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": 0.015458937198067632,
|
| 14 |
+
"grad_norm": 11.5,
|
| 15 |
+
"learning_rate": 7.627118644067798e-06,
|
| 16 |
+
"loss": 3.6877,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.030917874396135265,
|
| 21 |
+
"grad_norm": 7.25,
|
| 22 |
+
"learning_rate": 1.6101694915254237e-05,
|
| 23 |
+
"loss": 3.3724,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.0463768115942029,
|
| 28 |
+
"grad_norm": 7.125,
|
| 29 |
+
"learning_rate": 2.457627118644068e-05,
|
| 30 |
+
"loss": 3.143,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.06183574879227053,
|
| 35 |
+
"grad_norm": 6.15625,
|
| 36 |
+
"learning_rate": 3.305084745762712e-05,
|
| 37 |
+
"loss": 3.2346,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.07729468599033816,
|
| 42 |
+
"grad_norm": 6.0,
|
| 43 |
+
"learning_rate": 4.152542372881356e-05,
|
| 44 |
+
"loss": 2.9242,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.0927536231884058,
|
| 49 |
+
"grad_norm": 5.5625,
|
| 50 |
+
"learning_rate": 5e-05,
|
| 51 |
+
"loss": 2.9235,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.10821256038647344,
|
| 56 |
+
"grad_norm": 4.71875,
|
| 57 |
+
"learning_rate": 4.999651694417285e-05,
|
| 58 |
+
"loss": 3.0094,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.12367149758454106,
|
| 63 |
+
"grad_norm": 4.0625,
|
| 64 |
+
"learning_rate": 4.9986068747225644e-05,
|
| 65 |
+
"loss": 3.0419,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.1391304347826087,
|
| 70 |
+
"grad_norm": 3.96875,
|
| 71 |
+
"learning_rate": 4.9968658320490636e-05,
|
| 72 |
+
"loss": 2.8573,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.15458937198067632,
|
| 77 |
+
"grad_norm": 3.40625,
|
| 78 |
+
"learning_rate": 4.994429051528689e-05,
|
| 79 |
+
"loss": 2.9456,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.17004830917874397,
|
| 84 |
+
"grad_norm": 3.5,
|
| 85 |
+
"learning_rate": 4.991297212156848e-05,
|
| 86 |
+
"loss": 2.9164,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.1855072463768116,
|
| 91 |
+
"grad_norm": 3.328125,
|
| 92 |
+
"learning_rate": 4.9874711866032495e-05,
|
| 93 |
+
"loss": 2.9057,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.20096618357487922,
|
| 98 |
+
"grad_norm": 5.25,
|
| 99 |
+
"learning_rate": 4.982952040968744e-05,
|
| 100 |
+
"loss": 2.9002,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.21642512077294687,
|
| 105 |
+
"grad_norm": 3.40625,
|
| 106 |
+
"learning_rate": 4.977741034488251e-05,
|
| 107 |
+
"loss": 2.6435,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.2318840579710145,
|
| 112 |
+
"grad_norm": 3.15625,
|
| 113 |
+
"learning_rate": 4.971839619179892e-05,
|
| 114 |
+
"loss": 2.9411,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.24734299516908212,
|
| 119 |
+
"grad_norm": 4.28125,
|
| 120 |
+
"learning_rate": 4.965249439440384e-05,
|
| 121 |
+
"loss": 2.6526,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.26280193236714977,
|
| 126 |
+
"grad_norm": 7.0,
|
| 127 |
+
"learning_rate": 4.957972331586843e-05,
|
| 128 |
+
"loss": 2.6599,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.2782608695652174,
|
| 133 |
+
"grad_norm": 3.296875,
|
| 134 |
+
"learning_rate": 4.950010323345103e-05,
|
| 135 |
+
"loss": 2.7398,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.293719806763285,
|
| 140 |
+
"grad_norm": 4.9375,
|
| 141 |
+
"learning_rate": 4.941365633284699e-05,
|
| 142 |
+
"loss": 2.7388,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.30917874396135264,
|
| 147 |
+
"grad_norm": 4.125,
|
| 148 |
+
"learning_rate": 4.932040670200677e-05,
|
| 149 |
+
"loss": 2.8344,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.32463768115942027,
|
| 154 |
+
"grad_norm": 6.0,
|
| 155 |
+
"learning_rate": 4.9220380324424006e-05,
|
| 156 |
+
"loss": 2.9072,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.34009661835748795,
|
| 161 |
+
"grad_norm": 4.40625,
|
| 162 |
+
"learning_rate": 4.911360507189526e-05,
|
| 163 |
+
"loss": 2.6469,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.35555555555555557,
|
| 168 |
+
"grad_norm": 4.28125,
|
| 169 |
+
"learning_rate": 4.900011069675378e-05,
|
| 170 |
+
"loss": 2.6556,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.3710144927536232,
|
| 175 |
+
"grad_norm": 2.796875,
|
| 176 |
+
"learning_rate": 4.8879928823579136e-05,
|
| 177 |
+
"loss": 2.5898,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.3864734299516908,
|
| 182 |
+
"grad_norm": 3.484375,
|
| 183 |
+
"learning_rate": 4.875309294038523e-05,
|
| 184 |
+
"loss": 2.6991,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.40193236714975844,
|
| 189 |
+
"grad_norm": 4.25,
|
| 190 |
+
"learning_rate": 4.8619638389289026e-05,
|
| 191 |
+
"loss": 2.667,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.41739130434782606,
|
| 196 |
+
"grad_norm": 3.34375,
|
| 197 |
+
"learning_rate": 4.8479602356662665e-05,
|
| 198 |
+
