Commit
•
c90ae0e
1
Parent(s):
294a57c
Training in progress, step 9099, 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 654946216
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9557bc9f6cbee3ba0a056668f0757dc066d4f66def7e43b0358cae0ee9de6e67
|
| 3 |
size 654946216
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1310000698
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e2e156708c40a75d36123ea5629230da104e836b70e9cff0007571ec0cbb6f6a
|
| 3 |
size 1310000698
|
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:f44032b1543a896a398ba18329beb5bf8b1f128caeeac11d091c47bd6a192c0e
|
| 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,
|
|
@@ -14161,6 +14161,1770 @@
|
|
| 14161 |
"learning_rate": 9.988127040664887e-06,
|
| 14162 |
"loss": 2.4894,
|
| 14163 |
"step": 8088
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 14164 |
}
|
| 14165 |
],
|
| 14166 |
"logging_steps": 4,
|
|
@@ -14168,7 +15932,7 @@
|
|
| 14168 |
"num_input_tokens_seen": 0,
|
| 14169 |
"num_train_epochs": 1,
|
| 14170 |
"save_steps": 1011,
|
| 14171 |
-
"total_flos":
|
| 14172 |
"train_batch_size": 4,
|
| 14173 |
"trial_name": null,
|
| 14174 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.9002671415850401,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 9099,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 14161 |
"learning_rate": 9.988127040664887e-06,
|
| 14162 |
"loss": 2.4894,
|
| 14163 |
"step": 8088
|
| 14164 |
+
},
|
| 14165 |
+
{
|
| 14166 |
+
"epoch": 0.8,
|
| 14167 |
+
"grad_norm": 1.6651523113250732,
|
| 14168 |
+
"learning_rate": 9.968338775106362e-06,
|
| 14169 |
+
"loss": 2.5608,
|
| 14170 |
+
"step": 8092
|
| 14171 |
+
},
|
| 14172 |
+
{
|
| 14173 |
+
"epoch": 0.8,
|
| 14174 |
+
"grad_norm": 1.5435396432876587,
|
| 14175 |
+
"learning_rate": 9.948550509547838e-06,
|
| 14176 |
+
"loss": 2.5702,
|
| 14177 |
+
"step": 8096
|
| 14178 |
+
},
|
| 14179 |
+
{
|
| 14180 |
+
"epoch": 0.8,
|
| 14181 |
+
"grad_norm": 1.6652929782867432,
|
| 14182 |
+
"learning_rate": 9.928762243989314e-06,
|
| 14183 |
+
"loss": 2.5155,
|
| 14184 |
+
"step": 8100
|
| 14185 |
+
},
|
| 14186 |
+
{
|
| 14187 |
+
"epoch": 0.8,
|
| 14188 |
+
"grad_norm": 1.5997744798660278,
|
| 14189 |
+
"learning_rate": 9.90897397843079e-06,
|
| 14190 |
+
"loss": 2.412,
|
| 14191 |
+
"step": 8104
|
| 14192 |
+
},
|
| 14193 |
+
{
|
| 14194 |
+
"epoch": 0.8,
|
| 14195 |
+
"grad_norm": 1.518398404121399,
|
| 14196 |
+
"learning_rate": 9.889185712872267e-06,
|
| 14197 |
+
"loss": 2.4749,
|
| 14198 |
+
"step": 8108
|
| 14199 |
+
},
|
| 14200 |
+
{
|
| 14201 |
+
"epoch": 0.8,
|
| 14202 |
+
"grad_norm": 1.6166468858718872,
|
| 14203 |
+
"learning_rate": 9.869397447313745e-06,
|
| 14204 |
+
"loss": 2.528,
|
| 14205 |
+
"step": 8112
|
| 14206 |
+
},
|
| 14207 |
+
{
|
| 14208 |
+
"epoch": 0.8,
|
| 14209 |
+
"grad_norm": 1.7324260473251343,
|
| 14210 |
+
"learning_rate": 9.84960918175522e-06,
|
| 14211 |
+
"loss": 2.4854,
|
| 14212 |
+
"step": 8116
|
| 14213 |
+
},
|
| 14214 |
+
{
|
| 14215 |
+
"epoch": 0.8,
|
| 14216 |
+
"grad_norm": 1.6371716260910034,
|
| 14217 |
+
"learning_rate": 9.829820916196696e-06,
|
| 14218 |
+
"loss": 2.5037,
|
| 14219 |
+
"step": 8120
|
| 14220 |
+
},
|
| 14221 |
+
{
|
| 14222 |
+
"epoch": 0.8,
|
| 14223 |
+
"grad_norm": 1.711093544960022,
|
| 14224 |
+
"learning_rate": 9.810032650638172e-06,
|
| 14225 |
+
"loss": 2.6155,
|
| 14226 |
+
"step": 8124
|
| 14227 |
+
},
|
| 14228 |
+
{
|
| 14229 |
+
"epoch": 0.8,
|
| 14230 |
+
"grad_norm": 1.6455191373825073,
|
| 14231 |
+
"learning_rate": 9.790244385079648e-06,
|
| 14232 |
+
"loss": 2.4078,
|
| 14233 |
+
"step": 8128
|
| 14234 |
+
},
|
| 14235 |
+
{
|
| 14236 |
+
"epoch": 0.8,
|
| 14237 |
+
"grad_norm": 1.6366294622421265,
|
| 14238 |
+
"learning_rate": 9.770456119521125e-06,
|
| 14239 |
+
"loss": 2.4642,
|
| 14240 |
+
"step": 8132
|
| 14241 |
+
},
|
| 14242 |
+
{
|
| 14243 |
+
"epoch": 0.8,
|
| 14244 |
+
"grad_norm": 1.668990969657898,
|
| 14245 |
+
"learning_rate": 9.7506678539626e-06,
|
| 14246 |
+
"loss": 2.5074,
|
| 14247 |
+
"step": 8136
|
| 14248 |
+
},
|
| 14249 |
+
{
|
| 14250 |
+
"epoch": 0.81,
|
| 14251 |
+
"grad_norm": 1.5362697839736938,
|
| 14252 |
+
"learning_rate": 9.730879588404076e-06,
|
| 14253 |
+
"loss": 2.4896,
|
| 14254 |
+
"step": 8140
|
| 14255 |
+
},
|
| 14256 |
+
{
|
| 14257 |
+
"epoch": 0.81,
|
| 14258 |
+
"grad_norm": 1.8022223711013794,
|
| 14259 |
+
"learning_rate": 9.711091322845554e-06,
|
| 14260 |
+
"loss": 2.5353,
|
| 14261 |
+
"step": 8144
|
| 14262 |
+
},
|
| 14263 |
+
{
|
| 14264 |
+
"epoch": 0.81,
|
| 14265 |
+
"grad_norm": 1.6878408193588257,
|
| 14266 |
+
"learning_rate": 9.69130305728703e-06,
|
| 14267 |
+
"loss": 2.5586,
|
| 14268 |
+
"step": 8148
|
| 14269 |
+
},
|
| 14270 |
+
{
|
| 14271 |
+
"epoch": 0.81,
|
| 14272 |
+
"grad_norm": 1.689186692237854,
|
| 14273 |
+
"learning_rate": 9.671514791728506e-06,
|
| 14274 |
+
"loss": 2.655,
|
| 14275 |
+
"step": 8152
|
| 14276 |
+
},
|
| 14277 |
+
{
|
| 14278 |
+
"epoch": 0.81,
|
| 14279 |
+
"grad_norm": 1.6996088027954102,
|
| 14280 |
+
"learning_rate": 9.651726526169981e-06,
|
| 14281 |
+
"loss": 2.7126,
|
| 14282 |
+
"step": 8156
|
| 14283 |
+
},
|
| 14284 |
+
{
|
| 14285 |
+
"epoch": 0.81,
|
| 14286 |
+
"grad_norm": 2.8080925941467285,
|
| 14287 |
+
"learning_rate": 9.631938260611457e-06,
|
| 14288 |
+
"loss": 2.7099,
|
| 14289 |
+
"step": 8160
|
| 14290 |
+
},
|
| 14291 |
+
{
|
| 14292 |
+
"epoch": 0.81,
|
| 14293 |
+
"grad_norm": 1.7663508653640747,
|
| 14294 |
+
"learning_rate": 9.612149995052934e-06,
|
| 14295 |
+
"loss": 2.4897,
|
| 14296 |
+
"step": 8164
|
| 14297 |
+
},
|
| 14298 |
+
{
|
| 14299 |
+
"epoch": 0.81,
|
| 14300 |
+
"grad_norm": 1.7868987321853638,
|
| 14301 |
+
"learning_rate": 9.59236172949441e-06,
|
| 14302 |
+
"loss": 2.4844,
|
| 14303 |
+
"step": 8168
|
| 14304 |
+
},
|
| 14305 |
+
{
|
| 14306 |
+
"epoch": 0.81,
|
| 14307 |
+
"grad_norm": 1.6591540575027466,
|
| 14308 |
+
"learning_rate": 9.572573463935886e-06,
|
| 14309 |
+
"loss": 2.4604,
|
| 14310 |
+
"step": 8172
|
| 14311 |
+
},
|
| 14312 |
+
{
|
| 14313 |
+
"epoch": 0.81,
|
| 14314 |
+
"grad_norm": 1.7677662372589111,
|
| 14315 |
+
"learning_rate": 9.552785198377363e-06,
|
| 14316 |
+
"loss": 2.5354,
|
| 14317 |
+
"step": 8176
|
| 14318 |
+
},
|
| 14319 |
+
{
|
| 14320 |
+
"epoch": 0.81,
|
| 14321 |
+
"grad_norm": 1.6505751609802246,
|
| 14322 |
+
"learning_rate": 9.532996932818839e-06,
|
| 14323 |
+
"loss": 2.5834,
|
| 14324 |
+
"step": 8180
|
| 14325 |
+
},
|
| 14326 |
+
{
|
| 14327 |
+
"epoch": 0.81,
|
| 14328 |
+
"grad_norm": 1.7923264503479004,
|
| 14329 |
+
"learning_rate": 9.513208667260315e-06,
|
| 14330 |
+
"loss": 2.5445,
|
| 14331 |
+
"step": 8184
|
| 14332 |
+
},
|
| 14333 |
+
{
|
| 14334 |
+
"epoch": 0.81,
|
| 14335 |
+
"grad_norm": 1.6527979373931885,
|
| 14336 |
+
"learning_rate": 9.493420401701792e-06,
|
| 14337 |
+
"loss": 2.3743,
|
| 14338 |
+
"step": 8188
|
| 14339 |
+
},
|
| 14340 |
+
{
|
| 14341 |
+
"epoch": 0.81,
|
| 14342 |
+
"grad_norm": 1.490796446800232,
|
| 14343 |
+
"learning_rate": 9.473632136143268e-06,
|
| 14344 |
+
"loss": 2.4511,
|
| 14345 |
+
"step": 8192
|
