Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1834 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1834 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2560,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0390625,
|
| 13 |
+
"grad_norm": 8.492419242858887,
|
| 14 |
+
"learning_rate": 7.8125e-06,
|
| 15 |
+
"loss": 1.3342,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.078125,
|
| 20 |
+
"grad_norm": 3.8445992469787598,
|
| 21 |
+
"learning_rate": 1.5625e-05,
|
| 22 |
+
"loss": 1.0352,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.1171875,
|
| 27 |
+
"grad_norm": 2.211190700531006,
|
| 28 |
+
"learning_rate": 2.34375e-05,
|
| 29 |
+
"loss": 0.4873,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.15625,
|
| 34 |
+
"grad_norm": 1.9646821022033691,
|
| 35 |
+
"learning_rate": 3.125e-05,
|
| 36 |
+
"loss": 0.3898,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.1953125,
|
| 41 |
+
"grad_norm": 1.9914284944534302,
|
| 42 |
+
"learning_rate": 3.90625e-05,
|
| 43 |
+
"loss": 0.2554,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.234375,
|
| 48 |
+
"grad_norm": 0.9096956253051758,
|
| 49 |
+
"learning_rate": 4.6875e-05,
|
| 50 |
+
"loss": 0.1924,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.2734375,
|
| 55 |
+
"grad_norm": 0.6492406129837036,
|
| 56 |
+
"learning_rate": 5.46875e-05,
|
| 57 |
+
"loss": 0.1721,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.3125,
|
| 62 |
+
"grad_norm": 0.8035663962364197,
|
| 63 |
+
"learning_rate": 6.25e-05,
|
| 64 |
+
"loss": 0.1454,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.3515625,
|
| 69 |
+
"grad_norm": 0.5276547074317932,
|
| 70 |
+
"learning_rate": 7.031250000000001e-05,
|
| 71 |
+
"loss": 0.1343,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.390625,
|
| 76 |
+
"grad_norm": 0.735613226890564,
|
| 77 |
+
"learning_rate": 7.8125e-05,
|
| 78 |
+
"loss": 0.1318,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.4296875,
|
| 83 |
+
"grad_norm": 1.0286799669265747,
|
| 84 |
+
"learning_rate": 8.593750000000001e-05,
|
| 85 |
+
"loss": 0.111,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.46875,
|
| 90 |
+
"grad_norm": 0.6054307222366333,
|
| 91 |
+
"learning_rate": 9.375e-05,
|
| 92 |
+
"loss": 0.1102,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.5078125,
|
| 97 |
+
"grad_norm": 0.6528738737106323,
|
| 98 |
+
"learning_rate": 9.999983313226824e-05,
|
| 99 |
+
"loss": 0.1024,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.546875,
|
| 104 |
+
"grad_norm": 0.522591769695282,
|
| 105 |
+
"learning_rate": 9.999399287860388e-05,
|
| 106 |
+
"loss": 0.0946,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.5859375,
|
| 111 |
+
"grad_norm": 0.7753025889396667,
|
| 112 |
+
"learning_rate": 9.997981035211134e-05,
|
| 113 |
+
"loss": 0.093,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.625,
|
| 118 |
+
"grad_norm": 0.58796226978302,
|
| 119 |
+
"learning_rate": 9.995728791936504e-05,
|
| 120 |
+
"loss": 0.0881,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.6640625,
|
| 125 |
+
"grad_norm": 0.5107079744338989,
|
| 126 |
+
"learning_rate": 9.992642933858212e-05,
|
| 127 |
+
"loss": 0.0807,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.703125,
|
| 132 |
+
"grad_norm": 0.9606569409370422,
|
| 133 |
+
"learning_rate": 9.988723975899517e-05,
|
| 134 |
+
"loss": 0.0991,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.7421875,
|
| 139 |
+
"grad_norm": 0.7940914034843445,
|
| 140 |
+
"learning_rate": 9.983972571999318e-05,
|
| 141 |
+
"loss": 0.0858,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.78125,
|
| 146 |
+
"grad_norm": 0.805740237236023,
|
| 147 |
+
"learning_rate": 9.97838951500302e-05,
|
| 148 |
+
"loss": 0.0975,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.8203125,
|
| 153 |
+
"grad_norm": 0.8114387392997742,
|
| 154 |
+
"learning_rate": 9.971975736530243e-05,
|
| 155 |
+
"loss": 0.0815,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.859375,
|
| 160 |
+
"grad_norm": 0.5807178616523743,
|
| 161 |
+
"learning_rate": 9.96473230681937e-05,
|
| 162 |
+
"loss": 0.0768,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.8984375,
|
| 167 |
+
"grad_norm": 0.655323326587677,
|
| 168 |
+
"learning_rate": 9.956660434548948e-05,
|
| 169 |
+
"loss": 0.0858,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.9375,
|
| 174 |
+
"grad_norm": 0.513111412525177,
|
| 175 |
+
"learning_rate": 9.947761466636014e-05,
|
| 176 |
+
"loss": 0.0816,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.9765625,
|
| 181 |
+
"grad_norm": 0.7743618488311768,
|
| 182 |
+
"learning_rate": 9.938036888011337e-05,
|
| 183 |
+
"loss": 0.071,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 1.015625,
|
| 188 |
+
"grad_norm": 0.45275968313217163,
|
| 189 |
+
"learning_rate": 9.92748832137163e-05,
|
| 190 |
+
"loss": 0.073,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 1.0546875,
|
| 195 |
+
"grad_norm": 0.7138253450393677,
|
| 196 |
+
"learning_rate": 9.916117526908785e-05,
|
| 197 |
+
"loss": 0.0774,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 1.09375,
|
| 202 |
+
"grad_norm": 0.8007267117500305,
|
| 203 |
+
"learning_rate": 9.903926402016153e-05,
|
| 204 |
+
"loss": 0.0721,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 1.1328125,
|
| 209 |
+
"grad_norm": 0.6417205333709717,
|
| 210 |
+
"learning_rate": 9.890916980971936e-05,
|
| 211 |
+
"loss": 0.0692,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 1.171875,
|
| 216 |
+
"grad_norm": 0.605466902256012,
|
| 217 |
+
"learning_rate": 9.877091434599734e-05,
|
| 218 |
+
"loss": 0.0728,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 1.2109375,
|
| 223 |
+
"grad_norm": 0.6525840759277344,
|
| 224 |
+
"learning_rate": 9.862452069906317e-05,
|
| 225 |
+
"loss": 0.0758,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 1.25,
|
| 230 |
+
"grad_norm": 1.1812381744384766,
|
| 231 |
+
"learning_rate": 9.847001329696653e-05,
|
| 232 |
+
"loss": 0.0707,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.2890625,
|
| 237 |
+
"grad_norm": 0.7314594388008118,
|
| 238 |
+
"learning_rate": 9.830741792166295e-05,
|
| 239 |
+
"loss": 0.0693,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.328125,
|
| 244 |
+
"grad_norm": 0.721834659576416,
|
| 245 |
+
"learning_rate": 9.813676170471177e-05,
|
| 246 |
+
"loss": 0.0702,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.3671875,
|
| 251 |
+
"grad_norm": 0.5326610207557678,
