Commit
·
116bc8b
1
Parent(s):
0dd4fd0
Upload trainer_state.json
Browse files- trainer_state.json +2627 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,2627 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.6414651274681091,
|
| 3 |
+
"best_model_checkpoint": "output/checkpoint-170000",
|
| 4 |
+
"epoch": 15.0,
|
| 5 |
+
"global_step": 171660,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.04,
|
| 12 |
+
"learning_rate": 4.98e-06,
|
| 13 |
+
"loss": 2.0401,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.09,
|
| 18 |
+
"learning_rate": 9.980000000000001e-06,
|
| 19 |
+
"loss": 1.397,
|
| 20 |
+
"step": 1000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.13,
|
| 24 |
+
"learning_rate": 1.4979999999999999e-05,
|
| 25 |
+
"loss": 1.302,
|
| 26 |
+
"step": 1500
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.17,
|
| 30 |
+
"learning_rate": 1.9980000000000002e-05,
|
| 31 |
+
"loss": 1.242,
|
| 32 |
+
"step": 2000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.22,
|
| 36 |
+
"learning_rate": 2.4970000000000003e-05,
|
| 37 |
+
"loss": 1.1985,
|
| 38 |
+
"step": 2500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.22,
|
| 42 |
+
"eval_loss": 1.0940483808517456,
|
| 43 |
+
"eval_runtime": 3092.4183,
|
| 44 |
+
"eval_samples_per_second": 24.931,
|
| 45 |
+
"eval_steps_per_second": 1.558,
|
| 46 |
+
"step": 2500
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 0.26,
|
| 50 |
+
"learning_rate": 2.9970000000000003e-05,
|
| 51 |
+
"loss": 1.1712,
|
| 52 |
+
"step": 3000
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.31,
|
| 56 |
+
"learning_rate": 3.4970000000000006e-05,
|
| 57 |
+
"loss": 1.1419,
|
| 58 |
+
"step": 3500
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.35,
|
| 62 |
+
"learning_rate": 3.9970000000000005e-05,
|
| 63 |
+
"loss": 1.1259,
|
| 64 |
+
"step": 4000
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 0.39,
|
| 68 |
+
"learning_rate": 4.497e-05,
|
| 69 |
+
"loss": 1.1088,
|
| 70 |
+
"step": 4500
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 0.44,
|
| 74 |
+
"learning_rate": 4.997e-05,
|
| 75 |
+
"loss": 1.0937,
|
| 76 |
+
"step": 5000
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 0.44,
|
| 80 |
+
"eval_loss": 1.0032908916473389,
|
| 81 |
+
"eval_runtime": 3130.3678,
|
| 82 |
+
"eval_samples_per_second": 24.628,
|
| 83 |
+
"eval_steps_per_second": 1.539,
|
| 84 |
+
"step": 5000
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.48,
|
| 88 |
+
"learning_rate": 5.497e-05,
|
| 89 |
+
"loss": 1.0865,
|
| 90 |
+
"step": 5500
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 0.52,
|
| 94 |
+
"learning_rate": 5.9970000000000004e-05,
|
| 95 |
+
"loss": 1.0763,
|
| 96 |
+
"step": 6000
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 0.57,
|
| 100 |
+
"learning_rate": 6.497000000000001e-05,
|
| 101 |
+
"loss": 1.0666,
|
| 102 |
+
"step": 6500
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 0.61,
|
| 106 |
+
"learning_rate": 6.997e-05,
|
| 107 |
+
"loss": 1.0704,
|
| 108 |
+
"step": 7000
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.66,
|
| 112 |
+
"learning_rate": 7.497000000000001e-05,
|
| 113 |
+
"loss": 1.0675,
|
| 114 |
+
"step": 7500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.66,
|
| 118 |
+
"eval_loss": 0.9753177165985107,
|
| 119 |
+
"eval_runtime": 3110.5948,
|
| 120 |
+
"eval_samples_per_second": 24.785,
|
| 121 |
+
"eval_steps_per_second": 1.549,
|
| 122 |
+
"step": 7500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.7,
|
| 126 |
+
"learning_rate": 7.997e-05,
|
| 127 |
+
"loss": 1.0647,
|
| 128 |
+
"step": 8000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.74,
|
| 132 |
+
"learning_rate": 8.495e-05,
|
| 133 |
+
"loss": 1.0493,
|
| 134 |
+
"step": 8500
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.79,
|
| 138 |
+
"learning_rate": 8.995e-05,
|
| 139 |
+
"loss": 1.0583,
|
| 140 |
+
"step": 9000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.83,
|
| 144 |
+
"learning_rate": 9.495e-05,
|
| 145 |
+
"loss": 1.0582,
|
| 146 |
+
"step": 9500
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.87,
|
| 150 |
+
"learning_rate": 9.995e-05,
|
| 151 |
+
"loss": 1.0565,
|
| 152 |
+
"step": 10000
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.87,
|
| 156 |
+
"eval_loss": 0.9801222085952759,
|
| 157 |
+
"eval_runtime": 3136.3647,
|
| 158 |
+
"eval_samples_per_second": 24.581,
|
| 159 |
+
"eval_steps_per_second": 1.536,
|
| 160 |
+
"step": 10000
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 0.92,
|
| 164 |
+
"learning_rate": 9.895171537484117e-05,
|
| 165 |
+
"loss": 1.0507,
|
| 166 |
+
"step": 10500
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 0.96,
|
| 170 |
+
"learning_rate": 9.789284201609487e-05,
|
| 171 |
+
"loss": 1.0443,
|
| 172 |
+
"step": 11000
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 1.0,
|
| 176 |
+
"learning_rate": 9.683396865734859e-05,
|
| 177 |
+
"loss": 1.0318,
|
| 178 |
+
"step": 11500
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 1.05,
|
| 182 |
+
"learning_rate": 9.577721304531978e-05,
|
| 183 |
+
"loss": 1.0298,
|
| 184 |
+
"step": 12000
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 1.09,
|
| 188 |
+
"learning_rate": 9.471833968657349e-05,
|
| 189 |
+
"loss": 1.0244,
|
| 190 |
+
"step": 12500
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 1.09,
|
| 194 |
+
"eval_loss": 0.9525517225265503,
|
| 195 |
+
"eval_runtime": 3137.7232,
|
| 196 |
+
"eval_samples_per_second": 24.571,
|
| 197 |
+
"eval_steps_per_second": 1.536,
|
| 198 |
+
"step": 12500
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 1.14,
|
| 202 |
+
"learning_rate": 9.365946632782719e-05,
|
| 203 |
+
"loss": 1.024,
|
| 204 |
+
"step": 13000
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 1.18,
|
| 208 |
+
"learning_rate": 9.26005929690809e-05,
|
| 209 |
+
"loss": 1.0054,
|
| 210 |
+
"step": 13500
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 1.22,
|
| 214 |
+
"learning_rate": 9.154383735705209e-05,
|
| 215 |
+
"loss": 1.0105,
|
| 216 |
+
"step": 14000
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 1.27,
|
| 220 |
+
"learning_rate": 9.04849639983058e-05,
|
| 221 |
+
"loss": 1.0035,
|
| 222 |
+
"step": 14500
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 1.31,
|
| 226 |
+
"learning_rate": 8.942609063955952e-05,
|
| 227 |
+
"loss": 0.9943,
|
| 228 |
+
"step": 15000
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 1.31,
|
| 232 |
+
"eval_loss": 0.9298429489135742,
|
| 233 |
+
"eval_runtime": 3051.8254,
|
| 234 |
+
"eval_samples_per_second": 25.262,
|
| 235 |
+
"eval_steps_per_second": 1.579,
|
| 236 |
+
"step": 15000
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 1.35,
|
| 240 |
+
"learning_rate": 8.836721728081322e-05,
|
| 241 |
+
"loss": 0.9959,
|
| 242 |
+
"step": 15500
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 1.4,
|
| 246 |
+
"learning_rate": 8.731046166878442e-05,
|
| 247 |
+
"loss": 0.9874,
|
| 248 |
+
"step": 16000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 1.44,
|
| 252 |
+
"learning_rate": 8.625158831003812e-05,
|
| 253 |
+
"loss": 0.9834,
|
| 254 |
+
"step": 16500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.49,
|
| 258 |
+
"learning_rate": 8.519483269800933e-05,
|
| 259 |
+
"loss": 0.9764,
|
| 260 |
+
"step": 17000
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 1.53,
|
| 264 |
+
"learning_rate": 8.413595933926302e-05,
|
| 265 |
+
"loss": 0.9799,
|
| 266 |
+
"step": 17500
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 1.53,
|
| 270 |
+
"eval_loss": 0.9034547805786133,
|
| 271 |
+
"eval_runtime": 3114.4981,
|
| 272 |
+
"eval_samples_per_second": 24.754,
|
| 273 |
+
"eval_steps_per_second": 1.547,
|
| 274 |
+
"step": 17500
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 1.57,
|
| 278 |
+
"learning_rate": 8.307708598051674e-05,
|
| 279 |
+
"loss": 0.9642,
|
| 280 |
+
"step": 18000
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 1.62,
|
| 284 |
+
"learning_rate": 8.201821262177044e-05,
|
| 285 |
+
"loss": 0.9671,
|
| 286 |
+
"step": 18500
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 1.66,
|
| 290 |
+
"learning_rate": 8.095933926302415e-05,
|
| 291 |
+
"loss": 0.9637,
|
| 292 |
+
"step": 19000
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 1.7,
|
| 296 |
+
"learning_rate": 7.990046590427785e-05,
|
| 297 |
+
"loss": 0.9545,
|
| 298 |
+
"step": 19500
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 1.75,
|
| 302 |
+
"learning_rate": 7.884159254553156e-05,
|
| 303 |
+
"loss": 0.95,
|
| 304 |
+
"step": 20000
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 1.75,
|
| 308 |
+
"eval_loss": 0.8834716081619263,
|
| 309 |
+
"eval_runtime": 3125.2028,
|
| 310 |
+
"eval_samples_per_second": 24.669,
|
| 311 |
+
"eval_steps_per_second": 1.542,
|
| 312 |
+
"step": 20000
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 1.79,
|
| 316 |
+
"learning_rate": 7.778483693350275e-05,
|
| 317 |
+
"loss": 0.9556,
|
| 318 |
+
"step": 20500
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 1.84,
|
| 322 |
+
"learning_rate": 7.672596357475647e-05,
|
| 323 |
+
"loss": 0.9519,
|
| 324 |
+
"step": 21000
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.88,
|
| 328 |
+
"learning_rate": 7.566709021601018e-05,
|
| 329 |
+
"loss": 0.9448,
|
| 330 |
+
"step": 21500
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 1.92,
|
| 334 |
+
"learning_rate": 7.460821685726388e-05,
|
| 335 |
+
"loss": 0.9382,
|
| 336 |
+
"step": 22000
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 1.97,
|
| 340 |
+
