Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +3647 -0
trainer_state.json
ADDED
@@ -0,0 +1,3647 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 5.0,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 515,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.009708737864077669,
|
13 |
+
"grad_norm": 6.002007543902212,
|
14 |
+
"learning_rate": 1.5384615384615387e-06,
|
15 |
+
"loss": 1.0598,
|
16 |
+
"step": 1
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.019417475728155338,
|
20 |
+
"grad_norm": 5.991740133700607,
|
21 |
+
"learning_rate": 3.0769230769230774e-06,
|
22 |
+
"loss": 1.0579,
|
23 |
+
"step": 2
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.02912621359223301,
|
27 |
+
"grad_norm": 5.825979646199088,
|
28 |
+
"learning_rate": 4.615384615384616e-06,
|
29 |
+
"loss": 1.0544,
|
30 |
+
"step": 3
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.038834951456310676,
|
34 |
+
"grad_norm": 4.364941447544986,
|
35 |
+
"learning_rate": 6.153846153846155e-06,
|
36 |
+
"loss": 1.0106,
|
37 |
+
"step": 4
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.04854368932038835,
|
41 |
+
"grad_norm": 2.7678687055700784,
|
42 |
+
"learning_rate": 7.692307692307694e-06,
|
43 |
+
"loss": 0.9739,
|
44 |
+
"step": 5
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.05825242718446602,
|
48 |
+
"grad_norm": 2.5422026912598827,
|
49 |
+
"learning_rate": 9.230769230769232e-06,
|
50 |
+
"loss": 0.948,
|
51 |
+
"step": 6
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.06796116504854369,
|
55 |
+
"grad_norm": 3.961951944224408,
|
56 |
+
"learning_rate": 1.076923076923077e-05,
|
57 |
+
"loss": 0.9572,
|
58 |
+
"step": 7
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.07766990291262135,
|
62 |
+
"grad_norm": 3.5883110693606577,
|
63 |
+
"learning_rate": 1.230769230769231e-05,
|
64 |
+
"loss": 0.9112,
|
65 |
+
"step": 8
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.08737864077669903,
|
69 |
+
"grad_norm": 3.860767785745949,
|
70 |
+
"learning_rate": 1.3846153846153847e-05,
|
71 |
+
"loss": 0.9007,
|
72 |
+
"step": 9
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.0970873786407767,
|
76 |
+
"grad_norm": 2.7156118144668633,
|
77 |
+
"learning_rate": 1.5384615384615387e-05,
|
78 |
+
"loss": 0.8734,
|
79 |
+
"step": 10
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.10679611650485436,
|
83 |
+
"grad_norm": 1.9289111077886563,
|
84 |
+
"learning_rate": 1.6923076923076924e-05,
|
85 |
+
"loss": 0.8391,
|
86 |
+
"step": 11
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.11650485436893204,
|
90 |
+
"grad_norm": 1.8123247995331513,
|
91 |
+
"learning_rate": 1.8461538461538465e-05,
|
92 |
+
"loss": 0.8171,
|
93 |
+
"step": 12
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.1262135922330097,
|
97 |
+
"grad_norm": 1.428428989967321,
|
98 |
+
"learning_rate": 2e-05,
|
99 |
+
"loss": 0.7956,
|
100 |
+
"step": 13
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.13592233009708737,
|
104 |
+
"grad_norm": 1.3680143343907647,
|
105 |
+
"learning_rate": 2.153846153846154e-05,
|
106 |
+
"loss": 0.7813,
|
107 |
+
"step": 14
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 0.14563106796116504,
|
111 |
+
"grad_norm": 1.3879441764342295,
|
112 |
+
"learning_rate": 2.3076923076923076e-05,
|
113 |
+
"loss": 0.7681,
|
114 |
+
"step": 15
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.1553398058252427,
|
118 |
+
"grad_norm": 1.0796522703025953,
|
119 |
+
"learning_rate": 2.461538461538462e-05,
|
120 |
+
"loss": 0.7574,
|
121 |
+
"step": 16
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 0.1650485436893204,
|
125 |
+
"grad_norm": 1.371942931930126,
|
126 |
+
"learning_rate": 2.6153846153846157e-05,
|
127 |
+
"loss": 0.748,
|
128 |
+
"step": 17
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.17475728155339806,
|
132 |
+
"grad_norm": 1.410433870930946,
|
133 |
+
"learning_rate": 2.7692307692307694e-05,
|
134 |
+
"loss": 0.7432,
|
135 |
+
"step": 18
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 0.18446601941747573,
|
139 |
+
"grad_norm": 1.0096190041268163,
|
140 |
+
"learning_rate": 2.923076923076923e-05,
|
141 |
+
"loss": 0.7329,
|
142 |
+
"step": 19
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 0.1941747572815534,
|
146 |
+
"grad_norm": 1.7265065580167445,
|
147 |
+
"learning_rate": 3.0769230769230774e-05,
|
148 |
+
"loss": 0.7295,
|
149 |
+
"step": 20
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 0.20388349514563106,
|
153 |
+
"grad_norm": 1.1032179992161404,
|
154 |
+
"learning_rate": 3.230769230769231e-05,
|
155 |
+
"loss": 0.7245,
|
156 |
+
"step": 21
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 0.21359223300970873,
|
160 |
+
"grad_norm": 1.0901184558146035,
|
161 |
+
"learning_rate": 3.384615384615385e-05,
|
162 |
+
"loss": 0.7183,
|
163 |
+
"step": 22
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 0.22330097087378642,
|
167 |
+
"grad_norm": 1.7732487418217808,
|
168 |
+
"learning_rate": 3.538461538461539e-05,
|
169 |
+
"loss": 0.7136,
|
170 |
+
"step": 23
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 0.23300970873786409,
|
174 |
+
"grad_norm": 1.430366935608713,
|
175 |
+
"learning_rate": 3.692307692307693e-05,
|
176 |
+
"loss": 0.7105,
|
177 |
+
"step": 24
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 0.24271844660194175,
|
181 |
+
"grad_norm": 1.682581513255261,
|
182 |
+
"learning_rate": 3.846153846153846e-05,
|
183 |
+
"loss": 0.6988,
|
184 |
+
"step": 25
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 0.2524271844660194,
|
188 |
+
"grad_norm": 1.706823703289166,
|
189 |
+
"learning_rate": 4e-05,
|
190 |
+
"loss": 0.6991,
|
191 |
+
"step": 26
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 0.2621359223300971,
|
195 |
+
"grad_norm": 2.014798562695819,
|
196 |
+
"learning_rate": 4.1538461538461544e-05,
|
197 |
+
"loss": 0.6955,
|
198 |
+
"step": 27
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 0.27184466019417475,
|
202 |
+
"grad_norm": 1.4584117807771175,
|
203 |
+
"learning_rate": 4.307692307692308e-05,
|
204 |
+
"loss": 0.6901,
|
205 |
+
"step": 28
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 0.2815533980582524,
|
209 |
+
"grad_norm": 2.5240991865834883,
|
210 |
+
"learning_rate": 4.461538461538462e-05,
|
211 |
+
"loss": 0.6901,
|
212 |
+
"step": 29
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 0.2912621359223301,
|
216 |
+
"grad_norm": 1.7349683843216974,
|
217 |
+
"learning_rate": 4.615384615384615e-05,
|
218 |
+
"loss": 0.6912,
|
219 |
+
"step": 30
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 0.30097087378640774,
|
223 |
+
"grad_norm": 2.453207741393019,
|
224 |
+
"learning_rate": 4.76923076923077e-05,
|
225 |
+
"loss": 0.6837,
|
226 |
+
"step": 31
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 0.3106796116504854,
|
230 |
+
"grad_norm": 2.1891597642473517,
|
231 |
+
"learning_rate": 4.923076923076924e-05,
|
232 |
+
"loss": 0.6834,
|
233 |
+
"step": 32
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 0.32038834951456313,
|
237 |
+
"grad_norm": 2.0789856106085867,
|
238 |
+
"learning_rate": 5.076923076923077e-05,
|
239 |
+
"loss": 0.6854,
|
240 |
+
"step": 33
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 0.3300970873786408,
|
244 |
+
"grad_norm": 1.7347914028228881,
|
245 |
+
"learning_rate": 5.230769230769231e-05,
|
246 |
+
"loss": 0.6734,
|
247 |
+
"step": 34
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 0.33980582524271846,
|
251 |
+
"grad_norm": 2.1284302056196744,
|
252 |
+
"learning_rate": 5.3846153846153853e-05,
|
253 |
+
"loss": 0.6741,
|
254 |
+
"step": 35
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 0.34951456310679613,
|
258 |
+
"grad_norm": 1.9299657090664841,
|
259 |
+
"learning_rate": 5.538461538461539e-05,
|
260 |
+
"loss": 0.6737,
|
261 |
+
"step": 36
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 0.3592233009708738,
|
265 |
+
"grad_norm": 2.1216828159935135,
|
266 |
+
"learning_rate": 5.692307692307693e-05,
|
267 |
+
"loss": 0.6701,
|
268 |
+
"step": 37
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 0.36893203883495146,
|
272 |
+
"grad_norm": 1.2362157926807915,
|
273 |
+
"learning_rate": 5.846153846153846e-05,
|
274 |
+
"loss": 0.6639,
|
275 |
+
"step": 38
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 0.3786407766990291,
|
279 |
+
"grad_norm": 2.2344300721881525,
|
280 |
+
"learning_rate": 6.000000000000001e-05,
|
281 |
+
"loss": 0.666,
|
282 |
+
"step": 39
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 0.3883495145631068,
|
286 |
+
"grad_norm": 1.544248418606736,
|
287 |
+
"learning_rate": 6.153846153846155e-05,
|
288 |
+
"loss": 0.6656,
|
289 |
+
"step": 40
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 0.39805825242718446,
|
293 |
+
"grad_norm": 2.774120239479565,
|
294 |
+
"learning_rate": 6.307692307692308e-05,
|
295 |
+
"loss": 0.6683,
|
296 |
+
"step": 41
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 0.4077669902912621,
|
300 |
+
"grad_norm": 2.1993216366589663,
|
301 |
+
"learning_rate": 6.461538461538462e-05,
|
302 |
+
"loss": 0.6649,
|
303 |
+
"step": 42
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 0.4174757281553398,
|
307 |
+
"grad_norm": 1.8997195417294006,
|
308 |
+
"learning_rate": 6.615384615384616e-05,
|
309 |
+
"loss": 0.6501,
|
310 |
+
"step": 43
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 0.42718446601941745,
|
314 |
+
"grad_norm": 1.9017422216012896,
|
315 |
+
"learning_rate": 6.76923076923077e-05,
|
316 |
+
"loss": 0.6556,
|
317 |
+
"step": 44
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 0.4368932038834951,
|
321 |
+
"grad_norm": 2.2115485405167603,
|
322 |
+
"learning_rate": 6.923076923076924e-05,
|
323 |
+
"loss": 0.6582,
|
324 |
+
"step": 45
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 0.44660194174757284,
|
328 |
+
"grad_norm": 2.2558473718778282,
|
329 |
+
"learning_rate": 7.076923076923078e-05,
|
330 |
+
"loss": 0.6599,
|
331 |
+
"step": 46
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 0.4563106796116505,
|
335 |
+
"grad_norm": 1.8236441414000624,
|
336 |
+
"learning_rate": 7.230769230769232e-05,
|
337 |
+
"loss": 0.6502,
|
338 |
+
"step": 47
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 0.46601941747572817,
|
342 |
+
"grad_norm": 1.668233835252892,
|
343 |
+
"learning_rate": 7.384615384615386e-05,
|
344 |
+
"loss": 0.6525,
|
345 |
+
"step": 48
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 0.47572815533980584,
|
349 |
+
"grad_norm": 3.1753217004320975,
|
350 |
+
"learning_rate": 7.538461538461539e-05,
|
351 |
+
"loss": 0.6627,
|
352 |
+
"step": 49
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 0.4854368932038835,
|
356 |
+
"grad_norm": 1.621218570570416,
|
357 |
+
"learning_rate": 7.692307692307693e-05,
|
358 |
+
"loss": 0.6484,
|
359 |
+
"step": 50
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 0.49514563106796117,
|
363 |
+
"grad_norm": 4.339926226583821,
|
364 |
+
"learning_rate": 7.846153846153847e-05,
|
365 |
+
"loss": 0.6752,
|
366 |
+
"step": 51
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 0.5048543689320388,
|
370 |
+
"grad_norm": 3.113087384499717,
|
371 |
+
"learning_rate": 8e-05,
|
372 |
+
"loss": 0.6722,
|
373 |
+
"step": 52
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 0.5145631067961165,
|
377 |
+
"grad_norm": 2.714482442820161,
|
378 |
+
"learning_rate": 7.999907919834168e-05,
|
379 |
+
"loss": 0.6628,
|
380 |
+
"step": 53
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 0.5242718446601942,
|
384 |
+
"grad_norm": 2.914244429607087,
|
385 |
+
"learning_rate": 7.999631683576055e-05,
|
386 |
+
"loss": 0.6581,
|
387 |
+
"step": 54
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 0.5339805825242718,
|
391 |
+
"grad_norm": 1.884493343669805,
|
392 |
+
"learning_rate": 7.999171303943594e-05,
|
393 |
+
"loss": 0.6497,
|
394 |
+
"step": 55
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 0.5436893203883495,
|
398 |
+
"grad_norm": 3.1324813256148873,
|
399 |
+
"learning_rate": 7.998526802132707e-05,
|
400 |
+
"loss": 0.6539,
|
401 |
+
"step": 56
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 0.5533980582524272,
|
405 |
+
"grad_norm": 1.9123959562537083,
|
406 |
+
"learning_rate": 7.997698207816309e-05,
|
407 |
+
"loss": 0.6491,
|
408 |
+
"step": 57
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 0.5631067961165048,
|
412 |
+
"grad_norm": 2.223063366245145,
|
413 |
+
"learning_rate": 7.99668555914295e-05,
|
414 |
+
"loss": 0.6515,
|
415 |
+
"step": 58
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 0.5728155339805825,
|
419 |
+
"grad_norm": 2.268558088911201,
|
420 |
+
"learning_rate": 7.995488902735063e-05,
|
421 |
+
"loss": 0.6573,
|
422 |
+
"step": 59
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 0.5825242718446602,
|
426 |
+
"grad_norm": 1.5722495721856204,
|
427 |
+
"learning_rate": 7.994108293686804e-05,
|
428 |
+
"loss": 0.6441,
|
429 |
+
"step": 60
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 0.5922330097087378,
|
433 |
+
"grad_norm": 1.666894724940502,
|
434 |
+
"learning_rate": 7.992543795561527e-05,
|
435 |
+
"loss": 0.6408,
|
436 |
+
"step": 61
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"epoch": 0.6019417475728155,
|
440 |
+
"grad_norm": 1.0031568832099003,
|
441 |
+
"learning_rate": 7.990795480388861e-05,
|
442 |
+
"loss": 0.6342,
|
443 |
+
"step": 62
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 0.6116504854368932,
|
447 |
+
"grad_norm": 2.2162360491704067,
|
448 |
+
"learning_rate": 7.988863428661377e-05,
|
449 |
+
"loss": 0.6589,
|
450 |
+
"step": 63
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 0.6213592233009708,
|
454 |
+
"grad_norm": 1.7496212177283834,
|
455 |
+
"learning_rate": 7.9867477293309e-05,
|
456 |
+
"loss": 0.6484,
|
457 |
