Upload trainer_state.json
Browse files- trainer_state.json +2483 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,2483 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.1634913049801583,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3500,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.003324260871371881,
|
| 13 |
+
"grad_norm": 2.5143792629241943,
|
| 14 |
+
"learning_rate": 1.6622340425531916e-08,
|
| 15 |
+
"loss": 9.0836,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.006648521742743762,
|
| 20 |
+
"grad_norm": 2.6173431873321533,
|
| 21 |
+
"learning_rate": 3.324468085106383e-08,
|
| 22 |
+
"loss": 9.0829,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.009972782614115643,
|
| 27 |
+
"grad_norm": 2.5349628925323486,
|
| 28 |
+
"learning_rate": 4.9867021276595746e-08,
|
| 29 |
+
"loss": 9.0061,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.013297043485487523,
|
| 34 |
+
"grad_norm": 2.3916308879852295,
|
| 35 |
+
"learning_rate": 6.648936170212767e-08,
|
| 36 |
+
"loss": 8.9747,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.016621304356859403,
|
| 41 |
+
"grad_norm": 2.5103342533111572,
|
| 42 |
+
"learning_rate": 8.311170212765958e-08,
|
| 43 |
+
"loss": 9.0057,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.019945565228231286,
|
| 48 |
+
"grad_norm": 2.421079397201538,
|
| 49 |
+
"learning_rate": 9.973404255319149e-08,
|
| 50 |
+
"loss": 8.9885,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.023269826099603166,
|
| 55 |
+
"grad_norm": 2.6052393913269043,
|
| 56 |
+
"learning_rate": 1.163563829787234e-07,
|
| 57 |
+
"loss": 8.9706,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.026594086970975046,
|
| 62 |
+
"grad_norm": 2.376847505569458,
|
| 63 |
+
"learning_rate": 1.3297872340425533e-07,
|
| 64 |
+
"loss": 9.0211,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.02991834784234693,
|
| 69 |
+
"grad_norm": 2.6200971603393555,
|
| 70 |
+
"learning_rate": 1.4960106382978723e-07,
|
| 71 |
+
"loss": 8.9903,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.033242608713718806,
|
| 76 |
+
"grad_norm": 2.515320301055908,
|
| 77 |
+
"learning_rate": 1.6622340425531916e-07,
|
| 78 |
+
"loss": 8.9643,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.03656686958509069,
|
| 83 |
+
"grad_norm": 2.4840102195739746,
|
| 84 |
+
"learning_rate": 1.8284574468085108e-07,
|
| 85 |
+
"loss": 8.9761,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.03989113045646257,
|
| 90 |
+
"grad_norm": 2.5950074195861816,
|
| 91 |
+
"learning_rate": 1.9946808510638298e-07,
|
| 92 |
+
"loss": 9.0101,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.04321539132783445,
|
| 97 |
+
"grad_norm": 2.530604839324951,
|
| 98 |
+
"learning_rate": 2.160904255319149e-07,
|
| 99 |
+
"loss": 8.961,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.04653965219920633,
|
| 104 |
+
"grad_norm": 2.5579464435577393,
|
| 105 |
+
"learning_rate": 2.327127659574468e-07,
|
| 106 |
+
"loss": 8.8733,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.04986391307057821,
|
| 111 |
+
"grad_norm": 2.638901472091675,
|
| 112 |
+
"learning_rate": 2.4933510638297876e-07,
|
| 113 |
+
"loss": 8.9534,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.05318817394195009,
|
| 118 |
+
"grad_norm": 2.6817493438720703,
|
| 119 |
+
"learning_rate": 2.6595744680851066e-07,
|
| 120 |
+
"loss": 9.0014,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.05651243481332197,
|
| 125 |
+
"grad_norm": 2.6700024604797363,
|
| 126 |
+
"learning_rate": 2.8257978723404256e-07,
|
| 127 |
+
"loss": 8.8832,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.05983669568469386,
|
| 132 |
+
"grad_norm": 2.794243335723877,
|
| 133 |
+
"learning_rate": 2.9920212765957446e-07,
|
| 134 |
+
"loss": 8.9012,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.06316095655606574,
|
| 139 |
+
"grad_norm": 3.000873327255249,
|
| 140 |
+
"learning_rate": 3.1582446808510636e-07,
|
| 141 |
+
"loss": 8.7874,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.06648521742743761,
|
| 146 |
+
"grad_norm": 2.872612714767456,
|
| 147 |
+
"learning_rate": 3.324468085106383e-07,
|
| 148 |
+
"loss": 8.8558,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.0698094782988095,
|
| 153 |
+
"grad_norm": 2.9133315086364746,
|
| 154 |
+
"learning_rate": 3.490691489361702e-07,
|
| 155 |
+
"loss": 8.8333,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.07313373917018139,
|
| 160 |
+
"grad_norm": 3.1017534732818604,
|
| 161 |
+
"learning_rate": 3.6569148936170217e-07,
|
| 162 |
+
"loss": 8.8199,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.07645800004155326,
|
| 167 |
+
"grad_norm": 2.9153056144714355,
|
| 168 |
+
"learning_rate": 3.8231382978723407e-07,
|
| 169 |
+
"loss": 8.8266,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.07978226091292515,
|
| 174 |
+
"grad_norm": 3.0996434688568115,
|
| 175 |
+
"learning_rate": 3.9893617021276597e-07,
|
| 176 |
+
"loss": 8.7202,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.08310652178429702,
|
| 181 |
+
"grad_norm": 3.257809638977051,
|
| 182 |
+
"learning_rate": 4.1555851063829787e-07,
|
| 183 |
+
"loss": 8.6149,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.0864307826556689,
|
| 188 |
+
"grad_norm": 3.378631353378296,
|
| 189 |
+
"learning_rate": 4.321808510638298e-07,
|
| 190 |
+
"loss": 8.5092,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.08975504352704078,
|
| 195 |
+
"grad_norm": 3.3546876907348633,
|
| 196 |
+
"learning_rate": 4.488031914893618e-07,
|
| 197 |
+
"loss": 8.477,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.09307930439841267,
|
| 202 |
+
"grad_norm": 3.292569637298584,
|
| 203 |
+
"learning_rate": 4.654255319148936e-07,
|
| 204 |
+
"loss": 8.4594,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.09640356526978455,
|
| 209 |
+
"grad_norm": 3.190239906311035,
|
| 210 |
+
"learning_rate": 4.820478723404255e-07,
|
| 211 |
+
"loss": 8.3134,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.09972782614115643,
|
| 216 |
+
"grad_norm": 3.2521212100982666,
|
| 217 |
+
"learning_rate": 4.986702127659575e-07,
|
| 218 |
+
"loss": 8.2896,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.10305208701252831,
|
| 223 |
+
"grad_norm": 3.399919033050537,
|
| 224 |
+
"learning_rate": 5.152925531914893e-07,
|
| 225 |
+
"loss": 8.1656,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.10637634788390019,
|
| 230 |
+
"grad_norm": 3.412688970565796,
|
| 231 |
+
"learning_rate": 5.319148936170213e-07,
|
| 232 |
+
"loss": 7.9764,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.10970060875527207,
|
| 237 |
+
"grad_norm": 3.2669174671173096,
|
| 238 |
+
"learning_rate": 5.485372340425532e-07,
|
| 239 |
+
"loss": 7.9755,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.11302486962664395,
|
| 244 |
+
"grad_norm": 3.405444383621216,
|
| 245 |
+
"learning_rate": 5.651595744680851e-07,
|
| 246 |
+
"loss": 7.8587,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.11634913049801583,
|
| 251 |
+
"grad_norm": 3.2224161624908447,
|
| 252 |
+
"learning_rate": 5.81781914893617e-07,
|
| 253 |
+
"loss": 7.7357,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.11967339136938772,
|
| 258 |
+
"grad_norm": 3.230048894882202,
|
| 259 |
+
"learning_rate": 5.984042553191489e-07,
|
| 260 |
+
"loss": 7.5667,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.12299765224075959,
|
| 265 |
+
"grad_norm": 3.2728312015533447,
|
| 266 |
+
"learning_rate": 6.150265957446809e-07,
|
| 267 |
+
"loss": 7.4844,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.12632191311213148,
|
| 272 |
+
"grad_norm": 3.200800895690918,
|
| 273 |
+
"learning_rate": 6.316489361702127e-07,
|
| 274 |
+
"loss": 7.3059,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.12964617398350337,
|
| 279 |
+
"grad_norm": 3.075329065322876,
|
| 280 |
+
"learning_rate": 6.482712765957447e-07,
|
| 281 |
+
"loss": 7.1618,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.13297043485487522,
|
| 286 |
+
"grad_norm": 3.1853721141815186,
|
| 287 |
+
"learning_rate": 6.648936170212766e-07,
|
| 288 |
+
"loss": 7.1392,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.1362946957262471,
|
| 293 |
+
"grad_norm": 3.0336828231811523,
|
| 294 |
+
"learning_rate": 6.815159574468085e-07,
|
| 295 |
+
"loss": 6.974,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.139618956597619,
|
| 300 |
+
"grad_norm": 3.027355670928955,
|
| 301 |
+
"learning_rate": 6.981382978723404e-07,
|
| 302 |
+
"loss": 6.7714,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.14294321746899089,
|
| 307 |
+
"grad_norm": 2.99857497215271,
|
| 308 |
+
"learning_rate": 7.147606382978723e-07,
|
| 309 |
+
"loss": 6.6538,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.14626747834036277,
|
| 314 |
+
"grad_norm": 2.840437650680542,
|
| 315 |
+
"learning_rate": 7.313829787234043e-07,
|
| 316 |
+
"loss": 6.4758,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.14959173921173463,
|
| 321 |
+
"grad_norm": 3.076049566268921,
|
| 322 |
+
"learning_rate": 7.480053191489362e-07,
|
| 323 |
+
"loss": 6.3225,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.15291600008310652,
|
| 328 |
+
"grad_norm": 2.8588602542877197,
|
| 329 |
+
"learning_rate": 7.646276595744681e-07,
|
| 330 |
+
"loss": 6.1646,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.1562402609544784,
|
| 335 |
+
"grad_norm": 2.9168858528137207,
|
| 336 |
+