| 14346 |
+
},
|
| 14347 |
+
{
|
| 14348 |
+
"epoch": 0.81,
|
| 14349 |
+
"grad_norm": 1.6742923259735107,
|
| 14350 |
+
"learning_rate": 9.453843870584744e-06,
|
| 14351 |
+
"loss": 2.5265,
|
| 14352 |
+
"step": 8196
|
| 14353 |
+
},
|
| 14354 |
+
{
|
| 14355 |
+
"epoch": 0.81,
|
| 14356 |
+
"grad_norm": 1.6631675958633423,
|
| 14357 |
+
"learning_rate": 9.434055605026219e-06,
|
| 14358 |
+
"loss": 2.3355,
|
| 14359 |
+
"step": 8200
|
| 14360 |
+
},
|
| 14361 |
+
{
|
| 14362 |
+
"epoch": 0.81,
|
| 14363 |
+
"grad_norm": 1.9126391410827637,
|
| 14364 |
+
"learning_rate": 9.414267339467695e-06,
|
| 14365 |
+
"loss": 2.5204,
|
| 14366 |
+
"step": 8204
|
| 14367 |
+
},
|
| 14368 |
+
{
|
| 14369 |
+
"epoch": 0.81,
|
| 14370 |
+
"grad_norm": 1.6832706928253174,
|
| 14371 |
+
"learning_rate": 9.394479073909172e-06,
|
| 14372 |
+
"loss": 2.541,
|
| 14373 |
+
"step": 8208
|
| 14374 |
+
},
|
| 14375 |
+
{
|
| 14376 |
+
"epoch": 0.81,
|
| 14377 |
+
"grad_norm": 1.8622807264328003,
|
| 14378 |
+
"learning_rate": 9.37469080835065e-06,
|
| 14379 |
+
"loss": 2.6101,
|
| 14380 |
+
"step": 8212
|
| 14381 |
+
},
|
| 14382 |
+
{
|
| 14383 |
+
"epoch": 0.81,
|
| 14384 |
+
"grad_norm": 1.7272640466690063,
|
| 14385 |
+
"learning_rate": 9.354902542792126e-06,
|
| 14386 |
+
"loss": 2.7245,
|
| 14387 |
+
"step": 8216
|
| 14388 |
+
},
|
| 14389 |
+
{
|
| 14390 |
+
"epoch": 0.81,
|
| 14391 |
+
"grad_norm": 1.6699191331863403,
|
| 14392 |
+
"learning_rate": 9.3351142772336e-06,
|
| 14393 |
+
"loss": 2.4707,
|
| 14394 |
+
"step": 8220
|
| 14395 |
+
},
|
| 14396 |
+
{
|
| 14397 |
+
"epoch": 0.81,
|
| 14398 |
+
"grad_norm": 1.599859595298767,
|
| 14399 |
+
"learning_rate": 9.315326011675077e-06,
|
| 14400 |
+
"loss": 2.5186,
|
| 14401 |
+
"step": 8224
|
| 14402 |
+
},
|
| 14403 |
+
{
|
| 14404 |
+
"epoch": 0.81,
|
| 14405 |
+
"grad_norm": 1.6403051614761353,
|
| 14406 |
+
"learning_rate": 9.295537746116553e-06,
|
| 14407 |
+
"loss": 2.6446,
|
| 14408 |
+
"step": 8228
|
| 14409 |
+
},
|
| 14410 |
+
{
|
| 14411 |
+
"epoch": 0.81,
|
| 14412 |
+
"grad_norm": 1.6128116846084595,
|
| 14413 |
+
"learning_rate": 9.27574948055803e-06,
|
| 14414 |
+
"loss": 2.6126,
|
| 14415 |
+
"step": 8232
|
| 14416 |
+
},
|
| 14417 |
+
{
|
| 14418 |
+
"epoch": 0.81,
|
| 14419 |
+
"grad_norm": 1.588544487953186,
|
| 14420 |
+
"learning_rate": 9.255961214999506e-06,
|
| 14421 |
+
"loss": 2.6034,
|
| 14422 |
+
"step": 8236
|
| 14423 |
+
},
|
| 14424 |
+
{
|
| 14425 |
+
"epoch": 0.82,
|
| 14426 |
+
"grad_norm": 1.7381192445755005,
|
| 14427 |
+
"learning_rate": 9.236172949440982e-06,
|
| 14428 |
+
"loss": 2.6789,
|
| 14429 |
+
"step": 8240
|
| 14430 |
+
},
|
| 14431 |
+
{
|
| 14432 |
+
"epoch": 0.82,
|
| 14433 |
+
"grad_norm": 1.6580698490142822,
|
| 14434 |
+
"learning_rate": 9.216384683882458e-06,
|
| 14435 |
+
"loss": 2.6275,
|
| 14436 |
+
"step": 8244
|
| 14437 |
+
},
|
| 14438 |
+
{
|
| 14439 |
+
"epoch": 0.82,
|
| 14440 |
+
"grad_norm": 1.779470443725586,
|
| 14441 |
+
"learning_rate": 9.196596418323935e-06,
|
| 14442 |
+
"loss": 2.6094,
|
| 14443 |
+
"step": 8248
|
| 14444 |
+
},
|
| 14445 |
+
{
|
| 14446 |
+
"epoch": 0.82,
|
| 14447 |
+
"grad_norm": 1.6793197393417358,
|
| 14448 |
+
"learning_rate": 9.176808152765411e-06,
|
| 14449 |
+
"loss": 2.3446,
|
| 14450 |
+
"step": 8252
|
| 14451 |
+
},
|
| 14452 |
+
{
|
| 14453 |
+
"epoch": 0.82,
|
| 14454 |
+
"grad_norm": 1.7751526832580566,
|
| 14455 |
+
"learning_rate": 9.157019887206887e-06,
|
| 14456 |
+
"loss": 2.5427,
|
| 14457 |
+
"step": 8256
|
| 14458 |
+
},
|
| 14459 |
+
{
|
| 14460 |
+
"epoch": 0.82,
|
| 14461 |
+
"grad_norm": 1.687898874282837,
|
| 14462 |
+
"learning_rate": 9.137231621648364e-06,
|
| 14463 |
+
"loss": 2.5313,
|
| 14464 |
+
"step": 8260
|
| 14465 |
+
},
|
| 14466 |
+
{
|
| 14467 |
+
"epoch": 0.82,
|
| 14468 |
+
"grad_norm": 1.5011862516403198,
|
| 14469 |
+
"learning_rate": 9.117443356089838e-06,
|
| 14470 |
+
"loss": 2.4652,
|
| 14471 |
+
"step": 8264
|
| 14472 |
+
},
|
| 14473 |
+
{
|
| 14474 |
+
"epoch": 0.82,
|
| 14475 |
+
"grad_norm": 1.7039753198623657,
|
| 14476 |
+
"learning_rate": 9.097655090531315e-06,
|
| 14477 |
+
"loss": 2.5125,
|
| 14478 |
+
"step": 8268
|
| 14479 |
+
},
|
| 14480 |
+
{
|
| 14481 |
+
"epoch": 0.82,
|
| 14482 |
+
"grad_norm": 1.7907326221466064,
|
| 14483 |
+
"learning_rate": 9.077866824972791e-06,
|
| 14484 |
+
"loss": 2.4032,
|
| 14485 |
+
"step": 8272
|
| 14486 |
+
},
|
| 14487 |
+
{
|
| 14488 |
+
"epoch": 0.82,
|
| 14489 |
+
"grad_norm": 1.689347267150879,
|
| 14490 |
+
"learning_rate": 9.058078559414267e-06,
|
| 14491 |
+
"loss": 2.4636,
|
| 14492 |
+
"step": 8276
|
| 14493 |
+
},
|
| 14494 |
+
{
|
| 14495 |
+
"epoch": 0.82,
|
| 14496 |
+
"grad_norm": 1.7220097780227661,
|
| 14497 |
+
"learning_rate": 9.038290293855745e-06,
|
| 14498 |
+
"loss": 2.4757,
|
| 14499 |
+
"step": 8280
|
| 14500 |
+
},
|
| 14501 |
+
{
|
| 14502 |
+
"epoch": 0.82,
|
| 14503 |
+
"grad_norm": 1.8284951448440552,
|
| 14504 |
+
"learning_rate": 9.01850202829722e-06,
|
| 14505 |
+
"loss": 2.473,
|
| 14506 |
+
"step": 8284
|
| 14507 |
+
},
|
| 14508 |
+
{
|
| 14509 |
+
"epoch": 0.82,
|
| 14510 |
+
"grad_norm": 1.7803459167480469,
|
| 14511 |
+
"learning_rate": 8.998713762738696e-06,
|
| 14512 |
+
"loss": 2.4781,
|
| 14513 |
+
"step": 8288
|
| 14514 |
+
},
|
| 14515 |
+
{
|
| 14516 |
+
"epoch": 0.82,
|
| 14517 |
+
"grad_norm": 1.6893064975738525,
|
| 14518 |
+
"learning_rate": 8.978925497180173e-06,
|
| 14519 |
+
"loss": 2.5115,
|
| 14520 |
+
"step": 8292
|
| 14521 |
+
},
|
| 14522 |
+
{
|
| 14523 |
+
"epoch": 0.82,
|
| 14524 |
+
"grad_norm": 1.8316203355789185,
|
| 14525 |
+
"learning_rate": 8.959137231621649e-06,
|
| 14526 |
+
"loss": 2.5759,
|
| 14527 |
+
"step": 8296
|
| 14528 |
+
},
|
| 14529 |
+
{
|
| 14530 |
+
"epoch": 0.82,
|
| 14531 |
+
"grad_norm": 1.6714107990264893,
|
| 14532 |
+
"learning_rate": 8.939348966063125e-06,
|
| 14533 |
+
"loss": 2.5676,
|
| 14534 |
+
"step": 8300
|
| 14535 |
+
},
|
| 14536 |
+
{
|
| 14537 |
+
"epoch": 0.82,
|
| 14538 |
+
"grad_norm": 1.5850313901901245,
|
| 14539 |
+
"learning_rate": 8.9195607005046e-06,
|
| 14540 |
+
"loss": 2.3758,
|
| 14541 |
+
"step": 8304
|
| 14542 |
+
},
|
| 14543 |
+
{
|
| 14544 |
+
"epoch": 0.82,
|
| 14545 |
+
"grad_norm": 1.595004677772522,
|
| 14546 |
+
"learning_rate": 8.899772434946078e-06,
|
| 14547 |
+
"loss": 2.4782,
|
| 14548 |
+
"step": 8308
|
| 14549 |
+
},
|
| 14550 |
+
{
|
| 14551 |
+
"epoch": 0.82,
|
| 14552 |
+
"grad_norm": 1.756579041481018,
|
| 14553 |
+
"learning_rate": 8.879984169387554e-06,
|
| 14554 |
+
"loss": 2.4742,
|
| 14555 |
+
"step": 8312
|
| 14556 |
+
},
|
| 14557 |
+
{
|
| 14558 |
+
"epoch": 0.82,
|
| 14559 |
+
"grad_norm": 1.601770043373108,
|
| 14560 |
+
"learning_rate": 8.86019590382903e-06,
|
| 14561 |
+
"loss": 2.3767,
|
| 14562 |
+
"step": 8316
|
| 14563 |
+
},
|
| 14564 |
+
{
|
| 14565 |
+
"epoch": 0.82,
|
| 14566 |
+
"grad_norm": 1.8463127613067627,
|
| 14567 |
+
"learning_rate": 8.840407638270507e-06,
|
| 14568 |
+
"loss": 2.5466,
|
| 14569 |
+
"step": 8320
|
| 14570 |
+
},
|
| 14571 |
+
{
|
| 14572 |
+
"epoch": 0.82,
|
| 14573 |
+
"grad_norm": 1.8509941101074219,
|
