Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +3332 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,3332 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 470,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.010638297872340425,
|
| 13 |
+
"grad_norm": 5.62164644869217,
|
| 14 |
+
"learning_rate": 1.7021276595744682e-06,
|
| 15 |
+
"loss": 0.8559,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.02127659574468085,
|
| 20 |
+
"grad_norm": 5.541790955372414,
|
| 21 |
+
"learning_rate": 3.4042553191489363e-06,
|
| 22 |
+
"loss": 0.8537,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.031914893617021274,
|
| 27 |
+
"grad_norm": 5.407901882681235,
|
| 28 |
+
"learning_rate": 5.106382978723404e-06,
|
| 29 |
+
"loss": 0.8414,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.0425531914893617,
|
| 34 |
+
"grad_norm": 4.03578755597478,
|
| 35 |
+
"learning_rate": 6.808510638297873e-06,
|
| 36 |
+
"loss": 0.8026,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.05319148936170213,
|
| 41 |
+
"grad_norm": 2.235595604545287,
|
| 42 |
+
"learning_rate": 8.510638297872341e-06,
|
| 43 |
+
"loss": 0.7586,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.06382978723404255,
|
| 48 |
+
"grad_norm": 3.6713198410625205,
|
| 49 |
+
"learning_rate": 1.0212765957446808e-05,
|
| 50 |
+
"loss": 0.7542,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.07446808510638298,
|
| 55 |
+
"grad_norm": 3.8053717898299015,
|
| 56 |
+
"learning_rate": 1.1914893617021277e-05,
|
| 57 |
+
"loss": 0.7525,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.0851063829787234,
|
| 62 |
+
"grad_norm": 3.735929746217291,
|
| 63 |
+
"learning_rate": 1.3617021276595745e-05,
|
| 64 |
+
"loss": 0.7146,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.09574468085106383,
|
| 69 |
+
"grad_norm": 3.4388722472257522,
|
| 70 |
+
"learning_rate": 1.5319148936170214e-05,
|
| 71 |
+
"loss": 0.7076,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.10638297872340426,
|
| 76 |
+
"grad_norm": 2.0576402741081745,
|
| 77 |
+
"learning_rate": 1.7021276595744682e-05,
|
| 78 |
+
"loss": 0.6804,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.11702127659574468,
|
| 83 |
+
"grad_norm": 2.194245295451263,
|
| 84 |
+
"learning_rate": 1.872340425531915e-05,
|
| 85 |
+
"loss": 0.6574,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.1276595744680851,
|
| 90 |
+
"grad_norm": 1.970180537002589,
|
| 91 |
+
"learning_rate": 2.0425531914893616e-05,
|
| 92 |
+
"loss": 0.6351,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.13829787234042554,
|
| 97 |
+
"grad_norm": 1.1822804981997306,
|
| 98 |
+
"learning_rate": 2.2127659574468088e-05,
|
| 99 |
+
"loss": 0.6205,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.14893617021276595,
|
| 104 |
+
"grad_norm": 1.1602034908038812,
|
| 105 |
+
"learning_rate": 2.3829787234042553e-05,
|
| 106 |
+
"loss": 0.6102,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.1595744680851064,
|
| 111 |
+
"grad_norm": 1.6168637308896043,
|
| 112 |
+
"learning_rate": 2.5531914893617025e-05,
|
| 113 |
+
"loss": 0.6058,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.1702127659574468,
|
| 118 |
+
"grad_norm": 0.9268293511359277,
|
| 119 |
+
"learning_rate": 2.723404255319149e-05,
|
| 120 |
+
"loss": 0.5865,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.18085106382978725,
|
| 125 |
+
"grad_norm": 1.3995612819702326,
|
| 126 |
+
"learning_rate": 2.8936170212765963e-05,
|
| 127 |
+
"loss": 0.5828,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.19148936170212766,
|
| 132 |
+
"grad_norm": 0.9491984900355026,
|
| 133 |
+
"learning_rate": 3.063829787234043e-05,
|
| 134 |
+
"loss": 0.5814,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.20212765957446807,
|
| 139 |
+
"grad_norm": 1.2461341580332967,
|
| 140 |
+
"learning_rate": 3.234042553191489e-05,
|
| 141 |
+
"loss": 0.5732,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.2127659574468085,
|
| 146 |
+
"grad_norm": 0.9013655593666975,
|
| 147 |
+
"learning_rate": 3.4042553191489365e-05,
|
| 148 |
+
"loss": 0.572,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.22340425531914893,
|
| 153 |
+
"grad_norm": 1.4385376985061347,
|
| 154 |
+
"learning_rate": 3.574468085106383e-05,
|
| 155 |
+
"loss": 0.5775,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.23404255319148937,
|
| 160 |
+
"grad_norm": 1.1316548066081353,
|
| 161 |
+
"learning_rate": 3.74468085106383e-05,
|
| 162 |
+
"loss": 0.5589,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.24468085106382978,
|
| 167 |
+
"grad_norm": 1.1348968485093565,
|
| 168 |
+
"learning_rate": 3.914893617021277e-05,
|
| 169 |
+
"loss": 0.5555,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.2553191489361702,
|
| 174 |
+
"grad_norm": 0.9593256794712123,
|
| 175 |
+
"learning_rate": 4.085106382978723e-05,
|
| 176 |
+
"loss": 0.5577,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.26595744680851063,
|
| 181 |
+
"grad_norm": 1.5810352206230227,
|
| 182 |
+
"learning_rate": 4.2553191489361704e-05,
|
| 183 |
+
"loss": 0.5506,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.2765957446808511,
|
| 188 |
+
"grad_norm": 1.0552617330606424,
|
| 189 |
+
"learning_rate": 4.4255319148936176e-05,
|
| 190 |
+
"loss": 0.5373,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.2872340425531915,
|
| 195 |
+
"grad_norm": 1.651827285414071,
|
| 196 |
+
"learning_rate": 4.595744680851065e-05,
|
| 197 |
+
"loss": 0.5465,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.2978723404255319,
|
| 202 |
+
"grad_norm": 1.8485214296335475,
|
| 203 |
+
"learning_rate": 4.765957446808511e-05,
|
| 204 |
+
"loss": 0.5413,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.30851063829787234,
|
| 209 |
+
"grad_norm": 0.8693252455438564,
|
| 210 |
+
"learning_rate": 4.936170212765958e-05,
|
| 211 |
+
"loss": 0.5416,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.3191489361702128,
|
| 216 |
+
"grad_norm": 2.3515640310759522,
|
| 217 |
+
"learning_rate": 5.106382978723405e-05,
|
| 218 |
+
"loss": 0.547,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.32978723404255317,
|
| 223 |
+
"grad_norm": 1.4756724644968964,
|
| 224 |
+
"learning_rate": 5.276595744680851e-05,
|
| 225 |
+
"loss": 0.5442,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.3404255319148936,
|
| 230 |
+
"grad_norm": 1.9541127476117117,
|
| 231 |
+
"learning_rate": 5.446808510638298e-05,
|
| 232 |
+
"loss": 0.5376,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.35106382978723405,
|
| 237 |
+
"grad_norm": 1.4385140916597212,
|
| 238 |
+
"learning_rate": 5.617021276595745e-05,
|
| 239 |
+
"loss": 0.5347,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.3617021276595745,
|
| 244 |
+
"grad_norm": 1.6177174100025706,
|
| 245 |
+
"learning_rate": 5.7872340425531925e-05,
|
| 246 |
+
"loss": 0.5283,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.3723404255319149,
|
| 251 |
+
"grad_norm": 1.5517064428797538,
|
| 252 |
+
"learning_rate": 5.9574468085106384e-05,
|
| 253 |
+
"loss": 0.5304,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.3829787234042553,
|
| 258 |
+
"grad_norm": 1.3062103414138244,
|
| 259 |
+
"learning_rate": 6.127659574468086e-05,
|
| 260 |
+
"loss": 0.5277,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.39361702127659576,
|
| 265 |
+
"grad_norm": 1.1717376687944345,
|
| 266 |
+
"learning_rate": 6.297872340425533e-05,
|
| 267 |
+
"loss": 0.5246,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.40425531914893614,
|
| 272 |
+
"grad_norm": 1.5784372654316825,
|
| 273 |
+
"learning_rate": 6.468085106382979e-05,
|
| 274 |
+
"loss": 0.5211,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.4148936170212766,
|
| 279 |
+
"grad_norm": 1.2570855650815358,
|
| 280 |
+
"learning_rate": 6.638297872340426e-05,
|
| 281 |
+
"loss": 0.531,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.425531914893617,
|
| 286 |
+
"grad_norm": 1.1994392890244334,
|
| 287 |
+
"learning_rate": 6.808510638297873e-05,
|
| 288 |
+
"loss": 0.5098,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.43617021276595747,
|
| 293 |
+
"grad_norm": 1.608445496858305,
|
| 294 |
+
"learning_rate": 6.97872340425532e-05,
|
| 295 |
+
"loss": 0.5176,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.44680851063829785,
|
| 300 |
+
"grad_norm": 1.6126071175066283,
|
| 301 |
+
"learning_rate": 7.148936170212766e-05,
|
| 302 |
+
"loss": 0.5181,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.4574468085106383,
|
| 307 |
+
"grad_norm": 1.3431412401241174,
|
| 308 |
+
"learning_rate": 7.319148936170213e-05,
|
| 309 |
+
"loss": 0.5135,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.46808510638297873,
|
| 314 |
+
"grad_norm": 1.5807072225527692,
|
| 315 |
+
"learning_rate": 7.48936170212766e-05,
|
| 316 |
+
"loss": 0.5185,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.4787234042553192,
|
| 321 |
+
"grad_norm": 2.0147344702502554,
|
| 322 |
+
"learning_rate": 7.659574468085108e-05,
|
| 323 |
+
"loss": 0.5281,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.48936170212765956,
|
| 328 |
+
"grad_norm": 1.4169773369597476,
|
| 329 |
+
"learning_rate": 7.829787234042553e-05,
|
| 330 |
+
"loss": 0.5104,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.5,
|
| 335 |
+
"grad_norm": 1.327014601864401,
|
| 336 |
+
"learning_rate": 8e-05,
|
| 337 |
+
"loss": 0.507,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.5106382978723404,
|
| 342 |
+
"grad_norm": 1.463329858386686,
|
| 343 |
+
"learning_rate": 7.999889681839899e-05,
|
| 344 |
+
"loss": 0.513,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.5212765957446809,
|
| 349 |
+
"grad_norm": 1.7226533644834205,
|
| 350 |
+
"learning_rate": 7.999558733444641e-05,
|
| 351 |
+
"loss": 0.5241,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.5319148936170213,
|
| 356 |
+
"grad_norm": 1.4625043342314172,
|
| 357 |
+
"learning_rate": 7.999007173069037e-05,
|
| 358 |
+
"loss": 0.4992,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.5425531914893617,
|
| 363 |
+
"grad_norm": 1.74638144425408,
|
| 364 |
+
"learning_rate": 7.998235031136648e-05,
|
| 365 |
+
"loss": 0.5129,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.5531914893617021,
|
| 370 |
+
"grad_norm": 1.156051684687154,
|
| 371 |
+
"learning_rate": 7.997242350238117e-05,
|
| 372 |
+
"loss": 0.5017,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.5638297872340425,
|
| 377 |
+
"grad_norm": 1.1484647512074981,
|
| 378 |
+
"learning_rate": 7.996029185128804e-05,
|
| 379 |
+
"loss": 0.5021,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.574468085106383,
|
| 384 |
+
"grad_norm": 1.7889194838438403,
|
| 385 |
+
"learning_rate": 7.994595602725781e-05,
|
| 386 |
+
"loss": 0.5137,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.5851063829787234,
|
| 391 |
+
"grad_norm": 1.2512024906203,
|
| 392 |
+
"learning_rate": 7.992941682104139e-05,
|
| 393 |
+
"loss": 0.4935,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.5957446808510638,
|
| 398 |
+
"grad_norm": 1.9138223030633696,
|
| 399 |
+
"learning_rate": 7.991067514492614e-05,
|
| 400 |
+
"loss": 0.5084,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.6063829787234043,
|
| 405 |
+
"grad_norm": 1.2776429355884351,
|
| 406 |
+
"learning_rate": 7.988973203268567e-05,
|
| 407 |
+
"loss": 0.5045,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.6170212765957447,
|
| 412 |
+
"grad_norm": 2.141401973332067,
|
| 413 |
+
"learning_rate": 7.986658863952281e-05,
|
| 414 |
+
"loss": 0.4992,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.6276595744680851,
|
| 419 |
+
"grad_norm": 1.454535688794727,
|
| 420 |
+
"learning_rate": 7.984124624200583e-05,
|
| 421 |
+
"loss": 0.5058,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.6382978723404256,
|
| 426 |
+
"grad_norm": 1.7846076730422646,
|
| 427 |
+
"learning_rate": 7.981370623799803e-05,
|
| 428 |
+
"loss": 0.5013,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.648936170212766,
|
| 433 |
+
"grad_norm": 1.4310227485726414,
|
| 434 |
+
"learning_rate": 7.978397014658075e-05,
|
| 435 |
+
"loss": 0.5079,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.6595744680851063,
|
| 440 |
+
"grad_norm": 1.4094274678104701,
|
| 441 |
+
"learning_rate": 7.97520396079694e-05,
|
| 442 |
+
"loss": 0.4882,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.6702127659574468,
|
| 447 |
+
"grad_norm": 1.6288107207366431,
|
| 448 |
+
"learning_rate": 7.971791638342313e-05,
|
| 449 |
+
"loss": 0.4977,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.6808510638297872,
|
| 454 |
+
"grad_norm": 1.0461330955205717,
|
| 455 |
+
"learning_rate": 7.96816023551476e-05,
|
| 456 |
+
"loss": 0.4932,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.6914893617021277,
|
| 461 |
+
"grad_norm": 1.8472118603297956,
|
| 462 |
+
"learning_rate": 7.964309952619121e-05,
|
| 463 |
+
"loss": 0.4956,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.7021276595744681,
|
| 468 |
+
