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": 3.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 366,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.00819672131147541,
|
| 13 |
+
"grad_norm": 4.202638149261475,
|
| 14 |
+
"learning_rate": 1.8181818181818183e-06,
|
| 15 |
+
"loss": 1.0983,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.01639344262295082,
|
| 20 |
+
"grad_norm": 4.90077018737793,
|
| 21 |
+
"learning_rate": 3.6363636363636366e-06,
|
| 22 |
+
"loss": 1.1493,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.02459016393442623,
|
| 27 |
+
"grad_norm": 3.7728867530822754,
|
| 28 |
+
"learning_rate": 5.4545454545454545e-06,
|
| 29 |
+
"loss": 1.1002,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.03278688524590164,
|
| 34 |
+
"grad_norm": 3.952263593673706,
|
| 35 |
+
"learning_rate": 7.272727272727273e-06,
|
| 36 |
+
"loss": 1.0555,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.040983606557377046,
|
| 41 |
+
"grad_norm": 3.233116388320923,
|
| 42 |
+
"learning_rate": 9.090909090909091e-06,
|
| 43 |
+
"loss": 1.0049,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.04918032786885246,
|
| 48 |
+
"grad_norm": 2.412139892578125,
|
| 49 |
+
"learning_rate": 1.0909090909090909e-05,
|
| 50 |
+
"loss": 0.9923,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.05737704918032787,
|
| 55 |
+
"grad_norm": 1.4961522817611694,
|
| 56 |
+
"learning_rate": 1.2727272727272728e-05,
|
| 57 |
+
"loss": 0.9775,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.06557377049180328,
|
| 62 |
+
"grad_norm": 1.8131775856018066,
|
| 63 |
+
"learning_rate": 1.4545454545454546e-05,
|
| 64 |
+
"loss": 1.0089,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.07377049180327869,
|
| 69 |
+
"grad_norm": 1.7625133991241455,
|
| 70 |
+
"learning_rate": 1.6363636363636366e-05,
|
| 71 |
+
"loss": 0.9214,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.08196721311475409,
|
| 76 |
+
"grad_norm": 1.5888162851333618,
|
| 77 |
+
"learning_rate": 1.8181818181818182e-05,
|
| 78 |
+
"loss": 0.9976,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.09016393442622951,
|
| 83 |
+
"grad_norm": 1.6568323373794556,
|
| 84 |
+
"learning_rate": 2e-05,
|
| 85 |
+
"loss": 0.9872,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.09836065573770492,
|
| 90 |
+
"grad_norm": 1.7209982872009277,
|
| 91 |
+
"learning_rate": 1.9999608429280302e-05,
|
| 92 |
+
"loss": 0.9809,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.10655737704918032,
|
| 97 |
+
"grad_norm": 1.429029107093811,
|
| 98 |
+
"learning_rate": 1.999843374778673e-05,
|
| 99 |
+
"loss": 0.9346,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.11475409836065574,
|
| 104 |
+
"grad_norm": 1.2364733219146729,
|
| 105 |
+
"learning_rate": 1.9996476047513454e-05,
|
| 106 |
+
"loss": 0.9249,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.12295081967213115,
|
| 111 |
+
"grad_norm": 1.4021698236465454,
|
| 112 |
+
"learning_rate": 1.9993735481776105e-05,
|
| 113 |
+
"loss": 0.922,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.13114754098360656,
|
| 118 |
+
"grad_norm": 1.442407250404358,
|
| 119 |
+
"learning_rate": 1.9990212265199738e-05,
|
| 120 |
+
"loss": 0.8845,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.13934426229508196,
|
| 125 |
+
"grad_norm": 1.3294403553009033,
|
| 126 |
+
"learning_rate": 1.998590667370204e-05,
|
| 127 |
+
"loss": 0.926,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.14754098360655737,
|
| 132 |
+
"grad_norm": 1.3673713207244873,
|
| 133 |
+
"learning_rate": 1.9980819044471732e-05,
|
| 134 |
+
"loss": 0.9309,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.1557377049180328,
|
| 139 |
+
"grad_norm": 1.3765286207199097,
|
| 140 |
+
"learning_rate": 1.9974949775942134e-05,
|
| 141 |
+
"loss": 0.9608,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.16393442622950818,
|
| 146 |
+
"grad_norm": 1.20298433303833,
|
| 147 |
+
"learning_rate": 1.996829932775999e-05,
|
| 148 |
+
"loss": 0.875,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.1721311475409836,
|
| 153 |
+
"grad_norm": 1.2246984243392944,
|
| 154 |
+
"learning_rate": 1.996086822074945e-05,
|
| 155 |
+
"loss": 0.933,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.18032786885245902,
|
| 160 |
+
"grad_norm": 1.2818138599395752,
|
| 161 |
+
"learning_rate": 1.995265703687131e-05,
|
| 162 |
+
"loss": 0.929,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.1885245901639344,
|
| 167 |
+
"grad_norm": 1.302242636680603,
|
| 168 |
+
"learning_rate": 1.9943666419177392e-05,
|
| 169 |
+
"loss": 0.9397,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.19672131147540983,
|
| 174 |
+
"grad_norm": 1.3329646587371826,
|
| 175 |
+
"learning_rate": 1.9933897071760236e-05,
|
| 176 |
+
"loss": 0.8705,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.20491803278688525,
|
| 181 |
+
"grad_norm": 1.2400087118148804,
|
| 182 |
+
"learning_rate": 1.9923349759697915e-05,
|
| 183 |
+
"loss": 0.9778,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.21311475409836064,
|
| 188 |
+
"grad_norm": 1.3099018335342407,
|
| 189 |
+
"learning_rate": 1.9912025308994146e-05,
|
| 190 |
+
"loss": 0.9431,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.22131147540983606,
|
| 195 |
+
"grad_norm": 1.402883768081665,
|
| 196 |
+
"learning_rate": 1.9899924606513593e-05,
|
| 197 |
+
"loss": 0.9835,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.22950819672131148,
|
| 202 |
+
"grad_norm": 1.2836993932724,
|
| 203 |
+
"learning_rate": 1.9887048599912412e-05,
|
| 204 |
+
"loss": 0.9635,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.23770491803278687,
|
| 209 |
+
"grad_norm": 1.408251404762268,
|
| 210 |
+
"learning_rate": 1.9873398297564036e-05,
|
| 211 |
+
"loss": 1.0184,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.2459016393442623,
|
| 216 |
+
"grad_norm": 1.2877365350723267,
|
| 217 |
+
"learning_rate": 1.9858974768480206e-05,
|
| 218 |
+
"loss": 0.9376,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.2540983606557377,
|
| 223 |
+
"grad_norm": 1.1387895345687866,
|
| 224 |
+
"learning_rate": 1.9843779142227258e-05,
|
| 225 |
+
"loss": 0.8671,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.26229508196721313,
|
| 230 |
+
"grad_norm": 1.378002643585205,
|
| 231 |
+
"learning_rate": 1.982781260883765e-05,
|
| 232 |
+
"loss": 1.0214,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.27049180327868855,
|
| 237 |
+
"grad_norm": 1.2005506753921509,
|
| 238 |
+
"learning_rate": 1.981107641871678e-05,
|
| 239 |
+
"loss": 0.8976,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.2786885245901639,
|
| 244 |
+
"grad_norm": 1.345911979675293,
|
| 245 |
+
"learning_rate": 1.9793571882545048e-05,
|
| 246 |
+
"loss": 0.9229,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.28688524590163933,
|
| 251 |
+
"grad_norm": 1.2991893291473389,
|
| 252 |
+
"learning_rate": 1.9775300371175225e-05,
|
| 253 |
+
"loss": 0.9511,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.29508196721311475,
|
| 258 |
+
"grad_norm": 1.2064939737319946,
|
| 259 |
+
"learning_rate": 1.975626331552507e-05,
|
| 260 |
+
"loss": 0.906,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.30327868852459017,
|
| 265 |
+
"grad_norm": 1.2942262887954712,
|
| 266 |
+
"learning_rate": 1.973646220646531e-05,
|
| 267 |
+
"loss": 0.8906,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.3114754098360656,
|
| 272 |
+
"grad_norm": 1.2344310283660889,
|
| 273 |
+
"learning_rate": 1.9715898594702845e-05,
|
| 274 |
+
"loss": 0.959,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.319672131147541,
|
| 279 |
+
"grad_norm": 1.2887349128723145,
|
| 280 |
+
"learning_rate": 1.969457409065933e-05,
|
| 281 |
+
"loss": 0.8965,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.32786885245901637,
|
| 286 |
+
"grad_norm": 1.29547119140625,
|
| 287 |
+
"learning_rate": 1.967249036434504e-05,
|
| 288 |
+
"loss": 0.9255,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.3360655737704918,
|
| 293 |
+
"grad_norm": 1.3437081575393677,
|
| 294 |
+
"learning_rate": 1.96496491452281e-05,
|
| 295 |
+
"loss": 0.9526,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.3442622950819672,
|
| 300 |
+
"grad_norm": 1.3519320487976074,
|
| 301 |
+
"learning_rate": 1.962605222209903e-05,
|
| 302 |
+
"loss": 0.9633,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.3524590163934426,
|
| 307 |
+
"grad_norm": 1.2603236436843872,
|
| 308 |
+
"learning_rate": 1.9601701442930667e-05,
|
| 309 |
+
"loss": 0.9279,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.36065573770491804,
|
| 314 |
+
"grad_norm": 1.1612663269042969,
|
| 315 |
+
"learning_rate": 1.9576598714733434e-05,
|
| 316 |
+
"loss": 0.9603,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.36885245901639346,
|
| 321 |
+
"grad_norm": 1.3336308002471924,
|
| 322 |
+
"learning_rate": 1.9550746003405996e-05,
|
| 323 |
+
"loss": 0.9402,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.3770491803278688,
|
| 328 |
+
"grad_norm": 1.2551720142364502,
|
| 329 |
+
"learning_rate": 1.9524145333581315e-05,
|
| 330 |
+
"loss": 0.9047,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.38524590163934425,
|
| 335 |
+
"grad_norm": 1.181113600730896,
|
| 336 |
+
"learning_rate": 1.9496798788468077e-05,
|
| 337 |
+
"loss": 0.9295,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.39344262295081966,
|
| 342 |
+
"grad_norm": 1.2969979047775269,
|
| 343 |
+
"learning_rate": 1.9468708509687545e-05,
|
| 344 |
+
"loss": 0.9376,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.4016393442622951,
|
| 349 |
+
"grad_norm": 1.4306528568267822,
|
| 350 |
+
"learning_rate": 1.943987669710586e-05,
|
| 351 |
+
"loss": 0.9563,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.4098360655737705,
|
| 356 |
+
"grad_norm": 1.2334694862365723,
|
| 357 |
+
"learning_rate": 1.9410305608661745e-05,
|
| 358 |
+
"loss": 0.957,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.4180327868852459,
