Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +3178 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,3178 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.998885172798216,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 4484,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.011148272017837236,
|
| 13 |
+
"grad_norm": 7.207186698913574,
|
| 14 |
+
"learning_rate": 8.88888888888889e-06,
|
| 15 |
+
"loss": 1.1288,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.022296544035674472,
|
| 20 |
+
"grad_norm": 3.3555521965026855,
|
| 21 |
+
"learning_rate": 1.777777777777778e-05,
|
| 22 |
+
"loss": 0.8626,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.033444816053511704,
|
| 27 |
+
"grad_norm": 2.776094436645508,
|
| 28 |
+
"learning_rate": 2.6666666666666667e-05,
|
| 29 |
+
"loss": 0.4164,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.044593088071348944,
|
| 34 |
+
"grad_norm": 2.0430309772491455,
|
| 35 |
+
"learning_rate": 3.555555555555556e-05,
|
| 36 |
+
"loss": 0.2775,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.055741360089186176,
|
| 41 |
+
"grad_norm": 2.5862274169921875,
|
| 42 |
+
"learning_rate": 4.4444444444444447e-05,
|
| 43 |
+
"loss": 0.2283,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.06688963210702341,
|
| 48 |
+
"grad_norm": 2.1346654891967773,
|
| 49 |
+
"learning_rate": 5.333333333333333e-05,
|
| 50 |
+
"loss": 0.2196,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.07803790412486064,
|
| 55 |
+
"grad_norm": 1.602028489112854,
|
| 56 |
+
"learning_rate": 6.222222222222222e-05,
|
| 57 |
+
"loss": 0.162,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.08918617614269789,
|
| 62 |
+
"grad_norm": 1.2011414766311646,
|
| 63 |
+
"learning_rate": 7.111111111111112e-05,
|
| 64 |
+
"loss": 0.13,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.10033444816053512,
|
| 69 |
+
"grad_norm": 1.6477724313735962,
|
| 70 |
+
"learning_rate": 8e-05,
|
| 71 |
+
"loss": 0.1369,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.11148272017837235,
|
| 76 |
+
"grad_norm": 1.8019773960113525,
|
| 77 |
+
"learning_rate": 8.888888888888889e-05,
|
| 78 |
+
"loss": 0.1302,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.12263099219620958,
|
| 83 |
+
"grad_norm": 0.9655002355575562,
|
| 84 |
+
"learning_rate": 9.777777777777778e-05,
|
| 85 |
+
"loss": 0.131,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.13377926421404682,
|
| 90 |
+
"grad_norm": 1.120625376701355,
|
| 91 |
+
"learning_rate": 0.00010666666666666667,
|
| 92 |
+
"loss": 0.1277,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.14492753623188406,
|
| 97 |
+
"grad_norm": 1.1046862602233887,
|
| 98 |
+
"learning_rate": 0.00011555555555555555,
|
| 99 |
+
"loss": 0.1047,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.15607580824972128,
|
| 104 |
+
"grad_norm": 1.1683974266052246,
|
| 105 |
+
"learning_rate": 0.00012444444444444444,
|
| 106 |
+
"loss": 0.107,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.16722408026755853,
|
| 111 |
+
"grad_norm": 1.467659831047058,
|
| 112 |
+
"learning_rate": 0.00013333333333333334,
|
| 113 |
+
"loss": 0.1116,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.17837235228539577,
|
| 118 |
+
"grad_norm": 1.0501031875610352,
|
| 119 |
+
"learning_rate": 0.00014222222222222224,
|
| 120 |
+
"loss": 0.1004,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.189520624303233,
|
| 125 |
+
"grad_norm": 1.0245281457901,
|
| 126 |
+
"learning_rate": 0.0001511111111111111,
|
| 127 |
+
"loss": 0.1006,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.20066889632107024,
|
| 132 |
+
"grad_norm": 0.8765921592712402,
|
| 133 |
+
"learning_rate": 0.00016,
|
| 134 |
+
"loss": 0.0882,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.21181716833890746,
|
| 139 |
+
"grad_norm": 0.7301109433174133,
|
| 140 |
+
"learning_rate": 0.00016888888888888889,
|
| 141 |
+
"loss": 0.0915,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.2229654403567447,
|
| 146 |
+
"grad_norm": 0.9839191436767578,
|
| 147 |
+
"learning_rate": 0.00017777777777777779,
|
| 148 |
+
"loss": 0.1012,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.23411371237458195,
|
| 153 |
+
"grad_norm": 0.8372408151626587,
|
| 154 |
+
"learning_rate": 0.0001866666666666667,
|
| 155 |
+
"loss": 0.0964,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.24526198439241917,
|
| 160 |
+
"grad_norm": 1.0355066061019897,
|
| 161 |
+
"learning_rate": 0.00019555555555555556,
|
| 162 |
+
"loss": 0.1012,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.2564102564102564,
|
| 167 |
+
"grad_norm": 0.7156217098236084,
|
| 168 |
+
"learning_rate": 0.00019999931986678288,
|
| 169 |
+
"loss": 0.0831,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.26755852842809363,
|
| 174 |
+
"grad_norm": 0.7486444115638733,
|
| 175 |
+
"learning_rate": 0.00019999387885655527,
|
| 176 |
+
"loss": 0.0828,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.2787068004459309,
|
| 181 |
+
"grad_norm": 11.638164520263672,
|
| 182 |
+
"learning_rate": 0.00019998299713214806,
|
| 183 |
+
"loss": 0.0835,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.2898550724637681,
|
| 188 |
+
"grad_norm": 5.416531562805176,
|
| 189 |
+
"learning_rate": 0.00019996667528564093,
|
| 190 |
+
"loss": 0.0984,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.3010033444816054,
|
| 195 |
+
"grad_norm": 2.221506357192993,
|
| 196 |
+
"learning_rate": 0.00019994491420511335,
|
| 197 |
+
"loss": 0.1166,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.31215161649944256,
|
| 202 |
+
"grad_norm": 1.9484014511108398,
|
| 203 |
+
"learning_rate": 0.0001999177150745959,
|
| 204 |
+
"loss": 0.1086,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.3232998885172798,
|
| 209 |
+
"grad_norm": 1.6127281188964844,
|
| 210 |
+
"learning_rate": 0.00019988507937400613,
|
| 211 |
+
"loss": 0.1158,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.33444816053511706,
|
| 216 |
+
"grad_norm": 1.1419061422348022,
|
| 217 |
+
"learning_rate": 0.00019984700887906799,
|
| 218 |
+
"loss": 0.1303,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.3455964325529543,
|
| 223 |
+
"grad_norm": 1.017188549041748,
|
| 224 |
+
"learning_rate": 0.00019980350566121505,
|
| 225 |
+
"loss": 0.1168,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.35674470457079155,
|
| 230 |
+
"grad_norm": 1.1969743967056274,
|
| 231 |
+
"learning_rate": 0.0001997545720874779,
|
| 232 |
+
"loss": 0.1169,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.36789297658862874,
|
| 237 |
+
"grad_norm": 1.1392648220062256,
|
| 238 |
+
"learning_rate": 0.00019970021082035553,
|
| 239 |
+
"loss": 0.111,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.379041248606466,
|
| 244 |
+
"grad_norm": 1.3736199140548706,
|
| 245 |
+
"learning_rate": 0.00019964042481767003,
|
| 246 |
+
"loss": 0.1021,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.39018952062430323,
|
| 251 |
+
"grad_norm": 1.2694867849349976,
|
| 252 |
+
"learning_rate": 0.00019957521733240605,
|
| 253 |
+
"loss": 0.1033,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.4013377926421405,
|
| 258 |
+
"grad_norm": 1.0627532005310059,
|
| 259 |
+
"learning_rate": 0.00019950459191253375,
|
| 260 |
+
"loss": 0.1052,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.4124860646599777,
|
| 265 |
+
"grad_norm": 0.8336015939712524,
|
| 266 |
+
"learning_rate": 0.00019942855240081552,
|
| 267 |
+
"loss": 0.081,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.4236343366778149,
|
| 272 |
+
"grad_norm": 0.44888758659362793,
|
| 273 |
+
"learning_rate": 0.00019934710293459715,
|
| 274 |
+
"loss": 0.0902,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.43478260869565216,
|
| 279 |
+
"grad_norm": 0.639631450176239,
|
| 280 |
+
"learning_rate": 0.0001992602479455825,
|
| 281 |
+
"loss": 0.1012,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.4459308807134894,
|
| 286 |
+
"grad_norm": 0.9595453143119812,
|
| 287 |
+
"learning_rate": 0.00019916799215959264,
|
| 288 |
+
"loss": 0.0792,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.45707915273132665,
|
| 293 |
+
"grad_norm": 0.7724646329879761,
|
| 294 |
+
"learning_rate": 0.0001990703405963084,
|
| 295 |
+
"loss": 0.075,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.4682274247491639,
|
| 300 |
+
"grad_norm": 0.7783587574958801,
|
| 301 |
+
"learning_rate": 0.00019896729856899747,
|
| 302 |
+
"loss": 0.0905,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.4793756967670011,
|
| 307 |
+
"grad_norm": 0.954002320766449,
|
| 308 |
+
"learning_rate": 0.00019885887168422526,
|
| 309 |
+
"loss": 0.0914,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.49052396878483834,
|
| 314 |
+
"grad_norm": 0.509386420249939,
|
| 315 |
+
"learning_rate": 0.00019874506584154972,
|
| 316 |
+
"loss": 0.0871,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.5016722408026756,
|
| 321 |
+
"grad_norm": 1.112070083618164,
|
| 322 |
+
"learning_rate": 0.00019862588723320058,
|
| 323 |
+
"loss": 0.0927,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.5128205128205128,
|
| 328 |
+
"grad_norm": 1.083264946937561,
|
| 329 |
+
"learning_rate": 0.00019850134234374216,
|
| 330 |
+
"loss": 0.0777,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.5239687848383501,
|
| 335 |
+
"grad_norm": 0.6582397222518921,
|
| 336 |
+
"learning_rate": 0.00019837143794972076,
|
| 337 |
+
"loss": 0.0954,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.5351170568561873,
|
| 342 |
+
"grad_norm": 1.1856610774993896,
|
| 343 |
+
"learning_rate": 0.00019823618111929577,
|
| 344 |
+
"loss": 0.0893,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.5462653288740246,
|
| 349 |
+
"grad_norm": 0.9520995020866394,
|
| 350 |
+
"learning_rate": 0.00019809557921185523,
|
| 351 |
+
"loss": 0.089,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.5574136008918618,
|
| 356 |
+
"grad_norm": 0.6136443614959717,
|
| 357 |
+
"learning_rate": 0.0001979496398776154,
|
| 358 |
+
"loss": 0.08,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.568561872909699,
|
| 363 |
+
"grad_norm": 0.7204439043998718,
|
| 364 |
+
"learning_rate": 0.00019779837105720434,
|
| 365 |
+
"loss": 0.0753,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.5797101449275363,
|
| 370 |
+
"grad_norm": 0.4742894470691681,
|
| 371 |
+
"learning_rate": 0.00019764178098123002,
|
| 372 |
+
"loss": 0.0755,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.5908584169453734,
|
| 377 |
+
"grad_norm": 0.839692234992981,
|
| 378 |
+
"learning_rate": 0.00019747987816983242,
|
| 379 |
+
"loss": 0.086,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.6020066889632107,
|
| 384 |
+
"grad_norm": 0.5611550807952881,
|
| 385 |
+
"learning_rate": 0.00019731267143222,
|
| 386 |
+
"loss": 0.0795,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.6131549609810479,
|
| 391 |
+
"grad_norm": 0.7129683494567871,
|
| 392 |
+
"learning_rate": 0.00019714016986619034,
|
| 393 |
+
"loss": 0.065,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.6243032329988851,
|
| 398 |
+
"grad_norm": 0.5969811081886292,
|
| 399 |
+
"learning_rate": 0.0001969623828576351,
|
| 400 |
+
"loss": 0.0722,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.6354515050167224,
|
| 405 |
+
"grad_norm": 0.6012031435966492,
|
| 406 |
+
"learning_rate": 0.00019677932008002946,
|
| 407 |
+
"loss": 0.0902,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.6465997770345596,
|
| 412 |
+
"grad_norm": 0.9131368398666382,
|
| 413 |
+
"learning_rate": 0.00019659099149390552,
|
| 414 |
+
"loss": 0.0895,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.6577480490523969,
|
| 419 |
+
"grad_norm": 0.7205947637557983,
|
| 420 |
+
"learning_rate": 0.0001963974073463107,
|
| 421 |
+
"loss": 0.0798,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.6688963210702341,
|
| 426 |
+
"grad_norm": 0.697109043598175,
|
| 427 |
+
"learning_rate": 0.00019619857817024982,
|
| 428 |
+
"loss": 0.083,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.6800445930880713,
|
| 433 |
+
"grad_norm": 0.6167653203010559,
|
| 434 |
+
"learning_rate": 0.00019599451478411232,
|
| 435 |
+
"loss": 0.0899,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.6911928651059086,
