Commit
•
a22c09b
1
Parent(s):
d647c4a
Training in progress, step 9288, checkpoint
Browse files
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 439648328
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0015c35a2eddaec815db15463fdc85043d66de0a9c2962eda74d561b81414b6f
|
| 3 |
size 439648328
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 879415866
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eabcb53a0ad24de4121adf17313188b0c1f3343621a0e7c0c20bdb33797bd83f
|
| 3 |
size 879415866
|
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 14512
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fc1f62733ff0d8690c2a8797d153e8ec399303cbc92322dd118c7b88915a04be
|
| 3 |
size 14512
|
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 14512
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:da252e7b08747075dc2ad5c781c5deb49e64dd87681bbc7b4aaa89a30b6f9d5f
|
| 3 |
size 14512
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7a78424df3b6c3f885135b5bf0deeb48c09ab08dbdc18e3e5b017f82ee6747c9
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -14455,6 +14455,1812 @@
|
|
| 14455 |
"learning_rate": 9.98157843707582e-06,
|
| 14456 |
"loss": 5.2778,
|
| 14457 |
"step": 8256
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 14458 |
}
|
| 14459 |
],
|
| 14460 |
"logging_steps": 4,
|
|
@@ -14462,7 +16268,7 @@
|
|
| 14462 |
"num_input_tokens_seen": 0,
|
| 14463 |
"num_train_epochs": 1,
|
| 14464 |
"save_steps": 1032,
|
| 14465 |
-
"total_flos":
|
| 14466 |
"train_batch_size": 8,
|
| 14467 |
"trial_name": null,
|
| 14468 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.900523560209424,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 9288,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 14455 |
"learning_rate": 9.98157843707582e-06,
|
| 14456 |
"loss": 5.2778,
|
| 14457 |
"step": 8256
|
| 14458 |
+
},
|
| 14459 |
+
{
|
| 14460 |
+
"epoch": 0.8,
|
| 14461 |
+
"grad_norm": 1.036868691444397,
|
| 14462 |
+
"learning_rate": 9.962187318208263e-06,
|
| 14463 |
+
"loss": 5.31,
|
| 14464 |
+
"step": 8260
|
| 14465 |
+
},
|
| 14466 |
+
{
|
| 14467 |
+
"epoch": 0.8,
|
| 14468 |
+
"grad_norm": 1.0464235544204712,
|
| 14469 |
+
"learning_rate": 9.942796199340703e-06,
|
| 14470 |
+
"loss": 5.3512,
|
| 14471 |
+
"step": 8264
|
| 14472 |
+
},
|
| 14473 |
+
{
|
| 14474 |
+
"epoch": 0.8,
|
| 14475 |
+
"grad_norm": 1.001470685005188,
|
| 14476 |
+
"learning_rate": 9.923405080473144e-06,
|
| 14477 |
+
"loss": 5.2698,
|
| 14478 |
+
"step": 8268
|
| 14479 |
+
},
|
| 14480 |
+
{
|
| 14481 |
+
"epoch": 0.8,
|
| 14482 |
+
"grad_norm": 1.0910736322402954,
|
| 14483 |
+
"learning_rate": 9.904013961605584e-06,
|
| 14484 |
+
"loss": 5.363,
|
| 14485 |
+
"step": 8272
|
| 14486 |
+
},
|
| 14487 |
+
{
|
| 14488 |
+
"epoch": 0.8,
|
| 14489 |
+
"grad_norm": 1.087928056716919,
|
| 14490 |
+
"learning_rate": 9.884622842738026e-06,
|
| 14491 |
+
"loss": 5.3454,
|
| 14492 |
+
"step": 8276
|
| 14493 |
+
},
|
| 14494 |
+
{
|
| 14495 |
+
"epoch": 0.8,
|
| 14496 |
+
"grad_norm": 1.055014967918396,
|
| 14497 |
+
"learning_rate": 9.865231723870467e-06,
|
| 14498 |
+
"loss": 5.3134,
|
| 14499 |
+
"step": 8280
|
| 14500 |
+
},
|
| 14501 |
+
{
|
| 14502 |
+
"epoch": 0.8,
|
| 14503 |
+
"grad_norm": 1.1186180114746094,
|
| 14504 |
+
"learning_rate": 9.84584060500291e-06,
|
| 14505 |
+
"loss": 5.3016,
|
| 14506 |
+
"step": 8284
|
| 14507 |
+
},
|
| 14508 |
+
{
|
| 14509 |
+
"epoch": 0.8,
|
| 14510 |
+
"grad_norm": 1.0159074068069458,
|
| 14511 |
+
"learning_rate": 9.82644948613535e-06,
|
| 14512 |
+
"loss": 5.3974,
|
| 14513 |
+
"step": 8288
|
| 14514 |
+
},
|
| 14515 |
+
{
|
| 14516 |
+
"epoch": 0.8,
|
| 14517 |
+
"grad_norm": 1.1419733762741089,
|
| 14518 |
+
"learning_rate": 9.807058367267792e-06,
|
| 14519 |
+
"loss": 5.333,
|
| 14520 |
+
"step": 8292
|
| 14521 |
+
},
|
| 14522 |
+
{
|
| 14523 |
+
"epoch": 0.8,
|
| 14524 |
+
"grad_norm": 1.078598976135254,
|
| 14525 |
+
"learning_rate": 9.787667248400232e-06,
|
| 14526 |
+
"loss": 5.2157,
|
| 14527 |
+
"step": 8296
|
| 14528 |
+
},
|
| 14529 |
+
{
|
| 14530 |
+
"epoch": 0.8,
|
| 14531 |
+
"grad_norm": 1.0978525876998901,
|
| 14532 |
+
"learning_rate": 9.768276129532675e-06,
|
| 14533 |
+
"loss": 5.3084,
|
| 14534 |
+
"step": 8300
|
| 14535 |
+
},
|
| 14536 |
+
{
|
| 14537 |
+
"epoch": 0.81,
|
| 14538 |
+
"grad_norm": 1.0307817459106445,
|
| 14539 |
+
"learning_rate": 9.748885010665115e-06,
|
| 14540 |
+
"loss": 5.2962,
|
| 14541 |
+
"step": 8304
|
| 14542 |
+
},
|
| 14543 |
+
{
|
| 14544 |
+
"epoch": 0.81,
|
| 14545 |
+
"grad_norm": 0.9767160415649414,
|
| 14546 |
+
"learning_rate": 9.729493891797557e-06,
|
| 14547 |
+
"loss": 5.3765,
|
| 14548 |
+
"step": 8308
|
| 14549 |
+
},
|
| 14550 |
+
{
|
| 14551 |
+
"epoch": 0.81,
|
| 14552 |
+
"grad_norm": 1.048744797706604,
|
| 14553 |
+
"learning_rate": 9.710102772929998e-06,
|
| 14554 |
+
"loss": 5.3491,
|
| 14555 |
+
"step": 8312
|
| 14556 |
+
},
|
| 14557 |
+
{
|
| 14558 |
+
"epoch": 0.81,
|
| 14559 |
+
"grad_norm": 1.1452877521514893,
|
| 14560 |
+
"learning_rate": 9.69071165406244e-06,
|
| 14561 |
+
"loss": 5.3348,
|
| 14562 |
+
"step": 8316
|
| 14563 |
+
},
|
| 14564 |
+
{
|
| 14565 |
+
"epoch": 0.81,
|
| 14566 |
+
"grad_norm": 1.0742149353027344,
|
| 14567 |
+
"learning_rate": 9.671320535194882e-06,
|
| 14568 |
+
"loss": 5.4119,
|
| 14569 |
+
"step": 8320
|
| 14570 |
+
},
|
| 14571 |
+
{
|
| 14572 |
+
"epoch": 0.81,
|
| 14573 |
+
"grad_norm": 1.016554594039917,
|
| 14574 |
+
"learning_rate": 9.651929416327323e-06,
|
| 14575 |
+
"loss": 5.2832,
|
| 14576 |
+
"step": 8324
|
| 14577 |
+
},
|
| 14578 |
+
{
|
| 14579 |
+
"epoch": 0.81,
|
| 14580 |
+
"grad_norm": 1.1286094188690186,
|
| 14581 |
+
"learning_rate": 9.632538297459765e-06,
|
| 14582 |
+
"loss": 5.3436,
|
| 14583 |
+
"step": 8328
|
| 14584 |
+
},
|
| 14585 |
+
{
|
| 14586 |
+
"epoch": 0.81,
|
| 14587 |
+
"grad_norm": 1.0719375610351562,
|
| 14588 |
+
"learning_rate": 9.613147178592206e-06,
|
| 14589 |
+
"loss": 5.2561,
|
| 14590 |
+
"step": 8332
|
| 14591 |
+
},
|
| 14592 |
+
{
|
| 14593 |
+
"epoch": 0.81,
|
| 14594 |
+
"grad_norm": 1.0239946842193604,
|
| 14595 |
+
"learning_rate": 9.593756059724646e-06,
|
| 14596 |
+
"loss": 5.3411,
|
| 14597 |
+
"step": 8336
|
| 14598 |
+
},
|
| 14599 |
+
{
|
| 14600 |
+
"epoch": 0.81,
|
| 14601 |
+
"grad_norm": 1.16642165184021,
|
| 14602 |
+
"learning_rate": 9.574364940857087e-06,
|
| 14603 |
+
"loss": 5.2173,
|
| 14604 |
+
"step": 8340
|
| 14605 |
+
},
|
| 14606 |
+
{
|
| 14607 |
+
"epoch": 0.81,
|
| 14608 |
+
"grad_norm": 1.056943655014038,
|
| 14609 |
+
"learning_rate": 9.554973821989529e-06,
|
| 14610 |
+
"loss": 5.3138,
|
| 14611 |
+
"step": 8344
|
| 14612 |
+
},
|
| 14613 |
+
{
|
| 14614 |
+
"epoch": 0.81,
|
| 14615 |
+
"grad_norm": 1.0310717821121216,
|
| 14616 |
+
"learning_rate": 9.53558270312197e-06,
|
| 14617 |
+
"loss": 5.2147,
|
| 14618 |
+
"step": 8348
|
| 14619 |
+
},
|
| 14620 |
+
{
|
| 14621 |
+
"epoch": 0.81,
|
| 14622 |
+
"grad_norm": 1.0939549207687378,
|
| 14623 |
+
"learning_rate": 9.516191584254412e-06,
|
| 14624 |
+
"loss": 5.3849,
|
| 14625 |
+
"step": 8352
|
| 14626 |
+
},
|
| 14627 |
+
{
|
| 14628 |
+
"epoch": 0.81,
|
| 14629 |
+
"grad_norm": 1.0846009254455566,
|
| 14630 |
+
"learning_rate": 9.496800465386852e-06,
|
| 14631 |
+
"loss": 5.2793,
|
| 14632 |
+
"step": 8356
|
| 14633 |
+
},
|
| 14634 |
+
{
|
