Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1398 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1398 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"eval_steps": 80000000000,
|
| 6 |
+
"global_step": 757,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0,
|
| 13 |
+
"grad_norm": 39.25,
|
| 14 |
+
"learning_rate": 6.578947368421053e-08,
|
| 15 |
+
"logits/chosen": 1.1063932180404663,
|
| 16 |
+
"logits/rejected": 1.991615653038025,
|
| 17 |
+
"logps/chosen": -494.3671569824219,
|
| 18 |
+
"logps/rejected": -1003.34521484375,
|
| 19 |
+
"loss": 0.6931,
|
| 20 |
+
"rewards/accuracies": 0.0,
|
| 21 |
+
"rewards/chosen": 0.0,
|
| 22 |
+
"rewards/diff": 0.0,
|
| 23 |
+
"rewards/diff_abs": 0.0,
|
| 24 |
+
"rewards/rejected": 0.0,
|
| 25 |
+
"rewards/student_margin": 0.0,
|
| 26 |
+
"rewards/teacher_margin": 0.0,
|
| 27 |
+
"step": 1
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.01,
|
| 31 |
+
"grad_norm": 37.5,
|
| 32 |
+
"learning_rate": 6.578947368421053e-07,
|
| 33 |
+
"logits/chosen": 2.408324956893921,
|
| 34 |
+
"logits/rejected": 2.1886119842529297,
|
| 35 |
+
"logps/chosen": -703.4932861328125,
|
| 36 |
+
"logps/rejected": -911.1801147460938,
|
| 37 |
+
"loss": 0.7614,
|
| 38 |
+
"rewards/accuracies": 0.48148149251937866,
|
| 39 |
+
"rewards/chosen": -0.01438586413860321,
|
| 40 |
+
"rewards/diff": 0.0,
|
| 41 |
+
"rewards/diff_abs": 0.0,
|
| 42 |
+
"rewards/rejected": -0.03394882380962372,
|
| 43 |
+
"rewards/student_margin": 0.019562965258955956,
|
| 44 |
+
"rewards/teacher_margin": 0.0,
|
| 45 |
+
"step": 10
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.03,
|
| 49 |
+
"grad_norm": 38.0,
|
| 50 |
+
"learning_rate": 1.3157894736842106e-06,
|
| 51 |
+
"logits/chosen": 2.512688159942627,
|
| 52 |
+
"logits/rejected": 2.2939486503601074,
|
| 53 |
+
"logps/chosen": -730.9832763671875,
|
| 54 |
+
"logps/rejected": -981.3518676757812,
|
| 55 |
+
"loss": 0.6732,
|
| 56 |
+
"rewards/accuracies": 0.5666667222976685,
|
| 57 |
+
"rewards/chosen": 0.09581196308135986,
|
| 58 |
+
"rewards/diff": 0.0,
|
| 59 |
+
"rewards/diff_abs": 0.0,
|
| 60 |
+
"rewards/rejected": -0.09433543682098389,
|
| 61 |
+
"rewards/student_margin": 0.19014742970466614,
|
| 62 |
+
"rewards/teacher_margin": 0.0,
|
| 63 |
+
"step": 20
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.04,
|
| 67 |
+
"grad_norm": 34.5,
|
| 68 |
+
"learning_rate": 1.973684210526316e-06,
|
| 69 |
+
"logits/chosen": 2.7114477157592773,
|
| 70 |
+
"logits/rejected": 2.583392858505249,
|
| 71 |
+
"logps/chosen": -763.4949951171875,
|
| 72 |
+
"logps/rejected": -838.5108642578125,
|
| 73 |
+
"loss": 0.5979,
|
| 74 |
+
"rewards/accuracies": 0.8666666746139526,
|
| 75 |
+
"rewards/chosen": 0.19188283383846283,
|
| 76 |
+
"rewards/diff": 0.0,
|
| 77 |
+
"rewards/diff_abs": 0.0,
|
| 78 |
+
"rewards/rejected": -0.49494582414627075,
|
| 79 |
+
"rewards/student_margin": 0.68682861328125,
|
| 80 |
+
"rewards/teacher_margin": 0.0,
|
| 81 |
+
"step": 30
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.05,
|
| 85 |
+
"grad_norm": 26.875,
|
| 86 |
+
"learning_rate": 2.631578947368421e-06,
|
| 87 |
+
"logits/chosen": 2.125535726547241,
|
| 88 |
+
"logits/rejected": 2.1427719593048096,
|
| 89 |
+
"logps/chosen": -681.7012939453125,
|
| 90 |
+
"logps/rejected": -856.2781982421875,
|
| 91 |
+
"loss": 0.4696,
|
| 92 |
+
"rewards/accuracies": 0.8666666746139526,
|
| 93 |
+
"rewards/chosen": 0.25057679414749146,
|
| 94 |
+
"rewards/diff": 0.0,
|
| 95 |
+
"rewards/diff_abs": 0.0,
|
| 96 |
+
"rewards/rejected": -0.9931653738021851,
|
| 97 |
+
"rewards/student_margin": 1.2437422275543213,
|
| 98 |
+
"rewards/teacher_margin": 0.0,
|
| 99 |
+
"step": 40
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.07,
|
| 103 |
+
"grad_norm": 36.25,
|
| 104 |
+
"learning_rate": 3.289473684210527e-06,
|
| 105 |
+
"logits/chosen": 2.4738311767578125,
|
| 106 |
+
"logits/rejected": 2.2791666984558105,
|
| 107 |
+
"logps/chosen": -677.8941040039062,
|
| 108 |
+
"logps/rejected": -804.3392333984375,
|
| 109 |
+
"loss": 0.4189,
|
| 110 |
+
"rewards/accuracies": 0.8333333730697632,
|
| 111 |
+
"rewards/chosen": 0.4866231381893158,
|
| 112 |
+
"rewards/diff": 0.0,
|
| 113 |
+
"rewards/diff_abs": 0.0,
|
| 114 |
+
"rewards/rejected": -1.3408163785934448,
|
| 115 |
+
"rewards/student_margin": 1.8274396657943726,
|
| 116 |
+
"rewards/teacher_margin": 0.0,
|
| 117 |
+
"step": 50
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.08,
|
| 121 |
+
"grad_norm": 24.75,
|
| 122 |
+
"learning_rate": 3.947368421052632e-06,
|
| 123 |
+
"logits/chosen": 2.610652446746826,
|
| 124 |
+
"logits/rejected": 2.885200262069702,
|
| 125 |
+
"logps/chosen": -713.7745361328125,
|
| 126 |
+
"logps/rejected": -946.8289794921875,
|
| 127 |
+
"loss": 0.3675,
|
| 128 |
+
"rewards/accuracies": 0.8666666150093079,
|
| 129 |
+
"rewards/chosen": 0.2676035761833191,
|
| 130 |
+
"rewards/diff": 0.0,
|
| 131 |
+
"rewards/diff_abs": 0.0,
|
| 132 |
+
"rewards/rejected": -2.754769802093506,
|
| 133 |
+
"rewards/student_margin": 3.0223731994628906,
|
| 134 |
+
"rewards/teacher_margin": 0.0,
|
| 135 |
+
"step": 60
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.09,
|
| 139 |
+
"grad_norm": 17.125,
|
| 140 |
+
"learning_rate": 4.605263157894737e-06,
|
| 141 |
+
"logits/chosen": 2.8368079662323,
|
| 142 |
+
"logits/rejected": 2.663803815841675,
|
| 143 |
+
"logps/chosen": -791.5745239257812,
|
| 144 |
+
"logps/rejected": -844.3741455078125,
|
| 145 |
+
"loss": 0.3444,
|
| 146 |
+
"rewards/accuracies": 0.8666666150093079,
|
| 147 |
+
"rewards/chosen": 0.23359784483909607,
|
| 148 |
+
"rewards/diff": 0.0,
|
| 149 |
+
"rewards/diff_abs": 0.0,
|
| 150 |
+
"rewards/rejected": -2.148716449737549,
|
| 151 |
+
"rewards/student_margin": 2.3823142051696777,
|
| 152 |
+
"rewards/teacher_margin": 0.0,
|
| 153 |
+
"step": 70
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.11,
|
| 157 |
+
"grad_norm": 22.625,
|
| 158 |
+
"learning_rate": 4.999574378856833e-06,
|
| 159 |
+
"logits/chosen": 2.178849220275879,
|
| 160 |
+
"logits/rejected": 2.505402088165283,
|
| 161 |
+
"logps/chosen": -694.8255615234375,
|
| 162 |
+
"logps/rejected": -889.0216674804688,
|
| 163 |
+
"loss": 0.3232,
|
| 164 |
+
"rewards/accuracies": 0.8666666150093079,
|
| 165 |
+
"rewards/chosen": 0.25273844599723816,
|
| 166 |
+
"rewards/diff": 0.0,
|
| 167 |
+
"rewards/diff_abs": 0.0,
|
| 168 |
+
"rewards/rejected": -2.432995557785034,
|
| 169 |
+
"rewards/student_margin": 2.685734272003174,
|
| 170 |
+
"rewards/teacher_margin": 0.0,
|
| 171 |
+
"step": 80
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.12,
|
| 175 |
+
"grad_norm": 24.625,
|
| 176 |
+
"learning_rate": 4.994787805186825e-06,
|
| 177 |
+
"logits/chosen": 2.619114875793457,
|
| 178 |
+
"logits/rejected": 2.5830256938934326,
|
| 179 |
+
"logps/chosen": -722.3959350585938,
|
| 180 |
+
"logps/rejected": -835.6867065429688,
|
| 181 |
+
"loss": 0.3808,
|
| 182 |
+
