Commit
•
75ae99f
1
Parent(s):
3542d18
Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +2308 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,2308 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3250,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.03076923076923077,
|
| 13 |
+
"grad_norm": 1.4984049797058105,
|
| 14 |
+
"learning_rate": 0.0001999953280342959,
|
| 15 |
+
"loss": 2.6105,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.06153846153846154,
|
| 20 |
+
"grad_norm": 13.759217262268066,
|
| 21 |
+
"learning_rate": 0.00019998131257372876,
|
| 22 |
+
"loss": 2.1182,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.09230769230769231,
|
| 27 |
+
"grad_norm": 1.4669537544250488,
|
| 28 |
+
"learning_rate": 0.0001999579549278937,
|
| 29 |
+
"loss": 1.9057,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.12307692307692308,
|
| 34 |
+
"grad_norm": 1.4035037755966187,
|
| 35 |
+
"learning_rate": 0.00019992525727931303,
|
| 36 |
+
"loss": 1.6194,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.15384615384615385,
|
| 41 |
+
"grad_norm": 1.9111855030059814,
|
| 42 |
+
"learning_rate": 0.00019988322268323268,
|
| 43 |
+
"loss": 1.4734,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.18461538461538463,
|
| 48 |
+
"grad_norm": 1.690224051475525,
|
| 49 |
+
"learning_rate": 0.0001998318550673364,
|
| 50 |
+
"loss": 1.4012,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.2153846153846154,
|
| 55 |
+
"grad_norm": 1.4565370082855225,
|
| 56 |
+
"learning_rate": 0.00019977115923137912,
|
| 57 |
+
"loss": 1.3154,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.24615384615384617,
|
| 62 |
+
"grad_norm": 1.325324535369873,
|
| 63 |
+
"learning_rate": 0.00019970114084673796,
|
| 64 |
+
"loss": 1.2612,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.27692307692307694,
|
| 69 |
+
"grad_norm": 1.8117148876190186,
|
| 70 |
+
"learning_rate": 0.0001996218064558829,
|
| 71 |
+
"loss": 1.1993,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.3076923076923077,
|
| 76 |
+
"grad_norm": 1.4131654500961304,
|
| 77 |
+
"learning_rate": 0.00019953316347176488,
|
| 78 |
+
"loss": 1.1835,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.3384615384615385,
|
| 83 |
+
"grad_norm": 1.9338051080703735,
|
| 84 |
+
"learning_rate": 0.00019943522017712358,
|
| 85 |
+
"loss": 1.2223,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.36923076923076925,
|
| 90 |
+
"grad_norm": 1.7367668151855469,
|
| 91 |
+
"learning_rate": 0.0001993279857237133,
|
| 92 |
+
"loss": 1.1528,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.4,
|
| 97 |
+
"grad_norm": 1.5504117012023926,
|
| 98 |
+
"learning_rate": 0.0001992114701314478,
|
| 99 |
+
"loss": 1.1799,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.4307692307692308,
|
| 104 |
+
"grad_norm": 1.7593408823013306,
|
| 105 |
+
"learning_rate": 0.0001990856842874641,
|
| 106 |
+
"loss": 1.1915,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.46153846153846156,
|
| 111 |
+
"grad_norm": 1.5112876892089844,
|
| 112 |
+
"learning_rate": 0.0001989506399451051,
|
| 113 |
+
"loss": 1.1552,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.49230769230769234,
|
| 118 |
+
"grad_norm": 1.5315320491790771,
|
| 119 |
+
"learning_rate": 0.00019880634972282166,
|
| 120 |
+
"loss": 1.1271,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.5230769230769231,
|
| 125 |
+
"grad_norm": 1.7272446155548096,
|
| 126 |
+
"learning_rate": 0.0001986528271029931,
|
| 127 |
+
"loss": 1.1321,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.5538461538461539,
|
| 132 |
+
"grad_norm": 1.8666999340057373,
|
| 133 |
+
"learning_rate": 0.00019849008643066772,
|
| 134 |
+
"loss": 1.1367,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.5846153846153846,
|
| 139 |
+
"grad_norm": 1.8758796453475952,
|
| 140 |
+
"learning_rate": 0.00019831814291222232,
|
| 141 |
+
"loss": 1.1219,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.6153846153846154,
|
| 146 |
+
"grad_norm": 1.7113131284713745,
|
| 147 |
+
"learning_rate": 0.00019813701261394136,
|
| 148 |
+
"loss": 1.1101,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.6461538461538462,
|
| 153 |
+
"grad_norm": 1.506138801574707,
|
| 154 |
+
"learning_rate": 0.0001979467124605156,
|
| 155 |
+
"loss": 1.0801,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.676923076923077,
|
| 160 |
+
"grad_norm": 1.4458012580871582,
|
| 161 |
+
"learning_rate": 0.0001977472602334609,
|
| 162 |
+
"loss": 1.0838,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.7076923076923077,
|
| 167 |
+
"grad_norm": 1.615057349205017,
|
| 168 |
+
"learning_rate": 0.0001975386745694565,
|
| 169 |
+
"loss": 1.094,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.7384615384615385,
|
| 174 |
+
"grad_norm": 1.4620922803878784,
|
| 175 |
+
"learning_rate": 0.00019732097495860386,
|
| 176 |
+
"loss": 1.0689,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.7692307692307693,
|
| 181 |
+
"grad_norm": 1.5983144044876099,
|
| 182 |
+
"learning_rate": 0.0001970941817426052,
|
| 183 |
+
"loss": 1.072,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.8,
|
| 188 |
+
"grad_norm": 1.5557613372802734,
|
| 189 |
+
"learning_rate": 0.0001968583161128631,
|
| 190 |
+
"loss": 1.0974,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.8307692307692308,
|
| 195 |
+
"grad_norm": 1.674525499343872,
|
| 196 |
+
"learning_rate": 0.00019661340010850026,
|
| 197 |
+
"loss": 1.0958,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.8615384615384616,
|
| 202 |
+
"grad_norm": 1.9557188749313354,
|
| 203 |
+
"learning_rate": 0.00019635945661430006,
|
| 204 |
+
"loss": 1.0713,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.8923076923076924,
|
| 209 |
+
"grad_norm": 1.408217191696167,
|
| 210 |
+
"learning_rate": 0.00019609650935856844,
|
| 211 |
+
"loss": 1.0681,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.9230769230769231,
|
| 216 |
+
"grad_norm": 1.6660012006759644,
|
| 217 |
+
"learning_rate": 0.00019582458291091663,
|
| 218 |
+
"loss": 1.0639,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.9538461538461539,
|
| 223 |
+
"grad_norm": 1.5910792350769043,
|
| 224 |
+
"learning_rate": 0.00019554370267996538,
|
| 225 |
+
"loss": 1.0844,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.9846153846153847,
|
| 230 |
+
"grad_norm": 1.698133111000061,
|
| 231 |
+
"learning_rate": 0.0001952538949109708,
|
| 232 |
+
"loss": 1.1323,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.0153846153846153,
|
| 237 |
+
"grad_norm": 1.7201091051101685,
|
| 238 |
+
"learning_rate": 0.00019495518668337201,
|
| 239 |
+
"loss": 1.0761,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.0461538461538462,
|
| 244 |
+
"grad_norm": 1.7263455390930176,
|
| 245 |
+
"learning_rate": 0.00019464760590826098,
|
| 246 |
+
"loss": 1.0388,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.0769230769230769,
|
| 251 |
+
"grad_norm": 1.8437412977218628,
|
| 252 |
+
"learning_rate": 0.0001943311813257743,
|
| 253 |
+
"loss": 1.0229,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.1076923076923078,
|
| 258 |
+
"grad_norm": 1.4422318935394287,
|
| 259 |
+
"learning_rate": 0.00019400594250240798,
|
| 260 |
+
"loss": 1.0218,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.1384615384615384,
|
| 265 |
+
"grad_norm": 1.765499472618103,
|
| 266 |
+
"learning_rate": 0.0001936719198282545,
|
| 267 |
+
"loss": 1.0087,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.1692307692307693,
|
| 272 |
+
"grad_norm": 1.500490427017212,
|
| 273 |
+
"learning_rate": 0.00019332914451416347,
|
| 274 |
+
"loss": 1.0239,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.2,
|
| 279 |
+
"grad_norm": 1.5604429244995117,
|
| 280 |
+
"learning_rate": 0.00019297764858882514,
|
| 281 |
+
"loss": 1.0271,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.2307692307692308,
|
| 286 |
+
"grad_norm": 1.636246681213379,
|
| 287 |
+
"learning_rate": 0.00019261746489577765,
|
| 288 |
+
"loss": 1.0537,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.2615384615384615,
|
| 293 |
+
"grad_norm": 1.7009079456329346,
|
| 294 |
+
"learning_rate": 0.00019224862709033824,
|
| 295 |
+
"loss": 1.0267,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.2923076923076924,
|
| 300 |
+
"grad_norm": 1.7331833839416504,
|
| 301 |
+
"learning_rate": 0.00019187116963645842,
|
| 302 |
+
"loss": 0.962,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.323076923076923,
|
| 307 |
+
"grad_norm": 1.5424834489822388,
|
| 308 |
+
"learning_rate": 0.00019148512780350384,
|
| 309 |
+
"loss": 0.9834,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.353846153846154,
