Upload training info: trainer_state.json
Browse files- trainer_state.json +1972 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1972 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 138660,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.018029712966969567,
|
| 13 |
+
"grad_norm": 19.02682113647461,
|
| 14 |
+
"learning_rate": 4.981970287033031e-05,
|
| 15 |
+
"loss": 0.9419,
|
| 16 |
+
"step": 500
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.036059425933939133,
|
| 20 |
+
"grad_norm": 17.821584701538086,
|
| 21 |
+
"learning_rate": 4.963940574066061e-05,
|
| 22 |
+
"loss": 0.6135,
|
| 23 |
+
"step": 1000
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.0540891389009087,
|
| 27 |
+
"grad_norm": 12.752524375915527,
|
| 28 |
+
"learning_rate": 4.945910861099092e-05,
|
| 29 |
+
"loss": 0.5186,
|
| 30 |
+
"step": 1500
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.07211885186787827,
|
| 34 |
+
"grad_norm": 13.052861213684082,
|
| 35 |
+
"learning_rate": 4.927881148132122e-05,
|
| 36 |
+
"loss": 0.4466,
|
| 37 |
+
"step": 2000
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.09014856483484783,
|
| 41 |
+
"grad_norm": 19.749935150146484,
|
| 42 |
+
"learning_rate": 4.9098514351651524e-05,
|
| 43 |
+
"loss": 0.4104,
|
| 44 |
+
"step": 2500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.1081782778018174,
|
| 48 |
+
"grad_norm": 10.785118103027344,
|
| 49 |
+
"learning_rate": 4.8918217221981825e-05,
|
| 50 |
+
"loss": 0.3819,
|
| 51 |
+
"step": 3000
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.12620799076878697,
|
| 55 |
+
"grad_norm": 12.835149765014648,
|
| 56 |
+
"learning_rate": 4.873792009231213e-05,
|
| 57 |
+
"loss": 0.3584,
|
| 58 |
+
"step": 3500
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.14423770373575653,
|
| 62 |
+
"grad_norm": 12.419609069824219,
|
| 63 |
+
"learning_rate": 4.855762296264243e-05,
|
| 64 |
+
"loss": 0.3409,
|
| 65 |
+
"step": 4000
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.1622674167027261,
|
| 69 |
+
"grad_norm": 10.989585876464844,
|
| 70 |
+
"learning_rate": 4.837732583297274e-05,
|
| 71 |
+
"loss": 0.3163,
|
| 72 |
+
"step": 4500
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.18029712966969566,
|
| 76 |
+
"grad_norm": 10.476921081542969,
|
| 77 |
+
"learning_rate": 4.8197028703303046e-05,
|
| 78 |
+
"loss": 0.3156,
|
| 79 |
+
"step": 5000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.19832684263666522,
|
| 83 |
+
"grad_norm": 14.82525634765625,
|
| 84 |
+
"learning_rate": 4.8016731573633353e-05,
|
| 85 |
+
"loss": 0.3038,
|
| 86 |
+
"step": 5500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.2163565556036348,
|
| 90 |
+
"grad_norm": 7.5562334060668945,
|
| 91 |
+
"learning_rate": 4.7836434443963654e-05,
|
| 92 |
+
"loss": 0.2922,
|
| 93 |
+
"step": 6000
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.23438626857060435,
|
| 97 |
+
"grad_norm": 14.660656929016113,
|
| 98 |
+
"learning_rate": 4.765613731429396e-05,
|
| 99 |
+
"loss": 0.2773,
|
| 100 |
+
"step": 6500
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.25241598153757394,
|
| 104 |
+
"grad_norm": 8.908041000366211,
|
| 105 |
+
"learning_rate": 4.747584018462427e-05,
|
| 106 |
+
"loss": 0.2745,
|
| 107 |
+
"step": 7000
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.2704456945045435,
|
| 111 |
+
"grad_norm": 13.304757118225098,
|
| 112 |
+
"learning_rate": 4.729554305495457e-05,
|
| 113 |
+
"loss": 0.2653,
|
| 114 |
+
"step": 7500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.28847540747151307,
|
| 118 |
+
"grad_norm": 9.697709083557129,
|
| 119 |
+
"learning_rate": 4.7115245925284875e-05,
|
| 120 |
+
"loss": 0.2741,
|
| 121 |
+
"step": 8000
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.3065051204384826,
|
| 125 |
+
"grad_norm": 18.51068687438965,
|
| 126 |
+
"learning_rate": 4.6934948795615176e-05,
|
| 127 |
+
"loss": 0.256,
|
| 128 |
+
"step": 8500
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.3245348334054522,
|
| 132 |
+
"grad_norm": 10.845968246459961,
|
| 133 |
+
"learning_rate": 4.675465166594548e-05,
|
| 134 |
+
"loss": 0.2553,
|
| 135 |
+
"step": 9000
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.34256454637242173,
|
| 139 |
+
"grad_norm": 9.60972785949707,
|
| 140 |
+
"learning_rate": 4.657435453627578e-05,
|
| 141 |
+
"loss": 0.2395,
|
| 142 |
+
"step": 9500
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.3605942593393913,
|
| 146 |
+
"grad_norm": 9.215226173400879,
|
| 147 |
+
"learning_rate": 4.639405740660609e-05,
|
| 148 |
+
"loss": 0.2435,
|
| 149 |
+
"step": 10000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.37862397230636086,
|
| 153 |
+
"grad_norm": 6.0244221687316895,
|
| 154 |
+
"learning_rate": 4.621376027693639e-05,
|
| 155 |
+
"loss": 0.2404,
|
| 156 |
+
"step": 10500
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.39665368527333045,
|
| 160 |
+
"grad_norm": 6.418239116668701,
|
| 161 |
+
"learning_rate": 4.60334631472667e-05,
|
| 162 |
+
"loss": 0.2345,
|
| 163 |
+
"step": 11000
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.41468339824030004,
|
| 167 |
+
"grad_norm": 8.997867584228516,
|
| 168 |
+
"learning_rate": 4.5853166017597005e-05,
|
| 169 |
+
"loss": 0.2272,
|
| 170 |
+
"step": 11500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.4327131112072696,
|
| 174 |
+
"grad_norm": 7.10783576965332,
|
| 175 |
+
"learning_rate": 4.5672868887927305e-05,
|
| 176 |
+
"loss": 0.2265,
|
| 177 |
+
"step": 12000
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.45074282417423917,
|
| 181 |
+
"grad_norm": 11.675990104675293,
|
| 182 |
+
"learning_rate": 4.549257175825761e-05,
|
| 183 |
+
"loss": 0.2226,
|
| 184 |
+
"step": 12500
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.4687725371412087,
|
| 188 |
+
"grad_norm": 5.2179718017578125,
|
| 189 |
+
"learning_rate": 4.531227462858791e-05,
|
| 190 |
+
"loss": 0.2129,
|
| 191 |
+
"step": 13000
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.4868022501081783,
|
| 195 |
+
"grad_norm": 5.713563442230225,
|
| 196 |
+
"learning_rate": 4.513197749891822e-05,
|
| 197 |
+
"loss": 0.2102,
|
| 198 |
+
"step": 13500
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.5048319630751479,
|
| 202 |
+
"grad_norm": 8.981880187988281,
|
| 203 |
+
"learning_rate": 4.495168036924852e-05,
|
| 204 |
+
"loss": 0.2109,
|
| 205 |
+
"step": 14000
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.5228616760421174,
|
| 209 |
+
"grad_norm": 7.710575103759766,
|
| 210 |
+
"learning_rate": 4.477138323957883e-05,
|
| 211 |
+
"loss": 0.2027,
|
| 212 |
+
"step": 14500
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.540891389009087,
|
| 216 |
+
"grad_norm": 4.311069488525391,
|
| 217 |
+
"learning_rate": 4.459108610990913e-05,
|
| 218 |
+
"loss": 0.2104,
|
| 219 |
+
"step": 15000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.5589211019760565,
|
| 223 |
+
"grad_norm": 11.499236106872559,
|
| 224 |
+
"learning_rate": 4.4410788980239435e-05,
|
| 225 |
+
"loss": 0.1983,
|
| 226 |
+
"step": 15500
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.5769508149430261,
|
| 230 |
+
"grad_norm": 11.01004409790039,
|
| 231 |
+
"learning_rate": 4.423049185056974e-05,
|
| 232 |
+
"loss": 0.2002,
|
| 233 |
+
"step": 16000
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.5949805279099957,
|
| 237 |
+
"grad_norm": 12.05623722076416,
|
| 238 |
+
"learning_rate": 4.405019472090005e-05,
|
| 239 |
+
"loss": 0.196,
|
| 240 |
+
"step": 16500
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.6130102408769652,
|
| 244 |
+
"grad_norm": 9.926684379577637,
|
| 245 |
+
"learning_rate": 4.386989759123035e-05,
|
| 246 |
+
"loss": 0.1947,
|
| 247 |
+
"step": 17000
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.6310399538439349,
|
| 251 |
+
"grad_norm": 10.037376403808594,
|
| 252 |
+
"learning_rate": 4.3689600461560656e-05,
|
| 253 |
+
"loss": 0.1899,
|
| 254 |
+