"rewards/accuracies": 0.7666666507720947,
|
| 183 |
+
"rewards/chosen": 0.12093061208724976,
|
| 184 |
+
"rewards/diff": 0.0,
|
| 185 |
+
"rewards/diff_abs": 0.0,
|
| 186 |
+
"rewards/rejected": -2.4770522117614746,
|
| 187 |
+
"rewards/student_margin": 2.597982883453369,
|
| 188 |
+
"rewards/teacher_margin": 0.0,
|
| 189 |
+
"step": 90
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.13,
|
| 193 |
+
"grad_norm": 19.75,
|
| 194 |
+
"learning_rate": 4.984692850202173e-06,
|
| 195 |
+
"logits/chosen": 2.8101232051849365,
|
| 196 |
+
"logits/rejected": 2.5987629890441895,
|
| 197 |
+
"logps/chosen": -783.5296630859375,
|
| 198 |
+
"logps/rejected": -776.9186401367188,
|
| 199 |
+
"loss": 0.2936,
|
| 200 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 201 |
+
"rewards/chosen": 0.4671626091003418,
|
| 202 |
+
"rewards/diff": 0.0,
|
| 203 |
+
"rewards/diff_abs": 0.0,
|
| 204 |
+
"rewards/rejected": -1.996647834777832,
|
| 205 |
+
"rewards/student_margin": 2.4638102054595947,
|
| 206 |
+
"rewards/teacher_margin": 0.0,
|
| 207 |
+
"step": 100
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.15,
|
| 211 |
+
"grad_norm": 21.0,
|
| 212 |
+
"learning_rate": 4.9693109938340754e-06,
|
| 213 |
+
"logits/chosen": 2.310434341430664,
|
| 214 |
+
"logits/rejected": 2.1604161262512207,
|
| 215 |
+
"logps/chosen": -729.8660888671875,
|
| 216 |
+
"logps/rejected": -837.9103393554688,
|
| 217 |
+
"loss": 0.36,
|
| 218 |
+
"rewards/accuracies": 0.800000011920929,
|
| 219 |
+
"rewards/chosen": 0.6024355292320251,
|
| 220 |
+
"rewards/diff": 0.0,
|
| 221 |
+
"rewards/diff_abs": 0.0,
|
| 222 |
+
"rewards/rejected": -2.125154972076416,
|
| 223 |
+
"rewards/student_margin": 2.727590322494507,
|
| 224 |
+
"rewards/teacher_margin": 0.0,
|
| 225 |
+
"step": 110
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.16,
|
| 229 |
+
"grad_norm": 27.25,
|
| 230 |
+
"learning_rate": 4.948674965422784e-06,
|
| 231 |
+
"logits/chosen": 2.2550618648529053,
|
| 232 |
+
"logits/rejected": 2.4227263927459717,
|
| 233 |
+
"logps/chosen": -709.0026245117188,
|
| 234 |
+
"logps/rejected": -861.9893798828125,
|
| 235 |
+
"loss": 0.3055,
|
| 236 |
+
"rewards/accuracies": 0.9000000953674316,
|
| 237 |
+
"rewards/chosen": 0.4605357050895691,
|
| 238 |
+
"rewards/diff": 0.0,
|
| 239 |
+
"rewards/diff_abs": 0.0,
|
| 240 |
+
"rewards/rejected": -1.956944465637207,
|
| 241 |
+
"rewards/student_margin": 2.417480230331421,
|
| 242 |
+
"rewards/teacher_margin": 0.0,
|
| 243 |
+
"step": 120
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.17,
|
| 247 |
+
"grad_norm": 22.75,
|
| 248 |
+
"learning_rate": 4.922828674076479e-06,
|
| 249 |
+
"logits/chosen": 2.638427734375,
|
| 250 |
+
"logits/rejected": 2.6139321327209473,
|
| 251 |
+
"logps/chosen": -738.7854614257812,
|
| 252 |
+
"logps/rejected": -812.3598022460938,
|
| 253 |
+
"loss": 0.299,
|
| 254 |
+
"rewards/accuracies": 0.8666666150093079,
|
| 255 |
+
"rewards/chosen": 0.4465925693511963,
|
| 256 |
+
"rewards/diff": 0.0,
|
| 257 |
+
"rewards/diff_abs": 0.0,
|
| 258 |
+
"rewards/rejected": -2.559539318084717,
|
| 259 |
+
"rewards/student_margin": 3.006131887435913,
|
| 260 |
+
"rewards/teacher_margin": 0.0,
|
| 261 |
+
"step": 130
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.18,
|
| 265 |
+
"grad_norm": 24.75,
|
| 266 |
+
"learning_rate": 4.891827115241967e-06,
|
| 267 |
+
"logits/chosen": 2.491821765899658,
|
| 268 |
+
"logits/rejected": 2.1934759616851807,
|
| 269 |
+
"logps/chosen": -731.97998046875,
|
| 270 |
+
"logps/rejected": -768.702880859375,
|
| 271 |
+
"loss": 0.3217,
|
| 272 |
+
"rewards/accuracies": 0.8666666150093079,
|
| 273 |
+
"rewards/chosen": 1.0505127906799316,
|
| 274 |
+
"rewards/diff": 0.0,
|
| 275 |
+
"rewards/diff_abs": 0.0,
|
| 276 |
+
"rewards/rejected": -2.0951590538024902,
|
| 277 |
+
"rewards/student_margin": 3.145671844482422,
|
| 278 |
+
"rewards/teacher_margin": 0.0,
|
| 279 |
+
"step": 140
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.2,
|
| 283 |
+
"grad_norm": 18.75,
|
| 284 |
+
"learning_rate": 4.855736253685992e-06,
|
| 285 |
+
"logits/chosen": 2.818260908126831,
|
| 286 |
+
"logits/rejected": 2.7023606300354004,
|
| 287 |
+
"logps/chosen": -855.1539916992188,
|
| 288 |
+
"logps/rejected": -863.5784912109375,
|
| 289 |
+
"loss": 0.3102,
|
| 290 |
+
"rewards/accuracies": 0.9333333969116211,
|
| 291 |
+
"rewards/chosen": 0.9310780763626099,
|
| 292 |
+
"rewards/diff": 0.0,
|
| 293 |
+
"rewards/diff_abs": 0.0,
|
| 294 |
+
"rewards/rejected": -2.548107624053955,
|
| 295 |
+
"rewards/student_margin": 3.4791858196258545,
|
| 296 |
+
"rewards/teacher_margin": 0.0,
|
| 297 |
+
"step": 150
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.21,
|
| 301 |
+
"grad_norm": 15.9375,
|
| 302 |
+
"learning_rate": 4.8146328831361466e-06,
|
| 303 |
+
"logits/chosen": 2.5584826469421387,
|
| 304 |
+
"logits/rejected": 2.5176262855529785,
|
| 305 |
+
"logps/chosen": -710.6329345703125,
|
| 306 |
+
"logps/rejected": -765.013916015625,
|
| 307 |
+
"loss": 0.255,
|
| 308 |
+
"rewards/accuracies": 0.800000011920929,
|
| 309 |
+
"rewards/chosen": 0.6767626404762268,
|
| 310 |
+
"rewards/diff": 0.0,
|
| 311 |
+
"rewards/diff_abs": 0.0,
|
| 312 |
+
"rewards/rejected": -1.5031782388687134,
|
| 313 |
+
"rewards/student_margin": 2.179940700531006,
|
| 314 |
+
"rewards/teacher_margin": 0.0,
|
| 315 |
+
"step": 160
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 0.22,
|
| 319 |
+
"grad_norm": 21.75,
|
| 320 |
+
"learning_rate": 4.768604462880048e-06,
|
| 321 |
+
"logits/chosen": 2.6777164936065674,
|
| 322 |
+
"logits/rejected": 2.462739944458008,
|
| 323 |
+
"logps/chosen": -709.0971069335938,
|
| 324 |
+
"logps/rejected": -819.5133666992188,
|
| 325 |
+
"loss": 0.2703,
|
| 326 |
+
"rewards/accuracies": 0.966666579246521,
|
| 327 |
+
"rewards/chosen": 1.097994327545166,
|
| 328 |
+
"rewards/diff": 0.0,
|
| 329 |
+
"rewards/diff_abs": 0.0,
|
| 330 |
+
"rewards/rejected": -2.144744873046875,
|
| 331 |
+
"rewards/student_margin": 3.24273943901062,
|
| 332 |
+
"rewards/teacher_margin": 0.0,
|
| 333 |
+
"step": 170
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 0.24,
|
| 337 |
+
"grad_norm": 14.25,
|
| 338 |
+
"learning_rate": 4.717748931670449e-06,
|
| 339 |
+
"logits/chosen": 2.2394096851348877,
|
| 340 |
+
"logits/rejected": 2.3399100303649902,
|
| 341 |
+
"logps/chosen": -810.650634765625,
|
| 342 |
+
"logps/rejected": -890.9011840820312,
|
| 343 |
+
"loss": 0.2325,
|
| 344 |
+
"rewards/accuracies": 0.9333332777023315,
|
| 345 |
+
"rewards/chosen": 0.620319128036499,
|
| 346 |
+
"rewards/diff": 0.0,
|
| 347 |
+
"rewards/diff_abs": 0.0,
|
| 348 |
+
"rewards/rejected": -2.6976723670959473,
|
| 349 |
+
"rewards/student_margin": 3.3179917335510254,
|
| 350 |
+
"rewards/teacher_margin": 0.0,
|
| 351 |
+
"step": 180
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 0.25,
|
| 355 |
+
"grad_norm": 13.8125,
|
| 356 |
+
"learning_rate": 4.662174499332275e-06,