|
| 314 |
+
"grad_norm": 1.474749207496643,
|
| 315 |
+
"learning_rate": 0.0001910905376629585,
|
| 316 |
+
"loss": 1.0047,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.3846153846153846,
|
| 321 |
+
"grad_norm": 1.739359974861145,
|
| 322 |
+
"learning_rate": 0.00019068743608505455,
|
| 323 |
+
"loss": 1.0188,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.4153846153846155,
|
| 328 |
+
"grad_norm": 1.52376389503479,
|
| 329 |
+
"learning_rate": 0.0001902758607353269,
|
| 330 |
+
"loss": 1.0226,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.4461538461538461,
|
| 335 |
+
"grad_norm": 1.723073124885559,
|
| 336 |
+
"learning_rate": 0.0001898558500710939,
|
| 337 |
+
"loss": 1.0054,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.476923076923077,
|
| 342 |
+
"grad_norm": 1.7302210330963135,
|
| 343 |
+
"learning_rate": 0.00018942744333786397,
|
| 344 |
+
"loss": 1.0335,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.5076923076923077,
|
| 349 |
+
"grad_norm": 1.5802934169769287,
|
| 350 |
+
"learning_rate": 0.0001889906805656684,
|
| 351 |
+
"loss": 0.9924,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.5384615384615383,
|
| 356 |
+
"grad_norm": 1.383296012878418,
|
| 357 |
+
"learning_rate": 0.000188545602565321,
|
| 358 |
+
"loss": 1.0083,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.5692307692307692,
|
| 363 |
+
"grad_norm": 1.5573105812072754,
|
| 364 |
+
"learning_rate": 0.00018809225092460488,
|
| 365 |
+
"loss": 0.9982,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.6,
|
| 370 |
+
"grad_norm": 1.5339369773864746,
|
| 371 |
+
"learning_rate": 0.00018763066800438636,
|
| 372 |
+
"loss": 0.9827,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.6307692307692307,
|
| 377 |
+
"grad_norm": 1.5264880657196045,
|
| 378 |
+
"learning_rate": 0.00018716089693465696,
|
| 379 |
+
"loss": 0.9879,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.6615384615384614,
|
| 384 |
+
"grad_norm": 1.4193902015686035,
|
| 385 |
+
"learning_rate": 0.00018668298161050309,
|
| 386 |
+
"loss": 0.9825,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.6923076923076923,
|
| 391 |
+
"grad_norm": 1.6493202447891235,
|
| 392 |
+
"learning_rate": 0.00018619696668800492,
|
| 393 |
+
"loss": 1.0143,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.7230769230769232,
|
| 398 |
+
"grad_norm": 1.4842519760131836,
|
| 399 |
+
"learning_rate": 0.00018570289758006346,
|
| 400 |
+
"loss": 0.9721,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.7538461538461538,
|
| 405 |
+
"grad_norm": 1.617529034614563,
|
| 406 |
+
"learning_rate": 0.0001852008204521572,
|
| 407 |
+
"loss": 0.9826,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.7846153846153845,
|
| 412 |
+
"grad_norm": 1.6360862255096436,
|
| 413 |
+
"learning_rate": 0.0001846907822180286,
|
| 414 |
+
"loss": 0.9236,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.8153846153846154,
|
| 419 |
+
"grad_norm": 1.664921760559082,
|
| 420 |
+
"learning_rate": 0.00018417283053530044,
|
| 421 |
+
"loss": 1.0072,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.8461538461538463,
|
| 426 |
+
"grad_norm": 1.5803855657577515,
|
| 427 |
+
"learning_rate": 0.00018364701380102266,
|
| 428 |
+
"loss": 1.0154,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.876923076923077,
|
| 433 |
+
"grad_norm": 1.6201574802398682,
|
| 434 |
+
"learning_rate": 0.0001831133811471503,
|
| 435 |
+
"loss": 1.0299,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.9076923076923076,
|
| 440 |
+
"grad_norm": 1.5785369873046875,
|
| 441 |
+
"learning_rate": 0.0001825719824359524,
|
| 442 |
+
"loss": 1.015,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.9384615384615385,
|
| 447 |
+
"grad_norm": 1.6001813411712646,
|
| 448 |
+
"learning_rate": 0.0001820228682553533,
|
| 449 |
+
"loss": 0.967,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.9692307692307693,
|
| 454 |
+
"grad_norm": 1.5755499601364136,
|
| 455 |
+
"learning_rate": 0.00018146608991420534,
|
| 456 |
+
"loss": 0.9779,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 2.0,
|
| 461 |
+
"grad_norm": 1.593526840209961,
|
| 462 |
+
"learning_rate": 0.00018090169943749476,
|
| 463 |
+
"loss": 1.0293,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.0307692307692307,
|
| 468 |
+
"grad_norm": 2.3367538452148438,
|
| 469 |
+
"learning_rate": 0.00018032974956148063,
|
| 470 |
+
"loss": 0.9708,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 2.0615384615384613,
|
| 475 |
+
"grad_norm": 1.7174549102783203,
|
| 476 |
+
"learning_rate": 0.00017975029372876706,
|
| 477 |
+
"loss": 0.9657,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 2.0923076923076924,
|
| 482 |
+
"grad_norm": 1.5391074419021606,
|
| 483 |
+
"learning_rate": 0.0001791633860833096,
|
| 484 |
+
"loss": 0.9353,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 2.123076923076923,
|
| 489 |
+
"grad_norm": 1.7437571287155151,
|
| 490 |
+
"learning_rate": 0.00017856908146535603,
|
| 491 |
+
"loss": 0.9348,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 2.1538461538461537,
|
| 496 |
+
"grad_norm": 1.6768618822097778,
|
| 497 |
+
"learning_rate": 0.00017796743540632223,
|
| 498 |
+
"loss": 0.9376,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 2.184615384615385,
|
| 503 |
+
"grad_norm": 1.6524372100830078,
|
| 504 |
+
"learning_rate": 0.00017735850412360331,
|
| 505 |
+
"loss": 0.9116,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.2153846153846155,
|
| 510 |
+
"grad_norm": 1.7284245491027832,
|
| 511 |
+
"learning_rate": 0.00017674234451532065,
|
| 512 |
+
"loss": 0.9618,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.246153846153846,
|
| 517 |
+
"grad_norm": 1.5894191265106201,
|
| 518 |
+
"learning_rate": 0.00017611901415500535,
|
| 519 |
+
"loss": 0.924,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 2.276923076923077,
|
| 524 |
+
"grad_norm": 1.6842291355133057,
|
| 525 |
+
"learning_rate": 0.00017548857128621875,
|
| 526 |
+
"loss": 0.9143,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.3076923076923075,
|
| 531 |
+
"grad_norm": 1.7472116947174072,
|
| 532 |
+
"learning_rate": 0.00017485107481711012,
|
| 533 |
+
"loss": 0.9417,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.3384615384615386,
|
| 538 |
+
"grad_norm": 1.4523415565490723,
|
| 539 |
+
"learning_rate": 0.00017420658431491223,
|
| 540 |
+
"loss": 0.9279,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.3692307692307693,
|
| 545 |
+
"grad_norm": 1.6412981748580933,
|
| 546 |
+
"learning_rate": 0.00017355516000037554,
|
| 547 |
+
"loss": 0.9273,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.4,
|
| 552 |
+
"grad_norm": 1.49977445602417,
|
| 553 |
+
"learning_rate": 0.00017289686274214118,
|
| 554 |
+
"loss": 0.9011,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.430769230769231,
|
| 559 |
+
"grad_norm": 2.0299291610717773,
|
| 560 |
+
"learning_rate": 0.0001722317540510534,
|
| 561 |
+
"loss": 0.919,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.4615384615384617,
|
| 566 |
+
"grad_norm": 1.8127689361572266,
|
| 567 |
+
"learning_rate": 0.00017155989607441213,
|
| 568 |
+
"loss": 0.9446,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.4923076923076923,
|
| 573 |
+
"grad_norm": 1.7033181190490723,
|
| 574 |
+
"learning_rate": 0.00017088135159016584,
|
| 575 |
+
"loss": 0.9044,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.523076923076923,
|
| 580 |
+
"grad_norm": 1.5404083728790283,
|
| 581 |
+
"learning_rate": 0.00017019618400104572,
|
| 582 |
+
"loss": 0.8942,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.5538461538461537,
|
| 587 |
+
"grad_norm": 1.6343528032302856,
|
| 588 |
+
"learning_rate": 0.00016950445732864127,
|
| 589 |
+
"loss": 0.98,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.5846153846153848,
|
| 594 |
+
"grad_norm": 1.7268812656402588,
|
| 595 |
+
"learning_rate": 0.00016880623620741842,
|
| 596 |
+
"loss": 0.9107,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.6153846153846154,
|
| 601 |
+
"grad_norm": 1.6956366300582886,
|
| 602 |
+
"learning_rate": 0.00016810158587867973,
|
| 603 |
+
"loss": 0.9351,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.646153846153846,
|
| 608 |
+
"grad_norm": 1.7267690896987915,
|
| 609 |
+
"learning_rate": 0.0001673905721844686,
|
| 610 |
+
"loss": 0.9147,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.676923076923077,
|
| 615 |
+
"grad_norm": 1.5935823917388916,
|
| 616 |
+
"learning_rate": 0.00016667326156141692,
|
| 617 |
+
"loss": 0.9393,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.707692307692308,