| 14635 |
+
"epoch": 0.81,
|
| 14636 |
+
"grad_norm": 1.0984148979187012,
|
| 14637 |
+
"learning_rate": 9.477409346519295e-06,
|
| 14638 |
+
"loss": 5.3142,
|
| 14639 |
+
"step": 8360
|
| 14640 |
+
},
|
| 14641 |
+
{
|
| 14642 |
+
"epoch": 0.81,
|
| 14643 |
+
"grad_norm": 1.035758376121521,
|
| 14644 |
+
"learning_rate": 9.458018227651735e-06,
|
| 14645 |
+
"loss": 5.264,
|
| 14646 |
+
"step": 8364
|
| 14647 |
+
},
|
| 14648 |
+
{
|
| 14649 |
+
"epoch": 0.81,
|
| 14650 |
+
"grad_norm": 1.0837132930755615,
|
| 14651 |
+
"learning_rate": 9.438627108784177e-06,
|
| 14652 |
+
"loss": 5.432,
|
| 14653 |
+
"step": 8368
|
| 14654 |
+
},
|
| 14655 |
+
{
|
| 14656 |
+
"epoch": 0.81,
|
| 14657 |
+
"grad_norm": 1.0333995819091797,
|
| 14658 |
+
"learning_rate": 9.419235989916618e-06,
|
| 14659 |
+
"loss": 5.2622,
|
| 14660 |
+
"step": 8372
|
| 14661 |
+
},
|
| 14662 |
+
{
|
| 14663 |
+
"epoch": 0.81,
|
| 14664 |
+
"grad_norm": 1.054474949836731,
|
| 14665 |
+
"learning_rate": 9.39984487104906e-06,
|
| 14666 |
+
"loss": 5.2721,
|
| 14667 |
+
"step": 8376
|
| 14668 |
+
},
|
| 14669 |
+
{
|
| 14670 |
+
"epoch": 0.81,
|
| 14671 |
+
"grad_norm": 1.0750809907913208,
|
| 14672 |
+
"learning_rate": 9.380453752181502e-06,
|
| 14673 |
+
"loss": 5.2433,
|
| 14674 |
+
"step": 8380
|
| 14675 |
+
},
|
| 14676 |
+
{
|
| 14677 |
+
"epoch": 0.81,
|
| 14678 |
+
"grad_norm": 1.054914116859436,
|
| 14679 |
+
"learning_rate": 9.361062633313943e-06,
|
| 14680 |
+
"loss": 5.3407,
|
| 14681 |
+
"step": 8384
|
| 14682 |
+
},
|
| 14683 |
+
{
|
| 14684 |
+
"epoch": 0.81,
|
| 14685 |
+
"grad_norm": 1.1101247072219849,
|
| 14686 |
+
"learning_rate": 9.341671514446385e-06,
|
| 14687 |
+
"loss": 5.2774,
|
| 14688 |
+
"step": 8388
|
| 14689 |
+
},
|
| 14690 |
+
{
|
| 14691 |
+
"epoch": 0.81,
|
| 14692 |
+
"grad_norm": 1.1054069995880127,
|
| 14693 |
+
"learning_rate": 9.322280395578826e-06,
|
| 14694 |
+
"loss": 5.3978,
|
| 14695 |
+
"step": 8392
|
| 14696 |
+
},
|
| 14697 |
+
{
|
| 14698 |
+
"epoch": 0.81,
|
| 14699 |
+
"grad_norm": 1.0813637971878052,
|
| 14700 |
+
"learning_rate": 9.302889276711268e-06,
|
| 14701 |
+
"loss": 5.3735,
|
| 14702 |
+
"step": 8396
|
| 14703 |
+
},
|
| 14704 |
+
{
|
| 14705 |
+
"epoch": 0.81,
|
| 14706 |
+
"grad_norm": 1.151734709739685,
|
| 14707 |
+
"learning_rate": 9.283498157843708e-06,
|
| 14708 |
+
"loss": 5.4109,
|
| 14709 |
+
"step": 8400
|
| 14710 |
+
},
|
| 14711 |
+
{
|
| 14712 |
+
"epoch": 0.81,
|
| 14713 |
+
"grad_norm": 1.0774791240692139,
|
| 14714 |
+
"learning_rate": 9.264107038976149e-06,
|
| 14715 |
+
"loss": 5.2963,
|
| 14716 |
+
"step": 8404
|
| 14717 |
+
},
|
| 14718 |
+
{
|
| 14719 |
+
"epoch": 0.82,
|
| 14720 |
+
"grad_norm": 1.0468578338623047,
|
| 14721 |
+
"learning_rate": 9.24471592010859e-06,
|
| 14722 |
+
"loss": 5.334,
|
| 14723 |
+
"step": 8408
|
| 14724 |
+
},
|
| 14725 |
+
{
|
| 14726 |
+
"epoch": 0.82,
|
| 14727 |
+
"grad_norm": 1.0247350931167603,
|
| 14728 |
+
"learning_rate": 9.225324801241032e-06,
|
| 14729 |
+
"loss": 5.3307,
|
| 14730 |
+
"step": 8412
|
| 14731 |
+
},
|
| 14732 |
+
{
|
| 14733 |
+
"epoch": 0.82,
|
| 14734 |
+
"grad_norm": 1.021700143814087,
|
| 14735 |
+
"learning_rate": 9.205933682373472e-06,
|
| 14736 |
+
"loss": 5.3315,
|
| 14737 |
+
"step": 8416
|
| 14738 |
+
},
|
| 14739 |
+
{
|
| 14740 |
+
"epoch": 0.82,
|
| 14741 |
+
"grad_norm": 1.0852890014648438,
|
| 14742 |
+
"learning_rate": 9.186542563505915e-06,
|
| 14743 |
+
"loss": 5.23,
|
| 14744 |
+
"step": 8420
|
| 14745 |
+
},
|
| 14746 |
+
{
|
| 14747 |
+
"epoch": 0.82,
|
| 14748 |
+
"grad_norm": 1.0435699224472046,
|
| 14749 |
+
"learning_rate": 9.167151444638355e-06,
|
| 14750 |
+
"loss": 5.3353,
|
| 14751 |
+
"step": 8424
|
| 14752 |
+
},
|
| 14753 |
+
{
|
| 14754 |
+
"epoch": 0.82,
|
| 14755 |
+
"grad_norm": 1.0124831199645996,
|
| 14756 |
+
"learning_rate": 9.147760325770797e-06,
|
| 14757 |
+
"loss": 5.3257,
|
| 14758 |
+
"step": 8428
|
| 14759 |
+
},
|
| 14760 |
+
{
|
| 14761 |
+
"epoch": 0.82,
|
| 14762 |
+
"grad_norm": 1.0061957836151123,
|
| 14763 |
+
"learning_rate": 9.128369206903238e-06,
|
| 14764 |
+
"loss": 5.3437,
|
| 14765 |
+
"step": 8432
|
| 14766 |
+
},
|
| 14767 |
+
{
|
| 14768 |
+
"epoch": 0.82,
|
| 14769 |
+
"grad_norm": 1.0675718784332275,
|
| 14770 |
+
"learning_rate": 9.10897808803568e-06,
|
| 14771 |
+
"loss": 5.3543,
|
| 14772 |
+
"step": 8436
|
| 14773 |
+
},
|
| 14774 |
+
{
|
| 14775 |
+
"epoch": 0.82,
|
| 14776 |
+
"grad_norm": 1.0940582752227783,
|
| 14777 |
+
"learning_rate": 9.089586969168122e-06,
|
| 14778 |
+
"loss": 5.3205,
|
| 14779 |
+
"step": 8440
|
| 14780 |
+
},
|
| 14781 |
+
{
|
| 14782 |
+
"epoch": 0.82,
|
| 14783 |
+
"grad_norm": 1.0253922939300537,
|
| 14784 |
+
"learning_rate": 9.070195850300563e-06,
|
| 14785 |
+
"loss": 5.2836,
|
| 14786 |
+
"step": 8444
|
| 14787 |
+
},
|
| 14788 |
+
{
|
| 14789 |
+
"epoch": 0.82,
|
| 14790 |
+
"grad_norm": 1.0632801055908203,
|
| 14791 |
+
"learning_rate": 9.050804731433005e-06,
|
| 14792 |
+
"loss": 5.3634,
|
| 14793 |
+
"step": 8448
|
| 14794 |
+
},
|
| 14795 |
+
{
|
| 14796 |
+
"epoch": 0.82,
|
| 14797 |
+
"grad_norm": 1.151405692100525,
|
| 14798 |
+
"learning_rate": 9.031413612565446e-06,
|
| 14799 |
+
"loss": 5.266,
|
| 14800 |
+
"step": 8452
|
| 14801 |
+
},
|
| 14802 |
+
{
|
| 14803 |
+
"epoch": 0.82,
|
| 14804 |
+
"grad_norm": 1.1215803623199463,
|
| 14805 |
+
"learning_rate": 9.012022493697888e-06,
|
| 14806 |
+
"loss": 5.3605,
|
| 14807 |
+
"step": 8456
|
| 14808 |
+
},
|
| 14809 |
+
{
|
| 14810 |
+
"epoch": 0.82,
|
| 14811 |
+
"grad_norm": 1.0406687259674072,
|
| 14812 |
+
"learning_rate": 8.992631374830328e-06,
|
| 14813 |
+
"loss": 5.3037,
|
| 14814 |
+
"step": 8460
|
| 14815 |
+
},
|
| 14816 |
+
{
|
| 14817 |
+
"epoch": 0.82,
|
| 14818 |
+
"grad_norm": 1.0305143594741821,
|
| 14819 |
+
"learning_rate": 8.97324025596277e-06,
|
| 14820 |
+
"loss": 5.2419,
|
| 14821 |
+
"step": 8464
|
| 14822 |
+
},
|
| 14823 |
+
{
|
| 14824 |
+
"epoch": 0.82,
|
| 14825 |
+
"grad_norm": 1.0159248113632202,
|
| 14826 |
+
"learning_rate": 8.953849137095211e-06,
|
| 14827 |
+
"loss": 5.2576,
|
| 14828 |
+
"step": 8468
|
| 14829 |
+
},
|
| 14830 |
+
{
|
| 14831 |
+
"epoch": 0.82,
|
| 14832 |
+
"grad_norm": 1.0444971323013306,
|
| 14833 |
+
"learning_rate": 8.934458018227652e-06,
|
| 14834 |
+
"loss": 5.2461,
|
| 14835 |
+
"step": 8472
|
| 14836 |
+
},
|
| 14837 |
+
{
|
| 14838 |
+
"epoch": 0.82,
|
| 14839 |
+
"grad_norm": 1.062738299369812,
|
| 14840 |
+
"learning_rate": 8.915066899360092e-06,
|
| 14841 |
+
"loss": 5.2934,
|
| 14842 |
+
"step": 8476
|
| 14843 |
+
},
|
| 14844 |
+
{
|
| 14845 |
+
"epoch": 0.82,
|
| 14846 |
+
"grad_norm": 1.0445396900177002,
|
| 14847 |
+
"learning_rate": 8.895675780492534e-06,
|
| 14848 |
+
"loss": 5.3011,
|
| 14849 |
+
"step": 8480
|
| 14850 |
+
},
|
| 14851 |
+
{
|
| 14852 |
+
"epoch": 0.82,
|
| 14853 |
+
"grad_norm": 1.1299471855163574,
|
| 14854 |
+
"learning_rate": 8.876284661624975e-06,
|
| 14855 |
+
"loss": 5.1977,
|
| 14856 |
+
"step": 8484
|
| 14857 |
+
},
|
| 14858 |
+
{
|
| 14859 |
+
"epoch": 0.82,
|
| 14860 |
+
"grad_norm": 1.0241918563842773,
|
| 14861 |
+
"learning_rate": 8.856893542757417e-06,
|
| 14862 |
+
"loss": 5.2821,
|
| 14863 |
+
"step": 8488
|
| 14864 |
+
},
|
| 14865 |
+
{
|
| 14866 |
+
"epoch": 0.82,
|
| 14867 |
+
"grad_norm": 1.0978903770446777,
|
| 14868 |
+
"learning_rate": 8.837502423889858e-06,