|
| 357 |
+
"logits/chosen": 2.627516269683838,
|
| 358 |
+
"logits/rejected": 2.6304800510406494,
|
| 359 |
+
"logps/chosen": -743.6221923828125,
|
| 360 |
+
"logps/rejected": -846.0401611328125,
|
| 361 |
+
"loss": 0.2124,
|
| 362 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 363 |
+
"rewards/chosen": 0.11163077503442764,
|
| 364 |
+
"rewards/diff": 0.0,
|
| 365 |
+
"rewards/diff_abs": 0.0,
|
| 366 |
+
"rewards/rejected": -2.5682520866394043,
|
| 367 |
+
"rewards/student_margin": 2.6798830032348633,
|
| 368 |
+
"rewards/teacher_margin": 0.0,
|
| 369 |
+
"step": 190
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.26,
|
| 373 |
+
"grad_norm": 16.75,
|
| 374 |
+
"learning_rate": 4.601999416514973e-06,
|
| 375 |
+
"logits/chosen": 2.425690174102783,
|
| 376 |
+
"logits/rejected": 2.57381272315979,
|
| 377 |
+
"logps/chosen": -764.7454833984375,
|
| 378 |
+
"logps/rejected": -895.3714599609375,
|
| 379 |
+
"loss": 0.2577,
|
| 380 |
+
"rewards/accuracies": 0.9000000953674316,
|
| 381 |
+
"rewards/chosen": 1.1225372552871704,
|
| 382 |
+
"rewards/diff": 0.0,
|
| 383 |
+
"rewards/diff_abs": 0.0,
|
| 384 |
+
"rewards/rejected": -2.3252832889556885,
|
| 385 |
+
"rewards/student_margin": 3.4478206634521484,
|
| 386 |
+
"rewards/teacher_margin": 0.0,
|
| 387 |
+
"step": 200
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.28,
|
| 391 |
+
"grad_norm": 23.375,
|
| 392 |
+
"learning_rate": 4.537351723080117e-06,
|
| 393 |
+
"logits/chosen": 2.468726634979248,
|
| 394 |
+
"logits/rejected": 2.464418888092041,
|
| 395 |
+
"logps/chosen": -720.8771362304688,
|
| 396 |
+
"logps/rejected": -906.501953125,
|
| 397 |
+
"loss": 0.1797,
|
| 398 |
+
"rewards/accuracies": 1.0,
|
| 399 |
+
"rewards/chosen": 1.5130088329315186,
|
| 400 |
+
"rewards/diff": 0.0,
|
| 401 |
+
"rewards/diff_abs": 0.0,
|
| 402 |
+
"rewards/rejected": -2.8423776626586914,
|
| 403 |
+
"rewards/student_margin": 4.355386734008789,
|
| 404 |
+
"rewards/teacher_margin": 0.0,
|
| 405 |
+
"step": 210
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 0.29,
|
| 409 |
+
"grad_norm": 19.875,
|
| 410 |
+
"learning_rate": 4.468368975659638e-06,
|
| 411 |
+
"logits/chosen": 2.875859498977661,
|
| 412 |
+
"logits/rejected": 2.759331226348877,
|
| 413 |
+
"logps/chosen": -804.3804321289062,
|
| 414 |
+
"logps/rejected": -869.57958984375,
|
| 415 |
+
"loss": 0.1659,
|
| 416 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 417 |
+
"rewards/chosen": 1.109999179840088,
|
| 418 |
+
"rewards/diff": 0.0,
|
| 419 |
+
"rewards/diff_abs": 0.0,
|
| 420 |
+
"rewards/rejected": -3.0473203659057617,
|
| 421 |
+
"rewards/student_margin": 4.157319068908691,
|
| 422 |
+
"rewards/teacher_margin": 0.0,
|
| 423 |
+
"step": 220
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.3,
|
| 427 |
+
"grad_norm": 24.0,
|
| 428 |
+
"learning_rate": 4.395197954964377e-06,
|
| 429 |
+
"logits/chosen": 2.299351692199707,
|
| 430 |
+
"logits/rejected": 2.2807610034942627,
|
| 431 |
+
"logps/chosen": -668.815673828125,
|
| 432 |
+
"logps/rejected": -830.9869384765625,
|
| 433 |
+
"loss": 0.2658,
|
| 434 |
+
"rewards/accuracies": 0.9333332777023315,
|
| 435 |
+
"rewards/chosen": 1.0605835914611816,
|
| 436 |
+
"rewards/diff": 0.0,
|
| 437 |
+
"rewards/diff_abs": 0.0,
|
| 438 |
+
"rewards/rejected": -2.628436326980591,
|
| 439 |
+
"rewards/student_margin": 3.6890196800231934,
|
| 440 |
+
"rewards/teacher_margin": 0.0,
|
| 441 |
+
"step": 230
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.32,
|
| 445 |
+
"grad_norm": 17.0,
|
| 446 |
+
"learning_rate": 4.317994353465743e-06,
|
| 447 |
+
"logits/chosen": 2.4613468647003174,
|
| 448 |
+
"logits/rejected": 2.313901424407959,
|
| 449 |
+
"logps/chosen": -764.72119140625,
|
| 450 |
+
"logps/rejected": -858.3145751953125,
|
| 451 |
+
"loss": 0.2166,
|
| 452 |
+
"rewards/accuracies": 0.9333333969116211,
|
| 453 |
+
"rewards/chosen": 1.183833360671997,
|
| 454 |
+
"rewards/diff": 0.0,
|
| 455 |
+
"rewards/diff_abs": 0.0,
|
| 456 |
+
"rewards/rejected": -2.823739767074585,
|
| 457 |
+
"rewards/student_margin": 4.007573127746582,
|
| 458 |
+
"rewards/teacher_margin": 0.0,
|
| 459 |
+
"step": 240
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 0.33,
|
| 463 |
+
"grad_norm": 13.125,
|
| 464 |
+
"learning_rate": 4.236922444115033e-06,
|
| 465 |
+
"logits/chosen": 2.9010465145111084,
|
| 466 |
+
"logits/rejected": 2.975884199142456,
|
| 467 |
+
"logps/chosen": -766.8567504882812,
|
| 468 |
+
"logps/rejected": -921.1828002929688,
|
| 469 |
+
"loss": 0.1874,
|
| 470 |
+
"rewards/accuracies": 0.8333333730697632,
|
| 471 |
+
"rewards/chosen": 0.8090434074401855,
|
| 472 |
+
"rewards/diff": 0.0,
|
| 473 |
+
"rewards/diff_abs": 0.0,
|
| 474 |
+
"rewards/rejected": -1.870889663696289,
|
| 475 |
+
"rewards/student_margin": 2.6799330711364746,
|
| 476 |
+
"rewards/teacher_margin": 0.0,
|
| 477 |
+
"step": 250
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 0.34,
|
| 481 |
+
"grad_norm": 28.875,
|
| 482 |
+
"learning_rate": 4.152154730805301e-06,
|
| 483 |
+
"logits/chosen": 2.6264891624450684,
|
| 484 |
+
"logits/rejected": 2.745565891265869,
|
| 485 |
+
"logps/chosen": -748.6024169921875,
|
| 486 |
+
"logps/rejected": -891.1272583007812,
|
| 487 |
+
"loss": 0.2406,
|
| 488 |
+
"rewards/accuracies": 0.966666579246521,
|
| 489 |
+
"rewards/chosen": 1.5763810873031616,
|
| 490 |
+
"rewards/diff": 0.0,
|
| 491 |
+
"rewards/diff_abs": 0.0,
|
| 492 |
+
"rewards/rejected": -2.169581890106201,
|
| 493 |
+
"rewards/student_margin": 3.7459628582000732,
|
| 494 |
+
"rewards/teacher_margin": 0.0,
|
| 495 |
+
"step": 260
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 0.36,
|
| 499 |
+
"grad_norm": 10.0625,
|
| 500 |
+
"learning_rate": 4.063871581319526e-06,
|
| 501 |
+
"logits/chosen": 2.5747060775756836,
|
| 502 |
+
"logits/rejected": 2.6211907863616943,
|
| 503 |
+
"logps/chosen": -729.8704833984375,
|
| 504 |
+
"logps/rejected": -893.0794067382812,
|
| 505 |
+
"loss": 0.1921,
|
| 506 |
+
"rewards/accuracies": 0.9333333969116211,
|
| 507 |
+
"rewards/chosen": 0.9890060424804688,
|
| 508 |
+
"rewards/diff": 0.0,
|
| 509 |
+
"rewards/diff_abs": 0.0,
|
| 510 |
+
"rewards/rejected": -2.523777961730957,
|
| 511 |
+
"rewards/student_margin": 3.5127837657928467,
|
| 512 |
+
"rewards/teacher_margin": 0.0,
|
| 513 |
+
"step": 270
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.37,
|
| 517 |
+
"grad_norm": 10.875,
|
| 518 |
+
"learning_rate": 3.9722608435460716e-06,
|
| 519 |
+
"logits/chosen": 2.8572492599487305,
|
| 520 |
+
"logits/rejected": 2.787008762359619,
|
| 521 |
+
"logps/chosen": -744.4260864257812,
|
| 522 |
+
"logps/rejected": -903.74560546875,
|
| 523 |
+
"loss": 0.2131,
|
| 524 |
+
"rewards/accuracies": 0.9333333969116211,
|
| 525 |
+
"rewards/chosen": 1.8562577962875366,
|
| 526 |
+
"rewards/diff": 0.0,
|
| 527 |
+
"rewards/diff_abs": 0.0,
|
| 528 |
+
"rewards/rejected": -2.471896171569824,
|
| 529 |
+