"grad_norm": 1.3785935023957434,
|
| 469 |
+
"learning_rate": 7.96024100203346e-05,
|
| 470 |
+
"loss": 0.5072,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.7127659574468085,
|
| 475 |
+
"grad_norm": 1.321199071263235,
|
| 476 |
+
"learning_rate": 7.955953608197345e-05,
|
| 477 |
+
"loss": 0.4864,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.723404255319149,
|
| 482 |
+
"grad_norm": 1.2126321080139488,
|
| 483 |
+
"learning_rate": 7.951448007599478e-05,
|
| 484 |
+
"loss": 0.5016,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.7340425531914894,
|
| 489 |
+
"grad_norm": 1.1858808993782288,
|
| 490 |
+
"learning_rate": 7.946724448764644e-05,
|
| 491 |
+
"loss": 0.4943,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.7446808510638298,
|
| 496 |
+
"grad_norm": 1.0062356288958618,
|
| 497 |
+
"learning_rate": 7.94178319224e-05,
|
| 498 |
+
"loss": 0.491,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.7553191489361702,
|
| 503 |
+
"grad_norm": 1.9547579891672486,
|
| 504 |
+
"learning_rate": 7.936624510580712e-05,
|
| 505 |
+
"loss": 0.4879,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.7659574468085106,
|
| 510 |
+
"grad_norm": 1.0537143803568765,
|
| 511 |
+
"learning_rate": 7.931248688334915e-05,
|
| 512 |
+
"loss": 0.4916,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.776595744680851,
|
| 517 |
+
"grad_norm": 1.6913458685278957,
|
| 518 |
+
"learning_rate": 7.925656022028017e-05,
|
| 519 |
+
"loss": 0.4983,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.7872340425531915,
|
| 524 |
+
"grad_norm": 1.016587494594598,
|
| 525 |
+
"learning_rate": 7.919846820146348e-05,
|
| 526 |
+
"loss": 0.4964,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.7978723404255319,
|
| 531 |
+
"grad_norm": 1.6376648762774801,
|
| 532 |
+
"learning_rate": 7.913821403120139e-05,
|
| 533 |
+
"loss": 0.4916,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.8085106382978723,
|
| 538 |
+
"grad_norm": 1.2262934120029891,
|
| 539 |
+
"learning_rate": 7.90758010330585e-05,
|
| 540 |
+
"loss": 0.4851,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.8191489361702128,
|
| 545 |
+
"grad_norm": 1.381474776104582,
|
| 546 |
+
"learning_rate": 7.901123264967836e-05,
|
| 547 |
+
"loss": 0.4929,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.8297872340425532,
|
| 552 |
+
"grad_norm": 1.3905933735533444,
|
| 553 |
+
"learning_rate": 7.894451244259363e-05,
|
| 554 |
+
"loss": 0.4924,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.8404255319148937,
|
| 559 |
+
"grad_norm": 1.038270640845919,
|
| 560 |
+
"learning_rate": 7.887564409202953e-05,
|
| 561 |
+
"loss": 0.4893,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.851063829787234,
|
| 566 |
+
"grad_norm": 1.4032705044700307,
|
| 567 |
+
"learning_rate": 7.880463139670091e-05,
|
| 568 |
+
"loss": 0.485,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.8617021276595744,
|
| 573 |
+
"grad_norm": 1.1420897464578796,
|
| 574 |
+
"learning_rate": 7.873147827360273e-05,
|
| 575 |
+
"loss": 0.4867,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.8723404255319149,
|
| 580 |
+
"grad_norm": 1.667140313423623,
|
| 581 |
+
"learning_rate": 7.865618875779398e-05,
|
| 582 |
+
"loss": 0.4775,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.8829787234042553,
|
| 587 |
+
"grad_norm": 1.276151508655107,
|
| 588 |
+
"learning_rate": 7.857876700217508e-05,
|
| 589 |
+
"loss": 0.4786,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.8936170212765957,
|
| 594 |
+
"grad_norm": 1.1090496719024276,
|
| 595 |
+
"learning_rate": 7.849921727725882e-05,
|
| 596 |
+
"loss": 0.4722,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.9042553191489362,
|
| 601 |
+
"grad_norm": 0.991963411082399,
|
| 602 |
+
"learning_rate": 7.841754397093487e-05,
|
| 603 |
+
"loss": 0.478,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.9148936170212766,
|
| 608 |
+
"grad_norm": 1.731621852488109,
|
| 609 |
+
"learning_rate": 7.833375158822766e-05,
|
| 610 |
+
"loss": 0.4773,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.925531914893617,
|
| 615 |
+
"grad_norm": 1.3804671267005582,
|
| 616 |
+
"learning_rate": 7.824784475104795e-05,
|
| 617 |
+
"loss": 0.4787,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.9361702127659575,
|
| 622 |
+
"grad_norm": 1.1135118775414654,
|
| 623 |
+
"learning_rate": 7.815982819793784e-05,
|
| 624 |
+
"loss": 0.4772,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.9468085106382979,
|
| 629 |
+
"grad_norm": 1.9348194075183587,
|
| 630 |
+
"learning_rate": 7.806970678380943e-05,
|
| 631 |
+
"loss": 0.4768,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.9574468085106383,
|
| 636 |
+
"grad_norm": 0.9278131511467589,
|
| 637 |
+
"learning_rate": 7.797748547967701e-05,
|
| 638 |
+
"loss": 0.4768,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.9680851063829787,
|
| 643 |
+
"grad_norm": 1.5790755873743727,
|
| 644 |
+
"learning_rate": 7.788316937238287e-05,
|
| 645 |
+
"loss": 0.4777,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.9787234042553191,
|
| 650 |
+
"grad_norm": 1.1589932356056467,
|
| 651 |
+
"learning_rate": 7.778676366431676e-05,
|
| 652 |
+
"loss": 0.4793,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.9893617021276596,
|
| 657 |
+
"grad_norm": 0.9998066780828351,
|
| 658 |
+
"learning_rate": 7.768827367312882e-05,
|
| 659 |
+
"loss": 0.4736,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 1.0,
|
| 664 |
+
"grad_norm": 0.9964032491428947,
|
| 665 |
+
"learning_rate": 7.758770483143634e-05,
|
| 666 |
+
"loss": 0.4841,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 1.0106382978723405,
|
| 671 |
+
"grad_norm": 1.1948858983085593,
|
| 672 |
+
"learning_rate": 7.748506268652415e-05,
|
| 673 |
+
"loss": 0.461,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.0212765957446808,
|
| 678 |
+
"grad_norm": 0.7073134184929624,
|
| 679 |
+
"learning_rate": 7.73803529000385e-05,
|
| 680 |
+
"loss": 0.4536,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 1.0319148936170213,
|
| 685 |
+
"grad_norm": 1.1611831111477335,
|
| 686 |
+
"learning_rate": 7.727358124767491e-05,
|
| 687 |
+
"loss": 0.4612,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 1.0425531914893618,
|
| 692 |
+
"grad_norm": 1.0669064153598073,
|
| 693 |
+
"learning_rate": 7.71647536188595e-05,
|
| 694 |
+
"loss": 0.4653,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 1.053191489361702,
|
| 699 |
+
"grad_norm": 1.024613831141331,
|
| 700 |
+
"learning_rate": 7.705387601642416e-05,
|
| 701 |
+
"loss": 0.4584,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 1.0638297872340425,
|
| 706 |
+
"grad_norm": 1.4884061518621656,
|
| 707 |
+
"learning_rate": 7.694095455627542e-05,
|
| 708 |
+
"loss": 0.4608,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 1.074468085106383,
|
| 713 |
+
"grad_norm": 1.0414967862002833,
|
| 714 |
+
"learning_rate": 7.682599546705716e-05,
|
| 715 |
+
"loss": 0.4577,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.0851063829787233,
|
| 720 |
+
"grad_norm": 1.3598647542639628,
|
| 721 |
+
"learning_rate": 7.670900508980697e-05,
|
| 722 |
+
"loss": 0.4587,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 1.0957446808510638,
|
| 727 |
+
"grad_norm": 1.5748967124421995,
|
| 728 |
+
"learning_rate": 7.658998987760645e-05,
|
| 729 |
+
"loss": 0.4628,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 1.1063829787234043,
|
| 734 |
+
"grad_norm": 1.142076727768023,
|
| 735 |
+
"learning_rate": 7.646895639522518e-05,
|
| 736 |
+
"loss": 0.4551,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 1.1170212765957448,
|
| 741 |
+
"grad_norm": 1.3811816181823466,
|
| 742 |
+
"learning_rate": 7.634591131875875e-05,
|
| 743 |
+
"loss": 0.4594,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 1.127659574468085,
|
| 748 |
+
"grad_norm": 0.9160727711980033,
|
| 749 |
+
"learning_rate": 7.622086143526036e-05,
|
| 750 |
+
"loss": 0.4518,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 1.1382978723404256,
|
| 755 |
+
"grad_norm": 1.0917703855774061,
|
| 756 |
+
"learning_rate": 7.609381364236655e-05,
|
| 757 |
+
"loss": 0.4551,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.148936170212766,
|
| 762 |
+
"grad_norm": 1.43226233951233,
|
| 763 |
+
"learning_rate": 7.59647749479167e-05,
|
| 764 |
+
"loss": 0.4603,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 1.1595744680851063,
|
| 769 |
+
"grad_norm": 1.4134305371766467,
|
| 770 |
+
"learning_rate": 7.583375246956648e-05,
|
| 771 |
+
"loss": 0.4641,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 1.1702127659574468,
|
| 776 |
+
"grad_norm": 0.7629510146350382,
|
| 777 |
+
"learning_rate": 7.570075343439526e-05,
|
| 778 |
+
"loss": 0.4559,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 1.1808510638297873,
|
| 783 |
+
"grad_norm": 1.3551761060537464,
|
| 784 |
+
"learning_rate": 7.556578517850747e-05,
|
| 785 |
+
"loss": 0.4562,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 1.1914893617021276,
|
| 790 |
+
"grad_norm": 1.1143038754484502,
|
| 791 |
+
"learning_rate": 7.542885514662794e-05,
|
| 792 |
+
"loss": 0.454,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 1.202127659574468,
|
| 797 |
+
"grad_norm": 1.3365732739074656,
|
| 798 |
+
"learning_rate": 7.528997089169128e-05,
|
| 799 |
+
"loss": 0.4534,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.2127659574468086,
|
| 804 |
+
"grad_norm": 0.9865960113031025,
|
| 805 |
+
"learning_rate": 7.51491400744252e-05,
|
| 806 |
+
"loss": 0.4522,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 1.2234042553191489,
|
| 811 |
+
"grad_norm": 1.7137045691014816,
|
| 812 |
+
"learning_rate": 7.500637046292803e-05,
|
| 813 |
+
"loss": 0.4609,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 1.2340425531914894,
|
| 818 |
+
"grad_norm": 1.0423122076919606,
|
| 819 |
+
"learning_rate": 7.48616699322402e-05,
|
| 820 |
+
"loss": 0.4582,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 1.2446808510638299,
|
| 825 |
+
"grad_norm": 1.1897509411491205,
|
| 826 |
+
"learning_rate": 7.471504646390987e-05,
|
| 827 |
+
"loss": 0.4552,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 1.2553191489361701,
|
| 832 |
+
"grad_norm": 1.707159683909229,
|
| 833 |
+
"learning_rate": 7.456650814555267e-05,
|
| 834 |
+
"loss": 0.4668,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 1.2659574468085106,
|
| 839 |
+
"grad_norm": 0.7299856606632058,
|
| 840 |
+
"learning_rate": 7.441606317040558e-05,
|
| 841 |
+
"loss": 0.4523,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.2765957446808511,
|
| 846 |
+
"grad_norm": 1.7335201629368997,
|
| 847 |
+
"learning_rate": 7.426371983687503e-05,
|
| 848 |
+
"loss": 0.4645,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 1.2872340425531914,
|
| 853 |
+
"grad_norm": 1.0376836462353736,
|
| 854 |
+
"learning_rate": 7.410948654807916e-05,
|
| 855 |
+
"loss": 0.4567,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 1.297872340425532,
|
| 860 |
+
"grad_norm": 1.1899631811065923,
|
| 861 |
+
"learning_rate": 7.39533718113843e-05,
|
| 862 |
+
"loss": 0.4622,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 1.3085106382978724,
|
| 867 |
+
"grad_norm": 1.2350801182171,
|
| 868 |
+
"learning_rate": 7.379538423793568e-05,
|
| 869 |
+
"loss": 0.4557,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 1.3191489361702127,
|
| 874 |
+
"grad_norm": 0.8340292591613142,
|
| 875 |
+
"learning_rate": 7.363553254218253e-05,
|
| 876 |
+
"loss": 0.4492,
|
| 877 |
+
"step": 124
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 1.3297872340425532,
|
| 881 |
+
"grad_norm": 1.2214430144107857,
|
| 882 |
+
"learning_rate": 7.347382554139733e-05,
|
| 883 |
+
"loss": 0.4563,
|
| 884 |
+
"step": 125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.3404255319148937,
|
| 888 |
+
"grad_norm": 0.7515274378448177,
|
| 889 |
+
"learning_rate": 7.331027215518949e-05,
|
| 890 |
+
"loss": 0.4574,
|
| 891 |
+
"step": 126
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 1.351063829787234,
|
| 895 |
+
"grad_norm": 1.0593819816413108,
|
| 896 |
+
"learning_rate": 7.31448814050133e-05,
|
| 897 |
+
"loss": 0.4586,
|
| 898 |
+
"step": 127
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 1.3617021276595744,
|
| 902 |
+
"grad_norm": 0.8257535268103262,
|
| 903 |
+
"learning_rate": 7.297766241367041e-05,
|
| 904 |
+
"loss": 0.4525,
|
| 905 |
+
"step": 128
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 1.372340425531915,
|
| 909 |
+
"grad_norm": 0.8489466970658132,
|
| 910 |
+
"learning_rate": 7.280862440480658e-05,
|
| 911 |
+
"loss": 0.4486,
|
| 912 |
+
"step": 129
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 1.3829787234042552,
|
| 916 |
+
"grad_norm": 1.0407070508163005,
|
| 917 |
+
"learning_rate": 7.263777670240282e-05,
|
| 918 |
+
"loss": 0.4518,
|
| 919 |
+
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 1.3936170212765957,
|
| 923 |
+
"grad_norm": 1.0870662063487706,
|
| 924 |
+