|
| 440 |
+
"grad_norm": 0.7218028903007507,
|
| 441 |
+
"learning_rate": 0.00019578522829108344,
|
| 442 |
+
"loss": 0.076,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.7023411371237458,
|
| 447 |
+
"grad_norm": 0.5889600515365601,
|
| 448 |
+
"learning_rate": 0.00019557073007854005,
|
| 449 |
+
"loss": 0.0736,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.7134894091415831,
|
| 454 |
+
"grad_norm": 0.7686411142349243,
|
| 455 |
+
"learning_rate": 0.00019535103181743128,
|
| 456 |
+
"loss": 0.0698,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.7246376811594203,
|
| 461 |
+
"grad_norm": 0.7423682808876038,
|
| 462 |
+
"learning_rate": 0.00019512614546164325,
|
| 463 |
+
"loss": 0.0702,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.7357859531772575,
|
| 468 |
+
"grad_norm": 0.6501685976982117,
|
| 469 |
+
"learning_rate": 0.00019489608324734883,
|
| 470 |
+
"loss": 0.0728,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.7469342251950948,
|
| 475 |
+
"grad_norm": 0.4358183443546295,
|
| 476 |
+
"learning_rate": 0.00019466085769234175,
|
| 477 |
+
"loss": 0.06,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.758082497212932,
|
| 482 |
+
"grad_norm": 0.32865530252456665,
|
| 483 |
+
"learning_rate": 0.00019442048159535555,
|
| 484 |
+
"loss": 0.0794,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.7692307692307693,
|
| 489 |
+
"grad_norm": 0.9116402268409729,
|
| 490 |
+
"learning_rate": 0.00019417496803536723,
|
| 491 |
+
"loss": 0.0689,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.7803790412486065,
|
| 496 |
+
"grad_norm": 0.49878281354904175,
|
| 497 |
+
"learning_rate": 0.00019392433037088554,
|
| 498 |
+
"loss": 0.0716,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.7915273132664437,
|
| 503 |
+
"grad_norm": 0.8435157537460327,
|
| 504 |
+
"learning_rate": 0.0001936685822392242,
|
| 505 |
+
"loss": 0.0724,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.802675585284281,
|
| 510 |
+
"grad_norm": 0.581472635269165,
|
| 511 |
+
"learning_rate": 0.00019340773755575986,
|
| 512 |
+
"loss": 0.0703,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.8138238573021181,
|
| 517 |
+
"grad_norm": 0.5647342801094055,
|
| 518 |
+
"learning_rate": 0.00019314181051317494,
|
| 519 |
+
"loss": 0.0735,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.8249721293199554,
|
| 524 |
+
"grad_norm": 0.3640422523021698,
|
| 525 |
+
"learning_rate": 0.00019287081558068544,
|
| 526 |
+
"loss": 0.0739,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.8361204013377926,
|
| 531 |
+
"grad_norm": 0.6170327067375183,
|
| 532 |
+
"learning_rate": 0.00019259476750325364,
|
| 533 |
+
"loss": 0.0704,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.8472686733556298,
|
| 538 |
+
"grad_norm": 0.49137967824935913,
|
| 539 |
+
"learning_rate": 0.0001923136813007858,
|
| 540 |
+
"loss": 0.0773,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.8584169453734671,
|
| 545 |
+
"grad_norm": 0.616600513458252,
|
| 546 |
+
"learning_rate": 0.00019202757226731502,
|
| 547 |
+
"loss": 0.0696,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.8695652173913043,
|
| 552 |
+
"grad_norm": 0.6217297315597534,
|
| 553 |
+
"learning_rate": 0.00019173645597016893,
|
| 554 |
+
"loss": 0.0703,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.8807134894091416,
|
| 559 |
+
"grad_norm": 0.8147735595703125,
|
| 560 |
+
"learning_rate": 0.0001914403482491227,
|
| 561 |
+
"loss": 0.0627,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.8918617614269788,
|
| 566 |
+
"grad_norm": 0.4161452651023865,
|
| 567 |
+
"learning_rate": 0.00019113926521553736,
|
| 568 |
+
"loss": 0.0835,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.903010033444816,
|
| 573 |
+
"grad_norm": 0.3080677390098572,
|
| 574 |
+
"learning_rate": 0.00019083322325148298,
|
| 575 |
+
"loss": 0.0612,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.9141583054626533,
|
| 580 |
+
"grad_norm": 0.47890082001686096,
|
| 581 |
+
"learning_rate": 0.00019052223900884734,
|
| 582 |
+
"loss": 0.0769,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.9253065774804905,
|
| 587 |
+
"grad_norm": 0.5337966680526733,
|
| 588 |
+
"learning_rate": 0.00019020632940843,
|
| 589 |
+
"loss": 0.0659,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.9364548494983278,
|
| 594 |
+
"grad_norm": 0.5070458054542542,
|
| 595 |
+
"learning_rate": 0.00018988551163902158,
|
| 596 |
+
"loss": 0.0684,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.947603121516165,
|
| 601 |
+
"grad_norm": 0.42387887835502625,
|
| 602 |
+
"learning_rate": 0.00018955980315646834,
|
| 603 |
+
"loss": 0.0661,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.9587513935340022,
|
| 608 |
+
"grad_norm": 0.40716278553009033,
|
| 609 |
+
"learning_rate": 0.00018922922168272276,
|
| 610 |
+
"loss": 0.0735,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.9698996655518395,
|
| 615 |
+
"grad_norm": 0.5013402700424194,
|
| 616 |
+
"learning_rate": 0.00018889378520487894,
|
| 617 |
+
"loss": 0.0684,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.9810479375696767,
|
| 622 |
+
"grad_norm": 0.6920362114906311,
|
| 623 |
+
"learning_rate": 0.00018855351197419407,
|
| 624 |
+
"loss": 0.0769,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.992196209587514,
|
| 629 |
+
"grad_norm": 0.6236600279808044,
|
| 630 |
+
"learning_rate": 0.00018820842050509536,
|
| 631 |
+
"loss": 0.0627,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.0033444816053512,
|
| 636 |
+
"grad_norm": 0.5736573934555054,
|
| 637 |
+
"learning_rate": 0.00018785852957417263,
|
| 638 |
+
"loss": 0.0752,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 1.0144927536231885,
|
| 643 |
+
"grad_norm": 0.41500774025917053,
|
| 644 |
+
"learning_rate": 0.0001875038582191567,
|
| 645 |
+
"loss": 0.0621,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 1.0256410256410255,
|
| 650 |
+
"grad_norm": 0.5921109318733215,
|
| 651 |
+
"learning_rate": 0.00018714442573788357,
|
| 652 |
+
"loss": 0.0814,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 1.0367892976588629,
|
| 657 |
+
"grad_norm": 0.6504522562026978,
|
| 658 |
+
"learning_rate": 0.00018678025168724429,
|
| 659 |
+
"loss": 0.0905,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 1.0479375696767002,
|
| 664 |
+
"grad_norm": 0.7720198035240173,
|
| 665 |
+
"learning_rate": 0.00018641135588212093,
|
| 666 |
+
"loss": 0.0575,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 1.0590858416945372,
|
| 671 |
+
"grad_norm": 0.5303404927253723,
|
| 672 |
+
"learning_rate": 0.00018603775839430852,
|
| 673 |
+
"loss": 0.0698,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.0702341137123745,
|
| 678 |
+
"grad_norm": 0.5405688881874084,
|
| 679 |
+
"learning_rate": 0.0001856594795514229,
|
| 680 |
+
"loss": 0.0771,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 1.0813823857302118,
|
| 685 |
+
"grad_norm": 0.9393702149391174,
|
| 686 |
+
"learning_rate": 0.00018527653993579449,
|
| 687 |
+
"loss": 0.0717,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 1.0925306577480491,
|
| 692 |
+
"grad_norm": 0.4536750614643097,
|
| 693 |
+
"learning_rate": 0.0001848889603833487,
|
| 694 |
+
"loss": 0.0681,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 1.1036789297658862,
|
| 699 |
+
"grad_norm": 0.49481773376464844,
|
| 700 |
+
"learning_rate": 0.00018449676198247202,
|
| 701 |
+
"loss": 0.0702,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 1.1148272017837235,
|
| 706 |
+
"grad_norm": 0.44178834557533264,
|
| 707 |
+
"learning_rate": 0.00018409996607286476,
|
| 708 |
+
"loss": 0.0588,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 1.1259754738015608,
|
| 713 |
+
"grad_norm": 0.7120904326438904,
|
| 714 |
+
"learning_rate": 0.00018369859424437968,
|
| 715 |
+
"loss": 0.0839,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.137123745819398,
|
| 720 |
+
"grad_norm": 0.45710989832878113,
|
| 721 |
+
"learning_rate": 0.00018329266833584765,
|
| 722 |
+
"loss": 0.0641,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 1.1482720178372352,
|
| 727 |
+
"grad_norm": 0.4820857644081116,
|
| 728 |
+
"learning_rate": 0.00018288221043388905,
|
| 729 |
+
"loss": 0.0632,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 1.1594202898550725,
|
| 734 |
+
"grad_norm": 0.5055034160614014,
|
| 735 |
+
"learning_rate": 0.0001824672428717122,
|
| 736 |
+
"loss": 0.0641,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 1.1705685618729098,
|
| 741 |
+
"grad_norm": 0.5027738809585571,
|
| 742 |
+
"learning_rate": 0.00018204778822789814,
|
| 743 |
+
"loss": 0.0632,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 1.1817168338907469,
|
| 748 |
+
"grad_norm": 0.4935217499732971,
|
| 749 |
+
"learning_rate": 0.00018162386932517218,
|
| 750 |
+
"loss": 0.0747,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 1.1928651059085842,
|
| 755 |
+
"grad_norm": 0.6043739914894104,
|
| 756 |
+
"learning_rate": 0.00018119550922916208,
|
| 757 |
+
"loss": 0.059,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.2040133779264215,
|
| 762 |
+
"grad_norm": 0.556803286075592,
|
| 763 |
+
"learning_rate": 0.00018076273124714297,
|
| 764 |
+
"loss": 0.0609,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 1.2151616499442586,
|
| 769 |
+
"grad_norm": 0.41126689314842224,
|
| 770 |
+
"learning_rate": 0.00018032555892676927,
|
| 771 |
+
"loss": 0.0669,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 1.2263099219620959,
|
| 776 |
+
"grad_norm": 0.5530046224594116,
|
| 777 |
+
"learning_rate": 0.00017988401605479343,
|
| 778 |
+
"loss": 0.0592,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 1.2374581939799332,
|
| 783 |
+
"grad_norm": 0.7237766981124878,
|
| 784 |
+
"learning_rate": 0.0001794381266557717,
|
| 785 |
+
"loss": 0.0542,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 1.2486064659977703,
|
| 790 |
+
"grad_norm": 0.4300362765789032,
|
| 791 |
+
"learning_rate": 0.00017898791499075683,
|
| 792 |
+
"loss": 0.0607,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 1.2597547380156076,
|
| 797 |
+
"grad_norm": 0.5814384818077087,
|
| 798 |
+
"learning_rate": 0.00017853340555597818,
|
| 799 |
+
"loss": 0.0712,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.2709030100334449,
|
| 804 |
+
"grad_norm": 0.6296038627624512,
|
| 805 |
+
"learning_rate": 0.00017807462308150882,
|
| 806 |
+
"loss": 0.0571,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 1.282051282051282,
|
| 811 |
+
"grad_norm": 0.6297557353973389,
|
| 812 |
+
"learning_rate": 0.00017761159252991987,
|
| 813 |
+
"loss": 0.0745,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 1.2931995540691192,
|
| 818 |
+
"grad_norm": 0.7536351680755615,
|
| 819 |
+
"learning_rate": 0.00017714433909492234,
|
| 820 |
+
"loss": 0.0678,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 1.3043478260869565,
|
| 825 |
+
"grad_norm": 0.7061651349067688,
|
| 826 |
+
"learning_rate": 0.00017667288819999633,
|
| 827 |
+
"loss": 0.0681,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 1.3154960981047936,
|
| 832 |
+
"grad_norm": 0.5434197187423706,
|
| 833 |
+
"learning_rate": 0.00017619726549700774,
|
| 834 |
+
"loss": 0.0739,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 1.326644370122631,
|
| 839 |
+
"grad_norm": 0.7481071352958679,
|
| 840 |
+
"learning_rate": 0.0001757174968648125,
|
| 841 |
+
"loss": 0.0676,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.3377926421404682,
|
| 846 |
+
"grad_norm": 0.45640698075294495,
|
| 847 |
+
"learning_rate": 0.00017523360840784844,
|
| 848 |
+
"loss": 0.0687,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 1.3489409141583055,
|
| 853 |
+
"grad_norm": 0.38197511434555054,
|
| 854 |
+
"learning_rate": 0.0001747456264547151,
|
| 855 |
+
"loss": 0.0707,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 1.3600891861761428,
|
| 860 |
+
"grad_norm": 0.4496970772743225,
|
| 861 |
+
"learning_rate": 0.00017425357755674107,
|
| 862 |
+
"loss": 0.0627,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 1.37123745819398,
|
| 867 |
+
"grad_norm": 0.3308473527431488,
|
| 868 |
+
"learning_rate": 0.00017375748848653925,
|
| 869 |
+
"loss": 0.0631,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 1.3823857302118172,