| 14574 |
+
"learning_rate": 8.820619372711983e-06,
|
| 14575 |
+
"loss": 2.606,
|
| 14576 |
+
"step": 8324
|
| 14577 |
+
},
|
| 14578 |
+
{
|
| 14579 |
+
"epoch": 0.82,
|
| 14580 |
+
"grad_norm": 1.6294124126434326,
|
| 14581 |
+
"learning_rate": 8.800831107153458e-06,
|
| 14582 |
+
"loss": 2.4114,
|
| 14583 |
+
"step": 8328
|
| 14584 |
+
},
|
| 14585 |
+
{
|
| 14586 |
+
"epoch": 0.82,
|
| 14587 |
+
"grad_norm": 1.5632402896881104,
|
| 14588 |
+
"learning_rate": 8.781042841594934e-06,
|
| 14589 |
+
"loss": 2.3389,
|
| 14590 |
+
"step": 8332
|
| 14591 |
+
},
|
| 14592 |
+
{
|
| 14593 |
+
"epoch": 0.82,
|
| 14594 |
+
"grad_norm": 1.602977991104126,
|
| 14595 |
+
"learning_rate": 8.76125457603641e-06,
|
| 14596 |
+
"loss": 2.5443,
|
| 14597 |
+
"step": 8336
|
| 14598 |
+
},
|
| 14599 |
+
{
|
| 14600 |
+
"epoch": 0.83,
|
| 14601 |
+
"grad_norm": 1.8994452953338623,
|
| 14602 |
+
"learning_rate": 8.741466310477887e-06,
|
| 14603 |
+
"loss": 2.4973,
|
| 14604 |
+
"step": 8340
|
| 14605 |
+
},
|
| 14606 |
+
{
|
| 14607 |
+
"epoch": 0.83,
|
| 14608 |
+
"grad_norm": 1.703715205192566,
|
| 14609 |
+
"learning_rate": 8.721678044919363e-06,
|
| 14610 |
+
"loss": 2.3838,
|
| 14611 |
+
"step": 8344
|
| 14612 |
+
},
|
| 14613 |
+
{
|
| 14614 |
+
"epoch": 0.83,
|
| 14615 |
+
"grad_norm": 1.7744590044021606,
|
| 14616 |
+
"learning_rate": 8.70188977936084e-06,
|
| 14617 |
+
"loss": 2.2877,
|
| 14618 |
+
"step": 8348
|
| 14619 |
+
},
|
| 14620 |
+
{
|
| 14621 |
+
"epoch": 0.83,
|
| 14622 |
+
"grad_norm": 1.7415697574615479,
|
| 14623 |
+
"learning_rate": 8.682101513802316e-06,
|
| 14624 |
+
"loss": 2.5678,
|
| 14625 |
+
"step": 8352
|
| 14626 |
+
},
|
| 14627 |
+
{
|
| 14628 |
+
"epoch": 0.83,
|
| 14629 |
+
"grad_norm": 1.6912627220153809,
|
| 14630 |
+
"learning_rate": 8.662313248243792e-06,
|
| 14631 |
+
"loss": 2.3746,
|
| 14632 |
+
"step": 8356
|
| 14633 |
+
},
|
| 14634 |
+
{
|
| 14635 |
+
"epoch": 0.83,
|
| 14636 |
+
"grad_norm": 1.591155767440796,
|
| 14637 |
+
"learning_rate": 8.642524982685268e-06,
|
| 14638 |
+
"loss": 2.4637,
|
| 14639 |
+
"step": 8360
|
| 14640 |
+
},
|
| 14641 |
+
{
|
| 14642 |
+
"epoch": 0.83,
|
| 14643 |
+
"grad_norm": 1.625097393989563,
|
| 14644 |
+
"learning_rate": 8.622736717126745e-06,
|
| 14645 |
+
"loss": 2.4071,
|
| 14646 |
+
"step": 8364
|
| 14647 |
+
},
|
| 14648 |
+
{
|
| 14649 |
+
"epoch": 0.83,
|
| 14650 |
+
"grad_norm": 1.7765227556228638,
|
| 14651 |
+
"learning_rate": 8.60294845156822e-06,
|
| 14652 |
+
"loss": 2.4339,
|
| 14653 |
+
"step": 8368
|
| 14654 |
+
},
|
| 14655 |
+
{
|
| 14656 |
+
"epoch": 0.83,
|
| 14657 |
+
"grad_norm": 1.7714086771011353,
|
| 14658 |
+
"learning_rate": 8.583160186009696e-06,
|
| 14659 |
+
"loss": 2.4657,
|
| 14660 |
+
"step": 8372
|
| 14661 |
+
},
|
| 14662 |
+
{
|
| 14663 |
+
"epoch": 0.83,
|
| 14664 |
+
"grad_norm": 1.5393376350402832,
|
| 14665 |
+
"learning_rate": 8.563371920451174e-06,
|
| 14666 |
+
"loss": 2.4295,
|
| 14667 |
+
"step": 8376
|
| 14668 |
+
},
|
| 14669 |
+
{
|
| 14670 |
+
"epoch": 0.83,
|
| 14671 |
+
"grad_norm": 1.6363029479980469,
|
| 14672 |
+
"learning_rate": 8.54358365489265e-06,
|
| 14673 |
+
"loss": 2.3262,
|
| 14674 |
+
"step": 8380
|
| 14675 |
+
},
|
| 14676 |
+
{
|
| 14677 |
+
"epoch": 0.83,
|
| 14678 |
+
"grad_norm": 1.6590903997421265,
|
| 14679 |
+
"learning_rate": 8.523795389334126e-06,
|
| 14680 |
+
"loss": 2.5488,
|
| 14681 |
+
"step": 8384
|
| 14682 |
+
},
|
| 14683 |
+
{
|
| 14684 |
+
"epoch": 0.83,
|
| 14685 |
+
"grad_norm": 2.057309150695801,
|
| 14686 |
+
"learning_rate": 8.504007123775601e-06,
|
| 14687 |
+
"loss": 2.4571,
|
| 14688 |
+
"step": 8388
|
| 14689 |
+
},
|
| 14690 |
+
{
|
| 14691 |
+
"epoch": 0.83,
|
| 14692 |
+
"grad_norm": 1.643871784210205,
|
| 14693 |
+
"learning_rate": 8.484218858217077e-06,
|
| 14694 |
+
"loss": 2.5755,
|
| 14695 |
+
"step": 8392
|
| 14696 |
+
},
|
| 14697 |
+
{
|
| 14698 |
+
"epoch": 0.83,
|
| 14699 |
+
"grad_norm": 1.5784872770309448,
|
| 14700 |
+
"learning_rate": 8.464430592658554e-06,
|
| 14701 |
+
"loss": 2.3636,
|
| 14702 |
+
"step": 8396
|
| 14703 |
+
},
|
| 14704 |
+
{
|
| 14705 |
+
"epoch": 0.83,
|
| 14706 |
+
"grad_norm": 1.5051548480987549,
|
| 14707 |
+
"learning_rate": 8.44464232710003e-06,
|
| 14708 |
+
"loss": 2.3239,
|
| 14709 |
+
"step": 8400
|
| 14710 |
+
},
|
| 14711 |
+
{
|
| 14712 |
+
"epoch": 0.83,
|
| 14713 |
+
"grad_norm": 1.7668542861938477,
|
| 14714 |
+
"learning_rate": 8.424854061541506e-06,
|
| 14715 |
+
"loss": 2.4491,
|
| 14716 |
+
"step": 8404
|
| 14717 |
+
},
|
| 14718 |
+
{
|
| 14719 |
+
"epoch": 0.83,
|
| 14720 |
+
"grad_norm": 1.7257317304611206,
|
| 14721 |
+
"learning_rate": 8.405065795982982e-06,
|
| 14722 |
+
"loss": 2.4014,
|
| 14723 |
+
"step": 8408
|
| 14724 |
+
},
|
| 14725 |
+
{
|
| 14726 |
+
"epoch": 0.83,
|
| 14727 |
+
"grad_norm": 1.6491328477859497,
|
| 14728 |
+
"learning_rate": 8.385277530424459e-06,
|
| 14729 |
+
"loss": 2.5093,
|
| 14730 |
+
"step": 8412
|
| 14731 |
+
},
|
| 14732 |
+
{
|
| 14733 |
+
"epoch": 0.83,
|
| 14734 |
+
"grad_norm": 1.5918080806732178,
|
| 14735 |
+
"learning_rate": 8.365489264865935e-06,
|
| 14736 |
+
"loss": 2.7333,
|
| 14737 |
+
"step": 8416
|
| 14738 |
+
},
|
| 14739 |
+
{
|
| 14740 |
+
"epoch": 0.83,
|
| 14741 |
+
"grad_norm": 1.709262490272522,
|
| 14742 |
+
"learning_rate": 8.345700999307411e-06,
|
| 14743 |
+
"loss": 2.3297,
|
| 14744 |
+
"step": 8420
|
| 14745 |
+
},
|
| 14746 |
+
{
|
| 14747 |
+
"epoch": 0.83,
|
| 14748 |
+
"grad_norm": 1.5516706705093384,
|
| 14749 |
+
"learning_rate": 8.325912733748888e-06,
|
| 14750 |
+
"loss": 2.3839,
|
| 14751 |
+
"step": 8424
|
| 14752 |
+
},
|
| 14753 |
+
{
|
| 14754 |
+
"epoch": 0.83,
|
| 14755 |
+
"grad_norm": 1.711029291152954,
|
| 14756 |
+
"learning_rate": 8.306124468190364e-06,
|
| 14757 |
+
"loss": 2.4985,
|
| 14758 |
+
"step": 8428
|
| 14759 |
+
},
|
| 14760 |
+
{
|
| 14761 |
+
"epoch": 0.83,
|
| 14762 |
+
"grad_norm": 1.5779407024383545,
|
| 14763 |
+
"learning_rate": 8.286336202631839e-06,
|
| 14764 |
+
"loss": 2.4944,
|
| 14765 |
+
"step": 8432
|
| 14766 |
+
},
|
| 14767 |
+
{
|
| 14768 |
+
"epoch": 0.83,
|
| 14769 |
+
"grad_norm": 1.8659144639968872,
|
| 14770 |
+
"learning_rate": 8.266547937073315e-06,
|
| 14771 |
+
"loss": 2.4244,
|
| 14772 |
+
"step": 8436
|
| 14773 |
+
},
|
| 14774 |
+
{
|
| 14775 |
+
"epoch": 0.84,
|
| 14776 |
+
"grad_norm": 1.93182373046875,
|
| 14777 |
+
"learning_rate": 8.246759671514791e-06,
|
| 14778 |
+
"loss": 2.4782,
|
| 14779 |
+
"step": 8440
|
| 14780 |
+
},
|
| 14781 |
+
{
|
| 14782 |
+
"epoch": 0.84,
|
| 14783 |
+
"grad_norm": 1.5297750234603882,
|
| 14784 |
+
"learning_rate": 8.22697140595627e-06,
|
| 14785 |
+
"loss": 2.5646,
|
| 14786 |
+
"step": 8444
|
| 14787 |
+
},
|
| 14788 |
+
{
|
| 14789 |
+
"epoch": 0.84,
|
| 14790 |
+
"grad_norm": 1.7578099966049194,
|
| 14791 |
+
"learning_rate": 8.207183140397746e-06,
|
| 14792 |
+
"loss": 2.5986,
|
| 14793 |
+
"step": 8448
|
| 14794 |
+
},
|
| 14795 |
+
{
|
| 14796 |
+
"epoch": 0.84,
|
| 14797 |
+
"grad_norm": 1.7178421020507812,
|
| 14798 |
+
"learning_rate": 8.18739487483922e-06,
|
| 14799 |
+
"loss": 2.4336,
|
| 14800 |
+
"step": 8452
|
| 14801 |