|
| 252 |
+
"learning_rate": 9.795807312274862e-05,
|
| 253 |
+
"loss": 0.0684,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.40625,
|
| 258 |
+
"grad_norm": 0.6198597550392151,
|
| 259 |
+
"learning_rate": 9.777138199273384e-05,
|
| 260 |
+
"loss": 0.0675,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.4453125,
|
| 265 |
+
"grad_norm": 0.5808058977127075,
|
| 266 |
+
"learning_rate": 9.757671946697698e-05,
|
| 267 |
+
"loss": 0.0665,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.484375,
|
| 272 |
+
"grad_norm": 0.4104309678077698,
|
| 273 |
+
"learning_rate": 9.737411802793852e-05,
|
| 274 |
+
"loss": 0.0676,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.5234375,
|
| 279 |
+
"grad_norm": 0.618438720703125,
|
| 280 |
+
"learning_rate": 9.716361148280977e-05,
|
| 281 |
+
"loss": 0.0615,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.5625,
|
| 286 |
+
"grad_norm": 0.4504268765449524,
|
| 287 |
+
"learning_rate": 9.694523495787149e-05,
|
| 288 |
+
"loss": 0.0721,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.6015625,
|
| 293 |
+
"grad_norm": 0.333695650100708,
|
| 294 |
+
"learning_rate": 9.671902489263261e-05,
|
| 295 |
+
"loss": 0.0579,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.640625,
|
| 300 |
+
"grad_norm": 0.7360309362411499,
|
| 301 |
+
"learning_rate": 9.648501903374975e-05,
|
| 302 |
+
"loss": 0.0651,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.6796875,
|
| 307 |
+
"grad_norm": 0.4267025291919708,
|
| 308 |
+
"learning_rate": 9.624325642872853e-05,
|
| 309 |
+
"loss": 0.057,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.71875,
|
| 314 |
+
"grad_norm": 0.3154551684856415,
|
| 315 |
+
"learning_rate": 9.599377741940792e-05,
|
| 316 |
+
"loss": 0.0696,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.7578125,
|
| 321 |
+
"grad_norm": 0.4576047658920288,
|
| 322 |
+
"learning_rate": 9.57366236352286e-05,
|
| 323 |
+
"loss": 0.0595,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.796875,
|
| 328 |
+
"grad_norm": 0.5197241902351379,
|
| 329 |
+
"learning_rate": 9.547183798628641e-05,
|
| 330 |
+
"loss": 0.0661,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.8359375,
|
| 335 |
+
"grad_norm": 0.42932412028312683,
|
| 336 |
+
"learning_rate": 9.519946465617218e-05,
|
| 337 |
+
"loss": 0.0609,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.875,
|
| 342 |
+
"grad_norm": 0.4223524034023285,
|
| 343 |
+
"learning_rate": 9.491954909459895e-05,
|
| 344 |
+
"loss": 0.0574,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.9140625,
|
| 349 |
+
"grad_norm": 0.47859442234039307,
|
| 350 |
+
"learning_rate": 9.463213800981805e-05,
|
| 351 |
+
"loss": 0.0601,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.953125,
|
| 356 |
+
"grad_norm": 0.31995102763175964,
|
| 357 |
+
"learning_rate": 9.433727936082504e-05,
|
| 358 |
+
"loss": 0.0555,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.9921875,
|
| 363 |
+
"grad_norm": 0.3553142547607422,
|
| 364 |
+
"learning_rate": 9.403502234935706e-05,
|
| 365 |
+
"loss": 0.0582,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 2.03125,
|
| 370 |
+
"grad_norm": 0.6797146201133728,
|
| 371 |
+
"learning_rate": 9.372541741168272e-05,
|
| 372 |
+
"loss": 0.0528,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 2.0703125,
|
| 377 |
+
"grad_norm": 0.41948041319847107,
|
| 378 |
+
"learning_rate": 9.340851621018603e-05,
|
| 379 |
+
"loss": 0.0598,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 2.109375,
|
| 384 |
+
"grad_norm": 0.5185118317604065,
|
| 385 |
+
"learning_rate": 9.308437162474579e-05,
|
| 386 |
+
"loss": 0.0573,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 2.1484375,
|
| 391 |
+
"grad_norm": 0.5795381665229797,
|
| 392 |
+
"learning_rate": 9.275303774391165e-05,
|
| 393 |
+
"loss": 0.0582,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 2.1875,
|
| 398 |
+
"grad_norm": 0.44867536425590515,
|
| 399 |
+
"learning_rate": 9.241456985587868e-05,
|
| 400 |
+
"loss": 0.0616,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 2.2265625,
|
| 405 |
+
"grad_norm": 0.42551538348197937,
|
| 406 |
+
"learning_rate": 9.20690244392617e-05,
|
| 407 |
+
"loss": 0.0578,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 2.265625,
|
| 412 |
+
"grad_norm": 0.6047021746635437,
|
| 413 |
+
"learning_rate": 9.171645915367088e-05,
|
| 414 |
+
"loss": 0.0523,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 2.3046875,
|
| 419 |
+
"grad_norm": 0.5272088050842285,
|
| 420 |
+
"learning_rate": 9.135693283009035e-05,
|
| 421 |
+
"loss": 0.0648,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 2.34375,
|
| 426 |
+
"grad_norm": 0.396763414144516,
|
| 427 |
+
"learning_rate": 9.099050546106133e-05,
|
| 428 |
+
"loss": 0.0543,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 2.3828125,
|
| 433 |
+
"grad_norm": 0.4244937300682068,
|
| 434 |
+
"learning_rate": 9.061723819067154e-05,
|
| 435 |
+
"loss": 0.0506,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 2.421875,
|
| 440 |
+
"grad_norm": 0.37396669387817383,
|
| 441 |
+
"learning_rate": 9.023719330435223e-05,
|
| 442 |
+
"loss": 0.0566,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 2.4609375,
|
| 447 |
+
"grad_norm": 0.3504962623119354,
|
| 448 |
+
"learning_rate": 8.9850434218485e-05,
|
| 449 |
+
"loss": 0.0586,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 2.5,
|
| 454 |
+
"grad_norm": 0.5627210736274719,
|
| 455 |
+
"learning_rate": 8.945702546981969e-05,
|
| 456 |
+
"loss": 0.053,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 2.5390625,
|
| 461 |
+
"grad_norm": 0.6733695268630981,
|
| 462 |
+
"learning_rate": 8.905703270470549e-05,
|
| 463 |
+
"loss": 0.0566,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.578125,
|
| 468 |
+
"grad_norm": 0.28862297534942627,
|
| 469 |
+
"learning_rate": 8.865052266813685e-05,
|
| 470 |
+
"loss": 0.0484,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 2.6171875,
|
| 475 |
+
"grad_norm": 0.9207034111022949,
|
| 476 |
+
"learning_rate": 8.823756319261597e-05,
|
| 477 |
+
"loss": 0.0556,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 2.65625,
|
| 482 |
+
"grad_norm": 0.5370097756385803,
|
| 483 |
+
"learning_rate": 8.781822318683394e-05,
|
| 484 |
+
"loss": 0.0533,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 2.6953125,
|
| 489 |
+
"grad_norm": 0.3493472635746002,
|
| 490 |
+
"learning_rate": 8.739257262417225e-05,
|
| 491 |
+