|
| 874 |
+
"grad_norm": 0.5428873896598816,
|
| 875 |
+
"learning_rate": 0.00017325738623655032,
|
| 876 |
+
"loss": 0.0649,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 1.3935340022296545,
|
| 881 |
+
"grad_norm": 0.5461141467094421,
|
| 882 |
+
"learning_rate": 0.0001727532980175739,
|
| 883 |
+
"loss": 0.0599,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.4046822742474916,
|
| 888 |
+
"grad_norm": 0.581720769405365,
|
| 889 |
+
"learning_rate": 0.00017224525125728808,
|
| 890 |
+
"loss": 0.0577,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 1.415830546265329,
|
| 895 |
+
"grad_norm": 0.7777343392372131,
|
| 896 |
+
"learning_rate": 0.00017173327359875707,
|
| 897 |
+
"loss": 0.0689,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 1.4269788182831662,
|
| 902 |
+
"grad_norm": 0.42760902643203735,
|
| 903 |
+
"learning_rate": 0.00017121739289892708,
|
| 904 |
+
"loss": 0.0637,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 1.4381270903010033,
|
| 909 |
+
"grad_norm": 0.7080577611923218,
|
| 910 |
+
"learning_rate": 0.0001706976372271107,
|
| 911 |
+
"loss": 0.0687,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 1.4492753623188406,
|
| 916 |
+
"grad_norm": 0.6650500297546387,
|
| 917 |
+
"learning_rate": 0.00017017403486345946,
|
| 918 |
+
"loss": 0.0624,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 1.4604236343366779,
|
| 923 |
+
"grad_norm": 0.47004401683807373,
|
| 924 |
+
"learning_rate": 0.00016964661429742537,
|
| 925 |
+
"loss": 0.0684,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.471571906354515,
|
| 930 |
+
"grad_norm": 0.4500795304775238,
|
| 931 |
+
"learning_rate": 0.00016911540422621048,
|
| 932 |
+
"loss": 0.072,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 1.4827201783723523,
|
| 937 |
+
"grad_norm": 0.6200334429740906,
|
| 938 |
+
"learning_rate": 0.00016858043355320575,
|
| 939 |
+
"loss": 0.0719,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 1.4938684503901896,
|
| 944 |
+
"grad_norm": 0.595638632774353,
|
| 945 |
+
"learning_rate": 0.00016804173138641808,
|
| 946 |
+
"loss": 0.0738,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 1.5050167224080266,
|
| 951 |
+
"grad_norm": 0.40425845980644226,
|
| 952 |
+
"learning_rate": 0.0001674993270368869,
|
| 953 |
+
"loss": 0.0622,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 1.516164994425864,
|
| 958 |
+
"grad_norm": 0.3892674744129181,
|
| 959 |
+
"learning_rate": 0.00016695325001708905,
|
| 960 |
+
"loss": 0.0792,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 1.5273132664437012,
|
| 965 |
+
"grad_norm": 0.2616061270236969,
|
| 966 |
+
"learning_rate": 0.00016640353003933312,
|
| 967 |
+
"loss": 0.058,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 1.5384615384615383,
|
| 972 |
+
"grad_norm": 0.5053898692131042,
|
| 973 |
+
"learning_rate": 0.0001658501970141427,
|
| 974 |
+
"loss": 0.066,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 1.5496098104793758,
|
| 979 |
+
"grad_norm": 0.38966915011405945,
|
| 980 |
+
"learning_rate": 0.00016529328104862912,
|
| 981 |
+
"loss": 0.0574,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 1.560758082497213,
|
| 986 |
+
"grad_norm": 0.45522427558898926,
|
| 987 |
+
"learning_rate": 0.00016473281244485307,
|
| 988 |
+
"loss": 0.0551,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 1.57190635451505,
|
| 993 |
+
"grad_norm": 0.4286852777004242,
|
| 994 |
+
"learning_rate": 0.00016416882169817604,
|
| 995 |
+
"loss": 0.0756,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 1.5830546265328875,
|
| 1000 |
+
"grad_norm": 0.4656994938850403,
|
| 1001 |
+
"learning_rate": 0.00016360133949560094,
|
| 1002 |
+
"loss": 0.0573,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 1.5942028985507246,
|
| 1007 |
+
"grad_norm": 0.637538731098175,
|
| 1008 |
+
"learning_rate": 0.00016303039671410248,
|
| 1009 |
+
"loss": 0.062,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 1.605351170568562,
|
| 1014 |
+
"grad_norm": 0.722161054611206,
|
| 1015 |
+
"learning_rate": 0.00016245602441894704,
|
| 1016 |
+
"loss": 0.0657,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 1.6164994425863992,
|
| 1021 |
+
"grad_norm": 0.536309540271759,
|
| 1022 |
+
"learning_rate": 0.0001618782538620025,
|
| 1023 |
+
"loss": 0.0689,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 1.6276477146042363,
|
| 1028 |
+
"grad_norm": 0.27511873841285706,
|
| 1029 |
+
"learning_rate": 0.00016129711648003777,
|
| 1030 |
+
"loss": 0.044,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 1.6387959866220736,
|
| 1035 |
+
"grad_norm": 0.4901306629180908,
|
| 1036 |
+
"learning_rate": 0.00016071264389301234,
|
| 1037 |
+
"loss": 0.0548,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 1.649944258639911,
|
| 1042 |
+
"grad_norm": 0.29372069239616394,
|
| 1043 |
+
"learning_rate": 0.00016012486790235574,
|
| 1044 |
+
"loss": 0.0543,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 1.661092530657748,
|
| 1049 |
+
"grad_norm": 0.5928181409835815,
|
| 1050 |
+
"learning_rate": 0.00015953382048923719,
|
| 1051 |
+
"loss": 0.0618,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 1.6722408026755853,
|
| 1056 |
+
"grad_norm": 0.4672330319881439,
|
| 1057 |
+
"learning_rate": 0.00015893953381282563,
|
| 1058 |
+
"loss": 0.0603,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 1.6833890746934226,
|
| 1063 |
+
"grad_norm": 0.5659995675086975,
|
| 1064 |
+
"learning_rate": 0.00015834204020853988,
|
| 1065 |
+
"loss": 0.0521,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 1.6945373467112597,
|
| 1070 |
+
"grad_norm": 0.44387733936309814,
|
| 1071 |
+
"learning_rate": 0.0001577413721862891,
|
| 1072 |
+
"loss": 0.0626,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 1.705685618729097,
|
| 1077 |
+
"grad_norm": 0.5992341041564941,
|
| 1078 |
+
"learning_rate": 0.00015713756242870415,
|
| 1079 |
+
"loss": 0.0569,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 1.7168338907469343,
|
| 1084 |
+
"grad_norm": 0.4298606216907501,
|
| 1085 |
+
"learning_rate": 0.0001565306437893591,
|
| 1086 |
+
"loss": 0.0555,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 1.7279821627647713,
|
| 1091 |
+
"grad_norm": 0.47163352370262146,
|
| 1092 |
+
"learning_rate": 0.00015592064929098384,
|
| 1093 |
+
"loss": 0.0686,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 1.7391304347826086,
|
| 1098 |
+
"grad_norm": 0.5358185768127441,
|
| 1099 |
+
"learning_rate": 0.00015530761212366714,
|
| 1100 |
+
"loss": 0.0484,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 1.750278706800446,
|
| 1105 |
+
"grad_norm": 0.5074977278709412,
|
| 1106 |
+
"learning_rate": 0.00015469156564305075,
|
| 1107 |
+
"loss": 0.0507,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 1.761426978818283,
|
| 1112 |
+
"grad_norm": 0.5983792543411255,
|
| 1113 |
+
"learning_rate": 0.00015407254336851479,
|
| 1114 |
+
"loss": 0.0636,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 1.7725752508361206,
|
| 1119 |
+
"grad_norm": 0.699416995048523,
|
| 1120 |
+
"learning_rate": 0.00015345057898135355,
|
| 1121 |
+
"loss": 0.066,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 1.7837235228539576,
|
| 1126 |
+
"grad_norm": 0.7820698022842407,
|
| 1127 |
+
"learning_rate": 0.0001528257063229431,
|
| 1128 |
+
"loss": 0.059,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 1.7948717948717947,
|
| 1133 |
+
"grad_norm": 0.4525439143180847,
|
| 1134 |
+
"learning_rate": 0.00015219795939289997,
|
| 1135 |
+
"loss": 0.0662,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 1.8060200668896322,
|
| 1140 |
+
"grad_norm": 0.4681321382522583,
|
| 1141 |
+
"learning_rate": 0.0001515673723472311,
|
| 1142 |
+
"loss": 0.0625,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 1.8171683389074693,
|
| 1147 |
+
"grad_norm": 0.5068295001983643,
|
| 1148 |
+
"learning_rate": 0.00015093397949647548,
|
| 1149 |
+
"loss": 0.0611,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 1.8283166109253066,
|
| 1154 |
+
"grad_norm": 0.709778368473053,
|
| 1155 |
+
"learning_rate": 0.00015029781530383738,
|
| 1156 |
+
"loss": 0.0506,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 1.839464882943144,
|
| 1161 |
+
"grad_norm": 0.4446347653865814,
|
| 1162 |
+
"learning_rate": 0.0001496589143833109,
|
| 1163 |
+
"loss": 0.0515,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 1.850613154960981,
|
| 1168 |
+
"grad_norm": 0.5426603555679321,
|
| 1169 |
+
"learning_rate": 0.000149017311497797,
|
| 1170 |
+
"loss": 0.0644,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 1.8617614269788183,
|
| 1175 |
+
"grad_norm": 0.4359883964061737,
|
| 1176 |
+
"learning_rate": 0.00014837304155721165,
|
| 1177 |
+
"loss": 0.0583,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 1.8729096989966556,
|
| 1182 |
+
"grad_norm": 0.4507809579372406,
|
| 1183 |
+
"learning_rate": 0.00014772613961658673,
|
| 1184 |
+
"loss": 0.0615,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 1.8840579710144927,
|
| 1189 |
+
"grad_norm": 0.3186969459056854,
|
| 1190 |
+
"learning_rate": 0.00014707664087416228,
|
| 1191 |
+
"loss": 0.0669,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 1.89520624303233,
|
| 1196 |
+
"grad_norm": 0.3906411826610565,
|
| 1197 |
+
"learning_rate": 0.00014642458066947174,
|
| 1198 |
+
"loss": 0.0667,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 1.9063545150501673,
|
| 1203 |
+
"grad_norm": 1.0206559896469116,
|
| 1204 |
+
"learning_rate": 0.00014576999448141876,
|
| 1205 |
+
"loss": 0.0533,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 1.9175027870680044,
|
| 1210 |
+
"grad_norm": 0.5162254571914673,
|
| 1211 |
+
"learning_rate": 0.0001451129179263471,
|
| 1212 |
+
"loss": 0.0593,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 1.9286510590858417,
|
| 1217 |
+
"grad_norm": 0.5980589985847473,
|
| 1218 |
+
"learning_rate": 0.00014445338675610247,
|
| 1219 |
+
"loss": 0.0565,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 1.939799331103679,
|
| 1224 |
+
"grad_norm": 0.28657665848731995,
|
| 1225 |
+
"learning_rate": 0.00014379143685608735,
|
| 1226 |
+
"loss": 0.0603,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 1.950947603121516,
|
| 1231 |
+
"grad_norm": 0.3899282217025757,
|
| 1232 |
+
"learning_rate": 0.00014312710424330843,
|
| 1233 |
+
"loss": 0.0514,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 1.9620958751393534,
|
| 1238 |
+
"grad_norm": 0.36454200744628906,
|
| 1239 |
+
"learning_rate": 0.000142460425064417,
|
| 1240 |
+
"loss": 0.052,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 1.9732441471571907,
|
| 1245 |
+
"grad_norm": 0.608792245388031,
|
| 1246 |
+
"learning_rate": 0.00014179143559374212,
|
| 1247 |
+
"loss": 0.0649,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 1.9843924191750277,
|
| 1252 |
+
"grad_norm": 0.580376386642456,
|
| 1253 |
+
"learning_rate": 0.0001411201722313168,
|
| 1254 |
+
"loss": 0.0459,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 1.9955406911928653,
|
| 1259 |
+
"grad_norm": 0.7000721096992493,
|
| 1260 |
+
"learning_rate": 0.0001404466715008977,
|
| 1261 |
+
"loss": 0.0704,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 2.0066889632107023,
|
| 1266 |
+
"grad_norm": 0.5400611162185669,
|
| 1267 |
+
"learning_rate": 0.00013977097004797768,
|
| 1268 |
+
"loss": 0.0552,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 2.0178372352285394,
|
| 1273 |
+
"grad_norm": 0.5788174271583557,
|
| 1274 |
+
"learning_rate": 0.00013909310463779197,
|
| 1275 |
+
"loss": 0.0465,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 2.028985507246377,
|
| 1280 |
+
"grad_norm": 0.5689246654510498,
|
| 1281 |
+
"learning_rate": 0.00013841311215331768,
|
| 1282 |
+
"loss": 0.0479,
|
| 1283 |
+
"step": 1820
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 2.040133779264214,
|
| 1287 |
+
"grad_norm": 0.5124692320823669,
|
| 1288 |
+
"learning_rate": 0.00013773102959326709,
|
| 1289 |
+
"loss": 0.05,
|
| 1290 |
+
"step": 1830
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 2.051282051282051,
|
| 1294 |
+
"grad_norm": 0.46690040826797485,
|
| 1295 |
+
"learning_rate": 0.00013704689407007447,
|
| 1296 |
+
"loss": 0.0569,
|
| 1297 |
+
"step": 1840