"step": 17500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.6490696668109044,
|
| 258 |
+
"grad_norm": 10.394434928894043,
|
| 259 |
+
"learning_rate": 4.350930333189096e-05,
|
| 260 |
+
"loss": 0.1952,
|
| 261 |
+
"step": 18000
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.6670993797778739,
|
| 265 |
+
"grad_norm": 12.157942771911621,
|
| 266 |
+
"learning_rate": 4.3329006202221264e-05,
|
| 267 |
+
"loss": 0.1905,
|
| 268 |
+
"step": 18500
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.6851290927448435,
|
| 272 |
+
"grad_norm": 7.497695446014404,
|
| 273 |
+
"learning_rate": 4.314870907255157e-05,
|
| 274 |
+
"loss": 0.1931,
|
| 275 |
+
"step": 19000
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.7031588057118131,
|
| 279 |
+
"grad_norm": 12.07995891571045,
|
| 280 |
+
"learning_rate": 4.296841194288187e-05,
|
| 281 |
+
"loss": 0.1827,
|
| 282 |
+
"step": 19500
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.7211885186787826,
|
| 286 |
+
"grad_norm": 7.538449764251709,
|
| 287 |
+
"learning_rate": 4.278811481321218e-05,
|
| 288 |
+
"loss": 0.1796,
|
| 289 |
+
"step": 20000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.7392182316457522,
|
| 293 |
+
"grad_norm": 10.984667778015137,
|
| 294 |
+
"learning_rate": 4.260781768354248e-05,
|
| 295 |
+
"loss": 0.1771,
|
| 296 |
+
"step": 20500
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.7572479446127217,
|
| 300 |
+
"grad_norm": 10.688679695129395,
|
| 301 |
+
"learning_rate": 4.2427520553872786e-05,
|
| 302 |
+
"loss": 0.1896,
|
| 303 |
+
"step": 21000
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.7752776575796914,
|
| 307 |
+
"grad_norm": 7.357429504394531,
|
| 308 |
+
"learning_rate": 4.2247223424203086e-05,
|
| 309 |
+
"loss": 0.1711,
|
| 310 |
+
"step": 21500
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.7933073705466609,
|
| 314 |
+
"grad_norm": 5.935719013214111,
|
| 315 |
+
"learning_rate": 4.206692629453339e-05,
|
| 316 |
+
"loss": 0.1724,
|
| 317 |
+
"step": 22000
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.8113370835136304,
|
| 321 |
+
"grad_norm": 3.978806495666504,
|
| 322 |
+
"learning_rate": 4.18866291648637e-05,
|
| 323 |
+
"loss": 0.1699,
|
| 324 |
+
"step": 22500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.8293667964806001,
|
| 328 |
+
"grad_norm": 2.999682664871216,
|
| 329 |
+
"learning_rate": 4.1706332035194e-05,
|
| 330 |
+
"loss": 0.1663,
|
| 331 |
+
"step": 23000
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.8473965094475696,
|
| 335 |
+
"grad_norm": 5.842391014099121,
|
| 336 |
+
"learning_rate": 4.152603490552431e-05,
|
| 337 |
+
"loss": 0.1698,
|
| 338 |
+
"step": 23500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.8654262224145391,
|
| 342 |
+
"grad_norm": 9.631185531616211,
|
| 343 |
+
"learning_rate": 4.134573777585461e-05,
|
| 344 |
+
"loss": 0.1644,
|
| 345 |
+
"step": 24000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.8834559353815087,
|
| 349 |
+
"grad_norm": 5.9242753982543945,
|
| 350 |
+
"learning_rate": 4.1165440646184915e-05,
|
| 351 |
+
"loss": 0.1637,
|
| 352 |
+
"step": 24500
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.9014856483484783,
|
| 356 |
+
"grad_norm": 5.202564239501953,
|
| 357 |
+
"learning_rate": 4.0985143516515215e-05,
|
| 358 |
+
"loss": 0.1634,
|
| 359 |
+
"step": 25000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.9195153613154479,
|
| 363 |
+
"grad_norm": 17.529056549072266,
|
| 364 |
+
"learning_rate": 4.080484638684552e-05,
|
| 365 |
+
"loss": 0.1633,
|
| 366 |
+
"step": 25500
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.9375450742824174,
|
| 370 |
+
"grad_norm": 9.436565399169922,
|
| 371 |
+
"learning_rate": 4.062454925717582e-05,
|
| 372 |
+
"loss": 0.1596,
|
| 373 |
+
"step": 26000
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.9555747872493869,
|
| 377 |
+
"grad_norm": 3.6704201698303223,
|
| 378 |
+
"learning_rate": 4.044425212750613e-05,
|
| 379 |
+
"loss": 0.1619,
|
| 380 |
+
"step": 26500
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.9736045002163566,
|
| 384 |
+
"grad_norm": 5.051164150238037,
|
| 385 |
+
"learning_rate": 4.026395499783644e-05,
|
| 386 |
+
"loss": 0.1607,
|
| 387 |
+
"step": 27000
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.9916342131833261,
|
| 391 |
+
"grad_norm": 7.4777021408081055,
|
| 392 |
+
"learning_rate": 4.0083657868166744e-05,
|
| 393 |
+
"loss": 0.156,
|
| 394 |
+
"step": 27500
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.0096639261502958,
|
| 398 |
+
"grad_norm": 10.864873886108398,
|
| 399 |
+
"learning_rate": 3.9903360738497044e-05,
|
| 400 |
+
"loss": 0.1463,
|
| 401 |
+
"step": 28000
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.0276936391172653,
|
| 405 |
+
"grad_norm": 7.504316806793213,
|
| 406 |
+
"learning_rate": 3.972306360882735e-05,
|
| 407 |
+
"loss": 0.138,
|
| 408 |
+
"step": 28500
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.0457233520842348,
|
| 412 |
+
"grad_norm": 4.762522220611572,
|
| 413 |
+
"learning_rate": 3.954276647915766e-05,
|
| 414 |
+
"loss": 0.1385,
|
| 415 |
+
"step": 29000
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.0637530650512044,
|
| 419 |
+
"grad_norm": 8.836398124694824,
|
| 420 |
+
"learning_rate": 3.936246934948796e-05,
|
| 421 |
+
"loss": 0.137,
|
| 422 |
+
"step": 29500
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.081782778018174,
|
| 426 |
+
"grad_norm": 5.3541178703308105,
|
| 427 |
+
"learning_rate": 3.9182172219818266e-05,
|
| 428 |
+
"loss": 0.1325,
|
| 429 |
+
"step": 30000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.0998124909851434,
|
| 433 |
+
"grad_norm": 8.122745513916016,
|
| 434 |
+
"learning_rate": 3.9001875090148566e-05,
|
| 435 |
+
"loss": 0.141,
|
| 436 |
+
"step": 30500
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.117842203952113,
|
| 440 |
+
"grad_norm": 3.6183993816375732,
|
| 441 |
+
"learning_rate": 3.8821577960478873e-05,
|
| 442 |
+
"loss": 0.1431,
|
| 443 |
+
"step": 31000
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.1358719169190827,
|
| 447 |
+
"grad_norm": 11.851678848266602,
|
| 448 |
+
"learning_rate": 3.8641280830809174e-05,
|
| 449 |
+
"loss": 0.14,
|
| 450 |
+
"step": 31500
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.1539016298860523,
|
| 454 |
+
"grad_norm": 8.035588264465332,
|
| 455 |
+
"learning_rate": 3.846098370113948e-05,
|
| 456 |
+
"loss": 0.1316,
|
| 457 |
+
"step": 32000
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.1719313428530218,
|
| 461 |
+
"grad_norm": 18.247848510742188,
|
| 462 |
+
"learning_rate": 3.828068657146978e-05,
|
| 463 |
+
"loss": 0.1324,
|
| 464 |
+
"step": 32500
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.1899610558199913,
|
| 468 |
+
"grad_norm": 4.085416316986084,
|
| 469 |
+
"learning_rate": 3.810038944180009e-05,
|
| 470 |
+
"loss": 0.1289,
|
| 471 |
+
"step": 33000
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 1.2079907687869609,
|
| 475 |
+
"grad_norm": 6.628881454467773,
|
| 476 |
+
"learning_rate": 3.7920092312130395e-05,
|
| 477 |
+
"loss": 0.1317,
|
| 478 |
+
"step": 33500
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 1.2260204817539304,
|
| 482 |
+
"grad_norm": 9.051176071166992,
|
| 483 |
+
"learning_rate": 3.7739795182460696e-05,
|
| 484 |
+
"loss": 0.1338,
|
| 485 |
+
"step": 34000
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.2440501947209,
|
| 489 |
+
"grad_norm": 8.295066833496094,
|
| 490 |
+
"learning_rate": 3.7559498052791e-05,
|
| 491 |
+
"loss": 0.1288,
|
| 492 |
+
"step": 34500
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 1.2620799076878697,
|
| 496 |
+
"grad_norm": 4.8148345947265625,
|
| 497 |
+
"learning_rate": 3.73792009231213e-05,
|
| 498 |
+
"loss": 0.1284,
|
| 499 |
+
"step": 35000
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 1.2801096206548392,
|
| 503 |
+
"grad_norm": 1.635716199874878,
|
| 504 |
+
"learning_rate": 3.719890379345161e-05,