"learning_rate": 7.8125e-07,
|
| 337 |
+
"loss": 6.0828,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.1595645218258503,
|
| 342 |
+
"grad_norm": 2.795363187789917,
|
| 343 |
+
"learning_rate": 7.978723404255319e-07,
|
| 344 |
+
"loss": 5.9192,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.16288878269722215,
|
| 349 |
+
"grad_norm": 2.3897600173950195,
|
| 350 |
+
"learning_rate": 8.144946808510639e-07,
|
| 351 |
+
"loss": 5.6675,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.16621304356859404,
|
| 356 |
+
"grad_norm": 2.279939651489258,
|
| 357 |
+
"learning_rate": 8.311170212765957e-07,
|
| 358 |
+
"loss": 5.5443,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.16953730443996592,
|
| 363 |
+
"grad_norm": 2.394994020462036,
|
| 364 |
+
"learning_rate": 8.477393617021276e-07,
|
| 365 |
+
"loss": 5.4127,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.1728615653113378,
|
| 370 |
+
"grad_norm": 2.3148529529571533,
|
| 371 |
+
"learning_rate": 8.643617021276596e-07,
|
| 372 |
+
"loss": 5.2102,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.1761858261827097,
|
| 377 |
+
"grad_norm": 2.053243637084961,
|
| 378 |
+
"learning_rate": 8.809840425531915e-07,
|
| 379 |
+
"loss": 5.1204,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.17951008705408156,
|
| 384 |
+
"grad_norm": 2.0687060356140137,
|
| 385 |
+
"learning_rate": 8.976063829787235e-07,
|
| 386 |
+
"loss": 4.9786,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.18283434792545344,
|
| 391 |
+
"grad_norm": 1.8042306900024414,
|
| 392 |
+
"learning_rate": 9.142287234042553e-07,
|
| 393 |
+
"loss": 4.7735,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.18615860879682533,
|
| 398 |
+
"grad_norm": 1.8975441455841064,
|
| 399 |
+
"learning_rate": 9.308510638297872e-07,
|
| 400 |
+
"loss": 4.6749,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.18948286966819722,
|
| 405 |
+
"grad_norm": 1.6640989780426025,
|
| 406 |
+
"learning_rate": 9.474734042553192e-07,
|
| 407 |
+
"loss": 4.529,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.1928071305395691,
|
| 412 |
+
"grad_norm": 1.9563913345336914,
|
| 413 |
+
"learning_rate": 9.64095744680851e-07,
|
| 414 |
+
"loss": 4.4255,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.19613139141094096,
|
| 419 |
+
"grad_norm": 1.433192253112793,
|
| 420 |
+
"learning_rate": 9.80718085106383e-07,
|
| 421 |
+
"loss": 4.3369,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.19945565228231285,
|
| 426 |
+
"grad_norm": 1.6940258741378784,
|
| 427 |
+
"learning_rate": 9.97340425531915e-07,
|
| 428 |
+
"loss": 4.2264,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.20277991315368474,
|
| 433 |
+
"grad_norm": 1.3721990585327148,
|
| 434 |
+
"learning_rate": 1.0139627659574467e-06,
|
| 435 |
+
"loss": 4.0771,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.20610417402505662,
|
| 440 |
+
"grad_norm": 1.3481799364089966,
|
| 441 |
+
"learning_rate": 1.0305851063829786e-06,
|
| 442 |
+
"loss": 3.9652,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.20942843489642848,
|
| 447 |
+
"grad_norm": 1.3010597229003906,
|
| 448 |
+
"learning_rate": 1.0472074468085108e-06,
|
| 449 |
+
"loss": 3.9205,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.21275269576780037,
|
| 454 |
+
"grad_norm": 1.551216721534729,
|
| 455 |
+
"learning_rate": 1.0638297872340427e-06,
|
| 456 |
+
"loss": 3.8234,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.21607695663917226,
|
| 461 |
+
"grad_norm": 1.3280216455459595,
|
| 462 |
+
"learning_rate": 1.0804521276595746e-06,
|
| 463 |
+
"loss": 3.6898,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.21940121751054414,
|
| 468 |
+
"grad_norm": 1.0909334421157837,
|
| 469 |
+
"learning_rate": 1.0970744680851065e-06,
|
| 470 |
+
"loss": 3.6489,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.22272547838191603,
|
| 475 |
+
"grad_norm": 1.345831036567688,
|
| 476 |
+
"learning_rate": 1.1136968085106384e-06,
|
| 477 |
+
"loss": 3.5296,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.2260497392532879,
|
| 482 |
+
"grad_norm": 1.0882962942123413,
|
| 483 |
+
"learning_rate": 1.1303191489361703e-06,
|
| 484 |
+
"loss": 3.489,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.22937400012465978,
|
| 489 |
+
"grad_norm": 0.9840554594993591,
|
| 490 |
+
"learning_rate": 1.1469414893617022e-06,
|
| 491 |
+
"loss": 3.4164,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.23269826099603166,
|
| 496 |
+
"grad_norm": 1.0956693887710571,
|
| 497 |
+
"learning_rate": 1.163563829787234e-06,
|
| 498 |
+
"loss": 3.3211,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.23602252186740355,
|
| 503 |
+
"grad_norm": 0.8875247240066528,
|
| 504 |
+
"learning_rate": 1.1801861702127662e-06,
|
| 505 |
+
"loss": 3.2647,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.23934678273877544,
|
| 510 |
+
"grad_norm": 1.268930196762085,
|
| 511 |
+
"learning_rate": 1.1968085106382979e-06,
|
| 512 |
+
"loss": 3.2033,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.2426710436101473,
|
| 517 |
+
"grad_norm": 0.9430990815162659,
|
| 518 |
+
"learning_rate": 1.2134308510638298e-06,
|
| 519 |
+
"loss": 3.1317,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.24599530448151918,
|
| 524 |
+
"grad_norm": 0.9196615219116211,
|
| 525 |
+
"learning_rate": 1.2300531914893619e-06,
|
| 526 |
+
"loss": 3.0706,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.24931956535289107,
|
| 531 |
+
"grad_norm": 0.7046869993209839,
|
| 532 |
+
"learning_rate": 1.2466755319148936e-06,
|
| 533 |
+
"loss": 3.0142,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.25264382622426296,
|
| 538 |
+
"grad_norm": 0.9173153638839722,
|
| 539 |
+
"learning_rate": 1.2632978723404255e-06,
|
| 540 |
+
"loss": 2.949,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.25596808709563484,
|
| 545 |
+
"grad_norm": 0.8014841675758362,
|
| 546 |
+
"learning_rate": 1.2799202127659576e-06,
|
| 547 |
+
"loss": 2.9325,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.25929234796700673,
|
| 552 |
+
"grad_norm": 0.9520502686500549,
|
| 553 |
+
"learning_rate": 1.2965425531914895e-06,
|
| 554 |
+
"loss": 2.859,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.2626166088383786,
|
| 559 |
+
"grad_norm": 0.7679387331008911,
|
| 560 |
+
"learning_rate": 1.3131648936170214e-06,
|
| 561 |
+
"loss": 2.8509,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.26594086970975045,
|
| 566 |
+
"grad_norm": 0.7660825252532959,
|
| 567 |
+
"learning_rate": 1.3297872340425533e-06,
|
| 568 |
+
"loss": 2.7896,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.26926513058112234,
|
| 573 |
+
"grad_norm": 0.7754834294319153,
|
| 574 |
+
"learning_rate": 1.3464095744680852e-06,
|
| 575 |
+
"loss": 2.736,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.2725893914524942,
|
| 580 |
+
"grad_norm": 0.5802922248840332,
|
| 581 |
+
"learning_rate": 1.363031914893617e-06,
|
| 582 |
+
"loss": 2.6962,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.2759136523238661,
|
| 587 |
+
"grad_norm": 0.6394158601760864,
|
| 588 |
+
"learning_rate": 1.379654255319149e-06,
|
| 589 |
+
"loss": 2.6656,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.279237913195238,
|
| 594 |
+
"grad_norm": 0.6503139138221741,
|
| 595 |
+
"learning_rate": 1.3962765957446809e-06,
|
| 596 |
+
"loss": 2.616,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.2825621740666099,
|
| 601 |
+
"grad_norm": 0.6165557503700256,
|
| 602 |
+
"learning_rate": 1.412898936170213e-06,
|
| 603 |
+
"loss": 2.5971,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.28588643493798177,
|
| 608 |
+
"grad_norm": 0.6192012429237366,
|
| 609 |
+
"learning_rate": 1.4295212765957447e-06,
|
| 610 |
+
"loss": 2.5536,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.28921069580935366,
|
| 615 |
+
"grad_norm": 0.6266525983810425,
|
| 616 |
+
"learning_rate": 1.4461436170212766e-06,
|
| 617 |
+
"loss": 2.5036,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.29253495668072554,
|
| 622 |
+
"grad_norm": 0.5376760363578796,
|
| 623 |
+
"learning_rate": 1.4627659574468087e-06,
|
| 624 |
+
"loss": 2.5136,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.2958592175520974,
|
| 629 |
+
"grad_norm": 0.6490041613578796,
|
| 630 |
+
"learning_rate": 1.4793882978723404e-06,
|
| 631 |
+
"loss": 2.4638,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.29918347842346926,
|
| 636 |
+
"grad_norm": 0.6368073225021362,
|
| 637 |
+
"learning_rate": 1.4960106382978725e-06,
|
| 638 |
+
"loss": 2.4258,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.30250773929484115,
|
| 643 |
+
"grad_norm": 0.5121726989746094,
|
| 644 |
+
"learning_rate": 1.5126329787234044e-06,
|
| 645 |
+
"loss": 2.4016,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.30583200016621304,
|
| 650 |
+
"grad_norm": 0.5835744738578796,
|
| 651 |
+
"learning_rate": 1.5292553191489363e-06,
|
| 652 |
+
"loss": 2.4192,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.3091562610375849,
|
| 657 |
+
"grad_norm": 0.5275241732597351,
|
| 658 |
+
"learning_rate": 1.5458776595744682e-06,
|
| 659 |
+
"loss": 2.3687,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.3124805219089568,
|
| 664 |
+
"grad_norm": 0.4900510609149933,
|
| 665 |
+
"learning_rate": 1.5625e-06,
|
| 666 |
+
"loss": 2.3208,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.3158047827803287,
|
| 671 |
+
"grad_norm": 0.4609052240848541,
|
