Commit
•
ceff349
1
Parent(s):
85d8d99
Training in progress, step 8088, 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:bad36090d3fea6bf5c95ba42db0a60c3eebb874b00583f59320f54010495d068
|
| 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 sha256:6081f069ca372d13b29b3847db2ea0377d5705d67197467dc5eb4bba64d27645
|
| 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 sha256:57e93e92bbbccd397547149085c6875813b56fd6bea353a98cf0179f1892adb6
|
| 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,
|
|
@@ -12390,6 +12390,1777 @@
|
|
| 12390 |
"learning_rate": 1.4994558226971406e-05,
|
| 12391 |
"loss": 2.4362,
|
| 12392 |
"step": 7076
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 12393 |
}
|
| 12394 |
],
|
| 12395 |
"logging_steps": 4,
|
|
@@ -12397,7 +14168,7 @@
|
|
| 12397 |
"num_input_tokens_seen": 0,
|
| 12398 |
"num_train_epochs": 1,
|
| 12399 |
"save_steps": 1011,
|
| 12400 |
-
"total_flos":
|
| 12401 |
"train_batch_size": 4,
|
| 12402 |
"trial_name": null,
|
| 12403 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.8002374591867023,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 8088,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 12390 |
"learning_rate": 1.4994558226971406e-05,
|
| 12391 |
"loss": 2.4362,
|
| 12392 |
"step": 7076
|
| 12393 |
+
},
|
| 12394 |
+
{
|
| 12395 |
+
"epoch": 0.7,
|
| 12396 |
+
"grad_norm": 1.5943050384521484,
|
| 12397 |
+
"learning_rate": 1.4974769961412883e-05,
|
| 12398 |
+
"loss": 2.5309,
|
| 12399 |
+
"step": 7080
|
| 12400 |
+
},
|
| 12401 |
+
{
|
| 12402 |
+
"epoch": 0.7,
|
| 12403 |
+
"grad_norm": 1.7004754543304443,
|
| 12404 |
+
"learning_rate": 1.4954981695854359e-05,
|
| 12405 |
+
"loss": 2.573,
|
| 12406 |
+
"step": 7084
|
| 12407 |
+
},
|
| 12408 |
+
{
|
| 12409 |
+
"epoch": 0.7,
|
| 12410 |
+
"grad_norm": 1.6682156324386597,
|
| 12411 |
+
"learning_rate": 1.4935193430295835e-05,
|
| 12412 |
+
"loss": 2.5718,
|
| 12413 |
+
"step": 7088
|
| 12414 |
+
},
|
| 12415 |
+
{
|
| 12416 |
+
"epoch": 0.7,
|
| 12417 |
+
"grad_norm": 1.489130973815918,
|
| 12418 |
+
"learning_rate": 1.4915405164737312e-05,
|
| 12419 |
+
"loss": 2.7728,
|
| 12420 |
+
"step": 7092
|
| 12421 |
+
},
|
| 12422 |
+
{
|
| 12423 |
+
"epoch": 0.7,
|
| 12424 |
+
"grad_norm": 1.72616708278656,
|
| 12425 |
+
"learning_rate": 1.4895616899178788e-05,
|
| 12426 |
+
"loss": 2.6457,
|
| 12427 |
+
"step": 7096
|
| 12428 |
+
},
|
| 12429 |
+
{
|
| 12430 |
+
"epoch": 0.7,
|
| 12431 |
+
"grad_norm": 1.5968120098114014,
|
| 12432 |
+
"learning_rate": 1.4875828633620264e-05,
|
| 12433 |
+
"loss": 2.4886,
|
| 12434 |
+
"step": 7100
|
| 12435 |
+
},
|
| 12436 |
+
{
|
| 12437 |
+
"epoch": 0.7,
|
| 12438 |
+
"grad_norm": 1.7065930366516113,
|
| 12439 |
+
"learning_rate": 1.485604036806174e-05,
|
| 12440 |
+
"loss": 2.6894,
|
| 12441 |
+
"step": 7104
|
| 12442 |
+
},
|
| 12443 |
+
{
|
| 12444 |
+
"epoch": 0.7,
|
| 12445 |
+
"grad_norm": 1.8437420129776,
|
| 12446 |
+
"learning_rate": 1.4836252102503217e-05,
|
| 12447 |
+
"loss": 2.6103,
|
| 12448 |
+
"step": 7108
|
| 12449 |
+
},
|
| 12450 |
+
{
|
| 12451 |
+
"epoch": 0.7,
|
| 12452 |
+
"grad_norm": 1.5512233972549438,
|
| 12453 |
+
"learning_rate": 1.4816463836944692e-05,
|
| 12454 |
+
"loss": 2.4657,
|
| 12455 |
+
"step": 7112
|
| 12456 |
+
},
|
| 12457 |
+
{
|
| 12458 |
+
"epoch": 0.7,
|
| 12459 |
+
"grad_norm": 1.6278142929077148,
|
| 12460 |
+
"learning_rate": 1.4796675571386168e-05,
|
| 12461 |
+
"loss": 2.7127,
|
| 12462 |
+
"step": 7116
|
| 12463 |
+
},
|
| 12464 |
+
{
|
| 12465 |
+
"epoch": 0.7,
|
| 12466 |
+
"grad_norm": 1.5600184202194214,
|
| 12467 |
+
"learning_rate": 1.4776887305827644e-05,
|
| 12468 |
+
"loss": 2.4707,
|
| 12469 |
+
"step": 7120
|
| 12470 |
+
},
|
| 12471 |
+
{
|
| 12472 |
+
"epoch": 0.7,
|
| 12473 |
+
"grad_norm": 1.5605080127716064,
|
| 12474 |
+
"learning_rate": 1.475709904026912e-05,
|
| 12475 |
+
"loss": 2.5009,
|
| 12476 |
+
"step": 7124
|
| 12477 |
+
},
|
| 12478 |
+
{
|
| 12479 |
+
"epoch": 0.71,
|
| 12480 |
+
"grad_norm": 1.6376603841781616,
|
| 12481 |
+
"learning_rate": 1.4737310774710597e-05,
|
| 12482 |
+
"loss": 2.4116,
|
| 12483 |
+
"step": 7128
|
| 12484 |
+
},
|
| 12485 |
+
{
|
| 12486 |
+
"epoch": 0.71,
|
| 12487 |
+
"grad_norm": 1.5430164337158203,
|
| 12488 |
+
"learning_rate": 1.4717522509152073e-05,
|
| 12489 |
+
"loss": 2.4315,
|
| 12490 |
+
"step": 7132
|
| 12491 |
+
},
|
| 12492 |
+
{
|
| 12493 |
+
"epoch": 0.71,
|
| 12494 |
+
"grad_norm": 1.9249303340911865,
|
| 12495 |
+
"learning_rate": 1.469773424359355e-05,
|
| 12496 |
+
"loss": 2.5128,
|
| 12497 |
+
"step": 7136
|
| 12498 |
+
},
|
| 12499 |
+
{
|
| 12500 |
+
"epoch": 0.71,
|
| 12501 |
+
"grad_norm": 1.914553165435791,
|
| 12502 |
+
"learning_rate": 1.4677945978035026e-05,
|
| 12503 |
+
"loss": 2.5726,
|
| 12504 |
+
"step": 7140
|
| 12505 |
+
},
|
| 12506 |
+
{
|
| 12507 |
+
"epoch": 0.71,
|
| 12508 |
+
"grad_norm": 1.6890403032302856,
|
| 12509 |
+
"learning_rate": 1.4658157712476502e-05,
|
| 12510 |
+
"loss": 2.5233,
|
| 12511 |
+
"step": 7144
|
| 12512 |
+
},
|
| 12513 |
+
{
|
| 12514 |
+
"epoch": 0.71,
|
| 12515 |
+
"grad_norm": 1.6215218305587769,
|
| 12516 |
+
"learning_rate": 1.463836944691798e-05,
|
| 12517 |
+
"loss": 2.4541,
|
| 12518 |
+
"step": 7148
|
| 12519 |
+
},
|
| 12520 |
+
{
|
| 12521 |
+
"epoch": 0.71,
|
| 12522 |
+
"grad_norm": 1.6023436784744263,
|
| 12523 |
+
"learning_rate": 1.4618581181359457e-05,
|
| 12524 |
+
"loss": 2.5825,
|
| 12525 |
+
"step": 7152
|
| 12526 |
+
},
|
| 12527 |
+
{
|
| 12528 |
+
"epoch": 0.71,
|
| 12529 |
+
"grad_norm": 1.6800665855407715,
|
| 12530 |
+
"learning_rate": 1.459879291580093e-05,
|
| 12531 |
+
"loss": 2.5275,
|
| 12532 |
+
"step": 7156
|
| 12533 |
+
},
|
| 12534 |
+
{
|
| 12535 |
+
"epoch": 0.71,
|
| 12536 |
+
"grad_norm": 1.5266000032424927,
|
| 12537 |
+
"learning_rate": 1.4579004650242406e-05,
|
| 12538 |
+
"loss": 2.6414,
|
| 12539 |
+
"step": 7160
|
| 12540 |
+
},
|
| 12541 |
+
{
|
| 12542 |
+
"epoch": 0.71,
|
| 12543 |
+
"grad_norm": 1.660843014717102,
|
| 12544 |
+
"learning_rate": 1.4559216384683882e-05,
|
| 12545 |
+
"loss": 2.4934,
|
| 12546 |
+
"step": 7164
|
| 12547 |
+
},
|
| 12548 |
+
{
|
| 12549 |
+
"epoch": 0.71,
|
| 12550 |
+
"grad_norm": 1.5643970966339111,
|
| 12551 |
+
"learning_rate": 1.4539428119125358e-05,
|
| 12552 |
+
"loss": 2.4847,
|
| 12553 |
+
"step": 7168
|
| 12554 |
+
},
|
| 12555 |
+
{
|
| 12556 |
+
"epoch": 0.71,
|
| 12557 |
+
"grad_norm": 1.4681652784347534,
|
| 12558 |
+
"learning_rate": 1.4519639853566835e-05,
|
| 12559 |
+
"loss": 2.423,
|
| 12560 |
+
"step": 7172
|
| 12561 |
+
},
|
| 12562 |
+
{
|
| 12563 |
+
"epoch": 0.71,
|
| 12564 |
+
"grad_norm": 1.5900237560272217,
|
| 12565 |
+
"learning_rate": 1.4499851588008311e-05,
|
| 12566 |
+
"loss": 2.5645,
|
| 12567 |
+
"step": 7176
|
| 12568 |
+
},
|
| 12569 |
+
{
|
| 12570 |
+
"epoch": 0.71,
|
| 12571 |
+
"grad_norm": 1.6541297435760498,
|
| 12572 |
+
"learning_rate": 1.4480063322449789e-05,
|
| 12573 |
+
"loss": 2.5434,
|
| 12574 |
+
"step": 7180
|
| 12575 |
+
},
|
| 12576 |
+
{
|
| 12577 |
+
"epoch": 0.71,
|
| 12578 |
+
"grad_norm": 1.5105384588241577,
|
| 12579 |
+
"learning_rate": 1.4460275056891265e-05,
