Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1713 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1713 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 240,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.021052631578947368,
|
| 13 |
+
"grad_norm": 2.2153546810150146,
|
| 14 |
+
"learning_rate": 2.1276595744680852e-07,
|
| 15 |
+
"loss": 0.535,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.042105263157894736,
|
| 20 |
+
"grad_norm": 2.3668391704559326,
|
| 21 |
+
"learning_rate": 4.2553191489361704e-07,
|
| 22 |
+
"loss": 0.5907,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.06315789473684211,
|
| 27 |
+
"grad_norm": 2.0772104263305664,
|
| 28 |
+
"learning_rate": 6.382978723404255e-07,
|
| 29 |
+
"loss": 0.5031,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.08421052631578947,
|
| 34 |
+
"grad_norm": 2.2328884601593018,
|
| 35 |
+
"learning_rate": 8.510638297872341e-07,
|
| 36 |
+
"loss": 0.5602,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.10526315789473684,
|
| 41 |
+
"grad_norm": 2.2058968544006348,
|
| 42 |
+
"learning_rate": 1.0638297872340427e-06,
|
| 43 |
+
"loss": 0.5603,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.12631578947368421,
|
| 48 |
+
"grad_norm": 2.126004219055176,
|
| 49 |
+
"learning_rate": 1.276595744680851e-06,
|
| 50 |
+
"loss": 0.5517,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.14736842105263157,
|
| 55 |
+
"grad_norm": 2.0944392681121826,
|
| 56 |
+
"learning_rate": 1.4893617021276596e-06,
|
| 57 |
+
"loss": 0.512,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.16842105263157894,
|
| 62 |
+
"grad_norm": 1.9236035346984863,
|
| 63 |
+
"learning_rate": 1.7021276595744682e-06,
|
| 64 |
+
"loss": 0.4585,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.18947368421052632,
|
| 69 |
+
"grad_norm": 2.176151990890503,
|
| 70 |
+
"learning_rate": 1.9148936170212767e-06,
|
| 71 |
+
"loss": 0.5071,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.21052631578947367,
|
| 76 |
+
"grad_norm": 2.29136323928833,
|
| 77 |
+
"learning_rate": 2.1276595744680853e-06,
|
| 78 |
+
"loss": 0.5558,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.23157894736842105,
|
| 83 |
+
"grad_norm": 1.7607715129852295,
|
| 84 |
+
"learning_rate": 2.340425531914894e-06,
|
| 85 |
+
"loss": 0.4928,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.25263157894736843,
|
| 90 |
+
"grad_norm": 1.8522917032241821,
|
| 91 |
+
"learning_rate": 2.553191489361702e-06,
|
| 92 |
+
"loss": 0.5192,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.2736842105263158,
|
| 97 |
+
"grad_norm": 1.9679871797561646,
|
| 98 |
+
"learning_rate": 2.765957446808511e-06,
|
| 99 |
+
"loss": 0.5255,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.29473684210526313,
|
| 104 |
+
"grad_norm": 1.5592241287231445,
|
| 105 |
+
"learning_rate": 2.978723404255319e-06,
|
| 106 |
+
"loss": 0.5302,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.3157894736842105,
|
| 111 |
+
"grad_norm": 1.2780475616455078,
|
| 112 |
+
"learning_rate": 3.191489361702128e-06,
|
| 113 |
+
"loss": 0.4888,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.3368421052631579,
|
| 118 |
+
"grad_norm": 1.3540394306182861,
|
| 119 |
+
"learning_rate": 3.4042553191489363e-06,
|
| 120 |
+
"loss": 0.5174,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.35789473684210527,
|
| 125 |
+
"grad_norm": 1.111940860748291,
|
| 126 |
+
"learning_rate": 3.6170212765957453e-06,
|
| 127 |
+
"loss": 0.4876,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.37894736842105264,
|
| 132 |
+
"grad_norm": 0.9813674092292786,
|
| 133 |
+
"learning_rate": 3.8297872340425535e-06,
|
| 134 |
+
"loss": 0.4489,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.4,
|
| 139 |
+
"grad_norm": 0.6731572151184082,
|
| 140 |
+
"learning_rate": 4.042553191489362e-06,
|
| 141 |
+
"loss": 0.4572,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.42105263157894735,
|
| 146 |
+
"grad_norm": 0.6115604639053345,
|
| 147 |
+
"learning_rate": 4.255319148936171e-06,
|
| 148 |
+
"loss": 0.4152,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.4421052631578947,
|
| 153 |
+
"grad_norm": 0.7853341698646545,
|
| 154 |
+
"learning_rate": 4.468085106382979e-06,
|
| 155 |
+
"loss": 0.4825,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.4631578947368421,
|
| 160 |
+
"grad_norm": 0.7386396527290344,
|
| 161 |
+
"learning_rate": 4.680851063829788e-06,
|
| 162 |
+
"loss": 0.3861,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.4842105263157895,
|
| 167 |
+
"grad_norm": 0.822853684425354,
|
| 168 |
+
"learning_rate": 4.893617021276596e-06,
|
| 169 |
+
"loss": 0.41,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.5052631578947369,
|
| 174 |
+
"grad_norm": 0.8731152415275574,
|
| 175 |
+
"learning_rate": 5.106382978723404e-06,
|
| 176 |
+
"loss": 0.4646,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.5263157894736842,
|
| 181 |
+
"grad_norm": 0.7580763101577759,
|
| 182 |
+
"learning_rate": 5.319148936170213e-06,
|
| 183 |
+
"loss": 0.4295,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.5473684210526316,
|
| 188 |
+
"grad_norm": 0.6119412183761597,
|
| 189 |
+
"learning_rate": 5.531914893617022e-06,
|
| 190 |
+
"loss": 0.3869,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.5684210526315789,
|
| 195 |
+
"grad_norm": 0.6212654709815979,
|
| 196 |
+
"learning_rate": 5.744680851063831e-06,
|
| 197 |
+
"loss": 0.3993,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.5894736842105263,
|
| 202 |
+
"grad_norm": 0.6255398988723755,
|
| 203 |
+
"learning_rate": 5.957446808510638e-06,
|
| 204 |
+
"loss": 0.3891,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.6105263157894737,
|
| 209 |
+
"grad_norm": 0.6361572742462158,
|
| 210 |
+
"learning_rate": 6.170212765957447e-06,
|
| 211 |
+
"loss": 0.4479,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.631578947368421,
|
| 216 |
+
"grad_norm": 0.7150464653968811,
|
| 217 |
+
"learning_rate": 6.382978723404256e-06,
|
| 218 |
+
"loss": 0.4226,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.6526315789473685,
|
| 223 |
+
"grad_norm": 0.5256381034851074,
|
| 224 |
+
"learning_rate": 6.595744680851064e-06,
|
| 225 |
+
"loss": 0.3875,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.6736842105263158,
|
| 230 |
+
"grad_norm": 0.5503402352333069,
|
| 231 |
+
"learning_rate": 6.808510638297873e-06,
|
| 232 |