"learning_rate": 7.246512873026125e-05,
|
| 925 |
+
"loss": 0.4436,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.4042553191489362,
|
| 930 |
+
"grad_norm": 1.040969884122318,
|
| 931 |
+
"learning_rate": 7.229069001148518e-05,
|
| 932 |
+
"loss": 0.4376,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 1.4148936170212765,
|
| 937 |
+
"grad_norm": 0.8210287133544034,
|
| 938 |
+
"learning_rate": 7.211447016795388e-05,
|
| 939 |
+
"loss": 0.4459,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 1.425531914893617,
|
| 944 |
+
"grad_norm": 0.6924416024292802,
|
| 945 |
+
"learning_rate": 7.193647891979177e-05,
|
| 946 |
+
"loss": 0.4492,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 1.4361702127659575,
|
| 951 |
+
"grad_norm": 0.691504635298063,
|
| 952 |
+
"learning_rate": 7.17567260848324e-05,
|
| 953 |
+
"loss": 0.4382,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 1.4468085106382977,
|
| 958 |
+
"grad_norm": 0.557066728314828,
|
| 959 |
+
"learning_rate": 7.157522157807675e-05,
|
| 960 |
+
"loss": 0.4475,
|
| 961 |
+
"step": 136
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 1.4574468085106382,
|
| 965 |
+
"grad_norm": 0.578188119392695,
|
| 966 |
+
"learning_rate": 7.139197541114645e-05,
|
| 967 |
+
"loss": 0.4416,
|
| 968 |
+
"step": 137
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 1.4680851063829787,
|
| 972 |
+
"grad_norm": 0.8694472070990833,
|
| 973 |
+
"learning_rate": 7.120699769173149e-05,
|
| 974 |
+
"loss": 0.4492,
|
| 975 |
+
"step": 138
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 1.4787234042553192,
|
| 979 |
+
"grad_norm": 1.0414676752540923,
|
| 980 |
+
"learning_rate": 7.10202986230327e-05,
|
| 981 |
+
"loss": 0.4371,
|
| 982 |
+
"step": 139
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 1.4893617021276595,
|
| 986 |
+
"grad_norm": 0.9048472155799596,
|
| 987 |
+
"learning_rate": 7.083188850319895e-05,
|
| 988 |
+
"loss": 0.4411,
|
| 989 |
+
"step": 140
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 1.5,
|
| 993 |
+
"grad_norm": 0.9161726375221393,
|
| 994 |
+
"learning_rate": 7.064177772475912e-05,
|
| 995 |
+
"loss": 0.4529,
|
| 996 |
+
"step": 141
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 1.5106382978723403,
|
| 1000 |
+
"grad_norm": 1.1657596780724042,
|
| 1001 |
+
"learning_rate": 7.044997677404888e-05,
|
| 1002 |
+
"loss": 0.4433,
|
| 1003 |
+
"step": 142
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 1.521276595744681,
|
| 1007 |
+
"grad_norm": 0.8278836619484179,
|
| 1008 |
+
"learning_rate": 7.025649623063223e-05,
|
| 1009 |
+
"loss": 0.4432,
|
| 1010 |
+
"step": 143
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 1.5319148936170213,
|
| 1014 |
+
"grad_norm": 0.6982292221675673,
|
| 1015 |
+
"learning_rate": 7.006134676671791e-05,
|
| 1016 |
+
"loss": 0.4498,
|
| 1017 |
+
"step": 144
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 1.5425531914893615,
|
| 1021 |
+
"grad_norm": 1.1530549987417396,
|
| 1022 |
+
"learning_rate": 6.986453914657083e-05,
|
| 1023 |
+
"loss": 0.4484,
|
| 1024 |
+
"step": 145
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 1.5531914893617023,
|
| 1028 |
+
"grad_norm": 1.2290110194371844,
|
| 1029 |
+
"learning_rate": 6.96660842259183e-05,
|
| 1030 |
+
"loss": 0.4541,
|
| 1031 |
+
"step": 146
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 1.5638297872340425,
|
| 1035 |
+
"grad_norm": 0.6040374185478055,
|
| 1036 |
+
"learning_rate": 6.946599295135116e-05,
|
| 1037 |
+
"loss": 0.4356,
|
| 1038 |
+
"step": 147
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 1.574468085106383,
|
| 1042 |
+
"grad_norm": 0.5974414549578787,
|
| 1043 |
+
"learning_rate": 6.926427635972003e-05,
|
| 1044 |
+
"loss": 0.4374,
|
| 1045 |
+
"step": 148
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 1.5851063829787235,
|
| 1049 |
+
"grad_norm": 0.9321005443290319,
|
| 1050 |
+
"learning_rate": 6.906094557752654e-05,
|
| 1051 |
+
"loss": 0.4432,
|
| 1052 |
+
"step": 149
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 1.5957446808510638,
|
| 1056 |
+
"grad_norm": 1.0909193280528477,
|
| 1057 |
+
"learning_rate": 6.885601182030958e-05,
|
| 1058 |
+
"loss": 0.4474,
|
| 1059 |
+
"step": 150
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 1.6063829787234043,
|
| 1063 |
+
"grad_norm": 0.7679852263478857,
|
| 1064 |
+
"learning_rate": 6.864948639202667e-05,
|
| 1065 |
+
"loss": 0.4442,
|
| 1066 |
+
"step": 151
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 1.6170212765957448,
|
| 1070 |
+
"grad_norm": 0.7550553221749178,
|
| 1071 |
+
"learning_rate": 6.844138068443043e-05,
|
| 1072 |
+
"loss": 0.4452,
|
| 1073 |
+
"step": 152
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 1.627659574468085,
|
| 1077 |
+
"grad_norm": 1.1202478929189756,
|
| 1078 |
+
"learning_rate": 6.823170617644029e-05,
|
| 1079 |
+
"loss": 0.4391,
|
| 1080 |
+
"step": 153
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 1.6382978723404256,
|
| 1084 |
+
"grad_norm": 0.8957986675565583,
|
| 1085 |
+
"learning_rate": 6.802047443350915e-05,
|
| 1086 |
+
"loss": 0.44,
|
| 1087 |
+
"step": 154
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 1.648936170212766,
|
| 1091 |
+
"grad_norm": 0.7670887365074628,
|
| 1092 |
+
"learning_rate": 6.78076971069857e-05,
|
| 1093 |
+
"loss": 0.4427,
|
| 1094 |
+
"step": 155
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 1.6595744680851063,
|
| 1098 |
+
"grad_norm": 0.7780853531387013,
|
| 1099 |
+
"learning_rate": 6.759338593347148e-05,
|
| 1100 |
+
"loss": 0.4398,
|
| 1101 |
+
"step": 156
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 1.6702127659574468,
|
| 1105 |
+
"grad_norm": 0.7615691990716689,
|
| 1106 |
+
"learning_rate": 6.737755273417367e-05,
|
| 1107 |
+
"loss": 0.4456,
|
| 1108 |
+
"step": 157
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 1.6808510638297873,
|
| 1112 |
+
"grad_norm": 0.9908956798698196,
|
| 1113 |
+
"learning_rate": 6.716020941425302e-05,
|
| 1114 |
+
"loss": 0.4511,
|
| 1115 |
+
"step": 158
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 1.6914893617021276,
|
| 1119 |
+
"grad_norm": 1.1748417677944305,
|
| 1120 |
+
"learning_rate": 6.694136796216706e-05,
|
| 1121 |
+
"loss": 0.4442,
|
| 1122 |
+
"step": 159
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 1.702127659574468,
|
| 1126 |
+
"grad_norm": 0.7089766235204993,
|
| 1127 |
+
"learning_rate": 6.672104044900901e-05,
|
| 1128 |
+
"loss": 0.4437,
|
| 1129 |
+
"step": 160
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 1.7127659574468086,
|
| 1133 |
+
"grad_norm": 0.6183339909308402,
|
| 1134 |
+
"learning_rate": 6.649923902784178e-05,
|
| 1135 |
+
"loss": 0.4391,
|
| 1136 |
+
"step": 161
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 1.7234042553191489,
|
| 1140 |
+
"grad_norm": 0.6684140166542728,
|
| 1141 |
+
"learning_rate": 6.627597593302772e-05,
|
| 1142 |
+
"loss": 0.4543,
|
| 1143 |
+
"step": 162
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 1.7340425531914894,
|
| 1147 |
+
"grad_norm": 0.8241026926395437,
|
| 1148 |
+
"learning_rate": 6.605126347955376e-05,
|
| 1149 |
+
"loss": 0.446,
|
| 1150 |
+
"step": 163
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 1.7446808510638299,
|
| 1154 |
+
"grad_norm": 0.8675844769376447,
|
| 1155 |
+
"learning_rate": 6.58251140623521e-05,
|
| 1156 |
+
"loss": 0.4394,
|
| 1157 |
+
"step": 164
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 1.7553191489361701,
|
| 1161 |
+
"grad_norm": 0.7195091122160452,
|
| 1162 |
+
"learning_rate": 6.559754015561655e-05,
|
| 1163 |
+
"loss": 0.4361,
|
| 1164 |
+
"step": 165
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 1.7659574468085106,
|
| 1168 |
+
"grad_norm": 0.7586387307080512,
|
| 1169 |
+
"learning_rate": 6.536855431211445e-05,
|
| 1170 |
+
"loss": 0.4452,
|
| 1171 |
+
"step": 166
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 1.7765957446808511,
|
| 1175 |
+
"grad_norm": 0.8241109857261576,
|
| 1176 |
+
"learning_rate": 6.513816916249427e-05,
|
| 1177 |
+
"loss": 0.4462,
|
| 1178 |
+
"step": 167
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 1.7872340425531914,
|
| 1182 |
+
"grad_norm": 0.7828717831576161,
|
| 1183 |
+
"learning_rate": 6.490639741458891e-05,
|
| 1184 |
+
"loss": 0.4387,
|
| 1185 |
+
"step": 168
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 1.797872340425532,
|
| 1189 |
+
"grad_norm": 0.9841281333161127,
|
| 1190 |
+
"learning_rate": 6.46732518527148e-05,
|
| 1191 |
+
"loss": 0.4433,
|
| 1192 |
+
"step": 169
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 1.8085106382978724,
|
| 1196 |
+
"grad_norm": 0.8948930714592511,
|
| 1197 |
+
"learning_rate": 6.443874533696662e-05,
|
| 1198 |
+
"loss": 0.4329,
|
| 1199 |
+
"step": 170
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 1.8191489361702127,
|
| 1203 |
+
"grad_norm": 0.6167915147089398,
|
| 1204 |
+
"learning_rate": 6.420289080250804e-05,
|
| 1205 |
+
"loss": 0.4423,
|
| 1206 |
+
"step": 171
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 1.8297872340425532,
|
| 1210 |
+
"grad_norm": 0.7688689128970628,
|
| 1211 |
+
"learning_rate": 6.396570125885823e-05,
|
| 1212 |
+
"loss": 0.448,
|
| 1213 |
+
"step": 172
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 1.8404255319148937,
|
| 1217 |
+
"grad_norm": 0.7366333884582301,
|
| 1218 |
+
"learning_rate": 6.372718978917421e-05,
|
| 1219 |
+
"loss": 0.4421,
|
| 1220 |
+
"step": 173
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 1.851063829787234,
|
| 1224 |
+
"grad_norm": 0.6292983833226043,
|
| 1225 |
+
"learning_rate": 6.348736954952923e-05,
|
| 1226 |
+
"loss": 0.4345,
|
| 1227 |
+
"step": 174
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 1.8617021276595744,
|
| 1231 |
+
"grad_norm": 0.6094161942129656,
|
| 1232 |
+
"learning_rate": 6.324625376818707e-05,
|
| 1233 |
+
"loss": 0.4289,
|
| 1234 |
+
"step": 175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 1.872340425531915,
|
| 1238 |
+
"grad_norm": 0.5223920021177019,
|
| 1239 |
+
"learning_rate": 6.300385574487242e-05,
|
| 1240 |
+
"loss": 0.4281,
|
| 1241 |
+
"step": 176
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 1.8829787234042552,
|
| 1245 |
+
"grad_norm": 0.43105768007225675,
|
| 1246 |
+
"learning_rate": 6.276018885003727e-05,
|
| 1247 |
+
"loss": 0.43,
|
| 1248 |
+
"step": 177
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 1.8936170212765957,
|
| 1252 |
+
"grad_norm": 0.4154261253991378,
|
| 1253 |
+
"learning_rate": 6.251526652412335e-05,
|
| 1254 |
+
"loss": 0.4272,
|
| 1255 |
+
"step": 178
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 1.9042553191489362,
|
| 1259 |
+
"grad_norm": 0.5129738950204691,
|
| 1260 |
+
"learning_rate": 6.226910227682087e-05,
|
| 1261 |
+
"loss": 0.4404,
|
| 1262 |
+
"step": 179
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 1.9148936170212765,
|
| 1266 |
+
"grad_norm": 0.4069384888530334,
|
| 1267 |
+
"learning_rate": 6.202170968632324e-05,
|
| 1268 |
+
"loss": 0.4379,
|
| 1269 |
+
"step": 180
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 1.925531914893617,
|
| 1273 |
+
"grad_norm": 0.4346407097225191,
|
| 1274 |
+
"learning_rate": 6.177310239857815e-05,
|
| 1275 |
+
"loss": 0.4393,
|
| 1276 |
+
"step": 181
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 1.9361702127659575,
|
| 1280 |
+
"grad_norm": 0.7030752370713033,
|
| 1281 |
+
"learning_rate": 6.152329412653491e-05,
|
| 1282 |
+
"loss": 0.4359,
|
| 1283 |
+
"step": 182
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 1.9468085106382977,
|
| 1287 |
+
"grad_norm": 0.7931293701516278,
|
| 1288 |
+
"learning_rate": 6.127229864938798e-05,
|
| 1289 |
+
"loss": 0.4324,
|
| 1290 |
+
"step": 183
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 1.9574468085106385,
|
| 1294 |
+
"grad_norm": 0.8569387442215538,
|
| 1295 |
+
"learning_rate": 6.1020129811816985e-05,
|
| 1296 |
+
"loss": 0.439,
|
| 1297 |
+
"step": 184
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 1.9680851063829787,
|
| 1301 |
+
"grad_norm": 0.9366916666402957,
|
| 1302 |
+
"learning_rate": 6.076680152322302e-05,
|
| 1303 |
+
"loss": 0.4365,
|
| 1304 |
+
"step": 185
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 1.978723404255319,
|
| 1308 |
+
"grad_norm": 0.9953725617355069,
|
| 1309 |
+
"learning_rate": 6.051232775696143e-05,
|
| 1310 |
+
"loss": 0.436,
|
| 1311 |
+
"step": 186
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 1.9893617021276597,
|
| 1315 |
+
"grad_norm": 1.0794213458303916,
|
| 1316 |
+
"learning_rate": 6.025672254957106e-05,
|
| 1317 |
+
"loss": 0.4423,
|
| 1318 |
+
"step": 187
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 2.0,
|
| 1322 |
+
"grad_norm": 0.724853495699258,
|
| 1323 |
+
"learning_rate": 6.000000000000001e-05,
|
| 1324 |
+
"loss": 0.4317,
|
| 1325 |
+
"step": 188
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 2.0106382978723403,
|
| 1329 |
+
"grad_norm": 0.5489002192429712,
|
| 1330 |
+
"learning_rate": 5.9742174268827936e-05,