|
| 363 |
+
"grad_norm": 1.2543516159057617,
|
| 364 |
+
"learning_rate": 1.9379997560189677e-05,
|
| 365 |
+
"loss": 0.9015,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.4262295081967213,
|
| 370 |
+
"grad_norm": 1.2323187589645386,
|
| 371 |
+
"learning_rate": 1.934895492523852e-05,
|
| 372 |
+
"loss": 0.896,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.4344262295081967,
|
| 377 |
+
"grad_norm": 1.5682944059371948,
|
| 378 |
+
"learning_rate": 1.931718013488566e-05,
|
| 379 |
+
"loss": 0.9062,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.4426229508196721,
|
| 384 |
+
"grad_norm": 1.1828402280807495,
|
| 385 |
+
"learning_rate": 1.92846756775466e-05,
|
| 386 |
+
"loss": 0.9079,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.45081967213114754,
|
| 391 |
+
"grad_norm": 1.1668050289154053,
|
| 392 |
+
"learning_rate": 1.9251444098780095e-05,
|
| 393 |
+
"loss": 0.8981,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.45901639344262296,
|
| 398 |
+
"grad_norm": 1.198763370513916,
|
| 399 |
+
"learning_rate": 1.9217488001088784e-05,
|
| 400 |
+
"loss": 0.9422,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.4672131147540984,
|
| 405 |
+
"grad_norm": 1.3131557703018188,
|
| 406 |
+
"learning_rate": 1.9182810043715394e-05,
|
| 407 |
+
"loss": 0.964,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.47540983606557374,
|
| 412 |
+
"grad_norm": 1.2967156171798706,
|
| 413 |
+
"learning_rate": 1.9147412942434463e-05,
|
| 414 |
+
"loss": 0.9674,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.48360655737704916,
|
| 419 |
+
"grad_norm": 1.229436993598938,
|
| 420 |
+
"learning_rate": 1.911129946933968e-05,
|
| 421 |
+
"loss": 0.9925,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.4918032786885246,
|
| 426 |
+
"grad_norm": 1.1797034740447998,
|
| 427 |
+
"learning_rate": 1.907447245262678e-05,
|
| 428 |
+
"loss": 0.9273,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.5,
|
| 433 |
+
"grad_norm": 1.27089524269104,
|
| 434 |
+
"learning_rate": 1.903693477637204e-05,
|
| 435 |
+
"loss": 0.9208,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.5081967213114754,
|
| 440 |
+
"grad_norm": 1.2703099250793457,
|
| 441 |
+
"learning_rate": 1.899868938030645e-05,
|
| 442 |
+
"loss": 0.8863,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.5163934426229508,
|
| 447 |
+
"grad_norm": 1.207794189453125,
|
| 448 |
+
"learning_rate": 1.8959739259585458e-05,
|
| 449 |
+
"loss": 0.8923,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.5245901639344263,
|
| 454 |
+
"grad_norm": 1.1332623958587646,
|
| 455 |
+
"learning_rate": 1.8920087464554427e-05,
|
| 456 |
+
"loss": 0.927,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.5327868852459017,
|
| 461 |
+
"grad_norm": 1.1637699604034424,
|
| 462 |
+
"learning_rate": 1.887973710050974e-05,
|
| 463 |
+
"loss": 0.9053,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.5409836065573771,
|
| 468 |
+
"grad_norm": 1.2687480449676514,
|
| 469 |
+
"learning_rate": 1.883869132745561e-05,
|
| 470 |
+
"loss": 0.9481,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.5491803278688525,
|
| 475 |
+
"grad_norm": 1.2181267738342285,
|
| 476 |
+
"learning_rate": 1.8796953359856626e-05,
|
| 477 |
+
"loss": 0.9686,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.5573770491803278,
|
| 482 |
+
"grad_norm": 1.388643503189087,
|
| 483 |
+
"learning_rate": 1.8754526466385986e-05,
|
| 484 |
+
"loss": 0.9844,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.5655737704918032,
|
| 489 |
+
"grad_norm": 1.1547937393188477,
|
| 490 |
+
"learning_rate": 1.871141396966953e-05,
|
| 491 |
+
"loss": 0.9252,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.5737704918032787,
|
| 496 |
+
"grad_norm": 1.2230730056762695,
|
| 497 |
+
"learning_rate": 1.8667619246025527e-05,
|
| 498 |
+
"loss": 0.9481,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.5819672131147541,
|
| 503 |
+
"grad_norm": 1.191734790802002,
|
| 504 |
+
"learning_rate": 1.862314572520028e-05,
|
| 505 |
+
"loss": 0.9079,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.5901639344262295,
|
| 510 |
+
"grad_norm": 1.2768336534500122,
|
| 511 |
+
"learning_rate": 1.8577996890099494e-05,
|
| 512 |
+
"loss": 0.9624,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.5983606557377049,
|
| 517 |
+
"grad_norm": 1.3291285037994385,
|
| 518 |
+
"learning_rate": 1.8532176276515538e-05,
|
| 519 |
+
"loss": 0.9393,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.6065573770491803,
|
| 524 |
+
"grad_norm": 1.1865758895874023,
|
| 525 |
+
"learning_rate": 1.8485687472850538e-05,
|
| 526 |
+
"loss": 0.8999,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.6147540983606558,
|
| 531 |
+
"grad_norm": 1.272955060005188,
|
| 532 |
+
"learning_rate": 1.8438534119835365e-05,
|
| 533 |
+
"loss": 0.9224,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.6229508196721312,
|
| 538 |
+
"grad_norm": 1.0975733995437622,
|
| 539 |
+
"learning_rate": 1.8390719910244487e-05,
|
| 540 |
+
"loss": 0.8791,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.6311475409836066,
|
| 545 |
+
"grad_norm": 1.091784119606018,
|
| 546 |
+
"learning_rate": 1.83422485886068e-05,
|
| 547 |
+
"loss": 0.9104,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.639344262295082,
|
| 552 |
+
"grad_norm": 1.349613904953003,
|
| 553 |
+
"learning_rate": 1.8293123950912362e-05,
|
| 554 |
+
"loss": 0.8847,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.6475409836065574,
|
| 559 |
+
"grad_norm": 1.1367007493972778,
|
| 560 |
+
"learning_rate": 1.8243349844315116e-05,
|
| 561 |
+
"loss": 0.8731,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.6557377049180327,
|
| 566 |
+
"grad_norm": 1.1368755102157593,
|
| 567 |
+
"learning_rate": 1.8192930166831618e-05,
|
| 568 |
+
"loss": 0.9351,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.6639344262295082,
|
| 573 |
+
"grad_norm": 1.2286845445632935,
|
| 574 |
+
"learning_rate": 1.8141868867035745e-05,
|
| 575 |
+
"loss": 0.9791,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.6721311475409836,
|
| 580 |
+
"grad_norm": 1.2607823610305786,
|
| 581 |
+
"learning_rate": 1.8090169943749477e-05,
|
| 582 |
+
"loss": 0.9597,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.680327868852459,
|
| 587 |
+
"grad_norm": 1.2413727045059204,
|
| 588 |
+
"learning_rate": 1.8037837445729733e-05,
|
| 589 |
+
"loss": 0.9638,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.6885245901639344,
|
| 594 |
+
"grad_norm": 1.2126213312149048,
|
| 595 |
+
"learning_rate": 1.79848754713513e-05,
|
| 596 |
+
"loss": 0.9077,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.6967213114754098,
|
| 601 |
+
"grad_norm": 1.1528116464614868,
|
| 602 |
+
"learning_rate": 1.7931288168285863e-05,
|
| 603 |
+
"loss": 0.9249,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.7049180327868853,
|
| 608 |
+
"grad_norm": 1.279722809791565,
|
| 609 |
+
"learning_rate": 1.7877079733177185e-05,
|
| 610 |
+
"loss": 0.9793,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.7131147540983607,
|
| 615 |
+
"grad_norm": 1.2270166873931885,
|
| 616 |
+
"learning_rate": 1.7822254411312455e-05,
|
| 617 |
+
"loss": 0.8639,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.7213114754098361,
|
| 622 |
+
"grad_norm": 1.2468152046203613,
|
| 623 |
+
"learning_rate": 1.7766816496289822e-05,
|
| 624 |
+
"loss": 0.8633,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.7295081967213115,
|
| 629 |
+
"grad_norm": 1.075137972831726,
|
| 630 |
+
"learning_rate": 1.7710770329682145e-05,
|
| 631 |
+
"loss": 0.8652,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.7377049180327869,
|
| 636 |
+
"grad_norm": 1.332723617553711,
|
| 637 |
+
"learning_rate": 1.765412030069698e-05,
|
| 638 |
+
"loss": 0.9574,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.7459016393442623,
|
| 643 |
+
"grad_norm": 1.1367098093032837,
|
| 644 |
+
"learning_rate": 1.759687084583285e-05,
|
| 645 |
+
"loss": 0.9005,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.7540983606557377,
|
| 650 |
+
"grad_norm": 1.1797839403152466,
|
| 651 |
+
"learning_rate": 1.753902644853181e-05,
|
| 652 |
+
"loss": 0.8773,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.7622950819672131,
|
| 657 |
+
"grad_norm": 1.1969019174575806,
|
| 658 |
+
"learning_rate": 1.748059163882831e-05,
|
| 659 |
+
"loss": 0.9298,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.7704918032786885,
|
| 664 |
+
"grad_norm": 1.2756307125091553,
|
| 665 |
+
"learning_rate": 1.742157099299445e-05,
|
| 666 |
+
"loss": 0.8985,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.7786885245901639,
|
| 671 |
+
"grad_norm": 1.17881441116333,
|
| 672 |
+
"learning_rate": 1.7361969133181585e-05,
|
| 673 |
+
"loss": 0.8693,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.7868852459016393,
|
| 678 |
+
"grad_norm": 1.192720651626587,
|
| 679 |
+
"learning_rate": 1.7301790727058344e-05,
|
| 680 |
+
"loss": 0.9112,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.7950819672131147,
|
| 685 |
+
"grad_norm": 1.243441104888916,
|
| 686 |
+
"learning_rate": 1.7241040487445082e-05,
|
| 687 |
+
"loss": 1.0005,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.8032786885245902,
|
| 692 |
+
"grad_norm": 1.2219005823135376,
|
| 693 |
+
"learning_rate": 1.7179723171944812e-05,
|
| 694 |
+
"loss": 0.9575,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.8114754098360656,
|
| 699 |
+
"grad_norm": 1.169037103652954,
|
| 700 |
+
"learning_rate": 1.7117843582570608e-05,
|
| 701 |
+
"loss": 0.9366,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.819672131147541,
|
| 706 |
+
"grad_norm": 1.2007153034210205,
|
| 707 |
+
"learning_rate": 1.705540656536953e-05,
|
| 708 |
+
"loss": 0.9627,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.8278688524590164,
|
| 713 |
+
"grad_norm": 1.1723730564117432,
|
| 714 |
+
"learning_rate": 1.6992417010043144e-05,
|
| 715 |
+
"loss": 0.9362,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.8360655737704918,