+
"loss": 0.4036,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.6947368421052632,
|
| 237 |
+
"grad_norm": 0.5972853899002075,
|
| 238 |
+
"learning_rate": 7.021276595744682e-06,
|
| 239 |
+
"loss": 0.3886,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.7157894736842105,
|
| 244 |
+
"grad_norm": 0.47243115305900574,
|
| 245 |
+
"learning_rate": 7.234042553191491e-06,
|
| 246 |
+
"loss": 0.4119,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.7368421052631579,
|
| 251 |
+
"grad_norm": 0.5615103244781494,
|
| 252 |
+
"learning_rate": 7.446808510638298e-06,
|
| 253 |
+
"loss": 0.4343,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.7578947368421053,
|
| 258 |
+
"grad_norm": 0.4591762125492096,
|
| 259 |
+
"learning_rate": 7.659574468085107e-06,
|
| 260 |
+
"loss": 0.3711,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.7789473684210526,
|
| 265 |
+
"grad_norm": 0.5434284210205078,
|
| 266 |
+
"learning_rate": 7.872340425531916e-06,
|
| 267 |
+
"loss": 0.4223,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.8,
|
| 272 |
+
"grad_norm": 0.5143300294876099,
|
| 273 |
+
"learning_rate": 8.085106382978723e-06,
|
| 274 |
+
"loss": 0.3782,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.8210526315789474,
|
| 279 |
+
"grad_norm": 0.4913320243358612,
|
| 280 |
+
"learning_rate": 8.297872340425532e-06,
|
| 281 |
+
"loss": 0.4153,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.8421052631578947,
|
| 286 |
+
"grad_norm": 0.5144403576850891,
|
| 287 |
+
"learning_rate": 8.510638297872341e-06,
|
| 288 |
+
"loss": 0.3851,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.8631578947368421,
|
| 293 |
+
"grad_norm": 0.4538310170173645,
|
| 294 |
+
"learning_rate": 8.72340425531915e-06,
|
| 295 |
+
"loss": 0.3657,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.8842105263157894,
|
| 300 |
+
"grad_norm": 0.4959196448326111,
|
| 301 |
+
"learning_rate": 8.936170212765958e-06,
|
| 302 |
+
"loss": 0.4603,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.9052631578947369,
|
| 307 |
+
"grad_norm": 0.4344879984855652,
|
| 308 |
+
"learning_rate": 9.148936170212767e-06,
|
| 309 |
+
"loss": 0.3581,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.9263157894736842,
|
| 314 |
+
"grad_norm": 0.39444926381111145,
|
| 315 |
+
"learning_rate": 9.361702127659576e-06,
|
| 316 |
+
"loss": 0.3427,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.9473684210526315,
|
| 321 |
+
"grad_norm": 0.548050045967102,
|
| 322 |
+
"learning_rate": 9.574468085106385e-06,
|
| 323 |
+
"loss": 0.4282,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.968421052631579,
|
| 328 |
+
"grad_norm": 0.5126370787620544,
|
| 329 |
+
"learning_rate": 9.787234042553192e-06,
|
| 330 |
+
"loss": 0.4144,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.9894736842105263,
|
| 335 |
+
"grad_norm": 0.45015019178390503,
|
| 336 |
+
"learning_rate": 1e-05,
|
| 337 |
+
"loss": 0.425,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.0,
|
| 342 |
+
"grad_norm": 0.45015019178390503,
|
| 343 |
+
"learning_rate": 9.999862102299874e-06,
|
| 344 |
+
"loss": 0.3668,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.0210526315789474,
|
| 349 |
+
"grad_norm": 0.6312311887741089,
|
| 350 |
+
"learning_rate": 9.999448416805802e-06,
|
| 351 |
+
"loss": 0.3728,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.0421052631578946,
|
| 356 |
+
"grad_norm": 0.44266408681869507,
|
| 357 |
+
"learning_rate": 9.998758966336296e-06,
|
| 358 |
+
"loss": 0.3384,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.063157894736842,
|
| 363 |
+
"grad_norm": 0.4919297397136688,
|
| 364 |
+
"learning_rate": 9.99779378892081e-06,
|
| 365 |
+
"loss": 0.3753,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.0842105263157895,
|
| 370 |
+
"grad_norm": 0.41351792216300964,
|
| 371 |
+
"learning_rate": 9.996552937797646e-06,
|
| 372 |
+
"loss": 0.3159,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.1052631578947367,
|
| 377 |
+
"grad_norm": 0.533478856086731,
|
| 378 |
+
"learning_rate": 9.995036481411005e-06,
|
| 379 |
+
"loss": 0.3577,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.1263157894736842,
|
| 384 |
+
"grad_norm": 0.4766313433647156,
|
| 385 |
+
"learning_rate": 9.993244503407227e-06,
|
| 386 |
+
"loss": 0.2922,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.1473684210526316,
|
| 391 |
+
"grad_norm": 0.4464484453201294,
|
| 392 |
+
"learning_rate": 9.991177102630173e-06,
|
| 393 |
+
"loss": 0.3196,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.168421052631579,
|
| 398 |
+
"grad_norm": 0.502562403678894,
|
| 399 |
+
"learning_rate": 9.988834393115768e-06,
|
| 400 |
+
"loss": 0.369,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.1894736842105262,
|
| 405 |
+
"grad_norm": 0.5081046223640442,
|
| 406 |
+
"learning_rate": 9.986216504085709e-06,
|
| 407 |
+
"loss": 0.3079,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.2105263157894737,
|
| 412 |
+
"grad_norm": 0.47032320499420166,
|
| 413 |
+
"learning_rate": 9.983323579940351e-06,
|
| 414 |
+
"loss": 0.354,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.231578947368421,
|
| 419 |
+
"grad_norm": 0.4799109697341919,
|
| 420 |
+
"learning_rate": 9.980155780250728e-06,
|
| 421 |
+
"loss": 0.3072,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.2526315789473683,
|
| 426 |
+
"grad_norm": 0.4440138638019562,
|
| 427 |
+
"learning_rate": 9.976713279749754e-06,
|
| 428 |
+
"loss": 0.3567,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.2736842105263158,
|
| 433 |
+
"grad_norm": 0.559519350528717,
|
| 434 |
+
"learning_rate": 9.972996268322594e-06,
|
| 435 |
+
"loss": 0.3614,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.2947368421052632,
|
| 440 |
+
"grad_norm": 0.3834027647972107,
|
| 441 |
+
"learning_rate": 9.969004950996175e-06,
|
| 442 |
+
"loss": 0.3231,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.3157894736842106,
|
| 447 |
+
"grad_norm": 0.5550243258476257,
|
| 448 |
+
"learning_rate": 9.964739547927892e-06,
|
| 449 |
+
"loss": 0.32,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.3368421052631578,
|
| 454 |
+
"grad_norm": 0.490479439496994,
|
| 455 |
+
"learning_rate": 9.96020029439345e-06,
|
| 456 |
+
"loss": 0.3793,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.3578947368421053,
|
| 461 |
+
"grad_norm": 0.4827335774898529,
|
| 462 |
+
"learning_rate": 9.955387440773902e-06,
