Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +2833 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,2833 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.0639897616381379,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 400,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.00015997440409534473,
|
| 13 |
+
"grad_norm": 131.17303466796875,
|
| 14 |
+
"learning_rate": 0.0,
|
| 15 |
+
"loss": 2.02,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.00031994880819068947,
|
| 20 |
+
"grad_norm": 158.60177612304688,
|
| 21 |
+
"learning_rate": 1.3236987738623674e-07,
|
| 22 |
+
"loss": 1.8427,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.00047992321228603423,
|
| 27 |
+
"grad_norm": 159.34176635742188,
|
| 28 |
+
"learning_rate": 2.0980129188224264e-07,
|
| 29 |
+
"loss": 1.8302,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.0006398976163813789,
|
| 34 |
+
"grad_norm": 160.90457153320312,
|
| 35 |
+
"learning_rate": 2.647397547724735e-07,
|
| 36 |
+
"loss": 1.8208,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.0007998720204767237,
|
| 41 |
+
"grad_norm": 163.9342803955078,
|
| 42 |
+
"learning_rate": 3.073533372198984e-07,
|
| 43 |
+
"loss": 1.8017,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.0009598464245720685,
|
| 48 |
+
"grad_norm": 165.28424072265625,
|
| 49 |
+
"learning_rate": 3.421711692684794e-07,
|
| 50 |
+
"loss": 1.7632,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.0011198208286674132,
|
| 55 |
+
"grad_norm": 154.68109130859375,
|
| 56 |
+
"learning_rate": 3.7160922681241326e-07,
|
| 57 |
+
"loss": 1.6268,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.0012797952327627579,
|
| 62 |
+
"grad_norm": 91.99791717529297,
|
| 63 |
+
"learning_rate": 3.9710963215871023e-07,
|
| 64 |
+
"loss": 1.3067,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.0014397696368581027,
|
| 69 |
+
"grad_norm": 91.15142059326172,
|
| 70 |
+
"learning_rate": 4.1960258376448527e-07,
|
| 71 |
+
"loss": 1.2644,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.0015997440409534474,
|
| 76 |
+
"grad_norm": 95.00140380859375,
|
| 77 |
+
"learning_rate": 4.397232146061353e-07,
|
| 78 |
+
"loss": 1.2029,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.0017597184450487923,
|
| 83 |
+
"grad_norm": 100.01094055175781,
|
| 84 |
+
"learning_rate": 4.5792453918508963e-07,
|
| 85 |
+
"loss": 1.1511,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.001919692849144137,
|
| 90 |
+
"grad_norm": 75.83676147460938,
|
| 91 |
+
"learning_rate": 4.745410466547161e-07,
|
| 92 |
+
"loss": 0.7515,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.002079667253239482,
|
| 97 |
+
"grad_norm": 38.47466278076172,
|
| 98 |
+
"learning_rate": 4.898267517654968e-07,
|
| 99 |
+
"loss": 0.5764,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.0022396416573348264,
|
| 104 |
+
"grad_norm": 35.91501998901367,
|
| 105 |
+
"learning_rate": 5.0397910419865e-07,
|
| 106 |
+
"loss": 0.5374,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.002399616061430171,
|
| 111 |
+
"grad_norm": 35.598350524902344,
|
| 112 |
+
"learning_rate": 5.17154629102141e-07,
|
| 113 |
+
"loss": 0.5144,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.0025595904655255157,
|
| 118 |
+
"grad_norm": 38.198551177978516,
|
| 119 |
+
"learning_rate": 5.29479509544947e-07,
|
| 120 |
+
"loss": 0.4709,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.002719564869620861,
|
| 125 |
+
"grad_norm": 48.83864974975586,
|
| 126 |
+
"learning_rate": 5.410569551171063e-07,
|
| 127 |
+
"loss": 0.4102,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.0028795392737162055,
|
| 132 |
+
"grad_norm": 50.42375946044922,
|
| 133 |
+
"learning_rate": 5.519724611507219e-07,
|
| 134 |
+
"loss": 0.3637,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.00303951367781155,
|
| 139 |
+
"grad_norm": 26.880878448486328,
|
| 140 |
+
"learning_rate": 5.622976441001489e-07,
|
| 141 |
+
"loss": 0.325,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.003199488081906895,
|
| 146 |
+
"grad_norm": 50.65513229370117,
|
| 147 |
+
"learning_rate": 5.720930919923719e-07,
|
| 148 |
+
"loss": 0.2883,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.0033594624860022394,
|
| 153 |
+
"grad_norm": 12.885412216186523,
|
| 154 |
+
"learning_rate": 5.814105186946559e-07,
|
| 155 |
+
"loss": 0.237,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.0035194368900975845,
|
| 160 |
+
"grad_norm": 10.100977897644043,
|
| 161 |
+
"learning_rate": 5.902944165713264e-07,
|
| 162 |
+
"loss": 0.2293,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.003679411294192929,
|
| 167 |
+
"grad_norm": 7.237063407897949,
|
| 168 |
+
"learning_rate": 5.98783341472312e-07,
|
| 169 |
+
"loss": 0.2131,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.003839385698288274,
|
| 174 |
+
"grad_norm": 5.179779529571533,
|
| 175 |
+
"learning_rate": 6.069109240409529e-07,
|
| 176 |
+
"loss": 0.2051,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.0039993601023836185,
|
| 181 |
+
"grad_norm": 3.619863271713257,
|
| 182 |
+
"learning_rate": 6.147066744397969e-07,
|
| 183 |
+
"loss": 0.1905,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.004159334506478964,
|
| 188 |
+
"grad_norm": 3.3004238605499268,
|
| 189 |
+
"learning_rate": 6.221966291517337e-07,
|
| 190 |
+
"loss": 0.1855,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.004319308910574308,
|
| 195 |
+
"grad_norm": 2.556411027908325,
|
| 196 |
+
"learning_rate": 6.294038756467279e-07,
|
| 197 |
+
"loss": 0.179,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.004479283314669653,
|
| 202 |
+
"grad_norm": 3.1660890579223633,
|
| 203 |
+
"learning_rate": 6.363489815848868e-07,
|
| 204 |
+
"loss": 0.1762,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.004639257718764998,
|
| 209 |
+
"grad_norm": 2.2603609561920166,
|
| 210 |
+
"learning_rate": 6.430503486697051e-07,
|
| 211 |
+
"loss": 0.1743,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.004799232122860342,
|
| 216 |
+
"grad_norm": 3.1162281036376953,
|
| 217 |
+
"learning_rate": 6.495245064883779e-07,
|
| 218 |
+