| 672 |
+
"learning_rate": 1.5791223404255322e-06,
|
| 673 |
+
"loss": 2.3363,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.3191290436517006,
|
| 678 |
+
"grad_norm": 0.461566299200058,
|
| 679 |
+
"learning_rate": 1.5957446808510639e-06,
|
| 680 |
+
"loss": 2.2793,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.32245330452307247,
|
| 685 |
+
"grad_norm": 0.49795401096343994,
|
| 686 |
+
"learning_rate": 1.6123670212765958e-06,
|
| 687 |
+
"loss": 2.2845,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.3257775653944443,
|
| 692 |
+
"grad_norm": 0.4422404170036316,
|
| 693 |
+
"learning_rate": 1.6289893617021279e-06,
|
| 694 |
+
"loss": 2.2744,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.3291018262658162,
|
| 699 |
+
"grad_norm": 0.4161861538887024,
|
| 700 |
+
"learning_rate": 1.6456117021276596e-06,
|
| 701 |
+
"loss": 2.2463,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.3324260871371881,
|
| 706 |
+
"grad_norm": 0.46071523427963257,
|
| 707 |
+
"learning_rate": 1.6622340425531915e-06,
|
| 708 |
+
"loss": 2.2271,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.33575034800855996,
|
| 713 |
+
"grad_norm": 0.3772067129611969,
|
| 714 |
+
"learning_rate": 1.6788563829787236e-06,
|
| 715 |
+
"loss": 2.2119,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.33907460887993185,
|
| 720 |
+
"grad_norm": 0.44782117009162903,
|
| 721 |
+
"learning_rate": 1.6954787234042553e-06,
|
| 722 |
+
"loss": 2.2022,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.34239886975130374,
|
| 727 |
+
"grad_norm": 0.4486360251903534,
|
| 728 |
+
"learning_rate": 1.7121010638297872e-06,
|
| 729 |
+
"loss": 2.1723,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.3457231306226756,
|
| 734 |
+
"grad_norm": 0.47423475980758667,
|
| 735 |
+
"learning_rate": 1.7287234042553193e-06,
|
| 736 |
+
"loss": 2.1295,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.3490473914940475,
|
| 741 |
+
"grad_norm": 0.4199342131614685,
|
| 742 |
+
"learning_rate": 1.745345744680851e-06,
|
| 743 |
+
"loss": 2.1387,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.3523716523654194,
|
| 748 |
+
"grad_norm": 0.43744415044784546,
|
| 749 |
+
"learning_rate": 1.761968085106383e-06,
|
| 750 |
+
"loss": 2.1195,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.3556959132367913,
|
| 755 |
+
"grad_norm": 0.3780044913291931,
|
| 756 |
+
"learning_rate": 1.778590425531915e-06,
|
| 757 |
+
"loss": 2.1194,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.3590201741081631,
|
| 762 |
+
"grad_norm": 0.40349099040031433,
|
| 763 |
+
"learning_rate": 1.795212765957447e-06,
|
| 764 |
+
"loss": 2.1005,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.362344434979535,
|
| 769 |
+
"grad_norm": 0.378764271736145,
|
| 770 |
+
"learning_rate": 1.8118351063829788e-06,
|
| 771 |
+
"loss": 2.0757,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.3656686958509069,
|
| 776 |
+
"grad_norm": 0.34115588665008545,
|
| 777 |
+
"learning_rate": 1.8284574468085107e-06,
|
| 778 |
+
"loss": 2.0591,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.3689929567222788,
|
| 783 |
+
"grad_norm": 0.39553964138031006,
|
| 784 |
+
"learning_rate": 1.8450797872340428e-06,
|
| 785 |
+
"loss": 2.0298,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.37231721759365066,
|
| 790 |
+
"grad_norm": 0.36110466718673706,
|
| 791 |
+
"learning_rate": 1.8617021276595745e-06,
|
| 792 |
+
"loss": 2.0113,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.37564147846502255,
|
| 797 |
+
"grad_norm": 0.33477863669395447,
|
| 798 |
+
"learning_rate": 1.8783244680851066e-06,
|
| 799 |
+
"loss": 2.0197,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.37896573933639444,
|
| 804 |
+
"grad_norm": 0.43919846415519714,
|
| 805 |
+
"learning_rate": 1.8949468085106385e-06,
|
| 806 |
+
"loss": 1.9794,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.3822900002077663,
|
| 811 |
+
"grad_norm": 0.3243393898010254,
|
| 812 |
+
"learning_rate": 1.9115691489361704e-06,
|
| 813 |
+
"loss": 1.9667,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.3856142610791382,
|
| 818 |
+
"grad_norm": 0.3350262939929962,
|
| 819 |
+
"learning_rate": 1.928191489361702e-06,
|
| 820 |
+
"loss": 1.978,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.38893852195051004,
|
| 825 |
+
"grad_norm": 0.3365063965320587,
|
| 826 |
+
"learning_rate": 1.944813829787234e-06,
|
| 827 |
+
"loss": 1.9701,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.39226278282188193,
|
| 832 |
+
"grad_norm": 0.3240489661693573,
|
| 833 |
+
"learning_rate": 1.961436170212766e-06,
|
| 834 |
+
"loss": 1.9465,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.3955870436932538,
|
| 839 |
+
"grad_norm": 0.3239437937736511,
|
| 840 |
+
"learning_rate": 1.978058510638298e-06,
|
| 841 |
+
"loss": 1.9253,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.3989113045646257,
|
| 846 |
+
"grad_norm": 0.3397749364376068,
|
| 847 |
+
"learning_rate": 1.99468085106383e-06,
|
| 848 |
+
"loss": 1.9057,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.4022355654359976,
|
| 853 |
+
"grad_norm": 0.2915981113910675,
|
| 854 |
+
"learning_rate": 2.011303191489362e-06,
|
| 855 |
+
"loss": 1.9047,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.4055598263073695,
|
| 860 |
+
"grad_norm": 0.39456045627593994,
|
| 861 |
+
"learning_rate": 2.0279255319148935e-06,
|
| 862 |
+
"loss": 1.9144,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.40888408717874136,
|
| 867 |
+
"grad_norm": 0.2593387961387634,
|
| 868 |
+
"learning_rate": 2.0445478723404256e-06,
|
| 869 |
+
"loss": 1.8969,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.41220834805011325,
|
| 874 |
+
"grad_norm": 0.30935177206993103,
|
| 875 |
+
"learning_rate": 2.0611702127659573e-06,
|
| 876 |
+
"loss": 1.8931,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.41553260892148514,
|
| 881 |
+
"grad_norm": 0.27917250990867615,
|
| 882 |
+
"learning_rate": 2.0777925531914894e-06,
|
| 883 |
+
"loss": 1.8899,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.41885686979285697,
|
| 888 |
+
"grad_norm": 0.25976502895355225,
|
| 889 |
+
"learning_rate": 2.0944148936170215e-06,
|
| 890 |
+
"loss": 1.8503,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.42218113066422885,
|
| 895 |
+
"grad_norm": 0.31833794713020325,
|
| 896 |
+
"learning_rate": 2.111037234042553e-06,
|
| 897 |
+
"loss": 1.8527,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.42550539153560074,
|
| 902 |
+
"grad_norm": 0.2671976685523987,
|
| 903 |
+
"learning_rate": 2.1276595744680853e-06,
|
| 904 |
+
"loss": 1.8505,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.42882965240697263,
|
| 909 |
+
"grad_norm": 0.3245258629322052,
|
| 910 |
+
"learning_rate": 2.144281914893617e-06,
|
| 911 |
+
"loss": 1.864,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.4321539132783445,
|
| 916 |
+
"grad_norm": 0.2622531056404114,
|
| 917 |
+
"learning_rate": 2.160904255319149e-06,
|
| 918 |
+
"loss": 1.8301,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.4354781741497164,
|
| 923 |
+
"grad_norm": 0.3247709274291992,
|
| 924 |
+
"learning_rate": 2.177526595744681e-06,
|
| 925 |
+
"loss": 1.812,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.4388024350210883,
|
| 930 |
+
"grad_norm": 0.26424384117126465,
|
| 931 |
+
"learning_rate": 2.194148936170213e-06,
|
| 932 |
+
"loss": 1.7958,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.4421266958924602,
|
| 937 |
+
"grad_norm": 0.2569092810153961,
|
| 938 |
+
"learning_rate": 2.210771276595745e-06,
|
| 939 |
+
"loss": 1.8147,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.44545095676383206,
|
| 944 |
+
"grad_norm": 0.2393629103899002,
|
| 945 |
+
"learning_rate": 2.2273936170212767e-06,
|
| 946 |
+
"loss": 1.7976,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.44877521763520395,
|
| 951 |
+
"grad_norm": 0.232402965426445,
|
| 952 |
+
"learning_rate": 2.244015957446809e-06,
|
| 953 |
+
"loss": 1.7597,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.4520994785065758,
|
| 958 |
+
"grad_norm": 0.26385971903800964,
|
| 959 |
+
"learning_rate": 2.2606382978723405e-06,
|
| 960 |
+
"loss": 1.7781,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.45542373937794767,
|
| 965 |
+
"grad_norm": 0.2671038806438446,
|
| 966 |
+
"learning_rate": 2.277260638297872e-06,
|
| 967 |
+
"loss": 1.7583,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.45874800024931955,
|
| 972 |
+
"grad_norm": 0.27096447348594666,
|
| 973 |
+
"learning_rate": 2.2938829787234043e-06,
|
| 974 |
+
"loss": 1.7402,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.46207226112069144,
|
| 979 |
+
"grad_norm": 0.2245018631219864,
|
| 980 |
+
"learning_rate": 2.3105053191489364e-06,
|
| 981 |
+
"loss": 1.7644,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.46539652199206333,
|
| 986 |
+
"grad_norm": 0.20663714408874512,
|
| 987 |
+
"learning_rate": 2.327127659574468e-06,
|
| 988 |
+
"loss": 1.7519,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.4687207828634352,
|
| 993 |
+
"grad_norm": 0.26273128390312195,
|
| 994 |
+
"learning_rate": 2.3437500000000002e-06,
|
| 995 |
+
"loss": 1.7312,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.4720450437348071,
|
| 1000 |
+
"grad_norm": 0.24725256860256195,
|
| 1001 |
+
"learning_rate": 2.3603723404255323e-06,
|
| 1002 |
+
"loss": 1.7217,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.475369304606179,