|
| 14869 |
+
"loss": 5.3411,
|
| 14870 |
+
"step": 8492
|
| 14871 |
+
},
|
| 14872 |
+
{
|
| 14873 |
+
"epoch": 0.82,
|
| 14874 |
+
"grad_norm": 1.0630090236663818,
|
| 14875 |
+
"learning_rate": 8.8181113050223e-06,
|
| 14876 |
+
"loss": 5.2487,
|
| 14877 |
+
"step": 8496
|
| 14878 |
+
},
|
| 14879 |
+
{
|
| 14880 |
+
"epoch": 0.82,
|
| 14881 |
+
"grad_norm": 1.0468335151672363,
|
| 14882 |
+
"learning_rate": 8.798720186154742e-06,
|
| 14883 |
+
"loss": 5.2238,
|
| 14884 |
+
"step": 8500
|
| 14885 |
+
},
|
| 14886 |
+
{
|
| 14887 |
+
"epoch": 0.82,
|
| 14888 |
+
"grad_norm": 1.0701797008514404,
|
| 14889 |
+
"learning_rate": 8.779329067287183e-06,
|
| 14890 |
+
"loss": 5.2461,
|
| 14891 |
+
"step": 8504
|
| 14892 |
+
},
|
| 14893 |
+
{
|
| 14894 |
+
"epoch": 0.82,
|
| 14895 |
+
"grad_norm": 1.0644505023956299,
|
| 14896 |
+
"learning_rate": 8.759937948419625e-06,
|
| 14897 |
+
"loss": 5.2462,
|
| 14898 |
+
"step": 8508
|
| 14899 |
+
},
|
| 14900 |
+
{
|
| 14901 |
+
"epoch": 0.83,
|
| 14902 |
+
"grad_norm": 1.0857114791870117,
|
| 14903 |
+
"learning_rate": 8.740546829552066e-06,
|
| 14904 |
+
"loss": 5.3715,
|
| 14905 |
+
"step": 8512
|
| 14906 |
+
},
|
| 14907 |
+
{
|
| 14908 |
+
"epoch": 0.83,
|
| 14909 |
+
"grad_norm": 1.0911511182785034,
|
| 14910 |
+
"learning_rate": 8.721155710684508e-06,
|
| 14911 |
+
"loss": 5.3341,
|
| 14912 |
+
"step": 8516
|
| 14913 |
+
},
|
| 14914 |
+
{
|
| 14915 |
+
"epoch": 0.83,
|
| 14916 |
+
"grad_norm": 1.066658854484558,
|
| 14917 |
+
"learning_rate": 8.701764591816948e-06,
|
| 14918 |
+
"loss": 5.2962,
|
| 14919 |
+
"step": 8520
|
| 14920 |
+
},
|
| 14921 |
+
{
|
| 14922 |
+
"epoch": 0.83,
|
| 14923 |
+
"grad_norm": 1.0484204292297363,
|
| 14924 |
+
"learning_rate": 8.68237347294939e-06,
|
| 14925 |
+
"loss": 5.2937,
|
| 14926 |
+
"step": 8524
|
| 14927 |
+
},
|
| 14928 |
+
{
|
| 14929 |
+
"epoch": 0.83,
|
| 14930 |
+
"grad_norm": 1.1011348962783813,
|
| 14931 |
+
"learning_rate": 8.662982354081831e-06,
|
| 14932 |
+
"loss": 5.3372,
|
| 14933 |
+
"step": 8528
|
| 14934 |
+
},
|
| 14935 |
+
{
|
| 14936 |
+
"epoch": 0.83,
|
| 14937 |
+
"grad_norm": 1.1982569694519043,
|
| 14938 |
+
"learning_rate": 8.643591235214273e-06,
|
| 14939 |
+
"loss": 5.1987,
|
| 14940 |
+
"step": 8532
|
| 14941 |
+
},
|
| 14942 |
+
{
|
| 14943 |
+
"epoch": 0.83,
|
| 14944 |
+
"grad_norm": 1.1187928915023804,
|
| 14945 |
+
"learning_rate": 8.624200116346714e-06,
|
| 14946 |
+
"loss": 5.3326,
|
| 14947 |
+
"step": 8536
|
| 14948 |
+
},
|
| 14949 |
+
{
|
| 14950 |
+
"epoch": 0.83,
|
| 14951 |
+
"grad_norm": 1.0191898345947266,
|
| 14952 |
+
"learning_rate": 8.604808997479154e-06,
|
| 14953 |
+
"loss": 5.384,
|
| 14954 |
+
"step": 8540
|
| 14955 |
+
},
|
| 14956 |
+
{
|
| 14957 |
+
"epoch": 0.83,
|
| 14958 |
+
"grad_norm": 1.1834492683410645,
|
| 14959 |
+
"learning_rate": 8.585417878611597e-06,
|
| 14960 |
+
"loss": 5.3774,
|
| 14961 |
+
"step": 8544
|
| 14962 |
+
},
|
| 14963 |
+
{
|
| 14964 |
+
"epoch": 0.83,
|
| 14965 |
+
"grad_norm": 1.0459861755371094,
|
| 14966 |
+
"learning_rate": 8.566026759744037e-06,
|
| 14967 |
+
"loss": 5.3309,
|
| 14968 |
+
"step": 8548
|
| 14969 |
+
},
|
| 14970 |
+
{
|
| 14971 |
+
"epoch": 0.83,
|
| 14972 |
+
"grad_norm": 1.019656777381897,
|
| 14973 |
+
"learning_rate": 8.546635640876478e-06,
|
| 14974 |
+
"loss": 5.3024,
|
| 14975 |
+
"step": 8552
|
| 14976 |
+
},
|
| 14977 |
+
{
|
| 14978 |
+
"epoch": 0.83,
|
| 14979 |
+
"grad_norm": 1.1104713678359985,
|
| 14980 |
+
"learning_rate": 8.52724452200892e-06,
|
| 14981 |
+
"loss": 5.2309,
|
| 14982 |
+
"step": 8556
|
| 14983 |
+
},
|
| 14984 |
+
{
|
| 14985 |
+
"epoch": 0.83,
|
| 14986 |
+
"grad_norm": 1.0480828285217285,
|
| 14987 |
+
"learning_rate": 8.507853403141362e-06,
|
| 14988 |
+
"loss": 5.256,
|
| 14989 |
+
"step": 8560
|
| 14990 |
+
},
|
| 14991 |
+
{
|
| 14992 |
+
"epoch": 0.83,
|
| 14993 |
+
"grad_norm": 1.0698785781860352,
|
| 14994 |
+
"learning_rate": 8.488462284273803e-06,
|
| 14995 |
+
"loss": 5.2322,
|
| 14996 |
+
"step": 8564
|
| 14997 |
+
},
|
| 14998 |
+
{
|
| 14999 |
+
"epoch": 0.83,
|
| 15000 |
+
"grad_norm": 1.0998084545135498,
|
| 15001 |
+
"learning_rate": 8.469071165406245e-06,
|
| 15002 |
+
"loss": 5.2919,
|
| 15003 |
+
"step": 8568
|
| 15004 |
+
},
|
| 15005 |
+
{
|
| 15006 |
+
"epoch": 0.83,
|
| 15007 |
+
"grad_norm": 1.0241094827651978,
|
| 15008 |
+
"learning_rate": 8.449680046538685e-06,
|
| 15009 |
+
"loss": 5.3195,
|
| 15010 |
+
"step": 8572
|
| 15011 |
+
},
|
| 15012 |
+
{
|
| 15013 |
+
"epoch": 0.83,
|
| 15014 |
+
"grad_norm": 1.1492643356323242,
|
| 15015 |
+
"learning_rate": 8.430288927671128e-06,
|
| 15016 |
+
"loss": 5.2311,
|
| 15017 |
+
"step": 8576
|
| 15018 |
+
},
|
| 15019 |
+
{
|
| 15020 |
+
"epoch": 0.83,
|
| 15021 |
+
"grad_norm": 1.1048632860183716,
|
| 15022 |
+
"learning_rate": 8.410897808803568e-06,
|
| 15023 |
+
"loss": 5.3227,
|
| 15024 |
+
"step": 8580
|
| 15025 |
+
},
|
| 15026 |
+
{
|
| 15027 |
+
"epoch": 0.83,
|
| 15028 |
+
"grad_norm": 1.0702450275421143,
|
| 15029 |
+
"learning_rate": 8.39150668993601e-06,
|
| 15030 |
+
"loss": 5.3173,
|
| 15031 |
+
"step": 8584
|
| 15032 |
+
},
|
| 15033 |
+
{
|
| 15034 |
+
"epoch": 0.83,
|
| 15035 |
+
"grad_norm": 1.0409200191497803,
|
| 15036 |
+
"learning_rate": 8.372115571068451e-06,
|
| 15037 |
+
"loss": 5.3185,
|
| 15038 |
+
"step": 8588
|
| 15039 |
+
},
|
| 15040 |
+
{
|
| 15041 |
+
"epoch": 0.83,
|
| 15042 |
+
"grad_norm": 1.046713948249817,
|
| 15043 |
+
"learning_rate": 8.352724452200893e-06,
|
| 15044 |
+
"loss": 5.3996,
|
| 15045 |
+
"step": 8592
|
| 15046 |
+
},
|
| 15047 |
+
{
|
| 15048 |
+
"epoch": 0.83,
|
| 15049 |
+
"grad_norm": 1.039919376373291,
|
| 15050 |
+
"learning_rate": 8.333333333333334e-06,
|
| 15051 |
+
"loss": 5.2483,
|
| 15052 |
+
"step": 8596
|
| 15053 |
+
},
|
| 15054 |
+
{
|
| 15055 |
+
"epoch": 0.83,
|
| 15056 |
+
"grad_norm": 1.0520331859588623,
|
| 15057 |
+
"learning_rate": 8.313942214465776e-06,
|
| 15058 |
+
"loss": 5.3082,
|
| 15059 |
+
"step": 8600
|
| 15060 |
+
},
|
| 15061 |
+
{
|
| 15062 |
+
"epoch": 0.83,
|
| 15063 |
+
"grad_norm": 1.0794312953948975,
|
| 15064 |
+
"learning_rate": 8.294551095598217e-06,
|
| 15065 |
+
"loss": 5.382,
|
| 15066 |
+
"step": 8604
|
| 15067 |
+
},
|
| 15068 |
+
{
|
| 15069 |
+
"epoch": 0.83,
|
| 15070 |
+
"grad_norm": 1.0222445726394653,
|
| 15071 |
+
"learning_rate": 8.275159976730657e-06,
|
| 15072 |
+
"loss": 5.2186,
|
| 15073 |
+
"step": 8608
|
| 15074 |
+
},
|
| 15075 |
+
{
|
| 15076 |
+
"epoch": 0.83,
|
| 15077 |
+
"grad_norm": 1.043550729751587,
|
| 15078 |
+
"learning_rate": 8.2557688578631e-06,
|
| 15079 |
+
"loss": 5.2447,
|
| 15080 |
+
"step": 8612
|
| 15081 |
+
},
|
| 15082 |
+
{
|
| 15083 |
+
"epoch": 0.84,
|
| 15084 |
+
"grad_norm": 1.065027117729187,
|
| 15085 |
+
"learning_rate": 8.23637773899554e-06,
|
| 15086 |
+
"loss": 5.3149,
|
| 15087 |
+
"step": 8616
|
| 15088 |
+
},
|
| 15089 |
+
{
|
| 15090 |
+
"epoch": 0.84,
|
| 15091 |
+
"grad_norm": 1.089449405670166,
|
| 15092 |
+
"learning_rate": 8.216986620127982e-06,
|
| 15093 |
+
"loss": 5.231,
|
| 15094 |
+
"step": 8620
|
| 15095 |
+
},
|
| 15096 |
+
{
|
| 15097 |
+
"epoch": 0.84,
|
| 15098 |
+
"grad_norm": 1.0330005884170532,
|
| 15099 |
+
"learning_rate": 8.197595501260423e-06,
|
| 15100 |
+
"loss": 5.3099,
|
| 15101 |
+
"step": 8624
|
| 15102 |
+
},
|
| 15103 |
+
{
