| 1088 |
+
"step": 176000
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 1.03,
|
| 1092 |
+
"learning_rate": 9.565482757680415e-05,
|
| 1093 |
+
"loss": 0.6806,
|
| 1094 |
+
"step": 177000
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 1.03,
|
| 1098 |
+
"learning_rate": 9.558715892073323e-05,
|
| 1099 |
+
"loss": 0.6779,
|
| 1100 |
+
"step": 178000
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 1.03,
|
| 1104 |
+
"learning_rate": 9.551899173079607e-05,
|
| 1105 |
+
"loss": 0.6801,
|
| 1106 |
+
"step": 179000
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 1.04,
|
| 1110 |
+
"learning_rate": 9.545032675245813e-05,
|
| 1111 |
+
"loss": 0.678,
|
| 1112 |
+
"step": 180000
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 1.04,
|
| 1116 |
+
"learning_rate": 9.538116473662861e-05,
|
| 1117 |
+
"loss": 0.686,
|
| 1118 |
+
"step": 181000
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 1.04,
|
| 1122 |
+
"learning_rate": 9.531150643965223e-05,
|
| 1123 |
+
"loss": 0.6725,
|
| 1124 |
+
"step": 182000
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 1.04,
|
| 1128 |
+
"learning_rate": 9.524135262330098e-05,
|
| 1129 |
+
"loss": 0.6652,
|
| 1130 |
+
"step": 183000
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 1.04,
|
| 1134 |
+
"learning_rate": 9.517070405476575e-05,
|
| 1135 |
+
"loss": 0.6624,
|
| 1136 |
+
"step": 184000
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 1.04,
|
| 1140 |
+
"learning_rate": 9.509956150664796e-05,
|
| 1141 |
+
"loss": 0.6607,
|
| 1142 |
+
"step": 185000
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 1.04,
|
| 1146 |
+
"learning_rate": 9.502792575695112e-05,
|
| 1147 |
+
"loss": 0.6636,
|
| 1148 |
+
"step": 186000
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 1.04,
|
| 1152 |
+
"learning_rate": 9.49557975890723e-05,
|
| 1153 |
+
"loss": 0.6736,
|
| 1154 |
+
"step": 187000
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 1.04,
|
| 1158 |
+
"learning_rate": 9.488317779179361e-05,
|
| 1159 |
+
"loss": 0.6735,
|
| 1160 |
+
"step": 188000
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 1.04,
|
| 1164 |
+
"learning_rate": 9.481006715927351e-05,
|
| 1165 |
+
"loss": 0.6704,
|
| 1166 |
+
"step": 189000
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 1.05,
|
| 1170 |
+
"learning_rate": 9.473646649103818e-05,
|
| 1171 |
+
"loss": 0.6707,
|
| 1172 |
+
"step": 190000
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 1.05,
|
| 1176 |
+
"learning_rate": 9.46623765919727e-05,
|
| 1177 |
+
"loss": 0.6797,
|
| 1178 |
+
"step": 191000
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 1.05,
|
| 1182 |
+
"learning_rate": 9.458779827231237e-05,
|
| 1183 |
+
"loss": 0.6671,
|
| 1184 |
+
"step": 192000
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 1.05,
|
| 1188 |
+
"learning_rate": 9.451273234763371e-05,
|
| 1189 |
+
"loss": 0.6577,
|
| 1190 |
+
"step": 193000
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 1.05,
|
| 1194 |
+
"learning_rate": 9.443717963884569e-05,
|
| 1195 |
+
"loss": 0.6595,
|
| 1196 |
+
"step": 194000
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 1.05,
|
| 1200 |
+
"learning_rate": 9.43611409721806e-05,
|
| 1201 |
+
"loss": 0.6549,
|
| 1202 |
+
"step": 195000
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 1.05,
|
| 1206 |
+
"learning_rate": 9.428461717918511e-05,
|
| 1207 |
+
"loss": 0.6587,
|
| 1208 |
+
"step": 196000
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 1.05,
|
| 1212 |
+
"learning_rate": 9.420760909671118e-05,
|
| 1213 |
+
"loss": 0.6683,
|
| 1214 |
+
"step": 197000
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 1.05,
|
| 1218 |
+
"learning_rate": 9.413011756690685e-05,
|
| 1219 |
+
"loss": 0.67,
|
| 1220 |
+
"step": 198000
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 1.05,
|
| 1224 |
+
"learning_rate": 9.405214343720707e-05,
|
| 1225 |
+
"loss": 0.6677,
|
| 1226 |
+
"step": 199000
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 1.06,
|
| 1230 |
+
"learning_rate": 9.397368756032445e-05,
|
| 1231 |
+
"loss": 0.6686,
|
| 1232 |
+
"step": 200000
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 1.06,
|
| 1236 |
+
"eval_loss": 0.32631468772888184,
|
| 1237 |
+
"eval_runtime": 28.1755,
|
| 1238 |
+
"eval_samples_per_second": 177.459,
|
| 1239 |
+
"eval_steps_per_second": 2.804,
|
| 1240 |
+
"step": 200000
|
| 1241 |
+
},
|
| 1242 |
+
{
|
| 1243 |
+
"epoch": 1.06,
|
| 1244 |
+
"learning_rate": 9.389475079423988e-05,
|
| 1245 |
+
"loss": 0.672,
|
| 1246 |
+
"step": 201000
|
| 1247 |
+
},
|
| 1248 |
+
{
|
| 1249 |
+
"epoch": 1.06,
|
| 1250 |
+
"learning_rate": 9.381533400219318e-05,
|
| 1251 |
+
"loss": 0.6604,
|
| 1252 |
+
"step": 202000
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 1.06,
|
| 1256 |
+
"learning_rate": 9.373543805267368e-05,
|
| 1257 |
+
"loss": 0.6547,
|
| 1258 |
+
"step": 203000
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 1.06,
|
| 1262 |
+
"learning_rate": 9.365506381941066e-05,
|
| 1263 |
+
"loss": 0.6533,
|
| 1264 |
+
"step": 204000
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 1.06,
|
| 1268 |
+
"learning_rate": 9.357421218136386e-05,
|
| 1269 |
+
"loss": 0.6521,
|
| 1270 |
+
"step": 205000
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 1.06,
|
| 1274 |
+
"learning_rate": 9.349288402271388e-05,
|
| 1275 |
+
"loss": 0.652,
|
| 1276 |
+
"step": 206000
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 1.06,
|
| 1280 |
+
"learning_rate": 9.341108023285238e-05,
|
| 1281 |
+
"loss": 0.6637,
|
| 1282 |
+
"step": 207000
|
| 1283 |
+
},
|
| 1284 |
+
{
|
| 1285 |
+
"epoch": 1.06,
|
| 1286 |
+
"learning_rate": 9.332880170637252e-05,
|
| 1287 |
+
"loss": 0.6635,
|
| 1288 |
+
"step": 208000
|
| 1289 |
+
},
|
| 1290 |
+
{
|
| 1291 |
+
"epoch": 1.06,
|
| 1292 |
+
"learning_rate": 9.32460493430591e-05,
|
| 1293 |
+
"loss": 0.6617,
|
| 1294 |
+
"step": 209000
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 1.07,
|
| 1298 |
+
"learning_rate": 9.316282404787871e-05,
|
| 1299 |
+
"loss": 0.664,
|
| 1300 |
+
"step": 210000
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 1.07,
|
| 1304 |
+
"learning_rate": 9.30791267309698e-05,
|