"loss": 0.0539,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 2.734375,
|
| 496 |
+
"grad_norm": 0.36126288771629333,
|
| 497 |
+
"learning_rate": 8.696068253102665e-05,
|
| 498 |
+
"loss": 0.0557,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 2.7734375,
|
| 503 |
+
"grad_norm": 0.331190824508667,
|
| 504 |
+
"learning_rate": 8.652262497495527e-05,
|
| 505 |
+
"loss": 0.0514,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.8125,
|
| 510 |
+
"grad_norm": 0.4170578420162201,
|
| 511 |
+
"learning_rate": 8.60784730526531e-05,
|
| 512 |
+
"loss": 0.0469,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.8515625,
|
| 517 |
+
"grad_norm": 0.35276809334754944,
|
| 518 |
+
"learning_rate": 8.56283008777546e-05,
|
| 519 |
+
"loss": 0.0541,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 2.890625,
|
| 524 |
+
"grad_norm": 0.5037612318992615,
|
| 525 |
+
"learning_rate": 8.517218356846665e-05,
|
| 526 |
+
"loss": 0.0493,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.9296875,
|
| 531 |
+
"grad_norm": 0.3002952039241791,
|
| 532 |
+
"learning_rate": 8.471019723503411e-05,
|
| 533 |
+
"loss": 0.0526,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.96875,
|
| 538 |
+
"grad_norm": 0.33518242835998535,
|
| 539 |
+
"learning_rate": 8.424241896703936e-05,
|
| 540 |
+
"loss": 0.0522,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 3.0078125,
|
| 545 |
+
"grad_norm": 0.4655015766620636,
|
| 546 |
+
"learning_rate": 8.376892682053898e-05,
|
| 547 |
+
"loss": 0.0504,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 3.046875,
|
| 552 |
+
"grad_norm": 0.537617564201355,
|
| 553 |
+
"learning_rate": 8.328979980503873e-05,
|
| 554 |
+
"loss": 0.0543,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 3.0859375,
|
| 559 |
+
"grad_norm": 0.38395699858665466,
|
| 560 |
+
"learning_rate": 8.280511787030963e-05,
|
| 561 |
+
"loss": 0.0534,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 3.125,
|
| 566 |
+
"grad_norm": 0.44339320063591003,
|
| 567 |
+
"learning_rate": 8.231496189304704e-05,
|
| 568 |
+
"loss": 0.0474,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 3.1640625,
|
| 573 |
+
"grad_norm": 0.49106666445732117,
|
| 574 |
+
"learning_rate": 8.18194136633753e-05,
|
| 575 |
+
"loss": 0.0523,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 3.203125,
|
| 580 |
+
"grad_norm": 0.3647231161594391,
|
| 581 |
+
"learning_rate": 8.131855587119957e-05,
|
| 582 |
+
"loss": 0.0515,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 3.2421875,
|
| 587 |
+
"grad_norm": 0.43639010190963745,
|
| 588 |
+
"learning_rate": 8.081247209240788e-05,
|
| 589 |
+
"loss": 0.0503,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 3.28125,
|
| 594 |
+
"grad_norm": 0.3658983111381531,
|
| 595 |
+
"learning_rate": 8.030124677492512e-05,
|
| 596 |
+
"loss": 0.053,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 3.3203125,
|
| 601 |
+
"grad_norm": 0.23427674174308777,
|
| 602 |
+
"learning_rate": 7.978496522462167e-05,
|
| 603 |
+
"loss": 0.0497,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 3.359375,
|
| 608 |
+
"grad_norm": 0.3150970935821533,
|
| 609 |
+
"learning_rate": 7.926371359107871e-05,
|
| 610 |
+
"loss": 0.0468,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 3.3984375,
|
| 615 |
+
"grad_norm": 0.4161214828491211,
|
| 616 |
+
"learning_rate": 7.873757885321291e-05,
|
| 617 |
+
"loss": 0.0488,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 3.4375,
|
| 622 |
+
"grad_norm": 0.27835017442703247,
|
| 623 |
+
"learning_rate": 7.820664880476256e-05,
|
| 624 |
+
"loss": 0.0449,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 3.4765625,
|
| 629 |
+
"grad_norm": 0.31080129742622375,
|
| 630 |
+
"learning_rate": 7.767101203963786e-05,
|
| 631 |
+
"loss": 0.0459,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 3.515625,
|
| 636 |
+
"grad_norm": 0.32351627945899963,
|
| 637 |
+
"learning_rate": 7.713075793713774e-05,
|
| 638 |
+
"loss": 0.0414,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 3.5546875,
|
| 643 |
+
"grad_norm": 0.4844232499599457,
|
| 644 |
+
"learning_rate": 7.658597664703533e-05,
|
| 645 |
+
"loss": 0.0456,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 3.59375,
|
| 650 |
+
"grad_norm": 0.49269959330558777,
|
| 651 |
+
"learning_rate": 7.603675907453534e-05,
|
| 652 |
+
"loss": 0.0443,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 3.6328125,
|
| 657 |
+
"grad_norm": 0.29903239011764526,
|
| 658 |
+
"learning_rate": 7.548319686510487e-05,
|
| 659 |
+
"loss": 0.0481,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 3.671875,
|
| 664 |
+
"grad_norm": 0.34350693225860596,
|
| 665 |
+
"learning_rate": 7.492538238918116e-05,
|
| 666 |
+
"loss": 0.0456,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 3.7109375,
|
| 671 |
+
"grad_norm": 0.2822878956794739,
|
| 672 |
+
"learning_rate": 7.436340872675792e-05,
|
| 673 |
+
"loss": 0.0439,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 3.75,
|
| 678 |
+
"grad_norm": 0.3564528226852417,
|
| 679 |
+
"learning_rate": 7.379736965185368e-05,
|
| 680 |
+
"loss": 0.0495,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 3.7890625,
|
| 685 |
+
"grad_norm": 0.44323813915252686,
|
| 686 |
+
"learning_rate": 7.322735961686409e-05,
|
| 687 |
+
"loss": 0.0497,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 3.828125,
|
| 692 |
+
"grad_norm": 0.4935294985771179,
|
| 693 |
+
"learning_rate": 7.265347373680102e-05,
|
| 694 |
+
"loss": 0.0474,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 3.8671875,
|
| 699 |
+
"grad_norm": 0.3851205110549927,
|
| 700 |
+
"learning_rate": 7.20758077734212e-05,
|
| 701 |
+
"loss": 0.0475,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 3.90625,
|
| 706 |
+
"grad_norm": 0.4426199495792389,
|
| 707 |
+
"learning_rate": 7.149445811924685e-05,
|
| 708 |
+
"loss": 0.0467,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 3.9453125,
|
| 713 |
+
"grad_norm": 0.4967394769191742,
|
| 714 |
+
"learning_rate": 7.090952178148112e-05,
|
| 715 |
+
"loss": 0.0508,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 3.984375,
|
| 720 |
+
"grad_norm": 0.3541552722454071,
|
| 721 |
+
"learning_rate": 7.032109636582096e-05,
|
| 722 |
+
"loss": 0.0457,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 4.0234375,
|
| 727 |
+
"grad_norm": 0.4096279740333557,
|
| 728 |
+
"learning_rate": 6.972928006017001e-05,
|
| 729 |
+
"loss": 0.0446,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 4.0625,