|
| 15104 |
+
"epoch": 0.84,
|
| 15105 |
+
"grad_norm": 1.088131070137024,
|
| 15106 |
+
"learning_rate": 8.178204382392865e-06,
|
| 15107 |
+
"loss": 5.3941,
|
| 15108 |
+
"step": 8628
|
| 15109 |
+
},
|
| 15110 |
+
{
|
| 15111 |
+
"epoch": 0.84,
|
| 15112 |
+
"grad_norm": 1.0130773782730103,
|
| 15113 |
+
"learning_rate": 8.158813263525305e-06,
|
| 15114 |
+
"loss": 5.2442,
|
| 15115 |
+
"step": 8632
|
| 15116 |
+
},
|
| 15117 |
+
{
|
| 15118 |
+
"epoch": 0.84,
|
| 15119 |
+
"grad_norm": 1.035882592201233,
|
| 15120 |
+
"learning_rate": 8.139422144657748e-06,
|
| 15121 |
+
"loss": 5.21,
|
| 15122 |
+
"step": 8636
|
| 15123 |
+
},
|
| 15124 |
+
{
|
| 15125 |
+
"epoch": 0.84,
|
| 15126 |
+
"grad_norm": 1.0550565719604492,
|
| 15127 |
+
"learning_rate": 8.120031025790188e-06,
|
| 15128 |
+
"loss": 5.2888,
|
| 15129 |
+
"step": 8640
|
| 15130 |
+
},
|
| 15131 |
+
{
|
| 15132 |
+
"epoch": 0.84,
|
| 15133 |
+
"grad_norm": 1.1436634063720703,
|
| 15134 |
+
"learning_rate": 8.10063990692263e-06,
|
| 15135 |
+
"loss": 5.3587,
|
| 15136 |
+
"step": 8644
|
| 15137 |
+
},
|
| 15138 |
+
{
|
| 15139 |
+
"epoch": 0.84,
|
| 15140 |
+
"grad_norm": 1.1211497783660889,
|
| 15141 |
+
"learning_rate": 8.081248788055071e-06,
|
| 15142 |
+
"loss": 5.3188,
|
| 15143 |
+
"step": 8648
|
| 15144 |
+
},
|
| 15145 |
+
{
|
| 15146 |
+
"epoch": 0.84,
|
| 15147 |
+
"grad_norm": 1.1205918788909912,
|
| 15148 |
+
"learning_rate": 8.061857669187513e-06,
|
| 15149 |
+
"loss": 5.2688,
|
| 15150 |
+
"step": 8652
|
| 15151 |
+
},
|
| 15152 |
+
{
|
| 15153 |
+
"epoch": 0.84,
|
| 15154 |
+
"grad_norm": 1.0907244682312012,
|
| 15155 |
+
"learning_rate": 8.042466550319954e-06,
|
| 15156 |
+
"loss": 5.2518,
|
| 15157 |
+
"step": 8656
|
| 15158 |
+
},
|
| 15159 |
+
{
|
| 15160 |
+
"epoch": 0.84,
|
| 15161 |
+
"grad_norm": 1.0856692790985107,
|
| 15162 |
+
"learning_rate": 8.023075431452396e-06,
|
| 15163 |
+
"loss": 5.3384,
|
| 15164 |
+
"step": 8660
|
| 15165 |
+
},
|
| 15166 |
+
{
|
| 15167 |
+
"epoch": 0.84,
|
| 15168 |
+
"grad_norm": 1.0303173065185547,
|
| 15169 |
+
"learning_rate": 8.003684312584836e-06,
|
| 15170 |
+
"loss": 5.2906,
|
| 15171 |
+
"step": 8664
|
| 15172 |
+
},
|
| 15173 |
+
{
|
| 15174 |
+
"epoch": 0.84,
|
| 15175 |
+
"grad_norm": 1.0908282995224,
|
| 15176 |
+
"learning_rate": 7.984293193717279e-06,
|
| 15177 |
+
"loss": 5.3331,
|
| 15178 |
+
"step": 8668
|
| 15179 |
+
},
|
| 15180 |
+
{
|
| 15181 |
+
"epoch": 0.84,
|
| 15182 |
+
"grad_norm": 1.088040828704834,
|
| 15183 |
+
"learning_rate": 7.96490207484972e-06,
|
| 15184 |
+
"loss": 5.3521,
|
| 15185 |
+
"step": 8672
|
| 15186 |
+
},
|
| 15187 |
+
{
|
| 15188 |
+
"epoch": 0.84,
|
| 15189 |
+
"grad_norm": 1.0086363554000854,
|
| 15190 |
+
"learning_rate": 7.94551095598216e-06,
|
| 15191 |
+
"loss": 5.3294,
|
| 15192 |
+
"step": 8676
|
| 15193 |
+
},
|
| 15194 |
+
{
|
| 15195 |
+
"epoch": 0.84,
|
| 15196 |
+
"grad_norm": 1.1544169187545776,
|
| 15197 |
+
"learning_rate": 7.926119837114602e-06,
|
| 15198 |
+
"loss": 5.3184,
|
| 15199 |
+
"step": 8680
|
| 15200 |
+
},
|
| 15201 |
+
{
|
| 15202 |
+
"epoch": 0.84,
|
| 15203 |
+
"grad_norm": 1.0530931949615479,
|
| 15204 |
+
"learning_rate": 7.906728718247042e-06,
|
| 15205 |
+
"loss": 5.2866,
|
| 15206 |
+
"step": 8684
|
| 15207 |
+
},
|
| 15208 |
+
{
|
| 15209 |
+
"epoch": 0.84,
|
| 15210 |
+
"grad_norm": 1.080121397972107,
|
| 15211 |
+
"learning_rate": 7.887337599379485e-06,
|
| 15212 |
+
"loss": 5.2793,
|
| 15213 |
+
"step": 8688
|
| 15214 |
+
},
|
| 15215 |
+
{
|
| 15216 |
+
"epoch": 0.84,
|
| 15217 |
+
"grad_norm": 1.0175666809082031,
|
| 15218 |
+
"learning_rate": 7.867946480511925e-06,
|
| 15219 |
+
"loss": 5.3282,
|
| 15220 |
+
"step": 8692
|
| 15221 |
+
},
|
| 15222 |
+
{
|
| 15223 |
+
"epoch": 0.84,
|
| 15224 |
+
"grad_norm": 1.0283890962600708,
|
| 15225 |
+
"learning_rate": 7.848555361644367e-06,
|
| 15226 |
+
"loss": 5.3439,
|
| 15227 |
+
"step": 8696
|
| 15228 |
+
},
|
| 15229 |
+
{
|
| 15230 |
+
"epoch": 0.84,
|
| 15231 |
+
"grad_norm": 1.0480095148086548,
|
| 15232 |
+
"learning_rate": 7.829164242776808e-06,
|
| 15233 |
+
"loss": 5.3042,
|
| 15234 |
+
"step": 8700
|
| 15235 |
+
},
|
| 15236 |
+
{
|
| 15237 |
+
"epoch": 0.84,
|
| 15238 |
+
"grad_norm": 0.9947773814201355,
|
| 15239 |
+
"learning_rate": 7.80977312390925e-06,
|
| 15240 |
+
"loss": 5.251,
|
| 15241 |
+
"step": 8704
|
| 15242 |
+
},
|
| 15243 |
+
{
|
| 15244 |
+
"epoch": 0.84,
|
| 15245 |
+
"grad_norm": 1.033823847770691,
|
| 15246 |
+
"learning_rate": 7.79038200504169e-06,
|
| 15247 |
+
"loss": 5.4016,
|
| 15248 |
+
"step": 8708
|
| 15249 |
+
},
|
| 15250 |
+
{
|
| 15251 |
+
"epoch": 0.84,
|
| 15252 |
+
"grad_norm": 1.129824161529541,
|
| 15253 |
+
"learning_rate": 7.770990886174133e-06,
|
| 15254 |
+
"loss": 5.2176,
|
| 15255 |
+
"step": 8712
|
| 15256 |
+
},
|
| 15257 |
+
{
|
| 15258 |
+
"epoch": 0.85,
|
| 15259 |
+
"grad_norm": 1.0202304124832153,
|
| 15260 |
+
"learning_rate": 7.751599767306574e-06,
|
| 15261 |
+
"loss": 5.3219,
|
| 15262 |
+
"step": 8716
|
| 15263 |
+
},
|
| 15264 |
+
{
|
| 15265 |
+
"epoch": 0.85,
|
| 15266 |
+
"grad_norm": 1.0748639106750488,
|
| 15267 |
+
"learning_rate": 7.732208648439016e-06,
|
| 15268 |
+
"loss": 5.2945,
|
| 15269 |
+
"step": 8720
|
| 15270 |
+
},
|
| 15271 |
+
{
|
| 15272 |
+
"epoch": 0.85,
|
| 15273 |
+
"grad_norm": 1.0026463270187378,
|
| 15274 |
+
"learning_rate": 7.712817529571456e-06,
|
| 15275 |
+
"loss": 5.3474,
|
| 15276 |
+
"step": 8724
|
| 15277 |
+
},
|
| 15278 |
+
{
|
| 15279 |
+
"epoch": 0.85,
|
| 15280 |
+
"grad_norm": 0.98891282081604,
|
| 15281 |
+
"learning_rate": 7.693426410703899e-06,
|
| 15282 |
+
"loss": 5.2612,
|
| 15283 |
+
"step": 8728
|
| 15284 |
+
},
|
| 15285 |
+
{
|
| 15286 |
+
"epoch": 0.85,
|
| 15287 |
+
"grad_norm": 1.079750418663025,
|
| 15288 |
+
"learning_rate": 7.674035291836339e-06,
|
| 15289 |
+
"loss": 5.2398,
|
| 15290 |
+
"step": 8732
|
| 15291 |
+
},
|
| 15292 |
+
{
|
| 15293 |
+
"epoch": 0.85,
|
| 15294 |
+
"grad_norm": 1.057255744934082,
|
| 15295 |
+
"learning_rate": 7.654644172968781e-06,
|
| 15296 |
+
"loss": 5.4364,
|
| 15297 |
+
"step": 8736
|
| 15298 |
+
},
|
| 15299 |
+
{
|
| 15300 |
+
"epoch": 0.85,
|
| 15301 |
+
"grad_norm": 1.0570470094680786,
|
| 15302 |
+
"learning_rate": 7.635253054101222e-06,
|
| 15303 |
+
"loss": 5.2695,
|
| 15304 |
+
"step": 8740
|
| 15305 |
+
},
|
| 15306 |
+
{
|
| 15307 |
+
"epoch": 0.85,
|
| 15308 |
+
"grad_norm": 1.0664699077606201,
|
| 15309 |
+
"learning_rate": 7.615861935233663e-06,
|
| 15310 |
+
"loss": 5.3378,
|
| 15311 |
+
"step": 8744
|
| 15312 |
+
},
|
| 15313 |
+
{
|
| 15314 |
+
"epoch": 0.85,
|
| 15315 |
+
"grad_norm": 1.0890289545059204,
|
| 15316 |
+
"learning_rate": 7.5964708163661055e-06,
|
| 15317 |
+
"loss": 5.375,
|
| 15318 |
+
"step": 8748
|
| 15319 |
+
},
|
| 15320 |
+
{
|
| 15321 |
+
"epoch": 0.85,
|
| 15322 |
+
"grad_norm": 1.0607807636260986,
|
| 15323 |
+
"learning_rate": 7.577079697498546e-06,
|
| 15324 |
+
"loss": 5.3258,
|
| 15325 |
+
"step": 8752
|
| 15326 |
+
},
|
| 15327 |
+
{
|
| 15328 |
+
"epoch": 0.85,
|
| 15329 |
+
"grad_norm": 1.0911225080490112,
|
| 15330 |
+
"learning_rate": 7.557688578630988e-06,
|
| 15331 |
+
"loss": 5.242,
|
| 15332 |
+
"step": 8756
|
| 15333 |
+
},
|
| 15334 |
+
{
|
| 15335 |
+
"epoch": 0.85,
|
| 15336 |
+
"grad_norm": 1.1005817651748657,
|
| 15337 |
+
"learning_rate": 7.538297459763429e-06,