+
},
|
| 14802 |
+
{
|
| 14803 |
+
"epoch": 0.84,
|
| 14804 |
+
"grad_norm": 1.6756671667099,
|
| 14805 |
+
"learning_rate": 8.167606609280697e-06,
|
| 14806 |
+
"loss": 2.5326,
|
| 14807 |
+
"step": 8456
|
| 14808 |
+
},
|
| 14809 |
+
{
|
| 14810 |
+
"epoch": 0.84,
|
| 14811 |
+
"grad_norm": 1.681978464126587,
|
| 14812 |
+
"learning_rate": 8.147818343722173e-06,
|
| 14813 |
+
"loss": 2.4505,
|
| 14814 |
+
"step": 8460
|
| 14815 |
+
},
|
| 14816 |
+
{
|
| 14817 |
+
"epoch": 0.84,
|
| 14818 |
+
"grad_norm": 1.777003288269043,
|
| 14819 |
+
"learning_rate": 8.12803007816365e-06,
|
| 14820 |
+
"loss": 2.4595,
|
| 14821 |
+
"step": 8464
|
| 14822 |
+
},
|
| 14823 |
+
{
|
| 14824 |
+
"epoch": 0.84,
|
| 14825 |
+
"grad_norm": 1.7001906633377075,
|
| 14826 |
+
"learning_rate": 8.108241812605126e-06,
|
| 14827 |
+
"loss": 2.3165,
|
| 14828 |
+
"step": 8468
|
| 14829 |
+
},
|
| 14830 |
+
{
|
| 14831 |
+
"epoch": 0.84,
|
| 14832 |
+
"grad_norm": 1.6168255805969238,
|
| 14833 |
+
"learning_rate": 8.0884535470466e-06,
|
| 14834 |
+
"loss": 2.4513,
|
| 14835 |
+
"step": 8472
|
| 14836 |
+
},
|
| 14837 |
+
{
|
| 14838 |
+
"epoch": 0.84,
|
| 14839 |
+
"grad_norm": 1.7238038778305054,
|
| 14840 |
+
"learning_rate": 8.068665281488078e-06,
|
| 14841 |
+
"loss": 2.5386,
|
| 14842 |
+
"step": 8476
|
| 14843 |
+
},
|
| 14844 |
+
{
|
| 14845 |
+
"epoch": 0.84,
|
| 14846 |
+
"grad_norm": 1.7556039094924927,
|
| 14847 |
+
"learning_rate": 8.048877015929555e-06,
|
| 14848 |
+
"loss": 2.5683,
|
| 14849 |
+
"step": 8480
|
| 14850 |
+
},
|
| 14851 |
+
{
|
| 14852 |
+
"epoch": 0.84,
|
| 14853 |
+
"grad_norm": 1.8705509901046753,
|
| 14854 |
+
"learning_rate": 8.029088750371031e-06,
|
| 14855 |
+
"loss": 2.411,
|
| 14856 |
+
"step": 8484
|
| 14857 |
+
},
|
| 14858 |
+
{
|
| 14859 |
+
"epoch": 0.84,
|
| 14860 |
+
"grad_norm": 1.5241179466247559,
|
| 14861 |
+
"learning_rate": 8.009300484812507e-06,
|
| 14862 |
+
"loss": 2.424,
|
| 14863 |
+
"step": 8488
|
| 14864 |
+
},
|
| 14865 |
+
{
|
| 14866 |
+
"epoch": 0.84,
|
| 14867 |
+
"grad_norm": 1.6394929885864258,
|
| 14868 |
+
"learning_rate": 7.989512219253984e-06,
|
| 14869 |
+
"loss": 2.4855,
|
| 14870 |
+
"step": 8492
|
| 14871 |
+
},
|
| 14872 |
+
{
|
| 14873 |
+
"epoch": 0.84,
|
| 14874 |
+
"grad_norm": 1.7148469686508179,
|
| 14875 |
+
"learning_rate": 7.969723953695458e-06,
|
| 14876 |
+
"loss": 2.4967,
|
| 14877 |
+
"step": 8496
|
| 14878 |
+
},
|
| 14879 |
+
{
|
| 14880 |
+
"epoch": 0.84,
|
| 14881 |
+
"grad_norm": 1.7260617017745972,
|
| 14882 |
+
"learning_rate": 7.949935688136934e-06,
|
| 14883 |
+
"loss": 2.5231,
|
| 14884 |
+
"step": 8500
|
| 14885 |
+
},
|
| 14886 |
+
{
|
| 14887 |
+
"epoch": 0.84,
|
| 14888 |
+
"grad_norm": 1.84506356716156,
|
| 14889 |
+
"learning_rate": 7.93014742257841e-06,
|
| 14890 |
+
"loss": 2.3639,
|
| 14891 |
+
"step": 8504
|
| 14892 |
+
},
|
| 14893 |
+
{
|
| 14894 |
+
"epoch": 0.84,
|
| 14895 |
+
"grad_norm": 1.5255876779556274,
|
| 14896 |
+
"learning_rate": 7.910359157019887e-06,
|
| 14897 |
+
"loss": 2.4189,
|
| 14898 |
+
"step": 8508
|
| 14899 |
+
},
|
| 14900 |
+
{
|
| 14901 |
+
"epoch": 0.84,
|
| 14902 |
+
"grad_norm": 1.7980495691299438,
|
| 14903 |
+
"learning_rate": 7.890570891461365e-06,
|
| 14904 |
+
"loss": 2.4907,
|
| 14905 |
+
"step": 8512
|
| 14906 |
+
},
|
| 14907 |
+
{
|
| 14908 |
+
"epoch": 0.84,
|
| 14909 |
+
"grad_norm": 1.7311155796051025,
|
| 14910 |
+
"learning_rate": 7.87078262590284e-06,
|
| 14911 |
+
"loss": 2.5981,
|
| 14912 |
+
"step": 8516
|
| 14913 |
+
},
|
| 14914 |
+
{
|
| 14915 |
+
"epoch": 0.84,
|
| 14916 |
+
"grad_norm": 1.6992567777633667,
|
| 14917 |
+
"learning_rate": 7.850994360344316e-06,
|
| 14918 |
+
"loss": 2.5617,
|
| 14919 |
+
"step": 8520
|
| 14920 |
+
},
|
| 14921 |
+
{
|
| 14922 |
+
"epoch": 0.84,
|
| 14923 |
+
"grad_norm": 1.6677074432373047,
|
| 14924 |
+
"learning_rate": 7.831206094785792e-06,
|
| 14925 |
+
"loss": 2.5869,
|
| 14926 |
+
"step": 8524
|
| 14927 |
+
},
|
| 14928 |
+
{
|
| 14929 |
+
"epoch": 0.84,
|
| 14930 |
+
"grad_norm": 1.717405080795288,
|
| 14931 |
+
"learning_rate": 7.811417829227269e-06,
|
| 14932 |
+
"loss": 2.534,
|
| 14933 |
+
"step": 8528
|
| 14934 |
+
},
|
| 14935 |
+
{
|
| 14936 |
+
"epoch": 0.84,
|
| 14937 |
+
"grad_norm": 1.7169833183288574,
|
| 14938 |
+
"learning_rate": 7.791629563668745e-06,
|
| 14939 |
+
"loss": 2.5644,
|
| 14940 |
+
"step": 8532
|
| 14941 |
+
},
|
| 14942 |
+
{
|
| 14943 |
+
"epoch": 0.84,
|
| 14944 |
+
"grad_norm": 1.6315467357635498,
|
| 14945 |
+
"learning_rate": 7.77184129811022e-06,
|
| 14946 |
+
"loss": 2.5054,
|
| 14947 |
+
"step": 8536
|
| 14948 |
+
},
|
| 14949 |
+
{
|
| 14950 |
+
"epoch": 0.84,
|
| 14951 |
+
"grad_norm": 1.7004671096801758,
|
| 14952 |
+
"learning_rate": 7.752053032551696e-06,
|
| 14953 |
+
"loss": 2.3946,
|
| 14954 |
+
"step": 8540
|
| 14955 |
+
},
|
| 14956 |
+
{
|
| 14957 |
+
"epoch": 0.85,
|
| 14958 |
+
"grad_norm": 1.7185847759246826,
|
| 14959 |
+
"learning_rate": 7.732264766993174e-06,
|
| 14960 |
+
"loss": 2.6361,
|
| 14961 |
+
"step": 8544
|
| 14962 |
+
},
|
| 14963 |
+
{
|
| 14964 |
+
"epoch": 0.85,
|
| 14965 |
+
"grad_norm": 1.7826788425445557,
|
| 14966 |
+
"learning_rate": 7.71247650143465e-06,
|
| 14967 |
+
"loss": 2.4693,
|
| 14968 |
+
"step": 8548
|
| 14969 |
+
},
|
| 14970 |
+
{
|
| 14971 |
+
"epoch": 0.85,
|
| 14972 |
+
"grad_norm": 1.5494498014450073,
|
| 14973 |
+
"learning_rate": 7.692688235876127e-06,
|
| 14974 |
+
"loss": 2.475,
|
| 14975 |
+
"step": 8552
|
| 14976 |
+
},
|
| 14977 |
+
{
|
| 14978 |
+
"epoch": 0.85,
|
| 14979 |
+
"grad_norm": 1.6112264394760132,
|
| 14980 |
+
"learning_rate": 7.672899970317603e-06,
|
| 14981 |
+
"loss": 2.685,
|
| 14982 |
+
"step": 8556
|
| 14983 |
+
},
|
| 14984 |
+
{
|
| 14985 |
+
"epoch": 0.85,
|
| 14986 |
+
"grad_norm": 1.583173394203186,
|
| 14987 |
+
"learning_rate": 7.653111704759078e-06,
|
| 14988 |
+
"loss": 2.4246,
|
| 14989 |
+
"step": 8560
|
| 14990 |
+
},
|
| 14991 |
+
{
|
| 14992 |
+
"epoch": 0.85,
|
| 14993 |
+
"grad_norm": 1.6651097536087036,
|
| 14994 |
+
"learning_rate": 7.633323439200554e-06,
|
| 14995 |
+
"loss": 2.4226,
|
| 14996 |
+
"step": 8564
|
| 14997 |
+
},
|
| 14998 |
+
{
|
| 14999 |
+
"epoch": 0.85,
|
| 15000 |
+
"grad_norm": 1.6389166116714478,
|
| 15001 |
+
"learning_rate": 7.613535173642031e-06,
|
| 15002 |
+
"loss": 2.4979,
|
| 15003 |
+
"step": 8568
|
| 15004 |
+
},
|
| 15005 |
+
{
|
| 15006 |
+
"epoch": 0.85,
|
| 15007 |
+
"grad_norm": 1.7426066398620605,
|
| 15008 |
+
"learning_rate": 7.5937469080835074e-06,
|
| 15009 |
+
"loss": 2.5513,
|
| 15010 |
+
"step": 8572
|
| 15011 |
+
},
|
| 15012 |
+
{
|
| 15013 |
+
"epoch": 0.85,
|
| 15014 |
+
"grad_norm": 1.6744520664215088,
|
| 15015 |
+
"learning_rate": 7.573958642524984e-06,
|
| 15016 |
+
"loss": 2.4503,
|
| 15017 |
+
"step": 8576
|
| 15018 |
+
},
|
| 15019 |
+
{
|
| 15020 |
+
"epoch": 0.85,
|
| 15021 |
+
"grad_norm": 1.7158594131469727,
|
| 15022 |
+
"learning_rate": 7.554170376966458e-06,
|
| 15023 |
+
"loss": 2.4517,
|
| 15024 |
+
"step": 8580
|
| 15025 |
+
},
|
| 15026 |
+
{
|
| 15027 |
+
"epoch": 0.85,
|
| 15028 |
+
"grad_norm": 1.6604433059692383,
|
| 15029 |
+