|
| 463 |
+
"loss": 0.3257,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.3789473684210527,
|
| 468 |
+
"grad_norm": 0.42826348543167114,
|
| 469 |
+
"learning_rate": 9.950301252541824e-06,
|
| 470 |
+
"loss": 0.3631,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 1.4,
|
| 475 |
+
"grad_norm": 0.6259158849716187,
|
| 476 |
+
"learning_rate": 9.944942010246681e-06,
|
| 477 |
+
"loss": 0.2984,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 1.4210526315789473,
|
| 482 |
+
"grad_norm": 0.4576846659183502,
|
| 483 |
+
"learning_rate": 9.939310009499348e-06,
|
| 484 |
+
"loss": 0.3344,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.4421052631578948,
|
| 489 |
+
"grad_norm": 0.4172138571739197,
|
| 490 |
+
"learning_rate": 9.933405560955805e-06,
|
| 491 |
+
"loss": 0.3435,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 1.4631578947368422,
|
| 496 |
+
"grad_norm": 0.4487617611885071,
|
| 497 |
+
"learning_rate": 9.9272289903e-06,
|
| 498 |
+
"loss": 0.3315,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 1.4842105263157894,
|
| 503 |
+
"grad_norm": 0.4363557696342468,
|
| 504 |
+
"learning_rate": 9.92078063822589e-06,
|
| 505 |
+
"loss": 0.311,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.5052631578947369,
|
| 510 |
+
"grad_norm": 0.4071301817893982,
|
| 511 |
+
"learning_rate": 9.914060860418644e-06,
|
| 512 |
+
"loss": 0.2852,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 1.526315789473684,
|
| 517 |
+
"grad_norm": 0.4272218644618988,
|
| 518 |
+
"learning_rate": 9.907070027535022e-06,
|
| 519 |
+
"loss": 0.362,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.5473684210526315,
|
| 524 |
+
"grad_norm": 0.4792329668998718,
|
| 525 |
+
"learning_rate": 9.899808525182935e-06,
|
| 526 |
+
"loss": 0.3406,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.568421052631579,
|
| 531 |
+
"grad_norm": 0.49053433537483215,
|
| 532 |
+
"learning_rate": 9.892276753900173e-06,
|
| 533 |
+
"loss": 0.3356,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 1.5894736842105264,
|
| 538 |
+
"grad_norm": 0.43284040689468384,
|
| 539 |
+
"learning_rate": 9.884475129132312e-06,
|
| 540 |
+
"loss": 0.3358,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 1.6105263157894738,
|
| 545 |
+
"grad_norm": 0.4686710238456726,
|
| 546 |
+
"learning_rate": 9.876404081209796e-06,
|
| 547 |
+
"loss": 0.3104,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.631578947368421,
|
| 552 |
+
"grad_norm": 0.544070303440094,
|
| 553 |
+
"learning_rate": 9.868064055324204e-06,
|
| 554 |
+
"loss": 0.3503,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 1.6526315789473685,
|
| 559 |
+
"grad_norm": 0.4395733177661896,
|
| 560 |
+
"learning_rate": 9.859455511503691e-06,
|
| 561 |
+
"loss": 0.3194,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 1.6736842105263157,
|
| 566 |
+
"grad_norm": 0.48770004510879517,
|
| 567 |
+
"learning_rate": 9.850578924587614e-06,
|
| 568 |
+
"loss": 0.3339,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.694736842105263,
|
| 573 |
+
"grad_norm": 0.458671897649765,
|
| 574 |
+
"learning_rate": 9.841434784200341e-06,
|
| 575 |
+
"loss": 0.3219,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 1.7157894736842105,
|
| 580 |
+
"grad_norm": 0.4801121950149536,
|
| 581 |
+
"learning_rate": 9.832023594724248e-06,
|
| 582 |
+
"loss": 0.3479,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 1.736842105263158,
|
| 587 |
+
"grad_norm": 0.3471762239933014,
|
| 588 |
+
"learning_rate": 9.822345875271884e-06,
|
| 589 |
+
"loss": 0.3001,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.7578947368421054,
|
| 594 |
+
"grad_norm": 0.4898357391357422,
|
| 595 |
+
"learning_rate": 9.812402159657352e-06,
|
| 596 |
+
"loss": 0.347,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 1.7789473684210526,
|
| 601 |
+
"grad_norm": 0.4187287986278534,
|
| 602 |
+
"learning_rate": 9.802192996366859e-06,
|
| 603 |
+
"loss": 0.3679,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.8,
|
| 608 |
+
"grad_norm": 0.43426603078842163,
|
| 609 |
+
"learning_rate": 9.791718948528457e-06,
|
| 610 |
+
"loss": 0.3203,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 1.8210526315789473,
|
| 615 |
+
"grad_norm": 0.39182379841804504,
|
| 616 |
+
"learning_rate": 9.780980593880993e-06,
|
| 617 |
+
"loss": 0.3217,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 1.8421052631578947,
|
| 622 |
+
"grad_norm": 0.5164268016815186,
|
| 623 |
+
"learning_rate": 9.76997852474223e-06,
|
| 624 |
+
"loss": 0.3677,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 1.8631578947368421,
|
| 629 |
+
"grad_norm": 0.4525752365589142,
|
| 630 |
+
"learning_rate": 9.758713347976179e-06,
|
| 631 |
+
"loss": 0.3482,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.8842105263157896,
|
| 636 |
+
"grad_norm": 0.48504945635795593,
|
| 637 |
+
"learning_rate": 9.747185684959626e-06,
|
| 638 |
+
"loss": 0.2964,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 1.905263157894737,
|
| 643 |
+
"grad_norm": 0.4417821168899536,
|
| 644 |
+
"learning_rate": 9.735396171547859e-06,
|
| 645 |
+
"loss": 0.3399,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 1.9263157894736842,
|
| 650 |
+
"grad_norm": 0.45080071687698364,
|
| 651 |
+
"learning_rate": 9.723345458039595e-06,
|
| 652 |
+
"loss": 0.2995,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 1.9473684210526314,
|
| 657 |
+
"grad_norm": 0.36195772886276245,
|
| 658 |
+
"learning_rate": 9.711034209141102e-06,
|
| 659 |
+
"loss": 0.2853,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 1.9684210526315788,
|
| 664 |
+
"grad_norm": 0.43885594606399536,
|
| 665 |
+
"learning_rate": 9.698463103929542e-06,
|
| 666 |
+
"loss": 0.3519,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 1.9894736842105263,
|
| 671 |
+
"grad_norm": 0.4610655903816223,
|
| 672 |
+
"learning_rate": 9.685632835815519e-06,
|
| 673 |
+
"loss": 0.2969,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.0,
|
| 678 |
+
"grad_norm": 0.9271234273910522,
|
| 679 |
+
"learning_rate": 9.672544112504813e-06,
|
| 680 |
+
"loss": 0.3078,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 2.0210526315789474,
|
| 685 |
+
"grad_norm": 0.5120906829833984,
|
| 686 |
+
"learning_rate": 9.659197655959364e-06,
|
| 687 |
+
"loss": 0.2576,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 2.042105263157895,
|
| 692 |
+