|
| 720 |
+
"grad_norm": 1.0589674711227417,
|
| 721 |
+
"learning_rate": 1.692887984956454e-05,
|
| 722 |
+
"loss": 0.8873,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.8442622950819673,
|
| 727 |
+
"grad_norm": 1.1718519926071167,
|
| 728 |
+
"learning_rate": 1.6864800059792057e-05,
|
| 729 |
+
"loss": 0.8989,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.8524590163934426,
|
| 734 |
+
"grad_norm": 1.290388584136963,
|
| 735 |
+
"learning_rate": 1.6800182659079568e-05,
|
| 736 |
+
"loss": 0.9364,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.860655737704918,
|
| 741 |
+
"grad_norm": 1.2954158782958984,
|
| 742 |
+
"learning_rate": 1.6735032707883502e-05,
|
| 743 |
+
"loss": 0.8774,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.8688524590163934,
|
| 748 |
+
"grad_norm": 1.2934119701385498,
|
| 749 |
+
"learning_rate": 1.666935530836651e-05,
|
| 750 |
+
"loss": 0.9579,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.8770491803278688,
|
| 755 |
+
"grad_norm": 1.230776309967041,
|
| 756 |
+
"learning_rate": 1.6603155603997908e-05,
|
| 757 |
+
"loss": 0.9589,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.8852459016393442,
|
| 762 |
+
"grad_norm": 1.1555577516555786,
|
| 763 |
+
"learning_rate": 1.653643877915088e-05,
|
| 764 |
+
"loss": 0.898,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.8934426229508197,
|
| 769 |
+
"grad_norm": 1.1116188764572144,
|
| 770 |
+
"learning_rate": 1.6469210058696448e-05,
|
| 771 |
+
"loss": 0.8871,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.9016393442622951,
|
| 776 |
+
"grad_norm": 1.1556096076965332,
|
| 777 |
+
"learning_rate": 1.64014747075943e-05,
|
| 778 |
+
"loss": 0.9254,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.9098360655737705,
|
| 783 |
+
"grad_norm": 1.3375847339630127,
|
| 784 |
+
"learning_rate": 1.6333238030480473e-05,
|
| 785 |
+
"loss": 0.9108,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.9180327868852459,
|
| 790 |
+
"grad_norm": 1.2122970819473267,
|
| 791 |
+
"learning_rate": 1.626450537125192e-05,
|
| 792 |
+
"loss": 0.9578,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.9262295081967213,
|
| 797 |
+
"grad_norm": 1.3110710382461548,
|
| 798 |
+
"learning_rate": 1.6195282112648007e-05,
|
| 799 |
+
"loss": 0.9618,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.9344262295081968,
|
| 804 |
+
"grad_norm": 1.1797089576721191,
|
| 805 |
+
"learning_rate": 1.6125573675828984e-05,
|
| 806 |
+
"loss": 0.9279,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.9426229508196722,
|
| 811 |
+
"grad_norm": 1.2419110536575317,
|
| 812 |
+
"learning_rate": 1.605538551995139e-05,
|
| 813 |
+
"loss": 0.8881,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.9508196721311475,
|
| 818 |
+
"grad_norm": 1.2254469394683838,
|
| 819 |
+
"learning_rate": 1.5984723141740578e-05,
|
| 820 |
+
"loss": 0.8697,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.9590163934426229,
|
| 825 |
+
"grad_norm": 1.1255691051483154,
|
| 826 |
+
"learning_rate": 1.59135920750602e-05,
|
| 827 |
+
"loss": 0.8646,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.9672131147540983,
|
| 832 |
+
"grad_norm": 1.2165420055389404,
|
| 833 |
+
"learning_rate": 1.5841997890478853e-05,
|
| 834 |
+
"loss": 0.9259,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.9754098360655737,
|
| 839 |
+
"grad_norm": 1.205811858177185,
|
| 840 |
+
"learning_rate": 1.5769946194833816e-05,
|
| 841 |
+
"loss": 0.9419,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.9836065573770492,
|
| 846 |
+
"grad_norm": 1.3705413341522217,
|
| 847 |
+
"learning_rate": 1.569744263079195e-05,
|
| 848 |
+
"loss": 0.9238,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.9918032786885246,
|
| 853 |
+
"grad_norm": 1.0843297243118286,
|
| 854 |
+
"learning_rate": 1.562449287640781e-05,
|
| 855 |
+
"loss": 0.8372,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 1.0,
|
| 860 |
+
"grad_norm": 1.2068427801132202,
|
| 861 |
+
"learning_rate": 1.555110264467895e-05,
|
| 862 |
+
"loss": 0.8898,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 1.0081967213114753,
|
| 867 |
+
"grad_norm": 1.5948082208633423,
|
| 868 |
+
"learning_rate": 1.5477277683098555e-05,
|
| 869 |
+
"loss": 0.5819,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 1.0163934426229508,
|
| 874 |
+
"grad_norm": 1.470932126045227,
|
| 875 |
+
"learning_rate": 1.5403023773205286e-05,
|
| 876 |
+
"loss": 0.5502,
|
| 877 |
+
"step": 124
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 1.0245901639344261,
|
| 881 |
+
"grad_norm": 1.2653380632400513,
|
| 882 |
+
"learning_rate": 1.532834673013053e-05,
|
| 883 |
+
"loss": 0.5021,
|
| 884 |
+
"step": 125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.0327868852459017,
|
| 888 |
+
"grad_norm": 1.1609309911727905,
|
| 889 |
+
"learning_rate": 1.5253252402142989e-05,
|
| 890 |
+
"loss": 0.5464,
|
| 891 |
+
"step": 126
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 1.040983606557377,
|
| 895 |
+
"grad_norm": 1.3639906644821167,
|
| 896 |
+
"learning_rate": 1.5177746670190674e-05,
|
| 897 |
+
"loss": 0.502,
|
| 898 |
+
"step": 127
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 1.0491803278688525,
|
| 902 |
+
"grad_norm": 1.6784329414367676,
|
| 903 |
+
"learning_rate": 1.5101835447440346e-05,
|
| 904 |
+
"loss": 0.5305,
|
| 905 |
+
"step": 128
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 1.0573770491803278,
|
| 909 |
+
"grad_norm": 1.5389288663864136,
|
| 910 |
+
"learning_rate": 1.5025524678814428e-05,
|
| 911 |
+
"loss": 0.5165,
|
| 912 |
+
"step": 129
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 1.0655737704918034,
|
| 916 |
+
"grad_norm": 1.4360255002975464,
|
| 917 |
+
"learning_rate": 1.4948820340525439e-05,
|
| 918 |
+
"loss": 0.5415,
|
| 919 |
+
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 1.0737704918032787,
|
| 923 |
+
"grad_norm": 1.36026930809021,
|
| 924 |
+
"learning_rate": 1.4871728439607967e-05,
|
| 925 |
+
"loss": 0.5415,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.0819672131147542,
|
| 930 |
+
"grad_norm": 1.2092355489730835,
|
| 931 |
+
"learning_rate": 1.4794255013448242e-05,
|
| 932 |
+
"loss": 0.5053,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 1.0901639344262295,
|
| 937 |
+
"grad_norm": 1.2120226621627808,
|
| 938 |
+
"learning_rate": 1.4716406129311307e-05,
|
| 939 |
+
"loss": 0.4832,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 1.098360655737705,
|
| 944 |
+
"grad_norm": 1.3007643222808838,
|
| 945 |
+
"learning_rate": 1.4638187883865881e-05,
|
| 946 |
+
"loss": 0.4911,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 1.1065573770491803,
|
| 951 |
+
"grad_norm": 1.218314528465271,
|
| 952 |
+
"learning_rate": 1.45596064027069e-05,
|
| 953 |
+
"loss": 0.4783,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 1.1147540983606556,
|
| 958 |
+
"grad_norm": 1.0904790163040161,
|
| 959 |
+
"learning_rate": 1.4480667839875786e-05,
|
| 960 |
+
"loss": 0.4907,
|
| 961 |
+
"step": 136
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 1.1229508196721312,
|
| 965 |
+
"grad_norm": 1.2957193851470947,
|
| 966 |
+
"learning_rate": 1.4401378377378512e-05,
|
| 967 |
+
"loss": 0.5176,
|
| 968 |
+
"step": 137
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 1.1311475409836065,
|
| 972 |
+
"grad_norm": 1.4495291709899902,
|
| 973 |
+
"learning_rate": 1.432174422470146e-05,
|
| 974 |
+
"loss": 0.4754,
|
| 975 |
+
"step": 138
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 1.139344262295082,
|
| 979 |
+
"grad_norm": 1.339431881904602,
|
| 980 |
+
"learning_rate": 1.4241771618325123e-05,
|
| 981 |
+
"loss": 0.4744,
|
| 982 |
+
"step": 139
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 1.1475409836065573,
|
| 986 |
+
"grad_norm": 1.278767704963684,
|
| 987 |
+
"learning_rate": 1.4161466821235705e-05,
|
| 988 |
+
"loss": 0.4688,
|
| 989 |
+
"step": 140
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 1.1557377049180328,
|
| 993 |
+
"grad_norm": 1.2969565391540527,
|
| 994 |
+
"learning_rate": 1.408083612243465e-05,
|
| 995 |
+
"loss": 0.4792,
|
| 996 |
+
"step": 141
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 1.1639344262295082,
|
| 1000 |
+
"grad_norm": 1.364591360092163,
|
| 1001 |
+
"learning_rate": 1.3999885836446106e-05,
|
| 1002 |
+
"loss": 0.5112,
|
| 1003 |
+
"step": 142
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 1.1721311475409837,
|
| 1007 |
+
"grad_norm": 1.34937584400177,
|
| 1008 |
+
"learning_rate": 1.3918622302822425e-05,
|
| 1009 |
+
"loss": 0.4853,
|
| 1010 |
+
"step": 143
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 1.180327868852459,
|
| 1014 |
+
"grad_norm": 1.163983702659607,
|
| 1015 |
+
"learning_rate": 1.3837051885647673e-05,
|
| 1016 |
+
"loss": 0.449,
|
| 1017 |
+
"step": 144
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 1.1885245901639343,
|
| 1021 |
+
"grad_norm": 1.3048903942108154,
|
| 1022 |
+
"learning_rate": 1.3755180973039241e-05,
|
| 1023 |
+
"loss": 0.5604,
|
| 1024 |
+
"step": 145
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 1.1967213114754098,
|
| 1028 |
+
"grad_norm": 1.252719521522522,
|
| 1029 |
+
"learning_rate": 1.367301597664757e-05,
|
| 1030 |
+
"loss": 0.5214,
|
| 1031 |
+
"step": 146
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 1.2049180327868854,
|
| 1035 |
+
"grad_norm": 1.1311830282211304,
|
| 1036 |
+
"learning_rate": 1.3590563331154008e-05,
|
| 1037 |
+
"loss": 0.4601,
|
| 1038 |
+
"step": 147
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 1.2131147540983607,
|
| 1042 |
+
"grad_norm": 1.1108665466308594,
|
| 1043 |
+
"learning_rate": 1.3507829493766903e-05,
|
| 1044 |
+
"loss": 0.4873,
|
| 1045 |
+
"step": 148
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 1.221311475409836,
|
| 1049 |
+
"grad_norm": 1.1364582777023315,
|
| 1050 |
+
"learning_rate": 1.342482094371591e-05,
|
| 1051 |
+
"loss": 0.4507,
|
| 1052 |
+
"step": 149
