Upload trainer_state.json
Browse files- trainer_state.json +1401 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1401 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 3.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1941,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.015455950540958269,
|
| 14 |
+
"grad_norm": 9.875,
|
| 15 |
+
"learning_rate": 7.627118644067798e-06,
|
| 16 |
+
"loss": 3.7349,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.030911901081916538,
|
| 21 |
+
"grad_norm": 8.625,
|
| 22 |
+
"learning_rate": 1.6101694915254237e-05,
|
| 23 |
+
"loss": 3.5453,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.04636785162287481,
|
| 28 |
+
"grad_norm": 7.40625,
|
| 29 |
+
"learning_rate": 2.457627118644068e-05,
|
| 30 |
+
"loss": 3.441,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.061823802163833076,
|
| 35 |
+
"grad_norm": 5.71875,
|
| 36 |
+
"learning_rate": 3.305084745762712e-05,
|
| 37 |
+
"loss": 3.4128,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.07727975270479134,
|
| 42 |
+
"grad_norm": 6.21875,
|
| 43 |
+
"learning_rate": 4.152542372881356e-05,
|
| 44 |
+
"loss": 3.1485,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.09273570324574962,
|
| 49 |
+
"grad_norm": 5.0,
|
| 50 |
+
"learning_rate": 5e-05,
|
| 51 |
+
"loss": 3.0359,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.10819165378670788,
|
| 56 |
+
"grad_norm": 4.1875,
|
| 57 |
+
"learning_rate": 4.999651694417285e-05,
|
| 58 |
+
"loss": 3.2478,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.12364760432766615,
|
| 63 |
+
"grad_norm": 3.5,
|
| 64 |
+
"learning_rate": 4.9986068747225644e-05,
|
| 65 |
+
"loss": 3.1223,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.1391035548686244,
|
| 70 |
+
"grad_norm": 3.890625,
|
| 71 |
+
"learning_rate": 4.9968658320490636e-05,
|
| 72 |
+
"loss": 3.0477,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.1545595054095827,
|
| 77 |
+
"grad_norm": 3.859375,
|
| 78 |
+
"learning_rate": 4.994429051528689e-05,
|
| 79 |
+
"loss": 3.0829,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.17001545595054096,
|
| 84 |
+
"grad_norm": 3.734375,
|
| 85 |
+
"learning_rate": 4.991297212156848e-05,
|
| 86 |
+
"loss": 2.9565,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.18547140649149924,
|
| 91 |
+
"grad_norm": 3.671875,
|
| 92 |
+
"learning_rate": 4.9874711866032495e-05,
|
| 93 |
+
"loss": 3.142,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.2009273570324575,
|
| 98 |
+
"grad_norm": 5.15625,
|
| 99 |
+
"learning_rate": 4.982952040968744e-05,
|
| 100 |
+
"loss": 3.0807,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.21638330757341576,
|
| 105 |
+
"grad_norm": 4.625,
|
| 106 |
+
"learning_rate": 4.977741034488251e-05,
|
| 107 |
+
"loss": 2.9314,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.23183925811437403,
|
| 112 |
+
"grad_norm": 4.125,
|
| 113 |
+
"learning_rate": 4.971839619179892e-05,
|
| 114 |
+
"loss": 3.1325,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.2472952086553323,
|
| 119 |
+
"grad_norm": 4.65625,
|
| 120 |
+
"learning_rate": 4.965249439440384e-05,
|
| 121 |
+
"loss": 2.8936,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.26275115919629055,
|
| 126 |
+
"grad_norm": 5.0625,
|
| 127 |
+
"learning_rate": 4.957972331586843e-05,
|
| 128 |
+
"loss": 2.9205,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.2782071097372488,
|
| 133 |
+
"grad_norm": 4.34375,
|
| 134 |
+
"learning_rate": 4.950010323345103e-05,
|
| 135 |
+
"loss": 2.9776,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.2936630602782071,
|
| 140 |
+
"grad_norm": 4.53125,
|
| 141 |
+
"learning_rate": 4.941365633284699e-05,
|
| 142 |
+
"loss": 2.9432,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.3091190108191654,
|
| 147 |
+
"grad_norm": 4.25,
|
| 148 |
+
"learning_rate": 4.932040670200677e-05,
|
| 149 |
+
"loss": 2.9593,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.32457496136012365,
|
| 154 |
+
"grad_norm": 5.03125,
|
| 155 |
+
"learning_rate": 4.9220380324424006e-05,
|
| 156 |
+
"loss": 2.9826,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.3400309119010819,
|
| 161 |
+
"grad_norm": 4.28125,
|
| 162 |
+
"learning_rate": 4.911360507189526e-05,
|
| 163 |
+
"loss": 2.8009,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.3554868624420402,
|
| 168 |
+
"grad_norm": 4.5625,
|
| 169 |
+
"learning_rate": 4.900011069675378e-05,
|
| 170 |
+
"loss": 2.7733,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.37094281298299847,
|
| 175 |
+
"grad_norm": 3.609375,
|
| 176 |
+
"learning_rate": 4.8879928823579136e-05,
|
| 177 |
+
"loss": 2.8954,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.38639876352395675,
|
| 182 |
+
"grad_norm": 3.1875,
|
| 183 |
+
"learning_rate": 4.875309294038523e-05,
|
| 184 |
+
"loss": 2.8972,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.401854714064915,
|
| 189 |
+
"grad_norm": 4.5,
|
| 190 |
+
"learning_rate": 4.8619638389289026e-05,
|
| 191 |
+