|
| 1007 |
+
"grad_norm": 0.25341796875,
|
| 1008 |
+
"learning_rate": 2.376994680851064e-06,
|
| 1009 |
+
"loss": 1.7246,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.4786935654775509,
|
| 1014 |
+
"grad_norm": 0.21035414934158325,
|
| 1015 |
+
"learning_rate": 2.3936170212765957e-06,
|
| 1016 |
+
"loss": 1.7017,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.4820178263489227,
|
| 1021 |
+
"grad_norm": 0.21454143524169922,
|
| 1022 |
+
"learning_rate": 2.410239361702128e-06,
|
| 1023 |
+
"loss": 1.7049,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.4853420872202946,
|
| 1028 |
+
"grad_norm": 0.22413010895252228,
|
| 1029 |
+
"learning_rate": 2.4268617021276595e-06,
|
| 1030 |
+
"loss": 1.6809,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.4886663480916665,
|
| 1035 |
+
"grad_norm": 0.2039473056793213,
|
| 1036 |
+
"learning_rate": 2.4434840425531916e-06,
|
| 1037 |
+
"loss": 1.6873,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.49199060896303837,
|
| 1042 |
+
"grad_norm": 0.18895457684993744,
|
| 1043 |
+
"learning_rate": 2.4601063829787237e-06,
|
| 1044 |
+
"loss": 1.69,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.49531486983441025,
|
| 1049 |
+
"grad_norm": 0.21047964692115784,
|
| 1050 |
+
"learning_rate": 2.4767287234042554e-06,
|
| 1051 |
+
"loss": 1.681,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.49863913070578214,
|
| 1056 |
+
"grad_norm": 0.2226460874080658,
|
| 1057 |
+
"learning_rate": 2.493351063829787e-06,
|
| 1058 |
+
"loss": 1.6613,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.501963391577154,
|
| 1063 |
+
"grad_norm": 0.21892835199832916,
|
| 1064 |
+
"learning_rate": 2.5099734042553192e-06,
|
| 1065 |
+
"loss": 1.6376,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.5052876524485259,
|
| 1070 |
+
"grad_norm": 0.20363831520080566,
|
| 1071 |
+
"learning_rate": 2.526595744680851e-06,
|
| 1072 |
+
"loss": 1.6541,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.5086119133198977,
|
| 1077 |
+
"grad_norm": 0.1988699585199356,
|
| 1078 |
+
"learning_rate": 2.543218085106383e-06,
|
| 1079 |
+
"loss": 1.6422,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.5119361741912697,
|
| 1084 |
+
"grad_norm": 0.2050096094608307,
|
| 1085 |
+
"learning_rate": 2.559840425531915e-06,
|
| 1086 |
+
"loss": 1.6377,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.5152604350626415,
|
| 1091 |
+
"grad_norm": 0.23265878856182098,
|
| 1092 |
+
"learning_rate": 2.5764627659574472e-06,
|
| 1093 |
+
"loss": 1.6251,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.5185846959340135,
|
| 1098 |
+
"grad_norm": 0.2024969905614853,
|
| 1099 |
+
"learning_rate": 2.593085106382979e-06,
|
| 1100 |
+
"loss": 1.6048,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.5219089568053853,
|
| 1105 |
+
"grad_norm": 0.21343863010406494,
|
| 1106 |
+
"learning_rate": 2.6097074468085106e-06,
|
| 1107 |
+
"loss": 1.6195,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.5252332176767572,
|
| 1112 |
+
"grad_norm": 0.1862565129995346,
|
| 1113 |
+
"learning_rate": 2.6263297872340427e-06,
|
| 1114 |
+
"loss": 1.5991,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.5285574785481291,
|
| 1119 |
+
"grad_norm": 0.22765249013900757,
|
| 1120 |
+
"learning_rate": 2.6429521276595744e-06,
|
| 1121 |
+
"loss": 1.5957,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.5318817394195009,
|
| 1126 |
+
"grad_norm": 0.19874997437000275,
|
| 1127 |
+
"learning_rate": 2.6595744680851065e-06,
|
| 1128 |
+
"loss": 1.5847,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.5352060002908728,
|
| 1133 |
+
"grad_norm": 0.25979486107826233,
|
| 1134 |
+
"learning_rate": 2.6761968085106386e-06,
|
| 1135 |
+
"loss": 1.6046,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.5385302611622447,
|
| 1140 |
+
"grad_norm": 0.1831529289484024,
|
| 1141 |
+
"learning_rate": 2.6928191489361703e-06,
|
| 1142 |
+
"loss": 1.5835,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.5418545220336166,
|
| 1147 |
+
"grad_norm": 0.2680751085281372,
|
| 1148 |
+
"learning_rate": 2.7094414893617024e-06,
|
| 1149 |
+
"loss": 1.6009,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.5451787829049884,
|
| 1154 |
+
"grad_norm": 0.18160907924175262,
|
| 1155 |
+
"learning_rate": 2.726063829787234e-06,
|
| 1156 |
+
"loss": 1.5666,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.5485030437763604,
|
| 1161 |
+
"grad_norm": 0.22875571250915527,
|
| 1162 |
+
"learning_rate": 2.742686170212766e-06,
|
| 1163 |
+
"loss": 1.5614,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.5518273046477322,
|
| 1168 |
+
"grad_norm": 0.21110033988952637,
|
| 1169 |
+
"learning_rate": 2.759308510638298e-06,
|
| 1170 |
+
"loss": 1.5707,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.5551515655191042,
|
| 1175 |
+
"grad_norm": 0.1887374073266983,
|
| 1176 |
+
"learning_rate": 2.77593085106383e-06,
|
| 1177 |
+
"loss": 1.5781,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.558475826390476,
|
| 1182 |
+
"grad_norm": 0.1916954219341278,
|
| 1183 |
+
"learning_rate": 2.7925531914893617e-06,
|
| 1184 |
+
"loss": 1.563,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.5618000872618478,
|
| 1189 |
+
"grad_norm": 0.21001753211021423,
|
| 1190 |
+
"learning_rate": 2.809175531914894e-06,
|
| 1191 |
+
"loss": 1.5495,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 0.5651243481332198,
|
| 1196 |
+
"grad_norm": 0.1702377200126648,
|
| 1197 |
+
"learning_rate": 2.825797872340426e-06,
|
| 1198 |
+
"loss": 1.5427,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.5684486090045916,
|
| 1203 |
+
"grad_norm": 0.19061295688152313,
|
| 1204 |
+
"learning_rate": 2.8424202127659576e-06,
|
| 1205 |
+
"loss": 1.5387,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 0.5717728698759635,
|
| 1210 |
+
"grad_norm": 0.17503058910369873,
|
| 1211 |
+
"learning_rate": 2.8590425531914893e-06,
|
| 1212 |
+
"loss": 1.5154,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 0.5750971307473354,
|
| 1217 |
+
"grad_norm": 0.1703094244003296,
|
| 1218 |
+
"learning_rate": 2.8756648936170214e-06,
|
| 1219 |
+
"loss": 1.5209,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 0.5784213916187073,
|
| 1224 |
+
"grad_norm": 0.22713126242160797,
|
| 1225 |
+
"learning_rate": 2.892287234042553e-06,
|
| 1226 |
+
"loss": 1.529,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 0.5817456524900791,
|
| 1231 |
+
"grad_norm": 0.16218431293964386,
|
| 1232 |
+
"learning_rate": 2.9089095744680852e-06,
|
| 1233 |
+
"loss": 1.505,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 0.5850699133614511,
|
| 1238 |
+
"grad_norm": 0.16082778573036194,
|
| 1239 |
+
"learning_rate": 2.9255319148936174e-06,
|
| 1240 |
+
"loss": 1.5312,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.5883941742328229,
|
| 1245 |
+
"grad_norm": 0.19500340521335602,
|
| 1246 |
+
"learning_rate": 2.942154255319149e-06,
|
| 1247 |
+
"loss": 1.4971,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.5917184351041948,
|
| 1252 |
+
"grad_norm": 0.16831324994564056,
|
| 1253 |
+
"learning_rate": 2.9587765957446807e-06,
|
| 1254 |
+
"loss": 1.5172,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.5950426959755667,
|
| 1259 |
+
"grad_norm": 0.17963413894176483,
|
| 1260 |
+
"learning_rate": 2.975398936170213e-06,
|
| 1261 |
+
"loss": 1.5076,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.5983669568469385,
|
| 1266 |
+
"grad_norm": 0.17123515903949738,
|
| 1267 |
+
"learning_rate": 2.992021276595745e-06,
|
| 1268 |
+
"loss": 1.4941,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.6016912177183105,
|
| 1273 |
+
"grad_norm": 0.15727902948856354,
|
| 1274 |
+
"learning_rate": 3.0086436170212766e-06,
|
| 1275 |
+
"loss": 1.4609,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 0.6050154785896823,
|
| 1280 |
+
"grad_norm": 0.1833077073097229,
|
| 1281 |
+
"learning_rate": 3.0252659574468088e-06,
|
| 1282 |
+
"loss": 1.5042,
|
| 1283 |
+
"step": 1820
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.6083397394610542,
|
| 1287 |
+
"grad_norm": 0.16962528228759766,
|
| 1288 |
+
"learning_rate": 3.041888297872341e-06,
|
| 1289 |
+
"loss": 1.4651,
|
| 1290 |
+
"step": 1830
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 0.6116640003324261,
|
| 1294 |
+
"grad_norm": 0.17829731106758118,
|
| 1295 |
+
"learning_rate": 3.0585106382978726e-06,
|
| 1296 |
+
"loss": 1.4907,
|
| 1297 |
+
"step": 1840
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 0.614988261203798,
|
| 1301 |
+
"grad_norm": 0.16981306672096252,
|
| 1302 |
+
"learning_rate": 3.0751329787234042e-06,
|
| 1303 |
+
"loss": 1.4683,
|
| 1304 |
+
"step": 1850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.6183125220751698,
|
| 1308 |
+
"grad_norm": 0.20783671736717224,
|
| 1309 |
+
"learning_rate": 3.0917553191489363e-06,
|
| 1310 |
+
"loss": 1.463,
|
| 1311 |
+
"step": 1860
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.6216367829465417,
|
| 1315 |
+
"grad_norm": 0.20343361794948578,
|
| 1316 |
+
"learning_rate": 3.108377659574468e-06,
|
| 1317 |
+
"loss": 1.4632,
|
| 1318 |
+
"step": 1870
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 0.6249610438179136,
|
| 1322 |
+
"grad_norm": 0.18592675030231476,
|
| 1323 |
+
"learning_rate": 3.125e-06,
|
| 1324 |
+
"loss": 1.4887,
|
| 1325 |
+
"step": 1880
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 0.6282853046892855,