|
| 505 |
+
"loss": 0.1323,
|
| 506 |
+
"step": 35500
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.2981393336218088,
|
| 510 |
+
"grad_norm": 5.097489833831787,
|
| 511 |
+
"learning_rate": 3.701860666378191e-05,
|
| 512 |
+
"loss": 0.1221,
|
| 513 |
+
"step": 36000
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 1.3161690465887783,
|
| 517 |
+
"grad_norm": 6.574998378753662,
|
| 518 |
+
"learning_rate": 3.683830953411222e-05,
|
| 519 |
+
"loss": 0.124,
|
| 520 |
+
"step": 36500
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.3341987595557478,
|
| 524 |
+
"grad_norm": 7.00177001953125,
|
| 525 |
+
"learning_rate": 3.665801240444252e-05,
|
| 526 |
+
"loss": 0.1279,
|
| 527 |
+
"step": 37000
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.3522284725227174,
|
| 531 |
+
"grad_norm": 8.464075088500977,
|
| 532 |
+
"learning_rate": 3.6477715274772825e-05,
|
| 533 |
+
"loss": 0.1256,
|
| 534 |
+
"step": 37500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 1.370258185489687,
|
| 538 |
+
"grad_norm": 5.428006172180176,
|
| 539 |
+
"learning_rate": 3.629741814510313e-05,
|
| 540 |
+
"loss": 0.1322,
|
| 541 |
+
"step": 38000
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 1.3882878984566567,
|
| 545 |
+
"grad_norm": 7.391458988189697,
|
| 546 |
+
"learning_rate": 3.611712101543344e-05,
|
| 547 |
+
"loss": 0.1258,
|
| 548 |
+
"step": 38500
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.4063176114236262,
|
| 552 |
+
"grad_norm": 4.568892955780029,
|
| 553 |
+
"learning_rate": 3.593682388576374e-05,
|
| 554 |
+
"loss": 0.1237,
|
| 555 |
+
"step": 39000
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 1.4243473243905957,
|
| 559 |
+
"grad_norm": 7.763210773468018,
|
| 560 |
+
"learning_rate": 3.575652675609405e-05,
|
| 561 |
+
"loss": 0.1125,
|
| 562 |
+
"step": 39500
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 1.4423770373575653,
|
| 566 |
+
"grad_norm": 5.991769790649414,
|
| 567 |
+
"learning_rate": 3.5576229626424354e-05,
|
| 568 |
+
"loss": 0.1222,
|
| 569 |
+
"step": 40000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.4604067503245348,
|
| 573 |
+
"grad_norm": 12.884145736694336,
|
| 574 |
+
"learning_rate": 3.5395932496754654e-05,
|
| 575 |
+
"loss": 0.1259,
|
| 576 |
+
"step": 40500
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 1.4784364632915044,
|
| 580 |
+
"grad_norm": 18.740543365478516,
|
| 581 |
+
"learning_rate": 3.521563536708496e-05,
|
| 582 |
+
"loss": 0.1241,
|
| 583 |
+
"step": 41000
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 1.496466176258474,
|
| 587 |
+
"grad_norm": 14.884182929992676,
|
| 588 |
+
"learning_rate": 3.503533823741526e-05,
|
| 589 |
+
"loss": 0.1244,
|
| 590 |
+
"step": 41500
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.5144958892254436,
|
| 594 |
+
"grad_norm": 11.634753227233887,
|
| 595 |
+
"learning_rate": 3.485504110774557e-05,
|
| 596 |
+
"loss": 0.1147,
|
| 597 |
+
"step": 42000
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 1.532525602192413,
|
| 601 |
+
"grad_norm": 4.704160690307617,
|
| 602 |
+
"learning_rate": 3.467474397807587e-05,
|
| 603 |
+
"loss": 0.119,
|
| 604 |
+
"step": 42500
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.5505553151593827,
|
| 608 |
+
"grad_norm": 8.544620513916016,
|
| 609 |
+
"learning_rate": 3.4494446848406176e-05,
|
| 610 |
+
"loss": 0.117,
|
| 611 |
+
"step": 43000
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 1.5685850281263523,
|
| 615 |
+
"grad_norm": 4.399921894073486,
|
| 616 |
+
"learning_rate": 3.4314149718736476e-05,
|
| 617 |
+
"loss": 0.1108,
|
| 618 |
+
"step": 43500
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 1.5866147410933218,
|
| 622 |
+
"grad_norm": 0.47739681601524353,
|
| 623 |
+
"learning_rate": 3.4133852589066784e-05,
|
| 624 |
+
"loss": 0.1143,
|
| 625 |
+
"step": 44000
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 1.6046444540602913,
|
| 629 |
+
"grad_norm": 5.860989570617676,
|
| 630 |
+
"learning_rate": 3.395355545939709e-05,
|
| 631 |
+
"loss": 0.1106,
|
| 632 |
+
"step": 44500
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.6226741670272609,
|
| 636 |
+
"grad_norm": 5.640627861022949,
|
| 637 |
+
"learning_rate": 3.377325832972739e-05,
|
| 638 |
+
"loss": 0.1168,
|
| 639 |
+
"step": 45000
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 1.6407038799942306,
|
| 643 |
+
"grad_norm": 5.664581775665283,
|
| 644 |
+
"learning_rate": 3.35929612000577e-05,
|
| 645 |
+
"loss": 0.1142,
|
| 646 |
+
"step": 45500
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 1.6587335929612,
|
| 650 |
+
"grad_norm": 3.1675755977630615,
|
| 651 |
+
"learning_rate": 3.3412664070388e-05,
|
| 652 |
+
"loss": 0.1167,
|
| 653 |
+
"step": 46000
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 1.6767633059281697,
|
| 657 |
+
"grad_norm": 6.013214588165283,
|
| 658 |
+
"learning_rate": 3.3232366940718306e-05,
|
| 659 |
+
"loss": 0.112,
|
| 660 |
+
"step": 46500
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 1.6947930188951392,
|
| 664 |
+
"grad_norm": 4.673073768615723,
|
| 665 |
+
"learning_rate": 3.3052069811048606e-05,
|
| 666 |
+
"loss": 0.1116,
|
| 667 |
+
"step": 47000
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 1.7128227318621088,
|
| 671 |
+
"grad_norm": 0.9196059703826904,
|
| 672 |
+
"learning_rate": 3.287177268137891e-05,
|
| 673 |
+
"loss": 0.1108,
|
| 674 |
+
"step": 47500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.7308524448290783,
|
| 678 |
+
"grad_norm": 0.9425193667411804,
|
| 679 |
+
"learning_rate": 3.269147555170921e-05,
|
| 680 |
+
"loss": 0.1147,
|
| 681 |
+
"step": 48000
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 1.7488821577960478,
|
| 685 |
+
"grad_norm": 3.10310959815979,
|
| 686 |
+
"learning_rate": 3.251117842203952e-05,
|
| 687 |
+
"loss": 0.112,
|
| 688 |
+
"step": 48500
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 1.7669118707630176,
|
| 692 |
+
"grad_norm": 14.031669616699219,
|
| 693 |
+
"learning_rate": 3.233088129236983e-05,
|
| 694 |
+
"loss": 0.1055,
|
| 695 |
+
"step": 49000
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 1.784941583729987,
|
| 699 |
+
"grad_norm": 5.763240337371826,
|
| 700 |
+
"learning_rate": 3.2150584162700135e-05,
|
| 701 |
+
"loss": 0.1032,
|
| 702 |
+
"step": 49500
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 1.8029712966969567,
|
| 706 |
+
"grad_norm": 0.4475843608379364,
|
| 707 |
+
"learning_rate": 3.1970287033030435e-05,
|
| 708 |
+
"loss": 0.102,
|
| 709 |
+
"step": 50000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 1.8210010096639262,
|
| 713 |
+
"grad_norm": 7.822906017303467,
|
| 714 |
+
"learning_rate": 3.178998990336074e-05,
|
| 715 |
+
"loss": 0.1037,
|
| 716 |
+
"step": 50500
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.8390307226308957,
|
| 720 |
+
"grad_norm": 0.5396666526794434,
|
| 721 |
+
"learning_rate": 3.160969277369105e-05,
|
| 722 |
+
"loss": 0.1041,
|
| 723 |
+
"step": 51000
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 1.8570604355978653,
|
| 727 |
+
"grad_norm": 8.58728313446045,
|
| 728 |
+
"learning_rate": 3.142939564402135e-05,
|
| 729 |
+
"loss": 0.1074,
|
| 730 |
+
"step": 51500
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 1.8750901485648348,
|
| 734 |
+
"grad_norm": 3.0707802772521973,
|
| 735 |
+
"learning_rate": 3.1249098514351657e-05,
|
| 736 |
+
"loss": 0.0962,
|
| 737 |
+
"step": 52000
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 1.8931198615318046,
|
| 741 |
+
"grad_norm": 7.851358413696289,
|
| 742 |
+
"learning_rate": 3.106880138468196e-05,
|
| 743 |
+
"loss": 0.1075,
|
| 744 |
+
"step": 52500
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 1.9111495744987739,
|
| 748 |
+
"grad_norm": 2.6604301929473877,
|
| 749 |
+
"learning_rate": 3.0888504255012264e-05,
|
| 750 |
+
"loss": 0.1045,
|
| 751 |
+
"step": 53000
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 1.9291792874657436,
