Commit
•
8823408
1
Parent(s):
83afee1
Upload trainer_state.json
Browse files- trainer_state.json +2606 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,2606 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.9293885827064514,
|
| 3 |
+
"best_model_checkpoint": "C:\\Users\\User\\Downloads\\cache2\\checkpoint-7320000",
|
| 4 |
+
"epoch": 49.96556427327113,
|
| 5 |
+
"global_step": 7400000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.27,
|
| 12 |
+
"learning_rate": 5.39938690902216e-07,
|
| 13 |
+
"loss": 1.0744,
|
| 14 |
+
"step": 40000
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.27,
|
| 18 |
+
"eval_loss": 1.0361614227294922,
|
| 19 |
+
"eval_runtime": 731.8142,
|
| 20 |
+
"eval_samples_per_second": 42.359,
|
| 21 |
+
"eval_steps_per_second": 7.061,
|
| 22 |
+
"step": 40000
|
| 23 |
+
},
|
| 24 |
+
{
|
| 25 |
+
"epoch": 0.54,
|
| 26 |
+
"learning_rate": 1.0799178944241131e-06,
|
| 27 |
+
"loss": 1.0786,
|
| 28 |
+
"step": 80000
|
| 29 |
+
},
|
| 30 |
+
{
|
| 31 |
+
"epoch": 0.54,
|
| 32 |
+
"eval_loss": 1.0365289449691772,
|
| 33 |
+
"eval_runtime": 731.8318,
|
| 34 |
+
"eval_samples_per_second": 42.358,
|
| 35 |
+
"eval_steps_per_second": 7.06,
|
| 36 |
+
"step": 80000
|
| 37 |
+
},
|
| 38 |
+
{
|
| 39 |
+
"epoch": 0.81,
|
| 40 |
+
"learning_rate": 1.61988359373945e-06,
|
| 41 |
+
"loss": 1.0767,
|
| 42 |
+
"step": 120000
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"epoch": 0.81,
|
| 46 |
+
"eval_loss": 1.0318435430526733,
|
| 47 |
+
"eval_runtime": 731.7232,
|
| 48 |
+
"eval_samples_per_second": 42.364,
|
| 49 |
+
"eval_steps_per_second": 7.061,
|
| 50 |
+
"step": 120000
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.08,
|
| 54 |
+
"learning_rate": 2.1598357888482263e-06,
|
| 55 |
+
"loss": 1.0737,
|
| 56 |
+
"step": 160000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 1.08,
|
| 60 |
+
"eval_loss": 1.031322717666626,
|
| 61 |
+
"eval_runtime": 731.8002,
|
| 62 |
+
"eval_samples_per_second": 42.36,
|
| 63 |
+
"eval_steps_per_second": 7.061,
|
| 64 |
+
"step": 160000
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 1.35,
|
| 68 |
+
"learning_rate": 2.699828496576684e-06,
|
| 69 |
+
"loss": 1.0708,
|
| 70 |
+
"step": 200000
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.35,
|
| 74 |
+
"eval_loss": 1.0267112255096436,
|
| 75 |
+
"eval_runtime": 738.3448,
|
| 76 |
+
"eval_samples_per_second": 41.984,
|
| 77 |
+
"eval_steps_per_second": 6.998,
|
| 78 |
+
"step": 200000
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 1.62,
|
| 82 |
+
"learning_rate": 3.239807700098581e-06,
|
| 83 |
+
"loss": 1.0707,
|
| 84 |
+
"step": 240000
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.62,
|
| 88 |
+
"eval_loss": 1.0309606790542603,
|
| 89 |
+
"eval_runtime": 731.7252,
|
| 90 |
+
"eval_samples_per_second": 42.364,
|
| 91 |
+
"eval_steps_per_second": 7.061,
|
| 92 |
+
"step": 240000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 1.89,
|
| 96 |
+
"learning_rate": 3.779773399413918e-06,
|
| 97 |
+
"loss": 1.0699,
|
| 98 |
+
"step": 280000
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.89,
|
| 102 |
+
"eval_loss": 1.0270657539367676,
|
| 103 |
+
"eval_runtime": 731.6452,
|
| 104 |
+
"eval_samples_per_second": 42.369,
|
| 105 |
+
"eval_steps_per_second": 7.062,
|
| 106 |
+
"step": 280000
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 2.16,
|
| 110 |
+
"learning_rate": 4.319739098729255e-06,
|
| 111 |
+
"loss": 1.0683,
|
| 112 |
+
"step": 320000
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 2.16,
|
| 116 |
+
"eval_loss": 1.0262391567230225,
|
| 117 |
+
"eval_runtime": 735.9312,
|
| 118 |
+
"eval_samples_per_second": 42.122,
|
| 119 |
+
"eval_steps_per_second": 7.021,
|
| 120 |
+
"step": 320000
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 2.43,
|
| 124 |
+
"learning_rate": 4.859718302251152e-06,
|
| 125 |
+
"loss": 1.0686,
|
| 126 |
+
"step": 360000
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 2.43,
|
| 130 |
+
"eval_loss": 1.0267152786254883,
|
| 131 |
+
"eval_runtime": 731.9313,
|
| 132 |
+
"eval_samples_per_second": 42.352,
|
| 133 |
+
"eval_steps_per_second": 7.059,
|
| 134 |
+
"step": 360000
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 2.7,
|
| 138 |
+
"learning_rate": 5.3997110099796094e-06,
|
| 139 |
+
"loss": 1.0693,
|
| 140 |
+
"step": 400000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 2.7,
|
| 144 |
+
"eval_loss": 1.0279877185821533,
|
| 145 |
+
"eval_runtime": 731.9449,
|
| 146 |
+
"eval_samples_per_second": 42.352,
|
| 147 |
+
"eval_steps_per_second": 7.059,
|
| 148 |
+
"step": 400000
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 2.97,
|
| 152 |
+
"learning_rate": 5.939690213501507e-06,
|
| 153 |
+
"loss": 1.0714,
|
| 154 |
+
"step": 440000
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 2.97,
|
| 158 |
+
"eval_loss": 1.030158519744873,
|
| 159 |
+
"eval_runtime": 738.5138,
|
| 160 |
+
"eval_samples_per_second": 41.975,
|
| 161 |
+
"eval_steps_per_second": 6.996,
|
| 162 |
+
"step": 440000
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 3.24,
|
| 166 |
+
"learning_rate": 6.479655912816843e-06,
|
| 167 |
+
"loss": 1.07,
|
| 168 |
+
"step": 480000
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 3.24,
|
| 172 |
+
"eval_loss": 1.0300403833389282,
|
| 173 |
+
"eval_runtime": 738.4095,
|
| 174 |
+
"eval_samples_per_second": 41.981,
|
| 175 |
+
"eval_steps_per_second": 6.997,
|
| 176 |
+
"step": 480000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 3.51,
|
| 180 |
+
"learning_rate": 7.019648620545301e-06,
|
| 181 |
+
"loss": 1.0719,
|
| 182 |
+
"step": 520000
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 3.51,
|
| 186 |
+
"eval_loss": 1.0280787944793701,
|
| 187 |
+
"eval_runtime": 738.1732,
|
| 188 |
+
"eval_samples_per_second": 41.994,
|
| 189 |
+
"eval_steps_per_second": 7.0,
|
| 190 |
+
"step": 520000
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 3.78,
|
| 194 |
+
"learning_rate": 7.5596143198606375e-06,
|
| 195 |
+
"loss": 1.0735,
|
| 196 |
+
"step": 560000
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 3.78,
|
| 200 |
+
"eval_loss": 1.0311589241027832,
|
| 201 |
+
"eval_runtime": 741.9081,
|
| 202 |
+
"eval_samples_per_second": 41.783,
|
| 203 |
+
"eval_steps_per_second": 6.964,
|
| 204 |
+
"step": 560000
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 4.05,
|
| 208 |
+
"learning_rate": 8.099607027589095e-06,
|
| 209 |
+
"loss": 1.0781,
|
| 210 |
+
"step": 600000
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 4.05,
|
| 214 |
+
"eval_loss": 1.035496473312378,
|
| 215 |
+
"eval_runtime": 733.6394,
|
| 216 |
+
"eval_samples_per_second": 42.254,
|
| 217 |
+
"eval_steps_per_second": 7.043,
|
| 218 |
+
"step": 600000
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 4.32,
|
| 222 |
+
"learning_rate": 8.639559222697871e-06,
|
| 223 |
+
"loss": 1.0768,
|
| 224 |
+
"step": 640000
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 4.32,
|
| 228 |
+
"eval_loss": 1.0344969034194946,
|
| 229 |
+
"eval_runtime": 733.6669,
|
| 230 |
+
"eval_samples_per_second": 42.252,
|
| 231 |
+
"eval_steps_per_second": 7.043,
|
| 232 |
+
"step": 640000
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 4.59,
|
| 236 |
+
"learning_rate": 9.179538426219769e-06,
|
| 237 |
+
"loss": 1.0797,
|
| 238 |
+
"step": 680000
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 4.59,
|
| 242 |
+
"eval_loss": 1.0367223024368286,
|
| 243 |
+
"eval_runtime": 733.6495,
|
| 244 |
+
"eval_samples_per_second": 42.253,
|
| 245 |
+
"eval_steps_per_second": 7.043,
|
| 246 |
+
"step": 680000
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 4.86,
|
| 250 |
+
"learning_rate": 9.719517629741665e-06,
|
| 251 |
+
"loss": 1.0834,
|
| 252 |
+
"step": 720000
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 4.86,
|
| 256 |
+
"eval_loss": 1.0388565063476562,
|
| 257 |
+
"eval_runtime": 733.7938,
|
| 258 |
+
"eval_samples_per_second": 42.245,
|
| 259 |
+
"eval_steps_per_second": 7.041,
|
| 260 |
+
"step": 720000
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 5.13,
|
| 264 |
+
"learning_rate": 9.971168518993666e-06,
|
| 265 |
+
"loss": 1.0833,
|
| 266 |
+
"step": 760000
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 5.13,
|
| 270 |
+
"eval_loss": 1.0380399227142334,
|
| 271 |
+
"eval_runtime": 733.6877,
|
| 272 |
+
"eval_samples_per_second": 42.251,
|
| 273 |
+
"eval_steps_per_second": 7.043,
|
| 274 |
+
"step": 760000
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 5.4,
|
| 278 |
+
"learning_rate": 9.911172330180852e-06,
|
| 279 |
+
"loss": 1.0834,
|
| 280 |
+
"step": 800000
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 5.4,
|
| 284 |
+
"eval_loss": 1.0390028953552246,
|
| 285 |
+
"eval_runtime": 733.5168,
|
| 286 |
+
"eval_samples_per_second": 42.261,
|
| 287 |
+
"eval_steps_per_second": 7.044,
|
| 288 |
+
"step": 800000
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 5.67,
|
| 292 |
+
"learning_rate": 9.851176141368038e-06,
|
| 293 |
+
"loss": 1.0842,
|
| 294 |
+
"step": 840000
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 5.67,
|
| 298 |
+
"eval_loss": 1.03898286819458,
|
| 299 |
+
"eval_runtime": 733.1528,
|
| 300 |
+
"eval_samples_per_second": 42.282,
|
| 301 |
+
"eval_steps_per_second": 7.048,
|
| 302 |
+
"step": 840000
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 5.94,
|
| 306 |
+
"learning_rate": 9.791179952555222e-06,
|
| 307 |
+
"loss": 1.0854,
|
| 308 |
+
"step": 880000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 5.94,
|
| 312 |
+
"eval_loss": 1.0376673936843872,
|
| 313 |
+
"eval_runtime": 733.2924,
|
| 314 |
+
"eval_samples_per_second": 42.274,
|
| 315 |
+
"eval_steps_per_second": 7.046,
|
| 316 |
+
"step": 880000
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 6.21,
|
| 320 |
+
"learning_rate": 9.73117926234022e-06,
|
| 321 |
+
"loss": 1.0797,
|
| 322 |
+
"step": 920000
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 6.21,
|
| 326 |
+
"eval_loss": 1.0366719961166382,
|
| 327 |
+
"eval_runtime": 733.1418,
|
| 328 |
+
"eval_samples_per_second": 42.282,
|
| 329 |
+
"eval_steps_per_second": 7.048,
|
| 330 |
+