"learning_rate": 7.5343821114079355e-06,
|
| 15030 |
+
"loss": 2.4247,
|
| 15031 |
+
"step": 8584
|
| 15032 |
+
},
|
| 15033 |
+
{
|
| 15034 |
+
"epoch": 0.85,
|
| 15035 |
+
"grad_norm": 1.636860728263855,
|
| 15036 |
+
"learning_rate": 7.514593845849412e-06,
|
| 15037 |
+
"loss": 2.2244,
|
| 15038 |
+
"step": 8588
|
| 15039 |
+
},
|
| 15040 |
+
{
|
| 15041 |
+
"epoch": 0.85,
|
| 15042 |
+
"grad_norm": 1.6336443424224854,
|
| 15043 |
+
"learning_rate": 7.494805580290888e-06,
|
| 15044 |
+
"loss": 2.5156,
|
| 15045 |
+
"step": 8592
|
| 15046 |
+
},
|
| 15047 |
+
{
|
| 15048 |
+
"epoch": 0.85,
|
| 15049 |
+
"grad_norm": 1.8386069536209106,
|
| 15050 |
+
"learning_rate": 7.4750173147323645e-06,
|
| 15051 |
+
"loss": 2.4954,
|
| 15052 |
+
"step": 8596
|
| 15053 |
+
},
|
| 15054 |
+
{
|
| 15055 |
+
"epoch": 0.85,
|
| 15056 |
+
"grad_norm": 1.6216977834701538,
|
| 15057 |
+
"learning_rate": 7.45522904917384e-06,
|
| 15058 |
+
"loss": 2.4894,
|
| 15059 |
+
"step": 8600
|
| 15060 |
+
},
|
| 15061 |
+
{
|
| 15062 |
+
"epoch": 0.85,
|
| 15063 |
+
"grad_norm": 1.7700227499008179,
|
| 15064 |
+
"learning_rate": 7.435440783615316e-06,
|
| 15065 |
+
"loss": 2.6143,
|
| 15066 |
+
"step": 8604
|
| 15067 |
+
},
|
| 15068 |
+
{
|
| 15069 |
+
"epoch": 0.85,
|
| 15070 |
+
"grad_norm": 1.6691062450408936,
|
| 15071 |
+
"learning_rate": 7.415652518056793e-06,
|
| 15072 |
+
"loss": 2.4329,
|
| 15073 |
+
"step": 8608
|
| 15074 |
+
},
|
| 15075 |
+
{
|
| 15076 |
+
"epoch": 0.85,
|
| 15077 |
+
"grad_norm": 1.517760992050171,
|
| 15078 |
+
"learning_rate": 7.395864252498269e-06,
|
| 15079 |
+
"loss": 2.402,
|
| 15080 |
+
"step": 8612
|
| 15081 |
+
},
|
| 15082 |
+
{
|
| 15083 |
+
"epoch": 0.85,
|
| 15084 |
+
"grad_norm": 1.5939360857009888,
|
| 15085 |
+
"learning_rate": 7.376075986939745e-06,
|
| 15086 |
+
"loss": 2.4902,
|
| 15087 |
+
"step": 8616
|
| 15088 |
+
},
|
| 15089 |
+
{
|
| 15090 |
+
"epoch": 0.85,
|
| 15091 |
+
"grad_norm": 1.6323703527450562,
|
| 15092 |
+
"learning_rate": 7.356287721381221e-06,
|
| 15093 |
+
"loss": 2.6149,
|
| 15094 |
+
"step": 8620
|
| 15095 |
+
},
|
| 15096 |
+
{
|
| 15097 |
+
"epoch": 0.85,
|
| 15098 |
+
"grad_norm": 1.5932726860046387,
|
| 15099 |
+
"learning_rate": 7.336499455822697e-06,
|
| 15100 |
+
"loss": 2.4341,
|
| 15101 |
+
"step": 8624
|
| 15102 |
+
},
|
| 15103 |
+
{
|
| 15104 |
+
"epoch": 0.85,
|
| 15105 |
+
"grad_norm": 1.7558265924453735,
|
| 15106 |
+
"learning_rate": 7.316711190264173e-06,
|
| 15107 |
+
"loss": 2.6402,
|
| 15108 |
+
"step": 8628
|
| 15109 |
+
},
|
| 15110 |
+
{
|
| 15111 |
+
"epoch": 0.85,
|
| 15112 |
+
"grad_norm": 1.7289772033691406,
|
| 15113 |
+
"learning_rate": 7.29692292470565e-06,
|
| 15114 |
+
"loss": 2.3923,
|
| 15115 |
+
"step": 8632
|
| 15116 |
+
},
|
| 15117 |
+
{
|
| 15118 |
+
"epoch": 0.85,
|
| 15119 |
+
"grad_norm": 1.800723910331726,
|
| 15120 |
+
"learning_rate": 7.277134659147127e-06,
|
| 15121 |
+
"loss": 2.4707,
|
| 15122 |
+
"step": 8636
|
| 15123 |
+
},
|
| 15124 |
+
{
|
| 15125 |
+
"epoch": 0.85,
|
| 15126 |
+
"grad_norm": 1.672587275505066,
|
| 15127 |
+
"learning_rate": 7.257346393588603e-06,
|
| 15128 |
+
"loss": 2.526,
|
| 15129 |
+
"step": 8640
|
| 15130 |
+
},
|
| 15131 |
+
{
|
| 15132 |
+
"epoch": 0.86,
|
| 15133 |
+
"grad_norm": 1.686955451965332,
|
| 15134 |
+
"learning_rate": 7.237558128030078e-06,
|
| 15135 |
+
"loss": 2.4433,
|
| 15136 |
+
"step": 8644
|
| 15137 |
+
},
|
| 15138 |
+
{
|
| 15139 |
+
"epoch": 0.86,
|
| 15140 |
+
"grad_norm": 1.5995945930480957,
|
| 15141 |
+
"learning_rate": 7.217769862471554e-06,
|
| 15142 |
+
"loss": 2.609,
|
| 15143 |
+
"step": 8648
|
| 15144 |
+
},
|
| 15145 |
+
{
|
| 15146 |
+
"epoch": 0.86,
|
| 15147 |
+
"grad_norm": 1.6582813262939453,
|
| 15148 |
+
"learning_rate": 7.197981596913031e-06,
|
| 15149 |
+
"loss": 2.3675,
|
| 15150 |
+
"step": 8652
|
| 15151 |
+
},
|
| 15152 |
+
{
|
| 15153 |
+
"epoch": 0.86,
|
| 15154 |
+
"grad_norm": 1.6642727851867676,
|
| 15155 |
+
"learning_rate": 7.178193331354508e-06,
|
| 15156 |
+
"loss": 2.4762,
|
| 15157 |
+
"step": 8656
|
| 15158 |
+
},
|
| 15159 |
+
{
|
| 15160 |
+
"epoch": 0.86,
|
| 15161 |
+
"grad_norm": 1.6755313873291016,
|
| 15162 |
+
"learning_rate": 7.158405065795984e-06,
|
| 15163 |
+
"loss": 2.4811,
|
| 15164 |
+
"step": 8660
|
| 15165 |
+
},
|
| 15166 |
+
{
|
| 15167 |
+
"epoch": 0.86,
|
| 15168 |
+
"grad_norm": 1.787563681602478,
|
| 15169 |
+
"learning_rate": 7.138616800237459e-06,
|
| 15170 |
+
"loss": 2.4481,
|
| 15171 |
+
"step": 8664
|
| 15172 |
+
},
|
| 15173 |
+
{
|
| 15174 |
+
"epoch": 0.86,
|
| 15175 |
+
"grad_norm": 1.5694202184677124,
|
| 15176 |
+
"learning_rate": 7.118828534678936e-06,
|
| 15177 |
+
"loss": 2.44,
|
| 15178 |
+
"step": 8668
|
| 15179 |
+
},
|
| 15180 |
+
{
|
| 15181 |
+
"epoch": 0.86,
|
| 15182 |
+
"grad_norm": 1.6753268241882324,
|
| 15183 |
+
"learning_rate": 7.099040269120412e-06,
|
| 15184 |
+
"loss": 2.664,
|
| 15185 |
+
"step": 8672
|
| 15186 |
+
},
|
| 15187 |
+
{
|
| 15188 |
+
"epoch": 0.86,
|
| 15189 |
+
"grad_norm": 1.8283534049987793,
|
| 15190 |
+
"learning_rate": 7.079252003561888e-06,
|
| 15191 |
+
"loss": 2.4761,
|
| 15192 |
+
"step": 8676
|
| 15193 |
+
},
|
| 15194 |
+
{
|
| 15195 |
+
"epoch": 0.86,
|
| 15196 |
+
"grad_norm": 1.730433702468872,
|
| 15197 |
+
"learning_rate": 7.059463738003365e-06,
|
| 15198 |
+
"loss": 2.5623,
|
| 15199 |
+
"step": 8680
|
| 15200 |
+
},
|
| 15201 |
+
{
|
| 15202 |
+
"epoch": 0.86,
|
| 15203 |
+
"grad_norm": 1.7397688627243042,
|
| 15204 |
+
"learning_rate": 7.03967547244484e-06,
|
| 15205 |
+
"loss": 2.6005,
|
| 15206 |
+
"step": 8684
|
| 15207 |
+
},
|
| 15208 |
+
{
|
| 15209 |
+
"epoch": 0.86,
|
| 15210 |
+
"grad_norm": 1.7049192190170288,
|
| 15211 |
+
"learning_rate": 7.0198872068863165e-06,
|
| 15212 |
+
"loss": 2.4976,
|
| 15213 |
+
"step": 8688
|
| 15214 |
+
},
|
| 15215 |
+
{
|
| 15216 |
+
"epoch": 0.86,
|
| 15217 |
+
"grad_norm": 1.7921533584594727,
|
| 15218 |
+
"learning_rate": 7.000098941327793e-06,
|
| 15219 |
+
"loss": 2.5239,
|
| 15220 |
+
"step": 8692
|
| 15221 |
+
},
|
| 15222 |
+
{
|
| 15223 |
+
"epoch": 0.86,
|
| 15224 |
+
"grad_norm": 1.7059839963912964,
|
| 15225 |
+
"learning_rate": 6.980310675769269e-06,
|
| 15226 |
+
"loss": 2.403,
|
| 15227 |
+
"step": 8696
|
| 15228 |
+
},
|
| 15229 |
+
{
|
| 15230 |
+
"epoch": 0.86,
|
| 15231 |
+
"grad_norm": 1.636533260345459,
|
| 15232 |
+
"learning_rate": 6.9605224102107455e-06,
|
| 15233 |
+
"loss": 2.4023,
|
| 15234 |
+
"step": 8700
|
| 15235 |
+
},
|
| 15236 |
+
{
|
| 15237 |
+
"epoch": 0.86,
|
| 15238 |
+
"grad_norm": 1.6458524465560913,
|
| 15239 |
+
"learning_rate": 6.940734144652221e-06,
|
| 15240 |
+
"loss": 2.6021,
|
| 15241 |
+
"step": 8704
|
| 15242 |
+
},
|
| 15243 |
+
{
|
| 15244 |
+
"epoch": 0.86,
|
| 15245 |
+
"grad_norm": 1.8312097787857056,
|
| 15246 |
+
"learning_rate": 6.920945879093697e-06,
|
| 15247 |
+
"loss": 2.4073,
|
| 15248 |
+
"step": 8708
|
| 15249 |
+
},
|
| 15250 |
+
{
|
| 15251 |
+
"epoch": 0.86,
|
| 15252 |
+
"grad_norm": 1.6643646955490112,
|
| 15253 |
+
"learning_rate": 6.901157613535174e-06,
|
| 15254 |
+
"loss": 2.4315,
|
| 15255 |
+
"step": 8712
|
| 15256 |