"grad_norm": 0.44506344199180603,
|
| 693 |
+
"learning_rate": 9.645594202357438e-06,
|
| 694 |
+
"loss": 0.2279,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 2.0631578947368423,
|
| 699 |
+
"grad_norm": 0.517287015914917,
|
| 700 |
+
"learning_rate": 9.63173450205302e-06,
|
| 701 |
+
"loss": 0.2501,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 2.0842105263157893,
|
| 706 |
+
"grad_norm": 0.686529278755188,
|
| 707 |
+
"learning_rate": 9.617619319534427e-06,
|
| 708 |
+
"loss": 0.2252,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 2.1052631578947367,
|
| 713 |
+
"grad_norm": 0.44996634125709534,
|
| 714 |
+
"learning_rate": 9.603249433382145e-06,
|
| 715 |
+
"loss": 0.2325,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 2.126315789473684,
|
| 720 |
+
"grad_norm": 0.4678879678249359,
|
| 721 |
+
"learning_rate": 9.588625636225871e-06,
|
| 722 |
+
"loss": 0.2575,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 2.1473684210526316,
|
| 727 |
+
"grad_norm": 0.5605614185333252,
|
| 728 |
+
"learning_rate": 9.573748734700806e-06,
|
| 729 |
+
"loss": 0.2226,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 2.168421052631579,
|
| 734 |
+
"grad_norm": 0.5222639441490173,
|
| 735 |
+
"learning_rate": 9.558619549403148e-06,
|
| 736 |
+
"loss": 0.2276,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 2.1894736842105265,
|
| 741 |
+
"grad_norm": 0.4625437557697296,
|
| 742 |
+
"learning_rate": 9.543238914844844e-06,
|
| 743 |
+
"loss": 0.2264,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 2.2105263157894735,
|
| 748 |
+
"grad_norm": 0.3863196074962616,
|
| 749 |
+
"learning_rate": 9.527607679407545e-06,
|
| 750 |
+
"loss": 0.2285,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 2.231578947368421,
|
| 755 |
+
"grad_norm": 0.4566896855831146,
|
| 756 |
+
"learning_rate": 9.51172670529582e-06,
|
| 757 |
+
"loss": 0.2424,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 2.2526315789473683,
|
| 762 |
+
"grad_norm": 0.576726496219635,
|
| 763 |
+
"learning_rate": 9.495596868489588e-06,
|
| 764 |
+
"loss": 0.2801,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 2.2736842105263158,
|
| 769 |
+
"grad_norm": 0.5350817441940308,
|
| 770 |
+
"learning_rate": 9.47921905869581e-06,
|
| 771 |
+
"loss": 0.2382,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 2.294736842105263,
|
| 776 |
+
"grad_norm": 0.4740808308124542,
|
| 777 |
+
"learning_rate": 9.462594179299408e-06,
|
| 778 |
+
"loss": 0.2262,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 2.3157894736842106,
|
| 783 |
+
"grad_norm": 0.539569616317749,
|
| 784 |
+
"learning_rate": 9.445723147313434e-06,
|
| 785 |
+
"loss": 0.2478,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 2.336842105263158,
|
| 790 |
+
"grad_norm": 0.4873138964176178,
|
| 791 |
+
"learning_rate": 9.428606893328493e-06,
|
| 792 |
+
"loss": 0.2414,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 2.3578947368421055,
|
| 797 |
+
"grad_norm": 0.4593484699726105,
|
| 798 |
+
"learning_rate": 9.41124636146141e-06,
|
| 799 |
+
"loss": 0.2611,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 2.3789473684210525,
|
| 804 |
+
"grad_norm": 0.4127882421016693,
|
| 805 |
+
"learning_rate": 9.39364250930315e-06,
|
| 806 |
+
"loss": 0.2402,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 2.4,
|
| 811 |
+
"grad_norm": 0.4128904342651367,
|
| 812 |
+
"learning_rate": 9.375796307866003e-06,
|
| 813 |
+
"loss": 0.2037,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 2.4210526315789473,
|
| 818 |
+
"grad_norm": 0.49382951855659485,
|
| 819 |
+
"learning_rate": 9.357708741530025e-06,
|
| 820 |
+
"loss": 0.2363,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 2.442105263157895,
|
| 825 |
+
"grad_norm": 0.55715012550354,
|
| 826 |
+
"learning_rate": 9.339380807988734e-06,
|
| 827 |
+
"loss": 0.2791,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 2.463157894736842,
|
| 832 |
+
"grad_norm": 0.4708317816257477,
|
| 833 |
+
"learning_rate": 9.320813518194084e-06,
|
| 834 |
+
"loss": 0.2814,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 2.4842105263157896,
|
| 839 |
+
"grad_norm": 0.4327724277973175,
|
| 840 |
+
"learning_rate": 9.302007896300697e-06,
|
| 841 |
+
"loss": 0.234,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 2.5052631578947366,
|
| 846 |
+
"grad_norm": 0.44202759861946106,
|
| 847 |
+
"learning_rate": 9.28296497960938e-06,
|
| 848 |
+
"loss": 0.2239,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 2.526315789473684,
|
| 853 |
+
"grad_norm": 0.44984355568885803,
|
| 854 |
+
"learning_rate": 9.263685818509895e-06,
|
| 855 |
+
"loss": 0.2324,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 2.5473684210526315,
|
| 860 |
+
"grad_norm": 0.48632219433784485,
|
| 861 |
+
"learning_rate": 9.244171476423037e-06,
|
| 862 |
+
"loss": 0.2578,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 2.568421052631579,
|
| 867 |
+
"grad_norm": 0.42317625880241394,
|
| 868 |
+
"learning_rate": 9.22442302974196e-06,
|
| 869 |
+
"loss": 0.2181,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 2.5894736842105264,
|
| 874 |
+
"grad_norm": 0.5044139623641968,
|
| 875 |
+
"learning_rate": 9.204441567772817e-06,
|
| 876 |
+
"loss": 0.2412,
|
| 877 |
+
"step": 124
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 2.610526315789474,
|
| 881 |
+
"grad_norm": 0.46015551686286926,
|
| 882 |
+
"learning_rate": 9.184228192674667e-06,
|
| 883 |
+
"loss": 0.2701,
|
| 884 |
+
"step": 125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 2.6315789473684212,
|
| 888 |
+
"grad_norm": 0.4417663514614105,
|
| 889 |
+
"learning_rate": 9.163784019398686e-06,
|
| 890 |
+
"loss": 0.2578,
|
| 891 |
+
"step": 126
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 2.6526315789473687,
|
| 895 |
+
"grad_norm": 0.4231509268283844,
|
| 896 |
+
"learning_rate": 9.143110175626662e-06,
|
| 897 |
+
"loss": 0.2084,
|
| 898 |
+
"step": 127
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 2.6736842105263157,
|
| 902 |
+
"grad_norm": 0.5033539533615112,
|
| 903 |
+
"learning_rate": 9.122207801708802e-06,
|
| 904 |
+
"loss": 0.2696,
|
| 905 |
+
"step": 128
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 2.694736842105263,
|
| 909 |
+
"grad_norm": 0.3786817789077759,
|
| 910 |
+
"learning_rate": 9.101078050600823e-06,
|
| 911 |
+
"loss": 0.1774,
|
| 912 |
+
"step": 129