"rewards/student_margin": 4.32815408706665,
|
| 530 |
+
"rewards/teacher_margin": 0.0,
|
| 531 |
+
"step": 280
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 0.38,
|
| 535 |
+
"grad_norm": 18.25,
|
| 536 |
+
"learning_rate": 3.877517445778084e-06,
|
| 537 |
+
"logits/chosen": 2.7553367614746094,
|
| 538 |
+
"logits/rejected": 2.84175705909729,
|
| 539 |
+
"logps/chosen": -851.8234252929688,
|
| 540 |
+
"logps/rejected": -869.7921142578125,
|
| 541 |
+
"loss": 0.1889,
|
| 542 |
+
"rewards/accuracies": 0.9333333969116211,
|
| 543 |
+
"rewards/chosen": 1.2772903442382812,
|
| 544 |
+
"rewards/diff": 0.0,
|
| 545 |
+
"rewards/diff_abs": 0.0,
|
| 546 |
+
"rewards/rejected": -2.959285259246826,
|
| 547 |
+
"rewards/student_margin": 4.236575603485107,
|
| 548 |
+
"rewards/teacher_margin": 0.0,
|
| 549 |
+
"step": 290
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.4,
|
| 553 |
+
"grad_norm": 20.0,
|
| 554 |
+
"learning_rate": 3.779842981947257e-06,
|
| 555 |
+
"logits/chosen": 2.5193684101104736,
|
| 556 |
+
"logits/rejected": 2.8250792026519775,
|
| 557 |
+
"logps/chosen": -777.2213134765625,
|
| 558 |
+
"logps/rejected": -962.1539916992188,
|
| 559 |
+
"loss": 0.1268,
|
| 560 |
+
"rewards/accuracies": 0.966666579246521,
|
| 561 |
+
"rewards/chosen": 1.30393385887146,
|
| 562 |
+
"rewards/diff": 0.0,
|
| 563 |
+
"rewards/diff_abs": 0.0,
|
| 564 |
+
"rewards/rejected": -3.1054177284240723,
|
| 565 |
+
"rewards/student_margin": 4.409350872039795,
|
| 566 |
+
"rewards/teacher_margin": 0.0,
|
| 567 |
+
"step": 300
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 0.41,
|
| 571 |
+
"grad_norm": 19.625,
|
| 572 |
+
"learning_rate": 3.679445282674553e-06,
|
| 573 |
+
"logits/chosen": 2.3319785594940186,
|
| 574 |
+
"logits/rejected": 2.200749635696411,
|
| 575 |
+
"logps/chosen": -736.1558837890625,
|
| 576 |
+
"logps/rejected": -828.9473876953125,
|
| 577 |
+
"loss": 0.2129,
|
| 578 |
+
"rewards/accuracies": 0.9333332777023315,
|
| 579 |
+
"rewards/chosen": 1.365607738494873,
|
| 580 |
+
"rewards/diff": 0.0,
|
| 581 |
+
"rewards/diff_abs": 0.0,
|
| 582 |
+
"rewards/rejected": -3.0576624870300293,
|
| 583 |
+
"rewards/student_margin": 4.423270225524902,
|
| 584 |
+
"rewards/teacher_margin": 0.0,
|
| 585 |
+
"step": 310
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.42,
|
| 589 |
+
"grad_norm": 14.125,
|
| 590 |
+
"learning_rate": 3.5765379730505488e-06,
|
| 591 |
+
"logits/chosen": 2.6987969875335693,
|
| 592 |
+
"logits/rejected": 2.7594046592712402,
|
| 593 |
+
"logps/chosen": -770.1749877929688,
|
| 594 |
+
"logps/rejected": -893.2551879882812,
|
| 595 |
+
"loss": 0.1738,
|
| 596 |
+
"rewards/accuracies": 0.9333332777023315,
|
| 597 |
+
"rewards/chosen": 1.076539158821106,
|
| 598 |
+
"rewards/diff": 0.0,
|
| 599 |
+
"rewards/diff_abs": 0.0,
|
| 600 |
+
"rewards/rejected": -3.4383862018585205,
|
| 601 |
+
"rewards/student_margin": 4.5149245262146,
|
| 602 |
+
"rewards/teacher_margin": 0.0,
|
| 603 |
+
"step": 320
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 0.44,
|
| 607 |
+
"grad_norm": 23.5,
|
| 608 |
+
"learning_rate": 3.471340018086381e-06,
|
| 609 |
+
"logits/chosen": 2.5061514377593994,
|
| 610 |
+
"logits/rejected": 2.6357250213623047,
|
| 611 |
+
"logps/chosen": -764.2349243164062,
|
| 612 |
+
"logps/rejected": -1048.420166015625,
|
| 613 |
+
"loss": 0.2055,
|
| 614 |
+
"rewards/accuracies": 0.9666666984558105,
|
| 615 |
+
"rewards/chosen": 1.4593900442123413,
|
| 616 |
+
"rewards/diff": 0.0,
|
| 617 |
+
"rewards/diff_abs": 0.0,
|
| 618 |
+
"rewards/rejected": -2.759369373321533,
|
| 619 |
+
"rewards/student_margin": 4.218760013580322,
|
| 620 |
+
"rewards/teacher_margin": 0.0,
|
| 621 |
+
"step": 330
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 0.45,
|
| 625 |
+
"grad_norm": 11.625,
|
| 626 |
+
"learning_rate": 3.364075256802471e-06,
|
| 627 |
+
"logits/chosen": 2.7208733558654785,
|
| 628 |
+
"logits/rejected": 2.7141761779785156,
|
| 629 |
+
"logps/chosen": -805.8480834960938,
|
| 630 |
+
"logps/rejected": -931.7926025390625,
|
| 631 |
+
"loss": 0.2091,
|
| 632 |
+
"rewards/accuracies": 0.9333333969116211,
|
| 633 |
+
"rewards/chosen": 1.5951147079467773,
|
| 634 |
+
"rewards/diff": 0.0,
|
| 635 |
+
"rewards/diff_abs": 0.0,
|
| 636 |
+
"rewards/rejected": -2.8993289470672607,
|
| 637 |
+
"rewards/student_margin": 4.494443893432617,
|
| 638 |
+
"rewards/teacher_margin": 0.0,
|
| 639 |
+
"step": 340
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.46,
|
| 643 |
+
"grad_norm": 10.8125,
|
| 644 |
+
"learning_rate": 3.254971925946383e-06,
|
| 645 |
+
"logits/chosen": 2.7437257766723633,
|
| 646 |
+
"logits/rejected": 2.5028164386749268,
|
| 647 |
+
"logps/chosen": -769.4174194335938,
|
| 648 |
+
"logps/rejected": -856.7975463867188,
|
| 649 |
+
"loss": 0.1375,
|
| 650 |
+
"rewards/accuracies": 1.0,
|
| 651 |
+
"rewards/chosen": 1.8919599056243896,
|
| 652 |
+
"rewards/diff": 0.0,
|
| 653 |
+
"rewards/diff_abs": 0.0,
|
| 654 |
+
"rewards/rejected": -2.4549193382263184,
|
| 655 |
+
"rewards/student_margin": 4.346879005432129,
|
| 656 |
+
"rewards/teacher_margin": 0.0,
|
| 657 |
+
"step": 350
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 0.48,
|
| 661 |
+
"grad_norm": 8.3125,
|
| 662 |
+
"learning_rate": 3.1442621743532516e-06,
|
| 663 |
+
"logits/chosen": 2.987748622894287,
|
| 664 |
+
"logits/rejected": 2.8533196449279785,
|
| 665 |
+
"logps/chosen": -763.906494140625,
|
| 666 |
+
"logps/rejected": -766.51708984375,
|
| 667 |
+
"loss": 0.2198,
|
| 668 |
+
"rewards/accuracies": 0.8666666746139526,
|
| 669 |
+
"rewards/chosen": 1.3865398168563843,
|
| 670 |
+
"rewards/diff": 0.0,
|
| 671 |
+
"rewards/diff_abs": 0.0,
|
| 672 |
+
"rewards/rejected": -2.3393778800964355,
|
| 673 |
+
"rewards/student_margin": 3.725917339324951,
|
| 674 |
+
"rewards/teacher_margin": 0.0,
|
| 675 |
+
"step": 360
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.49,
|
| 679 |
+
"grad_norm": 4.28125,
|
| 680 |
+
"learning_rate": 3.0321815689821127e-06,
|
| 681 |
+
"logits/chosen": 2.4668517112731934,
|
| 682 |
+
"logits/rejected": 2.6435983180999756,
|
| 683 |
+
"logps/chosen": -738.1652221679688,
|
| 684 |
+
"logps/rejected": -1032.048583984375,
|
| 685 |
+
"loss": 0.1364,
|
| 686 |
+
"rewards/accuracies": 0.9333333969116211,
|
| 687 |
+
"rewards/chosen": 1.38614821434021,
|
| 688 |
+
"rewards/diff": 0.0,
|
| 689 |
+
"rewards/diff_abs": 0.0,
|
| 690 |
+
"rewards/rejected": -3.3692269325256348,
|
| 691 |
+
"rewards/student_margin": 4.755375385284424,
|
| 692 |
+
"rewards/teacher_margin": 0.0,
|
| 693 |
+
"step": 370
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 0.5,
|
| 697 |
+
"grad_norm": 26.875,
|
| 698 |
+
"learning_rate": 2.9189685936791856e-06,
|
| 699 |
+
"logits/chosen": 2.366330146789551,
|
| 700 |
+
"logits/rejected": 2.6206741333007812,
|
| 701 |
+
"logps/chosen": -717.1395263671875,
|
| 702 |
+
"logps/rejected": -890.14990234375,
|
| 703 |
+
"loss": 0.2086,
|
| 704 |
+