|
| 12580 |
+
"loss": 2.6391,
|
| 12581 |
+
"step": 7184
|
| 12582 |
+
},
|
| 12583 |
+
{
|
| 12584 |
+
"epoch": 0.71,
|
| 12585 |
+
"grad_norm": 1.6248685121536255,
|
| 12586 |
+
"learning_rate": 1.4440486791332742e-05,
|
| 12587 |
+
"loss": 2.6772,
|
| 12588 |
+
"step": 7188
|
| 12589 |
+
},
|
| 12590 |
+
{
|
| 12591 |
+
"epoch": 0.71,
|
| 12592 |
+
"grad_norm": 1.559591293334961,
|
| 12593 |
+
"learning_rate": 1.4420698525774218e-05,
|
| 12594 |
+
"loss": 2.5175,
|
| 12595 |
+
"step": 7192
|
| 12596 |
+
},
|
| 12597 |
+
{
|
| 12598 |
+
"epoch": 0.71,
|
| 12599 |
+
"grad_norm": 1.73567533493042,
|
| 12600 |
+
"learning_rate": 1.4400910260215691e-05,
|
| 12601 |
+
"loss": 2.5799,
|
| 12602 |
+
"step": 7196
|
| 12603 |
+
},
|
| 12604 |
+
{
|
| 12605 |
+
"epoch": 0.71,
|
| 12606 |
+
"grad_norm": 1.602149248123169,
|
| 12607 |
+
"learning_rate": 1.4381121994657167e-05,
|
| 12608 |
+
"loss": 2.5475,
|
| 12609 |
+
"step": 7200
|
| 12610 |
+
},
|
| 12611 |
+
{
|
| 12612 |
+
"epoch": 0.71,
|
| 12613 |
+
"grad_norm": 1.6269313097000122,
|
| 12614 |
+
"learning_rate": 1.4361333729098644e-05,
|
| 12615 |
+
"loss": 2.5729,
|
| 12616 |
+
"step": 7204
|
| 12617 |
+
},
|
| 12618 |
+
{
|
| 12619 |
+
"epoch": 0.71,
|
| 12620 |
+
"grad_norm": 1.5919511318206787,
|
| 12621 |
+
"learning_rate": 1.4341545463540122e-05,
|
| 12622 |
+
"loss": 2.4055,
|
| 12623 |
+
"step": 7208
|
| 12624 |
+
},
|
| 12625 |
+
{
|
| 12626 |
+
"epoch": 0.71,
|
| 12627 |
+
"grad_norm": 1.9625754356384277,
|
| 12628 |
+
"learning_rate": 1.4321757197981598e-05,
|
| 12629 |
+
"loss": 2.6867,
|
| 12630 |
+
"step": 7212
|
| 12631 |
+
},
|
| 12632 |
+
{
|
| 12633 |
+
"epoch": 0.71,
|
| 12634 |
+
"grad_norm": 1.7777873277664185,
|
| 12635 |
+
"learning_rate": 1.4301968932423074e-05,
|
| 12636 |
+
"loss": 2.7518,
|
| 12637 |
+
"step": 7216
|
| 12638 |
+
},
|
| 12639 |
+
{
|
| 12640 |
+
"epoch": 0.71,
|
| 12641 |
+
"grad_norm": 1.632881760597229,
|
| 12642 |
+
"learning_rate": 1.428218066686455e-05,
|
| 12643 |
+
"loss": 2.3954,
|
| 12644 |
+
"step": 7220
|
| 12645 |
+
},
|
| 12646 |
+
{
|
| 12647 |
+
"epoch": 0.71,
|
| 12648 |
+
"grad_norm": 1.7910826206207275,
|
| 12649 |
+
"learning_rate": 1.4262392401306027e-05,
|
| 12650 |
+
"loss": 2.5477,
|
| 12651 |
+
"step": 7224
|
| 12652 |
+
},
|
| 12653 |
+
{
|
| 12654 |
+
"epoch": 0.72,
|
| 12655 |
+
"grad_norm": 1.6172887086868286,
|
| 12656 |
+
"learning_rate": 1.4242604135747503e-05,
|
| 12657 |
+
"loss": 2.6665,
|
| 12658 |
+
"step": 7228
|
| 12659 |
+
},
|
| 12660 |
+
{
|
| 12661 |
+
"epoch": 0.72,
|
| 12662 |
+
"grad_norm": 1.7775293588638306,
|
| 12663 |
+
"learning_rate": 1.422281587018898e-05,
|
| 12664 |
+
"loss": 2.5237,
|
| 12665 |
+
"step": 7232
|
| 12666 |
+
},
|
| 12667 |
+
{
|
| 12668 |
+
"epoch": 0.72,
|
| 12669 |
+
"grad_norm": 1.6558274030685425,
|
| 12670 |
+
"learning_rate": 1.4203027604630456e-05,
|
| 12671 |
+
"loss": 2.5909,
|
| 12672 |
+
"step": 7236
|
| 12673 |
+
},
|
| 12674 |
+
{
|
| 12675 |
+
"epoch": 0.72,
|
| 12676 |
+
"grad_norm": 1.6271092891693115,
|
| 12677 |
+
"learning_rate": 1.418323933907193e-05,
|
| 12678 |
+
"loss": 2.5299,
|
| 12679 |
+
"step": 7240
|
| 12680 |
+
},
|
| 12681 |
+
{
|
| 12682 |
+
"epoch": 0.72,
|
| 12683 |
+
"grad_norm": 1.5768835544586182,
|
| 12684 |
+
"learning_rate": 1.4163451073513407e-05,
|
| 12685 |
+
"loss": 2.5376,
|
| 12686 |
+
"step": 7244
|
| 12687 |
+
},
|
| 12688 |
+
{
|
| 12689 |
+
"epoch": 0.72,
|
| 12690 |
+
"grad_norm": 1.641741156578064,
|
| 12691 |
+
"learning_rate": 1.4143662807954883e-05,
|
| 12692 |
+
"loss": 2.5802,
|
| 12693 |
+
"step": 7248
|
| 12694 |
+
},
|
| 12695 |
+
{
|
| 12696 |
+
"epoch": 0.72,
|
| 12697 |
+
"grad_norm": 1.5866329669952393,
|
| 12698 |
+
"learning_rate": 1.412387454239636e-05,
|
| 12699 |
+
"loss": 2.4445,
|
| 12700 |
+
"step": 7252
|
| 12701 |
+
},
|
| 12702 |
+
{
|
| 12703 |
+
"epoch": 0.72,
|
| 12704 |
+
"grad_norm": 1.6858747005462646,
|
| 12705 |
+
"learning_rate": 1.4104086276837836e-05,
|
| 12706 |
+
"loss": 2.51,
|
| 12707 |
+
"step": 7256
|
| 12708 |
+
},
|
| 12709 |
+
{
|
| 12710 |
+
"epoch": 0.72,
|
| 12711 |
+
"grad_norm": 1.6350436210632324,
|
| 12712 |
+
"learning_rate": 1.4084298011279312e-05,
|
| 12713 |
+
"loss": 2.6297,
|
| 12714 |
+
"step": 7260
|
| 12715 |
+
},
|
| 12716 |
+
{
|
| 12717 |
+
"epoch": 0.72,
|
| 12718 |
+
"grad_norm": 1.7628724575042725,
|
| 12719 |
+
"learning_rate": 1.4064509745720788e-05,
|
| 12720 |
+
"loss": 2.6294,
|
| 12721 |
+
"step": 7264
|
| 12722 |
+
},
|
| 12723 |
+
{
|
| 12724 |
+
"epoch": 0.72,
|
| 12725 |
+
"grad_norm": 1.6731958389282227,
|
| 12726 |
+
"learning_rate": 1.4044721480162265e-05,
|
| 12727 |
+
"loss": 2.5577,
|
| 12728 |
+
"step": 7268
|
| 12729 |
+
},
|
| 12730 |
+
{
|
| 12731 |
+
"epoch": 0.72,
|
| 12732 |
+
"grad_norm": 1.7393099069595337,
|
| 12733 |
+
"learning_rate": 1.4024933214603741e-05,
|
| 12734 |
+
"loss": 2.6079,
|
| 12735 |
+
"step": 7272
|
| 12736 |
+
},
|
| 12737 |
+
{
|
| 12738 |
+
"epoch": 0.72,
|
| 12739 |
+
"grad_norm": 1.7639967203140259,
|
| 12740 |
+
"learning_rate": 1.4005144949045217e-05,
|
| 12741 |
+
"loss": 2.7389,
|
| 12742 |
+
"step": 7276
|
| 12743 |
+
},
|
| 12744 |
+
{
|
| 12745 |
+
"epoch": 0.72,
|
| 12746 |
+
"grad_norm": 1.5961958169937134,
|
| 12747 |
+
"learning_rate": 1.3985356683486692e-05,
|
| 12748 |
+
"loss": 2.5846,
|
| 12749 |
+
"step": 7280
|
| 12750 |
+
},
|
| 12751 |
+
{
|
| 12752 |
+
"epoch": 0.72,
|
| 12753 |
+
"grad_norm": 1.6526871919631958,
|
| 12754 |
+
"learning_rate": 1.3965568417928168e-05,
|
| 12755 |
+
"loss": 2.5965,
|
| 12756 |
+
"step": 7284
|
| 12757 |
+
},
|
| 12758 |
+
{
|
| 12759 |
+
"epoch": 0.72,
|
| 12760 |
+
"grad_norm": 1.7771737575531006,
|
| 12761 |
+
"learning_rate": 1.3945780152369645e-05,
|
| 12762 |
+
"loss": 2.5905,
|
| 12763 |
+
"step": 7288
|
| 12764 |
+
},
|
| 12765 |
+
{
|
| 12766 |
+
"epoch": 0.72,
|
| 12767 |
+
"grad_norm": 1.713395118713379,
|
| 12768 |
+
"learning_rate": 1.3925991886811121e-05,
|
| 12769 |
+
"loss": 2.5063,
|
| 12770 |
+
"step": 7292
|
| 12771 |
+
},
|
| 12772 |
+
{
|
| 12773 |
+
"epoch": 0.72,
|
| 12774 |
+
"grad_norm": 1.5294148921966553,
|
| 12775 |
+
"learning_rate": 1.3906203621252597e-05,
|
| 12776 |
+
"loss": 2.4081,
|
| 12777 |
+
"step": 7296
|
| 12778 |
+
},
|
| 12779 |
+
{
|
| 12780 |
+
"epoch": 0.72,
|
| 12781 |
+
"grad_norm": 1.6980494260787964,
|
| 12782 |
+
"learning_rate": 1.3886415355694074e-05,
|
| 12783 |
+
"loss": 2.4693,
|
| 12784 |
+
"step": 7300
|
| 12785 |
+
},
|
| 12786 |
+
{
|
| 12787 |
+
"epoch": 0.72,
|
| 12788 |
+
"grad_norm": 1.703873634338379,
|
| 12789 |
+
"learning_rate": 1.386662709013555e-05,
|
| 12790 |
+
"loss": 2.4529,
|
| 12791 |
+
"step": 7304
|
| 12792 |
+
},
|
| 12793 |
+
{
|
| 12794 |
+
"epoch": 0.72,
|
| 12795 |
+
"grad_norm": 1.5093457698822021,
|
| 12796 |
+
"learning_rate": 1.3846838824577026e-05,
|
| 12797 |
+
"loss": 2.439,
|
| 12798 |
+
"step": 7308
|
| 12799 |
+
},
|
| 12800 |
+
{
|
| 12801 |
+
"epoch": 0.72,
|
| 12802 |
+
"grad_norm": 1.6407102346420288,
|
| 12803 |
+
"learning_rate": 1.3827050559018503e-05,
|
| 12804 |
+
"loss": 2.4581,
|
| 12805 |
+
"step": 7312
|
| 12806 |
+
},
|
| 12807 |
+
{
|
| 12808 |
+
"epoch": 0.72,