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 2.0624303232998886,
|
| 1301 |
+
"grad_norm": 0.785712718963623,
|
| 1302 |
+
"learning_rate": 0.00013636074280787681,
|
| 1303 |
+
"loss": 0.0575,
|
| 1304 |
+
"step": 1850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 2.0735785953177257,
|
| 1308 |
+
"grad_norm": 0.4149536192417145,
|
| 1309 |
+
"learning_rate": 0.00013567261314048834,
|
| 1310 |
+
"loss": 0.0566,
|
| 1311 |
+
"step": 1860
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 2.084726867335563,
|
| 1315 |
+
"grad_norm": 0.6050755381584167,
|
| 1316 |
+
"learning_rate": 0.0001349825425093693,
|
| 1317 |
+
"loss": 0.0557,
|
| 1318 |
+
"step": 1870
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 2.0958751393534003,
|
| 1322 |
+
"grad_norm": 0.34781813621520996,
|
| 1323 |
+
"learning_rate": 0.00013429056846158872,
|
| 1324 |
+
"loss": 0.0532,
|
| 1325 |
+
"step": 1880
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 2.1070234113712374,
|
| 1329 |
+
"grad_norm": 0.48258450627326965,
|
| 1330 |
+
"learning_rate": 0.0001335967286477813,
|
| 1331 |
+
"loss": 0.0574,
|
| 1332 |
+
"step": 1890
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 2.1181716833890745,
|
| 1336 |
+
"grad_norm": 0.43237388134002686,
|
| 1337 |
+
"learning_rate": 0.00013290106082009914,
|
| 1338 |
+
"loss": 0.0597,
|
| 1339 |
+
"step": 1900
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 2.129319955406912,
|
| 1343 |
+
"grad_norm": 0.3022902309894562,
|
| 1344 |
+
"learning_rate": 0.00013220360283015727,
|
| 1345 |
+
"loss": 0.0527,
|
| 1346 |
+
"step": 1910
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 2.140468227424749,
|
| 1350 |
+
"grad_norm": 0.5687480568885803,
|
| 1351 |
+
"learning_rate": 0.00013150439262697437,
|
| 1352 |
+
"loss": 0.0473,
|
| 1353 |
+
"step": 1920
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 2.1516164994425866,
|
| 1357 |
+
"grad_norm": 0.42970651388168335,
|
| 1358 |
+
"learning_rate": 0.00013080346825490789,
|
| 1359 |
+
"loss": 0.0531,
|
| 1360 |
+
"step": 1930
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 2.1627647714604237,
|
| 1364 |
+
"grad_norm": 0.5530820488929749,
|
| 1365 |
+
"learning_rate": 0.00013010086785158396,
|
| 1366 |
+
"loss": 0.0464,
|
| 1367 |
+
"step": 1940
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 2.1739130434782608,
|
| 1371 |
+
"grad_norm": 0.38890719413757324,
|
| 1372 |
+
"learning_rate": 0.00012939662964582234,
|
| 1373 |
+
"loss": 0.0586,
|
| 1374 |
+
"step": 1950
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 2.1850613154960983,
|
| 1378 |
+
"grad_norm": 0.4155096709728241,
|
| 1379 |
+
"learning_rate": 0.0001286907919555566,
|
| 1380 |
+
"loss": 0.0529,
|
| 1381 |
+
"step": 1960
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 2.1962095875139354,
|
| 1385 |
+
"grad_norm": 0.5036742091178894,
|
| 1386 |
+
"learning_rate": 0.00012798339318574871,
|
| 1387 |
+
"loss": 0.0548,
|
| 1388 |
+
"step": 1970
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 2.2073578595317724,
|
| 1392 |
+
"grad_norm": 0.4967355728149414,
|
| 1393 |
+
"learning_rate": 0.00012727447182629993,
|
| 1394 |
+
"loss": 0.0523,
|
| 1395 |
+
"step": 1980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 2.21850613154961,
|
| 1399 |
+
"grad_norm": 0.3667829632759094,
|
| 1400 |
+
"learning_rate": 0.00012656406644995623,
|
| 1401 |
+
"loss": 0.0541,
|
| 1402 |
+
"step": 1990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 2.229654403567447,
|
| 1406 |
+
"grad_norm": 0.5501829981803894,
|
| 1407 |
+
"learning_rate": 0.00012585221571020979,
|
| 1408 |
+
"loss": 0.0542,
|
| 1409 |
+
"step": 2000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 2.240802675585284,
|
| 1413 |
+
"grad_norm": 0.42376476526260376,
|
| 1414 |
+
"learning_rate": 0.0001251389583391955,
|
| 1415 |
+
"loss": 0.0572,
|
| 1416 |
+
"step": 2010
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 2.2519509476031216,
|
| 1420 |
+
"grad_norm": 0.49177291989326477,
|
| 1421 |
+
"learning_rate": 0.00012442433314558386,
|
| 1422 |
+
"loss": 0.0475,
|
| 1423 |
+
"step": 2020
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 2.2630992196209587,
|
| 1427 |
+
"grad_norm": 0.5596216917037964,
|
| 1428 |
+
"learning_rate": 0.00012370837901246918,
|
| 1429 |
+
"loss": 0.0471,
|
| 1430 |
+
"step": 2030
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 2.274247491638796,
|
| 1434 |
+
"grad_norm": 0.5342261791229248,
|
| 1435 |
+
"learning_rate": 0.00012299113489525404,
|
| 1436 |
+
"loss": 0.0516,
|
| 1437 |
+
"step": 2040
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 2.2853957636566333,
|
| 1441 |
+
"grad_norm": 0.8374896049499512,
|
| 1442 |
+
"learning_rate": 0.00012227263981952962,
|
| 1443 |
+
"loss": 0.0595,
|
| 1444 |
+
"step": 2050
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 2.2965440356744704,
|
| 1448 |
+
"grad_norm": 0.2919222116470337,
|
| 1449 |
+
"learning_rate": 0.00012155293287895237,
|
| 1450 |
+
"loss": 0.0528,
|
| 1451 |
+
"step": 2060
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 2.3076923076923075,
|
| 1455 |
+
"grad_norm": 0.29067209362983704,
|
| 1456 |
+
"learning_rate": 0.00012083205323311689,
|
| 1457 |
+
"loss": 0.0579,
|
| 1458 |
+
"step": 2070
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 2.318840579710145,
|
| 1462 |
+
"grad_norm": 0.2780458629131317,
|
| 1463 |
+
"learning_rate": 0.00012011004010542519,
|
| 1464 |
+
"loss": 0.0503,
|
| 1465 |
+
"step": 2080
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 2.329988851727982,
|
| 1469 |
+
"grad_norm": 0.5889238119125366,
|
| 1470 |
+
"learning_rate": 0.00011938693278095256,
|
| 1471 |
+
"loss": 0.0534,
|
| 1472 |
+
"step": 2090
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 2.3411371237458196,
|
| 1476 |
+
"grad_norm": 0.4364779591560364,
|
| 1477 |
+
"learning_rate": 0.00011866277060431015,
|
| 1478 |
+
"loss": 0.0558,
|
| 1479 |
+
"step": 2100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 2.3522853957636567,
|
| 1483 |
+
"grad_norm": 0.407135009765625,
|
| 1484 |
+
"learning_rate": 0.00011793759297750398,
|
| 1485 |
+
"loss": 0.0512,
|
| 1486 |
+
"step": 2110
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 2.3634336677814938,
|
| 1490 |
+
"grad_norm": 0.6752089262008667,
|
| 1491 |
+
"learning_rate": 0.00011721143935779124,
|
| 1492 |
+
"loss": 0.0599,
|
| 1493 |
+
"step": 2120
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 2.374581939799331,
|
| 1497 |
+
"grad_norm": 0.7100276947021484,
|
| 1498 |
+
"learning_rate": 0.00011648434925553342,
|
| 1499 |
+
"loss": 0.0534,
|
| 1500 |
+
"step": 2130
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 2.3857302118171684,
|
| 1504 |
+
"grad_norm": 0.5005069375038147,
|
| 1505 |
+
"learning_rate": 0.0001157563622320464,
|
| 1506 |
+
"loss": 0.055,
|
| 1507 |
+
"step": 2140
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 2.3968784838350055,
|
| 1511 |
+
"grad_norm": 0.48492226004600525,
|
| 1512 |
+
"learning_rate": 0.00011502751789744798,
|
| 1513 |
+
"loss": 0.0578,
|
| 1514 |
+
"step": 2150
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 2.408026755852843,
|
| 1518 |
+
"grad_norm": 0.34049931168556213,
|
| 1519 |
+
"learning_rate": 0.00011429785590850266,
|
| 1520 |
+
"loss": 0.0569,
|
| 1521 |
+
"step": 2160
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 2.41917502787068,
|
| 1525 |
+
"grad_norm": 0.40999093651771545,
|
| 1526 |
+
"learning_rate": 0.00011356741596646392,
|
| 1527 |
+
"loss": 0.0518,
|
| 1528 |
+
"step": 2170
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 2.430323299888517,
|
| 1532 |
+
"grad_norm": 0.2947206497192383,
|
| 1533 |
+
"learning_rate": 0.00011283623781491402,
|
| 1534 |
+
"loss": 0.0379,
|
| 1535 |
+
"step": 2180
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 2.4414715719063547,
|
| 1539 |
+
"grad_norm": 0.3870508074760437,
|
| 1540 |
+
"learning_rate": 0.00011210436123760155,
|
| 1541 |
+
"loss": 0.0472,
|
| 1542 |
+
"step": 2190
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 2.4526198439241917,
|
| 1546 |
+
"grad_norm": 0.594549834728241,
|
| 1547 |
+
"learning_rate": 0.00011137182605627677,
|
| 1548 |
+
"loss": 0.048,
|
| 1549 |
+
"step": 2200
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 2.463768115942029,
|
| 1553 |
+
"grad_norm": 0.6832001805305481,
|
| 1554 |
+
"learning_rate": 0.00011063867212852492,
|
| 1555 |
+
"loss": 0.0588,
|
| 1556 |
+
"step": 2210
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 2.4749163879598663,
|
| 1560 |
+
"grad_norm": 0.34398114681243896,
|
| 1561 |
+
"learning_rate": 0.00010990493934559746,
|
| 1562 |
+
"loss": 0.0439,
|
| 1563 |
+
"step": 2220
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 2.4860646599777034,
|
| 1567 |
+
"grad_norm": 0.4179747402667999,
|
| 1568 |
+
"learning_rate": 0.00010917066763024175,
|
| 1569 |
+
"loss": 0.0544,
|
| 1570 |
+
"step": 2230
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 2.4972129319955405,
|
| 1574 |
+
"grad_norm": 0.3994947373867035,
|
| 1575 |
+
"learning_rate": 0.0001084358969345286,
|
| 1576 |
+
"loss": 0.0442,
|
| 1577 |
+
"step": 2240
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 2.508361204013378,
|
| 1581 |
+
"grad_norm": 0.5612298846244812,
|
| 1582 |
+
"learning_rate": 0.00010770066723767862,
|
| 1583 |
+
"loss": 0.0541,
|
| 1584 |
+
"step": 2250
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 2.519509476031215,
|
| 1588 |
+
"grad_norm": 0.3757084608078003,
|
| 1589 |
+
"learning_rate": 0.00010696501854388692,
|
| 1590 |
+
"loss": 0.0596,
|
| 1591 |
+
"step": 2260
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 2.5306577480490526,
|
| 1595 |
+
"grad_norm": 0.43906617164611816,
|
| 1596 |
+
"learning_rate": 0.00010622899088014637,
|
| 1597 |
+
"loss": 0.054,
|
| 1598 |
+
"step": 2270
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 2.5418060200668897,
|
| 1602 |
+
"grad_norm": 0.3490893542766571,
|
| 1603 |
+
"learning_rate": 0.00010549262429406988,
|
| 1604 |
+
"loss": 0.0515,
|
| 1605 |
+
"step": 2280
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 2.552954292084727,
|
| 1609 |
+
"grad_norm": 0.6981559991836548,
|
| 1610 |
+
"learning_rate": 0.00010475595885171112,
|
| 1611 |
+
"loss": 0.0557,
|
| 1612 |
+
"step": 2290
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 2.564102564102564,
|
| 1616 |
+
"grad_norm": 0.5708040595054626,
|
| 1617 |
+
"learning_rate": 0.00010401903463538483,
|
| 1618 |
+
"loss": 0.0529,
|
| 1619 |
+
"step": 2300
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 2.5752508361204014,
|
| 1623 |
+
"grad_norm": 0.48492324352264404,
|
| 1624 |
+
"learning_rate": 0.00010328189174148568,
|
| 1625 |
+
"loss": 0.0493,
|
| 1626 |
+
"step": 2310
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 2.5863991081382385,
|
| 1630 |
+
"grad_norm": 0.5425471067428589,
|
| 1631 |
+
"learning_rate": 0.0001025445702783067,
|
| 1632 |
+
"loss": 0.051,
|
| 1633 |
+
"step": 2320
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 2.597547380156076,
|
| 1637 |
+
"grad_norm": 0.3145391345024109,
|
| 1638 |
+
"learning_rate": 0.00010180711036385698,
|
| 1639 |
+
"loss": 0.0477,
|
| 1640 |
+
"step": 2330
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 2.608695652173913,
|
| 1644 |
+
"grad_norm": 0.6779066920280457,
|
| 1645 |
+
"learning_rate": 0.00010106955212367881,
|
| 1646 |
+
"loss": 0.0475,
|
| 1647 |
+
"step": 2340
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 2.61984392419175,
|
| 1651 |
+
"grad_norm": 0.3190683126449585,
|
| 1652 |
+
"learning_rate": 0.00010033193568866437,
|
| 1653 |
+
"loss": 0.0528,
|
| 1654 |
+
"step": 2350
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 2.6309921962095872,
|
| 1658 |
+
"grad_norm": 0.41678014397621155,
|
| 1659 |
+
"learning_rate": 9.959430119287237e-05,
|
| 1660 |
+
"loss": 0.0472,
|
| 1661 |
+
"step": 2360
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 2.6421404682274248,
|
| 1665 |
+
"grad_norm": 0.32865625619888306,
|
| 1666 |
+
"learning_rate": 9.885668877134405e-05,
|
| 1667 |
+
"loss": 0.047,
|
| 1668 |
+
"step": 2370
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 2.653288740245262,
|
| 1672 |
+
"grad_norm": 0.5178107023239136,
|
| 1673 |
+