"loss": 0.1727,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.004959206526955687,
|
| 223 |
+
"grad_norm": 3.651181697845459,
|
| 224 |
+
"learning_rate": 6.557863581532573e-07,
|
| 225 |
+
"loss": 0.1671,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.0051191809310510315,
|
| 230 |
+
"grad_norm": 2.72200345993042,
|
| 231 |
+
"learning_rate": 6.618493869311837e-07,
|
| 232 |
+
"loss": 0.1652,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.005279155335146377,
|
| 237 |
+
"grad_norm": 2.1462044715881348,
|
| 238 |
+
"learning_rate": 6.677258310673321e-07,
|
| 239 |
+
"loss": 0.1698,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.005439129739241722,
|
| 244 |
+
"grad_norm": 2.3578836917877197,
|
| 245 |
+
"learning_rate": 6.734268325033431e-07,
|
| 246 |
+
"loss": 0.1678,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.005599104143337066,
|
| 251 |
+
"grad_norm": 3.18251633644104,
|
| 252 |
+
"learning_rate": 6.789625640323117e-07,
|
| 253 |
+
"loss": 0.1666,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.005759078547432411,
|
| 258 |
+
"grad_norm": 2.3386144638061523,
|
| 259 |
+
"learning_rate": 6.843423385369588e-07,
|
| 260 |
+
"loss": 0.1606,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.005919052951527755,
|
| 265 |
+
"grad_norm": 3.2586636543273926,
|
| 266 |
+
"learning_rate": 6.895747032576225e-07,
|
| 267 |
+
"loss": 0.1585,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.0060790273556231,
|
| 272 |
+
"grad_norm": 3.786763906478882,
|
| 273 |
+
"learning_rate": 6.946675214863857e-07,
|
| 274 |
+
"loss": 0.1624,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.006239001759718445,
|
| 279 |
+
"grad_norm": 2.901278495788574,
|
| 280 |
+
"learning_rate": 6.996280436477395e-07,
|
| 281 |
+
"loss": 0.16,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.00639897616381379,
|
| 286 |
+
"grad_norm": 1.7482857704162598,
|
| 287 |
+
"learning_rate": 7.044629693786087e-07,
|
| 288 |
+
"loss": 0.1596,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.006558950567909135,
|
| 293 |
+
"grad_norm": 2.393767833709717,
|
| 294 |
+
"learning_rate": 7.091785019416827e-07,
|
| 295 |
+
"loss": 0.1558,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.006718924972004479,
|
| 300 |
+
"grad_norm": 3.0040743350982666,
|
| 301 |
+
"learning_rate": 7.137803960808927e-07,
|
| 302 |
+
"loss": 0.154,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.006878899376099824,
|
| 307 |
+
"grad_norm": 2.8855950832366943,
|
| 308 |
+
"learning_rate": 7.182740002451747e-07,
|
| 309 |
+
"loss": 0.1568,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.007038873780195169,
|
| 314 |
+
"grad_norm": 2.0385258197784424,
|
| 315 |
+
"learning_rate": 7.22664293957563e-07,
|
| 316 |
+
"loss": 0.1579,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.007198848184290513,
|
| 321 |
+
"grad_norm": 1.4387277364730835,
|
| 322 |
+
"learning_rate": 7.269559209843836e-07,
|
| 323 |
+
"loss": 0.1487,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.007358822588385858,
|
| 328 |
+
"grad_norm": 2.1492249965667725,
|
| 329 |
+
"learning_rate": 7.311532188585489e-07,
|
| 330 |
+
"loss": 0.1538,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.007518796992481203,
|
| 335 |
+
"grad_norm": 1.820937991142273,
|
| 336 |
+
"learning_rate": 7.352602452275264e-07,
|
| 337 |
+
"loss": 0.1495,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.007678771396576548,
|
| 342 |
+
"grad_norm": 1.9386242628097534,
|
| 343 |
+
"learning_rate": 7.392808014271896e-07,
|
| 344 |
+
"loss": 0.1515,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.007838745800671893,
|
| 349 |
+
"grad_norm": 2.2448389530181885,
|
| 350 |
+
"learning_rate": 7.432184536248265e-07,
|
| 351 |
+
"loss": 0.1516,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.007998720204767237,
|
| 356 |
+
"grad_norm": 1.6324615478515625,
|
| 357 |
+
"learning_rate": 7.470765518260336e-07,
|
| 358 |
+
"loss": 0.1494,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.008158694608862581,
|
| 363 |
+
"grad_norm": 1.9566981792449951,
|
| 364 |
+
"learning_rate": 7.50858246999349e-07,
|
| 365 |
+
"loss": 0.1499,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.008318669012957927,
|
| 370 |
+
"grad_norm": 1.8566670417785645,
|
| 371 |
+
"learning_rate": 7.545665065379704e-07,
|
| 372 |
+
"loss": 0.147,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.008478643417053271,
|
| 377 |
+
"grad_norm": 1.4080177545547485,
|
| 378 |
+
"learning_rate": 7.582041282486483e-07,
|
| 379 |
+
"loss": 0.1484,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.008638617821148616,
|
| 384 |
+
"grad_norm": 2.2273147106170654,
|
| 385 |
+
"learning_rate": 7.617737530329646e-07,
|
| 386 |
+
"loss": 0.1505,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.008798592225243962,
|
| 391 |
+
"grad_norm": 1.559259057044983,
|
| 392 |
+
"learning_rate": 7.652778764049881e-07,
|
| 393 |
+
"loss": 0.1469,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.008958566629339306,
|
| 398 |
+
"grad_norm": 1.5281285047531128,
|
| 399 |
+
"learning_rate": 7.687188589711236e-07,
|
| 400 |
+
"loss": 0.1479,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.00911854103343465,
|
| 405 |
+
"grad_norm": 1.222518801689148,
|
| 406 |
+
"learning_rate": 7.720989359823916e-07,
|
| 407 |
+
"loss": 0.1431,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.009278515437529996,
|
| 412 |
+
"grad_norm": 1.6629137992858887,
|
| 413 |
+
"learning_rate": 7.754202260559418e-07,
|
| 414 |
+
"loss": 0.1463,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.00943848984162534,
|
| 419 |
+
"grad_norm": 1.7817844152450562,
|
| 420 |
+
"learning_rate": 7.786847391510249e-07,
|
| 421 |
+
"loss": 0.1461,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.009598464245720684,
|
| 426 |
+
"grad_norm": 1.7894457578659058,
|
| 427 |
+
"learning_rate": 7.818943838746145e-07,
|
| 428 |
+
"loss": 0.1475,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.009758438649816029,
|
| 433 |
+
"grad_norm": 1.6237293481826782,
|
| 434 |
+
"learning_rate": 7.850509741831755e-07,
|
| 435 |
+
"loss": 0.1441,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.009918413053911375,