+
"step": 64
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 0.6310679611650486,
|
461 |
+
"grad_norm": 1.5169651851287749,
|
462 |
+
"learning_rate": 7.984448479804398e-05,
|
463 |
+
"loss": 0.6394,
|
464 |
+
"step": 65
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 0.6407766990291263,
|
468 |
+
"grad_norm": 1.9120046573933336,
|
469 |
+
"learning_rate": 7.981965785939515e-05,
|
470 |
+
"loss": 0.635,
|
471 |
+
"step": 66
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 0.6504854368932039,
|
475 |
+
"grad_norm": 1.7557422251639143,
|
476 |
+
"learning_rate": 7.97929976203968e-05,
|
477 |
+
"loss": 0.647,
|
478 |
+
"step": 67
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 0.6601941747572816,
|
482 |
+
"grad_norm": 1.3556049828157026,
|
483 |
+
"learning_rate": 7.976450530848851e-05,
|
484 |
+
"loss": 0.6374,
|
485 |
+
"step": 68
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 0.6699029126213593,
|
489 |
+
"grad_norm": 2.244626845581171,
|
490 |
+
"learning_rate": 7.973418223545874e-05,
|
491 |
+
"loss": 0.6334,
|
492 |
+
"step": 69
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 0.6796116504854369,
|
496 |
+
"grad_norm": 1.3591820052552963,
|
497 |
+
"learning_rate": 7.970202979738426e-05,
|
498 |
+
"loss": 0.6376,
|
499 |
+
"step": 70
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 0.6893203883495146,
|
503 |
+
"grad_norm": 1.5805931826326196,
|
504 |
+
"learning_rate": 7.966804947456599e-05,
|
505 |
+
"loss": 0.632,
|
506 |
+
"step": 71
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 0.6990291262135923,
|
510 |
+
"grad_norm": 1.334661947485613,
|
511 |
+
"learning_rate": 7.96322428314608e-05,
|
512 |
+
"loss": 0.6269,
|
513 |
+
"step": 72
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 0.7087378640776699,
|
517 |
+
"grad_norm": 1.8903308561526113,
|
518 |
+
"learning_rate": 7.959461151660952e-05,
|
519 |
+
"loss": 0.6342,
|
520 |
+
"step": 73
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 0.7184466019417476,
|
524 |
+
"grad_norm": 1.1038829068258371,
|
525 |
+
"learning_rate": 7.955515726256101e-05,
|
526 |
+
"loss": 0.6275,
|
527 |
+
"step": 74
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 0.7281553398058253,
|
531 |
+
"grad_norm": 1.5525916361050445,
|
532 |
+
"learning_rate": 7.951388188579237e-05,
|
533 |
+
"loss": 0.6343,
|
534 |
+
"step": 75
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 0.7378640776699029,
|
538 |
+
"grad_norm": 1.5390664333547033,
|
539 |
+
"learning_rate": 7.94707872866254e-05,
|
540 |
+
"loss": 0.6315,
|
541 |
+
"step": 76
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 0.7475728155339806,
|
545 |
+
"grad_norm": 1.769589829850563,
|
546 |
+
"learning_rate": 7.942587544913901e-05,
|
547 |
+
"loss": 0.6329,
|
548 |
+
"step": 77
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 0.7572815533980582,
|
552 |
+
"grad_norm": 1.4148162374660378,
|
553 |
+
"learning_rate": 7.937914844107791e-05,
|
554 |
+
"loss": 0.6297,
|
555 |
+
"step": 78
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 0.7669902912621359,
|
559 |
+
"grad_norm": 1.7973463567105696,
|
560 |
+
"learning_rate": 7.933060841375745e-05,
|
561 |
+
"loss": 0.627,
|
562 |
+
"step": 79
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 0.7766990291262136,
|
566 |
+
"grad_norm": 1.184326171503996,
|
567 |
+
"learning_rate": 7.928025760196447e-05,
|
568 |
+
"loss": 0.6234,
|
569 |
+
"step": 80
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 0.7864077669902912,
|
573 |
+
"grad_norm": 1.6486867376699348,
|
574 |
+
"learning_rate": 7.922809832385456e-05,
|
575 |
+
"loss": 0.6224,
|
576 |
+
"step": 81
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 0.7961165048543689,
|
580 |
+
"grad_norm": 1.2312763534767475,
|
581 |
+
"learning_rate": 7.917413298084519e-05,
|
582 |
+
"loss": 0.6207,
|
583 |
+
"step": 82
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 0.8058252427184466,
|
587 |
+
"grad_norm": 1.319948768871293,
|
588 |
+
"learning_rate": 7.911836405750525e-05,
|
589 |
+
"loss": 0.618,
|
590 |
+
"step": 83
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 0.8155339805825242,
|
594 |
+
"grad_norm": 1.2459374631380746,
|
595 |
+
"learning_rate": 7.906079412144055e-05,
|
596 |
+
"loss": 0.6215,
|
597 |
+
"step": 84
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 0.8252427184466019,
|
601 |
+
"grad_norm": 1.2798874980653692,
|
602 |
+
"learning_rate": 7.900142582317576e-05,
|
603 |
+
"loss": 0.6172,
|
604 |
+
"step": 85
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 0.8349514563106796,
|
608 |
+
"grad_norm": 2.093389549775017,
|
609 |
+
"learning_rate": 7.894026189603225e-05,
|
610 |
+
"loss": 0.618,
|
611 |
+
"step": 86
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 0.8446601941747572,
|
615 |
+
"grad_norm": 1.1300093559740532,
|
616 |
+
"learning_rate": 7.887730515600227e-05,
|
617 |
+
"loss": 0.6173,
|
618 |
+
"step": 87
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 0.8543689320388349,
|
622 |
+
"grad_norm": 2.3541947389099094,
|
623 |
+
"learning_rate": 7.881255850161939e-05,
|
624 |
+
"loss": 0.6243,
|
625 |
+
"step": 88
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 0.8640776699029126,
|
629 |
+
"grad_norm": 1.5613220102186438,
|
630 |
+
"learning_rate": 7.87460249138249e-05,
|
631 |
+
"loss": 0.623,
|
632 |
+
"step": 89
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 0.8737864077669902,
|
636 |
+
"grad_norm": 1.9104766159006328,
|
637 |
+
"learning_rate": 7.867770745583074e-05,
|
638 |
+
"loss": 0.6241,
|
639 |
+
"step": 90
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 0.883495145631068,
|
643 |
+
"grad_norm": 1.419392875826999,
|
644 |
+
"learning_rate": 7.860760927297833e-05,
|
645 |
+
"loss": 0.62,
|
646 |
+
"step": 91
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 0.8932038834951457,
|
650 |
+
"grad_norm": 1.3282640676407322,
|
651 |
+
"learning_rate": 7.853573359259381e-05,
|
652 |
+
"loss": 0.6166,
|
653 |
+
"step": 92
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 0.9029126213592233,
|
657 |
+
"grad_norm": 1.3250464130522686,
|
658 |
+
"learning_rate": 7.846208372383947e-05,
|
659 |
+
"loss": 0.6179,
|
660 |
+
"step": 93
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 0.912621359223301,
|
664 |
+
"grad_norm": 1.1377495249178353,
|
665 |
+
"learning_rate": 7.838666305756138e-05,
|
666 |
+
"loss": 0.6122,
|
667 |
+
"step": 94
|
668 |
+
},
|
669 |
+
{
|
670 |
+
"epoch": 0.9223300970873787,
|
671 |
+
"grad_norm": 1.4646244528960923,
|
672 |
+
"learning_rate": 7.830947506613324e-05,
|
673 |
+
"loss": 0.6105,
|
674 |
+
"step": 95
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"epoch": 0.9320388349514563,
|
678 |
+
"grad_norm": 1.142509028160319,
|
679 |
+
"learning_rate": 7.823052330329663e-05,
|
680 |
+
"loss": 0.611,
|
681 |
+
"step": 96
|
682 |
+
},
|
683 |
+
{
|
684 |
+
"epoch": 0.941747572815534,
|
685 |
+
"grad_norm": 2.0274152032009494,
|
686 |
+
"learning_rate": 7.81498114039972e-05,
|
687 |
+
"loss": 0.616,
|
688 |
+
"step": 97
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"epoch": 0.9514563106796117,
|
692 |
+
"grad_norm": 1.0686288368525192,
|
693 |
+
"learning_rate": 7.806734308421753e-05,
|
694 |
+
"loss": 0.6101,
|
695 |
+
"step": 98
|
696 |
+
},
|
697 |
+
{
|
698 |
+
"epoch": 0.9611650485436893,
|
699 |
+
"grad_norm": 2.0849531670929626,
|
700 |
+
"learning_rate": 7.798312214080588e-05,
|
701 |
+
"loss": 0.6128,
|
702 |
+
"step": 99
|
703 |
+
},
|
704 |
+
{
|
705 |
+
"epoch": 0.970873786407767,
|
706 |
+
"grad_norm": 1.723568607059033,
|
707 |
+
"learning_rate": 7.789715245130148e-05,
|
708 |
+
"loss": 0.6156,
|
709 |
+
"step": 100
|
710 |
+
},
|
711 |
+
{
|
712 |
+
"epoch": 0.9805825242718447,
|
713 |
+
"grad_norm": 1.0991520002736015,
|
714 |
+
"learning_rate": 7.780943797375594e-05,
|
715 |
+
"loss": 0.6028,
|
716 |
+
"step": 101
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"epoch": 0.9902912621359223,
|
720 |
+
"grad_norm": 1.81475046862089,
|
721 |
+
"learning_rate": 7.77199827465511e-05,
|
722 |
+
"loss": 0.6176,
|
723 |
+
"step": 102
|
724 |
+
},
|
725 |
+
{
|
726 |
+
"epoch": 1.0,
|
727 |
+
"grad_norm": 1.0995121126222103,
|
728 |
+
"learning_rate": 7.762879088821302e-05,
|
729 |
+
"loss": 0.612,
|
730 |
+
"step": 103
|
731 |
+
},
|
732 |
+
{
|
733 |
+
"epoch": 1.0097087378640777,
|
734 |
+
"grad_norm": 1.1739346884276352,
|
735 |
+
"learning_rate": 7.753586659722243e-05,
|
736 |
+
"loss": 0.5928,
|
737 |
+
"step": 104
|
738 |
+
},
|
739 |
+
{
|
740 |
+
"epoch": 1.0194174757281553,
|
741 |
+
"grad_norm": 1.475480407854494,
|
742 |
+
"learning_rate": 7.74412141518214e-05,
|
743 |
+
"loss": 0.5989,
|
744 |
+
"step": 105
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"epoch": 1.029126213592233,
|
748 |
+
"grad_norm": 1.3719226999649305,
|
749 |
+
"learning_rate": 7.734483790981636e-05,
|
750 |
+
"loss": 0.5969,
|
751 |
+
"step": 106
|
752 |
+
},
|
753 |
+
{
|
754 |
+
"epoch": 1.0388349514563107,
|
755 |
+
"grad_norm": 1.0726737236370019,
|
756 |
+
"learning_rate": 7.724674230837747e-05,
|
757 |
+
"loss": 0.5825,
|
758 |
+
"step": 107
|
759 |
+
},
|
760 |
+
{
|
761 |
+
"epoch": 1.0485436893203883,
|
762 |
+
"grad_norm": 1.3362590355075374,
|
763 |
+
"learning_rate": 7.714693186383437e-05,
|
764 |
+
"loss": 0.594,
|
765 |
+
"step": 108
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"epoch": 1.058252427184466,
|
769 |
+
"grad_norm": 0.820555469002108,
|
770 |
+
"learning_rate": 7.704541117146819e-05,
|
771 |
+
"loss": 0.5874,
|
772 |
+
"step": 109
|
773 |
+
},
|
774 |
+
{
|
775 |
+
"epoch": 1.0679611650485437,
|
776 |
+
"grad_norm": 1.3323408029024517,
|
777 |
+
"learning_rate": 7.694218490530004e-05,
|
778 |
+
"loss": 0.5886,
|
779 |
+
"step": 110
|
780 |
+
},
|
781 |
+
{
|
782 |
+
"epoch": 1.0776699029126213,
|
783 |
+
"grad_norm": 1.066797450314409,
|
784 |
+
"learning_rate": 7.683725781787574e-05,
|
785 |
+
"loss": 0.5876,
|
786 |
+
"step": 111
|
787 |
+
},
|
788 |
+
{
|
789 |
+
"epoch": 1.087378640776699,
|
790 |
+
"grad_norm": 1.245072674676843,
|
791 |
+
"learning_rate": 7.673063474004715e-05,
|
792 |
+
"loss": 0.5879,
|
793 |
+
"step": 112
|
794 |
+
},
|
795 |
+
{
|
796 |
+
"epoch": 1.0970873786407767,
|
797 |
+
"grad_norm": 1.2819802358790717,
|
798 |
+
"learning_rate": 7.662232058074957e-05,
|
799 |
+
"loss": 0.5864,
|
800 |
+
"step": 113
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 1.1067961165048543,
|
804 |
+
"grad_norm": 1.436129834544921,
|
805 |
+
"learning_rate": 7.651232032677588e-05,
|
806 |
+
"loss": 0.5919,
|
807 |
+
"step": 114
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 1.116504854368932,
|
811 |
+
"grad_norm": 0.8175594211587869,
|
812 |
+
"learning_rate": 7.640063904254691e-05,
|
813 |
+
"loss": 0.5817,
|
814 |
+
"step": 115
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"epoch": 1.1262135922330097,
|
818 |
+
"grad_norm": 1.687072132305568,
|
819 |
+
"learning_rate": 7.628728186987824e-05,
|
820 |
+
"loss": 0.5848,
|
821 |
+
"step": 116
|
822 |
+
},
|
823 |
+
{
|
824 |
+
"epoch": 1.1359223300970873,
|
825 |
+
"grad_norm": 1.1326446741047016,
|
826 |
+
"learning_rate": 7.617225402774348e-05,
|
827 |
+
"loss": 0.5858,
|
828 |
+
"step": 117
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 1.145631067961165,
|
832 |
+
"grad_norm": 0.82496802522063,
|
833 |
+
"learning_rate": 7.605556081203405e-05,
|
834 |
+
"loss": 0.5812,
|
835 |
+
"step": 118
|
836 |
+
},
|
837 |
+
{
|
838 |
+
"epoch": 1.1553398058252426,
|
839 |
+
"grad_norm": 1.250641283128446,
|
840 |
+
"learning_rate": 7.593720759531526e-05,
|
841 |
+
"loss": 0.5876,
|
842 |
+
"step": 119
|
843 |
+
},
|
844 |
+
{
|
845 |
+
"epoch": 1.1650485436893203,
|
846 |
+
"grad_norm": 1.1921582278913663,
|
847 |
+
"learning_rate": 7.581719982657903e-05,
|
848 |
+
"loss": 0.5842,
|
849 |
+
"step": 120
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 1.174757281553398,
|
853 |
+
"grad_norm": 1.2734225443412055,
|
854 |
+
"learning_rate": 7.569554303099296e-05,
|
855 |
+
"loss": 0.579,
|
856 |
+
"step": 121
|
857 |
+
},
|
858 |
+
{
|
859 |
+
"epoch": 1.1844660194174756,
|
860 |
+
"grad_norm": 1.1853318091968652,
|
861 |
+
"learning_rate": 7.557224280964603e-05,
|
862 |
+
"loss": 0.582,
|
863 |
+
"step": 122
|
864 |
+
},
|
865 |
+
{
|
866 |
+
"epoch": 1.1941747572815533,
|
867 |
+
"grad_norm": 1.152376249051554,
|
868 |
+
"learning_rate": 7.544730483929065e-05,
|
869 |
+
"loss": 0.5836,
|
870 |
+
"step": 123
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"epoch": 1.203883495145631,
|
874 |
+
"grad_norm": 1.3041878863702414,
|
875 |
+
"learning_rate": 7.532073487208132e-05,
|
876 |
+
"loss": 0.5761,
|
877 |
+
"step": 124
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"epoch": 1.2135922330097086,
|
881 |
+
"grad_norm": 1.3112880078778733,
|
882 |
+
"learning_rate": 7.519253873530986e-05,
|
883 |
+
"loss": 0.5836,
|
884 |
+
"step": 125
|
885 |
+
},
|
886 |
+
{
|
887 |
+
"epoch": 1.2233009708737863,
|
888 |
+
"grad_norm": 1.3474986040772934,
|
889 |
+
"learning_rate": 7.5062722331137e-05,
|
890 |
+
"loss": 0.5801,
|
891 |
+
"step": 126
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 1.233009708737864,
|
895 |
+
"grad_norm": 0.8129513286807843,
|
896 |
+
"learning_rate": 7.493129163632076e-05,
|
897 |
+
"loss": 0.5748,
|
898 |
+
"step": 127
|
899 |
+
},
|
900 |
+
{
|
901 |
+
"epoch": 1.2427184466019416,
|
902 |
+
"grad_norm": 1.1010576487307613,
|
903 |
+
"learning_rate": 7.479825270194124e-05,
|
904 |
+
"loss": 0.5842,
|
905 |
+
"step": 128
|
906 |
+
},
|
907 |
+
{
|
908 |
+
"epoch": 1.2524271844660193,
|
909 |
+
"grad_norm": 1.398672981702769,
|
910 |
+
"learning_rate": 7.466361165312199e-05,
|
911 |
+
"loss": 0.5877,
|
912 |
+
"step": 129
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"epoch": 1.262135922330097,