"step": 920000
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 6.48,
|
| 334 |
+
"learning_rate": 9.671183073527405e-06,
|
| 335 |
+
"loss": 1.0788,
|
| 336 |
+
"step": 960000
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 6.48,
|
| 340 |
+
"eval_loss": 1.0325465202331543,
|
| 341 |
+
"eval_runtime": 733.2281,
|
| 342 |
+
"eval_samples_per_second": 42.277,
|
| 343 |
+
"eval_steps_per_second": 7.047,
|
| 344 |
+
"step": 960000
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 6.75,
|
| 348 |
+
"learning_rate": 9.611185384247194e-06,
|
| 349 |
+
"loss": 1.0798,
|
| 350 |
+
"step": 1000000
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 6.75,
|
| 354 |
+
"eval_loss": 1.0343610048294067,
|
| 355 |
+
"eval_runtime": 733.2347,
|
| 356 |
+
"eval_samples_per_second": 42.277,
|
| 357 |
+
"eval_steps_per_second": 7.047,
|
| 358 |
+
"step": 1000000
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 7.02,
|
| 362 |
+
"learning_rate": 9.551186194499587e-06,
|
| 363 |
+
"loss": 1.0784,
|
| 364 |
+
"step": 1040000
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 7.02,
|
| 368 |
+
"eval_loss": 1.0340007543563843,
|
| 369 |
+
"eval_runtime": 733.1036,
|
| 370 |
+
"eval_samples_per_second": 42.285,
|
| 371 |
+
"eval_steps_per_second": 7.048,
|
| 372 |
+
"step": 1040000
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 7.29,
|
| 376 |
+
"learning_rate": 9.491185504284586e-06,
|
| 377 |
+
"loss": 1.0723,
|
| 378 |
+
"step": 1080000
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 7.29,
|
| 382 |
+
"eval_loss": 1.0315660238265991,
|
| 383 |
+
"eval_runtime": 733.0253,
|
| 384 |
+
"eval_samples_per_second": 42.289,
|
| 385 |
+
"eval_steps_per_second": 7.049,
|
| 386 |
+
"step": 1080000
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 7.56,
|
| 390 |
+
"learning_rate": 9.43118931547177e-06,
|
| 391 |
+
"loss": 1.0732,
|
| 392 |
+
"step": 1120000
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 7.56,
|
| 396 |
+
"eval_loss": 1.0322542190551758,
|
| 397 |
+
"eval_runtime": 733.0537,
|
| 398 |
+
"eval_samples_per_second": 42.287,
|
| 399 |
+
"eval_steps_per_second": 7.049,
|
| 400 |
+
"step": 1120000
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 7.83,
|
| 404 |
+
"learning_rate": 9.371193126658954e-06,
|
| 405 |
+
"loss": 1.0744,
|
| 406 |
+
"step": 1160000
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 7.83,
|
| 410 |
+
"eval_loss": 1.029881477355957,
|
| 411 |
+
"eval_runtime": 733.0163,
|
| 412 |
+
"eval_samples_per_second": 42.29,
|
| 413 |
+
"eval_steps_per_second": 7.049,
|
| 414 |
+
"step": 1160000
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 8.1,
|
| 418 |
+
"learning_rate": 9.31119693784614e-06,
|
| 419 |
+
"loss": 1.0707,
|
| 420 |
+
"step": 1200000
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 8.1,
|
| 424 |
+
"eval_loss": 1.029407262802124,
|
| 425 |
+
"eval_runtime": 733.0808,
|
| 426 |
+
"eval_samples_per_second": 42.286,
|
| 427 |
+
"eval_steps_per_second": 7.048,
|
| 428 |
+
"step": 1200000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 8.37,
|
| 432 |
+
"learning_rate": 9.25120224950072e-06,
|
| 433 |
+
"loss": 1.0694,
|
| 434 |
+
"step": 1240000
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 8.37,
|
| 438 |
+
"eval_loss": 1.0290433168411255,
|
| 439 |
+
"eval_runtime": 732.9639,
|
| 440 |
+
"eval_samples_per_second": 42.293,
|
| 441 |
+
"eval_steps_per_second": 7.049,
|
| 442 |
+
"step": 1240000
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 8.64,
|
| 446 |
+
"learning_rate": 9.191201559285718e-06,
|
| 447 |
+
"loss": 1.0691,
|
| 448 |
+
"step": 1280000
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 8.64,
|
| 452 |
+
"eval_loss": 1.028395414352417,
|
| 453 |
+
"eval_runtime": 732.9086,
|
| 454 |
+
"eval_samples_per_second": 42.296,
|
| 455 |
+
"eval_steps_per_second": 7.05,
|
| 456 |
+
"step": 1280000
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 8.91,
|
| 460 |
+
"learning_rate": 9.131205370472902e-06,
|
| 461 |
+
"loss": 1.0699,
|
| 462 |
+
"step": 1320000
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 8.91,
|
| 466 |
+
"eval_loss": 1.0277873277664185,
|
| 467 |
+
"eval_runtime": 732.8691,
|
| 468 |
+
"eval_samples_per_second": 42.298,
|
| 469 |
+
"eval_steps_per_second": 7.05,
|
| 470 |
+
"step": 1320000
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 9.18,
|
| 474 |
+
"learning_rate": 9.071209181660088e-06,
|
| 475 |
+
"loss": 1.0643,
|
| 476 |
+
"step": 1360000
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 9.18,
|
| 480 |
+
"eval_loss": 1.0250962972640991,
|
| 481 |
+
"eval_runtime": 731.2424,
|
| 482 |
+
"eval_samples_per_second": 42.392,
|
| 483 |
+
"eval_steps_per_second": 7.066,
|
| 484 |
+
"step": 1360000
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 9.45,
|
| 488 |
+
"learning_rate": 9.011212992847273e-06,
|
| 489 |
+
"loss": 1.0642,
|
| 490 |
+
"step": 1400000
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 9.45,
|
| 494 |
+
"eval_loss": 1.0227874517440796,
|
| 495 |
+
"eval_runtime": 731.2164,
|
| 496 |
+
"eval_samples_per_second": 42.394,
|
| 497 |
+
"eval_steps_per_second": 7.066,
|
| 498 |
+
"step": 1400000
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 9.72,
|
| 502 |
+
"learning_rate": 8.951216804034457e-06,
|
| 503 |
+
"loss": 1.065,
|
| 504 |
+
"step": 1440000
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 9.72,
|
| 508 |
+
"eval_loss": 1.0216150283813477,
|
| 509 |
+
"eval_runtime": 730.9956,
|
| 510 |
+
"eval_samples_per_second": 42.407,
|
| 511 |
+
"eval_steps_per_second": 7.068,
|
| 512 |
+
"step": 1440000
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 9.99,
|
| 516 |
+
"learning_rate": 8.891219114754247e-06,
|
| 517 |
+
"loss": 1.0647,
|
| 518 |
+
"step": 1480000
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 9.99,
|
| 522 |
+
"eval_loss": 1.0240837335586548,
|
| 523 |
+
"eval_runtime": 731.0731,
|
| 524 |
+
"eval_samples_per_second": 42.402,
|
| 525 |
+
"eval_steps_per_second": 7.068,
|
| 526 |
+
"step": 1480000
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 10.26,
|
| 530 |
+
"learning_rate": 8.831224426408827e-06,
|
| 531 |
+
"loss": 1.0587,
|
| 532 |
+
"step": 1520000
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 10.26,
|
| 536 |
+
"eval_loss": 1.0212384462356567,
|
| 537 |
+
"eval_runtime": 733.0687,
|
| 538 |
+
"eval_samples_per_second": 42.287,
|
| 539 |
+
"eval_steps_per_second": 7.048,
|
| 540 |
+
"step": 1520000
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 10.53,
|
| 544 |
+
"learning_rate": 8.771226737128616e-06,
|
| 545 |
+
"loss": 1.0611,
|
| 546 |
+
"step": 1560000
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 10.53,
|
| 550 |
+
"eval_loss": 1.0210351943969727,
|
| 551 |
+
"eval_runtime": 732.9747,
|
| 552 |
+
"eval_samples_per_second": 42.292,
|
| 553 |
+
"eval_steps_per_second": 7.049,
|
| 554 |
+
"step": 1560000
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 10.8,
|
| 558 |
+
"learning_rate": 8.711230548315802e-06,
|
| 559 |
+
"loss": 1.06,
|
| 560 |
+
"step": 1600000
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 10.8,
|
| 564 |
+
"eval_loss": 1.0200951099395752,
|
| 565 |
+
"eval_runtime": 732.955,
|
| 566 |
+
"eval_samples_per_second": 42.293,
|
| 567 |
+
"eval_steps_per_second": 7.05,
|
| 568 |
+
"step": 1600000
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 11.07,
|
| 572 |
+
"learning_rate": 8.651234359502986e-06,
|
| 573 |
+
"loss": 1.058,
|
| 574 |
+
"step": 1640000
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 11.07,
|
| 578 |
+
"eval_loss": 1.018425703048706,
|
| 579 |
+
"eval_runtime": 732.944,
|
| 580 |
+
"eval_samples_per_second": 42.294,
|
| 581 |
+
"eval_steps_per_second": 7.05,
|
| 582 |
+
"step": 1640000
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 11.34,
|
| 586 |
+
"learning_rate": 8.591233669287984e-06,
|
| 587 |
+
"loss": 1.0543,
|
| 588 |
+
"step": 1680000
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 11.34,
|
| 592 |
+
"eval_loss": 1.0207031965255737,
|
| 593 |
+
"eval_runtime": 733.1758,
|
| 594 |
+
"eval_samples_per_second": 42.28,
|
| 595 |
+
"eval_steps_per_second": 7.047,
|
| 596 |
+
"step": 1680000
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 11.61,
|
| 600 |
+
"learning_rate": 8.531237480475169e-06,
|
| 601 |
+
"loss": 1.057,
|
| 602 |
+
"step": 1720000
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 11.61,
|
| 606 |
+
"eval_loss": 1.0180833339691162,
|
| 607 |
+
"eval_runtime": 733.0248,
|
| 608 |
+
"eval_samples_per_second": 42.289,
|
| 609 |
+
"eval_steps_per_second": 7.049,
|
| 610 |
+
"step": 1720000
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 11.88,
|
| 614 |
+
"learning_rate": 8.471236790260167e-06,
|
| 615 |
+
"loss": 1.0546,
|
| 616 |
+
"step": 1760000
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 11.88,
|
| 620 |
+
"eval_loss": 1.0137962102890015,
|
| 621 |
+
"eval_runtime": 733.3395,
|
| 622 |
+
"eval_samples_per_second": 42.271,
|
| 623 |
+
"eval_steps_per_second": 7.046,
|
| 624 |
+
"step": 1760000
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 12.15,
|
| 628 |
+
"learning_rate": 8.411240601447351e-06,
|
| 629 |
+
"loss": 1.0519,
|
| 630 |
+
"step": 1800000
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 12.15,
|
| 634 |
+
"eval_loss": 1.0171297788619995,
|
| 635 |
+
"eval_runtime": 731.3327,
|
| 636 |
+
"eval_samples_per_second": 42.387,
|
| 637 |
+
"eval_steps_per_second": 7.065,
|
| 638 |
+
"step": 1800000
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 12.42,
|
| 642 |
+
"learning_rate": 8.351245913101933e-06,
|
| 643 |
+
"loss": 1.0505,
|
| 644 |
+
"step": 1840000
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 12.42,
|
| 648 |
+
"eval_loss": 1.0128827095031738,
|
| 649 |
+
"eval_runtime": 730.8675,
|
| 650 |
+
"eval_samples_per_second": 42.414,
|
| 651 |
+
"eval_steps_per_second": 7.07,
|
| 652 |
+
"step": 1840000
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 12.69,
|
| 656 |
+
"learning_rate": 8.291246723354326e-06,
|
| 657 |
+
"loss": 1.0513,
|
| 658 |
+
"step": 1880000
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"epoch": 12.69,