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 1.2295081967213115,
|
| 1056 |
+
"grad_norm": 1.1517623662948608,
|
| 1057 |
+
"learning_rate": 1.3341544181744559e-05,
|
| 1058 |
+
"loss": 0.5004,
|
| 1059 |
+
"step": 150
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 1.2377049180327868,
|
| 1063 |
+
"grad_norm": 1.146333932876587,
|
| 1064 |
+
"learning_rate": 1.3258005729601178e-05,
|
| 1065 |
+
"loss": 0.4714,
|
| 1066 |
+
"step": 151
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 1.2459016393442623,
|
| 1070 |
+
"grad_norm": 1.3069875240325928,
|
| 1071 |
+
"learning_rate": 1.317421212952813e-05,
|
| 1072 |
+
"loss": 0.5107,
|
| 1073 |
+
"step": 152
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 1.2540983606557377,
|
| 1077 |
+
"grad_norm": 1.3697153329849243,
|
| 1078 |
+
"learning_rate": 1.3090169943749475e-05,
|
| 1079 |
+
"loss": 0.4996,
|
| 1080 |
+
"step": 153
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 1.2622950819672132,
|
| 1084 |
+
"grad_norm": 1.2243149280548096,
|
| 1085 |
+
"learning_rate": 1.3005885753957048e-05,
|
| 1086 |
+
"loss": 0.4929,
|
| 1087 |
+
"step": 154
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 1.2704918032786885,
|
| 1091 |
+
"grad_norm": 1.269228219985962,
|
| 1092 |
+
"learning_rate": 1.2921366160795017e-05,
|
| 1093 |
+
"loss": 0.4696,
|
| 1094 |
+
"step": 155
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 1.278688524590164,
|
| 1098 |
+
"grad_norm": 1.2698426246643066,
|
| 1099 |
+
"learning_rate": 1.2836617783342968e-05,
|
| 1100 |
+
"loss": 0.4948,
|
| 1101 |
+
"step": 156
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 1.2868852459016393,
|
| 1105 |
+
"grad_norm": 1.3285802602767944,
|
| 1106 |
+
"learning_rate": 1.2751647258597531e-05,
|
| 1107 |
+
"loss": 0.5147,
|
| 1108 |
+
"step": 157
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 1.2950819672131146,
|
| 1112 |
+
"grad_norm": 1.1310343742370605,
|
| 1113 |
+
"learning_rate": 1.2666461240952615e-05,
|
| 1114 |
+
"loss": 0.4979,
|
| 1115 |
+
"step": 158
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 1.3032786885245902,
|
| 1119 |
+
"grad_norm": 1.2382618188858032,
|
| 1120 |
+
"learning_rate": 1.2581066401678261e-05,
|
| 1121 |
+
"loss": 0.4411,
|
| 1122 |
+
"step": 159
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 1.3114754098360657,
|
| 1126 |
+
"grad_norm": 1.159347653388977,
|
| 1127 |
+
"learning_rate": 1.249546942839821e-05,
|
| 1128 |
+
"loss": 0.474,
|
| 1129 |
+
"step": 160
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 1.319672131147541,
|
| 1133 |
+
"grad_norm": 1.2726000547409058,
|
| 1134 |
+
"learning_rate": 1.2409677024566145e-05,
|
| 1135 |
+
"loss": 0.4959,
|
| 1136 |
+
"step": 161
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 1.3278688524590163,
|
| 1140 |
+
"grad_norm": 1.1478623151779175,
|
| 1141 |
+
"learning_rate": 1.2323695908940729e-05,
|
| 1142 |
+
"loss": 0.4486,
|
| 1143 |
+
"step": 162
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 1.3360655737704918,
|
| 1147 |
+
"grad_norm": 1.1755861043930054,
|
| 1148 |
+
"learning_rate": 1.2237532815059427e-05,
|
| 1149 |
+
"loss": 0.4866,
|
| 1150 |
+
"step": 163
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 1.3442622950819672,
|
| 1154 |
+
"grad_norm": 1.1865427494049072,
|
| 1155 |
+
"learning_rate": 1.2151194490711177e-05,
|
| 1156 |
+
"loss": 0.5105,
|
| 1157 |
+
"step": 164
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 1.3524590163934427,
|
| 1161 |
+
"grad_norm": 1.0980228185653687,
|
| 1162 |
+
"learning_rate": 1.2064687697407939e-05,
|
| 1163 |
+
"loss": 0.4791,
|
| 1164 |
+
"step": 165
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 1.360655737704918,
|
| 1168 |
+
"grad_norm": 1.2174019813537598,
|
| 1169 |
+
"learning_rate": 1.1978019209855174e-05,
|
| 1170 |
+
"loss": 0.5069,
|
| 1171 |
+
"step": 166
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 1.3688524590163935,
|
| 1175 |
+
"grad_norm": 1.2120226621627808,
|
| 1176 |
+
"learning_rate": 1.1891195815421293e-05,
|
| 1177 |
+
"loss": 0.5062,
|
| 1178 |
+
"step": 167
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 1.3770491803278688,
|
| 1182 |
+
"grad_norm": 1.1092822551727295,
|
| 1183 |
+
"learning_rate": 1.1804224313606104e-05,
|
| 1184 |
+
"loss": 0.4787,
|
| 1185 |
+
"step": 168
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 1.3852459016393444,
|
| 1189 |
+
"grad_norm": 1.2074278593063354,
|
| 1190 |
+
"learning_rate": 1.171711151550832e-05,
|
| 1191 |
+
"loss": 0.5181,
|
| 1192 |
+
"step": 169
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 1.3934426229508197,
|
| 1196 |
+
"grad_norm": 1.0840147733688354,
|
| 1197 |
+
"learning_rate": 1.1629864243292149e-05,
|
| 1198 |
+
"loss": 0.5098,
|
| 1199 |
+
"step": 170
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 1.401639344262295,
|
| 1203 |
+
"grad_norm": 1.2134615182876587,
|
| 1204 |
+
"learning_rate": 1.1542489329653024e-05,
|
| 1205 |
+
"loss": 0.5023,
|
| 1206 |
+
"step": 171
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 1.4098360655737705,
|
| 1210 |
+
"grad_norm": 1.2024002075195312,
|
| 1211 |
+
"learning_rate": 1.1454993617282513e-05,
|
| 1212 |
+
"loss": 0.489,
|
| 1213 |
+
"step": 172
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 1.418032786885246,
|
| 1217 |
+
"grad_norm": 1.4277188777923584,
|
| 1218 |
+
"learning_rate": 1.1367383958332427e-05,
|
| 1219 |
+
"loss": 0.4997,
|
| 1220 |
+
"step": 173
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 1.4262295081967213,
|
| 1224 |
+
"grad_norm": 1.2423990964889526,
|
| 1225 |
+
"learning_rate": 1.1279667213878206e-05,
|
| 1226 |
+
"loss": 0.5328,
|
| 1227 |
+
"step": 174
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 1.4344262295081966,
|
| 1231 |
+
"grad_norm": 1.1874176263809204,
|
| 1232 |
+
"learning_rate": 1.1191850253381602e-05,
|
| 1233 |
+
"loss": 0.5211,
|
| 1234 |
+
"step": 175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 1.4426229508196722,
|
| 1238 |
+
"grad_norm": 1.1690136194229126,
|
| 1239 |
+
"learning_rate": 1.11039399541527e-05,
|
| 1240 |
+
"loss": 0.4499,
|
| 1241 |
+
"step": 176
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 1.4508196721311475,
|
| 1245 |
+
"grad_norm": 1.3457874059677124,
|
| 1246 |
+
"learning_rate": 1.1015943200811325e-05,
|
| 1247 |
+
"loss": 0.5211,
|
| 1248 |
+
"step": 177
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 1.459016393442623,
|
| 1252 |
+
"grad_norm": 1.193812370300293,
|
| 1253 |
+
"learning_rate": 1.092786688474789e-05,
|
| 1254 |
+
"loss": 0.5259,
|
| 1255 |
+
"step": 178
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 1.4672131147540983,
|
| 1259 |
+
"grad_norm": 1.3439817428588867,
|
| 1260 |
+
"learning_rate": 1.0839717903583684e-05,
|
| 1261 |
+
"loss": 0.5,
|
| 1262 |
+
"step": 179
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 1.4754098360655736,
|
| 1266 |
+
"grad_norm": 1.1418300867080688,
|
| 1267 |
+
"learning_rate": 1.075150316063071e-05,
|
| 1268 |
+
"loss": 0.5322,
|
| 1269 |
+
"step": 180
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 1.4836065573770492,
|
| 1273 |
+
"grad_norm": 1.2361342906951904,
|
| 1274 |
+
"learning_rate": 1.066322956435104e-05,
|
| 1275 |
+
"loss": 0.4733,
|
| 1276 |
+
"step": 181
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 1.4918032786885247,
|
| 1280 |
+
"grad_norm": 1.2802929878234863,
|
| 1281 |
+
"learning_rate": 1.0574904027815802e-05,
|
| 1282 |
+
"loss": 0.4741,
|
| 1283 |
+
"step": 182
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 1.5,
|
| 1287 |
+
"grad_norm": 1.209621787071228,
|
| 1288 |
+
"learning_rate": 1.0486533468163782e-05,
|
| 1289 |
+
"loss": 0.5068,
|
| 1290 |
+
"step": 183
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 1.5081967213114753,
|
| 1294 |
+
"grad_norm": 1.2289994955062866,
|
| 1295 |
+
"learning_rate": 1.0398124806059702e-05,
|
| 1296 |
+
"loss": 0.5167,
|
| 1297 |
+
"step": 184
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 1.5163934426229508,
|
| 1301 |
+
"grad_norm": 1.4493712186813354,
|
| 1302 |
+
"learning_rate": 1.0309684965152254e-05,
|
| 1303 |
+
"loss": 0.4767,
|
| 1304 |
+
"step": 185
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 1.5245901639344264,
|
| 1308 |
+
"grad_norm": 1.2684192657470703,
|
| 1309 |
+
"learning_rate": 1.022122087153187e-05,
|
| 1310 |
+
"loss": 0.5124,
|
| 1311 |
+
"step": 186
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 1.5327868852459017,
|
| 1315 |
+
"grad_norm": 1.335960030555725,
|
| 1316 |
+
"learning_rate": 1.013273945318831e-05,
|
| 1317 |
+
"loss": 0.5138,
|
| 1318 |
+
"step": 187
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 1.540983606557377,
|
| 1322 |
+
"grad_norm": 1.157732605934143,
|
| 1323 |
+
"learning_rate": 1.0044247639468106e-05,
|
| 1324 |
+
"loss": 0.4959,
|
| 1325 |
+
"step": 188
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 1.5491803278688525,
|
| 1329 |
+
"grad_norm": 1.152340054512024,
|
| 1330 |
+
"learning_rate": 9.955752360531896e-06,
|
| 1331 |
+
"loss": 0.4647,
|
| 1332 |
+
"step": 189
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 1.5573770491803278,
|
| 1336 |
+
"grad_norm": 1.3083349466323853,
|
| 1337 |
+
"learning_rate": 9.867260546811694e-06,
|
| 1338 |
+
"loss": 0.5095,
|
| 1339 |
+
"step": 190
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 1.5655737704918034,
|
| 1343 |
+
"grad_norm": 1.3373643159866333,
|
| 1344 |
+
"learning_rate": 9.778779128468133e-06,
|
| 1345 |
+
"loss": 0.5087,
|
| 1346 |
+
"step": 191
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 1.5737704918032787,
|
| 1350 |
+
"grad_norm": 1.1853665113449097,
|
| 1351 |
+
"learning_rate": 9.690315034847749e-06,
|
| 1352 |
+
"loss": 0.45,
|
| 1353 |
+
"step": 192
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 1.581967213114754,
|
| 1357 |
+
"grad_norm": 1.1577143669128418,
|
| 1358 |
+
"learning_rate": 9.601875193940301e-06,
|
| 1359 |
+
"loss": 0.4622,
|
| 1360 |
+
"step": 193