|
916 |
+
"grad_norm": 1.2164255288325172,
|
917 |
+
"learning_rate": 7.452737468874809e-05,
|
918 |
+
"loss": 0.5826,
|
919 |
+
"step": 130
|
920 |
+
},
|
921 |
+
{
|
922 |
+
"epoch": 1.2718446601941746,
|
923 |
+
"grad_norm": 0.7866460429939296,
|
924 |
+
"learning_rate": 7.438954808118064e-05,
|
925 |
+
"loss": 0.5778,
|
926 |
+
"step": 131
|
927 |
+
},
|
928 |
+
{
|
929 |
+
"epoch": 1.2815533980582523,
|
930 |
+
"grad_norm": 1.2501923242149588,
|
931 |
+
"learning_rate": 7.425013817596812e-05,
|
932 |
+
"loss": 0.5797,
|
933 |
+
"step": 132
|
934 |
+
},
|
935 |
+
{
|
936 |
+
"epoch": 1.29126213592233,
|
937 |
+
"grad_norm": 1.3225455051065684,
|
938 |
+
"learning_rate": 7.41091513915541e-05,
|
939 |
+
"loss": 0.578,
|
940 |
+
"step": 133
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"epoch": 1.3009708737864076,
|
944 |
+
"grad_norm": 1.0105911545481185,
|
945 |
+
"learning_rate": 7.396659421898183e-05,
|
946 |
+
"loss": 0.5754,
|
947 |
+
"step": 134
|
948 |
+
},
|
949 |
+
{
|
950 |
+
"epoch": 1.3106796116504853,
|
951 |
+
"grad_norm": 1.2271731398121772,
|
952 |
+
"learning_rate": 7.382247322159534e-05,
|
953 |
+
"loss": 0.5808,
|
954 |
+
"step": 135
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 1.3203883495145632,
|
958 |
+
"grad_norm": 1.1215866709600937,
|
959 |
+
"learning_rate": 7.367679503473732e-05,
|
960 |
+
"loss": 0.5738,
|
961 |
+
"step": 136
|
962 |
+
},
|
963 |
+
{
|
964 |
+
"epoch": 1.3300970873786409,
|
965 |
+
"grad_norm": 1.0715712124069652,
|
966 |
+
"learning_rate": 7.352956636544358e-05,
|
967 |
+
"loss": 0.5754,
|
968 |
+
"step": 137
|
969 |
+
},
|
970 |
+
{
|
971 |
+
"epoch": 1.3398058252427185,
|
972 |
+
"grad_norm": 0.9077358541837017,
|
973 |
+
"learning_rate": 7.338079399213424e-05,
|
974 |
+
"loss": 0.5711,
|
975 |
+
"step": 138
|
976 |
+
},
|
977 |
+
{
|
978 |
+
"epoch": 1.3495145631067962,
|
979 |
+
"grad_norm": 1.064355488383383,
|
980 |
+
"learning_rate": 7.32304847643017e-05,
|
981 |
+
"loss": 0.5698,
|
982 |
+
"step": 139
|
983 |
+
},
|
984 |
+
{
|
985 |
+
"epoch": 1.3592233009708738,
|
986 |
+
"grad_norm": 0.9384952345701946,
|
987 |
+
"learning_rate": 7.30786456021953e-05,
|
988 |
+
"loss": 0.5739,
|
989 |
+
"step": 140
|
990 |
+
},
|
991 |
+
{
|
992 |
+
"epoch": 1.3689320388349515,
|
993 |
+
"grad_norm": 1.1821144606047003,
|
994 |
+
"learning_rate": 7.292528349650262e-05,
|
995 |
+
"loss": 0.5701,
|
996 |
+
"step": 141
|
997 |
+
},
|
998 |
+
{
|
999 |
+
"epoch": 1.3786407766990292,
|
1000 |
+
"grad_norm": 0.97575327357092,
|
1001 |
+
"learning_rate": 7.277040550802776e-05,
|
1002 |
+
"loss": 0.573,
|
1003 |
+
"step": 142
|
1004 |
+
},
|
1005 |
+
{
|
1006 |
+
"epoch": 1.3883495145631068,
|
1007 |
+
"grad_norm": 1.1580935526594842,
|
1008 |
+
"learning_rate": 7.261401876736611e-05,
|
1009 |
+
"loss": 0.5751,
|
1010 |
+
"step": 143
|
1011 |
+
},
|
1012 |
+
{
|
1013 |
+
"epoch": 1.3980582524271845,
|
1014 |
+
"grad_norm": 0.9292779139246362,
|
1015 |
+
"learning_rate": 7.245613047457621e-05,
|
1016 |
+
"loss": 0.5718,
|
1017 |
+
"step": 144
|
1018 |
+
},
|
1019 |
+
{
|
1020 |
+
"epoch": 1.4077669902912622,
|
1021 |
+
"grad_norm": 1.2215840715546447,
|
1022 |
+
"learning_rate": 7.229674789884813e-05,
|
1023 |
+
"loss": 0.5749,
|
1024 |
+
"step": 145
|
1025 |
+
},
|
1026 |
+
{
|
1027 |
+
"epoch": 1.4174757281553398,
|
1028 |
+
"grad_norm": 0.9138289618139581,
|
1029 |
+
"learning_rate": 7.213587837816889e-05,
|
1030 |
+
"loss": 0.5696,
|
1031 |
+
"step": 146
|
1032 |
+
},
|
1033 |
+
{
|
1034 |
+
"epoch": 1.4271844660194175,
|
1035 |
+
"grad_norm": 1.0533170144837214,
|
1036 |
+
"learning_rate": 7.197352931898454e-05,
|
1037 |
+
"loss": 0.5726,
|
1038 |
+
"step": 147
|
1039 |
+
},
|
1040 |
+
{
|
1041 |
+
"epoch": 1.4368932038834952,
|
1042 |
+
"grad_norm": 1.1454322229077762,
|
1043 |
+
"learning_rate": 7.180970819585923e-05,
|
1044 |
+
"loss": 0.5712,
|
1045 |
+
"step": 148
|
1046 |
+
},
|
1047 |
+
{
|
1048 |
+
"epoch": 1.4466019417475728,
|
1049 |
+
"grad_norm": 1.0300828181718542,
|
1050 |
+
"learning_rate": 7.164442255113107e-05,
|
1051 |
+
"loss": 0.5717,
|
1052 |
+
"step": 149
|
1053 |
+
},
|
1054 |
+
{
|
1055 |
+
"epoch": 1.4563106796116505,
|
1056 |
+
"grad_norm": 1.077162592235673,
|
1057 |
+
"learning_rate": 7.147767999456484e-05,
|
1058 |
+
"loss": 0.5785,
|
1059 |
+
"step": 150
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 1.4660194174757282,
|
1063 |
+
"grad_norm": 1.2410959167210567,
|
1064 |
+
"learning_rate": 7.130948820300166e-05,
|
1065 |
+
"loss": 0.5771,
|
1066 |
+
"step": 151
|
1067 |
+
},
|
1068 |
+
{
|
1069 |
+
"epoch": 1.4757281553398058,
|
1070 |
+
"grad_norm": 1.2329275408989433,
|
1071 |
+
"learning_rate": 7.113985492000558e-05,
|
1072 |
+
"loss": 0.5683,
|
1073 |
+
"step": 152
|
1074 |
+
},
|
1075 |
+
{
|
1076 |
+
"epoch": 1.4854368932038835,
|
1077 |
+
"grad_norm": 1.0776714897294795,
|
1078 |
+
"learning_rate": 7.0968787955507e-05,
|
1079 |
+
"loss": 0.5716,
|
1080 |
+
"step": 153
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"epoch": 1.4951456310679612,
|
1084 |
+
"grad_norm": 0.8984266405483924,
|
1085 |
+
"learning_rate": 7.079629518544312e-05,
|
1086 |
+
"loss": 0.5726,
|
1087 |
+
"step": 154
|
1088 |
+
},
|
1089 |
+
{
|
1090 |
+
"epoch": 1.5048543689320388,
|
1091 |
+
"grad_norm": 1.1047143317546186,
|
1092 |
+
"learning_rate": 7.062238455139544e-05,
|
1093 |
+
"loss": 0.5754,
|
1094 |
+
"step": 155
|
1095 |
+
},
|
1096 |
+
{
|
1097 |
+
"epoch": 1.5145631067961165,
|
1098 |
+
"grad_norm": 1.3200995574966732,
|
1099 |
+
"learning_rate": 7.044706406022393e-05,
|
1100 |
+
"loss": 0.565,
|
1101 |
+
"step": 156
|
1102 |
+
},
|
1103 |
+
{
|
1104 |
+
"epoch": 1.5242718446601942,
|
1105 |
+
"grad_norm": 0.795738070159204,
|
1106 |
+
"learning_rate": 7.027034178369853e-05,
|
1107 |
+
"loss": 0.57,
|
1108 |
+
"step": 157
|
1109 |
+
},
|
1110 |
+
{
|
1111 |
+
"epoch": 1.5339805825242718,
|
1112 |
+
"grad_norm": 0.9946136430312941,
|
1113 |
+
"learning_rate": 7.009222585812755e-05,
|
1114 |
+
"loss": 0.5733,
|
1115 |
+
"step": 158
|
1116 |
+
},
|
1117 |
+
{
|
1118 |
+
"epoch": 1.5436893203883495,
|
1119 |
+
"grad_norm": 1.376070028063213,
|
1120 |
+
"learning_rate": 6.991272448398291e-05,
|
1121 |
+
"loss": 0.5756,
|
1122 |
+
"step": 159
|
1123 |
+
},
|
1124 |
+
{
|
1125 |
+
"epoch": 1.5533980582524272,
|
1126 |
+
"grad_norm": 0.7388882390316432,
|
1127 |
+
"learning_rate": 6.973184592552283e-05,
|
1128 |
+
"loss": 0.5654,
|
1129 |
+
"step": 160
|
1130 |
+
},
|
1131 |
+
{
|
1132 |
+
"epoch": 1.5631067961165048,
|
1133 |
+
"grad_norm": 1.1905180684222885,
|
1134 |
+
"learning_rate": 6.95495985104111e-05,
|
1135 |
+
"loss": 0.5711,
|
1136 |
+
"step": 161
|
1137 |
+
},
|
1138 |
+
{
|
1139 |
+
"epoch": 1.5728155339805825,
|
1140 |
+
"grad_norm": 0.8020799108353651,
|
1141 |
+
"learning_rate": 6.93659906293338e-05,
|
1142 |
+
"loss": 0.5623,
|
1143 |
+
"step": 162
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 1.5825242718446602,
|
1147 |
+
"grad_norm": 1.0231881194650578,
|
1148 |
+
"learning_rate": 6.918103073561304e-05,
|
1149 |
+
"loss": 0.57,
|
1150 |
+
"step": 163
|
1151 |
+
},
|
1152 |
+
{
|
1153 |
+
"epoch": 1.5922330097087378,
|
1154 |
+
"grad_norm": 0.7935997277801292,
|
1155 |
+
"learning_rate": 6.899472734481765e-05,
|
1156 |
+
"loss": 0.5697,
|
1157 |
+
"step": 164
|
1158 |
+
},
|
1159 |
+
{
|
1160 |
+
"epoch": 1.6019417475728155,
|
1161 |
+
"grad_norm": 1.0863918666939771,
|
1162 |
+
"learning_rate": 6.880708903437116e-05,
|
1163 |
+
"loss": 0.5649,
|
1164 |
+
"step": 165
|
1165 |
+
},
|
1166 |
+
{
|
1167 |
+
"epoch": 1.6116504854368932,
|
1168 |
+
"grad_norm": 0.7946977064700003,
|
1169 |
+
"learning_rate": 6.8618124443157e-05,
|
1170 |
+
"loss": 0.5629,
|
1171 |
+
"step": 166
|
1172 |
+
},
|
1173 |
+
{
|
1174 |
+
"epoch": 1.6213592233009708,
|
1175 |
+
"grad_norm": 0.7600587207085567,
|
1176 |
+
"learning_rate": 6.842784227112057e-05,
|
1177 |
+
"loss": 0.5652,
|
1178 |
+
"step": 167
|
1179 |
+
},
|
1180 |
+
{
|
1181 |
+
"epoch": 1.6310679611650487,
|
1182 |
+
"grad_norm": 0.9222151118870435,
|
1183 |
+
"learning_rate": 6.823625127886888e-05,
|
1184 |
+
"loss": 0.5602,
|
1185 |
+
"step": 168
|
1186 |
+
},
|
1187 |
+
{
|
1188 |
+
"epoch": 1.6407766990291264,
|
1189 |
+
"grad_norm": 0.8500423865673382,
|
1190 |
+
"learning_rate": 6.804336028726706e-05,
|
1191 |
+
"loss": 0.5638,
|
1192 |
+
"step": 169
|
1193 |
+
},
|
1194 |
+
{
|
1195 |
+
"epoch": 1.650485436893204,
|
1196 |
+
"grad_norm": 1.009548584428779,
|
1197 |
+
"learning_rate": 6.78491781770324e-05,
|
1198 |
+
"loss": 0.5609,
|
1199 |
+
"step": 170
|
1200 |
+
},
|
1201 |
+
{
|
1202 |
+
"epoch": 1.6601941747572817,
|
1203 |
+
"grad_norm": 0.8596245784399692,
|
1204 |
+
"learning_rate": 6.765371388832531e-05,
|
1205 |
+
"loss": 0.5606,
|
1206 |
+
"step": 171
|
1207 |
+
},
|
1208 |
+
{
|
1209 |
+
"epoch": 1.6699029126213594,
|
1210 |
+
"grad_norm": 1.0991682854870917,
|
1211 |
+
"learning_rate": 6.745697642033791e-05,
|
1212 |
+
"loss": 0.5645,
|
1213 |
+
"step": 172
|
1214 |
+
},
|
1215 |
+
{
|
1216 |
+
"epoch": 1.679611650485437,
|
1217 |
+
"grad_norm": 0.9305228321364903,
|
1218 |
+
"learning_rate": 6.725897483087948e-05,
|
1219 |
+
"loss": 0.5572,
|
1220 |
+
"step": 173
|
1221 |
+
},
|
1222 |
+
{
|
1223 |
+
"epoch": 1.6893203883495147,
|
1224 |
+
"grad_norm": 0.9614358976051515,
|
1225 |
+
"learning_rate": 6.705971823595964e-05,
|
1226 |
+
"loss": 0.5616,
|
1227 |
+
"step": 174
|
1228 |
+
},
|
1229 |
+
{
|
1230 |
+
"epoch": 1.6990291262135924,
|
1231 |
+
"grad_norm": 1.0715923701630468,
|
1232 |
+
"learning_rate": 6.685921580936855e-05,
|
1233 |
+
"loss": 0.5629,
|
1234 |
+
"step": 175
|
1235 |
+
},
|
1236 |
+
{
|
1237 |
+
"epoch": 1.70873786407767,
|
1238 |
+
"grad_norm": 1.1630310321701869,
|
1239 |
+
"learning_rate": 6.665747678225454e-05,
|
1240 |
+
"loss": 0.565,
|
1241 |
+
"step": 176
|
1242 |
+
},
|
1243 |
+
{
|
1244 |
+
"epoch": 1.7184466019417477,
|
1245 |
+
"grad_norm": 0.8590469522925508,
|
1246 |
+
"learning_rate": 6.645451044269916e-05,
|
1247 |
+
"loss": 0.5585,
|
1248 |
+
"step": 177
|
1249 |
+
},
|
1250 |
+
{
|
1251 |
+
"epoch": 1.7281553398058254,
|
1252 |
+
"grad_norm": 0.7340387980261763,
|
1253 |
+
"learning_rate": 6.62503261352895e-05,
|
1254 |
+
"loss": 0.5613,
|
1255 |
+
"step": 178
|
1256 |
+
},
|
1257 |
+
{
|
1258 |
+
"epoch": 1.737864077669903,
|
1259 |
+
"grad_norm": 0.710845727118427,
|
1260 |
+
"learning_rate": 6.6044933260688e-05,
|
1261 |
+
"loss": 0.5562,
|
1262 |
+
"step": 179
|
1263 |
+
},
|
1264 |
+
{
|
1265 |
+
"epoch": 1.7475728155339807,
|
1266 |
+
"grad_norm": 0.8027866048243354,
|
1267 |
+
"learning_rate": 6.583834127519966e-05,
|
1268 |
+
"loss": 0.5616,
|
1269 |
+
"step": 180
|
1270 |
+
},
|
1271 |
+
{
|
1272 |
+
"epoch": 1.7572815533980584,
|
1273 |
+
"grad_norm": 1.0941839867674634,
|
1274 |
+
"learning_rate": 6.563055969033659e-05,
|
1275 |
+
"loss": 0.5591,
|
1276 |
+
"step": 181
|
1277 |
+
},
|
1278 |
+
{
|
1279 |
+
"epoch": 1.766990291262136,
|
1280 |
+
"grad_norm": 1.066378946036087,
|
1281 |
+
"learning_rate": 6.54215980723802e-05,
|
1282 |
+
"loss": 0.5612,
|
1283 |
+
"step": 182
|
1284 |
+
},
|
1285 |
+
{
|
1286 |
+
"epoch": 1.7766990291262137,
|
1287 |
+
"grad_norm": 0.7626588888095535,
|
1288 |
+
"learning_rate": 6.521146604194073e-05,
|
1289 |
+
"loss": 0.5528,
|
1290 |
+
"step": 183
|
1291 |
+
},
|
1292 |
+
{
|
1293 |
+
"epoch": 1.7864077669902914,
|
1294 |
+
"grad_norm": 0.8152238302302431,
|
1295 |
+
"learning_rate": 6.500017327351425e-05,
|
1296 |
+
"loss": 0.559,
|
1297 |
+
"step": 184
|
1298 |
+
},
|
1299 |
+
{
|
1300 |
+
"epoch": 1.796116504854369,
|
1301 |
+
"grad_norm": 1.0106535990003993,
|
1302 |
+
"learning_rate": 6.478772949503735e-05,
|
1303 |
+
"loss": 0.5586,
|
1304 |
+
"step": 185
|
1305 |
+
},
|
1306 |
+
{
|
1307 |
+
"epoch": 1.8058252427184467,
|
1308 |
+
"grad_norm": 1.3373530495717711,
|
1309 |
+
"learning_rate": 6.457414448743922e-05,
|
1310 |
+
"loss": 0.5602,
|
1311 |
+
"step": 186
|
1312 |
+
},
|
1313 |
+
{
|
1314 |
+
"epoch": 1.8155339805825244,
|
1315 |
+
"grad_norm": 0.6183996673247697,
|
1316 |
+
"learning_rate": 6.435942808419129e-05,
|
1317 |
+
"loss": 0.5563,
|
1318 |
+
"step": 187
|
1319 |
+
},
|
1320 |
+
{
|
1321 |
+
"epoch": 1.825242718446602,
|
1322 |
+
"grad_norm": 0.4513128772050125,
|
1323 |
+
"learning_rate": 6.41435901708546e-05,
|
1324 |
+
"loss": 0.5558,
|
1325 |
+
"step": 188
|
1326 |
+
},
|
1327 |
+
{
|
1328 |
+
"epoch": 1.8349514563106797,
|
1329 |
+
"grad_norm": 0.8068976523956712,
|
1330 |
+
"learning_rate": 6.392664068462455e-05,
|
1331 |
+
"loss": 0.5535,
|
1332 |
+
"step": 189
|
1333 |
+
},
|
1334 |
+
{
|
1335 |
+
"epoch": 1.8446601941747574,
|
1336 |
+
"grad_norm": 1.1435206608085113,
|
1337 |
+
"learning_rate": 6.370858961387348e-05,
|
1338 |
+
"loss": 0.5589,
|
1339 |
+
"step": 190
|
1340 |
+
},
|
1341 |
+
{
|
1342 |
+
"epoch": 1.854368932038835,
|
1343 |
+
"grad_norm": 0.9238122145618127,
|
1344 |
+
"learning_rate": 6.348944699769078e-05,
|
1345 |
+
"loss": 0.5521,
|
1346 |
+
"step": 191
|
1347 |
+
},
|
1348 |
+
{
|
1349 |
+
"epoch": 1.8640776699029127,
|
1350 |
+
"grad_norm": 0.7049348477803952,
|
1351 |
+
"learning_rate": 6.326922292542067e-05,
|
1352 |
+
"loss": 0.5533,
|
1353 |
+
"step": 192
|
1354 |
+
},
|
1355 |
+
{
|
1356 |
+
"epoch": 1.8737864077669903,
|
1357 |
+
"grad_norm": 0.4705716067710833,
|
1358 |
+
"learning_rate": 6.304792753619768e-05,
|
1359 |
+
"loss": 0.555,
|
1360 |
+
"step": 193
|
1361 |
+
},
|
1362 |
+
{
|
1363 |
+
"epoch": 1.883495145631068,
|
1364 |
+
"grad_norm": 0.5562958157730256,
|
1365 |
+
"learning_rate": 6.282557101847989e-05,
|
1366 |
+
"loss": 0.554,
|
1367 |
+
"step": 194
|
1368 |
+
},
|
1369 |
+
{
|
1370 |
+
"epoch": 1.8932038834951457,
|
1371 |
+
"grad_norm": 0.7776677254914264,
|
1372 |