|
| 755 |
+
"grad_norm": 5.727674961090088,
|
| 756 |
+
"learning_rate": 3.0708207125342564e-05,
|
| 757 |
+
"loss": 0.1016,
|
| 758 |
+
"step": 53500
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.9472090004327132,
|
| 762 |
+
"grad_norm": 2.06195068359375,
|
| 763 |
+
"learning_rate": 3.052790999567287e-05,
|
| 764 |
+
"loss": 0.1008,
|
| 765 |
+
"step": 54000
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 1.9652387133996827,
|
| 769 |
+
"grad_norm": 5.421629905700684,
|
| 770 |
+
"learning_rate": 3.0347612866003172e-05,
|
| 771 |
+
"loss": 0.0988,
|
| 772 |
+
"step": 54500
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 1.9832684263666522,
|
| 776 |
+
"grad_norm": 1.0911563634872437,
|
| 777 |
+
"learning_rate": 3.016731573633348e-05,
|
| 778 |
+
"loss": 0.0973,
|
| 779 |
+
"step": 55000
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 2.0012981393336218,
|
| 783 |
+
"grad_norm": 0.9827903509140015,
|
| 784 |
+
"learning_rate": 2.9987018606663786e-05,
|
| 785 |
+
"loss": 0.1033,
|
| 786 |
+
"step": 55500
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 2.0193278523005915,
|
| 790 |
+
"grad_norm": 1.9078912734985352,
|
| 791 |
+
"learning_rate": 2.9806721476994086e-05,
|
| 792 |
+
"loss": 0.0834,
|
| 793 |
+
"step": 56000
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 2.037357565267561,
|
| 797 |
+
"grad_norm": 6.1681084632873535,
|
| 798 |
+
"learning_rate": 2.9626424347324393e-05,
|
| 799 |
+
"loss": 0.0851,
|
| 800 |
+
"step": 56500
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 2.0553872782345306,
|
| 804 |
+
"grad_norm": 7.802475929260254,
|
| 805 |
+
"learning_rate": 2.9446127217654694e-05,
|
| 806 |
+
"loss": 0.085,
|
| 807 |
+
"step": 57000
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 2.0734169912015,
|
| 811 |
+
"grad_norm": 6.640796184539795,
|
| 812 |
+
"learning_rate": 2.9265830087985e-05,
|
| 813 |
+
"loss": 0.0899,
|
| 814 |
+
"step": 57500
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 2.0914467041684697,
|
| 818 |
+
"grad_norm": 2.4332051277160645,
|
| 819 |
+
"learning_rate": 2.9085532958315305e-05,
|
| 820 |
+
"loss": 0.0841,
|
| 821 |
+
"step": 58000
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 2.109476417135439,
|
| 825 |
+
"grad_norm": 2.4671130180358887,
|
| 826 |
+
"learning_rate": 2.890523582864561e-05,
|
| 827 |
+
"loss": 0.0833,
|
| 828 |
+
"step": 58500
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 2.1275061301024087,
|
| 832 |
+
"grad_norm": 3.583301067352295,
|
| 833 |
+
"learning_rate": 2.8724938698975912e-05,
|
| 834 |
+
"loss": 0.084,
|
| 835 |
+
"step": 59000
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 2.1455358430693785,
|
| 839 |
+
"grad_norm": 2.205716133117676,
|
| 840 |
+
"learning_rate": 2.854464156930622e-05,
|
| 841 |
+
"loss": 0.0826,
|
| 842 |
+
"step": 59500
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 2.163565556036348,
|
| 846 |
+
"grad_norm": 0.8957700133323669,
|
| 847 |
+
"learning_rate": 2.836434443963652e-05,
|
| 848 |
+
"loss": 0.0839,
|
| 849 |
+
"step": 60000
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 2.1815952690033176,
|
| 853 |
+
"grad_norm": 2.5046937465667725,
|
| 854 |
+
"learning_rate": 2.8184047309966827e-05,
|
| 855 |
+
"loss": 0.0774,
|
| 856 |
+
"step": 60500
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 2.199624981970287,
|
| 860 |
+
"grad_norm": 11.110339164733887,
|
| 861 |
+
"learning_rate": 2.8003750180297134e-05,
|
| 862 |
+
"loss": 0.0823,
|
| 863 |
+
"step": 61000
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 2.2176546949372566,
|
| 867 |
+
"grad_norm": 3.2354230880737305,
|
| 868 |
+
"learning_rate": 2.7823453050627434e-05,
|
| 869 |
+
"loss": 0.0842,
|
| 870 |
+
"step": 61500
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 2.235684407904226,
|
| 874 |
+
"grad_norm": 10.448586463928223,
|
| 875 |
+
"learning_rate": 2.764315592095774e-05,
|
| 876 |
+
"loss": 0.0809,
|
| 877 |
+
"step": 62000
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 2.2537141208711957,
|
| 881 |
+
"grad_norm": 5.9859514236450195,
|
| 882 |
+
"learning_rate": 2.746285879128804e-05,
|
| 883 |
+
"loss": 0.0875,
|
| 884 |
+
"step": 62500
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 2.2717438338381655,
|
| 888 |
+
"grad_norm": 5.923473834991455,
|
| 889 |
+
"learning_rate": 2.728256166161835e-05,
|
| 890 |
+
"loss": 0.0779,
|
| 891 |
+
"step": 63000
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 2.289773546805135,
|
| 895 |
+
"grad_norm": 5.31735897064209,
|
| 896 |
+
"learning_rate": 2.7102264531948652e-05,
|
| 897 |
+
"loss": 0.0862,
|
| 898 |
+
"step": 63500
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 2.3078032597721045,
|
| 902 |
+
"grad_norm": 9.569977760314941,
|
| 903 |
+
"learning_rate": 2.692196740227896e-05,
|
| 904 |
+
"loss": 0.0827,
|
| 905 |
+
"step": 64000
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 2.325832972739074,
|
| 909 |
+
"grad_norm": 2.9633631706237793,
|
| 910 |
+
"learning_rate": 2.674167027260926e-05,
|
| 911 |
+
"loss": 0.0844,
|
| 912 |
+
"step": 64500
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 2.3438626857060436,
|
| 916 |
+
"grad_norm": 5.674577236175537,
|
| 917 |
+
"learning_rate": 2.6561373142939567e-05,
|
| 918 |
+
"loss": 0.0832,
|
| 919 |
+
"step": 65000
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 2.361892398673013,
|
| 923 |
+
"grad_norm": 6.323156356811523,
|
| 924 |
+
"learning_rate": 2.6381076013269867e-05,
|
| 925 |
+
"loss": 0.0807,
|
| 926 |
+
"step": 65500
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 2.3799221116399827,
|
| 930 |
+
"grad_norm": 6.816585540771484,
|
| 931 |
+
"learning_rate": 2.6200778883600174e-05,
|
| 932 |
+
"loss": 0.0818,
|
| 933 |
+
"step": 66000
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 2.3979518246069524,
|
| 937 |
+
"grad_norm": 2.57076358795166,
|
| 938 |
+
"learning_rate": 2.602048175393048e-05,
|
| 939 |
+
"loss": 0.0814,
|
| 940 |
+
"step": 66500
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 2.4159815375739218,
|
| 944 |
+
"grad_norm": 7.286929130554199,
|
| 945 |
+
"learning_rate": 2.584018462426078e-05,
|
| 946 |
+
"loss": 0.079,
|
| 947 |
+
"step": 67000
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 2.4340112505408915,
|
| 951 |
+
"grad_norm": 7.754148006439209,
|
| 952 |
+
"learning_rate": 2.565988749459109e-05,
|
| 953 |
+
"loss": 0.0733,
|
| 954 |
+
"step": 67500
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 2.452040963507861,
|
| 958 |
+
"grad_norm": 6.270145416259766,
|
| 959 |
+
"learning_rate": 2.547959036492139e-05,
|
| 960 |
+
"loss": 0.079,
|
| 961 |
+
"step": 68000
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 2.4700706764748306,
|
| 965 |
+
"grad_norm": 2.421312093734741,
|
| 966 |
+
"learning_rate": 2.5299293235251696e-05,
|
| 967 |
+
"loss": 0.0754,
|
| 968 |
+
"step": 68500
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 2.4881003894418,
|
| 972 |
+
"grad_norm": 3.8859853744506836,
|
| 973 |
+
"learning_rate": 2.5118996105582e-05,
|
| 974 |
+
"loss": 0.0847,
|
| 975 |
+
"step": 69000
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 2.5061301024087697,
|
| 979 |
+
"grad_norm": 5.9454345703125,
|
| 980 |
+
"learning_rate": 2.4938698975912307e-05,
|
| 981 |
+
"loss": 0.0769,
|
| 982 |
+
"step": 69500
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 2.5241598153757394,
|
| 986 |
+
"grad_norm": 0.5154119729995728,
|
| 987 |
+
"learning_rate": 2.475840184624261e-05,
|
| 988 |
+
"loss": 0.0781,
|
| 989 |
+
"step": 70000
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 2.5421895283427087,
|
| 993 |
+
"grad_norm": 4.4899821281433105,
|
| 994 |
+
"learning_rate": 2.4578104716572914e-05,
|
| 995 |
+
"loss": 0.0736,
|
| 996 |
+
"step": 70500
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 2.5602192413096785,
|
| 1000 |
+
"grad_norm": 9.669061660766602,
|
| 1001 |
+