"loss": 2.7002,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.43285024154589374,
|
| 203 |
+
"grad_norm": 4.71875,
|
| 204 |
+
"learning_rate": 4.833302386277171e-05,
|
| 205 |
+
"loss": 2.7914,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.44830917874396137,
|
| 210 |
+
"grad_norm": 4.15625,
|
| 211 |
+
"learning_rate": 4.817994375090233e-05,
|
| 212 |
+
"loss": 2.6703,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.463768115942029,
|
| 217 |
+
"grad_norm": 3.140625,
|
| 218 |
+
"learning_rate": 4.80204046759806e-05,
|
| 219 |
+
"loss": 2.8089,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.4792270531400966,
|
| 224 |
+
"grad_norm": 4.90625,
|
| 225 |
+
"learning_rate": 4.785445109268687e-05,
|
| 226 |
+
"loss": 2.7172,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.49468599033816424,
|
| 231 |
+
"grad_norm": 2.921875,
|
| 232 |
+
"learning_rate": 4.768212924306877e-05,
|
| 233 |
+
"loss": 2.617,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.5101449275362319,
|
| 238 |
+
"grad_norm": 2.96875,
|
| 239 |
+
"learning_rate": 4.75034871436561e-05,
|
| 240 |
+
"loss": 2.7096,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.5256038647342995,
|
| 245 |
+
"grad_norm": 4.75,
|
| 246 |
+
"learning_rate": 4.7318574572081275e-05,
|
| 247 |
+
"loss": 2.6772,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.5410628019323671,
|
| 252 |
+
"grad_norm": 4.90625,
|
| 253 |
+
"learning_rate": 4.712744305320911e-05,
|
| 254 |
+
"loss": 2.7253,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.5565217391304348,
|
| 259 |
+
"grad_norm": 3.34375,
|
| 260 |
+
"learning_rate": 4.6930145844779626e-05,
|
| 261 |
+
"loss": 2.6389,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.5719806763285025,
|
| 266 |
+
"grad_norm": 2.84375,
|
| 267 |
+
"learning_rate": 4.672673792256816e-05,
|
| 268 |
+
"loss": 2.6831,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.58743961352657,
|
| 273 |
+
"grad_norm": 2.921875,
|
| 274 |
+
"learning_rate": 4.651727596506661e-05,
|
| 275 |
+
"loss": 2.7029,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.6028985507246377,
|
| 280 |
+
"grad_norm": 3.765625,
|
| 281 |
+
"learning_rate": 4.63018183376903e-05,
|
| 282 |
+
"loss": 2.7264,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.6183574879227053,
|
| 287 |
+
"grad_norm": 4.28125,
|
| 288 |
+
"learning_rate": 4.60804250765148e-05,
|
| 289 |
+
"loss": 2.6494,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.633816425120773,
|
| 294 |
+
"grad_norm": 3.0625,
|
| 295 |
+
"learning_rate": 4.5853157871547184e-05,
|
| 296 |
+
"loss": 2.6504,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.6492753623188405,
|
| 301 |
+
"grad_norm": 3.515625,
|
| 302 |
+
"learning_rate": 4.562008004953644e-05,
|
| 303 |
+
"loss": 2.817,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.6647342995169082,
|
| 308 |
+
"grad_norm": 4.0,
|
| 309 |
+
"learning_rate": 4.538125655632789e-05,
|
| 310 |
+
"loss": 2.5887,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.6801932367149759,
|
| 315 |
+
"grad_norm": 2.984375,
|
| 316 |
+
"learning_rate": 4.5136753938766286e-05,
|
| 317 |
+
"loss": 2.6033,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.6956521739130435,
|
| 322 |
+
"grad_norm": 3.8125,
|
| 323 |
+
"learning_rate": 4.4886640326152985e-05,
|
| 324 |
+
"loss": 2.6842,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.7111111111111111,
|
| 329 |
+
"grad_norm": 4.53125,
|
| 330 |
+
"learning_rate": 4.4630985411262054e-05,
|
| 331 |
+
"loss": 2.6597,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.7265700483091787,
|
| 336 |
+
"grad_norm": 3.9375,
|
| 337 |
+
"learning_rate": 4.4369860430920776e-05,
|
| 338 |
+
"loss": 2.5888,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.7420289855072464,
|
| 343 |
+
"grad_norm": 3.59375,
|
| 344 |
+
"learning_rate": 4.41033381461599e-05,
|
| 345 |
+
"loss": 2.6027,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.7574879227053141,
|
| 350 |
+
"grad_norm": 3.15625,
|
| 351 |
+
"learning_rate": 4.383149282193919e-05,
|
| 352 |
+
"loss": 2.4755,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.7729468599033816,
|
| 357 |
+
"grad_norm": 4.53125,
|
| 358 |
+
"learning_rate": 4.35544002064539e-05,
|
| 359 |
+
"loss": 2.5931,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.7884057971014493,
|
| 364 |
+
"grad_norm": 4.09375,
|
| 365 |
+
"learning_rate": 4.327213751002794e-05,
|
| 366 |
+
"loss": 2.6694,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.8038647342995169,
|
| 371 |
+
"grad_norm": 2.984375,
|
| 372 |
+
"learning_rate": 4.298478338359968e-05,
|
| 373 |
+
"loss": 2.6839,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.8193236714975846,
|
| 378 |
+
"grad_norm": 3.1875,
|
| 379 |
+
"learning_rate": 4.269241789680629e-05,
|
| 380 |
+
"loss": 2.4991,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.8347826086956521,
|
| 385 |
+
"grad_norm": 3.78125,
