Commit
•
4ff28de
1
Parent(s):
f3867c1
Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1433 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1433 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.2,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.001,
|
| 13 |
+
"grad_norm": 1.7063226699829102,
|
| 14 |
+
"learning_rate": 4.995e-05,
|
| 15 |
+
"loss": 9.6305,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.002,
|
| 20 |
+
"grad_norm": 1.467505693435669,
|
| 21 |
+
"learning_rate": 4.99e-05,
|
| 22 |
+
"loss": 8.8474,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.003,
|
| 27 |
+
"grad_norm": 1.3338744640350342,
|
| 28 |
+
"learning_rate": 4.9850000000000006e-05,
|
| 29 |
+
"loss": 8.4272,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.004,
|
| 34 |
+
"grad_norm": 1.194218635559082,
|
| 35 |
+
"learning_rate": 4.9800000000000004e-05,
|
| 36 |
+
"loss": 7.9969,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.005,
|
| 41 |
+
"grad_norm": 0.9542586207389832,
|
| 42 |
+
"learning_rate": 4.975e-05,
|
| 43 |
+
"loss": 7.8018,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.006,
|
| 48 |
+
"grad_norm": 0.8312947154045105,
|
| 49 |
+
"learning_rate": 4.97e-05,
|
| 50 |
+
"loss": 7.5303,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.007,
|
| 55 |
+
"grad_norm": 0.6978892683982849,
|
| 56 |
+
"learning_rate": 4.965e-05,
|
| 57 |
+
"loss": 7.3733,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.008,
|
| 62 |
+
"grad_norm": 0.6895764470100403,
|
| 63 |
+
"learning_rate": 4.96e-05,
|
| 64 |
+
"loss": 7.2434,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.009,
|
| 69 |
+
"grad_norm": 0.5555976033210754,
|
| 70 |
+
"learning_rate": 4.9550000000000005e-05,
|
| 71 |
+
"loss": 7.0877,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.01,
|
| 76 |
+
"grad_norm": 0.836391806602478,
|
| 77 |
+
"learning_rate": 4.9500000000000004e-05,
|
| 78 |
+
"loss": 7.0338,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.011,
|
| 83 |
+
"grad_norm": 0.782464861869812,
|
| 84 |
+
"learning_rate": 4.945e-05,
|
| 85 |
+
"loss": 6.878,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.012,
|
| 90 |
+
"grad_norm": 1.3705933094024658,
|
| 91 |
+
"learning_rate": 4.94e-05,
|
| 92 |
+
"loss": 6.5874,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.013,
|
| 97 |
+
"grad_norm": 0.7560775876045227,
|
| 98 |
+
"learning_rate": 4.935e-05,
|
| 99 |
+
"loss": 6.4978,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.014,
|
| 104 |
+
"grad_norm": 1.3238508701324463,
|
| 105 |
+
"learning_rate": 4.93e-05,
|
| 106 |
+
"loss": 6.3998,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.015,
|
| 111 |
+
"grad_norm": 0.7834548950195312,
|
| 112 |
+
"learning_rate": 4.9250000000000004e-05,
|
| 113 |
+
"loss": 6.2838,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.016,
|
| 118 |
+
"grad_norm": 0.762347400188446,
|
| 119 |
+
"learning_rate": 4.92e-05,
|
| 120 |
+
"loss": 6.0387,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.017,
|
| 125 |
+
"grad_norm": 0.7799501419067383,
|
| 126 |
+
"learning_rate": 4.915e-05,
|
| 127 |
+
"loss": 6.0241,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.018,
|
| 132 |
+
"grad_norm": 0.7948866486549377,
|
| 133 |
+
"learning_rate": 4.91e-05,
|
| 134 |
+
"loss": 5.8776,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.019,
|
| 139 |
+
"grad_norm": 0.9890483021736145,
|
| 140 |
+
"learning_rate": 4.905e-05,
|
| 141 |
+
"loss": 5.747,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.02,
|
| 146 |
+
"grad_norm": 0.9131263494491577,
|
| 147 |
+
"learning_rate": 4.9e-05,
|
| 148 |
+
"loss": 5.644,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.021,
|
| 153 |
+
"grad_norm": 1.7073436975479126,
|
| 154 |
+
"learning_rate": 4.8950000000000004e-05,
|
| 155 |
+
"loss": 5.778,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.022,
|
| 160 |
+
"grad_norm": 0.8059922456741333,
|
| 161 |
+
"learning_rate": 4.89e-05,
|
| 162 |
+
"loss": 5.4755,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.023,
|
| 167 |
+
"grad_norm": 1.2500686645507812,
|
| 168 |
+
"learning_rate": 4.885e-05,
|
| 169 |
+
"loss": 5.3769,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.024,
|
| 174 |
+
"grad_norm": 1.3848680257797241,
|
| 175 |
+
"learning_rate": 4.88e-05,
|
| 176 |
+
"loss": 5.2105,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.025,
|
| 181 |
+
"grad_norm": 1.2381746768951416,
|
| 182 |
+
"learning_rate": 4.875e-05,
|
| 183 |
+
"loss": 5.1444,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.026,