|
| 622 |
+
"grad_norm": 1.5465987920761108,
|
| 623 |
+
"learning_rate": 0.00016594972103453726,
|
| 624 |
+
"loss": 0.9158,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.7384615384615385,
|
| 629 |
+
"grad_norm": 1.7121493816375732,
|
| 630 |
+
"learning_rate": 0.0001652200182109602,
|
| 631 |
+
"loss": 0.934,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.769230769230769,
|
| 636 |
+
"grad_norm": 1.7504358291625977,
|
| 637 |
+
"learning_rate": 0.00016448422127361706,
|
| 638 |
+
"loss": 0.9277,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.8,
|
| 643 |
+
"grad_norm": 1.8299874067306519,
|
| 644 |
+
"learning_rate": 0.000163742398974869,
|
| 645 |
+
"loss": 0.9498,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.830769230769231,
|
| 650 |
+
"grad_norm": 1.7550499439239502,
|
| 651 |
+
"learning_rate": 0.00016299462063008272,
|
| 652 |
+
"loss": 0.9769,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 2.8615384615384616,
|
| 657 |
+
"grad_norm": 1.6253832578659058,
|
| 658 |
+
"learning_rate": 0.00016224095611115384,
|
| 659 |
+
"loss": 0.8633,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 2.8923076923076922,
|
| 664 |
+
"grad_norm": 1.7190197706222534,
|
| 665 |
+
"learning_rate": 0.00016148147583997812,
|
| 666 |
+
"loss": 0.9252,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 2.9230769230769234,
|
| 671 |
+
"grad_norm": 1.6005258560180664,
|
| 672 |
+
"learning_rate": 0.00016071625078187114,
|
| 673 |
+
"loss": 0.9606,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.953846153846154,
|
| 678 |
+
"grad_norm": 1.6922186613082886,
|
| 679 |
+
"learning_rate": 0.0001599453524389374,
|
| 680 |
+
"loss": 0.8922,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 2.9846153846153847,
|
| 685 |
+
"grad_norm": 1.7878159284591675,
|
| 686 |
+
"learning_rate": 0.00015916885284338937,
|
| 687 |
+
"loss": 0.9142,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 3.0153846153846153,
|
| 692 |
+
"grad_norm": 1.6686667203903198,
|
| 693 |
+
"learning_rate": 0.00015838682455081657,
|
| 694 |
+
"loss": 0.925,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 3.046153846153846,
|
| 699 |
+
"grad_norm": 1.6054987907409668,
|
| 700 |
+
"learning_rate": 0.00015759934063340627,
|
| 701 |
+
"loss": 0.9249,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 3.076923076923077,
|
| 706 |
+
"grad_norm": 1.7218083143234253,
|
| 707 |
+
"learning_rate": 0.00015680647467311557,
|
| 708 |
+
"loss": 0.8723,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 3.1076923076923078,
|
| 713 |
+
"grad_norm": 2.051055431365967,
|
| 714 |
+
"learning_rate": 0.00015600830075479603,
|
| 715 |
+
"loss": 0.8255,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 3.1384615384615384,
|
| 720 |
+
"grad_norm": 1.7609128952026367,
|
| 721 |
+
"learning_rate": 0.00015520489345927096,
|
| 722 |
+
"loss": 0.8821,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 3.169230769230769,
|
| 727 |
+
"grad_norm": 1.690056562423706,
|
| 728 |
+
"learning_rate": 0.00015439632785636706,
|
| 729 |
+
"loss": 0.9113,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 3.2,
|
| 734 |
+
"grad_norm": 1.9333645105361938,
|
| 735 |
+
"learning_rate": 0.00015358267949789966,
|
| 736 |
+
"loss": 0.9241,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 3.230769230769231,
|
| 741 |
+
"grad_norm": 1.7492021322250366,
|
| 742 |
+
"learning_rate": 0.0001527640244106133,
|
| 743 |
+
"loss": 0.8993,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 3.2615384615384615,
|
| 748 |
+
"grad_norm": 1.8243930339813232,
|
| 749 |
+
"learning_rate": 0.00015194043908907775,
|
| 750 |
+
"loss": 0.8928,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 3.292307692307692,
|
| 755 |
+
"grad_norm": 1.7545908689498901,
|
| 756 |
+
"learning_rate": 0.00015111200048854056,
|
| 757 |
+
"loss": 0.8879,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 3.3230769230769233,
|
| 762 |
+
"grad_norm": 1.7745007276535034,
|
| 763 |
+
"learning_rate": 0.00015027878601773633,
|
| 764 |
+
"loss": 0.869,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 3.353846153846154,
|
| 769 |
+
"grad_norm": 1.8173617124557495,
|
| 770 |
+
"learning_rate": 0.0001494408735316537,
|
| 771 |
+
"loss": 0.9033,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 3.3846153846153846,
|
| 776 |
+
"grad_norm": 1.9644584655761719,
|
| 777 |
+
"learning_rate": 0.0001485983413242606,
|
| 778 |
+
"loss": 0.8392,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 3.4153846153846152,
|
| 783 |
+
"grad_norm": 1.8126754760742188,
|
| 784 |
+
"learning_rate": 0.00014775126812118864,
|
| 785 |
+
"loss": 0.8928,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 3.4461538461538463,
|
| 790 |
+
"grad_norm": 1.8512457609176636,
|
| 791 |
+
"learning_rate": 0.00014689973307237687,
|
| 792 |
+
"loss": 0.887,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 3.476923076923077,
|
| 797 |
+
"grad_norm": 1.7760255336761475,
|
| 798 |
+
"learning_rate": 0.00014604381574467615,
|
| 799 |
+
"loss": 0.8951,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 3.5076923076923077,
|
| 804 |
+
"grad_norm": 1.7298815250396729,
|
| 805 |
+
"learning_rate": 0.0001451835961144145,
|
| 806 |
+
"loss": 0.8799,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 3.5384615384615383,
|
| 811 |
+
"grad_norm": 1.6533232927322388,
|
| 812 |
+
"learning_rate": 0.00014431915455992414,
|
| 813 |
+
"loss": 0.8652,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 3.569230769230769,
|
| 818 |
+
"grad_norm": 1.8500922918319702,
|
| 819 |
+
"learning_rate": 0.000143450571854031,
|
| 820 |
+
"loss": 0.8864,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 3.6,
|
| 825 |
+
"grad_norm": 1.719185471534729,
|
| 826 |
+
"learning_rate": 0.00014257792915650728,
|
| 827 |
+
"loss": 0.8602,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 3.6307692307692307,
|
| 832 |
+
"grad_norm": 2.0936434268951416,
|
| 833 |
+
"learning_rate": 0.00014170130800648814,
|
| 834 |
+
"loss": 0.8538,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 3.6615384615384614,
|
| 839 |
+
"grad_norm": 1.9639226198196411,
|
| 840 |
+
"learning_rate": 0.0001408207903148525,
|
| 841 |
+
"loss": 0.8279,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 3.6923076923076925,
|
| 846 |
+
"grad_norm": 2.0342531204223633,
|
| 847 |
+
"learning_rate": 0.00013993645835656953,
|
| 848 |
+
"loss": 0.8683,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 3.723076923076923,
|
| 853 |
+
"grad_norm": 1.8024394512176514,
|
| 854 |
+
"learning_rate": 0.0001390483947630109,
|
| 855 |
+
"loss": 0.888,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 3.753846153846154,
|
| 860 |
+
"grad_norm": 1.8016464710235596,
|
| 861 |
+
"learning_rate": 0.00013815668251422952,
|
| 862 |
+
"loss": 0.9126,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 3.7846153846153845,
|
| 867 |
+
"grad_norm": 1.8998101949691772,
|
| 868 |
+
"learning_rate": 0.0001372614049312064,
|
| 869 |
+
"loss": 0.8799,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 3.815384615384615,
|
| 874 |
+
"grad_norm": 1.674023151397705,
|
| 875 |
+
"learning_rate": 0.0001363626456680647,
|
| 876 |
+
"loss": 0.8822,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 3.8461538461538463,
|
| 881 |
+
"grad_norm": 1.7552895545959473,
|
| 882 |
+
"learning_rate": 0.00013546048870425356,
|
| 883 |
+
"loss": 0.8456,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 3.876923076923077,
|
| 888 |
+
"grad_norm": 1.790547490119934,
|
| 889 |
+
"learning_rate": 0.00013455501833670088,
|
| 890 |
+
"loss": 0.8149,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 3.9076923076923076,
|
| 895 |
+
"grad_norm": 1.8295682668685913,
|
| 896 |
+
"learning_rate": 0.0001336463191719367,
|
| 897 |
+
"loss": 0.8622,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 3.9384615384615387,
|
| 902 |
+
"grad_norm": 1.7753140926361084,
|
| 903 |
+
"learning_rate": 0.00013273447611818767,
|
| 904 |
+
"loss": 0.8678,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 3.9692307692307693,
|
| 909 |
+
"grad_norm": 1.808028221130371,
|
| 910 |
+
"learning_rate": 0.00013181957437744332,
|
| 911 |
+
"loss": 0.8578,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 4.0,
|
| 916 |
+
"grad_norm": 1.848429799079895,
|
| 917 |
+
"learning_rate": 0.00013090169943749476,
|
| 918 |
+
"loss": 0.8463,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 4.030769230769231,
|
| 923 |
+
"grad_norm": 1.9943642616271973,
|
| 924 |
+
"learning_rate": 0.00012998093706394675,
|
| 925 |
+
"loss": 0.8016,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 4.061538461538461,