|
| 440 |
+
"grad_norm": 1.5423122644424438,
|
| 441 |
+
"learning_rate": 7.881562355394939e-07,
|
| 442 |
+
"loss": 0.1448,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.010078387458006719,
|
| 447 |
+
"grad_norm": 1.381689429283142,
|
| 448 |
+
"learning_rate": 7.912118105768986e-07,
|
| 449 |
+
"loss": 0.1415,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.010238361862102063,
|
| 454 |
+
"grad_norm": 1.8088704347610474,
|
| 455 |
+
"learning_rate": 7.942192643174205e-07,
|
| 456 |
+
"loss": 0.1426,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.010398336266197409,
|
| 461 |
+
"grad_norm": 2.4634368419647217,
|
| 462 |
+
"learning_rate": 7.971800889853952e-07,
|
| 463 |
+
"loss": 0.1459,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.010558310670292753,
|
| 468 |
+
"grad_norm": 1.7363865375518799,
|
| 469 |
+
"learning_rate": 8.000957084535689e-07,
|
| 470 |
+
"loss": 0.1438,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.010718285074388097,
|
| 475 |
+
"grad_norm": 1.5139596462249756,
|
| 476 |
+
"learning_rate": 8.029674823548714e-07,
|
| 477 |
+
"loss": 0.1411,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.010878259478483443,
|
| 482 |
+
"grad_norm": 1.4979151487350464,
|
| 483 |
+
"learning_rate": 8.057967098895799e-07,
|
| 484 |
+
"loss": 0.142,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.011038233882578788,
|
| 489 |
+
"grad_norm": 1.7654547691345215,
|
| 490 |
+
"learning_rate": 8.085846333545548e-07,
|
| 491 |
+
"loss": 0.1424,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.011198208286674132,
|
| 496 |
+
"grad_norm": 1.9629215002059937,
|
| 497 |
+
"learning_rate": 8.113324414185485e-07,
|
| 498 |
+
"loss": 0.1458,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.011358182690769478,
|
| 503 |
+
"grad_norm": 1.9196921586990356,
|
| 504 |
+
"learning_rate": 8.140412721651974e-07,
|
| 505 |
+
"loss": 0.1465,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.011518157094864822,
|
| 510 |
+
"grad_norm": 2.098177909851074,
|
| 511 |
+
"learning_rate": 8.167122159231955e-07,
|
| 512 |
+
"loss": 0.1439,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.011678131498960166,
|
| 517 |
+
"grad_norm": 1.7505509853363037,
|
| 518 |
+
"learning_rate": 8.193463179012649e-07,
|
| 519 |
+
"loss": 0.1421,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.01183810590305551,
|
| 524 |
+
"grad_norm": 1.7455815076828003,
|
| 525 |
+
"learning_rate": 8.219445806438593e-07,
|
| 526 |
+
"loss": 0.1382,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.011998080307150856,
|
| 531 |
+
"grad_norm": 1.5361164808273315,
|
| 532 |
+
"learning_rate": 8.245079663220395e-07,
|
| 533 |
+
"loss": 0.142,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.0121580547112462,
|
| 538 |
+
"grad_norm": 2.7622509002685547,
|
| 539 |
+
"learning_rate": 8.270373988726225e-07,
|
| 540 |
+
"loss": 0.1474,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.012318029115341545,
|
| 545 |
+
"grad_norm": 2.590261936187744,
|
| 546 |
+
"learning_rate": 8.295337659975029e-07,
|
| 547 |
+
"loss": 0.1442,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.01247800351943689,
|
| 552 |
+
"grad_norm": 1.9166169166564941,
|
| 553 |
+
"learning_rate": 8.319979210339763e-07,
|
| 554 |
+
"loss": 0.1401,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.012637977923532235,
|
| 559 |
+
"grad_norm": 1.3086128234863281,
|
| 560 |
+
"learning_rate": 8.344306847059229e-07,
|
| 561 |
+
"loss": 0.1392,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.01279795232762758,
|
| 566 |
+
"grad_norm": 1.7761530876159668,
|
| 567 |
+
"learning_rate": 8.368328467648454e-07,
|
| 568 |
+
"loss": 0.1389,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.012957926731722925,
|
| 573 |
+
"grad_norm": 1.8375579118728638,
|
| 574 |
+
"learning_rate": 8.392051675289705e-07,
|
| 575 |
+
"loss": 0.1399,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.01311790113581827,
|
| 580 |
+
"grad_norm": 1.7359107732772827,
|
| 581 |
+
"learning_rate": 8.415483793279194e-07,
|
| 582 |
+
"loss": 0.1411,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.013277875539913614,
|
| 587 |
+
"grad_norm": 1.9930511713027954,
|
| 588 |
+
"learning_rate": 8.43863187859817e-07,
|
| 589 |
+
"loss": 0.135,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.013437849944008958,
|
| 594 |
+
"grad_norm": 1.8547563552856445,
|
| 595 |
+
"learning_rate": 8.461502734671294e-07,
|
| 596 |
+
"loss": 0.1365,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.013597824348104304,
|
| 601 |
+
"grad_norm": 1.7807646989822388,
|
| 602 |
+
"learning_rate": 8.484102923370048e-07,
|
| 603 |
+
"loss": 0.1377,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.013757798752199648,
|
| 608 |
+
"grad_norm": 2.3836963176727295,
|
| 609 |
+
"learning_rate": 8.506438776314115e-07,
|
| 610 |
+
"loss": 0.1392,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.013917773156294992,
|
| 615 |
+
"grad_norm": 1.9490735530853271,
|
| 616 |
+
"learning_rate": 8.528516405519477e-07,
|
| 617 |
+
"loss": 0.1352,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.014077747560390338,
|
| 622 |
+
"grad_norm": 1.6703108549118042,
|
| 623 |
+
"learning_rate": 8.550341713437998e-07,
|
| 624 |
+
"loss": 0.1404,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.014237721964485682,
|
| 629 |
+
"grad_norm": 1.2594908475875854,
|
| 630 |
+
"learning_rate": 8.571920402429763e-07,
|
| 631 |
+
"loss": 0.1343,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.014397696368581027,
|
| 636 |
+
"grad_norm": 1.4328629970550537,
|
| 637 |
+
"learning_rate": 8.593257983706205e-07,
|
| 638 |
+
"loss": 0.1399,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.014557670772676373,
|
| 643 |
+
"grad_norm": 1.3369605541229248,
|
| 644 |
+
"learning_rate": 8.6143597857791e-07,
|
| 645 |
+
"loss": 0.1411,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.014717645176771717,
|
| 650 |
+
"grad_norm": 1.4118566513061523,
|
| 651 |
+
"learning_rate": 8.635230962447856e-07,
|
| 652 |
+
"loss": 0.1378,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.014877619580867061,
|