"learning_rate": 7.354934349851758e-05,
|
| 341 |
+
"loss": 0.933,
|
| 342 |
+
"step": 22500
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 1.97,
|
| 346 |
+
"eval_loss": 0.8636178970336914,
|
| 347 |
+
"eval_runtime": 3087.2735,
|
| 348 |
+
"eval_samples_per_second": 24.972,
|
| 349 |
+
"eval_steps_per_second": 1.561,
|
| 350 |
+
"step": 22500
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 2.01,
|
| 354 |
+
"learning_rate": 7.249258788648878e-05,
|
| 355 |
+
"loss": 0.93,
|
| 356 |
+
"step": 23000
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 2.05,
|
| 360 |
+
"learning_rate": 7.14337145277425e-05,
|
| 361 |
+
"loss": 0.9193,
|
| 362 |
+
"step": 23500
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 2.1,
|
| 366 |
+
"learning_rate": 7.037484116899618e-05,
|
| 367 |
+
"loss": 0.9141,
|
| 368 |
+
"step": 24000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 2.14,
|
| 372 |
+
"learning_rate": 6.931596781024989e-05,
|
| 373 |
+
"loss": 0.9092,
|
| 374 |
+
"step": 24500
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 2.18,
|
| 378 |
+
"learning_rate": 6.82570944515036e-05,
|
| 379 |
+
"loss": 0.9079,
|
| 380 |
+
"step": 25000
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 2.18,
|
| 384 |
+
"eval_loss": 0.8507079482078552,
|
| 385 |
+
"eval_runtime": 3106.3211,
|
| 386 |
+
"eval_samples_per_second": 24.819,
|
| 387 |
+
"eval_steps_per_second": 1.551,
|
| 388 |
+
"step": 25000
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 2.23,
|
| 392 |
+
"learning_rate": 6.72003388394748e-05,
|
| 393 |
+
"loss": 0.9041,
|
| 394 |
+
"step": 25500
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 2.27,
|
| 398 |
+
"learning_rate": 6.614146548072851e-05,
|
| 399 |
+
"loss": 0.9004,
|
| 400 |
+
"step": 26000
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 2.32,
|
| 404 |
+
"learning_rate": 6.508259212198221e-05,
|
| 405 |
+
"loss": 0.8995,
|
| 406 |
+
"step": 26500
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 2.36,
|
| 410 |
+
"learning_rate": 6.402371876323592e-05,
|
| 411 |
+
"loss": 0.8922,
|
| 412 |
+
"step": 27000
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 2.4,
|
| 416 |
+
"learning_rate": 6.296484540448962e-05,
|
| 417 |
+
"loss": 0.8938,
|
| 418 |
+
"step": 27500
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 2.4,
|
| 422 |
+
"eval_loss": 0.8396568894386292,
|
| 423 |
+
"eval_runtime": 3033.9082,
|
| 424 |
+
"eval_samples_per_second": 25.411,
|
| 425 |
+
"eval_steps_per_second": 1.588,
|
| 426 |
+
"step": 27500
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 2.45,
|
| 430 |
+
"learning_rate": 6.190597204574333e-05,
|
| 431 |
+
"loss": 0.8897,
|
| 432 |
+
"step": 28000
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 2.49,
|
| 436 |
+
"learning_rate": 6.084709868699704e-05,
|
| 437 |
+
"loss": 0.8855,
|
| 438 |
+
"step": 28500
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 2.53,
|
| 442 |
+
"learning_rate": 5.979034307496824e-05,
|
| 443 |
+
"loss": 0.8795,
|
| 444 |
+
"step": 29000
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 2.58,
|
| 448 |
+
"learning_rate": 5.873146971622194e-05,
|
| 449 |
+
"loss": 0.8841,
|
| 450 |
+
"step": 29500
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 2.62,
|
| 454 |
+
"learning_rate": 5.767259635747565e-05,
|
| 455 |
+
"loss": 0.8781,
|
| 456 |
+
"step": 30000
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 2.62,
|
| 460 |
+
"eval_loss": 0.8194745182991028,
|
| 461 |
+
"eval_runtime": 3075.2096,
|
| 462 |
+
"eval_samples_per_second": 25.07,
|
| 463 |
+
"eval_steps_per_second": 1.567,
|
| 464 |
+
"step": 30000
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.67,
|
| 468 |
+
"learning_rate": 5.6613722998729355e-05,
|
| 469 |
+
"loss": 0.8776,
|
| 470 |
+
"step": 30500
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 2.71,
|
| 474 |
+
"learning_rate": 5.555484963998306e-05,
|
| 475 |
+
"loss": 0.8777,
|
| 476 |
+
"step": 31000
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 2.75,
|
| 480 |
+
"learning_rate": 5.4495976281236774e-05,
|
| 481 |
+
"loss": 0.865,
|
| 482 |
+
"step": 31500
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 2.8,
|
| 486 |
+
"learning_rate": 5.3437102922490467e-05,
|
| 487 |
+
"loss": 0.8752,
|
| 488 |
+
"step": 32000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 2.84,
|
| 492 |
+
"learning_rate": 5.237822956374417e-05,
|
| 493 |
+
"loss": 0.8647,
|
| 494 |
+
"step": 32500
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 2.84,
|
| 498 |
+
"eval_loss": 0.8087666630744934,
|
| 499 |
+
"eval_runtime": 3169.435,
|
| 500 |
+
"eval_samples_per_second": 24.325,
|
| 501 |
+
"eval_steps_per_second": 1.52,
|
| 502 |
+
"step": 32500
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 2.88,
|
| 506 |
+
"learning_rate": 5.1321473951715384e-05,
|
| 507 |
+
"loss": 0.8599,
|
| 508 |
+
"step": 33000
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 2.93,
|
| 512 |
+
"learning_rate": 5.0262600592969076e-05,
|
| 513 |
+
"loss": 0.8568,
|
| 514 |
+
"step": 33500
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 2.97,
|
| 518 |
+
"learning_rate": 4.920372723422279e-05,
|
| 519 |
+
"loss": 0.8528,
|
| 520 |
+
"step": 34000
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 3.01,
|
| 524 |
+
"learning_rate": 4.8144853875476495e-05,
|
| 525 |
+
"loss": 0.8463,
|
| 526 |
+
"step": 34500
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 3.06,
|
| 530 |
+
"learning_rate": 4.70859805167302e-05,
|
| 531 |
+
"loss": 0.8422,
|
| 532 |
+
"step": 35000
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 3.06,
|
| 536 |
+
"eval_loss": 0.7954000234603882,
|
| 537 |
+
"eval_runtime": 3127.0452,
|
| 538 |
+
"eval_samples_per_second": 24.655,
|
| 539 |
+
"eval_steps_per_second": 1.541,
|
| 540 |
+
"step": 35000
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 3.1,
|
| 544 |
+
"learning_rate": 4.602710715798391e-05,
|
| 545 |
+
"loss": 0.8373,
|
| 546 |
+
"step": 35500
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 3.15,
|
| 550 |
+
"learning_rate": 4.4968233799237613e-05,
|
| 551 |
+
"loss": 0.8268,
|
| 552 |
+
"step": 36000
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 3.19,
|
| 556 |
+
"learning_rate": 4.390936044049132e-05,
|
| 557 |
+
"loss": 0.8426,
|
| 558 |
+
"step": 36500
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 3.23,
|
| 562 |
+
"learning_rate": 4.2850487081745026e-05,
|
| 563 |
+
"loss": 0.837,
|
| 564 |
+
"step": 37000
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 3.28,
|
| 568 |
+
"learning_rate": 4.179373146971622e-05,
|
| 569 |
+
"loss": 0.831,
|
| 570 |
+
"step": 37500
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 3.28,
|
| 574 |
+
"eval_loss": 0.7871229648590088,
|
| 575 |
+
"eval_runtime": 3095.9065,
|
| 576 |
+
"eval_samples_per_second": 24.903,
|
| 577 |
+
"eval_steps_per_second": 1.557,
|
| 578 |
+
"step": 37500
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 3.32,
|
| 582 |
+
"learning_rate": 4.073485811096993e-05,
|
| 583 |
+
"loss": 0.8343,
|
| 584 |
+
"step": 38000
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 3.36,
|
| 588 |
+
"learning_rate": 3.9675984752223635e-05,
|
| 589 |
+
"loss": 0.825,
|
| 590 |
+
"step": 38500
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 3.41,
|
| 594 |
+
"learning_rate": 3.861711139347734e-05,
|
| 595 |
+
"loss": 0.8195,
|
| 596 |
+
"step": 39000
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 3.45,
|
| 600 |
+
"learning_rate": 3.755823803473105e-05,
|
| 601 |
+
"loss": 0.8219,
|
| 602 |
+
"step": 39500
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 3.5,
|
| 606 |
+
"learning_rate": 3.650148242270225e-05,
|
| 607 |
+
"loss": 0.8173,
|
| 608 |
+
"step": 40000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 3.5,
|
| 612 |
+
"eval_loss": 0.7720773220062256,
|
| 613 |
+
"eval_runtime": 3173.0037,
|
| 614 |
+
"eval_samples_per_second": 24.297,
|
| 615 |
+
"eval_steps_per_second": 1.519,
|
| 616 |
+
"step": 40000
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 3.54,
|
| 620 |
+
"learning_rate": 3.544260906395595e-05,
|
| 621 |
+
"loss": 0.8148,
|
| 622 |
+
"step": 40500
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 3.58,
|
| 626 |
+
"learning_rate": 3.4385853451927155e-05,
|
| 627 |
+
"loss": 0.8102,
|
| 628 |
+
"step": 41000
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 3.63,
|
| 632 |
+
"learning_rate": 3.3329097839898346e-05,
|
| 633 |
+
"loss": 0.8103,
|
| 634 |
+
"step": 41500
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 3.67,
|
| 638 |
+
"learning_rate": 3.227022448115206e-05,
|
| 639 |
+
"loss": 0.8116,
|
| 640 |
+
"step": 42000
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 3.71,
|
| 644 |
+
"learning_rate": 3.121346886912325e-05,
|
| 645 |
+
"loss": 0.8072,
|
| 646 |
+
"step": 42500
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 3.71,
|
| 650 |
+
"eval_loss": 0.7610893845558167,
|
| 651 |
+
"eval_runtime": 3130.5399,
|
| 652 |
+
"eval_samples_per_second": 24.627,
|
| 653 |
+
"eval_steps_per_second": 1.539,
|
| 654 |
+
"step": 42500
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 3.76,
|
| 658 |
+
"learning_rate": 3.015459551037696e-05,
|
| 659 |
+
"loss": 0.802,
|
| 660 |
+
"step": 43000
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 3.8,
|
| 664 |
+
"learning_rate": 2.9095722151630668e-05,
|
| 665 |
+
"loss": 0.7994,
|
| 666 |
+
"step": 43500
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 3.84,
|
| 670 |
+
"learning_rate": 2.8036848792884374e-05,
|
| 671 |
+
"loss": 0.8029,
|
| 672 |
+
"step": 44000