|
| 188 |
+
"grad_norm": 2.7005224227905273,
|
| 189 |
+
"learning_rate": 4.87e-05,
|
| 190 |
+
"loss": 5.1608,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.027,
|
| 195 |
+
"grad_norm": 1.1472671031951904,
|
| 196 |
+
"learning_rate": 4.8650000000000003e-05,
|
| 197 |
+
"loss": 4.9456,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.028,
|
| 202 |
+
"grad_norm": 1.9849270582199097,
|
| 203 |
+
"learning_rate": 4.86e-05,
|
| 204 |
+
"loss": 4.8466,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.029,
|
| 209 |
+
"grad_norm": 1.857001781463623,
|
| 210 |
+
"learning_rate": 4.855e-05,
|
| 211 |
+
"loss": 4.7323,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.03,
|
| 216 |
+
"grad_norm": 1.6731220483779907,
|
| 217 |
+
"learning_rate": 4.85e-05,
|
| 218 |
+
"loss": 4.5786,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.031,
|
| 223 |
+
"grad_norm": 1.7968906164169312,
|
| 224 |
+
"learning_rate": 4.845e-05,
|
| 225 |
+
"loss": 4.4588,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.032,
|
| 230 |
+
"grad_norm": 1.7908226251602173,
|
| 231 |
+
"learning_rate": 4.8400000000000004e-05,
|
| 232 |
+
"loss": 4.3645,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.033,
|
| 237 |
+
"grad_norm": 2.538881540298462,
|
| 238 |
+
"learning_rate": 4.835e-05,
|
| 239 |
+
"loss": 4.1489,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.034,
|
| 244 |
+
"grad_norm": 2.306257486343384,
|
| 245 |
+
"learning_rate": 4.83e-05,
|
| 246 |
+
"loss": 3.9798,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.035,
|
| 251 |
+
"grad_norm": 2.1730940341949463,
|
| 252 |
+
"learning_rate": 4.825e-05,
|
| 253 |
+
"loss": 4.0231,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.036,
|
| 258 |
+
"grad_norm": 2.4211463928222656,
|
| 259 |
+
"learning_rate": 4.82e-05,
|
| 260 |
+
"loss": 3.8495,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.037,
|
| 265 |
+
"grad_norm": 2.3698794841766357,
|
| 266 |
+
"learning_rate": 4.815e-05,
|
| 267 |
+
"loss": 3.6977,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.038,
|
| 272 |
+
"grad_norm": 2.147799491882324,
|
| 273 |
+
"learning_rate": 4.8100000000000004e-05,
|
| 274 |
+
"loss": 3.8008,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.039,
|
| 279 |
+
"grad_norm": 2.3577606678009033,
|
| 280 |
+
"learning_rate": 4.805e-05,
|
| 281 |
+
"loss": 3.6983,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.04,
|
| 286 |
+
"grad_norm": 2.065912961959839,
|
| 287 |
+
"learning_rate": 4.8e-05,
|
| 288 |
+
"loss": 3.5738,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.041,
|
| 293 |
+
"grad_norm": 2.930288314819336,
|
| 294 |
+
"learning_rate": 4.795e-05,
|
| 295 |
+
"loss": 3.5117,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.042,
|
| 300 |
+
"grad_norm": 2.3703155517578125,
|
| 301 |
+
"learning_rate": 4.79e-05,
|
| 302 |
+
"loss": 3.2483,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.043,
|
| 307 |
+
"grad_norm": 2.6050736904144287,
|
| 308 |
+
"learning_rate": 4.785e-05,
|
| 309 |
+
"loss": 3.2342,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.044,
|
| 314 |
+
"grad_norm": 2.0790674686431885,
|
| 315 |
+
"learning_rate": 4.78e-05,
|
| 316 |
+
"loss": 3.1452,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.045,
|
| 321 |
+
"grad_norm": 2.2497427463531494,
|
| 322 |
+
"learning_rate": 4.775e-05,
|
| 323 |
+
"loss": 3.0316,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.046,
|
| 328 |
+
"grad_norm": 2.507902145385742,
|
| 329 |
+
"learning_rate": 4.77e-05,
|
| 330 |
+
"loss": 2.8938,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.047,
|
| 335 |
+
"grad_norm": 2.517744541168213,
|
| 336 |
+
"learning_rate": 4.765e-05,
|
| 337 |
+
"loss": 2.8137,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.048,
|
| 342 |
+
"grad_norm": 3.9981460571289062,
|
| 343 |
+
"learning_rate": 4.76e-05,
|
| 344 |
+
"loss": 2.9864,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.049,
|
| 349 |
+
"grad_norm": 2.265026569366455,
|
| 350 |
+
"learning_rate": 4.755e-05,
|
| 351 |
+
"loss": 2.7839,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.05,
|
| 356 |
+
"grad_norm": 2.257293701171875,
|
| 357 |
+
"learning_rate": 4.75e-05,
|
| 358 |
+
"loss": 2.6834,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.051,
|
| 363 |
+
"grad_norm": 2.6932270526885986,
|
| 364 |
+
"learning_rate": 4.745e-05,
|
| 365 |
+
"loss": 2.5755,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.052,
|
| 370 |
+
"grad_norm": 1.7177081108093262,
|
| 371 |
+
"learning_rate": 4.74e-05,
|
| 372 |
+
