Training in progress, step 5000, checkpoint
Browse files- last-checkpoint/model.safetensors +1 -1
- last-checkpoint/optimizer.pt +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 +1777 -3
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 737632172
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5b16b8aa8e59ed005cab03437e858fc32c15da21710862fb7624c9d60fa09994
|
| 3 |
size 737632172
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1475354682
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7e9f520d78e8c2ec7d22e229de31961c7dc7141280c426b76ea47d2304f6e61e
|
| 3 |
size 1475354682
|
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 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:69ec6e3926fa071bede113523efa3dc6e630c3c7958c54a9ca321cf4d62ed145
|
| 3 |
size 15984
|
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 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f6127ee4f0c13500ec5038fce65af8f7beec63c137c7d4b7c157aa6303cf5879
|
| 3 |
size 15984
|
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 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:da01d1c5eb2cc3a323f97c1f590d13ccfac2a4c5b1479bd378b4e643304f5a4f
|
| 3 |
size 15984
|
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 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:49a3f04d76c0d3acc7d3dd95a04215f368f35a451ae8cba8a2fdba38cda9ca0a
|
| 3 |
size 15984
|
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 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:df7d2c9825dba80cb544920f8cc0c72122f96514e6cd259052a8765b034393e2
|
| 3 |
size 15984
|
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 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6a20a42d44ff48cc162224010190e898fe28598ddad8cd1896d330a3bb1d8ec3
|
| 3 |
size 15984
|
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 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:18ac0dc4f09f25179860561fcea7c5c8f997aabdc46a170665f9dc5a72bc27c6
|
| 3 |
size 15984
|
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 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6a16fcb5411ff961b47eff7378d85105fe9837e0492d19ea5ce3b7c4b77aa3b6
|
| 3 |
size 15984
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1000
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8b5ca0374fb823ceea8ffdfbdcc41c0cd34efad2563ba3211ec3385b1ecc3188
|
| 3 |
size 1000
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 1000,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -1773,6 +1773,1780 @@
|
|
| 1773 |
"learning_rate": 9.995666009518137e-06,
|
| 1774 |
"loss": 13.7492,
|
| 1775 |
"step": 2500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1776 |
}
|
| 1777 |
],
|
| 1778 |
"logging_steps": 10,
|
|
@@ -1792,7 +3566,7 @@
|
|
| 1792 |
"attributes": {}
|
| 1793 |
}
|
| 1794 |
},
|
| 1795 |
-
"total_flos":
|
| 1796 |
"train_batch_size": 4,
|
| 1797 |
"trial_name": null,
|
| 1798 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.22190046654573092,
|
| 5 |
"eval_steps": 1000,
|
| 6 |
+
"global_step": 5000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 1773 |
"learning_rate": 9.995666009518137e-06,
|
| 1774 |
"loss": 13.7492,
|
| 1775 |
"step": 2500
|
| 1776 |
+
},
|
| 1777 |
+
{
|
| 1778 |
+
"epoch": 0.11139403420595692,
|
| 1779 |
+
"grad_norm": 142.9310760498047,
|
| 1780 |
+
"learning_rate": 9.99564867355621e-06,
|
| 1781 |
+
"loss": 13.2862,
|
| 1782 |
+
"step": 2510
|
| 1783 |
+
},
|
| 1784 |
+
{
|
| 1785 |
+
"epoch": 0.11183783513904838,
|
| 1786 |
+
"grad_norm": 136.95596313476562,
|
| 1787 |
+
"learning_rate": 9.995631337594283e-06,
|
| 1788 |
+
"loss": 13.2403,
|
| 1789 |
+
"step": 2520
|
| 1790 |
+
},
|
| 1791 |
+
{
|
| 1792 |
+
"epoch": 0.11228163607213984,
|
| 1793 |
+
"grad_norm": 157.4971160888672,
|
| 1794 |
+
"learning_rate": 9.995614001632356e-06,
|
| 1795 |
+
"loss": 12.9846,
|
| 1796 |
+
"step": 2530
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 0.1127254370052313,
|
| 1800 |
+
"grad_norm": 157.4886474609375,
|
| 1801 |
+
"learning_rate": 9.995596665670427e-06,
|
| 1802 |
+
"loss": 13.6239,
|
| 1803 |
+
"step": 2540
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 0.11316923793832276,
|
| 1807 |
+
"grad_norm": 179.8428192138672,
|
| 1808 |
+
"learning_rate": 9.9955793297085e-06,
|
| 1809 |
+
"loss": 13.562,
|
| 1810 |
+
"step": 2550
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 0.11361303887141423,
|
| 1814 |
+
"grad_norm": 115.2100601196289,
|
| 1815 |
+
"learning_rate": 9.995561993746573e-06,
|
| 1816 |
+
"loss": 13.7602,
|
| 1817 |
+
"step": 2560
|
| 1818 |
+
},
|
| 1819 |
+
{
|
| 1820 |
+
"epoch": 0.11405683980450569,
|
| 1821 |
+
"grad_norm": 152.49363708496094,
|
| 1822 |
+
"learning_rate": 9.995544657784645e-06,
|
| 1823 |
+
"loss": 12.7727,
|
| 1824 |
+
"step": 2570
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"epoch": 0.11450064073759715,
|
| 1828 |
+
"grad_norm": 111.91357421875,
|
| 1829 |
+
"learning_rate": 9.995527321822718e-06,
|
| 1830 |
+
"loss": 13.2489,
|
| 1831 |
+
"step": 2580
|
| 1832 |
+
},
|
| 1833 |
+
{
|
| 1834 |
+
"epoch": 0.11494444167068861,
|
| 1835 |
+
"grad_norm": 149.418701171875,
|
| 1836 |
+
"learning_rate": 9.99550998586079e-06,
|
| 1837 |
+
"loss": 13.5211,
|
| 1838 |
+
"step": 2590
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 0.11538824260378007,
|
| 1842 |
+
"grad_norm": 158.1101531982422,
|
| 1843 |
+
"learning_rate": 9.995492649898862e-06,
|
| 1844 |
+
"loss": 13.5069,
|
| 1845 |
+
"step": 2600
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 0.11583204353687154,
|
| 1849 |
+
"grad_norm": 152.66778564453125,
|
| 1850 |
+
"learning_rate": 9.995475313936935e-06,
|
| 1851 |
+
"loss": 13.44,
|
| 1852 |
+
"step": 2610
|
| 1853 |
+
},
|
| 1854 |
+
{
|
| 1855 |
+
"epoch": 0.116275844469963,
|
| 1856 |
+
"grad_norm": 133.16883850097656,
|
| 1857 |
+
"learning_rate": 9.995457977975008e-06,
|
| 1858 |
+
"loss": 13.4715,
|
| 1859 |
+
"step": 2620
|
| 1860 |
+
},
|
| 1861 |
+
{
|
| 1862 |
+
"epoch": 0.11671964540305446,
|
| 1863 |
+
"grad_norm": 145.34190368652344,
|
| 1864 |
+
"learning_rate": 9.99544064201308e-06,
|
| 1865 |
+
"loss": 13.2012,
|
| 1866 |
+
"step": 2630
|
| 1867 |
+
},
|
| 1868 |
+
{
|
| 1869 |
+
"epoch": 0.11716344633614592,
|
| 1870 |
+
"grad_norm": 141.71560668945312,
|
| 1871 |
+
"learning_rate": 9.995423306051153e-06,
|
| 1872 |
+
"loss": 13.389,
|
| 1873 |
+
"step": 2640
|
| 1874 |
+
},
|
| 1875 |
+
{
|
| 1876 |
+
"epoch": 0.11760724726923738,
|
| 1877 |
+
"grad_norm": 158.38853454589844,
|
| 1878 |
+
"learning_rate": 9.995405970089226e-06,
|
| 1879 |
+
"loss": 13.6967,
|
| 1880 |
+
"step": 2650
|
| 1881 |
+
},
|
| 1882 |
+
{
|
| 1883 |
+
"epoch": 0.11805104820232884,
|
| 1884 |
+
"grad_norm": 153.3102264404297,
|
| 1885 |
+
"learning_rate": 9.995388634127297e-06,
|
| 1886 |
+
"loss": 12.9532,
|
| 1887 |
+
"step": 2660
|
| 1888 |
+
},
|
| 1889 |
+
{
|
| 1890 |
+
"epoch": 0.11849484913542031,
|
| 1891 |
+
"grad_norm": 168.13534545898438,
|
| 1892 |
+
"learning_rate": 9.99537129816537e-06,
|
| 1893 |
+
"loss": 12.9198,
|
| 1894 |
+
"step": 2670
|
| 1895 |
+
},
|
| 1896 |
+
{
|
| 1897 |
+
"epoch": 0.11893865006851177,
|
| 1898 |
+
"grad_norm": 152.37652587890625,
|
| 1899 |
+
"learning_rate": 9.995353962203443e-06,
|
| 1900 |
+
"loss": 12.762,
|
| 1901 |
+
"step": 2680
|
| 1902 |
+
},
|
| 1903 |
+
{
|
| 1904 |
+
"epoch": 0.11938245100160323,
|
| 1905 |
+
"grad_norm": 175.31565856933594,
|
| 1906 |
+
"learning_rate": 9.995336626241515e-06,
|
| 1907 |
+
"loss": 13.3768,
|
| 1908 |
+
"step": 2690
|
| 1909 |
+
},
|
| 1910 |
+
{
|
| 1911 |
+
"epoch": 0.11982625193469469,
|
| 1912 |
+
"grad_norm": 142.70001220703125,
|
| 1913 |
+
"learning_rate": 9.995319290279588e-06,