|
| 15338 |
+
"loss": 5.2734,
|
| 15339 |
+
"step": 8760
|
| 15340 |
+
},
|
| 15341 |
+
{
|
| 15342 |
+
"epoch": 0.85,
|
| 15343 |
+
"grad_norm": 1.0650907754898071,
|
| 15344 |
+
"learning_rate": 7.51890634089587e-06,
|
| 15345 |
+
"loss": 5.2939,
|
| 15346 |
+
"step": 8764
|
| 15347 |
+
},
|
| 15348 |
+
{
|
| 15349 |
+
"epoch": 0.85,
|
| 15350 |
+
"grad_norm": 1.0343334674835205,
|
| 15351 |
+
"learning_rate": 7.499515222028311e-06,
|
| 15352 |
+
"loss": 5.2601,
|
| 15353 |
+
"step": 8768
|
| 15354 |
+
},
|
| 15355 |
+
{
|
| 15356 |
+
"epoch": 0.85,
|
| 15357 |
+
"grad_norm": 1.092239260673523,
|
| 15358 |
+
"learning_rate": 7.480124103160753e-06,
|
| 15359 |
+
"loss": 5.266,
|
| 15360 |
+
"step": 8772
|
| 15361 |
+
},
|
| 15362 |
+
{
|
| 15363 |
+
"epoch": 0.85,
|
| 15364 |
+
"grad_norm": 1.140648603439331,
|
| 15365 |
+
"learning_rate": 7.4607329842931935e-06,
|
| 15366 |
+
"loss": 5.2353,
|
| 15367 |
+
"step": 8776
|
| 15368 |
+
},
|
| 15369 |
+
{
|
| 15370 |
+
"epoch": 0.85,
|
| 15371 |
+
"grad_norm": 1.0743423700332642,
|
| 15372 |
+
"learning_rate": 7.441341865425636e-06,
|
| 15373 |
+
"loss": 5.389,
|
| 15374 |
+
"step": 8780
|
| 15375 |
+
},
|
| 15376 |
+
{
|
| 15377 |
+
"epoch": 0.85,
|
| 15378 |
+
"grad_norm": 1.0870285034179688,
|
| 15379 |
+
"learning_rate": 7.421950746558076e-06,
|
| 15380 |
+
"loss": 5.3212,
|
| 15381 |
+
"step": 8784
|
| 15382 |
+
},
|
| 15383 |
+
{
|
| 15384 |
+
"epoch": 0.85,
|
| 15385 |
+
"grad_norm": 1.0318245887756348,
|
| 15386 |
+
"learning_rate": 7.4025596276905185e-06,
|
| 15387 |
+
"loss": 5.2854,
|
| 15388 |
+
"step": 8788
|
| 15389 |
+
},
|
| 15390 |
+
{
|
| 15391 |
+
"epoch": 0.85,
|
| 15392 |
+
"grad_norm": 1.0597593784332275,
|
| 15393 |
+
"learning_rate": 7.383168508822959e-06,
|
| 15394 |
+
"loss": 5.4204,
|
| 15395 |
+
"step": 8792
|
| 15396 |
+
},
|
| 15397 |
+
{
|
| 15398 |
+
"epoch": 0.85,
|
| 15399 |
+
"grad_norm": 1.0621132850646973,
|
| 15400 |
+
"learning_rate": 7.3637773899554e-06,
|
| 15401 |
+
"loss": 5.2903,
|
| 15402 |
+
"step": 8796
|
| 15403 |
+
},
|
| 15404 |
+
{
|
| 15405 |
+
"epoch": 0.85,
|
| 15406 |
+
"grad_norm": 1.057024598121643,
|
| 15407 |
+
"learning_rate": 7.344386271087843e-06,
|
| 15408 |
+
"loss": 5.3975,
|
| 15409 |
+
"step": 8800
|
| 15410 |
+
},
|
| 15411 |
+
{
|
| 15412 |
+
"epoch": 0.85,
|
| 15413 |
+
"grad_norm": 1.1537240743637085,
|
| 15414 |
+
"learning_rate": 7.324995152220283e-06,
|
| 15415 |
+
"loss": 5.3118,
|
| 15416 |
+
"step": 8804
|
| 15417 |
+
},
|
| 15418 |
+
{
|
| 15419 |
+
"epoch": 0.85,
|
| 15420 |
+
"grad_norm": 1.161657691001892,
|
| 15421 |
+
"learning_rate": 7.305604033352725e-06,
|
| 15422 |
+
"loss": 5.3516,
|
| 15423 |
+
"step": 8808
|
| 15424 |
+
},
|
| 15425 |
+
{
|
| 15426 |
+
"epoch": 0.85,
|
| 15427 |
+
"grad_norm": 1.0824769735336304,
|
| 15428 |
+
"learning_rate": 7.286212914485166e-06,
|
| 15429 |
+
"loss": 5.3601,
|
| 15430 |
+
"step": 8812
|
| 15431 |
+
},
|
| 15432 |
+
{
|
| 15433 |
+
"epoch": 0.85,
|
| 15434 |
+
"grad_norm": 1.0471476316452026,
|
| 15435 |
+
"learning_rate": 7.266821795617608e-06,
|
| 15436 |
+
"loss": 5.2858,
|
| 15437 |
+
"step": 8816
|
| 15438 |
+
},
|
| 15439 |
+
{
|
| 15440 |
+
"epoch": 0.86,
|
| 15441 |
+
"grad_norm": 1.0438990592956543,
|
| 15442 |
+
"learning_rate": 7.247430676750049e-06,
|
| 15443 |
+
"loss": 5.2688,
|
| 15444 |
+
"step": 8820
|
| 15445 |
+
},
|
| 15446 |
+
{
|
| 15447 |
+
"epoch": 0.86,
|
| 15448 |
+
"grad_norm": 1.0383694171905518,
|
| 15449 |
+
"learning_rate": 7.228039557882491e-06,
|
| 15450 |
+
"loss": 5.3404,
|
| 15451 |
+
"step": 8824
|
| 15452 |
+
},
|
| 15453 |
+
{
|
| 15454 |
+
"epoch": 0.86,
|
| 15455 |
+
"grad_norm": 1.039699673652649,
|
| 15456 |
+
"learning_rate": 7.2086484390149315e-06,
|
| 15457 |
+
"loss": 5.2846,
|
| 15458 |
+
"step": 8828
|
| 15459 |
+
},
|
| 15460 |
+
{
|
| 15461 |
+
"epoch": 0.86,
|
| 15462 |
+
"grad_norm": 1.1256271600723267,
|
| 15463 |
+
"learning_rate": 7.189257320147373e-06,
|
| 15464 |
+
"loss": 5.327,
|
| 15465 |
+
"step": 8832
|
| 15466 |
+
},
|
| 15467 |
+
{
|
| 15468 |
+
"epoch": 0.86,
|
| 15469 |
+
"grad_norm": 0.9789720177650452,
|
| 15470 |
+
"learning_rate": 7.169866201279813e-06,
|
| 15471 |
+
"loss": 5.3369,
|
| 15472 |
+
"step": 8836
|
| 15473 |
+
},
|
| 15474 |
+
{
|
| 15475 |
+
"epoch": 0.86,
|
| 15476 |
+
"grad_norm": 1.0387988090515137,
|
| 15477 |
+
"learning_rate": 7.150475082412256e-06,
|
| 15478 |
+
"loss": 5.3402,
|
| 15479 |
+
"step": 8840
|
| 15480 |
+
},
|
| 15481 |
+
{
|
| 15482 |
+
"epoch": 0.86,
|
| 15483 |
+
"grad_norm": 1.0737075805664062,
|
| 15484 |
+
"learning_rate": 7.131083963544696e-06,
|
| 15485 |
+
"loss": 5.2837,
|
| 15486 |
+
"step": 8844
|
| 15487 |
+
},
|
| 15488 |
+
{
|
| 15489 |
+
"epoch": 0.86,
|
| 15490 |
+
"grad_norm": 1.1215327978134155,
|
| 15491 |
+
"learning_rate": 7.111692844677138e-06,
|
| 15492 |
+
"loss": 5.2237,
|
| 15493 |
+
"step": 8848
|
| 15494 |
+
},
|
| 15495 |
+
{
|
| 15496 |
+
"epoch": 0.86,
|
| 15497 |
+
"grad_norm": 1.0533177852630615,
|
| 15498 |
+
"learning_rate": 7.092301725809579e-06,
|
| 15499 |
+
"loss": 5.3783,
|
| 15500 |
+
"step": 8852
|
| 15501 |
+
},
|
| 15502 |
+
{
|
| 15503 |
+
"epoch": 0.86,
|
| 15504 |
+
"grad_norm": 1.1658439636230469,
|
| 15505 |
+
"learning_rate": 7.072910606942021e-06,
|
| 15506 |
+
"loss": 5.3175,
|
| 15507 |
+
"step": 8856
|
| 15508 |
+
},
|
| 15509 |
+
{
|
| 15510 |
+
"epoch": 0.86,
|
| 15511 |
+
"grad_norm": 1.0966906547546387,
|
| 15512 |
+
"learning_rate": 7.0535194880744625e-06,
|
| 15513 |
+
"loss": 5.4216,
|
| 15514 |
+
"step": 8860
|
| 15515 |
+
},
|
| 15516 |
+
{
|
| 15517 |
+
"epoch": 0.86,
|
| 15518 |
+
"grad_norm": 1.0952768325805664,
|
| 15519 |
+
"learning_rate": 7.034128369206904e-06,
|
| 15520 |
+
"loss": 5.2748,
|
| 15521 |
+
"step": 8864
|
| 15522 |
+
},
|
| 15523 |
+
{
|
| 15524 |
+
"epoch": 0.86,
|
| 15525 |
+
"grad_norm": 1.096529483795166,
|
| 15526 |
+
"learning_rate": 7.014737250339345e-06,
|
| 15527 |
+
"loss": 5.1666,
|
| 15528 |
+
"step": 8868
|
| 15529 |
+
},
|
| 15530 |
+
{
|
| 15531 |
+
"epoch": 0.86,
|
| 15532 |
+
"grad_norm": 1.0736936330795288,
|
| 15533 |
+
"learning_rate": 6.995346131471786e-06,
|
| 15534 |
+
"loss": 5.24,
|
| 15535 |
+
"step": 8872
|
| 15536 |
+
},
|
| 15537 |
+
{
|
| 15538 |
+
"epoch": 0.86,
|
| 15539 |
+
"grad_norm": 1.0133376121520996,
|
| 15540 |
+
"learning_rate": 6.975955012604228e-06,
|
| 15541 |
+
"loss": 5.2827,
|
| 15542 |
+
"step": 8876
|
| 15543 |
+
},
|
| 15544 |
+
{
|
| 15545 |
+
"epoch": 0.86,
|
| 15546 |
+
"grad_norm": 1.0580708980560303,
|
| 15547 |
+
"learning_rate": 6.956563893736669e-06,
|
| 15548 |
+
"loss": 5.3801,
|
| 15549 |
+
"step": 8880
|
| 15550 |
+
},
|
| 15551 |
+
{
|
| 15552 |
+
"epoch": 0.86,
|
| 15553 |
+
"grad_norm": 1.1220327615737915,
|
| 15554 |
+
"learning_rate": 6.937172774869111e-06,
|
| 15555 |
+
"loss": 5.2949,
|
| 15556 |
+
"step": 8884
|
| 15557 |
+
},
|
| 15558 |
+
{
|
| 15559 |
+
"epoch": 0.86,
|
| 15560 |
+
"grad_norm": 1.136806845664978,
|
| 15561 |
+
"learning_rate": 6.917781656001551e-06,
|
| 15562 |
+
"loss": 5.347,
|
| 15563 |
+
"step": 8888
|
| 15564 |
+
},
|
| 15565 |
+
{
|
| 15566 |
+
"epoch": 0.86,
|
| 15567 |
+
"grad_norm": 1.1147714853286743,
|
| 15568 |
+
"learning_rate": 6.898390537133994e-06,
|
| 15569 |
+
"loss": 5.2648,
|
| 15570 |
+
"step": 8892
|
| 15571 |
+
},
|
| 15572 |
