Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +2604 -0
trainer_state.json
ADDED
@@ -0,0 +1,2604 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 20.0,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 3660,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.0546448087431694,
|
13 |
+
"grad_norm": 14.50916862487793,
|
14 |
+
"learning_rate": 5.46448087431694e-06,
|
15 |
+
"loss": 1.7462,
|
16 |
+
"step": 10
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.1092896174863388,
|
20 |
+
"grad_norm": 9.00194263458252,
|
21 |
+
"learning_rate": 1.092896174863388e-05,
|
22 |
+
"loss": 1.397,
|
23 |
+
"step": 20
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.16393442622950818,
|
27 |
+
"grad_norm": 2.302015781402588,
|
28 |
+
"learning_rate": 1.6393442622950818e-05,
|
29 |
+
"loss": 0.5593,
|
30 |
+
"step": 30
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.2185792349726776,
|
34 |
+
"grad_norm": 2.4071569442749023,
|
35 |
+
"learning_rate": 2.185792349726776e-05,
|
36 |
+
"loss": 0.3376,
|
37 |
+
"step": 40
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.273224043715847,
|
41 |
+
"grad_norm": 1.562657117843628,
|
42 |
+
"learning_rate": 2.7322404371584703e-05,
|
43 |
+
"loss": 0.2425,
|
44 |
+
"step": 50
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.32786885245901637,
|
48 |
+
"grad_norm": 1.5590847730636597,
|
49 |
+
"learning_rate": 3.2786885245901635e-05,
|
50 |
+
"loss": 0.2093,
|
51 |
+
"step": 60
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.3825136612021858,
|
55 |
+
"grad_norm": 1.3504713773727417,
|
56 |
+
"learning_rate": 3.825136612021858e-05,
|
57 |
+
"loss": 0.1947,
|
58 |
+
"step": 70
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.4371584699453552,
|
62 |
+
"grad_norm": 1.1426506042480469,
|
63 |
+
"learning_rate": 4.371584699453552e-05,
|
64 |
+
"loss": 0.1745,
|
65 |
+
"step": 80
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.4918032786885246,
|
69 |
+
"grad_norm": 1.2180607318878174,
|
70 |
+
"learning_rate": 4.918032786885246e-05,
|
71 |
+
"loss": 0.1541,
|
72 |
+
"step": 90
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.546448087431694,
|
76 |
+
"grad_norm": 1.4522769451141357,
|
77 |
+
"learning_rate": 5.4644808743169406e-05,
|
78 |
+
"loss": 0.1492,
|
79 |
+
"step": 100
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.6010928961748634,
|
83 |
+
"grad_norm": 1.8644955158233643,
|
84 |
+
"learning_rate": 6.010928961748634e-05,
|
85 |
+
"loss": 0.134,
|
86 |
+
"step": 110
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.6557377049180327,
|
90 |
+
"grad_norm": 0.8796963095664978,
|
91 |
+
"learning_rate": 6.557377049180327e-05,
|
92 |
+
"loss": 0.1363,
|
93 |
+
"step": 120
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.7103825136612022,
|
97 |
+
"grad_norm": 0.7654175758361816,
|
98 |
+
"learning_rate": 7.103825136612023e-05,
|
99 |
+
"loss": 0.1299,
|
100 |
+
"step": 130
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.7650273224043715,
|
104 |
+
"grad_norm": 0.8335488438606262,
|
105 |
+
"learning_rate": 7.650273224043716e-05,
|
106 |
+
"loss": 0.1308,
|
107 |
+
"step": 140
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 0.819672131147541,
|
111 |
+
"grad_norm": 1.0280165672302246,
|
112 |
+
"learning_rate": 8.19672131147541e-05,
|
113 |
+
"loss": 0.1134,
|
114 |
+
"step": 150
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.8743169398907104,
|
118 |
+
"grad_norm": 1.133649468421936,
|
119 |
+
"learning_rate": 8.743169398907104e-05,
|
120 |
+
"loss": 0.1128,
|
121 |
+
"step": 160
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 0.9289617486338798,
|
125 |
+
"grad_norm": 0.851076602935791,
|
126 |
+
"learning_rate": 9.289617486338798e-05,
|
127 |
+
"loss": 0.1044,
|
128 |
+
"step": 170
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.9836065573770492,
|
132 |
+
"grad_norm": 0.8838961720466614,
|
133 |
+
"learning_rate": 9.836065573770493e-05,
|
134 |
+
"loss": 0.1092,
|
135 |
+
"step": 180
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 1.0382513661202186,
|
139 |
+
"grad_norm": 0.7312606573104858,
|
140 |
+
"learning_rate": 9.999899994242224e-05,
|
141 |
+
"loss": 0.1036,
|
142 |
+
"step": 190
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 1.092896174863388,
|
146 |
+
"grad_norm": 0.835260808467865,
|
147 |
+
"learning_rate": 9.999410179752908e-05,
|
148 |
+
"loss": 0.0993,
|
149 |
+
"step": 200
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 1.1475409836065573,
|
153 |
+
"grad_norm": 0.8507461547851562,
|
154 |
+
"learning_rate": 9.998512228064598e-05,
|
155 |
+
"loss": 0.0938,
|
156 |
+
"step": 210
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 1.2021857923497268,
|
160 |
+
"grad_norm": 0.8367434740066528,
|
161 |
+
"learning_rate": 9.99720621248344e-05,
|
162 |
+
"loss": 0.0874,
|
163 |
+
"step": 220
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 1.2568306010928962,
|
167 |
+
"grad_norm": 0.8626418113708496,
|
168 |
+
"learning_rate": 9.995492239628717e-05,
|
169 |
+
"loss": 0.0921,
|
170 |
+
"step": 230
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 1.3114754098360657,
|
174 |
+
"grad_norm": 0.94256991147995,
|
175 |
+
"learning_rate": 9.993370449424153e-05,
|
176 |
+
"loss": 0.0874,
|
177 |
+
"step": 240
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 1.366120218579235,
|
181 |
+
"grad_norm": 0.7513673305511475,
|
182 |
+
"learning_rate": 9.990841015086483e-05,
|
183 |
+
"loss": 0.0856,
|
184 |
+
"step": 250
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 1.4207650273224044,
|
188 |
+
"grad_norm": 0.6465060114860535,
|
189 |
+
"learning_rate": 9.987904143111314e-05,
|
190 |
+
"loss": 0.0879,
|
191 |
+
"step": 260
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 1.4754098360655736,
|
195 |
+
"grad_norm": 0.8078054785728455,
|
196 |
+
"learning_rate": 9.984560073256271e-05,
|
197 |
+
"loss": 0.0913,
|
198 |
+
"step": 270
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 1.530054644808743,
|
202 |
+
"grad_norm": 0.4887768030166626,
|
203 |
+
"learning_rate": 9.980809078521416e-05,
|
204 |
+
"loss": 0.0848,
|
205 |
+
"step": 280
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 1.5846994535519126,
|
209 |
+
"grad_norm": 0.593806266784668,
|
210 |
+
"learning_rate": 9.976651465126973e-05,
|
211 |
+
"loss": 0.0805,
|
212 |
+
"step": 290
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 1.639344262295082,
|
216 |
+
"grad_norm": 0.7595932483673096,
|
217 |
+
"learning_rate": 9.972087572488315e-05,
|
218 |
+
"loss": 0.085,
|
219 |
+
"step": 300
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 1.6939890710382515,
|
223 |
+
"grad_norm": 0.5838643312454224,
|
224 |
+
"learning_rate": 9.967117773188265e-05,
|
225 |
+
"loss": 0.0802,
|
226 |
+
"step": 310
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 1.748633879781421,
|
230 |
+
"grad_norm": 1.192397117614746,
|
231 |
+
"learning_rate": 9.96174247294668e-05,
|
232 |
+
"loss": 0.0803,
|
233 |
+
"step": 320
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 1.8032786885245902,
|
237 |
+
"grad_norm": 0.6883787512779236,
|
238 |
+
"learning_rate": 9.955962110587318e-05,
|
239 |
+
"loss": 0.0771,
|
240 |
+
"step": 330
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 1.8579234972677594,
|
244 |
+
"grad_norm": 0.5919109582901001,
|
245 |
+
"learning_rate": 9.949777158002027e-05,
|
246 |
+
"loss": 0.071,
|
247 |
+
"step": 340
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 1.9125683060109289,
|
251 |
+
"grad_norm": 0.7835716605186462,
|
252 |
+
"learning_rate": 9.943188120112214e-05,
|
253 |
+
"loss": 0.0865,
|
254 |
+
"step": 350
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 1.9672131147540983,
|
258 |
+
"grad_norm": 0.6744383573532104,
|
259 |
+
"learning_rate": 9.936195534827629e-05,
|
260 |
+
"loss": 0.081,
|
261 |
+
"step": 360
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 2.021857923497268,
|
265 |
+
"grad_norm": 0.5261861681938171,
|
266 |
+
"learning_rate": 9.928799973002445e-05,
|
267 |
+
"loss": 0.0726,
|
268 |
+
"step": 370
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 2.0765027322404372,
|
272 |
+
"grad_norm": 0.7296725511550903,
|
273 |
+
"learning_rate": 9.92100203838866e-05,
|
274 |
+
"loss": 0.0723,
|
275 |
+
"step": 380
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 2.1311475409836067,
|
279 |
+
"grad_norm": 0.5582314729690552,
|
280 |
+
"learning_rate": 9.91280236758681e-05,
|
281 |
+
"loss": 0.0712,
|
282 |
+
"step": 390
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 2.185792349726776,
|
286 |
+
"grad_norm": 0.41796624660491943,
|
287 |
+
"learning_rate": 9.904201629993993e-05,
|
288 |
+
"loss": 0.0833,
|
289 |
+
"step": 400
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 2.240437158469945,
|
293 |
+
"grad_norm": 0.46099790930747986,
|
294 |
+
"learning_rate": 9.895200527749227e-05,
|
295 |
+
"loss": 0.0651,
|
296 |
+
"step": 410
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 2.2950819672131146,
|
300 |
+
"grad_norm": 0.5075456500053406,
|
301 |
+
"learning_rate": 9.885799795676126e-05,
|
302 |
+
"loss": 0.0675,
|
303 |
+
"step": 420
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 2.349726775956284,
|
307 |
+
"grad_norm": 0.5505486130714417,
|
308 |
+
"learning_rate": 9.876000201222912e-05,
|
309 |
+
"loss": 0.0672,
|
310 |
+
"step": 430
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 2.4043715846994536,
|
314 |
+
"grad_norm": 1.0104068517684937,
|
315 |
+
"learning_rate": 9.865802544399767e-05,
|
316 |
+
"loss": 0.0772,
|
317 |
+
"step": 440
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 2.459016393442623,
|
321 |
+
"grad_norm": 0.9975788593292236,
|
322 |
+
"learning_rate": 9.855207657713511e-05,
|
323 |
+
"loss": 0.0707,
|
324 |
+
"step": 450
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 2.5136612021857925,
|
328 |
+
"grad_norm": 0.5779078006744385,