|
| 1914 |
+
"loss": 14.0593,
|
| 1915 |
+
"step": 2700
|
| 1916 |
+
},
|
| 1917 |
+
{
|
| 1918 |
+
"epoch": 0.12027005286778615,
|
| 1919 |
+
"grad_norm": 138.56088256835938,
|
| 1920 |
+
"learning_rate": 9.995301954317661e-06,
|
| 1921 |
+
"loss": 12.9747,
|
| 1922 |
+
"step": 2710
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"epoch": 0.12071385380087761,
|
| 1926 |
+
"grad_norm": 159.65451049804688,
|
| 1927 |
+
"learning_rate": 9.995284618355732e-06,
|
| 1928 |
+
"loss": 13.163,
|
| 1929 |
+
"step": 2720
|
| 1930 |
+
},
|
| 1931 |
+
{
|
| 1932 |
+
"epoch": 0.12115765473396908,
|
| 1933 |
+
"grad_norm": 119.24352264404297,
|
| 1934 |
+
"learning_rate": 9.995267282393805e-06,
|
| 1935 |
+
"loss": 12.5898,
|
| 1936 |
+
"step": 2730
|
| 1937 |
+
},
|
| 1938 |
+
{
|
| 1939 |
+
"epoch": 0.12160145566706054,
|
| 1940 |
+
"grad_norm": 107.53023529052734,
|
| 1941 |
+
"learning_rate": 9.995249946431879e-06,
|
| 1942 |
+
"loss": 13.4115,
|
| 1943 |
+
"step": 2740
|
| 1944 |
+
},
|
| 1945 |
+
{
|
| 1946 |
+
"epoch": 0.122045256600152,
|
| 1947 |
+
"grad_norm": 141.00604248046875,
|
| 1948 |
+
"learning_rate": 9.995232610469952e-06,
|
| 1949 |
+
"loss": 13.4099,
|
| 1950 |
+
"step": 2750
|
| 1951 |
+
},
|
| 1952 |
+
{
|
| 1953 |
+
"epoch": 0.12248905753324346,
|
| 1954 |
+
"grad_norm": 121.9310302734375,
|
| 1955 |
+
"learning_rate": 9.995215274508023e-06,
|
| 1956 |
+
"loss": 13.0975,
|
| 1957 |
+
"step": 2760
|
| 1958 |
+
},
|
| 1959 |
+
{
|
| 1960 |
+
"epoch": 0.12293285846633492,
|
| 1961 |
+
"grad_norm": 170.2215576171875,
|
| 1962 |
+
"learning_rate": 9.995197938546096e-06,
|
| 1963 |
+
"loss": 13.1425,
|
| 1964 |
+
"step": 2770
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 0.12337665939942638,
|
| 1968 |
+
"grad_norm": 134.97998046875,
|
| 1969 |
+
"learning_rate": 9.99518060258417e-06,
|
| 1970 |
+
"loss": 13.0698,
|
| 1971 |
+
"step": 2780
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 0.12382046033251785,
|
| 1975 |
+
"grad_norm": 141.1931915283203,
|
| 1976 |
+
"learning_rate": 9.99516326662224e-06,
|
| 1977 |
+
"loss": 12.9536,
|
| 1978 |
+
"step": 2790
|
| 1979 |
+
},
|
| 1980 |
+
{
|
| 1981 |
+
"epoch": 0.12426426126560931,
|
| 1982 |
+
"grad_norm": 123.69408416748047,
|
| 1983 |
+
"learning_rate": 9.995145930660314e-06,
|
| 1984 |
+
"loss": 13.1385,
|
| 1985 |
+
"step": 2800
|
| 1986 |
+
},
|
| 1987 |
+
{
|
| 1988 |
+
"epoch": 0.12470806219870077,
|
| 1989 |
+
"grad_norm": 132.7433319091797,
|
| 1990 |
+
"learning_rate": 9.995128594698387e-06,
|
| 1991 |
+
"loss": 12.7706,
|
| 1992 |
+
"step": 2810
|
| 1993 |
+
},
|
| 1994 |
+
{
|
| 1995 |
+
"epoch": 0.12515186313179225,
|
| 1996 |
+
"grad_norm": 146.10305786132812,
|
| 1997 |
+
"learning_rate": 9.995111258736458e-06,
|
| 1998 |
+
"loss": 13.1492,
|
| 1999 |
+
"step": 2820
|
| 2000 |
+
},
|
| 2001 |
+
{
|
| 2002 |
+
"epoch": 0.1255956640648837,
|
| 2003 |
+
"grad_norm": 217.5339813232422,
|
| 2004 |
+
"learning_rate": 9.995093922774531e-06,
|
| 2005 |
+
"loss": 13.2251,
|
| 2006 |
+
"step": 2830
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 0.12603946499797516,
|
| 2010 |
+
"grad_norm": 178.1737518310547,
|
| 2011 |
+
"learning_rate": 9.995076586812604e-06,
|
| 2012 |
+
"loss": 12.8123,
|
| 2013 |
+
"step": 2840
|
| 2014 |
+
},
|
| 2015 |
+
{
|
| 2016 |
+
"epoch": 0.12648326593106662,
|
| 2017 |
+
"grad_norm": 123.01746368408203,
|
| 2018 |
+
"learning_rate": 9.995059250850676e-06,
|
| 2019 |
+
"loss": 12.8128,
|
| 2020 |
+
"step": 2850
|
| 2021 |
+
},
|
| 2022 |
+
{
|
| 2023 |
+
"epoch": 0.12692706686415808,
|
| 2024 |
+
"grad_norm": 173.1932830810547,
|
| 2025 |
+
"learning_rate": 9.995041914888749e-06,
|
| 2026 |
+
"loss": 13.3028,
|
| 2027 |
+
"step": 2860
|
| 2028 |
+
},
|
| 2029 |
+
{
|
| 2030 |
+
"epoch": 0.12737086779724954,
|
| 2031 |
+
"grad_norm": 129.58663940429688,
|
| 2032 |
+
"learning_rate": 9.995024578926822e-06,
|
| 2033 |
+
"loss": 12.685,
|
| 2034 |
+
"step": 2870
|
| 2035 |
+
},
|
| 2036 |
+
{
|
| 2037 |
+
"epoch": 0.127814668730341,
|
| 2038 |
+
"grad_norm": 163.02256774902344,
|
| 2039 |
+
"learning_rate": 9.995007242964893e-06,
|
| 2040 |
+
"loss": 12.8868,
|
| 2041 |
+
"step": 2880
|
| 2042 |
+
},
|
| 2043 |
+
{
|
| 2044 |
+
"epoch": 0.12825846966343246,
|
| 2045 |
+
"grad_norm": 129.8294677734375,
|
| 2046 |
+
"learning_rate": 9.994989907002966e-06,
|
| 2047 |
+
"loss": 12.6222,
|
| 2048 |
+
"step": 2890
|
| 2049 |
+
},
|
| 2050 |
+
{
|
| 2051 |
+
"epoch": 0.12870227059652392,
|
| 2052 |
+
"grad_norm": 133.46365356445312,
|
| 2053 |
+
"learning_rate": 9.99497257104104e-06,
|
| 2054 |
+
"loss": 12.5605,
|
| 2055 |
+
"step": 2900
|
| 2056 |
+
},
|
| 2057 |
+
{
|
| 2058 |
+
"epoch": 0.12914607152961538,
|
| 2059 |
+
"grad_norm": 141.15603637695312,
|
| 2060 |
+
"learning_rate": 9.99495523507911e-06,
|
| 2061 |
+
"loss": 12.972,
|
| 2062 |
+
"step": 2910
|
| 2063 |
+
},
|
| 2064 |
+
{
|
| 2065 |
+
"epoch": 0.12958987246270684,
|
| 2066 |
+
"grad_norm": 128.41879272460938,
|
| 2067 |
+
"learning_rate": 9.994937899117184e-06,
|
| 2068 |
+
"loss": 13.1774,
|
| 2069 |
+
"step": 2920
|
| 2070 |
+
},
|
| 2071 |
+
{
|
| 2072 |
+
"epoch": 0.13003367339579833,
|
| 2073 |
+
"grad_norm": 136.0733184814453,
|
| 2074 |
+
"learning_rate": 9.994920563155257e-06,
|
| 2075 |
+
"loss": 12.7431,
|
| 2076 |
+
"step": 2930
|
| 2077 |
+
},
|
| 2078 |
+
{
|
| 2079 |
+
"epoch": 0.1304774743288898,
|
| 2080 |
+
"grad_norm": 135.34146118164062,
|
| 2081 |
+
"learning_rate": 9.994903227193328e-06,
|
| 2082 |
+
"loss": 13.4061,
|
| 2083 |
+
"step": 2940
|
| 2084 |
+
},
|
| 2085 |
+
{
|
| 2086 |
+
"epoch": 0.13092127526198125,
|
| 2087 |
+
"grad_norm": 143.00442504882812,
|
| 2088 |
+
"learning_rate": 9.994885891231401e-06,
|
| 2089 |
+
"loss": 12.9928,
|
| 2090 |
+
"step": 2950
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"epoch": 0.1313650761950727,
|
| 2094 |
+
"grad_norm": 150.8383026123047,
|
| 2095 |
+
"learning_rate": 9.994868555269474e-06,
|
| 2096 |
+
"loss": 12.8895,
|
| 2097 |
+
"step": 2960
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 0.13180887712816416,
|
| 2101 |
+
"grad_norm": 144.29466247558594,
|
| 2102 |
+
"learning_rate": 9.994851219307547e-06,
|
| 2103 |
+
"loss": 13.2952,
|
| 2104 |
+
"step": 2970
|
| 2105 |
+
},
|
| 2106 |
+
{
|
| 2107 |
+
"epoch": 0.13225267806125562,
|
| 2108 |
+
"grad_norm": 130.09901428222656,
|
| 2109 |
+
"learning_rate": 9.994833883345619e-06,
|
| 2110 |
+
"loss": 13.1275,
|
| 2111 |
+
"step": 2980
|
| 2112 |
+
},
|
| 2113 |
+
{
|
| 2114 |
+
"epoch": 0.13269647899434708,
|
| 2115 |
+
"grad_norm": 161.23716735839844,
|
| 2116 |
+
"learning_rate": 9.994816547383692e-06,
|
| 2117 |
+
"loss": 12.7923,
|
| 2118 |
+
"step": 2990
|
| 2119 |
+
},
|
| 2120 |
+
{
|
| 2121 |
+
"epoch": 0.13314027992743854,
|
| 2122 |
+
"grad_norm": 126.25406646728516,
|
| 2123 |
+
"learning_rate": 9.994799211421765e-06,
|
| 2124 |
+
"loss": 12.9331,
|
| 2125 |
+
"step": 3000
|
| 2126 |
+
},
|
| 2127 |
+
{
|
| 2128 |
+
"epoch": 0.13314027992743854,
|
| 2129 |
+
"eval_loss": 0.40445423126220703,
|
| 2130 |
+
"eval_runtime": 673.2876,
|
| 2131 |
+
"eval_samples_per_second": 1803.673,
|
| 2132 |
+
"eval_steps_per_second": 56.365,
|
| 2133 |
+
"step": 3000
|
| 2134 |
+
},
|
| 2135 |
+
{
|
| 2136 |
+
"epoch": 0.13358408086053,
|
| 2137 |
+
"grad_norm": 121.59950256347656,
|
| 2138 |
+
"learning_rate": 9.994781875459836e-06,
|
| 2139 |
+
"loss": 12.1823,
|
| 2140 |
+
"step": 3010
|
| 2141 |
+
},
|
| 2142 |
+
{
|
| 2143 |
+
"epoch": 0.13402788179362146,
|
| 2144 |
+
"grad_norm": 112.05023193359375,
|
| 2145 |
+
"learning_rate": 9.99476453949791e-06,
|
| 2146 |
+
"loss": 12.886,
|
| 2147 |
+
"step": 3020