|
| 386 |
+
"learning_rate": 4.239512251567275e-05,
|
| 387 |
+
"loss": 2.5734,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.8502415458937198,
|
| 392 |
+
"grad_norm": 3.546875,
|
| 393 |
+
"learning_rate": 4.2092980079911836e-05,
|
| 394 |
+
"loss": 2.5269,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.8657004830917875,
|
| 399 |
+
"grad_norm": 4.25,
|
| 400 |
+
"learning_rate": 4.178607477984128e-05,
|
| 401 |
+
"loss": 2.4704,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.881159420289855,
|
| 406 |
+
"grad_norm": 4.59375,
|
| 407 |
+
"learning_rate": 4.147449213292457e-05,
|
| 408 |
+
"loss": 2.6069,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.8966183574879227,
|
| 413 |
+
"grad_norm": 4.03125,
|
| 414 |
+
"learning_rate": 4.115831895994203e-05,
|
| 415 |
+
"loss": 2.538,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.9120772946859903,
|
| 420 |
+
"grad_norm": 3.0625,
|
| 421 |
+
"learning_rate": 4.0837643360798666e-05,
|
| 422 |
+
"loss": 2.6185,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.927536231884058,
|
| 427 |
+
"grad_norm": 3.78125,
|
| 428 |
+
"learning_rate": 4.051255468997562e-05,
|
| 429 |
+
"loss": 2.4834,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.9429951690821256,
|
| 434 |
+
"grad_norm": 3.78125,
|
| 435 |
+
"learning_rate": 4.018314353163202e-05,
|
| 436 |
+
"loss": 2.504,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.9584541062801932,
|
| 441 |
+
"grad_norm": 3.59375,
|
| 442 |
+
"learning_rate": 3.984950167436423e-05,
|
| 443 |
+
"loss": 2.4611,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.9739130434782609,
|
| 448 |
+
"grad_norm": 2.765625,
|
| 449 |
+
"learning_rate": 3.9511722085629485e-05,
|
| 450 |
+
"loss": 2.5228,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.9893719806763285,
|
| 455 |
+
"grad_norm": 2.875,
|
| 456 |
+
"learning_rate": 3.916989888584095e-05,
|
| 457 |
+
"loss": 2.4755,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.0046376811594202,
|
| 462 |
+
"grad_norm": 3.015625,
|
| 463 |
+
"learning_rate": 3.882412732214167e-05,
|
| 464 |
+
"loss": 2.42,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.020096618357488,
|
| 469 |
+
"grad_norm": 2.890625,
|
| 470 |
+
"learning_rate": 3.847450374186441e-05,
|
| 471 |
+
"loss": 2.0797,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.0355555555555556,
|
| 476 |
+
"grad_norm": 3.046875,
|
| 477 |
+
"learning_rate": 3.812112556568508e-05,
|
| 478 |
+
"loss": 1.949,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.0510144927536231,
|
| 483 |
+
"grad_norm": 3.0625,
|
| 484 |
+
"learning_rate": 3.776409126047692e-05,
|
| 485 |
+
"loss": 1.9827,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.066473429951691,
|
| 490 |
+
"grad_norm": 4.03125,
|
| 491 |
+
"learning_rate": 3.740350031187332e-05,
|
| 492 |
+
"loss": 2.0228,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.0819323671497585,
|
| 497 |
+
"grad_norm": 3.0625,
|
| 498 |
+
"learning_rate": 3.7039453196546645e-05,
|
| 499 |
+
"loss": 1.9758,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.097391304347826,
|
| 504 |
+
"grad_norm": 4.25,
|
| 505 |
+
"learning_rate": 3.6672051354211035e-05,
|
| 506 |
+
"loss": 2.1762,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.1128502415458936,
|
| 511 |
+
"grad_norm": 2.453125,
|
| 512 |
+
"learning_rate": 3.6301397159356696e-05,
|
| 513 |
+
"loss": 2.0642,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.1283091787439614,
|
| 518 |
+
"grad_norm": 3.359375,
|
| 519 |
+
"learning_rate": 3.592759389272389e-05,
|
| 520 |
+
"loss": 2.0444,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.143768115942029,
|
| 525 |
+
"grad_norm": 3.46875,
|
| 526 |
+
"learning_rate": 3.555074571252431e-05,
|
| 527 |
+
"loss": 1.8933,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.1592270531400966,
|
| 532 |
+
"grad_norm": 3.53125,
|
| 533 |
+
"learning_rate": 3.517095762541795e-05,
|
| 534 |
+
"loss": 2.1575,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.1746859903381643,
|
| 539 |
+
"grad_norm": 3.25,
|
| 540 |
+
"learning_rate": 3.4788335457253595e-05,
|
| 541 |
+
"loss": 1.964,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.190144927536232,
|
| 546 |
+
"grad_norm": 5.09375,
|
| 547 |
+
"learning_rate": 3.4402985823581046e-05,
|
| 548 |
+
"loss": 1.9813,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.2056038647342995,
|
| 553 |
+
"grad_norm": 3.78125,
|
| 554 |
+
"learning_rate": 3.401501609994326e-05,
|
| 555 |
+
"loss": 1.8629,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.221062801932367,
|
| 560 |
+
"grad_norm": 5.09375,
|
| 561 |
+
"learning_rate": 3.362453439195679e-05,
|
| 562 |
+
"loss": 2.1078,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.2365217391304348,
|
| 567 |
+
"grad_norm": 3.078125,
|
| 568 |
+
"learning_rate": 3.323164950518868e-05,
|
| 569 |
+
"loss": 1.9695,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.2519806763285024,