"loss": 2.425,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.053,
|
| 377 |
+
"grad_norm": 2.2452073097229004,
|
| 378 |
+
"learning_rate": 4.735e-05,
|
| 379 |
+
"loss": 2.5261,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.054,
|
| 384 |
+
"grad_norm": 2.2109947204589844,
|
| 385 |
+
"learning_rate": 4.73e-05,
|
| 386 |
+
"loss": 2.3825,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.055,
|
| 391 |
+
"grad_norm": 2.574531078338623,
|
| 392 |
+
"learning_rate": 4.7249999999999997e-05,
|
| 393 |
+
"loss": 2.3087,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.056,
|
| 398 |
+
"grad_norm": 2.3631017208099365,
|
| 399 |
+
"learning_rate": 4.72e-05,
|
| 400 |
+
"loss": 2.3099,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.057,
|
| 405 |
+
"grad_norm": 2.3809709548950195,
|
| 406 |
+
"learning_rate": 4.715e-05,
|
| 407 |
+
"loss": 2.3001,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.058,
|
| 412 |
+
"grad_norm": 2.0683534145355225,
|
| 413 |
+
"learning_rate": 4.71e-05,
|
| 414 |
+
"loss": 2.0813,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.059,
|
| 419 |
+
"grad_norm": 2.5471837520599365,
|
| 420 |
+
"learning_rate": 4.705e-05,
|
| 421 |
+
"loss": 2.0378,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.06,
|
| 426 |
+
"grad_norm": 2.585564374923706,
|
| 427 |
+
"learning_rate": 4.7e-05,
|
| 428 |
+
"loss": 2.2062,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.061,
|
| 433 |
+
"grad_norm": 2.062100648880005,
|
| 434 |
+
"learning_rate": 4.695e-05,
|
| 435 |
+
"loss": 1.9914,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.062,
|
| 440 |
+
"grad_norm": 2.1019210815429688,
|
| 441 |
+
"learning_rate": 4.69e-05,
|
| 442 |
+
"loss": 1.9635,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.063,
|
| 447 |
+
"grad_norm": 2.630436658859253,
|
| 448 |
+
"learning_rate": 4.685000000000001e-05,
|
| 449 |
+
"loss": 1.9123,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.064,
|
| 454 |
+
"grad_norm": 2.1028494834899902,
|
| 455 |
+
"learning_rate": 4.6800000000000006e-05,
|
| 456 |
+
"loss": 1.7583,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.065,
|
| 461 |
+
"grad_norm": 2.392193078994751,
|
| 462 |
+
"learning_rate": 4.6750000000000005e-05,
|
| 463 |
+
"loss": 1.7532,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.066,
|
| 468 |
+
"grad_norm": 2.004413366317749,
|
| 469 |
+
"learning_rate": 4.6700000000000003e-05,
|
| 470 |
+
"loss": 1.6978,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.067,
|
| 475 |
+
"grad_norm": 2.210513114929199,
|
| 476 |
+
"learning_rate": 4.665e-05,
|
| 477 |
+
"loss": 1.6311,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.068,
|
| 482 |
+
"grad_norm": 1.8464936017990112,
|
| 483 |
+
"learning_rate": 4.660000000000001e-05,
|
| 484 |
+
"loss": 1.5507,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.069,
|
| 489 |
+
"grad_norm": 2.0246541500091553,
|
| 490 |
+
"learning_rate": 4.655000000000001e-05,
|
| 491 |
+
"loss": 1.5637,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.07,
|
| 496 |
+
"grad_norm": 2.199751138687134,
|
| 497 |
+
"learning_rate": 4.6500000000000005e-05,
|
| 498 |
+
"loss": 1.5603,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.071,
|
| 503 |
+
"grad_norm": 2.2002196311950684,
|
| 504 |
+
"learning_rate": 4.6450000000000004e-05,
|
| 505 |
+
"loss": 1.4558,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.072,
|
| 510 |
+
"grad_norm": 1.7826759815216064,
|
| 511 |
+
"learning_rate": 4.64e-05,
|
| 512 |
+
"loss": 1.4309,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.073,
|
| 517 |
+
"grad_norm": 1.760297417640686,
|
| 518 |
+
"learning_rate": 4.635e-05,
|
| 519 |
+
"loss": 1.3531,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.074,
|
| 524 |
+
"grad_norm": 2.0505475997924805,
|
| 525 |
+
"learning_rate": 4.630000000000001e-05,
|
| 526 |
+
"loss": 1.3641,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.075,
|
| 531 |
+
"grad_norm": 2.1375396251678467,
|
| 532 |
+
"learning_rate": 4.6250000000000006e-05,
|
| 533 |
+
"loss": 1.3259,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.076,
|
| 538 |
+
"grad_norm": 1.8252328634262085,
|
| 539 |
+
"learning_rate": 4.6200000000000005e-05,
|
| 540 |
+
"loss": 1.2026,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.077,
|
| 545 |
+
"grad_norm": 1.8945906162261963,
|
| 546 |
+
"learning_rate": 4.6150000000000004e-05,
|
| 547 |
+
"loss": 1.2878,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.078,
|
| 552 |
+
"grad_norm": 1.7990881204605103,
|
| 553 |
+