|
| 12809 |
+
"grad_norm": 1.485217809677124,
|
| 12810 |
+
"learning_rate": 1.380726229345998e-05,
|
| 12811 |
+
"loss": 2.5162,
|
| 12812 |
+
"step": 7316
|
| 12813 |
+
},
|
| 12814 |
+
{
|
| 12815 |
+
"epoch": 0.72,
|
| 12816 |
+
"grad_norm": 1.5687867403030396,
|
| 12817 |
+
"learning_rate": 1.3787474027901457e-05,
|
| 12818 |
+
"loss": 2.4868,
|
| 12819 |
+
"step": 7320
|
| 12820 |
+
},
|
| 12821 |
+
{
|
| 12822 |
+
"epoch": 0.72,
|
| 12823 |
+
"grad_norm": 1.5972037315368652,
|
| 12824 |
+
"learning_rate": 1.376768576234293e-05,
|
| 12825 |
+
"loss": 2.5713,
|
| 12826 |
+
"step": 7324
|
| 12827 |
+
},
|
| 12828 |
+
{
|
| 12829 |
+
"epoch": 0.73,
|
| 12830 |
+
"grad_norm": 1.6357609033584595,
|
| 12831 |
+
"learning_rate": 1.3747897496784406e-05,
|
| 12832 |
+
"loss": 2.5051,
|
| 12833 |
+
"step": 7328
|
| 12834 |
+
},
|
| 12835 |
+
{
|
| 12836 |
+
"epoch": 0.73,
|
| 12837 |
+
"grad_norm": 1.7274508476257324,
|
| 12838 |
+
"learning_rate": 1.3728109231225883e-05,
|
| 12839 |
+
"loss": 2.596,
|
| 12840 |
+
"step": 7332
|
| 12841 |
+
},
|
| 12842 |
+
{
|
| 12843 |
+
"epoch": 0.73,
|
| 12844 |
+
"grad_norm": 1.9209715127944946,
|
| 12845 |
+
"learning_rate": 1.3708320965667359e-05,
|
| 12846 |
+
"loss": 2.7628,
|
| 12847 |
+
"step": 7336
|
| 12848 |
+
},
|
| 12849 |
+
{
|
| 12850 |
+
"epoch": 0.73,
|
| 12851 |
+
"grad_norm": 1.5816470384597778,
|
| 12852 |
+
"learning_rate": 1.3688532700108835e-05,
|
| 12853 |
+
"loss": 2.4371,
|
| 12854 |
+
"step": 7340
|
| 12855 |
+
},
|
| 12856 |
+
{
|
| 12857 |
+
"epoch": 0.73,
|
| 12858 |
+
"grad_norm": 1.6190577745437622,
|
| 12859 |
+
"learning_rate": 1.3668744434550313e-05,
|
| 12860 |
+
"loss": 2.4256,
|
| 12861 |
+
"step": 7344
|
| 12862 |
+
},
|
| 12863 |
+
{
|
| 12864 |
+
"epoch": 0.73,
|
| 12865 |
+
"grad_norm": 1.627508282661438,
|
| 12866 |
+
"learning_rate": 1.364895616899179e-05,
|
| 12867 |
+
"loss": 2.6653,
|
| 12868 |
+
"step": 7348
|
| 12869 |
+
},
|
| 12870 |
+
{
|
| 12871 |
+
"epoch": 0.73,
|
| 12872 |
+
"grad_norm": 1.6422009468078613,
|
| 12873 |
+
"learning_rate": 1.3629167903433266e-05,
|
| 12874 |
+
"loss": 2.6699,
|
| 12875 |
+
"step": 7352
|
| 12876 |
+
},
|
| 12877 |
+
{
|
| 12878 |
+
"epoch": 0.73,
|
| 12879 |
+
"grad_norm": 1.546067476272583,
|
| 12880 |
+
"learning_rate": 1.3609379637874742e-05,
|
| 12881 |
+
"loss": 2.4263,
|
| 12882 |
+
"step": 7356
|
| 12883 |
+
},
|
| 12884 |
+
{
|
| 12885 |
+
"epoch": 0.73,
|
| 12886 |
+
"grad_norm": 1.621297836303711,
|
| 12887 |
+
"learning_rate": 1.3589591372316218e-05,
|
| 12888 |
+
"loss": 2.4296,
|
| 12889 |
+
"step": 7360
|
| 12890 |
+
},
|
| 12891 |
+
{
|
| 12892 |
+
"epoch": 0.73,
|
| 12893 |
+
"grad_norm": 1.643576741218567,
|
| 12894 |
+
"learning_rate": 1.3569803106757691e-05,
|
| 12895 |
+
"loss": 2.5806,
|
| 12896 |
+
"step": 7364
|
| 12897 |
+
},
|
| 12898 |
+
{
|
| 12899 |
+
"epoch": 0.73,
|
| 12900 |
+
"grad_norm": 1.6136977672576904,
|
| 12901 |
+
"learning_rate": 1.3550014841199168e-05,
|
| 12902 |
+
"loss": 2.4887,
|
| 12903 |
+
"step": 7368
|
| 12904 |
+
},
|
| 12905 |
+
{
|
| 12906 |
+
"epoch": 0.73,
|
| 12907 |
+
"grad_norm": 1.704002857208252,
|
| 12908 |
+
"learning_rate": 1.3530226575640644e-05,
|
| 12909 |
+
"loss": 2.583,
|
| 12910 |
+
"step": 7372
|
| 12911 |
+
},
|
| 12912 |
+
{
|
| 12913 |
+
"epoch": 0.73,
|
| 12914 |
+
"grad_norm": 1.6694648265838623,
|
| 12915 |
+
"learning_rate": 1.3510438310082122e-05,
|
| 12916 |
+
"loss": 2.4127,
|
| 12917 |
+
"step": 7376
|
| 12918 |
+
},
|
| 12919 |
+
{
|
| 12920 |
+
"epoch": 0.73,
|
| 12921 |
+
"grad_norm": 1.6585307121276855,
|
| 12922 |
+
"learning_rate": 1.3490650044523598e-05,
|
| 12923 |
+
"loss": 2.4128,
|
| 12924 |
+
"step": 7380
|
| 12925 |
+
},
|
| 12926 |
+
{
|
| 12927 |
+
"epoch": 0.73,
|
| 12928 |
+
"grad_norm": 1.7365046739578247,
|
| 12929 |
+
"learning_rate": 1.3470861778965075e-05,
|
| 12930 |
+
"loss": 2.6269,
|
| 12931 |
+
"step": 7384
|
| 12932 |
+
},
|
| 12933 |
+
{
|
| 12934 |
+
"epoch": 0.73,
|
| 12935 |
+
"grad_norm": 1.6612383127212524,
|
| 12936 |
+
"learning_rate": 1.3451073513406551e-05,
|
| 12937 |
+
"loss": 2.5896,
|
| 12938 |
+
"step": 7388
|
| 12939 |
+
},
|
| 12940 |
+
{
|
| 12941 |
+
"epoch": 0.73,
|
| 12942 |
+
"grad_norm": 1.7151292562484741,
|
| 12943 |
+
"learning_rate": 1.3431285247848027e-05,
|
| 12944 |
+
"loss": 2.4705,
|
| 12945 |
+
"step": 7392
|
| 12946 |
+
},
|
| 12947 |
+
{
|
| 12948 |
+
"epoch": 0.73,
|
| 12949 |
+
"grad_norm": 1.608148455619812,
|
| 12950 |
+
"learning_rate": 1.3411496982289504e-05,
|
| 12951 |
+
"loss": 2.4956,
|
| 12952 |
+
"step": 7396
|
| 12953 |
+
},
|
| 12954 |
+
{
|
| 12955 |
+
"epoch": 0.73,
|
| 12956 |
+
"grad_norm": 1.9409472942352295,
|
| 12957 |
+
"learning_rate": 1.339170871673098e-05,
|
| 12958 |
+
"loss": 2.5163,
|
| 12959 |
+
"step": 7400
|
| 12960 |
+
},
|
| 12961 |
+
{
|
| 12962 |
+
"epoch": 0.73,
|
| 12963 |
+
"grad_norm": 1.6790823936462402,
|
| 12964 |
+
"learning_rate": 1.3371920451172456e-05,
|
| 12965 |
+
"loss": 2.5319,
|
| 12966 |
+
"step": 7404
|
| 12967 |
+
},
|
| 12968 |
+
{
|
| 12969 |
+
"epoch": 0.73,
|
| 12970 |
+
"grad_norm": 1.833620548248291,
|
| 12971 |
+
"learning_rate": 1.3352132185613931e-05,
|
| 12972 |
+
"loss": 2.5467,
|
| 12973 |
+
"step": 7408
|
| 12974 |
+
},
|
| 12975 |
+
{
|
| 12976 |
+
"epoch": 0.73,
|
| 12977 |
+
"grad_norm": 1.7631961107254028,
|
| 12978 |
+
"learning_rate": 1.3332343920055407e-05,
|
| 12979 |
+
"loss": 2.5748,
|
| 12980 |
+
"step": 7412
|
| 12981 |
+
},
|
| 12982 |
+
{
|
| 12983 |
+
"epoch": 0.73,
|
| 12984 |
+
"grad_norm": 1.5768530368804932,
|
| 12985 |
+
"learning_rate": 1.3312555654496884e-05,
|
| 12986 |
+
"loss": 2.4908,
|
| 12987 |
+
"step": 7416
|
| 12988 |
+
},
|
| 12989 |
+
{
|
| 12990 |
+
"epoch": 0.73,
|
| 12991 |
+
"grad_norm": 1.6658759117126465,
|
| 12992 |
+
"learning_rate": 1.329276738893836e-05,
|
| 12993 |
+
"loss": 2.5698,
|
| 12994 |
+
"step": 7420
|
| 12995 |
+
},
|
| 12996 |
+
{
|
| 12997 |
+
"epoch": 0.73,
|
| 12998 |
+
"grad_norm": 1.7477085590362549,
|
| 12999 |
+
"learning_rate": 1.3272979123379836e-05,
|
| 13000 |
+
"loss": 2.3864,
|
| 13001 |
+
"step": 7424
|
| 13002 |
+
},
|
| 13003 |
+
{
|
| 13004 |
+
"epoch": 0.73,
|
| 13005 |
+
"grad_norm": 1.6787872314453125,
|
| 13006 |
+
"learning_rate": 1.3253190857821313e-05,
|
| 13007 |
+
"loss": 2.5065,
|
| 13008 |
+
"step": 7428
|
| 13009 |
+
},
|
| 13010 |
+
{
|
| 13011 |
+
"epoch": 0.74,
|
| 13012 |
+
"grad_norm": 1.6128181219100952,
|
| 13013 |
+
"learning_rate": 1.3233402592262789e-05,
|
| 13014 |
+
"loss": 2.6635,
|
| 13015 |
+
"step": 7432
|
| 13016 |
+
},
|
| 13017 |
+
{
|
| 13018 |
+
"epoch": 0.74,
|
| 13019 |
+
"grad_norm": 1.4591010808944702,
|
| 13020 |
+
"learning_rate": 1.3213614326704265e-05,
|
| 13021 |
+
"loss": 2.4067,
|
| 13022 |
+
"step": 7436
|
| 13023 |
+
},
|
| 13024 |
+
{
|
| 13025 |
+
"epoch": 0.74,
|
| 13026 |
+
"grad_norm": 1.7754063606262207,
|
| 13027 |
+
"learning_rate": 1.3193826061145742e-05,
|
| 13028 |
+
"loss": 2.5308,
|
| 13029 |
+
"step": 7440
|
| 13030 |
+
},
|
| 13031 |
+
{
|
| 13032 |
+
"epoch": 0.74,
|
| 13033 |
+
"grad_norm": 1.708154559135437,
|
| 13034 |
+
"learning_rate": 1.3174037795587218e-05,
|