| 657 |
+
"grad_norm": 1.9336967468261719,
|
| 658 |
+
"learning_rate": 8.655876500354998e-07,
|
| 659 |
+
"loss": 0.1367,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.015037593984962405,
|
| 664 |
+
"grad_norm": 1.9851653575897217,
|
| 665 |
+
"learning_rate": 8.676301226137632e-07,
|
| 666 |
+
"loss": 0.1398,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.015197568389057751,
|
| 671 |
+
"grad_norm": 1.2664474248886108,
|
| 672 |
+
"learning_rate": 8.696509813200474e-07,
|
| 673 |
+
"loss": 0.1358,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.015357542793153095,
|
| 678 |
+
"grad_norm": 2.4485650062561035,
|
| 679 |
+
"learning_rate": 8.716506788134264e-07,
|
| 680 |
+
"loss": 0.1409,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.01551751719724844,
|
| 685 |
+
"grad_norm": 1.413927435874939,
|
| 686 |
+
"learning_rate": 8.736296536801594e-07,
|
| 687 |
+
"loss": 0.1342,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.015677491601343786,
|
| 692 |
+
"grad_norm": 2.10976505279541,
|
| 693 |
+
"learning_rate": 8.755883310110635e-07,
|
| 694 |
+
"loss": 0.1428,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.01583746600543913,
|
| 699 |
+
"grad_norm": 1.454032063484192,
|
| 700 |
+
"learning_rate": 8.775271229495747e-07,
|
| 701 |
+
"loss": 0.1377,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.015997440409534474,
|
| 706 |
+
"grad_norm": 1.451311707496643,
|
| 707 |
+
"learning_rate": 8.794464292122706e-07,
|
| 708 |
+
"loss": 0.1366,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.016157414813629818,
|
| 713 |
+
"grad_norm": 1.2541290521621704,
|
| 714 |
+
"learning_rate": 8.813466375834887e-07,
|
| 715 |
+
"loss": 0.1371,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.016317389217725162,
|
| 720 |
+
"grad_norm": 1.826319694519043,
|
| 721 |
+
"learning_rate": 8.832281243855856e-07,
|
| 722 |
+
"loss": 0.1363,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.01647736362182051,
|
| 727 |
+
"grad_norm": 1.3739150762557983,
|
| 728 |
+
"learning_rate": 8.8509125492625e-07,
|
| 729 |
+
"loss": 0.1389,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.016637338025915854,
|
| 734 |
+
"grad_norm": 1.5730900764465332,
|
| 735 |
+
"learning_rate": 8.869363839242072e-07,
|
| 736 |
+
"loss": 0.1363,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.0167973124300112,
|
| 741 |
+
"grad_norm": 1.4488335847854614,
|
| 742 |
+
"learning_rate": 8.887638559145543e-07,
|
| 743 |
+
"loss": 0.1331,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.016957286834106543,
|
| 748 |
+
"grad_norm": 1.3076649904251099,
|
| 749 |
+
"learning_rate": 8.905740056348848e-07,
|
| 750 |
+
"loss": 0.1328,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.017117261238201887,
|
| 755 |
+
"grad_norm": 1.6239938735961914,
|
| 756 |
+
"learning_rate": 8.923671583932827e-07,
|
| 757 |
+
"loss": 0.1339,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.01727723564229723,
|
| 762 |
+
"grad_norm": 1.7959277629852295,
|
| 763 |
+
"learning_rate": 8.941436304192014e-07,
|
| 764 |
+
"loss": 0.1353,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.01743721004639258,
|
| 769 |
+
"grad_norm": 1.9266117811203003,
|
| 770 |
+
"learning_rate": 8.959037291981712e-07,
|
| 771 |
+
"loss": 0.1377,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.017597184450487923,
|
| 776 |
+
"grad_norm": 2.10133695602417,
|
| 777 |
+
"learning_rate": 8.976477537912248e-07,
|
| 778 |
+
"loss": 0.1359,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.017757158854583267,
|
| 783 |
+
"grad_norm": 1.285016417503357,
|
| 784 |
+
"learning_rate": 8.99375995139865e-07,
|
| 785 |
+
"loss": 0.134,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.01791713325867861,
|
| 790 |
+
"grad_norm": 1.2501912117004395,
|
| 791 |
+
"learning_rate": 9.010887363573603e-07,
|
| 792 |
+
"loss": 0.1353,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.018077107662773956,
|
| 797 |
+
"grad_norm": 1.2144287824630737,
|
| 798 |
+
"learning_rate": 9.027862530070898e-07,
|
| 799 |
+
"loss": 0.132,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.0182370820668693,
|
| 804 |
+
"grad_norm": 1.5820159912109375,
|
| 805 |
+
"learning_rate": 9.044688133686283e-07,
|
| 806 |
+
"loss": 0.1327,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.018397056470964644,
|
| 811 |
+
"grad_norm": 1.3742053508758545,
|
| 812 |
+
"learning_rate": 9.061366786922105e-07,
|
| 813 |
+
"loss": 0.1358,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.018557030875059992,
|
| 818 |
+
"grad_norm": 1.3718464374542236,
|
| 819 |
+
"learning_rate": 9.077901034421785e-07,
|
| 820 |
+
"loss": 0.136,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.018717005279155336,
|
| 825 |
+
"grad_norm": 1.246771216392517,
|
| 826 |
+
"learning_rate": 9.094293355299821e-07,
|
| 827 |
+
"loss": 0.1336,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.01887697968325068,
|
| 832 |
+
"grad_norm": 1.442410945892334,
|
| 833 |
+
"learning_rate": 9.110546165372615e-07,
|
| 834 |
+
"loss": 0.1335,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.019036954087346025,
|
| 839 |
+
"grad_norm": 1.5307565927505493,
|
| 840 |
+
"learning_rate": 9.126661819295197e-07,
|
| 841 |
+
"loss": 0.1338,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.01919692849144137,
|
| 846 |
+
"grad_norm": 1.809870958328247,
|
| 847 |
+
"learning_rate": 9.142642612608513e-07,
|
| 848 |
+
"loss": 0.1316,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.019356902895536713,
|
| 853 |
+
"grad_norm": 1.5888453722000122,
|
| 854 |
+
"learning_rate": 9.158490783701793e-07,
|
| 855 |
+
"loss": 0.1339,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.019516877299632057,
|
| 860 |
+
"grad_norm": 1.5183619260787964,
|
| 861 |
+
"learning_rate": 9.174208515694123e-07,
|
| 862 |
+
"loss": 0.1324,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.019676851703727405,
|
| 867 |
+
"grad_norm": 1.18596351146698,
|
| 868 |
+
"learning_rate": 9.189797938239252e-07,
|
| 869 |
+
"loss": 0.1335,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.01983682610782275,
|
| 874 |
+
"grad_norm": 0.9591464996337891,