| 1305 |
+
"loss": 0.6704,
|
| 1306 |
+
"step": 211000
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 1.07,
|
| 1310 |
+
"learning_rate": 9.299495830763286e-05,
|
| 1311 |
+
"loss": 0.6563,
|
| 1312 |
+
"step": 212000
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 1.07,
|
| 1316 |
+
"learning_rate": 9.291031969832026e-05,
|
| 1317 |
+
"loss": 0.651,
|
| 1318 |
+
"step": 213000
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 1.07,
|
| 1322 |
+
"learning_rate": 9.282521182862629e-05,
|
| 1323 |
+
"loss": 0.6487,
|
| 1324 |
+
"step": 214000
|
| 1325 |
+
},
|
| 1326 |
+
{
|
| 1327 |
+
"epoch": 1.07,
|
| 1328 |
+
"learning_rate": 9.273963562927695e-05,
|
| 1329 |
+
"loss": 0.6466,
|
| 1330 |
+
"step": 215000
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 1.07,
|
| 1334 |
+
"learning_rate": 9.265359203611987e-05,
|
| 1335 |
+
"loss": 0.6481,
|
| 1336 |
+
"step": 216000
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 1.07,
|
| 1340 |
+
"learning_rate": 9.256708199011401e-05,
|
| 1341 |
+
"loss": 0.6622,
|
| 1342 |
+
"step": 217000
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 1.07,
|
| 1346 |
+
"learning_rate": 9.248010643731935e-05,
|
| 1347 |
+
"loss": 0.6625,
|
| 1348 |
+
"step": 218000
|
| 1349 |
+
},
|
| 1350 |
+
{
|
| 1351 |
+
"epoch": 1.07,
|
| 1352 |
+
"learning_rate": 9.239266632888659e-05,
|
| 1353 |
+
"loss": 0.6595,
|
| 1354 |
+
"step": 219000
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 1.08,
|
| 1358 |
+
"learning_rate": 9.230476262104677e-05,
|
| 1359 |
+
"loss": 0.6614,
|
| 1360 |
+
"step": 220000
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 1.08,
|
| 1364 |
+
"learning_rate": 9.221639627510076e-05,
|
| 1365 |
+
"loss": 0.6689,
|
| 1366 |
+
"step": 221000
|
| 1367 |
+
},
|
| 1368 |
+
{
|
| 1369 |
+
"epoch": 1.08,
|
| 1370 |
+
"learning_rate": 9.212756825740873e-05,
|
| 1371 |
+
"loss": 0.6548,
|
| 1372 |
+
"step": 222000
|
| 1373 |
+
},
|
| 1374 |
+
{
|
| 1375 |
+
"epoch": 1.08,
|
| 1376 |
+
"learning_rate": 9.20382795393797e-05,
|
| 1377 |
+
"loss": 0.6505,
|
| 1378 |
+
"step": 223000
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 1.08,
|
| 1382 |
+
"learning_rate": 9.194853109746074e-05,
|
| 1383 |
+
"loss": 0.6469,
|
| 1384 |
+
"step": 224000
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 1.08,
|
| 1388 |
+
"learning_rate": 9.185832391312644e-05,
|
| 1389 |
+
"loss": 0.6457,
|
| 1390 |
+
"step": 225000
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 1.08,
|
| 1394 |
+
"learning_rate": 9.176765897286813e-05,
|
| 1395 |
+
"loss": 0.6457,
|
| 1396 |
+
"step": 226000
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 1.08,
|
| 1400 |
+
"learning_rate": 9.167653726818305e-05,
|
| 1401 |
+
"loss": 0.658,
|
| 1402 |
+
"step": 227000
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 1.08,
|
| 1406 |
+
"learning_rate": 9.158495979556358e-05,
|
| 1407 |
+
"loss": 0.6599,
|
| 1408 |
+
"step": 228000
|
| 1409 |
+
},
|
| 1410 |
+
{
|
| 1411 |
+
"epoch": 1.08,
|
| 1412 |
+
"learning_rate": 9.14929275564863e-05,
|
| 1413 |
+
"loss": 0.6557,
|
| 1414 |
+
"step": 229000
|
| 1415 |
+
},
|
| 1416 |
+
{
|
| 1417 |
+
"epoch": 1.09,
|
| 1418 |
+
"learning_rate": 9.140044155740101e-05,
|
| 1419 |
+
"loss": 0.6558,
|
| 1420 |
+
"step": 230000
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 1.09,
|
| 1424 |
+
"learning_rate": 9.130750280971978e-05,
|
| 1425 |
+
"loss": 0.6647,
|
| 1426 |
+
"step": 231000
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 1.09,
|
| 1430 |
+
"learning_rate": 9.121411232980588e-05,
|
| 1431 |
+
"loss": 0.6475,
|
| 1432 |
+
"step": 232000
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 1.09,
|
| 1436 |
+
"learning_rate": 9.112027113896262e-05,
|
| 1437 |
+
"loss": 0.6444,
|
| 1438 |
+
"step": 233000
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 1.09,
|
| 1442 |
+
"learning_rate": 9.102598026342222e-05,
|
| 1443 |
+
"loss": 0.6406,
|
| 1444 |
+
"step": 234000
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 1.09,
|
| 1448 |
+
"learning_rate": 9.093124073433463e-05,
|
| 1449 |
+
"loss": 0.6421,
|
| 1450 |
+
"step": 235000
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"epoch": 1.09,
|
| 1454 |
+
"learning_rate": 9.083605358775612e-05,
|
| 1455 |
+
"loss": 0.6437,
|
| 1456 |
+
"step": 236000
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 1.09,
|
| 1460 |
+
"learning_rate": 9.074041986463808e-05,
|
| 1461 |
+
"loss": 0.6509,
|
| 1462 |
+
"step": 237000
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 1.09,
|
| 1466 |
+
"learning_rate": 9.064434061081562e-05,
|
| 1467 |
+
"loss": 0.652,
|
| 1468 |
+
"step": 238000
|
| 1469 |
+
},
|
| 1470 |
+
{
|
| 1471 |
+
"epoch": 1.09,
|
| 1472 |
+
"learning_rate": 9.0547816876996e-05,
|
| 1473 |
+
"loss": 0.6486,
|
| 1474 |
+
"step": 239000
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 1.1,
|
| 1478 |
+
"learning_rate": 9.045084971874738e-05,
|
| 1479 |
+
"loss": 0.6538,
|
| 1480 |
+
"step": 240000
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 1.1,
|
| 1484 |
+
"learning_rate": 9.035344019648702e-05,
|
| 1485 |
+
"loss": 0.6612,
|
| 1486 |
+
"step": 241000
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 1.1,
|
| 1490 |
+
"learning_rate": 9.025558937546988e-05,
|
| 1491 |
+
"loss": 0.6411,
|
| 1492 |
+
"step": 242000
|
| 1493 |
+
},
|
| 1494 |
+
{
|
| 1495 |
+
"epoch": 1.1,
|
| 1496 |
+
"learning_rate": 9.015729832577681e-05,
|
| 1497 |
+
"loss": 0.6402,
|
| 1498 |
+
"step": 243000
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 1.1,
|
| 1502 |
+
"learning_rate": 9.005856812230304e-05,
|
| 1503 |
+
"loss": 0.636,
|
| 1504 |
+
"step": 244000
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 1.1,
|
| 1508 |
+
"learning_rate": 8.995939984474624e-05,
|
| 1509 |
+
"loss": 0.6333,
|
| 1510 |
+
"step": 245000
|
| 1511 |
+
},
|
| 1512 |
+
{
|
| 1513 |
+
"epoch": 1.1,
|
| 1514 |
+
"learning_rate": 8.98597945775948e-05,
|
| 1515 |
+
"loss": 0.6392,
|
| 1516 |
+
"step": 246000
|
| 1517 |
+
},
|
| 1518 |
+
{
|
| 1519 |
+
"epoch": 1.1,
|
| 1520 |
+
"learning_rate": 8.975975341011596e-05,
|
| 1521 |
+
"loss": 0.6493,
|
| 1522 |
+
"step": 247000