| 13035 |
+
"loss": 2.4604,
|
| 13036 |
+
"step": 7444
|
| 13037 |
+
},
|
| 13038 |
+
{
|
| 13039 |
+
"epoch": 0.74,
|
| 13040 |
+
"grad_norm": 1.539201021194458,
|
| 13041 |
+
"learning_rate": 1.3154249530028694e-05,
|
| 13042 |
+
"loss": 2.5372,
|
| 13043 |
+
"step": 7448
|
| 13044 |
+
},
|
| 13045 |
+
{
|
| 13046 |
+
"epoch": 0.74,
|
| 13047 |
+
"grad_norm": 1.8630115985870361,
|
| 13048 |
+
"learning_rate": 1.3134461264470169e-05,
|
| 13049 |
+
"loss": 2.5022,
|
| 13050 |
+
"step": 7452
|
| 13051 |
+
},
|
| 13052 |
+
{
|
| 13053 |
+
"epoch": 0.74,
|
| 13054 |
+
"grad_norm": 1.7771594524383545,
|
| 13055 |
+
"learning_rate": 1.3114672998911645e-05,
|
| 13056 |
+
"loss": 2.3826,
|
| 13057 |
+
"step": 7456
|
| 13058 |
+
},
|
| 13059 |
+
{
|
| 13060 |
+
"epoch": 0.74,
|
| 13061 |
+
"grad_norm": 1.8175123929977417,
|
| 13062 |
+
"learning_rate": 1.3094884733353121e-05,
|
| 13063 |
+
"loss": 2.5922,
|
| 13064 |
+
"step": 7460
|
| 13065 |
+
},
|
| 13066 |
+
{
|
| 13067 |
+
"epoch": 0.74,
|
| 13068 |
+
"grad_norm": 1.4767266511917114,
|
| 13069 |
+
"learning_rate": 1.3075096467794598e-05,
|
| 13070 |
+
"loss": 2.5334,
|
| 13071 |
+
"step": 7464
|
| 13072 |
+
},
|
| 13073 |
+
{
|
| 13074 |
+
"epoch": 0.74,
|
| 13075 |
+
"grad_norm": 1.7710050344467163,
|
| 13076 |
+
"learning_rate": 1.3055308202236074e-05,
|
| 13077 |
+
"loss": 2.5901,
|
| 13078 |
+
"step": 7468
|
| 13079 |
+
},
|
| 13080 |
+
{
|
| 13081 |
+
"epoch": 0.74,
|
| 13082 |
+
"grad_norm": 1.78278648853302,
|
| 13083 |
+
"learning_rate": 1.303551993667755e-05,
|
| 13084 |
+
"loss": 2.5621,
|
| 13085 |
+
"step": 7472
|
| 13086 |
+
},
|
| 13087 |
+
{
|
| 13088 |
+
"epoch": 0.74,
|
| 13089 |
+
"grad_norm": 1.553341269493103,
|
| 13090 |
+
"learning_rate": 1.3015731671119027e-05,
|
| 13091 |
+
"loss": 2.5296,
|
| 13092 |
+
"step": 7476
|
| 13093 |
+
},
|
| 13094 |
+
{
|
| 13095 |
+
"epoch": 0.74,
|
| 13096 |
+
"grad_norm": 1.4547873735427856,
|
| 13097 |
+
"learning_rate": 1.2995943405560505e-05,
|
| 13098 |
+
"loss": 2.5812,
|
| 13099 |
+
"step": 7480
|
| 13100 |
+
},
|
| 13101 |
+
{
|
| 13102 |
+
"epoch": 0.74,
|
| 13103 |
+
"grad_norm": 1.6034144163131714,
|
| 13104 |
+
"learning_rate": 1.2976155140001981e-05,
|
| 13105 |
+
"loss": 2.3507,
|
| 13106 |
+
"step": 7484
|
| 13107 |
+
},
|
| 13108 |
+
{
|
| 13109 |
+
"epoch": 0.74,
|
| 13110 |
+
"grad_norm": 1.6461652517318726,
|
| 13111 |
+
"learning_rate": 1.2956366874443457e-05,
|
| 13112 |
+
"loss": 2.6291,
|
| 13113 |
+
"step": 7488
|
| 13114 |
+
},
|
| 13115 |
+
{
|
| 13116 |
+
"epoch": 0.74,
|
| 13117 |
+
"grad_norm": 1.6804616451263428,
|
| 13118 |
+
"learning_rate": 1.293657860888493e-05,
|
| 13119 |
+
"loss": 2.4514,
|
| 13120 |
+
"step": 7492
|
| 13121 |
+
},
|
| 13122 |
+
{
|
| 13123 |
+
"epoch": 0.74,
|
| 13124 |
+
"grad_norm": 1.5887171030044556,
|
| 13125 |
+
"learning_rate": 1.2916790343326407e-05,
|
| 13126 |
+
"loss": 2.386,
|
| 13127 |
+
"step": 7496
|
| 13128 |
+
},
|
| 13129 |
+
{
|
| 13130 |
+
"epoch": 0.74,
|
| 13131 |
+
"grad_norm": 1.6213395595550537,
|
| 13132 |
+
"learning_rate": 1.2897002077767883e-05,
|
| 13133 |
+
"loss": 2.5569,
|
| 13134 |
+
"step": 7500
|
| 13135 |
+
},
|
| 13136 |
+
{
|
| 13137 |
+
"epoch": 0.74,
|
| 13138 |
+
"grad_norm": 1.7485177516937256,
|
| 13139 |
+
"learning_rate": 1.287721381220936e-05,
|
| 13140 |
+
"loss": 2.5454,
|
| 13141 |
+
"step": 7504
|
| 13142 |
+
},
|
| 13143 |
+
{
|
| 13144 |
+
"epoch": 0.74,
|
| 13145 |
+
"grad_norm": 1.660869836807251,
|
| 13146 |
+
"learning_rate": 1.2857425546650836e-05,
|
| 13147 |
+
"loss": 2.5177,
|
| 13148 |
+
"step": 7508
|
| 13149 |
+
},
|
| 13150 |
+
{
|
| 13151 |
+
"epoch": 0.74,
|
| 13152 |
+
"grad_norm": 1.674647331237793,
|
| 13153 |
+
"learning_rate": 1.2837637281092314e-05,
|
| 13154 |
+
"loss": 2.3855,
|
| 13155 |
+
"step": 7512
|
| 13156 |
+
},
|
| 13157 |
+
{
|
| 13158 |
+
"epoch": 0.74,
|
| 13159 |
+
"grad_norm": 2.053084373474121,
|
| 13160 |
+
"learning_rate": 1.281784901553379e-05,
|
| 13161 |
+
"loss": 2.7179,
|
| 13162 |
+
"step": 7516
|
| 13163 |
+
},
|
| 13164 |
+
{
|
| 13165 |
+
"epoch": 0.74,
|
| 13166 |
+
"grad_norm": 1.5778189897537231,
|
| 13167 |
+
"learning_rate": 1.2798060749975266e-05,
|
| 13168 |
+
"loss": 2.5341,
|
| 13169 |
+
"step": 7520
|
| 13170 |
+
},
|
| 13171 |
+
{
|
| 13172 |
+
"epoch": 0.74,
|
| 13173 |
+
"grad_norm": 1.670196294784546,
|
| 13174 |
+
"learning_rate": 1.2778272484416743e-05,
|
| 13175 |
+
"loss": 2.4186,
|
| 13176 |
+
"step": 7524
|
| 13177 |
+
},
|
| 13178 |
+
{
|
| 13179 |
+
"epoch": 0.74,
|
| 13180 |
+
"grad_norm": 1.7191470861434937,
|
| 13181 |
+
"learning_rate": 1.2758484218858219e-05,
|
| 13182 |
+
"loss": 2.4801,
|
| 13183 |
+
"step": 7528
|
| 13184 |
+
},
|
| 13185 |
+
{
|
| 13186 |
+
"epoch": 0.75,
|
| 13187 |
+
"grad_norm": 1.6857608556747437,
|
| 13188 |
+
"learning_rate": 1.2738695953299695e-05,
|
| 13189 |
+
"loss": 2.4867,
|
| 13190 |
+
"step": 7532
|
| 13191 |
+
},
|
| 13192 |
+
{
|
| 13193 |
+
"epoch": 0.75,
|
| 13194 |
+
"grad_norm": 1.6951642036437988,
|
| 13195 |
+
"learning_rate": 1.2718907687741168e-05,
|
| 13196 |
+
"loss": 2.5551,
|
| 13197 |
+
"step": 7536
|
| 13198 |
+
},
|
| 13199 |
+
{
|
| 13200 |
+
"epoch": 0.75,
|
| 13201 |
+
"grad_norm": 1.9836032390594482,
|
| 13202 |
+
"learning_rate": 1.2699119422182646e-05,
|
| 13203 |
+
"loss": 2.4202,
|
| 13204 |
+
"step": 7540
|
| 13205 |
+
},
|
| 13206 |
+
{
|
| 13207 |
+
"epoch": 0.75,
|
| 13208 |
+
"grad_norm": 1.6894361972808838,
|
| 13209 |
+
"learning_rate": 1.2679331156624122e-05,
|
| 13210 |
+
"loss": 2.5553,
|
| 13211 |
+
"step": 7544
|
| 13212 |
+
},
|
| 13213 |
+
{
|
| 13214 |
+
"epoch": 0.75,
|
| 13215 |
+
"grad_norm": 1.6373331546783447,
|
| 13216 |
+
"learning_rate": 1.2659542891065599e-05,
|
| 13217 |
+
"loss": 2.3549,
|
| 13218 |
+
"step": 7548
|
| 13219 |
+
},
|
| 13220 |
+
{
|
| 13221 |
+
"epoch": 0.75,
|
| 13222 |
+
"grad_norm": 1.5669854879379272,
|
| 13223 |
+
"learning_rate": 1.2639754625507075e-05,
|
| 13224 |
+
"loss": 2.4814,
|
| 13225 |
+
"step": 7552
|
| 13226 |
+
},
|
| 13227 |
+
{
|
| 13228 |
+
"epoch": 0.75,
|
| 13229 |
+
"grad_norm": 1.697510838508606,
|
| 13230 |
+
"learning_rate": 1.2619966359948551e-05,
|
| 13231 |
+
"loss": 2.5292,
|
| 13232 |
+
"step": 7556
|
| 13233 |
+
},
|
| 13234 |
+
{
|
| 13235 |
+
"epoch": 0.75,
|
| 13236 |
+
"grad_norm": 1.6051623821258545,
|
| 13237 |
+
"learning_rate": 1.2600178094390028e-05,
|
| 13238 |
+
"loss": 2.4635,
|
| 13239 |
+
"step": 7560
|
| 13240 |
+
},
|
| 13241 |
+
{
|
| 13242 |
+
"epoch": 0.75,
|
| 13243 |
+
"grad_norm": 1.550139307975769,
|
| 13244 |
+
"learning_rate": 1.2580389828831504e-05,
|
| 13245 |
+
"loss": 2.5052,
|
| 13246 |
+
"step": 7564
|
| 13247 |
+
},
|
| 13248 |
+
{
|
| 13249 |
+
"epoch": 0.75,
|
| 13250 |
+
"grad_norm": 1.8294285535812378,
|
| 13251 |
+
"learning_rate": 1.256060156327298e-05,
|
| 13252 |
+
"loss": 2.6251,
|
| 13253 |
+
"step": 7568
|
| 13254 |
+
},
|
| 13255 |
+
{
|
| 13256 |
+
"epoch": 0.75,
|
| 13257 |
+
"grad_norm": 1.731762170791626,
|
| 13258 |
+
"learning_rate": 1.2540813297714457e-05,
|
| 13259 |
+
"loss": 2.6812,
|
| 13260 |
+
"step": 7572
|
| 13261 |
+
},
|
| 13262 |
+
{
|
| 13263 |
+
"epoch": 0.75,
|