|
| 734 |
+
"grad_norm": 0.34364351630210876,
|
| 735 |
+
"learning_rate": 6.91341716182545e-05,
|
| 736 |
+
"loss": 0.0474,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 4.1015625,
|
| 741 |
+
"grad_norm": 0.7376656532287598,
|
| 742 |
+
"learning_rate": 6.853587034314463e-05,
|
| 743 |
+
"loss": 0.0452,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 4.140625,
|
| 748 |
+
"grad_norm": 0.46325570344924927,
|
| 749 |
+
"learning_rate": 6.793447607068437e-05,
|
| 750 |
+
"loss": 0.0426,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 4.1796875,
|
| 755 |
+
"grad_norm": 0.44051945209503174,
|
| 756 |
+
"learning_rate": 6.733008915283216e-05,
|
| 757 |
+
"loss": 0.0385,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 4.21875,
|
| 762 |
+
"grad_norm": 0.4191512167453766,
|
| 763 |
+
"learning_rate": 6.672281044091582e-05,
|
| 764 |
+
"loss": 0.0468,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 4.2578125,
|
| 769 |
+
"grad_norm": 0.23939189314842224,
|
| 770 |
+
"learning_rate": 6.611274126880378e-05,
|
| 771 |
+
"loss": 0.0417,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 4.296875,
|
| 776 |
+
"grad_norm": 0.2203720659017563,
|
| 777 |
+
"learning_rate": 6.549998343599601e-05,
|
| 778 |
+
"loss": 0.0457,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 4.3359375,
|
| 783 |
+
"grad_norm": 0.4468579888343811,
|
| 784 |
+
"learning_rate": 6.488463919063727e-05,
|
| 785 |
+
"loss": 0.0459,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 4.375,
|
| 790 |
+
"grad_norm": 0.39386874437332153,
|
| 791 |
+
"learning_rate": 6.426681121245527e-05,
|
| 792 |
+
"loss": 0.0439,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 4.4140625,
|
| 797 |
+
"grad_norm": 0.29376569390296936,
|
| 798 |
+
"learning_rate": 6.364660259562713e-05,
|
| 799 |
+
"loss": 0.0439,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 4.453125,
|
| 804 |
+
"grad_norm": 0.24146665632724762,
|
| 805 |
+
"learning_rate": 6.302411683157637e-05,
|
| 806 |
+
"loss": 0.0429,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 4.4921875,
|
| 811 |
+
"grad_norm": 0.3997792899608612,
|
| 812 |
+
"learning_rate": 6.239945779170389e-05,
|
| 813 |
+
"loss": 0.0463,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 4.53125,
|
| 818 |
+
"grad_norm": 0.302913635969162,
|
| 819 |
+
"learning_rate": 6.177272971005528e-05,
|
| 820 |
+
"loss": 0.0503,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 4.5703125,
|
| 825 |
+
"grad_norm": 0.4844031035900116,
|
| 826 |
+
"learning_rate": 6.114403716592794e-05,
|
| 827 |
+
"loss": 0.0489,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 4.609375,
|
| 832 |
+
"grad_norm": 0.34038305282592773,
|
| 833 |
+
"learning_rate": 6.051348506642017e-05,
|
| 834 |
+
"loss": 0.0436,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 4.6484375,
|
| 839 |
+
"grad_norm": 0.5346148610115051,
|
| 840 |
+
"learning_rate": 5.988117862892601e-05,
|
| 841 |
+
"loss": 0.0456,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 4.6875,
|
| 846 |
+
"grad_norm": 0.28704309463500977,
|
| 847 |
+
"learning_rate": 5.924722336357793e-05,
|
| 848 |
+
"loss": 0.0421,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 4.7265625,
|
| 853 |
+
"grad_norm": 0.2789936065673828,
|
| 854 |
+
"learning_rate": 5.861172505564091e-05,
|
| 855 |
+
"loss": 0.0414,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 4.765625,
|
| 860 |
+
"grad_norm": 0.29666388034820557,
|
| 861 |
+
"learning_rate": 5.797478974786055e-05,
|
| 862 |
+
"loss": 0.0476,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 4.8046875,
|
| 867 |
+
"grad_norm": 0.520137369632721,
|
| 868 |
+
"learning_rate": 5.733652372276809e-05,
|
| 869 |
+
"loss": 0.0399,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 4.84375,
|
| 874 |
+
"grad_norm": 0.29389891028404236,
|
| 875 |
+
"learning_rate": 5.669703348494564e-05,
|
| 876 |
+
"loss": 0.0426,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 4.8828125,
|
| 881 |
+
"grad_norm": 0.4732462465763092,
|
| 882 |
+
"learning_rate": 5.605642574325413e-05,
|
| 883 |
+
"loss": 0.0436,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 4.921875,
|
| 888 |
+
"grad_norm": 0.6422872543334961,
|
| 889 |
+
"learning_rate": 5.5414807393027376e-05,
|
| 890 |
+
"loss": 0.0385,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 4.9609375,
|
| 895 |
+
"grad_norm": 0.31786808371543884,
|
| 896 |
+
"learning_rate": 5.47722854982349e-05,
|
| 897 |
+
"loss": 0.0449,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 5.0,
|
| 902 |
+
"grad_norm": 0.7411214113235474,
|
| 903 |
+
"learning_rate": 5.4128967273616625e-05,
|
| 904 |
+
"loss": 0.0379,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 5.0390625,
|
| 909 |
+
"grad_norm": 0.440328449010849,
|
| 910 |
+
"learning_rate": 5.348496006679244e-05,
|
| 911 |
+
"loss": 0.0404,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 5.078125,
|
| 916 |
+
"grad_norm": 0.40569090843200684,
|
| 917 |
+
"learning_rate": 5.284037134034964e-05,
|
| 918 |
+
"loss": 0.0481,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 5.1171875,
|
| 923 |
+
"grad_norm": 0.23646467924118042,
|
| 924 |
+
"learning_rate": 5.2195308653911035e-05,
|
| 925 |
+
"loss": 0.0365,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 5.15625,
|
| 930 |
+
"grad_norm": 0.41665926575660706,
|
| 931 |
+
"learning_rate": 5.154987964618704e-05,
|
| 932 |
+
"loss": 0.0364,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 5.1953125,
|
| 937 |
+
"grad_norm": 0.24862350523471832,
|
| 938 |
+
"learning_rate": 5.090419201701445e-05,
|
| 939 |
+
"loss": 0.041,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 5.234375,
|
| 944 |
+
"grad_norm": 0.3748071491718292,
|
| 945 |
+
"learning_rate": 5.025835350938506e-05,
|
| 946 |
+
"loss": 0.0477,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 5.2734375,
|
| 951 |
+
"grad_norm": 0.40436264872550964,
|
| 952 |
+
"learning_rate": 4.961247189146707e-05,
|
| 953 |
+
"loss": 0.0348,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 5.3125,
|
| 958 |
+
"grad_norm": 0.4232139587402344,
|
| 959 |
+
"learning_rate": 4.8966654938622295e-05,
|
| 960 |
+
"loss": 0.0367,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 5.3515625,
|
| 965 |
+
"grad_norm": 0.26344922184944153,
|
| 966 |
+
"learning_rate": 4.832101041542219e-05,
|
| 967 |
+
"loss": 0.0355,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 5.390625,
|
| 972 |