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 3.89,
|
| 676 |
+
"learning_rate": 2.697797543413808e-05,
|
| 677 |
+
"loss": 0.7975,
|
| 678 |
+
"step": 44500
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 3.93,
|
| 682 |
+
"learning_rate": 2.5919102075391783e-05,
|
| 683 |
+
"loss": 0.8011,
|
| 684 |
+
"step": 45000
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 3.93,
|
| 688 |
+
"eval_loss": 0.7531821131706238,
|
| 689 |
+
"eval_runtime": 3062.0684,
|
| 690 |
+
"eval_samples_per_second": 25.178,
|
| 691 |
+
"eval_steps_per_second": 1.574,
|
| 692 |
+
"step": 45000
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 3.98,
|
| 696 |
+
"learning_rate": 2.486022871664549e-05,
|
| 697 |
+
"loss": 0.7935,
|
| 698 |
+
"step": 45500
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 4.02,
|
| 702 |
+
"learning_rate": 2.3801355357899195e-05,
|
| 703 |
+
"loss": 0.793,
|
| 704 |
+
"step": 46000
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 4.06,
|
| 708 |
+
"learning_rate": 2.2742481999152905e-05,
|
| 709 |
+
"loss": 0.788,
|
| 710 |
+
"step": 46500
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 4.11,
|
| 714 |
+
"learning_rate": 2.1683608640406607e-05,
|
| 715 |
+
"loss": 0.7853,
|
| 716 |
+
"step": 47000
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 4.15,
|
| 720 |
+
"learning_rate": 2.0624735281660313e-05,
|
| 721 |
+
"loss": 0.7828,
|
| 722 |
+
"step": 47500
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 4.15,
|
| 726 |
+
"eval_loss": 0.7431035041809082,
|
| 727 |
+
"eval_runtime": 3053.0232,
|
| 728 |
+
"eval_samples_per_second": 25.252,
|
| 729 |
+
"eval_steps_per_second": 1.578,
|
| 730 |
+
"step": 47500
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 4.19,
|
| 734 |
+
"learning_rate": 1.956586192291402e-05,
|
| 735 |
+
"loss": 0.7715,
|
| 736 |
+
"step": 48000
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 4.24,
|
| 740 |
+
"learning_rate": 1.8506988564167726e-05,
|
| 741 |
+
"loss": 0.7752,
|
| 742 |
+
"step": 48500
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"epoch": 4.28,
|
| 746 |
+
"learning_rate": 1.7450232952138926e-05,
|
| 747 |
+
"loss": 0.7714,
|
| 748 |
+
"step": 49000
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 4.33,
|
| 752 |
+
"learning_rate": 1.639135959339263e-05,
|
| 753 |
+
"loss": 0.7736,
|
| 754 |
+
"step": 49500
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 4.37,
|
| 758 |
+
"learning_rate": 1.533248623464634e-05,
|
| 759 |
+
"loss": 0.7691,
|
| 760 |
+
"step": 50000
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 4.37,
|
| 764 |
+
"eval_loss": 0.7367000579833984,
|
| 765 |
+
"eval_runtime": 3035.0963,
|
| 766 |
+
"eval_samples_per_second": 25.402,
|
| 767 |
+
"eval_steps_per_second": 1.588,
|
| 768 |
+
"step": 50000
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 4.41,
|
| 772 |
+
"learning_rate": 1.4273612875900045e-05,
|
| 773 |
+
"loss": 0.7685,
|
| 774 |
+
"step": 50500
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 4.46,
|
| 778 |
+
"learning_rate": 1.3214739517153749e-05,
|
| 779 |
+
"loss": 0.7639,
|
| 780 |
+
"step": 51000
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 4.5,
|
| 784 |
+
"learning_rate": 1.2157983905124948e-05,
|
| 785 |
+
"loss": 0.7654,
|
| 786 |
+
"step": 51500
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 4.54,
|
| 790 |
+
"learning_rate": 1.1099110546378654e-05,
|
| 791 |
+
"loss": 0.7677,
|
| 792 |
+
"step": 52000
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 4.59,
|
| 796 |
+
"learning_rate": 1.0040237187632359e-05,
|
| 797 |
+
"loss": 0.7659,
|
| 798 |
+
"step": 52500
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 4.59,
|
| 802 |
+
"eval_loss": 0.7291901111602783,
|
| 803 |
+
"eval_runtime": 3110.3355,
|
| 804 |
+
"eval_samples_per_second": 24.787,
|
| 805 |
+
"eval_steps_per_second": 1.549,
|
| 806 |
+
"step": 52500
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 4.63,
|
| 810 |
+
"learning_rate": 8.981363828886067e-06,
|
| 811 |
+
"loss": 0.763,
|
| 812 |
+
"step": 53000
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 4.67,
|
| 816 |
+
"learning_rate": 7.922490470139773e-06,
|
| 817 |
+
"loss": 0.759,
|
| 818 |
+
"step": 53500
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 4.72,
|
| 822 |
+
"learning_rate": 6.863617111393478e-06,
|
| 823 |
+
"loss": 0.7559,
|
| 824 |
+
"step": 54000
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 4.76,
|
| 828 |
+
"learning_rate": 5.804743752647183e-06,
|
| 829 |
+
"loss": 0.7547,
|
| 830 |
+
"step": 54500
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 4.81,
|
| 834 |
+
"learning_rate": 4.74587039390089e-06,
|
| 835 |
+
"loss": 0.7606,
|
| 836 |
+
"step": 55000
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 4.81,
|
| 840 |
+
"eval_loss": 0.7244983911514282,
|
| 841 |
+
"eval_runtime": 3076.3307,
|
| 842 |
+
"eval_samples_per_second": 25.061,
|
| 843 |
+
"eval_steps_per_second": 1.566,
|
| 844 |
+
"step": 55000
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 4.85,
|
| 848 |
+
"learning_rate": 5.6455381080045965e-05,
|
| 849 |
+
"loss": 0.7842,
|
| 850 |
+
"step": 55500
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 4.89,
|
| 854 |
+
"learning_rate": 5.597663730371505e-05,
|
| 855 |
+
"loss": 0.7983,
|
| 856 |
+
"step": 56000
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 4.94,
|
| 860 |
+
"learning_rate": 5.549789352738415e-05,
|
| 861 |
+
"loss": 0.8032,
|
| 862 |
+
"step": 56500
|
| 863 |
+
},
|
| 864 |
+
{
|
| 865 |
+
"epoch": 4.98,
|
| 866 |
+
"learning_rate": 5.5019149751053236e-05,
|
| 867 |
+
"loss": 0.8037,
|
| 868 |
+
"step": 57000
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 5.02,
|
| 872 |
+
"learning_rate": 5.454040597472233e-05,
|
| 873 |
+
"loss": 0.8082,
|
| 874 |
+
"step": 57500
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 5.02,
|
| 878 |
+
"eval_loss": 0.7695716023445129,
|
| 879 |
+
"eval_runtime": 3095.7063,
|
| 880 |
+
"eval_samples_per_second": 24.904,
|
| 881 |
+
"eval_steps_per_second": 1.557,
|
| 882 |
+
"step": 57500
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 5.07,
|
| 886 |
+
"learning_rate": 5.406166219839143e-05,
|
| 887 |
+
"loss": 0.8102,
|
| 888 |
+
"step": 58000
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 5.11,
|
| 892 |
+
"learning_rate": 5.358291842206051e-05,
|
| 893 |
+
"loss": 0.8067,
|
| 894 |
+
"step": 58500
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 5.16,
|
| 898 |
+
"learning_rate": 5.310417464572961e-05,
|
| 899 |
+
"loss": 0.8103,
|
| 900 |
+
"step": 59000
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 5.2,
|
| 904 |
+
"learning_rate": 5.2627345844504025e-05,
|
| 905 |
+
"loss": 0.8088,
|
| 906 |
+
"step": 59500
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 5.24,
|
| 910 |
+
"learning_rate": 5.214955955572578e-05,
|
| 911 |
+
"loss": 0.8114,
|
| 912 |
+
"step": 60000
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 5.24,
|
| 916 |
+
"eval_loss": 0.7694710493087769,
|
| 917 |
+
"eval_runtime": 3084.9083,
|
| 918 |
+
"eval_samples_per_second": 24.991,
|
| 919 |
+
"eval_steps_per_second": 1.562,
|
| 920 |
+
"step": 60000
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 5.29,
|
| 924 |
+
"learning_rate": 5.167081577939487e-05,
|
| 925 |
+
"loss": 0.8073,
|
| 926 |
+
"step": 60500
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 5.33,
|
| 930 |
+
"learning_rate": 5.119207200306396e-05,
|
| 931 |
+
"loss": 0.8019,
|
| 932 |
+
"step": 61000
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 5.37,
|
| 936 |
+
"learning_rate": 5.071332822673306e-05,
|
| 937 |
+
"loss": 0.7971,
|
| 938 |
+
"step": 61500
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 5.42,
|
| 942 |
+
"learning_rate": 5.0234584450402144e-05,
|
| 943 |
+
"loss": 0.8024,
|
| 944 |
+
"step": 62000
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 5.46,
|
| 948 |
+
"learning_rate": 4.97567981616239e-05,
|
| 949 |
+
"loss": 0.8022,
|
| 950 |
+
"step": 62500
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 5.46,
|
| 954 |
+
"eval_loss": 0.7612630128860474,
|
| 955 |
+
"eval_runtime": 3135.2609,
|
| 956 |
+
"eval_samples_per_second": 24.59,
|
| 957 |
+
"eval_steps_per_second": 1.537,
|
| 958 |
+
"step": 62500
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 5.51,
|
| 962 |
+
"learning_rate": 4.927805438529299e-05,
|
| 963 |
+
"loss": 0.8071,
|
| 964 |
+
"step": 63000
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 5.55,
|
| 968 |
+
"learning_rate": 4.8799310608962085e-05,
|
| 969 |
+
"loss": 0.793,
|
| 970 |
+
"step": 63500
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 5.59,
|
| 974 |
+
"learning_rate": 4.832152432018384e-05,
|
| 975 |
+
"loss": 0.798,
|
| 976 |
+
"step": 64000
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 5.64,
|
| 980 |
+
"learning_rate": 4.7842780543852934e-05,
|
| 981 |
+
"loss": 0.7976,
|
| 982 |
+
"step": 64500
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 5.68,
|
| 986 |
+
"learning_rate": 4.7364036767522026e-05,
|
| 987 |
+
"loss": 0.7986,
|
| 988 |
+
"step": 65000
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"epoch": 5.68,
|
| 992 |
+
"eval_loss": 0.7558260560035706,
|
| 993 |
+
"eval_runtime": 3154.5823,
|
| 994 |
+
"eval_samples_per_second": 24.439,
|
| 995 |
+
"eval_steps_per_second": 1.528,
|
| 996 |
+
"step": 65000
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 5.72,
|
| 1000 |
+
"learning_rate": 4.688529299119111e-05,
|
| 1001 |
+
"loss": 0.7918,
|
| 1002 |
+
"step": 65500
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 5.77,