| 13264 |
+
"grad_norm": 1.7276403903961182,
|
| 13265 |
+
"learning_rate": 1.2521025032155931e-05,
|
| 13266 |
+
"loss": 2.5587,
|
| 13267 |
+
"step": 7576
|
| 13268 |
+
},
|
| 13269 |
+
{
|
| 13270 |
+
"epoch": 0.75,
|
| 13271 |
+
"grad_norm": 1.8602039813995361,
|
| 13272 |
+
"learning_rate": 1.2501236766597408e-05,
|
| 13273 |
+
"loss": 2.5649,
|
| 13274 |
+
"step": 7580
|
| 13275 |
+
},
|
| 13276 |
+
{
|
| 13277 |
+
"epoch": 0.75,
|
| 13278 |
+
"grad_norm": 1.5579040050506592,
|
| 13279 |
+
"learning_rate": 1.2481448501038886e-05,
|
| 13280 |
+
"loss": 2.6513,
|
| 13281 |
+
"step": 7584
|
| 13282 |
+
},
|
| 13283 |
+
{
|
| 13284 |
+
"epoch": 0.75,
|
| 13285 |
+
"grad_norm": 1.7131482362747192,
|
| 13286 |
+
"learning_rate": 1.246166023548036e-05,
|
| 13287 |
+
"loss": 2.4914,
|
| 13288 |
+
"step": 7588
|
| 13289 |
+
},
|
| 13290 |
+
{
|
| 13291 |
+
"epoch": 0.75,
|
| 13292 |
+
"grad_norm": 1.68119215965271,
|
| 13293 |
+
"learning_rate": 1.2441871969921837e-05,
|
| 13294 |
+
"loss": 2.5319,
|
| 13295 |
+
"step": 7592
|
| 13296 |
+
},
|
| 13297 |
+
{
|
| 13298 |
+
"epoch": 0.75,
|
| 13299 |
+
"grad_norm": 1.8896154165267944,
|
| 13300 |
+
"learning_rate": 1.2422083704363313e-05,
|
| 13301 |
+
"loss": 2.5683,
|
| 13302 |
+
"step": 7596
|
| 13303 |
+
},
|
| 13304 |
+
{
|
| 13305 |
+
"epoch": 0.75,
|
| 13306 |
+
"grad_norm": 1.5681391954421997,
|
| 13307 |
+
"learning_rate": 1.240229543880479e-05,
|
| 13308 |
+
"loss": 2.4047,
|
| 13309 |
+
"step": 7600
|
| 13310 |
+
},
|
| 13311 |
+
{
|
| 13312 |
+
"epoch": 0.75,
|
| 13313 |
+
"grad_norm": 1.4728906154632568,
|
| 13314 |
+
"learning_rate": 1.2382507173246266e-05,
|
| 13315 |
+
"loss": 2.5071,
|
| 13316 |
+
"step": 7604
|
| 13317 |
+
},
|
| 13318 |
+
{
|
| 13319 |
+
"epoch": 0.75,
|
| 13320 |
+
"grad_norm": 1.5875272750854492,
|
| 13321 |
+
"learning_rate": 1.2362718907687742e-05,
|
| 13322 |
+
"loss": 2.5415,
|
| 13323 |
+
"step": 7608
|
| 13324 |
+
},
|
| 13325 |
+
{
|
| 13326 |
+
"epoch": 0.75,
|
| 13327 |
+
"grad_norm": 1.6239516735076904,
|
| 13328 |
+
"learning_rate": 1.2342930642129218e-05,
|
| 13329 |
+
"loss": 2.4362,
|
| 13330 |
+
"step": 7612
|
| 13331 |
+
},
|
| 13332 |
+
{
|
| 13333 |
+
"epoch": 0.75,
|
| 13334 |
+
"grad_norm": 1.812610149383545,
|
| 13335 |
+
"learning_rate": 1.2323142376570695e-05,
|
| 13336 |
+
"loss": 2.5034,
|
| 13337 |
+
"step": 7616
|
| 13338 |
+
},
|
| 13339 |
+
{
|
| 13340 |
+
"epoch": 0.75,
|
| 13341 |
+
"grad_norm": 1.6190868616104126,
|
| 13342 |
+
"learning_rate": 1.2303354111012171e-05,
|
| 13343 |
+
"loss": 2.4786,
|
| 13344 |
+
"step": 7620
|
| 13345 |
+
},
|
| 13346 |
+
{
|
| 13347 |
+
"epoch": 0.75,
|
| 13348 |
+
"grad_norm": 1.6609690189361572,
|
| 13349 |
+
"learning_rate": 1.2283565845453647e-05,
|
| 13350 |
+
"loss": 2.372,
|
| 13351 |
+
"step": 7624
|
| 13352 |
+
},
|
| 13353 |
+
{
|
| 13354 |
+
"epoch": 0.75,
|
| 13355 |
+
"grad_norm": 1.700492024421692,
|
| 13356 |
+
"learning_rate": 1.2263777579895122e-05,
|
| 13357 |
+
"loss": 2.4527,
|
| 13358 |
+
"step": 7628
|
| 13359 |
+
},
|
| 13360 |
+
{
|
| 13361 |
+
"epoch": 0.76,
|
| 13362 |
+
"grad_norm": 1.8237019777297974,
|
| 13363 |
+
"learning_rate": 1.2243989314336598e-05,
|
| 13364 |
+
"loss": 2.548,
|
| 13365 |
+
"step": 7632
|
| 13366 |
+
},
|
| 13367 |
+
{
|
| 13368 |
+
"epoch": 0.76,
|
| 13369 |
+
"grad_norm": 1.558048129081726,
|
| 13370 |
+
"learning_rate": 1.2224201048778074e-05,
|
| 13371 |
+
"loss": 2.5439,
|
| 13372 |
+
"step": 7636
|
| 13373 |
+
},
|
| 13374 |
+
{
|
| 13375 |
+
"epoch": 0.76,
|
| 13376 |
+
"grad_norm": 1.7001163959503174,
|
| 13377 |
+
"learning_rate": 1.220441278321955e-05,
|
| 13378 |
+
"loss": 2.5146,
|
| 13379 |
+
"step": 7640
|
| 13380 |
+
},
|
| 13381 |
+
{
|
| 13382 |
+
"epoch": 0.76,
|
| 13383 |
+
"grad_norm": 1.5829248428344727,
|
| 13384 |
+
"learning_rate": 1.2184624517661027e-05,
|
| 13385 |
+
"loss": 2.5473,
|
| 13386 |
+
"step": 7644
|
| 13387 |
+
},
|
| 13388 |
+
{
|
| 13389 |
+
"epoch": 0.76,
|
| 13390 |
+
"grad_norm": 1.5922554731369019,
|
| 13391 |
+
"learning_rate": 1.2164836252102505e-05,
|
| 13392 |
+
"loss": 2.5482,
|
| 13393 |
+
"step": 7648
|
| 13394 |
+
},
|
| 13395 |
+
{
|
| 13396 |
+
"epoch": 0.76,
|
| 13397 |
+
"grad_norm": 1.5552897453308105,
|
| 13398 |
+
"learning_rate": 1.214504798654398e-05,
|
| 13399 |
+
"loss": 2.7146,
|
| 13400 |
+
"step": 7652
|
| 13401 |
+
},
|
| 13402 |
+
{
|
| 13403 |
+
"epoch": 0.76,
|
| 13404 |
+
"grad_norm": 1.6641972064971924,
|
| 13405 |
+
"learning_rate": 1.2125259720985456e-05,
|
| 13406 |
+
"loss": 2.5317,
|
| 13407 |
+
"step": 7656
|
| 13408 |
+
},
|
| 13409 |
+
{
|
| 13410 |
+
"epoch": 0.76,
|
| 13411 |
+
"grad_norm": 1.727538824081421,
|
| 13412 |
+
"learning_rate": 1.2105471455426932e-05,
|
| 13413 |
+
"loss": 2.511,
|
| 13414 |
+
"step": 7660
|
| 13415 |
+
},
|
| 13416 |
+
{
|
| 13417 |
+
"epoch": 0.76,
|
| 13418 |
+
"grad_norm": 1.7466334104537964,
|
| 13419 |
+
"learning_rate": 1.2085683189868409e-05,
|
| 13420 |
+
"loss": 2.6592,
|
| 13421 |
+
"step": 7664
|
| 13422 |
+
},
|
| 13423 |
+
{
|
| 13424 |
+
"epoch": 0.76,
|
| 13425 |
+
"grad_norm": 1.578805923461914,
|
| 13426 |
+
"learning_rate": 1.2065894924309885e-05,
|
| 13427 |
+
"loss": 2.5405,
|
| 13428 |
+
"step": 7668
|
| 13429 |
+
},
|
| 13430 |
+
{
|
| 13431 |
+
"epoch": 0.76,
|
| 13432 |
+
"grad_norm": 1.7704498767852783,
|
| 13433 |
+
"learning_rate": 1.204610665875136e-05,
|
| 13434 |
+
"loss": 2.7104,
|
| 13435 |
+
"step": 7672
|
| 13436 |
+
},
|
| 13437 |
+
{
|
| 13438 |
+
"epoch": 0.76,
|
| 13439 |
+
"grad_norm": 1.901253342628479,
|
| 13440 |
+
"learning_rate": 1.2026318393192838e-05,
|
| 13441 |
+
"loss": 2.68,
|
| 13442 |
+
"step": 7676
|
| 13443 |
+
},
|
| 13444 |
+
{
|
| 13445 |
+
"epoch": 0.76,
|
| 13446 |
+
"grad_norm": 1.6251075267791748,
|
| 13447 |
+
"learning_rate": 1.2006530127634314e-05,
|
| 13448 |
+
"loss": 2.5327,
|
| 13449 |
+
"step": 7680
|
| 13450 |
+
},
|
| 13451 |
+
{
|
| 13452 |
+
"epoch": 0.76,
|
| 13453 |
+
"grad_norm": 1.9433711767196655,
|
| 13454 |
+
"learning_rate": 1.198674186207579e-05,
|
| 13455 |
+
"loss": 2.664,
|
| 13456 |
+
"step": 7684
|
| 13457 |
+
},
|
| 13458 |
+
{
|
| 13459 |
+
"epoch": 0.76,
|
| 13460 |
+
"grad_norm": 1.6232560873031616,
|
| 13461 |
+
"learning_rate": 1.1966953596517267e-05,
|
| 13462 |
+
"loss": 2.3881,
|
| 13463 |
+
"step": 7688
|
| 13464 |
+
},
|
| 13465 |
+
{
|
| 13466 |
+
"epoch": 0.76,
|
| 13467 |
+
"grad_norm": 1.625562310218811,
|
| 13468 |
+
"learning_rate": 1.1947165330958741e-05,
|
| 13469 |
+
"loss": 2.3981,
|
| 13470 |
+
"step": 7692
|
| 13471 |
+
},
|
| 13472 |
+
{
|
| 13473 |
+
"epoch": 0.76,
|
| 13474 |
+
"grad_norm": 1.8203229904174805,
|
| 13475 |
+
"learning_rate": 1.1927377065400218e-05,
|
| 13476 |
+
"loss": 2.5587,
|
| 13477 |
+
"step": 7696
|
| 13478 |
+
},
|
| 13479 |
+
{
|
| 13480 |
+
"epoch": 0.76,
|
| 13481 |
+
"grad_norm": 1.5125036239624023,
|
| 13482 |
+
"learning_rate": 1.1907588799841694e-05,
|
| 13483 |
+
"loss": 2.548,
|
| 13484 |
+
"step": 7700
|
| 13485 |
+
},
|
| 13486 |
+
{
|
| 13487 |
+
"epoch": 0.76,
|
| 13488 |
+
"grad_norm": 1.7927870750427246,
|
| 13489 |
+
"learning_rate": 1.188780053428317e-05,
|
| 13490 |
+