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 1.5901639344262295,
|
| 1364 |
+
"grad_norm": 1.1653783321380615,
|
| 1365 |
+
"learning_rate": 9.513466531836223e-06,
|
| 1366 |
+
"loss": 0.4463,
|
| 1367 |
+
"step": 194
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 1.598360655737705,
|
| 1371 |
+
"grad_norm": 1.350871205329895,
|
| 1372 |
+
"learning_rate": 9.4250959721842e-06,
|
| 1373 |
+
"loss": 0.5065,
|
| 1374 |
+
"step": 195
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 1.6065573770491803,
|
| 1378 |
+
"grad_norm": 1.194115161895752,
|
| 1379 |
+
"learning_rate": 9.336770435648963e-06,
|
| 1380 |
+
"loss": 0.4859,
|
| 1381 |
+
"step": 196
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 1.6147540983606556,
|
| 1385 |
+
"grad_norm": 1.209192156791687,
|
| 1386 |
+
"learning_rate": 9.248496839369293e-06,
|
| 1387 |
+
"loss": 0.4611,
|
| 1388 |
+
"step": 197
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 1.6229508196721312,
|
| 1392 |
+
"grad_norm": 1.228793740272522,
|
| 1393 |
+
"learning_rate": 9.160282096416317e-06,
|
| 1394 |
+
"loss": 0.4808,
|
| 1395 |
+
"step": 198
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 1.6311475409836067,
|
| 1399 |
+
"grad_norm": 1.4064335823059082,
|
| 1400 |
+
"learning_rate": 9.072133115252113e-06,
|
| 1401 |
+
"loss": 0.5201,
|
| 1402 |
+
"step": 199
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 1.639344262295082,
|
| 1406 |
+
"grad_norm": 1.3964109420776367,
|
| 1407 |
+
"learning_rate": 8.984056799188678e-06,
|
| 1408 |
+
"loss": 0.5148,
|
| 1409 |
+
"step": 200
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 1.6475409836065573,
|
| 1413 |
+
"grad_norm": 1.1933867931365967,
|
| 1414 |
+
"learning_rate": 8.896060045847305e-06,
|
| 1415 |
+
"loss": 0.4711,
|
| 1416 |
+
"step": 201
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 1.6557377049180326,
|
| 1420 |
+
"grad_norm": 1.16603422164917,
|
| 1421 |
+
"learning_rate": 8.808149746618403e-06,
|
| 1422 |
+
"loss": 0.5169,
|
| 1423 |
+
"step": 202
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 1.6639344262295082,
|
| 1427 |
+
"grad_norm": 1.174723505973816,
|
| 1428 |
+
"learning_rate": 8.7203327861218e-06,
|
| 1429 |
+
"loss": 0.4993,
|
| 1430 |
+
"step": 203
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 1.6721311475409837,
|
| 1434 |
+
"grad_norm": 1.1034168004989624,
|
| 1435 |
+
"learning_rate": 8.632616041667577e-06,
|
| 1436 |
+
"loss": 0.4484,
|
| 1437 |
+
"step": 204
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 1.680327868852459,
|
| 1441 |
+
"grad_norm": 1.1651763916015625,
|
| 1442 |
+
"learning_rate": 8.545006382717487e-06,
|
| 1443 |
+
"loss": 0.52,
|
| 1444 |
+
"step": 205
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 1.6885245901639343,
|
| 1448 |
+
"grad_norm": 1.218217134475708,
|
| 1449 |
+
"learning_rate": 8.457510670346976e-06,
|
| 1450 |
+
"loss": 0.4793,
|
| 1451 |
+
"step": 206
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 1.6967213114754098,
|
| 1455 |
+
"grad_norm": 1.2980520725250244,
|
| 1456 |
+
"learning_rate": 8.370135756707853e-06,
|
| 1457 |
+
"loss": 0.5435,
|
| 1458 |
+
"step": 207
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 1.7049180327868854,
|
| 1462 |
+
"grad_norm": 1.0842329263687134,
|
| 1463 |
+
"learning_rate": 8.282888484491682e-06,
|
| 1464 |
+
"loss": 0.4763,
|
| 1465 |
+
"step": 208
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 1.7131147540983607,
|
| 1469 |
+
"grad_norm": 1.2294825315475464,
|
| 1470 |
+
"learning_rate": 8.195775686393898e-06,
|
| 1471 |
+
"loss": 0.4812,
|
| 1472 |
+
"step": 209
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 1.721311475409836,
|
| 1476 |
+
"grad_norm": 1.2639816999435425,
|
| 1477 |
+
"learning_rate": 8.108804184578708e-06,
|
| 1478 |
+
"loss": 0.4887,
|
| 1479 |
+
"step": 210
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 1.7295081967213115,
|
| 1483 |
+
"grad_norm": 1.2210934162139893,
|
| 1484 |
+
"learning_rate": 8.021980790144828e-06,
|
| 1485 |
+
"loss": 0.4869,
|
| 1486 |
+
"step": 211
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 1.737704918032787,
|
| 1490 |
+
"grad_norm": 1.3877686262130737,
|
| 1491 |
+
"learning_rate": 7.935312302592062e-06,
|
| 1492 |
+
"loss": 0.4897,
|
| 1493 |
+
"step": 212
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 1.7459016393442623,
|
| 1497 |
+
"grad_norm": 1.2528547048568726,
|
| 1498 |
+
"learning_rate": 7.848805509288824e-06,
|
| 1499 |
+
"loss": 0.5066,
|
| 1500 |
+
"step": 213
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 1.7540983606557377,
|
| 1504 |
+
"grad_norm": 1.2030199766159058,
|
| 1505 |
+
"learning_rate": 7.762467184940574e-06,
|
| 1506 |
+
"loss": 0.4363,
|
| 1507 |
+
"step": 214
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 1.762295081967213,
|
| 1511 |
+
"grad_norm": 1.2085492610931396,
|
| 1512 |
+
"learning_rate": 7.676304091059273e-06,
|
| 1513 |
+
"loss": 0.4781,
|
| 1514 |
+
"step": 215
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 1.7704918032786885,
|
| 1518 |
+
"grad_norm": 1.2072522640228271,
|
| 1519 |
+
"learning_rate": 7.590322975433857e-06,
|
| 1520 |
+
"loss": 0.5077,
|
| 1521 |
+
"step": 216
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 1.778688524590164,
|
| 1525 |
+
"grad_norm": 1.2836977243423462,
|
| 1526 |
+
"learning_rate": 7.504530571601792e-06,
|
| 1527 |
+
"loss": 0.5191,
|
| 1528 |
+
"step": 217
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 1.7868852459016393,
|
| 1532 |
+
"grad_norm": 1.2558234930038452,
|
| 1533 |
+
"learning_rate": 7.418933598321741e-06,
|
| 1534 |
+
"loss": 0.4889,
|
| 1535 |
+
"step": 218
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 1.7950819672131146,
|
| 1539 |
+
"grad_norm": 1.2173186540603638,
|
| 1540 |
+
"learning_rate": 7.33353875904739e-06,
|
| 1541 |
+
"loss": 0.5121,
|
| 1542 |
+
"step": 219
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 1.8032786885245902,
|
| 1546 |
+
"grad_norm": 1.1933339834213257,
|
| 1547 |
+
"learning_rate": 7.2483527414024714e-06,
|
| 1548 |
+
"loss": 0.4579,
|
| 1549 |
+
"step": 220
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 1.8114754098360657,
|
| 1553 |
+
"grad_norm": 1.3315452337265015,
|
| 1554 |
+
"learning_rate": 7.163382216657033e-06,
|
| 1555 |
+
"loss": 0.4862,
|
| 1556 |
+
"step": 221
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 1.819672131147541,
|
| 1560 |
+
"grad_norm": 1.3630664348602295,
|
| 1561 |
+
"learning_rate": 7.078633839204984e-06,
|
| 1562 |
+
"loss": 0.5584,
|
| 1563 |
+
"step": 222
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 1.8278688524590163,
|
| 1567 |
+
"grad_norm": 1.1819186210632324,
|
| 1568 |
+
"learning_rate": 6.9941142460429555e-06,
|
| 1569 |
+
"loss": 0.4834,
|
| 1570 |
+
"step": 223
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 1.8360655737704918,
|
| 1574 |
+
"grad_norm": 1.1540125608444214,
|
| 1575 |
+
"learning_rate": 6.909830056250527e-06,
|
| 1576 |
+
"loss": 0.4533,
|
| 1577 |
+
"step": 224
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 1.8442622950819674,
|
| 1581 |
+
"grad_norm": 1.422378420829773,
|
| 1582 |
+
"learning_rate": 6.825787870471873e-06,
|
| 1583 |
+
"loss": 0.5382,
|
| 1584 |
+
"step": 225
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 1.8524590163934427,
|
| 1588 |
+
"grad_norm": 1.2260661125183105,
|
| 1589 |
+
"learning_rate": 6.741994270398826e-06,
|
| 1590 |
+
"loss": 0.4672,
|
| 1591 |
+
"step": 226
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 1.860655737704918,
|
| 1595 |
+
"grad_norm": 1.1611032485961914,
|
| 1596 |
+
"learning_rate": 6.658455818255445e-06,
|
| 1597 |
+
"loss": 0.5085,
|
| 1598 |
+
"step": 227
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 1.8688524590163933,
|
| 1602 |
+
"grad_norm": 1.3924514055252075,
|
| 1603 |
+
"learning_rate": 6.575179056284095e-06,
|
| 1604 |
+
"loss": 0.443,
|
| 1605 |
+
"step": 228
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 1.8770491803278688,
|
| 1609 |
+
"grad_norm": 1.0598273277282715,
|
| 1610 |
+
"learning_rate": 6.4921705062331e-06,
|
| 1611 |
+
"loss": 0.4768,
|
| 1612 |
+
"step": 229
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 1.8852459016393444,
|
| 1616 |
+
"grad_norm": 1.282805323600769,
|
| 1617 |
+
"learning_rate": 6.409436668845997e-06,
|
| 1618 |
+
"loss": 0.4935,
|
| 1619 |
+
"step": 230
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 1.8934426229508197,
|
| 1623 |
+
"grad_norm": 1.166425347328186,
|
| 1624 |
+
"learning_rate": 6.326984023352435e-06,
|
| 1625 |
+
"loss": 0.4613,
|
| 1626 |
+
"step": 231
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 1.901639344262295,
|
| 1630 |
+
"grad_norm": 1.2956987619400024,
|
| 1631 |
+
"learning_rate": 6.2448190269607625e-06,
|
| 1632 |
+
"loss": 0.46,
|
| 1633 |
+
"step": 232
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 1.9098360655737705,
|
| 1637 |
+
"grad_norm": 1.2106932401657104,
|
| 1638 |
+
"learning_rate": 6.162948114352328e-06,
|
| 1639 |
+
"loss": 0.483,
|
| 1640 |
+
"step": 233
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 1.918032786885246,
|
| 1644 |
+
"grad_norm": 1.1119098663330078,
|
| 1645 |
+
"learning_rate": 6.081377697177576e-06,
|
| 1646 |
+
"loss": 0.4535,
|
| 1647 |
+
"step": 234
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 1.9262295081967213,
|
| 1651 |
+
"grad_norm": 1.122563123703003,
|
| 1652 |
+
"learning_rate": 6.000114163553894e-06,
|
| 1653 |
+
"loss": 0.475,
|
| 1654 |
+
"step": 235
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 1.9344262295081966,
|
| 1658 |
+
"grad_norm": 1.2507013082504272,
|
| 1659 |
+
"learning_rate": 5.919163877565351e-06,
|
| 1660 |
+
"loss": 0.5149,
|
| 1661 |
+
"step": 236
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 1.9426229508196722,
|
| 1665 |
+
"grad_norm": 1.1644436120986938,
|
| 1666 |
+
"learning_rate": 5.8385331787642956e-06,
|
| 1667 |
+
"loss": 0.4654,
|
| 1668 |
+
"step": 237