|
| 1006 |
+
"learning_rate": 4.6406549214860205e-05,
|
| 1007 |
+
"loss": 0.7941,
|
| 1008 |
+
"step": 66000
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 5.81,
|
| 1012 |
+
"learning_rate": 4.5927805438529304e-05,
|
| 1013 |
+
"loss": 0.7976,
|
| 1014 |
+
"step": 66500
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 5.85,
|
| 1018 |
+
"learning_rate": 4.5449061662198396e-05,
|
| 1019 |
+
"loss": 0.7953,
|
| 1020 |
+
"step": 67000
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 5.9,
|
| 1024 |
+
"learning_rate": 4.497031788586749e-05,
|
| 1025 |
+
"loss": 0.8018,
|
| 1026 |
+
"step": 67500
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 5.9,
|
| 1030 |
+
"eval_loss": 0.7478091716766357,
|
| 1031 |
+
"eval_runtime": 3109.9687,
|
| 1032 |
+
"eval_samples_per_second": 24.79,
|
| 1033 |
+
"eval_steps_per_second": 1.55,
|
| 1034 |
+
"step": 67500
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 5.94,
|
| 1038 |
+
"learning_rate": 4.4491574109536574e-05,
|
| 1039 |
+
"loss": 0.7939,
|
| 1040 |
+
"step": 68000
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 5.99,
|
| 1044 |
+
"learning_rate": 4.401283033320567e-05,
|
| 1045 |
+
"loss": 0.7844,
|
| 1046 |
+
"step": 68500
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 6.03,
|
| 1050 |
+
"learning_rate": 4.3534086556874766e-05,
|
| 1051 |
+
"loss": 0.7853,
|
| 1052 |
+
"step": 69000
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 6.07,
|
| 1056 |
+
"learning_rate": 4.305534278054386e-05,
|
| 1057 |
+
"loss": 0.7776,
|
| 1058 |
+
"step": 69500
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 6.12,
|
| 1062 |
+
"learning_rate": 4.2576599004212944e-05,
|
| 1063 |
+
"loss": 0.782,
|
| 1064 |
+
"step": 70000
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 6.12,
|
| 1068 |
+
"eval_loss": 0.7434529662132263,
|
| 1069 |
+
"eval_runtime": 3111.4708,
|
| 1070 |
+
"eval_samples_per_second": 24.778,
|
| 1071 |
+
"eval_steps_per_second": 1.549,
|
| 1072 |
+
"step": 70000
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 6.16,
|
| 1076 |
+
"learning_rate": 4.209785522788204e-05,
|
| 1077 |
+
"loss": 0.7812,
|
| 1078 |
+
"step": 70500
|
| 1079 |
+
},
|
| 1080 |
+
{
|
| 1081 |
+
"epoch": 6.2,
|
| 1082 |
+
"learning_rate": 4.161911145155113e-05,
|
| 1083 |
+
"loss": 0.7816,
|
| 1084 |
+
"step": 71000
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 6.25,
|
| 1088 |
+
"learning_rate": 4.1141325162772885e-05,
|
| 1089 |
+
"loss": 0.7745,
|
| 1090 |
+
"step": 71500
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 6.29,
|
| 1094 |
+
"learning_rate": 4.066258138644198e-05,
|
| 1095 |
+
"loss": 0.7776,
|
| 1096 |
+
"step": 72000
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 6.34,
|
| 1100 |
+
"learning_rate": 4.018383761011107e-05,
|
| 1101 |
+
"loss": 0.7743,
|
| 1102 |
+
"step": 72500
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 6.34,
|
| 1106 |
+
"eval_loss": 0.7367435097694397,
|
| 1107 |
+
"eval_runtime": 3109.6155,
|
| 1108 |
+
"eval_samples_per_second": 24.793,
|
| 1109 |
+
"eval_steps_per_second": 1.55,
|
| 1110 |
+
"step": 72500
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 6.38,
|
| 1114 |
+
"learning_rate": 3.970509383378016e-05,
|
| 1115 |
+
"loss": 0.7762,
|
| 1116 |
+
"step": 73000
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 6.42,
|
| 1120 |
+
"learning_rate": 3.9226350057449255e-05,
|
| 1121 |
+
"loss": 0.7728,
|
| 1122 |
+
"step": 73500
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 6.47,
|
| 1126 |
+
"learning_rate": 3.874760628111835e-05,
|
| 1127 |
+
"loss": 0.7743,
|
| 1128 |
+
"step": 74000
|
| 1129 |
+
},
|
| 1130 |
+
{
|
| 1131 |
+
"epoch": 6.51,
|
| 1132 |
+
"learning_rate": 3.826886250478744e-05,
|
| 1133 |
+
"loss": 0.7715,
|
| 1134 |
+
"step": 74500
|
| 1135 |
+
},
|
| 1136 |
+
{
|
| 1137 |
+
"epoch": 6.55,
|
| 1138 |
+
"learning_rate": 3.779011872845653e-05,
|
| 1139 |
+
"loss": 0.774,
|
| 1140 |
+
"step": 75000
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 6.55,
|
| 1144 |
+
"eval_loss": 0.7312998175621033,
|
| 1145 |
+
"eval_runtime": 3143.7734,
|
| 1146 |
+
"eval_samples_per_second": 24.523,
|
| 1147 |
+
"eval_steps_per_second": 1.533,
|
| 1148 |
+
"step": 75000
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 6.6,
|
| 1152 |
+
"learning_rate": 3.7311374952125625e-05,
|
| 1153 |
+
"loss": 0.7656,
|
| 1154 |
+
"step": 75500
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 6.64,
|
| 1158 |
+
"learning_rate": 3.683263117579472e-05,
|
| 1159 |
+
"loss": 0.7653,
|
| 1160 |
+
"step": 76000
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 6.68,
|
| 1164 |
+
"learning_rate": 3.635388739946381e-05,
|
| 1165 |
+
"loss": 0.7619,
|
| 1166 |
+
"step": 76500
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 6.73,
|
| 1170 |
+
"learning_rate": 3.587610111068556e-05,
|
| 1171 |
+
"loss": 0.765,
|
| 1172 |
+
"step": 77000
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 6.77,
|
| 1176 |
+
"learning_rate": 3.539735733435465e-05,
|
| 1177 |
+
"loss": 0.7692,
|
| 1178 |
+
"step": 77500
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 6.77,
|
| 1182 |
+
"eval_loss": 0.7270153164863586,
|
| 1183 |
+
"eval_runtime": 3083.4104,
|
| 1184 |
+
"eval_samples_per_second": 25.003,
|
| 1185 |
+
"eval_steps_per_second": 1.563,
|
| 1186 |
+
"step": 77500
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 6.82,
|
| 1190 |
+
"learning_rate": 3.491957104557641e-05,
|
| 1191 |
+
"loss": 0.7676,
|
| 1192 |
+
"step": 78000
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 6.86,
|
| 1196 |
+
"learning_rate": 3.44408272692455e-05,
|
| 1197 |
+
"loss": 0.7617,
|
| 1198 |
+
"step": 78500
|
| 1199 |
+
},
|
| 1200 |
+
{
|
| 1201 |
+
"epoch": 6.9,
|
| 1202 |
+
"learning_rate": 3.396208349291459e-05,
|
| 1203 |
+
"loss": 0.765,
|
| 1204 |
+
"step": 79000
|
| 1205 |
+
},
|
| 1206 |
+
{
|
| 1207 |
+
"epoch": 6.95,
|
| 1208 |
+
"learning_rate": 3.3483339716583685e-05,
|
| 1209 |
+
"loss": 0.7609,
|
| 1210 |
+
"step": 79500
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 6.99,
|
| 1214 |
+
"learning_rate": 3.300459594025278e-05,
|
| 1215 |
+
"loss": 0.7604,
|
| 1216 |
+
"step": 80000
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"epoch": 6.99,
|
| 1220 |
+
"eval_loss": 0.7200314998626709,
|
| 1221 |
+
"eval_runtime": 3077.861,
|
| 1222 |
+
"eval_samples_per_second": 25.049,
|
| 1223 |
+
"eval_steps_per_second": 1.566,
|
| 1224 |
+
"step": 80000
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 7.03,
|
| 1228 |
+
"learning_rate": 3.252585216392187e-05,
|
| 1229 |
+
"loss": 0.7513,
|
| 1230 |
+
"step": 80500
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 7.08,
|
| 1234 |
+
"learning_rate": 3.204710838759096e-05,
|
| 1235 |
+
"loss": 0.7549,
|
| 1236 |
+
"step": 81000
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 7.12,
|
| 1240 |
+
"learning_rate": 3.156932209881271e-05,
|
| 1241 |
+
"loss": 0.7485,
|
| 1242 |
+
"step": 81500
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 7.17,
|
| 1246 |
+
"learning_rate": 3.109057832248181e-05,
|
| 1247 |
+
"loss": 0.7486,
|
| 1248 |
+
"step": 82000
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 7.21,
|
| 1252 |
+
"learning_rate": 3.0611834546150903e-05,
|
| 1253 |
+
"loss": 0.7468,
|
| 1254 |
+
"step": 82500
|
| 1255 |
+
},
|
| 1256 |
+
{
|
| 1257 |
+
"epoch": 7.21,
|
| 1258 |
+
"eval_loss": 0.7163689136505127,
|
| 1259 |
+
"eval_runtime": 3165.721,
|
| 1260 |
+
"eval_samples_per_second": 24.353,
|
| 1261 |
+
"eval_steps_per_second": 1.522,
|
| 1262 |
+
"step": 82500
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 7.25,
|
| 1266 |
+
"learning_rate": 3.0133090769819993e-05,
|
| 1267 |
+
"loss": 0.7471,
|
| 1268 |
+
"step": 83000
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 7.3,
|
| 1272 |
+
"learning_rate": 2.9654346993489085e-05,
|
| 1273 |
+
"loss": 0.7473,
|
| 1274 |
+
"step": 83500
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 7.34,
|
| 1278 |
+
"learning_rate": 2.9175603217158177e-05,
|
| 1279 |
+
"loss": 0.747,
|
| 1280 |
+
"step": 84000
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 7.38,
|
| 1284 |
+
"learning_rate": 2.8696859440827273e-05,
|
| 1285 |
+
"loss": 0.743,
|
| 1286 |
+
"step": 84500
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 7.43,
|
| 1290 |
+
"learning_rate": 2.8219073152049026e-05,
|
| 1291 |
+
"loss": 0.7486,
|
| 1292 |
+
"step": 85000
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 7.43,
|
| 1296 |
+
"eval_loss": 0.7117038369178772,
|
| 1297 |
+
"eval_runtime": 3102.6798,
|
| 1298 |
+
"eval_samples_per_second": 24.848,
|
| 1299 |
+
"eval_steps_per_second": 1.553,
|
| 1300 |
+
"step": 85000
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 7.47,
|
| 1304 |
+
"learning_rate": 2.7740329375718115e-05,
|
| 1305 |
+
"loss": 0.7411,
|
| 1306 |
+
"step": 85500
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 7.51,
|
| 1310 |
+
"learning_rate": 2.7261585599387208e-05,
|
| 1311 |
+
"loss": 0.7418,
|
| 1312 |
+
"step": 86000
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 7.56,
|
| 1316 |
+
"learning_rate": 2.6782841823056303e-05,
|
| 1317 |
+
"loss": 0.7443,
|
| 1318 |
+
"step": 86500
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 7.6,
|
| 1322 |
+
"learning_rate": 2.6305055534278056e-05,
|
| 1323 |
+
"loss": 0.7403,
|
| 1324 |
+
"step": 87000
|
| 1325 |
+
},
|
| 1326 |
+
{
|
| 1327 |
+
"epoch": 7.65,
|
| 1328 |
+
"learning_rate": 2.582631175794715e-05,
|
| 1329 |
+
"loss": 0.7399,
|
| 1330 |
+
"step": 87500
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 7.65,