+
"learning_rate": 6.260216360957982e-05,
|
1373 |
+
"loss": 0.5554,
|
1374 |
+
"step": 195
|
1375 |
+
},
|
1376 |
+
{
|
1377 |
+
"epoch": 1.9029126213592233,
|
1378 |
+
"grad_norm": 0.9626928256285646,
|
1379 |
+
"learning_rate": 6.237771559519309e-05,
|
1380 |
+
"loss": 0.557,
|
1381 |
+
"step": 196
|
1382 |
+
},
|
1383 |
+
{
|
1384 |
+
"epoch": 1.912621359223301,
|
1385 |
+
"grad_norm": 1.0125011253597305,
|
1386 |
+
"learning_rate": 6.215223730892488e-05,
|
1387 |
+
"loss": 0.5576,
|
1388 |
+
"step": 197
|
1389 |
+
},
|
1390 |
+
{
|
1391 |
+
"epoch": 1.9223300970873787,
|
1392 |
+
"grad_norm": 0.8791857222410507,
|
1393 |
+
"learning_rate": 6.192573913181423e-05,
|
1394 |
+
"loss": 0.5601,
|
1395 |
+
"step": 198
|
1396 |
+
},
|
1397 |
+
{
|
1398 |
+
"epoch": 1.9320388349514563,
|
1399 |
+
"grad_norm": 0.7995737202735174,
|
1400 |
+
"learning_rate": 6.169823149185594e-05,
|
1401 |
+
"loss": 0.558,
|
1402 |
+
"step": 199
|
1403 |
+
},
|
1404 |
+
{
|
1405 |
+
"epoch": 1.941747572815534,
|
1406 |
+
"grad_norm": 1.1194920590127924,
|
1407 |
+
"learning_rate": 6.146972486352062e-05,
|
1408 |
+
"loss": 0.5547,
|
1409 |
+
"step": 200
|
1410 |
+
},
|
1411 |
+
{
|
1412 |
+
"epoch": 1.9514563106796117,
|
1413 |
+
"grad_norm": 0.8287113779997599,
|
1414 |
+
"learning_rate": 6.124022976727246e-05,
|
1415 |
+
"loss": 0.5562,
|
1416 |
+
"step": 201
|
1417 |
+
},
|
1418 |
+
{
|
1419 |
+
"epoch": 1.9611650485436893,
|
1420 |
+
"grad_norm": 0.5720384230610551,
|
1421 |
+
"learning_rate": 6.1009756769084625e-05,
|
1422 |
+
"loss": 0.5549,
|
1423 |
+
"step": 202
|
1424 |
+
},
|
1425 |
+
{
|
1426 |
+
"epoch": 1.970873786407767,
|
1427 |
+
"grad_norm": 0.9412252549729389,
|
1428 |
+
"learning_rate": 6.077831647995312e-05,
|
1429 |
+
"loss": 0.5561,
|
1430 |
+
"step": 203
|
1431 |
+
},
|
1432 |
+
{
|
1433 |
+
"epoch": 1.9805825242718447,
|
1434 |
+
"grad_norm": 1.1651868503900755,
|
1435 |
+
"learning_rate": 6.0545919555408026e-05,
|
1436 |
+
"loss": 0.5549,
|
1437 |
+
"step": 204
|
1438 |
+
},
|
1439 |
+
{
|
1440 |
+
"epoch": 1.9902912621359223,
|
1441 |
+
"grad_norm": 0.6059833274896135,
|
1442 |
+
"learning_rate": 6.0312576695023015e-05,
|
1443 |
+
"loss": 0.5532,
|
1444 |
+
"step": 205
|
1445 |
+
},
|
1446 |
+
{
|
1447 |
+
"epoch": 2.0,
|
1448 |
+
"grad_norm": 0.7944875296787923,
|
1449 |
+
"learning_rate": 6.007829864192274e-05,
|
1450 |
+
"loss": 0.5471,
|
1451 |
+
"step": 206
|
1452 |
+
},
|
1453 |
+
{
|
1454 |
+
"epoch": 2.0097087378640777,
|
1455 |
+
"grad_norm": 0.9861185591611118,
|
1456 |
+
"learning_rate": 5.9843096182288184e-05,
|
1457 |
+
"loss": 0.5254,
|
1458 |
+
"step": 207
|
1459 |
+
},
|
1460 |
+
{
|
1461 |
+
"epoch": 2.0194174757281553,
|
1462 |
+
"grad_norm": 0.9193682360720784,
|
1463 |
+
"learning_rate": 5.960698014486009e-05,
|
1464 |
+
"loss": 0.5228,
|
1465 |
+
"step": 208
|
1466 |
+
},
|
1467 |
+
{
|
1468 |
+
"epoch": 2.029126213592233,
|
1469 |
+
"grad_norm": 0.8942942482878242,
|
1470 |
+
"learning_rate": 5.936996140044041e-05,
|
1471 |
+
"loss": 0.5213,
|
1472 |
+
"step": 209
|
1473 |
+
},
|
1474 |
+
{
|
1475 |
+
"epoch": 2.0388349514563107,
|
1476 |
+
"grad_norm": 0.9820673732498754,
|
1477 |
+
"learning_rate": 5.9132050861391774e-05,
|
1478 |
+
"loss": 0.5281,
|
1479 |
+
"step": 210
|
1480 |
+
},
|
1481 |
+
{
|
1482 |
+
"epoch": 2.0485436893203883,
|
1483 |
+
"grad_norm": 1.124245573357021,
|
1484 |
+
"learning_rate": 5.889325948113513e-05,
|
1485 |
+
"loss": 0.5252,
|
1486 |
+
"step": 211
|
1487 |
+
},
|
1488 |
+
{
|
1489 |
+
"epoch": 2.058252427184466,
|
1490 |
+
"grad_norm": 0.9569857392646174,
|
1491 |
+
"learning_rate": 5.865359825364543e-05,
|
1492 |
+
"loss": 0.5295,
|
1493 |
+
"step": 212
|
1494 |
+
},
|
1495 |
+
{
|
1496 |
+
"epoch": 2.0679611650485437,
|
1497 |
+
"grad_norm": 0.6979418510578903,
|
1498 |
+
"learning_rate": 5.841307821294546e-05,
|
1499 |
+
"loss": 0.5212,
|
1500 |
+
"step": 213
|
1501 |
+
},
|
1502 |
+
{
|
1503 |
+
"epoch": 2.0776699029126213,
|
1504 |
+
"grad_norm": 0.8030912411204089,
|
1505 |
+
"learning_rate": 5.8171710432597824e-05,
|
1506 |
+
"loss": 0.525,
|
1507 |
+
"step": 214
|
1508 |
+
},
|
1509 |
+
{
|
1510 |
+
"epoch": 2.087378640776699,
|
1511 |
+
"grad_norm": 1.0921718953136728,
|
1512 |
+
"learning_rate": 5.792950602519516e-05,
|
1513 |
+
"loss": 0.5249,
|
1514 |
+
"step": 215
|
1515 |
+
},
|
1516 |
+
{
|
1517 |
+
"epoch": 2.0970873786407767,
|
1518 |
+
"grad_norm": 0.6815361141499702,
|
1519 |
+
"learning_rate": 5.768647614184846e-05,
|
1520 |
+
"loss": 0.5256,
|
1521 |
+
"step": 216
|
1522 |
+
},
|
1523 |
+
{
|
1524 |
+
"epoch": 2.1067961165048543,
|
1525 |
+
"grad_norm": 0.5765577764382596,
|
1526 |
+
"learning_rate": 5.744263197167369e-05,
|
1527 |
+
"loss": 0.5196,
|
1528 |
+
"step": 217
|
1529 |
+
},
|
1530 |
+
{
|
1531 |
+
"epoch": 2.116504854368932,
|
1532 |
+
"grad_norm": 0.8497374889157474,
|
1533 |
+
"learning_rate": 5.719798474127668e-05,
|
1534 |
+
"loss": 0.5245,
|
1535 |
+
"step": 218
|
1536 |
+
},
|
1537 |
+
{
|
1538 |
+
"epoch": 2.1262135922330097,
|
1539 |
+
"grad_norm": 0.8473330942862705,
|
1540 |
+
"learning_rate": 5.69525457142362e-05,
|
1541 |
+
"loss": 0.5266,
|
1542 |
+
"step": 219
|
1543 |
+
},
|
1544 |
+
{
|
1545 |
+
"epoch": 2.1359223300970873,
|
1546 |
+
"grad_norm": 0.7383851235625609,
|
1547 |
+
"learning_rate": 5.6706326190585416e-05,
|
1548 |
+
"loss": 0.5241,
|
1549 |
+
"step": 220
|
1550 |
+
},
|
1551 |
+
{
|
1552 |
+
"epoch": 2.145631067961165,
|
1553 |
+
"grad_norm": 0.731625229205727,
|
1554 |
+
"learning_rate": 5.6459337506291594e-05,
|
1555 |
+
"loss": 0.5214,
|
1556 |
+
"step": 221
|
1557 |
+
},
|
1558 |
+
{
|
1559 |
+
"epoch": 2.1553398058252426,
|
1560 |
+
"grad_norm": 0.6257327662890664,
|
1561 |
+
"learning_rate": 5.621159103273424e-05,
|
1562 |
+
"loss": 0.521,
|
1563 |
+
"step": 222
|
1564 |
+
},
|
1565 |
+
{
|
1566 |
+
"epoch": 2.1650485436893203,
|
1567 |
+
"grad_norm": 0.5394134102082522,
|
1568 |
+
"learning_rate": 5.596309817618156e-05,
|
1569 |
+
"loss": 0.5202,
|
1570 |
+
"step": 223
|
1571 |
+
},
|
1572 |
+
{
|
1573 |
+
"epoch": 2.174757281553398,
|
1574 |
+
"grad_norm": 0.5108441929525305,
|
1575 |
+
"learning_rate": 5.571387037726524e-05,
|
1576 |
+
"loss": 0.5243,
|
1577 |
+
"step": 224
|
1578 |
+
},
|
1579 |
+
{
|
1580 |
+
"epoch": 2.1844660194174756,
|
1581 |
+
"grad_norm": 0.6078187548666663,
|
1582 |
+
"learning_rate": 5.5463919110453836e-05,
|
1583 |
+
"loss": 0.5196,
|
1584 |
+
"step": 225
|
1585 |
+
},
|
1586 |
+
{
|
1587 |
+
"epoch": 2.1941747572815533,
|
1588 |
+
"grad_norm": 0.663473200149725,
|
1589 |
+
"learning_rate": 5.521325588352437e-05,
|
1590 |
+
"loss": 0.5194,
|
1591 |
+
"step": 226
|
1592 |
+
},
|
1593 |
+
{
|
1594 |
+
"epoch": 2.203883495145631,
|
1595 |
+
"grad_norm": 0.6429831831672274,
|
1596 |
+
"learning_rate": 5.496189223703262e-05,
|
1597 |
+
"loss": 0.5173,
|
1598 |
+
"step": 227
|
1599 |
+
},
|
1600 |
+
{
|
1601 |
+
"epoch": 2.2135922330097086,
|
1602 |
+
"grad_norm": 0.6220750410182839,
|
1603 |
+
"learning_rate": 5.47098397437817e-05,
|
1604 |
+
"loss": 0.5209,
|
1605 |
+
"step": 228
|
1606 |
+
},
|
1607 |
+
{
|
1608 |
+
"epoch": 2.2233009708737863,
|
1609 |
+
"grad_norm": 0.5886484774766658,
|
1610 |
+
"learning_rate": 5.4457110008289306e-05,
|
1611 |
+
"loss": 0.5155,
|
1612 |
+
"step": 229
|
1613 |
+
},
|
1614 |
+
{
|
1615 |
+
"epoch": 2.233009708737864,
|
1616 |
+
"grad_norm": 0.4437653008171185,
|
1617 |
+
"learning_rate": 5.420371466625339e-05,
|
1618 |
+
"loss": 0.5185,
|
1619 |
+
"step": 230
|
1620 |
+
},
|
1621 |
+
{
|
1622 |
+
"epoch": 2.2427184466019416,
|
1623 |
+
"grad_norm": 0.39870906238532716,
|
1624 |
+
"learning_rate": 5.3949665384016556e-05,
|
1625 |
+
"loss": 0.5217,
|
1626 |
+
"step": 231
|
1627 |
+
},
|
1628 |
+
{
|
1629 |
+
"epoch": 2.2524271844660193,
|
1630 |
+
"grad_norm": 0.5717388200227635,
|
1631 |
+
"learning_rate": 5.369497385802877e-05,
|
1632 |
+
"loss": 0.5196,
|
1633 |
+
"step": 232
|
1634 |
+
},
|
1635 |
+
{
|
1636 |
+
"epoch": 2.262135922330097,
|
1637 |
+
"grad_norm": 0.6577624762016366,
|
1638 |
+
"learning_rate": 5.3439651814309044e-05,
|
1639 |
+
"loss": 0.5182,
|
1640 |
+
"step": 233
|
1641 |
+
},
|
1642 |
+
{
|
1643 |
+
"epoch": 2.2718446601941746,
|
1644 |
+
"grad_norm": 0.6946628514135419,
|
1645 |
+
"learning_rate": 5.3183711007905434e-05,
|
1646 |
+
"loss": 0.5188,
|
1647 |
+
"step": 234
|
1648 |
+
},
|
1649 |
+
{
|
1650 |
+
"epoch": 2.2815533980582523,
|
1651 |
+
"grad_norm": 0.6759552411316988,
|
1652 |
+
"learning_rate": 5.2927163222353876e-05,
|
1653 |
+
"loss": 0.5171,
|
1654 |
+
"step": 235
|
1655 |
+
},
|
1656 |
+
{
|
1657 |
+
"epoch": 2.29126213592233,
|
1658 |
+
"grad_norm": 0.6300902669161529,
|
1659 |
+
"learning_rate": 5.2670020269135703e-05,
|
1660 |
+
"loss": 0.5224,
|
1661 |
+
"step": 236
|
1662 |
+
},
|
1663 |
+
{
|
1664 |
+
"epoch": 2.3009708737864076,
|
1665 |
+
"grad_norm": 0.5185045096628276,
|
1666 |
+
"learning_rate": 5.241229398713379e-05,
|
1667 |
+
"loss": 0.5268,
|
1668 |
+
"step": 237
|
1669 |
+
},
|
1670 |
+
{
|
1671 |
+
"epoch": 2.3106796116504853,
|
1672 |
+
"grad_norm": 0.4449400609212431,
|
1673 |
+
"learning_rate": 5.2153996242087544e-05,
|
1674 |
+
"loss": 0.5207,
|
1675 |
+
"step": 238
|
1676 |
+
},
|
1677 |
+
{
|
1678 |
+
"epoch": 2.320388349514563,
|
1679 |
+
"grad_norm": 0.4056127362971999,
|
1680 |
+
"learning_rate": 5.1895138926046553e-05,
|
1681 |
+
"loss": 0.5207,
|
1682 |
+
"step": 239
|
1683 |
+
},
|
1684 |
+
{
|
1685 |
+
"epoch": 2.3300970873786406,
|
1686 |
+
"grad_norm": 0.31876595932444957,
|
1687 |
+
"learning_rate": 5.16357339568231e-05,
|
1688 |
+
"loss": 0.5233,
|
1689 |
+
"step": 240
|
1690 |
+
},
|
1691 |
+
{
|
1692 |
+
"epoch": 2.3398058252427183,
|
1693 |
+
"grad_norm": 0.3826214088490694,
|
1694 |
+
"learning_rate": 5.13757932774435e-05,
|
1695 |
+
"loss": 0.5229,
|
1696 |
+
"step": 241
|
1697 |
+
},
|
1698 |
+
{
|
1699 |
+
"epoch": 2.349514563106796,
|
1700 |
+
"grad_norm": 0.39293509694649387,
|
1701 |
+
"learning_rate": 5.111532885559816e-05,
|
1702 |
+
"loss": 0.517,
|
1703 |
+
"step": 242
|
1704 |
+
},
|
1705 |
+
{
|
1706 |
+
"epoch": 2.3592233009708736,
|
1707 |
+
"grad_norm": 0.3322451238923787,
|
1708 |
+
"learning_rate": 5.08543526830907e-05,
|
1709 |
+
"loss": 0.5183,
|
1710 |
+
"step": 243
|
1711 |
+
},
|
1712 |
+
{
|
1713 |
+
"epoch": 2.3689320388349513,
|
1714 |
+
"grad_norm": 0.35774381456905935,
|
1715 |
+
"learning_rate": 5.05928767752857e-05,
|
1716 |
+
"loss": 0.5207,
|
1717 |
+
"step": 244
|
1718 |
+
},
|
1719 |
+
{
|
1720 |
+
"epoch": 2.378640776699029,
|
1721 |
+
"grad_norm": 0.3490143073371343,
|
1722 |
+
"learning_rate": 5.033091317055565e-05,
|
1723 |
+
"loss": 0.5185,
|
1724 |
+
"step": 245
|
1725 |
+
},
|
1726 |
+
{
|
1727 |
+
"epoch": 2.3883495145631066,
|
1728 |
+
"grad_norm": 0.3488449914720745,
|
1729 |
+
"learning_rate": 5.006847392972664e-05,
|
1730 |
+
"loss": 0.5233,
|
1731 |
+
"step": 246
|
1732 |
+
},
|
1733 |
+
{
|
1734 |
+
"epoch": 2.3980582524271843,
|
1735 |
+
"grad_norm": 0.4080875571104722,
|
1736 |
+
"learning_rate": 4.9805571135523066e-05,
|
1737 |
+
"loss": 0.5196,
|
1738 |
+
"step": 247
|
1739 |
+
},
|
1740 |
+
{
|
1741 |
+
"epoch": 2.407766990291262,
|
1742 |
+
"grad_norm": 0.3959917551246165,
|
1743 |
+
"learning_rate": 4.954221689201138e-05,
|
1744 |
+
"loss": 0.5194,
|
1745 |
+
"step": 248
|
1746 |
+
},
|
1747 |
+
{
|
1748 |
+
"epoch": 2.4174757281553396,
|
1749 |
+
"grad_norm": 0.331330529722134,
|
1750 |
+
"learning_rate": 4.9278423324042776e-05,
|
1751 |
+
"loss": 0.5175,
|
1752 |
+
"step": 249
|
1753 |
+
},
|
1754 |
+
{
|
1755 |
+
"epoch": 2.4271844660194173,
|
1756 |
+
"grad_norm": 0.3809362865456529,
|
1757 |
+
"learning_rate": 4.901420257669501e-05,
|
1758 |
+
"loss": 0.5195,
|
1759 |
+
"step": 250
|
1760 |
+
},
|
1761 |
+
{
|
1762 |
+
"epoch": 2.436893203883495,
|
1763 |
+
"grad_norm": 0.3961884838888781,
|
1764 |
+
"learning_rate": 4.8749566814713204e-05,
|
1765 |
+
"loss": 0.5203,
|
1766 |
+
"step": 251
|
1767 |
+
},
|
1768 |
+
{
|
1769 |
+
"epoch": 2.4466019417475726,
|
1770 |
+
"grad_norm": 0.28927135799547965,
|
1771 |
+
"learning_rate": 4.848452822194977e-05,
|
1772 |
+
"loss": 0.522,
|
1773 |
+
"step": 252
|
1774 |
+
},
|
1775 |
+
{
|
1776 |
+
"epoch": 2.4563106796116507,
|
1777 |
+
"grad_norm": 0.32321223873962246,
|
1778 |
+
"learning_rate": 4.821909900080348e-05,
|
1779 |
+
"loss": 0.5192,
|
1780 |
+
"step": 253
|
1781 |
+
},
|
1782 |
+
{
|
1783 |
+
"epoch": 2.466019417475728,
|
1784 |
+
"grad_norm": 0.3034931174843224,
|
1785 |
+
"learning_rate": 4.7953291371657724e-05,
|
1786 |
+
"loss": 0.5214,
|
1787 |
+
"step": 254
|
1788 |
+
},
|
1789 |
+
{
|
1790 |
+
"epoch": 2.475728155339806,
|
1791 |
+
"grad_norm": 0.2955075395432886,
|
1792 |
+
"learning_rate": 4.768711757231775e-05,
|
1793 |
+
"loss": 0.5197,
|
1794 |
+
"step": 255
|
1795 |
+
},
|
1796 |
+
{
|
1797 |
+
"epoch": 2.4854368932038833,
|
1798 |
+
"grad_norm": 0.33319864095909213,
|
1799 |
+
"learning_rate": 4.742058985744738e-05,
|
1800 |
+
"loss": 0.5199,
|
1801 |
+
"step": 256
|
1802 |
+
},
|
1803 |
+
{
|
1804 |
+
"epoch": 2.4951456310679614,
|
1805 |
+
"grad_norm": 0.3018636044546277,
|
1806 |
+
"learning_rate": 4.715372049800467e-05,
|
1807 |
+
"loss": 0.5234,
|
1808 |
+
"step": 257
|
1809 |
+
},
|
1810 |
+
{
|
1811 |
+
"epoch": 2.5048543689320386,
|
1812 |
+
"grad_norm": 0.28068625121120355,
|
1813 |
+
"learning_rate": 4.688652178067708e-05,
|
1814 |
+
"loss": 0.5223,
|
1815 |
+
"step": 258
|
1816 |
+
},
|
1817 |
+
{
|
1818 |
+
"epoch": 2.5145631067961167,
|
1819 |
+
"grad_norm": 0.3229351457016303,
|
1820 |
+
"learning_rate": 4.661900600731571e-05,
|
1821 |
+
"loss": 0.5233,
|
1822 |
+
"step": 259
|
1823 |
+
},
|
1824 |
+
{
|
1825 |
+
"epoch": 2.524271844660194,
|
1826 |
+
"grad_norm": 0.3762838763816941,
|
1827 |
+