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 1.1,
|
| 1526 |
+
"learning_rate": 8.965927743634391e-05,
|
| 1527 |
+
"loss": 0.6501,
|
| 1528 |
+
"step": 248000
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 1.1,
|
| 1532 |
+
"learning_rate": 8.955836775506776e-05,
|
| 1533 |
+
"loss": 0.6451,
|
| 1534 |
+
"step": 249000
|
| 1535 |
+
},
|
| 1536 |
+
{
|
| 1537 |
+
"epoch": 1.11,
|
| 1538 |
+
"learning_rate": 8.945702546981969e-05,
|
| 1539 |
+
"loss": 0.6517,
|
| 1540 |
+
"step": 250000
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 1.11,
|
| 1544 |
+
"eval_loss": 0.31288740038871765,
|
| 1545 |
+
"eval_runtime": 25.786,
|
| 1546 |
+
"eval_samples_per_second": 193.904,
|
| 1547 |
+
"eval_steps_per_second": 3.064,
|
| 1548 |
+
"step": 250000
|
| 1549 |
+
},
|
| 1550 |
+
{
|
| 1551 |
+
"epoch": 1.11,
|
| 1552 |
+
"learning_rate": 8.935525168886262e-05,
|
| 1553 |
+
"loss": 0.6559,
|
| 1554 |
+
"step": 251000
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 1.11,
|
| 1558 |
+
"learning_rate": 8.92530475251784e-05,
|
| 1559 |
+
"loss": 0.6392,
|
| 1560 |
+
"step": 252000
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"epoch": 1.11,
|
| 1564 |
+
"learning_rate": 8.91504140964553e-05,
|
| 1565 |
+
"loss": 0.6357,
|
| 1566 |
+
"step": 253000
|
| 1567 |
+
},
|
| 1568 |
+
{
|
| 1569 |
+
"epoch": 1.11,
|
| 1570 |
+
"learning_rate": 8.90473525250761e-05,
|
| 1571 |
+
"loss": 0.6323,
|
| 1572 |
+
"step": 254000
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"epoch": 1.11,
|
| 1576 |
+
"learning_rate": 8.894386393810563e-05,
|
| 1577 |
+
"loss": 0.6344,
|
| 1578 |
+
"step": 255000
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 1.11,
|
| 1582 |
+
"learning_rate": 8.883994946727849e-05,
|
| 1583 |
+
"loss": 0.6396,
|
| 1584 |
+
"step": 256000
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 1.11,
|
| 1588 |
+
"learning_rate": 8.873561024898668e-05,
|
| 1589 |
+
"loss": 0.648,
|
| 1590 |
+
"step": 257000
|
| 1591 |
+
},
|
| 1592 |
+
{
|
| 1593 |
+
"epoch": 1.11,
|
| 1594 |
+
"learning_rate": 8.863084742426719e-05,
|
| 1595 |
+
"loss": 0.6504,
|
| 1596 |
+
"step": 258000
|
| 1597 |
+
},
|
| 1598 |
+
{
|
| 1599 |
+
"epoch": 1.11,
|
| 1600 |
+
"learning_rate": 8.852566213878947e-05,
|
| 1601 |
+
"loss": 0.6472,
|
| 1602 |
+
"step": 259000
|
| 1603 |
+
},
|
| 1604 |
+
{
|
| 1605 |
+
"epoch": 1.12,
|
| 1606 |
+
"learning_rate": 8.842005554284296e-05,
|
| 1607 |
+
"loss": 0.6483,
|
| 1608 |
+
"step": 260000
|
| 1609 |
+
},
|
| 1610 |
+
{
|
| 1611 |
+
"epoch": 1.12,
|
| 1612 |
+
"learning_rate": 8.831402879132446e-05,
|
| 1613 |
+
"loss": 0.6542,
|
| 1614 |
+
"step": 261000
|
| 1615 |
+
},
|
| 1616 |
+
{
|
| 1617 |
+
"epoch": 1.12,
|
| 1618 |
+
"learning_rate": 8.820758304372557e-05,
|
| 1619 |
+
"loss": 0.6348,
|
| 1620 |
+
"step": 262000
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 1.12,
|
| 1624 |
+
"learning_rate": 8.810071946411989e-05,
|
| 1625 |
+
"loss": 0.6371,
|
| 1626 |
+
"step": 263000
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 1.12,
|
| 1630 |
+
"learning_rate": 8.799343922115044e-05,
|
| 1631 |
+
"loss": 0.6306,
|
| 1632 |
+
"step": 264000
|
| 1633 |
+
},
|
| 1634 |
+
{
|
| 1635 |
+
"epoch": 1.12,
|
| 1636 |
+
"learning_rate": 8.788574348801675e-05,
|
| 1637 |
+
"loss": 0.629,
|
| 1638 |
+
"step": 265000
|
| 1639 |
+
},
|
| 1640 |
+
{
|
| 1641 |
+
"epoch": 1.12,
|
| 1642 |
+
"learning_rate": 8.77776334424621e-05,
|
| 1643 |
+
"loss": 0.6349,
|
| 1644 |
+
"step": 266000
|
| 1645 |
+
},
|
| 1646 |
+
{
|
| 1647 |
+
"epoch": 1.12,
|
| 1648 |
+
"learning_rate": 8.766911026676064e-05,
|
| 1649 |
+
"loss": 0.6428,
|
| 1650 |
+
"step": 267000
|
| 1651 |
+
},
|
| 1652 |
+
{
|
| 1653 |
+
"epoch": 1.12,
|
| 1654 |
+
"learning_rate": 8.756017514770443e-05,
|
| 1655 |
+
"loss": 0.6442,
|
| 1656 |
+
"step": 268000
|
| 1657 |
+
},
|
| 1658 |
+
{
|
| 1659 |
+
"epoch": 1.12,
|
| 1660 |
+
"learning_rate": 8.745082927659047e-05,
|
| 1661 |
+
"loss": 0.6393,
|
| 1662 |
+
"step": 269000
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 1.13,
|
| 1666 |
+
"learning_rate": 8.73410738492077e-05,
|
| 1667 |
+
"loss": 0.6421,
|
| 1668 |
+
"step": 270000
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 1.13,
|
| 1672 |
+
"learning_rate": 8.723091006582389e-05,
|
| 1673 |
+
"loss": 0.65,
|
| 1674 |
+
"step": 271000
|
| 1675 |
+
},
|
| 1676 |
+
{
|
| 1677 |
+
"epoch": 1.13,
|
| 1678 |
+
"learning_rate": 8.71203391311725e-05,
|
| 1679 |
+
"loss": 0.6294,
|
| 1680 |
+
"step": 272000
|
| 1681 |
+
},
|
| 1682 |
+
{
|
| 1683 |
+
"epoch": 1.13,
|
| 1684 |
+
"learning_rate": 8.700936225443959e-05,
|
| 1685 |
+
"loss": 0.6304,
|
| 1686 |
+
"step": 273000
|
| 1687 |
+
},
|
| 1688 |
+
{
|
| 1689 |
+
"epoch": 1.13,
|
| 1690 |
+
"learning_rate": 8.689798064925049e-05,
|
| 1691 |
+
"loss": 0.6259,
|
| 1692 |
+
"step": 274000
|
| 1693 |
+
},
|
| 1694 |
+
{
|
| 1695 |
+
"epoch": 1.13,
|
| 1696 |
+
"learning_rate": 8.678619553365659e-05,
|
| 1697 |
+
"loss": 0.626,
|
| 1698 |
+
"step": 275000
|
| 1699 |
+
},
|
| 1700 |
+
{
|
| 1701 |
+
"epoch": 1.13,
|
| 1702 |
+
"learning_rate": 8.6674008130122e-05,
|
| 1703 |
+
"loss": 0.6289,
|
| 1704 |
+
"step": 276000
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 1.13,
|
| 1708 |
+
"learning_rate": 8.656141966551019e-05,
|
| 1709 |
+
"loss": 0.6381,
|
| 1710 |
+
"step": 277000
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 1.13,
|
| 1714 |
+
"learning_rate": 8.644843137107059e-05,
|
| 1715 |
+
"loss": 0.64,
|
| 1716 |
+
"step": 278000
|
| 1717 |
+
},
|
| 1718 |
+
{
|
| 1719 |
+
"epoch": 1.13,
|
| 1720 |
+
"learning_rate": 8.633504448242505e-05,
|
| 1721 |
+
"loss": 0.6367,
|
| 1722 |
+
"step": 279000
|
| 1723 |
+
},
|
| 1724 |
+
{
|
| 1725 |
+
"epoch": 1.14,
|
| 1726 |
+
"learning_rate": 8.622126023955446e-05,
|
| 1727 |
+
"loss": 0.6409,
|
| 1728 |
+
"step": 280000
|
| 1729 |
+
},
|
| 1730 |
+
{
|
| 1731 |
+
"epoch": 1.14,
|
| 1732 |
+
"learning_rate": 8.610707988678503e-05,
|
| 1733 |
+
"loss": 0.6465,
|
| 1734 |
+
"step": 281000
|
| 1735 |
+
},
|
| 1736 |
+
{
|
| 1737 |
+
"epoch": 1.14,
|
| 1738 |
+
"learning_rate": 8.599250467277483e-05,
|
| 1739 |
+
"loss": 0.6268,
|
| 1740 |
+
"step": 282000