"learning_rate": 4.61e-05,
|
| 554 |
+
"loss": 1.1853,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.079,
|
| 559 |
+
"grad_norm": 1.4897470474243164,
|
| 560 |
+
"learning_rate": 4.605e-05,
|
| 561 |
+
"loss": 1.1279,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.08,
|
| 566 |
+
"grad_norm": 2.2804617881774902,
|
| 567 |
+
"learning_rate": 4.600000000000001e-05,
|
| 568 |
+
"loss": 1.0804,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.081,
|
| 573 |
+
"grad_norm": 1.4800664186477661,
|
| 574 |
+
"learning_rate": 4.5950000000000006e-05,
|
| 575 |
+
"loss": 1.0361,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.082,
|
| 580 |
+
"grad_norm": 1.3526049852371216,
|
| 581 |
+
"learning_rate": 4.5900000000000004e-05,
|
| 582 |
+
"loss": 1.0585,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.083,
|
| 587 |
+
"grad_norm": 1.534173607826233,
|
| 588 |
+
"learning_rate": 4.585e-05,
|
| 589 |
+
"loss": 1.0206,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.084,
|
| 594 |
+
"grad_norm": 1.4844435453414917,
|
| 595 |
+
"learning_rate": 4.58e-05,
|
| 596 |
+
"loss": 0.9758,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.085,
|
| 601 |
+
"grad_norm": 1.533679485321045,
|
| 602 |
+
"learning_rate": 4.575e-05,
|
| 603 |
+
"loss": 0.9168,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.086,
|
| 608 |
+
"grad_norm": 1.456162691116333,
|
| 609 |
+
"learning_rate": 4.5700000000000006e-05,
|
| 610 |
+
"loss": 0.8913,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.087,
|
| 615 |
+
"grad_norm": 1.7335631847381592,
|
| 616 |
+
"learning_rate": 4.5650000000000005e-05,
|
| 617 |
+
"loss": 0.9154,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.088,
|
| 622 |
+
"grad_norm": 1.3331761360168457,
|
| 623 |
+
"learning_rate": 4.5600000000000004e-05,
|
| 624 |
+
"loss": 0.8483,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.089,
|
| 629 |
+
"grad_norm": 1.6703053712844849,
|
| 630 |
+
"learning_rate": 4.555e-05,
|
| 631 |
+
"loss": 0.8116,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.09,
|
| 636 |
+
"grad_norm": 1.275975227355957,
|
| 637 |
+
"learning_rate": 4.55e-05,
|
| 638 |
+
"loss": 0.7869,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.091,
|
| 643 |
+
"grad_norm": 1.3800309896469116,
|
| 644 |
+
"learning_rate": 4.545000000000001e-05,
|
| 645 |
+
"loss": 0.7637,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.092,
|
| 650 |
+
"grad_norm": 1.9472386837005615,
|
| 651 |
+
"learning_rate": 4.5400000000000006e-05,
|
| 652 |
+
"loss": 0.7212,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.093,
|
| 657 |
+
"grad_norm": 1.3451333045959473,
|
| 658 |
+
"learning_rate": 4.5350000000000005e-05,
|
| 659 |
+
"loss": 0.6829,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.094,
|
| 664 |
+
"grad_norm": 1.5209784507751465,
|
| 665 |
+
"learning_rate": 4.53e-05,
|
| 666 |
+
"loss": 0.729,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.095,
|
| 671 |
+
"grad_norm": 1.3944469690322876,
|
| 672 |
+
"learning_rate": 4.525e-05,
|
| 673 |
+
"loss": 0.6732,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.096,
|
| 678 |
+
"grad_norm": 1.2177132368087769,
|
| 679 |
+
"learning_rate": 4.52e-05,
|
| 680 |
+
"loss": 0.6188,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.097,
|
| 685 |
+
"grad_norm": 1.5988528728485107,
|
| 686 |
+
"learning_rate": 4.5150000000000006e-05,
|
| 687 |
+
"loss": 0.6622,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.098,
|
| 692 |
+
"grad_norm": 1.3636531829833984,
|
| 693 |
+
"learning_rate": 4.5100000000000005e-05,
|
| 694 |
+
"loss": 0.5792,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.099,
|
| 699 |
+
"grad_norm": 1.377453088760376,
|
| 700 |
+
"learning_rate": 4.5050000000000004e-05,
|
| 701 |
+
"loss": 0.6062,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.1,
|
| 706 |
+
"grad_norm": 2.295713186264038,
|
| 707 |
+
"learning_rate": 4.5e-05,
|
| 708 |
+
"loss": 0.5709,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.101,
|
| 713 |
+
"grad_norm": 1.35196852684021,
|
| 714 |
+
"learning_rate": 4.495e-05,
|
| 715 |
+
"loss": 0.5521,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.102,
|
| 720 |
+
"grad_norm": 1.0617187023162842,
|
| 721 |
+
"learning_rate": 4.49e-05,
|
| 722 |
+
"loss": 0.5147,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.103,
|
| 727 |
+
"grad_norm": 1.3035167455673218,
|
| 728 |
+
"learning_rate": 4.4850000000000006e-05,
|
| 729 |
+
"loss": 0.5081,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.104,
|
| 734 |
+
"grad_norm": 1.2835568189620972,