|
| 2148 |
+
},
|
| 2149 |
+
{
|
| 2150 |
+
"epoch": 0.13447168272671292,
|
| 2151 |
+
"grad_norm": 129.76803588867188,
|
| 2152 |
+
"learning_rate": 9.994747203535983e-06,
|
| 2153 |
+
"loss": 12.7748,
|
| 2154 |
+
"step": 3030
|
| 2155 |
+
},
|
| 2156 |
+
{
|
| 2157 |
+
"epoch": 0.1349154836598044,
|
| 2158 |
+
"grad_norm": 124.96532440185547,
|
| 2159 |
+
"learning_rate": 9.994729867574054e-06,
|
| 2160 |
+
"loss": 13.1191,
|
| 2161 |
+
"step": 3040
|
| 2162 |
+
},
|
| 2163 |
+
{
|
| 2164 |
+
"epoch": 0.13535928459289587,
|
| 2165 |
+
"grad_norm": 141.1378631591797,
|
| 2166 |
+
"learning_rate": 9.994712531612127e-06,
|
| 2167 |
+
"loss": 13.4907,
|
| 2168 |
+
"step": 3050
|
| 2169 |
+
},
|
| 2170 |
+
{
|
| 2171 |
+
"epoch": 0.13580308552598733,
|
| 2172 |
+
"grad_norm": 160.61422729492188,
|
| 2173 |
+
"learning_rate": 9.9946951956502e-06,
|
| 2174 |
+
"loss": 13.6199,
|
| 2175 |
+
"step": 3060
|
| 2176 |
+
},
|
| 2177 |
+
{
|
| 2178 |
+
"epoch": 0.1362468864590788,
|
| 2179 |
+
"grad_norm": 136.35914611816406,
|
| 2180 |
+
"learning_rate": 9.994677859688271e-06,
|
| 2181 |
+
"loss": 13.2168,
|
| 2182 |
+
"step": 3070
|
| 2183 |
+
},
|
| 2184 |
+
{
|
| 2185 |
+
"epoch": 0.13669068739217025,
|
| 2186 |
+
"grad_norm": 127.0990982055664,
|
| 2187 |
+
"learning_rate": 9.994660523726345e-06,
|
| 2188 |
+
"loss": 12.9859,
|
| 2189 |
+
"step": 3080
|
| 2190 |
+
},
|
| 2191 |
+
{
|
| 2192 |
+
"epoch": 0.1371344883252617,
|
| 2193 |
+
"grad_norm": 138.41912841796875,
|
| 2194 |
+
"learning_rate": 9.994643187764418e-06,
|
| 2195 |
+
"loss": 12.4187,
|
| 2196 |
+
"step": 3090
|
| 2197 |
+
},
|
| 2198 |
+
{
|
| 2199 |
+
"epoch": 0.13757828925835316,
|
| 2200 |
+
"grad_norm": 140.53573608398438,
|
| 2201 |
+
"learning_rate": 9.994625851802489e-06,
|
| 2202 |
+
"loss": 13.0548,
|
| 2203 |
+
"step": 3100
|
| 2204 |
+
},
|
| 2205 |
+
{
|
| 2206 |
+
"epoch": 0.13802209019144462,
|
| 2207 |
+
"grad_norm": 140.9633331298828,
|
| 2208 |
+
"learning_rate": 9.994608515840562e-06,
|
| 2209 |
+
"loss": 12.8993,
|
| 2210 |
+
"step": 3110
|
| 2211 |
+
},
|
| 2212 |
+
{
|
| 2213 |
+
"epoch": 0.13846589112453608,
|
| 2214 |
+
"grad_norm": 161.15667724609375,
|
| 2215 |
+
"learning_rate": 9.994591179878635e-06,
|
| 2216 |
+
"loss": 12.7675,
|
| 2217 |
+
"step": 3120
|
| 2218 |
+
},
|
| 2219 |
+
{
|
| 2220 |
+
"epoch": 0.13890969205762754,
|
| 2221 |
+
"grad_norm": 131.9075469970703,
|
| 2222 |
+
"learning_rate": 9.994573843916707e-06,
|
| 2223 |
+
"loss": 13.4033,
|
| 2224 |
+
"step": 3130
|
| 2225 |
+
},
|
| 2226 |
+
{
|
| 2227 |
+
"epoch": 0.139353492990719,
|
| 2228 |
+
"grad_norm": 144.76168823242188,
|
| 2229 |
+
"learning_rate": 9.99455650795478e-06,
|
| 2230 |
+
"loss": 12.9296,
|
| 2231 |
+
"step": 3140
|
| 2232 |
+
},
|
| 2233 |
+
{
|
| 2234 |
+
"epoch": 0.13979729392381046,
|
| 2235 |
+
"grad_norm": 128.33094787597656,
|
| 2236 |
+
"learning_rate": 9.994539171992853e-06,
|
| 2237 |
+
"loss": 13.0926,
|
| 2238 |
+
"step": 3150
|
| 2239 |
+
},
|
| 2240 |
+
{
|
| 2241 |
+
"epoch": 0.14024109485690195,
|
| 2242 |
+
"grad_norm": 120.40109252929688,
|
| 2243 |
+
"learning_rate": 9.994521836030924e-06,
|
| 2244 |
+
"loss": 12.6591,
|
| 2245 |
+
"step": 3160
|
| 2246 |
+
},
|
| 2247 |
+
{
|
| 2248 |
+
"epoch": 0.1406848957899934,
|
| 2249 |
+
"grad_norm": 177.44715881347656,
|
| 2250 |
+
"learning_rate": 9.994504500068997e-06,
|
| 2251 |
+
"loss": 13.0056,
|
| 2252 |
+
"step": 3170
|
| 2253 |
+
},
|
| 2254 |
+
{
|
| 2255 |
+
"epoch": 0.14112869672308487,
|
| 2256 |
+
"grad_norm": 133.28228759765625,
|
| 2257 |
+
"learning_rate": 9.99448716410707e-06,
|
| 2258 |
+
"loss": 12.9662,
|
| 2259 |
+
"step": 3180
|
| 2260 |
+
},
|
| 2261 |
+
{
|
| 2262 |
+
"epoch": 0.14157249765617633,
|
| 2263 |
+
"grad_norm": 139.16880798339844,
|
| 2264 |
+
"learning_rate": 9.994469828145143e-06,
|
| 2265 |
+
"loss": 12.55,
|
| 2266 |
+
"step": 3190
|
| 2267 |
+
},
|
| 2268 |
+
{
|
| 2269 |
+
"epoch": 0.1420162985892678,
|
| 2270 |
+
"grad_norm": 125.42192077636719,
|
| 2271 |
+
"learning_rate": 9.994452492183215e-06,
|
| 2272 |
+
"loss": 13.1687,
|
| 2273 |
+
"step": 3200
|
| 2274 |
+
},
|
| 2275 |
+
{
|
| 2276 |
+
"epoch": 0.14246009952235925,
|
| 2277 |
+
"grad_norm": 167.23255920410156,
|
| 2278 |
+
"learning_rate": 9.994435156221288e-06,
|
| 2279 |
+
"loss": 13.1131,
|
| 2280 |
+
"step": 3210
|
| 2281 |
+
},
|
| 2282 |
+
{
|
| 2283 |
+
"epoch": 0.1429039004554507,
|
| 2284 |
+
"grad_norm": 160.59434509277344,
|
| 2285 |
+
"learning_rate": 9.994417820259361e-06,
|
| 2286 |
+
"loss": 12.6661,
|
| 2287 |
+
"step": 3220
|
| 2288 |
+
},
|
| 2289 |
+
{
|
| 2290 |
+
"epoch": 0.14334770138854216,
|
| 2291 |
+
"grad_norm": 157.50003051757812,
|
| 2292 |
+
"learning_rate": 9.994400484297432e-06,
|
| 2293 |
+
"loss": 13.0758,
|
| 2294 |
+
"step": 3230
|
| 2295 |
+
},
|
| 2296 |
+
{
|
| 2297 |
+
"epoch": 0.14379150232163362,
|
| 2298 |
+
"grad_norm": 152.7613067626953,
|
| 2299 |
+
"learning_rate": 9.994383148335505e-06,
|
| 2300 |
+
"loss": 12.6521,
|
| 2301 |
+
"step": 3240
|
| 2302 |
+
},
|
| 2303 |
+
{
|
| 2304 |
+
"epoch": 0.14423530325472508,
|
| 2305 |
+
"grad_norm": 120.58446502685547,
|
| 2306 |
+
"learning_rate": 9.994365812373578e-06,
|
| 2307 |
+
"loss": 12.9558,
|
| 2308 |
+
"step": 3250
|
| 2309 |
+
},
|
| 2310 |
+
{
|
| 2311 |
+
"epoch": 0.14467910418781654,
|
| 2312 |
+
"grad_norm": 130.59677124023438,
|
| 2313 |
+
"learning_rate": 9.99434847641165e-06,
|
| 2314 |
+
"loss": 12.919,
|
| 2315 |
+
"step": 3260
|
| 2316 |
+
},
|
| 2317 |
+
{
|
| 2318 |
+
"epoch": 0.14512290512090803,
|
| 2319 |
+
"grad_norm": 118.21358489990234,
|
| 2320 |
+
"learning_rate": 9.994331140449723e-06,
|
| 2321 |
+
"loss": 12.8244,
|
| 2322 |
+
"step": 3270
|
| 2323 |
+
},
|
| 2324 |
+
{
|
| 2325 |
+
"epoch": 0.1455667060539995,
|
| 2326 |
+
"grad_norm": 139.61123657226562,
|
| 2327 |
+
"learning_rate": 9.994313804487796e-06,
|
| 2328 |
+
"loss": 12.6272,
|
| 2329 |
+
"step": 3280
|
| 2330 |
+
},
|
| 2331 |
+
{
|
| 2332 |
+
"epoch": 0.14601050698709095,
|
| 2333 |
+
"grad_norm": 118.49354553222656,
|
| 2334 |
+
"learning_rate": 9.994296468525867e-06,
|
| 2335 |
+
"loss": 12.9009,
|
| 2336 |
+
"step": 3290
|
| 2337 |
+
},
|
| 2338 |
+
{
|
| 2339 |
+
"epoch": 0.1464543079201824,
|
| 2340 |
+
"grad_norm": 189.260498046875,
|
| 2341 |
+
"learning_rate": 9.99427913256394e-06,
|
| 2342 |
+
"loss": 12.7051,
|
| 2343 |
+
"step": 3300
|
| 2344 |
+
},
|
| 2345 |
+
{
|
| 2346 |
+
"epoch": 0.14689810885327387,
|
| 2347 |
+
"grad_norm": 146.55145263671875,
|
| 2348 |
+
"learning_rate": 9.994261796602014e-06,
|
| 2349 |
+
"loss": 12.7917,
|
| 2350 |
+
"step": 3310
|
| 2351 |
+
},
|
| 2352 |
+
{
|
| 2353 |
+
"epoch": 0.14734190978636533,
|
| 2354 |
+
"grad_norm": 140.71144104003906,
|
| 2355 |
+
"learning_rate": 9.994244460640085e-06,
|
| 2356 |
+
"loss": 12.2915,
|
| 2357 |
+
"step": 3320
|
| 2358 |
+
},
|
| 2359 |
+
{
|
| 2360 |
+
"epoch": 0.1477857107194568,
|
| 2361 |
+
"grad_norm": 153.54945373535156,
|
| 2362 |
+
"learning_rate": 9.994227124678158e-06,
|
| 2363 |
+
"loss": 12.4179,
|
| 2364 |
+
"step": 3330
|
| 2365 |
+
},
|
| 2366 |
+
{
|
| 2367 |
+
"epoch": 0.14822951165254825,
|
| 2368 |
+
"grad_norm": 113.84798431396484,
|
| 2369 |
+
"learning_rate": 9.994209788716231e-06,
|
| 2370 |
+
"loss": 12.8944,
|
| 2371 |
+
"step": 3340
|
| 2372 |
+
},
|
| 2373 |
+
{
|
| 2374 |
+
"epoch": 0.1486733125856397,
|
| 2375 |
+
"grad_norm": 140.952392578125,
|
| 2376 |
+
"learning_rate": 9.994192452754304e-06,
|
| 2377 |
+
"loss": 13.0441,
|
| 2378 |
+
"step": 3350
|
| 2379 |
+
},
|
| 2380 |
+
{
|
| 2381 |
+
"epoch": 0.14911711351873116,
|
| 2382 |
+
"grad_norm": 118.15951538085938,
|
| 2383 |
+