"loss": 2.5317,
|
| 13491 |
+
"step": 7704
|
| 13492 |
+
},
|
| 13493 |
+
{
|
| 13494 |
+
"epoch": 0.76,
|
| 13495 |
+
"grad_norm": 1.853130578994751,
|
| 13496 |
+
"learning_rate": 1.1868012268724647e-05,
|
| 13497 |
+
"loss": 2.5032,
|
| 13498 |
+
"step": 7708
|
| 13499 |
+
},
|
| 13500 |
+
{
|
| 13501 |
+
"epoch": 0.76,
|
| 13502 |
+
"grad_norm": 1.8347195386886597,
|
| 13503 |
+
"learning_rate": 1.1848224003166123e-05,
|
| 13504 |
+
"loss": 2.3406,
|
| 13505 |
+
"step": 7712
|
| 13506 |
+
},
|
| 13507 |
+
{
|
| 13508 |
+
"epoch": 0.76,
|
| 13509 |
+
"grad_norm": 1.6292403936386108,
|
| 13510 |
+
"learning_rate": 1.18284357376076e-05,
|
| 13511 |
+
"loss": 2.4874,
|
| 13512 |
+
"step": 7716
|
| 13513 |
+
},
|
| 13514 |
+
{
|
| 13515 |
+
"epoch": 0.76,
|
| 13516 |
+
"grad_norm": 1.6553095579147339,
|
| 13517 |
+
"learning_rate": 1.1808647472049076e-05,
|
| 13518 |
+
"loss": 2.5065,
|
| 13519 |
+
"step": 7720
|
| 13520 |
+
},
|
| 13521 |
+
{
|
| 13522 |
+
"epoch": 0.76,
|
| 13523 |
+
"grad_norm": 1.9831068515777588,
|
| 13524 |
+
"learning_rate": 1.1788859206490552e-05,
|
| 13525 |
+
"loss": 2.5079,
|
| 13526 |
+
"step": 7724
|
| 13527 |
+
},
|
| 13528 |
+
{
|
| 13529 |
+
"epoch": 0.76,
|
| 13530 |
+
"grad_norm": 1.7396299839019775,
|
| 13531 |
+
"learning_rate": 1.1769070940932028e-05,
|
| 13532 |
+
"loss": 2.5171,
|
| 13533 |
+
"step": 7728
|
| 13534 |
+
},
|
| 13535 |
+
{
|
| 13536 |
+
"epoch": 0.77,
|
| 13537 |
+
"grad_norm": 1.8842509984970093,
|
| 13538 |
+
"learning_rate": 1.1749282675373505e-05,
|
| 13539 |
+
"loss": 2.6561,
|
| 13540 |
+
"step": 7732
|
| 13541 |
+
},
|
| 13542 |
+
{
|
| 13543 |
+
"epoch": 0.77,
|
| 13544 |
+
"grad_norm": 1.5185445547103882,
|
| 13545 |
+
"learning_rate": 1.1729494409814979e-05,
|
| 13546 |
+
"loss": 2.6638,
|
| 13547 |
+
"step": 7736
|
| 13548 |
+
},
|
| 13549 |
+
{
|
| 13550 |
+
"epoch": 0.77,
|
| 13551 |
+
"grad_norm": 1.7892464399337769,
|
| 13552 |
+
"learning_rate": 1.1709706144256455e-05,
|
| 13553 |
+
"loss": 2.4509,
|
| 13554 |
+
"step": 7740
|
| 13555 |
+
},
|
| 13556 |
+
{
|
| 13557 |
+
"epoch": 0.77,
|
| 13558 |
+
"grad_norm": 1.6770362854003906,
|
| 13559 |
+
"learning_rate": 1.1689917878697933e-05,
|
| 13560 |
+
"loss": 2.536,
|
| 13561 |
+
"step": 7744
|
| 13562 |
+
},
|
| 13563 |
+
{
|
| 13564 |
+
"epoch": 0.77,
|
| 13565 |
+
"grad_norm": 1.6587915420532227,
|
| 13566 |
+
"learning_rate": 1.167012961313941e-05,
|
| 13567 |
+
"loss": 2.515,
|
| 13568 |
+
"step": 7748
|
| 13569 |
+
},
|
| 13570 |
+
{
|
| 13571 |
+
"epoch": 0.77,
|
| 13572 |
+
"grad_norm": 1.7976752519607544,
|
| 13573 |
+
"learning_rate": 1.1650341347580886e-05,
|
| 13574 |
+
"loss": 2.4524,
|
| 13575 |
+
"step": 7752
|
| 13576 |
+
},
|
| 13577 |
+
{
|
| 13578 |
+
"epoch": 0.77,
|
| 13579 |
+
"grad_norm": 1.7145442962646484,
|
| 13580 |
+
"learning_rate": 1.163055308202236e-05,
|
| 13581 |
+
"loss": 2.56,
|
| 13582 |
+
"step": 7756
|
| 13583 |
+
},
|
| 13584 |
+
{
|
| 13585 |
+
"epoch": 0.77,
|
| 13586 |
+
"grad_norm": 1.6737264394760132,
|
| 13587 |
+
"learning_rate": 1.1610764816463837e-05,
|
| 13588 |
+
"loss": 2.45,
|
| 13589 |
+
"step": 7760
|
| 13590 |
+
},
|
| 13591 |
+
{
|
| 13592 |
+
"epoch": 0.77,
|
| 13593 |
+
"grad_norm": 1.683347225189209,
|
| 13594 |
+
"learning_rate": 1.1590976550905313e-05,
|
| 13595 |
+
"loss": 2.4508,
|
| 13596 |
+
"step": 7764
|
| 13597 |
+
},
|
| 13598 |
+
{
|
| 13599 |
+
"epoch": 0.77,
|
| 13600 |
+
"grad_norm": 1.6320801973342896,
|
| 13601 |
+
"learning_rate": 1.157118828534679e-05,
|
| 13602 |
+
"loss": 2.6438,
|
| 13603 |
+
"step": 7768
|
| 13604 |
+
},
|
| 13605 |
+
{
|
| 13606 |
+
"epoch": 0.77,
|
| 13607 |
+
"grad_norm": 1.8501825332641602,
|
| 13608 |
+
"learning_rate": 1.1551400019788266e-05,
|
| 13609 |
+
"loss": 2.4262,
|
| 13610 |
+
"step": 7772
|
| 13611 |
+
},
|
| 13612 |
+
{
|
| 13613 |
+
"epoch": 0.77,
|
| 13614 |
+
"grad_norm": 1.8582115173339844,
|
| 13615 |
+
"learning_rate": 1.1531611754229742e-05,
|
| 13616 |
+
"loss": 2.4314,
|
| 13617 |
+
"step": 7776
|
| 13618 |
+
},
|
| 13619 |
+
{
|
| 13620 |
+
"epoch": 0.77,
|
| 13621 |
+
"grad_norm": 1.5969737768173218,
|
| 13622 |
+
"learning_rate": 1.1511823488671219e-05,
|
| 13623 |
+
"loss": 2.4917,
|
| 13624 |
+
"step": 7780
|
| 13625 |
+
},
|
| 13626 |
+
{
|
| 13627 |
+
"epoch": 0.77,
|
| 13628 |
+
"grad_norm": 1.5447652339935303,
|
| 13629 |
+
"learning_rate": 1.1492035223112695e-05,
|
| 13630 |
+
"loss": 2.5134,
|
| 13631 |
+
"step": 7784
|
| 13632 |
+
},
|
| 13633 |
+
{
|
| 13634 |
+
"epoch": 0.77,
|
| 13635 |
+
"grad_norm": 1.7579408884048462,
|
| 13636 |
+
"learning_rate": 1.1472246957554171e-05,
|
| 13637 |
+
"loss": 2.4526,
|
| 13638 |
+
"step": 7788
|
| 13639 |
+
},
|
| 13640 |
+
{
|
| 13641 |
+
"epoch": 0.77,
|
| 13642 |
+
"grad_norm": 1.6958955526351929,
|
| 13643 |
+
"learning_rate": 1.1452458691995648e-05,
|
| 13644 |
+
"loss": 2.5719,
|
| 13645 |
+
"step": 7792
|
| 13646 |
+
},
|
| 13647 |
+
{
|
| 13648 |
+
"epoch": 0.77,
|
| 13649 |
+
"grad_norm": 1.6478573083877563,
|
| 13650 |
+
"learning_rate": 1.1432670426437124e-05,
|
| 13651 |
+
"loss": 2.5227,
|
| 13652 |
+
"step": 7796
|
| 13653 |
+
},
|
| 13654 |
+
{
|
| 13655 |
+
"epoch": 0.77,
|
| 13656 |
+
"grad_norm": 1.5438200235366821,
|
| 13657 |
+
"learning_rate": 1.1412882160878599e-05,
|
| 13658 |
+
"loss": 2.3835,
|
| 13659 |
+
"step": 7800
|
| 13660 |
+
},
|
| 13661 |
+
{
|
| 13662 |
+
"epoch": 0.77,
|
| 13663 |
+
"grad_norm": 1.5718754529953003,
|
| 13664 |
+
"learning_rate": 1.1393093895320075e-05,
|
| 13665 |
+
"loss": 2.4568,
|
| 13666 |
+
"step": 7804
|
| 13667 |
+
},
|
| 13668 |
+
{
|
| 13669 |
+
"epoch": 0.77,
|
| 13670 |
+
"grad_norm": 1.6531394720077515,
|
| 13671 |
+
"learning_rate": 1.1373305629761551e-05,
|
| 13672 |
+
"loss": 2.3707,
|
| 13673 |
+
"step": 7808
|
| 13674 |
+
},
|
| 13675 |
+
{
|
| 13676 |
+
"epoch": 0.77,
|
| 13677 |
+
"grad_norm": 1.8568472862243652,
|
| 13678 |
+
"learning_rate": 1.135351736420303e-05,
|
| 13679 |
+
"loss": 2.583,
|
| 13680 |
+
"step": 7812
|
| 13681 |
+
},
|
| 13682 |
+
{
|
| 13683 |
+
"epoch": 0.77,
|
| 13684 |
+
"grad_norm": 1.7812650203704834,
|
| 13685 |
+
"learning_rate": 1.1333729098644506e-05,
|
| 13686 |
+
"loss": 2.5865,
|
| 13687 |
+
"step": 7816
|
| 13688 |
+
},
|
| 13689 |
+
{
|
| 13690 |
+
"epoch": 0.77,
|
| 13691 |
+
"grad_norm": 1.7314536571502686,
|
| 13692 |
+
"learning_rate": 1.131394083308598e-05,
|
| 13693 |
+
"loss": 2.5384,
|
| 13694 |
+
"step": 7820
|
| 13695 |
+
},
|
| 13696 |
+
{
|
| 13697 |
+
"epoch": 0.77,
|
| 13698 |
+
"grad_norm": 1.8775840997695923,
|
| 13699 |
+
"learning_rate": 1.1294152567527457e-05,
|
| 13700 |
+
"loss": 2.6276,
|
| 13701 |
+
"step": 7824
|
| 13702 |
+
},
|
| 13703 |
+
{
|
| 13704 |
+
"epoch": 0.77,
|
| 13705 |
+
"grad_norm": 1.3948659896850586,
|
| 13706 |
+
"learning_rate": 1.1274364301968933e-05,
|
| 13707 |
+
"loss": 2.5302,
|
| 13708 |
+
"step": 7828
|
| 13709 |
+
},
|
| 13710 |
+
{
|
| 13711 |
+
"epoch": 0.77,
|
| 13712 |
+
"grad_norm": 1.690691590309143,
|
| 13713 |
+
"learning_rate": 1.125457603641041e-05,
|
| 13714 |
+
"loss": 2.5569,
|
| 13715 |
+
"step": 7832
|
| 13716 |
+
},
|
| 13717 |
+
{
|
| 13718 |
+
"epoch": 0.78,
|
| 13719 |
+