|
| 735 |
+
"learning_rate": 4.4800000000000005e-05,
|
| 736 |
+
"loss": 0.5,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.105,
|
| 741 |
+
"grad_norm": 1.0403038263320923,
|
| 742 |
+
"learning_rate": 4.4750000000000004e-05,
|
| 743 |
+
"loss": 0.4825,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.106,
|
| 748 |
+
"grad_norm": 0.9538235068321228,
|
| 749 |
+
"learning_rate": 4.47e-05,
|
| 750 |
+
"loss": 0.4316,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.107,
|
| 755 |
+
"grad_norm": 1.4246289730072021,
|
| 756 |
+
"learning_rate": 4.465e-05,
|
| 757 |
+
"loss": 0.4304,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.108,
|
| 762 |
+
"grad_norm": 1.1217833757400513,
|
| 763 |
+
"learning_rate": 4.46e-05,
|
| 764 |
+
"loss": 0.4397,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.109,
|
| 769 |
+
"grad_norm": 1.0411335229873657,
|
| 770 |
+
"learning_rate": 4.4550000000000005e-05,
|
| 771 |
+
"loss": 0.4057,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.11,
|
| 776 |
+
"grad_norm": 0.8498069643974304,
|
| 777 |
+
"learning_rate": 4.4500000000000004e-05,
|
| 778 |
+
"loss": 0.3933,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.111,
|
| 783 |
+
"grad_norm": 1.1270406246185303,
|
| 784 |
+
"learning_rate": 4.445e-05,
|
| 785 |
+
"loss": 0.366,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.112,
|
| 790 |
+
"grad_norm": 1.189041256904602,
|
| 791 |
+
"learning_rate": 4.44e-05,
|
| 792 |
+
"loss": 0.3407,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.113,
|
| 797 |
+
"grad_norm": 0.9837467670440674,
|
| 798 |
+
"learning_rate": 4.435e-05,
|
| 799 |
+
"loss": 0.3511,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.114,
|
| 804 |
+
"grad_norm": 1.0432955026626587,
|
| 805 |
+
"learning_rate": 4.43e-05,
|
| 806 |
+
"loss": 0.3381,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.115,
|
| 811 |
+
"grad_norm": 0.9529951810836792,
|
| 812 |
+
"learning_rate": 4.4250000000000005e-05,
|
| 813 |
+
"loss": 0.3189,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.116,
|
| 818 |
+
"grad_norm": 1.008836030960083,
|
| 819 |
+
"learning_rate": 4.4200000000000004e-05,
|
| 820 |
+
"loss": 0.3077,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.117,
|
| 825 |
+
"grad_norm": 1.0005086660385132,
|
| 826 |
+
"learning_rate": 4.415e-05,
|
| 827 |
+
"loss": 0.3001,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.118,
|
| 832 |
+
"grad_norm": 1.1065175533294678,
|
| 833 |
+
"learning_rate": 4.41e-05,
|
| 834 |
+
"loss": 0.28,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.119,
|
| 839 |
+
"grad_norm": 0.6701949834823608,
|
| 840 |
+
"learning_rate": 4.405e-05,
|
| 841 |
+
"loss": 0.2692,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.12,
|
| 846 |
+
"grad_norm": 0.7154658436775208,
|
| 847 |
+
"learning_rate": 4.4000000000000006e-05,
|
| 848 |
+
"loss": 0.2663,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.121,
|
| 853 |
+
"grad_norm": 0.6997113823890686,
|
| 854 |
+
"learning_rate": 4.3950000000000004e-05,
|
| 855 |
+
"loss": 0.2595,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.122,
|
| 860 |
+
"grad_norm": 0.9047608971595764,
|
| 861 |
+
"learning_rate": 4.39e-05,
|
| 862 |
+
"loss": 0.2558,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.123,
|
| 867 |
+
"grad_norm": 0.8508415222167969,
|
| 868 |
+
"learning_rate": 4.385e-05,
|
| 869 |
+
"loss": 0.2459,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.124,
|
| 874 |
+
"grad_norm": 0.6505220532417297,
|
| 875 |
+
"learning_rate": 4.38e-05,
|
| 876 |
+
"loss": 0.2236,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.125,
|
| 881 |
+
"grad_norm": 0.5360460877418518,
|
| 882 |
+
"learning_rate": 4.375e-05,
|
| 883 |
+
"loss": 0.2189,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.126,
|
| 888 |
+
"grad_norm": 0.560817539691925,
|
| 889 |
+
"learning_rate": 4.3700000000000005e-05,
|
| 890 |
+
"loss": 0.2166,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.127,
|
| 895 |
+
"grad_norm": 0.7089666128158569,
|
| 896 |
+
"learning_rate": 4.3650000000000004e-05,
|
| 897 |
+
"loss": 0.2026,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.128,
|
| 902 |
+
"grad_norm": 0.5265817046165466,
|
| 903 |
+
"learning_rate": 4.36e-05,
|
| 904 |
+
"loss": 0.197,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.129,
|
| 909 |
+
"grad_norm": 0.6629377007484436,
|
| 910 |
+
"learning_rate": 4.355e-05,
|
| 911 |
+
"loss": 0.1934,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.13,
|
| 916 |
+
"grad_norm": 1.0730735063552856,