+
{
|
| 15573 |
+
"epoch": 0.86,
|
| 15574 |
+
"grad_norm": 1.0525692701339722,
|
| 15575 |
+
"learning_rate": 6.878999418266434e-06,
|
| 15576 |
+
"loss": 5.2665,
|
| 15577 |
+
"step": 8896
|
| 15578 |
+
},
|
| 15579 |
+
{
|
| 15580 |
+
"epoch": 0.86,
|
| 15581 |
+
"grad_norm": 1.0400636196136475,
|
| 15582 |
+
"learning_rate": 6.8596082993988755e-06,
|
| 15583 |
+
"loss": 5.3571,
|
| 15584 |
+
"step": 8900
|
| 15585 |
+
},
|
| 15586 |
+
{
|
| 15587 |
+
"epoch": 0.86,
|
| 15588 |
+
"grad_norm": 1.0699836015701294,
|
| 15589 |
+
"learning_rate": 6.840217180531316e-06,
|
| 15590 |
+
"loss": 5.3079,
|
| 15591 |
+
"step": 8904
|
| 15592 |
+
},
|
| 15593 |
+
{
|
| 15594 |
+
"epoch": 0.86,
|
| 15595 |
+
"grad_norm": 1.0171644687652588,
|
| 15596 |
+
"learning_rate": 6.820826061663758e-06,
|
| 15597 |
+
"loss": 5.3438,
|
| 15598 |
+
"step": 8908
|
| 15599 |
+
},
|
| 15600 |
+
{
|
| 15601 |
+
"epoch": 0.86,
|
| 15602 |
+
"grad_norm": 1.0426756143569946,
|
| 15603 |
+
"learning_rate": 6.801434942796199e-06,
|
| 15604 |
+
"loss": 5.417,
|
| 15605 |
+
"step": 8912
|
| 15606 |
+
},
|
| 15607 |
+
{
|
| 15608 |
+
"epoch": 0.86,
|
| 15609 |
+
"grad_norm": 1.1138461828231812,
|
| 15610 |
+
"learning_rate": 6.782043823928641e-06,
|
| 15611 |
+
"loss": 5.2769,
|
| 15612 |
+
"step": 8916
|
| 15613 |
+
},
|
| 15614 |
+
{
|
| 15615 |
+
"epoch": 0.86,
|
| 15616 |
+
"grad_norm": 1.0219694375991821,
|
| 15617 |
+
"learning_rate": 6.762652705061083e-06,
|
| 15618 |
+
"loss": 5.2811,
|
| 15619 |
+
"step": 8920
|
| 15620 |
+
},
|
| 15621 |
+
{
|
| 15622 |
+
"epoch": 0.87,
|
| 15623 |
+
"grad_norm": 1.1686511039733887,
|
| 15624 |
+
"learning_rate": 6.743261586193524e-06,
|
| 15625 |
+
"loss": 5.3218,
|
| 15626 |
+
"step": 8924
|
| 15627 |
+
},
|
| 15628 |
+
{
|
| 15629 |
+
"epoch": 0.87,
|
| 15630 |
+
"grad_norm": 1.0407147407531738,
|
| 15631 |
+
"learning_rate": 6.723870467325965e-06,
|
| 15632 |
+
"loss": 5.339,
|
| 15633 |
+
"step": 8928
|
| 15634 |
+
},
|
| 15635 |
+
{
|
| 15636 |
+
"epoch": 0.87,
|
| 15637 |
+
"grad_norm": 1.1056681871414185,
|
| 15638 |
+
"learning_rate": 6.704479348458407e-06,
|
| 15639 |
+
"loss": 5.2758,
|
| 15640 |
+
"step": 8932
|
| 15641 |
+
},
|
| 15642 |
+
{
|
| 15643 |
+
"epoch": 0.87,
|
| 15644 |
+
"grad_norm": 1.0969740152359009,
|
| 15645 |
+
"learning_rate": 6.685088229590848e-06,
|
| 15646 |
+
"loss": 5.2432,
|
| 15647 |
+
"step": 8936
|
| 15648 |
+
},
|
| 15649 |
+
{
|
| 15650 |
+
"epoch": 0.87,
|
| 15651 |
+
"grad_norm": 0.9841113090515137,
|
| 15652 |
+
"learning_rate": 6.6656971107232885e-06,
|
| 15653 |
+
"loss": 5.2596,
|
| 15654 |
+
"step": 8940
|
| 15655 |
+
},
|
| 15656 |
+
{
|
| 15657 |
+
"epoch": 0.87,
|
| 15658 |
+
"grad_norm": 1.1172292232513428,
|
| 15659 |
+
"learning_rate": 6.646305991855731e-06,
|
| 15660 |
+
"loss": 5.2884,
|
| 15661 |
+
"step": 8944
|
| 15662 |
+
},
|
| 15663 |
+
{
|
| 15664 |
+
"epoch": 0.87,
|
| 15665 |
+
"grad_norm": 0.9936596155166626,
|
| 15666 |
+
"learning_rate": 6.626914872988171e-06,
|
| 15667 |
+
"loss": 5.2968,
|
| 15668 |
+
"step": 8948
|
| 15669 |
+
},
|
| 15670 |
+
{
|
| 15671 |
+
"epoch": 0.87,
|
| 15672 |
+
"grad_norm": 1.0389301776885986,
|
| 15673 |
+
"learning_rate": 6.6075237541206135e-06,
|
| 15674 |
+
"loss": 5.2827,
|
| 15675 |
+
"step": 8952
|
| 15676 |
+
},
|
| 15677 |
+
{
|
| 15678 |
+
"epoch": 0.87,
|
| 15679 |
+
"grad_norm": 1.020494818687439,
|
| 15680 |
+
"learning_rate": 6.588132635253054e-06,
|
| 15681 |
+
"loss": 5.3882,
|
| 15682 |
+
"step": 8956
|
| 15683 |
+
},
|
| 15684 |
+
{
|
| 15685 |
+
"epoch": 0.87,
|
| 15686 |
+
"grad_norm": 1.0391160249710083,
|
| 15687 |
+
"learning_rate": 6.568741516385496e-06,
|
| 15688 |
+
"loss": 5.2172,
|
| 15689 |
+
"step": 8960
|
| 15690 |
+
},
|
| 15691 |
+
{
|
| 15692 |
+
"epoch": 0.87,
|
| 15693 |
+
"grad_norm": 1.0213825702667236,
|
| 15694 |
+
"learning_rate": 6.549350397517937e-06,
|
| 15695 |
+
"loss": 5.3307,
|
| 15696 |
+
"step": 8964
|
| 15697 |
+
},
|
| 15698 |
+
{
|
| 15699 |
+
"epoch": 0.87,
|
| 15700 |
+
"grad_norm": 1.0745649337768555,
|
| 15701 |
+
"learning_rate": 6.529959278650378e-06,
|
| 15702 |
+
"loss": 5.2638,
|
| 15703 |
+
"step": 8968
|
| 15704 |
+
},
|
| 15705 |
+
{
|
| 15706 |
+
"epoch": 0.87,
|
| 15707 |
+
"grad_norm": 1.0567609071731567,
|
| 15708 |
+
"learning_rate": 6.510568159782819e-06,
|
| 15709 |
+
"loss": 5.3164,
|
| 15710 |
+
"step": 8972
|
| 15711 |
+
},
|
| 15712 |
+
{
|
| 15713 |
+
"epoch": 0.87,
|
| 15714 |
+
"grad_norm": 1.0450811386108398,
|
| 15715 |
+
"learning_rate": 6.491177040915261e-06,
|
| 15716 |
+
"loss": 5.3648,
|
| 15717 |
+
"step": 8976
|
| 15718 |
+
},
|
| 15719 |
+
{
|
| 15720 |
+
"epoch": 0.87,
|
| 15721 |
+
"grad_norm": 1.0880790948867798,
|
| 15722 |
+
"learning_rate": 6.471785922047703e-06,
|
| 15723 |
+
"loss": 5.3617,
|
| 15724 |
+
"step": 8980
|
| 15725 |
+
},
|
| 15726 |
+
{
|
| 15727 |
+
"epoch": 0.87,
|
| 15728 |
+
"grad_norm": 1.0606417655944824,
|
| 15729 |
+
"learning_rate": 6.452394803180144e-06,
|
| 15730 |
+
"loss": 5.2201,
|
| 15731 |
+
"step": 8984
|
| 15732 |
+
},
|
| 15733 |
+
{
|
| 15734 |
+
"epoch": 0.87,
|
| 15735 |
+
"grad_norm": 1.0124664306640625,
|
| 15736 |
+
"learning_rate": 6.433003684312586e-06,
|
| 15737 |
+
"loss": 5.2499,
|
| 15738 |
+
"step": 8988
|
| 15739 |
+
},
|
| 15740 |
+
{
|
| 15741 |
+
"epoch": 0.87,
|
| 15742 |
+
"grad_norm": 1.1352604627609253,
|
| 15743 |
+
"learning_rate": 6.4136125654450265e-06,
|
| 15744 |
+
"loss": 5.3782,
|
| 15745 |
+
"step": 8992
|
| 15746 |
+
},
|
| 15747 |
+
{
|
| 15748 |
+
"epoch": 0.87,
|
| 15749 |
+
"grad_norm": 1.1061619520187378,
|
| 15750 |
+
"learning_rate": 6.394221446577468e-06,
|
| 15751 |
+
"loss": 5.3408,
|
| 15752 |
+
"step": 8996
|
| 15753 |
+
},
|
| 15754 |
+
{
|
| 15755 |
+
"epoch": 0.87,
|
| 15756 |
+
"grad_norm": 1.1135718822479248,
|
| 15757 |
+
"learning_rate": 6.374830327709909e-06,
|
| 15758 |
+
"loss": 5.2323,
|
| 15759 |
+
"step": 9000
|
| 15760 |
+
},
|
| 15761 |
+
{
|
| 15762 |
+
"epoch": 0.87,
|
| 15763 |
+
"grad_norm": 1.0155010223388672,
|
| 15764 |
+
"learning_rate": 6.355439208842351e-06,
|
| 15765 |
+
"loss": 5.2942,
|
| 15766 |
+
"step": 9004
|
| 15767 |
+
},
|
| 15768 |
+
{
|
| 15769 |
+
"epoch": 0.87,
|
| 15770 |
+
"grad_norm": 1.044931173324585,
|
| 15771 |
+
"learning_rate": 6.336048089974791e-06,
|
| 15772 |
+
"loss": 5.3267,
|
| 15773 |
+
"step": 9008
|
| 15774 |
+
},
|
| 15775 |
+
{
|
| 15776 |
+
"epoch": 0.87,
|
| 15777 |
+
"grad_norm": 1.0643398761749268,
|
| 15778 |
+
"learning_rate": 6.316656971107233e-06,
|
| 15779 |
+
"loss": 5.2651,
|
| 15780 |
+
"step": 9012
|
| 15781 |
+
},
|
| 15782 |
+
{
|
| 15783 |
+
"epoch": 0.87,
|
| 15784 |
+
"grad_norm": 1.0203381776809692,
|
| 15785 |
+
"learning_rate": 6.297265852239674e-06,
|
| 15786 |
+
"loss": 5.3493,
|
| 15787 |
+
"step": 9016
|
| 15788 |
+
},
|
| 15789 |
+
{
|
| 15790 |
+
"epoch": 0.87,
|
| 15791 |
+
"grad_norm": 1.0563126802444458,
|
| 15792 |
+
"learning_rate": 6.277874733372116e-06,
|
| 15793 |
+
"loss": 5.362,
|
| 15794 |
+
"step": 9020
|
| 15795 |
+
},
|
| 15796 |
+
{
|
| 15797 |
+
"epoch": 0.87,
|
| 15798 |
+
"grad_norm": 1.0667084455490112,
|
| 15799 |
+
"learning_rate": 6.258483614504557e-06,
|
| 15800 |
+
"loss": 5.2151,
|
| 15801 |
+
"step": 9024
|
| 15802 |
+
},
|
| 15803 |
+
{
|
| 15804 |
+
"epoch": 0.88,
|
| 15805 |
+
"grad_norm": 1.0885251760482788,
|
| 15806 |