"learning_rate": 9.811913855791974e-05,
|
| 1674 |
+
"loss": 0.0433,
|
| 1675 |
+
"step": 2380
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 2.6644370122630994,
|
| 1679 |
+
"grad_norm": 0.3974803686141968,
|
| 1680 |
+
"learning_rate": 9.738169068305487e-05,
|
| 1681 |
+
"loss": 0.0574,
|
| 1682 |
+
"step": 2390
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 2.6755852842809364,
|
| 1686 |
+
"grad_norm": 0.41880396008491516,
|
| 1687 |
+
"learning_rate": 9.664438527163666e-05,
|
| 1688 |
+
"loss": 0.0479,
|
| 1689 |
+
"step": 2400
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 2.6867335562987735,
|
| 1693 |
+
"grad_norm": 0.45363089442253113,
|
| 1694 |
+
"learning_rate": 9.590726244080084e-05,
|
| 1695 |
+
"loss": 0.0512,
|
| 1696 |
+
"step": 2410
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 2.697881828316611,
|
| 1700 |
+
"grad_norm": 0.27054697275161743,
|
| 1701 |
+
"learning_rate": 9.517036229774875e-05,
|
| 1702 |
+
"loss": 0.0443,
|
| 1703 |
+
"step": 2420
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 2.709030100334448,
|
| 1707 |
+
"grad_norm": 0.2927968502044678,
|
| 1708 |
+
"learning_rate": 9.443372493756531e-05,
|
| 1709 |
+
"loss": 0.0459,
|
| 1710 |
+
"step": 2430
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 2.7201783723522857,
|
| 1714 |
+
"grad_norm": 0.5281320810317993,
|
| 1715 |
+
"learning_rate": 9.369739044103713e-05,
|
| 1716 |
+
"loss": 0.0487,
|
| 1717 |
+
"step": 2440
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 2.7313266443701227,
|
| 1721 |
+
"grad_norm": 0.624927818775177,
|
| 1722 |
+
"learning_rate": 9.296139887247209e-05,
|
| 1723 |
+
"loss": 0.0589,
|
| 1724 |
+
"step": 2450
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 2.74247491638796,
|
| 1728 |
+
"grad_norm": 0.5565568804740906,
|
| 1729 |
+
"learning_rate": 9.222579027751895e-05,
|
| 1730 |
+
"loss": 0.0517,
|
| 1731 |
+
"step": 2460
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 2.753623188405797,
|
| 1735 |
+
"grad_norm": 0.34589365124702454,
|
| 1736 |
+
"learning_rate": 9.149060468098891e-05,
|
| 1737 |
+
"loss": 0.0442,
|
| 1738 |
+
"step": 2470
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 2.7647714604236344,
|
| 1742 |
+
"grad_norm": 0.4209279716014862,
|
| 1743 |
+
"learning_rate": 9.075588208467748e-05,
|
| 1744 |
+
"loss": 0.0541,
|
| 1745 |
+
"step": 2480
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 2.7759197324414715,
|
| 1749 |
+
"grad_norm": 0.6148848533630371,
|
| 1750 |
+
"learning_rate": 9.002166246518821e-05,
|
| 1751 |
+
"loss": 0.0469,
|
| 1752 |
+
"step": 2490
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 2.787068004459309,
|
| 1756 |
+
"grad_norm": 0.40948551893234253,
|
| 1757 |
+
"learning_rate": 8.928798577175735e-05,
|
| 1758 |
+
"loss": 0.0426,
|
| 1759 |
+
"step": 2500
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 2.798216276477146,
|
| 1763 |
+
"grad_norm": 0.3225531280040741,
|
| 1764 |
+
"learning_rate": 8.855489192408039e-05,
|
| 1765 |
+
"loss": 0.0456,
|
| 1766 |
+
"step": 2510
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 2.809364548494983,
|
| 1770 |
+
"grad_norm": 0.4164505898952484,
|
| 1771 |
+
"learning_rate": 8.782242081013979e-05,
|
| 1772 |
+
"loss": 0.0403,
|
| 1773 |
+
"step": 2520
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 2.8205128205128203,
|
| 1777 |
+
"grad_norm": 0.47506433725357056,
|
| 1778 |
+
"learning_rate": 8.709061228403487e-05,
|
| 1779 |
+
"loss": 0.0512,
|
| 1780 |
+
"step": 2530
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 2.831661092530658,
|
| 1784 |
+
"grad_norm": 0.32749128341674805,
|
| 1785 |
+
"learning_rate": 8.635950616381317e-05,
|
| 1786 |
+
"loss": 0.0553,
|
| 1787 |
+
"step": 2540
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 2.842809364548495,
|
| 1791 |
+
"grad_norm": 0.4301365911960602,
|
| 1792 |
+
"learning_rate": 8.562914222930408e-05,
|
| 1793 |
+
"loss": 0.0412,
|
| 1794 |
+
"step": 2550
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 2.8539576365663324,
|
| 1798 |
+
"grad_norm": 0.6139765381813049,
|
| 1799 |
+
"learning_rate": 8.489956021995418e-05,
|
| 1800 |
+
"loss": 0.0556,
|
| 1801 |
+
"step": 2560
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 2.8651059085841695,
|
| 1805 |
+
"grad_norm": 0.3337389826774597,
|
| 1806 |
+
"learning_rate": 8.417079983266527e-05,
|
| 1807 |
+
"loss": 0.0519,
|
| 1808 |
+
"step": 2570
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 2.8762541806020065,
|
| 1812 |
+
"grad_norm": 0.5801304578781128,
|
| 1813 |
+
"learning_rate": 8.344290071963425e-05,
|
| 1814 |
+
"loss": 0.0522,
|
| 1815 |
+
"step": 2580
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 2.887402452619844,
|
| 1819 |
+
"grad_norm": 0.5481535792350769,
|
| 1820 |
+
"learning_rate": 8.271590248619558e-05,
|
| 1821 |
+
"loss": 0.0482,
|
| 1822 |
+
"step": 2590
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 2.898550724637681,
|
| 1826 |
+
"grad_norm": 0.5965551733970642,
|
| 1827 |
+
"learning_rate": 8.198984468866665e-05,
|
| 1828 |
+
"loss": 0.0536,
|
| 1829 |
+
"step": 2600
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 2.9096989966555182,
|
| 1833 |
+
"grad_norm": 0.5813583731651306,
|
| 1834 |
+
"learning_rate": 8.126476683219509e-05,
|
| 1835 |
+
"loss": 0.0535,
|
| 1836 |
+
"step": 2610
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 2.9208472686733558,
|
| 1840 |
+
"grad_norm": 0.6612478494644165,
|
| 1841 |
+
"learning_rate": 8.054070836860959e-05,
|
| 1842 |
+
"loss": 0.0511,
|
| 1843 |
+
"step": 2620
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 2.931995540691193,
|
| 1847 |
+
"grad_norm": 0.6427285671234131,
|
| 1848 |
+
"learning_rate": 7.981770869427315e-05,
|
| 1849 |
+
"loss": 0.0496,
|
| 1850 |
+
"step": 2630
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 2.94314381270903,
|
| 1854 |
+
"grad_norm": 0.4178491234779358,
|
| 1855 |
+
"learning_rate": 7.909580714793954e-05,
|
| 1856 |
+
"loss": 0.0505,
|
| 1857 |
+
"step": 2640
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 2.9542920847268674,
|
| 1861 |
+
"grad_norm": 0.28486618399620056,
|
| 1862 |
+
"learning_rate": 7.837504300861289e-05,
|
| 1863 |
+
"loss": 0.0398,
|
| 1864 |
+
"step": 2650
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 2.9654403567447045,
|
| 1868 |
+
"grad_norm": 0.3555268347263336,
|
| 1869 |
+
"learning_rate": 7.765545549341047e-05,
|
| 1870 |
+
"loss": 0.0429,
|
| 1871 |
+
"step": 2660
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 2.976588628762542,
|
| 1875 |
+
"grad_norm": 0.4297826588153839,
|
| 1876 |
+
"learning_rate": 7.693708375542885e-05,
|
| 1877 |
+
"loss": 0.0526,
|
| 1878 |
+
"step": 2670
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 2.987736900780379,
|
| 1882 |
+
"grad_norm": 0.37716731429100037,
|
| 1883 |
+
"learning_rate": 7.621996688161364e-05,
|
| 1884 |
+
"loss": 0.0476,
|
| 1885 |
+
"step": 2680
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 2.998885172798216,
|
| 1889 |
+
"grad_norm": 0.34564629197120667,
|
| 1890 |
+
"learning_rate": 7.550414389063262e-05,
|
| 1891 |
+
"loss": 0.0413,
|
| 1892 |
+
"step": 2690
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 3.0100334448160537,
|
| 1896 |
+
"grad_norm": 0.41150006651878357,
|
| 1897 |
+
"learning_rate": 7.478965373075291e-05,
|
| 1898 |
+
"loss": 0.0471,
|
| 1899 |
+
"step": 2700
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 3.021181716833891,
|
| 1903 |
+
"grad_norm": 0.47604137659072876,
|
| 1904 |
+
"learning_rate": 7.407653527772152e-05,
|
| 1905 |
+
"loss": 0.0439,
|
| 1906 |
+
"step": 2710
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 3.032329988851728,
|
| 1910 |
+
"grad_norm": 0.29454073309898376,
|
| 1911 |
+
"learning_rate": 7.336482733265037e-05,
|
| 1912 |
+
"loss": 0.0442,
|
| 1913 |
+
"step": 2720
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 3.0434782608695654,
|
| 1917 |
+
"grad_norm": 0.3397221863269806,
|
| 1918 |
+
"learning_rate": 7.265456861990488e-05,
|
| 1919 |
+
"loss": 0.0333,
|
| 1920 |
+
"step": 2730
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 3.0546265328874025,
|
| 1924 |
+
"grad_norm": 0.4078027904033661,
|
| 1925 |
+
"learning_rate": 7.194579778499712e-05,
|
| 1926 |
+
"loss": 0.0367,
|
| 1927 |
+
"step": 2740
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 3.0657748049052396,
|
| 1931 |
+
"grad_norm": 0.3360117971897125,
|
| 1932 |
+
"learning_rate": 7.123855339248297e-05,
|
| 1933 |
+
"loss": 0.0475,
|
| 1934 |
+
"step": 2750
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 3.076923076923077,
|
| 1938 |
+
"grad_norm": 0.36956536769866943,
|
| 1939 |
+
"learning_rate": 7.053287392386393e-05,
|
| 1940 |
+
"loss": 0.0464,
|
| 1941 |
+
"step": 2760
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 3.088071348940914,
|
| 1945 |
+
"grad_norm": 0.3538147509098053,
|
| 1946 |
+
"learning_rate": 6.982879777549316e-05,
|
| 1947 |
+
"loss": 0.035,
|
| 1948 |
+
"step": 2770
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 3.0992196209587513,
|
| 1952 |
+
"grad_norm": 0.2888403534889221,
|
| 1953 |
+
"learning_rate": 6.91263632564865e-05,
|
| 1954 |
+
"loss": 0.0373,
|
| 1955 |
+
"step": 2780
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 3.1103678929765888,
|
| 1959 |
+
"grad_norm": 0.5293031930923462,
|
| 1960 |
+
"learning_rate": 6.842560858663785e-05,
|
| 1961 |
+
"loss": 0.0487,
|
| 1962 |
+
"step": 2790
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 3.121516164994426,
|
| 1966 |
+
"grad_norm": 0.46511995792388916,
|
| 1967 |
+
"learning_rate": 6.772657189433989e-05,
|
| 1968 |
+
"loss": 0.0488,
|
| 1969 |
+
"step": 2800
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 3.132664437012263,
|
| 1973 |
+
"grad_norm": 0.42650702595710754,
|
| 1974 |
+
"learning_rate": 6.702929121450913e-05,
|
| 1975 |
+
"loss": 0.0406,
|
| 1976 |
+
"step": 2810
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 3.1438127090301005,
|
| 1980 |
+
"grad_norm": 0.7566830515861511,
|
| 1981 |
+
"learning_rate": 6.633380448651678e-05,
|
| 1982 |
+
"loss": 0.0432,
|
| 1983 |
+
"step": 2820
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 3.1549609810479375,
|
| 1987 |
+
"grad_norm": 0.4249333441257477,
|
| 1988 |
+
"learning_rate": 6.564014955212416e-05,
|
| 1989 |
+
"loss": 0.0604,
|
| 1990 |
+
"step": 2830
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 3.1661092530657746,
|
| 1994 |
+
"grad_norm": 0.435446172952652,
|
| 1995 |
+
"learning_rate": 6.494836415342396e-05,
|
| 1996 |
+
"loss": 0.0487,
|
| 1997 |
+
"step": 2840
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 3.177257525083612,
|
| 2001 |
+
"grad_norm": 0.29008543491363525,
|
| 2002 |
+
"learning_rate": 6.425848593078643e-05,
|
| 2003 |
+
"loss": 0.0399,
|
| 2004 |
+
"step": 2850
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 3.1884057971014492,
|
| 2008 |
+
"grad_norm": 0.35164210200309753,
|
| 2009 |
+
"learning_rate": 6.357055242081159e-05,
|
| 2010 |
+
"loss": 0.0408,
|
| 2011 |
+
"step": 2860
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 3.1995540691192863,
|
| 2015 |
+
"grad_norm": 0.4294719099998474,
|
| 2016 |
+
"learning_rate": 6.288460105428655e-05,
|
| 2017 |
+
"loss": 0.0408,
|
| 2018 |
+
"step": 2870
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 3.210702341137124,
|
| 2022 |
+
"grad_norm": 0.5600726008415222,
|
| 2023 |
+
"learning_rate": 6.22006691541492e-05,
|
| 2024 |
+
"loss": 0.0446,
|
| 2025 |
+
"step": 2880
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 3.221850613154961,
|
| 2029 |
+
"grad_norm": 0.33129531145095825,
|
| 2030 |
+
"learning_rate": 6.15187939334573e-05,
|
| 2031 |
+
"loss": 0.0478,
|
| 2032 |
+
"step": 2890
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 3.2329988851727984,
|
| 2036 |
+
"grad_norm": 0.5079495906829834,
|
| 2037 |
+
"learning_rate": 6.083901249336369e-05,
|
| 2038 |
+
"loss": 0.0447,
|
| 2039 |
+
"step": 2900
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 3.2441471571906355,
|
| 2043 |
+
"grad_norm": 0.4107360243797302,
|
| 2044 |
+
"learning_rate": 6.016136182109758e-05,
|
| 2045 |
+
"loss": 0.0433,
|
| 2046 |
+
"step": 2910
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 3.2552954292084726,
|
| 2050 |