|
| 917 |
+
"learning_rate": 4.35e-05,
|
| 918 |
+
"loss": 0.1807,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.131,
|
| 923 |
+
"grad_norm": 0.6990699172019958,
|
| 924 |
+
"learning_rate": 4.345e-05,
|
| 925 |
+
"loss": 0.1845,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.132,
|
| 930 |
+
"grad_norm": 0.5047340393066406,
|
| 931 |
+
"learning_rate": 4.3400000000000005e-05,
|
| 932 |
+
"loss": 0.1725,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.133,
|
| 937 |
+
"grad_norm": 0.6830994486808777,
|
| 938 |
+
"learning_rate": 4.335e-05,
|
| 939 |
+
"loss": 0.1687,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.134,
|
| 944 |
+
"grad_norm": 0.5861710906028748,
|
| 945 |
+
"learning_rate": 4.33e-05,
|
| 946 |
+
"loss": 0.1671,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.135,
|
| 951 |
+
"grad_norm": 0.43594300746917725,
|
| 952 |
+
"learning_rate": 4.325e-05,
|
| 953 |
+
"loss": 0.1467,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.136,
|
| 958 |
+
"grad_norm": 0.44587692618370056,
|
| 959 |
+
"learning_rate": 4.32e-05,
|
| 960 |
+
"loss": 0.1509,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.137,
|
| 965 |
+
"grad_norm": 0.5523977875709534,
|
| 966 |
+
"learning_rate": 4.315e-05,
|
| 967 |
+
"loss": 0.1434,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.138,
|
| 972 |
+
"grad_norm": 0.6139170527458191,
|
| 973 |
+
"learning_rate": 4.3100000000000004e-05,
|
| 974 |
+
"loss": 0.1433,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.139,
|
| 979 |
+
"grad_norm": 0.6169497966766357,
|
| 980 |
+
"learning_rate": 4.305e-05,
|
| 981 |
+
"loss": 0.1365,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.14,
|
| 986 |
+
"grad_norm": 0.49120134115219116,
|
| 987 |
+
"learning_rate": 4.3e-05,
|
| 988 |
+
"loss": 0.1287,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.141,
|
| 993 |
+
"grad_norm": 0.451753169298172,
|
| 994 |
+
"learning_rate": 4.295e-05,
|
| 995 |
+
"loss": 0.1142,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.142,
|
| 1000 |
+
"grad_norm": 0.5429627895355225,
|
| 1001 |
+
"learning_rate": 4.29e-05,
|
| 1002 |
+
"loss": 0.134,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.143,
|
| 1007 |
+
"grad_norm": 0.7613041400909424,
|
| 1008 |
+
"learning_rate": 4.285e-05,
|
| 1009 |
+
"loss": 0.1391,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.144,
|
| 1014 |
+
"grad_norm": 0.4953358471393585,
|
| 1015 |
+
"learning_rate": 4.2800000000000004e-05,
|
| 1016 |
+
"loss": 0.1197,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.145,
|
| 1021 |
+
"grad_norm": 0.3657626509666443,
|
| 1022 |
+
"learning_rate": 4.275e-05,
|
| 1023 |
+
"loss": 0.1071,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.146,
|
| 1028 |
+
"grad_norm": 0.44240206480026245,
|
| 1029 |
+
"learning_rate": 4.27e-05,
|
| 1030 |
+
"loss": 0.1111,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.147,
|
| 1035 |
+
"grad_norm": 0.5007165670394897,
|
| 1036 |
+
"learning_rate": 4.265e-05,
|
| 1037 |
+
"loss": 0.1056,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.148,
|
| 1042 |
+
"grad_norm": 0.4580256938934326,
|
| 1043 |
+
"learning_rate": 4.26e-05,
|
| 1044 |
+
"loss": 0.1049,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.149,
|
| 1049 |
+
"grad_norm": 0.4970822036266327,
|
| 1050 |
+
"learning_rate": 4.2550000000000004e-05,
|
| 1051 |
+
"loss": 0.1032,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.15,
|
| 1056 |
+
"grad_norm": 0.4138182997703552,
|
| 1057 |
+
"learning_rate": 4.25e-05,
|
| 1058 |
+
"loss": 0.0961,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.151,
|
| 1063 |
+
"grad_norm": 0.4013712406158447,
|
| 1064 |
+
"learning_rate": 4.245e-05,
|
| 1065 |
+
"loss": 0.0949,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.152,
|
| 1070 |
+
"grad_norm": 0.3868940770626068,
|
| 1071 |
+
"learning_rate": 4.24e-05,
|
| 1072 |
+
"loss": 0.0837,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.153,
|
| 1077 |
+
"grad_norm": 0.3113015294075012,
|
| 1078 |
+
"learning_rate": 4.235e-05,
|
| 1079 |
+
"loss": 0.0909,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.154,
|
| 1084 |
+
"grad_norm": 0.3569623529911041,
|
| 1085 |
+
"learning_rate": 4.23e-05,
|
| 1086 |
+
"loss": 0.0908,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.155,
|
| 1091 |
+
"grad_norm": 0.3841746151447296,
|
| 1092 |
+
"learning_rate": 4.2250000000000004e-05,
|
| 1093 |
+
"loss": 0.0806,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.156,
|
| 1098 |
+
"grad_norm": 0.6565550565719604,
|