|
329 |
+
"learning_rate": 9.844216406099654e-05,
|
330 |
+
"loss": 0.073,
|
331 |
+
"step": 460
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 2.5683060109289615,
|
335 |
+
"grad_norm": 0.8848289847373962,
|
336 |
+
"learning_rate": 9.83282968685177e-05,
|
337 |
+
"loss": 0.0653,
|
338 |
+
"step": 470
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 2.6229508196721314,
|
342 |
+
"grad_norm": 0.6772857308387756,
|
343 |
+
"learning_rate": 9.821048429548258e-05,
|
344 |
+
"loss": 0.0659,
|
345 |
+
"step": 480
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 2.6775956284153004,
|
349 |
+
"grad_norm": 0.5181351900100708,
|
350 |
+
"learning_rate": 9.808873595976447e-05,
|
351 |
+
"loss": 0.0638,
|
352 |
+
"step": 490
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 2.73224043715847,
|
356 |
+
"grad_norm": 0.7094351053237915,
|
357 |
+
"learning_rate": 9.796306180054076e-05,
|
358 |
+
"loss": 0.0614,
|
359 |
+
"step": 500
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 2.7868852459016393,
|
363 |
+
"grad_norm": 0.5130027532577515,
|
364 |
+
"learning_rate": 9.783347207748158e-05,
|
365 |
+
"loss": 0.0625,
|
366 |
+
"step": 510
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 2.841530054644809,
|
370 |
+
"grad_norm": 0.7600369453430176,
|
371 |
+
"learning_rate": 9.769997736991226e-05,
|
372 |
+
"loss": 0.0671,
|
373 |
+
"step": 520
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 2.8961748633879782,
|
377 |
+
"grad_norm": 0.6283718943595886,
|
378 |
+
"learning_rate": 9.756258857594956e-05,
|
379 |
+
"loss": 0.0631,
|
380 |
+
"step": 530
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 2.9508196721311473,
|
384 |
+
"grad_norm": 0.5392641425132751,
|
385 |
+
"learning_rate": 9.74213169116121e-05,
|
386 |
+
"loss": 0.0652,
|
387 |
+
"step": 540
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 3.0054644808743167,
|
391 |
+
"grad_norm": 0.33834314346313477,
|
392 |
+
"learning_rate": 9.72761739099046e-05,
|
393 |
+
"loss": 0.0597,
|
394 |
+
"step": 550
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 3.060109289617486,
|
398 |
+
"grad_norm": 0.4568038284778595,
|
399 |
+
"learning_rate": 9.712717141987651e-05,
|
400 |
+
"loss": 0.0666,
|
401 |
+
"step": 560
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 3.1147540983606556,
|
405 |
+
"grad_norm": 0.7953147888183594,
|
406 |
+
"learning_rate": 9.69743216056545e-05,
|
407 |
+
"loss": 0.0631,
|
408 |
+
"step": 570
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 3.169398907103825,
|
412 |
+
"grad_norm": 0.8522589206695557,
|
413 |
+
"learning_rate": 9.681763694544957e-05,
|
414 |
+
"loss": 0.061,
|
415 |
+
"step": 580
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 3.2240437158469946,
|
419 |
+
"grad_norm": 0.8012239933013916,
|
420 |
+
"learning_rate": 9.66571302305383e-05,
|
421 |
+
"loss": 0.0626,
|
422 |
+
"step": 590
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 3.278688524590164,
|
426 |
+
"grad_norm": 0.6893075108528137,
|
427 |
+
"learning_rate": 9.649281456421862e-05,
|
428 |
+
"loss": 0.0546,
|
429 |
+
"step": 600
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 3.3333333333333335,
|
433 |
+
"grad_norm": 0.5718370079994202,
|
434 |
+
"learning_rate": 9.632470336074009e-05,
|
435 |
+
"loss": 0.0563,
|
436 |
+
"step": 610
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"epoch": 3.387978142076503,
|
440 |
+
"grad_norm": 0.8569139242172241,
|
441 |
+
"learning_rate": 9.61528103442088e-05,
|
442 |
+
"loss": 0.0613,
|
443 |
+
"step": 620
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 3.442622950819672,
|
447 |
+
"grad_norm": 0.685554563999176,
|
448 |
+
"learning_rate": 9.5977149547467e-05,
|
449 |
+
"loss": 0.0615,
|
450 |
+
"step": 630
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 3.4972677595628414,
|
454 |
+
"grad_norm": 0.7499738335609436,
|
455 |
+
"learning_rate": 9.579773531094742e-05,
|
456 |
+
"loss": 0.0635,
|
457 |
+
"step": 640
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 3.551912568306011,
|
461 |
+
"grad_norm": 0.9926331043243408,
|
462 |
+
"learning_rate": 9.561458228150266e-05,
|
463 |
+
"loss": 0.0627,
|
464 |
+
"step": 650
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 3.6065573770491803,
|
468 |
+
"grad_norm": 0.6195826530456543,
|
469 |
+
"learning_rate": 9.542770541120945e-05,
|
470 |
+
"loss": 0.0591,
|
471 |
+
"step": 660
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 3.66120218579235,
|
475 |
+
"grad_norm": 0.4464719891548157,
|
476 |
+
"learning_rate": 9.523711995614788e-05,
|
477 |
+
"loss": 0.0552,
|
478 |
+
"step": 670
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 3.7158469945355193,
|
482 |
+
"grad_norm": 1.1997320652008057,
|
483 |
+
"learning_rate": 9.504284147515615e-05,
|
484 |
+
"loss": 0.0593,
|
485 |
+
"step": 680
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 3.7704918032786887,
|
489 |
+
"grad_norm": 0.5045487284660339,
|
490 |
+
"learning_rate": 9.484488582856018e-05,
|
491 |
+
"loss": 0.0551,
|
492 |
+
"step": 690
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 3.8251366120218577,
|
496 |
+
"grad_norm": 0.5805765986442566,
|
497 |
+
"learning_rate": 9.464326917687898e-05,
|
498 |
+
"loss": 0.061,
|
499 |
+
"step": 700
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 3.879781420765027,
|
503 |
+
"grad_norm": 0.7676146030426025,
|
504 |
+
"learning_rate": 9.443800797950523e-05,
|
505 |
+
"loss": 0.055,
|
506 |
+
"step": 710
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 3.9344262295081966,
|
510 |
+
"grad_norm": 0.600481390953064,
|
511 |
+
"learning_rate": 9.422911899336174e-05,
|
512 |
+
"loss": 0.0566,
|
513 |
+
"step": 720
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 3.989071038251366,
|
517 |
+
"grad_norm": 0.6507868766784668,
|
518 |
+
"learning_rate": 9.401661927153323e-05,
|
519 |
+
"loss": 0.0598,
|
520 |
+
"step": 730
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 4.043715846994536,
|
524 |
+
"grad_norm": 0.6208195090293884,
|
525 |
+
"learning_rate": 9.380052616187441e-05,
|
526 |
+
"loss": 0.0622,
|
527 |
+
"step": 740
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 4.098360655737705,
|
531 |
+
"grad_norm": 0.45560985803604126,
|
532 |
+
"learning_rate": 9.35808573055936e-05,
|
533 |
+
"loss": 0.0572,
|
534 |
+
"step": 750
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 4.1530054644808745,
|
538 |
+
"grad_norm": 0.5124120116233826,
|
539 |
+
"learning_rate": 9.335763063581262e-05,
|
540 |
+
"loss": 0.0511,
|
541 |
+
"step": 760
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 4.2076502732240435,
|
545 |
+
"grad_norm": 0.6840647459030151,
|
546 |
+
"learning_rate": 9.313086437610273e-05,
|
547 |
+
"loss": 0.0549,
|
548 |
+
"step": 770
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 4.262295081967213,
|
552 |
+
"grad_norm": 0.5145549178123474,
|
553 |
+
"learning_rate": 9.290057703899697e-05,
|
554 |
+
"loss": 0.0568,
|
555 |
+
"step": 780
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 4.316939890710382,
|
559 |
+
"grad_norm": 0.6335499286651611,
|
560 |
+
"learning_rate": 9.266678742447878e-05,
|
561 |
+
"loss": 0.0505,
|
562 |
+
"step": 790
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 4.371584699453552,
|
566 |
+
"grad_norm": 0.3851253092288971,
|
567 |
+
"learning_rate": 9.242951461844734e-05,
|
568 |
+
"loss": 0.0491,
|
569 |
+
"step": 800
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 4.426229508196721,
|
573 |
+
"grad_norm": 0.5057039856910706,
|
574 |
+
"learning_rate": 9.218877799115928e-05,
|
575 |
+
"loss": 0.051,
|
576 |
+
"step": 810
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 4.48087431693989,
|
580 |
+
"grad_norm": 0.6589703559875488,
|
581 |
+
"learning_rate": 9.194459719564755e-05,
|
582 |
+
"loss": 0.0564,
|
583 |
+
"step": 820
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 4.53551912568306,
|
587 |
+
"grad_norm": 0.5856819152832031,
|
588 |
+
"learning_rate": 9.169699216611683e-05,
|
589 |
+
"loss": 0.0547,
|
590 |
+
"step": 830
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 4.590163934426229,
|
594 |
+
"grad_norm": 0.4095700681209564,
|
595 |
+
"learning_rate": 9.144598311631626e-05,
|
596 |
+
"loss": 0.0566,
|
597 |
+
"step": 840
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 4.644808743169399,
|
601 |
+
"grad_norm": 0.4853191673755646,
|
602 |
+
"learning_rate": 9.119159053788922e-05,
|
603 |
+
"loss": 0.0559,
|
604 |
+
"step": 850
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 4.699453551912568,
|
608 |
+
"grad_norm": 0.2956223487854004,
|
609 |
+
"learning_rate": 9.093383519870047e-05,
|
610 |
+
"loss": 0.055,
|
611 |
+
"step": 860
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 4.754098360655737,
|
615 |
+
"grad_norm": 0.8387729525566101,
|
616 |
+
"learning_rate": 9.067273814114066e-05,
|
617 |
+
"loss": 0.0483,
|
618 |
+
"step": 870
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 4.808743169398907,
|
622 |
+
"grad_norm": 0.5920037627220154,
|
623 |
+
"learning_rate": 9.040832068040857e-05,
|
624 |
+
"loss": 0.0483,
|
625 |
+
"step": 880
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 4.863387978142076,
|
629 |
+
"grad_norm": 0.6177905201911926,
|
630 |
+
"learning_rate": 9.014060440277097e-05,
|
631 |
+
"loss": 0.0506,
|
632 |
+
"step": 890
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 4.918032786885246,
|
636 |
+
"grad_norm": 0.5852020978927612,
|
637 |
+
"learning_rate": 8.986961116380037e-05,
|
638 |
+
"loss": 0.0494,
|
639 |
+
"step": 900
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 4.972677595628415,
|
643 |
+
"grad_norm": 0.6510049700737,
|
644 |
+
"learning_rate": 8.95953630865908e-05,
|
645 |
+
"loss": 0.0481,
|
646 |
+
"step": 910
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 5.027322404371585,
|
650 |
+
"grad_norm": 0.7334024906158447,
|
651 |
+