+
"grad_norm": 0.4235900044441223,
|
| 973 |
+
"learning_rate": 4.7675646057665644e-05,
|
| 974 |
+
"loss": 0.0432,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 5.4296875,
|
| 979 |
+
"grad_norm": 0.5052417516708374,
|
| 980 |
+
"learning_rate": 4.7030669554401596e-05,
|
| 981 |
+
"loss": 0.0353,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 5.46875,
|
| 986 |
+
"grad_norm": 0.38285690546035767,
|
| 987 |
+
"learning_rate": 4.638618852995947e-05,
|
| 988 |
+
"loss": 0.0401,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 5.5078125,
|
| 993 |
+
"grad_norm": 0.6338162422180176,
|
| 994 |
+
"learning_rate": 4.574231052599034e-05,
|
| 995 |
+
"loss": 0.0418,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 5.546875,
|
| 1000 |
+
"grad_norm": 0.3230713903903961,
|
| 1001 |
+
"learning_rate": 4.509914298352197e-05,
|
| 1002 |
+
"loss": 0.0372,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 5.5859375,
|
| 1007 |
+
"grad_norm": 0.3894837498664856,
|
| 1008 |
+
"learning_rate": 4.445679322503063e-05,
|
| 1009 |
+
"loss": 0.0387,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 5.625,
|
| 1014 |
+
"grad_norm": 0.3093172609806061,
|
| 1015 |
+
"learning_rate": 4.381536843653262e-05,
|
| 1016 |
+
"loss": 0.0363,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 5.6640625,
|
| 1021 |
+
"grad_norm": 0.4512802064418793,
|
| 1022 |
+
"learning_rate": 4.317497564969868e-05,
|
| 1023 |
+
"loss": 0.0382,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 5.703125,
|
| 1028 |
+
"grad_norm": 0.3944973051548004,
|
| 1029 |
+
"learning_rate": 4.253572172399407e-05,
|
| 1030 |
+
"loss": 0.0413,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 5.7421875,
|
| 1035 |
+
"grad_norm": 0.4658920466899872,
|
| 1036 |
+
"learning_rate": 4.18977133288474e-05,
|
| 1037 |
+
"loss": 0.0363,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 5.78125,
|
| 1042 |
+
"grad_norm": 0.3259652554988861,
|
| 1043 |
+
"learning_rate": 4.126105692585121e-05,
|
| 1044 |
+
"loss": 0.0372,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 5.8203125,
|
| 1049 |
+
"grad_norm": 0.3776373863220215,
|
| 1050 |
+
"learning_rate": 4.0625858750997184e-05,
|
| 1051 |
+
"loss": 0.0366,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 5.859375,
|
| 1056 |
+
"grad_norm": 0.22152118384838104,
|
| 1057 |
+
"learning_rate": 3.999222479694901e-05,
|
| 1058 |
+
"loss": 0.038,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 5.8984375,
|
| 1063 |
+
"grad_norm": 0.3621926009654999,
|
| 1064 |
+
"learning_rate": 3.936026079535593e-05,
|
| 1065 |
+
"loss": 0.0402,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 5.9375,
|
| 1070 |
+
"grad_norm": 0.3338565230369568,
|
| 1071 |
+
"learning_rate": 3.87300721992097e-05,
|
| 1072 |
+
"loss": 0.0417,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 5.9765625,
|
| 1077 |
+
"grad_norm": 0.5546092391014099,
|
| 1078 |
+
"learning_rate": 3.810176416524812e-05,
|
| 1079 |
+
"loss": 0.0357,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 6.015625,
|
| 1084 |
+
"grad_norm": 0.8308889865875244,
|
| 1085 |
+
"learning_rate": 3.747544153640809e-05,
|
| 1086 |
+
"loss": 0.0382,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 6.0546875,
|
| 1091 |
+
"grad_norm": 0.24918757379055023,
|
| 1092 |
+
"learning_rate": 3.6851208824330844e-05,
|
| 1093 |
+
"loss": 0.0361,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 6.09375,
|
| 1098 |
+
"grad_norm": 0.5025043487548828,
|
| 1099 |
+
"learning_rate": 3.622917019192261e-05,
|
| 1100 |
+
"loss": 0.0385,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 6.1328125,
|
| 1105 |
+
"grad_norm": 0.30080971121788025,
|
| 1106 |
+
"learning_rate": 3.560942943597344e-05,
|
| 1107 |
+
"loss": 0.0378,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 6.171875,
|
| 1112 |
+
"grad_norm": 0.2558280825614929,
|
| 1113 |
+
"learning_rate": 3.4992089969837014e-05,
|
| 1114 |
+
"loss": 0.0334,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 6.2109375,
|
| 1119 |
+
"grad_norm": 0.3407670259475708,
|
| 1120 |
+
"learning_rate": 3.437725480617462e-05,
|
| 1121 |
+
"loss": 0.0319,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 6.25,
|
| 1126 |
+
"grad_norm": 0.33768805861473083,
|
| 1127 |
+
"learning_rate": 3.3765026539765834e-05,
|
| 1128 |
+
"loss": 0.0372,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 6.2890625,
|
| 1133 |
+
"grad_norm": 0.3476319909095764,
|
| 1134 |
+
"learning_rate": 3.3155507330389e-05,
|
| 1135 |
+
"loss": 0.0369,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 6.328125,
|
| 1140 |
+
"grad_norm": 0.45832690596580505,
|
| 1141 |
+
"learning_rate": 3.254879888577432e-05,
|
| 1142 |
+
"loss": 0.0308,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 6.3671875,
|
| 1147 |
+
"grad_norm": 0.3107905387878418,
|
| 1148 |
+
"learning_rate": 3.194500244463229e-05,
|
| 1149 |
+
"loss": 0.0344,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 6.40625,
|
| 1154 |
+
"grad_norm": 0.30494946241378784,
|
| 1155 |
+
"learning_rate": 3.13442187597605e-05,
|
| 1156 |
+
"loss": 0.0365,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 6.4453125,
|
| 1161 |
+
"grad_norm": 0.4013335406780243,
|
| 1162 |
+
"learning_rate": 3.074654808123143e-05,
|
| 1163 |
+
"loss": 0.0357,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 6.484375,
|
| 1168 |
+
"grad_norm": 0.49742478132247925,
|
| 1169 |
+
"learning_rate": 3.0152090139664174e-05,
|
| 1170 |
+
"loss": 0.0339,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 6.5234375,
|
| 1175 |
+
"grad_norm": 0.3577755391597748,
|
| 1176 |
+
"learning_rate": 2.9560944129582896e-05,
|
| 1177 |
+
"loss": 0.0399,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 6.5625,
|
| 1182 |
+
"grad_norm": 0.40309587121009827,
|
| 1183 |
+
"learning_rate": 2.8973208692864624e-05,
|
| 1184 |
+
"loss": 0.0347,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 6.6015625,
|
| 1189 |
+
"grad_norm": 0.2728106677532196,
|
| 1190 |
+
"learning_rate": 2.838898190227931e-05,
|
| 1191 |
+
"loss": 0.0333,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 6.640625,
|
| 1196 |
+
"grad_norm": 0.2323320060968399,
|
| 1197 |
+
"learning_rate": 2.780836124512494e-05,
|
| 1198 |
+
"loss": 0.0362,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 6.6796875,
|
| 1203 |
+
"grad_norm": 0.3762952387332916,
|
| 1204 |
+
"learning_rate": 2.7231443606960238e-05,
|
| 1205 |
+