|
| 1331 |
+
"loss": 0.4137,
|
| 1332 |
+
"step": 189
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 2.021276595744681,
|
| 1336 |
+
"grad_norm": 0.5979267122970597,
|
| 1337 |
+
"learning_rate": 5.948325957748498e-05,
|
| 1338 |
+
"loss": 0.4107,
|
| 1339 |
+
"step": 190
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 2.0319148936170213,
|
| 1343 |
+
"grad_norm": 0.6769157942235892,
|
| 1344 |
+
"learning_rate": 5.9223270207467355e-05,
|
| 1345 |
+
"loss": 0.415,
|
| 1346 |
+
"step": 191
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 2.0425531914893615,
|
| 1350 |
+
"grad_norm": 0.694157788349472,
|
| 1351 |
+
"learning_rate": 5.896222049954951e-05,
|
| 1352 |
+
"loss": 0.4138,
|
| 1353 |
+
"step": 192
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 2.0531914893617023,
|
| 1357 |
+
"grad_norm": 0.7900658328855281,
|
| 1358 |
+
"learning_rate": 5.870012485299318e-05,
|
| 1359 |
+
"loss": 0.415,
|
| 1360 |
+
"step": 193
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 2.0638297872340425,
|
| 1364 |
+
"grad_norm": 0.6937916417075501,
|
| 1365 |
+
"learning_rate": 5.843699772475312e-05,
|
| 1366 |
+
"loss": 0.4132,
|
| 1367 |
+
"step": 194
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 2.074468085106383,
|
| 1371 |
+
"grad_norm": 0.6698631990082241,
|
| 1372 |
+
"learning_rate": 5.8172853628679676e-05,
|
| 1373 |
+
"loss": 0.4121,
|
| 1374 |
+
"step": 195
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 2.0851063829787235,
|
| 1378 |
+
"grad_norm": 0.6656108376135776,
|
| 1379 |
+
"learning_rate": 5.790770713471816e-05,
|
| 1380 |
+
"loss": 0.4127,
|
| 1381 |
+
"step": 196
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 2.095744680851064,
|
| 1385 |
+
"grad_norm": 0.5836102935781896,
|
| 1386 |
+
"learning_rate": 5.764157286810527e-05,
|
| 1387 |
+
"loss": 0.4068,
|
| 1388 |
+
"step": 197
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 2.106382978723404,
|
| 1392 |
+
"grad_norm": 0.6583841935134351,
|
| 1393 |
+
"learning_rate": 5.7374465508562324e-05,
|
| 1394 |
+
"loss": 0.4063,
|
| 1395 |
+
"step": 198
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 2.117021276595745,
|
| 1399 |
+
"grad_norm": 0.5654983891994843,
|
| 1400 |
+
"learning_rate": 5.710639978948555e-05,
|
| 1401 |
+
"loss": 0.412,
|
| 1402 |
+
"step": 199
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 2.127659574468085,
|
| 1406 |
+
"grad_norm": 0.423035149723838,
|
| 1407 |
+
"learning_rate": 5.6837390497133406e-05,
|
| 1408 |
+
"loss": 0.4038,
|
| 1409 |
+
"step": 200
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 2.1382978723404253,
|
| 1413 |
+
"grad_norm": 0.6023613024743625,
|
| 1414 |
+
"learning_rate": 5.6567452469810984e-05,
|
| 1415 |
+
"loss": 0.4072,
|
| 1416 |
+
"step": 201
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 2.148936170212766,
|
| 1420 |
+
"grad_norm": 0.5872005390949395,
|
| 1421 |
+
"learning_rate": 5.629660059705153e-05,
|
| 1422 |
+
"loss": 0.4064,
|
| 1423 |
+
"step": 202
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 2.1595744680851063,
|
| 1427 |
+
"grad_norm": 0.38488472443140903,
|
| 1428 |
+
"learning_rate": 5.602484981879519e-05,
|
| 1429 |
+
"loss": 0.4088,
|
| 1430 |
+
"step": 203
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 2.1702127659574466,
|
| 1434 |
+
"grad_norm": 0.34867056077161274,
|
| 1435 |
+
"learning_rate": 5.5752215124564895e-05,
|
| 1436 |
+
"loss": 0.4165,
|
| 1437 |
+
"step": 204
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 2.1808510638297873,
|
| 1441 |
+
"grad_norm": 0.5189695063587118,
|
| 1442 |
+
"learning_rate": 5.547871155263955e-05,
|
| 1443 |
+
"loss": 0.403,
|
| 1444 |
+
"step": 205
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 2.1914893617021276,
|
| 1448 |
+
"grad_norm": 0.5349240957528729,
|
| 1449 |
+
"learning_rate": 5.5204354189224596e-05,
|
| 1450 |
+
"loss": 0.4076,
|
| 1451 |
+
"step": 206
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 2.202127659574468,
|
| 1455 |
+
"grad_norm": 0.4097332185439771,
|
| 1456 |
+
"learning_rate": 5.492915816761979e-05,
|
| 1457 |
+
"loss": 0.4055,
|
| 1458 |
+
"step": 207
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 2.2127659574468086,
|
| 1462 |
+
"grad_norm": 0.29969289622943385,
|
| 1463 |
+
"learning_rate": 5.465313866738454e-05,
|
| 1464 |
+
"loss": 0.4102,
|
| 1465 |
+
"step": 208
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 2.223404255319149,
|
| 1469 |
+
"grad_norm": 0.3210843691588522,
|
| 1470 |
+
"learning_rate": 5.4376310913500514e-05,
|
| 1471 |
+
"loss": 0.4082,
|
| 1472 |
+
"step": 209
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 2.2340425531914896,
|
| 1476 |
+
"grad_norm": 0.3342881154938907,
|
| 1477 |
+
"learning_rate": 5.409869017553199e-05,
|
| 1478 |
+
"loss": 0.4043,
|
| 1479 |
+
"step": 210
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 2.24468085106383,
|
| 1483 |
+
"grad_norm": 0.3152754379746835,
|
| 1484 |
+
"learning_rate": 5.382029176678345e-05,
|
| 1485 |
+
"loss": 0.4091,
|
| 1486 |
+
"step": 211
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 2.25531914893617,
|
| 1490 |
+
"grad_norm": 0.27611840786578706,
|
| 1491 |
+
"learning_rate": 5.354113104345503e-05,
|
| 1492 |
+
"loss": 0.4039,
|
| 1493 |
+
"step": 212
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 2.2659574468085104,
|
| 1497 |
+
"grad_norm": 0.34803624809617745,
|
| 1498 |
+
"learning_rate": 5.326122340379539e-05,
|
| 1499 |
+
"loss": 0.4063,
|
| 1500 |
+
"step": 213
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 2.276595744680851,
|
| 1504 |
+
"grad_norm": 0.36164081157802974,
|
| 1505 |
+
"learning_rate": 5.2980584287252456e-05,
|
| 1506 |
+
"loss": 0.4054,
|
| 1507 |
+
"step": 214
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 2.2872340425531914,
|
| 1511 |
+
"grad_norm": 0.28080091164784143,
|
| 1512 |
+
"learning_rate": 5.269922917362171e-05,
|
| 1513 |
+
"loss": 0.4137,
|
| 1514 |
+
"step": 215
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 2.297872340425532,
|
| 1518 |
+
"grad_norm": 0.3123559500548175,
|
| 1519 |
+
"learning_rate": 5.241717358219239e-05,
|
| 1520 |
+
"loss": 0.4136,
|
| 1521 |
+
"step": 216
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 2.3085106382978724,
|
| 1525 |
+
"grad_norm": 0.33068468215663827,
|
| 1526 |
+
"learning_rate": 5.213443307089144e-05,
|
| 1527 |
+
"loss": 0.4067,
|
| 1528 |
+
"step": 217
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 2.3191489361702127,
|
| 1532 |
+
"grad_norm": 0.2984606793541823,
|
| 1533 |
+
"learning_rate": 5.1851023235425366e-05,
|
| 1534 |
+
"loss": 0.3982,
|
| 1535 |
+
"step": 218
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 2.329787234042553,
|
| 1539 |
+
"grad_norm": 0.3545208178999491,
|
| 1540 |
+
"learning_rate": 5.156695970841997e-05,
|
| 1541 |
+
"loss": 0.41,
|
| 1542 |
+
"step": 219
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 2.3404255319148937,
|
| 1546 |
+
"grad_norm": 0.3590473616023012,
|
| 1547 |
+
"learning_rate": 5.128225815855805e-05,
|
| 1548 |
+
"loss": 0.4119,
|
| 1549 |
+
"step": 220
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 2.351063829787234,
|
| 1553 |
+
"grad_norm": 0.3815124462451615,
|
| 1554 |
+
"learning_rate": 5.099693428971522e-05,
|
| 1555 |
+
"loss": 0.4117,
|
| 1556 |
+
"step": 221
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 2.3617021276595747,
|
| 1560 |
+
"grad_norm": 0.3777633888640618,
|
| 1561 |
+
"learning_rate": 5.0711003840093583e-05,
|
| 1562 |
+
"loss": 0.4064,
|
| 1563 |
+
"step": 222
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 2.372340425531915,
|
| 1567 |
+
"grad_norm": 0.42841655218775976,
|
| 1568 |
+
"learning_rate": 5.042448258135371e-05,
|
| 1569 |
+
"loss": 0.4059,
|
| 1570 |
+
"step": 223
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 2.382978723404255,
|
| 1574 |
+
"grad_norm": 0.3957041429267785,
|
| 1575 |
+
"learning_rate": 5.013738631774463e-05,
|
| 1576 |
+
"loss": 0.4082,
|
| 1577 |
+
"step": 224
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 2.393617021276596,
|
| 1581 |
+
"grad_norm": 0.23548416366582237,
|
| 1582 |
+
"learning_rate": 4.984973088523216e-05,
|
| 1583 |
+
"loss": 0.4069,
|
| 1584 |
+
"step": 225
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 2.404255319148936,
|
| 1588 |
+
"grad_norm": 0.3532835233128077,
|
| 1589 |
+
"learning_rate": 4.9561532150625305e-05,
|
| 1590 |
+
"loss": 0.4123,
|
| 1591 |
+
"step": 226
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 2.4148936170212765,
|
| 1595 |
+
"grad_norm": 0.41182185945267635,
|
| 1596 |
+
"learning_rate": 4.927280601070114e-05,
|
| 1597 |
+
"loss": 0.4043,
|
| 1598 |
+
"step": 227
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 2.425531914893617,
|
| 1602 |
+
"grad_norm": 0.3444115639955698,
|
| 1603 |
+
"learning_rate": 4.898356839132793e-05,
|
| 1604 |
+
"loss": 0.4085,
|
| 1605 |
+
"step": 228
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 2.4361702127659575,
|
| 1609 |
+
"grad_norm": 0.3641718763669911,
|
| 1610 |
+
"learning_rate": 4.869383524658668e-05,
|
| 1611 |
+
"loss": 0.4044,
|
| 1612 |
+
"step": 229
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 2.4468085106382977,
|
| 1616 |
+
"grad_norm": 0.31325645363197496,
|
| 1617 |
+
"learning_rate": 4.840362255789112e-05,
|
| 1618 |
+
"loss": 0.4088,
|
| 1619 |
+
"step": 230
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 2.4574468085106385,
|
| 1623 |
+
"grad_norm": 0.22787584781999606,
|
| 1624 |
+
"learning_rate": 4.811294633310617e-05,
|
| 1625 |
+
"loss": 0.4072,
|
| 1626 |
+
"step": 231
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 2.4680851063829787,
|
| 1630 |
+
"grad_norm": 0.34212691771738396,
|
| 1631 |
+
"learning_rate": 4.782182260566498e-05,
|
| 1632 |
+
"loss": 0.4029,
|
| 1633 |
+
"step": 232
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 2.478723404255319,
|
| 1637 |
+
"grad_norm": 0.4204486154126815,
|
| 1638 |
+
"learning_rate": 4.7530267433684546e-05,
|
| 1639 |
+
"loss": 0.4082,
|
| 1640 |
+
"step": 233
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 2.4893617021276597,
|
| 1644 |
+
"grad_norm": 0.4183586609670886,
|
| 1645 |
+
"learning_rate": 4.723829689907993e-05,
|
| 1646 |
+
"loss": 0.4069,
|
| 1647 |
+
"step": 234
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 2.5,
|
| 1651 |
+
"grad_norm": 0.361420219603079,
|
| 1652 |
+
"learning_rate": 4.694592710667723e-05,
|
| 1653 |
+
"loss": 0.4051,
|
| 1654 |
+
"step": 235
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 2.5106382978723403,
|
| 1658 |
+
"grad_norm": 0.3230255284520481,
|
| 1659 |
+
"learning_rate": 4.665317418332521e-05,
|
| 1660 |
+
"loss": 0.4032,
|
| 1661 |
+
"step": 236
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 2.521276595744681,
|
| 1665 |
+
"grad_norm": 0.370284885999409,
|
| 1666 |
+
"learning_rate": 4.6360054277005826e-05,
|
| 1667 |
+
"loss": 0.406,
|
| 1668 |
+
"step": 237
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 2.5319148936170213,
|
| 1672 |
+
"grad_norm": 0.30097725112497414,
|
| 1673 |
+
"learning_rate": 4.606658355594344e-05,
|
| 1674 |
+
"loss": 0.3993,
|
| 1675 |
+
"step": 238
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 2.5425531914893615,
|
| 1679 |
+
"grad_norm": 0.32435130844112303,
|
| 1680 |
+
"learning_rate": 4.577277820771307e-05,
|
| 1681 |
+
"loss": 0.4102,
|
| 1682 |
+
"step": 239
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 2.5531914893617023,
|
| 1686 |
+
"grad_norm": 0.3080820117631687,
|
| 1687 |
+
"learning_rate": 4.5478654438347414e-05,
|
| 1688 |
+
"loss": 0.4026,
|
| 1689 |
+
"step": 240
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 2.5638297872340425,
|
| 1693 |
+
"grad_norm": 0.39816170983067595,
|
| 1694 |
+
"learning_rate": 4.518422847144304e-05,
|
| 1695 |
+
"loss": 0.406,
|
| 1696 |
+
"step": 241
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 2.574468085106383,
|
| 1700 |
+
"grad_norm": 0.3800310483317905,
|
| 1701 |
+
"learning_rate": 4.488951654726539e-05,
|
| 1702 |
+
"loss": 0.4073,
|
| 1703 |
+
"step": 242
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 2.5851063829787235,
|
| 1707 |
+
"grad_norm": 0.2671985596247953,
|
| 1708 |
+
"learning_rate": 4.4594534921853096e-05,
|
| 1709 |
+
"loss": 0.4038,
|
| 1710 |
+
"step": 243
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 2.595744680851064,
|
| 1714 |
+
"grad_norm": 0.3417581375350411,
|
| 1715 |
+
"learning_rate": 4.429929986612125e-05,
|
| 1716 |
+
"loss": 0.4047,
|
| 1717 |
+
"step": 244
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 2.6063829787234045,
|
| 1721 |
+
"grad_norm": 0.3200819491909366,
|
| 1722 |
+
"learning_rate": 4.400382766496394e-05,
|
| 1723 |
+
"loss": 0.4051,
|
| 1724 |
+
"step": 245
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 2.617021276595745,
|
| 1728 |
+
"grad_norm": 0.2966057678537604,
|
| 1729 |
+
"learning_rate": 4.3708134616355934e-05,
|
| 1730 |
+
"loss": 0.4106,