"loss": 2.9377,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.41731066460587324,
|
| 196 |
+
"grad_norm": 3.796875,
|
| 197 |
+
"learning_rate": 4.8479602356662665e-05,
|
| 198 |
+
"loss": 2.8918,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.4327666151468315,
|
| 203 |
+
"grad_norm": 5.0625,
|
| 204 |
+
"learning_rate": 4.833302386277171e-05,
|
| 205 |
+
"loss": 2.8245,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.4482225656877898,
|
| 210 |
+
"grad_norm": 3.03125,
|
| 211 |
+
"learning_rate": 4.817994375090233e-05,
|
| 212 |
+
"loss": 2.8384,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.46367851622874806,
|
| 217 |
+
"grad_norm": 3.390625,
|
| 218 |
+
"learning_rate": 4.80204046759806e-05,
|
| 219 |
+
"loss": 2.9315,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.47913446676970634,
|
| 224 |
+
"grad_norm": 3.75,
|
| 225 |
+
"learning_rate": 4.785445109268687e-05,
|
| 226 |
+
"loss": 2.8416,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.4945904173106646,
|
| 231 |
+
"grad_norm": 3.734375,
|
| 232 |
+
"learning_rate": 4.768212924306877e-05,
|
| 233 |
+
"loss": 2.8615,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.5100463678516228,
|
| 238 |
+
"grad_norm": 3.28125,
|
| 239 |
+
"learning_rate": 4.75034871436561e-05,
|
| 240 |
+
"loss": 2.8885,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.5255023183925811,
|
| 245 |
+
"grad_norm": 3.96875,
|
| 246 |
+
"learning_rate": 4.7318574572081275e-05,
|
| 247 |
+
"loss": 2.8238,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.5409582689335394,
|
| 252 |
+
"grad_norm": 4.53125,
|
| 253 |
+
"learning_rate": 4.712744305320911e-05,
|
| 254 |
+
"loss": 2.816,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.5564142194744977,
|
| 259 |
+
"grad_norm": 2.9375,
|
| 260 |
+
"learning_rate": 4.6930145844779626e-05,
|
| 261 |
+
"loss": 2.6657,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.5718701700154559,
|
| 266 |
+
"grad_norm": 3.78125,
|
| 267 |
+
"learning_rate": 4.672673792256816e-05,
|
| 268 |
+
"loss": 2.8454,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.5873261205564142,
|
| 273 |
+
"grad_norm": 3.375,
|
| 274 |
+
"learning_rate": 4.651727596506661e-05,
|
| 275 |
+
"loss": 2.8263,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.6027820710973725,
|
| 280 |
+
"grad_norm": 4.03125,
|
| 281 |
+
"learning_rate": 4.63018183376903e-05,
|
| 282 |
+
"loss": 2.9417,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.6182380216383307,
|
| 287 |
+
"grad_norm": 3.453125,
|
| 288 |
+
"learning_rate": 4.60804250765148e-05,
|
| 289 |
+
"loss": 2.8449,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.633693972179289,
|
| 294 |
+
"grad_norm": 3.703125,
|
| 295 |
+
"learning_rate": 4.5853157871547184e-05,
|
| 296 |
+
"loss": 2.8171,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.6491499227202473,
|
| 301 |
+
"grad_norm": 3.25,
|
| 302 |
+
"learning_rate": 4.562008004953644e-05,
|
| 303 |
+
"loss": 2.9253,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.6646058732612056,
|
| 308 |
+
"grad_norm": 4.5,
|
| 309 |
+
"learning_rate": 4.538125655632789e-05,
|
| 310 |
+
"loss": 2.8203,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.6800618238021638,
|
| 315 |
+
"grad_norm": 3.71875,
|
| 316 |
+
"learning_rate": 4.5136753938766286e-05,
|
| 317 |
+
"loss": 2.8078,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.6955177743431221,
|
| 322 |
+
"grad_norm": 3.265625,
|
| 323 |
+
"learning_rate": 4.4886640326152985e-05,
|
| 324 |
+
"loss": 2.7962,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.7109737248840804,
|
| 329 |
+
"grad_norm": 4.1875,
|
| 330 |
+
"learning_rate": 4.4630985411262054e-05,
|
| 331 |
+
"loss": 2.7565,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.7264296754250387,
|
| 336 |
+
"grad_norm": 4.0,
|
| 337 |
+
"learning_rate": 4.4369860430920776e-05,
|
| 338 |
+
"loss": 2.7677,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.7418856259659969,
|
| 343 |
+
"grad_norm": 3.0,
|
| 344 |
+
"learning_rate": 4.41033381461599e-05,
|
| 345 |
+
"loss": 2.7879,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.7573415765069552,
|
| 350 |
+
"grad_norm": 4.71875,
|
| 351 |
+
"learning_rate": 4.383149282193919e-05,
|
| 352 |
+
"loss": 2.6332,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.7727975270479135,
|
| 357 |
+
"grad_norm": 4.96875,
|
| 358 |
+
"learning_rate": 4.35544002064539e-05,
|
| 359 |
+
"loss": 2.7762,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.7882534775888718,
|
| 364 |
+
"grad_norm": 4.375,
|
| 365 |
+
"learning_rate": 4.327213751002794e-05,
|
| 366 |
+
"loss": 2.8478,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.80370942812983,
|
| 371 |
+
"grad_norm": 4.0625,
|
| 372 |
+
"learning_rate": 4.298478338359968e-05,
|
| 373 |
+
"loss": 2.7518,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.8191653786707882,
|
| 378 |
+
"grad_norm": 4.40625,
|
| 379 |
+
"learning_rate": 4.269241789680629e-05,
|
| 380 |
+
"loss": 2.747,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.8346213292117465,