|
| 930 |
+
"grad_norm": 1.9723416566848755,
|
| 931 |
+
"learning_rate": 0.00012905737329220392,
|
| 932 |
+
"loss": 0.8379,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 4.092307692307692,
|
| 937 |
+
"grad_norm": 2.0291624069213867,
|
| 938 |
+
"learning_rate": 0.00012813109441943166,
|
| 939 |
+
"loss": 0.8041,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 4.123076923076923,
|
| 944 |
+
"grad_norm": 1.9695005416870117,
|
| 945 |
+
"learning_rate": 0.00012720218699649243,
|
| 946 |
+
"loss": 0.8507,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 4.153846153846154,
|
| 951 |
+
"grad_norm": 2.0986056327819824,
|
| 952 |
+
"learning_rate": 0.0001262707378198587,
|
| 953 |
+
"loss": 0.8786,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 4.184615384615385,
|
| 958 |
+
"grad_norm": 1.849661946296692,
|
| 959 |
+
"learning_rate": 0.00012533683392350263,
|
| 960 |
+
"loss": 0.8659,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 4.2153846153846155,
|
| 965 |
+
"grad_norm": 1.8133331537246704,
|
| 966 |
+
"learning_rate": 0.00012440056257076375,
|
| 967 |
+
"loss": 0.833,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 4.246153846153846,
|
| 972 |
+
"grad_norm": 2.1380434036254883,
|
| 973 |
+
"learning_rate": 0.00012346201124619502,
|
| 974 |
+
"loss": 0.8067,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 4.276923076923077,
|
| 979 |
+
"grad_norm": 1.7417466640472412,
|
| 980 |
+
"learning_rate": 0.00012252126764738844,
|
| 981 |
+
"loss": 0.8342,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 4.3076923076923075,
|
| 986 |
+
"grad_norm": 1.8836528062820435,
|
| 987 |
+
"learning_rate": 0.00012157841967678063,
|
| 988 |
+
"loss": 0.8251,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 4.338461538461538,
|
| 993 |
+
"grad_norm": 2.2744300365448,
|
| 994 |
+
"learning_rate": 0.00012063355543343924,
|
| 995 |
+
"loss": 0.8138,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 4.36923076923077,
|
| 1000 |
+
"grad_norm": 2.028259754180908,
|
| 1001 |
+
"learning_rate": 0.00011968676320483103,
|
| 1002 |
+
"loss": 0.8063,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 4.4,
|
| 1007 |
+
"grad_norm": 1.8546655178070068,
|
| 1008 |
+
"learning_rate": 0.00011873813145857249,
|
| 1009 |
+
"loss": 0.8415,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 4.430769230769231,
|
| 1014 |
+
"grad_norm": 1.8695989847183228,
|
| 1015 |
+
"learning_rate": 0.00011778774883416323,
|
| 1016 |
+
"loss": 0.8486,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 4.461538461538462,
|
| 1021 |
+
"grad_norm": 1.8322254419326782,
|
| 1022 |
+
"learning_rate": 0.00011683570413470383,
|
| 1023 |
+
"loss": 0.8131,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 4.492307692307692,
|
| 1028 |
+
"grad_norm": 2.126826047897339,
|
| 1029 |
+
"learning_rate": 0.00011588208631859807,
|
| 1030 |
+
"loss": 0.8307,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 4.523076923076923,
|
| 1035 |
+
"grad_norm": 1.9609365463256836,
|
| 1036 |
+
"learning_rate": 0.00011492698449124042,
|
| 1037 |
+
"loss": 0.8343,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 4.553846153846154,
|
| 1042 |
+
"grad_norm": 1.7528473138809204,
|
| 1043 |
+
"learning_rate": 0.0001139704878966906,
|
| 1044 |
+
"loss": 0.8788,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 4.584615384615384,
|
| 1049 |
+
"grad_norm": 1.9300537109375,
|
| 1050 |
+
"learning_rate": 0.00011301268590933434,
|
| 1051 |
+
"loss": 0.8691,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 4.615384615384615,
|
| 1056 |
+
"grad_norm": 2.0728390216827393,
|
| 1057 |
+
"learning_rate": 0.0001120536680255323,
|
| 1058 |
+
"loss": 0.8018,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 4.6461538461538465,
|
| 1063 |
+
"grad_norm": 2.0605084896087646,
|
| 1064 |
+
"learning_rate": 0.00011109352385525783,
|
| 1065 |
+
"loss": 0.8635,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 4.676923076923077,
|
| 1070 |
+
"grad_norm": 1.8925397396087646,
|
| 1071 |
+
"learning_rate": 0.00011013234311372353,
|
| 1072 |
+
"loss": 0.7742,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 4.707692307692308,
|
| 1077 |
+
"grad_norm": 1.926761507987976,
|
| 1078 |
+
"learning_rate": 0.00010917021561299863,
|
| 1079 |
+
"loss": 0.8611,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 4.7384615384615385,
|
| 1084 |
+
"grad_norm": 1.9038374423980713,
|
| 1085 |
+
"learning_rate": 0.00010820723125361684,
|
| 1086 |
+
"loss": 0.8712,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 4.769230769230769,
|
| 1091 |
+
"grad_norm": 2.0429656505584717,
|
| 1092 |
+
"learning_rate": 0.00010724348001617625,
|
| 1093 |
+
"loss": 0.8328,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 4.8,
|
| 1098 |
+
"grad_norm": 2.0092427730560303,
|
| 1099 |
+
"learning_rate": 0.00010627905195293135,
|
| 1100 |
+
"loss": 0.8373,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 4.8307692307692305,
|
| 1105 |
+
"grad_norm": 1.9809768199920654,
|
| 1106 |
+
"learning_rate": 0.00010531403717937887,
|
| 1107 |
+
"loss": 0.8246,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 4.861538461538462,
|
| 1112 |
+
"grad_norm": 1.7576910257339478,
|
| 1113 |
+
"learning_rate": 0.00010434852586583736,
|
| 1114 |
+
"loss": 0.8373,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 4.892307692307693,
|
| 1119 |
+
"grad_norm": 1.7677350044250488,
|
| 1120 |
+
"learning_rate": 0.00010338260822902167,
|
| 1121 |
+
"loss": 0.7948,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 4.923076923076923,
|
| 1126 |
+
"grad_norm": 1.678804636001587,
|
| 1127 |
+
"learning_rate": 0.00010241637452361323,
|
| 1128 |
+
"loss": 0.8287,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 4.953846153846154,
|
| 1133 |
+
"grad_norm": 1.6783541440963745,
|
| 1134 |
+
"learning_rate": 0.00010144991503382674,
|
| 1135 |
+
"loss": 0.7825,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 4.984615384615385,
|
| 1140 |
+
"grad_norm": 1.9260199069976807,
|
| 1141 |
+
"learning_rate": 0.00010048332006497406,
|
| 1142 |
+
"loss": 0.8559,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 5.015384615384615,
|
| 1147 |
+
"grad_norm": 1.8397397994995117,
|
| 1148 |
+
"learning_rate": 9.9516679935026e-05,
|
| 1149 |
+
"loss": 0.8318,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 5.046153846153846,
|
| 1154 |
+
"grad_norm": 2.1345362663269043,
|
| 1155 |
+
"learning_rate": 9.855008496617327e-05,
|
| 1156 |
+
"loss": 0.8066,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 5.076923076923077,
|
| 1161 |
+
"grad_norm": 1.9365698099136353,
|
| 1162 |
+
"learning_rate": 9.75836254763868e-05,
|
| 1163 |
+
"loss": 0.7916,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 5.107692307692307,
|
| 1168 |
+
"grad_norm": 2.0976245403289795,
|
| 1169 |
+
"learning_rate": 9.661739177097836e-05,
|
| 1170 |
+
"loss": 0.7567,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 5.138461538461539,
|
| 1175 |
+
"grad_norm": 1.8643560409545898,
|
| 1176 |
+
"learning_rate": 9.565147413416266e-05,
|
| 1177 |
+
"loss": 0.7796,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 5.1692307692307695,
|
| 1182 |
+
"grad_norm": 1.9309316873550415,
|
| 1183 |
+
"learning_rate": 9.468596282062114e-05,
|
| 1184 |
+
"loss": 0.7884,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 5.2,
|
| 1189 |
+
"grad_norm": 1.8694071769714355,
|
| 1190 |
+
"learning_rate": 9.372094804706867e-05,
|
| 1191 |
+
"loss": 0.805,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 5.230769230769231,
|
| 1196 |
+
"grad_norm": 1.7506272792816162,
|
| 1197 |
+
"learning_rate": 9.275651998382377e-05,
|
| 1198 |
+
"loss": 0.7667,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 5.2615384615384615,
|
| 1203 |
+
"grad_norm": 2.0087368488311768,
|
| 1204 |
+
"learning_rate": 9.179276874638315e-05,
|
| 1205 |
+
"loss": 0.8121,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 5.292307692307692,
|
| 1210 |
+
"grad_norm": 1.9830255508422852,
|
| 1211 |
+
"learning_rate": 9.082978438700138e-05,
|
| 1212 |
+
"loss": 0.7741,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 5.323076923076923,
|
| 1217 |
+
"grad_norm": 2.119537591934204,
|
| 1218 |
+
"learning_rate": 8.986765688627652e-05,
|
| 1219 |
+
"loss": 0.7927,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 5.3538461538461535,
|
| 1224 |
+
"grad_norm": 1.9642573595046997,
|
| 1225 |
+
"learning_rate": 8.890647614474224e-05,
|
| 1226 |
+
"loss": 0.7995,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 5.384615384615385,
|
| 1231 |
+
"grad_norm": 2.063324213027954,
|