|
| 662 |
+
"eval_loss": 1.0132421255111694,
|
| 663 |
+
"eval_runtime": 730.8331,
|
| 664 |
+
"eval_samples_per_second": 42.416,
|
| 665 |
+
"eval_steps_per_second": 7.07,
|
| 666 |
+
"step": 1880000
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 12.96,
|
| 670 |
+
"learning_rate": 8.23125053454151e-06,
|
| 671 |
+
"loss": 1.0514,
|
| 672 |
+
"step": 1920000
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 12.96,
|
| 676 |
+
"eval_loss": 1.0138520002365112,
|
| 677 |
+
"eval_runtime": 731.3619,
|
| 678 |
+
"eval_samples_per_second": 42.385,
|
| 679 |
+
"eval_steps_per_second": 7.065,
|
| 680 |
+
"step": 1920000
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 13.23,
|
| 684 |
+
"learning_rate": 8.1712528452613e-06,
|
| 685 |
+
"loss": 1.0473,
|
| 686 |
+
"step": 1960000
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 13.23,
|
| 690 |
+
"eval_loss": 1.0119198560714722,
|
| 691 |
+
"eval_runtime": 731.3481,
|
| 692 |
+
"eval_samples_per_second": 42.386,
|
| 693 |
+
"eval_steps_per_second": 7.065,
|
| 694 |
+
"step": 1960000
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 13.5,
|
| 698 |
+
"learning_rate": 8.111255155981088e-06,
|
| 699 |
+
"loss": 1.0461,
|
| 700 |
+
"step": 2000000
|
| 701 |
+
},
|
| 702 |
+
{
|
| 703 |
+
"epoch": 13.5,
|
| 704 |
+
"eval_loss": 1.0108563899993896,
|
| 705 |
+
"eval_runtime": 731.5139,
|
| 706 |
+
"eval_samples_per_second": 42.377,
|
| 707 |
+
"eval_steps_per_second": 7.063,
|
| 708 |
+
"step": 2000000
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 13.77,
|
| 712 |
+
"learning_rate": 8.051255966233482e-06,
|
| 713 |
+
"loss": 1.0496,
|
| 714 |
+
"step": 2040000
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 13.77,
|
| 718 |
+
"eval_loss": 1.007481336593628,
|
| 719 |
+
"eval_runtime": 731.5836,
|
| 720 |
+
"eval_samples_per_second": 42.372,
|
| 721 |
+
"eval_steps_per_second": 7.063,
|
| 722 |
+
"step": 2040000
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 14.04,
|
| 726 |
+
"learning_rate": 7.991261277888063e-06,
|
| 727 |
+
"loss": 1.047,
|
| 728 |
+
"step": 2080000
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 14.04,
|
| 732 |
+
"eval_loss": 1.0081983804702759,
|
| 733 |
+
"eval_runtime": 735.3379,
|
| 734 |
+
"eval_samples_per_second": 42.156,
|
| 735 |
+
"eval_steps_per_second": 7.027,
|
| 736 |
+
"step": 2080000
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 14.31,
|
| 740 |
+
"learning_rate": 7.931263588607852e-06,
|
| 741 |
+
"loss": 1.0435,
|
| 742 |
+
"step": 2120000
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"epoch": 14.31,
|
| 746 |
+
"eval_loss": 1.0105235576629639,
|
| 747 |
+
"eval_runtime": 731.6633,
|
| 748 |
+
"eval_samples_per_second": 42.368,
|
| 749 |
+
"eval_steps_per_second": 7.062,
|
| 750 |
+
"step": 2120000
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 14.58,
|
| 754 |
+
"learning_rate": 7.871264398860246e-06,
|
| 755 |
+
"loss": 1.0422,
|
| 756 |
+
"step": 2160000
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 14.58,
|
| 760 |
+
"eval_loss": 1.009293556213379,
|
| 761 |
+
"eval_runtime": 738.141,
|
| 762 |
+
"eval_samples_per_second": 41.996,
|
| 763 |
+
"eval_steps_per_second": 7.0,
|
| 764 |
+
"step": 2160000
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 14.85,
|
| 768 |
+
"learning_rate": 7.81126821004743e-06,
|
| 769 |
+
"loss": 1.0438,
|
| 770 |
+
"step": 2200000
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 14.85,
|
| 774 |
+
"eval_loss": 1.0082749128341675,
|
| 775 |
+
"eval_runtime": 731.6793,
|
| 776 |
+
"eval_samples_per_second": 42.367,
|
| 777 |
+
"eval_steps_per_second": 7.062,
|
| 778 |
+
"step": 2200000
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 15.12,
|
| 782 |
+
"learning_rate": 7.751269020299824e-06,
|
| 783 |
+
"loss": 1.0426,
|
| 784 |
+
"step": 2240000
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"epoch": 15.12,
|
| 788 |
+
"eval_loss": 1.0077896118164062,
|
| 789 |
+
"eval_runtime": 731.8248,
|
| 790 |
+
"eval_samples_per_second": 42.358,
|
| 791 |
+
"eval_steps_per_second": 7.06,
|
| 792 |
+
"step": 2240000
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 15.39,
|
| 796 |
+
"learning_rate": 7.691274331954405e-06,
|
| 797 |
+
"loss": 1.0402,
|
| 798 |
+
"step": 2280000
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 15.39,
|
| 802 |
+
"eval_loss": 1.0057011842727661,
|
| 803 |
+
"eval_runtime": 731.8433,
|
| 804 |
+
"eval_samples_per_second": 42.357,
|
| 805 |
+
"eval_steps_per_second": 7.06,
|
| 806 |
+
"step": 2280000
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 15.66,
|
| 810 |
+
"learning_rate": 7.631275142206797e-06,
|
| 811 |
+
"loss": 1.0403,
|
| 812 |
+
"step": 2320000
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 15.66,
|
| 816 |
+
"eval_loss": 1.003925085067749,
|
| 817 |
+
"eval_runtime": 732.0244,
|
| 818 |
+
"eval_samples_per_second": 42.347,
|
| 819 |
+
"eval_steps_per_second": 7.059,
|
| 820 |
+
"step": 2320000
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 15.93,
|
| 824 |
+
"learning_rate": 7.571277452926588e-06,
|
| 825 |
+
"loss": 1.0398,
|
| 826 |
+
"step": 2360000
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 15.93,
|
| 830 |
+
"eval_loss": 1.003111720085144,
|
| 831 |
+
"eval_runtime": 732.0283,
|
| 832 |
+
"eval_samples_per_second": 42.347,
|
| 833 |
+
"eval_steps_per_second": 7.058,
|
| 834 |
+
"step": 2360000
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 16.21,
|
| 838 |
+
"learning_rate": 7.511281264113772e-06,
|
| 839 |
+
"loss": 1.0368,
|
| 840 |
+
"step": 2400000
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 16.21,
|
| 844 |
+
"eval_loss": 1.003316879272461,
|
| 845 |
+
"eval_runtime": 731.9328,
|
| 846 |
+
"eval_samples_per_second": 42.352,
|
| 847 |
+
"eval_steps_per_second": 7.059,
|
| 848 |
+
"step": 2400000
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 16.48,
|
| 852 |
+
"learning_rate": 7.451283574833561e-06,
|
| 853 |
+
"loss": 1.036,
|
| 854 |
+
"step": 2440000
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 16.48,
|
| 858 |
+
"eval_loss": 1.0017095804214478,
|
| 859 |
+
"eval_runtime": 733.4383,
|
| 860 |
+
"eval_samples_per_second": 42.265,
|
| 861 |
+
"eval_steps_per_second": 7.045,
|
| 862 |
+
"step": 2440000
|
| 863 |
+
},
|
| 864 |
+
{
|
| 865 |
+
"epoch": 16.75,
|
| 866 |
+
"learning_rate": 7.391287386020746e-06,
|
| 867 |
+
"loss": 1.0369,
|
| 868 |
+
"step": 2480000
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 16.75,
|
| 872 |
+
"eval_loss": 0.9995156526565552,
|
| 873 |
+
"eval_runtime": 733.6542,
|
| 874 |
+
"eval_samples_per_second": 42.253,
|
| 875 |
+
"eval_steps_per_second": 7.043,
|
| 876 |
+
"step": 2480000
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 17.02,
|
| 880 |
+
"learning_rate": 7.33129119720793e-06,
|
| 881 |
+
"loss": 1.0383,
|
| 882 |
+
"step": 2520000
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 17.02,
|
| 886 |
+
"eval_loss": 1.0012413263320923,
|
| 887 |
+
"eval_runtime": 733.4897,
|
| 888 |
+
"eval_samples_per_second": 42.262,
|
| 889 |
+
"eval_steps_per_second": 7.044,
|
| 890 |
+
"step": 2520000
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 17.29,
|
| 894 |
+
"learning_rate": 7.2712965088625106e-06,
|
| 895 |
+
"loss": 1.0312,
|
| 896 |
+
"step": 2560000
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 17.29,
|
| 900 |
+
"eval_loss": 0.9984557032585144,
|
| 901 |
+
"eval_runtime": 733.8,
|
| 902 |
+
"eval_samples_per_second": 42.244,
|
| 903 |
+
"eval_steps_per_second": 7.041,
|
| 904 |
+
"step": 2560000
|
| 905 |
+
},
|
| 906 |
+
{
|
| 907 |
+
"epoch": 17.56,
|
| 908 |
+
"learning_rate": 7.211298819582301e-06,
|
| 909 |
+
"loss": 1.0344,
|
| 910 |
+
"step": 2600000
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 17.56,
|
| 914 |
+
"eval_loss": 1.0010262727737427,
|
| 915 |
+
"eval_runtime": 733.4795,
|
| 916 |
+
"eval_samples_per_second": 42.263,
|
| 917 |
+
"eval_steps_per_second": 7.045,
|
| 918 |
+
"step": 2600000
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 17.83,
|
| 922 |
+
"learning_rate": 7.151301130302089e-06,
|
| 923 |
+
"loss": 1.0333,
|
| 924 |
+
"step": 2640000
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 17.83,
|
| 928 |
+
"eval_loss": 0.995519757270813,
|
| 929 |
+
"eval_runtime": 733.5669,
|
| 930 |
+
"eval_samples_per_second": 42.258,
|
| 931 |
+
"eval_steps_per_second": 7.044,
|
| 932 |
+
"step": 2640000
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 18.1,
|
| 936 |
+
"learning_rate": 7.091303441021878e-06,
|
| 937 |
+
"loss": 1.0333,
|
| 938 |
+
"step": 2680000
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 18.1,
|
| 942 |
+
"eval_loss": 0.9943256974220276,
|
| 943 |
+
"eval_runtime": 733.6185,
|
| 944 |
+
"eval_samples_per_second": 42.255,
|
| 945 |
+
"eval_steps_per_second": 7.043,
|
| 946 |
+
"step": 2680000
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 18.37,
|
| 950 |
+
"learning_rate": 7.031307252209064e-06,
|
| 951 |
+
"loss": 1.0302,
|
| 952 |
+
"step": 2720000
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 18.37,
|
| 956 |
+
"eval_loss": 0.993632972240448,
|
| 957 |
+
"eval_runtime": 731.903,
|
| 958 |
+
"eval_samples_per_second": 42.354,
|
| 959 |
+
"eval_steps_per_second": 7.06,
|
| 960 |
+
"step": 2720000
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 18.64,
|
| 964 |
+
"learning_rate": 6.971312563863644e-06,
|
| 965 |
+
"loss": 1.0301,
|
| 966 |
+
"step": 2760000
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 18.64,
|
| 970 |
+
"eval_loss": 0.9943361282348633,
|
| 971 |
+
"eval_runtime": 731.9076,
|
| 972 |
+
"eval_samples_per_second": 42.354,
|
| 973 |
+
"eval_steps_per_second": 7.06,
|
| 974 |
+
"step": 2760000
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 18.91,
|
| 978 |
+
"learning_rate": 6.911311873648642e-06,
|
| 979 |
+
"loss": 1.0312,
|
| 980 |
+
"step": 2800000
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 18.91,
|
| 984 |
+
"eval_loss": 0.9974587559700012,
|
| 985 |
+
"eval_runtime": 731.8693,
|
| 986 |
+
"eval_samples_per_second": 42.356,
|
| 987 |
+
"eval_steps_per_second": 7.06,
|
| 988 |
+
"step": 2800000