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 2.7157894736842105,
|
| 916 |
+
"grad_norm": 0.5628529787063599,
|
| 917 |
+
"learning_rate": 9.079722087800353e-06,
|
| 918 |
+
"loss": 0.2408,
|
| 919 |
+
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 2.736842105263158,
|
| 923 |
+
"grad_norm": 0.4625787138938904,
|
| 924 |
+
"learning_rate": 9.058141091282656e-06,
|
| 925 |
+
"loss": 0.2434,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 2.7578947368421054,
|
| 930 |
+
"grad_norm": 0.44364896416664124,
|
| 931 |
+
"learning_rate": 9.036336251435647e-06,
|
| 932 |
+
"loss": 0.2492,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 2.7789473684210524,
|
| 937 |
+
"grad_norm": 0.4227466881275177,
|
| 938 |
+
"learning_rate": 9.014308770994235e-06,
|
| 939 |
+
"loss": 0.2362,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 2.8,
|
| 944 |
+
"grad_norm": 0.41738563776016235,
|
| 945 |
+
"learning_rate": 8.992059864973972e-06,
|
| 946 |
+
"loss": 0.2163,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 2.8210526315789473,
|
| 951 |
+
"grad_norm": 0.505607008934021,
|
| 952 |
+
"learning_rate": 8.96959076060405e-06,
|
| 953 |
+
"loss": 0.2502,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 2.8421052631578947,
|
| 958 |
+
"grad_norm": 0.40832582116127014,
|
| 959 |
+
"learning_rate": 8.946902697259593e-06,
|
| 960 |
+
"loss": 0.2257,
|
| 961 |
+
"step": 136
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 2.863157894736842,
|
| 965 |
+
"grad_norm": 0.508179247379303,
|
| 966 |
+
"learning_rate": 8.923996926393306e-06,
|
| 967 |
+
"loss": 0.2877,
|
| 968 |
+
"step": 137
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 2.8842105263157896,
|
| 972 |
+
"grad_norm": 0.47631341218948364,
|
| 973 |
+
"learning_rate": 8.900874711466436e-06,
|
| 974 |
+
"loss": 0.2482,
|
| 975 |
+
"step": 138
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 2.905263157894737,
|
| 979 |
+
"grad_norm": 0.5183837413787842,
|
| 980 |
+
"learning_rate": 8.877537327879087e-06,
|
| 981 |
+
"loss": 0.2352,
|
| 982 |
+
"step": 139
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 2.9263157894736844,
|
| 986 |
+
"grad_norm": 0.39236199855804443,
|
| 987 |
+
"learning_rate": 8.853986062899869e-06,
|
| 988 |
+
"loss": 0.2299,
|
| 989 |
+
"step": 140
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 2.9473684210526314,
|
| 993 |
+
"grad_norm": 0.4558231830596924,
|
| 994 |
+
"learning_rate": 8.83022221559489e-06,
|
| 995 |
+
"loss": 0.263,
|
| 996 |
+
"step": 141
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 2.968421052631579,
|
| 1000 |
+
"grad_norm": 0.5461449027061462,
|
| 1001 |
+
"learning_rate": 8.80624709675611e-06,
|
| 1002 |
+
"loss": 0.2606,
|
| 1003 |
+
"step": 142
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 2.9894736842105263,
|
| 1007 |
+
"grad_norm": 0.45027104020118713,
|
| 1008 |
+
"learning_rate": 8.782062028829028e-06,
|
| 1009 |
+
"loss": 0.2299,
|
| 1010 |
+
"step": 143
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 3.0,
|
| 1014 |
+
"grad_norm": 0.45027104020118713,
|
| 1015 |
+
"learning_rate": 8.757668345839739e-06,
|
| 1016 |
+
"loss": 0.2207,
|
| 1017 |
+
"step": 144
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 3.0210526315789474,
|
| 1021 |
+
"grad_norm": 0.8024611473083496,
|
| 1022 |
+
"learning_rate": 8.733067393321354e-06,
|
| 1023 |
+
"loss": 0.1749,
|
| 1024 |
+
"step": 145
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 3.042105263157895,
|
| 1028 |
+
"grad_norm": 0.6112058758735657,
|
| 1029 |
+
"learning_rate": 8.708260528239788e-06,
|
| 1030 |
+
"loss": 0.1303,
|
| 1031 |
+
"step": 146
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 3.0631578947368423,
|
| 1035 |
+
"grad_norm": 0.38725703954696655,
|
| 1036 |
+
"learning_rate": 8.683249118918895e-06,
|
| 1037 |
+
"loss": 0.1502,
|
| 1038 |
+
"step": 147
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 3.0842105263157893,
|
| 1042 |
+
"grad_norm": 0.4709077477455139,
|
| 1043 |
+
"learning_rate": 8.658034544965003e-06,
|
| 1044 |
+
"loss": 0.1561,
|
| 1045 |
+
"step": 148
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 3.1052631578947367,
|
| 1049 |
+
"grad_norm": 0.5597317814826965,
|
| 1050 |
+
"learning_rate": 8.632618197190817e-06,
|
| 1051 |
+
"loss": 0.1799,
|
| 1052 |
+
"step": 149
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 3.126315789473684,
|
| 1056 |
+
"grad_norm": 0.5309601426124573,
|
| 1057 |
+
"learning_rate": 8.607001477538697e-06,
|
| 1058 |
+
"loss": 0.1676,
|
| 1059 |
+
"step": 150
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 3.1473684210526316,
|
| 1063 |
+
"grad_norm": 0.5718960762023926,
|
| 1064 |
+
"learning_rate": 8.581185799003334e-06,
|
| 1065 |
+
"loss": 0.1581,
|
| 1066 |
+
"step": 151
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 3.168421052631579,
|
| 1070 |
+
"grad_norm": 0.46150919795036316,
|
| 1071 |
+
"learning_rate": 8.555172585553804e-06,
|
| 1072 |
+
"loss": 0.1324,
|
| 1073 |
+
"step": 152
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 3.1894736842105265,
|
| 1077 |
+
"grad_norm": 0.6925739049911499,
|
| 1078 |
+
"learning_rate": 8.528963272055036e-06,
|
| 1079 |
+
"loss": 0.1643,
|
| 1080 |
+
"step": 153
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 3.2105263157894735,
|
| 1084 |
+
"grad_norm": 0.5450615882873535,
|
| 1085 |
+
"learning_rate": 8.502559304188644e-06,
|
| 1086 |
+
"loss": 0.1436,
|
| 1087 |
+
"step": 154
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 3.231578947368421,
|
| 1091 |
+
"grad_norm": 0.5580042004585266,
|
| 1092 |
+
"learning_rate": 8.475962138373212e-06,
|
| 1093 |
+
"loss": 0.1421,
|
| 1094 |
+
"step": 155
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 3.2526315789473683,
|
| 1098 |
+
"grad_norm": 0.4670768976211548,
|
| 1099 |
+
"learning_rate": 8.449173241683934e-06,
|
| 1100 |
+
"loss": 0.1669,
|
| 1101 |
+
"step": 156
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 3.2736842105263158,
|
| 1105 |
+
"grad_norm": 0.5521131753921509,
|
| 1106 |
+
"learning_rate": 8.422194091771709e-06,
|
| 1107 |
+
"loss": 0.1772,
|
| 1108 |
+
"step": 157
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 3.294736842105263,
|
| 1112 |
+
"grad_norm": 0.538808286190033,
|
| 1113 |
+
"learning_rate": 8.395026176781627e-06,
|
| 1114 |
+
"loss": 0.1566,
|
| 1115 |
+
"step": 158
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 3.3157894736842106,