"learning_rate": 9.994175116792376e-06,
|
| 2384 |
+
"loss": 12.8967,
|
| 2385 |
+
"step": 3360
|
| 2386 |
+
},
|
| 2387 |
+
{
|
| 2388 |
+
"epoch": 0.14956091445182262,
|
| 2389 |
+
"grad_norm": 153.52896118164062,
|
| 2390 |
+
"learning_rate": 9.994157780830449e-06,
|
| 2391 |
+
"loss": 12.8702,
|
| 2392 |
+
"step": 3370
|
| 2393 |
+
},
|
| 2394 |
+
{
|
| 2395 |
+
"epoch": 0.15000471538491408,
|
| 2396 |
+
"grad_norm": 127.21448516845703,
|
| 2397 |
+
"learning_rate": 9.994140444868522e-06,
|
| 2398 |
+
"loss": 12.4004,
|
| 2399 |
+
"step": 3380
|
| 2400 |
+
},
|
| 2401 |
+
{
|
| 2402 |
+
"epoch": 0.15044851631800557,
|
| 2403 |
+
"grad_norm": 131.0957489013672,
|
| 2404 |
+
"learning_rate": 9.994123108906593e-06,
|
| 2405 |
+
"loss": 13.3359,
|
| 2406 |
+
"step": 3390
|
| 2407 |
+
},
|
| 2408 |
+
{
|
| 2409 |
+
"epoch": 0.15089231725109703,
|
| 2410 |
+
"grad_norm": 128.41253662109375,
|
| 2411 |
+
"learning_rate": 9.994105772944666e-06,
|
| 2412 |
+
"loss": 12.3271,
|
| 2413 |
+
"step": 3400
|
| 2414 |
+
},
|
| 2415 |
+
{
|
| 2416 |
+
"epoch": 0.1513361181841885,
|
| 2417 |
+
"grad_norm": 134.6806640625,
|
| 2418 |
+
"learning_rate": 9.99408843698274e-06,
|
| 2419 |
+
"loss": 12.5313,
|
| 2420 |
+
"step": 3410
|
| 2421 |
+
},
|
| 2422 |
+
{
|
| 2423 |
+
"epoch": 0.15177991911727995,
|
| 2424 |
+
"grad_norm": 119.5546646118164,
|
| 2425 |
+
"learning_rate": 9.99407110102081e-06,
|
| 2426 |
+
"loss": 12.3448,
|
| 2427 |
+
"step": 3420
|
| 2428 |
+
},
|
| 2429 |
+
{
|
| 2430 |
+
"epoch": 0.1522237200503714,
|
| 2431 |
+
"grad_norm": 138.1034393310547,
|
| 2432 |
+
"learning_rate": 9.994053765058884e-06,
|
| 2433 |
+
"loss": 13.0286,
|
| 2434 |
+
"step": 3430
|
| 2435 |
+
},
|
| 2436 |
+
{
|
| 2437 |
+
"epoch": 0.15266752098346287,
|
| 2438 |
+
"grad_norm": 110.58820343017578,
|
| 2439 |
+
"learning_rate": 9.994036429096957e-06,
|
| 2440 |
+
"loss": 12.455,
|
| 2441 |
+
"step": 3440
|
| 2442 |
+
},
|
| 2443 |
+
{
|
| 2444 |
+
"epoch": 0.15311132191655433,
|
| 2445 |
+
"grad_norm": 168.3693084716797,
|
| 2446 |
+
"learning_rate": 9.994019093135028e-06,
|
| 2447 |
+
"loss": 12.7619,
|
| 2448 |
+
"step": 3450
|
| 2449 |
+
},
|
| 2450 |
+
{
|
| 2451 |
+
"epoch": 0.1535551228496458,
|
| 2452 |
+
"grad_norm": 140.2900390625,
|
| 2453 |
+
"learning_rate": 9.994001757173101e-06,
|
| 2454 |
+
"loss": 12.8144,
|
| 2455 |
+
"step": 3460
|
| 2456 |
+
},
|
| 2457 |
+
{
|
| 2458 |
+
"epoch": 0.15399892378273725,
|
| 2459 |
+
"grad_norm": 108.32037353515625,
|
| 2460 |
+
"learning_rate": 9.993984421211174e-06,
|
| 2461 |
+
"loss": 12.6378,
|
| 2462 |
+
"step": 3470
|
| 2463 |
+
},
|
| 2464 |
+
{
|
| 2465 |
+
"epoch": 0.1544427247158287,
|
| 2466 |
+
"grad_norm": 137.5443115234375,
|
| 2467 |
+
"learning_rate": 9.993967085249247e-06,
|
| 2468 |
+
"loss": 12.2405,
|
| 2469 |
+
"step": 3480
|
| 2470 |
+
},
|
| 2471 |
+
{
|
| 2472 |
+
"epoch": 0.15488652564892016,
|
| 2473 |
+
"grad_norm": 157.03700256347656,
|
| 2474 |
+
"learning_rate": 9.993949749287319e-06,
|
| 2475 |
+
"loss": 12.654,
|
| 2476 |
+
"step": 3490
|
| 2477 |
+
},
|
| 2478 |
+
{
|
| 2479 |
+
"epoch": 0.15533032658201165,
|
| 2480 |
+
"grad_norm": 145.71405029296875,
|
| 2481 |
+
"learning_rate": 9.993932413325392e-06,
|
| 2482 |
+
"loss": 12.7073,
|
| 2483 |
+
"step": 3500
|
| 2484 |
+
},
|
| 2485 |
+
{
|
| 2486 |
+
"epoch": 0.1557741275151031,
|
| 2487 |
+
"grad_norm": 130.26255798339844,
|
| 2488 |
+
"learning_rate": 9.993915077363465e-06,
|
| 2489 |
+
"loss": 12.0506,
|
| 2490 |
+
"step": 3510
|
| 2491 |
+
},
|
| 2492 |
+
{
|
| 2493 |
+
"epoch": 0.15621792844819457,
|
| 2494 |
+
"grad_norm": 131.19674682617188,
|
| 2495 |
+
"learning_rate": 9.993897741401536e-06,
|
| 2496 |
+
"loss": 12.4799,
|
| 2497 |
+
"step": 3520
|
| 2498 |
+
},
|
| 2499 |
+
{
|
| 2500 |
+
"epoch": 0.15666172938128603,
|
| 2501 |
+
"grad_norm": 116.03231811523438,
|
| 2502 |
+
"learning_rate": 9.99388040543961e-06,
|
| 2503 |
+
"loss": 12.8077,
|
| 2504 |
+
"step": 3530
|
| 2505 |
+
},
|
| 2506 |
+
{
|
| 2507 |
+
"epoch": 0.1571055303143775,
|
| 2508 |
+
"grad_norm": 117.18372344970703,
|
| 2509 |
+
"learning_rate": 9.993863069477682e-06,
|
| 2510 |
+
"loss": 12.754,
|
| 2511 |
+
"step": 3540
|
| 2512 |
+
},
|
| 2513 |
+
{
|
| 2514 |
+
"epoch": 0.15754933124746895,
|
| 2515 |
+
"grad_norm": 141.51046752929688,
|
| 2516 |
+
"learning_rate": 9.993845733515754e-06,
|
| 2517 |
+
"loss": 12.9482,
|
| 2518 |
+
"step": 3550
|
| 2519 |
+
},
|
| 2520 |
+
{
|
| 2521 |
+
"epoch": 0.1579931321805604,
|
| 2522 |
+
"grad_norm": 145.37522888183594,
|
| 2523 |
+
"learning_rate": 9.993828397553827e-06,
|
| 2524 |
+
"loss": 12.9102,
|
| 2525 |
+
"step": 3560
|
| 2526 |
+
},
|
| 2527 |
+
{
|
| 2528 |
+
"epoch": 0.15843693311365187,
|
| 2529 |
+
"grad_norm": 121.03300476074219,
|
| 2530 |
+
"learning_rate": 9.9938110615919e-06,
|
| 2531 |
+
"loss": 12.5428,
|
| 2532 |
+
"step": 3570
|
| 2533 |
+
},
|
| 2534 |
+
{
|
| 2535 |
+
"epoch": 0.15888073404674333,
|
| 2536 |
+
"grad_norm": 97.84991455078125,
|
| 2537 |
+
"learning_rate": 9.993793725629971e-06,
|
| 2538 |
+
"loss": 12.2027,
|
| 2539 |
+
"step": 3580
|
| 2540 |
+
},
|
| 2541 |
+
{
|
| 2542 |
+
"epoch": 0.1593245349798348,
|
| 2543 |
+
"grad_norm": 139.6411590576172,
|
| 2544 |
+
"learning_rate": 9.993776389668044e-06,
|
| 2545 |
+
"loss": 12.6947,
|
| 2546 |
+
"step": 3590
|
| 2547 |
+
},
|
| 2548 |
+
{
|
| 2549 |
+
"epoch": 0.15976833591292625,
|
| 2550 |
+
"grad_norm": 128.70204162597656,
|
| 2551 |
+
"learning_rate": 9.993759053706118e-06,
|
| 2552 |
+
"loss": 12.7905,
|
| 2553 |
+
"step": 3600
|
| 2554 |
+
},
|
| 2555 |
+
{
|
| 2556 |
+
"epoch": 0.1602121368460177,
|
| 2557 |
+
"grad_norm": 118.03040313720703,
|
| 2558 |
+
"learning_rate": 9.99374171774419e-06,
|
| 2559 |
+
"loss": 12.8123,
|
| 2560 |
+
"step": 3610
|
| 2561 |
+
},
|
| 2562 |
+
{
|
| 2563 |
+
"epoch": 0.1606559377791092,
|
| 2564 |
+
"grad_norm": 111.89470672607422,
|
| 2565 |
+
"learning_rate": 9.993724381782262e-06,
|
| 2566 |
+
"loss": 12.5624,
|
| 2567 |
+
"step": 3620
|
| 2568 |
+
},
|
| 2569 |
+
{
|
| 2570 |
+
"epoch": 0.16109973871220065,
|
| 2571 |
+
"grad_norm": 129.9048614501953,
|
| 2572 |
+
"learning_rate": 9.993707045820335e-06,
|
| 2573 |
+
"loss": 12.6917,
|
| 2574 |
+
"step": 3630
|
| 2575 |
+
},
|
| 2576 |
+
{
|
| 2577 |
+
"epoch": 0.1615435396452921,
|
| 2578 |
+
"grad_norm": 127.17546844482422,
|
| 2579 |
+
"learning_rate": 9.993689709858408e-06,
|
| 2580 |
+
"loss": 13.1546,
|
| 2581 |
+
"step": 3640
|
| 2582 |
+
},
|
| 2583 |
+
{
|
| 2584 |
+
"epoch": 0.16198734057838357,
|
| 2585 |
+
"grad_norm": 117.04710388183594,
|
| 2586 |
+
"learning_rate": 9.99367237389648e-06,
|
| 2587 |
+
"loss": 12.452,
|
| 2588 |
+
"step": 3650
|
| 2589 |
+
},
|
| 2590 |
+
{
|
| 2591 |
+
"epoch": 0.16243114151147503,
|
| 2592 |
+
"grad_norm": 133.599853515625,
|
| 2593 |
+
"learning_rate": 9.993655037934553e-06,
|
| 2594 |
+
"loss": 12.5182,
|
| 2595 |
+
"step": 3660
|
| 2596 |
+
},
|
| 2597 |
+
{
|
| 2598 |
+
"epoch": 0.1628749424445665,
|
| 2599 |
+
"grad_norm": 158.9300994873047,
|
| 2600 |
+
"learning_rate": 9.993637701972626e-06,
|
| 2601 |
+
"loss": 12.0698,
|
| 2602 |
+
"step": 3670
|
| 2603 |
+
},
|
| 2604 |
+
{
|
| 2605 |
+
"epoch": 0.16331874337765795,
|
| 2606 |
+
"grad_norm": 125.10253143310547,
|
| 2607 |
+
"learning_rate": 9.993620366010697e-06,
|
| 2608 |
+
"loss": 12.5825,
|
| 2609 |
+
"step": 3680
|
| 2610 |
+
},
|
| 2611 |
+
{
|
| 2612 |
+
"epoch": 0.1637625443107494,
|
| 2613 |
+
"grad_norm": 129.07908630371094,
|
| 2614 |
+
"learning_rate": 9.99360303004877e-06,
|
| 2615 |
+
"loss": 12.5439,
|
| 2616 |
+
"step": 3690
|
| 2617 |
+
},
|
| 2618 |
+
{
|
| 2619 |
+