+
},
|
| 15257 |
+
{
|
| 15258 |
+
"epoch": 0.86,
|
| 15259 |
+
"grad_norm": 1.7501451969146729,
|
| 15260 |
+
"learning_rate": 6.88136934797665e-06,
|
| 15261 |
+
"loss": 2.4771,
|
| 15262 |
+
"step": 8716
|
| 15263 |
+
},
|
| 15264 |
+
{
|
| 15265 |
+
"epoch": 0.86,
|
| 15266 |
+
"grad_norm": 1.6172442436218262,
|
| 15267 |
+
"learning_rate": 6.861581082418127e-06,
|
| 15268 |
+
"loss": 2.4944,
|
| 15269 |
+
"step": 8720
|
| 15270 |
+
},
|
| 15271 |
+
{
|
| 15272 |
+
"epoch": 0.86,
|
| 15273 |
+
"grad_norm": 1.7708154916763306,
|
| 15274 |
+
"learning_rate": 6.841792816859603e-06,
|
| 15275 |
+
"loss": 2.5976,
|
| 15276 |
+
"step": 8724
|
| 15277 |
+
},
|
| 15278 |
+
{
|
| 15279 |
+
"epoch": 0.86,
|
| 15280 |
+
"grad_norm": 1.640822410583496,
|
| 15281 |
+
"learning_rate": 6.822004551301078e-06,
|
| 15282 |
+
"loss": 2.4854,
|
| 15283 |
+
"step": 8728
|
| 15284 |
+
},
|
| 15285 |
+
{
|
| 15286 |
+
"epoch": 0.86,
|
| 15287 |
+
"grad_norm": 1.7448416948318481,
|
| 15288 |
+
"learning_rate": 6.802216285742554e-06,
|
| 15289 |
+
"loss": 2.5023,
|
| 15290 |
+
"step": 8732
|
| 15291 |
+
},
|
| 15292 |
+
{
|
| 15293 |
+
"epoch": 0.86,
|
| 15294 |
+
"grad_norm": 1.7678117752075195,
|
| 15295 |
+
"learning_rate": 6.7824280201840315e-06,
|
| 15296 |
+
"loss": 2.3739,
|
| 15297 |
+
"step": 8736
|
| 15298 |
+
},
|
| 15299 |
+
{
|
| 15300 |
+
"epoch": 0.86,
|
| 15301 |
+
"grad_norm": 1.7742412090301514,
|
| 15302 |
+
"learning_rate": 6.762639754625508e-06,
|
| 15303 |
+
"loss": 2.5075,
|
| 15304 |
+
"step": 8740
|
| 15305 |
+
},
|
| 15306 |
+
{
|
| 15307 |
+
"epoch": 0.87,
|
| 15308 |
+
"grad_norm": 1.6153099536895752,
|
| 15309 |
+
"learning_rate": 6.742851489066984e-06,
|
| 15310 |
+
"loss": 2.3971,
|
| 15311 |
+
"step": 8744
|
| 15312 |
+
},
|
| 15313 |
+
{
|
| 15314 |
+
"epoch": 0.87,
|
| 15315 |
+
"grad_norm": 2.1363136768341064,
|
| 15316 |
+
"learning_rate": 6.72306322350846e-06,
|
| 15317 |
+
"loss": 2.5768,
|
| 15318 |
+
"step": 8748
|
| 15319 |
+
},
|
| 15320 |
+
{
|
| 15321 |
+
"epoch": 0.87,
|
| 15322 |
+
"grad_norm": 1.694061279296875,
|
| 15323 |
+
"learning_rate": 6.703274957949936e-06,
|
| 15324 |
+
"loss": 2.4628,
|
| 15325 |
+
"step": 8752
|
| 15326 |
+
},
|
| 15327 |
+
{
|
| 15328 |
+
"epoch": 0.87,
|
| 15329 |
+
"grad_norm": 1.5482895374298096,
|
| 15330 |
+
"learning_rate": 6.683486692391412e-06,
|
| 15331 |
+
"loss": 2.3695,
|
| 15332 |
+
"step": 8756
|
| 15333 |
+
},
|
| 15334 |
+
{
|
| 15335 |
+
"epoch": 0.87,
|
| 15336 |
+
"grad_norm": 1.6727862358093262,
|
| 15337 |
+
"learning_rate": 6.663698426832889e-06,
|
| 15338 |
+
"loss": 2.5452,
|
| 15339 |
+
"step": 8760
|
| 15340 |
+
},
|
| 15341 |
+
{
|
| 15342 |
+
"epoch": 0.87,
|
| 15343 |
+
"grad_norm": 2.0361154079437256,
|
| 15344 |
+
"learning_rate": 6.643910161274365e-06,
|
| 15345 |
+
"loss": 2.7449,
|
| 15346 |
+
"step": 8764
|
| 15347 |
+
},
|
| 15348 |
+
{
|
| 15349 |
+
"epoch": 0.87,
|
| 15350 |
+
"grad_norm": 1.6247440576553345,
|
| 15351 |
+
"learning_rate": 6.62412189571584e-06,
|
| 15352 |
+
"loss": 2.5772,
|
| 15353 |
+
"step": 8768
|
| 15354 |
+
},
|
| 15355 |
+
{
|
| 15356 |
+
"epoch": 0.87,
|
| 15357 |
+
"grad_norm": 1.6015691757202148,
|
| 15358 |
+
"learning_rate": 6.604333630157317e-06,
|
| 15359 |
+
"loss": 2.5729,
|
| 15360 |
+
"step": 8772
|
| 15361 |
+
},
|
| 15362 |
+
{
|
| 15363 |
+
"epoch": 0.87,
|
| 15364 |
+
"grad_norm": 1.7053395509719849,
|
| 15365 |
+
"learning_rate": 6.584545364598793e-06,
|
| 15366 |
+
"loss": 2.541,
|
| 15367 |
+
"step": 8776
|
| 15368 |
+
},
|
| 15369 |
+
{
|
| 15370 |
+
"epoch": 0.87,
|
| 15371 |
+
"grad_norm": 1.6078673601150513,
|
| 15372 |
+
"learning_rate": 6.564757099040269e-06,
|
| 15373 |
+
"loss": 2.5443,
|
| 15374 |
+
"step": 8780
|
| 15375 |
+
},
|
| 15376 |
+
{
|
| 15377 |
+
"epoch": 0.87,
|
| 15378 |
+
"grad_norm": 1.6850202083587646,
|
| 15379 |
+
"learning_rate": 6.544968833481746e-06,
|
| 15380 |
+
"loss": 2.6706,
|
| 15381 |
+
"step": 8784
|
| 15382 |
+
},
|
| 15383 |
+
{
|
| 15384 |
+
"epoch": 0.87,
|
| 15385 |
+
"grad_norm": 1.6219288110733032,
|
| 15386 |
+
"learning_rate": 6.525180567923223e-06,
|
| 15387 |
+
"loss": 2.5243,
|
| 15388 |
+
"step": 8788
|
| 15389 |
+
},
|
| 15390 |
+
{
|
| 15391 |
+
"epoch": 0.87,
|
| 15392 |
+
"grad_norm": 1.8155920505523682,
|
| 15393 |
+
"learning_rate": 6.5053923023646975e-06,
|
| 15394 |
+
"loss": 2.5544,
|
| 15395 |
+
"step": 8792
|
| 15396 |
+
},
|
| 15397 |
+
{
|
| 15398 |
+
"epoch": 0.87,
|
| 15399 |
+
"grad_norm": 1.7686444520950317,
|
| 15400 |
+
"learning_rate": 6.485604036806174e-06,
|
| 15401 |
+
"loss": 2.3905,
|
| 15402 |
+
"step": 8796
|
| 15403 |
+
},
|
| 15404 |
+
{
|
| 15405 |
+
"epoch": 0.87,
|
| 15406 |
+
"grad_norm": 1.9808334112167358,
|
| 15407 |
+
"learning_rate": 6.46581577124765e-06,
|
| 15408 |
+
"loss": 2.5068,
|
| 15409 |
+
"step": 8800
|
| 15410 |
+
},
|
| 15411 |
+
{
|
| 15412 |
+
"epoch": 0.87,
|
| 15413 |
+
"grad_norm": 1.7255302667617798,
|
| 15414 |
+
"learning_rate": 6.446027505689127e-06,
|
| 15415 |
+
"loss": 2.4938,
|
| 15416 |
+
"step": 8804
|
| 15417 |
+
},
|
| 15418 |
+
{
|
| 15419 |
+
"epoch": 0.87,
|
| 15420 |
+
"grad_norm": 1.6764934062957764,
|
| 15421 |
+
"learning_rate": 6.426239240130604e-06,
|
| 15422 |
+
"loss": 2.3194,
|
| 15423 |
+
"step": 8808
|
| 15424 |
+
},
|
| 15425 |
+
{
|
| 15426 |
+
"epoch": 0.87,
|
| 15427 |
+
"grad_norm": 1.5869959592819214,
|
| 15428 |
+
"learning_rate": 6.406450974572078e-06,
|
| 15429 |
+
"loss": 2.4587,
|
| 15430 |
+
"step": 8812
|
| 15431 |
+
},
|
| 15432 |
+
{
|
| 15433 |
+
"epoch": 0.87,
|
| 15434 |
+
"grad_norm": 1.5677440166473389,
|
| 15435 |
+
"learning_rate": 6.386662709013555e-06,
|
| 15436 |
+
"loss": 2.4172,
|
| 15437 |
+
"step": 8816
|
| 15438 |
+
},
|
| 15439 |
+
{
|
| 15440 |
+
"epoch": 0.87,
|
| 15441 |
+
"grad_norm": 1.6823347806930542,
|
| 15442 |
+
"learning_rate": 6.366874443455032e-06,
|
| 15443 |
+
"loss": 2.4559,
|
| 15444 |
+
"step": 8820
|
| 15445 |
+
},
|
| 15446 |
+
{
|
| 15447 |
+
"epoch": 0.87,
|
| 15448 |
+
"grad_norm": 1.6375516653060913,
|
| 15449 |
+
"learning_rate": 6.347086177896508e-06,
|
| 15450 |
+
"loss": 2.5198,
|
| 15451 |
+
"step": 8824
|
| 15452 |
+
},
|
| 15453 |
+
{
|
| 15454 |
+
"epoch": 0.87,
|
| 15455 |
+
"grad_norm": 1.629103422164917,
|
| 15456 |
+
"learning_rate": 6.327297912337984e-06,
|
| 15457 |
+
"loss": 2.5148,
|
| 15458 |
+
"step": 8828
|
| 15459 |
+
},
|
| 15460 |
+
{
|
| 15461 |
+
"epoch": 0.87,
|
| 15462 |
+
"grad_norm": 1.5278247594833374,
|
| 15463 |
+
"learning_rate": 6.30750964677946e-06,
|
| 15464 |
+
"loss": 2.4411,
|
| 15465 |
+
"step": 8832
|
| 15466 |
+
},
|
| 15467 |
+
{
|
| 15468 |
+
"epoch": 0.87,
|
| 15469 |
+
"grad_norm": 1.8897967338562012,
|
| 15470 |
+
"learning_rate": 6.287721381220936e-06,
|
| 15471 |
+
"loss": 2.4329,
|
| 15472 |
+
"step": 8836
|
| 15473 |
+
},
|
| 15474 |
+
{
|
| 15475 |
+
"epoch": 0.87,
|
| 15476 |
+
"grad_norm": 1.6970409154891968,
|
| 15477 |
+
"learning_rate": 6.2679331156624125e-06,
|
| 15478 |
+
"loss": 2.7446,
|
| 15479 |
+
"step": 8840
|
| 15480 |
+
},
|
| 15481 |
+
{
|
| 15482 |
+
"epoch": 0.88,
|
| 15483 |
+
"grad_norm": 1.5253138542175293,
|