"learning_rate": 8.931788255995175e-05,
|
652 |
+
"loss": 0.0465,
|
653 |
+
"step": 920
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 5.081967213114754,
|
657 |
+
"grad_norm": 0.8580073714256287,
|
658 |
+
"learning_rate": 8.903719223658038e-05,
|
659 |
+
"loss": 0.0502,
|
660 |
+
"step": 930
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 5.136612021857924,
|
664 |
+
"grad_norm": 0.7086485624313354,
|
665 |
+
"learning_rate": 8.875331503121231e-05,
|
666 |
+
"loss": 0.0514,
|
667 |
+
"step": 940
|
668 |
+
},
|
669 |
+
{
|
670 |
+
"epoch": 5.191256830601093,
|
671 |
+
"grad_norm": 0.49265387654304504,
|
672 |
+
"learning_rate": 8.846627411875081e-05,
|
673 |
+
"loss": 0.0499,
|
674 |
+
"step": 950
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"epoch": 5.245901639344262,
|
678 |
+
"grad_norm": 0.5628988742828369,
|
679 |
+
"learning_rate": 8.817609293237499e-05,
|
680 |
+
"loss": 0.0504,
|
681 |
+
"step": 960
|
682 |
+
},
|
683 |
+
{
|
684 |
+
"epoch": 5.300546448087432,
|
685 |
+
"grad_norm": 1.3476946353912354,
|
686 |
+
"learning_rate": 8.788279516162666e-05,
|
687 |
+
"loss": 0.0496,
|
688 |
+
"step": 970
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"epoch": 5.355191256830601,
|
692 |
+
"grad_norm": 0.4882979393005371,
|
693 |
+
"learning_rate": 8.758640475047649e-05,
|
694 |
+
"loss": 0.0511,
|
695 |
+
"step": 980
|
696 |
+
},
|
697 |
+
{
|
698 |
+
"epoch": 5.409836065573771,
|
699 |
+
"grad_norm": 0.3951756954193115,
|
700 |
+
"learning_rate": 8.72869458953692e-05,
|
701 |
+
"loss": 0.0529,
|
702 |
+
"step": 990
|
703 |
+
},
|
704 |
+
{
|
705 |
+
"epoch": 5.46448087431694,
|
706 |
+
"grad_norm": 0.4163540303707123,
|
707 |
+
"learning_rate": 8.698444304324835e-05,
|
708 |
+
"loss": 0.0491,
|
709 |
+
"step": 1000
|
710 |
+
},
|
711 |
+
{
|
712 |
+
"epoch": 5.51912568306011,
|
713 |
+
"grad_norm": 0.6048452258110046,
|
714 |
+
"learning_rate": 8.667892088956045e-05,
|
715 |
+
"loss": 0.0503,
|
716 |
+
"step": 1010
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"epoch": 5.573770491803279,
|
720 |
+
"grad_norm": 0.5346564054489136,
|
721 |
+
"learning_rate": 8.637040437623897e-05,
|
722 |
+
"loss": 0.0449,
|
723 |
+
"step": 1020
|
724 |
+
},
|
725 |
+
{
|
726 |
+
"epoch": 5.628415300546449,
|
727 |
+
"grad_norm": 0.35331907868385315,
|
728 |
+
"learning_rate": 8.60589186896681e-05,
|
729 |
+
"loss": 0.0435,
|
730 |
+
"step": 1030
|
731 |
+
},
|
732 |
+
{
|
733 |
+
"epoch": 5.683060109289618,
|
734 |
+
"grad_norm": 0.4379464387893677,
|
735 |
+
"learning_rate": 8.574448925862667e-05,
|
736 |
+
"loss": 0.0527,
|
737 |
+
"step": 1040
|
738 |
+
},
|
739 |
+
{
|
740 |
+
"epoch": 5.737704918032787,
|
741 |
+
"grad_norm": 0.45317304134368896,
|
742 |
+
"learning_rate": 8.542714175221216e-05,
|
743 |
+
"loss": 0.0463,
|
744 |
+
"step": 1050
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"epoch": 5.7923497267759565,
|
748 |
+
"grad_norm": 0.374124675989151,
|
749 |
+
"learning_rate": 8.510690207774517e-05,
|
750 |
+
"loss": 0.044,
|
751 |
+
"step": 1060
|
752 |
+
},
|
753 |
+
{
|
754 |
+
"epoch": 5.8469945355191255,
|
755 |
+
"grad_norm": 0.481268972158432,
|
756 |
+
"learning_rate": 8.47837963786544e-05,
|
757 |
+
"loss": 0.0492,
|
758 |
+
"step": 1070
|
759 |
+
},
|
760 |
+
{
|
761 |
+
"epoch": 5.901639344262295,
|
762 |
+
"grad_norm": 0.6521291732788086,
|
763 |
+
"learning_rate": 8.445785103234252e-05,
|
764 |
+
"loss": 0.0535,
|
765 |
+
"step": 1080
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"epoch": 5.956284153005464,
|
769 |
+
"grad_norm": 0.41385921835899353,
|
770 |
+
"learning_rate": 8.412909264803252e-05,
|
771 |
+
"loss": 0.0439,
|
772 |
+
"step": 1090
|
773 |
+
},
|
774 |
+
{
|
775 |
+
"epoch": 6.0109289617486334,
|
776 |
+
"grad_norm": 0.4521181285381317,
|
777 |
+
"learning_rate": 8.379754806459568e-05,
|
778 |
+
"loss": 0.0509,
|
779 |
+
"step": 1100
|
780 |
+
},
|
781 |
+
{
|
782 |
+
"epoch": 6.065573770491803,
|
783 |
+
"grad_norm": 0.295654296875,
|
784 |
+
"learning_rate": 8.346324434836038e-05,
|
785 |
+
"loss": 0.0425,
|
786 |
+
"step": 1110
|
787 |
+
},
|
788 |
+
{
|
789 |
+
"epoch": 6.120218579234972,
|
790 |
+
"grad_norm": 0.4982050359249115,
|
791 |
+
"learning_rate": 8.31262087909025e-05,
|
792 |
+
"loss": 0.0459,
|
793 |
+
"step": 1120
|
794 |
+
},
|
795 |
+
{
|
796 |
+
"epoch": 6.174863387978142,
|
797 |
+
"grad_norm": 0.9648367762565613,
|
798 |
+
"learning_rate": 8.278646890681746e-05,
|
799 |
+
"loss": 0.0501,
|
800 |
+
"step": 1130
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 6.229508196721311,
|
804 |
+
"grad_norm": 0.6268149018287659,
|
805 |
+
"learning_rate": 8.244405243147397e-05,
|
806 |
+
"loss": 0.0426,
|
807 |
+
"step": 1140
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 6.284153005464481,
|
811 |
+
"grad_norm": 1.036202311515808,
|
812 |
+
"learning_rate": 8.209898731874981e-05,
|
813 |
+
"loss": 0.047,
|
814 |
+
"step": 1150
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"epoch": 6.33879781420765,
|
818 |
+
"grad_norm": 0.6712212562561035,
|
819 |
+
"learning_rate": 8.175130173874975e-05,
|
820 |
+
"loss": 0.0485,
|
821 |
+
"step": 1160
|
822 |
+
},
|
823 |
+
{
|
824 |
+
"epoch": 6.39344262295082,
|
825 |
+
"grad_norm": 0.4752486050128937,
|
826 |
+
"learning_rate": 8.140102407550585e-05,
|
827 |
+
"loss": 0.0494,
|
828 |
+
"step": 1170
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 6.448087431693989,
|
832 |
+
"grad_norm": 0.7658908367156982,
|
833 |
+
"learning_rate": 8.104818292466022e-05,
|
834 |
+
"loss": 0.0422,
|
835 |
+
"step": 1180
|
836 |
+
},
|
837 |
+
{
|
838 |
+
"epoch": 6.502732240437158,
|
839 |
+
"grad_norm": 0.4015105366706848,
|
840 |
+
"learning_rate": 8.06928070911306e-05,
|
841 |
+
"loss": 0.041,
|
842 |
+
"step": 1190
|
843 |
+
},
|
844 |
+
{
|
845 |
+
"epoch": 6.557377049180328,
|
846 |
+
"grad_norm": 0.6179288029670715,
|
847 |
+
"learning_rate": 8.033492558675883e-05,
|
848 |
+
"loss": 0.0432,
|
849 |
+
"step": 1200
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 6.612021857923497,
|
853 |
+
"grad_norm": 0.4464608132839203,
|
854 |
+
"learning_rate": 7.997456762794232e-05,
|
855 |
+
"loss": 0.0415,
|
856 |
+
"step": 1210
|
857 |
+
},
|
858 |
+
{
|
859 |
+
"epoch": 6.666666666666667,
|
860 |
+
"grad_norm": 0.6367069482803345,
|
861 |
+
"learning_rate": 7.961176263324901e-05,
|
862 |
+
"loss": 0.045,
|
863 |
+
"step": 1220
|
864 |
+
},
|
865 |
+
{
|
866 |
+
"epoch": 6.721311475409836,
|
867 |
+
"grad_norm": 0.2885402739048004,
|
868 |
+
"learning_rate": 7.924654022101565e-05,
|
869 |
+
"loss": 0.0424,
|
870 |
+
"step": 1230
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"epoch": 6.775956284153006,
|
874 |
+
"grad_norm": 0.8153219819068909,
|
875 |
+
"learning_rate": 7.887893020692989e-05,
|
876 |
+
"loss": 0.047,
|
877 |
+
"step": 1240
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"epoch": 6.830601092896175,
|
881 |
+
"grad_norm": 0.312220960855484,
|
882 |
+
"learning_rate": 7.850896260159621e-05,
|
883 |
+
"loss": 0.0459,
|
884 |
+
"step": 1250
|
885 |
+
},
|
886 |
+
{
|
887 |
+
"epoch": 6.885245901639344,
|
888 |
+
"grad_norm": 0.48526033759117126,
|
889 |
+
"learning_rate": 7.81366676080859e-05,
|
890 |
+
"loss": 0.0452,
|
891 |
+
"step": 1260
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 6.939890710382514,
|
895 |
+
"grad_norm": 0.7081798315048218,
|
896 |
+
"learning_rate": 7.776207561947142e-05,
|
897 |
+
"loss": 0.0474,
|
898 |
+
"step": 1270
|
899 |
+
},
|
900 |
+
{
|
901 |
+
"epoch": 6.994535519125683,
|
902 |
+
"grad_norm": 0.5795289874076843,
|
903 |
+
"learning_rate": 7.738521721634522e-05,
|
904 |
+
"loss": 0.0367,
|
905 |
+
"step": 1280
|
906 |
+
},
|
907 |
+
{
|
908 |
+
"epoch": 7.049180327868853,
|
909 |
+
"grad_norm": 0.6600906848907471,
|
910 |
+
"learning_rate": 7.700612316432309e-05,
|
911 |
+
"loss": 0.0433,
|
912 |
+
"step": 1290
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"epoch": 7.103825136612022,
|
916 |
+
"grad_norm": 0.696719765663147,
|
917 |
+
"learning_rate": 7.662482441153274e-05,
|
918 |
+
"loss": 0.0461,
|
919 |
+
"step": 1300
|
920 |
+
},
|
921 |
+
{
|
922 |
+
"epoch": 7.158469945355192,
|
923 |
+
"grad_norm": 0.7991256713867188,
|
924 |
+
"learning_rate": 7.624135208608714e-05,
|
925 |
+
"loss": 0.039,
|
926 |
+
"step": 1310
|
927 |
+
},
|
928 |
+
{
|
929 |
+
"epoch": 7.213114754098361,
|
930 |
+
"grad_norm": 0.8350341320037842,
|
931 |
+
"learning_rate": 7.58557374935434e-05,
|
932 |
+
"loss": 0.0411,
|
933 |
+
"step": 1320
|
934 |
+
},
|
935 |
+
{
|
936 |
+
"epoch": 7.26775956284153,
|
937 |
+
"grad_norm": 0.8367980122566223,
|
938 |
+
"learning_rate": 7.546801211434702e-05,
|
939 |
+
"loss": 0.0457,
|
940 |
+
"step": 1330
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"epoch": 7.3224043715847,
|
944 |
+
"grad_norm": 0.7723026275634766,
|
945 |
+
"learning_rate": 7.507820760126193e-05,
|
946 |
+
"loss": 0.0449,
|
947 |
+
"step": 1340
|
948 |
+
},
|
949 |
+
{
|
950 |
+
"epoch": 7.377049180327869,
|
951 |
+
"grad_norm": 0.5425959229469299,
|
952 |
+
"learning_rate": 7.468635577678646e-05,
|
953 |
+
"loss": 0.0389,
|
954 |
+
"step": 1350
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 7.4316939890710385,
|
958 |
+
"grad_norm": 0.8735256195068359,
|
959 |
+
"learning_rate": 7.429248863055545e-05,
|
960 |
+
"loss": 0.0407,
|
961 |
+
"step": 1360
|
962 |
+
},
|
963 |
+
{
|
964 |
+
"epoch": 7.4863387978142075,
|
965 |
+
"grad_norm": 0.5877657532691956,
|
966 |
+
"learning_rate": 7.389663831672868e-05,
|
967 |
+
"loss": 0.0408,
|
968 |
+
"step": 1370
|
969 |
+
},
|
970 |
+
{
|
971 |
+
"epoch": 7.540983606557377,
|
972 |
+
"grad_norm": 0.40603581070899963,
|
973 |
+
"learning_rate": 7.3498837151366e-05,
|
974 |
+
"loss": 0.0377,
|
975 |
+
"step": 1380
|
976 |
+
},
|
977 |
+
{
|
978 |
+
"epoch": 7.595628415300546,
|
979 |
+
"grad_norm": 0.4147934019565582,
|
980 |
+