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"epoch": 1.14,
|
| 1744 |
+
"learning_rate": 8.587753585050004e-05,
|
| 1745 |
+
"loss": 0.6251,
|
| 1746 |
+
"step": 283000
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 1.14,
|
| 1750 |
+
"learning_rate": 8.576217467724128e-05,
|
| 1751 |
+
"loss": 0.6213,
|
| 1752 |
+
"step": 284000
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 1.14,
|
| 1756 |
+
"learning_rate": 8.564642241456986e-05,
|
| 1757 |
+
"loss": 0.6194,
|
| 1758 |
+
"step": 285000
|
| 1759 |
+
},
|
| 1760 |
+
{
|
| 1761 |
+
"epoch": 1.14,
|
| 1762 |
+
"learning_rate": 8.553028032833397e-05,
|
| 1763 |
+
"loss": 0.6259,
|
| 1764 |
+
"step": 286000
|
| 1765 |
+
},
|
| 1766 |
+
{
|
| 1767 |
+
"epoch": 1.14,
|
| 1768 |
+
"learning_rate": 8.541374968864487e-05,
|
| 1769 |
+
"loss": 0.6333,
|
| 1770 |
+
"step": 287000
|
| 1771 |
+
},
|
| 1772 |
+
{
|
| 1773 |
+
"epoch": 1.14,
|
| 1774 |
+
"learning_rate": 8.529683176986295e-05,
|
| 1775 |
+
"loss": 0.6338,
|
| 1776 |
+
"step": 288000
|
| 1777 |
+
},
|
| 1778 |
+
{
|
| 1779 |
+
"epoch": 2.0,
|
| 1780 |
+
"learning_rate": 8.517952785058385e-05,
|
| 1781 |
+
"loss": 0.6123,
|
| 1782 |
+
"step": 289000
|
| 1783 |
+
},
|
| 1784 |
+
{
|
| 1785 |
+
"epoch": 2.0,
|
| 1786 |
+
"learning_rate": 8.506183921362443e-05,
|
| 1787 |
+
"loss": 0.6328,
|
| 1788 |
+
"step": 290000
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 2.0,
|
| 1792 |
+
"learning_rate": 8.494376714600878e-05,
|
| 1793 |
+
"loss": 0.6317,
|
| 1794 |
+
"step": 291000
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 2.0,
|
| 1798 |
+
"learning_rate": 8.482531293895412e-05,
|
| 1799 |
+
"loss": 0.6357,
|
| 1800 |
+
"step": 292000
|
| 1801 |
+
},
|
| 1802 |
+
{
|
| 1803 |
+
"epoch": 2.0,
|
| 1804 |
+
"learning_rate": 8.470647788785665e-05,
|
| 1805 |
+
"loss": 0.6378,
|
| 1806 |
+
"step": 293000
|
| 1807 |
+
},
|
| 1808 |
+
{
|
| 1809 |
+
"epoch": 2.01,
|
| 1810 |
+
"learning_rate": 8.458726329227747e-05,
|
| 1811 |
+
"loss": 0.6184,
|
| 1812 |
+
"step": 294000
|
| 1813 |
+
},
|
| 1814 |
+
{
|
| 1815 |
+
"epoch": 2.01,
|
| 1816 |
+
"learning_rate": 8.44676704559283e-05,
|
| 1817 |
+
"loss": 0.6205,
|
| 1818 |
+
"step": 295000
|
| 1819 |
+
},
|
| 1820 |
+
{
|
| 1821 |
+
"epoch": 2.01,
|
| 1822 |
+
"learning_rate": 8.434770068665723e-05,
|
| 1823 |
+
"loss": 0.6144,
|
| 1824 |
+
"step": 296000
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"epoch": 2.01,
|
| 1828 |
+
"learning_rate": 8.422735529643444e-05,
|
| 1829 |
+
"loss": 0.6131,
|
| 1830 |
+
"step": 297000
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 2.01,
|
| 1834 |
+
"learning_rate": 8.410663560133784e-05,
|
| 1835 |
+
"loss": 0.6221,
|
| 1836 |
+
"step": 298000
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 2.01,
|
| 1840 |
+
"learning_rate": 8.398554292153866e-05,
|
| 1841 |
+
"loss": 0.6283,
|
| 1842 |
+
"step": 299000
|
| 1843 |
+
},
|
| 1844 |
+
{
|
| 1845 |
+
"epoch": 2.01,
|
| 1846 |
+
"learning_rate": 8.386407858128706e-05,
|
| 1847 |
+
"loss": 0.6292,
|
| 1848 |
+
"step": 300000
|
| 1849 |
+
},
|
| 1850 |
+
{
|
| 1851 |
+
"epoch": 2.01,
|
| 1852 |
+
"eval_loss": 0.29386115074157715,
|
| 1853 |
+
"eval_runtime": 25.9723,
|
| 1854 |
+
"eval_samples_per_second": 192.513,
|
| 1855 |
+
"eval_steps_per_second": 3.042,
|
| 1856 |
+
"step": 300000
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 2.01,
|
| 1860 |
+
"learning_rate": 8.37422439088976e-05,
|
| 1861 |
+
"loss": 0.6274,
|
| 1862 |
+
"step": 301000
|
| 1863 |
+
},
|
| 1864 |
+
{
|
| 1865 |
+
"epoch": 2.01,
|
| 1866 |
+
"learning_rate": 8.362004023673474e-05,
|
| 1867 |
+
"loss": 0.6359,
|
| 1868 |
+
"step": 302000
|
| 1869 |
+
},
|
| 1870 |
+
{
|
| 1871 |
+
"epoch": 2.01,
|
| 1872 |
+
"learning_rate": 8.349746890119826e-05,
|
| 1873 |
+
"loss": 0.6329,
|
| 1874 |
+
"step": 303000
|
| 1875 |
+
},
|
| 1876 |
+
{
|
| 1877 |
+
"epoch": 2.02,
|
| 1878 |
+
"learning_rate": 8.337453124270863e-05,
|
| 1879 |
+
"loss": 0.614,
|
| 1880 |
+
"step": 304000
|
| 1881 |
+
},
|
| 1882 |
+
{
|
| 1883 |
+
"epoch": 2.02,
|
| 1884 |
+
"learning_rate": 8.32512286056924e-05,
|
| 1885 |
+
"loss": 0.6146,
|
| 1886 |
+
"step": 305000
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 2.02,
|
| 1890 |
+
"learning_rate": 8.31275623385675e-05,
|
| 1891 |
+
"loss": 0.6135,
|
| 1892 |
+
"step": 306000
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 2.02,
|
| 1896 |
+
"learning_rate": 8.300353379372834e-05,
|
| 1897 |
+
"loss": 0.613,
|
| 1898 |
+
"step": 307000
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 2.02,
|
| 1902 |
+
"learning_rate": 8.287914432753123e-05,
|
| 1903 |
+
"loss": 0.6225,
|
| 1904 |
+
"step": 308000
|
| 1905 |
+
},
|
| 1906 |
+
{
|
| 1907 |
+
"epoch": 2.02,
|
| 1908 |
+
"learning_rate": 8.275439530027948e-05,
|
| 1909 |
+
"loss": 0.6267,
|
| 1910 |
+
"step": 309000
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"epoch": 2.02,
|
| 1914 |
+
"learning_rate": 8.262928807620843e-05,
|
| 1915 |
+
"loss": 0.6261,
|
| 1916 |
+
"step": 310000
|
| 1917 |
+
},
|
| 1918 |
+
{
|
| 1919 |
+
"epoch": 2.02,
|
| 1920 |
+
"learning_rate": 8.250382402347065e-05,
|
| 1921 |
+
"loss": 0.6269,
|
| 1922 |
+
"step": 311000
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"epoch": 2.02,
|
| 1926 |
+
"learning_rate": 8.237800451412095e-05,
|
| 1927 |
+
"loss": 0.6301,
|
| 1928 |
+
"step": 312000
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 2.02,
|
| 1932 |
+
"learning_rate": 8.225183092410128e-05,
|
| 1933 |
+
"loss": 0.6285,
|
| 1934 |
+
"step": 313000
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 2.03,
|
| 1938 |
+
"learning_rate": 8.212530463322583e-05,
|
| 1939 |
+
"loss": 0.6122,
|
| 1940 |
+
"step": 314000
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 2.03,
|
| 1944 |
+
"learning_rate": 8.199842702516583e-05,
|
| 1945 |
+
"loss": 0.6132,
|
| 1946 |
+
"step": 315000
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"epoch": 2.03,
|
| 1950 |
+
"learning_rate": 8.18711994874345e-05,
|
| 1951 |
+
"loss": 0.6114,
|
| 1952 |
+
"step": 316000
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 2.03,
|
| 1956 |
+
"learning_rate": 8.174362341137177e-05,