|
| 875 |
+
"learning_rate": 9.205261129257306e-07,
|
| 876 |
+
"loss": 0.1284,
|
| 877 |
+
"step": 124
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.019996800511918093,
|
| 881 |
+
"grad_norm": 1.1490577459335327,
|
| 882 |
+
"learning_rate": 9.220600116596953e-07,
|
| 883 |
+
"loss": 0.1337,
|
| 884 |
+
"step": 125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.020156774916013438,
|
| 888 |
+
"grad_norm": 1.3413796424865723,
|
| 889 |
+
"learning_rate": 9.235816879631353e-07,
|
| 890 |
+
"loss": 0.1318,
|
| 891 |
+
"step": 126
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.020316749320108782,
|
| 895 |
+
"grad_norm": 1.8173480033874512,
|
| 896 |
+
"learning_rate": 9.250913350791019e-07,
|
| 897 |
+
"loss": 0.1348,
|
| 898 |
+
"step": 127
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.020476723724204126,
|
| 902 |
+
"grad_norm": 1.4662258625030518,
|
| 903 |
+
"learning_rate": 9.265891417036572e-07,
|
| 904 |
+
"loss": 0.1333,
|
| 905 |
+
"step": 128
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.020636698128299474,
|
| 909 |
+
"grad_norm": 1.722621202468872,
|
| 910 |
+
"learning_rate": 9.280752921274173e-07,
|
| 911 |
+
"loss": 0.1337,
|
| 912 |
+
"step": 129
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.020796672532394818,
|
| 916 |
+
"grad_norm": 2.0989277362823486,
|
| 917 |
+
"learning_rate": 9.295499663716319e-07,
|
| 918 |
+
"loss": 0.1347,
|
| 919 |
+
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.020956646936490162,
|
| 923 |
+
"grad_norm": 1.9588260650634766,
|
| 924 |
+
"learning_rate": 9.310133403190494e-07,
|
| 925 |
+
"loss": 0.1346,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.021116621340585506,
|
| 930 |
+
"grad_norm": 1.5125082731246948,
|
| 931 |
+
"learning_rate": 9.324655858398056e-07,
|
| 932 |
+
"loss": 0.1376,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.02127659574468085,
|
| 937 |
+
"grad_norm": 2.068347454071045,
|
| 938 |
+
"learning_rate": 9.339068709125623e-07,
|
| 939 |
+
"loss": 0.1358,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.021436570148776195,
|
| 944 |
+
"grad_norm": 1.3576561212539673,
|
| 945 |
+
"learning_rate": 9.353373597411082e-07,
|
| 946 |
+
"loss": 0.1341,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.02159654455287154,
|
| 951 |
+
"grad_norm": 1.4498059749603271,
|
| 952 |
+
"learning_rate": 9.367572128666264e-07,
|
| 953 |
+
"loss": 0.1307,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.021756518956966887,
|
| 958 |
+
"grad_norm": 1.832219123840332,
|
| 959 |
+
"learning_rate": 9.381665872758167e-07,
|
| 960 |
+
"loss": 0.1322,
|
| 961 |
+
"step": 136
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.02191649336106223,
|
| 965 |
+
"grad_norm": 1.6702324151992798,
|
| 966 |
+
"learning_rate": 9.395656365050596e-07,
|
| 967 |
+
"loss": 0.1341,
|
| 968 |
+
"step": 137
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.022076467765157575,
|
| 972 |
+
"grad_norm": 2.3712918758392334,
|
| 973 |
+
"learning_rate": 9.409545107407915e-07,
|
| 974 |
+
"loss": 0.132,
|
| 975 |
+
"step": 138
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.02223644216925292,
|
| 979 |
+
"grad_norm": 2.47230863571167,
|
| 980 |
+
"learning_rate": 9.423333569162554e-07,
|
| 981 |
+
"loss": 0.1384,
|
| 982 |
+
"step": 139
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.022396416573348264,
|
| 986 |
+
"grad_norm": 1.5399249792099,
|
| 987 |
+
"learning_rate": 9.437023188047852e-07,
|
| 988 |
+
"loss": 0.1293,
|
| 989 |
+
"step": 140
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.022556390977443608,
|
| 993 |
+
"grad_norm": 2.937544822692871,
|
| 994 |
+
"learning_rate": 9.450615371097691e-07,
|
| 995 |
+
"loss": 0.1364,
|
| 996 |
+
"step": 141
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.022716365381538955,
|
| 1000 |
+
"grad_norm": 2.1480307579040527,
|
| 1001 |
+
"learning_rate": 9.464111495514342e-07,
|
| 1002 |
+
"loss": 0.1317,
|
| 1003 |
+
"step": 142
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.0228763397856343,
|
| 1007 |
+
"grad_norm": 1.650770902633667,
|
| 1008 |
+
"learning_rate": 9.477512909505864e-07,
|
| 1009 |
+
"loss": 0.1328,
|
| 1010 |
+
"step": 143
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.023036314189729644,
|
| 1014 |
+
"grad_norm": 1.477109432220459,
|
| 1015 |
+
"learning_rate": 9.490820933094322e-07,
|
| 1016 |
+
"loss": 0.1319,
|
| 1017 |
+
"step": 144
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.023196288593824988,
|
| 1021 |
+
"grad_norm": 2.036853551864624,
|
| 1022 |
+
"learning_rate": 9.504036858896035e-07,
|
| 1023 |
+
"loss": 0.1326,
|
| 1024 |
+
"step": 145
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.023356262997920332,
|
| 1028 |
+
"grad_norm": 1.4756255149841309,
|
| 1029 |
+
"learning_rate": 9.517161952875015e-07,
|
| 1030 |
+
"loss": 0.1324,
|
| 1031 |
+
"step": 146
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.023516237402015677,
|
| 1035 |
+
"grad_norm": 1.1842366456985474,
|
| 1036 |
+
"learning_rate": 9.53019745507069e-07,
|
| 1037 |
+
"loss": 0.1293,
|
| 1038 |
+
"step": 147
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.02367621180611102,
|
| 1042 |
+
"grad_norm": 1.311472773551941,
|
| 1043 |
+
"learning_rate": 9.54314458030096e-07,
|
| 1044 |
+
"loss": 0.1276,
|
| 1045 |
+
"step": 148
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.02383618621020637,
|
| 1049 |
+
"grad_norm": 1.8847229480743408,
|
| 1050 |
+
"learning_rate": 9.556004518841566e-07,
|
| 1051 |
+
"loss": 0.1314,
|
| 1052 |
+
"step": 149
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.023996160614301713,
|
| 1056 |
+
"grad_norm": 1.67835533618927,
|
| 1057 |
+
"learning_rate": 9.568778437082761e-07,
|
| 1058 |
+
"loss": 0.1321,
|
| 1059 |
+
"step": 150
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.024156135018397057,
|
| 1063 |
+
"grad_norm": 1.356938123703003,
|
| 1064 |
+
"learning_rate": 9.581467478164157e-07,
|
| 1065 |
+
"loss": 0.1309,
|
| 1066 |
+
"step": 151
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.0243161094224924,
|
| 1070 |
+
"grad_norm": 1.6580119132995605,
|
| 1071 |
+