| 1232 |
+
"learning_rate": 8.79463319744677e-05,
|
| 1233 |
+
"loss": 0.8108,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 5.415384615384616,
|
| 1238 |
+
"grad_norm": 2.1207165718078613,
|
| 1239 |
+
"learning_rate": 8.698731409066568e-05,
|
| 1240 |
+
"loss": 0.7708,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 5.446153846153846,
|
| 1245 |
+
"grad_norm": 2.086155891418457,
|
| 1246 |
+
"learning_rate": 8.602951210330942e-05,
|
| 1247 |
+
"loss": 0.8002,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 5.476923076923077,
|
| 1252 |
+
"grad_norm": 2.0774879455566406,
|
| 1253 |
+
"learning_rate": 8.50730155087596e-05,
|
| 1254 |
+
"loss": 0.7897,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 5.507692307692308,
|
| 1259 |
+
"grad_norm": 2.012019157409668,
|
| 1260 |
+
"learning_rate": 8.411791368140196e-05,
|
| 1261 |
+
"loss": 0.8106,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 5.538461538461538,
|
| 1266 |
+
"grad_norm": 2.1534152030944824,
|
| 1267 |
+
"learning_rate": 8.316429586529615e-05,
|
| 1268 |
+
"loss": 0.8051,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 5.569230769230769,
|
| 1273 |
+
"grad_norm": 1.9364471435546875,
|
| 1274 |
+
"learning_rate": 8.221225116583678e-05,
|
| 1275 |
+
"loss": 0.8359,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 5.6,
|
| 1280 |
+
"grad_norm": 2.1292238235473633,
|
| 1281 |
+
"learning_rate": 8.126186854142752e-05,
|
| 1282 |
+
"loss": 0.824,
|
| 1283 |
+
"step": 1820
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 5.63076923076923,
|
| 1287 |
+
"grad_norm": 1.9687916040420532,
|
| 1288 |
+
"learning_rate": 8.0313236795169e-05,
|
| 1289 |
+
"loss": 0.8499,
|
| 1290 |
+
"step": 1830
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 5.661538461538462,
|
| 1294 |
+
"grad_norm": 1.9281556606292725,
|
| 1295 |
+
"learning_rate": 7.936644456656081e-05,
|
| 1296 |
+
"loss": 0.7799,
|
| 1297 |
+
"step": 1840
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 5.6923076923076925,
|
| 1301 |
+
"grad_norm": 2.123990297317505,
|
| 1302 |
+
"learning_rate": 7.84215803232194e-05,
|
| 1303 |
+
"loss": 0.8438,
|
| 1304 |
+
"step": 1850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 5.723076923076923,
|
| 1308 |
+
"grad_norm": 1.9118835926055908,
|
| 1309 |
+
"learning_rate": 7.747873235261157e-05,
|
| 1310 |
+
"loss": 0.7557,
|
| 1311 |
+
"step": 1860
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 5.753846153846154,
|
| 1315 |
+
"grad_norm": 1.90312922000885,
|
| 1316 |
+
"learning_rate": 7.6537988753805e-05,
|
| 1317 |
+
"loss": 0.7715,
|
| 1318 |
+
"step": 1870
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 5.7846153846153845,
|
| 1322 |
+
"grad_norm": 1.9930269718170166,
|
| 1323 |
+
"learning_rate": 7.559943742923626e-05,
|
| 1324 |
+
"loss": 0.8177,
|
| 1325 |
+
"step": 1880
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 5.815384615384615,
|
| 1329 |
+
"grad_norm": 2.398965358734131,
|
| 1330 |
+
"learning_rate": 7.466316607649738e-05,
|
| 1331 |
+
"loss": 0.8063,
|
| 1332 |
+
"step": 1890
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 5.846153846153846,
|
| 1336 |
+
"grad_norm": 2.19911527633667,
|
| 1337 |
+
"learning_rate": 7.372926218014131e-05,
|
| 1338 |
+
"loss": 0.7767,
|
| 1339 |
+
"step": 1900
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 5.876923076923077,
|
| 1343 |
+
"grad_norm": 2.059420585632324,
|
| 1344 |
+
"learning_rate": 7.279781300350758e-05,
|
| 1345 |
+
"loss": 0.7907,
|
| 1346 |
+
"step": 1910
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 5.907692307692308,
|
| 1350 |
+
"grad_norm": 2.0258114337921143,
|
| 1351 |
+
"learning_rate": 7.186890558056836e-05,
|
| 1352 |
+
"loss": 0.7923,
|
| 1353 |
+
"step": 1920
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 5.938461538461539,
|
| 1357 |
+
"grad_norm": 1.8502027988433838,
|
| 1358 |
+
"learning_rate": 7.094262670779612e-05,
|
| 1359 |
+
"loss": 0.8067,
|
| 1360 |
+
"step": 1930
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 5.969230769230769,
|
| 1364 |
+
"grad_norm": 2.0791757106781006,
|
| 1365 |
+
"learning_rate": 7.00190629360533e-05,
|
| 1366 |
+
"loss": 0.8026,
|
| 1367 |
+
"step": 1940
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 6.0,
|
| 1371 |
+
"grad_norm": 2.09051513671875,
|
| 1372 |
+
"learning_rate": 6.909830056250527e-05,
|
| 1373 |
+
"loss": 0.8339,
|
| 1374 |
+
"step": 1950
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 6.030769230769231,
|
| 1378 |
+
"grad_norm": 1.9199817180633545,
|
| 1379 |
+
"learning_rate": 6.81804256225567e-05,
|
| 1380 |
+
"loss": 0.752,
|
| 1381 |
+
"step": 1960
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 6.061538461538461,
|
| 1385 |
+
"grad_norm": 2.218315362930298,
|
| 1386 |
+
"learning_rate": 6.726552388181233e-05,
|
| 1387 |
+
"loss": 0.7989,
|
| 1388 |
+
"step": 1970
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 6.092307692307692,
|
| 1392 |
+
"grad_norm": 1.7573823928833008,
|
| 1393 |
+
"learning_rate": 6.63536808280633e-05,
|
| 1394 |
+
"loss": 0.735,
|
| 1395 |
+
"step": 1980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 6.123076923076923,
|
| 1399 |
+
"grad_norm": 1.8894169330596924,
|
| 1400 |
+
"learning_rate": 6.544498166329913e-05,
|
| 1401 |
+
"loss": 0.7633,
|
| 1402 |
+
"step": 1990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 6.153846153846154,
|
| 1406 |
+
"grad_norm": 2.069439649581909,
|
| 1407 |
+
"learning_rate": 6.453951129574644e-05,
|
| 1408 |
+
"loss": 0.7882,
|
| 1409 |
+
"step": 2000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 6.184615384615385,
|
| 1413 |
+
"grad_norm": 2.2288105487823486,
|
| 1414 |
+
"learning_rate": 6.36373543319353e-05,
|
| 1415 |
+
"loss": 0.773,
|
| 1416 |
+
"step": 2010
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 6.2153846153846155,
|
| 1420 |
+
"grad_norm": 2.023712396621704,
|
| 1421 |
+
"learning_rate": 6.273859506879365e-05,
|
| 1422 |
+
"loss": 0.753,
|
| 1423 |
+
"step": 2020
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 6.246153846153846,
|
| 1427 |
+
"grad_norm": 2.0464038848876953,
|
| 1428 |
+
"learning_rate": 6.18433174857705e-05,
|
| 1429 |
+
"loss": 0.7616,
|
| 1430 |
+
"step": 2030
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 6.276923076923077,
|
| 1434 |
+
"grad_norm": 2.011988878250122,
|
| 1435 |
+
"learning_rate": 6.095160523698913e-05,
|
| 1436 |
+
"loss": 0.7611,
|
| 1437 |
+
"step": 2040
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 6.3076923076923075,
|
| 1441 |
+
"grad_norm": 2.3154494762420654,
|
| 1442 |
+
"learning_rate": 6.006354164343046e-05,
|
| 1443 |
+
"loss": 0.7943,
|
| 1444 |
+
"step": 2050
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 6.338461538461538,
|
| 1448 |
+
"grad_norm": 2.325244903564453,
|
| 1449 |
+
"learning_rate": 5.917920968514752e-05,
|
| 1450 |
+
"loss": 0.7529,
|
| 1451 |
+
"step": 2060
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 6.36923076923077,
|
| 1455 |
+
"grad_norm": 1.9565290212631226,
|
| 1456 |
+
"learning_rate": 5.829869199351188e-05,
|
| 1457 |
+
"loss": 0.7952,
|
| 1458 |
+
"step": 2070
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 6.4,
|
| 1462 |
+
"grad_norm": 1.8974802494049072,
|
| 1463 |
+
"learning_rate": 5.7422070843492734e-05,
|
| 1464 |
+
"loss": 0.7621,
|
| 1465 |
+
"step": 2080
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 6.430769230769231,
|
| 1469 |
+
"grad_norm": 2.1027376651763916,
|
| 1470 |
+
"learning_rate": 5.654942814596902e-05,
|
| 1471 |
+
"loss": 0.758,
|
| 1472 |
+
"step": 2090
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 6.461538461538462,
|
| 1476 |
+
"grad_norm": 1.9659016132354736,
|
| 1477 |
+
"learning_rate": 5.568084544007588e-05,
|
| 1478 |
+
"loss": 0.8167,
|
| 1479 |
+
"step": 2100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 6.492307692307692,
|
| 1483 |
+
"grad_norm": 2.048273801803589,
|
| 1484 |
+
"learning_rate": 5.481640388558551e-05,
|
| 1485 |
+
"loss": 0.7517,
|
| 1486 |
+
"step": 2110
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 6.523076923076923,
|
| 1490 |
+
"grad_norm": 2.1136772632598877,
|
| 1491 |
+
"learning_rate": 5.395618425532389e-05,
|
| 1492 |
+
"loss": 0.7406,
|
| 1493 |
+
"step": 2120
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 6.553846153846154,
|
| 1497 |
+
"grad_norm": 2.157108783721924,
|
| 1498 |
+
"learning_rate": 5.3100266927623156e-05,
|
| 1499 |
+
"loss": 0.8148,
|
| 1500 |
+
"step": 2130
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 6.584615384615384,
|
| 1504 |
+
"grad_norm": 2.106790781021118,
|
| 1505 |
+
"learning_rate": 5.2248731878811365e-05,