"epoch": 0.16420634524384087,
|
| 2620 |
+
"grad_norm": 119.1418685913086,
|
| 2621 |
+
"learning_rate": 9.993585694086843e-06,
|
| 2622 |
+
"loss": 12.564,
|
| 2623 |
+
"step": 3700
|
| 2624 |
+
},
|
| 2625 |
+
{
|
| 2626 |
+
"epoch": 0.16465014617693233,
|
| 2627 |
+
"grad_norm": 128.37750244140625,
|
| 2628 |
+
"learning_rate": 9.993568358124915e-06,
|
| 2629 |
+
"loss": 12.9106,
|
| 2630 |
+
"step": 3710
|
| 2631 |
+
},
|
| 2632 |
+
{
|
| 2633 |
+
"epoch": 0.1650939471100238,
|
| 2634 |
+
"grad_norm": 142.9105987548828,
|
| 2635 |
+
"learning_rate": 9.993551022162988e-06,
|
| 2636 |
+
"loss": 12.9379,
|
| 2637 |
+
"step": 3720
|
| 2638 |
+
},
|
| 2639 |
+
{
|
| 2640 |
+
"epoch": 0.16553774804311527,
|
| 2641 |
+
"grad_norm": 118.88358306884766,
|
| 2642 |
+
"learning_rate": 9.99353368620106e-06,
|
| 2643 |
+
"loss": 12.3977,
|
| 2644 |
+
"step": 3730
|
| 2645 |
+
},
|
| 2646 |
+
{
|
| 2647 |
+
"epoch": 0.16598154897620673,
|
| 2648 |
+
"grad_norm": 118.91588592529297,
|
| 2649 |
+
"learning_rate": 9.993516350239134e-06,
|
| 2650 |
+
"loss": 13.2059,
|
| 2651 |
+
"step": 3740
|
| 2652 |
+
},
|
| 2653 |
+
{
|
| 2654 |
+
"epoch": 0.1664253499092982,
|
| 2655 |
+
"grad_norm": 139.88050842285156,
|
| 2656 |
+
"learning_rate": 9.993499014277205e-06,
|
| 2657 |
+
"loss": 13.0386,
|
| 2658 |
+
"step": 3750
|
| 2659 |
+
},
|
| 2660 |
+
{
|
| 2661 |
+
"epoch": 0.16686915084238965,
|
| 2662 |
+
"grad_norm": 162.86024475097656,
|
| 2663 |
+
"learning_rate": 9.993481678315278e-06,
|
| 2664 |
+
"loss": 12.238,
|
| 2665 |
+
"step": 3760
|
| 2666 |
+
},
|
| 2667 |
+
{
|
| 2668 |
+
"epoch": 0.1673129517754811,
|
| 2669 |
+
"grad_norm": 127.96647644042969,
|
| 2670 |
+
"learning_rate": 9.993464342353351e-06,
|
| 2671 |
+
"loss": 13.2272,
|
| 2672 |
+
"step": 3770
|
| 2673 |
+
},
|
| 2674 |
+
{
|
| 2675 |
+
"epoch": 0.16775675270857257,
|
| 2676 |
+
"grad_norm": 112.12999725341797,
|
| 2677 |
+
"learning_rate": 9.993447006391423e-06,
|
| 2678 |
+
"loss": 12.5457,
|
| 2679 |
+
"step": 3780
|
| 2680 |
+
},
|
| 2681 |
+
{
|
| 2682 |
+
"epoch": 0.16820055364166403,
|
| 2683 |
+
"grad_norm": 155.6081085205078,
|
| 2684 |
+
"learning_rate": 9.993429670429496e-06,
|
| 2685 |
+
"loss": 12.3183,
|
| 2686 |
+
"step": 3790
|
| 2687 |
+
},
|
| 2688 |
+
{
|
| 2689 |
+
"epoch": 0.1686443545747555,
|
| 2690 |
+
"grad_norm": 138.3584442138672,
|
| 2691 |
+
"learning_rate": 9.993412334467569e-06,
|
| 2692 |
+
"loss": 12.4889,
|
| 2693 |
+
"step": 3800
|
| 2694 |
+
},
|
| 2695 |
+
{
|
| 2696 |
+
"epoch": 0.16908815550784695,
|
| 2697 |
+
"grad_norm": 115.07450866699219,
|
| 2698 |
+
"learning_rate": 9.99339499850564e-06,
|
| 2699 |
+
"loss": 12.9653,
|
| 2700 |
+
"step": 3810
|
| 2701 |
+
},
|
| 2702 |
+
{
|
| 2703 |
+
"epoch": 0.1695319564409384,
|
| 2704 |
+
"grad_norm": 126.59088897705078,
|
| 2705 |
+
"learning_rate": 9.993377662543713e-06,
|
| 2706 |
+
"loss": 12.9003,
|
| 2707 |
+
"step": 3820
|
| 2708 |
+
},
|
| 2709 |
+
{
|
| 2710 |
+
"epoch": 0.16997575737402987,
|
| 2711 |
+
"grad_norm": 108.5206069946289,
|
| 2712 |
+
"learning_rate": 9.993360326581786e-06,
|
| 2713 |
+
"loss": 13.1251,
|
| 2714 |
+
"step": 3830
|
| 2715 |
+
},
|
| 2716 |
+
{
|
| 2717 |
+
"epoch": 0.17041955830712133,
|
| 2718 |
+
"grad_norm": 124.42080688476562,
|
| 2719 |
+
"learning_rate": 9.993342990619858e-06,
|
| 2720 |
+
"loss": 12.9417,
|
| 2721 |
+
"step": 3840
|
| 2722 |
+
},
|
| 2723 |
+
{
|
| 2724 |
+
"epoch": 0.17086335924021281,
|
| 2725 |
+
"grad_norm": 106.24859619140625,
|
| 2726 |
+
"learning_rate": 9.993325654657931e-06,
|
| 2727 |
+
"loss": 12.3298,
|
| 2728 |
+
"step": 3850
|
| 2729 |
+
},
|
| 2730 |
+
{
|
| 2731 |
+
"epoch": 0.17130716017330427,
|
| 2732 |
+
"grad_norm": 112.69037628173828,
|
| 2733 |
+
"learning_rate": 9.993308318696004e-06,
|
| 2734 |
+
"loss": 12.3989,
|
| 2735 |
+
"step": 3860
|
| 2736 |
+
},
|
| 2737 |
+
{
|
| 2738 |
+
"epoch": 0.17175096110639573,
|
| 2739 |
+
"grad_norm": 154.77882385253906,
|
| 2740 |
+
"learning_rate": 9.993290982734077e-06,
|
| 2741 |
+
"loss": 13.245,
|
| 2742 |
+
"step": 3870
|
| 2743 |
+
},
|
| 2744 |
+
{
|
| 2745 |
+
"epoch": 0.1721947620394872,
|
| 2746 |
+
"grad_norm": 123.77738952636719,
|
| 2747 |
+
"learning_rate": 9.993273646772148e-06,
|
| 2748 |
+
"loss": 12.6657,
|
| 2749 |
+
"step": 3880
|
| 2750 |
+
},
|
| 2751 |
+
{
|
| 2752 |
+
"epoch": 0.17263856297257865,
|
| 2753 |
+
"grad_norm": 126.2992172241211,
|
| 2754 |
+
"learning_rate": 9.993256310810222e-06,
|
| 2755 |
+
"loss": 12.3853,
|
| 2756 |
+
"step": 3890
|
| 2757 |
+
},
|
| 2758 |
+
{
|
| 2759 |
+
"epoch": 0.1730823639056701,
|
| 2760 |
+
"grad_norm": 112.3527603149414,
|
| 2761 |
+
"learning_rate": 9.993238974848295e-06,
|
| 2762 |
+
"loss": 13.0207,
|
| 2763 |
+
"step": 3900
|
| 2764 |
+
},
|
| 2765 |
+
{
|
| 2766 |
+
"epoch": 0.17352616483876157,
|
| 2767 |
+
"grad_norm": 106.91303253173828,
|
| 2768 |
+
"learning_rate": 9.993221638886366e-06,
|
| 2769 |
+
"loss": 11.9467,
|
| 2770 |
+
"step": 3910
|
| 2771 |
+
},
|
| 2772 |
+
{
|
| 2773 |
+
"epoch": 0.17396996577185303,
|
| 2774 |
+
"grad_norm": 122.75757598876953,
|
| 2775 |
+
"learning_rate": 9.993204302924439e-06,
|
| 2776 |
+
"loss": 12.1821,
|
| 2777 |
+
"step": 3920
|
| 2778 |
+
},
|
| 2779 |
+
{
|
| 2780 |
+
"epoch": 0.1744137667049445,
|
| 2781 |
+
"grad_norm": 116.6715316772461,
|
| 2782 |
+
"learning_rate": 9.993186966962512e-06,
|
| 2783 |
+
"loss": 12.1554,
|
| 2784 |
+
"step": 3930
|
| 2785 |
+
},
|
| 2786 |
+
{
|
| 2787 |
+
"epoch": 0.17485756763803595,
|
| 2788 |
+
"grad_norm": 100.55429077148438,
|
| 2789 |
+
"learning_rate": 9.993169631000584e-06,
|
| 2790 |
+
"loss": 12.6786,
|
| 2791 |
+
"step": 3940
|
| 2792 |
+
},
|
| 2793 |
+
{
|
| 2794 |
+
"epoch": 0.1753013685711274,
|
| 2795 |
+
"grad_norm": 116.2847671508789,
|
| 2796 |
+
"learning_rate": 9.993152295038657e-06,
|
| 2797 |
+
"loss": 11.9186,
|
| 2798 |
+
"step": 3950
|
| 2799 |
+
},
|
| 2800 |
+
{
|
| 2801 |
+
"epoch": 0.1757451695042189,
|
| 2802 |
+
"grad_norm": 107.30577087402344,
|
| 2803 |
+
"learning_rate": 9.99313495907673e-06,
|
| 2804 |
+
"loss": 12.8207,
|
| 2805 |
+
"step": 3960
|
| 2806 |
+
},
|
| 2807 |
+
{
|
| 2808 |
+
"epoch": 0.17618897043731035,
|
| 2809 |
+
"grad_norm": 103.84146118164062,
|
| 2810 |
+
"learning_rate": 9.993117623114803e-06,
|
| 2811 |
+
"loss": 12.4032,
|
| 2812 |
+
"step": 3970
|
| 2813 |
+
},
|
| 2814 |
+
{
|
| 2815 |
+
"epoch": 0.17663277137040181,
|
| 2816 |
+
"grad_norm": 108.5525131225586,
|
| 2817 |
+
"learning_rate": 9.993100287152874e-06,
|
| 2818 |
+
"loss": 12.3449,
|
| 2819 |
+
"step": 3980
|
| 2820 |
+
},
|
| 2821 |
+
{
|
| 2822 |
+
"epoch": 0.17707657230349327,
|
| 2823 |
+
"grad_norm": 132.21353149414062,
|
| 2824 |
+
"learning_rate": 9.993082951190947e-06,
|
| 2825 |
+
"loss": 12.8847,
|
| 2826 |
+
"step": 3990
|
| 2827 |
+
},
|
| 2828 |
+
{
|
| 2829 |
+
"epoch": 0.17752037323658473,
|
| 2830 |
+
"grad_norm": 114.80290985107422,
|
| 2831 |
+
"learning_rate": 9.99306561522902e-06,
|
| 2832 |
+
"loss": 12.5004,
|
| 2833 |
+
"step": 4000
|
| 2834 |
+
},
|
| 2835 |
+
{
|
| 2836 |
+
"epoch": 0.17752037323658473,
|
| 2837 |
+
"eval_loss": 0.38956519961357117,
|
| 2838 |
+
"eval_runtime": 673.1566,
|
| 2839 |
+
"eval_samples_per_second": 1804.024,
|
| 2840 |
+
"eval_steps_per_second": 56.376,
|
| 2841 |
+
"step": 4000
|
| 2842 |
+
},
|
| 2843 |
+
{
|
| 2844 |
+
"epoch": 0.1779641741696762,
|
| 2845 |
+
"grad_norm": 123.59729766845703,
|
| 2846 |
+
"learning_rate": 9.993048279267092e-06,
|
| 2847 |
+
"loss": 11.8187,
|
| 2848 |
+
"step": 4010
|
| 2849 |
+
},
|
| 2850 |
+
{
|
| 2851 |
+
"epoch": 0.17840797510276765,
|
| 2852 |
+
"grad_norm": 121.13542175292969,
|
| 2853 |
+