|
| 385 |
+
"grad_norm": 3.4375,
|
| 386 |
+
"learning_rate": 4.239512251567275e-05,
|
| 387 |
+
"loss": 2.7868,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.8500772797527048,
|
| 392 |
+
"grad_norm": 3.859375,
|
| 393 |
+
"learning_rate": 4.2092980079911836e-05,
|
| 394 |
+
"loss": 2.694,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.865533230293663,
|
| 399 |
+
"grad_norm": 5.34375,
|
| 400 |
+
"learning_rate": 4.178607477984128e-05,
|
| 401 |
+
"loss": 2.693,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.8809891808346213,
|
| 406 |
+
"grad_norm": 6.53125,
|
| 407 |
+
"learning_rate": 4.147449213292457e-05,
|
| 408 |
+
"loss": 2.7596,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.8964451313755796,
|
| 413 |
+
"grad_norm": 3.875,
|
| 414 |
+
"learning_rate": 4.115831895994203e-05,
|
| 415 |
+
"loss": 2.7299,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.9119010819165378,
|
| 420 |
+
"grad_norm": 3.796875,
|
| 421 |
+
"learning_rate": 4.0837643360798666e-05,
|
| 422 |
+
"loss": 2.8138,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.9273570324574961,
|
| 427 |
+
"grad_norm": 4.21875,
|
| 428 |
+
"learning_rate": 4.051255468997562e-05,
|
| 429 |
+
"loss": 2.7069,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.9428129829984544,
|
| 434 |
+
"grad_norm": 4.125,
|
| 435 |
+
"learning_rate": 4.018314353163202e-05,
|
| 436 |
+
"loss": 2.6954,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.9582689335394127,
|
| 441 |
+
"grad_norm": 3.78125,
|
| 442 |
+
"learning_rate": 3.984950167436423e-05,
|
| 443 |
+
"loss": 2.5214,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.973724884080371,
|
| 448 |
+
"grad_norm": 3.859375,
|
| 449 |
+
"learning_rate": 3.9511722085629485e-05,
|
| 450 |
+
"loss": 2.7825,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.9891808346213292,
|
| 455 |
+
"grad_norm": 3.234375,
|
| 456 |
+
"learning_rate": 3.916989888584095e-05,
|
| 457 |
+
"loss": 2.7346,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.0046367851622875,
|
| 462 |
+
"grad_norm": 3.9375,
|
| 463 |
+
"learning_rate": 3.882412732214167e-05,
|
| 464 |
+
"loss": 2.5301,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.0200927357032457,
|
| 469 |
+
"grad_norm": 3.65625,
|
| 470 |
+
"learning_rate": 3.847450374186441e-05,
|
| 471 |
+
"loss": 2.1356,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.035548686244204,
|
| 476 |
+
"grad_norm": 3.78125,
|
| 477 |
+
"learning_rate": 3.812112556568508e-05,
|
| 478 |
+
"loss": 2.0337,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.0510046367851622,
|
| 483 |
+
"grad_norm": 4.96875,
|
| 484 |
+
"learning_rate": 3.776409126047692e-05,
|
| 485 |
+
"loss": 2.1394,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.0664605873261206,
|
| 490 |
+
"grad_norm": 5.15625,
|
| 491 |
+
"learning_rate": 3.740350031187332e-05,
|
| 492 |
+
"loss": 1.9865,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.0819165378670788,
|
| 497 |
+
"grad_norm": 3.671875,
|
| 498 |
+
"learning_rate": 3.7039453196546645e-05,
|
| 499 |
+
"loss": 2.0067,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.0973724884080371,
|
| 504 |
+
"grad_norm": 3.65625,
|
| 505 |
+
"learning_rate": 3.6672051354211035e-05,
|
| 506 |
+
"loss": 2.2291,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.1128284389489953,
|
| 511 |
+
"grad_norm": 3.421875,
|
| 512 |
+
"learning_rate": 3.6301397159356696e-05,
|
| 513 |
+
"loss": 2.0544,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.1282843894899537,
|
| 518 |
+
"grad_norm": 4.3125,
|
| 519 |
+
"learning_rate": 3.592759389272389e-05,
|
| 520 |
+
"loss": 1.9978,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.1437403400309119,
|
| 525 |
+
"grad_norm": 4.1875,
|
| 526 |
+
"learning_rate": 3.555074571252431e-05,
|
| 527 |
+
"loss": 1.9575,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.1591962905718702,
|
| 532 |
+
"grad_norm": 4.34375,
|
| 533 |
+
"learning_rate": 3.517095762541795e-05,
|
| 534 |
+
"loss": 2.0236,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.1746522411128284,
|
| 539 |
+
"grad_norm": 4.34375,
|
| 540 |
+
"learning_rate": 3.4788335457253595e-05,
|
| 541 |
+
"loss": 1.9811,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.1901081916537868,
|
| 546 |
+
"grad_norm": 5.25,
|
| 547 |
+
"learning_rate": 3.4402985823581046e-05,
|
| 548 |
+
"loss": 2.0005,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.205564142194745,
|
| 553 |
+
"grad_norm": 4.875,
|
| 554 |
+
"learning_rate": 3.401501609994326e-05,
|
| 555 |
+
"loss": 1.9412,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.2210200927357033,
|
| 560 |
+
"grad_norm": 4.5625,
|
| 561 |
+
"learning_rate": 3.362453439195679e-05,
|
| 562 |
+
"loss": 2.0814,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.2364760432766615,
|
| 567 |
+
"grad_norm": 4.03125,
|
| 568 |
+
"learning_rate": 3.323164950518868e-05,
|
| 569 |
+
"loss": 1.9947,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.2519319938176197,
|
| 574 |
+