+
"learning_rate": 6.239092495636999e-06,
|
| 15807 |
+
"loss": 5.3404,
|
| 15808 |
+
"step": 9028
|
| 15809 |
+
},
|
| 15810 |
+
{
|
| 15811 |
+
"epoch": 0.88,
|
| 15812 |
+
"grad_norm": 1.0210630893707275,
|
| 15813 |
+
"learning_rate": 6.21970137676944e-06,
|
| 15814 |
+
"loss": 5.2937,
|
| 15815 |
+
"step": 9032
|
| 15816 |
+
},
|
| 15817 |
+
{
|
| 15818 |
+
"epoch": 0.88,
|
| 15819 |
+
"grad_norm": 1.1303844451904297,
|
| 15820 |
+
"learning_rate": 6.200310257901881e-06,
|
| 15821 |
+
"loss": 5.3373,
|
| 15822 |
+
"step": 9036
|
| 15823 |
+
},
|
| 15824 |
+
{
|
| 15825 |
+
"epoch": 0.88,
|
| 15826 |
+
"grad_norm": 1.1023499965667725,
|
| 15827 |
+
"learning_rate": 6.180919139034322e-06,
|
| 15828 |
+
"loss": 5.2946,
|
| 15829 |
+
"step": 9040
|
| 15830 |
+
},
|
| 15831 |
+
{
|
| 15832 |
+
"epoch": 0.88,
|
| 15833 |
+
"grad_norm": 1.0469759702682495,
|
| 15834 |
+
"learning_rate": 6.161528020166764e-06,
|
| 15835 |
+
"loss": 5.3265,
|
| 15836 |
+
"step": 9044
|
| 15837 |
+
},
|
| 15838 |
+
{
|
| 15839 |
+
"epoch": 0.88,
|
| 15840 |
+
"grad_norm": 0.9917576313018799,
|
| 15841 |
+
"learning_rate": 6.142136901299205e-06,
|
| 15842 |
+
"loss": 5.2984,
|
| 15843 |
+
"step": 9048
|
| 15844 |
+
},
|
| 15845 |
+
{
|
| 15846 |
+
"epoch": 0.88,
|
| 15847 |
+
"grad_norm": 1.0630229711532593,
|
| 15848 |
+
"learning_rate": 6.122745782431646e-06,
|
| 15849 |
+
"loss": 5.2025,
|
| 15850 |
+
"step": 9052
|
| 15851 |
+
},
|
| 15852 |
+
{
|
| 15853 |
+
"epoch": 0.88,
|
| 15854 |
+
"grad_norm": 1.0385984182357788,
|
| 15855 |
+
"learning_rate": 6.103354663564088e-06,
|
| 15856 |
+
"loss": 5.2934,
|
| 15857 |
+
"step": 9056
|
| 15858 |
+
},
|
| 15859 |
+
{
|
| 15860 |
+
"epoch": 0.88,
|
| 15861 |
+
"grad_norm": 1.0480278730392456,
|
| 15862 |
+
"learning_rate": 6.083963544696529e-06,
|
| 15863 |
+
"loss": 5.2933,
|
| 15864 |
+
"step": 9060
|
| 15865 |
+
},
|
| 15866 |
+
{
|
| 15867 |
+
"epoch": 0.88,
|
| 15868 |
+
"grad_norm": 1.0395824909210205,
|
| 15869 |
+
"learning_rate": 6.0645724258289706e-06,
|
| 15870 |
+
"loss": 5.267,
|
| 15871 |
+
"step": 9064
|
| 15872 |
+
},
|
| 15873 |
+
{
|
| 15874 |
+
"epoch": 0.88,
|
| 15875 |
+
"grad_norm": 1.0996421575546265,
|
| 15876 |
+
"learning_rate": 6.045181306961412e-06,
|
| 15877 |
+
"loss": 5.3497,
|
| 15878 |
+
"step": 9068
|
| 15879 |
+
},
|
| 15880 |
+
{
|
| 15881 |
+
"epoch": 0.88,
|
| 15882 |
+
"grad_norm": 1.0920591354370117,
|
| 15883 |
+
"learning_rate": 6.025790188093853e-06,
|
| 15884 |
+
"loss": 5.2752,
|
| 15885 |
+
"step": 9072
|
| 15886 |
+
},
|
| 15887 |
+
{
|
| 15888 |
+
"epoch": 0.88,
|
| 15889 |
+
"grad_norm": 1.0730946063995361,
|
| 15890 |
+
"learning_rate": 6.006399069226295e-06,
|
| 15891 |
+
"loss": 5.3234,
|
| 15892 |
+
"step": 9076
|
| 15893 |
+
},
|
| 15894 |
+
{
|
| 15895 |
+
"epoch": 0.88,
|
| 15896 |
+
"grad_norm": 1.077646017074585,
|
| 15897 |
+
"learning_rate": 5.987007950358736e-06,
|
| 15898 |
+
"loss": 5.3009,
|
| 15899 |
+
"step": 9080
|
| 15900 |
+
},
|
| 15901 |
+
{
|
| 15902 |
+
"epoch": 0.88,
|
| 15903 |
+
"grad_norm": 1.0901986360549927,
|
| 15904 |
+
"learning_rate": 5.9676168314911775e-06,
|
| 15905 |
+
"loss": 5.2653,
|
| 15906 |
+
"step": 9084
|
| 15907 |
+
},
|
| 15908 |
+
{
|
| 15909 |
+
"epoch": 0.88,
|
| 15910 |
+
"grad_norm": 1.1307499408721924,
|
| 15911 |
+
"learning_rate": 5.948225712623619e-06,
|
| 15912 |
+
"loss": 5.393,
|
| 15913 |
+
"step": 9088
|
| 15914 |
+
},
|
| 15915 |
+
{
|
| 15916 |
+
"epoch": 0.88,
|
| 15917 |
+
"grad_norm": 1.1002899408340454,
|
| 15918 |
+
"learning_rate": 5.92883459375606e-06,
|
| 15919 |
+
"loss": 5.2785,
|
| 15920 |
+
"step": 9092
|
| 15921 |
+
},
|
| 15922 |
+
{
|
| 15923 |
+
"epoch": 0.88,
|
| 15924 |
+
"grad_norm": 1.1348730325698853,
|
| 15925 |
+
"learning_rate": 5.909443474888502e-06,
|
| 15926 |
+
"loss": 5.3749,
|
| 15927 |
+
"step": 9096
|
| 15928 |
+
},
|
| 15929 |
+
{
|
| 15930 |
+
"epoch": 0.88,
|
| 15931 |
+
"grad_norm": 1.2404577732086182,
|
| 15932 |
+
"learning_rate": 5.890052356020943e-06,
|
| 15933 |
+
"loss": 5.2968,
|
| 15934 |
+
"step": 9100
|
| 15935 |
+
},
|
| 15936 |
+
{
|
| 15937 |
+
"epoch": 0.88,
|
| 15938 |
+
"grad_norm": 1.1860145330429077,
|
| 15939 |
+
"learning_rate": 5.8706612371533835e-06,
|
| 15940 |
+
"loss": 5.2951,
|
| 15941 |
+
"step": 9104
|
| 15942 |
+
},
|
| 15943 |
+
{
|
| 15944 |
+
"epoch": 0.88,
|
| 15945 |
+
"grad_norm": 1.0747588872909546,
|
| 15946 |
+
"learning_rate": 5.851270118285825e-06,
|
| 15947 |
+
"loss": 5.2725,
|
| 15948 |
+
"step": 9108
|
| 15949 |
+
},
|
| 15950 |
+
{
|
| 15951 |
+
"epoch": 0.88,
|
| 15952 |
+
"grad_norm": 1.0642809867858887,
|
| 15953 |
+
"learning_rate": 5.831878999418266e-06,
|
| 15954 |
+
"loss": 5.3347,
|
| 15955 |
+
"step": 9112
|
| 15956 |
+
},
|
| 15957 |
+
{
|
| 15958 |
+
"epoch": 0.88,
|
| 15959 |
+
"grad_norm": 1.0361077785491943,
|
| 15960 |
+
"learning_rate": 5.812487880550708e-06,
|
| 15961 |
+
"loss": 5.2899,
|
| 15962 |
+
"step": 9116
|
| 15963 |
+
},
|
| 15964 |
+
{
|
| 15965 |
+
"epoch": 0.88,
|
| 15966 |
+
"grad_norm": 1.027950406074524,
|
| 15967 |
+
"learning_rate": 5.793096761683149e-06,
|
| 15968 |
+
"loss": 5.3016,
|
| 15969 |
+
"step": 9120
|
| 15970 |
+
},
|
| 15971 |
+
{
|
| 15972 |
+
"epoch": 0.88,
|
| 15973 |
+
"grad_norm": 1.0341978073120117,
|
| 15974 |
+
"learning_rate": 5.7737056428155905e-06,
|
| 15975 |
+
"loss": 5.3066,
|
| 15976 |
+
"step": 9124
|
| 15977 |
+
},
|
| 15978 |
+
{
|
| 15979 |
+
"epoch": 0.89,
|
| 15980 |
+
"grad_norm": 1.0980345010757446,
|
| 15981 |
+
"learning_rate": 5.754314523948033e-06,
|
| 15982 |
+
"loss": 5.2685,
|
| 15983 |
+
"step": 9128
|
| 15984 |
+
},
|
| 15985 |
+
{
|
| 15986 |
+
"epoch": 0.89,
|
| 15987 |
+
"grad_norm": 1.0036580562591553,
|
| 15988 |
+
"learning_rate": 5.734923405080474e-06,
|
| 15989 |
+
"loss": 5.2972,
|
| 15990 |
+
"step": 9132
|
| 15991 |
+
},
|
| 15992 |
+
{
|
| 15993 |
+
"epoch": 0.89,
|
| 15994 |
+
"grad_norm": 1.0951625108718872,
|
| 15995 |
+
"learning_rate": 5.715532286212915e-06,
|
| 15996 |
+
"loss": 5.2363,
|
| 15997 |
+
"step": 9136
|
| 15998 |
+
},
|
| 15999 |
+
{
|
| 16000 |
+
"epoch": 0.89,
|
| 16001 |
+
"grad_norm": 1.010358214378357,
|
| 16002 |
+
"learning_rate": 5.696141167345356e-06,
|
| 16003 |
+
"loss": 5.2715,
|
| 16004 |
+
"step": 9140
|
| 16005 |
+
},
|
| 16006 |
+
{
|
| 16007 |
+
"epoch": 0.89,
|
| 16008 |
+
"grad_norm": 1.0721516609191895,
|
| 16009 |
+
"learning_rate": 5.676750048477797e-06,
|
| 16010 |
+
"loss": 5.3006,
|
| 16011 |
+
"step": 9144
|
| 16012 |
+
},
|
| 16013 |
+
{
|
| 16014 |
+
"epoch": 0.89,
|
| 16015 |
+
"grad_norm": 0.9922645688056946,
|
| 16016 |
+
"learning_rate": 5.657358929610239e-06,
|
| 16017 |
+
"loss": 5.2985,
|
| 16018 |
+
"step": 9148
|
| 16019 |
+
},
|
| 16020 |
+
{
|
| 16021 |
+
"epoch": 0.89,
|
| 16022 |
+
"grad_norm": 1.0958447456359863,
|
| 16023 |
+
"learning_rate": 5.63796781074268e-06,
|
| 16024 |
+
"loss": 5.2388,
|
| 16025 |
+
"step": 9152
|
| 16026 |
+
},
|
| 16027 |
+
{
|
| 16028 |
+
"epoch": 0.89,
|
| 16029 |
+
"grad_norm": 0.9977266788482666,
|
| 16030 |
+
"learning_rate": 5.6185766918751215e-06,
|
| 16031 |
+
"loss": 5.4273,
|
| 16032 |
+
"step": 9156
|
| 16033 |
+
},
|
| 16034 |
+
{
|
| 16035 |
+
"epoch": 0.89,
|
| 16036 |
+
"grad_norm": 1.1025915145874023,
|
| 16037 |
+
"learning_rate": 5.599185573007563e-06,
|
| 16038 |
+
"loss": 5.3768,
|
| 16039 |
+