|
| 1119 |
+
"grad_norm": 0.47581419348716736,
|
| 1120 |
+
"learning_rate": 8.367670995270883e-06,
|
| 1121 |
+
"loss": 0.1671,
|
| 1122 |
+
"step": 159
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 3.336842105263158,
|
| 1126 |
+
"grad_norm": 0.6213884949684143,
|
| 1127 |
+
"learning_rate": 8.340130056126126e-06,
|
| 1128 |
+
"loss": 0.1602,
|
| 1129 |
+
"step": 160
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 3.3578947368421055,
|
| 1133 |
+
"grad_norm": 0.5021377801895142,
|
| 1134 |
+
"learning_rate": 8.312404878480222e-06,
|
| 1135 |
+
"loss": 0.1385,
|
| 1136 |
+
"step": 161
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 3.3789473684210525,
|
| 1140 |
+
"grad_norm": 0.38781818747520447,
|
| 1141 |
+
"learning_rate": 8.284496991628465e-06,
|
| 1142 |
+
"loss": 0.1508,
|
| 1143 |
+
"step": 162
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 3.4,
|
| 1147 |
+
"grad_norm": 0.4906439483165741,
|
| 1148 |
+
"learning_rate": 8.25640793494422e-06,
|
| 1149 |
+
"loss": 0.1283,
|
| 1150 |
+
"step": 163
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 3.4210526315789473,
|
| 1154 |
+
"grad_norm": 0.4907456934452057,
|
| 1155 |
+
"learning_rate": 8.228139257794012e-06,
|
| 1156 |
+
"loss": 0.1395,
|
| 1157 |
+
"step": 164
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 3.442105263157895,
|
| 1161 |
+
"grad_norm": 0.39403727650642395,
|
| 1162 |
+
"learning_rate": 8.19969251945207e-06,
|
| 1163 |
+
"loss": 0.1736,
|
| 1164 |
+
"step": 165
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 3.463157894736842,
|
| 1168 |
+
"grad_norm": 0.4710124433040619,
|
| 1169 |
+
"learning_rate": 8.171069289014307e-06,
|
| 1170 |
+
"loss": 0.1452,
|
| 1171 |
+
"step": 166
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 3.4842105263157896,
|
| 1175 |
+
"grad_norm": 0.40228015184402466,
|
| 1176 |
+
"learning_rate": 8.142271145311784e-06,
|
| 1177 |
+
"loss": 0.1286,
|
| 1178 |
+
"step": 167
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 3.5052631578947366,
|
| 1182 |
+
"grad_norm": 0.49649837613105774,
|
| 1183 |
+
"learning_rate": 8.113299676823614e-06,
|
| 1184 |
+
"loss": 0.1541,
|
| 1185 |
+
"step": 168
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 3.526315789473684,
|
| 1189 |
+
"grad_norm": 0.4689042866230011,
|
| 1190 |
+
"learning_rate": 8.08415648158935e-06,
|
| 1191 |
+
"loss": 0.154,
|
| 1192 |
+
"step": 169
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 3.5473684210526315,
|
| 1196 |
+
"grad_norm": 0.364935964345932,
|
| 1197 |
+
"learning_rate": 8.054843167120827e-06,
|
| 1198 |
+
"loss": 0.1775,
|
| 1199 |
+
"step": 170
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 3.568421052631579,
|
| 1203 |
+
"grad_norm": 0.4018644094467163,
|
| 1204 |
+
"learning_rate": 8.025361350313506e-06,
|
| 1205 |
+
"loss": 0.1401,
|
| 1206 |
+
"step": 171
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 3.5894736842105264,
|
| 1210 |
+
"grad_norm": 0.415256530046463,
|
| 1211 |
+
"learning_rate": 7.99571265735728e-06,
|
| 1212 |
+
"loss": 0.1655,
|
| 1213 |
+
"step": 172
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 3.610526315789474,
|
| 1217 |
+
"grad_norm": 0.45749431848526,
|
| 1218 |
+
"learning_rate": 7.965898723646777e-06,
|
| 1219 |
+
"loss": 0.1452,
|
| 1220 |
+
"step": 173
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 3.6315789473684212,
|
| 1224 |
+
"grad_norm": 0.4208436608314514,
|
| 1225 |
+
"learning_rate": 7.935921193691153e-06,
|
| 1226 |
+
"loss": 0.1729,
|
| 1227 |
+
"step": 174
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 3.6526315789473687,
|
| 1231 |
+
"grad_norm": 0.4315619170665741,
|
| 1232 |
+
"learning_rate": 7.905781721023384e-06,
|
| 1233 |
+
"loss": 0.1609,
|
| 1234 |
+
"step": 175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 3.6736842105263157,
|
| 1238 |
+
"grad_norm": 0.3571399748325348,
|
| 1239 |
+
"learning_rate": 7.875481968109052e-06,
|
| 1240 |
+
"loss": 0.1657,
|
| 1241 |
+
"step": 176
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 3.694736842105263,
|
| 1245 |
+
"grad_norm": 0.458696186542511,
|
| 1246 |
+
"learning_rate": 7.845023606254658e-06,
|
| 1247 |
+
"loss": 0.1658,
|
| 1248 |
+
"step": 177
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 3.7157894736842105,
|
| 1252 |
+
"grad_norm": 0.41482996940612793,
|
| 1253 |
+
"learning_rate": 7.814408315515419e-06,
|
| 1254 |
+
"loss": 0.1774,
|
| 1255 |
+
"step": 178
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 3.736842105263158,
|
| 1259 |
+
"grad_norm": 0.47281476855278015,
|
| 1260 |
+
"learning_rate": 7.783637784602608e-06,
|
| 1261 |
+
"loss": 0.1563,
|
| 1262 |
+
"step": 179
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 3.7578947368421054,
|
| 1266 |
+
"grad_norm": 0.4740898907184601,
|
| 1267 |
+
"learning_rate": 7.752713710790405e-06,
|
| 1268 |
+
"loss": 0.165,
|
| 1269 |
+
"step": 180
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 3.7789473684210524,
|
| 1273 |
+
"grad_norm": 0.5168819427490234,
|
| 1274 |
+
"learning_rate": 7.721637799822269e-06,
|
| 1275 |
+
"loss": 0.188,
|
| 1276 |
+
"step": 181
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 3.8,
|
| 1280 |
+
"grad_norm": 0.5252240896224976,
|
| 1281 |
+
"learning_rate": 7.690411765816864e-06,
|
| 1282 |
+
"loss": 0.154,
|
| 1283 |
+
"step": 182
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 3.8210526315789473,
|
| 1287 |
+
"grad_norm": 0.4278474450111389,
|
| 1288 |
+
"learning_rate": 7.659037331173498e-06,
|
| 1289 |
+
"loss": 0.1333,
|
| 1290 |
+
"step": 183
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 3.8421052631578947,
|
| 1294 |
+
"grad_norm": 0.3564305007457733,
|
| 1295 |
+
"learning_rate": 7.627516226477123e-06,
|
| 1296 |
+
"loss": 0.1373,
|
| 1297 |
+
"step": 184
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 3.863157894736842,
|
| 1301 |
+
"grad_norm": 0.44761034846305847,
|
| 1302 |
+
"learning_rate": 7.595850190402877e-06,
|
| 1303 |
+
"loss": 0.1877,
|
| 1304 |
+
"step": 185
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 3.8842105263157896,
|
| 1308 |
+
"grad_norm": 0.46550026535987854,
|
| 1309 |
+
"learning_rate": 7.564040969620179e-06,
|
| 1310 |
+
"loss": 0.1363,
|
| 1311 |
+
"step": 186
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 3.905263157894737,
|
| 1315 |
+
"grad_norm": 0.6242339611053467,
|
| 1316 |