|
| 1329 |
+
"grad_norm": 0.17272701859474182,
|
| 1330 |
+
"learning_rate": 3.141622340425532e-06,
|
| 1331 |
+
"loss": 1.4491,
|
| 1332 |
+
"step": 1890
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 0.6316095655606574,
|
| 1336 |
+
"grad_norm": 0.2021792083978653,
|
| 1337 |
+
"learning_rate": 3.1582446808510644e-06,
|
| 1338 |
+
"loss": 1.4537,
|
| 1339 |
+
"step": 1900
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 0.6349338264320292,
|
| 1343 |
+
"grad_norm": 0.16319766640663147,
|
| 1344 |
+
"learning_rate": 3.174867021276596e-06,
|
| 1345 |
+
"loss": 1.456,
|
| 1346 |
+
"step": 1910
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 0.6382580873034012,
|
| 1350 |
+
"grad_norm": 0.2344328761100769,
|
| 1351 |
+
"learning_rate": 3.1914893617021277e-06,
|
| 1352 |
+
"loss": 1.4801,
|
| 1353 |
+
"step": 1920
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 0.641582348174773,
|
| 1357 |
+
"grad_norm": 0.17495407164096832,
|
| 1358 |
+
"learning_rate": 3.20811170212766e-06,
|
| 1359 |
+
"loss": 1.4435,
|
| 1360 |
+
"step": 1930
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 0.6449066090461449,
|
| 1364 |
+
"grad_norm": 0.19222399592399597,
|
| 1365 |
+
"learning_rate": 3.2247340425531915e-06,
|
| 1366 |
+
"loss": 1.4391,
|
| 1367 |
+
"step": 1940
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 0.6482308699175168,
|
| 1371 |
+
"grad_norm": 0.24526530504226685,
|
| 1372 |
+
"learning_rate": 3.2413563829787232e-06,
|
| 1373 |
+
"loss": 1.4555,
|
| 1374 |
+
"step": 1950
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 0.6515551307888886,
|
| 1378 |
+
"grad_norm": 0.18150673806667328,
|
| 1379 |
+
"learning_rate": 3.2579787234042558e-06,
|
| 1380 |
+
"loss": 1.4396,
|
| 1381 |
+
"step": 1960
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 0.6548793916602605,
|
| 1385 |
+
"grad_norm": 0.18334811925888062,
|
| 1386 |
+
"learning_rate": 3.2746010638297875e-06,
|
| 1387 |
+
"loss": 1.4139,
|
| 1388 |
+
"step": 1970
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 0.6582036525316324,
|
| 1392 |
+
"grad_norm": 0.25186312198638916,
|
| 1393 |
+
"learning_rate": 3.291223404255319e-06,
|
| 1394 |
+
"loss": 1.439,
|
| 1395 |
+
"step": 1980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 0.6615279134030043,
|
| 1399 |
+
"grad_norm": 0.16558600962162018,
|
| 1400 |
+
"learning_rate": 3.3078457446808513e-06,
|
| 1401 |
+
"loss": 1.4383,
|
| 1402 |
+
"step": 1990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 0.6648521742743762,
|
| 1406 |
+
"grad_norm": 0.2373538315296173,
|
| 1407 |
+
"learning_rate": 3.324468085106383e-06,
|
| 1408 |
+
"loss": 1.4334,
|
| 1409 |
+
"step": 2000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 0.6681764351457481,
|
| 1413 |
+
"grad_norm": 0.2821474075317383,
|
| 1414 |
+
"learning_rate": 3.3410904255319146e-06,
|
| 1415 |
+
"loss": 1.4418,
|
| 1416 |
+
"step": 2010
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 0.6715006960171199,
|
| 1420 |
+
"grad_norm": 0.2443741410970688,
|
| 1421 |
+
"learning_rate": 3.357712765957447e-06,
|
| 1422 |
+
"loss": 1.4071,
|
| 1423 |
+
"step": 2020
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 0.6748249568884919,
|
| 1427 |
+
"grad_norm": 0.17468735575675964,
|
| 1428 |
+
"learning_rate": 3.374335106382979e-06,
|
| 1429 |
+
"loss": 1.4109,
|
| 1430 |
+
"step": 2030
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 0.6781492177598637,
|
| 1434 |
+
"grad_norm": 0.1655045598745346,
|
| 1435 |
+
"learning_rate": 3.3909574468085105e-06,
|
| 1436 |
+
"loss": 1.4049,
|
| 1437 |
+
"step": 2040
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 0.6814734786312356,
|
| 1441 |
+
"grad_norm": 0.17598801851272583,
|
| 1442 |
+
"learning_rate": 3.4075797872340427e-06,
|
| 1443 |
+
"loss": 1.4188,
|
| 1444 |
+
"step": 2050
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 0.6847977395026075,
|
| 1448 |
+
"grad_norm": 0.28528669476509094,
|
| 1449 |
+
"learning_rate": 3.4242021276595743e-06,
|
| 1450 |
+
"loss": 1.408,
|
| 1451 |
+
"step": 2060
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 0.6881220003739793,
|
| 1455 |
+
"grad_norm": 0.17654620110988617,
|
| 1456 |
+
"learning_rate": 3.440824468085106e-06,
|
| 1457 |
+
"loss": 1.4117,
|
| 1458 |
+
"step": 2070
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 0.6914462612453512,
|
| 1462 |
+
"grad_norm": 0.2636467516422272,
|
| 1463 |
+
"learning_rate": 3.4574468085106386e-06,
|
| 1464 |
+
"loss": 1.3947,
|
| 1465 |
+
"step": 2080
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 0.6947705221167231,
|
| 1469 |
+
"grad_norm": 0.26495933532714844,
|
| 1470 |
+
"learning_rate": 3.4740691489361703e-06,
|
| 1471 |
+
"loss": 1.398,
|
| 1472 |
+
"step": 2090
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 0.698094782988095,
|
| 1476 |
+
"grad_norm": 0.3873574435710907,
|
| 1477 |
+
"learning_rate": 3.490691489361702e-06,
|
| 1478 |
+
"loss": 1.4204,
|
| 1479 |
+
"step": 2100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 0.7014190438594669,
|
| 1483 |
+
"grad_norm": 0.327854186296463,
|
| 1484 |
+
"learning_rate": 3.5073138297872345e-06,
|
| 1485 |
+
"loss": 1.3744,
|
| 1486 |
+
"step": 2110
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 0.7047433047308388,
|
| 1490 |
+
"grad_norm": 0.308570921421051,
|
| 1491 |
+
"learning_rate": 3.523936170212766e-06,
|
| 1492 |
+
"loss": 1.4293,
|
| 1493 |
+
"step": 2120
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 0.7080675656022106,
|
| 1497 |
+
"grad_norm": 0.21123336255550385,
|
| 1498 |
+
"learning_rate": 3.5405585106382983e-06,
|
| 1499 |
+
"loss": 1.3878,
|
| 1500 |
+
"step": 2130
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 0.7113918264735826,
|
| 1504 |
+
"grad_norm": 0.18777534365653992,
|
| 1505 |
+
"learning_rate": 3.55718085106383e-06,
|
| 1506 |
+
"loss": 1.3882,
|
| 1507 |
+
"step": 2140
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 0.7147160873449544,
|
| 1511 |
+
"grad_norm": 0.2535350024700165,
|
| 1512 |
+
"learning_rate": 3.5738031914893617e-06,
|
| 1513 |
+
"loss": 1.3974,
|
| 1514 |
+
"step": 2150
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 0.7180403482163262,
|
| 1518 |
+
"grad_norm": 0.15405435860157013,
|
| 1519 |
+
"learning_rate": 3.590425531914894e-06,
|
| 1520 |
+
"loss": 1.3853,
|
| 1521 |
+
"step": 2160
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 0.7213646090876982,
|
| 1525 |
+
"grad_norm": 0.1863648146390915,
|
| 1526 |
+
"learning_rate": 3.607047872340426e-06,
|
| 1527 |
+
"loss": 1.3835,
|
| 1528 |
+
"step": 2170
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 0.72468886995907,
|
| 1532 |
+
"grad_norm": 0.18587157130241394,
|
| 1533 |
+
"learning_rate": 3.6236702127659576e-06,
|
| 1534 |
+
"loss": 1.3711,
|
| 1535 |
+
"step": 2180
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 0.728013130830442,
|
| 1539 |
+
"grad_norm": 0.18254730105400085,
|
| 1540 |
+
"learning_rate": 3.6402925531914897e-06,
|
| 1541 |
+
"loss": 1.3768,
|
| 1542 |
+
"step": 2190
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 0.7313373917018138,
|
| 1546 |
+
"grad_norm": 0.21665969491004944,
|
| 1547 |
+
"learning_rate": 3.6569148936170214e-06,
|
| 1548 |
+
"loss": 1.3638,
|
| 1549 |
+
"step": 2200
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 0.7346616525731857,
|
| 1553 |
+
"grad_norm": 0.15701924264431,
|
| 1554 |
+
"learning_rate": 3.673537234042553e-06,
|
| 1555 |
+
"loss": 1.3885,
|
| 1556 |
+
"step": 2210
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 0.7379859134445576,
|
| 1560 |
+
"grad_norm": 0.19307725131511688,
|
| 1561 |
+
"learning_rate": 3.6901595744680856e-06,
|
| 1562 |
+
"loss": 1.3933,
|
| 1563 |
+
"step": 2220
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 0.7413101743159295,
|
| 1567 |
+
"grad_norm": 0.16837100684642792,
|
| 1568 |
+
"learning_rate": 3.7067819148936173e-06,
|
| 1569 |
+
"loss": 1.3685,
|
| 1570 |
+
"step": 2230
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 0.7446344351873013,
|
| 1574 |
+
"grad_norm": 0.2914402484893799,
|
| 1575 |
+
"learning_rate": 3.723404255319149e-06,
|
| 1576 |
+
"loss": 1.3802,
|
| 1577 |
+
"step": 2240
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 0.7479586960586732,
|
| 1581 |
+
"grad_norm": 0.2770545184612274,
|
| 1582 |
+
"learning_rate": 3.7400265957446815e-06,
|
| 1583 |
+
"loss": 1.3575,
|
| 1584 |
+
"step": 2250
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 0.7512829569300451,
|
| 1588 |
+
"grad_norm": 0.19819234311580658,
|
| 1589 |
+
"learning_rate": 3.756648936170213e-06,
|
| 1590 |
+
"loss": 1.3695,
|
| 1591 |
+
"step": 2260
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 0.7546072178014169,
|
| 1595 |
+
"grad_norm": 0.15371359884738922,
|
| 1596 |
+
"learning_rate": 3.7732712765957445e-06,
|
| 1597 |
+
"loss": 1.3514,
|
| 1598 |
+
"step": 2270
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 0.7579314786727889,
|
| 1602 |
+
"grad_norm": 0.26700448989868164,
|
| 1603 |
+
"learning_rate": 3.789893617021277e-06,
|
| 1604 |
+
"loss": 1.3689,
|
| 1605 |
+
"step": 2280
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 0.7612557395441607,
|
| 1609 |
+
"grad_norm": 0.2938506007194519,
|
| 1610 |
+
"learning_rate": 3.8065159574468087e-06,
|
| 1611 |
+
"loss": 1.3518,
|
| 1612 |
+
"step": 2290
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 0.7645800004155326,
|
| 1616 |
+