"learning_rate": 7.309911760978899e-05,
|
981 |
+
"loss": 0.0455,
|
982 |
+
"step": 1390
|
983 |
+
},
|
984 |
+
{
|
985 |
+
"epoch": 7.6502732240437155,
|
986 |
+
"grad_norm": 0.3599822223186493,
|
987 |
+
"learning_rate": 7.269751232392983e-05,
|
988 |
+
"loss": 0.0442,
|
989 |
+
"step": 1400
|
990 |
+
},
|
991 |
+
{
|
992 |
+
"epoch": 7.704918032786885,
|
993 |
+
"grad_norm": 0.4773503243923187,
|
994 |
+
"learning_rate": 7.229405407966739e-05,
|
995 |
+
"loss": 0.0386,
|
996 |
+
"step": 1410
|
997 |
+
},
|
998 |
+
{
|
999 |
+
"epoch": 7.759562841530054,
|
1000 |
+
"grad_norm": 0.46479761600494385,
|
1001 |
+
"learning_rate": 7.188877581415059e-05,
|
1002 |
+
"loss": 0.0378,
|
1003 |
+
"step": 1420
|
1004 |
+
},
|
1005 |
+
{
|
1006 |
+
"epoch": 7.814207650273224,
|
1007 |
+
"grad_norm": 0.3012316823005676,
|
1008 |
+
"learning_rate": 7.148171061310963e-05,
|
1009 |
+
"loss": 0.0386,
|
1010 |
+
"step": 1430
|
1011 |
+
},
|
1012 |
+
{
|
1013 |
+
"epoch": 7.868852459016393,
|
1014 |
+
"grad_norm": 0.7883979082107544,
|
1015 |
+
"learning_rate": 7.107289170815481e-05,
|
1016 |
+
"loss": 0.0345,
|
1017 |
+
"step": 1440
|
1018 |
+
},
|
1019 |
+
{
|
1020 |
+
"epoch": 7.923497267759563,
|
1021 |
+
"grad_norm": 0.4798952341079712,
|
1022 |
+
"learning_rate": 7.06623524740637e-05,
|
1023 |
+
"loss": 0.0326,
|
1024 |
+
"step": 1450
|
1025 |
+
},
|
1026 |
+
{
|
1027 |
+
"epoch": 7.978142076502732,
|
1028 |
+
"grad_norm": 0.8079819083213806,
|
1029 |
+
"learning_rate": 7.02501264260565e-05,
|
1030 |
+
"loss": 0.0418,
|
1031 |
+
"step": 1460
|
1032 |
+
},
|
1033 |
+
{
|
1034 |
+
"epoch": 8.032786885245901,
|
1035 |
+
"grad_norm": 0.2653636634349823,
|
1036 |
+
"learning_rate": 6.983624721705999e-05,
|
1037 |
+
"loss": 0.0389,
|
1038 |
+
"step": 1470
|
1039 |
+
},
|
1040 |
+
{
|
1041 |
+
"epoch": 8.087431693989071,
|
1042 |
+
"grad_norm": 0.48421579599380493,
|
1043 |
+
"learning_rate": 6.94207486349601e-05,
|
1044 |
+
"loss": 0.0391,
|
1045 |
+
"step": 1480
|
1046 |
+
},
|
1047 |
+
{
|
1048 |
+
"epoch": 8.142076502732241,
|
1049 |
+
"grad_norm": 0.7526752352714539,
|
1050 |
+
"learning_rate": 6.900366459984374e-05,
|
1051 |
+
"loss": 0.0395,
|
1052 |
+
"step": 1490
|
1053 |
+
},
|
1054 |
+
{
|
1055 |
+
"epoch": 8.19672131147541,
|
1056 |
+
"grad_norm": 0.36561110615730286,
|
1057 |
+
"learning_rate": 6.858502916122943e-05,
|
1058 |
+
"loss": 0.0336,
|
1059 |
+
"step": 1500
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 8.251366120218579,
|
1063 |
+
"grad_norm": 0.4030620753765106,
|
1064 |
+
"learning_rate": 6.816487649528783e-05,
|
1065 |
+
"loss": 0.0414,
|
1066 |
+
"step": 1510
|
1067 |
+
},
|
1068 |
+
{
|
1069 |
+
"epoch": 8.306010928961749,
|
1070 |
+
"grad_norm": 0.4381183683872223,
|
1071 |
+
"learning_rate": 6.774324090205155e-05,
|
1072 |
+
"loss": 0.0388,
|
1073 |
+
"step": 1520
|
1074 |
+
},
|
1075 |
+
{
|
1076 |
+
"epoch": 8.360655737704919,
|
1077 |
+
"grad_norm": 0.4365619122982025,
|
1078 |
+
"learning_rate": 6.732015680261505e-05,
|
1079 |
+
"loss": 0.0408,
|
1080 |
+
"step": 1530
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"epoch": 8.415300546448087,
|
1084 |
+
"grad_norm": 0.6897148489952087,
|
1085 |
+
"learning_rate": 6.689565873632458e-05,
|
1086 |
+
"loss": 0.0377,
|
1087 |
+
"step": 1540
|
1088 |
+
},
|
1089 |
+
{
|
1090 |
+
"epoch": 8.469945355191257,
|
1091 |
+
"grad_norm": 0.3831244111061096,
|
1092 |
+
"learning_rate": 6.646978135795852e-05,
|
1093 |
+
"loss": 0.0327,
|
1094 |
+
"step": 1550
|
1095 |
+
},
|
1096 |
+
{
|
1097 |
+
"epoch": 8.524590163934427,
|
1098 |
+
"grad_norm": 0.3431888222694397,
|
1099 |
+
"learning_rate": 6.604255943489822e-05,
|
1100 |
+
"loss": 0.0368,
|
1101 |
+
"step": 1560
|
1102 |
+
},
|
1103 |
+
{
|
1104 |
+
"epoch": 8.579234972677595,
|
1105 |
+
"grad_norm": 0.6473280191421509,
|
1106 |
+
"learning_rate": 6.561402784428974e-05,
|
1107 |
+
"loss": 0.0399,
|
1108 |
+
"step": 1570
|
1109 |
+
},
|
1110 |
+
{
|
1111 |
+
"epoch": 8.633879781420765,
|
1112 |
+
"grad_norm": 0.3908807039260864,
|
1113 |
+
"learning_rate": 6.518422157019658e-05,
|
1114 |
+
"loss": 0.0354,
|
1115 |
+
"step": 1580
|
1116 |
+
},
|
1117 |
+
{
|
1118 |
+
"epoch": 8.688524590163935,
|
1119 |
+
"grad_norm": 0.4484104812145233,
|
1120 |
+
"learning_rate": 6.475317570074357e-05,
|
1121 |
+
"loss": 0.035,
|
1122 |
+
"step": 1590
|
1123 |
+
},
|
1124 |
+
{
|
1125 |
+
"epoch": 8.743169398907105,
|
1126 |
+
"grad_norm": 0.5294371247291565,
|
1127 |
+
"learning_rate": 6.432092542525263e-05,
|
1128 |
+
"loss": 0.0373,
|
1129 |
+
"step": 1600
|
1130 |
+
},
|
1131 |
+
{
|
1132 |
+
"epoch": 8.797814207650273,
|
1133 |
+
"grad_norm": 0.35410311818122864,
|
1134 |
+
"learning_rate": 6.388750603136973e-05,
|
1135 |
+
"loss": 0.0377,
|
1136 |
+
"step": 1610
|
1137 |
+
},
|
1138 |
+
{
|
1139 |
+
"epoch": 8.852459016393443,
|
1140 |
+
"grad_norm": 0.9991357326507568,
|
1141 |
+
"learning_rate": 6.345295290218427e-05,
|
1142 |
+
"loss": 0.0373,
|
1143 |
+
"step": 1620
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 8.907103825136613,
|
1147 |
+
"grad_norm": 0.47563156485557556,
|
1148 |
+
"learning_rate": 6.301730151334054e-05,
|
1149 |
+
"loss": 0.032,
|
1150 |
+
"step": 1630
|
1151 |
+
},
|
1152 |
+
{
|
1153 |
+
"epoch": 8.96174863387978,
|
1154 |
+
"grad_norm": 0.4704972803592682,
|
1155 |
+
"learning_rate": 6.258058743014145e-05,
|
1156 |
+
"loss": 0.038,
|
1157 |
+
"step": 1640
|
1158 |
+
},
|
1159 |
+
{
|
1160 |
+
"epoch": 9.01639344262295,
|
1161 |
+
"grad_norm": 0.5292912721633911,
|
1162 |
+
"learning_rate": 6.214284630464521e-05,
|
1163 |
+
"loss": 0.0362,
|
1164 |
+
"step": 1650
|
1165 |
+
},
|
1166 |
+
{
|
1167 |
+
"epoch": 9.07103825136612,
|
1168 |
+
"grad_norm": 0.6291676759719849,
|
1169 |
+
"learning_rate": 6.170411387275472e-05,
|
1170 |
+
"loss": 0.0366,
|
1171 |
+
"step": 1660
|
1172 |
+
},
|
1173 |
+
{
|
1174 |
+
"epoch": 9.12568306010929,
|
1175 |
+
"grad_norm": 0.38353613018989563,
|
1176 |
+
"learning_rate": 6.126442595130027e-05,
|
1177 |
+
"loss": 0.0382,
|
1178 |
+
"step": 1670
|
1179 |
+
},
|
1180 |
+
{
|
1181 |
+
"epoch": 9.180327868852459,
|
1182 |
+
"grad_norm": 0.45063936710357666,
|
1183 |
+
"learning_rate": 6.08238184351155e-05,
|
1184 |
+
"loss": 0.0335,
|
1185 |
+
"step": 1680
|
1186 |
+
},
|
1187 |
+
{
|
1188 |
+
"epoch": 9.234972677595628,
|
1189 |
+
"grad_norm": 0.3961513042449951,
|
1190 |
+
"learning_rate": 6.0382327294107065e-05,
|
1191 |
+
"loss": 0.0356,
|
1192 |
+
"step": 1690
|
1193 |
+
},
|
1194 |
+
{
|
1195 |
+
"epoch": 9.289617486338798,
|
1196 |
+
"grad_norm": 0.2651735544204712,
|
1197 |
+
"learning_rate": 5.993998857031814e-05,
|
1198 |
+
"loss": 0.029,
|
1199 |
+
"step": 1700
|
1200 |
+
},
|
1201 |
+
{
|
1202 |
+
"epoch": 9.344262295081966,
|
1203 |
+
"grad_norm": 1.1500636339187622,
|
1204 |
+
"learning_rate": 5.949683837498614e-05,
|
1205 |
+
"loss": 0.0337,
|
1206 |
+
"step": 1710
|
1207 |
+
},
|
1208 |
+
{
|
1209 |
+
"epoch": 9.398907103825136,
|
1210 |
+
"grad_norm": 0.4949517548084259,
|
1211 |
+
"learning_rate": 5.905291288559458e-05,
|
1212 |
+
"loss": 0.0335,
|
1213 |
+
"step": 1720
|
1214 |
+
},
|
1215 |
+
{
|
1216 |
+
"epoch": 9.453551912568306,
|
1217 |
+
"grad_norm": 0.34546446800231934,
|
1218 |
+
"learning_rate": 5.860824834291973e-05,
|
1219 |
+
"loss": 0.0292,
|
1220 |
+
"step": 1730
|
1221 |
+
},
|
1222 |
+
{
|
1223 |
+
"epoch": 9.508196721311476,
|
1224 |
+
"grad_norm": 0.3333743214607239,
|
1225 |
+
"learning_rate": 5.816288104807199e-05,
|
1226 |
+
"loss": 0.0394,
|
1227 |
+
"step": 1740
|
1228 |
+
},
|
1229 |
+
{
|
1230 |
+
"epoch": 9.562841530054644,
|
1231 |
+
"grad_norm": 0.3905448913574219,
|
1232 |
+
"learning_rate": 5.771684735953242e-05,
|
1233 |
+
"loss": 0.0369,
|
1234 |
+
"step": 1750
|
1235 |
+
},
|
1236 |
+
{
|
1237 |
+
"epoch": 9.617486338797814,
|
1238 |
+
"grad_norm": 0.4974684715270996,
|
1239 |
+
"learning_rate": 5.7270183690184495e-05,
|
1240 |
+
"loss": 0.0352,
|
1241 |
+
"step": 1760
|
1242 |
+
},
|
1243 |
+
{
|
1244 |
+
"epoch": 9.672131147540984,
|
1245 |
+
"grad_norm": 0.4600943624973297,
|
1246 |
+
"learning_rate": 5.682292650434147e-05,
|
1247 |
+
"loss": 0.0376,
|
1248 |
+
"step": 1770
|
1249 |
+
},
|
1250 |
+
{
|
1251 |
+
"epoch": 9.726775956284152,
|
1252 |
+
"grad_norm": 0.6638236045837402,
|
1253 |
+
"learning_rate": 5.6375112314769526e-05,
|
1254 |
+
"loss": 0.0349,
|
1255 |
+
"step": 1780
|
1256 |
+
},
|
1257 |
+
{
|
1258 |
+
"epoch": 9.781420765027322,
|
1259 |
+
"grad_norm": 0.8875242471694946,
|
1260 |
+
"learning_rate": 5.592677767970704e-05,
|
1261 |
+
"loss": 0.038,
|
1262 |
+
"step": 1790
|
1263 |
+
},
|
1264 |
+
{
|
1265 |
+
"epoch": 9.836065573770492,
|
1266 |
+
"grad_norm": 0.983162522315979,
|
1267 |
+
"learning_rate": 5.547795919988004e-05,
|
1268 |
+
"loss": 0.0345,
|
1269 |
+
"step": 1800
|
1270 |
+
},
|
1271 |
+
{
|
1272 |
+
"epoch": 9.890710382513662,
|
1273 |
+
"grad_norm": 0.26490020751953125,
|
1274 |
+
"learning_rate": 5.502869351551414e-05,
|
1275 |
+
"loss": 0.0343,
|
1276 |
+
"step": 1810
|
1277 |
+
},
|
1278 |
+
{
|
1279 |
+
"epoch": 9.94535519125683,
|
1280 |
+
"grad_norm": 0.5392178297042847,
|
1281 |
+
"learning_rate": 5.457901730334354e-05,
|
1282 |
+
"loss": 0.0334,
|
1283 |
+
"step": 1820
|
1284 |
+
},
|
1285 |
+
{
|
1286 |
+
"epoch": 10.0,
|
1287 |
+
"grad_norm": 0.36645573377609253,
|
1288 |
+
"learning_rate": 5.4128967273616625e-05,
|
1289 |
+
"loss": 0.0345,
|
1290 |
+
"step": 1830
|
1291 |
+
},
|
1292 |
+
{
|
1293 |
+
"epoch": 10.05464480874317,
|
1294 |
+
"grad_norm": 0.6238542199134827,
|
1295 |
+
"learning_rate": 5.3678580167099214e-05,
|
1296 |
+
"loss": 0.0299,
|
1297 |
+
"step": 1840
|
1298 |
+
},
|
1299 |
+
{
|
1300 |
+
"epoch": 10.109289617486338,
|
1301 |
+
"grad_norm": 0.3955978453159332,
|
1302 |
+
"learning_rate": 5.322789275207505e-05,
|
1303 |
+
"loss": 0.0321,
|
1304 |
+
"step": 1850
|
1305 |
+
},
|
1306 |
+
{
|
1307 |
+