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"epoch": 19.18,
|
| 992 |
+
"learning_rate": 6.8513171853032225e-06,
|
| 993 |
+
"loss": 1.0265,
|
| 994 |
+
"step": 2840000
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 19.18,
|
| 998 |
+
"eval_loss": 0.9932205677032471,
|
| 999 |
+
"eval_runtime": 731.8147,
|
| 1000 |
+
"eval_samples_per_second": 42.359,
|
| 1001 |
+
"eval_steps_per_second": 7.061,
|
| 1002 |
+
"step": 2840000
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 19.45,
|
| 1006 |
+
"learning_rate": 6.791320996490407e-06,
|
| 1007 |
+
"loss": 1.0271,
|
| 1008 |
+
"step": 2880000
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 19.45,
|
| 1012 |
+
"eval_loss": 0.9906313419342041,
|
| 1013 |
+
"eval_runtime": 732.268,
|
| 1014 |
+
"eval_samples_per_second": 42.333,
|
| 1015 |
+
"eval_steps_per_second": 7.056,
|
| 1016 |
+
"step": 2880000
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 19.72,
|
| 1020 |
+
"learning_rate": 6.7313233072101965e-06,
|
| 1021 |
+
"loss": 1.0275,
|
| 1022 |
+
"step": 2920000
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 19.72,
|
| 1026 |
+
"eval_loss": 0.9913482666015625,
|
| 1027 |
+
"eval_runtime": 731.9864,
|
| 1028 |
+
"eval_samples_per_second": 42.349,
|
| 1029 |
+
"eval_steps_per_second": 7.059,
|
| 1030 |
+
"step": 2920000
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 19.99,
|
| 1034 |
+
"learning_rate": 6.6713271183973816e-06,
|
| 1035 |
+
"loss": 1.0267,
|
| 1036 |
+
"step": 2960000
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"epoch": 19.99,
|
| 1040 |
+
"eval_loss": 0.9909727573394775,
|
| 1041 |
+
"eval_runtime": 731.9562,
|
| 1042 |
+
"eval_samples_per_second": 42.351,
|
| 1043 |
+
"eval_steps_per_second": 7.059,
|
| 1044 |
+
"step": 2960000
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 20.26,
|
| 1048 |
+
"learning_rate": 6.611332430051962e-06,
|
| 1049 |
+
"loss": 1.0238,
|
| 1050 |
+
"step": 3000000
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"epoch": 20.26,
|
| 1054 |
+
"eval_loss": 0.9900842308998108,
|
| 1055 |
+
"eval_runtime": 732.034,
|
| 1056 |
+
"eval_samples_per_second": 42.346,
|
| 1057 |
+
"eval_steps_per_second": 7.058,
|
| 1058 |
+
"step": 3000000
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 20.53,
|
| 1062 |
+
"learning_rate": 6.551330239369564e-06,
|
| 1063 |
+
"loss": 1.0236,
|
| 1064 |
+
"step": 3040000
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 20.53,
|
| 1068 |
+
"eval_loss": 0.9898263812065125,
|
| 1069 |
+
"eval_runtime": 731.9968,
|
| 1070 |
+
"eval_samples_per_second": 42.349,
|
| 1071 |
+
"eval_steps_per_second": 7.059,
|
| 1072 |
+
"step": 3040000
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 20.8,
|
| 1076 |
+
"learning_rate": 6.491334050556749e-06,
|
| 1077 |
+
"loss": 1.0237,
|
| 1078 |
+
"step": 3080000
|
| 1079 |
+
},
|
| 1080 |
+
{
|
| 1081 |
+
"epoch": 20.8,
|
| 1082 |
+
"eval_loss": 0.991007387638092,
|
| 1083 |
+
"eval_runtime": 731.9959,
|
| 1084 |
+
"eval_samples_per_second": 42.349,
|
| 1085 |
+
"eval_steps_per_second": 7.059,
|
| 1086 |
+
"step": 3080000
|
| 1087 |
+
},
|
| 1088 |
+
{
|
| 1089 |
+
"epoch": 21.07,
|
| 1090 |
+
"learning_rate": 6.4313378617439335e-06,
|
| 1091 |
+
"loss": 1.0238,
|
| 1092 |
+
"step": 3120000
|
| 1093 |
+
},
|
| 1094 |
+
{
|
| 1095 |
+
"epoch": 21.07,
|
| 1096 |
+
"eval_loss": 0.9906073808670044,
|
| 1097 |
+
"eval_runtime": 731.8574,
|
| 1098 |
+
"eval_samples_per_second": 42.357,
|
| 1099 |
+
"eval_steps_per_second": 7.06,
|
| 1100 |
+
"step": 3120000
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 21.34,
|
| 1104 |
+
"learning_rate": 6.3713401724637225e-06,
|
| 1105 |
+
"loss": 1.0193,
|
| 1106 |
+
"step": 3160000
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 21.34,
|
| 1110 |
+
"eval_loss": 0.9887716770172119,
|
| 1111 |
+
"eval_runtime": 731.675,
|
| 1112 |
+
"eval_samples_per_second": 42.367,
|
| 1113 |
+
"eval_steps_per_second": 7.062,
|
| 1114 |
+
"step": 3160000
|
| 1115 |
+
},
|
| 1116 |
+
{
|
| 1117 |
+
"epoch": 21.61,
|
| 1118 |
+
"learning_rate": 6.311342483183512e-06,
|
| 1119 |
+
"loss": 1.0197,
|
| 1120 |
+
"step": 3200000
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"epoch": 21.61,
|
| 1124 |
+
"eval_loss": 0.9868865013122559,
|
| 1125 |
+
"eval_runtime": 741.4509,
|
| 1126 |
+
"eval_samples_per_second": 41.809,
|
| 1127 |
+
"eval_steps_per_second": 6.969,
|
| 1128 |
+
"step": 3200000
|
| 1129 |
+
},
|
| 1130 |
+
{
|
| 1131 |
+
"epoch": 21.88,
|
| 1132 |
+
"learning_rate": 6.251343293435905e-06,
|
| 1133 |
+
"loss": 1.0211,
|
| 1134 |
+
"step": 3240000
|
| 1135 |
+
},
|
| 1136 |
+
{
|
| 1137 |
+
"epoch": 21.88,
|
| 1138 |
+
"eval_loss": 0.9857591986656189,
|
| 1139 |
+
"eval_runtime": 731.6308,
|
| 1140 |
+
"eval_samples_per_second": 42.37,
|
| 1141 |
+
"eval_steps_per_second": 7.062,
|
| 1142 |
+
"step": 3240000
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 22.15,
|
| 1146 |
+
"learning_rate": 6.191347104623091e-06,
|
| 1147 |
+
"loss": 1.0186,
|
| 1148 |
+
"step": 3280000
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 22.15,
|
| 1152 |
+
"eval_loss": 0.9871023893356323,
|
| 1153 |
+
"eval_runtime": 731.6768,
|
| 1154 |
+
"eval_samples_per_second": 42.367,
|
| 1155 |
+
"eval_steps_per_second": 7.062,
|
| 1156 |
+
"step": 3280000
|
| 1157 |
+
},
|
| 1158 |
+
{
|
| 1159 |
+
"epoch": 22.42,
|
| 1160 |
+
"learning_rate": 6.131346414408088e-06,
|
| 1161 |
+
"loss": 1.0155,
|
| 1162 |
+
"step": 3320000
|
| 1163 |
+
},
|
| 1164 |
+
{
|
| 1165 |
+
"epoch": 22.42,
|
| 1166 |
+
"eval_loss": 0.9836686849594116,
|
| 1167 |
+
"eval_runtime": 731.7409,
|
| 1168 |
+
"eval_samples_per_second": 42.363,
|
| 1169 |
+
"eval_steps_per_second": 7.061,
|
| 1170 |
+
"step": 3320000
|
| 1171 |
+
},
|
| 1172 |
+
{
|
| 1173 |
+
"epoch": 22.69,
|
| 1174 |
+
"learning_rate": 6.071350225595274e-06,
|
| 1175 |
+
"loss": 1.0179,
|
| 1176 |
+
"step": 3360000
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 22.69,
|
| 1180 |
+
"eval_loss": 0.9840074777603149,
|
| 1181 |
+
"eval_runtime": 731.7805,
|
| 1182 |
+
"eval_samples_per_second": 42.361,
|
| 1183 |
+
"eval_steps_per_second": 7.061,
|
| 1184 |
+
"step": 3360000
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 22.96,
|
| 1188 |
+
"learning_rate": 6.011354036782459e-06,
|
| 1189 |
+
"loss": 1.0193,
|
| 1190 |
+
"step": 3400000
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 22.96,
|
| 1194 |
+
"eval_loss": 0.9879838824272156,
|
| 1195 |
+
"eval_runtime": 731.7957,
|
| 1196 |
+
"eval_samples_per_second": 42.36,
|
| 1197 |
+
"eval_steps_per_second": 7.061,
|
| 1198 |
+
"step": 3400000
|
| 1199 |
+
},
|
| 1200 |
+
{
|
| 1201 |
+
"epoch": 23.23,
|
| 1202 |
+
"learning_rate": 5.951354847034852e-06,
|
| 1203 |
+
"loss": 1.0132,
|
| 1204 |
+
"step": 3440000
|
| 1205 |
+
},
|
| 1206 |
+
{
|
| 1207 |
+
"epoch": 23.23,
|
| 1208 |
+
"eval_loss": 0.9860868453979492,
|
| 1209 |
+
"eval_runtime": 731.3704,
|
| 1210 |
+
"eval_samples_per_second": 42.385,
|
| 1211 |
+
"eval_steps_per_second": 7.065,
|
| 1212 |
+
"step": 3440000
|
| 1213 |
+
},
|
| 1214 |
+
{
|
| 1215 |
+
"epoch": 23.5,
|
| 1216 |
+
"learning_rate": 5.8913571577546415e-06,
|
| 1217 |
+
"loss": 1.0131,
|
| 1218 |
+
"step": 3480000
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 23.5,
|
| 1222 |
+
"eval_loss": 0.980789840221405,
|
| 1223 |
+
"eval_runtime": 731.4402,
|
| 1224 |
+
"eval_samples_per_second": 42.381,
|
| 1225 |
+
"eval_steps_per_second": 7.064,
|
| 1226 |
+
"step": 3480000
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 23.77,
|
| 1230 |
+
"learning_rate": 5.8313594684744305e-06,
|
| 1231 |
+
"loss": 1.0122,
|
| 1232 |
+
"step": 3520000
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 23.77,
|
| 1236 |
+
"eval_loss": 0.97965407371521,
|
| 1237 |
+
"eval_runtime": 731.5077,
|
| 1238 |
+
"eval_samples_per_second": 42.377,
|
| 1239 |
+
"eval_steps_per_second": 7.063,
|
| 1240 |
+
"step": 3520000
|
| 1241 |
+
},
|
| 1242 |
+
{
|
| 1243 |
+
"epoch": 24.04,
|
| 1244 |
+
"learning_rate": 5.771361779194219e-06,
|
| 1245 |
+
"loss": 1.0157,
|
| 1246 |
+
"step": 3560000
|
| 1247 |
+
},
|
| 1248 |
+
{
|
| 1249 |
+
"epoch": 24.04,
|
| 1250 |
+
"eval_loss": 0.9817570447921753,
|
| 1251 |
+
"eval_runtime": 731.419,
|
| 1252 |
+
"eval_samples_per_second": 42.382,
|
| 1253 |
+
"eval_steps_per_second": 7.064,
|
| 1254 |
+
"step": 3560000
|
| 1255 |
+
},
|
| 1256 |
+
{
|
| 1257 |
+
"epoch": 24.31,
|
| 1258 |
+
"learning_rate": 5.711364089914008e-06,
|
| 1259 |
+
"loss": 1.0106,
|
| 1260 |
+
"step": 3600000
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"epoch": 24.31,
|
| 1264 |
+
"eval_loss": 0.9800501465797424,
|
| 1265 |
+
"eval_runtime": 731.5621,
|
| 1266 |
+
"eval_samples_per_second": 42.374,
|
| 1267 |
+
"eval_steps_per_second": 7.063,
|
| 1268 |
+
"step": 3600000
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 24.58,
|
| 1272 |
+
"learning_rate": 5.65136940156859e-06,
|
| 1273 |
+
"loss": 1.011,
|
| 1274 |
+
"step": 3640000
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 24.58,
|
| 1278 |
+
"eval_loss": 0.9801167249679565,
|
| 1279 |
+
"eval_runtime": 731.533,
|
| 1280 |
+
"eval_samples_per_second": 42.375,
|
| 1281 |
+
"eval_steps_per_second": 7.063,
|
| 1282 |
+
"step": 3640000
|
| 1283 |
+
},
|
| 1284 |
+
{
|
| 1285 |
+
"epoch": 24.85,
|
| 1286 |
+
"learning_rate": 5.591368711353587e-06,
|
| 1287 |
+
"loss": 1.0121,
|
| 1288 |
+
"step": 3680000
|
| 1289 |
+
},
|
| 1290 |
+
{
|
| 1291 |
+
"epoch": 24.85,
|
| 1292 |
+
"eval_loss": 0.9798985719680786,
|
| 1293 |
+
"eval_runtime": 731.4872,
|
| 1294 |
+
"eval_samples_per_second": 42.378,
|
| 1295 |
+
"eval_steps_per_second": 7.064,
|
| 1296 |
+
"step": 3680000
|
| 1297 |
+
},
|
| 1298 |
+
{
|
| 1299 |
+
"epoch": 25.12,
|
| 1300 |
+
"learning_rate": 5.531371022073376e-06,
|
| 1301 |
+
"loss": 1.0094,
|
| 1302 |
+
"step": 3720000
|
| 1303 |
+
},
|
| 1304 |
+
{
|
| 1305 |
+
"epoch": 25.12,
|
| 1306 |
+
"eval_loss": 0.9805187582969666,
|
| 1307 |
+
"eval_runtime": 731.6262,
|
| 1308 |
+
"eval_samples_per_second": 42.37,
|
| 1309 |
+
"eval_steps_per_second": 7.062,
|
| 1310 |
+
"step": 3720000