"rewards/accuracies": 1.0,
|
| 705 |
+
"rewards/chosen": 1.2325283288955688,
|
| 706 |
+
"rewards/diff": 0.0,
|
| 707 |
+
"rewards/diff_abs": 0.0,
|
| 708 |
+
"rewards/rejected": -3.064573049545288,
|
| 709 |
+
"rewards/student_margin": 4.297101020812988,
|
| 710 |
+
"rewards/teacher_margin": 0.0,
|
| 711 |
+
"step": 380
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 0.52,
|
| 715 |
+
"grad_norm": 9.75,
|
| 716 |
+
"learning_rate": 2.8048641417346447e-06,
|
| 717 |
+
"logits/chosen": 2.639984130859375,
|
| 718 |
+
"logits/rejected": 2.6091980934143066,
|
| 719 |
+
"logps/chosen": -689.2213134765625,
|
| 720 |
+
"logps/rejected": -898.9273681640625,
|
| 721 |
+
"loss": 0.1774,
|
| 722 |
+
"rewards/accuracies": 0.9333333969116211,
|
| 723 |
+
"rewards/chosen": 1.432947039604187,
|
| 724 |
+
"rewards/diff": 0.0,
|
| 725 |
+
"rewards/diff_abs": 0.0,
|
| 726 |
+
"rewards/rejected": -3.152224063873291,
|
| 727 |
+
"rewards/student_margin": 4.585171222686768,
|
| 728 |
+
"rewards/teacher_margin": 0.0,
|
| 729 |
+
"step": 390
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 0.53,
|
| 733 |
+
"grad_norm": 17.25,
|
| 734 |
+
"learning_rate": 2.69011100331259e-06,
|
| 735 |
+
"logits/chosen": 2.7592196464538574,
|
| 736 |
+
"logits/rejected": 2.9880573749542236,
|
| 737 |
+
"logps/chosen": -680.7615966796875,
|
| 738 |
+
"logps/rejected": -911.51708984375,
|
| 739 |
+
"loss": 0.1216,
|
| 740 |
+
"rewards/accuracies": 1.0,
|
| 741 |
+
"rewards/chosen": 1.259261131286621,
|
| 742 |
+
"rewards/diff": 0.0,
|
| 743 |
+
"rewards/diff_abs": 0.0,
|
| 744 |
+
"rewards/rejected": -3.410494565963745,
|
| 745 |
+
"rewards/student_margin": 4.669755458831787,
|
| 746 |
+
"rewards/teacher_margin": 0.0,
|
| 747 |
+
"step": 400
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 0.54,
|
| 751 |
+
"grad_norm": 5.15625,
|
| 752 |
+
"learning_rate": 2.574953348844884e-06,
|
| 753 |
+
"logits/chosen": 2.678232192993164,
|
| 754 |
+
"logits/rejected": 2.359405994415283,
|
| 755 |
+
"logps/chosen": -791.8428955078125,
|
| 756 |
+
"logps/rejected": -739.9708251953125,
|
| 757 |
+
"loss": 0.1245,
|
| 758 |
+
"rewards/accuracies": 1.0,
|
| 759 |
+
"rewards/chosen": 1.0176723003387451,
|
| 760 |
+
"rewards/diff": 0.0,
|
| 761 |
+
"rewards/diff_abs": 0.0,
|
| 762 |
+
"rewards/rejected": -2.4821863174438477,
|
| 763 |
+
"rewards/student_margin": 3.4998583793640137,
|
| 764 |
+
"rewards/teacher_margin": 0.0,
|
| 765 |
+
"step": 410
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.55,
|
| 769 |
+
"grad_norm": 7.5625,
|
| 770 |
+
"learning_rate": 2.4596362094880617e-06,
|
| 771 |
+
"logits/chosen": 2.144169569015503,
|
| 772 |
+
"logits/rejected": 1.9014570713043213,
|
| 773 |
+
"logps/chosen": -629.1177978515625,
|
| 774 |
+
"logps/rejected": -703.9340209960938,
|
| 775 |
+
"loss": 0.1786,
|
| 776 |
+
"rewards/accuracies": 0.9333333969116211,
|
| 777 |
+
"rewards/chosen": 1.782714605331421,
|
| 778 |
+
"rewards/diff": 0.0,
|
| 779 |
+
"rewards/diff_abs": 0.0,
|
| 780 |
+
"rewards/rejected": -2.654998302459717,
|
| 781 |
+
"rewards/student_margin": 4.437712669372559,
|
| 782 |
+
"rewards/teacher_margin": 0.0,
|
| 783 |
+
"step": 420
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 0.57,
|
| 787 |
+
"grad_norm": 15.625,
|
| 788 |
+
"learning_rate": 2.344404955748805e-06,
|
| 789 |
+
"logits/chosen": 2.4446656703948975,
|
| 790 |
+
"logits/rejected": 2.495633840560913,
|
| 791 |
+
"logps/chosen": -737.5274047851562,
|
| 792 |
+
"logps/rejected": -865.40771484375,
|
| 793 |
+
"loss": 0.2011,
|
| 794 |
+
"rewards/accuracies": 0.9333333969116211,
|
| 795 |
+
"rewards/chosen": 1.6026709079742432,
|
| 796 |
+
"rewards/diff": 0.0,
|
| 797 |
+
"rewards/diff_abs": 0.0,
|
| 798 |
+
"rewards/rejected": -3.152886152267456,
|
| 799 |
+
"rewards/student_margin": 4.755557060241699,
|
| 800 |
+
"rewards/teacher_margin": 0.0,
|
| 801 |
+
"step": 430
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.58,
|
| 805 |
+
"grad_norm": 10.0625,
|
| 806 |
+
"learning_rate": 2.2295047753873465e-06,
|
| 807 |
+
"logits/chosen": 2.3097705841064453,
|
| 808 |
+
"logits/rejected": 2.4514598846435547,
|
| 809 |
+
"logps/chosen": -707.8262939453125,
|
| 810 |
+
"logps/rejected": -975.5931396484375,
|
| 811 |
+
"loss": 0.0938,
|
| 812 |
+
"rewards/accuracies": 1.0,
|
| 813 |
+
"rewards/chosen": 1.9523252248764038,
|
| 814 |
+
"rewards/diff": 0.0,
|
| 815 |
+
"rewards/diff_abs": 0.0,
|
| 816 |
+
"rewards/rejected": -3.6663551330566406,
|
| 817 |
+
"rewards/student_margin": 5.618680000305176,
|
| 818 |
+
"rewards/teacher_margin": 0.0,
|
| 819 |
+
"step": 440
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 0.59,
|
| 823 |
+
"grad_norm": 9.1875,
|
| 824 |
+
"learning_rate": 2.1151801517097266e-06,
|
| 825 |
+
"logits/chosen": 2.646815776824951,
|
| 826 |
+
"logits/rejected": 2.6606507301330566,
|
| 827 |
+
"logps/chosen": -785.2379150390625,
|
| 828 |
+
"logps/rejected": -882.7388916015625,
|
| 829 |
+
"loss": 0.1304,
|
| 830 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 831 |
+
"rewards/chosen": 1.0314725637435913,
|
| 832 |
+
"rewards/diff": 0.0,
|
| 833 |
+
"rewards/diff_abs": 0.0,
|
| 834 |
+
"rewards/rejected": -3.4571354389190674,
|
| 835 |
+
"rewards/student_margin": 4.488607883453369,
|
| 836 |
+
"rewards/teacher_margin": 0.0,
|
| 837 |
+
"step": 450
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 0.61,
|
| 841 |
+
"grad_norm": 12.1875,
|
| 842 |
+
"learning_rate": 2.001674343358975e-06,
|
| 843 |
+
"logits/chosen": 2.504572629928589,
|
| 844 |
+
"logits/rejected": 2.6332459449768066,
|
| 845 |
+
"logps/chosen": -711.09375,
|
| 846 |
+
"logps/rejected": -883.10107421875,
|
| 847 |
+
"loss": 0.1243,
|
| 848 |
+
"rewards/accuracies": 0.9333332777023315,
|
| 849 |
+
"rewards/chosen": 1.1015676259994507,
|
| 850 |
+
"rewards/diff": 0.0,
|
| 851 |
+
"rewards/diff_abs": 0.0,
|
| 852 |
+
"rewards/rejected": -3.6316211223602295,
|
| 853 |
+
"rewards/student_margin": 4.733188629150391,
|
| 854 |
+
"rewards/teacher_margin": 0.0,
|
| 855 |
+
"step": 460
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 0.62,
|
| 859 |
+
"grad_norm": 4.3125,
|
| 860 |
+
"learning_rate": 1.8892288667121139e-06,
|
| 861 |
+
"logits/chosen": 2.414062738418579,
|
| 862 |
+
"logits/rejected": 2.228271484375,
|
| 863 |
+
"logps/chosen": -739.9596557617188,
|
| 864 |
+
"logps/rejected": -847.4876098632812,
|
| 865 |
+
"loss": 0.1183,
|
| 866 |
+
"rewards/accuracies": 0.9333333969116211,
|
| 867 |
+
"rewards/chosen": 1.0411046743392944,
|
| 868 |
+
"rewards/diff": 0.0,
|
| 869 |
+
"rewards/diff_abs": 0.0,
|
| 870 |
+
"rewards/rejected": -2.962963581085205,
|
| 871 |
+
"rewards/student_margin": 4.004067897796631,
|
| 872 |
+
"rewards/teacher_margin": 0.0,
|
| 873 |
+
"step": 470
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 0.63,
|
| 877 |
+
"grad_norm": 5.8125,
|
| 878 |
+
"learning_rate": 1.7780829819843458e-06,
|
| 879 |
+
"logits/chosen": 2.6683709621429443,
|