"learning_rate": 9.594072762588592e-07,
|
| 1072 |
+
"loss": 0.1316,
|
| 1073 |
+
"step": 152
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.024476083826587745,
|
| 1077 |
+
"grad_norm": 1.3679087162017822,
|
| 1078 |
+
"learning_rate": 9.606595388815916e-07,
|
| 1079 |
+
"loss": 0.131,
|
| 1080 |
+
"step": 153
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.02463605823068309,
|
| 1084 |
+
"grad_norm": 1.2950249910354614,
|
| 1085 |
+
"learning_rate": 9.619036433837396e-07,
|
| 1086 |
+
"loss": 0.1297,
|
| 1087 |
+
"step": 154
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.024796032634778434,
|
| 1091 |
+
"grad_norm": 1.3981151580810547,
|
| 1092 |
+
"learning_rate": 9.631396953731557e-07,
|
| 1093 |
+
"loss": 0.1288,
|
| 1094 |
+
"step": 155
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.02495600703887378,
|
| 1098 |
+
"grad_norm": 1.2822041511535645,
|
| 1099 |
+
"learning_rate": 9.64367798420213e-07,
|
| 1100 |
+
"loss": 0.1307,
|
| 1101 |
+
"step": 156
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.025115981442969126,
|
| 1105 |
+
"grad_norm": 1.3140027523040771,
|
| 1106 |
+
"learning_rate": 9.65588054109883e-07,
|
| 1107 |
+
"loss": 0.1331,
|
| 1108 |
+
"step": 157
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.02527595584706447,
|
| 1112 |
+
"grad_norm": 1.517618179321289,
|
| 1113 |
+
"learning_rate": 9.668005620921596e-07,
|
| 1114 |
+
"loss": 0.1325,
|
| 1115 |
+
"step": 158
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.025435930251159814,
|
| 1119 |
+
"grad_norm": 1.2211989164352417,
|
| 1120 |
+
"learning_rate": 9.680054201308908e-07,
|
| 1121 |
+
"loss": 0.1314,
|
| 1122 |
+
"step": 159
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.02559590465525516,
|
| 1126 |
+
"grad_norm": 1.1967936754226685,
|
| 1127 |
+
"learning_rate": 9.692027241510822e-07,
|
| 1128 |
+
"loss": 0.129,
|
| 1129 |
+
"step": 160
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.025755879059350503,
|
| 1133 |
+
"grad_norm": 1.2219634056091309,
|
| 1134 |
+
"learning_rate": 9.703925682847255e-07,
|
| 1135 |
+
"loss": 0.131,
|
| 1136 |
+
"step": 161
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.02591585346344585,
|
| 1140 |
+
"grad_norm": 1.4189143180847168,
|
| 1141 |
+
"learning_rate": 9.71575044915207e-07,
|
| 1142 |
+
"loss": 0.1287,
|
| 1143 |
+
"step": 162
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.026075827867541194,
|
| 1147 |
+
"grad_norm": 1.3790985345840454,
|
| 1148 |
+
"learning_rate": 9.727502447203536e-07,
|
| 1149 |
+
"loss": 0.1301,
|
| 1150 |
+
"step": 163
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.02623580227163654,
|
| 1154 |
+
"grad_norm": 1.5076289176940918,
|
| 1155 |
+
"learning_rate": 9.739182567141562e-07,
|
| 1156 |
+
"loss": 0.1314,
|
| 1157 |
+
"step": 164
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.026395776675731883,
|
| 1161 |
+
"grad_norm": 1.3452013731002808,
|
| 1162 |
+
"learning_rate": 9.750791682872306e-07,
|
| 1163 |
+
"loss": 0.131,
|
| 1164 |
+
"step": 165
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.026555751079827227,
|
| 1168 |
+
"grad_norm": 1.466188669204712,
|
| 1169 |
+
"learning_rate": 9.762330652460538e-07,
|
| 1170 |
+
"loss": 0.1303,
|
| 1171 |
+
"step": 166
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.02671572548392257,
|
| 1175 |
+
"grad_norm": 1.9016494750976562,
|
| 1176 |
+
"learning_rate": 9.773800318510204e-07,
|
| 1177 |
+
"loss": 0.1311,
|
| 1178 |
+
"step": 167
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.026875699888017916,
|
| 1182 |
+
"grad_norm": 1.223925232887268,
|
| 1183 |
+
"learning_rate": 9.78520150853366e-07,
|
| 1184 |
+
"loss": 0.1295,
|
| 1185 |
+
"step": 168
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.027035674292113263,
|
| 1189 |
+
"grad_norm": 1.5380191802978516,
|
| 1190 |
+
"learning_rate": 9.796535035309936e-07,
|
| 1191 |
+
"loss": 0.1278,
|
| 1192 |
+
"step": 169
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 0.027195648696208607,
|
| 1196 |
+
"grad_norm": 1.4442285299301147,
|
| 1197 |
+
"learning_rate": 9.807801697232416e-07,
|
| 1198 |
+
"loss": 0.1324,
|
| 1199 |
+
"step": 170
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.02735562310030395,
|
| 1203 |
+
"grad_norm": 1.7630947828292847,
|
| 1204 |
+
"learning_rate": 9.819002278646342e-07,
|
| 1205 |
+
"loss": 0.1322,
|
| 1206 |
+
"step": 171
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 0.027515597504399296,
|
| 1210 |
+
"grad_norm": 1.2535282373428345,
|
| 1211 |
+
"learning_rate": 9.830137550176481e-07,
|
| 1212 |
+
"loss": 0.1315,
|
| 1213 |
+
"step": 172
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 0.02767557190849464,
|
| 1217 |
+
"grad_norm": 1.4018826484680176,
|
| 1218 |
+
"learning_rate": 9.841208269045278e-07,
|
| 1219 |
+
"loss": 0.1332,
|
| 1220 |
+
"step": 173
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 0.027835546312589984,
|
| 1224 |
+
"grad_norm": 1.1194539070129395,
|
| 1225 |
+
"learning_rate": 9.852215179381846e-07,
|
| 1226 |
+
"loss": 0.1304,
|
| 1227 |
+
"step": 174
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 0.027995520716685332,
|
| 1231 |
+
"grad_norm": 1.3209128379821777,
|
| 1232 |
+
"learning_rate": 9.863159012522102e-07,
|
| 1233 |
+
"loss": 0.1318,
|
| 1234 |
+
"step": 175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 0.028155495120780676,
|
| 1238 |
+
"grad_norm": 1.637170672416687,
|
| 1239 |
+
"learning_rate": 9.874040487300364e-07,
|
| 1240 |
+
"loss": 0.128,
|
| 1241 |
+
"step": 176
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.02831546952487602,
|
| 1245 |
+
"grad_norm": 2.5420658588409424,
|
| 1246 |
+
"learning_rate": 9.884860310332674e-07,
|
| 1247 |
+
"loss": 0.1307,
|
| 1248 |
+
"step": 177
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.028475443928971365,
|
| 1252 |
+
"grad_norm": 1.7741079330444336,
|
| 1253 |
+
"learning_rate": 9.89561917629213e-07,
|
| 1254 |
+
"loss": 0.1309,
|
| 1255 |
+
"step": 178
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.02863541833306671,
|
| 1259 |
+
"grad_norm": 1.5465468168258667,
|
| 1260 |
+
"learning_rate": 9.906317768176538e-07,
|
| 1261 |
+