| 1099 |
+
"learning_rate": 4.22e-05,
|
| 1100 |
+
"loss": 0.075,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.157,
|
| 1105 |
+
"grad_norm": 0.4816874563694,
|
| 1106 |
+
"learning_rate": 4.215e-05,
|
| 1107 |
+
"loss": 0.0858,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.158,
|
| 1112 |
+
"grad_norm": 0.30408933758735657,
|
| 1113 |
+
"learning_rate": 4.21e-05,
|
| 1114 |
+
"loss": 0.0704,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.159,
|
| 1119 |
+
"grad_norm": 0.43388792872428894,
|
| 1120 |
+
"learning_rate": 4.205e-05,
|
| 1121 |
+
"loss": 0.0671,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.16,
|
| 1126 |
+
"grad_norm": 0.33304253220558167,
|
| 1127 |
+
"learning_rate": 4.2e-05,
|
| 1128 |
+
"loss": 0.07,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.161,
|
| 1133 |
+
"grad_norm": 0.4260387420654297,
|
| 1134 |
+
"learning_rate": 4.195e-05,
|
| 1135 |
+
"loss": 0.0691,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.162,
|
| 1140 |
+
"grad_norm": 0.37930798530578613,
|
| 1141 |
+
"learning_rate": 4.19e-05,
|
| 1142 |
+
"loss": 0.0715,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.163,
|
| 1147 |
+
"grad_norm": 0.3198983669281006,
|
| 1148 |
+
"learning_rate": 4.185e-05,
|
| 1149 |
+
"loss": 0.0651,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.164,
|
| 1154 |
+
"grad_norm": 0.3510359823703766,
|
| 1155 |
+
"learning_rate": 4.18e-05,
|
| 1156 |
+
"loss": 0.058,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.165,
|
| 1161 |
+
"grad_norm": 0.41047966480255127,
|
| 1162 |
+
"learning_rate": 4.175e-05,
|
| 1163 |
+
"loss": 0.065,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.166,
|
| 1168 |
+
"grad_norm": 0.3054174482822418,
|
| 1169 |
+
"learning_rate": 4.17e-05,
|
| 1170 |
+
"loss": 0.0564,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.167,
|
| 1175 |
+
"grad_norm": 0.29319772124290466,
|
| 1176 |
+
"learning_rate": 4.165e-05,
|
| 1177 |
+
"loss": 0.0599,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.168,
|
| 1182 |
+
"grad_norm": 0.257354736328125,
|
| 1183 |
+
"learning_rate": 4.16e-05,
|
| 1184 |
+
"loss": 0.0536,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.169,
|
| 1189 |
+
"grad_norm": 0.25215694308280945,
|
| 1190 |
+
"learning_rate": 4.155e-05,
|
| 1191 |
+
"loss": 0.0587,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 0.17,
|
| 1196 |
+
"grad_norm": 0.4573931097984314,
|
| 1197 |
+
"learning_rate": 4.15e-05,
|
| 1198 |
+
"loss": 0.0524,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.171,
|
| 1203 |
+
"grad_norm": 0.3514876663684845,
|
| 1204 |
+
"learning_rate": 4.145e-05,
|
| 1205 |
+
"loss": 0.0551,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 0.172,
|
| 1210 |
+
"grad_norm": 0.3239930272102356,
|
| 1211 |
+
"learning_rate": 4.14e-05,
|
| 1212 |
+
"loss": 0.0499,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 0.173,
|
| 1217 |
+
"grad_norm": 0.20213039219379425,
|
| 1218 |
+
"learning_rate": 4.135e-05,
|
| 1219 |
+
"loss": 0.0521,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 0.174,
|
| 1224 |
+
"grad_norm": 0.21831783652305603,
|
| 1225 |
+
"learning_rate": 4.13e-05,
|
| 1226 |
+
"loss": 0.0469,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 0.175,
|
| 1231 |
+
"grad_norm": 0.2585163712501526,
|
| 1232 |
+
"learning_rate": 4.125e-05,
|
| 1233 |
+
"loss": 0.0469,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 0.176,
|
| 1238 |
+
"grad_norm": 0.21717113256454468,
|
| 1239 |
+
"learning_rate": 4.12e-05,
|
| 1240 |
+
"loss": 0.0455,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.177,
|
| 1245 |
+
"grad_norm": 0.27248838543891907,
|
| 1246 |
+
"learning_rate": 4.115e-05,
|
| 1247 |
+
"loss": 0.046,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.178,
|
| 1252 |
+
"grad_norm": 0.2503461241722107,
|
| 1253 |
+
"learning_rate": 4.11e-05,
|
| 1254 |
+
"loss": 0.0447,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.179,
|
| 1259 |
+
"grad_norm": 0.27404382824897766,
|
| 1260 |
+
"learning_rate": 4.105e-05,
|
| 1261 |
+
"loss": 0.0437,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.18,
|
| 1266 |
+
"grad_norm": 0.23549066483974457,
|
| 1267 |
+
"learning_rate": 4.1e-05,
|
| 1268 |
+
"loss": 0.0423,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.181,
|
| 1273 |
+
"grad_norm": 0.19369937479496002,
|
| 1274 |
+
"learning_rate": 4.095e-05,
|
| 1275 |
+
"loss": 0.0408,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 0.182,
|
| 1280 |
+
"grad_norm": 0.20560242235660553,
|
| 1281 |
+
"learning_rate": 4.09e-05,