|
| 1506 |
+
"loss": 0.7784,
|
| 1507 |
+
"step": 2140
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 6.615384615384615,
|
| 1511 |
+
"grad_norm": 1.9885175228118896,
|
| 1512 |
+
"learning_rate": 5.14016586757394e-05,
|
| 1513 |
+
"loss": 0.787,
|
| 1514 |
+
"step": 2150
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 6.6461538461538465,
|
| 1518 |
+
"grad_norm": 2.2793591022491455,
|
| 1519 |
+
"learning_rate": 5.055912646834635e-05,
|
| 1520 |
+
"loss": 0.7643,
|
| 1521 |
+
"step": 2160
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 6.676923076923077,
|
| 1525 |
+
"grad_norm": 1.9444079399108887,
|
| 1526 |
+
"learning_rate": 4.972121398226371e-05,
|
| 1527 |
+
"loss": 0.7571,
|
| 1528 |
+
"step": 2170
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 6.707692307692308,
|
| 1532 |
+
"grad_norm": 2.114750385284424,
|
| 1533 |
+
"learning_rate": 4.888799951145948e-05,
|
| 1534 |
+
"loss": 0.7698,
|
| 1535 |
+
"step": 2180
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 6.7384615384615385,
|
| 1539 |
+
"grad_norm": 1.8953299522399902,
|
| 1540 |
+
"learning_rate": 4.805956091092227e-05,
|
| 1541 |
+
"loss": 0.7318,
|
| 1542 |
+
"step": 2190
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 6.769230769230769,
|
| 1546 |
+
"grad_norm": 2.02268385887146,
|
| 1547 |
+
"learning_rate": 4.723597558938672e-05,
|
| 1548 |
+
"loss": 0.7826,
|
| 1549 |
+
"step": 2200
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 6.8,
|
| 1553 |
+
"grad_norm": 2.0683560371398926,
|
| 1554 |
+
"learning_rate": 4.6417320502100316e-05,
|
| 1555 |
+
"loss": 0.7363,
|
| 1556 |
+
"step": 2210
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 6.8307692307692305,
|
| 1560 |
+
"grad_norm": 2.056417465209961,
|
| 1561 |
+
"learning_rate": 4.5603672143632944e-05,
|
| 1562 |
+
"loss": 0.7779,
|
| 1563 |
+
"step": 2220
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 6.861538461538462,
|
| 1567 |
+
"grad_norm": 2.0059967041015625,
|
| 1568 |
+
"learning_rate": 4.479510654072909e-05,
|
| 1569 |
+
"loss": 0.8147,
|
| 1570 |
+
"step": 2230
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 6.892307692307693,
|
| 1574 |
+
"grad_norm": 1.9279494285583496,
|
| 1575 |
+
"learning_rate": 4.399169924520403e-05,
|
| 1576 |
+
"loss": 0.7647,
|
| 1577 |
+
"step": 2240
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 6.923076923076923,
|
| 1581 |
+
"grad_norm": 2.5176281929016113,
|
| 1582 |
+
"learning_rate": 4.3193525326884435e-05,
|
| 1583 |
+
"loss": 0.8265,
|
| 1584 |
+
"step": 2250
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 6.953846153846154,
|
| 1588 |
+
"grad_norm": 1.9771342277526855,
|
| 1589 |
+
"learning_rate": 4.240065936659374e-05,
|
| 1590 |
+
"loss": 0.7755,
|
| 1591 |
+
"step": 2260
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 6.984615384615385,
|
| 1595 |
+
"grad_norm": 2.1674654483795166,
|
| 1596 |
+
"learning_rate": 4.161317544918345e-05,
|
| 1597 |
+
"loss": 0.7682,
|
| 1598 |
+
"step": 2270
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 7.015384615384615,
|
| 1602 |
+
"grad_norm": 1.9835222959518433,
|
| 1603 |
+
"learning_rate": 4.0831147156610684e-05,
|
| 1604 |
+
"loss": 0.7977,
|
| 1605 |
+
"step": 2280
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 7.046153846153846,
|
| 1609 |
+
"grad_norm": 2.224958658218384,
|
| 1610 |
+
"learning_rate": 4.005464756106262e-05,
|
| 1611 |
+
"loss": 0.7346,
|
| 1612 |
+
"step": 2290
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 7.076923076923077,
|
| 1616 |
+
"grad_norm": 2.0972986221313477,
|
| 1617 |
+
"learning_rate": 3.9283749218128885e-05,
|
| 1618 |
+
"loss": 0.7259,
|
| 1619 |
+
"step": 2300
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 7.107692307692307,
|
| 1623 |
+
"grad_norm": 2.0841434001922607,
|
| 1624 |
+
"learning_rate": 3.851852416002187e-05,
|
| 1625 |
+
"loss": 0.7083,
|
| 1626 |
+
"step": 2310
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 7.138461538461539,
|
| 1630 |
+
"grad_norm": 2.0690999031066895,
|
| 1631 |
+
"learning_rate": 3.775904388884618e-05,
|
| 1632 |
+
"loss": 0.7639,
|
| 1633 |
+
"step": 2320
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 7.1692307692307695,
|
| 1637 |
+
"grad_norm": 2.232002019882202,
|
| 1638 |
+
"learning_rate": 3.7005379369917325e-05,
|
| 1639 |
+
"loss": 0.7578,
|
| 1640 |
+
"step": 2330
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 7.2,
|
| 1644 |
+
"grad_norm": 2.051391363143921,
|
| 1645 |
+
"learning_rate": 3.6257601025131026e-05,
|
| 1646 |
+
"loss": 0.7454,
|
| 1647 |
+
"step": 2340
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 7.230769230769231,
|
| 1651 |
+
"grad_norm": 2.056314468383789,
|
| 1652 |
+
"learning_rate": 3.5515778726382966e-05,
|
| 1653 |
+
"loss": 0.7374,
|
| 1654 |
+
"step": 2350
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 7.2615384615384615,
|
| 1658 |
+
"grad_norm": 2.0361640453338623,
|
| 1659 |
+
"learning_rate": 3.477998178903982e-05,
|
| 1660 |
+
"loss": 0.7634,
|
| 1661 |
+
"step": 2360
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 7.292307692307692,
|
| 1665 |
+
"grad_norm": 2.221268653869629,
|
| 1666 |
+
"learning_rate": 3.4050278965462764e-05,
|
| 1667 |
+
"loss": 0.7884,
|
| 1668 |
+
"step": 2370
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 7.323076923076923,
|
| 1672 |
+
"grad_norm": 2.4283814430236816,
|
| 1673 |
+
"learning_rate": 3.3326738438583114e-05,
|
| 1674 |
+
"loss": 0.7827,
|
| 1675 |
+
"step": 2380
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 7.3538461538461535,
|
| 1679 |
+
"grad_norm": 2.1344385147094727,
|
| 1680 |
+
"learning_rate": 3.2609427815531426e-05,
|
| 1681 |
+
"loss": 0.7281,
|
| 1682 |
+
"step": 2390
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 7.384615384615385,
|
| 1686 |
+
"grad_norm": 2.1491506099700928,
|
| 1687 |
+
"learning_rate": 3.1898414121320276e-05,
|
| 1688 |
+
"loss": 0.7854,
|
| 1689 |
+
"step": 2400
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 7.415384615384616,
|
| 1693 |
+
"grad_norm": 2.2601113319396973,
|
| 1694 |
+
"learning_rate": 3.11937637925816e-05,
|
| 1695 |
+
"loss": 0.7765,
|
| 1696 |
+
"step": 2410
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 7.446153846153846,
|
| 1700 |
+
"grad_norm": 1.77631676197052,
|
| 1701 |
+
"learning_rate": 3.0495542671358746e-05,
|
| 1702 |
+
"loss": 0.7188,
|
| 1703 |
+
"step": 2420
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 7.476923076923077,
|
| 1707 |
+
"grad_norm": 1.9240316152572632,
|
| 1708 |
+
"learning_rate": 2.9803815998954332e-05,
|
| 1709 |
+
"loss": 0.7751,
|
| 1710 |
+
"step": 2430
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 7.507692307692308,
|
| 1714 |
+
"grad_norm": 1.9639230966567993,
|
| 1715 |
+
"learning_rate": 2.9118648409834205e-05,
|
| 1716 |
+
"loss": 0.7346,
|
| 1717 |
+
"step": 2440
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 7.538461538461538,
|
| 1721 |
+
"grad_norm": 1.9350874423980713,
|
| 1722 |
+
"learning_rate": 2.84401039255879e-05,
|
| 1723 |
+
"loss": 0.7412,
|
| 1724 |
+
"step": 2450
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 7.569230769230769,
|
| 1728 |
+
"grad_norm": 2.0040721893310547,
|
| 1729 |
+
"learning_rate": 2.7768245948946612e-05,
|
| 1730 |
+
"loss": 0.7466,
|
| 1731 |
+
"step": 2460
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 7.6,
|
| 1735 |
+
"grad_norm": 1.894864559173584,
|
| 1736 |
+
"learning_rate": 2.7103137257858868e-05,
|
| 1737 |
+
"loss": 0.7722,
|
| 1738 |
+
"step": 2470
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 7.63076923076923,
|
| 1742 |
+
"grad_norm": 2.5914828777313232,
|
| 1743 |
+
"learning_rate": 2.6444839999624494e-05,
|
| 1744 |
+
"loss": 0.7724,
|
| 1745 |
+
"step": 2480
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 7.661538461538462,
|
| 1749 |
+
"grad_norm": 2.3115148544311523,
|
| 1750 |
+
"learning_rate": 2.5793415685087797e-05,
|
| 1751 |
+
"loss": 0.8039,
|
| 1752 |
+
"step": 2490
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 7.6923076923076925,
|
| 1756 |
+
"grad_norm": 1.9655823707580566,
|
| 1757 |
+
"learning_rate": 2.514892518288988e-05,
|
| 1758 |
+
"loss": 0.7421,
|
| 1759 |
+
"step": 2500
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 7.723076923076923,
|
| 1763 |
+
"grad_norm": 2.0706140995025635,
|
| 1764 |
+
"learning_rate": 2.4511428713781238e-05,
|
| 1765 |
+
"loss": 0.7631,
|
| 1766 |
+
"step": 2510
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 7.753846153846154,
|
| 1770 |
+
"grad_norm": 1.910049557685852,
|
| 1771 |
+
"learning_rate": 2.3880985844994674e-05,
|
| 1772 |