"grad_norm": 0.2514606714248657,
|
| 1617 |
+
"learning_rate": 3.823138297872341e-06,
|
| 1618 |
+
"loss": 1.3655,
|
| 1619 |
+
"step": 2300
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 0.7679042612869045,
|
| 1623 |
+
"grad_norm": 0.2503184378147125,
|
| 1624 |
+
"learning_rate": 3.8397606382978725e-06,
|
| 1625 |
+
"loss": 1.3511,
|
| 1626 |
+
"step": 2310
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 0.7712285221582764,
|
| 1630 |
+
"grad_norm": 0.1815042346715927,
|
| 1631 |
+
"learning_rate": 3.856382978723404e-06,
|
| 1632 |
+
"loss": 1.383,
|
| 1633 |
+
"step": 2320
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 0.7745527830296483,
|
| 1637 |
+
"grad_norm": 0.25425419211387634,
|
| 1638 |
+
"learning_rate": 3.873005319148936e-06,
|
| 1639 |
+
"loss": 1.3354,
|
| 1640 |
+
"step": 2330
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 0.7778770439010201,
|
| 1644 |
+
"grad_norm": 0.18466657400131226,
|
| 1645 |
+
"learning_rate": 3.889627659574468e-06,
|
| 1646 |
+
"loss": 1.3514,
|
| 1647 |
+
"step": 2340
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 0.781201304772392,
|
| 1651 |
+
"grad_norm": 0.1782332807779312,
|
| 1652 |
+
"learning_rate": 3.90625e-06,
|
| 1653 |
+
"loss": 1.32,
|
| 1654 |
+
"step": 2350
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 0.7845255656437639,
|
| 1658 |
+
"grad_norm": 0.27637991309165955,
|
| 1659 |
+
"learning_rate": 3.922872340425532e-06,
|
| 1660 |
+
"loss": 1.3383,
|
| 1661 |
+
"step": 2360
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 0.7878498265151358,
|
| 1665 |
+
"grad_norm": 0.17314772307872772,
|
| 1666 |
+
"learning_rate": 3.939494680851064e-06,
|
| 1667 |
+
"loss": 1.3314,
|
| 1668 |
+
"step": 2370
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 0.7911740873865076,
|
| 1672 |
+
"grad_norm": 0.3641667068004608,
|
| 1673 |
+
"learning_rate": 3.956117021276596e-06,
|
| 1674 |
+
"loss": 1.3543,
|
| 1675 |
+
"step": 2380
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 0.7944983482578796,
|
| 1679 |
+
"grad_norm": 0.3088253438472748,
|
| 1680 |
+
"learning_rate": 3.972739361702128e-06,
|
| 1681 |
+
"loss": 1.3444,
|
| 1682 |
+
"step": 2390
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 0.7978226091292514,
|
| 1686 |
+
"grad_norm": 0.25276973843574524,
|
| 1687 |
+
"learning_rate": 3.98936170212766e-06,
|
| 1688 |
+
"loss": 1.3102,
|
| 1689 |
+
"step": 2400
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 0.8011468700006233,
|
| 1693 |
+
"grad_norm": 0.26414382457733154,
|
| 1694 |
+
"learning_rate": 4.005984042553192e-06,
|
| 1695 |
+
"loss": 1.3119,
|
| 1696 |
+
"step": 2410
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 0.8044711308719952,
|
| 1700 |
+
"grad_norm": 0.1684638261795044,
|
| 1701 |
+
"learning_rate": 4.022606382978724e-06,
|
| 1702 |
+
"loss": 1.3204,
|
| 1703 |
+
"step": 2420
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 0.807795391743367,
|
| 1707 |
+
"grad_norm": 0.18500946462154388,
|
| 1708 |
+
"learning_rate": 4.039228723404256e-06,
|
| 1709 |
+
"loss": 1.3251,
|
| 1710 |
+
"step": 2430
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 0.811119652614739,
|
| 1714 |
+
"grad_norm": 0.2754835784435272,
|
| 1715 |
+
"learning_rate": 4.055851063829787e-06,
|
| 1716 |
+
"loss": 1.3258,
|
| 1717 |
+
"step": 2440
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 0.8144439134861108,
|
| 1721 |
+
"grad_norm": 0.18949855864048004,
|
| 1722 |
+
"learning_rate": 4.072473404255319e-06,
|
| 1723 |
+
"loss": 1.3145,
|
| 1724 |
+
"step": 2450
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 0.8177681743574827,
|
| 1728 |
+
"grad_norm": 0.6927218437194824,
|
| 1729 |
+
"learning_rate": 4.089095744680851e-06,
|
| 1730 |
+
"loss": 1.3205,
|
| 1731 |
+
"step": 2460
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 0.8210924352288546,
|
| 1735 |
+
"grad_norm": 0.36098670959472656,
|
| 1736 |
+
"learning_rate": 4.105718085106383e-06,
|
| 1737 |
+
"loss": 1.3295,
|
| 1738 |
+
"step": 2470
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 0.8244166961002265,
|
| 1742 |
+
"grad_norm": 0.25839686393737793,
|
| 1743 |
+
"learning_rate": 4.1223404255319146e-06,
|
| 1744 |
+
"loss": 1.321,
|
| 1745 |
+
"step": 2480
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 0.8277409569715983,
|
| 1749 |
+
"grad_norm": 0.18720127642154694,
|
| 1750 |
+
"learning_rate": 4.138962765957447e-06,
|
| 1751 |
+
"loss": 1.2975,
|
| 1752 |
+
"step": 2490
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 0.8310652178429703,
|
| 1756 |
+
"grad_norm": 0.17975495755672455,
|
| 1757 |
+
"learning_rate": 4.155585106382979e-06,
|
| 1758 |
+
"loss": 1.318,
|
| 1759 |
+
"step": 2500
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 0.8343894787143421,
|
| 1763 |
+
"grad_norm": 0.3536490797996521,
|
| 1764 |
+
"learning_rate": 4.1722074468085105e-06,
|
| 1765 |
+
"loss": 1.312,
|
| 1766 |
+
"step": 2510
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 0.8377137395857139,
|
| 1770 |
+
"grad_norm": 0.4771016836166382,
|
| 1771 |
+
"learning_rate": 4.188829787234043e-06,
|
| 1772 |
+
"loss": 1.3242,
|
| 1773 |
+
"step": 2520
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 0.8410380004570859,
|
| 1777 |
+
"grad_norm": 0.6679111123085022,
|
| 1778 |
+
"learning_rate": 4.205452127659575e-06,
|
| 1779 |
+
"loss": 1.3108,
|
| 1780 |
+
"step": 2530
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 0.8443622613284577,
|
| 1784 |
+
"grad_norm": 0.2579514682292938,
|
| 1785 |
+
"learning_rate": 4.222074468085106e-06,
|
| 1786 |
+
"loss": 1.3266,
|
| 1787 |
+
"step": 2540
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 0.8476865221998297,
|
| 1791 |
+
"grad_norm": 0.17981646955013275,
|
| 1792 |
+
"learning_rate": 4.238696808510639e-06,
|
| 1793 |
+
"loss": 1.3182,
|
| 1794 |
+
"step": 2550
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 0.8510107830712015,
|
| 1798 |
+
"grad_norm": 0.19210171699523926,
|
| 1799 |
+
"learning_rate": 4.255319148936171e-06,
|
| 1800 |
+
"loss": 1.3022,
|
| 1801 |
+
"step": 2560
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 0.8543350439425734,
|
| 1805 |
+
"grad_norm": 0.18870197236537933,
|
| 1806 |
+
"learning_rate": 4.271941489361702e-06,
|
| 1807 |
+
"loss": 1.2911,
|
| 1808 |
+
"step": 2570
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 0.8576593048139453,
|
| 1812 |
+
"grad_norm": 0.24970248341560364,
|
| 1813 |
+
"learning_rate": 4.288563829787234e-06,
|
| 1814 |
+
"loss": 1.3006,
|
| 1815 |
+
"step": 2580
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 0.8609835656853172,
|
| 1819 |
+
"grad_norm": 0.18395459651947021,
|
| 1820 |
+
"learning_rate": 4.305186170212766e-06,
|
| 1821 |
+
"loss": 1.308,
|
| 1822 |
+
"step": 2590
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 0.864307826556689,
|
| 1826 |
+
"grad_norm": 0.19471628963947296,
|
| 1827 |
+
"learning_rate": 4.321808510638298e-06,
|
| 1828 |
+
"loss": 1.2775,
|
| 1829 |
+
"step": 2600
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 0.867632087428061,
|
| 1833 |
+
"grad_norm": 0.2713675796985626,
|
| 1834 |
+
"learning_rate": 4.33843085106383e-06,
|
| 1835 |
+
"loss": 1.2845,
|
| 1836 |
+
"step": 2610
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 0.8709563482994328,
|
| 1840 |
+
"grad_norm": 0.268235445022583,
|
| 1841 |
+
"learning_rate": 4.355053191489362e-06,
|
| 1842 |
+
"loss": 1.276,
|
| 1843 |
+
"step": 2620
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 0.8742806091708046,
|
| 1847 |
+
"grad_norm": 0.2281683087348938,
|
| 1848 |
+
"learning_rate": 4.371675531914894e-06,
|
| 1849 |
+
"loss": 1.2988,
|
| 1850 |
+
"step": 2630
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 0.8776048700421766,
|
| 1854 |
+
"grad_norm": 0.20509906113147736,
|
| 1855 |
+
"learning_rate": 4.388297872340426e-06,
|
| 1856 |
+
"loss": 1.2993,
|
| 1857 |
+
"step": 2640
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 0.8809291309135484,
|
| 1861 |
+
"grad_norm": 0.19255685806274414,
|
| 1862 |
+
"learning_rate": 4.4049202127659575e-06,
|
| 1863 |
+
"loss": 1.2816,
|
| 1864 |
+
"step": 2650
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 0.8842533917849204,
|
| 1868 |
+
"grad_norm": 0.1968984156847,
|
| 1869 |
+
"learning_rate": 4.42154255319149e-06,
|
| 1870 |
+
"loss": 1.2861,
|
| 1871 |
+
"step": 2660
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 0.8875776526562922,
|
| 1875 |
+
"grad_norm": 0.5413665771484375,
|
| 1876 |
+
"learning_rate": 4.438164893617022e-06,
|
| 1877 |
+
"loss": 1.2779,
|
| 1878 |
+
"step": 2670
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 0.8909019135276641,
|
| 1882 |
+
"grad_norm": 0.19381985068321228,
|
| 1883 |
+
"learning_rate": 4.454787234042553e-06,
|
| 1884 |
+
"loss": 1.2835,
|
| 1885 |
+
"step": 2680
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 0.894226174399036,
|
| 1889 |
+
"grad_norm": 0.19788819551467896,
|
| 1890 |
+
"learning_rate": 4.471409574468086e-06,
|
| 1891 |
+
"loss": 1.2788,
|
| 1892 |
+
"step": 2690
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 0.8975504352704079,
|
| 1896 |
+
"grad_norm": 0.19661852717399597,
|
| 1897 |
+
"learning_rate": 4.488031914893618e-06,
|
| 1898 |
+
"loss": 1.2782,
|
| 1899 |
+
"step": 2700
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 0.9008746961417797,
|
| 1903 |
+
"grad_norm": 0.1792154759168625,
|
| 1904 |
+
"learning_rate": 4.504654255319149e-06,