"loss": 0.1295,
|
| 1262 |
+
"step": 179
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.028795392737162053,
|
| 1266 |
+
"grad_norm": 1.5184584856033325,
|
| 1267 |
+
"learning_rate": 9.916956757568572e-07,
|
| 1268 |
+
"loss": 0.1302,
|
| 1269 |
+
"step": 180
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.028955367141257397,
|
| 1273 |
+
"grad_norm": 1.4520975351333618,
|
| 1274 |
+
"learning_rate": 9.92753680488876e-07,
|
| 1275 |
+
"loss": 0.1278,
|
| 1276 |
+
"step": 181
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 0.029115341545352745,
|
| 1280 |
+
"grad_norm": 1.275576114654541,
|
| 1281 |
+
"learning_rate": 9.93805855964147e-07,
|
| 1282 |
+
"loss": 0.1286,
|
| 1283 |
+
"step": 182
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.02927531594944809,
|
| 1287 |
+
"grad_norm": 1.4705110788345337,
|
| 1288 |
+
"learning_rate": 9.948522660654181e-07,
|
| 1289 |
+
"loss": 0.1278,
|
| 1290 |
+
"step": 183
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 0.029435290353543433,
|
| 1294 |
+
"grad_norm": 1.7193312644958496,
|
| 1295 |
+
"learning_rate": 9.958929736310225e-07,
|
| 1296 |
+
"loss": 0.1297,
|
| 1297 |
+
"step": 184
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 0.029595264757638778,
|
| 1301 |
+
"grad_norm": 1.483162522315979,
|
| 1302 |
+
"learning_rate": 9.969280404775208e-07,
|
| 1303 |
+
"loss": 0.1297,
|
| 1304 |
+
"step": 185
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.029755239161734122,
|
| 1308 |
+
"grad_norm": 1.191413402557373,
|
| 1309 |
+
"learning_rate": 9.979575274217366e-07,
|
| 1310 |
+
"loss": 0.1279,
|
| 1311 |
+
"step": 186
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.029915213565829466,
|
| 1315 |
+
"grad_norm": 1.3899563550949097,
|
| 1316 |
+
"learning_rate": 9.989814943021958e-07,
|
| 1317 |
+
"loss": 0.1308,
|
| 1318 |
+
"step": 187
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 0.03007518796992481,
|
| 1322 |
+
"grad_norm": 1.1342469453811646,
|
| 1323 |
+
"learning_rate": 1e-06,
|
| 1324 |
+
"loss": 0.1244,
|
| 1325 |
+
"step": 188
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 0.030235162374020158,
|
| 1329 |
+
"grad_norm": 1.0823456048965454,
|
| 1330 |
+
"learning_rate": 1e-06,
|
| 1331 |
+
"loss": 0.1296,
|
| 1332 |
+
"step": 189
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 0.030395136778115502,
|
| 1336 |
+
"grad_norm": 1.1805393695831299,
|
| 1337 |
+
"learning_rate": 1e-06,
|
| 1338 |
+
"loss": 0.1267,
|
| 1339 |
+
"step": 190
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 0.030555111182210846,
|
| 1343 |
+
"grad_norm": 1.355592131614685,
|
| 1344 |
+
"learning_rate": 1e-06,
|
| 1345 |
+
"loss": 0.1293,
|
| 1346 |
+
"step": 191
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 0.03071508558630619,
|
| 1350 |
+
"grad_norm": 1.7999428510665894,
|
| 1351 |
+
"learning_rate": 1e-06,
|
| 1352 |
+
"loss": 0.1313,
|
| 1353 |
+
"step": 192
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 0.030875059990401535,
|
| 1357 |
+
"grad_norm": 1.5406898260116577,
|
| 1358 |
+
"learning_rate": 1e-06,
|
| 1359 |
+
"loss": 0.1303,
|
| 1360 |
+
"step": 193
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 0.03103503439449688,
|
| 1364 |
+
"grad_norm": 1.6359665393829346,
|
| 1365 |
+
"learning_rate": 1e-06,
|
| 1366 |
+
"loss": 0.1297,
|
| 1367 |
+
"step": 194
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 0.031195008798592227,
|
| 1371 |
+
"grad_norm": 1.3248281478881836,
|
| 1372 |
+
"learning_rate": 1e-06,
|
| 1373 |
+
"loss": 0.1283,
|
| 1374 |
+
"step": 195
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 0.03135498320268757,
|
| 1378 |
+
"grad_norm": 1.4763351678848267,
|
| 1379 |
+
"learning_rate": 1e-06,
|
| 1380 |
+
"loss": 0.1251,
|
| 1381 |
+
"step": 196
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 0.03151495760678291,
|
| 1385 |
+
"grad_norm": 1.383988618850708,
|
| 1386 |
+
"learning_rate": 1e-06,
|
| 1387 |
+
"loss": 0.1327,
|
| 1388 |
+
"step": 197
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 0.03167493201087826,
|
| 1392 |
+
"grad_norm": 1.0943655967712402,
|
| 1393 |
+
"learning_rate": 1e-06,
|
| 1394 |
+
"loss": 0.1305,
|
| 1395 |
+
"step": 198
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 0.03183490641497361,
|
| 1399 |
+
"grad_norm": 1.2787405252456665,
|
| 1400 |
+
"learning_rate": 1e-06,
|
| 1401 |
+
"loss": 0.1273,
|
| 1402 |
+
"step": 199
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 0.03199488081906895,
|
| 1406 |
+
"grad_norm": 1.597632646560669,
|
| 1407 |
+
"learning_rate": 1e-06,
|
| 1408 |
+
"loss": 0.1261,
|
| 1409 |
+
"step": 200
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 0.032154855223164296,
|
| 1413 |
+
"grad_norm": 1.0324817895889282,
|
| 1414 |
+
"learning_rate": 1e-06,
|
| 1415 |
+
"loss": 0.1249,
|
| 1416 |
+
"step": 201
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 0.032314829627259636,
|
| 1420 |
+
"grad_norm": 1.453323245048523,
|
| 1421 |
+
"learning_rate": 1e-06,
|
| 1422 |
+
"loss": 0.1266,
|
| 1423 |
+
"step": 202
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 0.032474804031354984,
|
| 1427 |
+
"grad_norm": 1.1299526691436768,
|
| 1428 |
+
"learning_rate": 1e-06,
|
| 1429 |
+
"loss": 0.1237,
|
| 1430 |
+
"step": 203
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 0.032634778435450325,
|
| 1434 |
+
"grad_norm": 1.1185365915298462,
|
| 1435 |
+
"learning_rate": 1e-06,
|
| 1436 |
+
"loss": 0.129,
|
| 1437 |
+
"step": 204
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 0.03279475283954567,
|
| 1441 |
+
"grad_norm": 1.024652123451233,
|
| 1442 |
+
"learning_rate": 1e-06,
|
| 1443 |
+
"loss": 0.1256,
|
| 1444 |
+
"step": 205
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 0.03295472724364102,
|
| 1448 |
+
"grad_norm": 1.4425640106201172,
|
| 1449 |
+
"learning_rate": 1e-06,
|
| 1450 |
+
"loss": 0.1224,
|
| 1451 |
+
"step": 206
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 0.03311470164773636,
|
| 1455 |
+
"grad_norm": 1.5149112939834595,
|
| 1456 |
+
"learning_rate": 1e-06,
|
| 1457 |
+
"loss": 0.1275,
|
| 1458 |
+
"step": 207
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 0.03327467605183171,