"epoch": 10.163934426229508,
|
1308 |
+
"grad_norm": 0.43886300921440125,
|
1309 |
+
"learning_rate": 5.277694182134422e-05,
|
1310 |
+
"loss": 0.0297,
|
1311 |
+
"step": 1860
|
1312 |
+
},
|
1313 |
+
{
|
1314 |
+
"epoch": 10.218579234972678,
|
1315 |
+
"grad_norm": 0.3649499714374542,
|
1316 |
+
"learning_rate": 5.232576418921944e-05,
|
1317 |
+
"loss": 0.0342,
|
1318 |
+
"step": 1870
|
1319 |
+
},
|
1320 |
+
{
|
1321 |
+
"epoch": 10.273224043715848,
|
1322 |
+
"grad_norm": 0.2916063964366913,
|
1323 |
+
"learning_rate": 5.187439668852063e-05,
|
1324 |
+
"loss": 0.0272,
|
1325 |
+
"step": 1880
|
1326 |
+
},
|
1327 |
+
{
|
1328 |
+
"epoch": 10.327868852459016,
|
1329 |
+
"grad_norm": 0.5019255876541138,
|
1330 |
+
"learning_rate": 5.142287616756809e-05,
|
1331 |
+
"loss": 0.0308,
|
1332 |
+
"step": 1890
|
1333 |
+
},
|
1334 |
+
{
|
1335 |
+
"epoch": 10.382513661202186,
|
1336 |
+
"grad_norm": 0.28151676058769226,
|
1337 |
+
"learning_rate": 5.09712394871742e-05,
|
1338 |
+
"loss": 0.0383,
|
1339 |
+
"step": 1900
|
1340 |
+
},
|
1341 |
+
{
|
1342 |
+
"epoch": 10.437158469945356,
|
1343 |
+
"grad_norm": 0.7377211451530457,
|
1344 |
+
"learning_rate": 5.051952351763428e-05,
|
1345 |
+
"loss": 0.0322,
|
1346 |
+
"step": 1910
|
1347 |
+
},
|
1348 |
+
{
|
1349 |
+
"epoch": 10.491803278688524,
|
1350 |
+
"grad_norm": 1.962594985961914,
|
1351 |
+
"learning_rate": 5.006776513571655e-05,
|
1352 |
+
"loss": 0.032,
|
1353 |
+
"step": 1920
|
1354 |
+
},
|
1355 |
+
{
|
1356 |
+
"epoch": 10.546448087431694,
|
1357 |
+
"grad_norm": 0.3466651737689972,
|
1358 |
+
"learning_rate": 4.9616001221651693e-05,
|
1359 |
+
"loss": 0.0355,
|
1360 |
+
"step": 1930
|
1361 |
+
},
|
1362 |
+
{
|
1363 |
+
"epoch": 10.601092896174864,
|
1364 |
+
"grad_norm": 1.698729157447815,
|
1365 |
+
"learning_rate": 4.916426865612201e-05,
|
1366 |
+
"loss": 0.0345,
|
1367 |
+
"step": 1940
|
1368 |
+
},
|
1369 |
+
{
|
1370 |
+
"epoch": 10.655737704918034,
|
1371 |
+
"grad_norm": 0.9637352824211121,
|
1372 |
+
"learning_rate": 4.8712604317250576e-05,
|
1373 |
+
"loss": 0.0313,
|
1374 |
+
"step": 1950
|
1375 |
+
},
|
1376 |
+
{
|
1377 |
+
"epoch": 10.710382513661202,
|
1378 |
+
"grad_norm": 0.6569703817367554,
|
1379 |
+
"learning_rate": 4.8261045077590665e-05,
|
1380 |
+
"loss": 0.031,
|
1381 |
+
"step": 1960
|
1382 |
+
},
|
1383 |
+
{
|
1384 |
+
"epoch": 10.765027322404372,
|
1385 |
+
"grad_norm": 0.4290389120578766,
|
1386 |
+
"learning_rate": 4.780962780111554e-05,
|
1387 |
+
"loss": 0.0307,
|
1388 |
+
"step": 1970
|
1389 |
+
},
|
1390 |
+
{
|
1391 |
+
"epoch": 10.819672131147541,
|
1392 |
+
"grad_norm": 0.2776305079460144,
|
1393 |
+
"learning_rate": 4.735838934020903e-05,
|
1394 |
+
"loss": 0.0338,
|
1395 |
+
"step": 1980
|
1396 |
+
},
|
1397 |
+
{
|
1398 |
+
"epoch": 10.87431693989071,
|
1399 |
+
"grad_norm": 0.426277756690979,
|
1400 |
+
"learning_rate": 4.690736653265695e-05,
|
1401 |
+
"loss": 0.0271,
|
1402 |
+
"step": 1990
|
1403 |
+
},
|
1404 |
+
{
|
1405 |
+
"epoch": 10.92896174863388,
|
1406 |
+
"grad_norm": 0.890143096446991,
|
1407 |
+
"learning_rate": 4.6456596198639866e-05,
|
1408 |
+
"loss": 0.0325,
|
1409 |
+
"step": 2000
|
1410 |
+
},
|
1411 |
+
{
|
1412 |
+
"epoch": 10.98360655737705,
|
1413 |
+
"grad_norm": 0.27757057547569275,
|
1414 |
+
"learning_rate": 4.600611513772708e-05,
|
1415 |
+
"loss": 0.0313,
|
1416 |
+
"step": 2010
|
1417 |
+
},
|
1418 |
+
{
|
1419 |
+
"epoch": 11.03825136612022,
|
1420 |
+
"grad_norm": 0.5478549599647522,
|
1421 |
+
"learning_rate": 4.555596012587253e-05,
|
1422 |
+
"loss": 0.0349,
|
1423 |
+
"step": 2020
|
1424 |
+
},
|
1425 |
+
{
|
1426 |
+
"epoch": 11.092896174863387,
|
1427 |
+
"grad_norm": 0.5741052031517029,
|
1428 |
+
"learning_rate": 4.510616791241243e-05,
|
1429 |
+
"loss": 0.0351,
|
1430 |
+
"step": 2030
|
1431 |
+
},
|
1432 |
+
{
|
1433 |
+
"epoch": 11.147540983606557,
|
1434 |
+
"grad_norm": 0.6417993903160095,
|
1435 |
+
"learning_rate": 4.46567752170652e-05,
|
1436 |
+
"loss": 0.0311,
|
1437 |
+
"step": 2040
|
1438 |
+
},
|
1439 |
+
{
|
1440 |
+
"epoch": 11.202185792349727,
|
1441 |
+
"grad_norm": 0.5397584438323975,
|
1442 |
+
"learning_rate": 4.420781872693383e-05,
|
1443 |
+
"loss": 0.0355,
|
1444 |
+
"step": 2050
|
1445 |
+
},
|
1446 |
+
{
|
1447 |
+
"epoch": 11.256830601092895,
|
1448 |
+
"grad_norm": 0.19554707407951355,
|
1449 |
+
"learning_rate": 4.375933509351071e-05,
|
1450 |
+
"loss": 0.0306,
|
1451 |
+
"step": 2060
|
1452 |
+
},
|
1453 |
+
{
|
1454 |
+
"epoch": 11.311475409836065,
|
1455 |
+
"grad_norm": 0.5310783982276917,
|
1456 |
+
"learning_rate": 4.331136092968566e-05,
|
1457 |
+
"loss": 0.0321,
|
1458 |
+
"step": 2070
|
1459 |
+
},
|
1460 |
+
{
|
1461 |
+
"epoch": 11.366120218579235,
|
1462 |
+
"grad_norm": 0.25652116537094116,
|
1463 |
+
"learning_rate": 4.2863932806756845e-05,
|
1464 |
+
"loss": 0.0273,
|
1465 |
+
"step": 2080
|
1466 |
+
},
|
1467 |
+
{
|
1468 |
+
"epoch": 11.420765027322405,
|
1469 |
+
"grad_norm": 0.2970091998577118,
|
1470 |
+
"learning_rate": 4.241708725144529e-05,
|
1471 |
+
"loss": 0.0327,
|
1472 |
+
"step": 2090
|
1473 |
+
},
|
1474 |
+
{
|
1475 |
+
"epoch": 11.475409836065573,
|
1476 |
+
"grad_norm": 0.5687034726142883,
|
1477 |
+
"learning_rate": 4.197086074291285e-05,
|
1478 |
+
"loss": 0.0312,
|
1479 |
+
"step": 2100
|
1480 |
+
},
|
1481 |
+
{
|
1482 |
+
"epoch": 11.530054644808743,
|
1483 |
+
"grad_norm": 1.2950825691223145,
|
1484 |
+
"learning_rate": 4.152528970978432e-05,
|
1485 |
+
"loss": 0.0303,
|
1486 |
+
"step": 2110
|
1487 |
+
},
|
1488 |
+
{
|
1489 |
+
"epoch": 11.584699453551913,
|
1490 |
+
"grad_norm": 0.2822929620742798,
|
1491 |
+
"learning_rate": 4.108041052717329e-05,
|
1492 |
+
"loss": 0.0269,
|
1493 |
+
"step": 2120
|
1494 |
+
},
|
1495 |
+
{
|
1496 |
+
"epoch": 11.639344262295083,
|
1497 |
+
"grad_norm": 0.33775463700294495,
|
1498 |
+
"learning_rate": 4.063625951371278e-05,
|
1499 |
+
"loss": 0.0325,
|
1500 |
+
"step": 2130
|
1501 |
+
},
|
1502 |
+
{
|
1503 |
+
"epoch": 11.693989071038251,
|
1504 |
+
"grad_norm": 0.40372395515441895,
|
1505 |
+
"learning_rate": 4.019287292859016e-05,
|
1506 |
+
"loss": 0.0293,
|
1507 |
+
"step": 2140
|
1508 |
+
},
|
1509 |
+
{
|
1510 |
+
"epoch": 11.748633879781421,
|
1511 |
+
"grad_norm": 0.4413204491138458,
|
1512 |
+
"learning_rate": 3.975028696858713e-05,
|
1513 |
+
"loss": 0.0291,
|
1514 |
+
"step": 2150
|
1515 |
+
},
|
1516 |
+
{
|
1517 |
+
"epoch": 11.80327868852459,
|
1518 |
+
"grad_norm": 0.7298337817192078,
|
1519 |
+
"learning_rate": 3.930853776512473e-05,
|
1520 |
+
"loss": 0.0253,
|
1521 |
+
"step": 2160
|
1522 |
+
},
|
1523 |
+
{
|
1524 |
+
"epoch": 11.857923497267759,
|
1525 |
+
"grad_norm": 0.37207067012786865,
|
1526 |
+
"learning_rate": 3.8867661381313606e-05,
|
1527 |
+
"loss": 0.0299,
|
1528 |
+
"step": 2170
|
1529 |
+
},
|
1530 |
+
{
|
1531 |
+
"epoch": 11.912568306010929,
|
1532 |
+
"grad_norm": 0.658452570438385,
|
1533 |
+
"learning_rate": 3.8427693809009996e-05,
|
1534 |
+
"loss": 0.0294,
|
1535 |
+
"step": 2180
|
1536 |
+
},
|
1537 |
+
{
|
1538 |
+
"epoch": 11.967213114754099,
|
1539 |
+
"grad_norm": 0.632122814655304,
|
1540 |
+
"learning_rate": 3.798867096587739e-05,
|
1541 |
+
"loss": 0.0331,
|
1542 |
+
"step": 2190
|
1543 |
+
},
|
1544 |
+
{
|
1545 |
+
"epoch": 12.021857923497267,
|
1546 |
+
"grad_norm": 0.4645497798919678,
|
1547 |
+
"learning_rate": 3.755062869245441e-05,
|
1548 |
+
"loss": 0.0313,
|
1549 |
+
"step": 2200
|
1550 |
+
},
|
1551 |
+
{
|
1552 |
+
"epoch": 12.076502732240437,
|
1553 |
+
"grad_norm": 0.4805966019630432,
|
1554 |
+
"learning_rate": 3.711360274922876e-05,
|
1555 |
+
"loss": 0.0343,
|
1556 |
+
"step": 2210
|
1557 |
+
},
|
1558 |
+
{
|
1559 |
+
"epoch": 12.131147540983607,
|
1560 |
+
"grad_norm": 0.4689371585845947,
|
1561 |
+
"learning_rate": 3.667762881371804e-05,
|
1562 |
+
"loss": 0.0324,
|
1563 |
+
"step": 2220
|
1564 |
+
},
|
1565 |
+
{
|
1566 |
+
"epoch": 12.185792349726777,
|
1567 |
+
"grad_norm": 0.5546335577964783,
|
1568 |
+
"learning_rate": 3.6242742477556924e-05,
|
1569 |
+
"loss": 0.0316,
|
1570 |
+
"step": 2230
|
1571 |
+
},
|
1572 |
+
{
|
1573 |
+
"epoch": 12.240437158469945,
|
1574 |
+
"grad_norm": 0.5335983633995056,
|
1575 |
+
"learning_rate": 3.58089792435917e-05,
|
1576 |
+
"loss": 0.0322,
|
1577 |
+
"step": 2240
|
1578 |
+
},
|
1579 |
+
{
|
1580 |
+
"epoch": 12.295081967213115,
|
1581 |
+
"grad_norm": 0.9482993483543396,
|
1582 |
+
"learning_rate": 3.5376374522981883e-05,
|
1583 |
+
"loss": 0.031,
|
1584 |
+
"step": 2250
|
1585 |
+
},
|
1586 |
+
{
|
1587 |
+
"epoch": 12.349726775956285,
|
1588 |
+
"grad_norm": 0.41239190101623535,
|
1589 |
+
"learning_rate": 3.494496363230933e-05,
|
1590 |
+
"loss": 0.0323,
|
1591 |
+
"step": 2260
|
1592 |
+
},
|
1593 |
+
{
|
1594 |
+
"epoch": 12.404371584699454,
|
1595 |
+
"grad_norm": 0.3172128200531006,
|
1596 |
+
"learning_rate": 3.451478179069517e-05,
|
1597 |
+
"loss": 0.0247,
|
1598 |
+
"step": 2270
|
1599 |
+
},
|
1600 |
+
{
|
1601 |
+
"epoch": 12.459016393442623,
|
1602 |
+
"grad_norm": 0.5010955333709717,
|
1603 |
+
"learning_rate": 3.4085864116924516e-05,
|
1604 |
+
"loss": 0.0291,
|
1605 |
+
"step": 2280
|
1606 |
+
},
|
1607 |
+
{
|
1608 |
+
"epoch": 12.513661202185792,
|
1609 |
+
"grad_norm": 0.5820704102516174,
|
1610 |
+
"learning_rate": 3.365824562657956e-05,
|
1611 |
+
"loss": 0.0347,
|
1612 |
+
"step": 2290
|
1613 |
+
},
|
1614 |
+
{
|
1615 |
+
"epoch": 12.568306010928962,
|
1616 |
+
"grad_norm": 0.2892806828022003,
|
1617 |
+
"learning_rate": 3.323196122918091e-05,
|
1618 |
+
"loss": 0.0255,
|
1619 |
+
"step": 2300
|
1620 |
+
},
|
1621 |
+
{
|
1622 |
+
"epoch": 12.62295081967213,
|
1623 |
+
"grad_norm": 0.344103068113327,
|
1624 |
+
"learning_rate": 3.2807045725337815e-05,
|
1625 |
+
"loss": 0.0297,
|
1626 |
+
"step": 2310
|
1627 |
+
},
|
1628 |
+
{
|
1629 |
+
"epoch": 12.6775956284153,
|
1630 |
+
"grad_norm": 0.812961757183075,
|
1631 |
+