"grad_norm": 5.71875,
|
| 575 |
+
"learning_rate": 3.283647091483849e-05,
|
| 576 |
+
"loss": 1.9282,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.267387944358578,
|
| 581 |
+
"grad_norm": 4.40625,
|
| 582 |
+
"learning_rate": 3.243910873523356e-05,
|
| 583 |
+
"loss": 1.9283,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.2828438948995364,
|
| 588 |
+
"grad_norm": 6.84375,
|
| 589 |
+
"learning_rate": 3.203967368914631e-05,
|
| 590 |
+
"loss": 1.9607,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.2982998454404946,
|
| 595 |
+
"grad_norm": 6.15625,
|
| 596 |
+
"learning_rate": 3.163827707694193e-05,
|
| 597 |
+
"loss": 2.0587,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.3137557959814528,
|
| 602 |
+
"grad_norm": 4.84375,
|
| 603 |
+
"learning_rate": 3.123503074556513e-05,
|
| 604 |
+
"loss": 2.0544,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.3292117465224111,
|
| 609 |
+
"grad_norm": 4.40625,
|
| 610 |
+
"learning_rate": 3.083004705737468e-05,
|
| 611 |
+
"loss": 2.02,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.3446676970633695,
|
| 616 |
+
"grad_norm": 4.4375,
|
| 617 |
+
"learning_rate": 3.0423438858834174e-05,
|
| 618 |
+
"loss": 2.0169,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.3601236476043277,
|
| 623 |
+
"grad_norm": 5.15625,
|
| 624 |
+
"learning_rate": 3.0015319449068025e-05,
|
| 625 |
+
"loss": 2.0551,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.3755795981452859,
|
| 630 |
+
"grad_norm": 5.40625,
|
| 631 |
+
"learning_rate": 2.960580254829131e-05,
|
| 632 |
+
"loss": 1.8902,
|
| 633 |
+
"step": 890
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.3910355486862442,
|
| 637 |
+
"grad_norm": 5.28125,
|
| 638 |
+
"learning_rate": 2.919500226612224e-05,
|
| 639 |
+
"loss": 2.1326,
|
| 640 |
+
"step": 900
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.4064914992272024,
|
| 644 |
+
"grad_norm": 5.5625,
|
| 645 |
+
"learning_rate": 2.8783033069786132e-05,
|
| 646 |
+
"loss": 1.7767,
|
| 647 |
+
"step": 910
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.4219474497681608,
|
| 651 |
+
"grad_norm": 4.40625,
|
| 652 |
+
"learning_rate": 2.8370009752219788e-05,
|
| 653 |
+
"loss": 1.9215,
|
| 654 |
+
"step": 920
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.437403400309119,
|
| 658 |
+
"grad_norm": 5.78125,
|
| 659 |
+
"learning_rate": 2.795604740008504e-05,
|
| 660 |
+
"loss": 1.9508,
|
| 661 |
+
"step": 930
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.4528593508500773,
|
| 665 |
+
"grad_norm": 4.59375,
|
| 666 |
+
"learning_rate": 2.7541261361700514e-05,
|
| 667 |
+
"loss": 2.0011,
|
| 668 |
+
"step": 940
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.4683153013910355,
|
| 672 |
+
"grad_norm": 4.34375,
|
| 673 |
+
"learning_rate": 2.7125767214900455e-05,
|
| 674 |
+
"loss": 2.0049,
|
| 675 |
+
"step": 950
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.4837712519319939,
|
| 679 |
+
"grad_norm": 4.875,
|
| 680 |
+
"learning_rate": 2.67096807348296e-05,
|
| 681 |
+
"loss": 2.0328,
|
| 682 |
+
"step": 960
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.499227202472952,
|
| 686 |
+
"grad_norm": 5.375,
|
| 687 |
+
"learning_rate": 2.629311786168306e-05,
|
| 688 |
+
"loss": 2.0937,
|
| 689 |
+
"step": 970
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.5146831530139102,
|
| 693 |
+
"grad_norm": 3.984375,
|
| 694 |
+
"learning_rate": 2.5876194668400256e-05,
|
| 695 |
+
"loss": 2.0844,
|
| 696 |
+
"step": 980
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.5301391035548686,
|
| 700 |
+
"grad_norm": 4.46875,
|
| 701 |
+
"learning_rate": 2.545902732832181e-05,
|
| 702 |
+
"loss": 1.9383,
|
| 703 |
+
"step": 990
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.545595054095827,
|
| 707 |
+
"grad_norm": 4.90625,
|
| 708 |
+
"learning_rate": 2.5041732082818503e-05,
|
| 709 |
+
"loss": 1.963,
|
| 710 |
+
"step": 1000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.5610510046367851,
|
| 714 |
+
"grad_norm": 4.09375,
|
| 715 |
+
"learning_rate": 2.462442520890127e-05,
|
| 716 |
+
"loss": 2.0792,
|
| 717 |
+
"step": 1010
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.5765069551777433,
|
| 721 |
+
"grad_norm": 5.09375,
|
| 722 |
+
"learning_rate": 2.4207222986821194e-05,
|
| 723 |
+
"loss": 2.0617,
|
| 724 |
+
"step": 1020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.5919629057187017,
|
| 728 |
+
"grad_norm": 5.3125,
|
| 729 |
+
"learning_rate": 2.379024166766876e-05,
|
| 730 |
+
"loss": 1.9914,
|
| 731 |
+
"step": 1030
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.60741885625966,
|
| 735 |
+
"grad_norm": 5.0,
|
| 736 |
+
"learning_rate": 2.3373597440981027e-05,
|
| 737 |
+
"loss": 1.9685,
|
| 738 |
+
"step": 1040
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.6228748068006182,
|
| 742 |
+
"grad_norm": 5.96875,
|
| 743 |
+
"learning_rate": 2.295740640236614e-05,
|
| 744 |
+
"loss": 1.9599,
|
| 745 |
+
"step": 1050
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.6383307573415764,