"learning_rate": 9.993030943305165e-06,
|
| 2854 |
+
"loss": 12.569,
|
| 2855 |
+
"step": 4020
|
| 2856 |
+
},
|
| 2857 |
+
{
|
| 2858 |
+
"epoch": 0.1788517760358591,
|
| 2859 |
+
"grad_norm": 112.8544692993164,
|
| 2860 |
+
"learning_rate": 9.993013607343238e-06,
|
| 2861 |
+
"loss": 12.4908,
|
| 2862 |
+
"step": 4030
|
| 2863 |
+
},
|
| 2864 |
+
{
|
| 2865 |
+
"epoch": 0.17929557696895057,
|
| 2866 |
+
"grad_norm": 108.49791717529297,
|
| 2867 |
+
"learning_rate": 9.99299627138131e-06,
|
| 2868 |
+
"loss": 12.1717,
|
| 2869 |
+
"step": 4040
|
| 2870 |
+
},
|
| 2871 |
+
{
|
| 2872 |
+
"epoch": 0.17973937790204203,
|
| 2873 |
+
"grad_norm": 112.78793334960938,
|
| 2874 |
+
"learning_rate": 9.992978935419382e-06,
|
| 2875 |
+
"loss": 12.4246,
|
| 2876 |
+
"step": 4050
|
| 2877 |
+
},
|
| 2878 |
+
{
|
| 2879 |
+
"epoch": 0.1801831788351335,
|
| 2880 |
+
"grad_norm": 111.55317687988281,
|
| 2881 |
+
"learning_rate": 9.992961599457455e-06,
|
| 2882 |
+
"loss": 12.5279,
|
| 2883 |
+
"step": 4060
|
| 2884 |
+
},
|
| 2885 |
+
{
|
| 2886 |
+
"epoch": 0.18062697976822495,
|
| 2887 |
+
"grad_norm": 129.81776428222656,
|
| 2888 |
+
"learning_rate": 9.992944263495527e-06,
|
| 2889 |
+
"loss": 12.0691,
|
| 2890 |
+
"step": 4070
|
| 2891 |
+
},
|
| 2892 |
+
{
|
| 2893 |
+
"epoch": 0.18107078070131644,
|
| 2894 |
+
"grad_norm": 159.69786071777344,
|
| 2895 |
+
"learning_rate": 9.9929269275336e-06,
|
| 2896 |
+
"loss": 12.4083,
|
| 2897 |
+
"step": 4080
|
| 2898 |
+
},
|
| 2899 |
+
{
|
| 2900 |
+
"epoch": 0.1815145816344079,
|
| 2901 |
+
"grad_norm": 119.66799926757812,
|
| 2902 |
+
"learning_rate": 9.992909591571673e-06,
|
| 2903 |
+
"loss": 12.1316,
|
| 2904 |
+
"step": 4090
|
| 2905 |
+
},
|
| 2906 |
+
{
|
| 2907 |
+
"epoch": 0.18195838256749935,
|
| 2908 |
+
"grad_norm": 143.12181091308594,
|
| 2909 |
+
"learning_rate": 9.992892255609746e-06,
|
| 2910 |
+
"loss": 12.5121,
|
| 2911 |
+
"step": 4100
|
| 2912 |
+
},
|
| 2913 |
+
{
|
| 2914 |
+
"epoch": 0.18240218350059081,
|
| 2915 |
+
"grad_norm": 153.3807373046875,
|
| 2916 |
+
"learning_rate": 9.992874919647817e-06,
|
| 2917 |
+
"loss": 12.0202,
|
| 2918 |
+
"step": 4110
|
| 2919 |
+
},
|
| 2920 |
+
{
|
| 2921 |
+
"epoch": 0.18284598443368227,
|
| 2922 |
+
"grad_norm": 127.23503875732422,
|
| 2923 |
+
"learning_rate": 9.99285758368589e-06,
|
| 2924 |
+
"loss": 12.3906,
|
| 2925 |
+
"step": 4120
|
| 2926 |
+
},
|
| 2927 |
+
{
|
| 2928 |
+
"epoch": 0.18328978536677373,
|
| 2929 |
+
"grad_norm": 130.62158203125,
|
| 2930 |
+
"learning_rate": 9.992840247723964e-06,
|
| 2931 |
+
"loss": 12.0698,
|
| 2932 |
+
"step": 4130
|
| 2933 |
+
},
|
| 2934 |
+
{
|
| 2935 |
+
"epoch": 0.1837335862998652,
|
| 2936 |
+
"grad_norm": 126.1018295288086,
|
| 2937 |
+
"learning_rate": 9.992822911762035e-06,
|
| 2938 |
+
"loss": 12.4342,
|
| 2939 |
+
"step": 4140
|
| 2940 |
+
},
|
| 2941 |
+
{
|
| 2942 |
+
"epoch": 0.18417738723295665,
|
| 2943 |
+
"grad_norm": 131.96340942382812,
|
| 2944 |
+
"learning_rate": 9.992805575800108e-06,
|
| 2945 |
+
"loss": 12.6071,
|
| 2946 |
+
"step": 4150
|
| 2947 |
+
},
|
| 2948 |
+
{
|
| 2949 |
+
"epoch": 0.1846211881660481,
|
| 2950 |
+
"grad_norm": 130.8789520263672,
|
| 2951 |
+
"learning_rate": 9.992788239838181e-06,
|
| 2952 |
+
"loss": 12.323,
|
| 2953 |
+
"step": 4160
|
| 2954 |
+
},
|
| 2955 |
+
{
|
| 2956 |
+
"epoch": 0.18506498909913957,
|
| 2957 |
+
"grad_norm": 125.24624633789062,
|
| 2958 |
+
"learning_rate": 9.992770903876252e-06,
|
| 2959 |
+
"loss": 11.9601,
|
| 2960 |
+
"step": 4170
|
| 2961 |
+
},
|
| 2962 |
+
{
|
| 2963 |
+
"epoch": 0.18550879003223103,
|
| 2964 |
+
"grad_norm": 107.27322387695312,
|
| 2965 |
+
"learning_rate": 9.992753567914326e-06,
|
| 2966 |
+
"loss": 12.4659,
|
| 2967 |
+
"step": 4180
|
| 2968 |
+
},
|
| 2969 |
+
{
|
| 2970 |
+
"epoch": 0.18595259096532252,
|
| 2971 |
+
"grad_norm": 118.50186920166016,
|
| 2972 |
+
"learning_rate": 9.992736231952399e-06,
|
| 2973 |
+
"loss": 12.4699,
|
| 2974 |
+
"step": 4190
|
| 2975 |
+
},
|
| 2976 |
+
{
|
| 2977 |
+
"epoch": 0.18639639189841398,
|
| 2978 |
+
"grad_norm": 132.64605712890625,
|
| 2979 |
+
"learning_rate": 9.99271889599047e-06,
|
| 2980 |
+
"loss": 12.551,
|
| 2981 |
+
"step": 4200
|
| 2982 |
+
},
|
| 2983 |
+
{
|
| 2984 |
+
"epoch": 0.18684019283150544,
|
| 2985 |
+
"grad_norm": 129.10557556152344,
|
| 2986 |
+
"learning_rate": 9.992701560028543e-06,
|
| 2987 |
+
"loss": 12.5445,
|
| 2988 |
+
"step": 4210
|
| 2989 |
+
},
|
| 2990 |
+
{
|
| 2991 |
+
"epoch": 0.1872839937645969,
|
| 2992 |
+
"grad_norm": 118.41376495361328,
|
| 2993 |
+
"learning_rate": 9.992684224066616e-06,
|
| 2994 |
+
"loss": 12.4178,
|
| 2995 |
+
"step": 4220
|
| 2996 |
+
},
|
| 2997 |
+
{
|
| 2998 |
+
"epoch": 0.18772779469768835,
|
| 2999 |
+
"grad_norm": 115.65239715576172,
|
| 3000 |
+
"learning_rate": 9.992666888104688e-06,
|
| 3001 |
+
"loss": 12.4483,
|
| 3002 |
+
"step": 4230
|
| 3003 |
+
},
|
| 3004 |
+
{
|
| 3005 |
+
"epoch": 0.18817159563077981,
|
| 3006 |
+
"grad_norm": 125.76007080078125,
|
| 3007 |
+
"learning_rate": 9.99264955214276e-06,
|
| 3008 |
+
"loss": 13.088,
|
| 3009 |
+
"step": 4240
|
| 3010 |
+
},
|
| 3011 |
+
{
|
| 3012 |
+
"epoch": 0.18861539656387127,
|
| 3013 |
+
"grad_norm": 88.38092041015625,
|
| 3014 |
+
"learning_rate": 9.992632216180834e-06,
|
| 3015 |
+
"loss": 12.9315,
|
| 3016 |
+
"step": 4250
|
| 3017 |
+
},
|
| 3018 |
+
{
|
| 3019 |
+
"epoch": 0.18905919749696273,
|
| 3020 |
+
"grad_norm": 104.6368179321289,
|
| 3021 |
+
"learning_rate": 9.992614880218907e-06,
|
| 3022 |
+
"loss": 12.2731,
|
| 3023 |
+
"step": 4260
|
| 3024 |
+
},
|
| 3025 |
+
{
|
| 3026 |
+
"epoch": 0.1895029984300542,
|
| 3027 |
+
"grad_norm": 127.25080871582031,
|
| 3028 |
+
"learning_rate": 9.992597544256978e-06,
|
| 3029 |
+
"loss": 12.2454,
|
| 3030 |
+
"step": 4270
|
| 3031 |
+
},
|
| 3032 |
+
{
|
| 3033 |
+
"epoch": 0.18994679936314565,
|
| 3034 |
+
"grad_norm": 101.77681732177734,
|
| 3035 |
+
"learning_rate": 9.992580208295051e-06,
|
| 3036 |
+
"loss": 12.4424,
|
| 3037 |
+
"step": 4280
|
| 3038 |
+
},
|
| 3039 |
+
{
|
| 3040 |
+
"epoch": 0.1903906002962371,
|
| 3041 |
+
"grad_norm": 136.33221435546875,
|
| 3042 |
+
"learning_rate": 9.992562872333124e-06,
|
| 3043 |
+
"loss": 11.9918,
|
| 3044 |
+
"step": 4290
|
| 3045 |
+
},
|
| 3046 |
+
{
|
| 3047 |
+
"epoch": 0.1908344012293286,
|
| 3048 |
+
"grad_norm": 112.94047546386719,
|
| 3049 |
+
"learning_rate": 9.992545536371196e-06,
|
| 3050 |
+
"loss": 12.4194,
|
| 3051 |
+
"step": 4300
|
| 3052 |
+
},
|
| 3053 |
+
{
|
| 3054 |
+
"epoch": 0.19127820216242006,
|
| 3055 |
+
"grad_norm": 132.7883758544922,
|
| 3056 |
+
"learning_rate": 9.992528200409269e-06,
|
| 3057 |
+
"loss": 12.3539,
|
| 3058 |
+
"step": 4310
|
| 3059 |
+
},
|
| 3060 |
+
{
|
| 3061 |
+
"epoch": 0.19172200309551152,
|
| 3062 |
+
"grad_norm": 115.22777557373047,
|
| 3063 |
+
"learning_rate": 9.992510864447342e-06,
|
| 3064 |
+
"loss": 12.176,
|
| 3065 |
+
"step": 4320
|
| 3066 |
+
},
|
| 3067 |
+
{
|
| 3068 |
+
"epoch": 0.19216580402860298,
|
| 3069 |
+
"grad_norm": 133.0673828125,
|
| 3070 |
+
"learning_rate": 9.992493528485413e-06,
|
| 3071 |
+
"loss": 12.1488,
|
| 3072 |
+
"step": 4330
|
| 3073 |
+
},
|
| 3074 |
+
{
|
| 3075 |
+
"epoch": 0.19260960496169444,
|
| 3076 |
+
"grad_norm": 154.4739227294922,
|
| 3077 |
+
"learning_rate": 9.992476192523486e-06,
|
| 3078 |
+
"loss": 12.4385,
|
| 3079 |
+
"step": 4340
|
| 3080 |
+
},
|
| 3081 |
+
{
|
| 3082 |
+
"epoch": 0.1930534058947859,
|
| 3083 |
+
"grad_norm": 169.39413452148438,
|
| 3084 |
+
"learning_rate": 9.99245885656156e-06,
|
| 3085 |
+
"loss": 12.2436,
|
| 3086 |
+
"step": 4350
|
| 3087 |
+
},
|
| 3088 |
+
{
|
| 3089 |
+