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 25.39,
|
| 1314 |
+
"learning_rate": 5.471374833260561e-06,
|
| 1315 |
+
"loss": 1.0127,
|
| 1316 |
+
"step": 3760000
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 25.39,
|
| 1320 |
+
"eval_loss": 0.9804235696792603,
|
| 1321 |
+
"eval_runtime": 731.7948,
|
| 1322 |
+
"eval_samples_per_second": 42.36,
|
| 1323 |
+
"eval_steps_per_second": 7.061,
|
| 1324 |
+
"step": 3760000
|
| 1325 |
+
},
|
| 1326 |
+
{
|
| 1327 |
+
"epoch": 25.66,
|
| 1328 |
+
"learning_rate": 5.411377143980351e-06,
|
| 1329 |
+
"loss": 1.0087,
|
| 1330 |
+
"step": 3800000
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 25.66,
|
| 1334 |
+
"eval_loss": 0.9776991009712219,
|
| 1335 |
+
"eval_runtime": 731.6622,
|
| 1336 |
+
"eval_samples_per_second": 42.368,
|
| 1337 |
+
"eval_steps_per_second": 7.062,
|
| 1338 |
+
"step": 3800000
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 25.93,
|
| 1342 |
+
"learning_rate": 5.351380955167535e-06,
|
| 1343 |
+
"loss": 1.009,
|
| 1344 |
+
"step": 3840000
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 25.93,
|
| 1348 |
+
"eval_loss": 0.9773043990135193,
|
| 1349 |
+
"eval_runtime": 731.5756,
|
| 1350 |
+
"eval_samples_per_second": 42.373,
|
| 1351 |
+
"eval_steps_per_second": 7.063,
|
| 1352 |
+
"step": 3840000
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 26.2,
|
| 1356 |
+
"learning_rate": 5.291383265887324e-06,
|
| 1357 |
+
"loss": 1.005,
|
| 1358 |
+
"step": 3880000
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 26.2,
|
| 1362 |
+
"eval_loss": 0.9772471785545349,
|
| 1363 |
+
"eval_runtime": 731.6739,
|
| 1364 |
+
"eval_samples_per_second": 42.367,
|
| 1365 |
+
"eval_steps_per_second": 7.062,
|
| 1366 |
+
"step": 3880000
|
| 1367 |
+
},
|
| 1368 |
+
{
|
| 1369 |
+
"epoch": 26.47,
|
| 1370 |
+
"learning_rate": 5.2313885775419045e-06,
|
| 1371 |
+
"loss": 1.0049,
|
| 1372 |
+
"step": 3920000
|
| 1373 |
+
},
|
| 1374 |
+
{
|
| 1375 |
+
"epoch": 26.47,
|
| 1376 |
+
"eval_loss": 0.9808975458145142,
|
| 1377 |
+
"eval_runtime": 731.4892,
|
| 1378 |
+
"eval_samples_per_second": 42.378,
|
| 1379 |
+
"eval_steps_per_second": 7.064,
|
| 1380 |
+
"step": 3920000
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"epoch": 26.74,
|
| 1384 |
+
"learning_rate": 5.171389387794299e-06,
|
| 1385 |
+
"loss": 1.0059,
|
| 1386 |
+
"step": 3960000
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 26.74,
|
| 1390 |
+
"eval_loss": 0.9739375710487366,
|
| 1391 |
+
"eval_runtime": 731.4074,
|
| 1392 |
+
"eval_samples_per_second": 42.383,
|
| 1393 |
+
"eval_steps_per_second": 7.064,
|
| 1394 |
+
"step": 3960000
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 27.01,
|
| 1398 |
+
"learning_rate": 5.111393198981483e-06,
|
| 1399 |
+
"loss": 1.0052,
|
| 1400 |
+
"step": 4000000
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 27.01,
|
| 1404 |
+
"eval_loss": 0.9740833640098572,
|
| 1405 |
+
"eval_runtime": 730.8502,
|
| 1406 |
+
"eval_samples_per_second": 42.415,
|
| 1407 |
+
"eval_steps_per_second": 7.07,
|
| 1408 |
+
"step": 4000000
|
| 1409 |
+
},
|
| 1410 |
+
{
|
| 1411 |
+
"epoch": 27.28,
|
| 1412 |
+
"learning_rate": 5.051397010168668e-06,
|
| 1413 |
+
"loss": 1.0011,
|
| 1414 |
+
"step": 4040000
|
| 1415 |
+
},
|
| 1416 |
+
{
|
| 1417 |
+
"epoch": 27.28,
|
| 1418 |
+
"eval_loss": 0.9713684916496277,
|
| 1419 |
+
"eval_runtime": 730.9391,
|
| 1420 |
+
"eval_samples_per_second": 42.41,
|
| 1421 |
+
"eval_steps_per_second": 7.069,
|
| 1422 |
+
"step": 4040000
|
| 1423 |
+
},
|
| 1424 |
+
{
|
| 1425 |
+
"epoch": 27.55,
|
| 1426 |
+
"learning_rate": 4.991400821355853e-06,
|
| 1427 |
+
"loss": 1.0007,
|
| 1428 |
+
"step": 4080000
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 27.55,
|
| 1432 |
+
"eval_loss": 0.9728115200996399,
|
| 1433 |
+
"eval_runtime": 730.9125,
|
| 1434 |
+
"eval_samples_per_second": 42.411,
|
| 1435 |
+
"eval_steps_per_second": 7.069,
|
| 1436 |
+
"step": 4080000
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 27.82,
|
| 1440 |
+
"learning_rate": 4.931401631608246e-06,
|
| 1441 |
+
"loss": 1.0029,
|
| 1442 |
+
"step": 4120000
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 27.82,
|
| 1446 |
+
"eval_loss": 0.9748100638389587,
|
| 1447 |
+
"eval_runtime": 730.8598,
|
| 1448 |
+
"eval_samples_per_second": 42.414,
|
| 1449 |
+
"eval_steps_per_second": 7.07,
|
| 1450 |
+
"step": 4120000
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"epoch": 28.09,
|
| 1454 |
+
"learning_rate": 4.871408443730222e-06,
|
| 1455 |
+
"loss": 1.0009,
|
| 1456 |
+
"step": 4160000
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 28.09,
|
| 1460 |
+
"eval_loss": 0.969658374786377,
|
| 1461 |
+
"eval_runtime": 730.837,
|
| 1462 |
+
"eval_samples_per_second": 42.416,
|
| 1463 |
+
"eval_steps_per_second": 7.07,
|
| 1464 |
+
"step": 4160000
|
| 1465 |
+
},
|
| 1466 |
+
{
|
| 1467 |
+
"epoch": 28.36,
|
| 1468 |
+
"learning_rate": 4.811412254917407e-06,
|
| 1469 |
+
"loss": 0.9986,
|
| 1470 |
+
"step": 4200000
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 28.36,
|
| 1474 |
+
"eval_loss": 0.9716975092887878,
|
| 1475 |
+
"eval_runtime": 730.9826,
|
| 1476 |
+
"eval_samples_per_second": 42.407,
|
| 1477 |
+
"eval_steps_per_second": 7.069,
|
| 1478 |
+
"step": 4200000
|
| 1479 |
+
},
|
| 1480 |
+
{
|
| 1481 |
+
"epoch": 28.63,
|
| 1482 |
+
"learning_rate": 4.751413065169801e-06,
|
| 1483 |
+
"loss": 0.9983,
|
| 1484 |
+
"step": 4240000
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 28.63,
|
| 1488 |
+
"eval_loss": 0.9703741073608398,
|
| 1489 |
+
"eval_runtime": 730.8801,
|
| 1490 |
+
"eval_samples_per_second": 42.413,
|
| 1491 |
+
"eval_steps_per_second": 7.07,
|
| 1492 |
+
"step": 4240000
|
| 1493 |
+
},
|
| 1494 |
+
{
|
| 1495 |
+
"epoch": 28.9,
|
| 1496 |
+
"learning_rate": 4.69141537588959e-06,
|
| 1497 |
+
"loss": 1.0006,
|
| 1498 |
+
"step": 4280000
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 28.9,
|
| 1502 |
+
"eval_loss": 0.9680435657501221,
|
| 1503 |
+
"eval_runtime": 730.9241,
|
| 1504 |
+
"eval_samples_per_second": 42.411,
|
| 1505 |
+
"eval_steps_per_second": 7.069,
|
| 1506 |
+
"step": 4280000
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 29.17,
|
| 1510 |
+
"learning_rate": 4.631419187076775e-06,
|
| 1511 |
+
"loss": 0.9981,
|
| 1512 |
+
"step": 4320000
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 29.17,
|
| 1516 |
+
"eval_loss": 0.9698230028152466,
|
| 1517 |
+
"eval_runtime": 730.9102,
|
| 1518 |
+
"eval_samples_per_second": 42.412,
|
| 1519 |
+
"eval_steps_per_second": 7.069,
|
| 1520 |
+
"step": 4320000
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 29.44,
|
| 1524 |
+
"learning_rate": 4.571419997329168e-06,
|
| 1525 |
+
"loss": 0.9955,
|
| 1526 |
+
"step": 4360000
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 29.44,
|
| 1530 |
+
"eval_loss": 0.9697496891021729,
|
| 1531 |
+
"eval_runtime": 730.9617,
|
| 1532 |
+
"eval_samples_per_second": 42.409,
|
| 1533 |
+
"eval_steps_per_second": 7.069,
|
| 1534 |
+
"step": 4360000
|
| 1535 |
+
},
|
| 1536 |
+
{
|
| 1537 |
+
"epoch": 29.71,
|
| 1538 |
+
"learning_rate": 4.5114238085163535e-06,
|
| 1539 |
+
"loss": 0.9965,
|
| 1540 |
+
"step": 4400000
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 29.71,
|
| 1544 |
+
"eval_loss": 0.9706552028656006,
|
| 1545 |
+
"eval_runtime": 730.9543,
|
| 1546 |
+
"eval_samples_per_second": 42.409,
|
| 1547 |
+
"eval_steps_per_second": 7.069,
|
| 1548 |
+
"step": 4400000
|
| 1549 |
+
},
|
| 1550 |
+
{
|
| 1551 |
+
"epoch": 29.98,
|
| 1552 |
+
"learning_rate": 4.4514261192361425e-06,
|
| 1553 |
+
"loss": 0.9978,
|
| 1554 |
+
"step": 4440000
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 29.98,
|
| 1558 |
+
"eval_loss": 0.9674049615859985,
|
| 1559 |
+
"eval_runtime": 731.0402,
|
| 1560 |
+
"eval_samples_per_second": 42.404,
|
| 1561 |
+
"eval_steps_per_second": 7.068,
|
| 1562 |
+
"step": 4440000
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 30.25,
|
| 1566 |
+
"learning_rate": 4.391428429955931e-06,
|
| 1567 |
+
"loss": 0.9921,
|
| 1568 |
+
"step": 4480000
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 30.25,
|
| 1572 |
+
"eval_loss": 0.9650031328201294,
|
| 1573 |
+
"eval_runtime": 730.9913,
|
| 1574 |
+
"eval_samples_per_second": 42.407,
|
| 1575 |
+
"eval_steps_per_second": 7.068,
|
| 1576 |
+
"step": 4480000
|
| 1577 |
+
},
|
| 1578 |
+
{
|
| 1579 |
+
"epoch": 30.52,
|
| 1580 |
+
"learning_rate": 4.331429240208325e-06,
|
| 1581 |
+
"loss": 0.9934,
|
| 1582 |
+
"step": 4520000
|
| 1583 |
+
},
|
| 1584 |
+
{
|
| 1585 |
+
"epoch": 30.52,
|
| 1586 |
+
"eval_loss": 0.9674308896064758,
|
| 1587 |
+
"eval_runtime": 732.3441,
|
| 1588 |
+
"eval_samples_per_second": 42.328,
|
| 1589 |
+
"eval_steps_per_second": 7.055,
|
| 1590 |
+
"step": 4520000
|
| 1591 |
+
},
|
| 1592 |
+
{
|
| 1593 |
+
"epoch": 30.79,
|
| 1594 |
+
"learning_rate": 4.271430050460719e-06,
|
| 1595 |
+
"loss": 0.9943,
|
| 1596 |
+
"step": 4560000
|
| 1597 |
+
},
|
| 1598 |
+
{
|
| 1599 |
+
"epoch": 30.79,
|
| 1600 |
+
"eval_loss": 0.9649813771247864,
|
| 1601 |
+
"eval_runtime": 731.9631,
|
| 1602 |
+
"eval_samples_per_second": 42.35,
|
| 1603 |
+
"eval_steps_per_second": 7.059,
|
| 1604 |
+
"step": 4560000
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 31.06,
|
| 1608 |
+
"learning_rate": 4.2114308607131125e-06,
|
| 1609 |
+
"loss": 0.9946,
|
| 1610 |
+
"step": 4600000
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 31.06,
|
| 1614 |
+
"eval_loss": 0.9624907374382019,
|
| 1615 |
+
"eval_runtime": 731.9597,
|
| 1616 |
+
"eval_samples_per_second": 42.351,
|
| 1617 |
+
"eval_steps_per_second": 7.059,
|
| 1618 |
+
"step": 4600000
|
| 1619 |
+
},
|
| 1620 |
+
{
|
| 1621 |
+
"epoch": 31.33,
|
| 1622 |
+
"learning_rate": 4.151434671900298e-06,
|
| 1623 |
+
"loss": 0.9917,
|
| 1624 |
+
"step": 4640000
|
| 1625 |
+
},
|
| 1626 |
+
{
|
| 1627 |
+
"epoch": 31.33,
|
| 1628 |
+
"eval_loss": 0.9647811055183411,
|
| 1629 |
+
"eval_runtime": 731.9469,
|
| 1630 |
+
"eval_samples_per_second": 42.351,
|
| 1631 |
+
"eval_steps_per_second": 7.059,
|
| 1632 |
+
"step": 4640000
|
| 1633 |
+
},
|
| 1634 |
+
{
|
| 1635 |
+