|
| 1334 |
+
"eval_loss": 0.7042549252510071,
|
| 1335 |
+
"eval_runtime": 3110.6014,
|
| 1336 |
+
"eval_samples_per_second": 24.785,
|
| 1337 |
+
"eval_steps_per_second": 1.549,
|
| 1338 |
+
"step": 87500
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 7.69,
|
| 1342 |
+
"learning_rate": 2.5347567981616238e-05,
|
| 1343 |
+
"loss": 0.7393,
|
| 1344 |
+
"step": 88000
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 7.73,
|
| 1348 |
+
"learning_rate": 2.4868824205285333e-05,
|
| 1349 |
+
"loss": 0.7404,
|
| 1350 |
+
"step": 88500
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 7.78,
|
| 1354 |
+
"learning_rate": 2.4390080428954426e-05,
|
| 1355 |
+
"loss": 0.7356,
|
| 1356 |
+
"step": 89000
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 7.82,
|
| 1360 |
+
"learning_rate": 2.391133665262352e-05,
|
| 1361 |
+
"loss": 0.7395,
|
| 1362 |
+
"step": 89500
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 7.86,
|
| 1366 |
+
"learning_rate": 2.3432592876292607e-05,
|
| 1367 |
+
"loss": 0.7306,
|
| 1368 |
+
"step": 90000
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 7.86,
|
| 1372 |
+
"eval_loss": 0.6955912709236145,
|
| 1373 |
+
"eval_runtime": 3139.2914,
|
| 1374 |
+
"eval_samples_per_second": 24.558,
|
| 1375 |
+
"eval_steps_per_second": 1.535,
|
| 1376 |
+
"step": 90000
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 7.91,
|
| 1380 |
+
"learning_rate": 2.2953849099961703e-05,
|
| 1381 |
+
"loss": 0.735,
|
| 1382 |
+
"step": 90500
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 7.95,
|
| 1386 |
+
"learning_rate": 2.2475105323630792e-05,
|
| 1387 |
+
"loss": 0.7281,
|
| 1388 |
+
"step": 91000
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 8.0,
|
| 1392 |
+
"learning_rate": 2.1996361547299888e-05,
|
| 1393 |
+
"loss": 0.732,
|
| 1394 |
+
"step": 91500
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 8.04,
|
| 1398 |
+
"learning_rate": 2.1517617770968977e-05,
|
| 1399 |
+
"loss": 0.7219,
|
| 1400 |
+
"step": 92000
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 8.08,
|
| 1404 |
+
"learning_rate": 2.1039831482190733e-05,
|
| 1405 |
+
"loss": 0.7243,
|
| 1406 |
+
"step": 92500
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 8.08,
|
| 1410 |
+
"eval_loss": 0.695923924446106,
|
| 1411 |
+
"eval_runtime": 3092.5373,
|
| 1412 |
+
"eval_samples_per_second": 24.93,
|
| 1413 |
+
"eval_steps_per_second": 1.558,
|
| 1414 |
+
"step": 92500
|
| 1415 |
+
},
|
| 1416 |
+
{
|
| 1417 |
+
"epoch": 8.13,
|
| 1418 |
+
"learning_rate": 2.0561087705859826e-05,
|
| 1419 |
+
"loss": 0.7243,
|
| 1420 |
+
"step": 93000
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 8.17,
|
| 1424 |
+
"learning_rate": 2.0082343929528918e-05,
|
| 1425 |
+
"loss": 0.7258,
|
| 1426 |
+
"step": 93500
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 8.21,
|
| 1430 |
+
"learning_rate": 1.960360015319801e-05,
|
| 1431 |
+
"loss": 0.7211,
|
| 1432 |
+
"step": 94000
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 8.26,
|
| 1436 |
+
"learning_rate": 1.91248563768671e-05,
|
| 1437 |
+
"loss": 0.7214,
|
| 1438 |
+
"step": 94500
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 8.3,
|
| 1442 |
+
"learning_rate": 1.8646112600536196e-05,
|
| 1443 |
+
"loss": 0.7132,
|
| 1444 |
+
"step": 95000
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 8.3,
|
| 1448 |
+
"eval_loss": 0.6916212439537048,
|
| 1449 |
+
"eval_runtime": 3163.2312,
|
| 1450 |
+
"eval_samples_per_second": 24.373,
|
| 1451 |
+
"eval_steps_per_second": 1.523,
|
| 1452 |
+
"step": 95000
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 8.34,
|
| 1456 |
+
"learning_rate": 1.8167368824205285e-05,
|
| 1457 |
+
"loss": 0.7188,
|
| 1458 |
+
"step": 95500
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 8.39,
|
| 1462 |
+
"learning_rate": 1.768862504787438e-05,
|
| 1463 |
+
"loss": 0.717,
|
| 1464 |
+
"step": 96000
|
| 1465 |
+
},
|
| 1466 |
+
{
|
| 1467 |
+
"epoch": 8.43,
|
| 1468 |
+
"learning_rate": 1.7210838759096133e-05,
|
| 1469 |
+
"loss": 0.7189,
|
| 1470 |
+
"step": 96500
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 8.48,
|
| 1474 |
+
"learning_rate": 1.6732094982765226e-05,
|
| 1475 |
+
"loss": 0.7206,
|
| 1476 |
+
"step": 97000
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 8.52,
|
| 1480 |
+
"learning_rate": 1.625430869398698e-05,
|
| 1481 |
+
"loss": 0.71,
|
| 1482 |
+
"step": 97500
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 8.52,
|
| 1486 |
+
"eval_loss": 0.6853311061859131,
|
| 1487 |
+
"eval_runtime": 3146.2403,
|
| 1488 |
+
"eval_samples_per_second": 24.504,
|
| 1489 |
+
"eval_steps_per_second": 1.532,
|
| 1490 |
+
"step": 97500
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 8.56,
|
| 1494 |
+
"learning_rate": 1.577556491765607e-05,
|
| 1495 |
+
"loss": 0.7148,
|
| 1496 |
+
"step": 98000
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 8.61,
|
| 1500 |
+
"learning_rate": 1.5296821141325163e-05,
|
| 1501 |
+
"loss": 0.7169,
|
| 1502 |
+
"step": 98500
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 8.65,
|
| 1506 |
+
"learning_rate": 1.4818077364994256e-05,
|
| 1507 |
+
"loss": 0.7155,
|
| 1508 |
+
"step": 99000
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 8.69,
|
| 1512 |
+
"learning_rate": 1.4339333588663348e-05,
|
| 1513 |
+
"loss": 0.713,
|
| 1514 |
+
"step": 99500
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 8.74,
|
| 1518 |
+
"learning_rate": 1.3860589812332439e-05,
|
| 1519 |
+
"loss": 0.7128,
|
| 1520 |
+
"step": 100000
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 8.74,
|
| 1524 |
+
"eval_loss": 0.6855071783065796,
|
| 1525 |
+
"eval_runtime": 3110.846,
|
| 1526 |
+
"eval_samples_per_second": 24.783,
|
| 1527 |
+
"eval_steps_per_second": 1.549,
|
| 1528 |
+
"step": 100000
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 8.78,
|
| 1532 |
+
"learning_rate": 1.3381846036001533e-05,
|
| 1533 |
+
"loss": 0.7107,
|
| 1534 |
+
"step": 100500
|
| 1535 |
+
},
|
| 1536 |
+
{
|
| 1537 |
+
"epoch": 8.83,
|
| 1538 |
+
"learning_rate": 1.2903102259670624e-05,
|
| 1539 |
+
"loss": 0.7142,
|
| 1540 |
+
"step": 101000
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 8.87,
|
| 1544 |
+
"learning_rate": 1.2424358483339716e-05,
|
| 1545 |
+
"loss": 0.7035,
|
| 1546 |
+
"step": 101500
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 8.91,
|
| 1550 |
+
"learning_rate": 1.1945614707008809e-05,
|
| 1551 |
+
"loss": 0.7061,
|
| 1552 |
+
"step": 102000
|
| 1553 |
+
},
|
| 1554 |
+
{
|
| 1555 |
+
"epoch": 8.96,
|
| 1556 |
+
"learning_rate": 1.1467828418230563e-05,
|
| 1557 |
+
"loss": 0.7088,
|
| 1558 |
+
"step": 102500
|
| 1559 |
+
},
|
| 1560 |
+
{
|
| 1561 |
+
"epoch": 8.96,
|
| 1562 |
+
"eval_loss": 0.680884063243866,
|
| 1563 |
+
"eval_runtime": 3113.4696,
|
| 1564 |
+
"eval_samples_per_second": 24.762,
|
| 1565 |
+
"eval_steps_per_second": 1.548,
|
| 1566 |
+
"step": 102500
|
| 1567 |
+
},
|
| 1568 |
+
{
|
| 1569 |
+
"epoch": 9.0,
|
| 1570 |
+
"learning_rate": 1.0989084641899656e-05,
|
| 1571 |
+
"loss": 0.7039,
|
| 1572 |
+
"step": 103000
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"epoch": 9.04,
|
| 1576 |
+
"learning_rate": 1.0510340865568748e-05,
|
| 1577 |
+
"loss": 0.6983,
|
| 1578 |
+
"step": 103500
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 9.09,
|
| 1582 |
+
"learning_rate": 1.0032554576790503e-05,
|
| 1583 |
+
"loss": 0.7028,
|
| 1584 |
+
"step": 104000
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 9.13,
|
| 1588 |
+
"learning_rate": 9.553810800459595e-06,
|
| 1589 |
+
"loss": 0.7002,
|
| 1590 |
+
"step": 104500
|
| 1591 |
+
},
|
| 1592 |
+
{
|
| 1593 |
+
"epoch": 9.18,
|
| 1594 |
+
"learning_rate": 9.075067024128686e-06,
|
| 1595 |
+
"loss": 0.7002,
|
| 1596 |
+
"step": 105000
|
| 1597 |
+
},
|
| 1598 |
+
{
|
| 1599 |
+
"epoch": 9.18,
|
| 1600 |
+
"eval_loss": 0.6783619523048401,
|
| 1601 |
+
"eval_runtime": 3133.9738,
|
| 1602 |
+
"eval_samples_per_second": 24.6,
|
| 1603 |
+
"eval_steps_per_second": 1.538,
|
| 1604 |
+
"step": 105000
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 9.22,
|
| 1608 |
+
"learning_rate": 8.596323247797778e-06,
|
| 1609 |
+
"loss": 0.699,
|
| 1610 |
+
"step": 105500
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 9.26,
|
| 1614 |
+
"learning_rate": 8.11757947146687e-06,
|
| 1615 |
+
"loss": 0.6927,
|
| 1616 |
+
"step": 106000
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 9.31,
|
| 1620 |
+
"learning_rate": 7.638835695135963e-06,
|
| 1621 |
+
"loss": 0.6976,
|
| 1622 |
+
"step": 106500
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 9.35,
|
| 1626 |
+
"learning_rate": 7.1600919188050565e-06,
|
| 1627 |
+
"loss": 0.6978,
|
| 1628 |
+
"step": 107000
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 9.39,
|
| 1632 |
+
"learning_rate": 6.681348142474147e-06,
|
| 1633 |
+
"loss": 0.6953,
|
| 1634 |
+
"step": 107500
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 9.39,
|
| 1638 |
+
"eval_loss": 0.6736627221107483,
|
| 1639 |
+
"eval_runtime": 3133.4547,
|
| 1640 |
+
"eval_samples_per_second": 24.604,
|
| 1641 |
+
"eval_steps_per_second": 1.538,
|
| 1642 |
+
"step": 107500
|
| 1643 |
+
},
|
| 1644 |
+
{
|
| 1645 |
+
"epoch": 9.44,
|
| 1646 |
+
"learning_rate": 6.2026043661432406e-06,
|
| 1647 |
+
"loss": 0.6892,
|
| 1648 |
+
"step": 108000
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 9.48,
|
| 1652 |
+
"learning_rate": 5.723860589812333e-06,
|
| 1653 |
+
"loss": 0.6989,
|
| 1654 |
+
"step": 108500
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 9.52,
|
| 1658 |