|
| 749 |
+
"grad_norm": 6.4375,
|
| 750 |
+
"learning_rate": 2.2541784521153875e-05,
|
| 751 |
+
"loss": 2.0456,
|
| 752 |
+
"step": 1060
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.6537867078825348,
|
| 756 |
+
"grad_norm": 5.625,
|
| 757 |
+
"learning_rate": 2.2126847608081434e-05,
|
| 758 |
+
"loss": 1.9414,
|
| 759 |
+
"step": 1070
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 1.6692426584234932,
|
| 763 |
+
"grad_norm": 5.25,
|
| 764 |
+
"learning_rate": 2.1712711283023482e-05,
|
| 765 |
+
"loss": 2.0061,
|
| 766 |
+
"step": 1080
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 1.6846986089644513,
|
| 770 |
+
"grad_norm": 4.96875,
|
| 771 |
+
"learning_rate": 2.1299490942775213e-05,
|
| 772 |
+
"loss": 2.0977,
|
| 773 |
+
"step": 1090
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.7001545595054095,
|
| 777 |
+
"grad_norm": 5.125,
|
| 778 |
+
"learning_rate": 2.088730172889777e-05,
|
| 779 |
+
"loss": 2.0847,
|
| 780 |
+
"step": 1100
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.7156105100463679,
|
| 784 |
+
"grad_norm": 5.65625,
|
| 785 |
+
"learning_rate": 2.0476258495634594e-05,
|
| 786 |
+
"loss": 2.1084,
|
| 787 |
+
"step": 1110
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 1.7310664605873263,
|
| 791 |
+
"grad_norm": 4.125,
|
| 792 |
+
"learning_rate": 2.0066475777908005e-05,
|
| 793 |
+
"loss": 1.9382,
|
| 794 |
+
"step": 1120
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.7465224111282844,
|
| 798 |
+
"grad_norm": 6.40625,
|
| 799 |
+
"learning_rate": 1.9658067759404625e-05,
|
| 800 |
+
"loss": 2.0278,
|
| 801 |
+
"step": 1130
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 1.7619783616692426,
|
| 805 |
+
"grad_norm": 5.5625,
|
| 806 |
+
"learning_rate": 1.925114824075876e-05,
|
| 807 |
+
"loss": 2.0224,
|
| 808 |
+
"step": 1140
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 1.7774343122102008,
|
| 812 |
+
"grad_norm": 4.8125,
|
| 813 |
+
"learning_rate": 1.8845830607842438e-05,
|
| 814 |
+
"loss": 2.0613,
|
| 815 |
+
"step": 1150
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 1.7928902627511591,
|
| 819 |
+
"grad_norm": 5.375,
|
| 820 |
+
"learning_rate": 1.8442227800171125e-05,
|
| 821 |
+
"loss": 1.9505,
|
| 822 |
+
"step": 1160
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 1.8083462132921175,
|
| 826 |
+
"grad_norm": 5.46875,
|
| 827 |
+
"learning_rate": 1.8040452279433716e-05,
|
| 828 |
+
"loss": 1.9729,
|
| 829 |
+
"step": 1170
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 1.8238021638330757,
|
| 833 |
+
"grad_norm": 4.59375,
|
| 834 |
+
"learning_rate": 1.7640615998155693e-05,
|
| 835 |
+
"loss": 1.9394,
|
| 836 |
+
"step": 1180
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.8392581143740339,
|
| 840 |
+
"grad_norm": 4.875,
|
| 841 |
+
"learning_rate": 1.7242830368504227e-05,
|
| 842 |
+
"loss": 1.8939,
|
| 843 |
+
"step": 1190
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 1.8547140649149922,
|
| 847 |
+
"grad_norm": 3.984375,
|
| 848 |
+
"learning_rate": 1.6847206231243718e-05,
|
| 849 |
+
"loss": 1.8111,
|
| 850 |
+
"step": 1200
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 1.8701700154559506,
|
| 854 |
+
"grad_norm": 4.9375,
|
| 855 |
+
"learning_rate": 1.6453853824850726e-05,
|
| 856 |
+
"loss": 1.9524,
|
| 857 |
+
"step": 1210
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 1.8856259659969088,
|
| 861 |
+
"grad_norm": 3.953125,
|
| 862 |
+
"learning_rate": 1.606288275479652e-05,
|
| 863 |
+
"loss": 2.0018,
|
| 864 |
+
"step": 1220
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 1.901081916537867,
|
| 868 |
+
"grad_norm": 5.15625,
|
| 869 |
+
"learning_rate": 1.5674401963006235e-05,
|
| 870 |
+
"loss": 2.0096,
|
| 871 |
+
"step": 1230
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 1.9165378670788253,
|
| 875 |
+
"grad_norm": 5.5625,
|
| 876 |
+
"learning_rate": 1.5288519697502697e-05,
|
| 877 |
+
"loss": 1.9857,
|
| 878 |
+
"step": 1240
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.9319938176197837,
|
| 882 |
+
"grad_norm": 6.625,
|
| 883 |
+
"learning_rate": 1.4905343482243775e-05,
|
| 884 |
+
"loss": 1.9265,
|
| 885 |
+
"step": 1250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 1.947449768160742,
|
| 889 |
+
"grad_norm": 4.875,
|
| 890 |
+
"learning_rate": 1.4524980087161438e-05,
|
| 891 |
+
"loss": 2.0875,
|
| 892 |
+
"step": 1260
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 1.9629057187017,
|
| 896 |
+
"grad_norm": 3.484375,
|
| 897 |
+
"learning_rate": 1.4147535498410853e-05,
|
| 898 |
+
"loss": 2.0144,
|
| 899 |
+
"step": 1270
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 1.9783616692426584,
|
| 903 |
+
"grad_norm": 4.25,
|
| 904 |
+
"learning_rate": 1.3773114888837957e-05,
|
| 905 |
+
"loss": 1.9803,
|
| 906 |
+
"step": 1280
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 1.9938176197836168,
|
| 910 |
+
"grad_norm": 4.90625,
|
| 911 |
+
"learning_rate": 1.3401822588673636e-05,
|
| 912 |
+
"loss": 2.0252,
|
| 913 |