| 880 |
+
"logits/rejected": 2.7936007976531982,
|
| 881 |
+
"logps/chosen": -721.1678466796875,
|
| 882 |
+
"logps/rejected": -834.6106567382812,
|
| 883 |
+
"loss": 0.1543,
|
| 884 |
+
"rewards/accuracies": 0.966666579246521,
|
| 885 |
+
"rewards/chosen": 1.5083043575286865,
|
| 886 |
+
"rewards/diff": 0.0,
|
| 887 |
+
"rewards/diff_abs": 0.0,
|
| 888 |
+
"rewards/rejected": -3.2512154579162598,
|
| 889 |
+
"rewards/student_margin": 4.759519100189209,
|
| 890 |
+
"rewards/teacher_margin": 0.0,
|
| 891 |
+
"step": 480
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.65,
|
| 895 |
+
"grad_norm": 16.0,
|
| 896 |
+
"learning_rate": 1.668473184133862e-06,
|
| 897 |
+
"logits/chosen": 2.3921468257904053,
|
| 898 |
+
"logits/rejected": 2.376159429550171,
|
| 899 |
+
"logps/chosen": -738.181884765625,
|
| 900 |
+
"logps/rejected": -903.42919921875,
|
| 901 |
+
"loss": 0.1793,
|
| 902 |
+
"rewards/accuracies": 1.0,
|
| 903 |
+
"rewards/chosen": 1.5232467651367188,
|
| 904 |
+
"rewards/diff": 0.0,
|
| 905 |
+
"rewards/diff_abs": 0.0,
|
| 906 |
+
"rewards/rejected": -2.4825570583343506,
|
| 907 |
+
"rewards/student_margin": 4.00580358505249,
|
| 908 |
+
"rewards/teacher_margin": 0.0,
|
| 909 |
+
"step": 490
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 0.66,
|
| 913 |
+
"grad_norm": 23.125,
|
| 914 |
+
"learning_rate": 1.5606326996505478e-06,
|
| 915 |
+
"logits/chosen": 2.4245734214782715,
|
| 916 |
+
"logits/rejected": 2.066218614578247,
|
| 917 |
+
"logps/chosen": -789.9022827148438,
|
| 918 |
+
"logps/rejected": -760.2530517578125,
|
| 919 |
+
"loss": 0.1997,
|
| 920 |
+
"rewards/accuracies": 0.9666666984558105,
|
| 921 |
+
"rewards/chosen": 1.2094615697860718,
|
| 922 |
+
"rewards/diff": 0.0,
|
| 923 |
+
"rewards/diff_abs": 0.0,
|
| 924 |
+
"rewards/rejected": -2.621572256088257,
|
| 925 |
+
"rewards/student_margin": 3.831033706665039,
|
| 926 |
+
"rewards/teacher_margin": 0.0,
|
| 927 |
+
"step": 500
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 0.67,
|
| 931 |
+
"grad_norm": 20.625,
|
| 932 |
+
"learning_rate": 1.45479099029929e-06,
|
| 933 |
+
"logits/chosen": 2.312058687210083,
|
| 934 |
+
"logits/rejected": 2.6048178672790527,
|
| 935 |
+
"logps/chosen": -724.6693115234375,
|
| 936 |
+
"logps/rejected": -881.6843872070312,
|
| 937 |
+
"loss": 0.1882,
|
| 938 |
+
"rewards/accuracies": 0.966666579246521,
|
| 939 |
+
"rewards/chosen": 1.326018214225769,
|
| 940 |
+
"rewards/diff": 0.0,
|
| 941 |
+
"rewards/diff_abs": 0.0,
|
| 942 |
+
"rewards/rejected": -2.731388568878174,
|
| 943 |
+
"rewards/student_margin": 4.057406425476074,
|
| 944 |
+
"rewards/teacher_margin": 0.0,
|
| 945 |
+
"step": 510
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 0.69,
|
| 949 |
+
"grad_norm": 14.875,
|
| 950 |
+
"learning_rate": 1.3511732648738213e-06,
|
| 951 |
+
"logits/chosen": 2.2725682258605957,
|
| 952 |
+
"logits/rejected": 2.4900736808776855,
|
| 953 |
+
"logps/chosen": -779.0091552734375,
|
| 954 |
+
"logps/rejected": -983.3134765625,
|
| 955 |
+
"loss": 0.1265,
|
| 956 |
+
"rewards/accuracies": 0.9333333969116211,
|
| 957 |
+
"rewards/chosen": 1.638450026512146,
|
| 958 |
+
"rewards/diff": 0.0,
|
| 959 |
+
"rewards/diff_abs": 0.0,
|
| 960 |
+
"rewards/rejected": -2.900458812713623,
|
| 961 |
+
"rewards/student_margin": 4.538908958435059,
|
| 962 |
+
"rewards/teacher_margin": 0.0,
|
| 963 |
+
"step": 520
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 0.7,
|
| 967 |
+
"grad_norm": 20.0,
|
| 968 |
+
"learning_rate": 1.2500000000000007e-06,
|
| 969 |
+
"logits/chosen": 2.5716114044189453,
|
| 970 |
+
"logits/rejected": 2.5023512840270996,
|
| 971 |
+
"logps/chosen": -705.2183837890625,
|
| 972 |
+
"logps/rejected": -907.6337890625,
|
| 973 |
+
"loss": 0.1224,
|
| 974 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 975 |
+
"rewards/chosen": 1.617611289024353,
|
| 976 |
+
"rewards/diff": 0.0,
|
| 977 |
+
"rewards/diff_abs": 0.0,
|
| 978 |
+
"rewards/rejected": -2.396939992904663,
|
| 979 |
+
"rewards/student_margin": 4.014551639556885,
|
| 980 |
+
"rewards/teacher_margin": 0.0,
|
| 981 |
+
"step": 530
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 0.71,
|
| 985 |
+
"grad_norm": 20.125,
|
| 986 |
+
"learning_rate": 1.1514864710081295e-06,
|
| 987 |
+
"logits/chosen": 2.1336722373962402,
|
| 988 |
+
"logits/rejected": 2.49604868888855,
|
| 989 |
+
"logps/chosen": -596.4373779296875,
|
| 990 |
+
"logps/rejected": -815.4148559570312,
|
| 991 |
+
"loss": 0.1487,
|
| 992 |
+
"rewards/accuracies": 0.9666666984558105,
|
| 993 |
+
"rewards/chosen": 1.7167946100234985,
|
| 994 |
+
"rewards/diff": 0.0,
|
| 995 |
+
"rewards/diff_abs": 0.0,
|
| 996 |
+
"rewards/rejected": -2.7153236865997314,
|
| 997 |
+
"rewards/student_margin": 4.4321184158325195,
|
| 998 |
+
"rewards/teacher_margin": 0.0,
|
| 999 |
+
"step": 540
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 0.73,
|
| 1003 |
+
"grad_norm": 14.0,
|
| 1004 |
+
"learning_rate": 1.0558422938725404e-06,
|
| 1005 |
+
"logits/chosen": 2.081679344177246,
|
| 1006 |
+
"logits/rejected": 2.593083620071411,
|
| 1007 |
+
"logps/chosen": -605.888427734375,
|
| 1008 |
+
"logps/rejected": -937.4814453125,
|
| 1009 |
+
"loss": 0.1279,
|
| 1010 |
+
"rewards/accuracies": 1.0,
|
| 1011 |
+
"rewards/chosen": 1.3643019199371338,
|
| 1012 |
+
"rewards/diff": 0.0,
|
| 1013 |
+
"rewards/diff_abs": 0.0,
|
| 1014 |
+
"rewards/rejected": -3.542567491531372,
|
| 1015 |
+
"rewards/student_margin": 4.906869411468506,
|
| 1016 |
+
"rewards/teacher_margin": 0.0,
|
| 1017 |
+
"step": 550
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.74,
|
| 1021 |
+
"grad_norm": 10.3125,
|
| 1022 |
+
"learning_rate": 9.632709791930977e-07,
|
| 1023 |
+
"logits/chosen": 2.3516314029693604,
|
| 1024 |
+
"logits/rejected": 2.1533358097076416,
|
| 1025 |
+
"logps/chosen": -748.89892578125,
|
| 1026 |
+
"logps/rejected": -780.47998046875,
|
| 1027 |
+
"loss": 0.2078,
|
| 1028 |
+
"rewards/accuracies": 0.966666579246521,
|
| 1029 |
+
"rewards/chosen": 1.812962293624878,
|
| 1030 |
+
"rewards/diff": 0.0,
|
| 1031 |
+
"rewards/diff_abs": 0.0,
|
| 1032 |
+
"rewards/rejected": -2.3793187141418457,
|
| 1033 |
+
"rewards/student_margin": 4.1922807693481445,
|
| 1034 |
+
"rewards/teacher_margin": 0.0,
|
| 1035 |
+
"step": 560
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 0.75,
|
| 1039 |
+
"grad_norm": 9.1875,
|
| 1040 |
+
"learning_rate": 8.739694991676328e-07,
|
| 1041 |
+
"logits/chosen": 2.1663920879364014,
|
| 1042 |
+
"logits/rejected": 2.4739956855773926,
|
| 1043 |
+
"logps/chosen": -654.0028076171875,
|
| 1044 |
+
"logps/rejected": -846.7198486328125,
|
| 1045 |
+
"loss": 0.153,
|
| 1046 |
+
"rewards/accuracies": 0.9666666984558105,
|
| 1047 |
+
"rewards/chosen": 1.5124599933624268,
|
| 1048 |
+
"rewards/diff": 0.0,
|
| 1049 |
+
"rewards/diff_abs": 0.0,
|
| 1050 |
+
"rewards/rejected": -2.6132845878601074,
|
| 1051 |
+
"rewards/student_margin": 4.125744819641113,
|