| 15484 |
+
"learning_rate": 6.248144850103889e-06,
|
| 15485 |
+
"loss": 2.4237,
|
| 15486 |
+
"step": 8844
|
| 15487 |
+
},
|
| 15488 |
+
{
|
| 15489 |
+
"epoch": 0.88,
|
| 15490 |
+
"grad_norm": 1.6197317838668823,
|
| 15491 |
+
"learning_rate": 6.228356584545364e-06,
|
| 15492 |
+
"loss": 2.5055,
|
| 15493 |
+
"step": 8848
|
| 15494 |
+
},
|
| 15495 |
+
{
|
| 15496 |
+
"epoch": 0.88,
|
| 15497 |
+
"grad_norm": 1.5637788772583008,
|
| 15498 |
+
"learning_rate": 6.2085683189868415e-06,
|
| 15499 |
+
"loss": 2.3537,
|
| 15500 |
+
"step": 8852
|
| 15501 |
+
},
|
| 15502 |
+
{
|
| 15503 |
+
"epoch": 0.88,
|
| 15504 |
+
"grad_norm": 1.662288784980774,
|
| 15505 |
+
"learning_rate": 6.188780053428318e-06,
|
| 15506 |
+
"loss": 2.3261,
|
| 15507 |
+
"step": 8856
|
| 15508 |
+
},
|
| 15509 |
+
{
|
| 15510 |
+
"epoch": 0.88,
|
| 15511 |
+
"grad_norm": 1.642146348953247,
|
| 15512 |
+
"learning_rate": 6.168991787869793e-06,
|
| 15513 |
+
"loss": 2.4903,
|
| 15514 |
+
"step": 8860
|
| 15515 |
+
},
|
| 15516 |
+
{
|
| 15517 |
+
"epoch": 0.88,
|
| 15518 |
+
"grad_norm": 1.6560384035110474,
|
| 15519 |
+
"learning_rate": 6.1492035223112696e-06,
|
| 15520 |
+
"loss": 2.5563,
|
| 15521 |
+
"step": 8864
|
| 15522 |
+
},
|
| 15523 |
+
{
|
| 15524 |
+
"epoch": 0.88,
|
| 15525 |
+
"grad_norm": 1.7499784231185913,
|
| 15526 |
+
"learning_rate": 6.129415256752746e-06,
|
| 15527 |
+
"loss": 2.5078,
|
| 15528 |
+
"step": 8868
|
| 15529 |
+
},
|
| 15530 |
+
{
|
| 15531 |
+
"epoch": 0.88,
|
| 15532 |
+
"grad_norm": 1.8147586584091187,
|
| 15533 |
+
"learning_rate": 6.109626991194222e-06,
|
| 15534 |
+
"loss": 2.3585,
|
| 15535 |
+
"step": 8872
|
| 15536 |
+
},
|
| 15537 |
+
{
|
| 15538 |
+
"epoch": 0.88,
|
| 15539 |
+
"grad_norm": 1.6541246175765991,
|
| 15540 |
+
"learning_rate": 6.0898387256356985e-06,
|
| 15541 |
+
"loss": 2.5436,
|
| 15542 |
+
"step": 8876
|
| 15543 |
+
},
|
| 15544 |
+
{
|
| 15545 |
+
"epoch": 0.88,
|
| 15546 |
+
"grad_norm": 1.8309890031814575,
|
| 15547 |
+
"learning_rate": 6.070050460077174e-06,
|
| 15548 |
+
"loss": 2.6095,
|
| 15549 |
+
"step": 8880
|
| 15550 |
+
},
|
| 15551 |
+
{
|
| 15552 |
+
"epoch": 0.88,
|
| 15553 |
+
"grad_norm": 1.725372552871704,
|
| 15554 |
+
"learning_rate": 6.050262194518651e-06,
|
| 15555 |
+
"loss": 2.5652,
|
| 15556 |
+
"step": 8884
|
| 15557 |
+
},
|
| 15558 |
+
{
|
| 15559 |
+
"epoch": 0.88,
|
| 15560 |
+
"grad_norm": 1.6648577451705933,
|
| 15561 |
+
"learning_rate": 6.0304739289601275e-06,
|
| 15562 |
+
"loss": 2.4624,
|
| 15563 |
+
"step": 8888
|
| 15564 |
+
},
|
| 15565 |
+
{
|
| 15566 |
+
"epoch": 0.88,
|
| 15567 |
+
"grad_norm": 1.6224662065505981,
|
| 15568 |
+
"learning_rate": 6.010685663401603e-06,
|
| 15569 |
+
"loss": 2.4616,
|
| 15570 |
+
"step": 8892
|
| 15571 |
+
},
|
| 15572 |
+
{
|
| 15573 |
+
"epoch": 0.88,
|
| 15574 |
+
"grad_norm": 1.535581111907959,
|
| 15575 |
+
"learning_rate": 5.990897397843079e-06,
|
| 15576 |
+
"loss": 2.3335,
|
| 15577 |
+
"step": 8896
|
| 15578 |
+
},
|
| 15579 |
+
{
|
| 15580 |
+
"epoch": 0.88,
|
| 15581 |
+
"grad_norm": 1.5054175853729248,
|
| 15582 |
+
"learning_rate": 5.971109132284556e-06,
|
| 15583 |
+
"loss": 2.3097,
|
| 15584 |
+
"step": 8900
|
| 15585 |
+
},
|
| 15586 |
+
{
|
| 15587 |
+
"epoch": 0.88,
|
| 15588 |
+
"grad_norm": 1.7295256853103638,
|
| 15589 |
+
"learning_rate": 5.951320866726032e-06,
|
| 15590 |
+
"loss": 2.538,
|
| 15591 |
+
"step": 8904
|
| 15592 |
+
},
|
| 15593 |
+
{
|
| 15594 |
+
"epoch": 0.88,
|
| 15595 |
+
"grad_norm": 1.6823656558990479,
|
| 15596 |
+
"learning_rate": 5.931532601167508e-06,
|
| 15597 |
+
"loss": 2.4051,
|
| 15598 |
+
"step": 8908
|
| 15599 |
+
},
|
| 15600 |
+
{
|
| 15601 |
+
"epoch": 0.88,
|
| 15602 |
+
"grad_norm": 1.702989935874939,
|
| 15603 |
+
"learning_rate": 5.911744335608984e-06,
|
| 15604 |
+
"loss": 2.3926,
|
| 15605 |
+
"step": 8912
|
| 15606 |
+
},
|
| 15607 |
+
{
|
| 15608 |
+
"epoch": 0.88,
|
| 15609 |
+
"grad_norm": 1.6187750101089478,
|
| 15610 |
+
"learning_rate": 5.89195607005046e-06,
|
| 15611 |
+
"loss": 2.2627,
|
| 15612 |
+
"step": 8916
|
| 15613 |
+
},
|
| 15614 |
+
{
|
| 15615 |
+
"epoch": 0.88,
|
| 15616 |
+
"grad_norm": 1.6742513179779053,
|
| 15617 |
+
"learning_rate": 5.872167804491937e-06,
|
| 15618 |
+
"loss": 2.3604,
|
| 15619 |
+
"step": 8920
|
| 15620 |
+
},
|
| 15621 |
+
{
|
| 15622 |
+
"epoch": 0.88,
|
| 15623 |
+
"grad_norm": 1.7713173627853394,
|
| 15624 |
+
"learning_rate": 5.852379538933413e-06,
|
| 15625 |
+
"loss": 2.557,
|
| 15626 |
+
"step": 8924
|
| 15627 |
+
},
|
| 15628 |
+
{
|
| 15629 |
+
"epoch": 0.88,
|
| 15630 |
+
"grad_norm": 1.624006748199463,
|
| 15631 |
+
"learning_rate": 5.832591273374889e-06,
|
| 15632 |
+
"loss": 2.4546,
|
| 15633 |
+
"step": 8928
|
| 15634 |
+
},
|
| 15635 |
+
{
|
| 15636 |
+
"epoch": 0.88,
|
| 15637 |
+
"grad_norm": 1.8159652948379517,
|
| 15638 |
+
"learning_rate": 5.812803007816365e-06,
|
| 15639 |
+
"loss": 2.4153,
|
| 15640 |
+
"step": 8932
|
| 15641 |
+
},
|
| 15642 |
+
{
|
| 15643 |
+
"epoch": 0.88,
|
| 15644 |
+
"grad_norm": 1.6506266593933105,
|
| 15645 |
+
"learning_rate": 5.793014742257842e-06,
|
| 15646 |
+
"loss": 2.6401,
|
| 15647 |
+
"step": 8936
|
| 15648 |
+
},
|
| 15649 |
+
{
|
| 15650 |
+
"epoch": 0.88,
|
| 15651 |
+
"grad_norm": 1.7533698081970215,
|
| 15652 |
+
"learning_rate": 5.773226476699318e-06,
|
| 15653 |
+
"loss": 2.5884,
|
| 15654 |
+
"step": 8940
|
| 15655 |
+
},
|
| 15656 |
+
{
|
| 15657 |
+
"epoch": 0.88,
|
| 15658 |
+
"grad_norm": 1.6324323415756226,
|
| 15659 |
+
"learning_rate": 5.7534382111407935e-06,
|
| 15660 |
+
"loss": 2.2965,
|
| 15661 |
+
"step": 8944
|
| 15662 |
+
},
|
| 15663 |
+
{
|
| 15664 |
+
"epoch": 0.89,
|
| 15665 |
+
"grad_norm": 1.5913504362106323,
|
| 15666 |
+
"learning_rate": 5.73364994558227e-06,
|
| 15667 |
+
"loss": 2.4982,
|
| 15668 |
+
"step": 8948
|
| 15669 |
+
},
|
| 15670 |
+
{
|
| 15671 |
+
"epoch": 0.89,
|
| 15672 |
+
"grad_norm": 1.9502431154251099,
|
| 15673 |
+
"learning_rate": 5.713861680023746e-06,
|
| 15674 |
+
"loss": 2.3815,
|
| 15675 |
+
"step": 8952
|
| 15676 |
+
},
|
| 15677 |
+
{
|
| 15678 |
+
"epoch": 0.89,
|
| 15679 |
+
"grad_norm": 1.5532846450805664,
|
| 15680 |
+
"learning_rate": 5.694073414465222e-06,
|
| 15681 |
+
"loss": 2.4715,
|
| 15682 |
+
"step": 8956
|
| 15683 |
+
},
|
| 15684 |
+
{
|
| 15685 |
+
"epoch": 0.89,
|
| 15686 |
+
"grad_norm": 1.6778051853179932,
|
| 15687 |
+
"learning_rate": 5.674285148906699e-06,
|
| 15688 |
+
"loss": 2.2548,
|
| 15689 |
+
"step": 8960
|
| 15690 |
+
},
|
| 15691 |
+
{
|
| 15692 |
+
"epoch": 0.89,
|
| 15693 |
+
"grad_norm": 1.5361812114715576,
|
| 15694 |
+
"learning_rate": 5.654496883348174e-06,
|
| 15695 |
+
"loss": 2.4623,
|
| 15696 |
+
"step": 8964
|
| 15697 |
+
},
|
| 15698 |
+
{
|
| 15699 |
+
"epoch": 0.89,
|
| 15700 |
+
"grad_norm": 1.6158546209335327,
|
| 15701 |
+
"learning_rate": 5.634708617789651e-06,
|
| 15702 |
+
"loss": 2.4732,
|
| 15703 |
+
"step": 8968
|
| 15704 |
+
},
|
| 15705 |
+
{
|
| 15706 |
+
"epoch": 0.89,
|
| 15707 |
+
"grad_norm": 1.7343107461929321,
|
| 15708 |
+
"learning_rate": 5.614920352231128e-06,
|
| 15709 |
+
"loss": 2.4489,
|
| 15710 |
+