"grad_norm": 1.606754183769226,
|
| 13720 |
+
"learning_rate": 1.1234787770851885e-05,
|
| 13721 |
+
"loss": 2.3866,
|
| 13722 |
+
"step": 7836
|
| 13723 |
+
},
|
| 13724 |
+
{
|
| 13725 |
+
"epoch": 0.78,
|
| 13726 |
+
"grad_norm": 1.5848886966705322,
|
| 13727 |
+
"learning_rate": 1.1214999505293362e-05,
|
| 13728 |
+
"loss": 2.5407,
|
| 13729 |
+
"step": 7840
|
| 13730 |
+
},
|
| 13731 |
+
{
|
| 13732 |
+
"epoch": 0.78,
|
| 13733 |
+
"grad_norm": 1.5651729106903076,
|
| 13734 |
+
"learning_rate": 1.1195211239734838e-05,
|
| 13735 |
+
"loss": 2.2891,
|
| 13736 |
+
"step": 7844
|
| 13737 |
+
},
|
| 13738 |
+
{
|
| 13739 |
+
"epoch": 0.78,
|
| 13740 |
+
"grad_norm": 1.704177975654602,
|
| 13741 |
+
"learning_rate": 1.1175422974176314e-05,
|
| 13742 |
+
"loss": 2.6963,
|
| 13743 |
+
"step": 7848
|
| 13744 |
+
},
|
| 13745 |
+
{
|
| 13746 |
+
"epoch": 0.78,
|
| 13747 |
+
"grad_norm": 1.605334997177124,
|
| 13748 |
+
"learning_rate": 1.115563470861779e-05,
|
| 13749 |
+
"loss": 2.5576,
|
| 13750 |
+
"step": 7852
|
| 13751 |
+
},
|
| 13752 |
+
{
|
| 13753 |
+
"epoch": 0.78,
|
| 13754 |
+
"grad_norm": 1.530823826789856,
|
| 13755 |
+
"learning_rate": 1.1135846443059267e-05,
|
| 13756 |
+
"loss": 2.4859,
|
| 13757 |
+
"step": 7856
|
| 13758 |
+
},
|
| 13759 |
+
{
|
| 13760 |
+
"epoch": 0.78,
|
| 13761 |
+
"grad_norm": 1.6835219860076904,
|
| 13762 |
+
"learning_rate": 1.1116058177500742e-05,
|
| 13763 |
+
"loss": 2.462,
|
| 13764 |
+
"step": 7860
|
| 13765 |
+
},
|
| 13766 |
+
{
|
| 13767 |
+
"epoch": 0.78,
|
| 13768 |
+
"grad_norm": 1.6420457363128662,
|
| 13769 |
+
"learning_rate": 1.1096269911942218e-05,
|
| 13770 |
+
"loss": 2.6175,
|
| 13771 |
+
"step": 7864
|
| 13772 |
+
},
|
| 13773 |
+
{
|
| 13774 |
+
"epoch": 0.78,
|
| 13775 |
+
"grad_norm": 1.6330996751785278,
|
| 13776 |
+
"learning_rate": 1.1076481646383694e-05,
|
| 13777 |
+
"loss": 2.6308,
|
| 13778 |
+
"step": 7868
|
| 13779 |
+
},
|
| 13780 |
+
{
|
| 13781 |
+
"epoch": 0.78,
|
| 13782 |
+
"grad_norm": 2.031244993209839,
|
| 13783 |
+
"learning_rate": 1.105669338082517e-05,
|
| 13784 |
+
"loss": 2.4599,
|
| 13785 |
+
"step": 7872
|
| 13786 |
+
},
|
| 13787 |
+
{
|
| 13788 |
+
"epoch": 0.78,
|
| 13789 |
+
"grad_norm": 1.5154794454574585,
|
| 13790 |
+
"learning_rate": 1.1036905115266647e-05,
|
| 13791 |
+
"loss": 2.4949,
|
| 13792 |
+
"step": 7876
|
| 13793 |
+
},
|
| 13794 |
+
{
|
| 13795 |
+
"epoch": 0.78,
|
| 13796 |
+
"grad_norm": 1.7397664785385132,
|
| 13797 |
+
"learning_rate": 1.1017116849708125e-05,
|
| 13798 |
+
"loss": 2.3962,
|
| 13799 |
+
"step": 7880
|
| 13800 |
+
},
|
| 13801 |
+
{
|
| 13802 |
+
"epoch": 0.78,
|
| 13803 |
+
"grad_norm": 1.7765430212020874,
|
| 13804 |
+
"learning_rate": 1.09973285841496e-05,
|
| 13805 |
+
"loss": 2.4214,
|
| 13806 |
+
"step": 7884
|
| 13807 |
+
},
|
| 13808 |
+
{
|
| 13809 |
+
"epoch": 0.78,
|
| 13810 |
+
"grad_norm": 1.6177083253860474,
|
| 13811 |
+
"learning_rate": 1.0977540318591076e-05,
|
| 13812 |
+
"loss": 2.5221,
|
| 13813 |
+
"step": 7888
|
| 13814 |
+
},
|
| 13815 |
+
{
|
| 13816 |
+
"epoch": 0.78,
|
| 13817 |
+
"grad_norm": 1.6060839891433716,
|
| 13818 |
+
"learning_rate": 1.0957752053032552e-05,
|
| 13819 |
+
"loss": 2.6571,
|
| 13820 |
+
"step": 7892
|
| 13821 |
+
},
|
| 13822 |
+
{
|
| 13823 |
+
"epoch": 0.78,
|
| 13824 |
+
"grad_norm": 1.6707481145858765,
|
| 13825 |
+
"learning_rate": 1.0937963787474029e-05,
|
| 13826 |
+
"loss": 2.373,
|
| 13827 |
+
"step": 7896
|
| 13828 |
+
},
|
| 13829 |
+
{
|
| 13830 |
+
"epoch": 0.78,
|
| 13831 |
+
"grad_norm": 1.8542253971099854,
|
| 13832 |
+
"learning_rate": 1.0918175521915505e-05,
|
| 13833 |
+
"loss": 2.4029,
|
| 13834 |
+
"step": 7900
|
| 13835 |
+
},
|
| 13836 |
+
{
|
| 13837 |
+
"epoch": 0.78,
|
| 13838 |
+
"grad_norm": 1.6408201456069946,
|
| 13839 |
+
"learning_rate": 1.089838725635698e-05,
|
| 13840 |
+
"loss": 2.4065,
|
| 13841 |
+
"step": 7904
|
| 13842 |
+
},
|
| 13843 |
+
{
|
| 13844 |
+
"epoch": 0.78,
|
| 13845 |
+
"grad_norm": 1.5829615592956543,
|
| 13846 |
+
"learning_rate": 1.0878598990798458e-05,
|
| 13847 |
+
"loss": 2.6859,
|
| 13848 |
+
"step": 7908
|
| 13849 |
+
},
|
| 13850 |
+
{
|
| 13851 |
+
"epoch": 0.78,
|
| 13852 |
+
"grad_norm": 1.5252026319503784,
|
| 13853 |
+
"learning_rate": 1.0858810725239934e-05,
|
| 13854 |
+
"loss": 2.5326,
|
| 13855 |
+
"step": 7912
|
| 13856 |
+
},
|
| 13857 |
+
{
|
| 13858 |
+
"epoch": 0.78,
|
| 13859 |
+
"grad_norm": 1.5747402906417847,
|
| 13860 |
+
"learning_rate": 1.083902245968141e-05,
|
| 13861 |
+
"loss": 2.5311,
|
| 13862 |
+
"step": 7916
|
| 13863 |
+
},
|
| 13864 |
+
{
|
| 13865 |
+
"epoch": 0.78,
|
| 13866 |
+
"grad_norm": 1.592838168144226,
|
| 13867 |
+
"learning_rate": 1.0819234194122887e-05,
|
| 13868 |
+
"loss": 2.3582,
|
| 13869 |
+
"step": 7920
|
| 13870 |
+
},
|
| 13871 |
+
{
|
| 13872 |
+
"epoch": 0.78,
|
| 13873 |
+
"grad_norm": 1.5652263164520264,
|
| 13874 |
+
"learning_rate": 1.0799445928564361e-05,
|
| 13875 |
+
"loss": 2.5613,
|
| 13876 |
+
"step": 7924
|
| 13877 |
+
},
|
| 13878 |
+
{
|
| 13879 |
+
"epoch": 0.78,
|
| 13880 |
+
"grad_norm": 1.5245355367660522,
|
| 13881 |
+
"learning_rate": 1.0779657663005837e-05,
|
| 13882 |
+
"loss": 2.7002,
|
| 13883 |
+
"step": 7928
|
| 13884 |
+
},
|
| 13885 |
+
{
|
| 13886 |
+
"epoch": 0.78,
|
| 13887 |
+
"grad_norm": 1.724204421043396,
|
| 13888 |
+
"learning_rate": 1.0759869397447314e-05,
|
| 13889 |
+
"loss": 2.9039,
|
| 13890 |
+
"step": 7932
|
| 13891 |
+
},
|
| 13892 |
+
{
|
| 13893 |
+
"epoch": 0.79,
|
| 13894 |
+
"grad_norm": 1.7106647491455078,
|
| 13895 |
+
"learning_rate": 1.074008113188879e-05,
|
| 13896 |
+
"loss": 2.5086,
|
| 13897 |
+
"step": 7936
|
| 13898 |
+
},
|
| 13899 |
+
{
|
| 13900 |
+
"epoch": 0.79,
|
| 13901 |
+
"grad_norm": 1.5160934925079346,
|
| 13902 |
+
"learning_rate": 1.0720292866330266e-05,
|
| 13903 |
+
"loss": 2.447,
|
| 13904 |
+
"step": 7940
|
| 13905 |
+
},
|
| 13906 |
+
{
|
| 13907 |
+
"epoch": 0.79,
|
| 13908 |
+
"grad_norm": 1.833022952079773,
|
| 13909 |
+
"learning_rate": 1.0700504600771743e-05,
|
| 13910 |
+
"loss": 2.4753,
|
| 13911 |
+
"step": 7944
|
| 13912 |
+
},
|
| 13913 |
+
{
|
| 13914 |
+
"epoch": 0.79,
|
| 13915 |
+
"grad_norm": 1.7385578155517578,
|
| 13916 |
+
"learning_rate": 1.0680716335213219e-05,
|
| 13917 |
+
"loss": 2.6183,
|
| 13918 |
+
"step": 7948
|
| 13919 |
+
},
|
| 13920 |
+
{
|
| 13921 |
+
"epoch": 0.79,
|
| 13922 |
+
"grad_norm": 1.8667320013046265,
|
| 13923 |
+
"learning_rate": 1.0660928069654695e-05,
|
| 13924 |
+
"loss": 2.5605,
|
| 13925 |
+
"step": 7952
|
| 13926 |
+
},
|
| 13927 |
+
{
|
| 13928 |
+
"epoch": 0.79,
|
| 13929 |
+
"grad_norm": 1.5835785865783691,
|
| 13930 |
+
"learning_rate": 1.0641139804096172e-05,
|
| 13931 |
+
"loss": 2.4019,
|
| 13932 |
+
"step": 7956
|
| 13933 |
+
},
|
| 13934 |
+
{
|
| 13935 |
+
"epoch": 0.79,
|
| 13936 |
+
"grad_norm": 1.7108229398727417,
|
| 13937 |
+
"learning_rate": 1.0621351538537648e-05,
|
| 13938 |
+
"loss": 2.708,
|
| 13939 |
+
"step": 7960
|
| 13940 |
+
},
|
| 13941 |
+
{
|
| 13942 |
+
"epoch": 0.79,
|
| 13943 |
+
"grad_norm": 1.674804925918579,
|
| 13944 |
+
"learning_rate": 1.0601563272979124e-05,
|
| 13945 |
+
"loss": 2.5102,
|