+
"grad_norm": 0.40619489550590515,
|
| 2051 |
+
"learning_rate": 5.9485878787952174e-05,
|
| 2052 |
+
"loss": 0.0506,
|
| 2053 |
+
"step": 2920
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 3.26644370122631,
|
| 2057 |
+
"grad_norm": 0.6542393565177917,
|
| 2058 |
+
"learning_rate": 5.881260014727831e-05,
|
| 2059 |
+
"loss": 0.0404,
|
| 2060 |
+
"step": 2930
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 3.277591973244147,
|
| 2064 |
+
"grad_norm": 0.3707595765590668,
|
| 2065 |
+
"learning_rate": 5.8141562532484886e-05,
|
| 2066 |
+
"loss": 0.044,
|
| 2067 |
+
"step": 2940
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 3.2887402452619843,
|
| 2071 |
+
"grad_norm": 0.41909733414649963,
|
| 2072 |
+
"learning_rate": 5.747280245504545e-05,
|
| 2073 |
+
"loss": 0.0543,
|
| 2074 |
+
"step": 2950
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 3.299888517279822,
|
| 2078 |
+
"grad_norm": 0.5332537889480591,
|
| 2079 |
+
"learning_rate": 5.6806356302511697e-05,
|
| 2080 |
+
"loss": 0.0429,
|
| 2081 |
+
"step": 2960
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 3.311036789297659,
|
| 2085 |
+
"grad_norm": 0.526335597038269,
|
| 2086 |
+
"learning_rate": 5.614226033653365e-05,
|
| 2087 |
+
"loss": 0.0449,
|
| 2088 |
+
"step": 2970
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 3.322185061315496,
|
| 2092 |
+
"grad_norm": 0.5667539238929749,
|
| 2093 |
+
"learning_rate": 5.5480550690886376e-05,
|
| 2094 |
+
"loss": 0.0362,
|
| 2095 |
+
"step": 2980
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 3.3333333333333335,
|
| 2099 |
+
"grad_norm": 0.5061270594596863,
|
| 2100 |
+
"learning_rate": 5.482126336950433e-05,
|
| 2101 |
+
"loss": 0.0464,
|
| 2102 |
+
"step": 2990
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 3.3444816053511706,
|
| 2106 |
+
"grad_norm": 0.388734370470047,
|
| 2107 |
+
"learning_rate": 5.416443424452202e-05,
|
| 2108 |
+
"loss": 0.0516,
|
| 2109 |
+
"step": 3000
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 3.3556298773690076,
|
| 2113 |
+
"grad_norm": 0.3761380910873413,
|
| 2114 |
+
"learning_rate": 5.3510099054322485e-05,
|
| 2115 |
+
"loss": 0.0381,
|
| 2116 |
+
"step": 3010
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 3.366778149386845,
|
| 2120 |
+
"grad_norm": 0.4708857834339142,
|
| 2121 |
+
"learning_rate": 5.285829340159236e-05,
|
| 2122 |
+
"loss": 0.0429,
|
| 2123 |
+
"step": 3020
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 3.3779264214046822,
|
| 2127 |
+
"grad_norm": 0.31391504406929016,
|
| 2128 |
+
"learning_rate": 5.220905275138516e-05,
|
| 2129 |
+
"loss": 0.0369,
|
| 2130 |
+
"step": 3030
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 3.3890746934225193,
|
| 2134 |
+
"grad_norm": 0.5272077322006226,
|
| 2135 |
+
"learning_rate": 5.156241242919131e-05,
|
| 2136 |
+
"loss": 0.0471,
|
| 2137 |
+
"step": 3040
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 3.400222965440357,
|
| 2141 |
+
"grad_norm": 0.3571004271507263,
|
| 2142 |
+
"learning_rate": 5.091840761901621e-05,
|
| 2143 |
+
"loss": 0.0482,
|
| 2144 |
+
"step": 3050
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 3.411371237458194,
|
| 2148 |
+
"grad_norm": 0.2885373830795288,
|
| 2149 |
+
"learning_rate": 5.0277073361465686e-05,
|
| 2150 |
+
"loss": 0.0422,
|
| 2151 |
+
"step": 3060
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 3.4225195094760315,
|
| 2155 |
+
"grad_norm": 0.33935776352882385,
|
| 2156 |
+
"learning_rate": 4.963844455183968e-05,
|
| 2157 |
+
"loss": 0.0372,
|
| 2158 |
+
"step": 3070
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 3.4336677814938685,
|
| 2162 |
+
"grad_norm": 0.3822285532951355,
|
| 2163 |
+
"learning_rate": 4.900255593823335e-05,
|
| 2164 |
+
"loss": 0.046,
|
| 2165 |
+
"step": 3080
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 3.4448160535117056,
|
| 2169 |
+
"grad_norm": 0.5016531348228455,
|
| 2170 |
+
"learning_rate": 4.836944211964658e-05,
|
| 2171 |
+
"loss": 0.0417,
|
| 2172 |
+
"step": 3090
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 3.4559643255295427,
|
| 2176 |
+
"grad_norm": 0.3562139868736267,
|
| 2177 |
+
"learning_rate": 4.773913754410119e-05,
|
| 2178 |
+
"loss": 0.0464,
|
| 2179 |
+
"step": 3100
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 3.46711259754738,
|
| 2183 |
+
"grad_norm": 0.3429526090621948,
|
| 2184 |
+
"learning_rate": 4.711167650676689e-05,
|
| 2185 |
+
"loss": 0.0378,
|
| 2186 |
+
"step": 3110
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 3.4782608695652173,
|
| 2190 |
+
"grad_norm": 0.4539160430431366,
|
| 2191 |
+
"learning_rate": 4.64870931480951e-05,
|
| 2192 |
+
"loss": 0.0395,
|
| 2193 |
+
"step": 3120
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 3.489409141583055,
|
| 2197 |
+
"grad_norm": 0.3484383523464203,
|
| 2198 |
+
"learning_rate": 4.5865421451961446e-05,
|
| 2199 |
+
"loss": 0.0366,
|
| 2200 |
+
"step": 3130
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 3.500557413600892,
|
| 2204 |
+
"grad_norm": 0.3240243196487427,
|
| 2205 |
+
"learning_rate": 4.524669524381646e-05,
|
| 2206 |
+
"loss": 0.0372,
|
| 2207 |
+
"step": 3140
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 3.511705685618729,
|
| 2211 |
+
"grad_norm": 0.3189137578010559,
|
| 2212 |
+
"learning_rate": 4.463094818884539e-05,
|
| 2213 |
+
"loss": 0.0466,
|
| 2214 |
+
"step": 3150
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 3.522853957636566,
|
| 2218 |
+
"grad_norm": 0.46475180983543396,
|
| 2219 |
+
"learning_rate": 4.4018213790136355e-05,
|
| 2220 |
+
"loss": 0.0414,
|
| 2221 |
+
"step": 3160
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 3.5340022296544036,
|
| 2225 |
+
"grad_norm": 0.2477857619524002,
|
| 2226 |
+
"learning_rate": 4.34085253868574e-05,
|
| 2227 |
+
"loss": 0.0445,
|
| 2228 |
+
"step": 3170
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 3.5451505016722407,
|
| 2232 |
+
"grad_norm": 0.34155797958374023,
|
| 2233 |
+
"learning_rate": 4.280191615244258e-05,
|
| 2234 |
+
"loss": 0.0378,
|
| 2235 |
+
"step": 3180
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 3.556298773690078,
|
| 2239 |
+
"grad_norm": 0.37069597840309143,
|
| 2240 |
+
"learning_rate": 4.219841909278678e-05,
|
| 2241 |
+
"loss": 0.0381,
|
| 2242 |
+
"step": 3190
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 3.5674470457079153,
|
| 2246 |
+
"grad_norm": 0.4122008979320526,
|
| 2247 |
+
"learning_rate": 4.159806704445011e-05,
|
| 2248 |
+
"loss": 0.0467,
|
| 2249 |
+
"step": 3200
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 3.5785953177257523,
|
| 2253 |
+
"grad_norm": 0.3978338837623596,
|
| 2254 |
+
"learning_rate": 4.1000892672871095e-05,
|
| 2255 |
+
"loss": 0.0393,
|
| 2256 |
+
"step": 3210
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 3.58974358974359,
|
| 2260 |
+
"grad_norm": 0.33646389842033386,
|
| 2261 |
+
"learning_rate": 4.040692847058939e-05,
|
| 2262 |
+
"loss": 0.0387,
|
| 2263 |
+
"step": 3220
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 3.600891861761427,
|
| 2267 |
+
"grad_norm": 0.25111469626426697,
|
| 2268 |
+
"learning_rate": 3.9816206755477735e-05,
|
| 2269 |
+
"loss": 0.0358,
|
| 2270 |
+
"step": 3230
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 3.6120401337792645,
|
| 2274 |
+
"grad_norm": 0.3572184145450592,
|
| 2275 |
+
"learning_rate": 3.9228759668983695e-05,
|
| 2276 |
+
"loss": 0.0355,
|
| 2277 |
+
"step": 3240
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 3.6231884057971016,
|
| 2281 |
+
"grad_norm": 0.4244880676269531,
|
| 2282 |
+
"learning_rate": 3.8644619174380724e-05,
|
| 2283 |
+
"loss": 0.0387,
|
| 2284 |
+
"step": 3250
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 3.6343366778149386,
|
| 2288 |
+
"grad_norm": 0.30574020743370056,
|
| 2289 |
+
"learning_rate": 3.806381705502908e-05,
|
| 2290 |
+
"loss": 0.0389,
|
| 2291 |
+
"step": 3260
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 3.6454849498327757,
|
| 2295 |
+
"grad_norm": 0.39554497599601746,
|
| 2296 |
+
"learning_rate": 3.748638491264635e-05,
|
| 2297 |
+
"loss": 0.0485,
|
| 2298 |
+
"step": 3270
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 3.6566332218506132,
|
| 2302 |
+
"grad_norm": 0.48874419927597046,
|
| 2303 |
+
"learning_rate": 3.691235416558817e-05,
|
| 2304 |
+
"loss": 0.0477,
|
| 2305 |
+
"step": 3280
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 3.6677814938684503,
|
| 2309 |
+
"grad_norm": 0.5292513966560364,
|
| 2310 |
+
"learning_rate": 3.634175604713863e-05,
|
| 2311 |
+
"loss": 0.0454,
|
| 2312 |
+
"step": 3290
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 3.678929765886288,
|
| 2316 |
+
"grad_norm": 0.6230178475379944,
|
| 2317 |
+
"learning_rate": 3.57746216038109e-05,
|
| 2318 |
+
"loss": 0.0527,
|
| 2319 |
+
"step": 3300
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 3.690078037904125,
|
| 2323 |
+
"grad_norm": 0.45563289523124695,
|
| 2324 |
+
"learning_rate": 3.5210981693657866e-05,
|
| 2325 |
+
"loss": 0.0362,
|
| 2326 |
+
"step": 3310
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 3.701226309921962,
|
| 2330 |
+
"grad_norm": 0.37886667251586914,
|
| 2331 |
+
"learning_rate": 3.465086698459328e-05,
|
| 2332 |
+
"loss": 0.0355,
|
| 2333 |
+
"step": 3320
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 3.712374581939799,
|
| 2337 |
+
"grad_norm": 0.5162596702575684,
|
| 2338 |
+
"learning_rate": 3.409430795272305e-05,
|
| 2339 |
+
"loss": 0.042,
|
| 2340 |
+
"step": 3330
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 3.7235228539576366,
|
| 2344 |
+
"grad_norm": 0.3282869756221771,
|
| 2345 |
+
"learning_rate": 3.354133488068697e-05,
|
| 2346 |
+
"loss": 0.0414,
|
| 2347 |
+
"step": 3340
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 3.7346711259754737,
|
| 2351 |
+
"grad_norm": 0.2636353075504303,
|
| 2352 |
+
"learning_rate": 3.2991977856011126e-05,
|
| 2353 |
+
"loss": 0.0368,
|
| 2354 |
+
"step": 3350
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 3.745819397993311,
|
| 2358 |
+
"grad_norm": 0.5579274892807007,
|
| 2359 |
+
"learning_rate": 3.244626676947064e-05,
|
| 2360 |
+
"loss": 0.0395,
|
| 2361 |
+
"step": 3360
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 3.7569676700111483,
|
| 2365 |
+
"grad_norm": 0.426093190908432,
|
| 2366 |
+
"learning_rate": 3.1904231313463506e-05,
|
| 2367 |
+
"loss": 0.0545,
|
| 2368 |
+
"step": 3370
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 3.7681159420289854,
|
| 2372 |
+
"grad_norm": 0.5247809886932373,
|
| 2373 |
+
"learning_rate": 3.136590098039495e-05,
|
| 2374 |
+
"loss": 0.0449,
|
| 2375 |
+
"step": 3380
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 3.779264214046823,
|
| 2379 |
+
"grad_norm": 0.31705373525619507,
|
| 2380 |
+
"learning_rate": 3.0831305061072666e-05,
|
| 2381 |
+
"loss": 0.0414,
|
| 2382 |
+
"step": 3390
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 3.79041248606466,
|
| 2386 |
+
"grad_norm": 0.30089136958122253,
|
| 2387 |
+
"learning_rate": 3.0300472643113098e-05,
|
| 2388 |
+
"loss": 0.0372,
|
| 2389 |
+
"step": 3400
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 3.801560758082497,
|
| 2393 |
+
"grad_norm": 0.4322786331176758,
|
| 2394 |
+
"learning_rate": 2.9773432609358886e-05,
|
| 2395 |
+
"loss": 0.0337,
|
| 2396 |
+
"step": 3410
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 3.8127090301003346,
|
| 2400 |
+
"grad_norm": 0.391956627368927,
|
| 2401 |
+
"learning_rate": 2.9250213636307154e-05,
|
| 2402 |
+
"loss": 0.0432,
|
| 2403 |
+
"step": 3420
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 3.8238573021181717,
|
| 2407 |
+
"grad_norm": 0.249397873878479,
|
| 2408 |
+
"learning_rate": 2.873084419254939e-05,
|
| 2409 |
+
"loss": 0.0324,
|
| 2410 |
+
"step": 3430
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 3.8350055741360087,
|
| 2414 |
+
"grad_norm": 0.34695401787757874,
|
| 2415 |
+
"learning_rate": 2.8215352537222373e-05,
|
| 2416 |
+
"loss": 0.0474,
|
| 2417 |
+
"step": 3440
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 3.8461538461538463,
|
| 2421 |
+
"grad_norm": 0.3088880181312561,
|
| 2422 |
+
"learning_rate": 2.7703766718470504e-05,
|
| 2423 |
+
"loss": 0.0385,
|
| 2424 |
+
"step": 3450