|
| 1731 |
+
"step": 246
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 2.627659574468085,
|
| 1735 |
+
"grad_norm": 0.31666189726009564,
|
| 1736 |
+
"learning_rate": 4.341223703045379e-05,
|
| 1737 |
+
"loss": 0.4044,
|
| 1738 |
+
"step": 247
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 2.6382978723404253,
|
| 1742 |
+
"grad_norm": 0.2379399207090575,
|
| 1743 |
+
"learning_rate": 4.311615122869613e-05,
|
| 1744 |
+
"loss": 0.4049,
|
| 1745 |
+
"step": 248
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 2.648936170212766,
|
| 1749 |
+
"grad_norm": 0.33070194627221655,
|
| 1750 |
+
"learning_rate": 4.281989354290341e-05,
|
| 1751 |
+
"loss": 0.4104,
|
| 1752 |
+
"step": 249
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 2.6595744680851063,
|
| 1756 |
+
"grad_norm": 0.29152996582110796,
|
| 1757 |
+
"learning_rate": 4.2523480314376996e-05,
|
| 1758 |
+
"loss": 0.4083,
|
| 1759 |
+
"step": 250
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 2.670212765957447,
|
| 1763 |
+
"grad_norm": 0.2768236406010446,
|
| 1764 |
+
"learning_rate": 4.222692789299794e-05,
|
| 1765 |
+
"loss": 0.4005,
|
| 1766 |
+
"step": 251
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 2.6808510638297873,
|
| 1770 |
+
"grad_norm": 0.32822073352952436,
|
| 1771 |
+
"learning_rate": 4.193025263632495e-05,
|
| 1772 |
+
"loss": 0.4098,
|
| 1773 |
+
"step": 252
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 2.6914893617021276,
|
| 1777 |
+
"grad_norm": 0.20759537070759246,
|
| 1778 |
+
"learning_rate": 4.163347090869227e-05,
|
| 1779 |
+
"loss": 0.4015,
|
| 1780 |
+
"step": 253
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 2.702127659574468,
|
| 1784 |
+
"grad_norm": 0.2450986247008019,
|
| 1785 |
+
"learning_rate": 4.133659908030699e-05,
|
| 1786 |
+
"loss": 0.4011,
|
| 1787 |
+
"step": 254
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 2.7127659574468086,
|
| 1791 |
+
"grad_norm": 0.3094748307247204,
|
| 1792 |
+
"learning_rate": 4.103965352634604e-05,
|
| 1793 |
+
"loss": 0.4128,
|
| 1794 |
+
"step": 255
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 2.723404255319149,
|
| 1798 |
+
"grad_norm": 0.2811649837424367,
|
| 1799 |
+
"learning_rate": 4.0742650626053004e-05,
|
| 1800 |
+
"loss": 0.409,
|
| 1801 |
+
"step": 256
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 2.7340425531914896,
|
| 1805 |
+
"grad_norm": 0.2666126452683894,
|
| 1806 |
+
"learning_rate": 4.044560676183462e-05,
|
| 1807 |
+
"loss": 0.4052,
|
| 1808 |
+
"step": 257
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 2.74468085106383,
|
| 1812 |
+
"grad_norm": 0.3043401200996446,
|
| 1813 |
+
"learning_rate": 4.014853831835721e-05,
|
| 1814 |
+
"loss": 0.4075,
|
| 1815 |
+
"step": 258
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 2.75531914893617,
|
| 1819 |
+
"grad_norm": 0.26121736766165093,
|
| 1820 |
+
"learning_rate": 3.985146168164281e-05,
|
| 1821 |
+
"loss": 0.4114,
|
| 1822 |
+
"step": 259
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 2.7659574468085104,
|
| 1826 |
+
"grad_norm": 0.25762450388003205,
|
| 1827 |
+
"learning_rate": 3.9554393238165386e-05,
|
| 1828 |
+
"loss": 0.4066,
|
| 1829 |
+
"step": 260
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 2.776595744680851,
|
| 1833 |
+
"grad_norm": 0.2541891436719679,
|
| 1834 |
+
"learning_rate": 3.9257349373947016e-05,
|
| 1835 |
+
"loss": 0.4059,
|
| 1836 |
+
"step": 261
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 2.7872340425531914,
|
| 1840 |
+
"grad_norm": 0.2722839322667676,
|
| 1841 |
+
"learning_rate": 3.896034647365398e-05,
|
| 1842 |
+
"loss": 0.4058,
|
| 1843 |
+
"step": 262
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 2.797872340425532,
|
| 1847 |
+
"grad_norm": 0.28452361022067413,
|
| 1848 |
+
"learning_rate": 3.866340091969303e-05,
|
| 1849 |
+
"loss": 0.4077,
|
| 1850 |
+
"step": 263
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 2.8085106382978724,
|
| 1854 |
+
"grad_norm": 0.2530140333177483,
|
| 1855 |
+
"learning_rate": 3.836652909130774e-05,
|
| 1856 |
+
"loss": 0.4093,
|
| 1857 |
+
"step": 264
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 2.8191489361702127,
|
| 1861 |
+
"grad_norm": 0.45406695750367465,
|
| 1862 |
+
"learning_rate": 3.806974736367507e-05,
|
| 1863 |
+
"loss": 0.4117,
|
| 1864 |
+
"step": 265
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 2.829787234042553,
|
| 1868 |
+
"grad_norm": 0.2988144349921366,
|
| 1869 |
+
"learning_rate": 3.7773072107002084e-05,
|
| 1870 |
+
"loss": 0.4023,
|
| 1871 |
+
"step": 266
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 2.8404255319148937,
|
| 1875 |
+
"grad_norm": 0.22739583662187962,
|
| 1876 |
+
"learning_rate": 3.747651968562302e-05,
|
| 1877 |
+
"loss": 0.4054,
|
| 1878 |
+
"step": 267
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 2.851063829787234,
|
| 1882 |
+
"grad_norm": 0.2636532613327371,
|
| 1883 |
+
"learning_rate": 3.718010645709661e-05,
|
| 1884 |
+
"loss": 0.4106,
|
| 1885 |
+
"step": 268
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 2.8617021276595747,
|
| 1889 |
+
"grad_norm": 0.24726890660848583,
|
| 1890 |
+
"learning_rate": 3.688384877130388e-05,
|
| 1891 |
+
"loss": 0.4059,
|
| 1892 |
+
"step": 269
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 2.872340425531915,
|
| 1896 |
+
"grad_norm": 0.2944014064456227,
|
| 1897 |
+
"learning_rate": 3.658776296954622e-05,
|
| 1898 |
+
"loss": 0.4051,
|
| 1899 |
+
"step": 270
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 2.882978723404255,
|
| 1903 |
+
"grad_norm": 0.22859295611193128,
|
| 1904 |
+
"learning_rate": 3.629186538364408e-05,
|
| 1905 |
+
"loss": 0.4108,
|
| 1906 |
+
"step": 271
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 2.8936170212765955,
|
| 1910 |
+
"grad_norm": 0.2939557054489036,
|
| 1911 |
+
"learning_rate": 3.5996172335036065e-05,
|
| 1912 |
+
"loss": 0.4004,
|
| 1913 |
+
"step": 272
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 2.904255319148936,
|
| 1917 |
+
"grad_norm": 0.26529631331687625,
|
| 1918 |
+
"learning_rate": 3.570070013387876e-05,
|
| 1919 |
+
"loss": 0.4076,
|
| 1920 |
+
"step": 273
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 2.9148936170212765,
|
| 1924 |
+
"grad_norm": 0.25501324633080297,
|
| 1925 |
+
"learning_rate": 3.540546507814692e-05,
|
| 1926 |
+
"loss": 0.3984,
|
| 1927 |
+
"step": 274
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 2.925531914893617,
|
| 1931 |
+
"grad_norm": 0.21459730505126537,
|
| 1932 |
+
"learning_rate": 3.5110483452734633e-05,
|
| 1933 |
+
"loss": 0.4025,
|
| 1934 |
+
"step": 275
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 2.9361702127659575,
|
| 1938 |
+
"grad_norm": 0.2046281102038189,
|
| 1939 |
+
"learning_rate": 3.4815771528556976e-05,
|
| 1940 |
+
"loss": 0.4066,
|
| 1941 |
+
"step": 276
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 2.9468085106382977,
|
| 1945 |
+
"grad_norm": 0.25983079535813713,
|
| 1946 |
+
"learning_rate": 3.452134556165259e-05,
|
| 1947 |
+
"loss": 0.4062,
|
| 1948 |
+
"step": 277
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 2.9574468085106385,
|
| 1952 |
+
"grad_norm": 0.18624346254163252,
|
| 1953 |
+
"learning_rate": 3.4227221792286945e-05,
|
| 1954 |
+
"loss": 0.4101,
|
| 1955 |
+
"step": 278
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 2.9680851063829787,
|
| 1959 |
+
"grad_norm": 0.25858622274413134,
|
| 1960 |
+
"learning_rate": 3.393341644405657e-05,
|
| 1961 |
+
"loss": 0.3966,
|
| 1962 |
+
"step": 279
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 2.978723404255319,
|
| 1966 |
+
"grad_norm": 0.21292012777230374,
|
| 1967 |
+
"learning_rate": 3.363994572299418e-05,
|
| 1968 |
+
"loss": 0.4061,
|
| 1969 |
+
"step": 280
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 2.9893617021276597,
|
| 1973 |
+
"grad_norm": 0.2509387412075419,
|
| 1974 |
+
"learning_rate": 3.3346825816674804e-05,
|
| 1975 |
+
"loss": 0.4108,
|
| 1976 |
+
"step": 281
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 3.0,
|
| 1980 |
+
"grad_norm": 0.23091413343534087,
|
| 1981 |
+
"learning_rate": 3.305407289332279e-05,
|
| 1982 |
+
"loss": 0.3937,
|
| 1983 |
+
"step": 282
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 3.0106382978723403,
|
| 1987 |
+
"grad_norm": 0.20905420882080966,
|
| 1988 |
+
"learning_rate": 3.276170310092008e-05,
|
| 1989 |
+
"loss": 0.3778,
|
| 1990 |
+
"step": 283
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 3.021276595744681,
|
| 1994 |
+
"grad_norm": 0.23991076346040097,
|
| 1995 |
+
"learning_rate": 3.246973256631546e-05,
|
| 1996 |
+
"loss": 0.3798,
|
| 1997 |
+
"step": 284
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 3.0319148936170213,
|
| 2001 |
+
"grad_norm": 0.24449335128970107,
|
| 2002 |
+
"learning_rate": 3.217817739433502e-05,
|
| 2003 |
+
"loss": 0.3802,
|
| 2004 |
+
"step": 285
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 3.0425531914893615,
|
| 2008 |
+
"grad_norm": 0.245315113180501,
|
| 2009 |
+
"learning_rate": 3.1887053666893834e-05,
|
| 2010 |
+
"loss": 0.3875,
|
| 2011 |
+
"step": 286
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 3.0531914893617023,
|
| 2015 |
+
"grad_norm": 0.24057913704130401,
|
| 2016 |
+
"learning_rate": 3.159637744210888e-05,
|
| 2017 |
+
"loss": 0.3738,
|
| 2018 |
+
"step": 287
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 3.0638297872340425,
|
| 2022 |
+
"grad_norm": 0.2549241752014073,
|
| 2023 |
+
"learning_rate": 3.130616475341332e-05,
|
| 2024 |
+
"loss": 0.3814,
|
| 2025 |
+
"step": 288
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 3.074468085106383,
|
| 2029 |
+
"grad_norm": 0.21160294231186252,
|
| 2030 |
+
"learning_rate": 3.101643160867208e-05,
|
| 2031 |
+
"loss": 0.3815,
|
| 2032 |
+
"step": 289
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 3.0851063829787235,
|
| 2036 |
+
"grad_norm": 0.23805410062655133,
|
| 2037 |
+
"learning_rate": 3.072719398929887e-05,
|
| 2038 |
+
"loss": 0.3801,
|
| 2039 |
+
"step": 290
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 3.095744680851064,
|
| 2043 |
+
"grad_norm": 0.20195712835328383,
|
| 2044 |
+
"learning_rate": 3.0438467849374702e-05,
|
| 2045 |
+
"loss": 0.3797,
|
| 2046 |
+
"step": 291
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 3.106382978723404,
|
| 2050 |
+
"grad_norm": 0.23312007432367898,
|
| 2051 |
+
"learning_rate": 3.0150269114767862e-05,
|
| 2052 |
+
"loss": 0.3764,
|
| 2053 |
+
"step": 292
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 3.117021276595745,
|
| 2057 |
+
"grad_norm": 0.18274301420799274,
|
| 2058 |
+
"learning_rate": 2.9862613682255383e-05,
|
| 2059 |
+
"loss": 0.3754,
|
| 2060 |
+
"step": 293
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 3.127659574468085,
|
| 2064 |
+
"grad_norm": 0.2478292014777071,
|
| 2065 |
+
"learning_rate": 2.957551741864631e-05,
|
| 2066 |
+
"loss": 0.3767,
|
| 2067 |
+
"step": 294
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 3.1382978723404253,
|
| 2071 |
+
"grad_norm": 0.1927262119265681,
|
| 2072 |
+
"learning_rate": 2.928899615990643e-05,
|
| 2073 |
+
"loss": 0.375,
|
| 2074 |
+
"step": 295
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 3.148936170212766,
|
| 2078 |
+
"grad_norm": 0.2868229283783766,
|
| 2079 |
+
"learning_rate": 2.90030657102848e-05,
|
| 2080 |
+
"loss": 0.3808,
|
| 2081 |
+
"step": 296
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 3.1595744680851063,
|
| 2085 |
+
"grad_norm": 0.18547034376158922,
|
| 2086 |
+
"learning_rate": 2.8717741841441964e-05,
|
| 2087 |
+
"loss": 0.3767,
|
| 2088 |
+
"step": 297
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 3.1702127659574466,
|
| 2092 |
+
"grad_norm": 0.278737988337013,
|
| 2093 |
+
"learning_rate": 2.8433040291580053e-05,
|
| 2094 |
+
"loss": 0.3809,
|
| 2095 |
+
"step": 298
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 3.1808510638297873,
|
| 2099 |
+
"grad_norm": 0.1707915913816605,
|
| 2100 |
+
"learning_rate": 2.8148976764574648e-05,
|
| 2101 |
+
"loss": 0.377,
|
| 2102 |
+
"step": 299
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 3.1914893617021276,
|
| 2106 |
+
"grad_norm": 0.2487011269688432,
|
| 2107 |
+
"learning_rate": 2.7865566929108573e-05,
|
| 2108 |
+
"loss": 0.3838,
|
| 2109 |
+
"step": 300
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 3.202127659574468,
|
| 2113 |
+
"grad_norm": 0.17967682539404906,
|
| 2114 |
+
"learning_rate": 2.758282641780762e-05,
|
| 2115 |
+
"loss": 0.382,
|
| 2116 |
+
"step": 301
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 3.2127659574468086,
|
| 2120 |
+
"grad_norm": 0.21709328897403127,
|
| 2121 |
+
"learning_rate": 2.7300770826378302e-05,
|
| 2122 |
+
"loss": 0.3816,
|
| 2123 |
+
"step": 302
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 3.223404255319149,
|
| 2127 |
+
"grad_norm": 0.178520628301021,
|
| 2128 |
+
"learning_rate": 2.7019415712747558e-05,
|
| 2129 |
+
"loss": 0.3834,
|
| 2130 |
+