"learning_rate": 2.4397807586903218e-05,
|
| 1002 |
+
"loss": 0.0709,
|
| 1003 |
+
"step": 71000
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 2.578248954276648,
|
| 1007 |
+
"grad_norm": 3.6468615531921387,
|
| 1008 |
+
"learning_rate": 2.4217510457233522e-05,
|
| 1009 |
+
"loss": 0.0691,
|
| 1010 |
+
"step": 71500
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 2.5962786672436176,
|
| 1014 |
+
"grad_norm": 6.391994953155518,
|
| 1015 |
+
"learning_rate": 2.4037213327563826e-05,
|
| 1016 |
+
"loss": 0.0777,
|
| 1017 |
+
"step": 72000
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 2.614308380210587,
|
| 1021 |
+
"grad_norm": 6.9082112312316895,
|
| 1022 |
+
"learning_rate": 2.385691619789413e-05,
|
| 1023 |
+
"loss": 0.0742,
|
| 1024 |
+
"step": 72500
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 2.6323380931775566,
|
| 1028 |
+
"grad_norm": 6.8455681800842285,
|
| 1029 |
+
"learning_rate": 2.3676619068224433e-05,
|
| 1030 |
+
"loss": 0.0764,
|
| 1031 |
+
"step": 73000
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 2.6503678061445264,
|
| 1035 |
+
"grad_norm": 4.3308329582214355,
|
| 1036 |
+
"learning_rate": 2.349632193855474e-05,
|
| 1037 |
+
"loss": 0.0747,
|
| 1038 |
+
"step": 73500
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 2.6683975191114957,
|
| 1042 |
+
"grad_norm": 2.893226146697998,
|
| 1043 |
+
"learning_rate": 2.3316024808885044e-05,
|
| 1044 |
+
"loss": 0.0706,
|
| 1045 |
+
"step": 74000
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 2.6864272320784655,
|
| 1049 |
+
"grad_norm": 3.5622036457061768,
|
| 1050 |
+
"learning_rate": 2.313572767921535e-05,
|
| 1051 |
+
"loss": 0.0765,
|
| 1052 |
+
"step": 74500
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 2.7044569450454348,
|
| 1056 |
+
"grad_norm": 4.554545879364014,
|
| 1057 |
+
"learning_rate": 2.2955430549545655e-05,
|
| 1058 |
+
"loss": 0.0756,
|
| 1059 |
+
"step": 75000
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 2.7224866580124045,
|
| 1063 |
+
"grad_norm": 5.788700103759766,
|
| 1064 |
+
"learning_rate": 2.277513341987596e-05,
|
| 1065 |
+
"loss": 0.0693,
|
| 1066 |
+
"step": 75500
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 2.740516370979374,
|
| 1070 |
+
"grad_norm": 0.5161257982254028,
|
| 1071 |
+
"learning_rate": 2.2594836290206262e-05,
|
| 1072 |
+
"loss": 0.0673,
|
| 1073 |
+
"step": 76000
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 2.7585460839463436,
|
| 1077 |
+
"grad_norm": 10.974600791931152,
|
| 1078 |
+
"learning_rate": 2.2414539160536566e-05,
|
| 1079 |
+
"loss": 0.0715,
|
| 1080 |
+
"step": 76500
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 2.7765757969133134,
|
| 1084 |
+
"grad_norm": 1.995335340499878,
|
| 1085 |
+
"learning_rate": 2.223424203086687e-05,
|
| 1086 |
+
"loss": 0.0703,
|
| 1087 |
+
"step": 77000
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 2.7946055098802827,
|
| 1091 |
+
"grad_norm": 1.148324966430664,
|
| 1092 |
+
"learning_rate": 2.2053944901197173e-05,
|
| 1093 |
+
"loss": 0.0667,
|
| 1094 |
+
"step": 77500
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 2.8126352228472524,
|
| 1098 |
+
"grad_norm": 4.751012802124023,
|
| 1099 |
+
"learning_rate": 2.1873647771527477e-05,
|
| 1100 |
+
"loss": 0.0631,
|
| 1101 |
+
"step": 78000
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 2.8306649358142217,
|
| 1105 |
+
"grad_norm": 1.9793184995651245,
|
| 1106 |
+
"learning_rate": 2.169335064185778e-05,
|
| 1107 |
+
"loss": 0.0685,
|
| 1108 |
+
"step": 78500
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 2.8486946487811915,
|
| 1112 |
+
"grad_norm": 11.45518970489502,
|
| 1113 |
+
"learning_rate": 2.1513053512188088e-05,
|
| 1114 |
+
"loss": 0.0653,
|
| 1115 |
+
"step": 79000
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 2.866724361748161,
|
| 1119 |
+
"grad_norm": 3.078552484512329,
|
| 1120 |
+
"learning_rate": 2.133275638251839e-05,
|
| 1121 |
+
"loss": 0.0706,
|
| 1122 |
+
"step": 79500
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 2.8847540747151306,
|
| 1126 |
+
"grad_norm": 3.5840249061584473,
|
| 1127 |
+
"learning_rate": 2.11524592528487e-05,
|
| 1128 |
+
"loss": 0.0662,
|
| 1129 |
+
"step": 80000
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 2.9027837876821003,
|
| 1133 |
+
"grad_norm": 4.622115135192871,
|
| 1134 |
+
"learning_rate": 2.0972162123179002e-05,
|
| 1135 |
+
"loss": 0.061,
|
| 1136 |
+
"step": 80500
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 2.9208135006490696,
|
| 1140 |
+
"grad_norm": 2.47527813911438,
|
| 1141 |
+
"learning_rate": 2.0791864993509306e-05,
|
| 1142 |
+
"loss": 0.0662,
|
| 1143 |
+
"step": 81000
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 2.9388432136160394,
|
| 1147 |
+
"grad_norm": 2.809105634689331,
|
| 1148 |
+
"learning_rate": 2.061156786383961e-05,
|
| 1149 |
+
"loss": 0.0633,
|
| 1150 |
+
"step": 81500
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 2.9568729265830087,
|
| 1154 |
+
"grad_norm": 1.644235372543335,
|
| 1155 |
+
"learning_rate": 2.0431270734169913e-05,
|
| 1156 |
+
"loss": 0.062,
|
| 1157 |
+
"step": 82000
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 2.9749026395499785,
|
| 1161 |
+
"grad_norm": 9.02820110321045,
|
| 1162 |
+
"learning_rate": 2.0250973604500217e-05,
|
| 1163 |
+
"loss": 0.0635,
|
| 1164 |
+
"step": 82500
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 2.992932352516948,
|
| 1168 |
+
"grad_norm": 5.800693988800049,
|
| 1169 |
+
"learning_rate": 2.007067647483052e-05,
|
| 1170 |
+
"loss": 0.0651,
|
| 1171 |
+
"step": 83000
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 3.0109620654839175,
|
| 1175 |
+
"grad_norm": 2.655238151550293,
|
| 1176 |
+
"learning_rate": 1.9890379345160825e-05,
|
| 1177 |
+
"loss": 0.0579,
|
| 1178 |
+
"step": 83500
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 3.028991778450887,
|
| 1182 |
+
"grad_norm": 3.668182373046875,
|
| 1183 |
+
"learning_rate": 1.9710082215491128e-05,
|
| 1184 |
+
"loss": 0.056,
|
| 1185 |
+
"step": 84000
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 3.0470214914178566,
|
| 1189 |
+
"grad_norm": 4.268499374389648,
|
| 1190 |
+
"learning_rate": 1.9529785085821435e-05,
|
| 1191 |
+
"loss": 0.058,
|
| 1192 |
+
"step": 84500
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 3.0650512043848264,
|
| 1196 |
+
"grad_norm": 10.706989288330078,
|
| 1197 |
+
"learning_rate": 1.934948795615174e-05,
|
| 1198 |
+
"loss": 0.0567,
|
| 1199 |
+
"step": 85000
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 3.0830809173517957,
|
| 1203 |
+
"grad_norm": 2.8376147747039795,
|
| 1204 |
+
"learning_rate": 1.9169190826482046e-05,
|
| 1205 |
+
"loss": 0.052,
|
| 1206 |
+
"step": 85500
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 3.1011106303187654,
|
| 1210 |
+
"grad_norm": 1.0124831199645996,
|
| 1211 |
+
"learning_rate": 1.898889369681235e-05,
|
| 1212 |
+
"loss": 0.0569,
|
| 1213 |
+
"step": 86000
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 3.1191403432857348,
|
| 1217 |
+
"grad_norm": 1.6948038339614868,
|
| 1218 |
+
"learning_rate": 1.8808596567142654e-05,
|
| 1219 |
+
"loss": 0.0556,
|
| 1220 |
+
"step": 86500
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 3.1371700562527045,
|
| 1224 |
+
"grad_norm": 4.397452354431152,
|
| 1225 |
+
"learning_rate": 1.8628299437472957e-05,
|
| 1226 |
+
"loss": 0.0548,
|
| 1227 |
+
"step": 87000
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 3.155199769219674,
|
| 1231 |
+
"grad_norm": 0.15065419673919678,
|
| 1232 |
+
"learning_rate": 1.844800230780326e-05,
|
| 1233 |
+
"loss": 0.0509,
|
| 1234 |
+
"step": 87500
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 3.1732294821866436,
|
| 1238 |
+
"grad_norm": 7.567072868347168,
|
| 1239 |
+
"learning_rate": 1.8267705178133565e-05,
|
| 1240 |
+
"loss": 0.056,
|
| 1241 |
+
"step": 88000
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 3.1912591951536133,