"epoch": 31.6,
|
| 1636 |
+
"learning_rate": 4.091436982620087e-06,
|
| 1637 |
+
"loss": 0.9902,
|
| 1638 |
+
"step": 4680000
|
| 1639 |
+
},
|
| 1640 |
+
{
|
| 1641 |
+
"epoch": 31.6,
|
| 1642 |
+
"eval_loss": 0.962338387966156,
|
| 1643 |
+
"eval_runtime": 731.9973,
|
| 1644 |
+
"eval_samples_per_second": 42.349,
|
| 1645 |
+
"eval_steps_per_second": 7.059,
|
| 1646 |
+
"step": 4680000
|
| 1647 |
+
},
|
| 1648 |
+
{
|
| 1649 |
+
"epoch": 31.87,
|
| 1650 |
+
"learning_rate": 4.031442294274667e-06,
|
| 1651 |
+
"loss": 0.9918,
|
| 1652 |
+
"step": 4720000
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 31.87,
|
| 1656 |
+
"eval_loss": 0.9640004634857178,
|
| 1657 |
+
"eval_runtime": 723.6956,
|
| 1658 |
+
"eval_samples_per_second": 42.834,
|
| 1659 |
+
"eval_steps_per_second": 7.14,
|
| 1660 |
+
"step": 4720000
|
| 1661 |
+
},
|
| 1662 |
+
{
|
| 1663 |
+
"epoch": 32.14,
|
| 1664 |
+
"learning_rate": 3.9714416040596645e-06,
|
| 1665 |
+
"loss": 0.9897,
|
| 1666 |
+
"step": 4760000
|
| 1667 |
+
},
|
| 1668 |
+
{
|
| 1669 |
+
"epoch": 32.14,
|
| 1670 |
+
"eval_loss": 0.961030125617981,
|
| 1671 |
+
"eval_runtime": 723.561,
|
| 1672 |
+
"eval_samples_per_second": 42.842,
|
| 1673 |
+
"eval_steps_per_second": 7.141,
|
| 1674 |
+
"step": 4760000
|
| 1675 |
+
},
|
| 1676 |
+
{
|
| 1677 |
+
"epoch": 32.41,
|
| 1678 |
+
"learning_rate": 3.91144541524685e-06,
|
| 1679 |
+
"loss": 0.9886,
|
| 1680 |
+
"step": 4800000
|
| 1681 |
+
},
|
| 1682 |
+
{
|
| 1683 |
+
"epoch": 32.41,
|
| 1684 |
+
"eval_loss": 0.9641813039779663,
|
| 1685 |
+
"eval_runtime": 723.7468,
|
| 1686 |
+
"eval_samples_per_second": 42.831,
|
| 1687 |
+
"eval_steps_per_second": 7.139,
|
| 1688 |
+
"step": 4800000
|
| 1689 |
+
},
|
| 1690 |
+
{
|
| 1691 |
+
"epoch": 32.68,
|
| 1692 |
+
"learning_rate": 3.8514477259666385e-06,
|
| 1693 |
+
"loss": 0.9892,
|
| 1694 |
+
"step": 4840000
|
| 1695 |
+
},
|
| 1696 |
+
{
|
| 1697 |
+
"epoch": 32.68,
|
| 1698 |
+
"eval_loss": 0.9588494896888733,
|
| 1699 |
+
"eval_runtime": 723.6649,
|
| 1700 |
+
"eval_samples_per_second": 42.836,
|
| 1701 |
+
"eval_steps_per_second": 7.14,
|
| 1702 |
+
"step": 4840000
|
| 1703 |
+
},
|
| 1704 |
+
{
|
| 1705 |
+
"epoch": 32.95,
|
| 1706 |
+
"learning_rate": 3.7914485362190323e-06,
|
| 1707 |
+
"loss": 0.9885,
|
| 1708 |
+
"step": 4880000
|
| 1709 |
+
},
|
| 1710 |
+
{
|
| 1711 |
+
"epoch": 32.95,
|
| 1712 |
+
"eval_loss": 0.959790050983429,
|
| 1713 |
+
"eval_runtime": 731.8929,
|
| 1714 |
+
"eval_samples_per_second": 42.355,
|
| 1715 |
+
"eval_steps_per_second": 7.06,
|
| 1716 |
+
"step": 4880000
|
| 1717 |
+
},
|
| 1718 |
+
{
|
| 1719 |
+
"epoch": 33.22,
|
| 1720 |
+
"learning_rate": 3.7314508469388216e-06,
|
| 1721 |
+
"loss": 0.9867,
|
| 1722 |
+
"step": 4920000
|
| 1723 |
+
},
|
| 1724 |
+
{
|
| 1725 |
+
"epoch": 33.22,
|
| 1726 |
+
"eval_loss": 0.9605200290679932,
|
| 1727 |
+
"eval_runtime": 731.9065,
|
| 1728 |
+
"eval_samples_per_second": 42.354,
|
| 1729 |
+
"eval_steps_per_second": 7.06,
|
| 1730 |
+
"step": 4920000
|
| 1731 |
+
},
|
| 1732 |
+
{
|
| 1733 |
+
"epoch": 33.49,
|
| 1734 |
+
"learning_rate": 3.6714546581260063e-06,
|
| 1735 |
+
"loss": 0.9851,
|
| 1736 |
+
"step": 4960000
|
| 1737 |
+
},
|
| 1738 |
+
{
|
| 1739 |
+
"epoch": 33.49,
|
| 1740 |
+
"eval_loss": 0.9623289704322815,
|
| 1741 |
+
"eval_runtime": 731.9518,
|
| 1742 |
+
"eval_samples_per_second": 42.351,
|
| 1743 |
+
"eval_steps_per_second": 7.059,
|
| 1744 |
+
"step": 4960000
|
| 1745 |
+
},
|
| 1746 |
+
{
|
| 1747 |
+
"epoch": 33.76,
|
| 1748 |
+
"learning_rate": 3.6114539679110043e-06,
|
| 1749 |
+
"loss": 0.9865,
|
| 1750 |
+
"step": 5000000
|
| 1751 |
+
},
|
| 1752 |
+
{
|
| 1753 |
+
"epoch": 33.76,
|
| 1754 |
+
"eval_loss": 0.9566452503204346,
|
| 1755 |
+
"eval_runtime": 731.9347,
|
| 1756 |
+
"eval_samples_per_second": 42.352,
|
| 1757 |
+
"eval_steps_per_second": 7.059,
|
| 1758 |
+
"step": 5000000
|
| 1759 |
+
},
|
| 1760 |
+
{
|
| 1761 |
+
"epoch": 34.03,
|
| 1762 |
+
"learning_rate": 3.5514562786307937e-06,
|
| 1763 |
+
"loss": 0.9847,
|
| 1764 |
+
"step": 5040000
|
| 1765 |
+
},
|
| 1766 |
+
{
|
| 1767 |
+
"epoch": 34.03,
|
| 1768 |
+
"eval_loss": 0.9604076743125916,
|
| 1769 |
+
"eval_runtime": 731.8939,
|
| 1770 |
+
"eval_samples_per_second": 42.354,
|
| 1771 |
+
"eval_steps_per_second": 7.06,
|
| 1772 |
+
"step": 5040000
|
| 1773 |
+
},
|
| 1774 |
+
{
|
| 1775 |
+
"epoch": 34.3,
|
| 1776 |
+
"learning_rate": 3.4914570888831874e-06,
|
| 1777 |
+
"loss": 0.9813,
|
| 1778 |
+
"step": 5080000
|
| 1779 |
+
},
|
| 1780 |
+
{
|
| 1781 |
+
"epoch": 34.3,
|
| 1782 |
+
"eval_loss": 0.9602544903755188,
|
| 1783 |
+
"eval_runtime": 731.8897,
|
| 1784 |
+
"eval_samples_per_second": 42.355,
|
| 1785 |
+
"eval_steps_per_second": 7.06,
|
| 1786 |
+
"step": 5080000
|
| 1787 |
+
},
|
| 1788 |
+
{
|
| 1789 |
+
"epoch": 34.57,
|
| 1790 |
+
"learning_rate": 3.431457899135581e-06,
|
| 1791 |
+
"loss": 0.9838,
|
| 1792 |
+
"step": 5120000
|
| 1793 |
+
},
|
| 1794 |
+
{
|
| 1795 |
+
"epoch": 34.57,
|
| 1796 |
+
"eval_loss": 0.9588419198989868,
|
| 1797 |
+
"eval_runtime": 731.8728,
|
| 1798 |
+
"eval_samples_per_second": 42.356,
|
| 1799 |
+
"eval_steps_per_second": 7.06,
|
| 1800 |
+
"step": 5120000
|
| 1801 |
+
},
|
| 1802 |
+
{
|
| 1803 |
+
"epoch": 34.84,
|
| 1804 |
+
"learning_rate": 3.37146020985537e-06,
|
| 1805 |
+
"loss": 0.9842,
|
| 1806 |
+
"step": 5160000
|
| 1807 |
+
},
|
| 1808 |
+
{
|
| 1809 |
+
"epoch": 34.84,
|
| 1810 |
+
"eval_loss": 0.956282377243042,
|
| 1811 |
+
"eval_runtime": 731.8168,
|
| 1812 |
+
"eval_samples_per_second": 42.359,
|
| 1813 |
+
"eval_steps_per_second": 7.061,
|
| 1814 |
+
"step": 5160000
|
| 1815 |
+
},
|
| 1816 |
+
{
|
| 1817 |
+
"epoch": 35.11,
|
| 1818 |
+
"learning_rate": 3.311464021042555e-06,
|
| 1819 |
+
"loss": 0.982,
|
| 1820 |
+
"step": 5200000
|
| 1821 |
+
},
|
| 1822 |
+
{
|
| 1823 |
+
"epoch": 35.11,
|
| 1824 |
+
"eval_loss": 0.9570952653884888,
|
| 1825 |
+
"eval_runtime": 731.9288,
|
| 1826 |
+
"eval_samples_per_second": 42.352,
|
| 1827 |
+
"eval_steps_per_second": 7.059,
|
| 1828 |
+
"step": 5200000
|
| 1829 |
+
},
|
| 1830 |
+
{
|
| 1831 |
+
"epoch": 35.38,
|
| 1832 |
+
"learning_rate": 3.251464831294949e-06,
|
| 1833 |
+
"loss": 0.9803,
|
| 1834 |
+
"step": 5240000
|
| 1835 |
+
},
|
| 1836 |
+
{
|
| 1837 |
+
"epoch": 35.38,
|
| 1838 |
+
"eval_loss": 0.9592018127441406,
|
| 1839 |
+
"eval_runtime": 731.9144,
|
| 1840 |
+
"eval_samples_per_second": 42.353,
|
| 1841 |
+
"eval_steps_per_second": 7.06,
|
| 1842 |
+
"step": 5240000
|
| 1843 |
+
},
|
| 1844 |
+
{
|
| 1845 |
+
"epoch": 35.65,
|
| 1846 |
+
"learning_rate": 3.1914671420147383e-06,
|
| 1847 |
+
"loss": 0.9804,
|
| 1848 |
+
"step": 5280000
|
| 1849 |
+
},
|
| 1850 |
+
{
|
| 1851 |
+
"epoch": 35.65,
|
| 1852 |
+
"eval_loss": 0.9540479779243469,
|
| 1853 |
+
"eval_runtime": 731.9484,
|
| 1854 |
+
"eval_samples_per_second": 42.351,
|
| 1855 |
+
"eval_steps_per_second": 7.059,
|
| 1856 |
+
"step": 5280000
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 35.92,
|
| 1860 |
+
"learning_rate": 3.131469452734527e-06,
|
| 1861 |
+
"loss": 0.9799,
|
| 1862 |
+
"step": 5320000
|
| 1863 |
+
},
|
| 1864 |
+
{
|
| 1865 |
+
"epoch": 35.92,
|
| 1866 |
+
"eval_loss": 0.9547669887542725,
|
| 1867 |
+
"eval_runtime": 731.8909,
|
| 1868 |
+
"eval_samples_per_second": 42.355,
|
| 1869 |
+
"eval_steps_per_second": 7.06,
|
| 1870 |
+
"step": 5320000
|
| 1871 |
+
},
|
| 1872 |
+
{
|
| 1873 |
+
"epoch": 36.19,
|
| 1874 |
+
"learning_rate": 3.0714732639217123e-06,
|
| 1875 |
+
"loss": 0.9793,
|
| 1876 |
+
"step": 5360000
|
| 1877 |
+
},
|
| 1878 |
+
{
|
| 1879 |
+
"epoch": 36.19,
|
| 1880 |
+
"eval_loss": 0.9536721706390381,
|
| 1881 |
+
"eval_runtime": 731.8644,
|
| 1882 |
+
"eval_samples_per_second": 42.356,
|
| 1883 |
+
"eval_steps_per_second": 7.06,
|
| 1884 |
+
"step": 5360000
|
| 1885 |
+
},
|
| 1886 |
+
{
|
| 1887 |
+
"epoch": 36.46,
|
| 1888 |
+
"learning_rate": 3.011475574641501e-06,
|
| 1889 |
+
"loss": 0.9783,
|
| 1890 |
+
"step": 5400000
|
| 1891 |
+
},
|
| 1892 |
+
{
|
| 1893 |
+
"epoch": 36.46,
|
| 1894 |
+
"eval_loss": 0.9515364766120911,
|
| 1895 |
+
"eval_runtime": 731.9101,
|
| 1896 |
+
"eval_samples_per_second": 42.354,
|
| 1897 |
+
"eval_steps_per_second": 7.06,
|
| 1898 |
+
"step": 5400000
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 36.73,
|
| 1902 |
+
"learning_rate": 2.951476384893895e-06,
|
| 1903 |
+
"loss": 0.9772,
|
| 1904 |
+
"step": 5440000
|
| 1905 |
+
},
|
| 1906 |
+
{
|
| 1907 |
+
"epoch": 36.73,
|
| 1908 |
+
"eval_loss": 0.95276939868927,
|
| 1909 |
+
"eval_runtime": 731.8184,
|
| 1910 |
+
"eval_samples_per_second": 42.359,
|
| 1911 |
+
"eval_steps_per_second": 7.06,
|
| 1912 |
+
"step": 5440000
|
| 1913 |
+
},
|
| 1914 |
+
{
|
| 1915 |
+
"epoch": 37.0,
|
| 1916 |
+
"learning_rate": 2.8914801960810796e-06,
|
| 1917 |
+
"loss": 0.9804,
|
| 1918 |
+
"step": 5480000
|
| 1919 |
+
},
|
| 1920 |
+
{
|
| 1921 |
+
"epoch": 37.0,
|
| 1922 |
+
"eval_loss": 0.9519599676132202,
|
| 1923 |
+
"eval_runtime": 731.9107,
|
| 1924 |
+
"eval_samples_per_second": 42.354,
|
| 1925 |
+
"eval_steps_per_second": 7.06,
|
| 1926 |
+
"step": 5480000
|
| 1927 |
+
},
|
| 1928 |
+
{
|
| 1929 |
+
"epoch": 37.27,
|
| 1930 |
+
"learning_rate": 2.83148550773566e-06,
|
| 1931 |
+
"loss": 0.9745,
|
| 1932 |
+
"step": 5520000
|
| 1933 |
+
},
|
| 1934 |
+
{
|
| 1935 |
+
"epoch": 37.27,
|
| 1936 |
+
"eval_loss": 0.9534033536911011,
|
| 1937 |
+
"eval_runtime": 731.9292,
|
| 1938 |
+
"eval_samples_per_second": 42.352,
|
| 1939 |
+
"eval_steps_per_second": 7.059,
|
| 1940 |
+
"step": 5520000
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 37.54,
|
| 1944 |
+
"learning_rate": 2.7714863179880536e-06,
|
| 1945 |
+
"loss": 0.9757,
|
| 1946 |
+
"step": 5560000
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"epoch": 37.54,
|
| 1950 |
+
"eval_loss": 0.9531847238540649,
|
| 1951 |
+
"eval_runtime": 731.8616,
|
| 1952 |
+
"eval_samples_per_second": 42.356,
|
| 1953 |
+
"eval_steps_per_second": 7.06,
|
| 1954 |
+
"step": 5560000
|
| 1955 |
+
},
|
| 1956 |
+
{
|
| 1957 |
+
"epoch": 37.81,
|
| 1958 |
+
"learning_rate": 2.7114901291752383e-06,