"step": 303
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 3.2340425531914896,
|
| 2134 |
+
"grad_norm": 0.21223809263671378,
|
| 2135 |
+
"learning_rate": 2.6738776596204624e-05,
|
| 2136 |
+
"loss": 0.3832,
|
| 2137 |
+
"step": 304
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 3.24468085106383,
|
| 2141 |
+
"grad_norm": 0.15649445787021687,
|
| 2142 |
+
"learning_rate": 2.6458868956544984e-05,
|
| 2143 |
+
"loss": 0.3726,
|
| 2144 |
+
"step": 305
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 3.25531914893617,
|
| 2148 |
+
"grad_norm": 0.1764738407571,
|
| 2149 |
+
"learning_rate": 2.6179708233216557e-05,
|
| 2150 |
+
"loss": 0.3831,
|
| 2151 |
+
"step": 306
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 3.2659574468085104,
|
| 2155 |
+
"grad_norm": 0.15542729747310566,
|
| 2156 |
+
"learning_rate": 2.590130982446802e-05,
|
| 2157 |
+
"loss": 0.3736,
|
| 2158 |
+
"step": 307
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 3.276595744680851,
|
| 2162 |
+
"grad_norm": 0.1747611677030782,
|
| 2163 |
+
"learning_rate": 2.5623689086499496e-05,
|
| 2164 |
+
"loss": 0.3793,
|
| 2165 |
+
"step": 308
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 3.2872340425531914,
|
| 2169 |
+
"grad_norm": 0.13212244563255485,
|
| 2170 |
+
"learning_rate": 2.5346861332615476e-05,
|
| 2171 |
+
"loss": 0.3767,
|
| 2172 |
+
"step": 309
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 3.297872340425532,
|
| 2176 |
+
"grad_norm": 0.17023514875738952,
|
| 2177 |
+
"learning_rate": 2.5070841832380212e-05,
|
| 2178 |
+
"loss": 0.3879,
|
| 2179 |
+
"step": 310
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 3.3085106382978724,
|
| 2183 |
+
"grad_norm": 0.15961770967934297,
|
| 2184 |
+
"learning_rate": 2.4795645810775414e-05,
|
| 2185 |
+
"loss": 0.3769,
|
| 2186 |
+
"step": 311
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 3.3191489361702127,
|
| 2190 |
+
"grad_norm": 0.18863144944962443,
|
| 2191 |
+
"learning_rate": 2.4521288447360457e-05,
|
| 2192 |
+
"loss": 0.3814,
|
| 2193 |
+
"step": 312
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 3.329787234042553,
|
| 2197 |
+
"grad_norm": 0.17296200115217117,
|
| 2198 |
+
"learning_rate": 2.424778487543512e-05,
|
| 2199 |
+
"loss": 0.38,
|
| 2200 |
+
"step": 313
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 3.3404255319148937,
|
| 2204 |
+
"grad_norm": 0.18291540188393604,
|
| 2205 |
+
"learning_rate": 2.3975150181204817e-05,
|
| 2206 |
+
"loss": 0.3736,
|
| 2207 |
+
"step": 314
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 3.351063829787234,
|
| 2211 |
+
"grad_norm": 0.1539347041742154,
|
| 2212 |
+
"learning_rate": 2.370339940294848e-05,
|
| 2213 |
+
"loss": 0.374,
|
| 2214 |
+
"step": 315
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 3.3617021276595747,
|
| 2218 |
+
"grad_norm": 0.17067354232177032,
|
| 2219 |
+
"learning_rate": 2.3432547530189033e-05,
|
| 2220 |
+
"loss": 0.375,
|
| 2221 |
+
"step": 316
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 3.372340425531915,
|
| 2225 |
+
"grad_norm": 0.16125243954444898,
|
| 2226 |
+
"learning_rate": 2.316260950286661e-05,
|
| 2227 |
+
"loss": 0.386,
|
| 2228 |
+
"step": 317
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 3.382978723404255,
|
| 2232 |
+
"grad_norm": 0.17706581816471947,
|
| 2233 |
+
"learning_rate": 2.2893600210514464e-05,
|
| 2234 |
+
"loss": 0.3859,
|
| 2235 |
+
"step": 318
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 3.393617021276596,
|
| 2239 |
+
"grad_norm": 0.19387187739363282,
|
| 2240 |
+
"learning_rate": 2.2625534491437672e-05,
|
| 2241 |
+
"loss": 0.3818,
|
| 2242 |
+
"step": 319
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 3.404255319148936,
|
| 2246 |
+
"grad_norm": 0.16113410459885513,
|
| 2247 |
+
"learning_rate": 2.2358427131894732e-05,
|
| 2248 |
+
"loss": 0.3762,
|
| 2249 |
+
"step": 320
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 3.4148936170212765,
|
| 2253 |
+
"grad_norm": 0.20116908168809555,
|
| 2254 |
+
"learning_rate": 2.2092292865281845e-05,
|
| 2255 |
+
"loss": 0.3803,
|
| 2256 |
+
"step": 321
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 3.425531914893617,
|
| 2260 |
+
"grad_norm": 0.16099148978300246,
|
| 2261 |
+
"learning_rate": 2.1827146371320334e-05,
|
| 2262 |
+
"loss": 0.3822,
|
| 2263 |
+
"step": 322
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 3.4361702127659575,
|
| 2267 |
+
"grad_norm": 0.18990059248934593,
|
| 2268 |
+
"learning_rate": 2.156300227524688e-05,
|
| 2269 |
+
"loss": 0.3825,
|
| 2270 |
+
"step": 323
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 3.4468085106382977,
|
| 2274 |
+
"grad_norm": 0.16540658424680887,
|
| 2275 |
+
"learning_rate": 2.1299875147006838e-05,
|
| 2276 |
+
"loss": 0.3781,
|
| 2277 |
+
"step": 324
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 3.4574468085106385,
|
| 2281 |
+
"grad_norm": 0.16752244552028583,
|
| 2282 |
+
"learning_rate": 2.10377795004505e-05,
|
| 2283 |
+
"loss": 0.3809,
|
| 2284 |
+
"step": 325
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 3.4680851063829787,
|
| 2288 |
+
"grad_norm": 0.1810597095974272,
|
| 2289 |
+
"learning_rate": 2.0776729792532652e-05,
|
| 2290 |
+
"loss": 0.3836,
|
| 2291 |
+
"step": 326
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 3.478723404255319,
|
| 2295 |
+
"grad_norm": 0.1417871274980451,
|
| 2296 |
+
"learning_rate": 2.0516740422515022e-05,
|
| 2297 |
+
"loss": 0.3878,
|
| 2298 |
+
"step": 327
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 3.4893617021276597,
|
| 2302 |
+
"grad_norm": 0.15754394939129449,
|
| 2303 |
+
"learning_rate": 2.0257825731172077e-05,
|
| 2304 |
+
"loss": 0.3833,
|
| 2305 |
+
"step": 328
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 3.5,
|
| 2309 |
+
"grad_norm": 0.12745856435079908,
|
| 2310 |
+
"learning_rate": 2.0000000000000012e-05,
|
| 2311 |
+
"loss": 0.3736,
|
| 2312 |
+
"step": 329
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 3.5106382978723403,
|
| 2316 |
+
"grad_norm": 0.13505830495843163,
|
| 2317 |
+
"learning_rate": 1.9743277450428962e-05,
|
| 2318 |
+
"loss": 0.3807,
|
| 2319 |
+
"step": 330
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 3.521276595744681,
|
| 2323 |
+
"grad_norm": 0.12857257976193204,
|
| 2324 |
+
"learning_rate": 1.9487672243038594e-05,
|
| 2325 |
+
"loss": 0.381,
|
| 2326 |
+
"step": 331
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 3.5319148936170213,
|
| 2330 |
+
"grad_norm": 0.13541243433649724,
|
| 2331 |
+
"learning_rate": 1.9233198476777003e-05,
|
| 2332 |
+
"loss": 0.385,
|
| 2333 |
+
"step": 332
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 3.5425531914893615,
|
| 2337 |
+
"grad_norm": 0.13836684138720814,
|
| 2338 |
+
"learning_rate": 1.897987018818302e-05,
|
| 2339 |
+
"loss": 0.3808,
|
| 2340 |
+
"step": 333
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 3.5531914893617023,
|
| 2344 |
+
"grad_norm": 0.12028403797052283,
|
| 2345 |
+
"learning_rate": 1.8727701350612026e-05,
|
| 2346 |
+
"loss": 0.3787,
|
| 2347 |
+
"step": 334
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 3.5638297872340425,
|
| 2351 |
+
"grad_norm": 0.13853549831902906,
|
| 2352 |
+
"learning_rate": 1.8476705873465097e-05,
|
| 2353 |
+
"loss": 0.3845,
|
| 2354 |
+
"step": 335
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 3.574468085106383,
|
| 2358 |
+
"grad_norm": 0.12529122214780986,
|
| 2359 |
+
"learning_rate": 1.8226897601421858e-05,
|
| 2360 |
+
"loss": 0.3773,
|
| 2361 |
+
"step": 336
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 3.5851063829787235,
|
| 2365 |
+
"grad_norm": 0.13718627686941767,
|
| 2366 |
+
"learning_rate": 1.7978290313676774e-05,
|
| 2367 |
+
"loss": 0.3771,
|
| 2368 |
+
"step": 337
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 3.595744680851064,
|
| 2372 |
+
"grad_norm": 0.13452451362504975,
|
| 2373 |
+
"learning_rate": 1.7730897723179144e-05,
|
| 2374 |
+
"loss": 0.3799,
|
| 2375 |
+
"step": 338
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 3.6063829787234045,
|
| 2379 |
+
"grad_norm": 0.13390152462264926,
|
| 2380 |
+
"learning_rate": 1.748473347587666e-05,
|
| 2381 |
+
"loss": 0.3784,
|
| 2382 |
+
"step": 339
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 3.617021276595745,
|
| 2386 |
+
"grad_norm": 0.1381149397634833,
|
| 2387 |
+
"learning_rate": 1.7239811149962756e-05,
|
| 2388 |
+
"loss": 0.3759,
|
| 2389 |
+
"step": 340
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 3.627659574468085,
|
| 2393 |
+
"grad_norm": 0.12941518047233325,
|
| 2394 |
+
"learning_rate": 1.6996144255127586e-05,
|
| 2395 |
+
"loss": 0.377,
|
| 2396 |
+
"step": 341
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 3.6382978723404253,
|
| 2400 |
+
"grad_norm": 0.14340217919626994,
|
| 2401 |
+
"learning_rate": 1.675374623181294e-05,
|
| 2402 |
+
"loss": 0.3799,
|
| 2403 |
+
"step": 342
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 3.648936170212766,
|
| 2407 |
+
"grad_norm": 0.11836717659820724,
|
| 2408 |
+
"learning_rate": 1.6512630450470784e-05,
|
| 2409 |
+
"loss": 0.3768,
|
| 2410 |
+
"step": 343
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 3.6595744680851063,
|
| 2414 |
+
"grad_norm": 0.15778520770347512,
|
| 2415 |
+
"learning_rate": 1.6272810210825794e-05,
|
| 2416 |
+
"loss": 0.3832,
|
| 2417 |
+
"step": 344
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 3.670212765957447,
|
| 2421 |
+
"grad_norm": 0.1204529845934934,
|
| 2422 |
+
"learning_rate": 1.6034298741141768e-05,
|
| 2423 |
+
"loss": 0.382,
|
| 2424 |
+
"step": 345
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 3.6808510638297873,
|
| 2428 |
+
"grad_norm": 0.15750579133093734,
|
| 2429 |
+
"learning_rate": 1.579710919749196e-05,
|
| 2430 |
+
"loss": 0.3798,
|
| 2431 |
+
"step": 346
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 3.6914893617021276,
|
| 2435 |
+
"grad_norm": 0.1386144145229538,
|
| 2436 |
+
"learning_rate": 1.5561254663033393e-05,
|
| 2437 |
+
"loss": 0.3764,
|
| 2438 |
+
"step": 347
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 3.702127659574468,
|
| 2442 |
+
"grad_norm": 0.1239231423498722,
|
| 2443 |
+
"learning_rate": 1.53267481472852e-05,
|
| 2444 |
+
"loss": 0.3827,
|
| 2445 |
+
"step": 348
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 3.7127659574468086,
|
| 2449 |
+
"grad_norm": 0.14157764395949052,
|
| 2450 |
+
"learning_rate": 1.5093602585411078e-05,
|
| 2451 |
+
"loss": 0.3773,
|
| 2452 |
+
"step": 349
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 3.723404255319149,
|
| 2456 |
+
"grad_norm": 0.140686345314232,
|
| 2457 |
+
"learning_rate": 1.4861830837505733e-05,
|
| 2458 |
+
"loss": 0.3784,
|
| 2459 |
+
"step": 350
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 3.7340425531914896,
|
| 2463 |
+
"grad_norm": 0.12320059023506465,
|
| 2464 |
+
"learning_rate": 1.4631445687885553e-05,
|
| 2465 |
+
"loss": 0.3844,
|
| 2466 |
+
"step": 351
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 3.74468085106383,
|
| 2470 |
+
"grad_norm": 0.14431820751164123,
|
| 2471 |
+
"learning_rate": 1.4402459844383451e-05,
|
| 2472 |
+
"loss": 0.3803,
|
| 2473 |
+
"step": 352
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 3.75531914893617,
|
| 2477 |
+
"grad_norm": 0.12143326106626687,
|
| 2478 |
+
"learning_rate": 1.4174885937647905e-05,
|
| 2479 |
+
"loss": 0.383,
|
| 2480 |
+
"step": 353
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 3.7659574468085104,
|
| 2484 |
+
"grad_norm": 0.11735894401221722,
|
| 2485 |
+
"learning_rate": 1.3948736520446246e-05,
|
| 2486 |
+
"loss": 0.3751,
|
| 2487 |
+
"step": 354
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 3.776595744680851,
|
| 2491 |
+
"grad_norm": 0.13185033729281154,
|
| 2492 |
+
"learning_rate": 1.372402406697229e-05,
|
| 2493 |
+
"loss": 0.3737,
|
| 2494 |
+
"step": 355
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 3.7872340425531914,
|
| 2498 |
+
"grad_norm": 0.1154428148825697,
|
| 2499 |
+
"learning_rate": 1.3500760972158223e-05,
|
| 2500 |
+
"loss": 0.378,
|
| 2501 |
+
"step": 356
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 3.797872340425532,
|
| 2505 |
+
"grad_norm": 0.12842073282310573,
|
| 2506 |
+
"learning_rate": 1.3278959550991011e-05,
|
| 2507 |
+
"loss": 0.3775,
|
| 2508 |
+
"step": 357
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 3.8085106382978724,
|
| 2512 |
+
"grad_norm": 0.12670592325854088,
|
| 2513 |
+
"learning_rate": 1.3058632037832957e-05,
|
| 2514 |
+
"loss": 0.3828,
|
| 2515 |
+
"step": 358
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 3.8191489361702127,
|
| 2519 |
+
"grad_norm": 0.11334725671446295,
|
| 2520 |
+
"learning_rate": 1.2839790585747008e-05,
|
| 2521 |
+
"loss": 0.3784,
|
| 2522 |
+
"step": 359
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 3.829787234042553,
|
| 2526 |
+
"grad_norm": 0.11759887567137367,
|
| 2527 |
+
"learning_rate": 1.2622447265826345e-05,
|
| 2528 |
+
"loss": 0.3823,