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 1.9508196721311475,
|
| 1672 |
+
"grad_norm": 1.1831374168395996,
|
| 1673 |
+
"learning_rate": 5.75822838167488e-06,
|
| 1674 |
+
"loss": 0.4529,
|
| 1675 |
+
"step": 238
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 1.959016393442623,
|
| 1679 |
+
"grad_norm": 1.3696836233139038,
|
| 1680 |
+
"learning_rate": 5.678255775298542e-06,
|
| 1681 |
+
"loss": 0.5144,
|
| 1682 |
+
"step": 239
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 1.9672131147540983,
|
| 1686 |
+
"grad_norm": 1.1835501194000244,
|
| 1687 |
+
"learning_rate": 5.598621622621489e-06,
|
| 1688 |
+
"loss": 0.4885,
|
| 1689 |
+
"step": 240
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 1.9754098360655736,
|
| 1693 |
+
"grad_norm": 1.3461436033248901,
|
| 1694 |
+
"learning_rate": 5.519332160124215e-06,
|
| 1695 |
+
"loss": 0.4925,
|
| 1696 |
+
"step": 241
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 1.9836065573770492,
|
| 1700 |
+
"grad_norm": 1.1397427320480347,
|
| 1701 |
+
"learning_rate": 5.440393597293103e-06,
|
| 1702 |
+
"loss": 0.4976,
|
| 1703 |
+
"step": 242
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 1.9918032786885247,
|
| 1707 |
+
"grad_norm": 1.2887569665908813,
|
| 1708 |
+
"learning_rate": 5.361812116134122e-06,
|
| 1709 |
+
"loss": 0.4561,
|
| 1710 |
+
"step": 243
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 2.0,
|
| 1714 |
+
"grad_norm": 1.1181073188781738,
|
| 1715 |
+
"learning_rate": 5.283593870688697e-06,
|
| 1716 |
+
"loss": 0.4845,
|
| 1717 |
+
"step": 244
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 2.0081967213114753,
|
| 1721 |
+
"grad_norm": 1.6684272289276123,
|
| 1722 |
+
"learning_rate": 5.205744986551763e-06,
|
| 1723 |
+
"loss": 0.2702,
|
| 1724 |
+
"step": 245
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 2.0163934426229506,
|
| 1728 |
+
"grad_norm": 1.4320805072784424,
|
| 1729 |
+
"learning_rate": 5.128271560392037e-06,
|
| 1730 |
+
"loss": 0.2446,
|
| 1731 |
+
"step": 246
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 2.0245901639344264,
|
| 1735 |
+
"grad_norm": 1.3470230102539062,
|
| 1736 |
+
"learning_rate": 5.051179659474568e-06,
|
| 1737 |
+
"loss": 0.2465,
|
| 1738 |
+
"step": 247
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 2.0327868852459017,
|
| 1742 |
+
"grad_norm": 1.2290575504302979,
|
| 1743 |
+
"learning_rate": 4.974475321185578e-06,
|
| 1744 |
+
"loss": 0.2473,
|
| 1745 |
+
"step": 248
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 2.040983606557377,
|
| 1749 |
+
"grad_norm": 1.1487990617752075,
|
| 1750 |
+
"learning_rate": 4.89816455255966e-06,
|
| 1751 |
+
"loss": 0.2277,
|
| 1752 |
+
"step": 249
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 2.0491803278688523,
|
| 1756 |
+
"grad_norm": 1.0775490999221802,
|
| 1757 |
+
"learning_rate": 4.82225332980933e-06,
|
| 1758 |
+
"loss": 0.2504,
|
| 1759 |
+
"step": 250
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 2.057377049180328,
|
| 1763 |
+
"grad_norm": 1.0720856189727783,
|
| 1764 |
+
"learning_rate": 4.746747597857014e-06,
|
| 1765 |
+
"loss": 0.2316,
|
| 1766 |
+
"step": 251
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 2.0655737704918034,
|
| 1770 |
+
"grad_norm": 1.1619036197662354,
|
| 1771 |
+
"learning_rate": 4.671653269869474e-06,
|
| 1772 |
+
"loss": 0.2151,
|
| 1773 |
+
"step": 252
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 2.0737704918032787,
|
| 1777 |
+
"grad_norm": 1.5073219537734985,
|
| 1778 |
+
"learning_rate": 4.596976226794718e-06,
|
| 1779 |
+
"loss": 0.2421,
|
| 1780 |
+
"step": 253
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 2.081967213114754,
|
| 1784 |
+
"grad_norm": 1.6046669483184814,
|
| 1785 |
+
"learning_rate": 4.5227223169014455e-06,
|
| 1786 |
+
"loss": 0.2066,
|
| 1787 |
+
"step": 254
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 2.0901639344262297,
|
| 1791 |
+
"grad_norm": 1.9477217197418213,
|
| 1792 |
+
"learning_rate": 4.448897355321049e-06,
|
| 1793 |
+
"loss": 0.2563,
|
| 1794 |
+
"step": 255
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 2.098360655737705,
|
| 1798 |
+
"grad_norm": 2.05069899559021,
|
| 1799 |
+
"learning_rate": 4.375507123592194e-06,
|
| 1800 |
+
"loss": 0.2497,
|
| 1801 |
+
"step": 256
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 2.1065573770491803,
|
| 1805 |
+
"grad_norm": 1.7480195760726929,
|
| 1806 |
+
"learning_rate": 4.3025573692080516e-06,
|
| 1807 |
+
"loss": 0.2553,
|
| 1808 |
+
"step": 257
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 2.1147540983606556,
|
| 1812 |
+
"grad_norm": 1.5176650285720825,
|
| 1813 |
+
"learning_rate": 4.230053805166186e-06,
|
| 1814 |
+
"loss": 0.2454,
|
| 1815 |
+
"step": 258
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 2.122950819672131,
|
| 1819 |
+
"grad_norm": 1.147923231124878,
|
| 1820 |
+
"learning_rate": 4.158002109521149e-06,
|
| 1821 |
+
"loss": 0.1963,
|
| 1822 |
+
"step": 259
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 2.1311475409836067,
|
| 1826 |
+
"grad_norm": 1.1350828409194946,
|
| 1827 |
+
"learning_rate": 4.086407924939803e-06,
|
| 1828 |
+
"loss": 0.2134,
|
| 1829 |
+
"step": 260
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 2.139344262295082,
|
| 1833 |
+
"grad_norm": 1.0722763538360596,
|
| 1834 |
+
"learning_rate": 4.015276858259427e-06,
|
| 1835 |
+
"loss": 0.205,
|
| 1836 |
+
"step": 261
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 2.1475409836065573,
|
| 1840 |
+
"grad_norm": 1.2143354415893555,
|
| 1841 |
+
"learning_rate": 3.944614480048614e-06,
|
| 1842 |
+
"loss": 0.2233,
|
| 1843 |
+
"step": 262
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 2.1557377049180326,
|
| 1847 |
+
"grad_norm": 1.3119699954986572,
|
| 1848 |
+
"learning_rate": 3.874426324171019e-06,
|
| 1849 |
+
"loss": 0.2435,
|
| 1850 |
+
"step": 263
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 2.1639344262295084,
|
| 1854 |
+
"grad_norm": 1.1167805194854736,
|
| 1855 |
+
"learning_rate": 3.804717887351991e-06,
|
| 1856 |
+
"loss": 0.2274,
|
| 1857 |
+
"step": 264
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 2.1721311475409837,
|
| 1861 |
+
"grad_norm": 1.3277482986450195,
|
| 1862 |
+
"learning_rate": 3.735494628748082e-06,
|
| 1863 |
+
"loss": 0.2541,
|
| 1864 |
+
"step": 265
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 2.180327868852459,
|
| 1868 |
+
"grad_norm": 1.1086663007736206,
|
| 1869 |
+
"learning_rate": 3.6667619695195287e-06,
|
| 1870 |
+
"loss": 0.2388,
|
| 1871 |
+
"step": 266
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 2.1885245901639343,
|
| 1875 |
+
"grad_norm": 1.0080981254577637,
|
| 1876 |
+
"learning_rate": 3.5985252924057023e-06,
|
| 1877 |
+
"loss": 0.217,
|
| 1878 |
+
"step": 267
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 2.19672131147541,
|
| 1882 |
+
"grad_norm": 1.1589125394821167,
|
| 1883 |
+
"learning_rate": 3.530789941303554e-06,
|
| 1884 |
+
"loss": 0.2195,
|
| 1885 |
+
"step": 268
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 2.2049180327868854,
|
| 1889 |
+
"grad_norm": 1.045792579650879,
|
| 1890 |
+
"learning_rate": 3.4635612208491197e-06,
|
| 1891 |
+
"loss": 0.2067,
|
| 1892 |
+
"step": 269
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 2.2131147540983607,
|
| 1896 |
+
"grad_norm": 1.1592333316802979,
|
| 1897 |
+
"learning_rate": 3.3968443960020913e-06,
|
| 1898 |
+
"loss": 0.2214,
|
| 1899 |
+
"step": 270
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 2.221311475409836,
|
| 1903 |
+
"grad_norm": 1.150566816329956,
|
| 1904 |
+
"learning_rate": 3.330644691633492e-06,
|
| 1905 |
+
"loss": 0.2088,
|
| 1906 |
+
"step": 271
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 2.2295081967213113,
|
| 1910 |
+
"grad_norm": 1.2285228967666626,
|
| 1911 |
+
"learning_rate": 3.2649672921165e-06,
|
| 1912 |
+
"loss": 0.2092,
|
| 1913 |
+
"step": 272
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 2.237704918032787,
|
| 1917 |
+
"grad_norm": 1.13582444190979,
|
| 1918 |
+
"learning_rate": 3.1998173409204326e-06,
|
| 1919 |
+
"loss": 0.2168,
|
| 1920 |
+
"step": 273
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 2.2459016393442623,
|
| 1924 |
+
"grad_norm": 1.0956363677978516,
|
| 1925 |
+
"learning_rate": 3.135199940207947e-06,
|
| 1926 |
+
"loss": 0.2089,
|
| 1927 |
+
"step": 274
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 2.2540983606557377,
|
| 1931 |
+
"grad_norm": 1.0761518478393555,
|
| 1932 |
+
"learning_rate": 3.0711201504354628e-06,
|
| 1933 |
+
"loss": 0.2024,
|
| 1934 |
+
"step": 275
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 2.262295081967213,
|
| 1938 |
+
"grad_norm": 1.1069421768188477,
|
| 1939 |
+
"learning_rate": 3.00758298995686e-06,
|
| 1940 |
+
"loss": 0.2154,
|
| 1941 |
+
"step": 276
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 2.2704918032786887,
|
| 1945 |
+
"grad_norm": 1.2374476194381714,
|
| 1946 |
+
"learning_rate": 2.9445934346304706e-06,
|
| 1947 |
+
"loss": 0.244,
|
| 1948 |
+
"step": 277
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 2.278688524590164,
|
| 1952 |
+
"grad_norm": 1.233173131942749,
|
| 1953 |
+
"learning_rate": 2.882156417429396e-06,
|
| 1954 |
+
"loss": 0.2281,
|
| 1955 |
+
"step": 278
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 2.2868852459016393,
|
| 1959 |
+
"grad_norm": 1.3890464305877686,
|
| 1960 |
+
"learning_rate": 2.8202768280551894e-06,
|
| 1961 |
+
"loss": 0.2489,
|
| 1962 |
+
"step": 279
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 2.2950819672131146,
|
| 1966 |
+
"grad_norm": 1.2962629795074463,
|
| 1967 |
+
"learning_rate": 2.7589595125549195e-06,
|
| 1968 |
+
"loss": 0.2104,
|
| 1969 |
+
"step": 280
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 2.30327868852459,
|
| 1973 |
+
"grad_norm": 1.2199065685272217,
|
| 1974 |
+
"learning_rate": 2.698209272941659e-06,
|
| 1975 |
+
"loss": 0.2021,
|
| 1976 |
+
"step": 281