"step": 9160
|
| 16040 |
+
},
|
| 16041 |
+
{
|
| 16042 |
+
"epoch": 0.89,
|
| 16043 |
+
"grad_norm": 1.0521866083145142,
|
| 16044 |
+
"learning_rate": 5.579794454140004e-06,
|
| 16045 |
+
"loss": 5.2555,
|
| 16046 |
+
"step": 9164
|
| 16047 |
+
},
|
| 16048 |
+
{
|
| 16049 |
+
"epoch": 0.89,
|
| 16050 |
+
"grad_norm": 1.0546320676803589,
|
| 16051 |
+
"learning_rate": 5.560403335272446e-06,
|
| 16052 |
+
"loss": 5.259,
|
| 16053 |
+
"step": 9168
|
| 16054 |
+
},
|
| 16055 |
+
{
|
| 16056 |
+
"epoch": 0.89,
|
| 16057 |
+
"grad_norm": 1.084153413772583,
|
| 16058 |
+
"learning_rate": 5.541012216404887e-06,
|
| 16059 |
+
"loss": 5.2733,
|
| 16060 |
+
"step": 9172
|
| 16061 |
+
},
|
| 16062 |
+
{
|
| 16063 |
+
"epoch": 0.89,
|
| 16064 |
+
"grad_norm": 1.1298420429229736,
|
| 16065 |
+
"learning_rate": 5.521621097537328e-06,
|
| 16066 |
+
"loss": 5.2499,
|
| 16067 |
+
"step": 9176
|
| 16068 |
+
},
|
| 16069 |
+
{
|
| 16070 |
+
"epoch": 0.89,
|
| 16071 |
+
"grad_norm": 1.0296047925949097,
|
| 16072 |
+
"learning_rate": 5.502229978669769e-06,
|
| 16073 |
+
"loss": 5.3293,
|
| 16074 |
+
"step": 9180
|
| 16075 |
+
},
|
| 16076 |
+
{
|
| 16077 |
+
"epoch": 0.89,
|
| 16078 |
+
"grad_norm": 1.0231281518936157,
|
| 16079 |
+
"learning_rate": 5.48283885980221e-06,
|
| 16080 |
+
"loss": 5.2614,
|
| 16081 |
+
"step": 9184
|
| 16082 |
+
},
|
| 16083 |
+
{
|
| 16084 |
+
"epoch": 0.89,
|
| 16085 |
+
"grad_norm": 1.1063759326934814,
|
| 16086 |
+
"learning_rate": 5.463447740934653e-06,
|
| 16087 |
+
"loss": 5.3569,
|
| 16088 |
+
"step": 9188
|
| 16089 |
+
},
|
| 16090 |
+
{
|
| 16091 |
+
"epoch": 0.89,
|
| 16092 |
+
"grad_norm": 1.0762827396392822,
|
| 16093 |
+
"learning_rate": 5.444056622067094e-06,
|
| 16094 |
+
"loss": 5.3734,
|
| 16095 |
+
"step": 9192
|
| 16096 |
+
},
|
| 16097 |
+
{
|
| 16098 |
+
"epoch": 0.89,
|
| 16099 |
+
"grad_norm": 1.0667394399642944,
|
| 16100 |
+
"learning_rate": 5.424665503199535e-06,
|
| 16101 |
+
"loss": 5.3136,
|
| 16102 |
+
"step": 9196
|
| 16103 |
+
},
|
| 16104 |
+
{
|
| 16105 |
+
"epoch": 0.89,
|
| 16106 |
+
"grad_norm": 1.0663567781448364,
|
| 16107 |
+
"learning_rate": 5.405274384331977e-06,
|
| 16108 |
+
"loss": 5.2178,
|
| 16109 |
+
"step": 9200
|
| 16110 |
+
},
|
| 16111 |
+
{
|
| 16112 |
+
"epoch": 0.89,
|
| 16113 |
+
"grad_norm": 1.1014020442962646,
|
| 16114 |
+
"learning_rate": 5.385883265464417e-06,
|
| 16115 |
+
"loss": 5.3121,
|
| 16116 |
+
"step": 9204
|
| 16117 |
+
},
|
| 16118 |
+
{
|
| 16119 |
+
"epoch": 0.89,
|
| 16120 |
+
"grad_norm": 1.0944761037826538,
|
| 16121 |
+
"learning_rate": 5.366492146596859e-06,
|
| 16122 |
+
"loss": 5.3321,
|
| 16123 |
+
"step": 9208
|
| 16124 |
+
},
|
| 16125 |
+
{
|
| 16126 |
+
"epoch": 0.89,
|
| 16127 |
+
"grad_norm": 1.0576825141906738,
|
| 16128 |
+
"learning_rate": 5.3471010277293e-06,
|
| 16129 |
+
"loss": 5.187,
|
| 16130 |
+
"step": 9212
|
| 16131 |
+
},
|
| 16132 |
+
{
|
| 16133 |
+
"epoch": 0.89,
|
| 16134 |
+
"grad_norm": 1.102414608001709,
|
| 16135 |
+
"learning_rate": 5.3277099088617414e-06,
|
| 16136 |
+
"loss": 5.2198,
|
| 16137 |
+
"step": 9216
|
| 16138 |
+
},
|
| 16139 |
+
{
|
| 16140 |
+
"epoch": 0.89,
|
| 16141 |
+
"grad_norm": 1.0515443086624146,
|
| 16142 |
+
"learning_rate": 5.308318789994183e-06,
|
| 16143 |
+
"loss": 5.3509,
|
| 16144 |
+
"step": 9220
|
| 16145 |
+
},
|
| 16146 |
+
{
|
| 16147 |
+
"epoch": 0.89,
|
| 16148 |
+
"grad_norm": 1.0225639343261719,
|
| 16149 |
+
"learning_rate": 5.288927671126624e-06,
|
| 16150 |
+
"loss": 5.3264,
|
| 16151 |
+
"step": 9224
|
| 16152 |
+
},
|
| 16153 |
+
{
|
| 16154 |
+
"epoch": 0.89,
|
| 16155 |
+
"grad_norm": 1.0482409000396729,
|
| 16156 |
+
"learning_rate": 5.269536552259066e-06,
|
| 16157 |
+
"loss": 5.314,
|
| 16158 |
+
"step": 9228
|
| 16159 |
+
},
|
| 16160 |
+
{
|
| 16161 |
+
"epoch": 0.9,
|
| 16162 |
+
"grad_norm": 1.0829366445541382,
|
| 16163 |
+
"learning_rate": 5.250145433391507e-06,
|
| 16164 |
+
"loss": 5.3288,
|
| 16165 |
+
"step": 9232
|
| 16166 |
+
},
|
| 16167 |
+
{
|
| 16168 |
+
"epoch": 0.9,
|
| 16169 |
+
"grad_norm": 1.018813967704773,
|
| 16170 |
+
"learning_rate": 5.230754314523948e-06,
|
| 16171 |
+
"loss": 5.245,
|
| 16172 |
+
"step": 9236
|
| 16173 |
+
},
|
| 16174 |
+
{
|
| 16175 |
+
"epoch": 0.9,
|
| 16176 |
+
"grad_norm": 1.1042280197143555,
|
| 16177 |
+
"learning_rate": 5.21136319565639e-06,
|
| 16178 |
+
"loss": 5.3871,
|
| 16179 |
+
"step": 9240
|
| 16180 |
+
},
|
| 16181 |
+
{
|
| 16182 |
+
"epoch": 0.9,
|
| 16183 |
+
"grad_norm": 1.0457403659820557,
|
| 16184 |
+
"learning_rate": 5.19197207678883e-06,
|
| 16185 |
+
"loss": 5.3858,
|
| 16186 |
+
"step": 9244
|
| 16187 |
+
},
|
| 16188 |
+
{
|
| 16189 |
+
"epoch": 0.9,
|
| 16190 |
+
"grad_norm": 1.0873547792434692,
|
| 16191 |
+
"learning_rate": 5.1725809579212725e-06,
|
| 16192 |
+
"loss": 5.2907,
|
| 16193 |
+
"step": 9248
|
| 16194 |
+
},
|
| 16195 |
+
{
|
| 16196 |
+
"epoch": 0.9,
|
| 16197 |
+
"grad_norm": 1.0615798234939575,
|
| 16198 |
+
"learning_rate": 5.153189839053714e-06,
|
| 16199 |
+
"loss": 5.2729,
|
| 16200 |
+
"step": 9252
|
| 16201 |
+
},
|
| 16202 |
+
{
|
| 16203 |
+
"epoch": 0.9,
|
| 16204 |
+
"grad_norm": 1.1086784601211548,
|
| 16205 |
+
"learning_rate": 5.133798720186155e-06,
|
| 16206 |
+
"loss": 5.388,
|
| 16207 |
+
"step": 9256
|
| 16208 |
+
},
|
| 16209 |
+
{
|
| 16210 |
+
"epoch": 0.9,
|
| 16211 |
+
"grad_norm": 1.0895764827728271,
|
| 16212 |
+
"learning_rate": 5.114407601318597e-06,
|
| 16213 |
+
"loss": 5.2892,
|
| 16214 |
+
"step": 9260
|
| 16215 |
+
},
|
| 16216 |
+
{
|
| 16217 |
+
"epoch": 0.9,
|
| 16218 |
+
"grad_norm": 1.0369954109191895,
|
| 16219 |
+
"learning_rate": 5.095016482451038e-06,
|
| 16220 |
+
"loss": 5.2989,
|
| 16221 |
+
"step": 9264
|
| 16222 |
+
},
|
| 16223 |
+
{
|
| 16224 |
+
"epoch": 0.9,
|
| 16225 |
+
"grad_norm": 1.049892544746399,
|
| 16226 |
+
"learning_rate": 5.0756253635834794e-06,
|
| 16227 |
+
"loss": 5.3419,
|
| 16228 |
+
"step": 9268
|
| 16229 |
+
},
|
| 16230 |
+
{
|
| 16231 |
+
"epoch": 0.9,
|
| 16232 |
+
"grad_norm": 1.0920242071151733,
|
| 16233 |
+
"learning_rate": 5.056234244715921e-06,
|
| 16234 |
+
"loss": 5.3116,
|
| 16235 |
+
"step": 9272
|
| 16236 |
+
},
|
| 16237 |
+
{
|
| 16238 |
+
"epoch": 0.9,
|
| 16239 |
+
"grad_norm": 1.0999177694320679,
|
| 16240 |
+
"learning_rate": 5.036843125848361e-06,
|
| 16241 |
+
"loss": 5.2779,
|
| 16242 |
+
"step": 9276
|
| 16243 |
+
},
|
| 16244 |
+
{
|
| 16245 |
+
"epoch": 0.9,
|
| 16246 |
+
"grad_norm": 1.0239474773406982,
|
| 16247 |
+
"learning_rate": 5.017452006980803e-06,
|
| 16248 |
+
"loss": 5.35,
|
| 16249 |
+
"step": 9280
|
| 16250 |
+
},
|
| 16251 |
+
{
|
| 16252 |
+
"epoch": 0.9,
|
| 16253 |
+
"grad_norm": 1.0571128129959106,
|
| 16254 |
+
"learning_rate": 4.998060888113244e-06,
|
| 16255 |
+
"loss": 5.3139,
|
| 16256 |
+
"step": 9284
|
| 16257 |
+
},
|
| 16258 |
+
{
|
| 16259 |
+
"epoch": 0.9,
|
| 16260 |
+
"grad_norm": 1.1082773208618164,
|
| 16261 |
+
"learning_rate": 4.9786697692456855e-06,
|
| 16262 |
+
"loss": 5.2779,
|
| 16263 |
+
"step": 9288
|
| 16264 |
}
|
| 16265 |
],
|
| 16266 |
"logging_steps": 4,
|
|
|
|
| 16268 |
"num_input_tokens_seen": 0,
|
| 16269 |
"num_train_epochs": 1,
|
| 16270 |
"save_steps": 1032,
|
| 16271 |
+
"total_flos": 7.822868224986317e+16,
|
| 16272 |
"train_batch_size": 8,
|
| 16273 |
"trial_name": null,
|
| 16274 |
"trial_params": null
|