"epoch": 0.19349720682787735,
|
| 3090 |
+
"grad_norm": 127.55028533935547,
|
| 3091 |
+
"learning_rate": 9.99244152059963e-06,
|
| 3092 |
+
"loss": 12.1869,
|
| 3093 |
+
"step": 4360
|
| 3094 |
+
},
|
| 3095 |
+
{
|
| 3096 |
+
"epoch": 0.19394100776096881,
|
| 3097 |
+
"grad_norm": 124.5598373413086,
|
| 3098 |
+
"learning_rate": 9.992424184637704e-06,
|
| 3099 |
+
"loss": 12.6008,
|
| 3100 |
+
"step": 4370
|
| 3101 |
+
},
|
| 3102 |
+
{
|
| 3103 |
+
"epoch": 0.19438480869406027,
|
| 3104 |
+
"grad_norm": 140.38067626953125,
|
| 3105 |
+
"learning_rate": 9.992406848675777e-06,
|
| 3106 |
+
"loss": 12.8844,
|
| 3107 |
+
"step": 4380
|
| 3108 |
+
},
|
| 3109 |
+
{
|
| 3110 |
+
"epoch": 0.19482860962715173,
|
| 3111 |
+
"grad_norm": 111.26557922363281,
|
| 3112 |
+
"learning_rate": 9.992389512713848e-06,
|
| 3113 |
+
"loss": 12.7417,
|
| 3114 |
+
"step": 4390
|
| 3115 |
+
},
|
| 3116 |
+
{
|
| 3117 |
+
"epoch": 0.1952724105602432,
|
| 3118 |
+
"grad_norm": 109.17289733886719,
|
| 3119 |
+
"learning_rate": 9.992372176751921e-06,
|
| 3120 |
+
"loss": 12.5357,
|
| 3121 |
+
"step": 4400
|
| 3122 |
+
},
|
| 3123 |
+
{
|
| 3124 |
+
"epoch": 0.19571621149333465,
|
| 3125 |
+
"grad_norm": 93.85588073730469,
|
| 3126 |
+
"learning_rate": 9.992354840789995e-06,
|
| 3127 |
+
"loss": 13.057,
|
| 3128 |
+
"step": 4410
|
| 3129 |
+
},
|
| 3130 |
+
{
|
| 3131 |
+
"epoch": 0.19616001242642614,
|
| 3132 |
+
"grad_norm": 132.5755157470703,
|
| 3133 |
+
"learning_rate": 9.992337504828066e-06,
|
| 3134 |
+
"loss": 11.8174,
|
| 3135 |
+
"step": 4420
|
| 3136 |
+
},
|
| 3137 |
+
{
|
| 3138 |
+
"epoch": 0.1966038133595176,
|
| 3139 |
+
"grad_norm": 137.10018920898438,
|
| 3140 |
+
"learning_rate": 9.992320168866139e-06,
|
| 3141 |
+
"loss": 12.8481,
|
| 3142 |
+
"step": 4430
|
| 3143 |
+
},
|
| 3144 |
+
{
|
| 3145 |
+
"epoch": 0.19704761429260906,
|
| 3146 |
+
"grad_norm": 114.44795989990234,
|
| 3147 |
+
"learning_rate": 9.992302832904212e-06,
|
| 3148 |
+
"loss": 12.6655,
|
| 3149 |
+
"step": 4440
|
| 3150 |
+
},
|
| 3151 |
+
{
|
| 3152 |
+
"epoch": 0.19749141522570052,
|
| 3153 |
+
"grad_norm": 101.33007049560547,
|
| 3154 |
+
"learning_rate": 9.992285496942283e-06,
|
| 3155 |
+
"loss": 11.2619,
|
| 3156 |
+
"step": 4450
|
| 3157 |
+
},
|
| 3158 |
+
{
|
| 3159 |
+
"epoch": 0.19793521615879198,
|
| 3160 |
+
"grad_norm": 121.8523941040039,
|
| 3161 |
+
"learning_rate": 9.992268160980357e-06,
|
| 3162 |
+
"loss": 11.8743,
|
| 3163 |
+
"step": 4460
|
| 3164 |
+
},
|
| 3165 |
+
{
|
| 3166 |
+
"epoch": 0.19837901709188344,
|
| 3167 |
+
"grad_norm": 133.41293334960938,
|
| 3168 |
+
"learning_rate": 9.99225082501843e-06,
|
| 3169 |
+
"loss": 12.4099,
|
| 3170 |
+
"step": 4470
|
| 3171 |
+
},
|
| 3172 |
+
{
|
| 3173 |
+
"epoch": 0.1988228180249749,
|
| 3174 |
+
"grad_norm": 121.90374755859375,
|
| 3175 |
+
"learning_rate": 9.992233489056503e-06,
|
| 3176 |
+
"loss": 12.8135,
|
| 3177 |
+
"step": 4480
|
| 3178 |
+
},
|
| 3179 |
+
{
|
| 3180 |
+
"epoch": 0.19926661895806635,
|
| 3181 |
+
"grad_norm": 108.74071502685547,
|
| 3182 |
+
"learning_rate": 9.992216153094574e-06,
|
| 3183 |
+
"loss": 11.9765,
|
| 3184 |
+
"step": 4490
|
| 3185 |
+
},
|
| 3186 |
+
{
|
| 3187 |
+
"epoch": 0.19971041989115781,
|
| 3188 |
+
"grad_norm": 112.89215850830078,
|
| 3189 |
+
"learning_rate": 9.992198817132647e-06,
|
| 3190 |
+
"loss": 12.9006,
|
| 3191 |
+
"step": 4500
|
| 3192 |
+
},
|
| 3193 |
+
{
|
| 3194 |
+
"epoch": 0.20015422082424927,
|
| 3195 |
+
"grad_norm": 132.44119262695312,
|
| 3196 |
+
"learning_rate": 9.99218148117072e-06,
|
| 3197 |
+
"loss": 12.6348,
|
| 3198 |
+
"step": 4510
|
| 3199 |
+
},
|
| 3200 |
+
{
|
| 3201 |
+
"epoch": 0.20059802175734073,
|
| 3202 |
+
"grad_norm": 132.33868408203125,
|
| 3203 |
+
"learning_rate": 9.992164145208792e-06,
|
| 3204 |
+
"loss": 12.3398,
|
| 3205 |
+
"step": 4520
|
| 3206 |
+
},
|
| 3207 |
+
{
|
| 3208 |
+
"epoch": 0.20104182269043222,
|
| 3209 |
+
"grad_norm": 125.20413970947266,
|
| 3210 |
+
"learning_rate": 9.992146809246865e-06,
|
| 3211 |
+
"loss": 12.8135,
|
| 3212 |
+
"step": 4530
|
| 3213 |
+
},
|
| 3214 |
+
{
|
| 3215 |
+
"epoch": 0.20148562362352368,
|
| 3216 |
+
"grad_norm": 104.17672729492188,
|
| 3217 |
+
"learning_rate": 9.992129473284938e-06,
|
| 3218 |
+
"loss": 12.1691,
|
| 3219 |
+
"step": 4540
|
| 3220 |
+
},
|
| 3221 |
+
{
|
| 3222 |
+
"epoch": 0.20192942455661514,
|
| 3223 |
+
"grad_norm": 145.2282257080078,
|
| 3224 |
+
"learning_rate": 9.992112137323009e-06,
|
| 3225 |
+
"loss": 12.2538,
|
| 3226 |
+
"step": 4550
|
| 3227 |
+
},
|
| 3228 |
+
{
|
| 3229 |
+
"epoch": 0.2023732254897066,
|
| 3230 |
+
"grad_norm": 130.9412078857422,
|
| 3231 |
+
"learning_rate": 9.992094801361082e-06,
|
| 3232 |
+
"loss": 12.1611,
|
| 3233 |
+
"step": 4560
|
| 3234 |
+
},
|
| 3235 |
+
{
|
| 3236 |
+
"epoch": 0.20281702642279806,
|
| 3237 |
+
"grad_norm": 104.62641143798828,
|
| 3238 |
+
"learning_rate": 9.992077465399155e-06,
|
| 3239 |
+
"loss": 12.3785,
|
| 3240 |
+
"step": 4570
|
| 3241 |
+
},
|
| 3242 |
+
{
|
| 3243 |
+
"epoch": 0.20326082735588952,
|
| 3244 |
+
"grad_norm": 106.15484619140625,
|
| 3245 |
+
"learning_rate": 9.992060129437227e-06,
|
| 3246 |
+
"loss": 12.139,
|
| 3247 |
+
"step": 4580
|
| 3248 |
+
},
|
| 3249 |
+
{
|
| 3250 |
+
"epoch": 0.20370462828898098,
|
| 3251 |
+
"grad_norm": 135.23989868164062,
|
| 3252 |
+
"learning_rate": 9.9920427934753e-06,
|
| 3253 |
+
"loss": 11.6559,
|
| 3254 |
+
"step": 4590
|
| 3255 |
+
},
|
| 3256 |
+
{
|
| 3257 |
+
"epoch": 0.20414842922207244,
|
| 3258 |
+
"grad_norm": 129.04571533203125,
|
| 3259 |
+
"learning_rate": 9.992025457513373e-06,
|
| 3260 |
+
"loss": 12.2296,
|
| 3261 |
+
"step": 4600
|
| 3262 |
+
},
|
| 3263 |
+
{
|
| 3264 |
+
"epoch": 0.2045922301551639,
|
| 3265 |
+
"grad_norm": 100.73985290527344,
|
| 3266 |
+
"learning_rate": 9.992008121551444e-06,
|
| 3267 |
+
"loss": 12.429,
|
| 3268 |
+
"step": 4610
|
| 3269 |
+
},
|
| 3270 |
+
{
|
| 3271 |
+
"epoch": 0.20503603108825535,
|
| 3272 |
+
"grad_norm": 120.50385284423828,
|
| 3273 |
+
"learning_rate": 9.991990785589517e-06,
|
| 3274 |
+
"loss": 12.6906,
|
| 3275 |
+
"step": 4620
|
| 3276 |
+
},
|
| 3277 |
+
{
|
| 3278 |
+
"epoch": 0.20547983202134681,
|
| 3279 |
+
"grad_norm": 119.28755187988281,
|
| 3280 |
+
"learning_rate": 9.99197344962759e-06,
|
| 3281 |
+
"loss": 12.0725,
|
| 3282 |
+
"step": 4630
|
| 3283 |
+
},
|
| 3284 |
+
{
|
| 3285 |
+
"epoch": 0.20592363295443827,
|
| 3286 |
+
"grad_norm": 122.34222412109375,
|
| 3287 |
+
"learning_rate": 9.991956113665662e-06,
|
| 3288 |
+
"loss": 12.5552,
|
| 3289 |
+
"step": 4640
|
| 3290 |
+
},
|
| 3291 |
+
{
|
| 3292 |
+
"epoch": 0.20636743388752976,
|
| 3293 |
+
"grad_norm": 123.09967041015625,
|
| 3294 |
+
"learning_rate": 9.991938777703735e-06,
|
| 3295 |
+
"loss": 11.6526,
|
| 3296 |
+
"step": 4650
|
| 3297 |
+
},
|
| 3298 |
+
{
|
| 3299 |
+
"epoch": 0.20681123482062122,
|
| 3300 |
+
"grad_norm": 97.26423645019531,
|
| 3301 |
+
"learning_rate": 9.991921441741808e-06,
|
| 3302 |
+
"loss": 12.1968,
|
| 3303 |
+
"step": 4660
|
| 3304 |
+
},
|
| 3305 |
+
{
|
| 3306 |
+
"epoch": 0.20725503575371268,
|
| 3307 |
+
"grad_norm": 114.7996826171875,
|
| 3308 |
+
"learning_rate": 9.99190410577988e-06,
|
| 3309 |
+
"loss": 12.6007,
|
| 3310 |
+
"step": 4670
|
| 3311 |
+
},
|
| 3312 |
+
{
|
| 3313 |
+
"epoch": 0.20769883668680414,
|
| 3314 |
+
"grad_norm": 140.515869140625,
|
| 3315 |
+
"learning_rate": 9.991886769817952e-06,
|
| 3316 |
+
"loss": 12.2736,
|
| 3317 |
+
"step": 4680
|
| 3318 |
+
},
|
| 3319 |
+
{
|
| 3320 |
+
"epoch": 0.2081426376198956,
|
| 3321 |
+
"grad_norm": 125.37648010253906,
|
| 3322 |
+
"learning_rate": 9.991869433856025e-06,
|