|
| 574 |
+
"grad_norm": 4.53125,
|
| 575 |
+
"learning_rate": 3.283647091483849e-05,
|
| 576 |
+
"loss": 1.9425,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.26743961352657,
|
| 581 |
+
"grad_norm": 2.78125,
|
| 582 |
+
"learning_rate": 3.243910873523356e-05,
|
| 583 |
+
"loss": 1.9954,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.2828985507246378,
|
| 588 |
+
"grad_norm": 4.21875,
|
| 589 |
+
"learning_rate": 3.203967368914631e-05,
|
| 590 |
+
"loss": 1.8656,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.2983574879227053,
|
| 595 |
+
"grad_norm": 5.0,
|
| 596 |
+
"learning_rate": 3.163827707694193e-05,
|
| 597 |
+
"loss": 1.9414,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.313816425120773,
|
| 602 |
+
"grad_norm": 3.75,
|
| 603 |
+
"learning_rate": 3.123503074556513e-05,
|
| 604 |
+
"loss": 2.0901,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.3292753623188407,
|
| 609 |
+
"grad_norm": 3.5625,
|
| 610 |
+
"learning_rate": 3.083004705737468e-05,
|
| 611 |
+
"loss": 1.9897,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.3447342995169083,
|
| 616 |
+
"grad_norm": 3.65625,
|
| 617 |
+
"learning_rate": 3.0423438858834174e-05,
|
| 618 |
+
"loss": 1.996,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.3601932367149758,
|
| 623 |
+
"grad_norm": 5.09375,
|
| 624 |
+
"learning_rate": 3.0015319449068025e-05,
|
| 625 |
+
"loss": 2.0757,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.3756521739130434,
|
| 630 |
+
"grad_norm": 4.1875,
|
| 631 |
+
"learning_rate": 2.960580254829131e-05,
|
| 632 |
+
"loss": 1.9013,
|
| 633 |
+
"step": 890
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.3911111111111112,
|
| 637 |
+
"grad_norm": 3.65625,
|
| 638 |
+
"learning_rate": 2.919500226612224e-05,
|
| 639 |
+
"loss": 2.0255,
|
| 640 |
+
"step": 900
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.4065700483091788,
|
| 644 |
+
"grad_norm": 4.1875,
|
| 645 |
+
"learning_rate": 2.8783033069786132e-05,
|
| 646 |
+
"loss": 1.8022,
|
| 647 |
+
"step": 910
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.4220289855072463,
|
| 651 |
+
"grad_norm": 2.859375,
|
| 652 |
+
"learning_rate": 2.8370009752219788e-05,
|
| 653 |
+
"loss": 1.8961,
|
| 654 |
+
"step": 920
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.437487922705314,
|
| 658 |
+
"grad_norm": 3.8125,
|
| 659 |
+
"learning_rate": 2.795604740008504e-05,
|
| 660 |
+
"loss": 1.927,
|
| 661 |
+
"step": 930
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.4529468599033817,
|
| 665 |
+
"grad_norm": 4.0625,
|
| 666 |
+
"learning_rate": 2.7541261361700514e-05,
|
| 667 |
+
"loss": 1.9894,
|
| 668 |
+
"step": 940
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.4684057971014493,
|
| 672 |
+
"grad_norm": 3.15625,
|
| 673 |
+
"learning_rate": 2.7125767214900455e-05,
|
| 674 |
+
"loss": 1.9681,
|
| 675 |
+
"step": 950
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.4838647342995168,
|
| 679 |
+
"grad_norm": 3.53125,
|
| 680 |
+
"learning_rate": 2.67096807348296e-05,
|
| 681 |
+
"loss": 2.0076,
|
| 682 |
+
"step": 960
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.4993236714975846,
|
| 686 |
+
"grad_norm": 3.890625,
|
| 687 |
+
"learning_rate": 2.629311786168306e-05,
|
| 688 |
+
"loss": 2.0221,
|
| 689 |
+
"step": 970
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.5147826086956522,
|
| 693 |
+
"grad_norm": 3.09375,
|
| 694 |
+
"learning_rate": 2.5876194668400256e-05,
|
| 695 |
+
"loss": 2.1633,
|
| 696 |
+
"step": 980
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.5302415458937197,
|
| 700 |
+
"grad_norm": 3.40625,
|
| 701 |
+
"learning_rate": 2.545902732832181e-05,
|
| 702 |
+
"loss": 2.1009,
|
| 703 |
+
"step": 990
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.5457004830917875,
|
| 707 |
+
"grad_norm": 3.390625,
|
| 708 |
+
"learning_rate": 2.5041732082818503e-05,
|
| 709 |
+
"loss": 1.9915,
|
| 710 |
+
"step": 1000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.561159420289855,
|
| 714 |
+
"grad_norm": 3.0625,
|
| 715 |
+
"learning_rate": 2.462442520890127e-05,
|
| 716 |
+
"loss": 2.0244,
|
| 717 |
+
"step": 1010
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.5766183574879227,
|
| 721 |
+
"grad_norm": 7.3125,
|
| 722 |
+
"learning_rate": 2.4207222986821194e-05,
|
| 723 |
+
"loss": 2.0306,
|
| 724 |
+
"step": 1020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.5920772946859905,
|
| 728 |
+
"grad_norm": 3.890625,
|
| 729 |
+
"learning_rate": 2.379024166766876e-05,
|
| 730 |
+
"loss": 1.986,
|
| 731 |
+
"step": 1030
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.6075362318840578,
|
| 735 |
+
"grad_norm": 3.359375,
|
| 736 |
+
"learning_rate": 2.3373597440981027e-05,
|
| 737 |
+
"loss": 2.0218,
|
| 738 |
+
"step": 1040
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.6229951690821256,
|
| 742 |
+
"grad_norm": 4.09375,
|
| 743 |
+