+
"learning_rate": 5.2451168134814254e-06,
|
| 1659 |
+
"loss": 0.6975,
|
| 1660 |
+
"step": 109000
|
| 1661 |
+
},
|
| 1662 |
+
{
|
| 1663 |
+
"epoch": 9.57,
|
| 1664 |
+
"learning_rate": 4.767330524703179e-06,
|
| 1665 |
+
"loss": 0.6977,
|
| 1666 |
+
"step": 109500
|
| 1667 |
+
},
|
| 1668 |
+
{
|
| 1669 |
+
"epoch": 9.61,
|
| 1670 |
+
"learning_rate": 4.288586748372271e-06,
|
| 1671 |
+
"loss": 0.695,
|
| 1672 |
+
"step": 110000
|
| 1673 |
+
},
|
| 1674 |
+
{
|
| 1675 |
+
"epoch": 9.61,
|
| 1676 |
+
"eval_loss": 0.6714410185813904,
|
| 1677 |
+
"eval_runtime": 3139.9991,
|
| 1678 |
+
"eval_samples_per_second": 24.553,
|
| 1679 |
+
"eval_steps_per_second": 1.535,
|
| 1680 |
+
"step": 110000
|
| 1681 |
+
},
|
| 1682 |
+
{
|
| 1683 |
+
"epoch": 9.66,
|
| 1684 |
+
"learning_rate": 3.810800459594025e-06,
|
| 1685 |
+
"loss": 0.6947,
|
| 1686 |
+
"step": 110500
|
| 1687 |
+
},
|
| 1688 |
+
{
|
| 1689 |
+
"epoch": 9.7,
|
| 1690 |
+
"learning_rate": 3.3320566832631176e-06,
|
| 1691 |
+
"loss": 0.6967,
|
| 1692 |
+
"step": 111000
|
| 1693 |
+
},
|
| 1694 |
+
{
|
| 1695 |
+
"epoch": 9.74,
|
| 1696 |
+
"learning_rate": 2.85331290693221e-06,
|
| 1697 |
+
"loss": 0.6904,
|
| 1698 |
+
"step": 111500
|
| 1699 |
+
},
|
| 1700 |
+
{
|
| 1701 |
+
"epoch": 9.79,
|
| 1702 |
+
"learning_rate": 2.3745691306013025e-06,
|
| 1703 |
+
"loss": 0.6949,
|
| 1704 |
+
"step": 112000
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 9.83,
|
| 1708 |
+
"learning_rate": 1.8958253542703947e-06,
|
| 1709 |
+
"loss": 0.6871,
|
| 1710 |
+
"step": 112500
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 9.83,
|
| 1714 |
+
"eval_loss": 0.6687204837799072,
|
| 1715 |
+
"eval_runtime": 3111.3683,
|
| 1716 |
+
"eval_samples_per_second": 24.779,
|
| 1717 |
+
"eval_steps_per_second": 1.549,
|
| 1718 |
+
"step": 112500
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 9.87,
|
| 1722 |
+
"learning_rate": 3.6311394284300384e-05,
|
| 1723 |
+
"loss": 0.7034,
|
| 1724 |
+
"step": 113000
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 9.92,
|
| 1728 |
+
"learning_rate": 3.600210317951256e-05,
|
| 1729 |
+
"loss": 0.7149,
|
| 1730 |
+
"step": 113500
|
| 1731 |
+
},
|
| 1732 |
+
{
|
| 1733 |
+
"epoch": 9.96,
|
| 1734 |
+
"learning_rate": 3.5692812074724736e-05,
|
| 1735 |
+
"loss": 0.716,
|
| 1736 |
+
"step": 114000
|
| 1737 |
+
},
|
| 1738 |
+
{
|
| 1739 |
+
"epoch": 10.01,
|
| 1740 |
+
"learning_rate": 3.5383520969936905e-05,
|
| 1741 |
+
"loss": 0.7236,
|
| 1742 |
+
"step": 114500
|
| 1743 |
+
},
|
| 1744 |
+
{
|
| 1745 |
+
"epoch": 10.05,
|
| 1746 |
+
"learning_rate": 3.5074229865149074e-05,
|
| 1747 |
+
"loss": 0.7161,
|
| 1748 |
+
"step": 115000
|
| 1749 |
+
},
|
| 1750 |
+
{
|
| 1751 |
+
"epoch": 10.05,
|
| 1752 |
+
"eval_loss": 0.6960982084274292,
|
| 1753 |
+
"eval_runtime": 3127.0335,
|
| 1754 |
+
"eval_samples_per_second": 24.655,
|
| 1755 |
+
"eval_steps_per_second": 1.541,
|
| 1756 |
+
"step": 115000
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 10.09,
|
| 1760 |
+
"learning_rate": 3.476493876036126e-05,
|
| 1761 |
+
"loss": 0.7149,
|
| 1762 |
+
"step": 115500
|
| 1763 |
+
},
|
| 1764 |
+
{
|
| 1765 |
+
"epoch": 10.14,
|
| 1766 |
+
"learning_rate": 3.4455647655573426e-05,
|
| 1767 |
+
"loss": 0.7203,
|
| 1768 |
+
"step": 116000
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 10.18,
|
| 1772 |
+
"learning_rate": 3.41463565507856e-05,
|
| 1773 |
+
"loss": 0.7259,
|
| 1774 |
+
"step": 116500
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 10.22,
|
| 1778 |
+
"learning_rate": 3.383706544599777e-05,
|
| 1779 |
+
"loss": 0.7213,
|
| 1780 |
+
"step": 117000
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 10.27,
|
| 1784 |
+
"learning_rate": 3.35290115056291e-05,
|
| 1785 |
+
"loss": 0.7265,
|
| 1786 |
+
"step": 117500
|
| 1787 |
+
},
|
| 1788 |
+
{
|
| 1789 |
+
"epoch": 10.27,
|
| 1790 |
+
"eval_loss": 0.7006255984306335,
|
| 1791 |
+
"eval_runtime": 3139.1903,
|
| 1792 |
+
"eval_samples_per_second": 24.559,
|
| 1793 |
+
"eval_steps_per_second": 1.535,
|
| 1794 |
+
"step": 117500
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 10.31,
|
| 1798 |
+
"learning_rate": 3.321972040084128e-05,
|
| 1799 |
+
"loss": 0.7228,
|
| 1800 |
+
"step": 118000
|
| 1801 |
+
},
|
| 1802 |
+
{
|
| 1803 |
+
"epoch": 10.35,
|
| 1804 |
+
"learning_rate": 3.2910429296053447e-05,
|
| 1805 |
+
"loss": 0.7214,
|
| 1806 |
+
"step": 118500
|
| 1807 |
+
},
|
| 1808 |
+
{
|
| 1809 |
+
"epoch": 10.4,
|
| 1810 |
+
"learning_rate": 3.260113819126562e-05,
|
| 1811 |
+
"loss": 0.7309,
|
| 1812 |
+
"step": 119000
|
| 1813 |
+
},
|
| 1814 |
+
{
|
| 1815 |
+
"epoch": 10.44,
|
| 1816 |
+
"learning_rate": 3.229184708647779e-05,
|
| 1817 |
+
"loss": 0.7208,
|
| 1818 |
+
"step": 119500
|
| 1819 |
+
},
|
| 1820 |
+
{
|
| 1821 |
+
"epoch": 10.49,
|
| 1822 |
+
"learning_rate": 3.198255598168997e-05,
|
| 1823 |
+
"loss": 0.7284,
|
| 1824 |
+
"step": 120000
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"epoch": 10.49,
|
| 1828 |
+
"eval_loss": 0.6941153407096863,
|
| 1829 |
+
"eval_runtime": 3132.0931,
|
| 1830 |
+
"eval_samples_per_second": 24.615,
|
| 1831 |
+
"eval_steps_per_second": 1.539,
|
| 1832 |
+
"step": 120000
|
| 1833 |
+
},
|
| 1834 |
+
{
|
| 1835 |
+
"epoch": 10.53,
|
| 1836 |
+
"learning_rate": 3.167326487690214e-05,
|
| 1837 |
+
"loss": 0.7233,
|
| 1838 |
+
"step": 120500
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 10.57,
|
| 1842 |
+
"learning_rate": 3.136397377211431e-05,
|
| 1843 |
+
"loss": 0.7219,
|
| 1844 |
+
"step": 121000
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 10.62,
|
| 1848 |
+
"learning_rate": 3.105468266732649e-05,
|
| 1849 |
+
"loss": 0.7163,
|
| 1850 |
+
"step": 121500
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 10.66,
|
| 1854 |
+
"learning_rate": 3.0745391562538664e-05,
|
| 1855 |
+
"loss": 0.7223,
|
| 1856 |
+
"step": 122000
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 10.7,
|
| 1860 |
+
"learning_rate": 3.0436100457750836e-05,
|
| 1861 |
+
"loss": 0.724,
|
| 1862 |
+
"step": 122500
|
| 1863 |
+
},
|
| 1864 |
+
{
|
| 1865 |
+
"epoch": 10.7,
|
| 1866 |
+
"eval_loss": 0.6886956095695496,
|
| 1867 |
+
"eval_runtime": 3117.6812,
|
| 1868 |
+
"eval_samples_per_second": 24.729,
|
| 1869 |
+
"eval_steps_per_second": 1.546,
|
| 1870 |
+
"step": 122500
|
| 1871 |
+
},
|
| 1872 |
+
{
|
| 1873 |
+
"epoch": 10.75,
|
| 1874 |
+
"learning_rate": 3.012680935296301e-05,
|
| 1875 |
+
"loss": 0.7181,
|
| 1876 |
+
"step": 123000
|
| 1877 |
+
},
|
| 1878 |
+
{
|
| 1879 |
+
"epoch": 10.79,
|
| 1880 |
+
"learning_rate": 2.9817518248175185e-05,
|
| 1881 |
+
"loss": 0.7215,
|
| 1882 |
+
"step": 123500
|
| 1883 |
+
},
|
| 1884 |
+
{
|
| 1885 |
+
"epoch": 10.84,
|
| 1886 |
+
"learning_rate": 2.9508845725596935e-05,
|
| 1887 |
+
"loss": 0.726,
|
| 1888 |
+
"step": 124000
|
| 1889 |
+
},
|
| 1890 |
+
{
|
| 1891 |
+
"epoch": 10.88,
|
| 1892 |
+
"learning_rate": 2.9199554620809104e-05,
|
| 1893 |
+
"loss": 0.7205,
|
| 1894 |
+
"step": 124500
|
| 1895 |
+
},
|
| 1896 |
+
{
|
| 1897 |
+
"epoch": 10.92,
|
| 1898 |
+
"learning_rate": 2.8890882098230854e-05,
|
| 1899 |
+
"loss": 0.7266,
|
| 1900 |
+
"step": 125000
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 10.92,
|
| 1904 |
+
"eval_loss": 0.6931244134902954,
|
| 1905 |
+
"eval_runtime": 3180.6149,
|
| 1906 |
+
"eval_samples_per_second": 24.239,
|
| 1907 |
+
"eval_steps_per_second": 1.515,
|
| 1908 |
+
"step": 125000
|
| 1909 |
+
},
|
| 1910 |
+
{
|
| 1911 |
+
"epoch": 10.97,
|
| 1912 |
+
"learning_rate": 2.858159099344303e-05,
|
| 1913 |
+
"loss": 0.7214,
|
| 1914 |
+
"step": 125500
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 11.01,
|
| 1918 |
+
"learning_rate": 2.8272299888655202e-05,
|
| 1919 |
+
"loss": 0.7165,
|
| 1920 |
+
"step": 126000
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 11.05,
|
| 1924 |
+
"learning_rate": 2.7963627366076955e-05,
|
| 1925 |
+
"loss": 0.7044,
|
| 1926 |
+
"step": 126500
|
| 1927 |
+
},
|
| 1928 |
+
{
|
| 1929 |
+
"epoch": 11.1,
|
| 1930 |
+
"learning_rate": 2.7654954843498705e-05,
|
| 1931 |
+
"loss": 0.7071,
|
| 1932 |
+
"step": 127000
|
| 1933 |
+
},
|
| 1934 |
+
{
|
| 1935 |
+
"epoch": 11.14,
|
| 1936 |
+
"learning_rate": 2.7345663738710874e-05,
|
| 1937 |
+
"loss": 0.7051,
|
| 1938 |
+
"step": 127500
|
| 1939 |
+
},
|
| 1940 |
+
{
|
| 1941 |
+
"epoch": 11.14,
|
| 1942 |
+
"eval_loss": 0.6846074461936951,
|
| 1943 |
+
"eval_runtime": 3127.4042,
|
| 1944 |
+
"eval_samples_per_second": 24.652,
|
| 1945 |
+
"eval_steps_per_second": 1.541,
|
| 1946 |
+
"step": 127500
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"epoch": 11.18,
|
| 1950 |
+
"learning_rate": 2.703637263392305e-05,
|
| 1951 |
+
"loss": 0.7139,
|
| 1952 |
+
"step": 128000
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 11.23,
|
| 1956 |
+
"learning_rate": 2.6727081529135223e-05,
|
| 1957 |
+
"loss": 0.7057,
|
| 1958 |
+
"step": 128500
|
| 1959 |
+
},
|
| 1960 |
+
{
|
| 1961 |
+
"epoch": 11.27,
|
| 1962 |
+
"learning_rate": 2.64177904243474e-05,
|
| 1963 |
+
"loss": 0.7104,
|
| 1964 |
+
"step": 129000
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 11.32,
|
| 1968 |
+
"learning_rate": 2.610849931955957e-05,
|
| 1969 |
+
"loss": 0.7036,
|
| 1970 |
+
"step": 129500
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 11.36,
|
| 1974 |
+
"learning_rate": 2.5799208214771743e-05,
|
| 1975 |
+
"loss": 0.7106,
|
| 1976 |
+
"step": 130000
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 11.36,
|
| 1980 |
+
"eval_loss": 0.68162602186203,
|
| 1981 |
+
"eval_runtime": 3071.6142,
|