"learning_rate": 3.2383533803906986e-05,
|
1632 |
+
"loss": 0.0296,
|
1633 |
+
"step": 2320
|
1634 |
+
},
|
1635 |
+
{
|
1636 |
+
"epoch": 12.73224043715847,
|
1637 |
+
"grad_norm": 0.8150091767311096,
|
1638 |
+
"learning_rate": 3.196146003916084e-05,
|
1639 |
+
"loss": 0.029,
|
1640 |
+
"step": 2330
|
1641 |
+
},
|
1642 |
+
{
|
1643 |
+
"epoch": 12.78688524590164,
|
1644 |
+
"grad_norm": 0.38141801953315735,
|
1645 |
+
"learning_rate": 3.154085888796486e-05,
|
1646 |
+
"loss": 0.0276,
|
1647 |
+
"step": 2340
|
1648 |
+
},
|
1649 |
+
{
|
1650 |
+
"epoch": 12.841530054644808,
|
1651 |
+
"grad_norm": 0.5500876307487488,
|
1652 |
+
"learning_rate": 3.11217646869647e-05,
|
1653 |
+
"loss": 0.0272,
|
1654 |
+
"step": 2350
|
1655 |
+
},
|
1656 |
+
{
|
1657 |
+
"epoch": 12.896174863387978,
|
1658 |
+
"grad_norm": 0.534770667552948,
|
1659 |
+
"learning_rate": 3.0704211649782975e-05,
|
1660 |
+
"loss": 0.0262,
|
1661 |
+
"step": 2360
|
1662 |
+
},
|
1663 |
+
{
|
1664 |
+
"epoch": 12.950819672131148,
|
1665 |
+
"grad_norm": 0.8099629282951355,
|
1666 |
+
"learning_rate": 3.0288233864226235e-05,
|
1667 |
+
"loss": 0.0311,
|
1668 |
+
"step": 2370
|
1669 |
+
},
|
1670 |
+
{
|
1671 |
+
"epoch": 13.005464480874316,
|
1672 |
+
"grad_norm": 0.8436483144760132,
|
1673 |
+
"learning_rate": 2.9873865289502112e-05,
|
1674 |
+
"loss": 0.031,
|
1675 |
+
"step": 2380
|
1676 |
+
},
|
1677 |
+
{
|
1678 |
+
"epoch": 13.060109289617486,
|
1679 |
+
"grad_norm": 0.4314166307449341,
|
1680 |
+
"learning_rate": 2.946113975344694e-05,
|
1681 |
+
"loss": 0.0267,
|
1682 |
+
"step": 2390
|
1683 |
+
},
|
1684 |
+
{
|
1685 |
+
"epoch": 13.114754098360656,
|
1686 |
+
"grad_norm": 0.30424031615257263,
|
1687 |
+
"learning_rate": 2.9050090949764218e-05,
|
1688 |
+
"loss": 0.0289,
|
1689 |
+
"step": 2400
|
1690 |
+
},
|
1691 |
+
{
|
1692 |
+
"epoch": 13.169398907103826,
|
1693 |
+
"grad_norm": 0.3659568727016449,
|
1694 |
+
"learning_rate": 2.8640752435273942e-05,
|
1695 |
+
"loss": 0.0306,
|
1696 |
+
"step": 2410
|
1697 |
+
},
|
1698 |
+
{
|
1699 |
+
"epoch": 13.224043715846994,
|
1700 |
+
"grad_norm": 0.36199069023132324,
|
1701 |
+
"learning_rate": 2.8233157627173062e-05,
|
1702 |
+
"loss": 0.0248,
|
1703 |
+
"step": 2420
|
1704 |
+
},
|
1705 |
+
{
|
1706 |
+
"epoch": 13.278688524590164,
|
1707 |
+
"grad_norm": 0.29980045557022095,
|
1708 |
+
"learning_rate": 2.7827339800307466e-05,
|
1709 |
+
"loss": 0.0321,
|
1710 |
+
"step": 2430
|
1711 |
+
},
|
1712 |
+
{
|
1713 |
+
"epoch": 13.333333333333334,
|
1714 |
+
"grad_norm": 1.3584239482879639,
|
1715 |
+
"learning_rate": 2.7423332084455544e-05,
|
1716 |
+
"loss": 0.0263,
|
1717 |
+
"step": 2440
|
1718 |
+
},
|
1719 |
+
{
|
1720 |
+
"epoch": 13.387978142076502,
|
1721 |
+
"grad_norm": 0.372349351644516,
|
1722 |
+
"learning_rate": 2.7021167461623454e-05,
|
1723 |
+
"loss": 0.0259,
|
1724 |
+
"step": 2450
|
1725 |
+
},
|
1726 |
+
{
|
1727 |
+
"epoch": 13.442622950819672,
|
1728 |
+
"grad_norm": 0.32903772592544556,
|
1729 |
+
"learning_rate": 2.6620878763352714e-05,
|
1730 |
+
"loss": 0.0296,
|
1731 |
+
"step": 2460
|
1732 |
+
},
|
1733 |
+
{
|
1734 |
+
"epoch": 13.497267759562842,
|
1735 |
+
"grad_norm": 0.4593813419342041,
|
1736 |
+
"learning_rate": 2.6222498668039798e-05,
|
1737 |
+
"loss": 0.0254,
|
1738 |
+
"step": 2470
|
1739 |
+
},
|
1740 |
+
{
|
1741 |
+
"epoch": 13.551912568306012,
|
1742 |
+
"grad_norm": 0.2875046133995056,
|
1743 |
+
"learning_rate": 2.5826059698268433e-05,
|
1744 |
+
"loss": 0.0242,
|
1745 |
+
"step": 2480
|
1746 |
+
},
|
1747 |
+
{
|
1748 |
+
"epoch": 13.60655737704918,
|
1749 |
+
"grad_norm": 0.7870773673057556,
|
1750 |
+
"learning_rate": 2.5431594218154585e-05,
|
1751 |
+
"loss": 0.0228,
|
1752 |
+
"step": 2490
|
1753 |
+
},
|
1754 |
+
{
|
1755 |
+
"epoch": 13.66120218579235,
|
1756 |
+
"grad_norm": 0.6613791584968567,
|
1757 |
+
"learning_rate": 2.5039134430704292e-05,
|
1758 |
+
"loss": 0.0316,
|
1759 |
+
"step": 2500
|
1760 |
+
},
|
1761 |
+
{
|
1762 |
+
"epoch": 13.71584699453552,
|
1763 |
+
"grad_norm": 0.6127548813819885,
|
1764 |
+
"learning_rate": 2.4648712375184695e-05,
|
1765 |
+
"loss": 0.0249,
|
1766 |
+
"step": 2510
|
1767 |
+
},
|
1768 |
+
{
|
1769 |
+
"epoch": 13.770491803278688,
|
1770 |
+
"grad_norm": 0.5010942220687866,
|
1771 |
+
"learning_rate": 2.426035992450848e-05,
|
1772 |
+
"loss": 0.0248,
|
1773 |
+
"step": 2520
|
1774 |
+
},
|
1775 |
+
{
|
1776 |
+
"epoch": 13.825136612021858,
|
1777 |
+
"grad_norm": 0.3922024071216583,
|
1778 |
+
"learning_rate": 2.3874108782631847e-05,
|
1779 |
+
"loss": 0.0271,
|
1780 |
+
"step": 2530
|
1781 |
+
},
|
1782 |
+
{
|
1783 |
+
"epoch": 13.879781420765028,
|
1784 |
+
"grad_norm": 0.43783921003341675,
|
1785 |
+
"learning_rate": 2.348999048196634e-05,
|
1786 |
+
"loss": 0.0251,
|
1787 |
+
"step": 2540
|
1788 |
+
},
|
1789 |
+
{
|
1790 |
+
"epoch": 13.934426229508198,
|
1791 |
+
"grad_norm": 0.6914840936660767,
|
1792 |
+
"learning_rate": 2.310803638080458e-05,
|
1793 |
+
"loss": 0.0274,
|
1794 |
+
"step": 2550
|
1795 |
+
},
|
1796 |
+
{
|
1797 |
+
"epoch": 13.989071038251366,
|
1798 |
+
"grad_norm": 0.5838907957077026,
|
1799 |
+
"learning_rate": 2.2728277660760273e-05,
|
1800 |
+
"loss": 0.0268,
|
1801 |
+
"step": 2560
|
1802 |
+
},
|
1803 |
+
{
|
1804 |
+
"epoch": 14.043715846994536,
|
1805 |
+
"grad_norm": 0.2598058581352234,
|
1806 |
+
"learning_rate": 2.2350745324222632e-05,
|
1807 |
+
"loss": 0.0252,
|
1808 |
+
"step": 2570
|
1809 |
+
},
|
1810 |
+
{
|
1811 |
+
"epoch": 14.098360655737705,
|
1812 |
+
"grad_norm": 0.26290881633758545,
|
1813 |
+
"learning_rate": 2.197547019182545e-05,
|
1814 |
+
"loss": 0.0273,
|
1815 |
+
"step": 2580
|
1816 |
+
},
|
1817 |
+
{
|
1818 |
+
"epoch": 14.153005464480874,
|
1819 |
+
"grad_norm": 0.18337023258209229,
|
1820 |
+
"learning_rate": 2.1602482899930998e-05,
|
1821 |
+
"loss": 0.0238,
|
1822 |
+
"step": 2590
|
1823 |
+
},
|
1824 |
+
{
|
1825 |
+
"epoch": 14.207650273224044,
|
1826 |
+
"grad_norm": 0.7507574558258057,
|
1827 |
+
"learning_rate": 2.123181389812897e-05,
|
1828 |
+
"loss": 0.0286,
|
1829 |
+
"step": 2600
|
1830 |
+
},
|
1831 |
+
{
|
1832 |
+
"epoch": 14.262295081967213,
|
1833 |
+
"grad_norm": 0.4715905785560608,
|
1834 |
+
"learning_rate": 2.086349344675062e-05,
|
1835 |
+
"loss": 0.023,
|
1836 |
+
"step": 2610
|
1837 |
+
},
|
1838 |
+
{
|
1839 |
+
"epoch": 14.316939890710383,
|
1840 |
+
"grad_norm": 0.3056642413139343,
|
1841 |
+
"learning_rate": 2.0497551614398402e-05,
|
1842 |
+
"loss": 0.0276,
|
1843 |
+
"step": 2620
|
1844 |
+
},
|
1845 |
+
{
|
1846 |
+
"epoch": 14.371584699453551,
|
1847 |
+
"grad_norm": 0.39559224247932434,
|
1848 |
+
"learning_rate": 2.0134018275491357e-05,
|
1849 |
+
"loss": 0.0239,
|
1850 |
+
"step": 2630
|
1851 |
+
},
|
1852 |
+
{
|
1853 |
+
"epoch": 14.426229508196721,
|
1854 |
+
"grad_norm": 0.7475664019584656,
|
1855 |
+
"learning_rate": 1.9772923107826113e-05,
|
1856 |
+
"loss": 0.0285,
|
1857 |
+
"step": 2640
|
1858 |
+
},
|
1859 |
+
{
|
1860 |
+
"epoch": 14.480874316939891,
|
1861 |
+
"grad_norm": 0.3495936095714569,
|
1862 |
+
"learning_rate": 1.941429559015415e-05,
|
1863 |
+
"loss": 0.0257,
|
1864 |
+
"step": 2650
|
1865 |
+
},
|
1866 |
+
{
|
1867 |
+
"epoch": 14.53551912568306,
|
1868 |
+
"grad_norm": 0.40958720445632935,
|
1869 |
+
"learning_rate": 1.905816499977528e-05,
|
1870 |
+
"loss": 0.029,
|
1871 |
+
"step": 2660
|
1872 |
+
},
|
1873 |
+
{
|
1874 |
+
"epoch": 14.59016393442623,
|
1875 |
+
"grad_norm": 0.7965494394302368,
|
1876 |
+
"learning_rate": 1.8704560410147415e-05,
|
1877 |
+
"loss": 0.0224,
|
1878 |
+
"step": 2670
|
1879 |
+
},
|
1880 |
+
{
|
1881 |
+
"epoch": 14.6448087431694,
|
1882 |
+
"grad_norm": 0.30071818828582764,
|
1883 |
+
"learning_rate": 1.8353510688513235e-05,
|
1884 |
+
"loss": 0.0246,
|
1885 |
+
"step": 2680
|
1886 |
+
},
|
1887 |
+
{
|
1888 |
+
"epoch": 14.699453551912569,
|
1889 |
+
"grad_norm": 0.36322134733200073,
|
1890 |
+
"learning_rate": 1.8005044493543444e-05,
|
1891 |
+
"loss": 0.0236,
|
1892 |
+
"step": 2690
|
1893 |
+
},
|
1894 |
+
{
|
1895 |
+
"epoch": 14.754098360655737,
|
1896 |
+
"grad_norm": 0.3105396330356598,
|
1897 |
+
"learning_rate": 1.7659190272997185e-05,
|
1898 |
+
"loss": 0.0277,
|
1899 |
+
"step": 2700
|
1900 |
+
},
|
1901 |
+
{
|
1902 |
+
"epoch": 14.808743169398907,
|
1903 |
+
"grad_norm": 0.9227954149246216,
|
1904 |
+
"learning_rate": 1.7315976261399696e-05,
|
1905 |
+
"loss": 0.0243,
|
1906 |
+
"step": 2710
|
1907 |
+
},
|
1908 |
+
{
|
1909 |
+
"epoch": 14.863387978142077,
|
1910 |
+
"grad_norm": 0.254707008600235,
|
1911 |
+
"learning_rate": 1.6975430477737308e-05,
|
1912 |
+
"loss": 0.0225,
|
1913 |
+
"step": 2720
|
1914 |
+
},
|
1915 |
+
{
|
1916 |
+
"epoch": 14.918032786885245,
|
1917 |
+
"grad_norm": 0.30804020166397095,
|
1918 |
+
"learning_rate": 1.6637580723169988e-05,
|
1919 |
+
"loss": 0.0279,
|
1920 |
+
"step": 2730
|
1921 |
+
},
|
1922 |
+
{
|
1923 |
+
"epoch": 14.972677595628415,
|
1924 |
+
"grad_norm": 0.27004608511924744,
|
1925 |
+
"learning_rate": 1.630245457876181e-05,
|
1926 |
+
"loss": 0.0269,
|
1927 |
+
"step": 2740
|
1928 |
+
},
|
1929 |
+
{
|
1930 |
+
"epoch": 15.027322404371585,
|
1931 |
+
"grad_norm": 0.95454341173172,
|
1932 |
+
"learning_rate": 1.5970079403229286e-05,
|
1933 |
+
"loss": 0.0217,
|
1934 |
+
"step": 2750
|
1935 |
+
},
|
1936 |
+
{
|
1937 |
+
"epoch": 15.081967213114755,
|
1938 |
+
"grad_norm": 0.19432038068771362,
|
1939 |
+
"learning_rate": 1.564048233070792e-05,
|
1940 |
+
"loss": 0.0229,
|
1941 |
+
"step": 2760
|
1942 |
+
},
|
1943 |
+
{
|
1944 |
+
"epoch": 15.136612021857923,
|
1945 |
+
"grad_norm": 0.27142611145973206,
|
1946 |
+
"learning_rate": 1.5313690268537008e-05,
|
1947 |
+
"loss": 0.0243,
|
1948 |
+
"step": 2770
|
1949 |
+
},
|
1950 |
+
{
|
1951 |
+
"epoch": 15.191256830601093,
|
1952 |
+
"grad_norm": 0.2761251628398895,
|
1953 |
+
"learning_rate": 1.498972989506301e-05,
|
1954 |
+