"learning_rate": 2.295740640236614e-05,
|
| 744 |
+
"loss": 1.843,
|
| 745 |
+
"step": 1050
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.6384541062801934,
|
| 749 |
+
"grad_norm": 3.90625,
|
| 750 |
+
"learning_rate": 2.2541784521153875e-05,
|
| 751 |
+
"loss": 2.0191,
|
| 752 |
+
"step": 1060
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.6539130434782607,
|
| 756 |
+
"grad_norm": 4.46875,
|
| 757 |
+
"learning_rate": 2.2126847608081434e-05,
|
| 758 |
+
"loss": 1.959,
|
| 759 |
+
"step": 1070
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 1.6693719806763285,
|
| 763 |
+
"grad_norm": 4.1875,
|
| 764 |
+
"learning_rate": 2.1712711283023482e-05,
|
| 765 |
+
"loss": 1.9462,
|
| 766 |
+
"step": 1080
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 1.684830917874396,
|
| 770 |
+
"grad_norm": 3.53125,
|
| 771 |
+
"learning_rate": 2.1299490942775213e-05,
|
| 772 |
+
"loss": 2.007,
|
| 773 |
+
"step": 1090
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.7002898550724637,
|
| 777 |
+
"grad_norm": 4.0625,
|
| 778 |
+
"learning_rate": 2.088730172889777e-05,
|
| 779 |
+
"loss": 1.9861,
|
| 780 |
+
"step": 1100
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.7157487922705315,
|
| 784 |
+
"grad_norm": 3.59375,
|
| 785 |
+
"learning_rate": 2.0476258495634594e-05,
|
| 786 |
+
"loss": 2.057,
|
| 787 |
+
"step": 1110
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 1.731207729468599,
|
| 791 |
+
"grad_norm": 3.71875,
|
| 792 |
+
"learning_rate": 2.0066475777908005e-05,
|
| 793 |
+
"loss": 2.0005,
|
| 794 |
+
"step": 1120
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.7466666666666666,
|
| 798 |
+
"grad_norm": 4.65625,
|
| 799 |
+
"learning_rate": 1.9658067759404625e-05,
|
| 800 |
+
"loss": 1.9735,
|
| 801 |
+
"step": 1130
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 1.7621256038647344,
|
| 805 |
+
"grad_norm": 4.15625,
|
| 806 |
+
"learning_rate": 1.925114824075876e-05,
|
| 807 |
+
"loss": 2.0445,
|
| 808 |
+
"step": 1140
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 1.777584541062802,
|
| 812 |
+
"grad_norm": 3.890625,
|
| 813 |
+
"learning_rate": 1.8845830607842438e-05,
|
| 814 |
+
"loss": 2.0739,
|
| 815 |
+
"step": 1150
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 1.7930434782608695,
|
| 819 |
+
"grad_norm": 4.0,
|
| 820 |
+
"learning_rate": 1.8442227800171125e-05,
|
| 821 |
+
"loss": 1.9375,
|
| 822 |
+
"step": 1160
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 1.8085024154589373,
|
| 826 |
+
"grad_norm": 3.921875,
|
| 827 |
+
"learning_rate": 1.8040452279433716e-05,
|
| 828 |
+
"loss": 1.9623,
|
| 829 |
+
"step": 1170
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 1.8239613526570049,
|
| 833 |
+
"grad_norm": 2.765625,
|
| 834 |
+
"learning_rate": 1.7640615998155693e-05,
|
| 835 |
+
"loss": 1.9219,
|
| 836 |
+
"step": 1180
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.8394202898550724,
|
| 840 |
+
"grad_norm": 4.15625,
|
| 841 |
+
"learning_rate": 1.7242830368504227e-05,
|
| 842 |
+
"loss": 1.8705,
|
| 843 |
+
"step": 1190
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 1.8548792270531402,
|
| 847 |
+
"grad_norm": 2.984375,
|
| 848 |
+
"learning_rate": 1.6847206231243718e-05,
|
| 849 |
+
"loss": 1.9085,
|
| 850 |
+
"step": 1200
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 1.8703381642512076,
|
| 854 |
+
"grad_norm": 4.03125,
|
| 855 |
+
"learning_rate": 1.6453853824850726e-05,
|
| 856 |
+
"loss": 1.8975,
|
| 857 |
+
"step": 1210
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 1.8857971014492754,
|
| 861 |
+
"grad_norm": 3.4375,
|
| 862 |
+
"learning_rate": 1.606288275479652e-05,
|
| 863 |
+
"loss": 1.905,
|
| 864 |
+
"step": 1220
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 1.9012560386473432,
|
| 868 |
+
"grad_norm": 3.71875,
|
| 869 |
+
"learning_rate": 1.5674401963006235e-05,
|
| 870 |
+
"loss": 2.1359,
|
| 871 |
+
"step": 1230
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 1.9167149758454105,
|
| 875 |
+
"grad_norm": 4.25,
|
| 876 |
+
"learning_rate": 1.5288519697502697e-05,
|
| 877 |
+
"loss": 1.9069,
|
| 878 |
+
"step": 1240
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.9321739130434783,
|
| 882 |
+
"grad_norm": 5.34375,
|
| 883 |
+
"learning_rate": 1.4905343482243775e-05,
|
| 884 |
+
"loss": 1.8041,
|
| 885 |
+
"step": 1250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 1.9476328502415459,
|
| 889 |
+
"grad_norm": 2.96875,
|
| 890 |
+
"learning_rate": 1.4524980087161438e-05,
|
| 891 |
+
"loss": 2.1233,
|
| 892 |
+
"step": 1260
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 1.9630917874396134,
|
| 896 |
+
"grad_norm": 2.625,
|
| 897 |
+
"learning_rate": 1.4147535498410853e-05,
|
| 898 |
+
"loss": 1.9942,
|
| 899 |
+
"step": 1270
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 1.9785507246376812,
|
| 903 |
+
"grad_norm": 2.890625,
|
| 904 |
+
"learning_rate": 1.3773114888837957e-05,
|
| 905 |
+