| 13946 |
+
"step": 7964
|
| 13947 |
+
},
|
| 13948 |
+
{
|
| 13949 |
+
"epoch": 0.79,
|
| 13950 |
+
"grad_norm": 1.955705165863037,
|
| 13951 |
+
"learning_rate": 1.0581775007420599e-05,
|
| 13952 |
+
"loss": 2.5221,
|
| 13953 |
+
"step": 7968
|
| 13954 |
+
},
|
| 13955 |
+
{
|
| 13956 |
+
"epoch": 0.79,
|
| 13957 |
+
"grad_norm": 1.740157961845398,
|
| 13958 |
+
"learning_rate": 1.0561986741862075e-05,
|
| 13959 |
+
"loss": 2.5232,
|
| 13960 |
+
"step": 7972
|
| 13961 |
+
},
|
| 13962 |
+
{
|
| 13963 |
+
"epoch": 0.79,
|
| 13964 |
+
"grad_norm": 1.6809014081954956,
|
| 13965 |
+
"learning_rate": 1.0542198476303553e-05,
|
| 13966 |
+
"loss": 2.3372,
|
| 13967 |
+
"step": 7976
|
| 13968 |
+
},
|
| 13969 |
+
{
|
| 13970 |
+
"epoch": 0.79,
|
| 13971 |
+
"grad_norm": 1.7483688592910767,
|
| 13972 |
+
"learning_rate": 1.052241021074503e-05,
|
| 13973 |
+
"loss": 2.4913,
|
| 13974 |
+
"step": 7980
|
| 13975 |
+
},
|
| 13976 |
+
{
|
| 13977 |
+
"epoch": 0.79,
|
| 13978 |
+
"grad_norm": 1.5279000997543335,
|
| 13979 |
+
"learning_rate": 1.0502621945186506e-05,
|
| 13980 |
+
"loss": 2.6386,
|
| 13981 |
+
"step": 7984
|
| 13982 |
+
},
|
| 13983 |
+
{
|
| 13984 |
+
"epoch": 0.79,
|
| 13985 |
+
"grad_norm": 1.7373197078704834,
|
| 13986 |
+
"learning_rate": 1.048283367962798e-05,
|
| 13987 |
+
"loss": 2.5823,
|
| 13988 |
+
"step": 7988
|
| 13989 |
+
},
|
| 13990 |
+
{
|
| 13991 |
+
"epoch": 0.79,
|
| 13992 |
+
"grad_norm": 1.6561895608901978,
|
| 13993 |
+
"learning_rate": 1.0463045414069457e-05,
|
| 13994 |
+
"loss": 2.4531,
|
| 13995 |
+
"step": 7992
|
| 13996 |
+
},
|
| 13997 |
+
{
|
| 13998 |
+
"epoch": 0.79,
|
| 13999 |
+
"grad_norm": 1.6324809789657593,
|
| 14000 |
+
"learning_rate": 1.0443257148510933e-05,
|
| 14001 |
+
"loss": 2.4943,
|
| 14002 |
+
"step": 7996
|
| 14003 |
+
},
|
| 14004 |
+
{
|
| 14005 |
+
"epoch": 0.79,
|
| 14006 |
+
"grad_norm": 1.4932702779769897,
|
| 14007 |
+
"learning_rate": 1.042346888295241e-05,
|
| 14008 |
+
"loss": 2.5745,
|
| 14009 |
+
"step": 8000
|
| 14010 |
+
},
|
| 14011 |
+
{
|
| 14012 |
+
"epoch": 0.79,
|
| 14013 |
+
"grad_norm": 1.6975860595703125,
|
| 14014 |
+
"learning_rate": 1.0403680617393886e-05,
|
| 14015 |
+
"loss": 2.5195,
|
| 14016 |
+
"step": 8004
|
| 14017 |
+
},
|
| 14018 |
+
{
|
| 14019 |
+
"epoch": 0.79,
|
| 14020 |
+
"grad_norm": 1.5818710327148438,
|
| 14021 |
+
"learning_rate": 1.0383892351835362e-05,
|
| 14022 |
+
"loss": 2.5146,
|
| 14023 |
+
"step": 8008
|
| 14024 |
+
},
|
| 14025 |
+
{
|
| 14026 |
+
"epoch": 0.79,
|
| 14027 |
+
"grad_norm": 1.6648439168930054,
|
| 14028 |
+
"learning_rate": 1.0364104086276839e-05,
|
| 14029 |
+
"loss": 2.5038,
|
| 14030 |
+
"step": 8012
|
| 14031 |
+
},
|
| 14032 |
+
{
|
| 14033 |
+
"epoch": 0.79,
|
| 14034 |
+
"grad_norm": 1.8664870262145996,
|
| 14035 |
+
"learning_rate": 1.0344315820718315e-05,
|
| 14036 |
+
"loss": 2.4395,
|
| 14037 |
+
"step": 8016
|
| 14038 |
+
},
|
| 14039 |
+
{
|
| 14040 |
+
"epoch": 0.79,
|
| 14041 |
+
"grad_norm": 1.6299059391021729,
|
| 14042 |
+
"learning_rate": 1.0324527555159791e-05,
|
| 14043 |
+
"loss": 2.632,
|
| 14044 |
+
"step": 8020
|
| 14045 |
+
},
|
| 14046 |
+
{
|
| 14047 |
+
"epoch": 0.79,
|
| 14048 |
+
"grad_norm": 1.6017898321151733,
|
| 14049 |
+
"learning_rate": 1.0304739289601268e-05,
|
| 14050 |
+
"loss": 2.523,
|
| 14051 |
+
"step": 8024
|
| 14052 |
+
},
|
| 14053 |
+
{
|
| 14054 |
+
"epoch": 0.79,
|
| 14055 |
+
"grad_norm": 1.8422410488128662,
|
| 14056 |
+
"learning_rate": 1.0284951024042744e-05,
|
| 14057 |
+
"loss": 2.5135,
|
| 14058 |
+
"step": 8028
|
| 14059 |
+
},
|
| 14060 |
+
{
|
| 14061 |
+
"epoch": 0.79,
|
| 14062 |
+
"grad_norm": 1.7916709184646606,
|
| 14063 |
+
"learning_rate": 1.0265162758484218e-05,
|
| 14064 |
+
"loss": 2.3953,
|
| 14065 |
+
"step": 8032
|
| 14066 |
+
},
|
| 14067 |
+
{
|
| 14068 |
+
"epoch": 0.8,
|
| 14069 |
+
"grad_norm": 1.6924244165420532,
|
| 14070 |
+
"learning_rate": 1.0245374492925695e-05,
|
| 14071 |
+
"loss": 2.5083,
|
| 14072 |
+
"step": 8036
|
| 14073 |
+
},
|
| 14074 |
+
{
|
| 14075 |
+
"epoch": 0.8,
|
| 14076 |
+
"grad_norm": 1.6407674551010132,
|
| 14077 |
+
"learning_rate": 1.0225586227367171e-05,
|
| 14078 |
+
"loss": 2.5489,
|
| 14079 |
+
"step": 8040
|
| 14080 |
+
},
|
| 14081 |
+
{
|
| 14082 |
+
"epoch": 0.8,
|
| 14083 |
+
"grad_norm": 2.751863718032837,
|
| 14084 |
+
"learning_rate": 1.0205797961808649e-05,
|
| 14085 |
+
"loss": 2.5538,
|
| 14086 |
+
"step": 8044
|
| 14087 |
+
},
|
| 14088 |
+
{
|
| 14089 |
+
"epoch": 0.8,
|
| 14090 |
+
"grad_norm": 1.8018790483474731,
|
| 14091 |
+
"learning_rate": 1.0186009696250125e-05,
|
| 14092 |
+
"loss": 2.4931,
|
| 14093 |
+
"step": 8048
|
| 14094 |
+
},
|
| 14095 |
+
{
|
| 14096 |
+
"epoch": 0.8,
|
| 14097 |
+
"grad_norm": 1.6749464273452759,
|
| 14098 |
+
"learning_rate": 1.01662214306916e-05,
|
| 14099 |
+
"loss": 2.5116,
|
| 14100 |
+
"step": 8052
|
| 14101 |
+
},
|
| 14102 |
+
{
|
| 14103 |
+
"epoch": 0.8,
|
| 14104 |
+
"grad_norm": 1.8620166778564453,
|
| 14105 |
+
"learning_rate": 1.0146433165133076e-05,
|
| 14106 |
+
"loss": 2.4857,
|
| 14107 |
+
"step": 8056
|
| 14108 |
+
},
|
| 14109 |
+
{
|
| 14110 |
+
"epoch": 0.8,
|
| 14111 |
+
"grad_norm": 1.8505146503448486,
|
| 14112 |
+
"learning_rate": 1.0126644899574553e-05,
|
| 14113 |
+
"loss": 2.535,
|
| 14114 |
+
"step": 8060
|
| 14115 |
+
},
|
| 14116 |
+
{
|
| 14117 |
+
"epoch": 0.8,
|
| 14118 |
+
"grad_norm": 1.4946340322494507,
|
| 14119 |
+
"learning_rate": 1.0106856634016029e-05,
|
| 14120 |
+
"loss": 2.3432,
|
| 14121 |
+
"step": 8064
|
| 14122 |
+
},
|
| 14123 |
+
{
|
| 14124 |
+
"epoch": 0.8,
|
| 14125 |
+
"grad_norm": 1.8005412817001343,
|
| 14126 |
+
"learning_rate": 1.0087068368457505e-05,
|
| 14127 |
+
"loss": 2.4651,
|
| 14128 |
+
"step": 8068
|
| 14129 |
+
},
|
| 14130 |
+
{
|
| 14131 |
+
"epoch": 0.8,
|
| 14132 |
+
"grad_norm": 1.786773443222046,
|
| 14133 |
+
"learning_rate": 1.006728010289898e-05,
|
| 14134 |
+
"loss": 2.4242,
|
| 14135 |
+
"step": 8072
|
| 14136 |
+
},
|
| 14137 |
+
{
|
| 14138 |
+
"epoch": 0.8,
|
| 14139 |
+
"grad_norm": 1.7355870008468628,
|
| 14140 |
+
"learning_rate": 1.0047491837340458e-05,
|
| 14141 |
+
"loss": 2.4785,
|
| 14142 |
+
"step": 8076
|
| 14143 |
+
},
|
| 14144 |
+
{
|
| 14145 |
+
"epoch": 0.8,
|
| 14146 |
+
"grad_norm": 1.677388310432434,
|
| 14147 |
+
"learning_rate": 1.0027703571781934e-05,
|
| 14148 |
+
"loss": 2.5572,
|
| 14149 |
+
"step": 8080
|
| 14150 |
+
},
|
| 14151 |
+
{
|
| 14152 |
+
"epoch": 0.8,
|
| 14153 |
+
"grad_norm": 1.7865822315216064,
|
| 14154 |
+
"learning_rate": 1.000791530622341e-05,
|
| 14155 |
+
"loss": 2.771,
|
| 14156 |
+
"step": 8084
|
| 14157 |
+
},
|
| 14158 |
+
{
|
| 14159 |
+
"epoch": 0.8,
|
| 14160 |
+
"grad_norm": 1.8583182096481323,
|
| 14161 |
+
"learning_rate": 9.988127040664887e-06,
|
| 14162 |
+
"loss": 2.4894,
|
| 14163 |
+
"step": 8088
|
| 14164 |
}
|
| 14165 |
],
|
| 14166 |
"logging_steps": 4,
|
|
|
|
| 14168 |
"num_input_tokens_seen": 0,
|
| 14169 |
"num_train_epochs": 1,
|
| 14170 |
"save_steps": 1011,
|
| 14171 |
+
"total_flos": 9.891653776677274e+16,
|
| 14172 |
"train_batch_size": 4,
|
| 14173 |
"trial_name": null,
|
| 14174 |
"trial_params": null
|