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 3.8573021181716833,
|
| 2428 |
+
"grad_norm": 0.2468530386686325,
|
| 2429 |
+
"learning_rate": 2.719611457191985e-05,
|
| 2430 |
+
"loss": 0.0412,
|
| 2431 |
+
"step": 3460
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 3.868450390189521,
|
| 2435 |
+
"grad_norm": 0.37490421533584595,
|
| 2436 |
+
"learning_rate": 2.6692423719163483e-05,
|
| 2437 |
+
"loss": 0.0401,
|
| 2438 |
+
"step": 3470
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 3.879598662207358,
|
| 2442 |
+
"grad_norm": 0.41058897972106934,
|
| 2443 |
+
"learning_rate": 2.619272156625866e-05,
|
| 2444 |
+
"loss": 0.0376,
|
| 2445 |
+
"step": 3480
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 3.890746934225195,
|
| 2449 |
+
"grad_norm": 0.49217766523361206,
|
| 2450 |
+
"learning_rate": 2.5697035302235485e-05,
|
| 2451 |
+
"loss": 0.0421,
|
| 2452 |
+
"step": 3490
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 3.901895206243032,
|
| 2456 |
+
"grad_norm": 0.6945267915725708,
|
| 2457 |
+
"learning_rate": 2.5205391897617738e-05,
|
| 2458 |
+
"loss": 0.0312,
|
| 2459 |
+
"step": 3500
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 3.9130434782608696,
|
| 2463 |
+
"grad_norm": 0.39039209485054016,
|
| 2464 |
+
"learning_rate": 2.471781810295529e-05,
|
| 2465 |
+
"loss": 0.0373,
|
| 2466 |
+
"step": 3510
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 3.9241917502787067,
|
| 2470 |
+
"grad_norm": 0.44833821058273315,
|
| 2471 |
+
"learning_rate": 2.4234340447368642e-05,
|
| 2472 |
+
"loss": 0.0445,
|
| 2473 |
+
"step": 3520
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 3.9353400222965442,
|
| 2477 |
+
"grad_norm": 0.3711733818054199,
|
| 2478 |
+
"learning_rate": 2.3754985237105333e-05,
|
| 2479 |
+
"loss": 0.0407,
|
| 2480 |
+
"step": 3530
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 3.9464882943143813,
|
| 2484 |
+
"grad_norm": 0.7864418625831604,
|
| 2485 |
+
"learning_rate": 2.3279778554108812e-05,
|
| 2486 |
+
"loss": 0.0457,
|
| 2487 |
+
"step": 3540
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 3.9576365663322184,
|
| 2491 |
+
"grad_norm": 0.4766187369823456,
|
| 2492 |
+
"learning_rate": 2.2808746254599144e-05,
|
| 2493 |
+
"loss": 0.0366,
|
| 2494 |
+
"step": 3550
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 3.9687848383500555,
|
| 2498 |
+
"grad_norm": 0.3816820979118347,
|
| 2499 |
+
"learning_rate": 2.2341913967666284e-05,
|
| 2500 |
+
"loss": 0.0391,
|
| 2501 |
+
"step": 3560
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 3.979933110367893,
|
| 2505 |
+
"grad_norm": 0.455503910779953,
|
| 2506 |
+
"learning_rate": 2.187930709387539e-05,
|
| 2507 |
+
"loss": 0.0423,
|
| 2508 |
+
"step": 3570
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 3.99108138238573,
|
| 2512 |
+
"grad_norm": 0.40320706367492676,
|
| 2513 |
+
"learning_rate": 2.142095080388502e-05,
|
| 2514 |
+
"loss": 0.0397,
|
| 2515 |
+
"step": 3580
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 4.002229654403568,
|
| 2519 |
+
"grad_norm": 0.3067777752876282,
|
| 2520 |
+
"learning_rate": 2.0966870037077392e-05,
|
| 2521 |
+
"loss": 0.0356,
|
| 2522 |
+
"step": 3590
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 4.013377926421405,
|
| 2526 |
+
"grad_norm": 0.2318330556154251,
|
| 2527 |
+
"learning_rate": 2.0517089500201548e-05,
|
| 2528 |
+
"loss": 0.0425,
|
| 2529 |
+
"step": 3600
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 4.024526198439242,
|
| 2533 |
+
"grad_norm": 0.34464046359062195,
|
| 2534 |
+
"learning_rate": 2.0071633666028887e-05,
|
| 2535 |
+
"loss": 0.0408,
|
| 2536 |
+
"step": 3610
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 4.035674470457079,
|
| 2540 |
+
"grad_norm": 0.34659889340400696,
|
| 2541 |
+
"learning_rate": 1.9630526772021785e-05,
|
| 2542 |
+
"loss": 0.0381,
|
| 2543 |
+
"step": 3620
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 4.046822742474917,
|
| 2547 |
+
"grad_norm": 0.37813758850097656,
|
| 2548 |
+
"learning_rate": 1.919379281901469e-05,
|
| 2549 |
+
"loss": 0.0374,
|
| 2550 |
+
"step": 3630
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 4.057971014492754,
|
| 2554 |
+
"grad_norm": 0.3929744362831116,
|
| 2555 |
+
"learning_rate": 1.8761455569908314e-05,
|
| 2556 |
+
"loss": 0.0295,
|
| 2557 |
+
"step": 3640
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 4.069119286510591,
|
| 2561 |
+
"grad_norm": 0.36455968022346497,
|
| 2562 |
+
"learning_rate": 1.8333538548376517e-05,
|
| 2563 |
+
"loss": 0.0391,
|
| 2564 |
+
"step": 3650
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 4.080267558528428,
|
| 2568 |
+
"grad_norm": 0.5188841223716736,
|
| 2569 |
+
"learning_rate": 1.791006503758661e-05,
|
| 2570 |
+
"loss": 0.0375,
|
| 2571 |
+
"step": 3660
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 4.091415830546265,
|
| 2575 |
+
"grad_norm": 0.3800041675567627,
|
| 2576 |
+
"learning_rate": 1.749105807893232e-05,
|
| 2577 |
+
"loss": 0.0387,
|
| 2578 |
+
"step": 3670
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 4.102564102564102,
|
| 2582 |
+
"grad_norm": 0.5408110618591309,
|
| 2583 |
+
"learning_rate": 1.7076540470780222e-05,
|
| 2584 |
+
"loss": 0.0368,
|
| 2585 |
+
"step": 3680
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 4.11371237458194,
|
| 2589 |
+
"grad_norm": 0.5009745955467224,
|
| 2590 |
+
"learning_rate": 1.6666534767229135e-05,
|
| 2591 |
+
"loss": 0.0344,
|
| 2592 |
+
"step": 3690
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 4.124860646599777,
|
| 2596 |
+
"grad_norm": 0.44900792837142944,
|
| 2597 |
+
"learning_rate": 1.6261063276883027e-05,
|
| 2598 |
+
"loss": 0.0414,
|
| 2599 |
+
"step": 3700
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 4.136008918617614,
|
| 2603 |
+
"grad_norm": 0.49352550506591797,
|
| 2604 |
+
"learning_rate": 1.5860148061637224e-05,
|
| 2605 |
+
"loss": 0.0378,
|
| 2606 |
+
"step": 3710
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 4.147157190635451,
|
| 2610 |
+
"grad_norm": 0.4194272756576538,
|
| 2611 |
+
"learning_rate": 1.5463810935477973e-05,
|
| 2612 |
+
"loss": 0.0398,
|
| 2613 |
+
"step": 3720
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 4.1583054626532885,
|
| 2617 |
+
"grad_norm": 0.3689301013946533,
|
| 2618 |
+
"learning_rate": 1.5072073463295521e-05,
|
| 2619 |
+
"loss": 0.0366,
|
| 2620 |
+
"step": 3730
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 4.169453734671126,
|
| 2624 |
+
"grad_norm": 0.40687835216522217,
|
| 2625 |
+
"learning_rate": 1.4684956959710739e-05,
|
| 2626 |
+
"loss": 0.0327,
|
| 2627 |
+
"step": 3740
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 4.1806020066889635,
|
| 2631 |
+
"grad_norm": 0.31265124678611755,
|
| 2632 |
+
"learning_rate": 1.4302482487915436e-05,
|
| 2633 |
+
"loss": 0.0359,
|
| 2634 |
+
"step": 3750
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 4.191750278706801,
|
| 2638 |
+
"grad_norm": 0.7661409974098206,
|
| 2639 |
+
"learning_rate": 1.392467085852629e-05,
|
| 2640 |
+
"loss": 0.0357,
|
| 2641 |
+
"step": 3760
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 4.202898550724638,
|
| 2645 |
+
"grad_norm": 0.3778116703033447,
|
| 2646 |
+
"learning_rate": 1.355154262845253e-05,
|
| 2647 |
+
"loss": 0.0324,
|
| 2648 |
+
"step": 3770
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 4.214046822742475,
|
| 2652 |
+
"grad_norm": 0.406654417514801,
|
| 2653 |
+
"learning_rate": 1.3183118099777359e-05,
|
| 2654 |
+
"loss": 0.0342,
|
| 2655 |
+
"step": 3780
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 4.225195094760312,
|
| 2659 |
+
"grad_norm": 0.36551547050476074,
|
| 2660 |
+
"learning_rate": 1.2819417318653427e-05,
|
| 2661 |
+
"loss": 0.0301,
|
| 2662 |
+
"step": 3790
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 4.236343366778149,
|
| 2666 |
+
"grad_norm": 0.3642226755619049,
|
| 2667 |
+
"learning_rate": 1.246046007421201e-05,
|
| 2668 |
+
"loss": 0.0359,
|
| 2669 |
+
"step": 3800
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 4.247491638795987,
|
| 2673 |
+
"grad_norm": 0.3174770772457123,
|
| 2674 |
+
"learning_rate": 1.2106265897486379e-05,
|
| 2675 |
+
"loss": 0.0394,
|
| 2676 |
+
"step": 3810
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 4.258639910813824,
|
| 2680 |
+
"grad_norm": 0.517148494720459,
|
| 2681 |
+
"learning_rate": 1.175685406034891e-05,
|
| 2682 |
+
"loss": 0.0438,
|
| 2683 |
+
"step": 3820
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 4.269788182831661,
|
| 2687 |
+
"grad_norm": 0.7560349106788635,
|
| 2688 |
+
"learning_rate": 1.1412243574462756e-05,
|
| 2689 |
+
"loss": 0.0415,
|
| 2690 |
+
"step": 3830
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 4.280936454849498,
|
| 2694 |
+
"grad_norm": 0.33101922273635864,
|
| 2695 |
+
"learning_rate": 1.1072453190247211e-05,
|
| 2696 |
+
"loss": 0.0409,
|
| 2697 |
+
"step": 3840
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 4.292084726867335,
|
| 2701 |
+
"grad_norm": 0.34227287769317627,
|
| 2702 |
+
"learning_rate": 1.0737501395857574e-05,
|
| 2703 |
+
"loss": 0.0297,
|
| 2704 |
+
"step": 3850
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 4.303232998885173,
|
| 2708 |
+
"grad_norm": 0.44605475664138794,
|
| 2709 |
+
"learning_rate": 1.0407406416179221e-05,
|
| 2710 |
+
"loss": 0.042,
|
| 2711 |
+
"step": 3860
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 4.31438127090301,
|
| 2715 |
+
"grad_norm": 0.44734978675842285,
|
| 2716 |
+
"learning_rate": 1.0082186211835864e-05,
|
| 2717 |
+
"loss": 0.0381,
|
| 2718 |
+
"step": 3870
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 4.325529542920847,
|
| 2722 |
+
"grad_norm": 0.29438066482543945,
|
| 2723 |
+
"learning_rate": 9.76185847821246e-06,
|
| 2724 |
+
"loss": 0.0367,
|
| 2725 |
+
"step": 3880
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 4.336677814938684,
|
| 2729 |
+
"grad_norm": 0.23462656140327454,
|
| 2730 |
+
"learning_rate": 9.446440644492293e-06,
|
| 2731 |
+
"loss": 0.0377,
|
| 2732 |
+
"step": 3890
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 4.3478260869565215,
|
| 2736 |
+
"grad_norm": 0.4501112699508667,
|
| 2737 |
+
"learning_rate": 9.135949872708704e-06,
|
| 2738 |
+
"loss": 0.0416,
|
| 2739 |
+
"step": 3900
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 4.358974358974359,
|
| 2743 |
+
"grad_norm": 0.1577267348766327,
|
| 2744 |
+
"learning_rate": 8.830403056811199e-06,
|
| 2745 |
+
"loss": 0.0281,
|
| 2746 |
+
"step": 3910
|
| 2747 |
+
},
|
| 2748 |
+
{
|
| 2749 |
+
"epoch": 4.3701226309921966,
|
| 2750 |
+
"grad_norm": 0.2782362401485443,
|
| 2751 |
+
"learning_rate": 8.529816821746384e-06,
|
| 2752 |
+
"loss": 0.0338,
|
| 2753 |
+
"step": 3920
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 4.381270903010034,
|
| 2757 |
+
"grad_norm": 0.4171885848045349,
|
| 2758 |
+
"learning_rate": 8.234207522553272e-06,
|
| 2759 |
+
"loss": 0.0393,
|
| 2760 |
+
"step": 3930
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 4.392419175027871,
|
| 2764 |
+
"grad_norm": 0.3518731892108917,
|
| 2765 |
+
"learning_rate": 7.943591243473492e-06,
|
| 2766 |
+
"loss": 0.0382,
|
| 2767 |
+
"step": 3940
|
| 2768 |
+
},
|
| 2769 |
+
{
|
| 2770 |
+
"epoch": 4.403567447045708,
|
| 2771 |
+
"grad_norm": 0.44675251841545105,
|
| 2772 |
+
"learning_rate": 7.657983797076018e-06,
|
| 2773 |
+
"loss": 0.039,
|
| 2774 |
+
"step": 3950
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"epoch": 4.414715719063545,
|
| 2778 |
+
"grad_norm": 0.5793998837471008,
|
| 2779 |
+
"learning_rate": 7.377400723396943e-06,
|
| 2780 |
+
"loss": 0.0473,
|
| 2781 |
+
"step": 3960
|
| 2782 |
+
},
|
| 2783 |
+
{
|
| 2784 |
+
"epoch": 4.425863991081382,
|
| 2785 |
+
"grad_norm": 0.25570735335350037,
|
| 2786 |
+
"learning_rate": 7.101857289093805e-06,
|
| 2787 |
+
"loss": 0.0389,
|
| 2788 |
+
"step": 3970
|
| 2789 |
+
},
|
| 2790 |
+
{
|
| 2791 |
+
"epoch": 4.43701226309922,
|
| 2792 |
+
"grad_norm": 0.35994723439216614,
|
| 2793 |
+
"learning_rate": 6.83136848661503e-06,
|
| 2794 |
+
"loss": 0.0337,
|
| 2795 |
+
"step": 3980
|
| 2796 |
+
},
|
| 2797 |
+
{
|
| 2798 |
+
"epoch": 4.448160535117057,
|
| 2799 |
+
"grad_norm": 0.6161410808563232,
|
| 2800 |
+
"learning_rate": 6.565949033384144e-06,