| 1052 |
+
"rewards/teacher_margin": 0.0,
|
| 1053 |
+
"step": 570
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 0.77,
|
| 1057 |
+
"grad_norm": 20.125,
|
| 1058 |
+
"learning_rate": 7.881278684767354e-07,
|
| 1059 |
+
"logits/chosen": 2.249051809310913,
|
| 1060 |
+
"logits/rejected": 2.377701997756958,
|
| 1061 |
+
"logps/chosen": -744.0155029296875,
|
| 1062 |
+
"logps/rejected": -837.7652587890625,
|
| 1063 |
+
"loss": 0.1775,
|
| 1064 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 1065 |
+
"rewards/chosen": 1.166346788406372,
|
| 1066 |
+
"rewards/diff": 0.0,
|
| 1067 |
+
"rewards/diff_abs": 0.0,
|
| 1068 |
+
"rewards/rejected": -2.4967403411865234,
|
| 1069 |
+
"rewards/student_margin": 3.6630866527557373,
|
| 1070 |
+
"rewards/teacher_margin": 0.0,
|
| 1071 |
+
"step": 580
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 0.78,
|
| 1075 |
+
"grad_norm": 19.5,
|
| 1076 |
+
"learning_rate": 7.059287399726558e-07,
|
| 1077 |
+
"logits/chosen": 2.0429625511169434,
|
| 1078 |
+
"logits/rejected": 2.2974820137023926,
|
| 1079 |
+
"logps/chosen": -675.0552368164062,
|
| 1080 |
+
"logps/rejected": -815.6538696289062,
|
| 1081 |
+
"loss": 0.1259,
|
| 1082 |
+
"rewards/accuracies": 1.0,
|
| 1083 |
+
"rewards/chosen": 1.4241259098052979,
|
| 1084 |
+
"rewards/diff": 0.0,
|
| 1085 |
+
"rewards/diff_abs": 0.0,
|
| 1086 |
+
"rewards/rejected": -2.6004066467285156,
|
| 1087 |
+
"rewards/student_margin": 4.024532794952393,
|
| 1088 |
+
"rewards/teacher_margin": 0.0,
|
| 1089 |
+
"step": 590
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 0.79,
|
| 1093 |
+
"grad_norm": 15.25,
|
| 1094 |
+
"learning_rate": 6.275470160326308e-07,
|
| 1095 |
+
"logits/chosen": 2.4218051433563232,
|
| 1096 |
+
"logits/rejected": 2.626558780670166,
|
| 1097 |
+
"logps/chosen": -694.3748168945312,
|
| 1098 |
+
"logps/rejected": -825.705078125,
|
| 1099 |
+
"loss": 0.1295,
|
| 1100 |
+
"rewards/accuracies": 0.8333333134651184,
|
| 1101 |
+
"rewards/chosen": 1.1830830574035645,
|
| 1102 |
+
"rewards/diff": 0.0,
|
| 1103 |
+
"rewards/diff_abs": 0.0,
|
| 1104 |
+
"rewards/rejected": -2.594542980194092,
|
| 1105 |
+
"rewards/student_margin": 3.7776265144348145,
|
| 1106 |
+
"rewards/teacher_margin": 0.0,
|
| 1107 |
+
"step": 600
|
| 1108 |
+
},
|
| 1109 |
+
{
|
| 1110 |
+
"epoch": 0.81,
|
| 1111 |
+
"grad_norm": 14.0625,
|
| 1112 |
+
"learning_rate": 5.531494764035842e-07,
|
| 1113 |
+
"logits/chosen": 2.403853416442871,
|
| 1114 |
+
"logits/rejected": 2.4846677780151367,
|
| 1115 |
+
"logps/chosen": -745.2608642578125,
|
| 1116 |
+
"logps/rejected": -816.1573486328125,
|
| 1117 |
+
"loss": 0.1549,
|
| 1118 |
+
"rewards/accuracies": 0.800000011920929,
|
| 1119 |
+
"rewards/chosen": 0.9859026670455933,
|
| 1120 |
+
"rewards/diff": 0.0,
|
| 1121 |
+
"rewards/diff_abs": 0.0,
|
| 1122 |
+
"rewards/rejected": -2.085474729537964,
|
| 1123 |
+
"rewards/student_margin": 3.0713775157928467,
|
| 1124 |
+
"rewards/teacher_margin": 0.0,
|
| 1125 |
+
"step": 610
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 0.82,
|
| 1129 |
+
"grad_norm": 10.1875,
|
| 1130 |
+
"learning_rate": 4.828944233300581e-07,
|
| 1131 |
+
"logits/chosen": 2.5432651042938232,
|
| 1132 |
+
"logits/rejected": 2.3120391368865967,
|
| 1133 |
+
"logps/chosen": -758.4690551757812,
|
| 1134 |
+
"logps/rejected": -820.4471435546875,
|
| 1135 |
+
"loss": 0.0827,
|
| 1136 |
+
"rewards/accuracies": 1.0,
|
| 1137 |
+
"rewards/chosen": 1.7055679559707642,
|
| 1138 |
+
"rewards/diff": 0.0,
|
| 1139 |
+
"rewards/diff_abs": 0.0,
|
| 1140 |
+
"rewards/rejected": -2.691899538040161,
|
| 1141 |
+
"rewards/student_margin": 4.397467613220215,
|
| 1142 |
+
"rewards/teacher_margin": 0.0,
|
| 1143 |
+
"step": 620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.83,
|
| 1147 |
+
"grad_norm": 18.875,
|
| 1148 |
+
"learning_rate": 4.1693134472049254e-07,
|
| 1149 |
+
"logits/chosen": 2.4983761310577393,
|
| 1150 |
+
"logits/rejected": 2.4922070503234863,
|
| 1151 |
+
"logps/chosen": -759.421875,
|
| 1152 |
+
"logps/rejected": -880.2047119140625,
|
| 1153 |
+
"loss": 0.1595,
|
| 1154 |
+
"rewards/accuracies": 0.966666579246521,
|
| 1155 |
+
"rewards/chosen": 1.5521160364151,
|
| 1156 |
+
"rewards/diff": 0.0,
|
| 1157 |
+
"rewards/diff_abs": 0.0,
|
| 1158 |
+
"rewards/rejected": -2.921255588531494,
|
| 1159 |
+
"rewards/student_margin": 4.473371505737305,
|
| 1160 |
+
"rewards/teacher_margin": 0.0,
|
| 1161 |
+
"step": 630
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 0.85,
|
| 1165 |
+
"grad_norm": 12.875,
|
| 1166 |
+
"learning_rate": 3.554005960685364e-07,
|
| 1167 |
+
"logits/chosen": 2.4154155254364014,
|
| 1168 |
+
"logits/rejected": 2.3739418983459473,
|
| 1169 |
+
"logps/chosen": -722.2140502929688,
|
| 1170 |
+
"logps/rejected": -797.0501708984375,
|
| 1171 |
+
"loss": 0.1201,
|
| 1172 |
+
"rewards/accuracies": 1.0,
|
| 1173 |
+
"rewards/chosen": 2.029052495956421,
|
| 1174 |
+
"rewards/diff": 0.0,
|
| 1175 |
+
"rewards/diff_abs": 0.0,
|
| 1176 |
+
"rewards/rejected": -2.8493266105651855,
|
| 1177 |
+
"rewards/student_margin": 4.8783793449401855,
|
| 1178 |
+
"rewards/teacher_margin": 0.0,
|
| 1179 |
+
"step": 640
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 0.86,
|
| 1183 |
+
"grad_norm": 28.75,
|
| 1184 |
+
"learning_rate": 2.984331018062148e-07,
|
| 1185 |
+
"logits/chosen": 2.273947238922119,
|
| 1186 |
+
"logits/rejected": 2.346890926361084,
|
| 1187 |
+
"logps/chosen": -617.5139770507812,
|
| 1188 |
+
"logps/rejected": -829.6735229492188,
|
| 1189 |
+
"loss": 0.1981,
|
| 1190 |
+
"rewards/accuracies": 0.9000000953674316,
|
| 1191 |
+
"rewards/chosen": 1.2198824882507324,
|
| 1192 |
+
"rewards/diff": 0.0,
|
| 1193 |
+
"rewards/diff_abs": 0.0,
|
| 1194 |
+
"rewards/rejected": -2.615978717803955,
|
| 1195 |
+
"rewards/student_margin": 3.8358612060546875,
|
| 1196 |
+
"rewards/teacher_margin": 0.0,
|
| 1197 |
+
"step": 650
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 0.87,
|
| 1201 |
+
"grad_norm": 3.953125,
|
| 1202 |
+
"learning_rate": 2.461500767244049e-07,
|
| 1203 |
+
"logits/chosen": 2.328996181488037,
|
| 1204 |
+
"logits/rejected": 2.3885445594787598,
|
| 1205 |
+
"logps/chosen": -684.5684814453125,
|
| 1206 |
+
"logps/rejected": -871.5986328125,
|
| 1207 |
+
"loss": 0.1078,
|
| 1208 |
+
"rewards/accuracies": 0.9666666984558105,
|
| 1209 |
+
"rewards/chosen": 1.9273674488067627,
|
| 1210 |
+
"rewards/diff": 0.0,
|
| 1211 |
+
"rewards/diff_abs": 0.0,
|
| 1212 |
+
"rewards/rejected": -2.5571177005767822,
|
| 1213 |
+
"rewards/student_margin": 4.484485149383545,
|
| 1214 |
+
"rewards/teacher_margin": 0.0,
|
| 1215 |
+
"step": 660
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 0.89,
|
| 1219 |
+
"grad_norm": 12.9375,
|
| 1220 |
+
"learning_rate": 1.9866276805336733e-07,
|
| 1221 |
+
"logits/chosen": 2.427218198776245,
|
| 1222 |
+
"logits/rejected": 2.4423775672912598,
|
| 1223 |
+
"logps/chosen": -671.7244262695312,
|
| 1224 |
+