"loss": 2.0154,
|
| 906 |
+
"step": 1280
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 1.9940096618357488,
|
| 910 |
+
"grad_norm": 3.875,
|
| 911 |
+
"learning_rate": 1.3401822588673636e-05,
|
| 912 |
+
"loss": 1.9747,
|
| 913 |
+
"step": 1290
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 2.0092753623188404,
|
| 917 |
+
"grad_norm": 3.625,
|
| 918 |
+
"learning_rate": 1.3033762056462654e-05,
|
| 919 |
+
"loss": 1.8271,
|
| 920 |
+
"step": 1300
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 2.024734299516908,
|
| 924 |
+
"grad_norm": 2.5,
|
| 925 |
+
"learning_rate": 1.2669035850235536e-05,
|
| 926 |
+
"loss": 1.5612,
|
| 927 |
+
"step": 1310
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 2.040193236714976,
|
| 931 |
+
"grad_norm": 4.3125,
|
| 932 |
+
"learning_rate": 1.230774559893131e-05,
|
| 933 |
+
"loss": 1.6105,
|
| 934 |
+
"step": 1320
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 2.0556521739130433,
|
| 938 |
+
"grad_norm": 5.1875,
|
| 939 |
+
"learning_rate": 1.1949991974079197e-05,
|
| 940 |
+
"loss": 1.6631,
|
| 941 |
+
"step": 1330
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 2.071111111111111,
|
| 945 |
+
"grad_norm": 3.84375,
|
| 946 |
+
"learning_rate": 1.1595874661746986e-05,
|
| 947 |
+
"loss": 1.7435,
|
| 948 |
+
"step": 1340
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 2.086570048309179,
|
| 952 |
+
"grad_norm": 3.484375,
|
| 953 |
+
"learning_rate": 1.1245492334764169e-05,
|
| 954 |
+
"loss": 1.5578,
|
| 955 |
+
"step": 1350
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 2.1020289855072463,
|
| 959 |
+
"grad_norm": 5.84375,
|
| 960 |
+
"learning_rate": 1.0898942625227168e-05,
|
| 961 |
+
"loss": 1.7355,
|
| 962 |
+
"step": 1360
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 2.117487922705314,
|
| 966 |
+
"grad_norm": 5.15625,
|
| 967 |
+
"learning_rate": 1.0556322097294835e-05,
|
| 968 |
+
"loss": 1.6674,
|
| 969 |
+
"step": 1370
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 2.132946859903382,
|
| 973 |
+
"grad_norm": 3.921875,
|
| 974 |
+
"learning_rate": 1.0217726220281243e-05,
|
| 975 |
+
"loss": 1.6355,
|
| 976 |
+
"step": 1380
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 2.148405797101449,
|
| 980 |
+
"grad_norm": 3.34375,
|
| 981 |
+
"learning_rate": 9.883249342053821e-06,
|
| 982 |
+
"loss": 1.7309,
|
| 983 |
+
"step": 1390
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 2.163864734299517,
|
| 987 |
+
"grad_norm": 3.71875,
|
| 988 |
+
"learning_rate": 9.552984662743716e-06,
|
| 989 |
+
"loss": 1.6491,
|
| 990 |
+
"step": 1400
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 2.1793236714975848,
|
| 994 |
+
"grad_norm": 3.6875,
|
| 995 |
+
"learning_rate": 9.227024208776224e-06,
|
| 996 |
+
"loss": 1.5166,
|
| 997 |
+
"step": 1410
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 2.194782608695652,
|
| 1001 |
+
"grad_norm": 4.0,
|
| 1002 |
+
"learning_rate": 8.905458807228013e-06,
|
| 1003 |
+
"loss": 1.7866,
|
| 1004 |
+
"step": 1420
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 2.21024154589372,
|
| 1008 |
+
"grad_norm": 3.25,
|
| 1009 |
+
"learning_rate": 8.588378060518738e-06,
|
| 1010 |
+
"loss": 1.6775,
|
| 1011 |
+
"step": 1430
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 2.2257004830917873,
|
| 1015 |
+
"grad_norm": 4.34375,
|
| 1016 |
+
"learning_rate": 8.275870321443802e-06,
|
| 1017 |
+
"loss": 1.5462,
|
| 1018 |
+
"step": 1440
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 2.241159420289855,
|
| 1022 |
+
"grad_norm": 3.15625,
|
| 1023 |
+
"learning_rate": 7.968022668555334e-06,
|
| 1024 |
+
"loss": 1.7529,
|
| 1025 |
+
"step": 1450
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 2.256618357487923,
|
| 1029 |
+
"grad_norm": 3.796875,
|
| 1030 |
+
"learning_rate": 7.664920881898236e-06,
|
| 1031 |
+
"loss": 1.5787,
|
| 1032 |
+
"step": 1460
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 2.27207729468599,
|
| 1036 |
+
"grad_norm": 3.875,
|
| 1037 |
+
"learning_rate": 7.3666494191080455e-06,
|
| 1038 |
+
"loss": 1.5908,
|
| 1039 |
+
"step": 1470
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 2.287536231884058,
|
| 1043 |
+
"grad_norm": 3.375,
|
| 1044 |
+
"learning_rate": 7.073291391877288e-06,
|
| 1045 |
+
"loss": 1.5812,
|
| 1046 |
+
"step": 1480
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 2.3029951690821258,
|
| 1050 |
+
"grad_norm": 4.46875,
|
| 1051 |
+
"learning_rate": 6.7849285427968586e-06,
|
| 1052 |
+
"loss": 1.5513,
|
| 1053 |
+
"step": 1490
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 2.318454106280193,
|
| 1057 |
+
"grad_norm": 4.75,
|
| 1058 |
+
"learning_rate": 6.5016412225789e-06,
|
| 1059 |
+
"loss": 1.6197,
|
| 1060 |
+
"step": 1500
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 2.333913043478261,
|
| 1064 |
+
"grad_norm": 3.859375,
|
| 1065 |
+
"learning_rate": 6.223508367667538e-06,
|
| 1066 |
+
"loss": 1.624,