"step": 8972
|
| 15711 |
+
},
|
| 15712 |
+
{
|
| 15713 |
+
"epoch": 0.89,
|
| 15714 |
+
"grad_norm": 1.6918219327926636,
|
| 15715 |
+
"learning_rate": 5.595132086672603e-06,
|
| 15716 |
+
"loss": 2.4923,
|
| 15717 |
+
"step": 8976
|
| 15718 |
+
},
|
| 15719 |
+
{
|
| 15720 |
+
"epoch": 0.89,
|
| 15721 |
+
"grad_norm": 1.725468397140503,
|
| 15722 |
+
"learning_rate": 5.5753438211140795e-06,
|
| 15723 |
+
"loss": 2.5126,
|
| 15724 |
+
"step": 8980
|
| 15725 |
+
},
|
| 15726 |
+
{
|
| 15727 |
+
"epoch": 0.89,
|
| 15728 |
+
"grad_norm": 1.6052392721176147,
|
| 15729 |
+
"learning_rate": 5.555555555555556e-06,
|
| 15730 |
+
"loss": 2.3489,
|
| 15731 |
+
"step": 8984
|
| 15732 |
+
},
|
| 15733 |
+
{
|
| 15734 |
+
"epoch": 0.89,
|
| 15735 |
+
"grad_norm": 1.6850762367248535,
|
| 15736 |
+
"learning_rate": 5.535767289997032e-06,
|
| 15737 |
+
"loss": 2.5302,
|
| 15738 |
+
"step": 8988
|
| 15739 |
+
},
|
| 15740 |
+
{
|
| 15741 |
+
"epoch": 0.89,
|
| 15742 |
+
"grad_norm": 2.009579658508301,
|
| 15743 |
+
"learning_rate": 5.5159790244385085e-06,
|
| 15744 |
+
"loss": 2.4038,
|
| 15745 |
+
"step": 8992
|
| 15746 |
+
},
|
| 15747 |
+
{
|
| 15748 |
+
"epoch": 0.89,
|
| 15749 |
+
"grad_norm": 1.787927269935608,
|
| 15750 |
+
"learning_rate": 5.496190758879984e-06,
|
| 15751 |
+
"loss": 2.4092,
|
| 15752 |
+
"step": 8996
|
| 15753 |
+
},
|
| 15754 |
+
{
|
| 15755 |
+
"epoch": 0.89,
|
| 15756 |
+
"grad_norm": 1.583129644393921,
|
| 15757 |
+
"learning_rate": 5.476402493321461e-06,
|
| 15758 |
+
"loss": 2.5511,
|
| 15759 |
+
"step": 9000
|
| 15760 |
+
},
|
| 15761 |
+
{
|
| 15762 |
+
"epoch": 0.89,
|
| 15763 |
+
"grad_norm": 1.5953203439712524,
|
| 15764 |
+
"learning_rate": 5.4566142277629374e-06,
|
| 15765 |
+
"loss": 2.3661,
|
| 15766 |
+
"step": 9004
|
| 15767 |
+
},
|
| 15768 |
+
{
|
| 15769 |
+
"epoch": 0.89,
|
| 15770 |
+
"grad_norm": 1.6569699048995972,
|
| 15771 |
+
"learning_rate": 5.436825962204413e-06,
|
| 15772 |
+
"loss": 2.3866,
|
| 15773 |
+
"step": 9008
|
| 15774 |
+
},
|
| 15775 |
+
{
|
| 15776 |
+
"epoch": 0.89,
|
| 15777 |
+
"grad_norm": 1.5289307832717896,
|
| 15778 |
+
"learning_rate": 5.417037696645889e-06,
|
| 15779 |
+
"loss": 2.4527,
|
| 15780 |
+
"step": 9012
|
| 15781 |
+
},
|
| 15782 |
+
{
|
| 15783 |
+
"epoch": 0.89,
|
| 15784 |
+
"grad_norm": 1.5741091966629028,
|
| 15785 |
+
"learning_rate": 5.3972494310873655e-06,
|
| 15786 |
+
"loss": 2.5182,
|
| 15787 |
+
"step": 9016
|
| 15788 |
+
},
|
| 15789 |
+
{
|
| 15790 |
+
"epoch": 0.89,
|
| 15791 |
+
"grad_norm": 1.672529697418213,
|
| 15792 |
+
"learning_rate": 5.377461165528842e-06,
|
| 15793 |
+
"loss": 2.4246,
|
| 15794 |
+
"step": 9020
|
| 15795 |
+
},
|
| 15796 |
+
{
|
| 15797 |
+
"epoch": 0.89,
|
| 15798 |
+
"grad_norm": 1.5960274934768677,
|
| 15799 |
+
"learning_rate": 5.357672899970318e-06,
|
| 15800 |
+
"loss": 2.3121,
|
| 15801 |
+
"step": 9024
|
| 15802 |
+
},
|
| 15803 |
+
{
|
| 15804 |
+
"epoch": 0.89,
|
| 15805 |
+
"grad_norm": 1.9494187831878662,
|
| 15806 |
+
"learning_rate": 5.337884634411794e-06,
|
| 15807 |
+
"loss": 2.3272,
|
| 15808 |
+
"step": 9028
|
| 15809 |
+
},
|
| 15810 |
+
{
|
| 15811 |
+
"epoch": 0.89,
|
| 15812 |
+
"grad_norm": 1.8283107280731201,
|
| 15813 |
+
"learning_rate": 5.31809636885327e-06,
|
| 15814 |
+
"loss": 2.4936,
|
| 15815 |
+
"step": 9032
|
| 15816 |
+
},
|
| 15817 |
+
{
|
| 15818 |
+
"epoch": 0.89,
|
| 15819 |
+
"grad_norm": 1.6064289808273315,
|
| 15820 |
+
"learning_rate": 5.298308103294747e-06,
|
| 15821 |
+
"loss": 2.5352,
|
| 15822 |
+
"step": 9036
|
| 15823 |
+
},
|
| 15824 |
+
{
|
| 15825 |
+
"epoch": 0.89,
|
| 15826 |
+
"grad_norm": 1.6828787326812744,
|
| 15827 |
+
"learning_rate": 5.278519837736223e-06,
|
| 15828 |
+
"loss": 2.2623,
|
| 15829 |
+
"step": 9040
|
| 15830 |
+
},
|
| 15831 |
+
{
|
| 15832 |
+
"epoch": 0.89,
|
| 15833 |
+
"grad_norm": 1.93602454662323,
|
| 15834 |
+
"learning_rate": 5.258731572177699e-06,
|
| 15835 |
+
"loss": 2.4931,
|
| 15836 |
+
"step": 9044
|
| 15837 |
+
},
|
| 15838 |
+
{
|
| 15839 |
+
"epoch": 0.9,
|
| 15840 |
+
"grad_norm": 1.6569011211395264,
|
| 15841 |
+
"learning_rate": 5.238943306619174e-06,
|
| 15842 |
+
"loss": 2.4178,
|
| 15843 |
+
"step": 9048
|
| 15844 |
+
},
|
| 15845 |
+
{
|
| 15846 |
+
"epoch": 0.9,
|
| 15847 |
+
"grad_norm": 1.6328504085540771,
|
| 15848 |
+
"learning_rate": 5.219155041060652e-06,
|
| 15849 |
+
"loss": 2.6022,
|
| 15850 |
+
"step": 9052
|
| 15851 |
+
},
|
| 15852 |
+
{
|
| 15853 |
+
"epoch": 0.9,
|
| 15854 |
+
"grad_norm": 1.7538514137268066,
|
| 15855 |
+
"learning_rate": 5.199366775502128e-06,
|
| 15856 |
+
"loss": 2.3894,
|
| 15857 |
+
"step": 9056
|
| 15858 |
+
},
|
| 15859 |
+
{
|
| 15860 |
+
"epoch": 0.9,
|
| 15861 |
+
"grad_norm": 1.7447439432144165,
|
| 15862 |
+
"learning_rate": 5.179578509943603e-06,
|
| 15863 |
+
"loss": 2.3315,
|
| 15864 |
+
"step": 9060
|
| 15865 |
+
},
|
| 15866 |
+
{
|
| 15867 |
+
"epoch": 0.9,
|
| 15868 |
+
"grad_norm": 1.793074607849121,
|
| 15869 |
+
"learning_rate": 5.15979024438508e-06,
|
| 15870 |
+
"loss": 2.4952,
|
| 15871 |
+
"step": 9064
|
| 15872 |
+
},
|
| 15873 |
+
{
|
| 15874 |
+
"epoch": 0.9,
|
| 15875 |
+
"grad_norm": 1.6562350988388062,
|
| 15876 |
+
"learning_rate": 5.140001978826556e-06,
|
| 15877 |
+
"loss": 2.3996,
|
| 15878 |
+
"step": 9068
|
| 15879 |
+
},
|
| 15880 |
+
{
|
| 15881 |
+
"epoch": 0.9,
|
| 15882 |
+
"grad_norm": 1.7558027505874634,
|
| 15883 |
+
"learning_rate": 5.120213713268032e-06,
|
| 15884 |
+
"loss": 2.4778,
|
| 15885 |
+
"step": 9072
|
| 15886 |
+
},
|
| 15887 |
+
{
|
| 15888 |
+
"epoch": 0.9,
|
| 15889 |
+
"grad_norm": 1.6544588804244995,
|
| 15890 |
+
"learning_rate": 5.100425447709509e-06,
|
| 15891 |
+
"loss": 2.5229,
|
| 15892 |
+
"step": 9076
|
| 15893 |
+
},
|
| 15894 |
+
{
|
| 15895 |
+
"epoch": 0.9,
|
| 15896 |
+
"grad_norm": 1.775094985961914,
|
| 15897 |
+
"learning_rate": 5.080637182150984e-06,
|
| 15898 |
+
"loss": 2.3536,
|
| 15899 |
+
"step": 9080
|
| 15900 |
+
},
|
| 15901 |
+
{
|
| 15902 |
+
"epoch": 0.9,
|
| 15903 |
+
"grad_norm": 1.9515008926391602,
|
| 15904 |
+
"learning_rate": 5.060848916592461e-06,
|
| 15905 |
+
"loss": 2.4411,
|
| 15906 |
+
"step": 9084
|
| 15907 |
+
},
|
| 15908 |
+
{
|
| 15909 |
+
"epoch": 0.9,
|
| 15910 |
+
"grad_norm": 1.8084925413131714,
|
| 15911 |
+
"learning_rate": 5.041060651033938e-06,
|
| 15912 |
+
"loss": 2.3756,
|
| 15913 |
+
"step": 9088
|
| 15914 |
+
},
|
| 15915 |
+
{
|
| 15916 |
+
"epoch": 0.9,
|
| 15917 |
+
"grad_norm": 1.6303074359893799,
|
| 15918 |
+
"learning_rate": 5.021272385475413e-06,
|
| 15919 |
+
"loss": 2.3407,
|
| 15920 |
+
"step": 9092
|
| 15921 |
+
},
|
| 15922 |
+
{
|
| 15923 |
+
"epoch": 0.9,
|
| 15924 |
+
"grad_norm": 1.6556874513626099,
|
| 15925 |
+
"learning_rate": 5.0014841199168894e-06,
|
| 15926 |
+
"loss": 2.4766,
|
| 15927 |
+
"step": 9096
|
| 15928 |
}
|
| 15929 |
],
|
| 15930 |
"logging_steps": 4,
|
|
|
|
| 15932 |
"num_input_tokens_seen": 0,
|
| 15933 |
"num_train_epochs": 1,
|
| 15934 |
"save_steps": 1011,
|
| 15935 |
+
"total_flos": 1.1128110498761933e+17,
|
| 15936 |
"train_batch_size": 4,
|
| 15937 |
"trial_name": null,
|
| 15938 |
"trial_params": null
|