"loss": 0.0352,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 6.71875,
|
| 1210 |
+
"grad_norm": 0.31488674879074097,
|
| 1211 |
+
"learning_rate": 2.6658325255437744e-05,
|
| 1212 |
+
"loss": 0.0318,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 6.7578125,
|
| 1217 |
+
"grad_norm": 0.3722898066043854,
|
| 1218 |
+
"learning_rate": 2.6089101824240138e-05,
|
| 1219 |
+
"loss": 0.0341,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 6.796875,
|
| 1224 |
+
"grad_norm": 0.41492217779159546,
|
| 1225 |
+
"learning_rate": 2.5523868297122265e-05,
|
| 1226 |
+
"loss": 0.0446,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 6.8359375,
|
| 1231 |
+
"grad_norm": 0.2935633361339569,
|
| 1232 |
+
"learning_rate": 2.4962718992061634e-05,
|
| 1233 |
+
"loss": 0.0381,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 6.875,
|
| 1238 |
+
"grad_norm": 0.30372825264930725,
|
| 1239 |
+
"learning_rate": 2.4405747545519963e-05,
|
| 1240 |
+
"loss": 0.0369,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 6.9140625,
|
| 1245 |
+
"grad_norm": 0.265733540058136,
|
| 1246 |
+
"learning_rate": 2.385304689681847e-05,
|
| 1247 |
+
"loss": 0.0319,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 6.953125,
|
| 1252 |
+
"grad_norm": 0.22491760551929474,
|
| 1253 |
+
"learning_rate": 2.3304709272629584e-05,
|
| 1254 |
+
"loss": 0.0333,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 6.9921875,
|
| 1259 |
+
"grad_norm": 0.3895363211631775,
|
| 1260 |
+
"learning_rate": 2.276082617158743e-05,
|
| 1261 |
+
"loss": 0.0367,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 7.03125,
|
| 1266 |
+
"grad_norm": 0.24833568930625916,
|
| 1267 |
+
"learning_rate": 2.2221488349019903e-05,
|
| 1268 |
+
"loss": 0.0292,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 7.0703125,
|
| 1273 |
+
"grad_norm": 0.3045443594455719,
|
| 1274 |
+
"learning_rate": 2.1686785801804653e-05,
|
| 1275 |
+
"loss": 0.0322,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 7.109375,
|
| 1280 |
+
"grad_norm": 0.2556830644607544,
|
| 1281 |
+
"learning_rate": 2.115680775335184e-05,
|
| 1282 |
+
"loss": 0.0325,
|
| 1283 |
+
"step": 1820
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 7.1484375,
|
| 1287 |
+
"grad_norm": 0.28755396604537964,
|
| 1288 |
+
"learning_rate": 2.063164263871573e-05,
|
| 1289 |
+
"loss": 0.0333,
|
| 1290 |
+
"step": 1830
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 7.1875,
|
| 1294 |
+
"grad_norm": 0.4773646593093872,
|
| 1295 |
+
"learning_rate": 2.0111378089837956e-05,
|
| 1296 |
+
"loss": 0.0333,
|
| 1297 |
+
"step": 1840
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 7.2265625,
|
| 1301 |
+
"grad_norm": 0.375969260931015,
|
| 1302 |
+
"learning_rate": 1.959610092092479e-05,
|
| 1303 |
+
"loss": 0.0322,
|
| 1304 |
+
"step": 1850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 7.265625,
|
| 1308 |
+
"grad_norm": 0.2566347122192383,
|
| 1309 |
+
"learning_rate": 1.9085897113960793e-05,
|
| 1310 |
+
"loss": 0.0312,
|
| 1311 |
+
"step": 1860
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 7.3046875,
|
| 1315 |
+
"grad_norm": 0.513367235660553,
|
| 1316 |
+
"learning_rate": 1.858085180436147e-05,
|
| 1317 |
+
"loss": 0.0337,
|
| 1318 |
+
"step": 1870
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 7.34375,
|
| 1322 |
+
"grad_norm": 0.31472358107566833,
|
| 1323 |
+
"learning_rate": 1.8081049266767046e-05,
|
| 1324 |
+
"loss": 0.0302,
|
| 1325 |
+
"step": 1880
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 7.3828125,
|
| 1329 |
+
"grad_norm": 0.31810522079467773,
|
| 1330 |
+
"learning_rate": 1.758657290097997e-05,
|
| 1331 |
+
"loss": 0.03,
|
| 1332 |
+
"step": 1890
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 7.421875,
|
| 1336 |
+
"grad_norm": 0.3877921998500824,
|
| 1337 |
+
"learning_rate": 1.709750521804831e-05,
|
| 1338 |
+
"loss": 0.0384,
|
| 1339 |
+
"step": 1900
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 7.4609375,
|
| 1343 |
+
"grad_norm": 0.45386990904808044,
|
| 1344 |
+
"learning_rate": 1.661392782649755e-05,
|
| 1345 |
+
"loss": 0.037,
|
| 1346 |
+
"step": 1910
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 7.5,
|
| 1350 |
+
"grad_norm": 0.33699849247932434,
|
| 1351 |
+
"learning_rate": 1.6135921418712956e-05,
|
| 1352 |
+
"loss": 0.0316,
|
| 1353 |
+
"step": 1920
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 7.5390625,
|
| 1357 |
+
"grad_norm": 0.31667983531951904,
|
| 1358 |
+
"learning_rate": 1.5663565757474773e-05,
|
| 1359 |
+
"loss": 0.029,
|
| 1360 |
+
"step": 1930
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 7.578125,
|
| 1364 |
+
"grad_norm": 0.3529506325721741,
|
| 1365 |
+
"learning_rate": 1.5196939662648569e-05,
|
| 1366 |
+
"loss": 0.0353,
|
| 1367 |
+
"step": 1940
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 7.6171875,
|
| 1371 |
+
"grad_norm": 0.3233952820301056,
|
| 1372 |
+
"learning_rate": 1.4736120998032949e-05,
|
| 1373 |
+
"loss": 0.0331,
|
| 1374 |
+
"step": 1950
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 7.65625,
|
| 1378 |
+
"grad_norm": 0.439943790435791,
|
| 1379 |
+
"learning_rate": 1.428118665836668e-05,
|
| 1380 |
+
"loss": 0.0296,
|
| 1381 |
+
"step": 1960
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 7.6953125,
|
| 1385 |
+
"grad_norm": 0.25135713815689087,
|
| 1386 |
+
"learning_rate": 1.3832212556497748e-05,
|
| 1387 |
+
"loss": 0.0312,
|
| 1388 |
+
"step": 1970
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 7.734375,
|
| 1392 |
+
"grad_norm": 0.4036009609699249,
|
| 1393 |
+
"learning_rate": 1.338927361071603e-05,
|
| 1394 |
+
"loss": 0.0324,
|
| 1395 |
+
"step": 1980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 7.7734375,
|
| 1399 |
+
"grad_norm": 0.27127811312675476,
|
| 1400 |
+
"learning_rate": 1.2952443732252057e-05,
|
| 1401 |
+
"loss": 0.0305,
|
| 1402 |
+
"step": 1990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 7.8125,
|
| 1406 |
+
"grad_norm": 0.569515585899353,
|
| 1407 |
+
"learning_rate": 1.2521795812943704e-05,
|
| 1408 |
+
"loss": 0.0286,
|
| 1409 |
+
"step": 2000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 7.8515625,
|
| 1413 |
+
"grad_norm": 0.7388462424278259,
|
| 1414 |
+
"learning_rate": 1.2097401713073175e-05,
|
| 1415 |
+
"loss": 0.0329,
|
| 1416 |
+
"step": 2010
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 7.890625,
|
| 1420 |
+
"grad_norm": 0.35602420568466187,
|
| 1421 |
+
"learning_rate": 1.1679332249375923e-05,
|