"learning_rate": 4.635118549436895e-05,
|
1828 |
+
"loss": 0.5238,
|
1829 |
+
"step": 260
|
1830 |
+
},
|
1831 |
+
{
|
1832 |
+
"epoch": 2.533980582524272,
|
1833 |
+
"grad_norm": 0.36897391891079456,
|
1834 |
+
"learning_rate": 4.608307257231541e-05,
|
1835 |
+
"loss": 0.5184,
|
1836 |
+
"step": 261
|
1837 |
+
},
|
1838 |
+
{
|
1839 |
+
"epoch": 2.5436893203883493,
|
1840 |
+
"grad_norm": 0.3195793897829173,
|
1841 |
+
"learning_rate": 4.5814679585096265e-05,
|
1842 |
+
"loss": 0.5169,
|
1843 |
+
"step": 262
|
1844 |
+
},
|
1845 |
+
{
|
1846 |
+
"epoch": 2.5533980582524274,
|
1847 |
+
"grad_norm": 0.2773810615279961,
|
1848 |
+
"learning_rate": 4.5546018889546876e-05,
|
1849 |
+
"loss": 0.5168,
|
1850 |
+
"step": 263
|
1851 |
+
},
|
1852 |
+
{
|
1853 |
+
"epoch": 2.5631067961165046,
|
1854 |
+
"grad_norm": 0.3441559401937381,
|
1855 |
+
"learning_rate": 4.527710285482799e-05,
|
1856 |
+
"loss": 0.517,
|
1857 |
+
"step": 264
|
1858 |
+
},
|
1859 |
+
{
|
1860 |
+
"epoch": 2.5728155339805827,
|
1861 |
+
"grad_norm": 0.37273218540100866,
|
1862 |
+
"learning_rate": 4.500794386185609e-05,
|
1863 |
+
"loss": 0.5185,
|
1864 |
+
"step": 265
|
1865 |
+
},
|
1866 |
+
{
|
1867 |
+
"epoch": 2.58252427184466,
|
1868 |
+
"grad_norm": 0.26503820958744123,
|
1869 |
+
"learning_rate": 4.473855430273355e-05,
|
1870 |
+
"loss": 0.5164,
|
1871 |
+
"step": 266
|
1872 |
+
},
|
1873 |
+
{
|
1874 |
+
"epoch": 2.592233009708738,
|
1875 |
+
"grad_norm": 0.2934088823294493,
|
1876 |
+
"learning_rate": 4.4468946580178026e-05,
|
1877 |
+
"loss": 0.5127,
|
1878 |
+
"step": 267
|
1879 |
+
},
|
1880 |
+
{
|
1881 |
+
"epoch": 2.6019417475728153,
|
1882 |
+
"grad_norm": 0.3186295434544236,
|
1883 |
+
"learning_rate": 4.4199133106951407e-05,
|
1884 |
+
"loss": 0.5173,
|
1885 |
+
"step": 268
|
1886 |
+
},
|
1887 |
+
{
|
1888 |
+
"epoch": 2.6116504854368934,
|
1889 |
+
"grad_norm": 0.3309244613515348,
|
1890 |
+
"learning_rate": 4.3929126305288364e-05,
|
1891 |
+
"loss": 0.5229,
|
1892 |
+
"step": 269
|
1893 |
+
},
|
1894 |
+
{
|
1895 |
+
"epoch": 2.6213592233009706,
|
1896 |
+
"grad_norm": 0.26814510063287106,
|
1897 |
+
"learning_rate": 4.365893860632444e-05,
|
1898 |
+
"loss": 0.5167,
|
1899 |
+
"step": 270
|
1900 |
+
},
|
1901 |
+
{
|
1902 |
+
"epoch": 2.6310679611650487,
|
1903 |
+
"grad_norm": 0.3074091286659034,
|
1904 |
+
"learning_rate": 4.338858244952369e-05,
|
1905 |
+
"loss": 0.5156,
|
1906 |
+
"step": 271
|
1907 |
+
},
|
1908 |
+
{
|
1909 |
+
"epoch": 2.6407766990291264,
|
1910 |
+
"grad_norm": 0.3823340679989687,
|
1911 |
+
"learning_rate": 4.3118070282106e-05,
|
1912 |
+
"loss": 0.5168,
|
1913 |
+
"step": 272
|
1914 |
+
},
|
1915 |
+
{
|
1916 |
+
"epoch": 2.650485436893204,
|
1917 |
+
"grad_norm": 0.47904147679754805,
|
1918 |
+
"learning_rate": 4.2847414558473987e-05,
|
1919 |
+
"loss": 0.5184,
|
1920 |
+
"step": 273
|
1921 |
+
},
|
1922 |
+
{
|
1923 |
+
"epoch": 2.6601941747572817,
|
1924 |
+
"grad_norm": 0.4269268816899063,
|
1925 |
+
"learning_rate": 4.257662773963961e-05,
|
1926 |
+
"loss": 0.5173,
|
1927 |
+
"step": 274
|
1928 |
+
},
|
1929 |
+
{
|
1930 |
+
"epoch": 2.6699029126213594,
|
1931 |
+
"grad_norm": 0.3094464875254195,
|
1932 |
+
"learning_rate": 4.230572229265045e-05,
|
1933 |
+
"loss": 0.5142,
|
1934 |
+
"step": 275
|
1935 |
+
},
|
1936 |
+
{
|
1937 |
+
"epoch": 2.679611650485437,
|
1938 |
+
"grad_norm": 0.31791889845655724,
|
1939 |
+
"learning_rate": 4.2034710690015766e-05,
|
1940 |
+
"loss": 0.517,
|
1941 |
+
"step": 276
|
1942 |
+
},
|
1943 |
+
{
|
1944 |
+
"epoch": 2.6893203883495147,
|
1945 |
+
"grad_norm": 0.4064644387432894,
|
1946 |
+
"learning_rate": 4.17636054091322e-05,
|
1947 |
+
"loss": 0.516,
|
1948 |
+
"step": 277
|
1949 |
+
},
|
1950 |
+
{
|
1951 |
+
"epoch": 2.6990291262135924,
|
1952 |
+
"grad_norm": 0.3053310406953197,
|
1953 |
+
"learning_rate": 4.1492418931709366e-05,
|
1954 |
+
"loss": 0.5175,
|
1955 |
+
"step": 278
|
1956 |
+
},
|
1957 |
+
{
|
1958 |
+
"epoch": 2.70873786407767,
|
1959 |
+
"grad_norm": 0.2765928946593284,
|
1960 |
+
"learning_rate": 4.1221163743195175e-05,
|
1961 |
+
"loss": 0.5185,
|
1962 |
+
"step": 279
|
1963 |
+
},
|
1964 |
+
{
|
1965 |
+
"epoch": 2.7184466019417477,
|
1966 |
+
"grad_norm": 0.3378318482897848,
|
1967 |
+
"learning_rate": 4.094985233220098e-05,
|
1968 |
+
"loss": 0.5211,
|
1969 |
+
"step": 280
|
1970 |
+
},
|
1971 |
+
{
|
1972 |
+
"epoch": 2.7281553398058254,
|
1973 |
+
"grad_norm": 0.3143734127353884,
|
1974 |
+
"learning_rate": 4.067849718992665e-05,
|
1975 |
+
"loss": 0.5197,
|
1976 |
+
"step": 281
|
1977 |
+
},
|
1978 |
+
{
|
1979 |
+
"epoch": 2.737864077669903,
|
1980 |
+
"grad_norm": 0.259820524121846,
|
1981 |
+
"learning_rate": 4.040711080958547e-05,
|
1982 |
+
"loss": 0.5259,
|
1983 |
+
"step": 282
|
1984 |
+
},
|
1985 |
+
{
|
1986 |
+
"epoch": 2.7475728155339807,
|
1987 |
+
"grad_norm": 0.32611605738060934,
|
1988 |
+
"learning_rate": 4.013570568582883e-05,
|
1989 |
+
"loss": 0.5174,
|
1990 |
+
"step": 283
|
1991 |
+
},
|
1992 |
+
{
|
1993 |
+
"epoch": 2.7572815533980584,
|
1994 |
+
"grad_norm": 0.29760137014246574,
|
1995 |
+
"learning_rate": 3.986429431417118e-05,
|
1996 |
+
"loss": 0.5124,
|
1997 |
+
"step": 284
|
1998 |
+
},
|
1999 |
+
{
|
2000 |
+
"epoch": 2.766990291262136,
|
2001 |
+
"grad_norm": 0.26849482707048117,
|
2002 |
+
"learning_rate": 3.959288919041455e-05,
|
2003 |
+
"loss": 0.5116,
|
2004 |
+
"step": 285
|
2005 |
+
},
|
2006 |
+
{
|
2007 |
+
"epoch": 2.7766990291262137,
|
2008 |
+
"grad_norm": 0.28358089368587186,
|
2009 |
+
"learning_rate": 3.9321502810073354e-05,
|
2010 |
+
"loss": 0.5179,
|
2011 |
+
"step": 286
|
2012 |
+
},
|
2013 |
+
{
|
2014 |
+
"epoch": 2.7864077669902914,
|
2015 |
+
"grad_norm": 0.3098687865760963,
|
2016 |
+
"learning_rate": 3.905014766779904e-05,
|
2017 |
+
"loss": 0.5148,
|
2018 |
+
"step": 287
|
2019 |
+
},
|
2020 |
+
{
|
2021 |
+
"epoch": 2.796116504854369,
|
2022 |
+
"grad_norm": 0.3018433729775123,
|
2023 |
+
"learning_rate": 3.8778836256804845e-05,
|
2024 |
+
"loss": 0.5165,
|
2025 |
+
"step": 288
|
2026 |
+
},
|
2027 |
+
{
|
2028 |
+
"epoch": 2.8058252427184467,
|
2029 |
+
"grad_norm": 0.23893414660568268,
|
2030 |
+
"learning_rate": 3.850758106829065e-05,
|
2031 |
+
"loss": 0.5161,
|
2032 |
+
"step": 289
|
2033 |
+
},
|
2034 |
+
{
|
2035 |
+
"epoch": 2.8155339805825244,
|
2036 |
+
"grad_norm": 0.2524448827230724,
|
2037 |
+
"learning_rate": 3.823639459086781e-05,
|
2038 |
+
"loss": 0.5152,
|
2039 |
+
"step": 290
|
2040 |
+
},
|
2041 |
+
{
|
2042 |
+
"epoch": 2.825242718446602,
|
2043 |
+
"grad_norm": 0.2770033187618851,
|
2044 |
+
"learning_rate": 3.796528930998425e-05,
|
2045 |
+
"loss": 0.5145,
|
2046 |
+
"step": 291
|
2047 |
+
},
|
2048 |
+
{
|
2049 |
+
"epoch": 2.8349514563106797,
|
2050 |
+
"grad_norm": 0.25224488768181713,
|
2051 |
+
"learning_rate": 3.769427770734955e-05,
|
2052 |
+
"loss": 0.5195,
|
2053 |
+
"step": 292
|
2054 |
+
},
|
2055 |
+
{
|
2056 |
+
"epoch": 2.8446601941747574,
|
2057 |
+
"grad_norm": 0.27486298363736394,
|
2058 |
+
"learning_rate": 3.742337226036041e-05,
|
2059 |
+
"loss": 0.5157,
|
2060 |
+
"step": 293
|
2061 |
+
},
|
2062 |
+
{
|
2063 |
+
"epoch": 2.854368932038835,
|
2064 |
+
"grad_norm": 0.23960082819762807,
|
2065 |
+
"learning_rate": 3.715258544152603e-05,
|
2066 |
+
"loss": 0.5154,
|
2067 |
+
"step": 294
|
2068 |
+
},
|
2069 |
+
{
|
2070 |
+
"epoch": 2.8640776699029127,
|
2071 |
+
"grad_norm": 0.2710779329439077,
|
2072 |
+
"learning_rate": 3.688192971789401e-05,
|
2073 |
+
"loss": 0.5199,
|
2074 |
+
"step": 295
|
2075 |
+
},
|
2076 |
+
{
|
2077 |
+
"epoch": 2.8737864077669903,
|
2078 |
+
"grad_norm": 0.256469645800299,
|
2079 |
+
"learning_rate": 3.6611417550476324e-05,
|
2080 |
+
"loss": 0.5149,
|
2081 |
+
"step": 296
|
2082 |
+
},
|
2083 |
+
{
|
2084 |
+
"epoch": 2.883495145631068,
|
2085 |
+
"grad_norm": 0.3010842522468919,
|
2086 |
+
"learning_rate": 3.6341061393675574e-05,
|
2087 |
+
"loss": 0.5188,
|
2088 |
+
"step": 297
|
2089 |
+
},
|
2090 |
+
{
|
2091 |
+
"epoch": 2.8932038834951457,
|
2092 |
+
"grad_norm": 0.24615248139015927,
|
2093 |
+
"learning_rate": 3.607087369471164e-05,
|
2094 |
+
"loss": 0.5142,
|
2095 |
+
"step": 298
|
2096 |
+
},
|
2097 |
+
{
|
2098 |
+
"epoch": 2.9029126213592233,
|
2099 |
+
"grad_norm": 0.3025647213003904,
|
2100 |
+
"learning_rate": 3.580086689304861e-05,
|
2101 |
+
"loss": 0.5188,
|
2102 |
+
"step": 299
|
2103 |
+
},
|
2104 |
+
{
|
2105 |
+
"epoch": 2.912621359223301,
|
2106 |
+
"grad_norm": 0.2506413395712758,
|
2107 |
+
"learning_rate": 3.553105341982198e-05,
|
2108 |
+
"loss": 0.5127,
|
2109 |
+
"step": 300
|
2110 |
+
},
|
2111 |
+
{
|
2112 |
+
"epoch": 2.9223300970873787,
|
2113 |
+
"grad_norm": 0.26488744144075266,
|
2114 |
+
"learning_rate": 3.526144569726647e-05,
|
2115 |
+
"loss": 0.5148,
|
2116 |
+
"step": 301
|
2117 |
+
},
|
2118 |
+
{
|
2119 |
+
"epoch": 2.9320388349514563,
|
2120 |
+
"grad_norm": 0.23180961021643326,
|
2121 |
+
"learning_rate": 3.499205613814393e-05,
|
2122 |
+
"loss": 0.518,
|
2123 |
+
"step": 302
|
2124 |
+
},
|
2125 |
+
{
|
2126 |
+
"epoch": 2.941747572815534,
|
2127 |
+
"grad_norm": 0.24614944764274665,
|
2128 |
+
"learning_rate": 3.472289714517203e-05,
|
2129 |
+
"loss": 0.5119,
|
2130 |
+
"step": 303
|
2131 |
+
},
|
2132 |
+
{
|
2133 |
+
"epoch": 2.9514563106796117,
|
2134 |
+
"grad_norm": 0.22977263011082621,
|
2135 |
+
"learning_rate": 3.445398111045313e-05,
|
2136 |
+
"loss": 0.5184,
|
2137 |
+
"step": 304
|
2138 |
+
},
|
2139 |
+
{
|
2140 |
+
"epoch": 2.9611650485436893,
|
2141 |
+
"grad_norm": 0.21286238167178875,
|
2142 |
+
"learning_rate": 3.418532041490375e-05,
|
2143 |
+
"loss": 0.517,
|
2144 |
+
"step": 305
|
2145 |
+
},
|
2146 |
+
{
|
2147 |
+
"epoch": 2.970873786407767,
|
2148 |
+
"grad_norm": 0.2548130433665424,
|
2149 |
+
"learning_rate": 3.3916927427684595e-05,
|
2150 |
+
"loss": 0.5186,
|
2151 |
+
"step": 306
|
2152 |
+
},
|
2153 |
+
{
|
2154 |
+
"epoch": 2.9805825242718447,
|
2155 |
+
"grad_norm": 0.31989149824666413,
|
2156 |
+
"learning_rate": 3.364881450563106e-05,
|
2157 |
+
"loss": 0.5153,
|
2158 |
+
"step": 307
|
2159 |
+
},
|
2160 |
+
{
|
2161 |
+
"epoch": 2.9902912621359223,
|
2162 |
+
"grad_norm": 0.2762238599236643,
|
2163 |
+
"learning_rate": 3.338099399268429e-05,
|
2164 |
+
"loss": 0.5167,
|
2165 |
+
"step": 308
|
2166 |
+
},
|
2167 |
+
{
|
2168 |
+
"epoch": 3.0,
|
2169 |
+
"grad_norm": 0.20481446931191682,
|
2170 |
+
"learning_rate": 3.311347821932292e-05,
|
2171 |
+
"loss": 0.5092,
|
2172 |
+
"step": 309
|
2173 |
+
},
|
2174 |
+
{
|
2175 |
+
"epoch": 3.0097087378640777,
|
2176 |
+
"grad_norm": 0.22631177358316332,
|
2177 |
+
"learning_rate": 3.284627950199535e-05,
|
2178 |
+
"loss": 0.4909,
|
2179 |
+
"step": 310
|
2180 |
+
},
|
2181 |
+
{
|
2182 |
+
"epoch": 3.0194174757281553,
|
2183 |
+
"grad_norm": 0.21905855651481868,
|
2184 |
+
"learning_rate": 3.2579410142552646e-05,
|
2185 |
+
"loss": 0.4889,
|
2186 |
+
"step": 311
|
2187 |
+
},
|
2188 |
+
{
|
2189 |
+
"epoch": 3.029126213592233,
|
2190 |
+
"grad_norm": 0.268856232171971,
|
2191 |
+
"learning_rate": 3.231288242768226e-05,
|
2192 |
+
"loss": 0.4882,
|
2193 |
+
"step": 312
|
2194 |
+
},
|
2195 |
+
{
|
2196 |
+
"epoch": 3.0388349514563107,
|
2197 |
+
"grad_norm": 0.2631480748788351,
|
2198 |
+
"learning_rate": 3.204670862834228e-05,
|
2199 |
+
"loss": 0.4822,
|
2200 |
+
"step": 313
|
2201 |
+
},
|
2202 |
+
{
|
2203 |
+
"epoch": 3.0485436893203883,
|
2204 |
+
"grad_norm": 0.28121213579534965,
|
2205 |
+
"learning_rate": 3.178090099919653e-05,
|
2206 |
+
"loss": 0.4848,
|
2207 |
+
"step": 314
|
2208 |
+
},
|
2209 |
+
{
|
2210 |
+
"epoch": 3.058252427184466,
|
2211 |
+
"grad_norm": 0.2553829402200111,
|
2212 |
+
"learning_rate": 3.1515471778050246e-05,
|
2213 |
+
"loss": 0.4853,
|
2214 |
+
"step": 315
|
2215 |
+
},
|
2216 |
+
{
|
2217 |
+
"epoch": 3.0679611650485437,
|
2218 |
+
"grad_norm": 0.29584988541778207,
|
2219 |
+
"learning_rate": 3.12504331852868e-05,
|
2220 |
+
"loss": 0.4835,
|
2221 |
+
"step": 316
|
2222 |
+
},
|
2223 |
+
{
|
2224 |
+
"epoch": 3.0776699029126213,
|
2225 |
+
"grad_norm": 0.21912251520340542,
|
2226 |
+
"learning_rate": 3.098579742330499e-05,
|
2227 |
+
"loss": 0.4866,
|
2228 |
+
"step": 317
|
2229 |
+
},
|
2230 |
+
{
|
2231 |
+
"epoch": 3.087378640776699,
|
2232 |
+
"grad_norm": 0.2816186665047795,
|
2233 |
+
"learning_rate": 3.0721576675957224e-05,
|
2234 |
+
"loss": 0.4856,
|
2235 |
+
"step": 318
|
2236 |
+
},
|
2237 |
+
{
|
2238 |
+
"epoch": 3.0970873786407767,
|
2239 |
+
"grad_norm": 0.24175134230414194,
|
2240 |
+
"learning_rate": 3.0457783107988642e-05,
|
2241 |
+
"loss": 0.4886,
|
2242 |
+
"step": 319
|
2243 |
+
},
|
2244 |
+
{
|
2245 |
+
"epoch": 3.1067961165048543,
|
2246 |
+
"grad_norm": 0.22350452741221052,
|
2247 |
+
"learning_rate": 3.0194428864476947e-05,
|
2248 |
+
"loss": 0.4836,
|
2249 |
+
"step": 320
|
2250 |
+
},
|
2251 |
+
{
|
2252 |
+
"epoch": 3.116504854368932,
|
2253 |
+
"grad_norm": 0.25073180546811125,
|
2254 |
+
"learning_rate": 2.9931526070273374e-05,
|
2255 |
+
"loss": 0.4884,
|
2256 |
+
"step": 321
|
2257 |
+
},
|
2258 |
+
{
|
2259 |
+
"epoch": 3.1262135922330097,
|
2260 |
+
"grad_norm": 0.19079613465758094,
|
2261 |
+
"learning_rate": 2.9669086829444364e-05,
|
2262 |
+
"loss": 0.485,
|
2263 |
+
"step": 322
|
2264 |
+
},
|
2265 |
+
{
|
2266 |
+
"epoch": 3.1359223300970873,
|
2267 |
+
"grad_norm": 0.23960288437553956,
|
2268 |
+
"learning_rate": 2.9407123224714312e-05,
|
2269 |
+
"loss": 0.4856,
|
2270 |
+
"step": 323
|
2271 |
+
},
|
2272 |
+
{
|
2273 |
+
"epoch": 3.145631067961165,
|
2274 |
+
"grad_norm": 0.22627618578886288,
|
2275 |
+
"learning_rate": 2.9145647316909306e-05,
|
2276 |
+
"loss": 0.4862,
|
2277 |
+
"step": 324
|
2278 |
+
},
|
2279 |
+
{
|
2280 |
+
"epoch": 3.1553398058252426,
|
2281 |
+
"grad_norm": 0.1960335183955326,
|
2282 |
+