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 2.3114754098360657,
|
| 1980 |
+
"grad_norm": 1.3473303318023682,
|
| 1981 |
+
"learning_rate": 2.6380308668184163e-06,
|
| 1982 |
+
"loss": 0.2463,
|
| 1983 |
+
"step": 282
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 2.319672131147541,
|
| 1987 |
+
"grad_norm": 1.0375776290893555,
|
| 1988 |
+
"learning_rate": 2.578429007005552e-06,
|
| 1989 |
+
"loss": 0.194,
|
| 1990 |
+
"step": 283
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 2.3278688524590163,
|
| 1994 |
+
"grad_norm": 1.1980507373809814,
|
| 1995 |
+
"learning_rate": 2.5194083611716937e-06,
|
| 1996 |
+
"loss": 0.2184,
|
| 1997 |
+
"step": 284
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 2.3360655737704916,
|
| 2001 |
+
"grad_norm": 1.1433260440826416,
|
| 2002 |
+
"learning_rate": 2.460973551468194e-06,
|
| 2003 |
+
"loss": 0.2181,
|
| 2004 |
+
"step": 285
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 2.3442622950819674,
|
| 2008 |
+
"grad_norm": 1.1053398847579956,
|
| 2009 |
+
"learning_rate": 2.403129154167153e-06,
|
| 2010 |
+
"loss": 0.2299,
|
| 2011 |
+
"step": 286
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 2.3524590163934427,
|
| 2015 |
+
"grad_norm": 1.1965283155441284,
|
| 2016 |
+
"learning_rate": 2.345879699303025e-06,
|
| 2017 |
+
"loss": 0.2059,
|
| 2018 |
+
"step": 287
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 2.360655737704918,
|
| 2022 |
+
"grad_norm": 1.2773933410644531,
|
| 2023 |
+
"learning_rate": 2.2892296703178596e-06,
|
| 2024 |
+
"loss": 0.2019,
|
| 2025 |
+
"step": 288
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 2.3688524590163933,
|
| 2029 |
+
"grad_norm": 1.1140257120132446,
|
| 2030 |
+
"learning_rate": 2.2331835037101825e-06,
|
| 2031 |
+
"loss": 0.2052,
|
| 2032 |
+
"step": 289
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 2.3770491803278686,
|
| 2036 |
+
"grad_norm": 1.525154709815979,
|
| 2037 |
+
"learning_rate": 2.17774558868755e-06,
|
| 2038 |
+
"loss": 0.2639,
|
| 2039 |
+
"step": 290
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 2.3852459016393444,
|
| 2043 |
+
"grad_norm": 1.1313905715942383,
|
| 2044 |
+
"learning_rate": 2.1229202668228197e-06,
|
| 2045 |
+
"loss": 0.2248,
|
| 2046 |
+
"step": 291
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 2.3934426229508197,
|
| 2050 |
+
"grad_norm": 1.3205841779708862,
|
| 2051 |
+
"learning_rate": 2.0687118317141407e-06,
|
| 2052 |
+
"loss": 0.23,
|
| 2053 |
+
"step": 292
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 2.401639344262295,
|
| 2057 |
+
"grad_norm": 1.3589439392089844,
|
| 2058 |
+
"learning_rate": 2.0151245286486998e-06,
|
| 2059 |
+
"loss": 0.2478,
|
| 2060 |
+
"step": 293
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 2.4098360655737707,
|
| 2064 |
+
"grad_norm": 1.1922129392623901,
|
| 2065 |
+
"learning_rate": 1.962162554270267e-06,
|
| 2066 |
+
"loss": 0.2189,
|
| 2067 |
+
"step": 294
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 2.418032786885246,
|
| 2071 |
+
"grad_norm": 1.0274560451507568,
|
| 2072 |
+
"learning_rate": 1.9098300562505266e-06,
|
| 2073 |
+
"loss": 0.2127,
|
| 2074 |
+
"step": 295
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 2.4262295081967213,
|
| 2078 |
+
"grad_norm": 1.2022790908813477,
|
| 2079 |
+
"learning_rate": 1.8581311329642592e-06,
|
| 2080 |
+
"loss": 0.2096,
|
| 2081 |
+
"step": 296
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 2.4344262295081966,
|
| 2085 |
+
"grad_norm": 1.1015734672546387,
|
| 2086 |
+
"learning_rate": 1.8070698331683844e-06,
|
| 2087 |
+
"loss": 0.2181,
|
| 2088 |
+
"step": 297
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 2.442622950819672,
|
| 2092 |
+
"grad_norm": 1.0971348285675049,
|
| 2093 |
+
"learning_rate": 1.7566501556848858e-06,
|
| 2094 |
+
"loss": 0.2032,
|
| 2095 |
+
"step": 298
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 2.4508196721311477,
|
| 2099 |
+
"grad_norm": 1.1589995622634888,
|
| 2100 |
+
"learning_rate": 1.7068760490876425e-06,
|
| 2101 |
+
"loss": 0.2324,
|
| 2102 |
+
"step": 299
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 2.459016393442623,
|
| 2106 |
+
"grad_norm": 1.3047738075256348,
|
| 2107 |
+
"learning_rate": 1.6577514113932036e-06,
|
| 2108 |
+
"loss": 0.2301,
|
| 2109 |
+
"step": 300
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 2.4672131147540983,
|
| 2113 |
+
"grad_norm": 1.2343322038650513,
|
| 2114 |
+
"learning_rate": 1.609280089755515e-06,
|
| 2115 |
+
"loss": 0.226,
|
| 2116 |
+
"step": 301
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 2.4754098360655736,
|
| 2120 |
+
"grad_norm": 1.0935797691345215,
|
| 2121 |
+
"learning_rate": 1.5614658801646354e-06,
|
| 2122 |
+
"loss": 0.2311,
|
| 2123 |
+
"step": 302
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 2.4836065573770494,
|
| 2127 |
+
"grad_norm": 1.0060436725616455,
|
| 2128 |
+
"learning_rate": 1.5143125271494607e-06,
|
| 2129 |
+
"loss": 0.1992,
|
| 2130 |
+
"step": 303
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 2.4918032786885247,
|
| 2134 |
+
"grad_norm": 1.211691975593567,
|
| 2135 |
+
"learning_rate": 1.467823723484465e-06,
|
| 2136 |
+
"loss": 0.2132,
|
| 2137 |
+
"step": 304
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 2.5,
|
| 2141 |
+
"grad_norm": 1.2444989681243896,
|
| 2142 |
+
"learning_rate": 1.4220031099005094e-06,
|
| 2143 |
+
"loss": 0.2181,
|
| 2144 |
+
"step": 305
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 2.5081967213114753,
|
| 2148 |
+
"grad_norm": 1.5254817008972168,
|
| 2149 |
+
"learning_rate": 1.3768542747997215e-06,
|
| 2150 |
+
"loss": 0.2447,
|
| 2151 |
+
"step": 306
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 2.5163934426229506,
|
| 2155 |
+
"grad_norm": 1.20688796043396,
|
| 2156 |
+
"learning_rate": 1.3323807539744726e-06,
|
| 2157 |
+
"loss": 0.2145,
|
| 2158 |
+
"step": 307
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 2.5245901639344264,
|
| 2162 |
+
"grad_norm": 1.1355769634246826,
|
| 2163 |
+
"learning_rate": 1.2885860303304743e-06,
|
| 2164 |
+
"loss": 0.2014,
|
| 2165 |
+
"step": 308
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 2.5327868852459017,
|
| 2169 |
+
"grad_norm": 1.2493382692337036,
|
| 2170 |
+
"learning_rate": 1.2454735336140167e-06,
|
| 2171 |
+
"loss": 0.2142,
|
| 2172 |
+
"step": 309
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 2.540983606557377,
|
| 2176 |
+
"grad_norm": 1.1296652555465698,
|
| 2177 |
+
"learning_rate": 1.2030466401433749e-06,
|
| 2178 |
+
"loss": 0.2081,
|
| 2179 |
+
"step": 310
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 2.5491803278688527,
|
| 2183 |
+
"grad_norm": 1.142025351524353,
|
| 2184 |
+
"learning_rate": 1.161308672544389e-06,
|
| 2185 |
+
"loss": 0.2097,
|
| 2186 |
+
"step": 311
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 2.557377049180328,
|
| 2190 |
+
"grad_norm": 1.2121732234954834,
|
| 2191 |
+
"learning_rate": 1.1202628994902632e-06,
|
| 2192 |
+
"loss": 0.2026,
|
| 2193 |
+
"step": 312
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 2.5655737704918034,
|
| 2197 |
+
"grad_norm": 1.3817976713180542,
|
| 2198 |
+
"learning_rate": 1.0799125354455752e-06,
|
| 2199 |
+
"loss": 0.2207,
|
| 2200 |
+
"step": 313
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 2.5737704918032787,
|
| 2204 |
+
"grad_norm": 1.2908271551132202,
|
| 2205 |
+
"learning_rate": 1.0402607404145448e-06,
|
| 2206 |
+
"loss": 0.2152,
|
| 2207 |
+
"step": 314
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 2.581967213114754,
|
| 2211 |
+
"grad_norm": 1.213399052619934,
|
| 2212 |
+
"learning_rate": 1.0013106196935528e-06,
|
| 2213 |
+
"loss": 0.2277,
|
| 2214 |
+
"step": 315
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 2.5901639344262293,
|
| 2218 |
+
"grad_norm": 1.2260936498641968,
|
| 2219 |
+
"learning_rate": 9.630652236279626e-07,
|
| 2220 |
+
"loss": 0.2333,
|
| 2221 |
+
"step": 316
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 2.598360655737705,
|
| 2225 |
+
"grad_norm": 1.0845645666122437,
|
| 2226 |
+
"learning_rate": 9.255275473732239e-07,
|
| 2227 |
+
"loss": 0.2012,
|
| 2228 |
+
"step": 317
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 2.6065573770491803,
|
| 2232 |
+
"grad_norm": 1.1773573160171509,
|
| 2233 |
+
"learning_rate": 8.887005306603203e-07,
|
| 2234 |
+
"loss": 0.2017,
|
| 2235 |
+
"step": 318
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 2.6147540983606556,
|
| 2239 |
+
"grad_norm": 1.1452627182006836,
|
| 2240 |
+
"learning_rate": 8.525870575655393e-07,
|
| 2241 |
+
"loss": 0.2146,
|
| 2242 |
+
"step": 319
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 2.6229508196721314,
|
| 2246 |
+
"grad_norm": 1.1558417081832886,
|
| 2247 |
+
"learning_rate": 8.171899562846097e-07,
|
| 2248 |
+
"loss": 0.23,
|
| 2249 |
+
"step": 320
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 2.6311475409836067,
|
| 2253 |
+
"grad_norm": 1.1180776357650757,
|
| 2254 |
+
"learning_rate": 7.825119989112173e-07,
|
| 2255 |
+
"loss": 0.1994,
|
| 2256 |
+
"step": 321
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 2.639344262295082,
|
| 2260 |
+
"grad_norm": 1.2736459970474243,
|
| 2261 |
+
"learning_rate": 7.485559012199062e-07,
|
| 2262 |
+
"loss": 0.2086,
|
| 2263 |
+
"step": 322
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 2.6475409836065573,
|
| 2267 |
+
"grad_norm": 1.2114237546920776,
|
| 2268 |
+
"learning_rate": 7.153243224534001e-07,
|
| 2269 |
+
"loss": 0.2089,
|
| 2270 |
+
"step": 323
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 2.6557377049180326,
|
| 2274 |
+
"grad_norm": 1.1722540855407715,
|
| 2275 |
+
"learning_rate": 6.828198651143425e-07,
|
| 2276 |
+
"loss": 0.2133,
|
| 2277 |
+
"step": 324
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 2.663934426229508,
|
| 2281 |
+
"grad_norm": 1.128095030784607,
|
| 2282 |
+
"learning_rate": 6.510450747614816e-07,
|
| 2283 |
+
"loss": 0.2055,
|
| 2284 |
+
"step": 325