|
| 1905 |
+
"loss": 1.2561,
|
| 1906 |
+
"step": 2710
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 0.9041989570131516,
|
| 1910 |
+
"grad_norm": 0.19385012984275818,
|
| 1911 |
+
"learning_rate": 4.521276595744681e-06,
|
| 1912 |
+
"loss": 1.2743,
|
| 1913 |
+
"step": 2720
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 0.9075232178845235,
|
| 1917 |
+
"grad_norm": 0.19243328273296356,
|
| 1918 |
+
"learning_rate": 4.537898936170213e-06,
|
| 1919 |
+
"loss": 1.2536,
|
| 1920 |
+
"step": 2730
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 0.9108474787558953,
|
| 1924 |
+
"grad_norm": 0.1733260452747345,
|
| 1925 |
+
"learning_rate": 4.554521276595744e-06,
|
| 1926 |
+
"loss": 1.2497,
|
| 1927 |
+
"step": 2740
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 0.9141717396272673,
|
| 1931 |
+
"grad_norm": 0.1626838594675064,
|
| 1932 |
+
"learning_rate": 4.571143617021277e-06,
|
| 1933 |
+
"loss": 1.2663,
|
| 1934 |
+
"step": 2750
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 0.9174960004986391,
|
| 1938 |
+
"grad_norm": 0.22138752043247223,
|
| 1939 |
+
"learning_rate": 4.587765957446809e-06,
|
| 1940 |
+
"loss": 1.2627,
|
| 1941 |
+
"step": 2760
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 0.920820261370011,
|
| 1945 |
+
"grad_norm": 0.2485809475183487,
|
| 1946 |
+
"learning_rate": 4.60438829787234e-06,
|
| 1947 |
+
"loss": 1.2625,
|
| 1948 |
+
"step": 2770
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 0.9241445222413829,
|
| 1952 |
+
"grad_norm": 0.19163627922534943,
|
| 1953 |
+
"learning_rate": 4.621010638297873e-06,
|
| 1954 |
+
"loss": 1.2637,
|
| 1955 |
+
"step": 2780
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 0.9274687831127548,
|
| 1959 |
+
"grad_norm": 0.20371121168136597,
|
| 1960 |
+
"learning_rate": 4.6376329787234045e-06,
|
| 1961 |
+
"loss": 1.2468,
|
| 1962 |
+
"step": 2790
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 0.9307930439841267,
|
| 1966 |
+
"grad_norm": 0.207363098859787,
|
| 1967 |
+
"learning_rate": 4.654255319148936e-06,
|
| 1968 |
+
"loss": 1.2533,
|
| 1969 |
+
"step": 2800
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 0.9341173048554985,
|
| 1973 |
+
"grad_norm": 0.1894274204969406,
|
| 1974 |
+
"learning_rate": 4.670877659574469e-06,
|
| 1975 |
+
"loss": 1.2573,
|
| 1976 |
+
"step": 2810
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 0.9374415657268704,
|
| 1980 |
+
"grad_norm": 0.19954580068588257,
|
| 1981 |
+
"learning_rate": 4.6875000000000004e-06,
|
| 1982 |
+
"loss": 1.2416,
|
| 1983 |
+
"step": 2820
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 0.9407658265982423,
|
| 1987 |
+
"grad_norm": 0.21874602138996124,
|
| 1988 |
+
"learning_rate": 4.704122340425532e-06,
|
| 1989 |
+
"loss": 1.2639,
|
| 1990 |
+
"step": 2830
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 0.9440900874696142,
|
| 1994 |
+
"grad_norm": 0.20103342831134796,
|
| 1995 |
+
"learning_rate": 4.720744680851065e-06,
|
| 1996 |
+
"loss": 1.2291,
|
| 1997 |
+
"step": 2840
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 0.947414348340986,
|
| 2001 |
+
"grad_norm": 0.1904035359621048,
|
| 2002 |
+
"learning_rate": 4.737367021276596e-06,
|
| 2003 |
+
"loss": 1.2405,
|
| 2004 |
+
"step": 2850
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 0.950738609212358,
|
| 2008 |
+
"grad_norm": 0.1963120400905609,
|
| 2009 |
+
"learning_rate": 4.753989361702128e-06,
|
| 2010 |
+
"loss": 1.2358,
|
| 2011 |
+
"step": 2860
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 0.9540628700837298,
|
| 2015 |
+
"grad_norm": 0.18605996668338776,
|
| 2016 |
+
"learning_rate": 4.77061170212766e-06,
|
| 2017 |
+
"loss": 1.2436,
|
| 2018 |
+
"step": 2870
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 0.9573871309551018,
|
| 2022 |
+
"grad_norm": 0.3102928698062897,
|
| 2023 |
+
"learning_rate": 4.787234042553191e-06,
|
| 2024 |
+
"loss": 1.2512,
|
| 2025 |
+
"step": 2880
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 0.9607113918264736,
|
| 2029 |
+
"grad_norm": 0.19887375831604004,
|
| 2030 |
+
"learning_rate": 4.803856382978723e-06,
|
| 2031 |
+
"loss": 1.252,
|
| 2032 |
+
"step": 2890
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 0.9640356526978454,
|
| 2036 |
+
"grad_norm": 0.19327284395694733,
|
| 2037 |
+
"learning_rate": 4.820478723404256e-06,
|
| 2038 |
+
"loss": 1.2266,
|
| 2039 |
+
"step": 2900
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 0.9673599135692174,
|
| 2043 |
+
"grad_norm": 0.19663333892822266,
|
| 2044 |
+
"learning_rate": 4.837101063829787e-06,
|
| 2045 |
+
"loss": 1.2413,
|
| 2046 |
+
"step": 2910
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 0.9706841744405892,
|
| 2050 |
+
"grad_norm": 0.23062920570373535,
|
| 2051 |
+
"learning_rate": 4.853723404255319e-06,
|
| 2052 |
+
"loss": 1.2314,
|
| 2053 |
+
"step": 2920
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 0.9740084353119611,
|
| 2057 |
+
"grad_norm": 0.2594512403011322,
|
| 2058 |
+
"learning_rate": 4.8703457446808515e-06,
|
| 2059 |
+
"loss": 1.2444,
|
| 2060 |
+
"step": 2930
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 0.977332696183333,
|
| 2064 |
+
"grad_norm": 0.19132867455482483,
|
| 2065 |
+
"learning_rate": 4.886968085106383e-06,
|
| 2066 |
+
"loss": 1.2386,
|
| 2067 |
+
"step": 2940
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 0.9806569570547049,
|
| 2071 |
+
"grad_norm": 0.2195463478565216,
|
| 2072 |
+
"learning_rate": 4.903590425531915e-06,
|
| 2073 |
+
"loss": 1.2257,
|
| 2074 |
+
"step": 2950
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 0.9839812179260767,
|
| 2078 |
+
"grad_norm": 0.19802413880825043,
|
| 2079 |
+
"learning_rate": 4.9202127659574475e-06,
|
| 2080 |
+
"loss": 1.2367,
|
| 2081 |
+
"step": 2960
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 0.9873054787974487,
|
| 2085 |
+
"grad_norm": 0.22345732152462006,
|
| 2086 |
+
"learning_rate": 4.936835106382979e-06,
|
| 2087 |
+
"loss": 1.1885,
|
| 2088 |
+
"step": 2970
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 0.9906297396688205,
|
| 2092 |
+
"grad_norm": 0.29191696643829346,
|
| 2093 |
+
"learning_rate": 4.953457446808511e-06,
|
| 2094 |
+
"loss": 1.213,
|
| 2095 |
+
"step": 2980
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 0.9939540005401923,
|
| 2099 |
+
"grad_norm": 0.2331935614347458,
|
| 2100 |
+
"learning_rate": 4.9700797872340425e-06,
|
| 2101 |
+
"loss": 1.2078,
|
| 2102 |
+
"step": 2990
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 0.9972782614115643,
|
| 2106 |
+
"grad_norm": 0.20795047283172607,
|
| 2107 |
+
"learning_rate": 4.986702127659574e-06,
|
| 2108 |
+
"loss": 1.2167,
|
| 2109 |
+
"step": 3000
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 1.0006025222829362,
|
| 2113 |
+
"grad_norm": 0.19333745539188385,
|
| 2114 |
+
"learning_rate": 5.003324468085106e-06,
|
| 2115 |
+
"loss": 1.2232,
|
| 2116 |
+
"step": 3010
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 1.003926783154308,
|
| 2120 |
+
"grad_norm": 0.21302959322929382,
|
| 2121 |
+
"learning_rate": 5.0199468085106384e-06,
|
| 2122 |
+
"loss": 1.2129,
|
| 2123 |
+
"step": 3020
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 1.0072510440256799,
|
| 2127 |
+
"grad_norm": 0.21599692106246948,
|
| 2128 |
+
"learning_rate": 5.03656914893617e-06,
|
| 2129 |
+
"loss": 1.1955,
|
| 2130 |
+
"step": 3030
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 1.0105753048970518,
|
| 2134 |
+
"grad_norm": 0.22842222452163696,
|
| 2135 |
+
"learning_rate": 5.053191489361702e-06,
|
| 2136 |
+
"loss": 1.2085,
|
| 2137 |
+
"step": 3040
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 1.0138995657684238,
|
| 2141 |
+
"grad_norm": 0.19472521543502808,
|
| 2142 |
+
"learning_rate": 5.069813829787234e-06,
|
| 2143 |
+
"loss": 1.2029,
|
| 2144 |
+
"step": 3050
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 1.0172238266397955,
|
| 2148 |
+
"grad_norm": 0.22058968245983124,
|
| 2149 |
+
"learning_rate": 5.086436170212766e-06,
|
| 2150 |
+
"loss": 1.2058,
|
| 2151 |
+
"step": 3060
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 1.0205480875111674,
|
| 2155 |
+
"grad_norm": 0.23935015499591827,
|
| 2156 |
+
"learning_rate": 5.1030585106382986e-06,
|
| 2157 |
+
"loss": 1.2052,
|
| 2158 |
+
"step": 3070
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 1.0238723483825394,
|
| 2162 |
+
"grad_norm": 0.2543661594390869,
|
| 2163 |
+
"learning_rate": 5.11968085106383e-06,
|
| 2164 |
+
"loss": 1.1964,
|
| 2165 |
+
"step": 3080
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 1.027196609253911,
|
| 2169 |
+
"grad_norm": 0.22757141292095184,
|
| 2170 |
+
"learning_rate": 5.136303191489362e-06,
|
| 2171 |
+
"loss": 1.2131,
|
| 2172 |
+
"step": 3090
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 1.030520870125283,
|
| 2176 |
+
"grad_norm": 0.2462802529335022,
|
| 2177 |
+
"learning_rate": 5.1529255319148945e-06,
|
| 2178 |
+
"loss": 1.2,
|
| 2179 |
+
"step": 3100
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 1.033845130996655,
|
| 2183 |
+
"grad_norm": 0.3865806758403778,
|
| 2184 |
+
"learning_rate": 5.169547872340426e-06,
|
| 2185 |
+
"loss": 1.1976,
|
| 2186 |
+
"step": 3110
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 1.037169391868027,
|
| 2190 |
+
"grad_norm": 0.2478027641773224,
|
| 2191 |
+
"learning_rate": 5.186170212765958e-06,
|
| 2192 |
+
"loss": 1.1892,
|
| 2193 |