"loss": 0.0282,
|
1955 |
+
"step": 2780
|
1956 |
+
},
|
1957 |
+
{
|
1958 |
+
"epoch": 15.245901639344263,
|
1959 |
+
"grad_norm": 0.25468575954437256,
|
1960 |
+
"learning_rate": 1.4668627657461632e-05,
|
1961 |
+
"loss": 0.0245,
|
1962 |
+
"step": 2790
|
1963 |
+
},
|
1964 |
+
{
|
1965 |
+
"epoch": 15.300546448087431,
|
1966 |
+
"grad_norm": 0.5863296389579773,
|
1967 |
+
"learning_rate": 1.4350409769578705e-05,
|
1968 |
+
"loss": 0.0237,
|
1969 |
+
"step": 2800
|
1970 |
+
},
|
1971 |
+
{
|
1972 |
+
"epoch": 15.3551912568306,
|
1973 |
+
"grad_norm": 0.598770022392273,
|
1974 |
+
"learning_rate": 1.4035102209790252e-05,
|
1975 |
+
"loss": 0.0236,
|
1976 |
+
"step": 2810
|
1977 |
+
},
|
1978 |
+
{
|
1979 |
+
"epoch": 15.40983606557377,
|
1980 |
+
"grad_norm": 1.1667654514312744,
|
1981 |
+
"learning_rate": 1.3722730718881633e-05,
|
1982 |
+
"loss": 0.0227,
|
1983 |
+
"step": 2820
|
1984 |
+
},
|
1985 |
+
{
|
1986 |
+
"epoch": 15.46448087431694,
|
1987 |
+
"grad_norm": 0.22816218435764313,
|
1988 |
+
"learning_rate": 1.3413320797946104e-05,
|
1989 |
+
"loss": 0.0201,
|
1990 |
+
"step": 2830
|
1991 |
+
},
|
1992 |
+
{
|
1993 |
+
"epoch": 15.519125683060109,
|
1994 |
+
"grad_norm": 0.1558607667684555,
|
1995 |
+
"learning_rate": 1.3106897706303051e-05,
|
1996 |
+
"loss": 0.0232,
|
1997 |
+
"step": 2840
|
1998 |
+
},
|
1999 |
+
{
|
2000 |
+
"epoch": 15.573770491803279,
|
2001 |
+
"grad_norm": 0.23202569782733917,
|
2002 |
+
"learning_rate": 1.2803486459435887e-05,
|
2003 |
+
"loss": 0.0209,
|
2004 |
+
"step": 2850
|
2005 |
+
},
|
2006 |
+
{
|
2007 |
+
"epoch": 15.628415300546449,
|
2008 |
+
"grad_norm": 0.23527105152606964,
|
2009 |
+
"learning_rate": 1.2503111826949809e-05,
|
2010 |
+
"loss": 0.0202,
|
2011 |
+
"step": 2860
|
2012 |
+
},
|
2013 |
+
{
|
2014 |
+
"epoch": 15.683060109289617,
|
2015 |
+
"grad_norm": 0.5382643938064575,
|
2016 |
+
"learning_rate": 1.2205798330549712e-05,
|
2017 |
+
"loss": 0.0265,
|
2018 |
+
"step": 2870
|
2019 |
+
},
|
2020 |
+
{
|
2021 |
+
"epoch": 15.737704918032787,
|
2022 |
+
"grad_norm": 0.5895487070083618,
|
2023 |
+
"learning_rate": 1.1911570242038356e-05,
|
2024 |
+
"loss": 0.0247,
|
2025 |
+
"step": 2880
|
2026 |
+
},
|
2027 |
+
{
|
2028 |
+
"epoch": 15.792349726775956,
|
2029 |
+
"grad_norm": 0.17227207124233246,
|
2030 |
+
"learning_rate": 1.1620451581334784e-05,
|
2031 |
+
"loss": 0.0247,
|
2032 |
+
"step": 2890
|
2033 |
+
},
|
2034 |
+
{
|
2035 |
+
"epoch": 15.846994535519126,
|
2036 |
+
"grad_norm": 0.30139684677124023,
|
2037 |
+
"learning_rate": 1.1332466114513512e-05,
|
2038 |
+
"loss": 0.0252,
|
2039 |
+
"step": 2900
|
2040 |
+
},
|
2041 |
+
{
|
2042 |
+
"epoch": 15.901639344262295,
|
2043 |
+
"grad_norm": 0.2650488018989563,
|
2044 |
+
"learning_rate": 1.1047637351864254e-05,
|
2045 |
+
"loss": 0.0231,
|
2046 |
+
"step": 2910
|
2047 |
+
},
|
2048 |
+
{
|
2049 |
+
"epoch": 15.956284153005464,
|
2050 |
+
"grad_norm": 0.8009262084960938,
|
2051 |
+
"learning_rate": 1.0765988545972639e-05,
|
2052 |
+
"loss": 0.0234,
|
2053 |
+
"step": 2920
|
2054 |
+
},
|
2055 |
+
{
|
2056 |
+
"epoch": 16.010928961748633,
|
2057 |
+
"grad_norm": 0.2774312198162079,
|
2058 |
+
"learning_rate": 1.0487542689821978e-05,
|
2059 |
+
"loss": 0.0244,
|
2060 |
+
"step": 2930
|
2061 |
+
},
|
2062 |
+
{
|
2063 |
+
"epoch": 16.065573770491802,
|
2064 |
+
"grad_norm": 0.3976193964481354,
|
2065 |
+
"learning_rate": 1.0212322514916139e-05,
|
2066 |
+
"loss": 0.0226,
|
2067 |
+
"step": 2940
|
2068 |
+
},
|
2069 |
+
{
|
2070 |
+
"epoch": 16.120218579234972,
|
2071 |
+
"grad_norm": 0.604283332824707,
|
2072 |
+
"learning_rate": 9.940350489423811e-06,
|
2073 |
+
"loss": 0.0258,
|
2074 |
+
"step": 2950
|
2075 |
+
},
|
2076 |
+
{
|
2077 |
+
"epoch": 16.174863387978142,
|
2078 |
+
"grad_norm": 0.7713361978530884,
|
2079 |
+
"learning_rate": 9.671648816344282e-06,
|
2080 |
+
"loss": 0.0227,
|
2081 |
+
"step": 2960
|
2082 |
+
},
|
2083 |
+
{
|
2084 |
+
"epoch": 16.229508196721312,
|
2085 |
+
"grad_norm": 0.4896545112133026,
|
2086 |
+
"learning_rate": 9.406239431694841e-06,
|
2087 |
+
"loss": 0.0209,
|
2088 |
+
"step": 2970
|
2089 |
+
},
|
2090 |
+
{
|
2091 |
+
"epoch": 16.284153005464482,
|
2092 |
+
"grad_norm": 0.3895213305950165,
|
2093 |
+
"learning_rate": 9.14414400272003e-06,
|
2094 |
+
"loss": 0.0218,
|
2095 |
+
"step": 2980
|
2096 |
+
},
|
2097 |
+
{
|
2098 |
+
"epoch": 16.338797814207652,
|
2099 |
+
"grad_norm": 0.37582123279571533,
|
2100 |
+
"learning_rate": 8.885383926122748e-06,
|
2101 |
+
"loss": 0.0214,
|
2102 |
+
"step": 2990
|
2103 |
+
},
|
2104 |
+
{
|
2105 |
+
"epoch": 16.39344262295082,
|
2106 |
+
"grad_norm": 0.7232876420021057,
|
2107 |
+
"learning_rate": 8.629980326317489e-06,
|
2108 |
+
"loss": 0.0214,
|
2109 |
+
"step": 3000
|
2110 |
+
},
|
2111 |
+
{
|
2112 |
+
"epoch": 16.44808743169399,
|
2113 |
+
"grad_norm": 0.3176191747188568,
|
2114 |
+
"learning_rate": 8.377954053705805e-06,
|
2115 |
+
"loss": 0.0256,
|
2116 |
+
"step": 3010
|
2117 |
+
},
|
2118 |
+
{
|
2119 |
+
"epoch": 16.502732240437158,
|
2120 |
+
"grad_norm": 0.3513713479042053,
|
2121 |
+
"learning_rate": 8.129325682974165e-06,
|
2122 |
+
"loss": 0.0268,
|
2123 |
+
"step": 3020
|
2124 |
+
},
|
2125 |
+
{
|
2126 |
+
"epoch": 16.557377049180328,
|
2127 |
+
"grad_norm": 0.32964545488357544,
|
2128 |
+
"learning_rate": 7.884115511414286e-06,
|
2129 |
+
"loss": 0.0219,
|
2130 |
+
"step": 3030
|
2131 |
+
},
|
2132 |
+
{
|
2133 |
+
"epoch": 16.612021857923498,
|
2134 |
+
"grad_norm": 0.2857699394226074,
|
2135 |
+
"learning_rate": 7.642343557266134e-06,
|
2136 |
+
"loss": 0.024,
|
2137 |
+
"step": 3040
|
2138 |
+
},
|
2139 |
+
{
|
2140 |
+
"epoch": 16.666666666666668,
|
2141 |
+
"grad_norm": 0.5647152662277222,
|
2142 |
+
"learning_rate": 7.404029558083653e-06,
|
2143 |
+
"loss": 0.0295,
|
2144 |
+
"step": 3050
|
2145 |
+
},
|
2146 |
+
{
|
2147 |
+
"epoch": 16.721311475409838,
|
2148 |
+
"grad_norm": 0.3039594888687134,
|
2149 |
+
"learning_rate": 7.169192969123467e-06,
|
2150 |
+
"loss": 0.0226,
|
2151 |
+
"step": 3060
|
2152 |
+
},
|
2153 |
+
{
|
2154 |
+
"epoch": 16.775956284153004,
|
2155 |
+
"grad_norm": 0.4706748127937317,
|
2156 |
+
"learning_rate": 6.937852961756669e-06,
|
2157 |
+
"loss": 0.0248,
|
2158 |
+
"step": 3070
|
2159 |
+
},
|
2160 |
+
{
|
2161 |
+
"epoch": 16.830601092896174,
|
2162 |
+
"grad_norm": 0.35213521122932434,
|
2163 |
+
"learning_rate": 6.710028421903625e-06,
|
2164 |
+
"loss": 0.0237,
|
2165 |
+
"step": 3080
|
2166 |
+
},
|
2167 |
+
{
|
2168 |
+
"epoch": 16.885245901639344,
|
2169 |
+
"grad_norm": 0.3645097017288208,
|
2170 |
+
"learning_rate": 6.4857379484922375e-06,
|
2171 |
+
"loss": 0.0209,
|
2172 |
+
"step": 3090
|
2173 |
+
},
|
2174 |
+
{
|
2175 |
+
"epoch": 16.939890710382514,
|
2176 |
+
"grad_norm": 0.5215752720832825,
|
2177 |
+
"learning_rate": 6.2649998519396235e-06,
|
2178 |
+
"loss": 0.0207,
|
2179 |
+
"step": 3100
|
2180 |
+
},
|
2181 |
+
{
|
2182 |
+
"epoch": 16.994535519125684,
|
2183 |
+
"grad_norm": 0.7735849618911743,
|
2184 |
+
"learning_rate": 6.047832152657207e-06,
|
2185 |
+
"loss": 0.0221,
|
2186 |
+
"step": 3110
|
2187 |
+
},
|
2188 |
+
{
|
2189 |
+
"epoch": 17.049180327868854,
|
2190 |
+
"grad_norm": 0.3790043294429779,
|
2191 |
+
"learning_rate": 5.834252579579685e-06,
|
2192 |
+
"loss": 0.0242,
|
2193 |
+
"step": 3120
|
2194 |
+
},
|
2195 |
+
{
|
2196 |
+
"epoch": 17.103825136612024,
|
2197 |
+
"grad_norm": 0.29151690006256104,
|
2198 |
+
"learning_rate": 5.624278568717611e-06,
|
2199 |
+
"loss": 0.0207,
|
2200 |
+
"step": 3130
|
2201 |
+
},
|
2202 |
+
{
|
2203 |
+
"epoch": 17.15846994535519,
|
2204 |
+
"grad_norm": 0.4057449996471405,
|
2205 |
+
"learning_rate": 5.417927261734007e-06,
|
2206 |
+
"loss": 0.0221,
|
2207 |
+
"step": 3140
|
2208 |
+
},
|
2209 |
+
{
|
2210 |
+
"epoch": 17.21311475409836,
|
2211 |
+
"grad_norm": 0.3481403887271881,
|
2212 |
+
"learning_rate": 5.215215504544995e-06,
|
2213 |
+
"loss": 0.0202,
|
2214 |
+
"step": 3150
|
2215 |
+
},
|
2216 |
+
{
|
2217 |
+
"epoch": 17.26775956284153,
|
2218 |
+
"grad_norm": 0.23598790168762207,
|
2219 |
+
"learning_rate": 5.016159845944473e-06,
|
2220 |
+
"loss": 0.0298,
|
2221 |
+
"step": 3160
|
2222 |
+
},
|
2223 |
+
{
|
2224 |
+
"epoch": 17.3224043715847,
|
2225 |
+
"grad_norm": 0.3719569146633148,
|
2226 |
+
"learning_rate": 4.820776536253202e-06,
|
2227 |
+
"loss": 0.0195,
|
2228 |
+
"step": 3170
|
2229 |
+
},
|
2230 |
+
{
|
2231 |
+
"epoch": 17.37704918032787,
|
2232 |
+
"grad_norm": 0.3433237373828888,
|
2233 |
+
"learning_rate": 4.629081525992118e-06,
|
2234 |
+
"loss": 0.0227,
|
2235 |
+
"step": 3180
|
2236 |
+
},
|
2237 |
+
{
|
2238 |
+
"epoch": 17.43169398907104,
|
2239 |
+
"grad_norm": 0.3161409795284271,
|
2240 |
+
"learning_rate": 4.441090464580178e-06,
|
2241 |
+
"loss": 0.023,
|
2242 |
+
"step": 3190
|
2243 |
+
},
|
2244 |
+
{
|
2245 |
+
"epoch": 17.48633879781421,
|
2246 |
+
"grad_norm": 0.8182094097137451,
|
2247 |
+
"learning_rate": 4.256818699056847e-06,
|
2248 |
+
"loss": 0.0235,
|
2249 |
+
"step": 3200
|
2250 |
+
},
|
2251 |
+
{
|
2252 |
+
"epoch": 17.540983606557376,
|
2253 |
+
"grad_norm": 0.4030332863330841,
|
2254 |
+
"learning_rate": 4.076281272829158e-06,
|
2255 |
+
"loss": 0.023,
|
2256 |
+
"step": 3210
|
2257 |
+
},
|
2258 |
+
{
|
2259 |
+
"epoch": 17.595628415300546,
|
2260 |
+
"grad_norm": 0.252172589302063,
|
2261 |
+
"learning_rate": 3.89949292444361e-06,
|
2262 |
+
"loss": 0.0221,
|
2263 |
+
"step": 3220
|
2264 |
+
},
|
2265 |
+
{
|
2266 |
+
"epoch": 17.650273224043715,
|
2267 |
+
"grad_norm": 1.1839289665222168,
|
2268 |
+
"learning_rate": 3.726468086382989e-06,
|
2269 |
+
"loss": 0.0222,
|
2270 |
+
"step": 3230
|
2271 |
+
},
|
2272 |
+
{
|
2273 |
+
"epoch": 17.704918032786885,
|
2274 |
+
"grad_norm": 0.3704149127006531,
|
2275 |
+
"learning_rate": 3.5572208838881073e-06,
|
2276 |
+
"loss": 0.02,
|
2277 |
+
"step": 3240
|
2278 |
+
},
|
2279 |
+
{
|
2280 |
+