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 2.6721311475409837,
|
| 2288 |
+
"grad_norm": 1.1521621942520142,
|
| 2289 |
+
"learning_rate": 6.200024398103255e-07,
|
| 2290 |
+
"loss": 0.2165,
|
| 2291 |
+
"step": 326
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 2.680327868852459,
|
| 2295 |
+
"grad_norm": 1.1144236326217651,
|
| 2296 |
+
"learning_rate": 5.896943913382547e-07,
|
| 2297 |
+
"loss": 0.2158,
|
| 2298 |
+
"step": 327
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 2.6885245901639343,
|
| 2302 |
+
"grad_norm": 1.1580475568771362,
|
| 2303 |
+
"learning_rate": 5.601233028941389e-07,
|
| 2304 |
+
"loss": 0.2071,
|
| 2305 |
+
"step": 328
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 2.69672131147541,
|
| 2309 |
+
"grad_norm": 1.1493444442749023,
|
| 2310 |
+
"learning_rate": 5.312914903124566e-07,
|
| 2311 |
+
"loss": 0.2307,
|
| 2312 |
+
"step": 329
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 2.7049180327868854,
|
| 2316 |
+
"grad_norm": 1.261407494544983,
|
| 2317 |
+
"learning_rate": 5.032012115319273e-07,
|
| 2318 |
+
"loss": 0.2233,
|
| 2319 |
+
"step": 330
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 2.7131147540983607,
|
| 2323 |
+
"grad_norm": 1.1146408319473267,
|
| 2324 |
+
"learning_rate": 4.7585466641868696e-07,
|
| 2325 |
+
"loss": 0.2104,
|
| 2326 |
+
"step": 331
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 2.721311475409836,
|
| 2330 |
+
"grad_norm": 1.152559518814087,
|
| 2331 |
+
"learning_rate": 4.492539965940057e-07,
|
| 2332 |
+
"loss": 0.1936,
|
| 2333 |
+
"step": 332
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 2.7295081967213113,
|
| 2337 |
+
"grad_norm": 1.1715896129608154,
|
| 2338 |
+
"learning_rate": 4.234012852665703e-07,
|
| 2339 |
+
"loss": 0.2255,
|
| 2340 |
+
"step": 333
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 2.737704918032787,
|
| 2344 |
+
"grad_norm": 1.2842741012573242,
|
| 2345 |
+
"learning_rate": 3.982985570693354e-07,
|
| 2346 |
+
"loss": 0.2126,
|
| 2347 |
+
"step": 334
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 2.7459016393442623,
|
| 2351 |
+
"grad_norm": 1.290299892425537,
|
| 2352 |
+
"learning_rate": 3.739477779009704e-07,
|
| 2353 |
+
"loss": 0.2468,
|
| 2354 |
+
"step": 335
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 2.7540983606557377,
|
| 2358 |
+
"grad_norm": 1.1366935968399048,
|
| 2359 |
+
"learning_rate": 3.5035085477190143e-07,
|
| 2360 |
+
"loss": 0.2083,
|
| 2361 |
+
"step": 336
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 2.762295081967213,
|
| 2365 |
+
"grad_norm": 1.2218340635299683,
|
| 2366 |
+
"learning_rate": 3.275096356549612e-07,
|
| 2367 |
+
"loss": 0.2514,
|
| 2368 |
+
"step": 337
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 2.7704918032786887,
|
| 2372 |
+
"grad_norm": 1.1977328062057495,
|
| 2373 |
+
"learning_rate": 3.054259093406731e-07,
|
| 2374 |
+
"loss": 0.1986,
|
| 2375 |
+
"step": 338
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 2.778688524590164,
|
| 2379 |
+
"grad_norm": 1.2030324935913086,
|
| 2380 |
+
"learning_rate": 2.8410140529715803e-07,
|
| 2381 |
+
"loss": 0.2547,
|
| 2382 |
+
"step": 339
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 2.7868852459016393,
|
| 2386 |
+
"grad_norm": 1.1166476011276245,
|
| 2387 |
+
"learning_rate": 2.6353779353469387e-07,
|
| 2388 |
+
"loss": 0.2035,
|
| 2389 |
+
"step": 340
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 2.7950819672131146,
|
| 2393 |
+
"grad_norm": 1.3586655855178833,
|
| 2394 |
+
"learning_rate": 2.4373668447493225e-07,
|
| 2395 |
+
"loss": 0.2395,
|
| 2396 |
+
"step": 341
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 2.80327868852459,
|
| 2400 |
+
"grad_norm": 1.1721775531768799,
|
| 2401 |
+
"learning_rate": 2.2469962882478046e-07,
|
| 2402 |
+
"loss": 0.2127,
|
| 2403 |
+
"step": 342
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 2.8114754098360657,
|
| 2407 |
+
"grad_norm": 1.3731749057769775,
|
| 2408 |
+
"learning_rate": 2.0642811745495206e-07,
|
| 2409 |
+
"loss": 0.2245,
|
| 2410 |
+
"step": 343
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 2.819672131147541,
|
| 2414 |
+
"grad_norm": 1.2181392908096313,
|
| 2415 |
+
"learning_rate": 1.889235812832202e-07,
|
| 2416 |
+
"loss": 0.2171,
|
| 2417 |
+
"step": 344
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 2.8278688524590163,
|
| 2421 |
+
"grad_norm": 1.1868208646774292,
|
| 2422 |
+
"learning_rate": 1.7218739116235061e-07,
|
| 2423 |
+
"loss": 0.2165,
|
| 2424 |
+
"step": 345
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 2.836065573770492,
|
| 2428 |
+
"grad_norm": 1.1789835691452026,
|
| 2429 |
+
"learning_rate": 1.562208577727442e-07,
|
| 2430 |
+
"loss": 0.2226,
|
| 2431 |
+
"step": 346
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 2.8442622950819674,
|
| 2435 |
+
"grad_norm": 1.1302640438079834,
|
| 2436 |
+
"learning_rate": 1.4102523151979574e-07,
|
| 2437 |
+
"loss": 0.2172,
|
| 2438 |
+
"step": 347
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 2.8524590163934427,
|
| 2442 |
+
"grad_norm": 1.1117348670959473,
|
| 2443 |
+
"learning_rate": 1.266017024359656e-07,
|
| 2444 |
+
"loss": 0.216,
|
| 2445 |
+
"step": 348
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 2.860655737704918,
|
| 2449 |
+
"grad_norm": 1.1936352252960205,
|
| 2450 |
+
"learning_rate": 1.1295140008758864e-07,
|
| 2451 |
+
"loss": 0.2325,
|
| 2452 |
+
"step": 349
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 2.8688524590163933,
|
| 2456 |
+
"grad_norm": 1.349346399307251,
|
| 2457 |
+
"learning_rate": 1.0007539348640738e-07,
|
| 2458 |
+
"loss": 0.2522,
|
| 2459 |
+
"step": 350
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 2.8770491803278686,
|
| 2463 |
+
"grad_norm": 1.176551103591919,
|
| 2464 |
+
"learning_rate": 8.797469100585432e-08,
|
| 2465 |
+
"loss": 0.2404,
|
| 2466 |
+
"step": 351
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 2.8852459016393444,
|
| 2470 |
+
"grad_norm": 1.3001667261123657,
|
| 2471 |
+
"learning_rate": 7.665024030208635e-08,
|
| 2472 |
+
"loss": 0.2435,
|
| 2473 |
+
"step": 352
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 2.8934426229508197,
|
| 2477 |
+
"grad_norm": 1.1511950492858887,
|
| 2478 |
+
"learning_rate": 6.610292823976628e-08,
|
| 2479 |
+
"loss": 0.2327,
|
| 2480 |
+
"step": 353
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 2.901639344262295,
|
| 2484 |
+
"grad_norm": 1.1168369054794312,
|
| 2485 |
+
"learning_rate": 5.633358082260953e-08,
|
| 2486 |
+
"loss": 0.195,
|
| 2487 |
+
"step": 354
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 2.9098360655737707,
|
| 2491 |
+
"grad_norm": 1.0813003778457642,
|
| 2492 |
+
"learning_rate": 4.73429631286948e-08,
|
| 2493 |
+
"loss": 0.209,
|
| 2494 |
+
"step": 355
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 2.918032786885246,
|
| 2498 |
+
"grad_norm": 1.225760579109192,
|
| 2499 |
+
"learning_rate": 3.913177925055189e-08,
|
| 2500 |
+
"loss": 0.2013,
|
| 2501 |
+
"step": 356
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 2.9262295081967213,
|
| 2505 |
+
"grad_norm": 1.1697818040847778,
|
| 2506 |
+
"learning_rate": 3.170067224001483e-08,
|
| 2507 |
+
"loss": 0.2148,
|
| 2508 |
+
"step": 357
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 2.9344262295081966,
|
| 2512 |
+
"grad_norm": 1.2408915758132935,
|
| 2513 |
+
"learning_rate": 2.5050224057868723e-08,
|
| 2514 |
+
"loss": 0.2302,
|
| 2515 |
+
"step": 358
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 2.942622950819672,
|
| 2519 |
+
"grad_norm": 1.202064037322998,
|
| 2520 |
+
"learning_rate": 1.9180955528270706e-08,
|
| 2521 |
+
"loss": 0.2068,
|
| 2522 |
+
"step": 359
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 2.9508196721311473,
|
| 2526 |
+
"grad_norm": 1.1742192506790161,
|
| 2527 |
+
"learning_rate": 1.4093326297959232e-08,
|
| 2528 |
+
"loss": 0.1977,
|
| 2529 |
+
"step": 360
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 2.959016393442623,
|
| 2533 |
+
"grad_norm": 1.194492220878601,
|
| 2534 |
+
"learning_rate": 9.78773480026396e-09,
|
| 2535 |
+
"loss": 0.2333,
|
| 2536 |
+
"step": 361
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 2.9672131147540983,
|
| 2540 |
+
"grad_norm": 1.1205930709838867,
|
| 2541 |
+
"learning_rate": 6.264518223896288e-09,
|
| 2542 |
+
"loss": 0.2191,
|
| 2543 |
+
"step": 362
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 2.9754098360655736,
|
| 2547 |
+
"grad_norm": 1.2019057273864746,
|
| 2548 |
+
"learning_rate": 3.5239524865460227e-09,
|
| 2549 |
+
"loss": 0.2074,
|
| 2550 |
+
"step": 363
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 2.9836065573770494,
|
| 2554 |
+
"grad_norm": 1.1968348026275635,
|
| 2555 |
+
"learning_rate": 1.566252213274222e-09,
|
| 2556 |
+
"loss": 0.2316,
|
| 2557 |
+
"step": 364
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 2.9918032786885247,
|
| 2561 |
+
"grad_norm": 1.0804020166397095,
|
| 2562 |
+
"learning_rate": 3.9157071969997407e-10,
|
| 2563 |
+
"loss": 0.2412,
|
| 2564 |
+
"step": 365
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 3.0,
|
| 2568 |
+
"grad_norm": 1.0318905115127563,
|
| 2569 |
+
"learning_rate": 0.0,
|
| 2570 |
+
"loss": 0.1958,
|
| 2571 |
+
"step": 366
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 3.0,
|
| 2575 |
+
"step": 366,
|
| 2576 |
+
"total_flos": 9.143633223446692e+17,
|
| 2577 |
+
"train_loss": 0.5501216588531687,
|
| 2578 |
+
"train_runtime": 2923.287,
|
| 2579 |
+
"train_samples_per_second": 16.01,
|
| 2580 |
+
"train_steps_per_second": 0.125
|
| 2581 |
+
}
|
| 2582 |
+
],
|
| 2583 |
+
"logging_steps": 1,
|
| 2584 |
+
"max_steps": 366,
|
| 2585 |
+
"num_input_tokens_seen": 0,
|
| 2586 |
+
"num_train_epochs": 3,
|
| 2587 |
+
"save_steps": 0,
|
| 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": 9.143633223446692e+17,
|
| 2601 |
+
"train_batch_size": 32,
|
| 2602 |
+
"trial_name": null,
|
| 2603 |
+
"trial_params": null
|
| 2604 |
+
}
|