+
"step": 1290
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 2.009273570324575,
|
| 917 |
+
"grad_norm": 4.4375,
|
| 918 |
+
"learning_rate": 1.3033762056462654e-05,
|
| 919 |
+
"loss": 1.7441,
|
| 920 |
+
"step": 1300
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 2.024729520865533,
|
| 924 |
+
"grad_norm": 3.84375,
|
| 925 |
+
"learning_rate": 1.2669035850235536e-05,
|
| 926 |
+
"loss": 1.58,
|
| 927 |
+
"step": 1310
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 2.0401854714064913,
|
| 931 |
+
"grad_norm": 4.6875,
|
| 932 |
+
"learning_rate": 1.230774559893131e-05,
|
| 933 |
+
"loss": 1.5253,
|
| 934 |
+
"step": 1320
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 2.05564142194745,
|
| 938 |
+
"grad_norm": 5.46875,
|
| 939 |
+
"learning_rate": 1.1949991974079197e-05,
|
| 940 |
+
"loss": 1.5205,
|
| 941 |
+
"step": 1330
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 2.071097372488408,
|
| 945 |
+
"grad_norm": 5.1875,
|
| 946 |
+
"learning_rate": 1.1595874661746986e-05,
|
| 947 |
+
"loss": 1.6223,
|
| 948 |
+
"step": 1340
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 2.0865533230293662,
|
| 952 |
+
"grad_norm": 4.90625,
|
| 953 |
+
"learning_rate": 1.1245492334764169e-05,
|
| 954 |
+
"loss": 1.5583,
|
| 955 |
+
"step": 1350
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 2.1020092735703244,
|
| 959 |
+
"grad_norm": 4.90625,
|
| 960 |
+
"learning_rate": 1.0898942625227168e-05,
|
| 961 |
+
"loss": 1.5896,
|
| 962 |
+
"step": 1360
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 2.117465224111283,
|
| 966 |
+
"grad_norm": 6.78125,
|
| 967 |
+
"learning_rate": 1.0556322097294835e-05,
|
| 968 |
+
"loss": 1.6192,
|
| 969 |
+
"step": 1370
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 2.132921174652241,
|
| 973 |
+
"grad_norm": 5.0,
|
| 974 |
+
"learning_rate": 1.0217726220281243e-05,
|
| 975 |
+
"loss": 1.6102,
|
| 976 |
+
"step": 1380
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 2.1483771251931993,
|
| 980 |
+
"grad_norm": 5.8125,
|
| 981 |
+
"learning_rate": 9.883249342053821e-06,
|
| 982 |
+
"loss": 1.5518,
|
| 983 |
+
"step": 1390
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 2.1638330757341575,
|
| 987 |
+
"grad_norm": 5.6875,
|
| 988 |
+
"learning_rate": 9.552984662743716e-06,
|
| 989 |
+
"loss": 1.4824,
|
| 990 |
+
"step": 1400
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 2.179289026275116,
|
| 994 |
+
"grad_norm": 5.09375,
|
| 995 |
+
"learning_rate": 9.227024208776224e-06,
|
| 996 |
+
"loss": 1.4147,
|
| 997 |
+
"step": 1410
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 2.1947449768160743,
|
| 1001 |
+
"grad_norm": 5.5,
|
| 1002 |
+
"learning_rate": 8.905458807228013e-06,
|
| 1003 |
+
"loss": 1.5841,
|
| 1004 |
+
"step": 1420
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 2.2102009273570324,
|
| 1008 |
+
"grad_norm": 5.09375,
|
| 1009 |
+
"learning_rate": 8.588378060518738e-06,
|
| 1010 |
+
"loss": 1.6099,
|
| 1011 |
+
"step": 1430
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 2.2256568778979906,
|
| 1015 |
+
"grad_norm": 6.6875,
|
| 1016 |
+
"learning_rate": 8.275870321443802e-06,
|
| 1017 |
+
"loss": 1.5302,
|
| 1018 |
+
"step": 1440
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 2.2411128284389488,
|
| 1022 |
+
"grad_norm": 5.21875,
|
| 1023 |
+
"learning_rate": 7.968022668555334e-06,
|
| 1024 |
+
"loss": 1.5991,
|
| 1025 |
+
"step": 1450
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 2.2565687789799074,
|
| 1029 |
+
"grad_norm": 6.8125,
|
| 1030 |
+
"learning_rate": 7.664920881898236e-06,
|
| 1031 |
+
"loss": 1.6388,
|
| 1032 |
+
"step": 1460
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 2.2720247295208655,
|
| 1036 |
+
"grad_norm": 6.78125,
|
| 1037 |
+
"learning_rate": 7.3666494191080455e-06,
|
| 1038 |
+
"loss": 1.5273,
|
| 1039 |
+
"step": 1470
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 2.2874806800618237,
|
| 1043 |
+
"grad_norm": 5.5625,
|
| 1044 |
+
"learning_rate": 7.073291391877288e-06,
|
| 1045 |
+
"loss": 1.502,
|
| 1046 |
+
"step": 1480
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 2.3029366306027823,
|
| 1050 |
+
"grad_norm": 5.9375,
|
| 1051 |
+
"learning_rate": 6.7849285427968586e-06,
|
| 1052 |
+
"loss": 1.4714,
|
| 1053 |
+
"step": 1490
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 2.3183925811437405,
|
| 1057 |
+
"grad_norm": 6.4375,
|
| 1058 |
+
"learning_rate": 6.5016412225789e-06,
|
| 1059 |
+
"loss": 1.4667,
|
| 1060 |
+
"step": 1500
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 2.3338485316846986,
|
| 1064 |
+
"grad_norm": 6.625,
|
| 1065 |
+
"learning_rate": 6.223508367667538e-06,
|
| 1066 |
+
"loss": 1.604,
|
| 1067 |
+
"step": 1510
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 2.349304482225657,
|
| 1071 |
+
"grad_norm": 5.4375,
|
| 1072 |
+
"learning_rate": 5.9506074782436275e-06,
|
| 1073 |
+
"loss": 1.6265,
|
| 1074 |
+
"step": 1520
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 2.364760432766615,
|
| 1078 |