+
"step": 3120
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 1.0404936527393986,
|
| 2197 |
+
"grad_norm": 0.27189669013023376,
|
| 2198 |
+
"learning_rate": 5.2027925531914895e-06,
|
| 2199 |
+
"loss": 1.1915,
|
| 2200 |
+
"step": 3130
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 1.0438179136107706,
|
| 2204 |
+
"grad_norm": 0.24409879744052887,
|
| 2205 |
+
"learning_rate": 5.219414893617021e-06,
|
| 2206 |
+
"loss": 1.193,
|
| 2207 |
+
"step": 3140
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 1.0471421744821425,
|
| 2211 |
+
"grad_norm": 0.2317332923412323,
|
| 2212 |
+
"learning_rate": 5.236037234042553e-06,
|
| 2213 |
+
"loss": 1.209,
|
| 2214 |
+
"step": 3150
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 1.0504664353535145,
|
| 2218 |
+
"grad_norm": 0.2069033831357956,
|
| 2219 |
+
"learning_rate": 5.2526595744680855e-06,
|
| 2220 |
+
"loss": 1.1926,
|
| 2221 |
+
"step": 3160
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 1.0537906962248862,
|
| 2225 |
+
"grad_norm": 0.233448788523674,
|
| 2226 |
+
"learning_rate": 5.269281914893617e-06,
|
| 2227 |
+
"loss": 1.195,
|
| 2228 |
+
"step": 3170
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 1.0571149570962581,
|
| 2232 |
+
"grad_norm": 0.20409508049488068,
|
| 2233 |
+
"learning_rate": 5.285904255319149e-06,
|
| 2234 |
+
"loss": 1.1984,
|
| 2235 |
+
"step": 3180
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 1.06043921796763,
|
| 2239 |
+
"grad_norm": 0.24429172277450562,
|
| 2240 |
+
"learning_rate": 5.302526595744681e-06,
|
| 2241 |
+
"loss": 1.1608,
|
| 2242 |
+
"step": 3190
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 1.0637634788390018,
|
| 2246 |
+
"grad_norm": 0.22169041633605957,
|
| 2247 |
+
"learning_rate": 5.319148936170213e-06,
|
| 2248 |
+
"loss": 1.1735,
|
| 2249 |
+
"step": 3200
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 1.0670877397103737,
|
| 2253 |
+
"grad_norm": 0.25986573100090027,
|
| 2254 |
+
"learning_rate": 5.335771276595745e-06,
|
| 2255 |
+
"loss": 1.1706,
|
| 2256 |
+
"step": 3210
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 1.0704120005817457,
|
| 2260 |
+
"grad_norm": 0.301274836063385,
|
| 2261 |
+
"learning_rate": 5.352393617021277e-06,
|
| 2262 |
+
"loss": 1.1588,
|
| 2263 |
+
"step": 3220
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 1.0737362614531176,
|
| 2267 |
+
"grad_norm": 0.21944095194339752,
|
| 2268 |
+
"learning_rate": 5.369015957446809e-06,
|
| 2269 |
+
"loss": 1.201,
|
| 2270 |
+
"step": 3230
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 1.0770605223244893,
|
| 2274 |
+
"grad_norm": 0.2160642445087433,
|
| 2275 |
+
"learning_rate": 5.385638297872341e-06,
|
| 2276 |
+
"loss": 1.189,
|
| 2277 |
+
"step": 3240
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 1.0803847831958613,
|
| 2281 |
+
"grad_norm": 0.28684699535369873,
|
| 2282 |
+
"learning_rate": 5.402260638297873e-06,
|
| 2283 |
+
"loss": 1.1844,
|
| 2284 |
+
"step": 3250
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 1.0837090440672332,
|
| 2288 |
+
"grad_norm": 0.2690293788909912,
|
| 2289 |
+
"learning_rate": 5.418882978723405e-06,
|
| 2290 |
+
"loss": 1.1671,
|
| 2291 |
+
"step": 3260
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 1.087033304938605,
|
| 2295 |
+
"grad_norm": 0.24337069690227509,
|
| 2296 |
+
"learning_rate": 5.4355053191489366e-06,
|
| 2297 |
+
"loss": 1.177,
|
| 2298 |
+
"step": 3270
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 1.090357565809977,
|
| 2302 |
+
"grad_norm": 0.2460750788450241,
|
| 2303 |
+
"learning_rate": 5.452127659574468e-06,
|
| 2304 |
+
"loss": 1.1843,
|
| 2305 |
+
"step": 3280
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 1.0936818266813488,
|
| 2309 |
+
"grad_norm": 0.23255892097949982,
|
| 2310 |
+
"learning_rate": 5.46875e-06,
|
| 2311 |
+
"loss": 1.1669,
|
| 2312 |
+
"step": 3290
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 1.0970060875527208,
|
| 2316 |
+
"grad_norm": 0.2593073844909668,
|
| 2317 |
+
"learning_rate": 5.485372340425532e-06,
|
| 2318 |
+
"loss": 1.1703,
|
| 2319 |
+
"step": 3300
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 1.1003303484240925,
|
| 2323 |
+
"grad_norm": 0.26776090264320374,
|
| 2324 |
+
"learning_rate": 5.501994680851064e-06,
|
| 2325 |
+
"loss": 1.1595,
|
| 2326 |
+
"step": 3310
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 1.1036546092954644,
|
| 2330 |
+
"grad_norm": 0.2763229012489319,
|
| 2331 |
+
"learning_rate": 5.518617021276596e-06,
|
| 2332 |
+
"loss": 1.1687,
|
| 2333 |
+
"step": 3320
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 1.1069788701668364,
|
| 2337 |
+
"grad_norm": 0.22993025183677673,
|
| 2338 |
+
"learning_rate": 5.5352393617021275e-06,
|
| 2339 |
+
"loss": 1.1599,
|
| 2340 |
+
"step": 3330
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 1.1103031310382083,
|
| 2344 |
+
"grad_norm": 0.33553963899612427,
|
| 2345 |
+
"learning_rate": 5.55186170212766e-06,
|
| 2346 |
+
"loss": 1.1313,
|
| 2347 |
+
"step": 3340
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 1.11362739190958,
|
| 2351 |
+
"grad_norm": 0.26335400342941284,
|
| 2352 |
+
"learning_rate": 5.568484042553192e-06,
|
| 2353 |
+
"loss": 1.1617,
|
| 2354 |
+
"step": 3350
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 1.116951652780952,
|
| 2358 |
+
"grad_norm": 0.2727535367012024,
|
| 2359 |
+
"learning_rate": 5.5851063829787235e-06,
|
| 2360 |
+
"loss": 1.1502,
|
| 2361 |
+
"step": 3360
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 1.120275913652324,
|
| 2365 |
+
"grad_norm": 0.2764947712421417,
|
| 2366 |
+
"learning_rate": 5.601728723404256e-06,
|
| 2367 |
+
"loss": 1.1632,
|
| 2368 |
+
"step": 3370
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 1.1236001745236956,
|
| 2372 |
+
"grad_norm": 0.31052595376968384,
|
| 2373 |
+
"learning_rate": 5.618351063829788e-06,
|
| 2374 |
+
"loss": 1.1566,
|
| 2375 |
+
"step": 3380
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 1.1269244353950676,
|
| 2379 |
+
"grad_norm": 0.2513100504875183,
|
| 2380 |
+
"learning_rate": 5.634973404255319e-06,
|
| 2381 |
+
"loss": 1.1419,
|
| 2382 |
+
"step": 3390
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 1.1302486962664395,
|
| 2386 |
+
"grad_norm": 0.24479760229587555,
|
| 2387 |
+
"learning_rate": 5.651595744680852e-06,
|
| 2388 |
+
"loss": 1.1444,
|
| 2389 |
+
"step": 3400
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 1.1335729571378115,
|
| 2393 |
+
"grad_norm": 0.49369576573371887,
|
| 2394 |
+
"learning_rate": 5.668218085106384e-06,
|
| 2395 |
+
"loss": 1.1358,
|
| 2396 |
+
"step": 3410
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 1.1368972180091832,
|
| 2400 |
+
"grad_norm": 0.24435286223888397,
|
| 2401 |
+
"learning_rate": 5.684840425531915e-06,
|
| 2402 |
+
"loss": 1.1361,
|
| 2403 |
+
"step": 3420
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 1.1402214788805551,
|
| 2407 |
+
"grad_norm": 0.260768860578537,
|
| 2408 |
+
"learning_rate": 5.701462765957447e-06,
|
| 2409 |
+
"loss": 1.149,
|
| 2410 |
+
"step": 3430
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 1.143545739751927,
|
| 2414 |
+
"grad_norm": 0.19861842691898346,
|
| 2415 |
+
"learning_rate": 5.718085106382979e-06,
|
| 2416 |
+
"loss": 1.1449,
|
| 2417 |
+
"step": 3440
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 1.146870000623299,
|
| 2421 |
+
"grad_norm": 0.3177703022956848,
|
| 2422 |
+
"learning_rate": 5.73470744680851e-06,
|
| 2423 |
+
"loss": 1.1437,
|
| 2424 |
+
"step": 3450
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 1.1501942614946707,
|
| 2428 |
+
"grad_norm": 0.258980929851532,
|
| 2429 |
+
"learning_rate": 5.751329787234043e-06,
|
| 2430 |
+
"loss": 1.1695,
|
| 2431 |
+
"step": 3460
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 1.1535185223660427,
|
| 2435 |
+
"grad_norm": 0.27876991033554077,
|
| 2436 |
+
"learning_rate": 5.7679521276595746e-06,
|
| 2437 |
+
"loss": 1.151,
|
| 2438 |
+
"step": 3470
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 1.1568427832374146,
|
| 2442 |
+
"grad_norm": 0.26816457509994507,
|
| 2443 |
+
"learning_rate": 5.784574468085106e-06,
|
| 2444 |
+
"loss": 1.1378,
|
| 2445 |
+
"step": 3480
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 1.1601670441087863,
|
| 2449 |
+
"grad_norm": 0.26024243235588074,
|
| 2450 |
+
"learning_rate": 5.801196808510639e-06,
|
| 2451 |
+
"loss": 1.1502,
|
| 2452 |
+
"step": 3490
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 1.1634913049801583,
|
| 2456 |
+
"grad_norm": 0.27794960141181946,
|
| 2457 |
+
"learning_rate": 5.8178191489361705e-06,
|
| 2458 |
+
"loss": 1.12,
|
| 2459 |
+
"step": 3500
|
| 2460 |
+
}
|
| 2461 |
+
],
|
| 2462 |
+
"logging_steps": 10,
|
| 2463 |
+
"max_steps": 150400,
|
| 2464 |
+
"num_input_tokens_seen": 0,
|
| 2465 |
+
"num_train_epochs": 50,
|
| 2466 |
+
"save_steps": 500,
|
| 2467 |
+
"stateful_callbacks": {
|
| 2468 |
+
"TrainerControl": {
|
| 2469 |
+
"args": {
|
| 2470 |
+
"should_epoch_stop": false,
|
| 2471 |
+
"should_evaluate": false,
|
| 2472 |
+
"should_log": false,
|
| 2473 |
+
"should_save": true,
|
| 2474 |
+
"should_training_stop": false
|
| 2475 |
+
},
|
| 2476 |
+
"attributes": {}
|
| 2477 |
+
}
|
| 2478 |
+
},
|
| 2479 |
+
"total_flos": 2.02023714816e+18,
|
| 2480 |
+
"train_batch_size": 5,
|
| 2481 |
+
"trial_name": null,
|
| 2482 |
+
"trial_params": null
|
| 2483 |
+
}
|