|
| 1245 |
+
"grad_norm": 0.5679315328598022,
|
| 1246 |
+
"learning_rate": 1.808740804846387e-05,
|
| 1247 |
+
"loss": 0.0494,
|
| 1248 |
+
"step": 88500
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 3.2092889081205827,
|
| 1252 |
+
"grad_norm": 3.4336838722229004,
|
| 1253 |
+
"learning_rate": 1.7907110918794172e-05,
|
| 1254 |
+
"loss": 0.0534,
|
| 1255 |
+
"step": 89000
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 3.2273186210875524,
|
| 1259 |
+
"grad_norm": 5.46131706237793,
|
| 1260 |
+
"learning_rate": 1.7726813789124476e-05,
|
| 1261 |
+
"loss": 0.0516,
|
| 1262 |
+
"step": 89500
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 3.2453483340545217,
|
| 1266 |
+
"grad_norm": 4.545089244842529,
|
| 1267 |
+
"learning_rate": 1.7546516659454783e-05,
|
| 1268 |
+
"loss": 0.0558,
|
| 1269 |
+
"step": 90000
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 3.2633780470214915,
|
| 1273 |
+
"grad_norm": 0.6450003385543823,
|
| 1274 |
+
"learning_rate": 1.7366219529785087e-05,
|
| 1275 |
+
"loss": 0.0523,
|
| 1276 |
+
"step": 90500
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 3.281407759988461,
|
| 1280 |
+
"grad_norm": 1.9566916227340698,
|
| 1281 |
+
"learning_rate": 1.7185922400115394e-05,
|
| 1282 |
+
"loss": 0.0489,
|
| 1283 |
+
"step": 91000
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 3.2994374729554306,
|
| 1287 |
+
"grad_norm": 2.5970304012298584,
|
| 1288 |
+
"learning_rate": 1.7005625270445698e-05,
|
| 1289 |
+
"loss": 0.054,
|
| 1290 |
+
"step": 91500
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 3.3174671859224003,
|
| 1294 |
+
"grad_norm": 5.6770501136779785,
|
| 1295 |
+
"learning_rate": 1.6825328140776e-05,
|
| 1296 |
+
"loss": 0.0453,
|
| 1297 |
+
"step": 92000
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 3.3354968988893696,
|
| 1301 |
+
"grad_norm": 5.3049139976501465,
|
| 1302 |
+
"learning_rate": 1.6645031011106305e-05,
|
| 1303 |
+
"loss": 0.0543,
|
| 1304 |
+
"step": 92500
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 3.3535266118563394,
|
| 1308 |
+
"grad_norm": 1.5339233875274658,
|
| 1309 |
+
"learning_rate": 1.646473388143661e-05,
|
| 1310 |
+
"loss": 0.05,
|
| 1311 |
+
"step": 93000
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 3.3715563248233087,
|
| 1315 |
+
"grad_norm": 6.024913311004639,
|
| 1316 |
+
"learning_rate": 1.6284436751766912e-05,
|
| 1317 |
+
"loss": 0.0507,
|
| 1318 |
+
"step": 93500
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 3.3895860377902784,
|
| 1322 |
+
"grad_norm": 3.0228421688079834,
|
| 1323 |
+
"learning_rate": 1.6104139622097216e-05,
|
| 1324 |
+
"loss": 0.0483,
|
| 1325 |
+
"step": 94000
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 3.4076157507572478,
|
| 1329 |
+
"grad_norm": 2.426029682159424,
|
| 1330 |
+
"learning_rate": 1.592384249242752e-05,
|
| 1331 |
+
"loss": 0.0464,
|
| 1332 |
+
"step": 94500
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 3.4256454637242175,
|
| 1336 |
+
"grad_norm": 1.8258663415908813,
|
| 1337 |
+
"learning_rate": 1.5743545362757824e-05,
|
| 1338 |
+
"loss": 0.0486,
|
| 1339 |
+
"step": 95000
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 3.443675176691187,
|
| 1343 |
+
"grad_norm": 0.862032413482666,
|
| 1344 |
+
"learning_rate": 1.556324823308813e-05,
|
| 1345 |
+
"loss": 0.0512,
|
| 1346 |
+
"step": 95500
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 3.4617048896581566,
|
| 1350 |
+
"grad_norm": 4.423261642456055,
|
| 1351 |
+
"learning_rate": 1.5382951103418434e-05,
|
| 1352 |
+
"loss": 0.0508,
|
| 1353 |
+
"step": 96000
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 3.4797346026251263,
|
| 1357 |
+
"grad_norm": 4.624811172485352,
|
| 1358 |
+
"learning_rate": 1.520265397374874e-05,
|
| 1359 |
+
"loss": 0.0476,
|
| 1360 |
+
"step": 96500
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 3.4977643155920957,
|
| 1364 |
+
"grad_norm": 1.2744200229644775,
|
| 1365 |
+
"learning_rate": 1.5022356844079044e-05,
|
| 1366 |
+
"loss": 0.051,
|
| 1367 |
+
"step": 97000
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 3.5157940285590654,
|
| 1371 |
+
"grad_norm": 4.255450248718262,
|
| 1372 |
+
"learning_rate": 1.4842059714409349e-05,
|
| 1373 |
+
"loss": 0.0491,
|
| 1374 |
+
"step": 97500
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 3.533823741526035,
|
| 1378 |
+
"grad_norm": 0.2981502115726471,
|
| 1379 |
+
"learning_rate": 1.4661762584739653e-05,
|
| 1380 |
+
"loss": 0.0496,
|
| 1381 |
+
"step": 98000
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 3.5518534544930045,
|
| 1385 |
+
"grad_norm": 6.509685039520264,
|
| 1386 |
+
"learning_rate": 1.4481465455069956e-05,
|
| 1387 |
+
"loss": 0.0491,
|
| 1388 |
+
"step": 98500
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 3.569883167459974,
|
| 1392 |
+
"grad_norm": 3.7302205562591553,
|
| 1393 |
+
"learning_rate": 1.430116832540026e-05,
|
| 1394 |
+
"loss": 0.0485,
|
| 1395 |
+
"step": 99000
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 3.5879128804269436,
|
| 1399 |
+
"grad_norm": 3.449720621109009,
|
| 1400 |
+
"learning_rate": 1.4120871195730564e-05,
|
| 1401 |
+
"loss": 0.0462,
|
| 1402 |
+
"step": 99500
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 3.6059425933939133,
|
| 1406 |
+
"grad_norm": 0.168724924325943,
|
| 1407 |
+
"learning_rate": 1.394057406606087e-05,
|
| 1408 |
+
"loss": 0.0454,
|
| 1409 |
+
"step": 100000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 3.6239723063608826,
|
| 1413 |
+
"grad_norm": 2.3195059299468994,
|
| 1414 |
+
"learning_rate": 1.3760276936391173e-05,
|
| 1415 |
+
"loss": 0.0492,
|
| 1416 |
+
"step": 100500
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 3.6420020193278524,
|
| 1420 |
+
"grad_norm": 1.8226135969161987,
|
| 1421 |
+
"learning_rate": 1.3579979806721477e-05,
|
| 1422 |
+
"loss": 0.0445,
|
| 1423 |
+
"step": 101000
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 3.6600317322948217,
|
| 1427 |
+
"grad_norm": 4.066940784454346,
|
| 1428 |
+
"learning_rate": 1.339968267705178e-05,
|
| 1429 |
+
"loss": 0.0479,
|
| 1430 |
+
"step": 101500
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 3.6780614452617915,
|
| 1434 |
+
"grad_norm": 5.052645206451416,
|
| 1435 |
+
"learning_rate": 1.3219385547382087e-05,
|
| 1436 |
+
"loss": 0.0428,
|
| 1437 |
+
"step": 102000
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 3.6960911582287608,
|
| 1441 |
+
"grad_norm": 8.159462928771973,
|
| 1442 |
+
"learning_rate": 1.3039088417712391e-05,
|
| 1443 |
+
"loss": 0.0474,
|
| 1444 |
+
"step": 102500
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 3.7141208711957305,
|
| 1448 |
+
"grad_norm": 2.3855013847351074,
|
| 1449 |
+
"learning_rate": 1.2858791288042697e-05,
|
| 1450 |
+
"loss": 0.0435,
|
| 1451 |
+
"step": 103000
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 3.7321505841627003,
|
| 1455 |
+
"grad_norm": 4.628015518188477,
|
| 1456 |
+
"learning_rate": 1.2678494158373e-05,
|
| 1457 |
+
"loss": 0.0417,
|
| 1458 |
+
"step": 103500
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 3.7501802971296696,
|
| 1462 |
+
"grad_norm": 4.056214332580566,
|
| 1463 |
+
"learning_rate": 1.2498197028703304e-05,
|
| 1464 |
+
"loss": 0.0432,
|
| 1465 |
+
"step": 104000
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 3.7682100100966394,
|
| 1469 |
+
"grad_norm": 0.9179813861846924,
|
| 1470 |
+
"learning_rate": 1.2317899899033608e-05,
|
| 1471 |
+
"loss": 0.042,
|
| 1472 |
+
"step": 104500
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 3.7862397230636087,
|
| 1476 |
+
"grad_norm": 4.917830944061279,
|
| 1477 |
+
"learning_rate": 1.2137602769363911e-05,
|
| 1478 |
+
"loss": 0.0409,
|
| 1479 |
+
"step": 105000
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 3.8042694360305784,
|
| 1483 |
+
"grad_norm": 0.4413645565509796,
|
| 1484 |
+
"learning_rate": 1.1957305639694217e-05,
|
| 1485 |
+
"loss": 0.0463,
|
| 1486 |
+