|
| 1959 |
+
"loss": 0.9774,
|
| 1960 |
+
"step": 5600000
|
| 1961 |
+
},
|
| 1962 |
+
{
|
| 1963 |
+
"epoch": 37.81,
|
| 1964 |
+
"eval_loss": 0.9502018690109253,
|
| 1965 |
+
"eval_runtime": 731.8317,
|
| 1966 |
+
"eval_samples_per_second": 42.358,
|
| 1967 |
+
"eval_steps_per_second": 7.06,
|
| 1968 |
+
"step": 5600000
|
| 1969 |
+
},
|
| 1970 |
+
{
|
| 1971 |
+
"epoch": 38.08,
|
| 1972 |
+
"learning_rate": 2.6514909394276316e-06,
|
| 1973 |
+
"loss": 0.9753,
|
| 1974 |
+
"step": 5640000
|
| 1975 |
+
},
|
| 1976 |
+
{
|
| 1977 |
+
"epoch": 38.08,
|
| 1978 |
+
"eval_loss": 0.9478103518486023,
|
| 1979 |
+
"eval_runtime": 731.8606,
|
| 1980 |
+
"eval_samples_per_second": 42.356,
|
| 1981 |
+
"eval_steps_per_second": 7.06,
|
| 1982 |
+
"step": 5640000
|
| 1983 |
+
},
|
| 1984 |
+
{
|
| 1985 |
+
"epoch": 38.35,
|
| 1986 |
+
"learning_rate": 2.591494750614817e-06,
|
| 1987 |
+
"loss": 0.9718,
|
| 1988 |
+
"step": 5680000
|
| 1989 |
+
},
|
| 1990 |
+
{
|
| 1991 |
+
"epoch": 38.35,
|
| 1992 |
+
"eval_loss": 0.95122891664505,
|
| 1993 |
+
"eval_runtime": 731.8255,
|
| 1994 |
+
"eval_samples_per_second": 42.358,
|
| 1995 |
+
"eval_steps_per_second": 7.06,
|
| 1996 |
+
"step": 5680000
|
| 1997 |
+
},
|
| 1998 |
+
{
|
| 1999 |
+
"epoch": 38.62,
|
| 2000 |
+
"learning_rate": 2.5314970613346056e-06,
|
| 2001 |
+
"loss": 0.9734,
|
| 2002 |
+
"step": 5720000
|
| 2003 |
+
},
|
| 2004 |
+
{
|
| 2005 |
+
"epoch": 38.62,
|
| 2006 |
+
"eval_loss": 0.9479803442955017,
|
| 2007 |
+
"eval_runtime": 731.8289,
|
| 2008 |
+
"eval_samples_per_second": 42.358,
|
| 2009 |
+
"eval_steps_per_second": 7.06,
|
| 2010 |
+
"step": 5720000
|
| 2011 |
+
},
|
| 2012 |
+
{
|
| 2013 |
+
"epoch": 38.89,
|
| 2014 |
+
"learning_rate": 2.4715023729891864e-06,
|
| 2015 |
+
"loss": 0.9745,
|
| 2016 |
+
"step": 5760000
|
| 2017 |
+
},
|
| 2018 |
+
{
|
| 2019 |
+
"epoch": 38.89,
|
| 2020 |
+
"eval_loss": 0.9491930603981018,
|
| 2021 |
+
"eval_runtime": 731.856,
|
| 2022 |
+
"eval_samples_per_second": 42.357,
|
| 2023 |
+
"eval_steps_per_second": 7.06,
|
| 2024 |
+
"step": 5760000
|
| 2025 |
+
},
|
| 2026 |
+
{
|
| 2027 |
+
"epoch": 39.16,
|
| 2028 |
+
"learning_rate": 2.4115046837089757e-06,
|
| 2029 |
+
"loss": 0.9712,
|
| 2030 |
+
"step": 5800000
|
| 2031 |
+
},
|
| 2032 |
+
{
|
| 2033 |
+
"epoch": 39.16,
|
| 2034 |
+
"eval_loss": 0.949318528175354,
|
| 2035 |
+
"eval_runtime": 731.8365,
|
| 2036 |
+
"eval_samples_per_second": 42.358,
|
| 2037 |
+
"eval_steps_per_second": 7.06,
|
| 2038 |
+
"step": 5800000
|
| 2039 |
+
},
|
| 2040 |
+
{
|
| 2041 |
+
"epoch": 39.43,
|
| 2042 |
+
"learning_rate": 2.3515084948961604e-06,
|
| 2043 |
+
"loss": 0.97,
|
| 2044 |
+
"step": 5840000
|
| 2045 |
+
},
|
| 2046 |
+
{
|
| 2047 |
+
"epoch": 39.43,
|
| 2048 |
+
"eval_loss": 0.9494725465774536,
|
| 2049 |
+
"eval_runtime": 731.7796,
|
| 2050 |
+
"eval_samples_per_second": 42.361,
|
| 2051 |
+
"eval_steps_per_second": 7.061,
|
| 2052 |
+
"step": 5840000
|
| 2053 |
+
},
|
| 2054 |
+
{
|
| 2055 |
+
"epoch": 39.7,
|
| 2056 |
+
"learning_rate": 2.291512306083345e-06,
|
| 2057 |
+
"loss": 0.9719,
|
| 2058 |
+
"step": 5880000
|
| 2059 |
+
},
|
| 2060 |
+
{
|
| 2061 |
+
"epoch": 39.7,
|
| 2062 |
+
"eval_loss": 0.9493642449378967,
|
| 2063 |
+
"eval_runtime": 731.7838,
|
| 2064 |
+
"eval_samples_per_second": 42.361,
|
| 2065 |
+
"eval_steps_per_second": 7.061,
|
| 2066 |
+
"step": 5880000
|
| 2067 |
+
},
|
| 2068 |
+
{
|
| 2069 |
+
"epoch": 39.97,
|
| 2070 |
+
"learning_rate": 2.2315131163357387e-06,
|
| 2071 |
+
"loss": 0.9725,
|
| 2072 |
+
"step": 5920000
|
| 2073 |
+
},
|
| 2074 |
+
{
|
| 2075 |
+
"epoch": 39.97,
|
| 2076 |
+
"eval_loss": 0.945613443851471,
|
| 2077 |
+
"eval_runtime": 731.8091,
|
| 2078 |
+
"eval_samples_per_second": 42.359,
|
| 2079 |
+
"eval_steps_per_second": 7.061,
|
| 2080 |
+
"step": 5920000
|
| 2081 |
+
},
|
| 2082 |
+
{
|
| 2083 |
+
"epoch": 40.24,
|
| 2084 |
+
"learning_rate": 2.171515427055528e-06,
|
| 2085 |
+
"loss": 0.9692,
|
| 2086 |
+
"step": 5960000
|
| 2087 |
+
},
|
| 2088 |
+
{
|
| 2089 |
+
"epoch": 40.24,
|
| 2090 |
+
"eval_loss": 0.9440425634384155,
|
| 2091 |
+
"eval_runtime": 731.7685,
|
| 2092 |
+
"eval_samples_per_second": 42.362,
|
| 2093 |
+
"eval_steps_per_second": 7.061,
|
| 2094 |
+
"step": 5960000
|
| 2095 |
+
},
|
| 2096 |
+
{
|
| 2097 |
+
"epoch": 40.51,
|
| 2098 |
+
"learning_rate": 2.111517737775317e-06,
|
| 2099 |
+
"loss": 0.9686,
|
| 2100 |
+
"step": 6000000
|
| 2101 |
+
},
|
| 2102 |
+
{
|
| 2103 |
+
"epoch": 40.51,
|
| 2104 |
+
"eval_loss": 0.9432018399238586,
|
| 2105 |
+
"eval_runtime": 731.8797,
|
| 2106 |
+
"eval_samples_per_second": 42.355,
|
| 2107 |
+
"eval_steps_per_second": 7.06,
|
| 2108 |
+
"step": 6000000
|
| 2109 |
+
},
|
| 2110 |
+
{
|
| 2111 |
+
"epoch": 40.78,
|
| 2112 |
+
"learning_rate": 2.051523049429898e-06,
|
| 2113 |
+
"loss": 0.9691,
|
| 2114 |
+
"step": 6040000
|
| 2115 |
+
},
|
| 2116 |
+
{
|
| 2117 |
+
"epoch": 40.78,
|
| 2118 |
+
"eval_loss": 0.9459187388420105,
|
| 2119 |
+
"eval_runtime": 731.8926,
|
| 2120 |
+
"eval_samples_per_second": 42.355,
|
| 2121 |
+
"eval_steps_per_second": 7.06,
|
| 2122 |
+
"step": 6040000
|
| 2123 |
+
},
|
| 2124 |
+
{
|
| 2125 |
+
"epoch": 41.05,
|
| 2126 |
+
"learning_rate": 1.991523859682291e-06,
|
| 2127 |
+
"loss": 0.9677,
|
| 2128 |
+
"step": 6080000
|
| 2129 |
+
},
|
| 2130 |
+
{
|
| 2131 |
+
"epoch": 41.05,
|
| 2132 |
+
"eval_loss": 0.9420450329780579,
|
| 2133 |
+
"eval_runtime": 731.8945,
|
| 2134 |
+
"eval_samples_per_second": 42.354,
|
| 2135 |
+
"eval_steps_per_second": 7.06,
|
| 2136 |
+
"step": 6080000
|
| 2137 |
+
},
|
| 2138 |
+
{
|
| 2139 |
+
"epoch": 41.32,
|
| 2140 |
+
"learning_rate": 1.931527670869476e-06,
|
| 2141 |
+
"loss": 0.9677,
|
| 2142 |
+
"step": 6120000
|
| 2143 |
+
},
|
| 2144 |
+
{
|
| 2145 |
+
"epoch": 41.32,
|
| 2146 |
+
"eval_loss": 0.945288360118866,
|
| 2147 |
+
"eval_runtime": 731.7905,
|
| 2148 |
+
"eval_samples_per_second": 42.36,
|
| 2149 |
+
"eval_steps_per_second": 7.061,
|
| 2150 |
+
"step": 6120000
|
| 2151 |
+
},
|
| 2152 |
+
{
|
| 2153 |
+
"epoch": 41.59,
|
| 2154 |
+
"learning_rate": 1.8715299815892652e-06,
|
| 2155 |
+
"loss": 0.9661,
|
| 2156 |
+
"step": 6160000
|
| 2157 |
+
},
|
| 2158 |
+
{
|
| 2159 |
+
"epoch": 41.59,
|
| 2160 |
+
"eval_loss": 0.9455643892288208,
|
| 2161 |
+
"eval_runtime": 731.7869,
|
| 2162 |
+
"eval_samples_per_second": 42.361,
|
| 2163 |
+
"eval_steps_per_second": 7.061,
|
| 2164 |
+
"step": 6160000
|
| 2165 |
+
},
|
| 2166 |
+
{
|
| 2167 |
+
"epoch": 41.86,
|
| 2168 |
+
"learning_rate": 1.8115352932438457e-06,
|
| 2169 |
+
"loss": 0.9664,
|
| 2170 |
+
"step": 6200000
|
| 2171 |
+
},
|
| 2172 |
+
{
|
| 2173 |
+
"epoch": 41.86,
|
| 2174 |
+
"eval_loss": 0.9422553181648254,
|
| 2175 |
+
"eval_runtime": 731.8325,
|
| 2176 |
+
"eval_samples_per_second": 42.358,
|
| 2177 |
+
"eval_steps_per_second": 7.06,
|
| 2178 |
+
"step": 6200000
|
| 2179 |
+
},
|
| 2180 |
+
{
|
| 2181 |
+
"epoch": 42.13,
|
| 2182 |
+
"learning_rate": 1.7515361034962392e-06,
|
| 2183 |
+
"loss": 0.9653,
|
| 2184 |
+
"step": 6240000
|
| 2185 |
+
},
|
| 2186 |
+
{
|
| 2187 |
+
"epoch": 42.13,
|
| 2188 |
+
"eval_loss": 0.9434319734573364,
|
| 2189 |
+
"eval_runtime": 731.8054,
|
| 2190 |
+
"eval_samples_per_second": 42.36,
|
| 2191 |
+
"eval_steps_per_second": 7.061,
|
| 2192 |
+
"step": 6240000
|
| 2193 |
+
},
|
| 2194 |
+
{
|
| 2195 |
+
"epoch": 42.4,
|
| 2196 |
+
"learning_rate": 1.6915429156182154e-06,
|
| 2197 |
+
"loss": 0.9633,
|
| 2198 |
+
"step": 6280000
|
| 2199 |
+
},
|
| 2200 |
+
{
|
| 2201 |
+
"epoch": 42.4,
|
| 2202 |
+
"eval_loss": 0.9410218000411987,
|
| 2203 |
+
"eval_runtime": 731.7099,
|
| 2204 |
+
"eval_samples_per_second": 42.365,
|
| 2205 |
+
"eval_steps_per_second": 7.062,
|
| 2206 |
+
"step": 6280000
|
| 2207 |
+
},
|
| 2208 |
+
{
|
| 2209 |
+
"epoch": 42.67,
|
| 2210 |
+
"learning_rate": 1.6315452263380044e-06,
|
| 2211 |
+
"loss": 0.9699,
|
| 2212 |
+
"step": 6320000
|
| 2213 |
+
},
|
| 2214 |
+
{
|
| 2215 |
+
"epoch": 42.67,
|
| 2216 |
+
"eval_loss": 0.9450401663780212,
|
| 2217 |
+
"eval_runtime": 731.4804,
|
| 2218 |
+
"eval_samples_per_second": 42.378,
|
| 2219 |
+
"eval_steps_per_second": 7.064,
|
| 2220 |
+
"step": 6320000
|
| 2221 |
+
},
|
| 2222 |
+
{
|
| 2223 |
+
"epoch": 42.94,
|
| 2224 |
+
"learning_rate": 1.571546036590398e-06,
|
| 2225 |
+
"loss": 0.9673,
|
| 2226 |
+
"step": 6360000
|
| 2227 |
+
},
|
| 2228 |
+
{
|
| 2229 |
+
"epoch": 42.94,
|
| 2230 |
+
"eval_loss": 0.9430739283561707,
|
| 2231 |
+
"eval_runtime": 731.5218,
|
| 2232 |
+
"eval_samples_per_second": 42.376,
|
| 2233 |
+
"eval_steps_per_second": 7.063,
|
| 2234 |
+
"step": 6360000
|
| 2235 |
+
},
|
| 2236 |
+
{
|
| 2237 |
+
"epoch": 43.21,
|
| 2238 |
+
"learning_rate": 1.5115513482449784e-06,
|
| 2239 |
+
"loss": 0.9635,
|
| 2240 |
+
"step": 6400000
|
| 2241 |
+
},
|
| 2242 |
+
{
|
| 2243 |
+
"epoch": 43.21,
|
| 2244 |
+
"eval_loss": 0.9403471350669861,
|
| 2245 |
+
"eval_runtime": 731.5739,
|
| 2246 |
+
"eval_samples_per_second": 42.373,
|
| 2247 |
+
"eval_steps_per_second": 7.063,
|
| 2248 |
+
"step": 6400000
|
| 2249 |
+
},
|
| 2250 |
+
{
|
| 2251 |
+
"epoch": 43.48,
|
| 2252 |
+
"learning_rate": 1.4515536589647678e-06,
|
| 2253 |
+
"loss": 0.9643,
|
| 2254 |
+
"step": 6440000
|
| 2255 |
+
},
|
| 2256 |
+
{
|
| 2257 |
+
"epoch": 43.48,
|
| 2258 |
+
"eval_loss": 0.9411180019378662,
|
| 2259 |
+
"eval_runtime": 731.5857,
|
| 2260 |
+
"eval_samples_per_second": 42.372,
|
| 2261 |
+
"eval_steps_per_second": 7.063,
|
| 2262 |
+
"step": 6440000
|
| 2263 |
+
},
|
| 2264 |
+
{
|
| 2265 |
+
"epoch": 43.75,
|
| 2266 |
+
"learning_rate": 1.3915604710867436e-06,
|
| 2267 |
+
"loss": 0.9629,
|
| 2268 |
+
"step": 6480000
|
| 2269 |
+
},
|
| 2270 |
+
{
|
| 2271 |
+
"epoch": 43.75,
|
| 2272 |
+
"eval_loss": 0.942000150680542,
|
| 2273 |
+
"eval_runtime": 731.7059,
|
| 2274 |
+
"eval_samples_per_second": 42.365,
|
| 2275 |
+
"eval_steps_per_second": 7.062,
|
| 2276 |
+
"step": 6480000
|
| 2277 |
+
},
|
| 2278 |
+
{
|
| 2279 |
+
"epoch": 44.02,
|
| 2280 |
+
"learning_rate": 1.3315612813391373e-06,
|
| 2281 |
+
"loss": 0.9636,
|
| 2282 |
+
"step": 6520000
|