+
"grad_norm": 5.1875,
|
| 1079 |
+
"learning_rate": 5.683014596629857e-06,
|
| 1080 |
+
"loss": 1.4042,
|
| 1081 |
+
"step": 1530
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 2.3802163833075736,
|
| 1085 |
+
"grad_norm": 4.40625,
|
| 1086 |
+
"learning_rate": 5.420804286101838e-06,
|
| 1087 |
+
"loss": 1.6431,
|
| 1088 |
+
"step": 1540
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 2.3956723338485317,
|
| 1092 |
+
"grad_norm": 4.84375,
|
| 1093 |
+
"learning_rate": 5.164049610111607e-06,
|
| 1094 |
+
"loss": 1.6615,
|
| 1095 |
+
"step": 1550
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 2.41112828438949,
|
| 1099 |
+
"grad_norm": 4.15625,
|
| 1100 |
+
"learning_rate": 4.912822111928767e-06,
|
| 1101 |
+
"loss": 1.6007,
|
| 1102 |
+
"step": 1560
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 2.426584234930448,
|
| 1106 |
+
"grad_norm": 6.8125,
|
| 1107 |
+
"learning_rate": 4.667191794705444e-06,
|
| 1108 |
+
"loss": 1.5768,
|
| 1109 |
+
"step": 1570
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 2.4420401854714067,
|
| 1113 |
+
"grad_norm": 7.09375,
|
| 1114 |
+
"learning_rate": 4.427227101970261e-06,
|
| 1115 |
+
"loss": 1.6504,
|
| 1116 |
+
"step": 1580
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 2.457496136012365,
|
| 1120 |
+
"grad_norm": 4.125,
|
| 1121 |
+
"learning_rate": 4.192994898556921e-06,
|
| 1122 |
+
"loss": 1.5617,
|
| 1123 |
+
"step": 1590
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 2.472952086553323,
|
| 1127 |
+
"grad_norm": 4.9375,
|
| 1128 |
+
"learning_rate": 3.964560451972705e-06,
|
| 1129 |
+
"loss": 1.4937,
|
| 1130 |
+
"step": 1600
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 2.488408037094281,
|
| 1134 |
+
"grad_norm": 4.34375,
|
| 1135 |
+
"learning_rate": 3.7419874142120363e-06,
|
| 1136 |
+
"loss": 1.5398,
|
| 1137 |
+
"step": 1610
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 2.5038639876352393,
|
| 1141 |
+
"grad_norm": 7.125,
|
| 1142 |
+
"learning_rate": 3.525337804020212e-06,
|
| 1143 |
+
"loss": 1.6331,
|
| 1144 |
+
"step": 1620
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 2.519319938176198,
|
| 1148 |
+
"grad_norm": 4.65625,
|
| 1149 |
+
"learning_rate": 3.314671989612195e-06,
|
| 1150 |
+
"loss": 1.5092,
|
| 1151 |
+
"step": 1630
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 2.534775888717156,
|
| 1155 |
+
"grad_norm": 6.0,
|
| 1156 |
+
"learning_rate": 3.110048671851404e-06,
|
| 1157 |
+
"loss": 1.6244,
|
| 1158 |
+
"step": 1640
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 2.5502318392581143,
|
| 1162 |
+
"grad_norm": 5.875,
|
| 1163 |
+
"learning_rate": 2.911524867892956e-06,
|
| 1164 |
+
"loss": 1.5014,
|
| 1165 |
+
"step": 1650
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 2.565687789799073,
|
| 1169 |
+
"grad_norm": 5.71875,
|
| 1170 |
+
"learning_rate": 2.719155895296255e-06,
|
| 1171 |
+
"loss": 1.5084,
|
| 1172 |
+
"step": 1660
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 2.581143740340031,
|
| 1176 |
+
"grad_norm": 5.40625,
|
| 1177 |
+
"learning_rate": 2.5329953566109537e-06,
|
| 1178 |
+
"loss": 1.6218,
|
| 1179 |
+
"step": 1670
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 2.596599690880989,
|
| 1183 |
+
"grad_norm": 6.28125,
|
| 1184 |
+
"learning_rate": 2.353095124440999e-06,
|
| 1185 |
+
"loss": 1.5937,
|
| 1186 |
+
"step": 1680
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 2.6120556414219473,
|
| 1190 |
+
"grad_norm": 5.78125,
|
| 1191 |
+
"learning_rate": 2.17950532699053e-06,
|
| 1192 |
+
"loss": 1.5456,
|
| 1193 |
+
"step": 1690
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 2.6275115919629055,
|
| 1197 |
+
"grad_norm": 4.15625,
|
| 1198 |
+
"learning_rate": 2.012274334095998e-06,
|
| 1199 |
+
"loss": 1.6362,
|
| 1200 |
+
"step": 1700
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 2.642967542503864,
|
| 1204 |
+
"grad_norm": 5.28125,
|
| 1205 |
+
"learning_rate": 1.8514487437481436e-06,
|
| 1206 |
+
"loss": 1.5103,
|
| 1207 |
+
"step": 1710
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 2.6584234930448223,
|
| 1211 |
+
"grad_norm": 5.53125,
|
| 1212 |
+
"learning_rate": 1.6970733691077378e-06,
|
| 1213 |
+
"loss": 1.5271,
|
| 1214 |
+
"step": 1720
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 2.6738794435857804,
|
| 1218 |
+
"grad_norm": 5.34375,
|
| 1219 |
+
"learning_rate": 1.549191226018637e-06,
|
| 1220 |
+
"loss": 1.5694,
|
| 1221 |
+
"step": 1730
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 2.689335394126739,
|
| 1225 |
+
"grad_norm": 5.21875,
|
| 1226 |
+
"learning_rate": 1.4078435210216568e-06,
|
| 1227 |
+
"loss": 1.564,
|
| 1228 |
+
"step": 1740
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 2.704791344667697,
|
| 1232 |
+
"grad_norm": 4.8125,
|
| 1233 |
+
"learning_rate": 1.273069639872601e-06,
|
| 1234 |
+
"loss": 1.6323,
|
| 1235 |
+
"step": 1750
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 2.7202472952086554,
|
| 1239 |
+
"grad_norm": 4.75,
|
| 1240 |
+