|
| 1957 |
+
"loss": 0.6084,
|
| 1958 |
+
"step": 317000
|
| 1959 |
+
},
|
| 1960 |
+
{
|
| 1961 |
+
"epoch": 2.03,
|
| 1962 |
+
"learning_rate": 8.161570019212921e-05,
|
| 1963 |
+
"loss": 0.6162,
|
| 1964 |
+
"step": 318000
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 2.03,
|
| 1968 |
+
"learning_rate": 8.148743122865463e-05,
|
| 1969 |
+
"loss": 0.6225,
|
| 1970 |
+
"step": 319000
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 2.03,
|
| 1974 |
+
"learning_rate": 8.135881792367686e-05,
|
| 1975 |
+
"loss": 0.6211,
|
| 1976 |
+
"step": 320000
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 2.03,
|
| 1980 |
+
"learning_rate": 8.12298616836904e-05,
|
| 1981 |
+
"loss": 0.6201,
|
| 1982 |
+
"step": 321000
|
| 1983 |
+
},
|
| 1984 |
+
{
|
| 1985 |
+
"epoch": 2.03,
|
| 1986 |
+
"learning_rate": 8.110056391894005e-05,
|
| 1987 |
+
"loss": 0.6266,
|
| 1988 |
+
"step": 322000
|
| 1989 |
+
},
|
| 1990 |
+
{
|
| 1991 |
+
"epoch": 2.03,
|
| 1992 |
+
"learning_rate": 8.097092604340542e-05,
|
| 1993 |
+
"loss": 0.6257,
|
| 1994 |
+
"step": 323000
|
| 1995 |
+
},
|
| 1996 |
+
{
|
| 1997 |
+
"epoch": 2.04,
|
| 1998 |
+
"learning_rate": 8.084094947478556e-05,
|
| 1999 |
+
"loss": 0.6087,
|
| 2000 |
+
"step": 324000
|
| 2001 |
+
},
|
| 2002 |
+
{
|
| 2003 |
+
"epoch": 2.04,
|
| 2004 |
+
"learning_rate": 8.07106356344834e-05,
|
| 2005 |
+
"loss": 0.6109,
|
| 2006 |
+
"step": 325000
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 2.04,
|
| 2010 |
+
"learning_rate": 8.057998594759022e-05,
|
| 2011 |
+
"loss": 0.6044,
|
| 2012 |
+
"step": 326000
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 2.04,
|
| 2016 |
+
"learning_rate": 8.044900184287007e-05,
|
| 2017 |
+
"loss": 0.6034,
|
| 2018 |
+
"step": 327000
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 2.04,
|
| 2022 |
+
"learning_rate": 8.031768475274413e-05,
|
| 2023 |
+
"loss": 0.6144,
|
| 2024 |
+
"step": 328000
|
| 2025 |
+
},
|
| 2026 |
+
{
|
| 2027 |
+
"epoch": 2.04,
|
| 2028 |
+
"learning_rate": 8.018603611327504e-05,
|
| 2029 |
+
"loss": 0.6189,
|
| 2030 |
+
"step": 329000
|
| 2031 |
+
},
|
| 2032 |
+
{
|
| 2033 |
+
"epoch": 2.04,
|
| 2034 |
+
"learning_rate": 8.005405736415126e-05,
|
| 2035 |
+
"loss": 0.618,
|
| 2036 |
+
"step": 330000
|
| 2037 |
+
},
|
| 2038 |
+
{
|
| 2039 |
+
"epoch": 2.04,
|
| 2040 |
+
"learning_rate": 7.992174994867123e-05,
|
| 2041 |
+
"loss": 0.6175,
|
| 2042 |
+
"step": 331000
|
| 2043 |
+
},
|
| 2044 |
+
{
|
| 2045 |
+
"epoch": 2.04,
|
| 2046 |
+
"learning_rate": 7.978911531372765e-05,
|
| 2047 |
+
"loss": 0.6242,
|
| 2048 |
+
"step": 332000
|
| 2049 |
+
},
|
| 2050 |
+
{
|
| 2051 |
+
"epoch": 2.04,
|
| 2052 |
+
"learning_rate": 7.965615490979163e-05,
|
| 2053 |
+
"loss": 0.619,
|
| 2054 |
+
"step": 333000
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 2.05,
|
| 2058 |
+
"learning_rate": 7.952287019089685e-05,
|
| 2059 |
+
"loss": 0.604,
|
| 2060 |
+
"step": 334000
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 2.05,
|
| 2064 |
+
"learning_rate": 7.938926261462366e-05,
|
| 2065 |
+
"loss": 0.6039,
|
| 2066 |
+
"step": 335000
|
| 2067 |
+
},
|
| 2068 |
+
{
|
| 2069 |
+
"epoch": 2.05,
|
| 2070 |
+
"learning_rate": 7.925533364208309e-05,
|
| 2071 |
+
"loss": 0.6025,
|
| 2072 |
+
"step": 336000
|
| 2073 |
+
},
|
| 2074 |
+
{
|
| 2075 |
+
"epoch": 2.05,
|
| 2076 |
+
"learning_rate": 7.912108473790092e-05,
|
| 2077 |
+
"loss": 0.5991,
|
| 2078 |
+
"step": 337000
|
| 2079 |
+
},
|
| 2080 |
+
{
|
| 2081 |
+
"epoch": 2.05,
|
| 2082 |
+
"learning_rate": 7.898651737020166e-05,
|
| 2083 |
+
"loss": 0.6119,
|
| 2084 |
+
"step": 338000
|
| 2085 |
+
},
|
| 2086 |
+
{
|
| 2087 |
+
"epoch": 2.05,
|
| 2088 |
+
"learning_rate": 7.88516330105925e-05,
|
| 2089 |
+
"loss": 0.6138,
|
| 2090 |
+
"step": 339000
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"epoch": 2.05,
|
| 2094 |
+
"learning_rate": 7.871643313414718e-05,
|
| 2095 |
+
"loss": 0.6134,
|
| 2096 |
+
"step": 340000
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 2.05,
|
| 2100 |
+
"learning_rate": 7.858091921938988e-05,
|
| 2101 |
+
"loss": 0.6137,
|
| 2102 |
+
"step": 341000
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 2.05,
|
| 2106 |
+
"learning_rate": 7.844509274827907e-05,
|
| 2107 |
+
"loss": 0.6216,
|
| 2108 |
+
"step": 342000
|
| 2109 |
+
},
|
| 2110 |
+
{
|
| 2111 |
+
"epoch": 2.05,
|
| 2112 |
+
"learning_rate": 7.830895520619128e-05,
|
| 2113 |
+
"loss": 0.616,
|
| 2114 |
+
"step": 343000
|
| 2115 |
+
},
|
| 2116 |
+
{
|
| 2117 |
+
"epoch": 2.06,
|
| 2118 |
+
"learning_rate": 7.817250808190483e-05,
|
| 2119 |
+
"loss": 0.6017,
|
| 2120 |
+
"step": 344000
|
| 2121 |
+
},
|
| 2122 |
+
{
|
| 2123 |
+
"epoch": 2.06,
|
| 2124 |
+
"learning_rate": 7.803575286758364e-05,
|
| 2125 |
+
"loss": 0.5984,
|
| 2126 |
+
"step": 345000
|
| 2127 |
+
},
|
| 2128 |
+
{
|
| 2129 |
+
"epoch": 2.06,
|
| 2130 |
+
"learning_rate": 7.789869105876083e-05,
|
| 2131 |
+
"loss": 0.5992,
|
| 2132 |
+
"step": 346000
|
| 2133 |
+
},
|
| 2134 |
+
{
|
| 2135 |
+
"epoch": 2.06,
|
| 2136 |
+
"learning_rate": 7.776132415432234e-05,
|
| 2137 |
+
"loss": 0.5962,
|
| 2138 |
+
"step": 347000
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 2.06,
|
| 2142 |
+
"learning_rate": 7.762365365649067e-05,
|
| 2143 |
+
"loss": 0.6069,
|
| 2144 |
+
"step": 348000
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 2.06,
|
| 2148 |
+
"learning_rate": 7.748568107080832e-05,
|
| 2149 |
+
"loss": 0.6131,
|
| 2150 |
+
"step": 349000
|
| 2151 |
+
},
|
| 2152 |
+
{
|
| 2153 |
+
"epoch": 2.06,
|
| 2154 |
+
"learning_rate": 7.734740790612136e-05,
|
| 2155 |
+
"loss": 0.6099,
|
| 2156 |
+
"step": 350000
|
| 2157 |
+
},
|
| 2158 |
+
{
|
| 2159 |
+
"epoch": 2.06,
|
| 2160 |
+
"eval_loss": 0.28194719552993774,
|
| 2161 |
+
"eval_runtime": 25.6338,
|
| 2162 |
+
"eval_samples_per_second": 195.055,
|
| 2163 |
+
"eval_steps_per_second": 3.082,
|
| 2164 |
+
"step": 350000
|
| 2165 |
+
},
|
| 2166 |
+
{
|
| 2167 |
+
"epoch": 2.06,
|
| 2168 |
+
"learning_rate": 7.720883567456298e-05,
|
| 2169 |
+
"loss": 0.6113,
|
| 2170 |
+
"step": 351000
|
| 2171 |
+
},
|
| 2172 |
+
{
|
| 2173 |
+
"epoch": 2.06,
|
| 2174 |
+