| 1982 |
+
"eval_samples_per_second": 25.1,
|
| 1983 |
+
"eval_steps_per_second": 1.569,
|
| 1984 |
+
"step": 130000
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 11.4,
|
| 1988 |
+
"learning_rate": 2.548991710998392e-05,
|
| 1989 |
+
"loss": 0.7072,
|
| 1990 |
+
"step": 130500
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 11.45,
|
| 1994 |
+
"learning_rate": 2.5180626005196088e-05,
|
| 1995 |
+
"loss": 0.7073,
|
| 1996 |
+
"step": 131000
|
| 1997 |
+
},
|
| 1998 |
+
{
|
| 1999 |
+
"epoch": 11.49,
|
| 2000 |
+
"learning_rate": 2.4871334900408268e-05,
|
| 2001 |
+
"loss": 0.7109,
|
| 2002 |
+
"step": 131500
|
| 2003 |
+
},
|
| 2004 |
+
{
|
| 2005 |
+
"epoch": 11.53,
|
| 2006 |
+
"learning_rate": 2.456204379562044e-05,
|
| 2007 |
+
"loss": 0.7041,
|
| 2008 |
+
"step": 132000
|
| 2009 |
+
},
|
| 2010 |
+
{
|
| 2011 |
+
"epoch": 11.58,
|
| 2012 |
+
"learning_rate": 2.4252752690832612e-05,
|
| 2013 |
+
"loss": 0.7011,
|
| 2014 |
+
"step": 132500
|
| 2015 |
+
},
|
| 2016 |
+
{
|
| 2017 |
+
"epoch": 11.58,
|
| 2018 |
+
"eval_loss": 0.6830089688301086,
|
| 2019 |
+
"eval_runtime": 3073.7046,
|
| 2020 |
+
"eval_samples_per_second": 25.082,
|
| 2021 |
+
"eval_steps_per_second": 1.568,
|
| 2022 |
+
"step": 132500
|
| 2023 |
+
},
|
| 2024 |
+
{
|
| 2025 |
+
"epoch": 11.62,
|
| 2026 |
+
"learning_rate": 2.3943461586044785e-05,
|
| 2027 |
+
"loss": 0.7041,
|
| 2028 |
+
"step": 133000
|
| 2029 |
+
},
|
| 2030 |
+
{
|
| 2031 |
+
"epoch": 11.67,
|
| 2032 |
+
"learning_rate": 2.3634789063466535e-05,
|
| 2033 |
+
"loss": 0.7057,
|
| 2034 |
+
"step": 133500
|
| 2035 |
+
},
|
| 2036 |
+
{
|
| 2037 |
+
"epoch": 11.71,
|
| 2038 |
+
"learning_rate": 2.332549795867871e-05,
|
| 2039 |
+
"loss": 0.7068,
|
| 2040 |
+
"step": 134000
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 11.75,
|
| 2044 |
+
"learning_rate": 2.3016825436100457e-05,
|
| 2045 |
+
"loss": 0.7073,
|
| 2046 |
+
"step": 134500
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 11.8,
|
| 2050 |
+
"learning_rate": 2.2707534331312633e-05,
|
| 2051 |
+
"loss": 0.6997,
|
| 2052 |
+
"step": 135000
|
| 2053 |
+
},
|
| 2054 |
+
{
|
| 2055 |
+
"epoch": 11.8,
|
| 2056 |
+
"eval_loss": 0.6784160137176514,
|
| 2057 |
+
"eval_runtime": 3076.1607,
|
| 2058 |
+
"eval_samples_per_second": 25.062,
|
| 2059 |
+
"eval_steps_per_second": 1.567,
|
| 2060 |
+
"step": 135000
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 11.84,
|
| 2064 |
+
"learning_rate": 2.2398243226524806e-05,
|
| 2065 |
+
"loss": 0.7062,
|
| 2066 |
+
"step": 135500
|
| 2067 |
+
},
|
| 2068 |
+
{
|
| 2069 |
+
"epoch": 11.88,
|
| 2070 |
+
"learning_rate": 2.208895212173698e-05,
|
| 2071 |
+
"loss": 0.6982,
|
| 2072 |
+
"step": 136000
|
| 2073 |
+
},
|
| 2074 |
+
{
|
| 2075 |
+
"epoch": 11.93,
|
| 2076 |
+
"learning_rate": 2.1779661016949154e-05,
|
| 2077 |
+
"loss": 0.7,
|
| 2078 |
+
"step": 136500
|
| 2079 |
+
},
|
| 2080 |
+
{
|
| 2081 |
+
"epoch": 11.97,
|
| 2082 |
+
"learning_rate": 2.147036991216133e-05,
|
| 2083 |
+
"loss": 0.6981,
|
| 2084 |
+
"step": 137000
|
| 2085 |
+
},
|
| 2086 |
+
{
|
| 2087 |
+
"epoch": 12.02,
|
| 2088 |
+
"learning_rate": 2.1161697389583076e-05,
|
| 2089 |
+
"loss": 0.6969,
|
| 2090 |
+
"step": 137500
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"epoch": 12.02,
|
| 2094 |
+
"eval_loss": 0.673393189907074,
|
| 2095 |
+
"eval_runtime": 3053.9536,
|
| 2096 |
+
"eval_samples_per_second": 25.245,
|
| 2097 |
+
"eval_steps_per_second": 1.578,
|
| 2098 |
+
"step": 137500
|
| 2099 |
+
},
|
| 2100 |
+
{
|
| 2101 |
+
"epoch": 12.06,
|
| 2102 |
+
"learning_rate": 2.0852406284795252e-05,
|
| 2103 |
+
"loss": 0.6961,
|
| 2104 |
+
"step": 138000
|
| 2105 |
+
},
|
| 2106 |
+
{
|
| 2107 |
+
"epoch": 12.1,
|
| 2108 |
+
"learning_rate": 2.0543115180007425e-05,
|
| 2109 |
+
"loss": 0.6941,
|
| 2110 |
+
"step": 138500
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 12.15,
|
| 2114 |
+
"learning_rate": 2.0233824075219597e-05,
|
| 2115 |
+
"loss": 0.6938,
|
| 2116 |
+
"step": 139000
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 12.19,
|
| 2120 |
+
"learning_rate": 1.9924532970431773e-05,
|
| 2121 |
+
"loss": 0.6968,
|
| 2122 |
+
"step": 139500
|
| 2123 |
+
},
|
| 2124 |
+
{
|
| 2125 |
+
"epoch": 12.23,
|
| 2126 |
+
"learning_rate": 1.9615241865643945e-05,
|
| 2127 |
+
"loss": 0.6968,
|
| 2128 |
+
"step": 140000
|
| 2129 |
+
},
|
| 2130 |
+
{
|
| 2131 |
+
"epoch": 12.23,
|
| 2132 |
+
"eval_loss": 0.6708864569664001,
|
| 2133 |
+
"eval_runtime": 3020.4771,
|
| 2134 |
+
"eval_samples_per_second": 25.524,
|
| 2135 |
+
"eval_steps_per_second": 1.595,
|
| 2136 |
+
"step": 140000
|
| 2137 |
+
},
|
| 2138 |
+
{
|
| 2139 |
+
"epoch": 12.28,
|
| 2140 |
+
"learning_rate": 1.9305950760856118e-05,
|
| 2141 |
+
"loss": 0.6928,
|
| 2142 |
+
"step": 140500
|
| 2143 |
+
},
|
| 2144 |
+
{
|
| 2145 |
+
"epoch": 12.32,
|
| 2146 |
+
"learning_rate": 1.899665965606829e-05,
|
| 2147 |
+
"loss": 0.6975,
|
| 2148 |
+
"step": 141000
|
| 2149 |
+
},
|
| 2150 |
+
{
|
| 2151 |
+
"epoch": 12.36,
|
| 2152 |
+
"learning_rate": 1.8687368551280466e-05,
|
| 2153 |
+
"loss": 0.6912,
|
| 2154 |
+
"step": 141500
|
| 2155 |
+
},
|
| 2156 |
+
{
|
| 2157 |
+
"epoch": 12.41,
|
| 2158 |
+
"learning_rate": 1.837807744649264e-05,
|
| 2159 |
+
"loss": 0.6923,
|
| 2160 |
+
"step": 142000
|
| 2161 |
+
},
|
| 2162 |
+
{
|
| 2163 |
+
"epoch": 12.45,
|
| 2164 |
+
"learning_rate": 1.8068786341704814e-05,
|
| 2165 |
+
"loss": 0.6867,
|
| 2166 |
+
"step": 142500
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 12.45,
|
| 2170 |
+
"eval_loss": 0.6655944585800171,
|
| 2171 |
+
"eval_runtime": 3033.3682,
|
| 2172 |
+
"eval_samples_per_second": 25.416,
|
| 2173 |
+
"eval_steps_per_second": 1.589,
|
| 2174 |
+
"step": 142500
|
| 2175 |
+
},
|
| 2176 |
+
{
|
| 2177 |
+
"epoch": 12.5,
|
| 2178 |
+
"learning_rate": 1.7759495236916987e-05,
|
| 2179 |
+
"loss": 0.6894,
|
| 2180 |
+
"step": 143000
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 12.54,
|
| 2184 |
+
"learning_rate": 1.745020413212916e-05,
|
| 2185 |
+
"loss": 0.6861,
|
| 2186 |
+
"step": 143500
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 12.58,
|
| 2190 |
+
"learning_rate": 1.714153160955091e-05,
|
| 2191 |
+
"loss": 0.6917,
|
| 2192 |
+
"step": 144000
|
| 2193 |
+
},
|
| 2194 |
+
{
|
| 2195 |
+
"epoch": 12.63,
|
| 2196 |
+
"learning_rate": 1.6832240504763082e-05,
|
| 2197 |
+
"loss": 0.688,
|
| 2198 |
+
"step": 144500
|
| 2199 |
+
},
|
| 2200 |
+
{
|
| 2201 |
+
"epoch": 12.67,
|
| 2202 |
+
"learning_rate": 1.6522949399975258e-05,
|
| 2203 |
+
"loss": 0.6925,
|
| 2204 |
+
"step": 145000
|
| 2205 |
+
},
|
| 2206 |
+
{
|
| 2207 |
+
"epoch": 12.67,
|
| 2208 |
+
"eval_loss": 0.6661481261253357,
|
| 2209 |
+
"eval_runtime": 3083.2268,
|
| 2210 |
+
"eval_samples_per_second": 25.005,
|
| 2211 |
+
"eval_steps_per_second": 1.563,
|
| 2212 |
+
"step": 145000
|
| 2213 |
+
},
|
| 2214 |
+
{
|
| 2215 |
+
"epoch": 12.71,
|
| 2216 |
+
"learning_rate": 1.621365829518743e-05,
|
| 2217 |
+
"loss": 0.6943,
|
| 2218 |
+
"step": 145500
|
| 2219 |
+
},
|
| 2220 |
+
{
|
| 2221 |
+
"epoch": 12.76,
|
| 2222 |
+
"learning_rate": 1.5904367190399606e-05,
|
| 2223 |
+
"loss": 0.6891,
|
| 2224 |
+
"step": 146000
|
| 2225 |
+
},
|
| 2226 |
+
{
|
| 2227 |
+
"epoch": 12.8,
|
| 2228 |
+
"learning_rate": 1.559507608561178e-05,
|
| 2229 |
+
"loss": 0.6848,
|
| 2230 |
+
"step": 146500
|
| 2231 |
+
},
|
| 2232 |
+
{
|
| 2233 |
+
"epoch": 12.85,
|
| 2234 |
+
"learning_rate": 1.5285784980823954e-05,
|
| 2235 |
+
"loss": 0.6816,
|
| 2236 |
+
"step": 147000
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 12.89,
|
| 2240 |
+
"learning_rate": 1.49771124582457e-05,
|
| 2241 |
+
"loss": 0.6795,
|
| 2242 |
+
"step": 147500
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 12.89,
|
| 2246 |
+
"eval_loss": 0.6605859398841858,
|
| 2247 |
+
"eval_runtime": 3089.4045,
|
| 2248 |
+
"eval_samples_per_second": 24.955,
|
| 2249 |
+
"eval_steps_per_second": 1.56,
|
| 2250 |
+
"step": 147500
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 12.93,
|
| 2254 |
+
"learning_rate": 1.4667821353457875e-05,
|
| 2255 |
+
"loss": 0.6779,
|
| 2256 |
+
"step": 148000
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 12.98,
|
| 2260 |
+
"learning_rate": 1.4358530248670049e-05,
|
| 2261 |
+
"loss": 0.6846,
|
| 2262 |
+
"step": 148500
|
| 2263 |
+
},
|
| 2264 |
+
{
|
| 2265 |
+
"epoch": 13.02,
|
| 2266 |
+
"learning_rate": 1.4049239143882223e-05,
|
| 2267 |
+
"loss": 0.6739,
|
| 2268 |
+
"step": 149000
|
| 2269 |
+
},
|
| 2270 |
+
{
|
| 2271 |
+
"epoch": 13.06,
|
| 2272 |
+
"learning_rate": 1.3740566621303971e-05,
|
| 2273 |
+
"loss": 0.6772,
|
| 2274 |
+
"step": 149500
|
| 2275 |
+
},
|
| 2276 |
+
{
|
| 2277 |
+
"epoch": 13.11,
|
| 2278 |
+
"learning_rate": 1.3431275516516146e-05,
|
| 2279 |
+
"loss": 0.6774,
|
| 2280 |
+
"step": 150000
|
| 2281 |
+
},
|
| 2282 |
+
{
|
| 2283 |
+
"epoch": 13.11,
|
| 2284 |
+
"eval_loss": 0.6616868376731873,
|
| 2285 |
+
"eval_runtime": 3137.3755,
|
| 2286 |
+
"eval_samples_per_second": 24.573,
|
| 2287 |
+
"eval_steps_per_second": 1.536,
|
| 2288 |
+
"step": 150000
|
| 2289 |
+
},
|
| 2290 |
+
{
|
| 2291 |
+
"epoch": 13.15,
|
| 2292 |
+
"learning_rate": 1.3122602993937894e-05,
|
| 2293 |
+
"loss": 0.6768,
|
| 2294 |
+
"step": 150500
|
| 2295 |
+
},
|
| 2296 |
+
{
|
| 2297 |
+
"epoch": 13.19,
|
| 2298 |
+
"learning_rate": 1.2813311889150068e-05,
|
| 2299 |
+
"loss": 0.6794,
|
| 2300 |
+
"step": 151000
|
| 2301 |
+
},
|
| 2302 |
+
{
|
| 2303 |
+
"epoch": 13.24,
|
| 2304 |
+
"learning_rate": 1.2504020784362242e-05,
|
| 2305 |
+