"step": 105500
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 3.8222991489975477,
|
| 1490 |
+
"grad_norm": 4.744266033172607,
|
| 1491 |
+
"learning_rate": 1.177700851002452e-05,
|
| 1492 |
+
"loss": 0.0436,
|
| 1493 |
+
"step": 106000
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 3.8403288619645175,
|
| 1497 |
+
"grad_norm": 0.22985690832138062,
|
| 1498 |
+
"learning_rate": 1.1596711380354826e-05,
|
| 1499 |
+
"loss": 0.04,
|
| 1500 |
+
"step": 106500
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 3.8583585749314873,
|
| 1504 |
+
"grad_norm": 5.225405216217041,
|
| 1505 |
+
"learning_rate": 1.141641425068513e-05,
|
| 1506 |
+
"loss": 0.0412,
|
| 1507 |
+
"step": 107000
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 3.8763882878984566,
|
| 1511 |
+
"grad_norm": 1.378030776977539,
|
| 1512 |
+
"learning_rate": 1.1236117121015433e-05,
|
| 1513 |
+
"loss": 0.041,
|
| 1514 |
+
"step": 107500
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 3.8944180008654263,
|
| 1518 |
+
"grad_norm": 0.7432510852813721,
|
| 1519 |
+
"learning_rate": 1.1055819991345739e-05,
|
| 1520 |
+
"loss": 0.044,
|
| 1521 |
+
"step": 108000
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 3.9124477138323956,
|
| 1525 |
+
"grad_norm": 0.2676824927330017,
|
| 1526 |
+
"learning_rate": 1.0875522861676043e-05,
|
| 1527 |
+
"loss": 0.0375,
|
| 1528 |
+
"step": 108500
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 3.9304774267993654,
|
| 1532 |
+
"grad_norm": 0.13629376888275146,
|
| 1533 |
+
"learning_rate": 1.0695225732006348e-05,
|
| 1534 |
+
"loss": 0.0395,
|
| 1535 |
+
"step": 109000
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 3.9485071397663347,
|
| 1539 |
+
"grad_norm": 4.708954811096191,
|
| 1540 |
+
"learning_rate": 1.0514928602336652e-05,
|
| 1541 |
+
"loss": 0.0368,
|
| 1542 |
+
"step": 109500
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 3.9665368527333045,
|
| 1546 |
+
"grad_norm": 4.771941184997559,
|
| 1547 |
+
"learning_rate": 1.0334631472666955e-05,
|
| 1548 |
+
"loss": 0.0432,
|
| 1549 |
+
"step": 110000
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 3.9845665657002742,
|
| 1553 |
+
"grad_norm": 10.987969398498535,
|
| 1554 |
+
"learning_rate": 1.0154334342997259e-05,
|
| 1555 |
+
"loss": 0.0393,
|
| 1556 |
+
"step": 110500
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 4.0025962786672435,
|
| 1560 |
+
"grad_norm": 0.05221960321068764,
|
| 1561 |
+
"learning_rate": 9.974037213327564e-06,
|
| 1562 |
+
"loss": 0.0413,
|
| 1563 |
+
"step": 111000
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 4.020625991634213,
|
| 1567 |
+
"grad_norm": 2.2666168212890625,
|
| 1568 |
+
"learning_rate": 9.793740083657868e-06,
|
| 1569 |
+
"loss": 0.0327,
|
| 1570 |
+
"step": 111500
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 4.038655704601183,
|
| 1574 |
+
"grad_norm": 0.029105104506015778,
|
| 1575 |
+
"learning_rate": 9.613442953988174e-06,
|
| 1576 |
+
"loss": 0.0335,
|
| 1577 |
+
"step": 112000
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 4.056685417568152,
|
| 1581 |
+
"grad_norm": 5.034819602966309,
|
| 1582 |
+
"learning_rate": 9.433145824318477e-06,
|
| 1583 |
+
"loss": 0.0332,
|
| 1584 |
+
"step": 112500
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 4.074715130535122,
|
| 1588 |
+
"grad_norm": 2.1591038703918457,
|
| 1589 |
+
"learning_rate": 9.252848694648781e-06,
|
| 1590 |
+
"loss": 0.036,
|
| 1591 |
+
"step": 113000
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 4.092744843502091,
|
| 1595 |
+
"grad_norm": 0.35260650515556335,
|
| 1596 |
+
"learning_rate": 9.072551564979086e-06,
|
| 1597 |
+
"loss": 0.036,
|
| 1598 |
+
"step": 113500
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 4.110774556469061,
|
| 1602 |
+
"grad_norm": 3.1788244247436523,
|
| 1603 |
+
"learning_rate": 8.89225443530939e-06,
|
| 1604 |
+
"loss": 0.035,
|
| 1605 |
+
"step": 114000
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 4.1288042694360305,
|
| 1609 |
+
"grad_norm": 0.9687299132347107,
|
| 1610 |
+
"learning_rate": 8.711957305639696e-06,
|
| 1611 |
+
"loss": 0.0337,
|
| 1612 |
+
"step": 114500
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 4.146833982403,
|
| 1616 |
+
"grad_norm": 3.0961289405822754,
|
| 1617 |
+
"learning_rate": 8.53166017597e-06,
|
| 1618 |
+
"loss": 0.0363,
|
| 1619 |
+
"step": 115000
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 4.16486369536997,
|
| 1623 |
+
"grad_norm": 2.062838554382324,
|
| 1624 |
+
"learning_rate": 8.351363046300303e-06,
|
| 1625 |
+
"loss": 0.0353,
|
| 1626 |
+
"step": 115500
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 4.182893408336939,
|
| 1630 |
+
"grad_norm": 0.016814429312944412,
|
| 1631 |
+
"learning_rate": 8.171065916630607e-06,
|
| 1632 |
+
"loss": 0.0344,
|
| 1633 |
+
"step": 116000
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 4.200923121303909,
|
| 1637 |
+
"grad_norm": 1.94704270362854,
|
| 1638 |
+
"learning_rate": 7.990768786960912e-06,
|
| 1639 |
+
"loss": 0.0343,
|
| 1640 |
+
"step": 116500
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 4.218952834270878,
|
| 1644 |
+
"grad_norm": 2.006126642227173,
|
| 1645 |
+
"learning_rate": 7.810471657291216e-06,
|
| 1646 |
+
"loss": 0.0343,
|
| 1647 |
+
"step": 117000
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 4.236982547237848,
|
| 1651 |
+
"grad_norm": 0.010784619487822056,
|
| 1652 |
+
"learning_rate": 7.630174527621521e-06,
|
| 1653 |
+
"loss": 0.0325,
|
| 1654 |
+
"step": 117500
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 4.2550122602048175,
|
| 1658 |
+
"grad_norm": 2.098203182220459,
|
| 1659 |
+
"learning_rate": 7.449877397951825e-06,
|
| 1660 |
+
"loss": 0.0335,
|
| 1661 |
+
"step": 118000
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 4.273041973171787,
|
| 1665 |
+
"grad_norm": 0.4809391498565674,
|
| 1666 |
+
"learning_rate": 7.2695802682821296e-06,
|
| 1667 |
+
"loss": 0.0337,
|
| 1668 |
+
"step": 118500
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 4.291071686138757,
|
| 1672 |
+
"grad_norm": 0.22787824273109436,
|
| 1673 |
+
"learning_rate": 7.089283138612433e-06,
|
| 1674 |
+
"loss": 0.0319,
|
| 1675 |
+
"step": 119000
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 4.309101399105726,
|
| 1679 |
+
"grad_norm": 1.1637790203094482,
|
| 1680 |
+
"learning_rate": 6.908986008942738e-06,
|
| 1681 |
+
"loss": 0.0341,
|
| 1682 |
+
"step": 119500
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 4.327131112072696,
|
| 1686 |
+
"grad_norm": 2.828528881072998,
|
| 1687 |
+
"learning_rate": 6.728688879273043e-06,
|
| 1688 |
+
"loss": 0.0323,
|
| 1689 |
+
"step": 120000
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 4.345160825039665,
|
| 1693 |
+
"grad_norm": 1.3863029479980469,
|
| 1694 |
+
"learning_rate": 6.548391749603347e-06,
|
| 1695 |
+
"loss": 0.0343,
|
| 1696 |
+
"step": 120500
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 4.363190538006635,
|
| 1700 |
+
"grad_norm": 4.092630863189697,
|
| 1701 |
+
"learning_rate": 6.368094619933651e-06,
|
| 1702 |
+
"loss": 0.0334,
|
| 1703 |
+
"step": 121000
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 4.3812202509736045,
|
| 1707 |
+
"grad_norm": 0.7305290102958679,
|
| 1708 |
+
"learning_rate": 6.187797490263955e-06,
|
| 1709 |
+
"loss": 0.032,
|
| 1710 |
+
"step": 121500
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 4.399249963940574,
|
| 1714 |
+
"grad_norm": 2.802487850189209,
|
| 1715 |
+
"learning_rate": 6.00750036059426e-06,
|
| 1716 |
+
"loss": 0.0342,
|
| 1717 |
+
"step": 122000
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 4.417279676907544,
|
| 1721 |
+
"grad_norm": 2.806305170059204,
|
| 1722 |
+
"learning_rate": 5.827203230924564e-06,
|
| 1723 |
+
"loss": 0.0289,
|
| 1724 |
+
"step": 122500
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 4.435309389874513,
|
| 1728 |
+
"grad_norm": 4.486428260803223,
|
| 1729 |
+