"learning_rate": 7.70699658915369e-05,
|
| 2175 |
+
"loss": 0.6161,
|
| 2176 |
+
"step": 352000
|
| 2177 |
+
},
|
| 2178 |
+
{
|
| 2179 |
+
"epoch": 2.06,
|
| 2180 |
+
"learning_rate": 7.693080007570084e-05,
|
| 2181 |
+
"loss": 0.6113,
|
| 2182 |
+
"step": 353000
|
| 2183 |
+
},
|
| 2184 |
+
{
|
| 2185 |
+
"epoch": 2.07,
|
| 2186 |
+
"learning_rate": 7.679133974894983e-05,
|
| 2187 |
+
"loss": 0.5998,
|
| 2188 |
+
"step": 354000
|
| 2189 |
+
},
|
| 2190 |
+
{
|
| 2191 |
+
"epoch": 2.07,
|
| 2192 |
+
"learning_rate": 7.66515864363997e-05,
|
| 2193 |
+
"loss": 0.5955,
|
| 2194 |
+
"step": 355000
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 2.07,
|
| 2198 |
+
"learning_rate": 7.651154166637025e-05,
|
| 2199 |
+
"loss": 0.5976,
|
| 2200 |
+
"step": 356000
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 2.07,
|
| 2204 |
+
"learning_rate": 7.637120697036866e-05,
|
| 2205 |
+
"loss": 0.5941,
|
| 2206 |
+
"step": 357000
|
| 2207 |
+
},
|
| 2208 |
+
{
|
| 2209 |
+
"epoch": 2.07,
|
| 2210 |
+
"learning_rate": 7.623058388307269e-05,
|
| 2211 |
+
"loss": 0.6049,
|
| 2212 |
+
"step": 358000
|
| 2213 |
+
},
|
| 2214 |
+
{
|
| 2215 |
+
"epoch": 2.07,
|
| 2216 |
+
"learning_rate": 7.608967394231387e-05,
|
| 2217 |
+
"loss": 0.6096,
|
| 2218 |
+
"step": 359000
|
| 2219 |
+
},
|
| 2220 |
+
{
|
| 2221 |
+
"epoch": 2.07,
|
| 2222 |
+
"learning_rate": 7.594847868906076e-05,
|
| 2223 |
+
"loss": 0.6059,
|
| 2224 |
+
"step": 360000
|
| 2225 |
+
},
|
| 2226 |
+
{
|
| 2227 |
+
"epoch": 2.07,
|
| 2228 |
+
"learning_rate": 7.580699966740201e-05,
|
| 2229 |
+
"loss": 0.6094,
|
| 2230 |
+
"step": 361000
|
| 2231 |
+
},
|
| 2232 |
+
{
|
| 2233 |
+
"epoch": 2.07,
|
| 2234 |
+
"learning_rate": 7.566523842452958e-05,
|
| 2235 |
+
"loss": 0.6142,
|
| 2236 |
+
"step": 362000
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 2.07,
|
| 2240 |
+
"learning_rate": 7.552319651072164e-05,
|
| 2241 |
+
"loss": 0.6084,
|
| 2242 |
+
"step": 363000
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 2.08,
|
| 2246 |
+
"learning_rate": 7.538087547932585e-05,
|
| 2247 |
+
"loss": 0.5961,
|
| 2248 |
+
"step": 364000
|
| 2249 |
+
},
|
| 2250 |
+
{
|
| 2251 |
+
"epoch": 2.08,
|
| 2252 |
+
"learning_rate": 7.52382768867422e-05,
|
| 2253 |
+
"loss": 0.595,
|
| 2254 |
+
"step": 365000
|
| 2255 |
+
},
|
| 2256 |
+
{
|
| 2257 |
+
"epoch": 2.08,
|
| 2258 |
+
"learning_rate": 7.509540229240601e-05,
|
| 2259 |
+
"loss": 0.5939,
|
| 2260 |
+
"step": 366000
|
| 2261 |
+
},
|
| 2262 |
+
{
|
| 2263 |
+
"epoch": 2.08,
|
| 2264 |
+
"learning_rate": 7.495225325877103e-05,
|
| 2265 |
+
"loss": 0.5945,
|
| 2266 |
+
"step": 367000
|
| 2267 |
+
},
|
| 2268 |
+
{
|
| 2269 |
+
"epoch": 2.08,
|
| 2270 |
+
"learning_rate": 7.480883135129211e-05,
|
| 2271 |
+
"loss": 0.6041,
|
| 2272 |
+
"step": 368000
|
| 2273 |
+
},
|
| 2274 |
+
{
|
| 2275 |
+
"epoch": 2.08,
|
| 2276 |
+
"learning_rate": 7.466513813840825e-05,
|
| 2277 |
+
"loss": 0.6074,
|
| 2278 |
+
"step": 369000
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 2.08,
|
| 2282 |
+
"learning_rate": 7.452117519152542e-05,
|
| 2283 |
+
"loss": 0.6038,
|
| 2284 |
+
"step": 370000
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 2.08,
|
| 2288 |
+
"learning_rate": 7.437694408499933e-05,
|
| 2289 |
+
"loss": 0.6068,
|
| 2290 |
+
"step": 371000
|
| 2291 |
+
},
|
| 2292 |
+
{
|
| 2293 |
+
"epoch": 2.08,
|
| 2294 |
+
"learning_rate": 7.423244639611826e-05,
|
| 2295 |
+
"loss": 0.6137,
|
| 2296 |
+
"step": 372000
|
| 2297 |
+
},
|
| 2298 |
+
{
|
| 2299 |
+
"epoch": 2.08,
|
| 2300 |
+
"learning_rate": 7.408768370508576e-05,
|
| 2301 |
+
"loss": 0.6081,
|
| 2302 |
+
"step": 373000
|
| 2303 |
+
},
|
| 2304 |
+
{
|
| 2305 |
+
"epoch": 2.09,
|
| 2306 |
+
"learning_rate": 7.394265759500348e-05,
|
| 2307 |
+
"loss": 0.5918,
|
| 2308 |
+
"step": 374000
|
| 2309 |
+
},
|
| 2310 |
+
{
|
| 2311 |
+
"epoch": 2.09,
|
| 2312 |
+
"learning_rate": 7.379736965185368e-05,
|
| 2313 |
+
"loss": 0.5927,
|
| 2314 |
+
"step": 375000
|
| 2315 |
+
},
|
| 2316 |
+
{
|
| 2317 |
+
"epoch": 2.09,
|
| 2318 |
+
"learning_rate": 7.365182146448205e-05,
|
| 2319 |
+
"loss": 0.5905,
|
| 2320 |
+
"step": 376000
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 2.09,
|
| 2324 |
+
"learning_rate": 7.350601462458024e-05,
|
| 2325 |
+
"loss": 0.5909,
|
| 2326 |
+
"step": 377000
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 2.09,
|
| 2330 |
+
"learning_rate": 7.335995072666848e-05,
|
| 2331 |
+
"loss": 0.5998,
|
| 2332 |
+
"step": 378000
|
| 2333 |
+
},
|
| 2334 |
+
{
|
| 2335 |
+
"epoch": 2.09,
|
| 2336 |
+
"learning_rate": 7.32136313680782e-05,
|
| 2337 |
+
"loss": 0.6044,
|
| 2338 |
+
"step": 379000
|
| 2339 |
+
},
|
| 2340 |
+
{
|
| 2341 |
+
"epoch": 2.09,
|
| 2342 |
+
"learning_rate": 7.30670581489344e-05,
|
| 2343 |
+
"loss": 0.6039,
|
| 2344 |
+
"step": 380000
|
| 2345 |
+
},
|
| 2346 |
+
{
|
| 2347 |
+
"epoch": 2.09,
|
| 2348 |
+
"learning_rate": 7.292023267213835e-05,
|
| 2349 |
+
"loss": 0.6009,
|
| 2350 |
+
"step": 381000
|
| 2351 |
+
},
|
| 2352 |
+
{
|
| 2353 |
+
"epoch": 2.09,
|
| 2354 |
+
"learning_rate": 7.277315654334997e-05,
|
| 2355 |
+
"loss": 0.6093,
|
| 2356 |
+
"step": 382000
|
| 2357 |
+
},
|
| 2358 |
+
{
|
| 2359 |
+
"epoch": 2.09,
|
| 2360 |
+
"learning_rate": 7.262583137097018e-05,
|
| 2361 |
+
"loss": 0.6027,
|
| 2362 |
+
"step": 383000
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 2.1,
|
| 2366 |
+
"learning_rate": 7.247825876612353e-05,
|
| 2367 |
+
"loss": 0.5911,
|
| 2368 |
+
"step": 384000
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 2.1,
|
| 2372 |
+
"learning_rate": 7.233044034264034e-05,
|
| 2373 |
+
"loss": 0.5913,
|
| 2374 |
+
"step": 385000
|
| 2375 |
+
},
|
| 2376 |
+
{
|
| 2377 |
+
"epoch": 2.1,
|
| 2378 |
+
"learning_rate": 7.218237771703921e-05,
|
| 2379 |
+
"loss": 0.5853,
|
| 2380 |
+
"step": 386000
|
| 2381 |
+
},
|
| 2382 |
+
{
|
| 2383 |
+
"epoch": 2.1,
|
| 2384 |
+
"learning_rate": 7.203407250850928e-05,
|
| 2385 |
+
"loss": 0.5875,
|
| 2386 |
+
"step": 387000
|
| 2387 |
+
},
|
| 2388 |
+
{
|
| 2389 |
+
"epoch": 2.1,
|
| 2390 |
+
"learning_rate": 7.188552633889259e-05,
|
| 2391 |
+
"loss": 0.5973,
|
| 2392 |
+
"step": 388000
|
| 2393 |