|
| 1067 |
+
"step": 1510
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 2.3493719806763287,
|
| 1071 |
+
"grad_norm": 3.796875,
|
| 1072 |
+
"learning_rate": 5.9506074782436275e-06,
|
| 1073 |
+
"loss": 1.6575,
|
| 1074 |
+
"step": 1520
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 2.364830917874396,
|
| 1078 |
+
"grad_norm": 4.03125,
|
| 1079 |
+
"learning_rate": 5.683014596629857e-06,
|
| 1080 |
+
"loss": 1.501,
|
| 1081 |
+
"step": 1530
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 2.380289855072464,
|
| 1085 |
+
"grad_norm": 3.453125,
|
| 1086 |
+
"learning_rate": 5.420804286101838e-06,
|
| 1087 |
+
"loss": 1.7582,
|
| 1088 |
+
"step": 1540
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 2.395748792270531,
|
| 1092 |
+
"grad_norm": 3.96875,
|
| 1093 |
+
"learning_rate": 5.164049610111607e-06,
|
| 1094 |
+
"loss": 1.8236,
|
| 1095 |
+
"step": 1550
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 2.411207729468599,
|
| 1099 |
+
"grad_norm": 2.96875,
|
| 1100 |
+
"learning_rate": 4.912822111928767e-06,
|
| 1101 |
+
"loss": 1.7672,
|
| 1102 |
+
"step": 1560
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 2.4266666666666667,
|
| 1106 |
+
"grad_norm": 5.8125,
|
| 1107 |
+
"learning_rate": 4.667191794705444e-06,
|
| 1108 |
+
"loss": 1.6043,
|
| 1109 |
+
"step": 1570
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 2.442125603864734,
|
| 1113 |
+
"grad_norm": 9.0625,
|
| 1114 |
+
"learning_rate": 4.427227101970261e-06,
|
| 1115 |
+
"loss": 1.6339,
|
| 1116 |
+
"step": 1580
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 2.457584541062802,
|
| 1120 |
+
"grad_norm": 3.90625,
|
| 1121 |
+
"learning_rate": 4.192994898556921e-06,
|
| 1122 |
+
"loss": 1.685,
|
| 1123 |
+
"step": 1590
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 2.4730434782608697,
|
| 1127 |
+
"grad_norm": 3.6875,
|
| 1128 |
+
"learning_rate": 3.964560451972705e-06,
|
| 1129 |
+
"loss": 1.6333,
|
| 1130 |
+
"step": 1600
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 2.488502415458937,
|
| 1134 |
+
"grad_norm": 2.890625,
|
| 1135 |
+
"learning_rate": 3.7419874142120363e-06,
|
| 1136 |
+
"loss": 1.661,
|
| 1137 |
+
"step": 1610
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 2.503961352657005,
|
| 1141 |
+
"grad_norm": 5.1875,
|
| 1142 |
+
"learning_rate": 3.525337804020212e-06,
|
| 1143 |
+
"loss": 1.6976,
|
| 1144 |
+
"step": 1620
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 2.5194202898550726,
|
| 1148 |
+
"grad_norm": 2.953125,
|
| 1149 |
+
"learning_rate": 3.314671989612195e-06,
|
| 1150 |
+
"loss": 1.6764,
|
| 1151 |
+
"step": 1630
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 2.53487922705314,
|
| 1155 |
+
"grad_norm": 5.03125,
|
| 1156 |
+
"learning_rate": 3.110048671851404e-06,
|
| 1157 |
+
"loss": 1.658,
|
| 1158 |
+
"step": 1640
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 2.5503381642512077,
|
| 1162 |
+
"grad_norm": 4.03125,
|
| 1163 |
+
"learning_rate": 2.911524867892956e-06,
|
| 1164 |
+
"loss": 1.5693,
|
| 1165 |
+
"step": 1650
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 2.5657971014492755,
|
| 1169 |
+
"grad_norm": 5.1875,
|
| 1170 |
+
"learning_rate": 2.719155895296255e-06,
|
| 1171 |
+
"loss": 1.6833,
|
| 1172 |
+
"step": 1660
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 2.581256038647343,
|
| 1176 |
+
"grad_norm": 4.40625,
|
| 1177 |
+
"learning_rate": 2.5329953566109537e-06,
|
| 1178 |
+
"loss": 1.7026,
|
| 1179 |
+
"step": 1670
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 2.5967149758454107,
|
| 1183 |
+
"grad_norm": 4.59375,
|
| 1184 |
+
"learning_rate": 2.353095124440999e-06,
|
| 1185 |
+
"loss": 1.6888,
|
| 1186 |
+
"step": 1680
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 2.6121739130434785,
|
| 1190 |
+
"grad_norm": 3.296875,
|
| 1191 |
+
"learning_rate": 2.17950532699053e-06,
|
| 1192 |
+
"loss": 1.5656,
|
| 1193 |
+
"step": 1690
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 2.627632850241546,
|
| 1197 |
+
"grad_norm": 3.78125,
|
| 1198 |
+
"learning_rate": 2.012274334095998e-06,
|
| 1199 |
+
"loss": 1.732,
|
| 1200 |
+
"step": 1700
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 2.6430917874396136,
|
| 1204 |
+
"grad_norm": 3.15625,
|
| 1205 |
+
"learning_rate": 1.8514487437481436e-06,
|
| 1206 |
+
"loss": 1.4964,
|
| 1207 |
+
"step": 1710
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 2.6585507246376814,
|
| 1211 |
+
"grad_norm": 3.640625,
|
| 1212 |
+
"learning_rate": 1.6970733691077378e-06,
|
| 1213 |
+
"loss": 1.5922,
|
| 1214 |
+
"step": 1720
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 2.6740096618357487,
|
| 1218 |
+
"grad_norm": 3.609375,
|
| 1219 |
+
"learning_rate": 1.549191226018637e-06,
|
| 1220 |
+
"loss": 1.6481,
|
| 1221 |
+
"step": 1730
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 2.6894685990338165,
|
| 1225 |
+
"grad_norm": 4.0,
|
| 1226 |
+