"learning_rate": 5.646906101254868e-06,
|
| 1730 |
+
"loss": 0.0276,
|
| 1731 |
+
"step": 123000
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 4.453339102841483,
|
| 1735 |
+
"grad_norm": 2.269925594329834,
|
| 1736 |
+
"learning_rate": 5.466608971585173e-06,
|
| 1737 |
+
"loss": 0.032,
|
| 1738 |
+
"step": 123500
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 4.471368815808452,
|
| 1742 |
+
"grad_norm": 2.5022552013397217,
|
| 1743 |
+
"learning_rate": 5.286311841915477e-06,
|
| 1744 |
+
"loss": 0.0303,
|
| 1745 |
+
"step": 124000
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 4.489398528775422,
|
| 1749 |
+
"grad_norm": 3.7070631980895996,
|
| 1750 |
+
"learning_rate": 5.106014712245781e-06,
|
| 1751 |
+
"loss": 0.0309,
|
| 1752 |
+
"step": 124500
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 4.507428241742391,
|
| 1756 |
+
"grad_norm": 5.330507278442383,
|
| 1757 |
+
"learning_rate": 4.9257175825760855e-06,
|
| 1758 |
+
"loss": 0.0282,
|
| 1759 |
+
"step": 125000
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 4.525457954709361,
|
| 1763 |
+
"grad_norm": 0.08433328568935394,
|
| 1764 |
+
"learning_rate": 4.74542045290639e-06,
|
| 1765 |
+
"loss": 0.0303,
|
| 1766 |
+
"step": 125500
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 4.543487667676331,
|
| 1770 |
+
"grad_norm": 2.25901198387146,
|
| 1771 |
+
"learning_rate": 4.565123323236694e-06,
|
| 1772 |
+
"loss": 0.0318,
|
| 1773 |
+
"step": 126000
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 4.5615173806433,
|
| 1777 |
+
"grad_norm": 1.9923254251480103,
|
| 1778 |
+
"learning_rate": 4.384826193566999e-06,
|
| 1779 |
+
"loss": 0.0305,
|
| 1780 |
+
"step": 126500
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 4.57954709361027,
|
| 1784 |
+
"grad_norm": 0.01763979159295559,
|
| 1785 |
+
"learning_rate": 4.204529063897303e-06,
|
| 1786 |
+
"loss": 0.0288,
|
| 1787 |
+
"step": 127000
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 4.597576806577239,
|
| 1791 |
+
"grad_norm": 1.686081051826477,
|
| 1792 |
+
"learning_rate": 4.0242319342276074e-06,
|
| 1793 |
+
"loss": 0.0307,
|
| 1794 |
+
"step": 127500
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 4.615606519544209,
|
| 1798 |
+
"grad_norm": 2.052227020263672,
|
| 1799 |
+
"learning_rate": 3.843934804557912e-06,
|
| 1800 |
+
"loss": 0.0276,
|
| 1801 |
+
"step": 128000
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 4.633636232511178,
|
| 1805 |
+
"grad_norm": 0.24964001774787903,
|
| 1806 |
+
"learning_rate": 3.6636376748882157e-06,
|
| 1807 |
+
"loss": 0.0288,
|
| 1808 |
+
"step": 128500
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 4.651665945478148,
|
| 1812 |
+
"grad_norm": 0.9556944370269775,
|
| 1813 |
+
"learning_rate": 3.4833405452185207e-06,
|
| 1814 |
+
"loss": 0.0273,
|
| 1815 |
+
"step": 129000
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 4.669695658445118,
|
| 1819 |
+
"grad_norm": 6.254054546356201,
|
| 1820 |
+
"learning_rate": 3.3030434155488244e-06,
|
| 1821 |
+
"loss": 0.0265,
|
| 1822 |
+
"step": 129500
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 4.687725371412087,
|
| 1826 |
+
"grad_norm": 0.08789811283349991,
|
| 1827 |
+
"learning_rate": 3.122746285879129e-06,
|
| 1828 |
+
"loss": 0.0303,
|
| 1829 |
+
"step": 130000
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 4.7057550843790565,
|
| 1833 |
+
"grad_norm": 2.2659811973571777,
|
| 1834 |
+
"learning_rate": 2.942449156209433e-06,
|
| 1835 |
+
"loss": 0.0295,
|
| 1836 |
+
"step": 130500
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 4.723784797346026,
|
| 1840 |
+
"grad_norm": 0.11348801106214523,
|
| 1841 |
+
"learning_rate": 2.7621520265397377e-06,
|
| 1842 |
+
"loss": 0.0261,
|
| 1843 |
+
"step": 131000
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 4.741814510312996,
|
| 1847 |
+
"grad_norm": 0.13747121393680573,
|
| 1848 |
+
"learning_rate": 2.581854896870042e-06,
|
| 1849 |
+
"loss": 0.0269,
|
| 1850 |
+
"step": 131500
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 4.759844223279965,
|
| 1854 |
+
"grad_norm": 0.1785968840122223,
|
| 1855 |
+
"learning_rate": 2.4015577672003464e-06,
|
| 1856 |
+
"loss": 0.0283,
|
| 1857 |
+
"step": 132000
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 4.777873936246935,
|
| 1861 |
+
"grad_norm": 3.6021809577941895,
|
| 1862 |
+
"learning_rate": 2.2212606375306505e-06,
|
| 1863 |
+
"loss": 0.0284,
|
| 1864 |
+
"step": 132500
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 4.795903649213905,
|
| 1868 |
+
"grad_norm": 1.1172057390213013,
|
| 1869 |
+
"learning_rate": 2.040963507860955e-06,
|
| 1870 |
+
"loss": 0.0273,
|
| 1871 |
+
"step": 133000
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 4.813933362180874,
|
| 1875 |
+
"grad_norm": 6.031317234039307,
|
| 1876 |
+
"learning_rate": 1.8606663781912594e-06,
|
| 1877 |
+
"loss": 0.0292,
|
| 1878 |
+
"step": 133500
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 4.8319630751478435,
|
| 1882 |
+
"grad_norm": 0.16859100759029388,
|
| 1883 |
+
"learning_rate": 1.6803692485215638e-06,
|
| 1884 |
+
"loss": 0.0276,
|
| 1885 |
+
"step": 134000
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 4.849992788114813,
|
| 1889 |
+
"grad_norm": 0.5369877219200134,
|
| 1890 |
+
"learning_rate": 1.500072118851868e-06,
|
| 1891 |
+
"loss": 0.029,
|
| 1892 |
+
"step": 134500
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 4.868022501081783,
|
| 1896 |
+
"grad_norm": 3.227905750274658,
|
| 1897 |
+
"learning_rate": 1.3197749891821723e-06,
|
| 1898 |
+
"loss": 0.0266,
|
| 1899 |
+
"step": 135000
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 4.886052214048752,
|
| 1903 |
+
"grad_norm": 2.0745456218719482,
|
| 1904 |
+
"learning_rate": 1.1394778595124766e-06,
|
| 1905 |
+
"loss": 0.03,
|
| 1906 |
+
"step": 135500
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 4.904081927015722,
|
| 1910 |
+
"grad_norm": 2.4073119163513184,
|
| 1911 |
+
"learning_rate": 9.59180729842781e-07,
|
| 1912 |
+
"loss": 0.0291,
|
| 1913 |
+
"step": 136000
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 4.922111639982692,
|
| 1917 |
+
"grad_norm": 0.05728856474161148,
|
| 1918 |
+
"learning_rate": 7.788836001730853e-07,
|
| 1919 |
+
"loss": 0.0266,
|
| 1920 |
+
"step": 136500
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 4.940141352949661,
|
| 1924 |
+
"grad_norm": 1.301661491394043,
|
| 1925 |
+
"learning_rate": 5.985864705033896e-07,
|
| 1926 |
+
"loss": 0.0261,
|
| 1927 |
+
"step": 137000
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 4.9581710659166305,
|
| 1931 |
+
"grad_norm": 1.6154828071594238,
|
| 1932 |
+
"learning_rate": 4.1828934083369394e-07,
|
| 1933 |
+
"loss": 0.0267,
|
| 1934 |
+
"step": 137500
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 4.9762007788836,
|
| 1938 |
+
"grad_norm": 2.9492568969726562,
|
| 1939 |
+
"learning_rate": 2.3799221116399829e-07,
|
| 1940 |
+
"loss": 0.0294,
|
| 1941 |
+
"step": 138000
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 4.99423049185057,
|
| 1945 |
+
"grad_norm": 0.5173937082290649,
|
| 1946 |
+
"learning_rate": 5.769508149430262e-08,
|
| 1947 |
+
"loss": 0.0267,
|
| 1948 |
+
"step": 138500
|
| 1949 |
+
}
|
| 1950 |
+
],
|
| 1951 |
+
"logging_steps": 500,
|
| 1952 |
+
"max_steps": 138660,
|
| 1953 |
+
"num_input_tokens_seen": 0,
|
| 1954 |
+
"num_train_epochs": 5,
|
| 1955 |
+
"save_steps": 500,
|
| 1956 |
+
"stateful_callbacks": {
|
| 1957 |
+
"TrainerControl": {
|
| 1958 |
+
"args": {
|
| 1959 |
+
"should_epoch_stop": false,
|
| 1960 |
+
"should_evaluate": false,
|
| 1961 |
+
"should_log": false,
|
| 1962 |
+
"should_save": true,
|
| 1963 |
+
"should_training_stop": true
|
| 1964 |
+
},
|
| 1965 |
+
"attributes": {}
|
| 1966 |
+
}
|
| 1967 |
+
},
|
| 1968 |
+
"total_flos": 7.981400913961166e+17,
|
| 1969 |
+
"train_batch_size": 32,
|
| 1970 |
+
"trial_name": null,
|
| 1971 |
+
"trial_params": null
|
| 1972 |
+
}
|