| 3323 |
+
"loss": 12.1014,
|
| 3324 |
+
"step": 4690
|
| 3325 |
+
},
|
| 3326 |
+
{
|
| 3327 |
+
"epoch": 0.20858643855298706,
|
| 3328 |
+
"grad_norm": 128.2080078125,
|
| 3329 |
+
"learning_rate": 9.991852097894099e-06,
|
| 3330 |
+
"loss": 11.7794,
|
| 3331 |
+
"step": 4700
|
| 3332 |
+
},
|
| 3333 |
+
{
|
| 3334 |
+
"epoch": 0.20903023948607852,
|
| 3335 |
+
"grad_norm": 115.89141082763672,
|
| 3336 |
+
"learning_rate": 9.99183476193217e-06,
|
| 3337 |
+
"loss": 12.2364,
|
| 3338 |
+
"step": 4710
|
| 3339 |
+
},
|
| 3340 |
+
{
|
| 3341 |
+
"epoch": 0.20947404041916998,
|
| 3342 |
+
"grad_norm": 118.23058319091797,
|
| 3343 |
+
"learning_rate": 9.991817425970243e-06,
|
| 3344 |
+
"loss": 11.7807,
|
| 3345 |
+
"step": 4720
|
| 3346 |
+
},
|
| 3347 |
+
{
|
| 3348 |
+
"epoch": 0.20991784135226144,
|
| 3349 |
+
"grad_norm": 120.32998657226562,
|
| 3350 |
+
"learning_rate": 9.991800090008316e-06,
|
| 3351 |
+
"loss": 12.3347,
|
| 3352 |
+
"step": 4730
|
| 3353 |
+
},
|
| 3354 |
+
{
|
| 3355 |
+
"epoch": 0.2103616422853529,
|
| 3356 |
+
"grad_norm": 131.60520935058594,
|
| 3357 |
+
"learning_rate": 9.991782754046387e-06,
|
| 3358 |
+
"loss": 12.5857,
|
| 3359 |
+
"step": 4740
|
| 3360 |
+
},
|
| 3361 |
+
{
|
| 3362 |
+
"epoch": 0.21080544321844435,
|
| 3363 |
+
"grad_norm": 123.4472885131836,
|
| 3364 |
+
"learning_rate": 9.99176541808446e-06,
|
| 3365 |
+
"loss": 12.5721,
|
| 3366 |
+
"step": 4750
|
| 3367 |
+
},
|
| 3368 |
+
{
|
| 3369 |
+
"epoch": 0.21124924415153584,
|
| 3370 |
+
"grad_norm": 115.33013153076172,
|
| 3371 |
+
"learning_rate": 9.991748082122534e-06,
|
| 3372 |
+
"loss": 12.2645,
|
| 3373 |
+
"step": 4760
|
| 3374 |
+
},
|
| 3375 |
+
{
|
| 3376 |
+
"epoch": 0.2116930450846273,
|
| 3377 |
+
"grad_norm": 128.06332397460938,
|
| 3378 |
+
"learning_rate": 9.991730746160605e-06,
|
| 3379 |
+
"loss": 11.9115,
|
| 3380 |
+
"step": 4770
|
| 3381 |
+
},
|
| 3382 |
+
{
|
| 3383 |
+
"epoch": 0.21213684601771876,
|
| 3384 |
+
"grad_norm": 133.6379852294922,
|
| 3385 |
+
"learning_rate": 9.991713410198678e-06,
|
| 3386 |
+
"loss": 12.2299,
|
| 3387 |
+
"step": 4780
|
| 3388 |
+
},
|
| 3389 |
+
{
|
| 3390 |
+
"epoch": 0.21258064695081022,
|
| 3391 |
+
"grad_norm": 99.40995025634766,
|
| 3392 |
+
"learning_rate": 9.991696074236751e-06,
|
| 3393 |
+
"loss": 12.232,
|
| 3394 |
+
"step": 4790
|
| 3395 |
+
},
|
| 3396 |
+
{
|
| 3397 |
+
"epoch": 0.21302444788390168,
|
| 3398 |
+
"grad_norm": 128.07489013671875,
|
| 3399 |
+
"learning_rate": 9.991678738274823e-06,
|
| 3400 |
+
"loss": 12.5043,
|
| 3401 |
+
"step": 4800
|
| 3402 |
+
},
|
| 3403 |
+
{
|
| 3404 |
+
"epoch": 0.21346824881699314,
|
| 3405 |
+
"grad_norm": 106.18391418457031,
|
| 3406 |
+
"learning_rate": 9.991661402312896e-06,
|
| 3407 |
+
"loss": 12.515,
|
| 3408 |
+
"step": 4810
|
| 3409 |
+
},
|
| 3410 |
+
{
|
| 3411 |
+
"epoch": 0.2139120497500846,
|
| 3412 |
+
"grad_norm": 103.52220153808594,
|
| 3413 |
+
"learning_rate": 9.991644066350969e-06,
|
| 3414 |
+
"loss": 12.5614,
|
| 3415 |
+
"step": 4820
|
| 3416 |
+
},
|
| 3417 |
+
{
|
| 3418 |
+
"epoch": 0.21435585068317606,
|
| 3419 |
+
"grad_norm": 144.36888122558594,
|
| 3420 |
+
"learning_rate": 9.99162673038904e-06,
|
| 3421 |
+
"loss": 12.0411,
|
| 3422 |
+
"step": 4830
|
| 3423 |
+
},
|
| 3424 |
+
{
|
| 3425 |
+
"epoch": 0.21479965161626752,
|
| 3426 |
+
"grad_norm": 119.41954803466797,
|
| 3427 |
+
"learning_rate": 9.991609394427113e-06,
|
| 3428 |
+
"loss": 12.0378,
|
| 3429 |
+
"step": 4840
|
| 3430 |
+
},
|
| 3431 |
+
{
|
| 3432 |
+
"epoch": 0.21524345254935898,
|
| 3433 |
+
"grad_norm": 97.38760375976562,
|
| 3434 |
+
"learning_rate": 9.991592058465186e-06,
|
| 3435 |
+
"loss": 12.0504,
|
| 3436 |
+
"step": 4850
|
| 3437 |
+
},
|
| 3438 |
+
{
|
| 3439 |
+
"epoch": 0.21568725348245044,
|
| 3440 |
+
"grad_norm": 123.15618896484375,
|
| 3441 |
+
"learning_rate": 9.991574722503258e-06,
|
| 3442 |
+
"loss": 12.4191,
|
| 3443 |
+
"step": 4860
|
| 3444 |
+
},
|
| 3445 |
+
{
|
| 3446 |
+
"epoch": 0.2161310544155419,
|
| 3447 |
+
"grad_norm": 121.77255249023438,
|
| 3448 |
+
"learning_rate": 9.99155738654133e-06,
|
| 3449 |
+
"loss": 11.9067,
|
| 3450 |
+
"step": 4870
|
| 3451 |
+
},
|
| 3452 |
+
{
|
| 3453 |
+
"epoch": 0.21657485534863338,
|
| 3454 |
+
"grad_norm": 128.0386962890625,
|
| 3455 |
+
"learning_rate": 9.991540050579404e-06,
|
| 3456 |
+
"loss": 12.9843,
|
| 3457 |
+
"step": 4880
|
| 3458 |
+
},
|
| 3459 |
+
{
|
| 3460 |
+
"epoch": 0.21701865628172484,
|
| 3461 |
+
"grad_norm": 102.69422149658203,
|
| 3462 |
+
"learning_rate": 9.991522714617475e-06,
|
| 3463 |
+
"loss": 12.2082,
|
| 3464 |
+
"step": 4890
|
| 3465 |
+
},
|
| 3466 |
+
{
|
| 3467 |
+
"epoch": 0.2174624572148163,
|
| 3468 |
+
"grad_norm": 125.6636734008789,
|
| 3469 |
+
"learning_rate": 9.991505378655548e-06,
|
| 3470 |
+
"loss": 12.5348,
|
| 3471 |
+
"step": 4900
|
| 3472 |
+
},
|
| 3473 |
+
{
|
| 3474 |
+
"epoch": 0.21790625814790776,
|
| 3475 |
+
"grad_norm": 135.658203125,
|
| 3476 |
+
"learning_rate": 9.991488042693621e-06,
|
| 3477 |
+
"loss": 11.9152,
|
| 3478 |
+
"step": 4910
|
| 3479 |
+
},
|
| 3480 |
+
{
|
| 3481 |
+
"epoch": 0.21835005908099922,
|
| 3482 |
+
"grad_norm": 99.29556274414062,
|
| 3483 |
+
"learning_rate": 9.991470706731694e-06,
|
| 3484 |
+
"loss": 12.5844,
|
| 3485 |
+
"step": 4920
|
| 3486 |
+
},
|
| 3487 |
+
{
|
| 3488 |
+
"epoch": 0.21879386001409068,
|
| 3489 |
+
"grad_norm": 95.45423126220703,
|
| 3490 |
+
"learning_rate": 9.991453370769766e-06,
|
| 3491 |
+
"loss": 12.926,
|
| 3492 |
+
"step": 4930
|
| 3493 |
+
},
|
| 3494 |
+
{
|
| 3495 |
+
"epoch": 0.21923766094718214,
|
| 3496 |
+
"grad_norm": 118.17921447753906,
|
| 3497 |
+
"learning_rate": 9.991436034807839e-06,
|
| 3498 |
+
"loss": 12.2757,
|
| 3499 |
+
"step": 4940
|
| 3500 |
+
},
|
| 3501 |
+
{
|
| 3502 |
+
"epoch": 0.2196814618802736,
|
| 3503 |
+
"grad_norm": 120.35387420654297,
|
| 3504 |
+
"learning_rate": 9.991418698845912e-06,
|
| 3505 |
+
"loss": 12.4541,
|
| 3506 |
+
"step": 4950
|
| 3507 |
+
},
|
| 3508 |
+
{
|
| 3509 |
+
"epoch": 0.22012526281336506,
|
| 3510 |
+
"grad_norm": 109.4905776977539,
|
| 3511 |
+
"learning_rate": 9.991401362883983e-06,
|
| 3512 |
+
"loss": 11.7813,
|
| 3513 |
+
"step": 4960
|
| 3514 |
+
},
|
| 3515 |
+
{
|
| 3516 |
+
"epoch": 0.22056906374645652,
|
| 3517 |
+
"grad_norm": 102.00797271728516,
|
| 3518 |
+
"learning_rate": 9.991384026922056e-06,
|
| 3519 |
+
"loss": 11.9456,
|
| 3520 |
+
"step": 4970
|
| 3521 |
+
},
|
| 3522 |
+
{
|
| 3523 |
+
"epoch": 0.22101286467954798,
|
| 3524 |
+
"grad_norm": 104.43452453613281,
|
| 3525 |
+
"learning_rate": 9.99136669096013e-06,
|
| 3526 |
+
"loss": 11.8744,
|
| 3527 |
+
"step": 4980
|
| 3528 |
+
},
|
| 3529 |
+
{
|
| 3530 |
+
"epoch": 0.22145666561263946,
|
| 3531 |
+
"grad_norm": 91.9101791381836,
|
| 3532 |
+
"learning_rate": 9.991349354998201e-06,
|
| 3533 |
+
"loss": 12.1663,
|
| 3534 |
+
"step": 4990
|
| 3535 |
+
},
|
| 3536 |
+
{
|
| 3537 |
+
"epoch": 0.22190046654573092,
|
| 3538 |
+
"grad_norm": 139.61109924316406,
|
| 3539 |
+
"learning_rate": 9.991332019036274e-06,
|
| 3540 |
+
"loss": 12.5729,
|
| 3541 |
+
"step": 5000
|
| 3542 |
+
},
|
| 3543 |
+
{
|
| 3544 |
+
"epoch": 0.22190046654573092,
|
| 3545 |
+
"eval_loss": 0.37864938378334045,
|
| 3546 |
+
"eval_runtime": 675.391,
|
| 3547 |
+
"eval_samples_per_second": 1798.056,
|
| 3548 |
+
"eval_steps_per_second": 56.19,
|
| 3549 |
+
"step": 5000
|
| 3550 |
}
|
| 3551 |
],
|
| 3552 |
"logging_steps": 10,
|
|
|
|
| 3566 |
"attributes": {}
|
| 3567 |
}
|
| 3568 |
},
|
| 3569 |
+
"total_flos": 1.744861602512896e+18,
|
| 3570 |
"train_batch_size": 4,
|
| 3571 |
"trial_name": null,
|
| 3572 |
"trial_params": null
|