"learning_rate": 1.4078435210216568e-06,
|
| 1227 |
+
"loss": 1.5963,
|
| 1228 |
+
"step": 1740
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 2.7049275362318843,
|
| 1232 |
+
"grad_norm": 3.9375,
|
| 1233 |
+
"learning_rate": 1.273069639872601e-06,
|
| 1234 |
+
"loss": 1.6844,
|
| 1235 |
+
"step": 1750
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 2.7203864734299517,
|
| 1239 |
+
"grad_norm": 3.375,
|
| 1240 |
+
"learning_rate": 1.1449071365676356e-06,
|
| 1241 |
+
"loss": 1.7353,
|
| 1242 |
+
"step": 1760
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 2.7358454106280194,
|
| 1246 |
+
"grad_norm": 6.0,
|
| 1247 |
+
"learning_rate": 1.0233917228790828e-06,
|
| 1248 |
+
"loss": 1.6287,
|
| 1249 |
+
"step": 1770
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 2.751304347826087,
|
| 1253 |
+
"grad_norm": 3.46875,
|
| 1254 |
+
"learning_rate": 9.08557258404516e-07,
|
| 1255 |
+
"loss": 1.6166,
|
| 1256 |
+
"step": 1780
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 2.7667632850241546,
|
| 1260 |
+
"grad_norm": 4.21875,
|
| 1261 |
+
"learning_rate": 8.004357411319857e-07,
|
| 1262 |
+
"loss": 1.6725,
|
| 1263 |
+
"step": 1790
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 2.7822222222222224,
|
| 1267 |
+
"grad_norm": 4.6875,
|
| 1268 |
+
"learning_rate": 6.990572985239612e-07,
|
| 1269 |
+
"loss": 1.5655,
|
| 1270 |
+
"step": 1800
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 2.7976811594202897,
|
| 1274 |
+
"grad_norm": 3.703125,
|
| 1275 |
+
"learning_rate": 6.044501791224539e-07,
|
| 1276 |
+
"loss": 1.6394,
|
| 1277 |
+
"step": 1810
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 2.8131400966183575,
|
| 1281 |
+
"grad_norm": 4.21875,
|
| 1282 |
+
"learning_rate": 5.166407446777572e-07,
|
| 1283 |
+
"loss": 1.7043,
|
| 1284 |
+
"step": 1820
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 2.828599033816425,
|
| 1288 |
+
"grad_norm": 6.5,
|
| 1289 |
+
"learning_rate": 4.3565346280284305e-07,
|
| 1290 |
+
"loss": 1.5092,
|
| 1291 |
+
"step": 1830
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 2.8440579710144926,
|
| 1295 |
+
"grad_norm": 4.21875,
|
| 1296 |
+
"learning_rate": 3.6151090015565103e-07,
|
| 1297 |
+
"loss": 1.7447,
|
| 1298 |
+
"step": 1840
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 2.8595169082125604,
|
| 1302 |
+
"grad_norm": 3.984375,
|
| 1303 |
+
"learning_rate": 2.9423371615096005e-07,
|
| 1304 |
+
"loss": 1.66,
|
| 1305 |
+
"step": 1850
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 2.874975845410628,
|
| 1309 |
+
"grad_norm": 3.296875,
|
| 1310 |
+
"learning_rate": 2.3384065720379878e-07,
|
| 1311 |
+
"loss": 1.6567,
|
| 1312 |
+
"step": 1860
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 2.8904347826086956,
|
| 1316 |
+
"grad_norm": 3.71875,
|
| 1317 |
+
"learning_rate": 1.803485515058323e-07,
|
| 1318 |
+
"loss": 1.5354,
|
| 1319 |
+
"step": 1870
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 2.9058937198067634,
|
| 1323 |
+
"grad_norm": 5.28125,
|
| 1324 |
+
"learning_rate": 1.3377230433630205e-07,
|
| 1325 |
+
"loss": 1.6183,
|
| 1326 |
+
"step": 1880
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 2.9213526570048307,
|
| 1330 |
+
"grad_norm": 4.1875,
|
| 1331 |
+
"learning_rate": 9.412489390873414e-08,
|
| 1332 |
+
"loss": 1.7047,
|
| 1333 |
+
"step": 1890
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 2.9368115942028985,
|
| 1337 |
+
"grad_norm": 4.78125,
|
| 1338 |
+
"learning_rate": 6.141736775464313e-08,
|
| 1339 |
+
"loss": 1.5235,
|
| 1340 |
+
"step": 1900
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 2.9522705314009663,
|
| 1344 |
+
"grad_norm": 4.375,
|
| 1345 |
+
"learning_rate": 3.5658839645194565e-08,
|
| 1346 |
+
"loss": 1.6394,
|
| 1347 |
+
"step": 1910
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 2.9677294685990336,
|
| 1351 |
+
"grad_norm": 4.3125,
|
| 1352 |
+
"learning_rate": 1.6856487051700177e-08,
|
| 1353 |
+
"loss": 1.5718,
|
| 1354 |
+
"step": 1920
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 2.9831884057971014,
|
| 1358 |
+
"grad_norm": 3.296875,
|
| 1359 |
+
"learning_rate": 5.015549145664933e-09,
|
| 1360 |
+
"loss": 1.6534,
|
| 1361 |
+
"step": 1930
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 2.998647342995169,
|
| 1365 |
+
"grad_norm": 4.65625,
|
| 1366 |
+
"learning_rate": 1.393253389103677e-10,
|
| 1367 |
+
"loss": 1.6622,
|
| 1368 |
+
"step": 1940
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 3.0,
|
| 1372 |
+
"step": 1941,
|
| 1373 |
+
"total_flos": 7.76867645696736e+16,
|
| 1374 |
+
"train_loss": 2.121323910710248,
|
| 1375 |
+
"train_runtime": 3108.1718,
|
| 1376 |
+
"train_samples_per_second": 9.99,
|
| 1377 |
+
"train_steps_per_second": 0.624
|
| 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": 7.76867645696736e+16,
|
| 1398 |
+
"train_batch_size": 2,
|
| 1399 |
+
"trial_name": null,
|
| 1400 |
+
"trial_params": null
|
| 1401 |
+
}
|