"epoch": 17.759562841530055,
|
2281 |
+
"grad_norm": 0.25189173221588135,
|
2282 |
+
"learning_rate": 3.391765133804692e-06,
|
2283 |
+
"loss": 0.0294,
|
2284 |
+
"step": 3250
|
2285 |
+
},
|
2286 |
+
{
|
2287 |
+
"epoch": 17.814207650273225,
|
2288 |
+
"grad_norm": 0.31096377968788147,
|
2289 |
+
"learning_rate": 3.2301143434554147e-06,
|
2290 |
+
"loss": 0.0216,
|
2291 |
+
"step": 3260
|
2292 |
+
},
|
2293 |
+
{
|
2294 |
+
"epoch": 17.868852459016395,
|
2295 |
+
"grad_norm": 0.37782198190689087,
|
2296 |
+
"learning_rate": 3.072281709537156e-06,
|
2297 |
+
"loss": 0.0223,
|
2298 |
+
"step": 3270
|
2299 |
+
},
|
2300 |
+
{
|
2301 |
+
"epoch": 17.92349726775956,
|
2302 |
+
"grad_norm": 0.3171507716178894,
|
2303 |
+
"learning_rate": 2.918280117043709e-06,
|
2304 |
+
"loss": 0.0253,
|
2305 |
+
"step": 3280
|
2306 |
+
},
|
2307 |
+
{
|
2308 |
+
"epoch": 17.97814207650273,
|
2309 |
+
"grad_norm": 0.46874311566352844,
|
2310 |
+
"learning_rate": 2.768122138213858e-06,
|
2311 |
+
"loss": 0.0215,
|
2312 |
+
"step": 3290
|
2313 |
+
},
|
2314 |
+
{
|
2315 |
+
"epoch": 18.0327868852459,
|
2316 |
+
"grad_norm": 0.7982349395751953,
|
2317 |
+
"learning_rate": 2.621820031505051e-06,
|
2318 |
+
"loss": 0.0211,
|
2319 |
+
"step": 3300
|
2320 |
+
},
|
2321 |
+
{
|
2322 |
+
"epoch": 18.08743169398907,
|
2323 |
+
"grad_norm": 0.6364455819129944,
|
2324 |
+
"learning_rate": 2.4793857405926044e-06,
|
2325 |
+
"loss": 0.0191,
|
2326 |
+
"step": 3310
|
2327 |
+
},
|
2328 |
+
{
|
2329 |
+
"epoch": 18.14207650273224,
|
2330 |
+
"grad_norm": 0.3849509060382843,
|
2331 |
+
"learning_rate": 2.3408308933947244e-06,
|
2332 |
+
"loss": 0.0212,
|
2333 |
+
"step": 3320
|
2334 |
+
},
|
2335 |
+
{
|
2336 |
+
"epoch": 18.19672131147541,
|
2337 |
+
"grad_norm": 0.6707078814506531,
|
2338 |
+
"learning_rate": 2.2061668011231496e-06,
|
2339 |
+
"loss": 0.0244,
|
2340 |
+
"step": 3330
|
2341 |
+
},
|
2342 |
+
{
|
2343 |
+
"epoch": 18.25136612021858,
|
2344 |
+
"grad_norm": 0.377101331949234,
|
2345 |
+
"learning_rate": 2.0754044573598306e-06,
|
2346 |
+
"loss": 0.0197,
|
2347 |
+
"step": 3340
|
2348 |
+
},
|
2349 |
+
{
|
2350 |
+
"epoch": 18.306010928961747,
|
2351 |
+
"grad_norm": 0.531783938407898,
|
2352 |
+
"learning_rate": 1.9485545371593807e-06,
|
2353 |
+
"loss": 0.0227,
|
2354 |
+
"step": 3350
|
2355 |
+
},
|
2356 |
+
{
|
2357 |
+
"epoch": 18.360655737704917,
|
2358 |
+
"grad_norm": 0.5045121312141418,
|
2359 |
+
"learning_rate": 1.8256273961776093e-06,
|
2360 |
+
"loss": 0.0202,
|
2361 |
+
"step": 3360
|
2362 |
+
},
|
2363 |
+
{
|
2364 |
+
"epoch": 18.415300546448087,
|
2365 |
+
"grad_norm": 0.6575759053230286,
|
2366 |
+
"learning_rate": 1.7066330698261467e-06,
|
2367 |
+
"loss": 0.0236,
|
2368 |
+
"step": 3370
|
2369 |
+
},
|
2370 |
+
{
|
2371 |
+
"epoch": 18.469945355191257,
|
2372 |
+
"grad_norm": 0.454512357711792,
|
2373 |
+
"learning_rate": 1.591581272453141e-06,
|
2374 |
+
"loss": 0.0176,
|
2375 |
+
"step": 3380
|
2376 |
+
},
|
2377 |
+
{
|
2378 |
+
"epoch": 18.524590163934427,
|
2379 |
+
"grad_norm": 0.2828448712825775,
|
2380 |
+
"learning_rate": 1.4804813965502374e-06,
|
2381 |
+
"loss": 0.0219,
|
2382 |
+
"step": 3390
|
2383 |
+
},
|
2384 |
+
{
|
2385 |
+
"epoch": 18.579234972677597,
|
2386 |
+
"grad_norm": 0.38783037662506104,
|
2387 |
+
"learning_rate": 1.3733425119857812e-06,
|
2388 |
+
"loss": 0.0281,
|
2389 |
+
"step": 3400
|
2390 |
+
},
|
2391 |
+
{
|
2392 |
+
"epoch": 18.633879781420767,
|
2393 |
+
"grad_norm": 0.42596179246902466,
|
2394 |
+
"learning_rate": 1.2701733652643933e-06,
|
2395 |
+
"loss": 0.0212,
|
2396 |
+
"step": 3410
|
2397 |
+
},
|
2398 |
+
{
|
2399 |
+
"epoch": 18.688524590163933,
|
2400 |
+
"grad_norm": 0.2547903060913086,
|
2401 |
+
"learning_rate": 1.1709823788129292e-06,
|
2402 |
+
"loss": 0.0233,
|
2403 |
+
"step": 3420
|
2404 |
+
},
|
2405 |
+
{
|
2406 |
+
"epoch": 18.743169398907103,
|
2407 |
+
"grad_norm": 0.463740736246109,
|
2408 |
+
"learning_rate": 1.0757776502929084e-06,
|
2409 |
+
"loss": 0.0217,
|
2410 |
+
"step": 3430
|
2411 |
+
},
|
2412 |
+
{
|
2413 |
+
"epoch": 18.797814207650273,
|
2414 |
+
"grad_norm": 0.9804412722587585,
|
2415 |
+
"learning_rate": 9.84566951939414e-07,
|
2416 |
+
"loss": 0.0242,
|
2417 |
+
"step": 3440
|
2418 |
+
},
|
2419 |
+
{
|
2420 |
+
"epoch": 18.852459016393443,
|
2421 |
+
"grad_norm": 0.23706400394439697,
|
2422 |
+
"learning_rate": 8.973577299266178e-07,
|
2423 |
+
"loss": 0.0228,
|
2424 |
+
"step": 3450
|
2425 |
+
},
|
2426 |
+
{
|
2427 |
+
"epoch": 18.907103825136613,
|
2428 |
+
"grad_norm": 0.2117275893688202,
|
2429 |
+
"learning_rate": 8.141571037598828e-07,
|
2430 |
+
"loss": 0.0212,
|
2431 |
+
"step": 3460
|
2432 |
+
},
|
2433 |
+
{
|
2434 |
+
"epoch": 18.961748633879782,
|
2435 |
+
"grad_norm": 0.7138676047325134,
|
2436 |
+
"learning_rate": 7.349718656945504e-07,
|
2437 |
+
"loss": 0.0217,
|
2438 |
+
"step": 3470
|
2439 |
+
},
|
2440 |
+
{
|
2441 |
+
"epoch": 19.016393442622952,
|
2442 |
+
"grad_norm": 0.38242462277412415,
|
2443 |
+
"learning_rate": 6.598084801814563e-07,
|
2444 |
+
"loss": 0.0234,
|
2445 |
+
"step": 3480
|
2446 |
+
},
|
2447 |
+
{
|
2448 |
+
"epoch": 19.07103825136612,
|
2449 |
+
"grad_norm": 0.30162400007247925,
|
2450 |
+
"learning_rate": 5.88673083339164e-07,
|
2451 |
+
"loss": 0.0226,
|
2452 |
+
"step": 3490
|
2453 |
+
},
|
2454 |
+
{
|
2455 |
+
"epoch": 19.12568306010929,
|
2456 |
+
"grad_norm": 0.3553517460823059,
|
2457 |
+
"learning_rate": 5.215714824530427e-07,
|
2458 |
+
"loss": 0.0234,
|
2459 |
+
"step": 3500
|
2460 |
+
},
|
2461 |
+
{
|
2462 |
+
"epoch": 19.18032786885246,
|
2463 |
+
"grad_norm": 0.2980111539363861,
|
2464 |
+
"learning_rate": 4.585091555011856e-07,
|
2465 |
+
"loss": 0.0223,
|
2466 |
+
"step": 3510
|
2467 |
+
},
|
2468 |
+
{
|
2469 |
+
"epoch": 19.23497267759563,
|
2470 |
+
"grad_norm": 0.3606838881969452,
|
2471 |
+
"learning_rate": 3.994912507072013e-07,
|
2472 |
+
"loss": 0.0193,
|
2473 |
+
"step": 3520
|
2474 |
+
},
|
2475 |
+
{
|
2476 |
+
"epoch": 19.2896174863388,
|
2477 |
+
"grad_norm": 0.2700374722480774,
|
2478 |
+
"learning_rate": 3.4452258611991083e-07,
|
2479 |
+
"loss": 0.0209,
|
2480 |
+
"step": 3530
|
2481 |
+
},
|
2482 |
+
{
|
2483 |
+
"epoch": 19.34426229508197,
|
2484 |
+
"grad_norm": 0.44229865074157715,
|
2485 |
+
"learning_rate": 2.9360764922004014e-07,
|
2486 |
+
"loss": 0.0194,
|
2487 |
+
"step": 3540
|
2488 |
+
},
|
2489 |
+
{
|
2490 |
+
"epoch": 19.398907103825138,
|
2491 |
+
"grad_norm": 0.3039417266845703,
|
2492 |
+
"learning_rate": 2.467505965538519e-07,
|
2493 |
+
"loss": 0.0202,
|
2494 |
+
"step": 3550
|
2495 |
+
},
|
2496 |
+
{
|
2497 |
+
"epoch": 19.453551912568305,
|
2498 |
+
"grad_norm": 0.4226384460926056,
|
2499 |
+
"learning_rate": 2.0395525339383937e-07,
|
2500 |
+
"loss": 0.0221,
|
2501 |
+
"step": 3560
|
2502 |
+
},
|
2503 |
+
{
|
2504 |
+
"epoch": 19.508196721311474,
|
2505 |
+
"grad_norm": 0.7776270508766174,
|
2506 |
+
"learning_rate": 1.6522511342643155e-07,
|
2507 |
+
"loss": 0.0203,
|
2508 |
+
"step": 3570
|
2509 |
+
},
|
2510 |
+
{
|
2511 |
+
"epoch": 19.562841530054644,
|
2512 |
+
"grad_norm": 0.23865941166877747,
|
2513 |
+
"learning_rate": 1.3056333846677705e-07,
|
2514 |
+
"loss": 0.0195,
|
2515 |
+
"step": 3580
|
2516 |
+
},
|
2517 |
+
{
|
2518 |
+
"epoch": 19.617486338797814,
|
2519 |
+
"grad_norm": 0.5467799305915833,
|
2520 |
+
"learning_rate": 9.997275820062268e-08,
|
2521 |
+
"loss": 0.0193,
|
2522 |
+
"step": 3590
|
2523 |
+
},
|
2524 |
+
{
|
2525 |
+
"epoch": 19.672131147540984,
|
2526 |
+
"grad_norm": 0.24738432466983795,
|
2527 |
+
"learning_rate": 7.3455869953315e-08,
|
2528 |
+
"loss": 0.0209,
|
2529 |
+
"step": 3600
|
2530 |
+
},
|
2531 |
+
{
|
2532 |
+
"epoch": 19.726775956284154,
|
2533 |
+
"grad_norm": 1.1878104209899902,
|
2534 |
+
"learning_rate": 5.101483848591881e-08,
|
2535 |
+
"loss": 0.0205,
|
2536 |
+
"step": 3610
|
2537 |
+
},
|
2538 |
+
{
|
2539 |
+
"epoch": 19.781420765027324,
|
2540 |
+
"grad_norm": 0.37522733211517334,
|
2541 |
+
"learning_rate": 3.2651495818486476e-08,
|
2542 |
+
"loss": 0.0213,
|
2543 |
+
"step": 3620
|
2544 |
+
},
|
2545 |
+
{
|
2546 |
+
"epoch": 19.83606557377049,
|
2547 |
+
"grad_norm": 0.3089437782764435,
|
2548 |
+
"learning_rate": 1.8367341080510747e-08,
|
2549 |
+
"loss": 0.0181,
|
2550 |
+
"step": 3630
|
2551 |
+
},
|
2552 |
+
{
|
2553 |
+
"epoch": 19.89071038251366,
|
2554 |
+
"grad_norm": 0.45112279057502747,
|
2555 |
+
"learning_rate": 8.163540388539393e-09,
|
2556 |
+
"loss": 0.0207,
|
2557 |
+
"step": 3640
|
2558 |
+
},
|
2559 |
+
{
|
2560 |
+
"epoch": 19.94535519125683,
|
2561 |
+
"grad_norm": 0.6257203817367554,
|
2562 |
+
"learning_rate": 2.0409267509569062e-09,
|
2563 |
+
"loss": 0.0234,
|
2564 |
+
"step": 3650
|
2565 |
+
},
|
2566 |
+
{
|
2567 |
+
"epoch": 20.0,
|
2568 |
+
"grad_norm": 0.20247182250022888,
|
2569 |
+
"learning_rate": 0.0,
|
2570 |
+
"loss": 0.0241,
|
2571 |
+
"step": 3660
|
2572 |
+
},
|
2573 |
+
{
|
2574 |
+
"epoch": 20.0,
|
2575 |
+
"step": 3660,
|
2576 |
+
"total_flos": 3.885044834331936e+17,
|
2577 |
+
"train_loss": 0.05339513023287221,
|
2578 |
+
"train_runtime": 3949.2706,
|
2579 |
+
"train_samples_per_second": 45.401,
|
2580 |
+
"train_steps_per_second": 0.927
|
2581 |
+
}
|
2582 |
+
],
|
2583 |
+
"logging_steps": 10,
|
2584 |
+
"max_steps": 3660,
|
2585 |
+
"num_input_tokens_seen": 0,
|
2586 |
+
"num_train_epochs": 20,
|
2587 |
+
"save_steps": 10000,
|
2588 |
+
"stateful_callbacks": {
|
2589 |
+
"TrainerControl": {
|
2590 |
+
"args": {
|
2591 |
+
"should_epoch_stop": false,
|
2592 |
+
"should_evaluate": false,
|
2593 |
+
"should_log": false,
|
2594 |
+
"should_save": true,
|
2595 |
+
"should_training_stop": true
|
2596 |
+
},
|
2597 |
+
"attributes": {}
|
2598 |
+
}
|
2599 |
+
},
|
2600 |
+
"total_flos": 3.885044834331936e+17,
|
2601 |
+
"train_batch_size": 49,
|
2602 |
+
"trial_name": null,
|
2603 |
+
"trial_params": null
|
2604 |
+
}
|