+
},
|
| 2394 |
+
{
|
| 2395 |
+
"epoch": 2.1,
|
| 2396 |
+
"learning_rate": 7.173674083266624e-05,
|
| 2397 |
+
"loss": 0.5992,
|
| 2398 |
+
"step": 389000
|
| 2399 |
+
},
|
| 2400 |
+
{
|
| 2401 |
+
"epoch": 2.1,
|
| 2402 |
+
"learning_rate": 7.158771761692464e-05,
|
| 2403 |
+
"loss": 0.5965,
|
| 2404 |
+
"step": 390000
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 2.1,
|
| 2408 |
+
"learning_rate": 7.143845832136188e-05,
|
| 2409 |
+
"loss": 0.5996,
|
| 2410 |
+
"step": 391000
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 2.1,
|
| 2414 |
+
"learning_rate": 7.128896457825364e-05,
|
| 2415 |
+
"loss": 0.6079,
|
| 2416 |
+
"step": 392000
|
| 2417 |
+
},
|
| 2418 |
+
{
|
| 2419 |
+
"epoch": 2.1,
|
| 2420 |
+
"learning_rate": 7.113923802243957e-05,
|
| 2421 |
+
"loss": 0.5973,
|
| 2422 |
+
"step": 393000
|
| 2423 |
+
},
|
| 2424 |
+
{
|
| 2425 |
+
"epoch": 2.11,
|
| 2426 |
+
"learning_rate": 7.09892802913053e-05,
|
| 2427 |
+
"loss": 0.5879,
|
| 2428 |
+
"step": 394000
|
| 2429 |
+
},
|
| 2430 |
+
{
|
| 2431 |
+
"epoch": 2.11,
|
| 2432 |
+
"learning_rate": 7.083909302476453e-05,
|
| 2433 |
+
"loss": 0.5844,
|
| 2434 |
+
"step": 395000
|
| 2435 |
+
},
|
| 2436 |
+
{
|
| 2437 |
+
"epoch": 2.11,
|
| 2438 |
+
"learning_rate": 7.068867786524116e-05,
|
| 2439 |
+
"loss": 0.5838,
|
| 2440 |
+
"step": 396000
|
| 2441 |
+
},
|
| 2442 |
+
{
|
| 2443 |
+
"epoch": 2.11,
|
| 2444 |
+
"learning_rate": 7.053803645765128e-05,
|
| 2445 |
+
"loss": 0.5838,
|
| 2446 |
+
"step": 397000
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 2.11,
|
| 2450 |
+
"learning_rate": 7.038717044938519e-05,
|
| 2451 |
+
"loss": 0.5941,
|
| 2452 |
+
"step": 398000
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 2.11,
|
| 2456 |
+
"learning_rate": 7.023608149028937e-05,
|
| 2457 |
+
"loss": 0.5989,
|
| 2458 |
+
"step": 399000
|
| 2459 |
+
},
|
| 2460 |
+
{
|
| 2461 |
+
"epoch": 2.11,
|
| 2462 |
+
"learning_rate": 7.008477123264848e-05,
|
| 2463 |
+
"loss": 0.5956,
|
| 2464 |
+
"step": 400000
|
| 2465 |
+
},
|
| 2466 |
+
{
|
| 2467 |
+
"epoch": 2.11,
|
| 2468 |
+
"eval_loss": 0.2689692974090576,
|
| 2469 |
+
"eval_runtime": 25.7546,
|
| 2470 |
+
"eval_samples_per_second": 194.14,
|
| 2471 |
+
"eval_steps_per_second": 3.067,
|
| 2472 |
+
"step": 400000
|
| 2473 |
+
},
|
| 2474 |
+
{
|
| 2475 |
+
"epoch": 2.11,
|
| 2476 |
+
"learning_rate": 6.993324133116726e-05,
|
| 2477 |
+
"loss": 0.5963,
|
| 2478 |
+
"step": 401000
|
| 2479 |
+
},
|
| 2480 |
+
{
|
| 2481 |
+
"epoch": 2.11,
|
| 2482 |
+
"learning_rate": 6.978149344295242e-05,
|
| 2483 |
+
"loss": 0.6046,
|
| 2484 |
+
"step": 402000
|
| 2485 |
+
},
|
| 2486 |
+
{
|
| 2487 |
+
"epoch": 2.11,
|
| 2488 |
+
"learning_rate": 6.962952922749457e-05,
|
| 2489 |
+
"loss": 0.5946,
|
| 2490 |
+
"step": 403000
|
| 2491 |
+
},
|
| 2492 |
+
{
|
| 2493 |
+
"epoch": 2.12,
|
| 2494 |
+
"learning_rate": 6.947735034665002e-05,
|
| 2495 |
+
"loss": 0.5836,
|
| 2496 |
+
"step": 404000
|
| 2497 |
+
},
|
| 2498 |
+
{
|
| 2499 |
+
"epoch": 2.12,
|
| 2500 |
+
"learning_rate": 6.932495846462261e-05,
|
| 2501 |
+
"loss": 0.58,
|
| 2502 |
+
"step": 405000
|
| 2503 |
+
},
|
| 2504 |
+
{
|
| 2505 |
+
"epoch": 2.12,
|
| 2506 |
+
"learning_rate": 6.917235524794558e-05,
|
| 2507 |
+
"loss": 0.5782,
|
| 2508 |
+
"step": 406000
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 2.12,
|
| 2512 |
+
"learning_rate": 6.901954236546323e-05,
|
| 2513 |
+
"loss": 0.582,
|
| 2514 |
+
"step": 407000
|
| 2515 |
+
},
|
| 2516 |
+
{
|
| 2517 |
+
"epoch": 2.12,
|
| 2518 |
+
"learning_rate": 6.886652148831279e-05,
|
| 2519 |
+
"loss": 0.5916,
|
| 2520 |
+
"step": 408000
|
| 2521 |
+
},
|
| 2522 |
+
{
|
| 2523 |
+
"epoch": 2.12,
|
| 2524 |
+
"learning_rate": 6.871329428990602e-05,
|
| 2525 |
+
"loss": 0.5928,
|
| 2526 |
+
"step": 409000
|
| 2527 |
+
},
|
| 2528 |
+
{
|
| 2529 |
+
"epoch": 2.12,
|
| 2530 |
+
"learning_rate": 6.855986244591104e-05,
|
| 2531 |
+
"loss": 0.5901,
|
| 2532 |
+
"step": 410000
|
| 2533 |
+
},
|
| 2534 |
+
{
|
| 2535 |
+
"epoch": 2.12,
|
| 2536 |
+
"learning_rate": 6.840622763423391e-05,
|
| 2537 |
+
"loss": 0.5912,
|
| 2538 |
+
"step": 411000
|
| 2539 |
+
},
|
| 2540 |
+
{
|
| 2541 |
+
"epoch": 2.12,
|
| 2542 |
+
"learning_rate": 6.825239153500029e-05,
|
| 2543 |
+
"loss": 0.6018,
|
| 2544 |
+
"step": 412000
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 2.12,
|
| 2548 |
+
"learning_rate": 6.809835583053715e-05,
|
| 2549 |
+
"loss": 0.5887,
|
| 2550 |
+
"step": 413000
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 2.13,
|
| 2554 |
+
"learning_rate": 6.794412220535426e-05,
|
| 2555 |
+
"loss": 0.5816,
|
| 2556 |
+
"step": 414000
|
| 2557 |
+
},
|
| 2558 |
+
{
|
| 2559 |
+
"epoch": 2.13,
|
| 2560 |
+
"learning_rate": 6.778969234612584e-05,
|
| 2561 |
+
"loss": 0.5777,
|
| 2562 |
+
"step": 415000
|
| 2563 |
+
},
|
| 2564 |
+
{
|
| 2565 |
+
"epoch": 2.13,
|
| 2566 |
+
"learning_rate": 6.763506794167208e-05,
|
| 2567 |
+
"loss": 0.5774,
|
| 2568 |
+
"step": 416000
|
| 2569 |
+
},
|
| 2570 |
+
{
|
| 2571 |
+
"epoch": 2.13,
|
| 2572 |
+
"learning_rate": 6.748025068294067e-05,
|
| 2573 |
+
"loss": 0.5778,
|
| 2574 |
+
"step": 417000
|
| 2575 |
+
},
|
| 2576 |
+
{
|
| 2577 |
+
"epoch": 2.13,
|
| 2578 |
+
"learning_rate": 6.732524226298841e-05,
|
| 2579 |
+
"loss": 0.5885,
|
| 2580 |
+
"step": 418000
|
| 2581 |
+
},
|
| 2582 |
+
{
|
| 2583 |
+
"epoch": 2.13,
|
| 2584 |
+
"learning_rate": 6.71700443769625e-05,
|
| 2585 |
+
"loss": 0.5902,
|
| 2586 |
+
"step": 419000
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 2.13,
|
| 2590 |
+
"learning_rate": 6.701465872208216e-05,
|
| 2591 |
+
"loss": 0.5851,
|
| 2592 |
+
"step": 420000
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 2.13,
|
| 2596 |
+
"learning_rate": 6.685908699762002e-05,
|
| 2597 |
+
"loss": 0.5884,
|
| 2598 |
+
"step": 421000
|
| 2599 |
+
},
|
| 2600 |
+
{
|
| 2601 |
+
"epoch": 2.13,
|
| 2602 |
+
"learning_rate": 6.670333090488356e-05,
|
| 2603 |
+
"loss": 0.5973,
|
| 2604 |
+
"step": 422000
|
| 2605 |
+
},
|
| 2606 |
+
{
|
| 2607 |
+
"epoch": 2.13,
|
| 2608 |
+
"learning_rate": 6.654739214719641e-05,
|
| 2609 |
+
"loss": 0.5865,
|
| 2610 |
+
"step": 423000
|