"learning_rate": 2.8884671144401833e-05,
|
2283 |
+
"loss": 0.4869,
|
2284 |
+
"step": 325
|
2285 |
+
},
|
2286 |
+
{
|
2287 |
+
"epoch": 3.1650485436893203,
|
2288 |
+
"grad_norm": 0.2079951634137142,
|
2289 |
+
"learning_rate": 2.8624206722556508e-05,
|
2290 |
+
"loss": 0.491,
|
2291 |
+
"step": 326
|
2292 |
+
},
|
2293 |
+
{
|
2294 |
+
"epoch": 3.174757281553398,
|
2295 |
+
"grad_norm": 0.18132268629371445,
|
2296 |
+
"learning_rate": 2.8364266043176897e-05,
|
2297 |
+
"loss": 0.4876,
|
2298 |
+
"step": 327
|
2299 |
+
},
|
2300 |
+
{
|
2301 |
+
"epoch": 3.1844660194174756,
|
2302 |
+
"grad_norm": 0.18068511996455502,
|
2303 |
+
"learning_rate": 2.810486107395347e-05,
|
2304 |
+
"loss": 0.4858,
|
2305 |
+
"step": 328
|
2306 |
+
},
|
2307 |
+
{
|
2308 |
+
"epoch": 3.1941747572815533,
|
2309 |
+
"grad_norm": 0.19264695463261375,
|
2310 |
+
"learning_rate": 2.7846003757912473e-05,
|
2311 |
+
"loss": 0.4891,
|
2312 |
+
"step": 329
|
2313 |
+
},
|
2314 |
+
{
|
2315 |
+
"epoch": 3.203883495145631,
|
2316 |
+
"grad_norm": 0.16897146599425877,
|
2317 |
+
"learning_rate": 2.7587706012866227e-05,
|
2318 |
+
"loss": 0.4868,
|
2319 |
+
"step": 330
|
2320 |
+
},
|
2321 |
+
{
|
2322 |
+
"epoch": 3.2135922330097086,
|
2323 |
+
"grad_norm": 0.2264234031229311,
|
2324 |
+
"learning_rate": 2.7329979730864313e-05,
|
2325 |
+
"loss": 0.4844,
|
2326 |
+
"step": 331
|
2327 |
+
},
|
2328 |
+
{
|
2329 |
+
"epoch": 3.2233009708737863,
|
2330 |
+
"grad_norm": 0.2433703396011944,
|
2331 |
+
"learning_rate": 2.707283677764613e-05,
|
2332 |
+
"loss": 0.4855,
|
2333 |
+
"step": 332
|
2334 |
+
},
|
2335 |
+
{
|
2336 |
+
"epoch": 3.233009708737864,
|
2337 |
+
"grad_norm": 0.1780724623798919,
|
2338 |
+
"learning_rate": 2.6816288992094573e-05,
|
2339 |
+
"loss": 0.4853,
|
2340 |
+
"step": 333
|
2341 |
+
},
|
2342 |
+
{
|
2343 |
+
"epoch": 3.2427184466019416,
|
2344 |
+
"grad_norm": 0.2550763857085153,
|
2345 |
+
"learning_rate": 2.6560348185690956e-05,
|
2346 |
+
"loss": 0.4865,
|
2347 |
+
"step": 334
|
2348 |
+
},
|
2349 |
+
{
|
2350 |
+
"epoch": 3.2524271844660193,
|
2351 |
+
"grad_norm": 0.18943111426329542,
|
2352 |
+
"learning_rate": 2.6305026141971227e-05,
|
2353 |
+
"loss": 0.4869,
|
2354 |
+
"step": 335
|
2355 |
+
},
|
2356 |
+
{
|
2357 |
+
"epoch": 3.262135922330097,
|
2358 |
+
"grad_norm": 0.21284295474198855,
|
2359 |
+
"learning_rate": 2.6050334615983467e-05,
|
2360 |
+
"loss": 0.4872,
|
2361 |
+
"step": 336
|
2362 |
+
},
|
2363 |
+
{
|
2364 |
+
"epoch": 3.2718446601941746,
|
2365 |
+
"grad_norm": 0.22131826031463572,
|
2366 |
+
"learning_rate": 2.5796285333746615e-05,
|
2367 |
+
"loss": 0.4816,
|
2368 |
+
"step": 337
|
2369 |
+
},
|
2370 |
+
{
|
2371 |
+
"epoch": 3.2815533980582523,
|
2372 |
+
"grad_norm": 0.182135484368646,
|
2373 |
+
"learning_rate": 2.554288999171072e-05,
|
2374 |
+
"loss": 0.4915,
|
2375 |
+
"step": 338
|
2376 |
+
},
|
2377 |
+
{
|
2378 |
+
"epoch": 3.29126213592233,
|
2379 |
+
"grad_norm": 0.23761469952772257,
|
2380 |
+
"learning_rate": 2.5290160256218313e-05,
|
2381 |
+
"loss": 0.4853,
|
2382 |
+
"step": 339
|
2383 |
+
},
|
2384 |
+
{
|
2385 |
+
"epoch": 3.3009708737864076,
|
2386 |
+
"grad_norm": 0.1840898568254873,
|
2387 |
+
"learning_rate": 2.5038107762967393e-05,
|
2388 |
+
"loss": 0.4883,
|
2389 |
+
"step": 340
|
2390 |
+
},
|
2391 |
+
{
|
2392 |
+
"epoch": 3.3106796116504853,
|
2393 |
+
"grad_norm": 0.21359795567551282,
|
2394 |
+
"learning_rate": 2.4786744116475638e-05,
|
2395 |
+
"loss": 0.4871,
|
2396 |
+
"step": 341
|
2397 |
+
},
|
2398 |
+
{
|
2399 |
+
"epoch": 3.320388349514563,
|
2400 |
+
"grad_norm": 0.24890926664546134,
|
2401 |
+
"learning_rate": 2.4536080889546177e-05,
|
2402 |
+
"loss": 0.489,
|
2403 |
+
"step": 342
|
2404 |
+
},
|
2405 |
+
{
|
2406 |
+
"epoch": 3.3300970873786406,
|
2407 |
+
"grad_norm": 0.20782018056580667,
|
2408 |
+
"learning_rate": 2.4286129622734764e-05,
|
2409 |
+
"loss": 0.4844,
|
2410 |
+
"step": 343
|
2411 |
+
},
|
2412 |
+
{
|
2413 |
+
"epoch": 3.3398058252427183,
|
2414 |
+
"grad_norm": 0.21924373313456091,
|
2415 |
+
"learning_rate": 2.4036901823818454e-05,
|
2416 |
+
"loss": 0.4863,
|
2417 |
+
"step": 344
|
2418 |
+
},
|
2419 |
+
{
|
2420 |
+
"epoch": 3.349514563106796,
|
2421 |
+
"grad_norm": 0.19867052689304365,
|
2422 |
+
"learning_rate": 2.378840896726577e-05,
|
2423 |
+
"loss": 0.4908,
|
2424 |
+
"step": 345
|
2425 |
+
},
|
2426 |
+
{
|
2427 |
+
"epoch": 3.3592233009708736,
|
2428 |
+
"grad_norm": 0.2327076797489768,
|
2429 |
+
"learning_rate": 2.3540662493708423e-05,
|
2430 |
+
"loss": 0.4869,
|
2431 |
+
"step": 346
|
2432 |
+
},
|
2433 |
+
{
|
2434 |
+
"epoch": 3.3689320388349513,
|
2435 |
+
"grad_norm": 0.17819481518798666,
|
2436 |
+
"learning_rate": 2.3293673809414598e-05,
|
2437 |
+
"loss": 0.4816,
|
2438 |
+
"step": 347
|
2439 |
+
},
|
2440 |
+
{
|
2441 |
+
"epoch": 3.378640776699029,
|
2442 |
+
"grad_norm": 0.17940622231016476,
|
2443 |
+
"learning_rate": 2.3047454285763793e-05,
|
2444 |
+
"loss": 0.4874,
|
2445 |
+
"step": 348
|
2446 |
+
},
|
2447 |
+
{
|
2448 |
+
"epoch": 3.3883495145631066,
|
2449 |
+
"grad_norm": 0.16039118497959268,
|
2450 |
+
"learning_rate": 2.2802015258723324e-05,
|
2451 |
+
"loss": 0.4869,
|
2452 |
+
"step": 349
|
2453 |
+
},
|
2454 |
+
{
|
2455 |
+
"epoch": 3.3980582524271843,
|
2456 |
+
"grad_norm": 0.17049865419413107,
|
2457 |
+
"learning_rate": 2.2557368028326324e-05,
|
2458 |
+
"loss": 0.4845,
|
2459 |
+
"step": 350
|
2460 |
+
},
|
2461 |
+
{
|
2462 |
+
"epoch": 3.407766990291262,
|
2463 |
+
"grad_norm": 0.18902552117069968,
|
2464 |
+
"learning_rate": 2.2313523858151554e-05,
|
2465 |
+
"loss": 0.4872,
|
2466 |
+
"step": 351
|
2467 |
+
},
|
2468 |
+
{
|
2469 |
+
"epoch": 3.4174757281553396,
|
2470 |
+
"grad_norm": 0.15937546411881842,
|
2471 |
+
"learning_rate": 2.207049397480485e-05,
|
2472 |
+
"loss": 0.4845,
|
2473 |
+
"step": 352
|
2474 |
+
},
|
2475 |
+
{
|
2476 |
+
"epoch": 3.4271844660194173,
|
2477 |
+
"grad_norm": 0.19899647930250075,
|
2478 |
+
"learning_rate": 2.1828289567402173e-05,
|
2479 |
+
"loss": 0.4874,
|
2480 |
+
"step": 353
|
2481 |
+
},
|
2482 |
+
{
|
2483 |
+
"epoch": 3.436893203883495,
|
2484 |
+
"grad_norm": 0.17339320655418355,
|
2485 |
+
"learning_rate": 2.1586921787054564e-05,
|
2486 |
+
"loss": 0.4871,
|
2487 |
+
"step": 354
|
2488 |
+
},
|
2489 |
+
{
|
2490 |
+
"epoch": 3.4466019417475726,
|
2491 |
+
"grad_norm": 0.17543747218166306,
|
2492 |
+
"learning_rate": 2.1346401746354576e-05,
|
2493 |
+
"loss": 0.4828,
|
2494 |
+
"step": 355
|
2495 |
+
},
|
2496 |
+
{
|
2497 |
+
"epoch": 3.4563106796116507,
|
2498 |
+
"grad_norm": 0.17798441146417476,
|
2499 |
+
"learning_rate": 2.110674051886488e-05,
|
2500 |
+
"loss": 0.4838,
|
2501 |
+
"step": 356
|
2502 |
+
},
|
2503 |
+
{
|
2504 |
+
"epoch": 3.466019417475728,
|
2505 |
+
"grad_norm": 0.16029275511201987,
|
2506 |
+
"learning_rate": 2.0867949138608242e-05,
|
2507 |
+
"loss": 0.4841,
|
2508 |
+
"step": 357
|
2509 |
+
},
|
2510 |
+
{
|
2511 |
+
"epoch": 3.475728155339806,
|
2512 |
+
"grad_norm": 0.1663439154413797,
|
2513 |
+
"learning_rate": 2.06300385995596e-05,
|
2514 |
+
"loss": 0.4865,
|
2515 |
+
"step": 358
|
2516 |
+
},
|
2517 |
+
{
|
2518 |
+
"epoch": 3.4854368932038833,
|
2519 |
+
"grad_norm": 0.16712634012536365,
|
2520 |
+
"learning_rate": 2.0393019855139915e-05,
|
2521 |
+
"loss": 0.4833,
|
2522 |
+
"step": 359
|
2523 |
+
},
|
2524 |
+
{
|
2525 |
+
"epoch": 3.4951456310679614,
|
2526 |
+
"grad_norm": 0.17251097817233577,
|
2527 |
+
"learning_rate": 2.0156903817711812e-05,
|
2528 |
+
"loss": 0.485,
|
2529 |
+
"step": 360
|
2530 |
+
},
|
2531 |
+
{
|
2532 |
+
"epoch": 3.5048543689320386,
|
2533 |
+
"grad_norm": 0.17262950471482985,
|
2534 |
+
"learning_rate": 1.9921701358077265e-05,
|
2535 |
+
"loss": 0.4846,
|
2536 |
+
"step": 361
|
2537 |
+
},
|
2538 |
+
{
|
2539 |
+
"epoch": 3.5145631067961167,
|
2540 |
+
"grad_norm": 0.17329043714528009,
|
2541 |
+
"learning_rate": 1.9687423304976994e-05,
|
2542 |
+
"loss": 0.4835,
|
2543 |
+
"step": 362
|
2544 |
+
},
|
2545 |
+
{
|
2546 |
+
"epoch": 3.524271844660194,
|
2547 |
+
"grad_norm": 0.16219535331159693,
|
2548 |
+
"learning_rate": 1.9454080444591998e-05,
|
2549 |
+
"loss": 0.4849,
|
2550 |
+
"step": 363
|
2551 |
+
},
|
2552 |
+
{
|
2553 |
+
"epoch": 3.533980582524272,
|
2554 |
+
"grad_norm": 0.16515501782595504,
|
2555 |
+
"learning_rate": 1.9221683520046892e-05,
|
2556 |
+
"loss": 0.4857,
|
2557 |
+
"step": 364
|
2558 |
+
},
|
2559 |
+
{
|
2560 |
+
"epoch": 3.5436893203883493,
|
2561 |
+
"grad_norm": 0.19734035971441835,
|
2562 |
+
"learning_rate": 1.899024323091539e-05,
|
2563 |
+
"loss": 0.4836,
|
2564 |
+
"step": 365
|
2565 |
+
},
|
2566 |
+
{
|
2567 |
+
"epoch": 3.5533980582524274,
|
2568 |
+
"grad_norm": 0.16333328213007223,
|
2569 |
+
"learning_rate": 1.875977023272757e-05,
|
2570 |
+
"loss": 0.485,
|
2571 |
+
"step": 366
|
2572 |
+
},
|
2573 |
+
{
|
2574 |
+
"epoch": 3.5631067961165046,
|
2575 |
+
"grad_norm": 0.18228849361538985,
|
2576 |
+
"learning_rate": 1.853027513647937e-05,
|
2577 |
+
"loss": 0.4869,
|
2578 |
+
"step": 367
|
2579 |
+
},
|
2580 |
+
{
|
2581 |
+
"epoch": 3.5728155339805827,
|
2582 |
+
"grad_norm": 0.14382298808613353,
|
2583 |
+
"learning_rate": 1.8301768508144078e-05,
|
2584 |
+
"loss": 0.4837,
|
2585 |
+
"step": 368
|
2586 |
+
},
|
2587 |
+
{
|
2588 |
+
"epoch": 3.58252427184466,
|
2589 |
+
"grad_norm": 0.1436393142572651,
|
2590 |
+
"learning_rate": 1.8074260868185784e-05,
|
2591 |
+
"loss": 0.4828,
|
2592 |
+
"step": 369
|
2593 |
+
},
|
2594 |
+
{
|
2595 |
+
"epoch": 3.592233009708738,
|
2596 |
+
"grad_norm": 0.16780032690913915,
|
2597 |
+
"learning_rate": 1.7847762691075115e-05,
|
2598 |
+
"loss": 0.487,
|
2599 |
+
"step": 370
|
2600 |
+
},
|
2601 |
+
{
|
2602 |
+
"epoch": 3.6019417475728153,
|
2603 |
+
"grad_norm": 0.14947935740806928,
|
2604 |
+
"learning_rate": 1.762228440480692e-05,
|
2605 |
+
"loss": 0.4845,
|
2606 |
+
"step": 371
|
2607 |
+
},
|
2608 |
+
{
|
2609 |
+
"epoch": 3.6116504854368934,
|
2610 |
+
"grad_norm": 0.14559731703552367,
|
2611 |
+
"learning_rate": 1.7397836390420192e-05,
|
2612 |
+
"loss": 0.4836,
|
2613 |
+
"step": 372
|
2614 |
+
},
|
2615 |
+
{
|
2616 |
+
"epoch": 3.6213592233009706,
|
2617 |
+
"grad_norm": 0.1506762657862038,
|
2618 |
+
"learning_rate": 1.717442898152012e-05,
|
2619 |
+
"loss": 0.4855,
|
2620 |
+
"step": 373
|
2621 |
+
},
|
2622 |
+
{
|
2623 |
+
"epoch": 3.6310679611650487,
|
2624 |
+
"grad_norm": 0.13248764059764206,
|
2625 |
+
"learning_rate": 1.6952072463802326e-05,
|
2626 |
+
"loss": 0.4848,
|
2627 |
+
"step": 374
|
2628 |
+
},
|
2629 |
+
{
|
2630 |
+
"epoch": 3.6407766990291264,
|
2631 |
+
"grad_norm": 0.14524262090031537,
|
2632 |
+
"learning_rate": 1.6730777074579346e-05,
|
2633 |
+
"loss": 0.4841,
|
2634 |
+
"step": 375
|
2635 |
+
},
|
2636 |
+
{
|
2637 |
+
"epoch": 3.650485436893204,
|
2638 |
+
"grad_norm": 0.13279946211767724,
|
2639 |
+
"learning_rate": 1.651055300230922e-05,
|
2640 |
+
"loss": 0.4847,
|
2641 |
+
"step": 376
|
2642 |
+
},
|
2643 |
+
{
|
2644 |
+
"epoch": 3.6601941747572817,
|
2645 |
+
"grad_norm": 0.14766932876767538,
|
2646 |
+
"learning_rate": 1.6291410386126524e-05,
|
2647 |
+
"loss": 0.4854,
|
2648 |
+
"step": 377
|
2649 |
+
},
|
2650 |
+
{
|
2651 |
+
"epoch": 3.6699029126213594,
|
2652 |
+
"grad_norm": 0.14421279285412036,
|
2653 |
+
"learning_rate": 1.607335931537547e-05,
|
2654 |
+
"loss": 0.4834,
|
2655 |
+
"step": 378
|
2656 |
+
},
|
2657 |
+
{
|
2658 |
+
"epoch": 3.679611650485437,
|
2659 |
+
"grad_norm": 0.1332433537811279,
|
2660 |
+
"learning_rate": 1.585640982914541e-05,
|
2661 |
+
"loss": 0.4881,
|
2662 |
+
"step": 379
|
2663 |
+
},
|
2664 |
+
{
|
2665 |
+
"epoch": 3.6893203883495147,
|
2666 |
+
"grad_norm": 0.13964291558360487,
|
2667 |
+
"learning_rate": 1.564057191580873e-05,
|
2668 |
+
"loss": 0.4812,
|
2669 |
+
"step": 380
|
2670 |
+
},
|
2671 |
+
{
|
2672 |
+
"epoch": 3.6990291262135924,
|
2673 |
+
"grad_norm": 0.13379924755179787,
|
2674 |
+
"learning_rate": 1.54258555125608e-05,
|
2675 |
+
"loss": 0.4865,
|
2676 |
+
"step": 381
|
2677 |
+
},
|
2678 |
+
{
|
2679 |
+
"epoch": 3.70873786407767,
|
2680 |
+
"grad_norm": 0.13942795631282168,
|
2681 |
+
"learning_rate": 1.521227050496266e-05,
|
2682 |
+
"loss": 0.4862,
|
2683 |
+
"step": 382
|
2684 |
+
},
|
2685 |
+
{
|
2686 |
+
"epoch": 3.7184466019417477,
|
2687 |
+
"grad_norm": 0.14829905133246946,
|
2688 |
+
"learning_rate": 1.4999826726485754e-05,
|
2689 |
+
"loss": 0.4841,
|
2690 |
+
"step": 383
|
2691 |
+
},
|
2692 |
+
{
|
2693 |
+
"epoch": 3.7281553398058254,
|
2694 |
+
"grad_norm": 0.13774937243212151,
|
2695 |
+
"learning_rate": 1.4788533958059281e-05,
|
2696 |
+
"loss": 0.4873,
|
2697 |
+
"step": 384
|
2698 |
+
},
|
2699 |
+
{
|
2700 |
+
"epoch": 3.737864077669903,
|
2701 |
+
"grad_norm": 0.14632219044552267,
|
2702 |
+
"learning_rate": 1.457840192761979e-05,
|
2703 |
+
"loss": 0.4854,
|
2704 |
+
"step": 385
|
2705 |
+
},
|
2706 |
+
{
|
2707 |
+
"epoch": 3.7475728155339807,
|
2708 |
+
"grad_norm": 0.15302510684664683,
|
2709 |
+
"learning_rate": 1.4369440309663412e-05,
|
2710 |
+
"loss": 0.4833,
|
2711 |
+
"step": 386
|
2712 |
+
},
|
2713 |
+
{
|
2714 |
+
"epoch": 3.7572815533980584,
|
2715 |
+
"grad_norm": 0.14352691900463502,
|
2716 |
+
"learning_rate": 1.4161658724800357e-05,
|
2717 |
+
"loss": 0.4846,
|
2718 |
+
"step": 387
|
2719 |
+
},
|
2720 |
+
{
|
2721 |
+
"epoch": 3.766990291262136,
|
2722 |
+
"grad_norm": 0.1517023916720968,
|
2723 |
+
"learning_rate": 1.3955066739312e-05,
|
2724 |
+
"loss": 0.4867,
|
2725 |
+
"step": 388
|
2726 |
+
},
|
2727 |
+
{
|
2728 |
+
"epoch": 3.7766990291262137,
|
2729 |
+
"grad_norm": 0.15194309330245784,
|
2730 |
+
"learning_rate": 1.3749673864710524e-05,
|
2731 |
+
"loss": 0.4865,
|
2732 |
+
"step": 389
|
2733 |
+
},
|
2734 |
+
{
|
2735 |
+
"epoch": 3.7864077669902914,
|
2736 |