+
"loss": 0.7399,
|
| 1773 |
+
"step": 2520
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 7.7846153846153845,
|
| 1777 |
+
"grad_norm": 2.1228690147399902,
|
| 1778 |
+
"learning_rate": 2.3257655484679374e-05,
|
| 1779 |
+
"loss": 0.7572,
|
| 1780 |
+
"step": 2530
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 7.815384615384615,
|
| 1784 |
+
"grad_norm": 2.168492555618286,
|
| 1785 |
+
"learning_rate": 2.2641495876396713e-05,
|
| 1786 |
+
"loss": 0.7703,
|
| 1787 |
+
"step": 2540
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 7.846153846153846,
|
| 1791 |
+
"grad_norm": 1.8387569189071655,
|
| 1792 |
+
"learning_rate": 2.2032564593677774e-05,
|
| 1793 |
+
"loss": 0.7217,
|
| 1794 |
+
"step": 2550
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 7.876923076923077,
|
| 1798 |
+
"grad_norm": 2.0391128063201904,
|
| 1799 |
+
"learning_rate": 2.1430918534643996e-05,
|
| 1800 |
+
"loss": 0.752,
|
| 1801 |
+
"step": 2560
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 7.907692307692308,
|
| 1805 |
+
"grad_norm": 2.573230743408203,
|
| 1806 |
+
"learning_rate": 2.0836613916690428e-05,
|
| 1807 |
+
"loss": 0.7175,
|
| 1808 |
+
"step": 2570
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 7.938461538461539,
|
| 1812 |
+
"grad_norm": 2.0587637424468994,
|
| 1813 |
+
"learning_rate": 2.024970627123295e-05,
|
| 1814 |
+
"loss": 0.8003,
|
| 1815 |
+
"step": 2580
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 7.969230769230769,
|
| 1819 |
+
"grad_norm": 2.126275062561035,
|
| 1820 |
+
"learning_rate": 1.967025043851939e-05,
|
| 1821 |
+
"loss": 0.7484,
|
| 1822 |
+
"step": 2590
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 8.0,
|
| 1826 |
+
"grad_norm": 2.07289719581604,
|
| 1827 |
+
"learning_rate": 1.9098300562505266e-05,
|
| 1828 |
+
"loss": 0.7706,
|
| 1829 |
+
"step": 2600
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 8.03076923076923,
|
| 1833 |
+
"grad_norm": 2.0858993530273438,
|
| 1834 |
+
"learning_rate": 1.8533910085794713e-05,
|
| 1835 |
+
"loss": 0.7456,
|
| 1836 |
+
"step": 2610
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 8.061538461538461,
|
| 1840 |
+
"grad_norm": 2.192556619644165,
|
| 1841 |
+
"learning_rate": 1.7977131744646724e-05,
|
| 1842 |
+
"loss": 0.7149,
|
| 1843 |
+
"step": 2620
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 8.092307692307692,
|
| 1847 |
+
"grad_norm": 1.8930575847625732,
|
| 1848 |
+
"learning_rate": 1.7428017564047594e-05,
|
| 1849 |
+
"loss": 0.7205,
|
| 1850 |
+
"step": 2630
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 8.123076923076923,
|
| 1854 |
+
"grad_norm": 2.337496042251587,
|
| 1855 |
+
"learning_rate": 1.6886618852849724e-05,
|
| 1856 |
+
"loss": 0.7782,
|
| 1857 |
+
"step": 2640
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 8.153846153846153,
|
| 1861 |
+
"grad_norm": 2.018669366836548,
|
| 1862 |
+
"learning_rate": 1.6352986198977325e-05,
|
| 1863 |
+
"loss": 0.7889,
|
| 1864 |
+
"step": 2650
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 8.184615384615384,
|
| 1868 |
+
"grad_norm": 2.315338134765625,
|
| 1869 |
+
"learning_rate": 1.5827169464699576e-05,
|
| 1870 |
+
"loss": 0.7248,
|
| 1871 |
+
"step": 2660
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 8.215384615384615,
|
| 1875 |
+
"grad_norm": 2.2808115482330322,
|
| 1876 |
+
"learning_rate": 1.530921778197142e-05,
|
| 1877 |
+
"loss": 0.7778,
|
| 1878 |
+
"step": 2670
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 8.246153846153845,
|
| 1882 |
+
"grad_norm": 2.2129008769989014,
|
| 1883 |
+
"learning_rate": 1.4799179547842822e-05,
|
| 1884 |
+
"loss": 0.7097,
|
| 1885 |
+
"step": 2680
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 8.276923076923078,
|
| 1889 |
+
"grad_norm": 2.071539878845215,
|
| 1890 |
+
"learning_rate": 1.429710241993656e-05,
|
| 1891 |
+
"loss": 0.7472,
|
| 1892 |
+
"step": 2690
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 8.307692307692308,
|
| 1896 |
+
"grad_norm": 2.134145736694336,
|
| 1897 |
+
"learning_rate": 1.3803033311995072e-05,
|
| 1898 |
+
"loss": 0.7591,
|
| 1899 |
+
"step": 2700
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 8.338461538461539,
|
| 1903 |
+
"grad_norm": 2.066234827041626,
|
| 1904 |
+
"learning_rate": 1.3317018389496927e-05,
|
| 1905 |
+
"loss": 0.7537,
|
| 1906 |
+
"step": 2710
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 8.36923076923077,
|
| 1910 |
+
"grad_norm": 2.0239179134368896,
|
| 1911 |
+
"learning_rate": 1.2839103065343083e-05,
|
| 1912 |
+
"loss": 0.8152,
|
| 1913 |
+
"step": 2720
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 8.4,
|
| 1917 |
+
"grad_norm": 2.076111316680908,
|
| 1918 |
+
"learning_rate": 1.2369331995613665e-05,
|
| 1919 |
+
"loss": 0.7242,
|
| 1920 |
+
"step": 2730
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 8.430769230769231,
|
| 1924 |
+
"grad_norm": 2.0224554538726807,
|
| 1925 |
+
"learning_rate": 1.1907749075395147e-05,
|
| 1926 |
+
"loss": 0.7407,
|
| 1927 |
+
"step": 2740
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 8.461538461538462,
|
| 1931 |
+
"grad_norm": 2.344095230102539,
|
| 1932 |
+
"learning_rate": 1.1454397434679021e-05,
|
| 1933 |
+
"loss": 0.8089,
|
| 1934 |
+
"step": 2750
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 8.492307692307692,
|
| 1938 |
+
"grad_norm": 2.1623709201812744,
|
| 1939 |
+
"learning_rate": 1.1009319434331622e-05,
|
| 1940 |
+
"loss": 0.7209,
|
| 1941 |
+
"step": 2760
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 8.523076923076923,
|
| 1945 |
+
"grad_norm": 2.238278865814209,
|
| 1946 |
+
"learning_rate": 1.0572556662136035e-05,
|
| 1947 |
+
"loss": 0.6911,
|
| 1948 |
+
"step": 2770
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 8.553846153846154,
|
| 1952 |
+
"grad_norm": 2.4050676822662354,
|
| 1953 |
+
"learning_rate": 1.014414992890611e-05,
|
| 1954 |
+
"loss": 0.7138,
|
| 1955 |
+
"step": 2780
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 8.584615384615384,
|
| 1959 |
+
"grad_norm": 2.0228419303894043,
|
| 1960 |
+
"learning_rate": 9.724139264673116e-06,
|
| 1961 |
+
"loss": 0.7452,
|
| 1962 |
+
"step": 2790
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 8.615384615384615,
|
| 1966 |
+
"grad_norm": 2.2917470932006836,
|
| 1967 |
+
"learning_rate": 9.31256391494546e-06,
|
| 1968 |
+
"loss": 0.6921,
|
| 1969 |
+
"step": 2800
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 8.646153846153846,
|
| 1973 |
+
"grad_norm": 1.8075125217437744,
|
| 1974 |
+
"learning_rate": 8.909462337041507e-06,
|
| 1975 |
+
"loss": 0.734,
|
| 1976 |
+
"step": 2810
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 8.676923076923076,
|
| 1980 |
+
"grad_norm": 2.01468825340271,
|
| 1981 |
+
"learning_rate": 8.514872196496183e-06,
|
| 1982 |
+
"loss": 0.7335,
|
| 1983 |
+
"step": 2820
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 8.707692307692307,
|
| 1987 |
+
"grad_norm": 2.298112630844116,
|
| 1988 |
+
"learning_rate": 8.128830363541574e-06,
|
| 1989 |
+
"loss": 0.7094,
|
| 1990 |
+
"step": 2830
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 8.73846153846154,
|
| 1994 |
+
"grad_norm": 2.1774020195007324,
|
| 1995 |
+
"learning_rate": 7.751372909661769e-06,
|
| 1996 |
+
"loss": 0.8051,
|
| 1997 |
+
"step": 2840
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 8.76923076923077,
|
| 2001 |
+
"grad_norm": 2.1409170627593994,
|
| 2002 |
+
"learning_rate": 7.382535104222366e-06,
|
| 2003 |
+
"loss": 0.7679,
|
| 2004 |
+
"step": 2850
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 8.8,
|
| 2008 |
+
"grad_norm": 2.082650899887085,
|
| 2009 |
+
"learning_rate": 7.022351411174866e-06,
|
| 2010 |
+
"loss": 0.7277,
|
| 2011 |
+
"step": 2860
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 8.830769230769231,
|
| 2015 |
+
"grad_norm": 2.0364997386932373,
|
| 2016 |
+
"learning_rate": 6.670855485836525e-06,
|
| 2017 |
+
"loss": 0.7121,
|
| 2018 |
+
"step": 2870
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 8.861538461538462,
|
| 2022 |
+
"grad_norm": 2.206631898880005,
|
| 2023 |
+
"learning_rate": 6.32808017174551e-06,
|
| 2024 |
+
"loss": 0.741,
|
| 2025 |
+
"step": 2880
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 8.892307692307693,
|
| 2029 |
+
"grad_norm": 2.1782455444335938,
|
| 2030 |
+
"learning_rate": 5.994057497592031e-06,
|
| 2031 |
+
"loss": 0.7468,
|
| 2032 |
+
"step": 2890
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 8.923076923076923,
|
| 2036 |
+
"grad_norm": 2.374114751815796,
|
| 2037 |
+
"learning_rate": 5.668818674225685e-06,
|
| 2038 |
+
"loss": 0.7532,
|
| 2039 |
+
"step": 2900