| 1422 |
+
"loss": 0.032,
|
| 1423 |
+
"step": 2020
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 7.9296875,
|
| 1427 |
+
"grad_norm": 0.23437829315662384,
|
| 1428 |
+
"learning_rate": 1.1267657183223812e-05,
|
| 1429 |
+
"loss": 0.0325,
|
| 1430 |
+
"step": 2030
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 7.96875,
|
| 1434 |
+
"grad_norm": 0.8530910015106201,
|
| 1435 |
+
"learning_rate": 1.086244520898428e-05,
|
| 1436 |
+
"loss": 0.0365,
|
| 1437 |
+
"step": 2040
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 8.0078125,
|
| 1441 |
+
"grad_norm": 0.44752374291419983,
|
| 1442 |
+
"learning_rate": 1.046376394255773e-05,
|
| 1443 |
+
"loss": 0.0309,
|
| 1444 |
+
"step": 2050
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 8.046875,
|
| 1448 |
+
"grad_norm": 0.5457649230957031,
|
| 1449 |
+
"learning_rate": 1.00716799100947e-05,
|
| 1450 |
+
"loss": 0.0317,
|
| 1451 |
+
"step": 2060
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 8.0859375,
|
| 1455 |
+
"grad_norm": 0.4462254047393799,
|
| 1456 |
+
"learning_rate": 9.686258536894948e-06,
|
| 1457 |
+
"loss": 0.0286,
|
| 1458 |
+
"step": 2070
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 8.125,
|
| 1462 |
+
"grad_norm": 0.2501213252544403,
|
| 1463 |
+
"learning_rate": 9.307564136490254e-06,
|
| 1464 |
+
"loss": 0.0351,
|
| 1465 |
+
"step": 2080
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 8.1640625,
|
| 1469 |
+
"grad_norm": 0.49755340814590454,
|
| 1470 |
+
"learning_rate": 8.935659899912658e-06,
|
| 1471 |
+
"loss": 0.0328,
|
| 1472 |
+
"step": 2090
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 8.203125,
|
| 1476 |
+
"grad_norm": 0.4353477954864502,
|
| 1477 |
+
"learning_rate": 8.5706078851501e-06,
|
| 1478 |
+
"loss": 0.0306,
|
| 1479 |
+
"step": 2100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 8.2421875,
|
| 1483 |
+
"grad_norm": 0.2827202379703522,
|
| 1484 |
+
"learning_rate": 8.21246900679109e-06,
|
| 1485 |
+
"loss": 0.0316,
|
| 1486 |
+
"step": 2110
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 8.28125,
|
| 1490 |
+
"grad_norm": 0.26606473326683044,
|
| 1491 |
+
"learning_rate": 7.861303025860135e-06,
|
| 1492 |
+
"loss": 0.0254,
|
| 1493 |
+
"step": 2120
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 8.3203125,
|
| 1497 |
+
"grad_norm": 0.251438170671463,
|
| 1498 |
+
"learning_rate": 7.517168539845709e-06,
|
| 1499 |
+
"loss": 0.0258,
|
| 1500 |
+
"step": 2130
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 8.359375,
|
| 1504 |
+
"grad_norm": 0.38992422819137573,
|
| 1505 |
+
"learning_rate": 7.180122972922304e-06,
|
| 1506 |
+
"loss": 0.03,
|
| 1507 |
+
"step": 2140
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 8.3984375,
|
| 1511 |
+
"grad_norm": 0.4747330844402313,
|
| 1512 |
+
"learning_rate": 6.8502225663683975e-06,
|
| 1513 |
+
"loss": 0.0305,
|
| 1514 |
+
"step": 2150
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 8.4375,
|
| 1518 |
+
"grad_norm": 0.29881390929222107,
|
| 1519 |
+
"learning_rate": 6.527522369181655e-06,
|
| 1520 |
+
"loss": 0.0255,
|
| 1521 |
+
"step": 2160
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 8.4765625,
|
| 1525 |
+
"grad_norm": 0.696563720703125,
|
| 1526 |
+
"learning_rate": 6.212076228893205e-06,
|
| 1527 |
+
"loss": 0.0286,
|
| 1528 |
+
"step": 2170
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 8.515625,
|
| 1532 |
+
"grad_norm": 0.26265567541122437,
|
| 1533 |
+
"learning_rate": 5.903936782582253e-06,
|
| 1534 |
+
"loss": 0.033,
|
| 1535 |
+
"step": 2180
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 8.5546875,
|
| 1539 |
+
"grad_norm": 0.42761778831481934,
|
| 1540 |
+
"learning_rate": 5.603155448092895e-06,
|
| 1541 |
+
"loss": 0.0296,
|
| 1542 |
+
"step": 2190
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 8.59375,
|
| 1546 |
+
"grad_norm": 0.2030072659254074,
|
| 1547 |
+
"learning_rate": 5.309782415454151e-06,
|
| 1548 |
+
"loss": 0.0289,
|
| 1549 |
+
"step": 2200
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 8.6328125,
|
| 1553 |
+
"grad_norm": 0.2712879776954651,
|
| 1554 |
+
"learning_rate": 5.023866638505031e-06,
|
| 1555 |
+
"loss": 0.0354,
|
| 1556 |
+
"step": 2210
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 8.671875,
|
| 1560 |
+
"grad_norm": 0.26241105794906616,
|
| 1561 |
+
"learning_rate": 4.745455826725825e-06,
|
| 1562 |
+
"loss": 0.0316,
|
| 1563 |
+
"step": 2220
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 8.7109375,
|
| 1567 |
+
"grad_norm": 0.2872694134712219,
|
| 1568 |
+
"learning_rate": 4.474596437276996e-06,
|
| 1569 |
+
"loss": 0.03,
|
| 1570 |
+
"step": 2230
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 8.75,
|
| 1574 |
+
"grad_norm": 0.28568607568740845,
|
| 1575 |
+
"learning_rate": 4.2113336672471245e-06,
|
| 1576 |
+
"loss": 0.0294,
|
| 1577 |
+
"step": 2240
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 8.7890625,
|
| 1581 |
+
"grad_norm": 0.3385384976863861,
|
| 1582 |
+
"learning_rate": 3.955711446111082e-06,
|
| 1583 |
+
"loss": 0.0304,
|
| 1584 |
+
"step": 2250
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 8.828125,
|
| 1588 |
+
"grad_norm": 0.675881564617157,
|
| 1589 |
+
"learning_rate": 3.707772428399681e-06,
|
| 1590 |
+
"loss": 0.0334,
|
| 1591 |
+
"step": 2260
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 8.8671875,
|
| 1595 |
+
"grad_norm": 0.5481550097465515,
|
| 1596 |
+
"learning_rate": 3.467557986582104e-06,
|
| 1597 |
+
"loss": 0.0297,
|
| 1598 |
+
"step": 2270
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 8.90625,
|
| 1602 |
+
"grad_norm": 0.26067742705345154,
|
| 1603 |
+
"learning_rate": 3.2351082041623125e-06,
|
| 1604 |
+
"loss": 0.0287,
|
| 1605 |
+
"step": 2280
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 8.9453125,
|
| 1609 |
+
"grad_norm": 0.5334944128990173,
|
| 1610 |
+
"learning_rate": 3.0104618689904386e-06,
|
| 1611 |
+
"loss": 0.0322,
|
| 1612 |
+
"step": 2290
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 8.984375,
|
| 1616 |
+
"grad_norm": 0.37815043330192566,
|
| 1617 |
+
"learning_rate": 2.793656466790473e-06,
|
| 1618 |
+
"loss": 0.0255,
|
| 1619 |
+
"step": 2300
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 9.0234375,
|
| 1623 |
+
"grad_norm": 0.1792028248310089,
|
| 1624 |
+
"learning_rate": 2.584728174905171e-06,
|
| 1625 |
+
"loss": 0.029,
|
| 1626 |
+
"step": 2310
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 9.0625,
|