|
| 1282 |
+
"loss": 0.0379,
|
| 1283 |
+
"step": 1820
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.183,
|
| 1287 |
+
"grad_norm": 0.34989863634109497,
|
| 1288 |
+
"learning_rate": 4.085e-05,
|
| 1289 |
+
"loss": 0.0364,
|
| 1290 |
+
"step": 1830
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 0.184,
|
| 1294 |
+
"grad_norm": 0.2310326248407364,
|
| 1295 |
+
"learning_rate": 4.08e-05,
|
| 1296 |
+
"loss": 0.0385,
|
| 1297 |
+
"step": 1840
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 0.185,
|
| 1301 |
+
"grad_norm": 0.21055462956428528,
|
| 1302 |
+
"learning_rate": 4.075e-05,
|
| 1303 |
+
"loss": 0.0351,
|
| 1304 |
+
"step": 1850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.186,
|
| 1308 |
+
"grad_norm": 0.3251895308494568,
|
| 1309 |
+
"learning_rate": 4.07e-05,
|
| 1310 |
+
"loss": 0.0381,
|
| 1311 |
+
"step": 1860
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.187,
|
| 1315 |
+
"grad_norm": 0.2887445390224457,
|
| 1316 |
+
"learning_rate": 4.065e-05,
|
| 1317 |
+
"loss": 0.0341,
|
| 1318 |
+
"step": 1870
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 0.188,
|
| 1322 |
+
"grad_norm": 0.15948843955993652,
|
| 1323 |
+
"learning_rate": 4.0600000000000004e-05,
|
| 1324 |
+
"loss": 0.0313,
|
| 1325 |
+
"step": 1880
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 0.189,
|
| 1329 |
+
"grad_norm": 0.2413359135389328,
|
| 1330 |
+
"learning_rate": 4.055e-05,
|
| 1331 |
+
"loss": 0.0338,
|
| 1332 |
+
"step": 1890
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 0.19,
|
| 1336 |
+
"grad_norm": 0.2132706493139267,
|
| 1337 |
+
"learning_rate": 4.05e-05,
|
| 1338 |
+
"loss": 0.0339,
|
| 1339 |
+
"step": 1900
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 0.191,
|
| 1343 |
+
"grad_norm": 0.17968431115150452,
|
| 1344 |
+
"learning_rate": 4.045000000000001e-05,
|
| 1345 |
+
"loss": 0.0317,
|
| 1346 |
+
"step": 1910
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 0.192,
|
| 1350 |
+
"grad_norm": 0.15828929841518402,
|
| 1351 |
+
"learning_rate": 4.0400000000000006e-05,
|
| 1352 |
+
"loss": 0.0302,
|
| 1353 |
+
"step": 1920
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 0.193,
|
| 1357 |
+
"grad_norm": 0.18106874823570251,
|
| 1358 |
+
"learning_rate": 4.0350000000000005e-05,
|
| 1359 |
+
"loss": 0.0331,
|
| 1360 |
+
"step": 1930
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 0.194,
|
| 1364 |
+
"grad_norm": 0.34827324748039246,
|
| 1365 |
+
"learning_rate": 4.0300000000000004e-05,
|
| 1366 |
+
"loss": 0.032,
|
| 1367 |
+
"step": 1940
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 0.195,
|
| 1371 |
+
"grad_norm": 0.21621111035346985,
|
| 1372 |
+
"learning_rate": 4.025e-05,
|
| 1373 |
+
"loss": 0.0317,
|
| 1374 |
+
"step": 1950
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 0.196,
|
| 1378 |
+
"grad_norm": 0.2159423679113388,
|
| 1379 |
+
"learning_rate": 4.02e-05,
|
| 1380 |
+
"loss": 0.0296,
|
| 1381 |
+
"step": 1960
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 0.197,
|
| 1385 |
+
"grad_norm": 0.17750391364097595,
|
| 1386 |
+
"learning_rate": 4.015000000000001e-05,
|
| 1387 |
+
"loss": 0.0297,
|
| 1388 |
+
"step": 1970
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 0.198,
|
| 1392 |
+
"grad_norm": 0.13952311873435974,
|
| 1393 |
+
"learning_rate": 4.0100000000000006e-05,
|
| 1394 |
+
"loss": 0.0279,
|
| 1395 |
+
"step": 1980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 0.199,
|
| 1399 |
+
"grad_norm": 0.19622887670993805,
|
| 1400 |
+
"learning_rate": 4.0050000000000004e-05,
|
| 1401 |
+
"loss": 0.0278,
|
| 1402 |
+
"step": 1990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 0.2,
|
| 1406 |
+
"grad_norm": 0.14959514141082764,
|
| 1407 |
+
"learning_rate": 4e-05,
|
| 1408 |
+
"loss": 0.0251,
|
| 1409 |
+
"step": 2000
|
| 1410 |
+
}
|
| 1411 |
+
],
|
| 1412 |
+
"logging_steps": 10,
|
| 1413 |
+
"max_steps": 10000,
|
| 1414 |
+
"num_input_tokens_seen": 0,
|
| 1415 |
+
"num_train_epochs": 9223372036854775807,
|
| 1416 |
+
"save_steps": 1000,
|
| 1417 |
+
"stateful_callbacks": {
|
| 1418 |
+
"TrainerControl": {
|
| 1419 |
+
"args": {
|
| 1420 |
+
"should_epoch_stop": false,
|
| 1421 |
+
"should_evaluate": false,
|
| 1422 |
+
"should_log": false,
|
| 1423 |
+
"should_save": true,
|
| 1424 |
+
"should_training_stop": false
|
| 1425 |
+
},
|
| 1426 |
+
"attributes": {}
|
| 1427 |
+
}
|
| 1428 |
+
},
|
| 1429 |
+
"total_flos": 7.3924406673408e+16,
|
| 1430 |
+
"train_batch_size": 1,
|
| 1431 |
+
"trial_name": null,
|
| 1432 |
+
"trial_params": null
|
| 1433 |
+
}
|