+
"learning_rate": 7.532090318696382e-06,
|
| 1317 |
+
"loss": 0.1729,
|
| 1318 |
+
"step": 187
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 3.9263157894736844,
|
| 1322 |
+
"grad_norm": 0.38208359479904175,
|
| 1323 |
+
"learning_rate": 7.500000000000001e-06,
|
| 1324 |
+
"loss": 0.1523,
|
| 1325 |
+
"step": 188
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 3.9473684210526314,
|
| 1329 |
+
"grad_norm": 0.42417314648628235,
|
| 1330 |
+
"learning_rate": 7.467771783603492e-06,
|
| 1331 |
+
"loss": 0.1624,
|
| 1332 |
+
"step": 189
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 3.968421052631579,
|
| 1336 |
+
"grad_norm": 0.4905320703983307,
|
| 1337 |
+
"learning_rate": 7.435407447185623e-06,
|
| 1338 |
+
"loss": 0.1686,
|
| 1339 |
+
"step": 190
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 3.9894736842105263,
|
| 1343 |
+
"grad_norm": 0.47805875539779663,
|
| 1344 |
+
"learning_rate": 7.402908775933419e-06,
|
| 1345 |
+
"loss": 0.1732,
|
| 1346 |
+
"step": 191
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 4.0,
|
| 1350 |
+
"grad_norm": 0.6476956605911255,
|
| 1351 |
+
"learning_rate": 7.370277562443689e-06,
|
| 1352 |
+
"loss": 0.0892,
|
| 1353 |
+
"step": 192
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 4.021052631578947,
|
| 1357 |
+
"grad_norm": 0.4977462589740753,
|
| 1358 |
+
"learning_rate": 7.337515606624148e-06,
|
| 1359 |
+
"loss": 0.1068,
|
| 1360 |
+
"step": 193
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 4.042105263157895,
|
| 1364 |
+
"grad_norm": 0.39349666237831116,
|
| 1365 |
+
"learning_rate": 7.30462471559414e-06,
|
| 1366 |
+
"loss": 0.0691,
|
| 1367 |
+
"step": 194
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 4.063157894736842,
|
| 1371 |
+
"grad_norm": 0.35715895891189575,
|
| 1372 |
+
"learning_rate": 7.2716067035849595e-06,
|
| 1373 |
+
"loss": 0.0853,
|
| 1374 |
+
"step": 195
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 4.08421052631579,
|
| 1378 |
+
"grad_norm": 0.39891958236694336,
|
| 1379 |
+
"learning_rate": 7.23846339183977e-06,
|
| 1380 |
+
"loss": 0.1031,
|
| 1381 |
+
"step": 196
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 4.105263157894737,
|
| 1385 |
+
"grad_norm": 0.5504170060157776,
|
| 1386 |
+
"learning_rate": 7.2051966085131584e-06,
|
| 1387 |
+
"loss": 0.096,
|
| 1388 |
+
"step": 197
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 4.126315789473685,
|
| 1392 |
+
"grad_norm": 0.45905762910842896,
|
| 1393 |
+
"learning_rate": 7.1718081885702905e-06,
|
| 1394 |
+
"loss": 0.0974,
|
| 1395 |
+
"step": 198
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 4.147368421052631,
|
| 1399 |
+
"grad_norm": 0.521188497543335,
|
| 1400 |
+
"learning_rate": 7.138299973685694e-06,
|
| 1401 |
+
"loss": 0.0884,
|
| 1402 |
+
"step": 199
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 4.168421052631579,
|
| 1406 |
+
"grad_norm": 0.4836471676826477,
|
| 1407 |
+
"learning_rate": 7.104673812141676e-06,
|
| 1408 |
+
"loss": 0.0984,
|
| 1409 |
+
"step": 200
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 4.189473684210526,
|
| 1413 |
+
"grad_norm": 0.4355693459510803,
|
| 1414 |
+
"learning_rate": 7.070931558726373e-06,
|
| 1415 |
+
"loss": 0.1012,
|
| 1416 |
+
"step": 201
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 4.2105263157894735,
|
| 1420 |
+
"grad_norm": 0.43118616938591003,
|
| 1421 |
+
"learning_rate": 7.037075074631441e-06,
|
| 1422 |
+
"loss": 0.0892,
|
| 1423 |
+
"step": 202
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 4.231578947368421,
|
| 1427 |
+
"grad_norm": 0.41400811076164246,
|
| 1428 |
+
"learning_rate": 7.003106227349399e-06,
|
| 1429 |
+
"loss": 0.0908,
|
| 1430 |
+
"step": 203
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 4.252631578947368,
|
| 1434 |
+
"grad_norm": 0.41520023345947266,
|
| 1435 |
+
"learning_rate": 6.969026890570612e-06,
|
| 1436 |
+
"loss": 0.0818,
|
| 1437 |
+
"step": 204
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 4.273684210526316,
|
| 1441 |
+
"grad_norm": 0.4351101219654083,
|
| 1442 |
+
"learning_rate": 6.934838944079944e-06,
|
| 1443 |
+
"loss": 0.079,
|
| 1444 |
+
"step": 205
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 4.294736842105263,
|
| 1448 |
+
"grad_norm": 0.4256405234336853,
|
| 1449 |
+
"learning_rate": 6.9005442736530745e-06,
|
| 1450 |
+
"loss": 0.0818,
|
| 1451 |
+
"step": 206
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 4.315789473684211,
|
| 1455 |
+
"grad_norm": 0.43969833850860596,
|
| 1456 |
+
"learning_rate": 6.866144770952474e-06,
|
| 1457 |
+
"loss": 0.0907,
|
| 1458 |
+
"step": 207
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 4.336842105263158,
|
| 1462 |
+
"grad_norm": 0.40439504384994507,
|
| 1463 |
+
"learning_rate": 6.831642333423068e-06,
|
| 1464 |
+
"loss": 0.0905,
|
| 1465 |
+
"step": 208
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 4.3578947368421055,
|
| 1469 |
+
"grad_norm": 0.4105828106403351,
|
| 1470 |
+
"learning_rate": 6.797038864187564e-06,
|
| 1471 |
+
"loss": 0.0964,
|
| 1472 |
+
"step": 209
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 4.378947368421053,
|
| 1476 |
+
"grad_norm": 0.42454293370246887,
|
| 1477 |
+
"learning_rate": 6.762336271941499e-06,
|
| 1478 |
+
"loss": 0.0862,
|
| 1479 |
+
"step": 210
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 4.4,
|
| 1483 |
+
"grad_norm": 0.3944256603717804,
|
| 1484 |
+
"learning_rate": 6.7275364708479316e-06,
|
| 1485 |
+
"loss": 0.1105,
|
| 1486 |
+
"step": 211
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 4.421052631578947,
|
| 1490 |
+
"grad_norm": 0.44638609886169434,
|
| 1491 |
+
"learning_rate": 6.692641380431879e-06,
|
| 1492 |
+
"loss": 0.0882,
|
| 1493 |
+
"step": 212
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 4.442105263157894,
|
| 1497 |
+
"grad_norm": 0.4451676607131958,
|
| 1498 |
+
"learning_rate": 6.657652925474424e-06,
|
| 1499 |
+
"loss": 0.0947,
|
| 1500 |
+
"step": 213
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 4.463157894736842,
|
| 1504 |
+
"grad_norm": 0.39187949895858765,
|
| 1505 |
+
"learning_rate": 6.622573035906557e-06,
|
| 1506 |
+
"loss": 0.0715,
|
| 1507 |
+
"step": 214
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 4.484210526315789,
|
| 1511 |
+
"grad_norm": 0.40735164284706116,
|
| 1512 |
+
"learning_rate": 6.5874036467027135e-06,
|
| 1513 |
+
"loss": 0.0921,