|
| 2801 |
+
"loss": 0.0324,
|
| 2802 |
+
"step": 3990
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 4.459308807134894,
|
| 2806 |
+
"grad_norm": 0.36720171570777893,
|
| 2807 |
+
"learning_rate": 6.305613370998931e-06,
|
| 2808 |
+
"loss": 0.0368,
|
| 2809 |
+
"step": 4000
|
| 2810 |
+
},
|
| 2811 |
+
{
|
| 2812 |
+
"epoch": 4.470457079152731,
|
| 2813 |
+
"grad_norm": 0.16264738142490387,
|
| 2814 |
+
"learning_rate": 6.0503756644457625e-06,
|
| 2815 |
+
"loss": 0.0334,
|
| 2816 |
+
"step": 4010
|
| 2817 |
+
},
|
| 2818 |
+
{
|
| 2819 |
+
"epoch": 4.481605351170568,
|
| 2820 |
+
"grad_norm": 0.32380685210227966,
|
| 2821 |
+
"learning_rate": 5.800249801328805e-06,
|
| 2822 |
+
"loss": 0.0324,
|
| 2823 |
+
"step": 4020
|
| 2824 |
+
},
|
| 2825 |
+
{
|
| 2826 |
+
"epoch": 4.492753623188406,
|
| 2827 |
+
"grad_norm": 0.39603450894355774,
|
| 2828 |
+
"learning_rate": 5.555249391114448e-06,
|
| 2829 |
+
"loss": 0.0388,
|
| 2830 |
+
"step": 4030
|
| 2831 |
+
},
|
| 2832 |
+
{
|
| 2833 |
+
"epoch": 4.503901895206243,
|
| 2834 |
+
"grad_norm": 0.21810968220233917,
|
| 2835 |
+
"learning_rate": 5.315387764390689e-06,
|
| 2836 |
+
"loss": 0.0396,
|
| 2837 |
+
"step": 4040
|
| 2838 |
+
},
|
| 2839 |
+
{
|
| 2840 |
+
"epoch": 4.51505016722408,
|
| 2841 |
+
"grad_norm": 0.3549976348876953,
|
| 2842 |
+
"learning_rate": 5.080677972141945e-06,
|
| 2843 |
+
"loss": 0.0448,
|
| 2844 |
+
"step": 4050
|
| 2845 |
+
},
|
| 2846 |
+
{
|
| 2847 |
+
"epoch": 4.5261984392419174,
|
| 2848 |
+
"grad_norm": 0.3499640226364136,
|
| 2849 |
+
"learning_rate": 4.851132785038892e-06,
|
| 2850 |
+
"loss": 0.0383,
|
| 2851 |
+
"step": 4060
|
| 2852 |
+
},
|
| 2853 |
+
{
|
| 2854 |
+
"epoch": 4.5373467112597545,
|
| 2855 |
+
"grad_norm": 0.3587091863155365,
|
| 2856 |
+
"learning_rate": 4.6267646927435905e-06,
|
| 2857 |
+
"loss": 0.0485,
|
| 2858 |
+
"step": 4070
|
| 2859 |
+
},
|
| 2860 |
+
{
|
| 2861 |
+
"epoch": 4.548494983277592,
|
| 2862 |
+
"grad_norm": 0.3215045928955078,
|
| 2863 |
+
"learning_rate": 4.407585903229894e-06,
|
| 2864 |
+
"loss": 0.0336,
|
| 2865 |
+
"step": 4080
|
| 2866 |
+
},
|
| 2867 |
+
{
|
| 2868 |
+
"epoch": 4.55964325529543,
|
| 2869 |
+
"grad_norm": 0.32968947291374207,
|
| 2870 |
+
"learning_rate": 4.193608342119271e-06,
|
| 2871 |
+
"loss": 0.0385,
|
| 2872 |
+
"step": 4090
|
| 2873 |
+
},
|
| 2874 |
+
{
|
| 2875 |
+
"epoch": 4.570791527313267,
|
| 2876 |
+
"grad_norm": 0.3530113101005554,
|
| 2877 |
+
"learning_rate": 3.9848436520319e-06,
|
| 2878 |
+
"loss": 0.0382,
|
| 2879 |
+
"step": 4100
|
| 2880 |
+
},
|
| 2881 |
+
{
|
| 2882 |
+
"epoch": 4.581939799331104,
|
| 2883 |
+
"grad_norm": 0.3487902283668518,
|
| 2884 |
+
"learning_rate": 3.7813031919531894e-06,
|
| 2885 |
+
"loss": 0.0332,
|
| 2886 |
+
"step": 4110
|
| 2887 |
+
},
|
| 2888 |
+
{
|
| 2889 |
+
"epoch": 4.593088071348941,
|
| 2890 |
+
"grad_norm": 0.4483624994754791,
|
| 2891 |
+
"learning_rate": 3.582998036615681e-06,
|
| 2892 |
+
"loss": 0.0372,
|
| 2893 |
+
"step": 4120
|
| 2894 |
+
},
|
| 2895 |
+
{
|
| 2896 |
+
"epoch": 4.604236343366778,
|
| 2897 |
+
"grad_norm": 0.20653927326202393,
|
| 2898 |
+
"learning_rate": 3.389938975896567e-06,
|
| 2899 |
+
"loss": 0.0299,
|
| 2900 |
+
"step": 4130
|
| 2901 |
+
},
|
| 2902 |
+
{
|
| 2903 |
+
"epoch": 4.615384615384615,
|
| 2904 |
+
"grad_norm": 0.4962182939052582,
|
| 2905 |
+
"learning_rate": 3.2021365142305136e-06,
|
| 2906 |
+
"loss": 0.0341,
|
| 2907 |
+
"step": 4140
|
| 2908 |
+
},
|
| 2909 |
+
{
|
| 2910 |
+
"epoch": 4.626532887402453,
|
| 2911 |
+
"grad_norm": 0.43579134345054626,
|
| 2912 |
+
"learning_rate": 3.019600870038197e-06,
|
| 2913 |
+
"loss": 0.041,
|
| 2914 |
+
"step": 4150
|
| 2915 |
+
},
|
| 2916 |
+
{
|
| 2917 |
+
"epoch": 4.63768115942029,
|
| 2918 |
+
"grad_norm": 0.3810863792896271,
|
| 2919 |
+
"learning_rate": 2.8423419751702043e-06,
|
| 2920 |
+
"loss": 0.0304,
|
| 2921 |
+
"step": 4160
|
| 2922 |
+
},
|
| 2923 |
+
{
|
| 2924 |
+
"epoch": 4.648829431438127,
|
| 2925 |
+
"grad_norm": 0.34862056374549866,
|
| 2926 |
+
"learning_rate": 2.6703694743667406e-06,
|
| 2927 |
+
"loss": 0.0282,
|
| 2928 |
+
"step": 4170
|
| 2929 |
+
},
|
| 2930 |
+
{
|
| 2931 |
+
"epoch": 4.659977703455964,
|
| 2932 |
+
"grad_norm": 0.2332887500524521,
|
| 2933 |
+
"learning_rate": 2.503692724732809e-06,
|
| 2934 |
+
"loss": 0.0364,
|
| 2935 |
+
"step": 4180
|
| 2936 |
+
},
|
| 2937 |
+
{
|
| 2938 |
+
"epoch": 4.671125975473801,
|
| 2939 |
+
"grad_norm": 0.2178041636943817,
|
| 2940 |
+
"learning_rate": 2.3423207952291027e-06,
|
| 2941 |
+
"loss": 0.0332,
|
| 2942 |
+
"step": 4190
|
| 2943 |
+
},
|
| 2944 |
+
{
|
| 2945 |
+
"epoch": 4.682274247491639,
|
| 2946 |
+
"grad_norm": 0.5643392205238342,
|
| 2947 |
+
"learning_rate": 2.1862624661784903e-06,
|
| 2948 |
+
"loss": 0.0271,
|
| 2949 |
+
"step": 4200
|
| 2950 |
+
},
|
| 2951 |
+
{
|
| 2952 |
+
"epoch": 4.693422519509476,
|
| 2953 |
+
"grad_norm": 0.3044584393501282,
|
| 2954 |
+
"learning_rate": 2.035526228788387e-06,
|
| 2955 |
+
"loss": 0.0333,
|
| 2956 |
+
"step": 4210
|
| 2957 |
+
},
|
| 2958 |
+
{
|
| 2959 |
+
"epoch": 4.704570791527313,
|
| 2960 |
+
"grad_norm": 0.3905907869338989,
|
| 2961 |
+
"learning_rate": 1.8901202846886679e-06,
|
| 2962 |
+
"loss": 0.0366,
|
| 2963 |
+
"step": 4220
|
| 2964 |
+
},
|
| 2965 |
+
{
|
| 2966 |
+
"epoch": 4.7157190635451505,
|
| 2967 |
+
"grad_norm": 0.35689982771873474,
|
| 2968 |
+
"learning_rate": 1.7500525454854366e-06,
|
| 2969 |
+
"loss": 0.0316,
|
| 2970 |
+
"step": 4230
|
| 2971 |
+
},
|
| 2972 |
+
{
|
| 2973 |
+
"epoch": 4.7268673355629875,
|
| 2974 |
+
"grad_norm": 0.29505422711372375,
|
| 2975 |
+
"learning_rate": 1.6153306323305474e-06,
|
| 2976 |
+
"loss": 0.0295,
|
| 2977 |
+
"step": 4240
|
| 2978 |
+
},
|
| 2979 |
+
{
|
| 2980 |
+
"epoch": 4.738015607580825,
|
| 2981 |
+
"grad_norm": 0.8075879812240601,
|
| 2982 |
+
"learning_rate": 1.4859618755069138e-06,
|
| 2983 |
+
"loss": 0.0388,
|
| 2984 |
+
"step": 4250
|
| 2985 |
+
},
|
| 2986 |
+
{
|
| 2987 |
+
"epoch": 4.749163879598662,
|
| 2988 |
+
"grad_norm": 0.5042214393615723,
|
| 2989 |
+
"learning_rate": 1.3619533140296958e-06,
|
| 2990 |
+
"loss": 0.0364,
|
| 2991 |
+
"step": 4260
|
| 2992 |
+
},
|
| 2993 |
+
{
|
| 2994 |
+
"epoch": 4.7603121516165,
|
| 2995 |
+
"grad_norm": 0.39261937141418457,
|
| 2996 |
+
"learning_rate": 1.243311695263316e-06,
|
| 2997 |
+
"loss": 0.0358,
|
| 2998 |
+
"step": 4270
|
| 2999 |
+
},
|
| 3000 |
+
{
|
| 3001 |
+
"epoch": 4.771460423634337,
|
| 3002 |
+
"grad_norm": 0.4254363477230072,
|
| 3003 |
+
"learning_rate": 1.1300434745542877e-06,
|
| 3004 |
+
"loss": 0.0312,
|
| 3005 |
+
"step": 4280
|
| 3006 |
+
},
|
| 3007 |
+
{
|
| 3008 |
+
"epoch": 4.782608695652174,
|
| 3009 |
+
"grad_norm": 0.47518375515937805,
|
| 3010 |
+
"learning_rate": 1.0221548148799943e-06,
|
| 3011 |
+
"loss": 0.0481,
|
| 3012 |
+
"step": 4290
|
| 3013 |
+
},
|
| 3014 |
+
{
|
| 3015 |
+
"epoch": 4.793756967670011,
|
| 3016 |
+
"grad_norm": 0.2959277033805847,
|
| 3017 |
+
"learning_rate": 9.196515865133704e-07,
|
| 3018 |
+
"loss": 0.0292,
|
| 3019 |
+
"step": 4300
|
| 3020 |
+
},
|
| 3021 |
+
{
|
| 3022 |
+
"epoch": 4.804905239687848,
|
| 3023 |
+
"grad_norm": 0.40768375992774963,
|
| 3024 |
+
"learning_rate": 8.225393667034787e-07,
|
| 3025 |
+
"loss": 0.0327,
|
| 3026 |
+
"step": 4310
|
| 3027 |
+
},
|
| 3028 |
+
{
|
| 3029 |
+
"epoch": 4.816053511705686,
|
| 3030 |
+
"grad_norm": 0.7143698334693909,
|
| 3031 |
+
"learning_rate": 7.308234393720859e-07,
|
| 3032 |
+
"loss": 0.037,
|
| 3033 |
+
"step": 4320
|
| 3034 |
+
},
|
| 3035 |
+
{
|
| 3036 |
+
"epoch": 4.827201783723523,
|
| 3037 |
+
"grad_norm": 0.3526574969291687,
|
| 3038 |
+
"learning_rate": 6.445087948261263e-07,
|
| 3039 |
+
"loss": 0.0331,
|
| 3040 |
+
"step": 4330
|
| 3041 |
+
},
|
| 3042 |
+
{
|
| 3043 |
+
"epoch": 4.83835005574136,
|
| 3044 |
+
"grad_norm": 0.2807491421699524,
|
| 3045 |
+
"learning_rate": 5.636001294861415e-07,
|
| 3046 |
+
"loss": 0.0307,
|
| 3047 |
+
"step": 4340
|
| 3048 |
+
},
|
| 3049 |
+
{
|
| 3050 |
+
"epoch": 4.849498327759197,
|
| 3051 |
+
"grad_norm": 0.5166495442390442,
|
| 3052 |
+
"learning_rate": 4.881018456308617e-07,
|
| 3053 |
+
"loss": 0.0352,
|
| 3054 |
+
"step": 4350
|
| 3055 |
+
},
|
| 3056 |
+
{
|
| 3057 |
+
"epoch": 4.860646599777034,
|
| 3058 |
+
"grad_norm": 0.41393277049064636,
|
| 3059 |
+
"learning_rate": 4.1801805115756486e-07,
|
| 3060 |
+
"loss": 0.0361,
|
| 3061 |
+
"step": 4360
|
| 3062 |
+
},
|
| 3063 |
+
{
|
| 3064 |
+
"epoch": 4.871794871794872,
|
| 3065 |
+
"grad_norm": 0.30996108055114746,
|
| 3066 |
+
"learning_rate": 3.533525593586218e-07,
|
| 3067 |
+
"loss": 0.0326,
|
| 3068 |
+
"step": 4370
|
| 3069 |
+
},
|
| 3070 |
+
{
|
| 3071 |
+
"epoch": 4.882943143812709,
|
| 3072 |
+
"grad_norm": 0.27660462260246277,
|
| 3073 |
+
"learning_rate": 2.9410888871396203e-07,
|
| 3074 |
+
"loss": 0.0366,
|
| 3075 |
+
"step": 4380
|
| 3076 |
+
},
|
| 3077 |
+
{
|
| 3078 |
+
"epoch": 4.894091415830546,
|
| 3079 |
+
"grad_norm": 0.29506394267082214,
|
| 3080 |
+
"learning_rate": 2.402902626996939e-07,
|
| 3081 |
+
"loss": 0.0375,
|
| 3082 |
+
"step": 4390
|
| 3083 |
+
},
|
| 3084 |
+
{
|
| 3085 |
+
"epoch": 4.9052396878483835,
|
| 3086 |
+
"grad_norm": 0.34554314613342285,
|
| 3087 |
+
"learning_rate": 1.9189960961268906e-07,
|
| 3088 |
+
"loss": 0.0361,
|
| 3089 |
+
"step": 4400
|
| 3090 |
+
},
|
| 3091 |
+
{
|
| 3092 |
+
"epoch": 4.916387959866221,
|
| 3093 |
+
"grad_norm": 0.45658591389656067,
|
| 3094 |
+
"learning_rate": 1.489395624112433e-07,
|
| 3095 |
+
"loss": 0.0379,
|
| 3096 |
+
"step": 4410
|
| 3097 |
+
},
|
| 3098 |
+
{
|
| 3099 |
+
"epoch": 4.927536231884058,
|
| 3100 |
+
"grad_norm": 0.3536657989025116,
|
| 3101 |
+
"learning_rate": 1.1141245857180238e-07,
|
| 3102 |
+
"loss": 0.0355,
|
| 3103 |
+
"step": 4420
|
| 3104 |
+
},
|
| 3105 |
+
{
|
| 3106 |
+
"epoch": 4.938684503901895,
|
| 3107 |
+
"grad_norm": 0.5148051977157593,
|
| 3108 |
+
"learning_rate": 7.93203399618192e-08,
|
| 3109 |
+
"loss": 0.0323,
|
| 3110 |
+
"step": 4430
|
| 3111 |
+
},
|
| 3112 |
+
{
|
| 3113 |
+
"epoch": 4.949832775919733,
|
| 3114 |
+
"grad_norm": 0.5716732144355774,
|
| 3115 |
+
"learning_rate": 5.266495272862049e-08,
|
| 3116 |
+
"loss": 0.0332,
|
| 3117 |
+
"step": 4440
|
| 3118 |
+
},
|
| 3119 |
+
{
|
| 3120 |
+
"epoch": 4.96098104793757,
|
| 3121 |
+
"grad_norm": 0.2653854787349701,
|
| 3122 |
+
"learning_rate": 3.1447747204427226e-08,
|
| 3123 |
+
"loss": 0.0381,
|
| 3124 |
+
"step": 4450
|
| 3125 |
+
},
|
| 3126 |
+
{
|
| 3127 |
+
"epoch": 4.972129319955407,
|
| 3128 |
+
"grad_norm": 0.47235697507858276,
|
| 3129 |
+
"learning_rate": 1.5669877827395595e-08,
|
| 3130 |
+
"loss": 0.0345,
|
| 3131 |
+
"step": 4460
|
| 3132 |
+
},
|
| 3133 |
+
{
|
| 3134 |
+
"epoch": 4.983277591973244,
|
| 3135 |
+
"grad_norm": 0.35938119888305664,
|
| 3136 |
+
"learning_rate": 5.33220307886717e-09,
|
| 3137 |
+
"loss": 0.0394,
|
| 3138 |
+
"step": 4470
|
| 3139 |
+
},
|
| 3140 |
+
{
|
| 3141 |
+
"epoch": 4.994425863991081,
|
| 3142 |
+
"grad_norm": 0.3504440188407898,
|
| 3143 |
+
"learning_rate": 4.3528543660631417e-10,
|
| 3144 |
+
"loss": 0.0314,
|
| 3145 |
+
"step": 4480
|
| 3146 |
+
},
|
| 3147 |
+
{
|
| 3148 |
+
"epoch": 4.998885172798216,
|
| 3149 |
+
"step": 4484,
|
| 3150 |
+
"total_flos": 1.5541912765028506e+17,
|
| 3151 |
+
"train_loss": 0.06215949216317535,
|
| 3152 |
+
"train_runtime": 2002.6506,
|
| 3153 |
+
"train_samples_per_second": 35.825,
|
| 3154 |
+
"train_steps_per_second": 2.239
|
| 3155 |
+
}
|
| 3156 |
+
],
|
| 3157 |
+
"logging_steps": 10,
|
| 3158 |
+
"max_steps": 4484,
|
| 3159 |
+
"num_input_tokens_seen": 0,
|
| 3160 |
+
"num_train_epochs": 5,
|
| 3161 |
+
"save_steps": 500,
|
| 3162 |
+
"stateful_callbacks": {
|
| 3163 |
+
"TrainerControl": {
|
| 3164 |
+
"args": {
|
| 3165 |
+
"should_epoch_stop": false,
|
| 3166 |
+
"should_evaluate": false,
|
| 3167 |
+
"should_log": false,
|
| 3168 |
+
"should_save": true,
|
| 3169 |
+
"should_training_stop": true
|
| 3170 |
+
},
|
| 3171 |
+
"attributes": {}
|
| 3172 |
+
}
|
| 3173 |
+
},
|
| 3174 |
+
"total_flos": 1.5541912765028506e+17,
|
| 3175 |
+
"train_batch_size": 16,
|
| 3176 |
+
"trial_name": null,
|
| 3177 |
+
"trial_params": null
|
| 3178 |
+
}
|