| 2283 |
+
},
|
| 2284 |
+
{
|
| 2285 |
+
"epoch": 44.02,
|
| 2286 |
+
"eval_loss": 0.9400261640548706,
|
| 2287 |
+
"eval_runtime": 731.6205,
|
| 2288 |
+
"eval_samples_per_second": 42.37,
|
| 2289 |
+
"eval_steps_per_second": 7.062,
|
| 2290 |
+
"step": 6520000
|
| 2291 |
+
},
|
| 2292 |
+
{
|
| 2293 |
+
"epoch": 44.29,
|
| 2294 |
+
"learning_rate": 1.2715620915915308e-06,
|
| 2295 |
+
"loss": 0.9603,
|
| 2296 |
+
"step": 6560000
|
| 2297 |
+
},
|
| 2298 |
+
{
|
| 2299 |
+
"epoch": 44.29,
|
| 2300 |
+
"eval_loss": 0.9395481944084167,
|
| 2301 |
+
"eval_runtime": 731.6787,
|
| 2302 |
+
"eval_samples_per_second": 42.367,
|
| 2303 |
+
"eval_steps_per_second": 7.062,
|
| 2304 |
+
"step": 6560000
|
| 2305 |
+
},
|
| 2306 |
+
{
|
| 2307 |
+
"epoch": 44.56,
|
| 2308 |
+
"learning_rate": 1.2115659027787157e-06,
|
| 2309 |
+
"loss": 0.9623,
|
| 2310 |
+
"step": 6600000
|
| 2311 |
+
},
|
| 2312 |
+
{
|
| 2313 |
+
"epoch": 44.56,
|
| 2314 |
+
"eval_loss": 0.9395203590393066,
|
| 2315 |
+
"eval_runtime": 731.6894,
|
| 2316 |
+
"eval_samples_per_second": 42.366,
|
| 2317 |
+
"eval_steps_per_second": 7.062,
|
| 2318 |
+
"step": 6600000
|
| 2319 |
+
},
|
| 2320 |
+
{
|
| 2321 |
+
"epoch": 44.83,
|
| 2322 |
+
"learning_rate": 1.1515682134985048e-06,
|
| 2323 |
+
"loss": 0.9631,
|
| 2324 |
+
"step": 6640000
|
| 2325 |
+
},
|
| 2326 |
+
{
|
| 2327 |
+
"epoch": 44.83,
|
| 2328 |
+
"eval_loss": 0.9380507469177246,
|
| 2329 |
+
"eval_runtime": 731.5615,
|
| 2330 |
+
"eval_samples_per_second": 42.374,
|
| 2331 |
+
"eval_steps_per_second": 7.063,
|
| 2332 |
+
"step": 6640000
|
| 2333 |
+
},
|
| 2334 |
+
{
|
| 2335 |
+
"epoch": 45.1,
|
| 2336 |
+
"learning_rate": 1.0915720246856897e-06,
|
| 2337 |
+
"loss": 0.9592,
|
| 2338 |
+
"step": 6680000
|
| 2339 |
+
},
|
| 2340 |
+
{
|
| 2341 |
+
"epoch": 45.1,
|
| 2342 |
+
"eval_loss": 0.9388097524642944,
|
| 2343 |
+
"eval_runtime": 731.6793,
|
| 2344 |
+
"eval_samples_per_second": 42.367,
|
| 2345 |
+
"eval_steps_per_second": 7.062,
|
| 2346 |
+
"step": 6680000
|
| 2347 |
+
},
|
| 2348 |
+
{
|
| 2349 |
+
"epoch": 45.37,
|
| 2350 |
+
"learning_rate": 1.0315758358728746e-06,
|
| 2351 |
+
"loss": 0.9597,
|
| 2352 |
+
"step": 6720000
|
| 2353 |
+
},
|
| 2354 |
+
{
|
| 2355 |
+
"epoch": 45.37,
|
| 2356 |
+
"eval_loss": 0.9375278949737549,
|
| 2357 |
+
"eval_runtime": 731.6622,
|
| 2358 |
+
"eval_samples_per_second": 42.368,
|
| 2359 |
+
"eval_steps_per_second": 7.062,
|
| 2360 |
+
"step": 6720000
|
| 2361 |
+
},
|
| 2362 |
+
{
|
| 2363 |
+
"epoch": 45.64,
|
| 2364 |
+
"learning_rate": 9.715796470600592e-07,
|
| 2365 |
+
"loss": 0.9597,
|
| 2366 |
+
"step": 6760000
|
| 2367 |
+
},
|
| 2368 |
+
{
|
| 2369 |
+
"epoch": 45.64,
|
| 2370 |
+
"eval_loss": 0.938642680644989,
|
| 2371 |
+
"eval_runtime": 731.6777,
|
| 2372 |
+
"eval_samples_per_second": 42.367,
|
| 2373 |
+
"eval_steps_per_second": 7.062,
|
| 2374 |
+
"step": 6760000
|
| 2375 |
+
},
|
| 2376 |
+
{
|
| 2377 |
+
"epoch": 45.91,
|
| 2378 |
+
"learning_rate": 9.115819577798485e-07,
|
| 2379 |
+
"loss": 0.9572,
|
| 2380 |
+
"step": 6800000
|
| 2381 |
+
},
|
| 2382 |
+
{
|
| 2383 |
+
"epoch": 45.91,
|
| 2384 |
+
"eval_loss": 0.9379501342773438,
|
| 2385 |
+
"eval_runtime": 731.8028,
|
| 2386 |
+
"eval_samples_per_second": 42.36,
|
| 2387 |
+
"eval_steps_per_second": 7.061,
|
| 2388 |
+
"step": 6800000
|
| 2389 |
+
},
|
| 2390 |
+
{
|
| 2391 |
+
"epoch": 46.18,
|
| 2392 |
+
"learning_rate": 8.515827680322421e-07,
|
| 2393 |
+
"loss": 0.9591,
|
| 2394 |
+
"step": 6840000
|
| 2395 |
+
},
|
| 2396 |
+
{
|
| 2397 |
+
"epoch": 46.18,
|
| 2398 |
+
"eval_loss": 0.9362252950668335,
|
| 2399 |
+
"eval_runtime": 731.8252,
|
| 2400 |
+
"eval_samples_per_second": 42.358,
|
| 2401 |
+
"eval_steps_per_second": 7.06,
|
| 2402 |
+
"step": 6840000
|
| 2403 |
+
},
|
| 2404 |
+
{
|
| 2405 |
+
"epoch": 46.45,
|
| 2406 |
+
"learning_rate": 7.915865792194268e-07,
|
| 2407 |
+
"loss": 0.9566,
|
| 2408 |
+
"step": 6880000
|
| 2409 |
+
},
|
| 2410 |
+
{
|
| 2411 |
+
"epoch": 46.45,
|
| 2412 |
+
"eval_loss": 0.9348224997520447,
|
| 2413 |
+
"eval_runtime": 731.8651,
|
| 2414 |
+
"eval_samples_per_second": 42.356,
|
| 2415 |
+
"eval_steps_per_second": 7.06,
|
| 2416 |
+
"step": 6880000
|
| 2417 |
+
},
|
| 2418 |
+
{
|
| 2419 |
+
"epoch": 46.72,
|
| 2420 |
+
"learning_rate": 7.315903904066117e-07,
|
| 2421 |
+
"loss": 0.9574,
|
| 2422 |
+
"step": 6920000
|
| 2423 |
+
},
|
| 2424 |
+
{
|
| 2425 |
+
"epoch": 46.72,
|
| 2426 |
+
"eval_loss": 0.9346414804458618,
|
| 2427 |
+
"eval_runtime": 731.9242,
|
| 2428 |
+
"eval_samples_per_second": 42.353,
|
| 2429 |
+
"eval_steps_per_second": 7.059,
|
| 2430 |
+
"step": 6920000
|
| 2431 |
+
},
|
| 2432 |
+
{
|
| 2433 |
+
"epoch": 46.99,
|
| 2434 |
+
"learning_rate": 6.715957020611921e-07,
|
| 2435 |
+
"loss": 0.9563,
|
| 2436 |
+
"step": 6960000
|
| 2437 |
+
},
|
| 2438 |
+
{
|
| 2439 |
+
"epoch": 46.99,
|
| 2440 |
+
"eval_loss": 0.9362134337425232,
|
| 2441 |
+
"eval_runtime": 731.9416,
|
| 2442 |
+
"eval_samples_per_second": 42.352,
|
| 2443 |
+
"eval_steps_per_second": 7.059,
|
| 2444 |
+
"step": 6960000
|
| 2445 |
+
},
|
| 2446 |
+
{
|
| 2447 |
+
"epoch": 47.26,
|
| 2448 |
+
"learning_rate": 6.116025141831681e-07,
|
| 2449 |
+
"loss": 0.9546,
|
| 2450 |
+
"step": 7000000
|
| 2451 |
+
},
|
| 2452 |
+
{
|
| 2453 |
+
"epoch": 47.26,
|
| 2454 |
+
"eval_loss": 0.9321693778038025,
|
| 2455 |
+
"eval_runtime": 731.9394,
|
| 2456 |
+
"eval_samples_per_second": 42.352,
|
| 2457 |
+
"eval_steps_per_second": 7.059,
|
| 2458 |
+
"step": 7000000
|
| 2459 |
+
},
|
| 2460 |
+
{
|
| 2461 |
+
"epoch": 47.53,
|
| 2462 |
+
"learning_rate": 5.516048249029573e-07,
|
| 2463 |
+
"loss": 0.955,
|
| 2464 |
+
"step": 7040000
|
| 2465 |
+
},
|
| 2466 |
+
{
|
| 2467 |
+
"epoch": 47.53,
|
| 2468 |
+
"eval_loss": 0.9357685446739197,
|
| 2469 |
+
"eval_runtime": 731.9226,
|
| 2470 |
+
"eval_samples_per_second": 42.353,
|
| 2471 |
+
"eval_steps_per_second": 7.059,
|
| 2472 |
+
"step": 7040000
|
| 2473 |
+
},
|
| 2474 |
+
{
|
| 2475 |
+
"epoch": 47.8,
|
| 2476 |
+
"learning_rate": 4.916071356227465e-07,
|
| 2477 |
+
"loss": 0.9558,
|
| 2478 |
+
"step": 7080000
|
| 2479 |
+
},
|
| 2480 |
+
{
|
| 2481 |
+
"epoch": 47.8,
|
| 2482 |
+
"eval_loss": 0.9322512149810791,
|
| 2483 |
+
"eval_runtime": 731.9386,
|
| 2484 |
+
"eval_samples_per_second": 42.352,
|
| 2485 |
+
"eval_steps_per_second": 7.059,
|
| 2486 |
+
"step": 7080000
|
| 2487 |
+
},
|
| 2488 |
+
{
|
| 2489 |
+
"epoch": 48.07,
|
| 2490 |
+
"learning_rate": 4.3160794587514014e-07,
|
| 2491 |
+
"loss": 0.9547,
|
| 2492 |
+
"step": 7120000
|
| 2493 |
+
},
|
| 2494 |
+
{
|
| 2495 |
+
"epoch": 48.07,
|
| 2496 |
+
"eval_loss": 0.9324430227279663,
|
| 2497 |
+
"eval_runtime": 731.9445,
|
| 2498 |
+
"eval_samples_per_second": 42.352,
|
| 2499 |
+
"eval_steps_per_second": 7.059,
|
| 2500 |
+
"step": 7120000
|
| 2501 |
+
},
|
| 2502 |
+
{
|
| 2503 |
+
"epoch": 48.35,
|
| 2504 |
+
"learning_rate": 3.7161025659492937e-07,
|
| 2505 |
+
"loss": 0.9536,
|
| 2506 |
+
"step": 7160000
|
| 2507 |
+
},
|
| 2508 |
+
{
|
| 2509 |
+
"epoch": 48.35,
|
| 2510 |
+
"eval_loss": 0.9327294230461121,
|
| 2511 |
+
"eval_runtime": 731.9579,
|
| 2512 |
+
"eval_samples_per_second": 42.351,
|
| 2513 |
+
"eval_steps_per_second": 7.059,
|
| 2514 |
+
"step": 7160000
|
| 2515 |
+
},
|
| 2516 |
+
{
|
| 2517 |
+
"epoch": 48.62,
|
| 2518 |
+
"learning_rate": 3.116140677821142e-07,
|
| 2519 |
+
"loss": 0.9533,
|
| 2520 |
+
"step": 7200000
|
| 2521 |
+
},
|
| 2522 |
+
{
|
| 2523 |
+
"epoch": 48.62,
|
| 2524 |
+
"eval_loss": 0.933174192905426,
|
| 2525 |
+
"eval_runtime": 732.1291,
|
| 2526 |
+
"eval_samples_per_second": 42.341,
|
| 2527 |
+
"eval_steps_per_second": 7.057,
|
| 2528 |
+
"step": 7200000
|
| 2529 |
+
},
|
| 2530 |
+
{
|
| 2531 |
+
"epoch": 48.89,
|
| 2532 |
+
"learning_rate": 2.5161637850190335e-07,
|
| 2533 |
+
"loss": 0.9536,
|
| 2534 |
+
"step": 7240000
|
| 2535 |
+
},
|
| 2536 |
+
{
|
| 2537 |
+
"epoch": 48.89,
|
| 2538 |
+
"eval_loss": 0.9329264163970947,
|
| 2539 |
+
"eval_runtime": 732.2008,
|
| 2540 |
+
"eval_samples_per_second": 42.337,
|
| 2541 |
+
"eval_steps_per_second": 7.057,
|
| 2542 |
+
"step": 7240000
|
| 2543 |
+
},
|
| 2544 |
+
{
|
| 2545 |
+
"epoch": 49.16,
|
| 2546 |
+
"learning_rate": 1.9162018968908816e-07,
|
| 2547 |
+
"loss": 0.9517,
|
| 2548 |
+
"step": 7280000
|
| 2549 |
+
},
|
| 2550 |
+
{
|
| 2551 |
+
"epoch": 49.16,
|
| 2552 |
+
"eval_loss": 0.9311585426330566,
|
| 2553 |
+
"eval_runtime": 732.166,
|
| 2554 |
+
"eval_samples_per_second": 42.339,
|
| 2555 |
+
"eval_steps_per_second": 7.057,
|
| 2556 |
+
"step": 7280000
|
| 2557 |
+
},
|
| 2558 |
+
{
|
| 2559 |
+
"epoch": 49.43,
|
| 2560 |
+
"learning_rate": 1.3162550134366856e-07,
|
| 2561 |
+
"loss": 0.9525,
|
| 2562 |
+
"step": 7320000
|
| 2563 |
+
},
|
| 2564 |
+
{
|
| 2565 |
+
"epoch": 49.43,
|
| 2566 |
+
"eval_loss": 0.9293885827064514,
|
| 2567 |
+
"eval_runtime": 732.1103,
|
| 2568 |
+
"eval_samples_per_second": 42.342,
|
| 2569 |
+
"eval_steps_per_second": 7.058,
|
| 2570 |
+
"step": 7320000
|
| 2571 |
+
},
|
| 2572 |
+
{
|
| 2573 |
+
"epoch": 49.7,
|
| 2574 |
+
"learning_rate": 7.162931253085337e-08,
|
| 2575 |
+
"loss": 0.9519,
|
| 2576 |
+
"step": 7360000
|
| 2577 |
+
},
|
| 2578 |
+
{
|
| 2579 |
+
"epoch": 49.7,
|
| 2580 |
+
"eval_loss": 0.9308800101280212,
|
| 2581 |
+
"eval_runtime": 732.0979,
|
| 2582 |
+
"eval_samples_per_second": 42.343,
|
| 2583 |
+
"eval_steps_per_second": 7.058,
|
| 2584 |
+
"step": 7360000
|
| 2585 |
+
},
|
| 2586 |
+
{
|
| 2587 |
+
"epoch": 49.97,
|
| 2588 |
+
"learning_rate": 1.1630122783246982e-08,
|
| 2589 |
+
"loss": 0.9518,
|
| 2590 |
+
"step": 7400000
|
| 2591 |
+
},
|
| 2592 |
+
{
|
| 2593 |
+
"epoch": 49.97,
|
| 2594 |
+
"eval_loss": 0.9334636926651001,
|
| 2595 |
+
"eval_runtime": 731.9724,
|
| 2596 |
+
"eval_samples_per_second": 42.35,
|
| 2597 |
+
"eval_steps_per_second": 7.059,
|
| 2598 |
+
"step": 7400000
|
| 2599 |
+
}
|
| 2600 |
+
],
|
| 2601 |
+
"max_steps": 7405100,
|
| 2602 |
+
"num_train_epochs": 50,
|
| 2603 |
+
"total_flos": 7.798931788086835e+18,
|
| 2604 |
+
"trial_name": null,
|
| 2605 |
+
"trial_params": null
|
| 2606 |
+
}
|