|
| 1514 |
+
"step": 215
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 4.505263157894737,
|
| 1518 |
+
"grad_norm": 0.4299719035625458,
|
| 1519 |
+
"learning_rate": 6.552146697774049e-06,
|
| 1520 |
+
"loss": 0.0943,
|
| 1521 |
+
"step": 216
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 4.526315789473684,
|
| 1525 |
+
"grad_norm": 0.41552671790122986,
|
| 1526 |
+
"learning_rate": 6.51680413386143e-06,
|
| 1527 |
+
"loss": 0.0978,
|
| 1528 |
+
"step": 217
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 4.5473684210526315,
|
| 1532 |
+
"grad_norm": 0.36326003074645996,
|
| 1533 |
+
"learning_rate": 6.481377904428171e-06,
|
| 1534 |
+
"loss": 0.0824,
|
| 1535 |
+
"step": 218
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 4.568421052631579,
|
| 1539 |
+
"grad_norm": 0.36748915910720825,
|
| 1540 |
+
"learning_rate": 6.445869963552496e-06,
|
| 1541 |
+
"loss": 0.0754,
|
| 1542 |
+
"step": 219
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 4.589473684210526,
|
| 1546 |
+
"grad_norm": 0.4343811869621277,
|
| 1547 |
+
"learning_rate": 6.410282269819756e-06,
|
| 1548 |
+
"loss": 0.0799,
|
| 1549 |
+
"step": 220
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 4.610526315789474,
|
| 1553 |
+
"grad_norm": 0.33867624402046204,
|
| 1554 |
+
"learning_rate": 6.374616786214402e-06,
|
| 1555 |
+
"loss": 0.0939,
|
| 1556 |
+
"step": 221
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 4.631578947368421,
|
| 1560 |
+
"grad_norm": 0.3912070095539093,
|
| 1561 |
+
"learning_rate": 6.338875480011698e-06,
|
| 1562 |
+
"loss": 0.1031,
|
| 1563 |
+
"step": 222
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 4.652631578947369,
|
| 1567 |
+
"grad_norm": 0.413815438747406,
|
| 1568 |
+
"learning_rate": 6.303060322669214e-06,
|
| 1569 |
+
"loss": 0.0846,
|
| 1570 |
+
"step": 223
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 4.673684210526316,
|
| 1574 |
+
"grad_norm": 0.41708818078041077,
|
| 1575 |
+
"learning_rate": 6.267173289718079e-06,
|
| 1576 |
+
"loss": 0.0749,
|
| 1577 |
+
"step": 224
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 4.6947368421052635,
|
| 1581 |
+
"grad_norm": 0.3973020613193512,
|
| 1582 |
+
"learning_rate": 6.23121636065402e-06,
|
| 1583 |
+
"loss": 0.0904,
|
| 1584 |
+
"step": 225
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 4.715789473684211,
|
| 1588 |
+
"grad_norm": 0.35116341710090637,
|
| 1589 |
+
"learning_rate": 6.195191518828163e-06,
|
| 1590 |
+
"loss": 0.0759,
|
| 1591 |
+
"step": 226
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 4.7368421052631575,
|
| 1595 |
+
"grad_norm": 0.429392009973526,
|
| 1596 |
+
"learning_rate": 6.1591007513376425e-06,
|
| 1597 |
+
"loss": 0.0636,
|
| 1598 |
+
"step": 227
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 4.757894736842105,
|
| 1602 |
+
"grad_norm": 0.37354663014411926,
|
| 1603 |
+
"learning_rate": 6.122946048915991e-06,
|
| 1604 |
+
"loss": 0.0715,
|
| 1605 |
+
"step": 228
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 4.778947368421052,
|
| 1609 |
+
"grad_norm": 0.46815237402915955,
|
| 1610 |
+
"learning_rate": 6.086729405823335e-06,
|
| 1611 |
+
"loss": 0.1122,
|
| 1612 |
+
"step": 229
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 4.8,
|
| 1616 |
+
"grad_norm": 0.38619089126586914,
|
| 1617 |
+
"learning_rate": 6.05045281973639e-06,
|
| 1618 |
+
"loss": 0.0802,
|
| 1619 |
+
"step": 230
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 4.821052631578947,
|
| 1623 |
+
"grad_norm": 0.3449643552303314,
|
| 1624 |
+
"learning_rate": 6.014118291638272e-06,
|
| 1625 |
+
"loss": 0.0999,
|
| 1626 |
+
"step": 231
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 4.842105263157895,
|
| 1630 |
+
"grad_norm": 0.4162727892398834,
|
| 1631 |
+
"learning_rate": 5.977727825708123e-06,
|
| 1632 |
+
"loss": 0.0847,
|
| 1633 |
+
"step": 232
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 4.863157894736842,
|
| 1637 |
+
"grad_norm": 0.4373238682746887,
|
| 1638 |
+
"learning_rate": 5.941283429210568e-06,
|
| 1639 |
+
"loss": 0.0996,
|
| 1640 |
+
"step": 233
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 4.88421052631579,
|
| 1644 |
+
"grad_norm": 0.43960970640182495,
|
| 1645 |
+
"learning_rate": 5.904787112384991e-06,
|
| 1646 |
+
"loss": 0.1044,
|
| 1647 |
+
"step": 234
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 4.905263157894737,
|
| 1651 |
+
"grad_norm": 0.40882420539855957,
|
| 1652 |
+
"learning_rate": 5.8682408883346535e-06,
|
| 1653 |
+
"loss": 0.078,
|
| 1654 |
+
"step": 235
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 4.926315789473684,
|
| 1658 |
+
"grad_norm": 0.4359929859638214,
|
| 1659 |
+
"learning_rate": 5.831646772915651e-06,
|
| 1660 |
+
"loss": 0.0966,
|
| 1661 |
+
"step": 236
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 4.947368421052632,
|
| 1665 |
+
"grad_norm": 0.44604310393333435,
|
| 1666 |
+
"learning_rate": 5.795006784625728e-06,
|
| 1667 |
+
"loss": 0.09,
|
| 1668 |
+
"step": 237
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 4.968421052631579,
|
| 1672 |
+
"grad_norm": 0.4074229896068573,
|
| 1673 |
+
"learning_rate": 5.75832294449293e-06,
|
| 1674 |
+
"loss": 0.0847,
|
| 1675 |
+
"step": 238
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 4.989473684210527,
|
| 1679 |
+
"grad_norm": 0.3633570969104767,
|
| 1680 |
+
"learning_rate": 5.7215972759641335e-06,
|
| 1681 |
+
"loss": 0.0766,
|
| 1682 |
+
"step": 239
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 5.0,
|
| 1686 |
+
"grad_norm": 0.3633570969104767,
|
| 1687 |
+
"learning_rate": 5.684831804793427e-06,
|
| 1688 |
+
"loss": 0.063,
|
| 1689 |
+
"step": 240
|
| 1690 |
+
}
|
| 1691 |
+
],
|
| 1692 |
+
"logging_steps": 1,
|
| 1693 |
+
"max_steps": 470,
|
| 1694 |
+
"num_input_tokens_seen": 0,
|
| 1695 |
+
"num_train_epochs": 10,
|
| 1696 |
+
"save_steps": 20,
|
| 1697 |
+
"stateful_callbacks": {
|
| 1698 |
+
"TrainerControl": {
|
| 1699 |
+
"args": {
|
| 1700 |
+
"should_epoch_stop": false,
|
| 1701 |
+
"should_evaluate": false,
|
| 1702 |
+
"should_log": false,
|
| 1703 |
+
"should_save": true,
|
| 1704 |
+
"should_training_stop": false
|
| 1705 |
+
},
|
| 1706 |
+
"attributes": {}
|
| 1707 |
+
}
|
| 1708 |
+
},
|
| 1709 |
+
"total_flos": 158948547461120.0,
|
| 1710 |
+
"train_batch_size": 1,
|
| 1711 |
+
"trial_name": null,
|
| 1712 |
+
"trial_params": null
|
| 1713 |
+
}
|