| 1630 |
+
"grad_norm": 0.8325735926628113,
|
| 1631 |
+
"learning_rate": 2.3837118562592797e-06,
|
| 1632 |
+
"loss": 0.0301,
|
| 1633 |
+
"step": 2320
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 9.1015625,
|
| 1637 |
+
"grad_norm": 0.28852543234825134,
|
| 1638 |
+
"learning_rate": 2.1906410535421697e-06,
|
| 1639 |
+
"loss": 0.0303,
|
| 1640 |
+
"step": 2330
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 9.140625,
|
| 1644 |
+
"grad_norm": 0.5437615513801575,
|
| 1645 |
+
"learning_rate": 2.005547983610684e-06,
|
| 1646 |
+
"loss": 0.0319,
|
| 1647 |
+
"step": 2340
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 9.1796875,
|
| 1651 |
+
"grad_norm": 0.2050412893295288,
|
| 1652 |
+
"learning_rate": 1.828463532113256e-06,
|
| 1653 |
+
"loss": 0.029,
|
| 1654 |
+
"step": 2350
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 9.21875,
|
| 1658 |
+
"grad_norm": 0.38466140627861023,
|
| 1659 |
+
"learning_rate": 1.6594172483361758e-06,
|
| 1660 |
+
"loss": 0.0261,
|
| 1661 |
+
"step": 2360
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 9.2578125,
|
| 1665 |
+
"grad_norm": 0.30740851163864136,
|
| 1666 |
+
"learning_rate": 1.4984373402728014e-06,
|
| 1667 |
+
"loss": 0.0258,
|
| 1668 |
+
"step": 2370
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 9.296875,
|
| 1672 |
+
"grad_norm": 0.33520451188087463,
|
| 1673 |
+
"learning_rate": 1.3455506699166732e-06,
|
| 1674 |
+
"loss": 0.0305,
|
| 1675 |
+
"step": 2380
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 9.3359375,
|
| 1679 |
+
"grad_norm": 0.2344502955675125,
|
| 1680 |
+
"learning_rate": 1.2007827487791112e-06,
|
| 1681 |
+
"loss": 0.0307,
|
| 1682 |
+
"step": 2390
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 9.375,
|
| 1686 |
+
"grad_norm": 0.4900089502334595,
|
| 1687 |
+
"learning_rate": 1.064157733632276e-06,
|
| 1688 |
+
"loss": 0.0305,
|
| 1689 |
+
"step": 2400
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 9.4140625,
|
| 1693 |
+
"grad_norm": 0.34539440274238586,
|
| 1694 |
+
"learning_rate": 9.35698422478204e-07,
|
| 1695 |
+
"loss": 0.0272,
|
| 1696 |
+
"step": 2410
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 9.453125,
|
| 1700 |
+
"grad_norm": 0.4583103358745575,
|
| 1701 |
+
"learning_rate": 8.154262507446109e-07,
|
| 1702 |
+
"loss": 0.0279,
|
| 1703 |
+
"step": 2420
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 9.4921875,
|
| 1707 |
+
"grad_norm": 0.39252835512161255,
|
| 1708 |
+
"learning_rate": 7.033612877080819e-07,
|
| 1709 |
+
"loss": 0.0339,
|
| 1710 |
+
"step": 2430
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 9.53125,
|
| 1714 |
+
"grad_norm": 0.6091126799583435,
|
| 1715 |
+
"learning_rate": 5.995222331451722e-07,
|
| 1716 |
+
"loss": 0.028,
|
| 1717 |
+
"step": 2440
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 9.5703125,
|
| 1721 |
+
"grad_norm": 0.36491745710372925,
|
| 1722 |
+
"learning_rate": 5.039264142120803e-07,
|
| 1723 |
+
"loss": 0.0295,
|
| 1724 |
+
"step": 2450
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 9.609375,
|
| 1728 |
+
"grad_norm": 0.31240883469581604,
|
| 1729 |
+
"learning_rate": 4.165897825533227e-07,
|
| 1730 |
+
"loss": 0.028,
|
| 1731 |
+
"step": 2460
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 9.6484375,
|
| 1735 |
+
"grad_norm": 0.35602691769599915,
|
| 1736 |
+
"learning_rate": 3.375269116399793e-07,
|
| 1737 |
+
"loss": 0.0359,
|
| 1738 |
+
"step": 2470
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 9.6875,
|
| 1742 |
+
"grad_norm": 0.2611185312271118,
|
| 1743 |
+
"learning_rate": 2.667509943378721e-07,
|
| 1744 |
+
"loss": 0.0244,
|
| 1745 |
+
"step": 2480
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 9.7265625,
|
| 1749 |
+
"grad_norm": 0.3753460943698883,
|
| 1750 |
+
"learning_rate": 2.0427384070612065e-07,
|
| 1751 |
+
"loss": 0.0278,
|
| 1752 |
+
"step": 2490
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 9.765625,
|
| 1756 |
+
"grad_norm": 0.4795132577419281,
|
| 1757 |
+
"learning_rate": 1.5010587602647397e-07,
|
| 1758 |
+
"loss": 0.0324,
|
| 1759 |
+
"step": 2500
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 9.8046875,
|
| 1763 |
+
"grad_norm": 0.4804089069366455,
|
| 1764 |
+
"learning_rate": 1.0425613906365761e-07,
|
| 1765 |
+
"loss": 0.0289,
|
| 1766 |
+
"step": 2510
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 9.84375,
|
| 1770 |
+
"grad_norm": 0.3166308104991913,
|
| 1771 |
+
"learning_rate": 6.673228055715241e-08,
|
| 1772 |
+
"loss": 0.0289,
|
| 1773 |
+
"step": 2520
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 9.8828125,
|
| 1777 |
+
"grad_norm": 0.3317022919654846,
|
| 1778 |
+
"learning_rate": 3.754056194452682e-08,
|
| 1779 |
+
"loss": 0.0327,
|
| 1780 |
+
"step": 2530
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 9.921875,
|
| 1784 |
+
"grad_norm": 0.2820194363594055,
|
| 1785 |
+
"learning_rate": 1.668585431663372e-08,
|
| 1786 |
+
"loss": 0.0268,
|
| 1787 |
+
"step": 2540
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 9.9609375,
|
| 1791 |
+
"grad_norm": 0.3767685890197754,
|
| 1792 |
+
"learning_rate": 4.171637604760692e-09,
|
| 1793 |
+
"loss": 0.0298,
|
| 1794 |
+
"step": 2550
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 10.0,
|
| 1798 |
+
"grad_norm": 0.375841349363327,
|
| 1799 |
+
"learning_rate": 0.0,
|
| 1800 |
+
"loss": 0.0277,
|
| 1801 |
+
"step": 2560
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 10.0,
|
| 1805 |
+
"step": 2560,
|
| 1806 |
+
"total_flos": 2.6088527768247245e+17,
|
| 1807 |
+
"train_loss": 0.060091388958971946,
|
| 1808 |
+
"train_runtime": 2712.9359,
|
| 1809 |
+
"train_samples_per_second": 46.127,
|
| 1810 |
+
"train_steps_per_second": 0.944
|
| 1811 |
+
}
|
| 1812 |
+
],
|
| 1813 |
+
"logging_steps": 10,
|
| 1814 |
+
"max_steps": 2560,
|
| 1815 |
+
"num_input_tokens_seen": 0,
|
| 1816 |
+
"num_train_epochs": 10,
|
| 1817 |
+
"save_steps": 10000,
|
| 1818 |
+
"stateful_callbacks": {
|
| 1819 |
+
"TrainerControl": {
|
| 1820 |
+
"args": {
|
| 1821 |
+
"should_epoch_stop": false,
|
| 1822 |
+
"should_evaluate": false,
|
| 1823 |
+
"should_log": false,
|
| 1824 |
+
"should_save": true,
|
| 1825 |
+
"should_training_stop": true
|
| 1826 |
+
},
|
| 1827 |
+
"attributes": {}
|
| 1828 |
+
}
|
| 1829 |
+
},
|
| 1830 |
+
"total_flos": 2.6088527768247245e+17,
|
| 1831 |
+
"train_batch_size": 49,
|
| 1832 |
+
"trial_name": null,
|
| 1833 |
+
"trial_params": null
|
| 1834 |
+
}
|