"loss": 0.6816,
|
| 2306 |
+
"step": 151500
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 13.28,
|
| 2310 |
+
"learning_rate": 1.2194729679574415e-05,
|
| 2311 |
+
"loss": 0.6748,
|
| 2312 |
+
"step": 152000
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 13.33,
|
| 2316 |
+
"learning_rate": 1.1886057156996165e-05,
|
| 2317 |
+
"loss": 0.6756,
|
| 2318 |
+
"step": 152500
|
| 2319 |
+
},
|
| 2320 |
+
{
|
| 2321 |
+
"epoch": 13.33,
|
| 2322 |
+
"eval_loss": 0.6562890410423279,
|
| 2323 |
+
"eval_runtime": 3110.9217,
|
| 2324 |
+
"eval_samples_per_second": 24.782,
|
| 2325 |
+
"eval_steps_per_second": 1.549,
|
| 2326 |
+
"step": 152500
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 13.37,
|
| 2330 |
+
"learning_rate": 1.1576766052208339e-05,
|
| 2331 |
+
"loss": 0.672,
|
| 2332 |
+
"step": 153000
|
| 2333 |
+
},
|
| 2334 |
+
{
|
| 2335 |
+
"epoch": 13.41,
|
| 2336 |
+
"learning_rate": 1.1267474947420513e-05,
|
| 2337 |
+
"loss": 0.6711,
|
| 2338 |
+
"step": 153500
|
| 2339 |
+
},
|
| 2340 |
+
{
|
| 2341 |
+
"epoch": 13.46,
|
| 2342 |
+
"learning_rate": 1.0958183842632687e-05,
|
| 2343 |
+
"loss": 0.6731,
|
| 2344 |
+
"step": 154000
|
| 2345 |
+
},
|
| 2346 |
+
{
|
| 2347 |
+
"epoch": 13.5,
|
| 2348 |
+
"learning_rate": 1.064889273784486e-05,
|
| 2349 |
+
"loss": 0.6707,
|
| 2350 |
+
"step": 154500
|
| 2351 |
+
},
|
| 2352 |
+
{
|
| 2353 |
+
"epoch": 13.54,
|
| 2354 |
+
"learning_rate": 1.0339601633057034e-05,
|
| 2355 |
+
"loss": 0.6728,
|
| 2356 |
+
"step": 155000
|
| 2357 |
+
},
|
| 2358 |
+
{
|
| 2359 |
+
"epoch": 13.54,
|
| 2360 |
+
"eval_loss": 0.6547101140022278,
|
| 2361 |
+
"eval_runtime": 3126.1621,
|
| 2362 |
+
"eval_samples_per_second": 24.662,
|
| 2363 |
+
"eval_steps_per_second": 1.542,
|
| 2364 |
+
"step": 155000
|
| 2365 |
+
},
|
| 2366 |
+
{
|
| 2367 |
+
"epoch": 13.59,
|
| 2368 |
+
"learning_rate": 1.0030929110478784e-05,
|
| 2369 |
+
"loss": 0.6727,
|
| 2370 |
+
"step": 155500
|
| 2371 |
+
},
|
| 2372 |
+
{
|
| 2373 |
+
"epoch": 13.63,
|
| 2374 |
+
"learning_rate": 9.721638005690958e-06,
|
| 2375 |
+
"loss": 0.6679,
|
| 2376 |
+
"step": 156000
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 13.68,
|
| 2380 |
+
"learning_rate": 9.41234690090313e-06,
|
| 2381 |
+
"loss": 0.6626,
|
| 2382 |
+
"step": 156500
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 13.72,
|
| 2386 |
+
"learning_rate": 9.103055796115304e-06,
|
| 2387 |
+
"loss": 0.6703,
|
| 2388 |
+
"step": 157000
|
| 2389 |
+
},
|
| 2390 |
+
{
|
| 2391 |
+
"epoch": 13.76,
|
| 2392 |
+
"learning_rate": 8.793764691327477e-06,
|
| 2393 |
+
"loss": 0.6732,
|
| 2394 |
+
"step": 157500
|
| 2395 |
+
},
|
| 2396 |
+
{
|
| 2397 |
+
"epoch": 13.76,
|
| 2398 |
+
"eval_loss": 0.6520426273345947,
|
| 2399 |
+
"eval_runtime": 3118.7072,
|
| 2400 |
+
"eval_samples_per_second": 24.72,
|
| 2401 |
+
"eval_steps_per_second": 1.545,
|
| 2402 |
+
"step": 157500
|
| 2403 |
+
},
|
| 2404 |
+
{
|
| 2405 |
+
"epoch": 13.81,
|
| 2406 |
+
"learning_rate": 8.484473586539651e-06,
|
| 2407 |
+
"loss": 0.6734,
|
| 2408 |
+
"step": 158000
|
| 2409 |
+
},
|
| 2410 |
+
{
|
| 2411 |
+
"epoch": 13.85,
|
| 2412 |
+
"learning_rate": 8.175182481751825e-06,
|
| 2413 |
+
"loss": 0.6707,
|
| 2414 |
+
"step": 158500
|
| 2415 |
+
},
|
| 2416 |
+
{
|
| 2417 |
+
"epoch": 13.89,
|
| 2418 |
+
"learning_rate": 7.865891376964e-06,
|
| 2419 |
+
"loss": 0.667,
|
| 2420 |
+
"step": 159000
|
| 2421 |
+
},
|
| 2422 |
+
{
|
| 2423 |
+
"epoch": 13.94,
|
| 2424 |
+
"learning_rate": 7.557218854385749e-06,
|
| 2425 |
+
"loss": 0.6667,
|
| 2426 |
+
"step": 159500
|
| 2427 |
+
},
|
| 2428 |
+
{
|
| 2429 |
+
"epoch": 13.98,
|
| 2430 |
+
"learning_rate": 7.2485463318074974e-06,
|
| 2431 |
+
"loss": 0.6704,
|
| 2432 |
+
"step": 160000
|
| 2433 |
+
},
|
| 2434 |
+
{
|
| 2435 |
+
"epoch": 13.98,
|
| 2436 |
+
"eval_loss": 0.6492029428482056,
|
| 2437 |
+
"eval_runtime": 3114.0251,
|
| 2438 |
+
"eval_samples_per_second": 24.758,
|
| 2439 |
+
"eval_steps_per_second": 1.548,
|
| 2440 |
+
"step": 160000
|
| 2441 |
+
},
|
| 2442 |
+
{
|
| 2443 |
+
"epoch": 14.02,
|
| 2444 |
+
"learning_rate": 6.939255227019672e-06,
|
| 2445 |
+
"loss": 0.6631,
|
| 2446 |
+
"step": 160500
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 14.07,
|
| 2450 |
+
"learning_rate": 6.629964122231845e-06,
|
| 2451 |
+
"loss": 0.663,
|
| 2452 |
+
"step": 161000
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 14.11,
|
| 2456 |
+
"learning_rate": 6.320673017444019e-06,
|
| 2457 |
+
"loss": 0.6661,
|
| 2458 |
+
"step": 161500
|
| 2459 |
+
},
|
| 2460 |
+
{
|
| 2461 |
+
"epoch": 14.16,
|
| 2462 |
+
"learning_rate": 6.011381912656192e-06,
|
| 2463 |
+
"loss": 0.6641,
|
| 2464 |
+
"step": 162000
|
| 2465 |
+
},
|
| 2466 |
+
{
|
| 2467 |
+
"epoch": 14.2,
|
| 2468 |
+
"learning_rate": 5.7020908078683665e-06,
|
| 2469 |
+
"loss": 0.6666,
|
| 2470 |
+
"step": 162500
|
| 2471 |
+
},
|
| 2472 |
+
{
|
| 2473 |
+
"epoch": 14.2,
|
| 2474 |
+
"eval_loss": 0.6445870995521545,
|
| 2475 |
+
"eval_runtime": 3137.0802,
|
| 2476 |
+
"eval_samples_per_second": 24.576,
|
| 2477 |
+
"eval_steps_per_second": 1.536,
|
| 2478 |
+
"step": 162500
|
| 2479 |
+
},
|
| 2480 |
+
{
|
| 2481 |
+
"epoch": 14.24,
|
| 2482 |
+
"learning_rate": 5.392799703080539e-06,
|
| 2483 |
+
"loss": 0.6612,
|
| 2484 |
+
"step": 163000
|
| 2485 |
+
},
|
| 2486 |
+
{
|
| 2487 |
+
"epoch": 14.29,
|
| 2488 |
+
"learning_rate": 5.083508598292713e-06,
|
| 2489 |
+
"loss": 0.6645,
|
| 2490 |
+
"step": 163500
|
| 2491 |
+
},
|
| 2492 |
+
{
|
| 2493 |
+
"epoch": 14.33,
|
| 2494 |
+
"learning_rate": 4.774217493504887e-06,
|
| 2495 |
+
"loss": 0.6613,
|
| 2496 |
+
"step": 164000
|
| 2497 |
+
},
|
| 2498 |
+
{
|
| 2499 |
+
"epoch": 14.37,
|
| 2500 |
+
"learning_rate": 4.464926388717061e-06,
|
| 2501 |
+
"loss": 0.6608,
|
| 2502 |
+
"step": 164500
|
| 2503 |
+
},
|
| 2504 |
+
{
|
| 2505 |
+
"epoch": 14.42,
|
| 2506 |
+
"learning_rate": 4.155635283929235e-06,
|
| 2507 |
+
"loss": 0.6615,
|
| 2508 |
+
"step": 165000
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 14.42,
|
| 2512 |
+
"eval_loss": 0.6487849950790405,
|
| 2513 |
+
"eval_runtime": 3095.1776,
|
| 2514 |
+
"eval_samples_per_second": 24.908,
|
| 2515 |
+
"eval_steps_per_second": 1.557,
|
| 2516 |
+
"step": 165000
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 14.46,
|
| 2520 |
+
"learning_rate": 3.846344179141408e-06,
|
| 2521 |
+
"loss": 0.6598,
|
| 2522 |
+
"step": 165500
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 14.51,
|
| 2526 |
+
"learning_rate": 3.537671656563157e-06,
|
| 2527 |
+
"loss": 0.6569,
|
| 2528 |
+
"step": 166000
|
| 2529 |
+
},
|
| 2530 |
+
{
|
| 2531 |
+
"epoch": 14.55,
|
| 2532 |
+
"learning_rate": 3.2289991339849062e-06,
|
| 2533 |
+
"loss": 0.6587,
|
| 2534 |
+
"step": 166500
|
| 2535 |
+
},
|
| 2536 |
+
{
|
| 2537 |
+
"epoch": 14.59,
|
| 2538 |
+
"learning_rate": 2.920326611406656e-06,
|
| 2539 |
+
"loss": 0.6575,
|
| 2540 |
+
"step": 167000
|
| 2541 |
+
},
|
| 2542 |
+
{
|
| 2543 |
+
"epoch": 14.64,
|
| 2544 |
+
"learning_rate": 2.61103550661883e-06,
|
| 2545 |
+
"loss": 0.6638,
|
| 2546 |
+
"step": 167500
|
| 2547 |
+
},
|
| 2548 |
+
{
|
| 2549 |
+
"epoch": 14.64,
|
| 2550 |
+
"eval_loss": 0.6522655487060547,
|
| 2551 |
+
"eval_runtime": 3117.0073,
|
| 2552 |
+
"eval_samples_per_second": 24.734,
|
| 2553 |
+
"eval_steps_per_second": 1.546,
|
| 2554 |
+
"step": 167500
|
| 2555 |
+
},
|
| 2556 |
+
{
|
| 2557 |
+
"epoch": 14.68,
|
| 2558 |
+
"learning_rate": 2.3017444018310032e-06,
|
| 2559 |
+
"loss": 0.6553,
|
| 2560 |
+
"step": 168000
|
| 2561 |
+
},
|
| 2562 |
+
{
|
| 2563 |
+
"epoch": 14.72,
|
| 2564 |
+
"learning_rate": 1.9930718792527527e-06,
|
| 2565 |
+
"loss": 0.6617,
|
| 2566 |
+
"step": 168500
|
| 2567 |
+
},
|
| 2568 |
+
{
|
| 2569 |
+
"epoch": 14.77,
|
| 2570 |
+
"learning_rate": 1.6837807744649262e-06,
|
| 2571 |
+
"loss": 0.6583,
|
| 2572 |
+
"step": 169000
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 14.81,
|
| 2576 |
+
"learning_rate": 1.3744896696771002e-06,
|
| 2577 |
+
"loss": 0.6579,
|
| 2578 |
+
"step": 169500
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 14.85,
|
| 2582 |
+
"learning_rate": 1.0651985648892737e-06,
|
| 2583 |
+
"loss": 0.6588,
|
| 2584 |
+
"step": 170000
|
| 2585 |
+
},
|
| 2586 |
+
{
|
| 2587 |
+
"epoch": 14.85,
|
| 2588 |
+
"eval_loss": 0.6414651274681091,
|
| 2589 |
+
"eval_runtime": 3135.6411,
|
| 2590 |
+
"eval_samples_per_second": 24.587,
|
| 2591 |
+
"eval_steps_per_second": 1.537,
|
| 2592 |
+
"step": 170000
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 14.9,
|
| 2596 |
+
"learning_rate": 7.559074601014475e-07,
|
| 2597 |
+
"loss": 0.6589,
|
| 2598 |
+
"step": 170500
|
| 2599 |
+
},
|
| 2600 |
+
{
|
| 2601 |
+
"epoch": 14.94,
|
| 2602 |
+
"learning_rate": 4.4661635531362117e-07,
|
| 2603 |
+
"loss": 0.6582,
|
| 2604 |
+
"step": 171000
|
| 2605 |
+
},
|
| 2606 |
+
{
|
| 2607 |
+
"epoch": 14.99,
|
| 2608 |
+
"learning_rate": 1.3732525052579488e-07,
|
| 2609 |
+
"loss": 0.6569,
|
| 2610 |
+
"step": 171500
|
| 2611 |
+
},
|
| 2612 |
+
{
|
| 2613 |
+
"epoch": 15.0,
|
| 2614 |
+
"step": 171660,
|
| 2615 |
+
"total_flos": 2.5739338132512957e+18,
|
| 2616 |
+
"train_loss": 0.23807151880656507,
|
| 2617 |
+
"train_runtime": 725915.7505,
|
| 2618 |
+
"train_samples_per_second": 30.268,
|
| 2619 |
+
"train_steps_per_second": 0.236
|
| 2620 |
+
}
|
| 2621 |
+
],
|
| 2622 |
+
"max_steps": 171660,
|
| 2623 |
+
"num_train_epochs": 15,
|
| 2624 |
+
"total_flos": 2.5739338132512957e+18,
|
| 2625 |
+
"trial_name": null,
|
| 2626 |
+
"trial_params": null
|
| 2627 |
+
}
|