"logps/rejected": -919.7501220703125,
|
| 1225 |
+
"loss": 0.1608,
|
| 1226 |
+
"rewards/accuracies": 0.8666666746139526,
|
| 1227 |
+
"rewards/chosen": 1.4714218378067017,
|
| 1228 |
+
"rewards/diff": 0.0,
|
| 1229 |
+
"rewards/diff_abs": 0.0,
|
| 1230 |
+
"rewards/rejected": -2.4355459213256836,
|
| 1231 |
+
"rewards/student_margin": 3.9069676399230957,
|
| 1232 |
+
"rewards/teacher_margin": 0.0,
|
| 1233 |
+
"step": 670
|
| 1234 |
+
},
|
| 1235 |
+
{
|
| 1236 |
+
"epoch": 0.9,
|
| 1237 |
+
"grad_norm": 13.375,
|
| 1238 |
+
"learning_rate": 1.560722187521496e-07,
|
| 1239 |
+
"logits/chosen": 2.2423176765441895,
|
| 1240 |
+
"logits/rejected": 1.9140275716781616,
|
| 1241 |
+
"logps/chosen": -724.7633056640625,
|
| 1242 |
+
"logps/rejected": -754.9696655273438,
|
| 1243 |
+
"loss": 0.1739,
|
| 1244 |
+
"rewards/accuracies": 0.9666666984558105,
|
| 1245 |
+
"rewards/chosen": 1.4883036613464355,
|
| 1246 |
+
"rewards/diff": 0.0,
|
| 1247 |
+
"rewards/diff_abs": 0.0,
|
| 1248 |
+
"rewards/rejected": -2.9442994594573975,
|
| 1249 |
+
"rewards/student_margin": 4.432602882385254,
|
| 1250 |
+
"rewards/teacher_margin": 0.0,
|
| 1251 |
+
"step": 680
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 0.91,
|
| 1255 |
+
"grad_norm": 14.5625,
|
| 1256 |
+
"learning_rate": 1.1846905251051894e-07,
|
| 1257 |
+
"logits/chosen": 2.494938850402832,
|
| 1258 |
+
"logits/rejected": 2.532592296600342,
|
| 1259 |
+
"logps/chosen": -734.0770263671875,
|
| 1260 |
+
"logps/rejected": -898.4133911132812,
|
| 1261 |
+
"loss": 0.156,
|
| 1262 |
+
"rewards/accuracies": 0.9333333969116211,
|
| 1263 |
+
"rewards/chosen": 1.3606359958648682,
|
| 1264 |
+
"rewards/diff": 0.0,
|
| 1265 |
+
"rewards/diff_abs": 0.0,
|
| 1266 |
+
"rewards/rejected": -2.778499126434326,
|
| 1267 |
+
"rewards/student_margin": 4.139134883880615,
|
| 1268 |
+
"rewards/teacher_margin": 0.0,
|
| 1269 |
+
"step": 690
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.92,
|
| 1273 |
+
"grad_norm": 19.125,
|
| 1274 |
+
"learning_rate": 8.593328092089737e-08,
|
| 1275 |
+
"logits/chosen": 2.027740001678467,
|
| 1276 |
+
"logits/rejected": 1.8771311044692993,
|
| 1277 |
+
"logps/chosen": -752.6876831054688,
|
| 1278 |
+
"logps/rejected": -762.588134765625,
|
| 1279 |
+
"loss": 0.1369,
|
| 1280 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 1281 |
+
"rewards/chosen": 1.0343955755233765,
|
| 1282 |
+
"rewards/diff": 0.0,
|
| 1283 |
+
"rewards/diff_abs": 0.0,
|
| 1284 |
+
"rewards/rejected": -2.3003334999084473,
|
| 1285 |
+
"rewards/student_margin": 3.3347294330596924,
|
| 1286 |
+
"rewards/teacher_margin": 0.0,
|
| 1287 |
+
"step": 700
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 0.94,
|
| 1291 |
+
"grad_norm": 10.125,
|
| 1292 |
+
"learning_rate": 5.8534133230599974e-08,
|
| 1293 |
+
"logits/chosen": 2.5199155807495117,
|
| 1294 |
+
"logits/rejected": 2.8889317512512207,
|
| 1295 |
+
"logps/chosen": -701.169921875,
|
| 1296 |
+
"logps/rejected": -910.6904296875,
|
| 1297 |
+
"loss": 0.1635,
|
| 1298 |
+
"rewards/accuracies": 0.966666579246521,
|
| 1299 |
+
"rewards/chosen": 1.4893516302108765,
|
| 1300 |
+
"rewards/diff": 0.0,
|
| 1301 |
+
"rewards/diff_abs": 0.0,
|
| 1302 |
+
"rewards/rejected": -2.9409499168395996,
|
| 1303 |
+
"rewards/student_margin": 4.430301189422607,
|
| 1304 |
+
"rewards/teacher_margin": 0.0,
|
| 1305 |
+
"step": 710
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 0.95,
|
| 1309 |
+
"grad_norm": 13.5,
|
| 1310 |
+
"learning_rate": 3.632990903662398e-08,
|
| 1311 |
+
"logits/chosen": 2.757976770401001,
|
| 1312 |
+
"logits/rejected": 2.7887349128723145,
|
| 1313 |
+
"logps/chosen": -732.608642578125,
|
| 1314 |
+
"logps/rejected": -873.4423828125,
|
| 1315 |
+
"loss": 0.1751,
|
| 1316 |
+
"rewards/accuracies": 0.9333333969116211,
|
| 1317 |
+
"rewards/chosen": 1.6327383518218994,
|
| 1318 |
+
"rewards/diff": 0.0,
|
| 1319 |
+
"rewards/diff_abs": 0.0,
|
| 1320 |
+
"rewards/rejected": -3.2143402099609375,
|
| 1321 |
+
"rewards/student_margin": 4.847078800201416,
|
| 1322 |
+
"rewards/teacher_margin": 0.0,
|
| 1323 |
+
"step": 720
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"epoch": 0.96,
|
| 1327 |
+
"grad_norm": 11.875,
|
| 1328 |
+
"learning_rate": 1.9367854236425065e-08,
|
| 1329 |
+
"logits/chosen": 2.469646453857422,
|
| 1330 |
+
"logits/rejected": 2.3219096660614014,
|
| 1331 |
+
"logps/chosen": -724.3168334960938,
|
| 1332 |
+
"logps/rejected": -875.6312255859375,
|
| 1333 |
+
"loss": 0.1225,
|
| 1334 |
+
"rewards/accuracies": 1.0,
|
| 1335 |
+
"rewards/chosen": 1.6782375574111938,
|
| 1336 |
+
"rewards/diff": 0.0,
|
| 1337 |
+
"rewards/diff_abs": 0.0,
|
| 1338 |
+
"rewards/rejected": -2.8410890102386475,
|
| 1339 |
+
"rewards/student_margin": 4.519326686859131,
|
| 1340 |
+
"rewards/teacher_margin": 0.0,
|
| 1341 |
+
"step": 730
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 0.98,
|
| 1345 |
+
"grad_norm": 3.140625,
|
| 1346 |
+
"learning_rate": 7.684060498630575e-09,
|
| 1347 |
+
"logits/chosen": 2.5574631690979004,
|
| 1348 |
+
"logits/rejected": 2.5168144702911377,
|
| 1349 |
+
"logps/chosen": -690.5845336914062,
|
| 1350 |
+
"logps/rejected": -850.1302490234375,
|
| 1351 |
+
"loss": 0.1444,
|
| 1352 |
+
"rewards/accuracies": 1.0,
|
| 1353 |
+
"rewards/chosen": 2.156329393386841,
|
| 1354 |
+
"rewards/diff": 0.0,
|
| 1355 |
+
"rewards/diff_abs": 0.0,
|
| 1356 |
+
"rewards/rejected": -2.7730555534362793,
|
| 1357 |
+
"rewards/student_margin": 4.929385185241699,
|
| 1358 |
+
"rewards/teacher_margin": 0.0,
|
| 1359 |
+
"step": 740
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 0.99,
|
| 1363 |
+
"grad_norm": 7.34375,
|
| 1364 |
+
"learning_rate": 1.3033884675933695e-09,
|
| 1365 |
+
"logits/chosen": 2.7296438217163086,
|
| 1366 |
+
"logits/rejected": 2.305525779724121,
|
| 1367 |
+
"logps/chosen": -714.5927734375,
|
| 1368 |
+
"logps/rejected": -803.5744018554688,
|
| 1369 |
+
"loss": 0.1035,
|
| 1370 |
+
"rewards/accuracies": 0.9333333969116211,
|
| 1371 |
+
"rewards/chosen": 1.8840978145599365,
|
| 1372 |
+
"rewards/diff": 0.0,
|
| 1373 |
+
"rewards/diff_abs": 0.0,
|
| 1374 |
+
"rewards/rejected": -2.6910955905914307,
|
| 1375 |
+
"rewards/student_margin": 4.575193405151367,
|
| 1376 |
+
"rewards/teacher_margin": 0.0,
|
| 1377 |
+
"step": 750
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 1.0,
|
| 1381 |
+
"step": 757,
|
| 1382 |
+
"total_flos": 0.0,
|
| 1383 |
+
"train_loss": 0.2174788669361936,
|
| 1384 |
+
"train_runtime": 882.7004,
|
| 1385 |
+
"train_samples_per_second": 20.572,
|
| 1386 |
+
"train_steps_per_second": 0.858
|
| 1387 |
+
}
|
| 1388 |
+
],
|
| 1389 |
+
"logging_steps": 10,
|
| 1390 |
+
"max_steps": 757,
|
| 1391 |
+
"num_input_tokens_seen": 0,
|
| 1392 |
+
"num_train_epochs": 1,
|
| 1393 |
+
"save_steps": 400,
|
| 1394 |
+
"total_flos": 0.0,
|
| 1395 |
+
"train_batch_size": 3,
|
| 1396 |
+
"trial_name": null,
|
| 1397 |
+
"trial_params": null
|
| 1398 |
+
}
|