"learning_rate": 1.1449071365676356e-06,
|
| 1241 |
+
"loss": 1.6231,
|
| 1242 |
+
"step": 1760
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 2.7357032457496135,
|
| 1246 |
+
"grad_norm": 6.125,
|
| 1247 |
+
"learning_rate": 1.0233917228790828e-06,
|
| 1248 |
+
"loss": 1.5671,
|
| 1249 |
+
"step": 1770
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 2.7511591962905717,
|
| 1253 |
+
"grad_norm": 5.53125,
|
| 1254 |
+
"learning_rate": 9.08557258404516e-07,
|
| 1255 |
+
"loss": 1.5081,
|
| 1256 |
+
"step": 1780
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 2.76661514683153,
|
| 1260 |
+
"grad_norm": 4.9375,
|
| 1261 |
+
"learning_rate": 8.004357411319857e-07,
|
| 1262 |
+
"loss": 1.5764,
|
| 1263 |
+
"step": 1790
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 2.7820710973724885,
|
| 1267 |
+
"grad_norm": 6.46875,
|
| 1268 |
+
"learning_rate": 6.990572985239612e-07,
|
| 1269 |
+
"loss": 1.5529,
|
| 1270 |
+
"step": 1800
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 2.7975270479134466,
|
| 1274 |
+
"grad_norm": 5.6875,
|
| 1275 |
+
"learning_rate": 6.044501791224539e-07,
|
| 1276 |
+
"loss": 1.5893,
|
| 1277 |
+
"step": 1810
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 2.812982998454405,
|
| 1281 |
+
"grad_norm": 7.125,
|
| 1282 |
+
"learning_rate": 5.166407446777572e-07,
|
| 1283 |
+
"loss": 1.6085,
|
| 1284 |
+
"step": 1820
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 2.8284389489953634,
|
| 1288 |
+
"grad_norm": 5.25,
|
| 1289 |
+
"learning_rate": 4.3565346280284305e-07,
|
| 1290 |
+
"loss": 1.4917,
|
| 1291 |
+
"step": 1830
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 2.8438948995363216,
|
| 1295 |
+
"grad_norm": 5.3125,
|
| 1296 |
+
"learning_rate": 3.6151090015565103e-07,
|
| 1297 |
+
"loss": 1.5961,
|
| 1298 |
+
"step": 1840
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 2.8593508500772797,
|
| 1302 |
+
"grad_norm": 5.0,
|
| 1303 |
+
"learning_rate": 2.9423371615096005e-07,
|
| 1304 |
+
"loss": 1.621,
|
| 1305 |
+
"step": 1850
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 2.874806800618238,
|
| 1309 |
+
"grad_norm": 4.375,
|
| 1310 |
+
"learning_rate": 2.3384065720379878e-07,
|
| 1311 |
+
"loss": 1.5848,
|
| 1312 |
+
"step": 1860
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 2.890262751159196,
|
| 1316 |
+
"grad_norm": 4.90625,
|
| 1317 |
+
"learning_rate": 1.803485515058323e-07,
|
| 1318 |
+
"loss": 1.47,
|
| 1319 |
+
"step": 1870
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 2.9057187017001547,
|
| 1323 |
+
"grad_norm": 7.21875,
|
| 1324 |
+
"learning_rate": 1.3377230433630205e-07,
|
| 1325 |
+
"loss": 1.5546,
|
| 1326 |
+
"step": 1880
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 2.921174652241113,
|
| 1330 |
+
"grad_norm": 5.6875,
|
| 1331 |
+
"learning_rate": 9.412489390873414e-08,
|
| 1332 |
+
"loss": 1.6426,
|
| 1333 |
+
"step": 1890
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 2.936630602782071,
|
| 1337 |
+
"grad_norm": 5.78125,
|
| 1338 |
+
"learning_rate": 6.141736775464313e-08,
|
| 1339 |
+
"loss": 1.5036,
|
| 1340 |
+
"step": 1900
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 2.9520865533230296,
|
| 1344 |
+
"grad_norm": 7.125,
|
| 1345 |
+
"learning_rate": 3.5658839645194565e-08,
|
| 1346 |
+
"loss": 1.5698,
|
| 1347 |
+
"step": 1910
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 2.9675425038639878,
|
| 1351 |
+
"grad_norm": 5.75,
|
| 1352 |
+
"learning_rate": 1.6856487051700177e-08,
|
| 1353 |
+
"loss": 1.5624,
|
| 1354 |
+
"step": 1920
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 2.982998454404946,
|
| 1358 |
+
"grad_norm": 5.6875,
|
| 1359 |
+
"learning_rate": 5.015549145664933e-09,
|
| 1360 |
+
"loss": 1.7195,
|
| 1361 |
+
"step": 1930
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 2.998454404945904,
|
| 1365 |
+
"grad_norm": 6.15625,
|
| 1366 |
+
"learning_rate": 1.393253389103677e-10,
|
| 1367 |
+
"loss": 1.6643,
|
| 1368 |
+
"step": 1940
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 3.0,
|
| 1372 |
+
"step": 1941,
|
| 1373 |
+
"total_flos": 7.627263225499238e+16,
|
| 1374 |
+
"train_loss": 2.1592409425884838,
|
| 1375 |
+
"train_runtime": 1942.4867,
|
| 1376 |
+
"train_samples_per_second": 15.985,
|
| 1377 |
+
"train_steps_per_second": 0.999
|
| 1378 |
+
}
|
| 1379 |
+
],
|
| 1380 |
+
"logging_steps": 10,
|
| 1381 |
+
"max_steps": 1941,
|
| 1382 |
+
"num_input_tokens_seen": 0,
|
| 1383 |
+
"num_train_epochs": 3,
|
| 1384 |
+
"save_steps": 500,
|
| 1385 |
+
"stateful_callbacks": {
|
| 1386 |
+
"TrainerControl": {
|
| 1387 |
+
"args": {
|
| 1388 |
+
"should_epoch_stop": false,
|
| 1389 |
+
"should_evaluate": false,
|
| 1390 |
+
"should_log": false,
|
| 1391 |
+
"should_save": true,
|
| 1392 |
+
"should_training_stop": true
|
| 1393 |
+
},
|
| 1394 |
+
"attributes": {}
|
| 1395 |
+
}
|
| 1396 |
+
},
|
| 1397 |
+
"total_flos": 7.627263225499238e+16,
|
| 1398 |
+
"train_batch_size": 4,
|
| 1399 |
+
"trial_name": null,
|
| 1400 |
+
"trial_params": null
|
| 1401 |
+
}
|