+
"grad_norm": 0.1468613357237534,
|
2737 |
+
"learning_rate": 1.3545489557300853e-05,
|
2738 |
+
"loss": 0.4846,
|
2739 |
+
"step": 390
|
2740 |
+
},
|
2741 |
+
{
|
2742 |
+
"epoch": 3.796116504854369,
|
2743 |
+
"grad_norm": 0.16395511168934657,
|
2744 |
+
"learning_rate": 1.3342523217745473e-05,
|
2745 |
+
"loss": 0.4869,
|
2746 |
+
"step": 391
|
2747 |
+
},
|
2748 |
+
{
|
2749 |
+
"epoch": 3.8058252427184467,
|
2750 |
+
"grad_norm": 0.15078193434327586,
|
2751 |
+
"learning_rate": 1.3140784190631459e-05,
|
2752 |
+
"loss": 0.4825,
|
2753 |
+
"step": 392
|
2754 |
+
},
|
2755 |
+
{
|
2756 |
+
"epoch": 3.8155339805825244,
|
2757 |
+
"grad_norm": 0.14510817920359698,
|
2758 |
+
"learning_rate": 1.2940281764040368e-05,
|
2759 |
+
"loss": 0.4825,
|
2760 |
+
"step": 393
|
2761 |
+
},
|
2762 |
+
{
|
2763 |
+
"epoch": 3.825242718446602,
|
2764 |
+
"grad_norm": 0.17265976668387825,
|
2765 |
+
"learning_rate": 1.2741025169120539e-05,
|
2766 |
+
"loss": 0.4872,
|
2767 |
+
"step": 394
|
2768 |
+
},
|
2769 |
+
{
|
2770 |
+
"epoch": 3.8349514563106797,
|
2771 |
+
"grad_norm": 0.14620030187493488,
|
2772 |
+
"learning_rate": 1.2543023579662106e-05,
|
2773 |
+
"loss": 0.4845,
|
2774 |
+
"step": 395
|
2775 |
+
},
|
2776 |
+
{
|
2777 |
+
"epoch": 3.8446601941747574,
|
2778 |
+
"grad_norm": 0.1418045201790065,
|
2779 |
+
"learning_rate": 1.234628611167469e-05,
|
2780 |
+
"loss": 0.4845,
|
2781 |
+
"step": 396
|
2782 |
+
},
|
2783 |
+
{
|
2784 |
+
"epoch": 3.854368932038835,
|
2785 |
+
"grad_norm": 0.14833568549734966,
|
2786 |
+
"learning_rate": 1.2150821822967611e-05,
|
2787 |
+
"loss": 0.4882,
|
2788 |
+
"step": 397
|
2789 |
+
},
|
2790 |
+
{
|
2791 |
+
"epoch": 3.8640776699029127,
|
2792 |
+
"grad_norm": 0.147355117137231,
|
2793 |
+
"learning_rate": 1.1956639712732958e-05,
|
2794 |
+
"loss": 0.4845,
|
2795 |
+
"step": 398
|
2796 |
+
},
|
2797 |
+
{
|
2798 |
+
"epoch": 3.8737864077669903,
|
2799 |
+
"grad_norm": 0.13630141790437844,
|
2800 |
+
"learning_rate": 1.1763748721131142e-05,
|
2801 |
+
"loss": 0.4819,
|
2802 |
+
"step": 399
|
2803 |
+
},
|
2804 |
+
{
|
2805 |
+
"epoch": 3.883495145631068,
|
2806 |
+
"grad_norm": 0.13485086713221084,
|
2807 |
+
"learning_rate": 1.1572157728879444e-05,
|
2808 |
+
"loss": 0.485,
|
2809 |
+
"step": 400
|
2810 |
+
},
|
2811 |
+
{
|
2812 |
+
"epoch": 3.8932038834951457,
|
2813 |
+
"grad_norm": 0.14115351641144186,
|
2814 |
+
"learning_rate": 1.1381875556843007e-05,
|
2815 |
+
"loss": 0.4874,
|
2816 |
+
"step": 401
|
2817 |
+
},
|
2818 |
+
{
|
2819 |
+
"epoch": 3.9029126213592233,
|
2820 |
+
"grad_norm": 0.13251154708545737,
|
2821 |
+
"learning_rate": 1.119291096562884e-05,
|
2822 |
+
"loss": 0.4879,
|
2823 |
+
"step": 402
|
2824 |
+
},
|
2825 |
+
{
|
2826 |
+
"epoch": 3.912621359223301,
|
2827 |
+
"grad_norm": 0.14538526251672798,
|
2828 |
+
"learning_rate": 1.1005272655182378e-05,
|
2829 |
+
"loss": 0.487,
|
2830 |
+
"step": 403
|
2831 |
+
},
|
2832 |
+
{
|
2833 |
+
"epoch": 3.9223300970873787,
|
2834 |
+
"grad_norm": 0.14640871835491487,
|
2835 |
+
"learning_rate": 1.0818969264386973e-05,
|
2836 |
+
"loss": 0.4843,
|
2837 |
+
"step": 404
|
2838 |
+
},
|
2839 |
+
{
|
2840 |
+
"epoch": 3.9320388349514563,
|
2841 |
+
"grad_norm": 0.13392456360817495,
|
2842 |
+
"learning_rate": 1.0634009370666214e-05,
|
2843 |
+
"loss": 0.4826,
|
2844 |
+
"step": 405
|
2845 |
+
},
|
2846 |
+
{
|
2847 |
+
"epoch": 3.941747572815534,
|
2848 |
+
"grad_norm": 0.15291675444361674,
|
2849 |
+
"learning_rate": 1.045040148958893e-05,
|
2850 |
+
"loss": 0.488,
|
2851 |
+
"step": 406
|
2852 |
+
},
|
2853 |
+
{
|
2854 |
+
"epoch": 3.9514563106796117,
|
2855 |
+
"grad_norm": 0.13901859144613157,
|
2856 |
+
"learning_rate": 1.0268154074477188e-05,
|
2857 |
+
"loss": 0.4853,
|
2858 |
+
"step": 407
|
2859 |
+
},
|
2860 |
+
{
|
2861 |
+
"epoch": 3.9611650485436893,
|
2862 |
+
"grad_norm": 0.13797907262373485,
|
2863 |
+
"learning_rate": 1.0087275516017083e-05,
|
2864 |
+
"loss": 0.4834,
|
2865 |
+
"step": 408
|
2866 |
+
},
|
2867 |
+
{
|
2868 |
+
"epoch": 3.970873786407767,
|
2869 |
+
"grad_norm": 0.13762327771479307,
|
2870 |
+
"learning_rate": 9.907774141872468e-06,
|
2871 |
+
"loss": 0.488,
|
2872 |
+
"step": 409
|
2873 |
+
},
|
2874 |
+
{
|
2875 |
+
"epoch": 3.9805825242718447,
|
2876 |
+
"grad_norm": 0.14986633803246888,
|
2877 |
+
"learning_rate": 9.729658216301479e-06,
|
2878 |
+
"loss": 0.4828,
|
2879 |
+
"step": 410
|
2880 |
+
},
|
2881 |
+
{
|
2882 |
+
"epoch": 3.9902912621359223,
|
2883 |
+
"grad_norm": 0.13435313597563736,
|
2884 |
+
"learning_rate": 9.552935939776083e-06,
|
2885 |
+
"loss": 0.4871,
|
2886 |
+
"step": 411
|
2887 |
+
},
|
2888 |
+
{
|
2889 |
+
"epoch": 4.0,
|
2890 |
+
"grad_norm": 0.14759467236298449,
|
2891 |
+
"learning_rate": 9.377615448604574e-06,
|
2892 |
+
"loss": 0.4831,
|
2893 |
+
"step": 412
|
2894 |
+
},
|
2895 |
+
{
|
2896 |
+
"epoch": 4.009708737864078,
|
2897 |
+
"grad_norm": 0.17306287750975244,
|
2898 |
+
"learning_rate": 9.203704814556871e-06,
|
2899 |
+
"loss": 0.4669,
|
2900 |
+
"step": 413
|
2901 |
+
},
|
2902 |
+
{
|
2903 |
+
"epoch": 4.019417475728155,
|
2904 |
+
"grad_norm": 0.14023047375285927,
|
2905 |
+
"learning_rate": 9.031212044493016e-06,
|
2906 |
+
"loss": 0.4659,
|
2907 |
+
"step": 414
|
2908 |
+
},
|
2909 |
+
{
|
2910 |
+
"epoch": 4.029126213592233,
|
2911 |
+
"grad_norm": 0.14153603129051684,
|
2912 |
+
"learning_rate": 8.860145079994433e-06,
|
2913 |
+
"loss": 0.4656,
|
2914 |
+
"step": 415
|
2915 |
+
},
|
2916 |
+
{
|
2917 |
+
"epoch": 4.038834951456311,
|
2918 |
+
"grad_norm": 0.14948734290428067,
|
2919 |
+
"learning_rate": 8.690511796998344e-06,
|
2920 |
+
"loss": 0.4653,
|
2921 |
+
"step": 416
|
2922 |
+
},
|
2923 |
+
{
|
2924 |
+
"epoch": 4.048543689320389,
|
2925 |
+
"grad_norm": 0.16197544725648444,
|
2926 |
+
"learning_rate": 8.522320005435162e-06,
|
2927 |
+
"loss": 0.4686,
|
2928 |
+
"step": 417
|
2929 |
+
},
|
2930 |
+
{
|
2931 |
+
"epoch": 4.058252427184466,
|
2932 |
+
"grad_norm": 0.16013381839658847,
|
2933 |
+
"learning_rate": 8.355577448868933e-06,
|
2934 |
+
"loss": 0.4709,
|
2935 |
+
"step": 418
|
2936 |
+
},
|
2937 |
+
{
|
2938 |
+
"epoch": 4.067961165048544,
|
2939 |
+
"grad_norm": 0.13999136843500865,
|
2940 |
+
"learning_rate": 8.190291804140775e-06,
|
2941 |
+
"loss": 0.4648,
|
2942 |
+
"step": 419
|
2943 |
+
},
|
2944 |
+
{
|
2945 |
+
"epoch": 4.077669902912621,
|
2946 |
+
"grad_norm": 0.1470240140496139,
|
2947 |
+
"learning_rate": 8.02647068101547e-06,
|
2948 |
+
"loss": 0.4681,
|
2949 |
+
"step": 420
|
2950 |
+
},
|
2951 |
+
{
|
2952 |
+
"epoch": 4.087378640776699,
|
2953 |
+
"grad_norm": 0.1355700665658678,
|
2954 |
+
"learning_rate": 7.864121621831126e-06,
|
2955 |
+
"loss": 0.467,
|
2956 |
+
"step": 421
|
2957 |
+
},
|
2958 |
+
{
|
2959 |
+
"epoch": 4.097087378640777,
|
2960 |
+
"grad_norm": 0.14144514644659487,
|
2961 |
+
"learning_rate": 7.703252101151873e-06,
|
2962 |
+
"loss": 0.4613,
|
2963 |
+
"step": 422
|
2964 |
+
},
|
2965 |
+
{
|
2966 |
+
"epoch": 4.106796116504855,
|
2967 |
+
"grad_norm": 0.129930341121314,
|
2968 |
+
"learning_rate": 7.5438695254238e-06,
|
2969 |
+
"loss": 0.4638,
|
2970 |
+
"step": 423
|
2971 |
+
},
|
2972 |
+
{
|
2973 |
+
"epoch": 4.116504854368932,
|
2974 |
+
"grad_norm": 0.14218321035952994,
|
2975 |
+
"learning_rate": 7.385981232633894e-06,
|
2976 |
+
"loss": 0.465,
|
2977 |
+
"step": 424
|
2978 |
+
},
|
2979 |
+
{
|
2980 |
+
"epoch": 4.12621359223301,
|
2981 |
+
"grad_norm": 0.1420896044955223,
|
2982 |
+
"learning_rate": 7.229594491972256e-06,
|
2983 |
+
"loss": 0.4659,
|
2984 |
+
"step": 425
|
2985 |
+
},
|
2986 |
+
{
|
2987 |
+
"epoch": 4.135922330097087,
|
2988 |
+
"grad_norm": 0.13076263084557904,
|
2989 |
+
"learning_rate": 7.07471650349739e-06,
|
2990 |
+
"loss": 0.4648,
|
2991 |
+
"step": 426
|
2992 |
+
},
|
2993 |
+
{
|
2994 |
+
"epoch": 4.145631067961165,
|
2995 |
+
"grad_norm": 0.13079701173247169,
|
2996 |
+
"learning_rate": 6.921354397804712e-06,
|
2997 |
+
"loss": 0.4651,
|
2998 |
+
"step": 427
|
2999 |
+
},
|
3000 |
+
{
|
3001 |
+
"epoch": 4.155339805825243,
|
3002 |
+
"grad_norm": 0.1299754901865967,
|
3003 |
+
"learning_rate": 6.7695152356983054e-06,
|
3004 |
+
"loss": 0.4647,
|
3005 |
+
"step": 428
|
3006 |
+
},
|
3007 |
+
{
|
3008 |
+
"epoch": 4.165048543689321,
|
3009 |
+
"grad_norm": 0.12764332296015807,
|
3010 |
+
"learning_rate": 6.619206007865768e-06,
|
3011 |
+
"loss": 0.4607,
|
3012 |
+
"step": 429
|
3013 |
+
},
|
3014 |
+
{
|
3015 |
+
"epoch": 4.174757281553398,
|
3016 |
+
"grad_norm": 0.12376699505437842,
|
3017 |
+
"learning_rate": 6.47043363455643e-06,
|
3018 |
+
"loss": 0.4674,
|
3019 |
+
"step": 430
|
3020 |
+
},
|
3021 |
+
{
|
3022 |
+
"epoch": 4.184466019417476,
|
3023 |
+
"grad_norm": 0.11650682736889516,
|
3024 |
+
"learning_rate": 6.323204965262686e-06,
|
3025 |
+
"loss": 0.463,
|
3026 |
+
"step": 431
|
3027 |
+
},
|
3028 |
+
{
|
3029 |
+
"epoch": 4.194174757281553,
|
3030 |
+
"grad_norm": 0.1217502825230698,
|
3031 |
+
"learning_rate": 6.177526778404663e-06,
|
3032 |
+
"loss": 0.4679,
|
3033 |
+
"step": 432
|
3034 |
+
},
|
3035 |
+
{
|
3036 |
+
"epoch": 4.203883495145631,
|
3037 |
+
"grad_norm": 0.11887732797821776,
|
3038 |
+
"learning_rate": 6.033405781018195e-06,
|
3039 |
+
"loss": 0.4655,
|
3040 |
+
"step": 433
|
3041 |
+
},
|
3042 |
+
{
|
3043 |
+
"epoch": 4.213592233009709,
|
3044 |
+
"grad_norm": 0.12496115598562435,
|
3045 |
+
"learning_rate": 5.8908486084459134e-06,
|
3046 |
+
"loss": 0.4635,
|
3047 |
+
"step": 434
|
3048 |
+
},
|
3049 |
+
{
|
3050 |
+
"epoch": 4.223300970873787,
|
3051 |
+
"grad_norm": 0.12470991178018197,
|
3052 |
+
"learning_rate": 5.74986182403189e-06,
|
3053 |
+
"loss": 0.4646,
|
3054 |
+
"step": 435
|
3055 |
+
},
|
3056 |
+
{
|
3057 |
+
"epoch": 4.233009708737864,
|
3058 |
+
"grad_norm": 0.1292310303978258,
|
3059 |
+
"learning_rate": 5.610451918819357e-06,
|
3060 |
+
"loss": 0.4651,
|
3061 |
+
"step": 436
|
3062 |
+
},
|
3063 |
+
{
|
3064 |
+
"epoch": 4.242718446601942,
|
3065 |
+
"grad_norm": 0.13694613972829348,
|
3066 |
+
"learning_rate": 5.472625311251918e-06,
|
3067 |
+
"loss": 0.4667,
|
3068 |
+
"step": 437
|
3069 |
+
},
|
3070 |
+
{
|
3071 |
+
"epoch": 4.252427184466019,
|
3072 |
+
"grad_norm": 0.12260186181424092,
|
3073 |
+
"learning_rate": 5.336388346878006e-06,
|
3074 |
+
"loss": 0.4654,
|
3075 |
+
"step": 438
|
3076 |
+
},
|
3077 |
+
{
|
3078 |
+
"epoch": 4.262135922330097,
|
3079 |
+
"grad_norm": 0.12523666490053748,
|
3080 |
+
"learning_rate": 5.201747298058765e-06,
|
3081 |
+
"loss": 0.4604,
|
3082 |
+
"step": 439
|
3083 |
+
},
|
3084 |
+
{
|
3085 |
+
"epoch": 4.271844660194175,
|
3086 |
+
"grad_norm": 0.12451383039630047,
|
3087 |
+
"learning_rate": 5.068708363679249e-06,
|
3088 |
+
"loss": 0.4645,
|
3089 |
+
"step": 440
|
3090 |
+
},
|
3091 |
+
{
|
3092 |
+
"epoch": 4.281553398058253,
|
3093 |
+
"grad_norm": 0.10990810773913537,
|
3094 |
+
"learning_rate": 4.937277668863014e-06,
|
3095 |
+
"loss": 0.4647,
|
3096 |
+
"step": 441
|
3097 |
+
},
|
3098 |
+
{
|
3099 |
+
"epoch": 4.29126213592233,
|
3100 |
+
"grad_norm": 0.10982607372347657,
|
3101 |
+
"learning_rate": 4.807461264690157e-06,
|
3102 |
+
"loss": 0.465,
|
3103 |
+
"step": 442
|
3104 |
+
},
|
3105 |
+
{
|
3106 |
+
"epoch": 4.300970873786408,
|
3107 |
+
"grad_norm": 0.12022614154318102,
|
3108 |
+
"learning_rate": 4.67926512791868e-06,
|
3109 |
+
"loss": 0.4654,
|
3110 |
+
"step": 443
|
3111 |
+
},
|
3112 |
+
{
|
3113 |
+
"epoch": 4.310679611650485,
|
3114 |
+
"grad_norm": 0.10923349467562726,
|
3115 |
+
"learning_rate": 4.552695160709362e-06,
|
3116 |
+
"loss": 0.466,
|
3117 |
+
"step": 444
|
3118 |
+
},
|
3119 |
+
{
|
3120 |
+
"epoch": 4.320388349514563,
|
3121 |
+
"grad_norm": 0.10653612984033727,
|
3122 |
+
"learning_rate": 4.427757190353976e-06,
|
3123 |
+
"loss": 0.4684,
|
3124 |
+
"step": 445
|
3125 |
+
},
|
3126 |
+
{
|
3127 |
+
"epoch": 4.330097087378641,
|
3128 |
+
"grad_norm": 0.10730397458255132,
|
3129 |
+
"learning_rate": 4.304456969007049e-06,
|
3130 |
+
"loss": 0.4657,
|
3131 |
+
"step": 446
|
3132 |
+
},
|
3133 |
+
{
|
3134 |
+
"epoch": 4.339805825242719,
|
3135 |
+
"grad_norm": 0.1003612678568525,
|
3136 |
+
"learning_rate": 4.182800173420991e-06,
|
3137 |
+
"loss": 0.4649,
|
3138 |
+
"step": 447
|
3139 |
+
},
|
3140 |
+
{
|
3141 |
+
"epoch": 4.349514563106796,
|
3142 |
+
"grad_norm": 0.11090894330673261,
|
3143 |
+
"learning_rate": 4.06279240468475e-06,
|
3144 |
+
"loss": 0.4631,
|
3145 |
+
"step": 448
|
3146 |
+
},
|
3147 |
+
{
|
3148 |
+
"epoch": 4.359223300970874,
|
3149 |
+
"grad_norm": 0.1167058523870307,
|
3150 |
+
"learning_rate": 3.9444391879659604e-06,
|
3151 |
+
"loss": 0.4665,
|
3152 |
+
"step": 449
|
3153 |
+
},
|
3154 |
+
{
|
3155 |
+
"epoch": 4.368932038834951,
|
3156 |
+
"grad_norm": 0.10433976260435962,
|
3157 |
+
"learning_rate": 3.827745972256529e-06,
|
3158 |
+
"loss": 0.4659,
|
3159 |
+
"step": 450
|
3160 |
+
},
|
3161 |
+
{
|
3162 |
+
"epoch": 4.378640776699029,
|
3163 |
+
"grad_norm": 0.10011905909384386,
|
3164 |
+
"learning_rate": 3.7127181301217817e-06,
|
3165 |
+
"loss": 0.4641,
|
3166 |
+
"step": 451
|
3167 |
+
},
|
3168 |
+
{
|
3169 |
+
"epoch": 4.388349514563107,
|
3170 |
+
"grad_norm": 0.10655509962726566,
|
3171 |
+
"learning_rate": 3.599360957453102e-06,
|
3172 |
+
"loss": 0.467,
|
3173 |
+
"step": 452
|
3174 |
+
},
|
3175 |
+
{
|
3176 |
+
"epoch": 4.398058252427185,
|
3177 |
+
"grad_norm": 0.10938664062060237,
|
3178 |
+
"learning_rate": 3.487679673224129e-06,
|
3179 |
+
"loss": 0.4607,
|
3180 |
+
"step": 453
|
3181 |
+
},
|
3182 |
+
{
|
3183 |
+
"epoch": 4.407766990291262,
|
3184 |
+
"grad_norm": 0.10191086413678936,
|
3185 |
+
"learning_rate": 3.3776794192504412e-06,
|
3186 |
+
"loss": 0.4623,
|
3187 |
+
"step": 454
|
3188 |
+
},
|
3189 |
+
{
|
3190 |
+
"epoch": 4.41747572815534,
|
3191 |
+