|
| 2529 |
+
"step": 360
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 3.8404255319148937,
|
| 2533 |
+
"grad_norm": 0.11200560690429617,
|
| 2534 |
+
"learning_rate": 1.2406614066528543e-05,
|
| 2535 |
+
"loss": 0.3839,
|
| 2536 |
+
"step": 361
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 3.851063829787234,
|
| 2540 |
+
"grad_norm": 0.10929402999457798,
|
| 2541 |
+
"learning_rate": 1.219230289301431e-05,
|
| 2542 |
+
"loss": 0.38,
|
| 2543 |
+
"step": 362
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 3.8617021276595747,
|
| 2547 |
+
"grad_norm": 0.12132649925473797,
|
| 2548 |
+
"learning_rate": 1.1979525566490845e-05,
|
| 2549 |
+
"loss": 0.3752,
|
| 2550 |
+
"step": 363
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 3.872340425531915,
|
| 2554 |
+
"grad_norm": 0.11204838908548817,
|
| 2555 |
+
"learning_rate": 1.176829382355973e-05,
|
| 2556 |
+
"loss": 0.3752,
|
| 2557 |
+
"step": 364
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 3.882978723404255,
|
| 2561 |
+
"grad_norm": 0.11317597743783404,
|
| 2562 |
+
"learning_rate": 1.1558619315569572e-05,
|
| 2563 |
+
"loss": 0.3815,
|
| 2564 |
+
"step": 365
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 3.8936170212765955,
|
| 2568 |
+
"grad_norm": 0.13136090336497436,
|
| 2569 |
+
"learning_rate": 1.1350513607973351e-05,
|
| 2570 |
+
"loss": 0.3839,
|
| 2571 |
+
"step": 366
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 3.904255319148936,
|
| 2575 |
+
"grad_norm": 0.11332184292598117,
|
| 2576 |
+
"learning_rate": 1.1143988179690441e-05,
|
| 2577 |
+
"loss": 0.3784,
|
| 2578 |
+
"step": 367
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 3.9148936170212765,
|
| 2582 |
+
"grad_norm": 0.12049799122378656,
|
| 2583 |
+
"learning_rate": 1.093905442247348e-05,
|
| 2584 |
+
"loss": 0.3737,
|
| 2585 |
+
"step": 368
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 3.925531914893617,
|
| 2589 |
+
"grad_norm": 0.12112505146396743,
|
| 2590 |
+
"learning_rate": 1.073572364027999e-05,
|
| 2591 |
+
"loss": 0.3784,
|
| 2592 |
+
"step": 369
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 3.9361702127659575,
|
| 2596 |
+
"grad_norm": 0.11152808514251403,
|
| 2597 |
+
"learning_rate": 1.0534007048648846e-05,
|
| 2598 |
+
"loss": 0.3836,
|
| 2599 |
+
"step": 370
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 3.9468085106382977,
|
| 2603 |
+
"grad_norm": 0.11026956462916394,
|
| 2604 |
+
"learning_rate": 1.0333915774081698e-05,
|
| 2605 |
+
"loss": 0.382,
|
| 2606 |
+
"step": 371
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 3.9574468085106385,
|
| 2610 |
+
"grad_norm": 0.10840535830420614,
|
| 2611 |
+
"learning_rate": 1.0135460853429166e-05,
|
| 2612 |
+
"loss": 0.3734,
|
| 2613 |
+
"step": 372
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 3.9680851063829787,
|
| 2617 |
+
"grad_norm": 0.1236335149703349,
|
| 2618 |
+
"learning_rate": 9.938653233282105e-06,
|
| 2619 |
+
"loss": 0.3763,
|
| 2620 |
+
"step": 373
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 3.978723404255319,
|
| 2624 |
+
"grad_norm": 0.12252596041380151,
|
| 2625 |
+
"learning_rate": 9.74350376936779e-06,
|
| 2626 |
+
"loss": 0.3716,
|
| 2627 |
+
"step": 374
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 3.9893617021276597,
|
| 2631 |
+
"grad_norm": 0.10540461633402395,
|
| 2632 |
+
"learning_rate": 9.550023225951124e-06,
|
| 2633 |
+
"loss": 0.3725,
|
| 2634 |
+
"step": 375
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 4.0,
|
| 2638 |
+
"grad_norm": 0.12520513195330457,
|
| 2639 |
+
"learning_rate": 9.358222275240884e-06,
|
| 2640 |
+
"loss": 0.3755,
|
| 2641 |
+
"step": 376
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 4.01063829787234,
|
| 2645 |
+
"grad_norm": 0.14847057489702098,
|
| 2646 |
+
"learning_rate": 9.168111496801071e-06,
|
| 2647 |
+
"loss": 0.3652,
|
| 2648 |
+
"step": 377
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 4.0212765957446805,
|
| 2652 |
+
"grad_norm": 0.11332943867336272,
|
| 2653 |
+
"learning_rate": 8.979701376967313e-06,
|
| 2654 |
+
"loss": 0.364,
|
| 2655 |
+
"step": 378
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 4.031914893617022,
|
| 2659 |
+
"grad_norm": 0.1232583568124227,
|
| 2660 |
+
"learning_rate": 8.793002308268521e-06,
|
| 2661 |
+
"loss": 0.3577,
|
| 2662 |
+
"step": 379
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 4.042553191489362,
|
| 2666 |
+
"grad_norm": 0.12780375441362188,
|
| 2667 |
+
"learning_rate": 8.608024588853561e-06,
|
| 2668 |
+
"loss": 0.3635,
|
| 2669 |
+
"step": 380
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 4.053191489361702,
|
| 2673 |
+
"grad_norm": 0.12259673347799972,
|
| 2674 |
+
"learning_rate": 8.424778421923258e-06,
|
| 2675 |
+
"loss": 0.3622,
|
| 2676 |
+
"step": 381
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 4.0638297872340425,
|
| 2680 |
+
"grad_norm": 0.11562199484437485,
|
| 2681 |
+
"learning_rate": 8.24327391516761e-06,
|
| 2682 |
+
"loss": 0.3619,
|
| 2683 |
+
"step": 382
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 4.074468085106383,
|
| 2687 |
+
"grad_norm": 0.11649371701938849,
|
| 2688 |
+
"learning_rate": 8.06352108020823e-06,
|
| 2689 |
+
"loss": 0.367,
|
| 2690 |
+
"step": 383
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 4.085106382978723,
|
| 2694 |
+
"grad_norm": 0.10570361375232744,
|
| 2695 |
+
"learning_rate": 7.885529832046134e-06,
|
| 2696 |
+
"loss": 0.3582,
|
| 2697 |
+
"step": 384
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 4.095744680851064,
|
| 2701 |
+
"grad_norm": 0.10461915141244181,
|
| 2702 |
+
"learning_rate": 7.709309988514824e-06,
|
| 2703 |
+
"loss": 0.3608,
|
| 2704 |
+
"step": 385
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 4.1063829787234045,
|
| 2708 |
+
"grad_norm": 0.11511232586029703,
|
| 2709 |
+
"learning_rate": 7.534871269738753e-06,
|
| 2710 |
+
"loss": 0.3647,
|
| 2711 |
+
"step": 386
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 4.117021276595745,
|
| 2715 |
+
"grad_norm": 0.10867329200636058,
|
| 2716 |
+
"learning_rate": 7.362223297597184e-06,
|
| 2717 |
+
"loss": 0.3629,
|
| 2718 |
+
"step": 387
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 4.127659574468085,
|
| 2722 |
+
"grad_norm": 0.10982240516813774,
|
| 2723 |
+
"learning_rate": 7.191375595193433e-06,
|
| 2724 |
+
"loss": 0.3691,
|
| 2725 |
+
"step": 388
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 4.138297872340425,
|
| 2729 |
+
"grad_norm": 0.11776263009120118,
|
| 2730 |
+
"learning_rate": 7.022337586329597e-06,
|
| 2731 |
+
"loss": 0.3617,
|
| 2732 |
+
"step": 389
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 4.148936170212766,
|
| 2736 |
+
"grad_norm": 0.10569398652654016,
|
| 2737 |
+
"learning_rate": 6.855118594986718e-06,
|
| 2738 |
+
"loss": 0.3567,
|
| 2739 |
+
"step": 390
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 4.159574468085107,
|
| 2743 |
+
"grad_norm": 0.10505309486877212,
|
| 2744 |
+
"learning_rate": 6.6897278448105405e-06,
|
| 2745 |
+
"loss": 0.3639,
|
| 2746 |
+
"step": 391
|
| 2747 |
+
},
|
| 2748 |
+
{
|
| 2749 |
+
"epoch": 4.170212765957447,
|
| 2750 |
+
"grad_norm": 0.1079113458903354,
|
| 2751 |
+
"learning_rate": 6.526174458602681e-06,
|
| 2752 |
+
"loss": 0.368,
|
| 2753 |
+
"step": 392
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 4.180851063829787,
|
| 2757 |
+
"grad_norm": 0.11197666687109296,
|
| 2758 |
+
"learning_rate": 6.364467457817482e-06,
|
| 2759 |
+
"loss": 0.3587,
|
| 2760 |
+
"step": 393
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 4.191489361702128,
|
| 2764 |
+
"grad_norm": 0.11278688179148387,
|
| 2765 |
+
"learning_rate": 6.20461576206433e-06,
|
| 2766 |
+
"loss": 0.3611,
|
| 2767 |
+
"step": 394
|
| 2768 |
+
},
|
| 2769 |
+
{
|
| 2770 |
+
"epoch": 4.202127659574468,
|
| 2771 |
+
"grad_norm": 0.11202771282958791,
|
| 2772 |
+
"learning_rate": 6.046628188615718e-06,
|
| 2773 |
+
"loss": 0.3632,
|
| 2774 |
+
"step": 395
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"epoch": 4.212765957446808,
|
| 2778 |
+
"grad_norm": 0.098500856006095,
|
| 2779 |
+
"learning_rate": 5.890513451920843e-06,
|
| 2780 |
+
"loss": 0.3538,
|
| 2781 |
+
"step": 396
|
| 2782 |
+
},
|
| 2783 |
+
{
|
| 2784 |
+
"epoch": 4.223404255319149,
|
| 2785 |
+
"grad_norm": 0.09784156477819876,
|
| 2786 |
+
"learning_rate": 5.736280163124974e-06,
|
| 2787 |
+
"loss": 0.365,
|
| 2788 |
+
"step": 397
|
| 2789 |
+
},
|
| 2790 |
+
{
|
| 2791 |
+
"epoch": 4.23404255319149,
|
| 2792 |
+
"grad_norm": 0.10953538933768776,
|
| 2793 |
+
"learning_rate": 5.583936829594434e-06,
|
| 2794 |
+
"loss": 0.3603,
|
| 2795 |
+
"step": 398
|
| 2796 |
+
},
|
| 2797 |
+
{
|
| 2798 |
+
"epoch": 4.24468085106383,
|
| 2799 |
+
"grad_norm": 0.10639360562326262,
|
| 2800 |
+
"learning_rate": 5.4334918544473436e-06,
|
| 2801 |
+
"loss": 0.3659,
|
| 2802 |
+
"step": 399
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 4.25531914893617,
|
| 2806 |
+
"grad_norm": 0.10211474883073758,
|
| 2807 |
+
"learning_rate": 5.284953536090131e-06,
|
| 2808 |
+
"loss": 0.3641,
|
| 2809 |
+
"step": 400
|
| 2810 |
+
},
|
| 2811 |
+
{
|
| 2812 |
+
"epoch": 4.26595744680851,
|
| 2813 |
+
"grad_norm": 0.10574255358672346,
|
| 2814 |
+
"learning_rate": 5.1383300677598024e-06,
|
| 2815 |
+
"loss": 0.3643,
|
| 2816 |
+
"step": 401
|
| 2817 |
+
},
|
| 2818 |
+
{
|
| 2819 |
+
"epoch": 4.276595744680851,
|
| 2820 |
+
"grad_norm": 0.11115010018265371,
|
| 2821 |
+
"learning_rate": 4.993629537071978e-06,
|
| 2822 |
+
"loss": 0.3701,
|
| 2823 |
+
"step": 402
|
| 2824 |
+
},
|
| 2825 |
+
{
|
| 2826 |
+
"epoch": 4.287234042553192,
|
| 2827 |
+
"grad_norm": 0.09127212979081487,
|
| 2828 |
+
"learning_rate": 4.850859925574809e-06,
|
| 2829 |
+
"loss": 0.3662,
|
| 2830 |
+
"step": 403
|
| 2831 |
+
},
|
| 2832 |
+
{
|
| 2833 |
+
"epoch": 4.297872340425532,
|
| 2834 |
+
"grad_norm": 0.08625180913725822,
|
| 2835 |
+
"learning_rate": 4.710029108308733e-06,
|
| 2836 |
+
"loss": 0.3548,
|
| 2837 |
+
"step": 404
|
| 2838 |
+
},
|
| 2839 |
+
{
|
| 2840 |
+
"epoch": 4.308510638297872,
|
| 2841 |
+
"grad_norm": 0.10510607272882863,
|
| 2842 |
+
"learning_rate": 4.571144853372063e-06,
|
| 2843 |
+
"loss": 0.363,
|
| 2844 |
+
"step": 405
|
| 2845 |
+
},
|
| 2846 |
+
{
|
| 2847 |
+
"epoch": 4.319148936170213,
|
| 2848 |
+
"grad_norm": 0.09294640053501395,
|
| 2849 |
+
"learning_rate": 4.434214821492542e-06,
|
| 2850 |
+
"loss": 0.3659,
|
| 2851 |
+
"step": 406
|
| 2852 |
+
},
|
| 2853 |
+
{
|
| 2854 |
+
"epoch": 4.329787234042553,
|
| 2855 |
+
"grad_norm": 0.10322126590894673,
|
| 2856 |
+
"learning_rate": 4.299246565604756e-06,
|
| 2857 |
+
"loss": 0.3636,
|
| 2858 |
+
"step": 407
|
| 2859 |
+
},
|
| 2860 |
+
{
|
| 2861 |
+
"epoch": 4.340425531914893,
|
| 2862 |
+
"grad_norm": 0.08513844389462451,
|
| 2863 |
+
"learning_rate": 4.166247530433531e-06,
|
| 2864 |
+
"loss": 0.361,
|
| 2865 |
+
"step": 408
|
| 2866 |
+
},
|
| 2867 |
+
{
|
| 2868 |
+
"epoch": 4.351063829787234,
|
| 2869 |
+
"grad_norm": 0.09456400637267966,
|
| 2870 |
+
"learning_rate": 4.035225052083309e-06,
|
| 2871 |
+
"loss": 0.3631,
|
| 2872 |
+
"step": 409
|
| 2873 |
+
},
|
| 2874 |
+
{
|
| 2875 |
+
"epoch": 4.361702127659575,
|
| 2876 |
+
"grad_norm": 0.09912377148014549,
|
| 2877 |
+
"learning_rate": 3.906186357633455e-06,
|
| 2878 |
+
"loss": 0.3627,
|
| 2879 |
+
"step": 410
|
| 2880 |
+
},
|
| 2881 |
+
{
|
| 2882 |
+
"epoch": 4.372340425531915,
|
| 2883 |
+
"grad_norm": 0.1007261033413922,
|
| 2884 |
+
"learning_rate": 3.779138564739646e-06,
|
| 2885 |
+
"loss": 0.3625,
|
| 2886 |
+
"step": 411
|
| 2887 |
+
},
|
| 2888 |
+
{
|
| 2889 |
+
"epoch": 4.382978723404255,
|
| 2890 |
+
"grad_norm": 0.08934724234816895,
|
| 2891 |
+
"learning_rate": 3.6540886812412547e-06,
|
| 2892 |
+
"loss": 0.3648,
|
| 2893 |
+
"step": 412
|
| 2894 |
+
},
|
| 2895 |
+
{
|
| 2896 |
+
"epoch": 4.3936170212765955,
|
| 2897 |
+
"grad_norm": 0.0925468177451363,
|
| 2898 |
+
"learning_rate": 3.5310436047748263e-06,
|
| 2899 |
+
"loss": 0.3581,
|
| 2900 |
+
"step": 413
|
| 2901 |
+
},
|
| 2902 |
+
{
|
| 2903 |
+
"epoch": 4.404255319148936,
|
| 2904 |
+
"grad_norm": 0.08919561633872664,
|
| 2905 |
+
"learning_rate": 3.4100101223935743e-06,
|
| 2906 |
+
"loss": 0.3571,
|
| 2907 |
+
"step": 414
|
| 2908 |
+
},
|
| 2909 |
+
{
|
| 2910 |
+
"epoch": 4.414893617021277,
|
| 2911 |
+
"grad_norm": 0.09499084946810057,
|
| 2912 |
+
"learning_rate": 3.290994910193037e-06,
|
| 2913 |
+
"loss": 0.3579,
|
| 2914 |
+
"step": 415
|
| 2915 |
+
},
|
| 2916 |
+
{
|
| 2917 |
+
"epoch": 4.425531914893617,
|
| 2918 |
+
"grad_norm": 0.09014647001183278,
|
| 2919 |
+
"learning_rate": 3.174004532942845e-06,
|
| 2920 |
+
"loss": 0.3689,
|
| 2921 |
+
"step": 416
|
| 2922 |
+
},
|
| 2923 |
+
{
|
| 2924 |
+
"epoch": 4.4361702127659575,
|
| 2925 |
+
"grad_norm": 0.08581817647958025,
|
| 2926 |
+
"learning_rate": 3.059045443724582e-06,
|
| 2927 |
+
"loss": 0.3623,
|
| 2928 |
+
"step": 417
|
| 2929 |
+
},