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 8.953846153846154,
|
| 2043 |
+
"grad_norm": 2.0386927127838135,
|
| 2044 |
+
"learning_rate": 5.3523940917390215e-06,
|
| 2045 |
+
"loss": 0.7524,
|
| 2046 |
+
"step": 2910
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 8.984615384615385,
|
| 2050 |
+
"grad_norm": 1.9349490404129028,
|
| 2051 |
+
"learning_rate": 5.0448133166279944e-06,
|
| 2052 |
+
"loss": 0.7254,
|
| 2053 |
+
"step": 2920
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 9.015384615384615,
|
| 2057 |
+
"grad_norm": 2.218142032623291,
|
| 2058 |
+
"learning_rate": 4.746105089029229e-06,
|
| 2059 |
+
"loss": 0.744,
|
| 2060 |
+
"step": 2930
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 9.046153846153846,
|
| 2064 |
+
"grad_norm": 2.208949565887451,
|
| 2065 |
+
"learning_rate": 4.4562973200346416e-06,
|
| 2066 |
+
"loss": 0.7338,
|
| 2067 |
+
"step": 2940
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 9.076923076923077,
|
| 2071 |
+
"grad_norm": 1.8909201622009277,
|
| 2072 |
+
"learning_rate": 4.175417089083378e-06,
|
| 2073 |
+
"loss": 0.7232,
|
| 2074 |
+
"step": 2950
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 9.107692307692307,
|
| 2078 |
+
"grad_norm": 2.163405179977417,
|
| 2079 |
+
"learning_rate": 3.903490641431573e-06,
|
| 2080 |
+
"loss": 0.7349,
|
| 2081 |
+
"step": 2960
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 9.138461538461538,
|
| 2085 |
+
"grad_norm": 2.3600995540618896,
|
| 2086 |
+
"learning_rate": 3.6405433856999684e-06,
|
| 2087 |
+
"loss": 0.7678,
|
| 2088 |
+
"step": 2970
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 9.169230769230769,
|
| 2092 |
+
"grad_norm": 2.188930034637451,
|
| 2093 |
+
"learning_rate": 3.3865998914997643e-06,
|
| 2094 |
+
"loss": 0.7418,
|
| 2095 |
+
"step": 2980
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 9.2,
|
| 2099 |
+
"grad_norm": 2.2357707023620605,
|
| 2100 |
+
"learning_rate": 3.1416838871368924e-06,
|
| 2101 |
+
"loss": 0.7426,
|
| 2102 |
+
"step": 2990
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 9.23076923076923,
|
| 2106 |
+
"grad_norm": 1.9671388864517212,
|
| 2107 |
+
"learning_rate": 2.905818257394799e-06,
|
| 2108 |
+
"loss": 0.7298,
|
| 2109 |
+
"step": 3000
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 9.261538461538462,
|
| 2113 |
+
"grad_norm": 2.478001594543457,
|
| 2114 |
+
"learning_rate": 2.679025041396155e-06,
|
| 2115 |
+
"loss": 0.7185,
|
| 2116 |
+
"step": 3010
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 9.292307692307693,
|
| 2120 |
+
"grad_norm": 2.290837287902832,
|
| 2121 |
+
"learning_rate": 2.461325430543482e-06,
|
| 2122 |
+
"loss": 0.7444,
|
| 2123 |
+
"step": 3020
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 9.323076923076924,
|
| 2127 |
+
"grad_norm": 2.3114941120147705,
|
| 2128 |
+
"learning_rate": 2.2527397665391027e-06,
|
| 2129 |
+
"loss": 0.7101,
|
| 2130 |
+
"step": 3030
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 9.353846153846154,
|
| 2134 |
+
"grad_norm": 1.9803047180175781,
|
| 2135 |
+
"learning_rate": 2.053287539484405e-06,
|
| 2136 |
+
"loss": 0.6882,
|
| 2137 |
+
"step": 3040
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 9.384615384615385,
|
| 2141 |
+
"grad_norm": 2.2022945880889893,
|
| 2142 |
+
"learning_rate": 1.8629873860586566e-06,
|
| 2143 |
+
"loss": 0.6975,
|
| 2144 |
+
"step": 3050
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 9.415384615384616,
|
| 2148 |
+
"grad_norm": 1.9524860382080078,
|
| 2149 |
+
"learning_rate": 1.6818570877776718e-06,
|
| 2150 |
+
"loss": 0.6846,
|
| 2151 |
+
"step": 3060
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 9.446153846153846,
|
| 2155 |
+
"grad_norm": 2.2137529850006104,
|
| 2156 |
+
"learning_rate": 1.5099135693322774e-06,
|
| 2157 |
+
"loss": 0.7273,
|
| 2158 |
+
"step": 3070
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 9.476923076923077,
|
| 2162 |
+
"grad_norm": 2.1356773376464844,
|
| 2163 |
+
"learning_rate": 1.3471728970068987e-06,
|
| 2164 |
+
"loss": 0.7458,
|
| 2165 |
+
"step": 3080
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 9.507692307692308,
|
| 2169 |
+
"grad_norm": 2.098583221435547,
|
| 2170 |
+
"learning_rate": 1.1936502771783486e-06,
|
| 2171 |
+
"loss": 0.7143,
|
| 2172 |
+
"step": 3090
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 9.538461538461538,
|
| 2176 |
+
"grad_norm": 2.1301488876342773,
|
| 2177 |
+
"learning_rate": 1.0493600548948878e-06,
|
| 2178 |
+
"loss": 0.7496,
|
| 2179 |
+
"step": 3100
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 9.569230769230769,
|
| 2183 |
+
"grad_norm": 2.2320845127105713,
|
| 2184 |
+
"learning_rate": 9.143157125359514e-07,
|
| 2185 |
+
"loss": 0.7591,
|
| 2186 |
+
"step": 3110
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 9.6,
|
| 2190 |
+
"grad_norm": 2.2147562503814697,
|
| 2191 |
+
"learning_rate": 7.885298685522235e-07,
|
| 2192 |
+
"loss": 0.7303,
|
| 2193 |
+
"step": 3120
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 9.63076923076923,
|
| 2197 |
+
"grad_norm": 2.2397658824920654,
|
| 2198 |
+
"learning_rate": 6.720142762867032e-07,
|
| 2199 |
+
"loss": 0.7555,
|
| 2200 |
+
"step": 3130
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 9.661538461538461,
|
| 2204 |
+
"grad_norm": 2.177968740463257,
|
| 2205 |
+
"learning_rate": 5.647798228764156e-07,
|
| 2206 |
+
"loss": 0.7479,
|
| 2207 |
+
"step": 3140
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 9.692307692307692,
|
| 2211 |
+
"grad_norm": 2.243819236755371,
|
| 2212 |
+
"learning_rate": 4.668365282351372e-07,
|
| 2213 |
+
"loss": 0.7534,
|
| 2214 |
+
"step": 3150
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 9.723076923076922,
|
| 2218 |
+
"grad_norm": 2.2024483680725098,
|
| 2219 |
+
"learning_rate": 3.781935441171336e-07,
|
| 2220 |
+
"loss": 0.7508,
|
| 2221 |
+
"step": 3160
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 9.753846153846155,
|
| 2225 |
+
"grad_norm": 2.2095537185668945,
|
| 2226 |
+
"learning_rate": 2.988591532620322e-07,
|
| 2227 |
+
"loss": 0.7463,
|
| 2228 |
+
"step": 3170
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 9.784615384615385,
|
| 2232 |
+
"grad_norm": 2.1779325008392334,
|
| 2233 |
+
"learning_rate": 2.288407686208971e-07,
|
| 2234 |
+
"loss": 0.7814,
|
| 2235 |
+
"step": 3180
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 9.815384615384616,
|
| 2239 |
+
"grad_norm": 2.161087989807129,
|
| 2240 |
+
"learning_rate": 1.68144932663572e-07,
|
| 2241 |
+
"loss": 0.7342,
|
| 2242 |
+
"step": 3190
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 9.846153846153847,
|
| 2246 |
+
"grad_norm": 2.0373594760894775,
|
| 2247 |
+
"learning_rate": 1.1677731676733584e-07,
|
| 2248 |
+
"loss": 0.7054,
|
| 2249 |
+
"step": 3200
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 9.876923076923077,
|
| 2253 |
+
"grad_norm": 2.077526330947876,
|
| 2254 |
+
"learning_rate": 7.474272068698218e-08,
|
| 2255 |
+
"loss": 0.7355,
|
| 2256 |
+
"step": 3210
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 9.907692307692308,
|
| 2260 |
+
"grad_norm": 2.0211634635925293,
|
| 2261 |
+
"learning_rate": 4.2045072106333684e-08,
|
| 2262 |
+
"loss": 0.763,
|
| 2263 |
+
"step": 3220
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 9.938461538461539,
|
| 2267 |
+
"grad_norm": 2.205732822418213,
|
| 2268 |
+
"learning_rate": 1.8687426271246645e-08,
|
| 2269 |
+
"loss": 0.7732,
|
| 2270 |
+
"step": 3230
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 9.96923076923077,
|
| 2274 |
+
"grad_norm": 2.1264724731445312,
|
| 2275 |
+
"learning_rate": 4.6719657041283114e-09,
|
| 2276 |
+
"loss": 0.7328,
|
| 2277 |
+
"step": 3240
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 10.0,
|
| 2281 |
+
"grad_norm": 2.179966449737549,
|
| 2282 |
+
"learning_rate": 0.0,
|
| 2283 |
+
"loss": 0.7768,
|
| 2284 |
+
"step": 3250
|
| 2285 |
+
}
|
| 2286 |
+
],
|
| 2287 |
+
"logging_steps": 10,
|
| 2288 |
+
"max_steps": 3250,
|
| 2289 |
+
"num_input_tokens_seen": 0,
|
| 2290 |
+
"num_train_epochs": 10,
|
| 2291 |
+
"save_steps": 500,
|
| 2292 |
+
"stateful_callbacks": {
|
| 2293 |
+
"TrainerControl": {
|
| 2294 |
+
"args": {
|
| 2295 |
+
"should_epoch_stop": false,
|
| 2296 |
+
"should_evaluate": false,
|
| 2297 |
+
"should_log": false,
|
| 2298 |
+
"should_save": true,
|
| 2299 |
+
"should_training_stop": true
|
| 2300 |
+
},
|
| 2301 |
+
"attributes": {}
|
| 2302 |
+
}
|
| 2303 |
+
},
|
| 2304 |
+
"total_flos": 2.2869986411741184e+16,
|
| 2305 |
+
"train_batch_size": 4,
|
| 2306 |
+
"trial_name": null,
|
| 2307 |
+
"trial_params": null
|
| 2308 |
+
}
|