| 2611 |
+
},
|
| 2612 |
+
{
|
| 2613 |
+
"epoch": 2.14,
|
| 2614 |
+
"learning_rate": 6.639127242987988e-05,
|
| 2615 |
+
"loss": 0.578,
|
| 2616 |
+
"step": 424000
|
| 2617 |
+
},
|
| 2618 |
+
{
|
| 2619 |
+
"epoch": 2.14,
|
| 2620 |
+
"learning_rate": 6.623497346023418e-05,
|
| 2621 |
+
"loss": 0.5757,
|
| 2622 |
+
"step": 425000
|
| 2623 |
+
},
|
| 2624 |
+
{
|
| 2625 |
+
"epoch": 2.14,
|
| 2626 |
+
"learning_rate": 6.607849694751977e-05,
|
| 2627 |
+
"loss": 0.5741,
|
| 2628 |
+
"step": 426000
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 2.14,
|
| 2632 |
+
"learning_rate": 6.592184460293877e-05,
|
| 2633 |
+
"loss": 0.5755,
|
| 2634 |
+
"step": 427000
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 2.14,
|
| 2638 |
+
"learning_rate": 6.576501813961609e-05,
|
| 2639 |
+
"loss": 0.5847,
|
| 2640 |
+
"step": 428000
|
| 2641 |
+
},
|
| 2642 |
+
{
|
| 2643 |
+
"epoch": 2.14,
|
| 2644 |
+
"learning_rate": 6.56080192725808e-05,
|
| 2645 |
+
"loss": 0.5872,
|
| 2646 |
+
"step": 429000
|
| 2647 |
+
},
|
| 2648 |
+
{
|
| 2649 |
+
"epoch": 2.14,
|
| 2650 |
+
"learning_rate": 6.545084971874738e-05,
|
| 2651 |
+
"loss": 0.5827,
|
| 2652 |
+
"step": 430000
|
| 2653 |
+
},
|
| 2654 |
+
{
|
| 2655 |
+
"epoch": 2.14,
|
| 2656 |
+
"learning_rate": 6.529351119689688e-05,
|
| 2657 |
+
"loss": 0.587,
|
| 2658 |
+
"step": 431000
|
| 2659 |
+
},
|
| 2660 |
+
{
|
| 2661 |
+
"epoch": 2.14,
|
| 2662 |
+
"learning_rate": 6.513600542765817e-05,
|
| 2663 |
+
"loss": 0.5948,
|
| 2664 |
+
"step": 432000
|
| 2665 |
+
},
|
| 2666 |
+
{
|
| 2667 |
+
"epoch": 3.0,
|
| 2668 |
+
"learning_rate": 6.497833413348909e-05,
|
| 2669 |
+
"loss": 0.5557,
|
| 2670 |
+
"step": 433000
|
| 2671 |
+
},
|
| 2672 |
+
{
|
| 2673 |
+
"epoch": 3.0,
|
| 2674 |
+
"learning_rate": 6.48204990386577e-05,
|
| 2675 |
+
"loss": 0.5954,
|
| 2676 |
+
"step": 434000
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 3.0,
|
| 2680 |
+
"learning_rate": 6.466250186922325e-05,
|
| 2681 |
+
"loss": 0.577,
|
| 2682 |
+
"step": 435000
|
| 2683 |
+
},
|
| 2684 |
+
{
|
| 2685 |
+
"epoch": 3.0,
|
| 2686 |
+
"learning_rate": 6.450434435301751e-05,
|
| 2687 |
+
"loss": 0.5707,
|
| 2688 |
+
"step": 436000
|
| 2689 |
+
},
|
| 2690 |
+
{
|
| 2691 |
+
"epoch": 3.0,
|
| 2692 |
+
"learning_rate": 6.43460282196257e-05,
|
| 2693 |
+
"loss": 0.5732,
|
| 2694 |
+
"step": 437000
|
| 2695 |
+
},
|
| 2696 |
+
{
|
| 2697 |
+
"epoch": 3.01,
|
| 2698 |
+
"learning_rate": 6.418755520036775e-05,
|
| 2699 |
+
"loss": 0.5677,
|
| 2700 |
+
"step": 438000
|
| 2701 |
+
},
|
| 2702 |
+
{
|
| 2703 |
+
"epoch": 3.01,
|
| 2704 |
+
"learning_rate": 6.402892702827916e-05,
|
| 2705 |
+
"loss": 0.5745,
|
| 2706 |
+
"step": 439000
|
| 2707 |
+
},
|
| 2708 |
+
{
|
| 2709 |
+
"epoch": 3.01,
|
| 2710 |
+
"learning_rate": 6.387014543809223e-05,
|
| 2711 |
+
"loss": 0.5822,
|
| 2712 |
+
"step": 440000
|
| 2713 |
+
},
|
| 2714 |
+
{
|
| 2715 |
+
"epoch": 3.01,
|
| 2716 |
+
"learning_rate": 6.371121216621698e-05,
|
| 2717 |
+
"loss": 0.582,
|
| 2718 |
+
"step": 441000
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 3.01,
|
| 2722 |
+
"learning_rate": 6.355212895072223e-05,
|
| 2723 |
+
"loss": 0.5809,
|
| 2724 |
+
"step": 442000
|
| 2725 |
+
},
|
| 2726 |
+
{
|
| 2727 |
+
"epoch": 3.01,
|
| 2728 |
+
"learning_rate": 6.339289753131649e-05,
|
| 2729 |
+
"loss": 0.5832,
|
| 2730 |
+
"step": 443000
|
| 2731 |
+
},
|
| 2732 |
+
{
|
| 2733 |
+
"epoch": 3.01,
|
| 2734 |
+
"learning_rate": 6.323351964932908e-05,
|
| 2735 |
+
"loss": 0.5928,
|
| 2736 |
+
"step": 444000
|
| 2737 |
+
},
|
| 2738 |
+
{
|
| 2739 |
+
"epoch": 3.01,
|
| 2740 |
+
"learning_rate": 6.307399704769099e-05,
|
| 2741 |
+
"loss": 0.5725,
|
| 2742 |
+
"step": 445000
|
| 2743 |
+
},
|
| 2744 |
+
{
|
| 2745 |
+
"epoch": 3.01,
|
| 2746 |
+
"learning_rate": 6.291433147091583e-05,
|
| 2747 |
+
"loss": 0.5717,
|
| 2748 |
+
"step": 446000
|
| 2749 |
+
},
|
| 2750 |
+
{
|
| 2751 |
+
"epoch": 3.01,
|
| 2752 |
+
"learning_rate": 6.275452466508077e-05,
|
| 2753 |
+
"loss": 0.5675,
|
| 2754 |
+
"step": 447000
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 3.02,
|
| 2758 |
+
"learning_rate": 6.259457837780742e-05,
|
| 2759 |
+
"loss": 0.5653,
|
| 2760 |
+
"step": 448000
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 3.02,
|
| 2764 |
+
"learning_rate": 6.243449435824276e-05,
|
| 2765 |
+
"loss": 0.5689,
|
| 2766 |
+
"step": 449000
|
| 2767 |
+
},
|
| 2768 |
+
{
|
| 2769 |
+
"epoch": 3.02,
|
| 2770 |
+
"learning_rate": 6.227427435703997e-05,
|
| 2771 |
+
"loss": 0.5773,
|
| 2772 |
+
"step": 450000
|
| 2773 |
+
},
|
| 2774 |
+
{
|
| 2775 |
+
"epoch": 3.02,
|
| 2776 |
+
"eval_loss": 0.26493486762046814,
|
| 2777 |
+
"eval_runtime": 35.3283,
|
| 2778 |
+
"eval_samples_per_second": 141.529,
|
| 2779 |
+
"eval_steps_per_second": 2.236,
|
| 2780 |
+
"step": 450000
|
| 2781 |
}
|
| 2782 |
],
|
| 2783 |
"max_steps": 1000000,
|
| 2784 |
"num_train_epochs": 9223372036854775807,
|
| 2785 |
+
"total_flos": 2.684246915457378e+19,
|
| 2786 |
"trial_name": null,
|
| 2787 |
"trial_params": null
|
| 2788 |
}
|
last-checkpoint/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3503
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:01dd3096f314faf9269c4ac68dfb98c88092420e70aa87490e26292446fd05d9
|
| 3 |
size 3503
|
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1346893675
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b2688ccc6d66b88b6fcc4941eb91cb8f99c587c38ad79a52f32dec77cce83f0c
|
| 3 |
size 1346893675
|
runs/Mar31_08-12-57_t1v-n-7a44a9fa-w-0/1680250902.9870226/events.out.tfevents.1680250902.t1v-n-7a44a9fa-w-0.3132829.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:132f7afd5b11e16c74c6c85a6a96eef67873ddede45d2ccb62fb8c3502983956
|
| 3 |
+
size 5499
|
runs/Mar31_08-12-57_t1v-n-7a44a9fa-w-0/events.out.tfevents.1680250902.t1v-n-7a44a9fa-w-0.3132829.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:df0f0753373da766cbb012f7cb4ea80c8482e5a55180c6a9e513cd8b785cd1ad
|
| 3 |
+
size 12112
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3503
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:01dd3096f314faf9269c4ac68dfb98c88092420e70aa87490e26292446fd05d9
|
| 3 |
size 3503
|