|
| 2930 |
+
{
|
| 2931 |
+
"epoch": 4.446808510638298,
|
| 2932 |
+
"grad_norm": 0.09478322839033952,
|
| 2933 |
+
"learning_rate": 2.946123983575846e-06,
|
| 2934 |
+
"loss": 0.3621,
|
| 2935 |
+
"step": 418
|
| 2936 |
+
},
|
| 2937 |
+
{
|
| 2938 |
+
"epoch": 4.457446808510638,
|
| 2939 |
+
"grad_norm": 0.09127292341411823,
|
| 2940 |
+
"learning_rate": 2.8352463811404952e-06,
|
| 2941 |
+
"loss": 0.352,
|
| 2942 |
+
"step": 419
|
| 2943 |
+
},
|
| 2944 |
+
{
|
| 2945 |
+
"epoch": 4.468085106382979,
|
| 2946 |
+
"grad_norm": 0.09314064563107814,
|
| 2947 |
+
"learning_rate": 2.726418752325084e-06,
|
| 2948 |
+
"loss": 0.3604,
|
| 2949 |
+
"step": 420
|
| 2950 |
+
},
|
| 2951 |
+
{
|
| 2952 |
+
"epoch": 4.4787234042553195,
|
| 2953 |
+
"grad_norm": 0.08258953047242434,
|
| 2954 |
+
"learning_rate": 2.619647099961502e-06,
|
| 2955 |
+
"loss": 0.3613,
|
| 2956 |
+
"step": 421
|
| 2957 |
+
},
|
| 2958 |
+
{
|
| 2959 |
+
"epoch": 4.48936170212766,
|
| 2960 |
+
"grad_norm": 0.08671649480194962,
|
| 2961 |
+
"learning_rate": 2.514937313475865e-06,
|
| 2962 |
+
"loss": 0.3619,
|
| 2963 |
+
"step": 422
|
| 2964 |
+
},
|
| 2965 |
+
{
|
| 2966 |
+
"epoch": 4.5,
|
| 2967 |
+
"grad_norm": 0.08348749730392281,
|
| 2968 |
+
"learning_rate": 2.4122951685636674e-06,
|
| 2969 |
+
"loss": 0.362,
|
| 2970 |
+
"step": 423
|
| 2971 |
+
},
|
| 2972 |
+
{
|
| 2973 |
+
"epoch": 4.51063829787234,
|
| 2974 |
+
"grad_norm": 0.08847421831587975,
|
| 2975 |
+
"learning_rate": 2.3117263268712e-06,
|
| 2976 |
+
"loss": 0.3643,
|
| 2977 |
+
"step": 424
|
| 2978 |
+
},
|
| 2979 |
+
{
|
| 2980 |
+
"epoch": 4.5212765957446805,
|
| 2981 |
+
"grad_norm": 0.08541200353819858,
|
| 2982 |
+
"learning_rate": 2.2132363356832532e-06,
|
| 2983 |
+
"loss": 0.3617,
|
| 2984 |
+
"step": 425
|
| 2985 |
+
},
|
| 2986 |
+
{
|
| 2987 |
+
"epoch": 4.531914893617021,
|
| 2988 |
+
"grad_norm": 0.09072202451972457,
|
| 2989 |
+
"learning_rate": 2.11683062761713e-06,
|
| 2990 |
+
"loss": 0.3637,
|
| 2991 |
+
"step": 426
|
| 2992 |
+
},
|
| 2993 |
+
{
|
| 2994 |
+
"epoch": 4.542553191489362,
|
| 2995 |
+
"grad_norm": 0.08771913713301659,
|
| 2996 |
+
"learning_rate": 2.0225145203230044e-06,
|
| 2997 |
+
"loss": 0.3635,
|
| 2998 |
+
"step": 427
|
| 2999 |
+
},
|
| 3000 |
+
{
|
| 3001 |
+
"epoch": 4.553191489361702,
|
| 3002 |
+
"grad_norm": 0.08227630624973846,
|
| 3003 |
+
"learning_rate": 1.930293216190586e-06,
|
| 3004 |
+
"loss": 0.3677,
|
| 3005 |
+
"step": 428
|
| 3006 |
+
},
|
| 3007 |
+
{
|
| 3008 |
+
"epoch": 4.5638297872340425,
|
| 3009 |
+
"grad_norm": 0.08142816323777968,
|
| 3010 |
+
"learning_rate": 1.8401718020621694e-06,
|
| 3011 |
+
"loss": 0.3653,
|
| 3012 |
+
"step": 429
|
| 3013 |
+
},
|
| 3014 |
+
{
|
| 3015 |
+
"epoch": 4.574468085106383,
|
| 3016 |
+
"grad_norm": 0.08626743211827775,
|
| 3017 |
+
"learning_rate": 1.7521552489520566e-06,
|
| 3018 |
+
"loss": 0.3685,
|
| 3019 |
+
"step": 430
|
| 3020 |
+
},
|
| 3021 |
+
{
|
| 3022 |
+
"epoch": 4.585106382978723,
|
| 3023 |
+
"grad_norm": 0.08468387346146096,
|
| 3024 |
+
"learning_rate": 1.666248411772342e-06,
|
| 3025 |
+
"loss": 0.3651,
|
| 3026 |
+
"step": 431
|
| 3027 |
+
},
|
| 3028 |
+
{
|
| 3029 |
+
"epoch": 4.595744680851064,
|
| 3030 |
+
"grad_norm": 0.08278070326775537,
|
| 3031 |
+
"learning_rate": 1.5824560290651404e-06,
|
| 3032 |
+
"loss": 0.3638,
|
| 3033 |
+
"step": 432
|
| 3034 |
+
},
|
| 3035 |
+
{
|
| 3036 |
+
"epoch": 4.6063829787234045,
|
| 3037 |
+
"grad_norm": 0.08542627322845557,
|
| 3038 |
+
"learning_rate": 1.5007827227411942e-06,
|
| 3039 |
+
"loss": 0.3608,
|
| 3040 |
+
"step": 433
|
| 3041 |
+
},
|
| 3042 |
+
{
|
| 3043 |
+
"epoch": 4.617021276595745,
|
| 3044 |
+
"grad_norm": 0.0850283799369069,
|
| 3045 |
+
"learning_rate": 1.4212329978249417e-06,
|
| 3046 |
+
"loss": 0.3694,
|
| 3047 |
+
"step": 434
|
| 3048 |
+
},
|
| 3049 |
+
{
|
| 3050 |
+
"epoch": 4.627659574468085,
|
| 3051 |
+
"grad_norm": 0.08375767787904646,
|
| 3052 |
+
"learning_rate": 1.3438112422060256e-06,
|
| 3053 |
+
"loss": 0.3572,
|
| 3054 |
+
"step": 435
|
| 3055 |
+
},
|
| 3056 |
+
{
|
| 3057 |
+
"epoch": 4.638297872340425,
|
| 3058 |
+
"grad_norm": 0.08181511297244645,
|
| 3059 |
+
"learning_rate": 1.2685217263972693e-06,
|
| 3060 |
+
"loss": 0.3625,
|
| 3061 |
+
"step": 436
|
| 3062 |
+
},
|
| 3063 |
+
{
|
| 3064 |
+
"epoch": 4.648936170212766,
|
| 3065 |
+
"grad_norm": 0.08280046685526832,
|
| 3066 |
+
"learning_rate": 1.1953686032990964e-06,
|
| 3067 |
+
"loss": 0.3604,
|
| 3068 |
+
"step": 437
|
| 3069 |
+
},
|
| 3070 |
+
{
|
| 3071 |
+
"epoch": 4.659574468085106,
|
| 3072 |
+
"grad_norm": 0.07768111306576052,
|
| 3073 |
+
"learning_rate": 1.124355907970487e-06,
|
| 3074 |
+
"loss": 0.3642,
|
| 3075 |
+
"step": 438
|
| 3076 |
+
},
|
| 3077 |
+
{
|
| 3078 |
+
"epoch": 4.670212765957447,
|
| 3079 |
+
"grad_norm": 0.08142474629210497,
|
| 3080 |
+
"learning_rate": 1.0554875574063784e-06,
|
| 3081 |
+
"loss": 0.3645,
|
| 3082 |
+
"step": 439
|
| 3083 |
+
},
|
| 3084 |
+
{
|
| 3085 |
+
"epoch": 4.680851063829787,
|
| 3086 |
+
"grad_norm": 0.0795668987754362,
|
| 3087 |
+
"learning_rate": 9.887673503216422e-07,
|
| 3088 |
+
"loss": 0.3663,
|
| 3089 |
+
"step": 440
|
| 3090 |
+
},
|
| 3091 |
+
{
|
| 3092 |
+
"epoch": 4.691489361702128,
|
| 3093 |
+
"grad_norm": 0.07508730919255964,
|
| 3094 |
+
"learning_rate": 9.241989669415097e-07,
|
| 3095 |
+
"loss": 0.3661,
|
| 3096 |
+
"step": 441
|
| 3097 |
+
},
|
| 3098 |
+
{
|
| 3099 |
+
"epoch": 4.702127659574468,
|
| 3100 |
+
"grad_norm": 0.08240799525472663,
|
| 3101 |
+
"learning_rate": 8.61785968798623e-07,
|
| 3102 |
+
"loss": 0.3608,
|
| 3103 |
+
"step": 442
|
| 3104 |
+
},
|
| 3105 |
+
{
|
| 3106 |
+
"epoch": 4.712765957446808,
|
| 3107 |
+
"grad_norm": 0.07849402490031887,
|
| 3108 |
+
"learning_rate": 8.015317985365301e-07,
|
| 3109 |
+
"loss": 0.3616,
|
| 3110 |
+
"step": 443
|
| 3111 |
+
},
|
| 3112 |
+
{
|
| 3113 |
+
"epoch": 4.723404255319149,
|
| 3114 |
+
"grad_norm": 0.07995230536661296,
|
| 3115 |
+
"learning_rate": 7.434397797198367e-07,
|
| 3116 |
+
"loss": 0.3599,
|
| 3117 |
+
"step": 444
|
| 3118 |
+
},
|
| 3119 |
+
{
|
| 3120 |
+
"epoch": 4.73404255319149,
|
| 3121 |
+
"grad_norm": 0.08161632016821302,
|
| 3122 |
+
"learning_rate": 6.875131166508553e-07,
|
| 3123 |
+
"loss": 0.36,
|
| 3124 |
+
"step": 445
|
| 3125 |
+
},
|
| 3126 |
+
{
|
| 3127 |
+
"epoch": 4.74468085106383,
|
| 3128 |
+
"grad_norm": 0.07946253233083479,
|
| 3129 |
+
"learning_rate": 6.337548941928839e-07,
|
| 3130 |
+
"loss": 0.3626,
|
| 3131 |
+
"step": 446
|
| 3132 |
+
},
|
| 3133 |
+
{
|
| 3134 |
+
"epoch": 4.75531914893617,
|
| 3135 |
+
"grad_norm": 0.07786315052046529,
|
| 3136 |
+
"learning_rate": 5.821680776000049e-07,
|
| 3137 |
+
"loss": 0.3619,
|
| 3138 |
+
"step": 447
|
| 3139 |
+
},
|
| 3140 |
+
{
|
| 3141 |
+
"epoch": 4.76595744680851,
|
| 3142 |
+
"grad_norm": 0.07728112424595031,
|
| 3143 |
+
"learning_rate": 5.32755512353571e-07,
|
| 3144 |
+
"loss": 0.3607,
|
| 3145 |
+
"step": 448
|
| 3146 |
+
},
|
| 3147 |
+
{
|
| 3148 |
+
"epoch": 4.776595744680851,
|
| 3149 |
+
"grad_norm": 0.07848088350738662,
|
| 3150 |
+
"learning_rate": 4.8551992400522e-07,
|
| 3151 |
+
"loss": 0.361,
|
| 3152 |
+
"step": 449
|
| 3153 |
+
},
|
| 3154 |
+
{
|
| 3155 |
+
"epoch": 4.787234042553192,
|
| 3156 |
+
"grad_norm": 0.07839104640568627,
|
| 3157 |
+
"learning_rate": 4.4046391802655463e-07,
|
| 3158 |
+
"loss": 0.3597,
|
| 3159 |
+
"step": 450
|
| 3160 |
+
},
|
| 3161 |
+
{
|
| 3162 |
+
"epoch": 4.797872340425532,
|
| 3163 |
+
"grad_norm": 0.0757611098723252,
|
| 3164 |
+
"learning_rate": 3.975899796654137e-07,
|
| 3165 |
+
"loss": 0.3671,
|
| 3166 |
+
"step": 451
|
| 3167 |
+
},
|
| 3168 |
+
{
|
| 3169 |
+
"epoch": 4.808510638297872,
|
| 3170 |
+
"grad_norm": 0.07660526674986573,
|
| 3171 |
+
"learning_rate": 3.569004738087989e-07,
|
| 3172 |
+
"loss": 0.3589,
|
| 3173 |
+
"step": 452
|
| 3174 |
+
},
|
| 3175 |
+
{
|
| 3176 |
+
"epoch": 4.819148936170213,
|
| 3177 |
+
"grad_norm": 0.07303253050833616,
|
| 3178 |
+
"learning_rate": 3.183976448524106e-07,
|
| 3179 |
+
"loss": 0.363,
|
| 3180 |
+
"step": 453
|
| 3181 |
+
},
|
| 3182 |
+
{
|
| 3183 |
+
"epoch": 4.829787234042553,
|
| 3184 |
+
"grad_norm": 0.0754318214355068,
|
| 3185 |
+
"learning_rate": 2.8208361657688474e-07,
|
| 3186 |
+
"loss": 0.3643,
|
| 3187 |
+
"step": 454
|
| 3188 |
+
},
|
| 3189 |
+
{
|
| 3190 |
+
"epoch": 4.840425531914894,
|
| 3191 |
+
"grad_norm": 0.07876372271022951,
|
| 3192 |
+
"learning_rate": 2.479603920306106e-07,
|
| 3193 |
+
"loss": 0.3631,
|
| 3194 |
+
"step": 455
|
| 3195 |
+
},
|
| 3196 |
+
{
|
| 3197 |
+
"epoch": 4.851063829787234,
|
| 3198 |
+
"grad_norm": 0.07633688369964613,
|
| 3199 |
+
"learning_rate": 2.1602985341925953e-07,
|
| 3200 |
+
"loss": 0.3667,
|
| 3201 |
+
"step": 456
|
| 3202 |
+
},
|
| 3203 |
+
{
|
| 3204 |
+
"epoch": 4.861702127659575,
|
| 3205 |
+
"grad_norm": 0.07789291049580865,
|
| 3206 |
+
"learning_rate": 1.8629376200197004e-07,
|
| 3207 |
+
"loss": 0.3626,
|
| 3208 |
+
"step": 457
|
| 3209 |
+
},
|
| 3210 |
+
{
|
| 3211 |
+
"epoch": 4.872340425531915,
|
| 3212 |
+
"grad_norm": 0.07875357482220077,
|
| 3213 |
+
"learning_rate": 1.5875375799419e-07,
|
| 3214 |
+
"loss": 0.3634,
|
| 3215 |
+
"step": 458
|
| 3216 |
+
},
|
| 3217 |
+
{
|
| 3218 |
+
"epoch": 4.882978723404255,
|
| 3219 |
+
"grad_norm": 0.0786747469968429,
|
| 3220 |
+
"learning_rate": 1.3341136047719805e-07,
|
| 3221 |
+
"loss": 0.3643,
|
| 3222 |
+
"step": 459
|
| 3223 |
+
},
|
| 3224 |
+
{
|
| 3225 |
+
"epoch": 4.8936170212765955,
|
| 3226 |
+
"grad_norm": 0.07411973494338685,
|
| 3227 |
+
"learning_rate": 1.1026796731433475e-07,
|
| 3228 |
+
"loss": 0.3619,
|
| 3229 |
+
"step": 460
|
| 3230 |
+
},
|
| 3231 |
+
{
|
| 3232 |
+
"epoch": 4.904255319148936,
|
| 3233 |
+
"grad_norm": 0.07325052452508983,
|
| 3234 |
+
"learning_rate": 8.932485507387345e-08,
|
| 3235 |
+
"loss": 0.3538,
|
| 3236 |
+
"step": 461
|
| 3237 |
+
},
|
| 3238 |
+
{
|
| 3239 |
+
"epoch": 4.914893617021277,
|
| 3240 |
+
"grad_norm": 0.07867411963780692,
|
| 3241 |
+
"learning_rate": 7.058317895861866e-08,
|
| 3242 |
+
"loss": 0.3627,
|
| 3243 |
+
"step": 462
|
| 3244 |
+
},
|
| 3245 |
+
{
|
| 3246 |
+
"epoch": 4.925531914893617,
|
| 3247 |
+
"grad_norm": 0.07601171107784903,
|
| 3248 |
+
"learning_rate": 5.40439727421882e-08,
|
| 3249 |
+
"loss": 0.3645,
|
| 3250 |
+
"step": 463
|
| 3251 |
+
},
|
| 3252 |
+
{
|
| 3253 |
+
"epoch": 4.9361702127659575,
|
| 3254 |
+
"grad_norm": 0.07741148103950671,
|
| 3255 |
+
"learning_rate": 3.970814871197437e-08,
|
| 3256 |
+
"loss": 0.3611,
|
| 3257 |
+
"step": 464
|
| 3258 |
+
},
|
| 3259 |
+
{
|
| 3260 |
+
"epoch": 4.946808510638298,
|
| 3261 |
+
"grad_norm": 0.07590962518579049,
|
| 3262 |
+
"learning_rate": 2.75764976188464e-08,
|
| 3263 |
+
"loss": 0.3594,
|
| 3264 |
+
"step": 465
|
| 3265 |
+
},
|
| 3266 |
+
{
|
| 3267 |
+
"epoch": 4.957446808510638,
|
| 3268 |
+
"grad_norm": 0.07705705973848143,
|
| 3269 |
+
"learning_rate": 1.764968863351424e-08,
|
| 3270 |
+
"loss": 0.3619,
|
| 3271 |
+
"step": 466
|
| 3272 |
+
},
|
| 3273 |
+
{
|
| 3274 |
+
"epoch": 4.968085106382979,
|
| 3275 |
+
"grad_norm": 0.0738876803957471,
|
| 3276 |
+
"learning_rate": 9.928269309638083e-09,
|
| 3277 |
+
"loss": 0.3604,
|
| 3278 |
+
"step": 467
|
| 3279 |
+
},
|
| 3280 |
+
{
|
| 3281 |
+
"epoch": 4.9787234042553195,
|
| 3282 |
+
"grad_norm": 0.07315503801724814,
|
| 3283 |
+
"learning_rate": 4.412665553594764e-09,
|
| 3284 |
+
"loss": 0.3638,
|
| 3285 |
+
"step": 468
|
| 3286 |
+
},
|
| 3287 |
+
{
|
| 3288 |
+
"epoch": 4.98936170212766,
|
| 3289 |
+
"grad_norm": 0.07954351258156074,
|
| 3290 |
+
"learning_rate": 1.103181601020964e-09,
|
| 3291 |
+
"loss": 0.3628,
|
| 3292 |
+
"step": 469
|
| 3293 |
+
},
|
| 3294 |
+
{
|
| 3295 |
+
"epoch": 5.0,
|
| 3296 |
+
"grad_norm": 0.07533000239349877,
|
| 3297 |
+
"learning_rate": 0.0,
|
| 3298 |
+
"loss": 0.3558,
|
| 3299 |
+
"step": 470
|
| 3300 |
+
},
|
| 3301 |
+
{
|
| 3302 |
+
"epoch": 5.0,
|
| 3303 |
+
"step": 470,
|
| 3304 |
+
"total_flos": 7884786861342720.0,
|
| 3305 |
+
"train_loss": 0.4284587557011462,
|
| 3306 |
+
"train_runtime": 6272.7597,
|
| 3307 |
+
"train_samples_per_second": 38.218,
|
| 3308 |
+
"train_steps_per_second": 0.075
|
| 3309 |
+
}
|
| 3310 |
+
],
|
| 3311 |
+
"logging_steps": 1,
|
| 3312 |
+
"max_steps": 470,
|
| 3313 |
+
"num_input_tokens_seen": 0,
|
| 3314 |
+
"num_train_epochs": 5,
|
| 3315 |
+
"save_steps": 500,
|
| 3316 |
+
"stateful_callbacks": {
|
| 3317 |
+
"TrainerControl": {
|
| 3318 |
+
"args": {
|
| 3319 |
+
"should_epoch_stop": false,
|
| 3320 |
+
"should_evaluate": false,
|
| 3321 |
+
"should_log": false,
|
| 3322 |
+
"should_save": true,
|
| 3323 |
+
"should_training_stop": true
|
| 3324 |
+
},
|
| 3325 |
+
"attributes": {}
|
| 3326 |
+
}
|
| 3327 |
+
},
|
| 3328 |
+
"total_flos": 7884786861342720.0,
|
| 3329 |
+
"train_batch_size": 1,
|
| 3330 |
+
"trial_name": null,
|
| 3331 |
+
"trial_params": null
|
| 3332 |
+
}
|