"grad_norm": 0.09548255692426758,
|
3192 |
+
"learning_rate": 3.269365259952859e-06,
|
3193 |
+
"loss": 0.4626,
|
3194 |
+
"step": 455
|
3195 |
+
},
|
3196 |
+
{
|
3197 |
+
"epoch": 4.427184466019417,
|
3198 |
+
"grad_norm": 0.10109872749133958,
|
3199 |
+
"learning_rate": 3.1627421821242586e-06,
|
3200 |
+
"loss": 0.4627,
|
3201 |
+
"step": 456
|
3202 |
+
},
|
3203 |
+
{
|
3204 |
+
"epoch": 4.436893203883495,
|
3205 |
+
"grad_norm": 0.1003445240191667,
|
3206 |
+
"learning_rate": 3.0578150946999695e-06,
|
3207 |
+
"loss": 0.4637,
|
3208 |
+
"step": 457
|
3209 |
+
},
|
3210 |
+
{
|
3211 |
+
"epoch": 4.446601941747573,
|
3212 |
+
"grad_norm": 0.10094079472060287,
|
3213 |
+
"learning_rate": 2.954588828531817e-06,
|
3214 |
+
"loss": 0.4671,
|
3215 |
+
"step": 458
|
3216 |
+
},
|
3217 |
+
{
|
3218 |
+
"epoch": 4.456310679611651,
|
3219 |
+
"grad_norm": 0.09877743367390623,
|
3220 |
+
"learning_rate": 2.8530681361656422e-06,
|
3221 |
+
"loss": 0.4649,
|
3222 |
+
"step": 459
|
3223 |
+
},
|
3224 |
+
{
|
3225 |
+
"epoch": 4.466019417475728,
|
3226 |
+
"grad_norm": 0.09360681988866616,
|
3227 |
+
"learning_rate": 2.7532576916225395e-06,
|
3228 |
+
"loss": 0.4678,
|
3229 |
+
"step": 460
|
3230 |
+
},
|
3231 |
+
{
|
3232 |
+
"epoch": 4.475728155339806,
|
3233 |
+
"grad_norm": 0.09976388673239875,
|
3234 |
+
"learning_rate": 2.6551620901836515e-06,
|
3235 |
+
"loss": 0.4645,
|
3236 |
+
"step": 461
|
3237 |
+
},
|
3238 |
+
{
|
3239 |
+
"epoch": 4.485436893203883,
|
3240 |
+
"grad_norm": 0.0965607126962889,
|
3241 |
+
"learning_rate": 2.5587858481786086e-06,
|
3242 |
+
"loss": 0.4664,
|
3243 |
+
"step": 462
|
3244 |
+
},
|
3245 |
+
{
|
3246 |
+
"epoch": 4.495145631067961,
|
3247 |
+
"grad_norm": 0.09897406466321508,
|
3248 |
+
"learning_rate": 2.4641334027775755e-06,
|
3249 |
+
"loss": 0.467,
|
3250 |
+
"step": 463
|
3251 |
+
},
|
3252 |
+
{
|
3253 |
+
"epoch": 4.504854368932039,
|
3254 |
+
"grad_norm": 0.10396092442681994,
|
3255 |
+
"learning_rate": 2.371209111786987e-06,
|
3256 |
+
"loss": 0.468,
|
3257 |
+
"step": 464
|
3258 |
+
},
|
3259 |
+
{
|
3260 |
+
"epoch": 4.514563106796117,
|
3261 |
+
"grad_norm": 0.0948373657066504,
|
3262 |
+
"learning_rate": 2.280017253448916e-06,
|
3263 |
+
"loss": 0.4642,
|
3264 |
+
"step": 465
|
3265 |
+
},
|
3266 |
+
{
|
3267 |
+
"epoch": 4.524271844660194,
|
3268 |
+
"grad_norm": 0.09507655846542853,
|
3269 |
+
"learning_rate": 2.190562026244072e-06,
|
3270 |
+
"loss": 0.4669,
|
3271 |
+
"step": 466
|
3272 |
+
},
|
3273 |
+
{
|
3274 |
+
"epoch": 4.533980582524272,
|
3275 |
+
"grad_norm": 0.09659880423736872,
|
3276 |
+
"learning_rate": 2.102847548698539e-06,
|
3277 |
+
"loss": 0.4642,
|
3278 |
+
"step": 467
|
3279 |
+
},
|
3280 |
+
{
|
3281 |
+
"epoch": 4.543689320388349,
|
3282 |
+
"grad_norm": 0.09162921938645756,
|
3283 |
+
"learning_rate": 2.0168778591941242e-06,
|
3284 |
+
"loss": 0.4694,
|
3285 |
+
"step": 468
|
3286 |
+
},
|
3287 |
+
{
|
3288 |
+
"epoch": 4.553398058252427,
|
3289 |
+
"grad_norm": 0.08943623447949108,
|
3290 |
+
"learning_rate": 1.9326569157824736e-06,
|
3291 |
+
"loss": 0.4654,
|
3292 |
+
"step": 469
|
3293 |
+
},
|
3294 |
+
{
|
3295 |
+
"epoch": 4.563106796116505,
|
3296 |
+
"grad_norm": 0.0958809139961805,
|
3297 |
+
"learning_rate": 1.850188596002802e-06,
|
3298 |
+
"loss": 0.4638,
|
3299 |
+
"step": 470
|
3300 |
+
},
|
3301 |
+
{
|
3302 |
+
"epoch": 4.572815533980583,
|
3303 |
+
"grad_norm": 0.0928325035695772,
|
3304 |
+
"learning_rate": 1.7694766967033805e-06,
|
3305 |
+
"loss": 0.4699,
|
3306 |
+
"step": 471
|
3307 |
+
},
|
3308 |
+
{
|
3309 |
+
"epoch": 4.58252427184466,
|
3310 |
+
"grad_norm": 0.09326809079774372,
|
3311 |
+
"learning_rate": 1.6905249338667617e-06,
|
3312 |
+
"loss": 0.4654,
|
3313 |
+
"step": 472
|
3314 |
+
},
|
3315 |
+
{
|
3316 |
+
"epoch": 4.592233009708738,
|
3317 |
+
"grad_norm": 0.08894714059682778,
|
3318 |
+
"learning_rate": 1.613336942438637e-06,
|
3319 |
+
"loss": 0.4678,
|
3320 |
+
"step": 473
|
3321 |
+
},
|
3322 |
+
{
|
3323 |
+
"epoch": 4.601941747572815,
|
3324 |
+
"grad_norm": 0.09702666150761258,
|
3325 |
+
"learning_rate": 1.5379162761605427e-06,
|
3326 |
+
"loss": 0.4642,
|
3327 |
+
"step": 474
|
3328 |
+
},
|
3329 |
+
{
|
3330 |
+
"epoch": 4.611650485436893,
|
3331 |
+
"grad_norm": 0.0925613181306804,
|
3332 |
+
"learning_rate": 1.4642664074061962e-06,
|
3333 |
+
"loss": 0.463,
|
3334 |
+
"step": 475
|
3335 |
+
},
|
3336 |
+
{
|
3337 |
+
"epoch": 4.621359223300971,
|
3338 |
+
"grad_norm": 0.0934935785749229,
|
3339 |
+
"learning_rate": 1.3923907270216819e-06,
|
3340 |
+
"loss": 0.4636,
|
3341 |
+
"step": 476
|
3342 |
+
},
|
3343 |
+
{
|
3344 |
+
"epoch": 4.631067961165049,
|
3345 |
+
"grad_norm": 0.08692636915664595,
|
3346 |
+
"learning_rate": 1.3222925441692635e-06,
|
3347 |
+
"loss": 0.4656,
|
3348 |
+
"step": 477
|
3349 |
+
},
|
3350 |
+
{
|
3351 |
+
"epoch": 4.640776699029126,
|
3352 |
+
"grad_norm": 0.08592627332074977,
|
3353 |
+
"learning_rate": 1.2539750861751031e-06,
|
3354 |
+
"loss": 0.464,
|
3355 |
+
"step": 478
|
3356 |
+
},
|
3357 |
+
{
|
3358 |
+
"epoch": 4.650485436893204,
|
3359 |
+
"grad_norm": 0.08826705670035949,
|
3360 |
+
"learning_rate": 1.1874414983806283e-06,
|
3361 |
+
"loss": 0.4669,
|
3362 |
+
"step": 479
|
3363 |
+
},
|
3364 |
+
{
|
3365 |
+
"epoch": 4.660194174757281,
|
3366 |
+
"grad_norm": 0.08820160855473339,
|
3367 |
+
"learning_rate": 1.1226948439977314e-06,
|
3368 |
+
"loss": 0.4631,
|
3369 |
+
"step": 480
|
3370 |
+
},
|
3371 |
+
{
|
3372 |
+
"epoch": 4.669902912621359,
|
3373 |
+
"grad_norm": 0.09622872398963238,
|
3374 |
+
"learning_rate": 1.0597381039677646e-06,
|
3375 |
+
"loss": 0.4686,
|
3376 |
+
"step": 481
|
3377 |
+
},
|
3378 |
+
{
|
3379 |
+
"epoch": 4.679611650485437,
|
3380 |
+
"grad_norm": 0.08913032971687475,
|
3381 |
+
"learning_rate": 9.985741768242429e-07,
|
3382 |
+
"loss": 0.4647,
|
3383 |
+
"step": 482
|
3384 |
+
},
|
3385 |
+
{
|
3386 |
+
"epoch": 4.689320388349515,
|
3387 |
+
"grad_norm": 0.09028397610583812,
|
3388 |
+
"learning_rate": 9.392058785594504e-07,
|
3389 |
+
"loss": 0.4623,
|
3390 |
+
"step": 483
|
3391 |
+
},
|
3392 |
+
{
|
3393 |
+
"epoch": 4.699029126213592,
|
3394 |
+
"grad_norm": 0.09375615377365946,
|
3395 |
+
"learning_rate": 8.816359424947652e-07,
|
3396 |
+
"loss": 0.4629,
|
3397 |
+
"step": 484
|
3398 |
+
},
|
3399 |
+
{
|
3400 |
+
"epoch": 4.70873786407767,
|
3401 |
+
"grad_norm": 0.09058552987790736,
|
3402 |
+
"learning_rate": 8.258670191548135e-07,
|
3403 |
+
"loss": 0.4633,
|
3404 |
+
"step": 485
|
3405 |
+
},
|
3406 |
+
{
|
3407 |
+
"epoch": 4.718446601941747,
|
3408 |
+
"grad_norm": 0.08822401104082032,
|
3409 |
+
"learning_rate": 7.719016761454479e-07,
|
3410 |
+
"loss": 0.4668,
|
3411 |
+
"step": 486
|
3412 |
+
},
|
3413 |
+
{
|
3414 |
+
"epoch": 4.728155339805825,
|
3415 |
+
"grad_norm": 0.0922329791045878,
|
3416 |
+
"learning_rate": 7.197423980355344e-07,
|
3417 |
+
"loss": 0.4665,
|
3418 |
+
"step": 487
|
3419 |
+
},
|
3420 |
+
{
|
3421 |
+
"epoch": 4.737864077669903,
|
3422 |
+
"grad_norm": 0.08744767491607641,
|
3423 |
+
"learning_rate": 6.693915862425692e-07,
|
3424 |
+
"loss": 0.4657,
|
3425 |
+
"step": 488
|
3426 |
+
},
|
3427 |
+
{
|
3428 |
+
"epoch": 4.747572815533981,
|
3429 |
+
"grad_norm": 0.09683409497038403,
|
3430 |
+
"learning_rate": 6.20851558922091e-07,
|
3431 |
+
"loss": 0.4615,
|
3432 |
+
"step": 489
|
3433 |
+
},
|
3434 |
+
{
|
3435 |
+
"epoch": 4.757281553398058,
|
3436 |
+
"grad_norm": 0.08456750174183916,
|
3437 |
+
"learning_rate": 5.741245508609972e-07,
|
3438 |
+
"loss": 0.4649,
|
3439 |
+
"step": 490
|
3440 |
+
},
|
3441 |
+
{
|
3442 |
+
"epoch": 4.766990291262136,
|
3443 |
+
"grad_norm": 0.08773860548864587,
|
3444 |
+
"learning_rate": 5.292127133746005e-07,
|
3445 |
+
"loss": 0.4653,
|
3446 |
+
"step": 491
|
3447 |
+
},
|
3448 |
+
{
|
3449 |
+
"epoch": 4.776699029126213,
|
3450 |
+
"grad_norm": 0.0917555007099176,
|
3451 |
+
"learning_rate": 4.861181142076276e-07,
|
3452 |
+
"loss": 0.4656,
|
3453 |
+
"step": 492
|
3454 |
+
},
|
3455 |
+
{
|
3456 |
+
"epoch": 4.786407766990291,
|
3457 |
+
"grad_norm": 0.08845762170540458,
|
3458 |
+
"learning_rate": 4.448427374389974e-07,
|
3459 |
+
"loss": 0.4692,
|
3460 |
+
"step": 493
|
3461 |
+
},
|
3462 |
+
{
|
3463 |
+
"epoch": 4.796116504854369,
|
3464 |
+
"grad_norm": 0.08827155953592115,
|
3465 |
+
"learning_rate": 4.053884833904809e-07,
|
3466 |
+
"loss": 0.4662,
|
3467 |
+
"step": 494
|
3468 |
+
},
|
3469 |
+
{
|
3470 |
+
"epoch": 4.805825242718447,
|
3471 |
+
"grad_norm": 0.08804630863083254,
|
3472 |
+
"learning_rate": 3.677571685392023e-07,
|
3473 |
+
"loss": 0.4653,
|
3474 |
+
"step": 495
|
3475 |
+
},
|
3476 |
+
{
|
3477 |
+
"epoch": 4.815533980582524,
|
3478 |
+
"grad_norm": 0.09273872362298237,
|
3479 |
+
"learning_rate": 3.319505254340172e-07,
|
3480 |
+
"loss": 0.4674,
|
3481 |
+
"step": 496
|
3482 |
+
},
|
3483 |
+
{
|
3484 |
+
"epoch": 4.825242718446602,
|
3485 |
+
"grad_norm": 0.08937217299785047,
|
3486 |
+
"learning_rate": 2.9797020261574494e-07,
|
3487 |
+
"loss": 0.4675,
|
3488 |
+
"step": 497
|
3489 |
+
},
|
3490 |
+
{
|
3491 |
+
"epoch": 4.834951456310679,
|
3492 |
+
"grad_norm": 0.08269997531774376,
|
3493 |
+
"learning_rate": 2.6581776454126075e-07,
|
3494 |
+
"loss": 0.4683,
|
3495 |
+
"step": 498
|
3496 |
+
},
|
3497 |
+
{
|
3498 |
+
"epoch": 4.844660194174757,
|
3499 |
+
"grad_norm": 0.08810498069517812,
|
3500 |
+
"learning_rate": 2.3549469151149085e-07,
|
3501 |
+
"loss": 0.4658,
|
3502 |
+
"step": 499
|
3503 |
+
},
|
3504 |
+
{
|
3505 |
+
"epoch": 4.854368932038835,
|
3506 |
+
"grad_norm": 0.08729625064317513,
|
3507 |
+
"learning_rate": 2.0700237960322279e-07,
|
3508 |
+
"loss": 0.4696,
|
3509 |
+
"step": 500
|
3510 |
+
},
|
3511 |
+
{
|
3512 |
+
"epoch": 4.864077669902913,
|
3513 |
+
"grad_norm": 0.08416226889517374,
|
3514 |
+
"learning_rate": 1.803421406048589e-07,
|
3515 |
+
"loss": 0.4636,
|
3516 |
+
"step": 501
|
3517 |
+
},
|
3518 |
+
{
|
3519 |
+
"epoch": 4.87378640776699,
|
3520 |
+
"grad_norm": 0.08684248344197969,
|
3521 |
+
"learning_rate": 1.5551520195601577e-07,
|
3522 |
+
"loss": 0.4647,
|
3523 |
+
"step": 502
|
3524 |
+
},
|
3525 |
+
{
|
3526 |
+
"epoch": 4.883495145631068,
|
3527 |
+
"grad_norm": 0.08935580567045806,
|
3528 |
+
"learning_rate": 1.3252270669100953e-07,
|
3529 |
+
"loss": 0.4642,
|
3530 |
+
"step": 503
|
3531 |
+
},
|
3532 |
+
{
|
3533 |
+
"epoch": 4.893203883495145,
|
3534 |
+
"grad_norm": 0.0846095709740149,
|
3535 |
+
"learning_rate": 1.113657133862267e-07,
|
3536 |
+
"loss": 0.4626,
|
3537 |
+
"step": 504
|
3538 |
+
},
|
3539 |
+
{
|
3540 |
+
"epoch": 4.902912621359223,
|
3541 |
+
"grad_norm": 0.08317972288259207,
|
3542 |
+
"learning_rate": 9.204519611138995e-08,
|
3543 |
+
"loss": 0.4628,
|
3544 |
+
"step": 505
|
3545 |
+
},
|
3546 |
+
{
|
3547 |
+
"epoch": 4.9126213592233015,
|
3548 |
+
"grad_norm": 0.0845289066820202,
|
3549 |
+
"learning_rate": 7.45620443847228e-08,
|
3550 |
+
"loss": 0.4657,
|
3551 |
+
"step": 506
|
3552 |
+
},
|
3553 |
+
{
|
3554 |
+
"epoch": 4.922330097087379,
|
3555 |
+
"grad_norm": 0.08611674149409979,
|
3556 |
+
"learning_rate": 5.891706313197354e-08,
|
3557 |
+
"loss": 0.4643,
|
3558 |
+
"step": 507
|
3559 |
+
},
|
3560 |
+
{
|
3561 |
+
"epoch": 4.932038834951456,
|
3562 |
+
"grad_norm": 0.08707702652246953,
|
3563 |
+
"learning_rate": 4.511097264938258e-08,
|
3564 |
+
"loss": 0.4653,
|
3565 |
+
"step": 508
|
3566 |
+
},
|
3567 |
+
{
|
3568 |
+
"epoch": 4.941747572815534,
|
3569 |
+
"grad_norm": 0.08653800248757759,
|
3570 |
+
"learning_rate": 3.314440857049572e-08,
|
3571 |
+
"loss": 0.4679,
|
3572 |
+
"step": 509
|
3573 |
+
},
|
3574 |
+
{
|
3575 |
+
"epoch": 4.951456310679612,
|
3576 |
+
"grad_norm": 0.08635221643773852,
|
3577 |
+
"learning_rate": 2.3017921836916425e-08,
|
3578 |
+
"loss": 0.4627,
|
3579 |
+
"step": 510
|
3580 |
+
},
|
3581 |
+
{
|
3582 |
+
"epoch": 4.961165048543689,
|
3583 |
+
"grad_norm": 0.08744848791920268,
|
3584 |
+
"learning_rate": 1.4731978672939407e-08,
|
3585 |
+
"loss": 0.4664,
|
3586 |
+
"step": 511
|
3587 |
+
},
|
3588 |
+
{
|
3589 |
+
"epoch": 4.970873786407767,
|
3590 |
+
"grad_norm": 0.08802432305414154,
|
3591 |
+
"learning_rate": 8.286960564065639e-09,
|
3592 |
+
"loss": 0.4628,
|
3593 |
+
"step": 512
|
3594 |
+
},
|
3595 |
+
{
|
3596 |
+
"epoch": 4.980582524271845,
|
3597 |
+
"grad_norm": 0.08116167073862661,
|
3598 |
+
"learning_rate": 3.683164239469683e-09,
|
3599 |
+
"loss": 0.4637,
|
3600 |
+
"step": 513
|
3601 |
+
},
|
3602 |
+
{
|
3603 |
+
"epoch": 4.990291262135923,
|
3604 |
+
"grad_norm": 0.08686945147014129,
|
3605 |
+
"learning_rate": 9.208016583128754e-10,
|
3606 |
+
"loss": 0.4606,
|
3607 |
+
"step": 514
|
3608 |
+
},
|
3609 |
+
{
|
3610 |
+
"epoch": 5.0,
|
3611 |
+
"grad_norm": 0.08703184321831914,
|
3612 |
+
"learning_rate": 0.0,
|
3613 |
+
"loss": 0.4639,
|
3614 |
+
"step": 515
|
3615 |
+
},
|
3616 |
+
{
|
3617 |
+
"epoch": 5.0,
|
3618 |
+
"step": 515,
|
3619 |
+
"total_flos": 8639713262960640.0,
|
3620 |
+
"train_loss": 0.0,
|
3621 |
+
"train_runtime": 13.1123,
|
3622 |
+
"train_samples_per_second": 20070.162,
|
3623 |
+
"train_steps_per_second": 39.276
|
3624 |
+
}
|
3625 |
+
],
|
3626 |
+
"logging_steps": 1,
|
3627 |
+
"max_steps": 515,
|
3628 |
+
"num_input_tokens_seen": 0,
|
3629 |
+
"num_train_epochs": 5,
|
3630 |
+
"save_steps": 500,
|
3631 |
+
"stateful_callbacks": {
|
3632 |
+
"TrainerControl": {
|
3633 |
+
"args": {
|
3634 |
+
"should_epoch_stop": false,
|
3635 |
+
"should_evaluate": false,
|
3636 |
+
"should_log": false,
|
3637 |
+
"should_save": true,
|
3638 |
+
"should_training_stop": true
|
3639 |
+
},
|
3640 |
+
"attributes": {}
|
3641 |
+
}
|
3642 |
+
},
|
3643 |
+
"total_flos": 8639713262960640.0,
|
3644 |
+
"train_batch_size": 1,
|
3645 |
+
"trial_name": null,
|
3646 |
+
"trial_params": null
|
3647 |
+
}
|