|
| 1462 |
+
"grad_norm": 1.4966243505477905,
|
| 1463 |
+
"learning_rate": 1e-06,
|
| 1464 |
+
"loss": 0.1273,
|
| 1465 |
+
"step": 208
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 0.03343465045592705,
|
| 1469 |
+
"grad_norm": 1.7751713991165161,
|
| 1470 |
+
"learning_rate": 1e-06,
|
| 1471 |
+
"loss": 0.1313,
|
| 1472 |
+
"step": 209
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 0.0335946248600224,
|
| 1476 |
+
"grad_norm": 1.3089805841445923,
|
| 1477 |
+
"learning_rate": 1e-06,
|
| 1478 |
+
"loss": 0.1223,
|
| 1479 |
+
"step": 210
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 0.03375459926411774,
|
| 1483 |
+
"grad_norm": 1.116400957107544,
|
| 1484 |
+
"learning_rate": 1e-06,
|
| 1485 |
+
"loss": 0.1288,
|
| 1486 |
+
"step": 211
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 0.033914573668213085,
|
| 1490 |
+
"grad_norm": 1.0689202547073364,
|
| 1491 |
+
"learning_rate": 1e-06,
|
| 1492 |
+
"loss": 0.1266,
|
| 1493 |
+
"step": 212
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 0.03407454807230843,
|
| 1497 |
+
"grad_norm": 1.4078688621520996,
|
| 1498 |
+
"learning_rate": 1e-06,
|
| 1499 |
+
"loss": 0.1247,
|
| 1500 |
+
"step": 213
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 0.034234522476403774,
|
| 1504 |
+
"grad_norm": 1.030319333076477,
|
| 1505 |
+
"learning_rate": 1e-06,
|
| 1506 |
+
"loss": 0.1292,
|
| 1507 |
+
"step": 214
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 0.03439449688049912,
|
| 1511 |
+
"grad_norm": 1.2724783420562744,
|
| 1512 |
+
"learning_rate": 1e-06,
|
| 1513 |
+
"loss": 0.1261,
|
| 1514 |
+
"step": 215
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 0.03455447128459446,
|
| 1518 |
+
"grad_norm": 1.1030813455581665,
|
| 1519 |
+
"learning_rate": 1e-06,
|
| 1520 |
+
"loss": 0.1304,
|
| 1521 |
+
"step": 216
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 0.03471444568868981,
|
| 1525 |
+
"grad_norm": 1.212585210800171,
|
| 1526 |
+
"learning_rate": 1e-06,
|
| 1527 |
+
"loss": 0.1279,
|
| 1528 |
+
"step": 217
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 0.03487442009278516,
|
| 1532 |
+
"grad_norm": 1.146895408630371,
|
| 1533 |
+
"learning_rate": 1e-06,
|
| 1534 |
+
"loss": 0.124,
|
| 1535 |
+
"step": 218
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 0.0350343944968805,
|
| 1539 |
+
"grad_norm": 1.102738857269287,
|
| 1540 |
+
"learning_rate": 1e-06,
|
| 1541 |
+
"loss": 0.1278,
|
| 1542 |
+
"step": 219
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 0.035194368900975846,
|
| 1546 |
+
"grad_norm": 1.302136778831482,
|
| 1547 |
+
"learning_rate": 1e-06,
|
| 1548 |
+
"loss": 0.1279,
|
| 1549 |
+
"step": 220
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 0.03535434330507119,
|
| 1553 |
+
"grad_norm": 1.228232502937317,
|
| 1554 |
+
"learning_rate": 1e-06,
|
| 1555 |
+
"loss": 0.128,
|
| 1556 |
+
"step": 221
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 0.035514317709166535,
|
| 1560 |
+
"grad_norm": 0.8978828191757202,
|
| 1561 |
+
"learning_rate": 1e-06,
|
| 1562 |
+
"loss": 0.1232,
|
| 1563 |
+
"step": 222
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 0.035674292113261875,
|
| 1567 |
+
"grad_norm": 1.0016555786132812,
|
| 1568 |
+
"learning_rate": 1e-06,
|
| 1569 |
+
"loss": 0.1261,
|
| 1570 |
+
"step": 223
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 0.03583426651735722,
|
| 1574 |
+
"grad_norm": 1.0512871742248535,
|
| 1575 |
+
"learning_rate": 1e-06,
|
| 1576 |
+
"loss": 0.1239,
|
| 1577 |
+
"step": 224
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 0.03599424092145257,
|
| 1581 |
+
"grad_norm": 1.1789789199829102,
|
| 1582 |
+
"learning_rate": 1e-06,
|
| 1583 |
+
"loss": 0.1224,
|
| 1584 |
+
"step": 225
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 0.03615421532554791,
|
| 1588 |
+
"grad_norm": 1.0191493034362793,
|
| 1589 |
+
"learning_rate": 1e-06,
|
| 1590 |
+
"loss": 0.1218,
|
| 1591 |
+
"step": 226
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 0.03631418972964326,
|
| 1595 |
+
"grad_norm": 1.345475196838379,
|
| 1596 |
+
"learning_rate": 1e-06,
|
| 1597 |
+
"loss": 0.1243,
|
| 1598 |
+
"step": 227
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 0.0364741641337386,
|
| 1602 |
+
"grad_norm": 2.024815082550049,
|
| 1603 |
+
"learning_rate": 1e-06,
|
| 1604 |
+
"loss": 0.1266,
|
| 1605 |
+
"step": 228
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 0.03663413853783395,
|
| 1609 |
+
"grad_norm": 1.0629597902297974,
|
| 1610 |
+
"learning_rate": 1e-06,
|
| 1611 |
+
"loss": 0.1244,
|
| 1612 |
+
"step": 229
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 0.03679411294192929,
|
| 1616 |
+
"grad_norm": 0.9979191422462463,
|
| 1617 |
+
"learning_rate": 1e-06,
|
| 1618 |
+
"loss": 0.1242,
|
| 1619 |
+
"step": 230
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 0.036954087346024636,
|
| 1623 |
+
"grad_norm": 1.2679247856140137,
|
| 1624 |
+
"learning_rate": 1e-06,
|
| 1625 |
+
"loss": 0.1224,
|
| 1626 |
+
"step": 231
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 0.037114061750119984,
|
| 1630 |
+
"grad_norm": 1.0288176536560059,
|
| 1631 |
+
"learning_rate": 1e-06,
|
| 1632 |
+
"loss": 0.1256,
|
| 1633 |
+
"step": 232
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 0.037274036154215325,
|
| 1637 |
+
"grad_norm": 1.3102436065673828,
|
| 1638 |
+
"learning_rate": 1e-06,
|
| 1639 |
+
"loss": 0.1262,
|
| 1640 |
+
"step": 233
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 0.03743401055831067,
|
| 1644 |
+
"grad_norm": 1.1388421058654785,
|
| 1645 |
+
"learning_rate": 1e-06,
|
| 1646 |
+
"loss": 0.1236,
|
| 1647 |
+
"step": 234
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 0.03759398496240601,
|
| 1651 |
+
"grad_norm": 1.0879204273223877,
|
| 1652 |
+
"learning_rate": 1e-06,
|
| 1653 |
+
"loss": 0.1268,
|
| 1654 |
+
"step": 235
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 0.03775395936650136,
|
| 1658 |
+
"grad_norm": 1.7273410558700562,
|
| 1659 |
+
"learning_rate": 1e-06,
|
| 1660 |
+
"loss": 0.1279,
|
| 1661 |
+
"step": 236
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 0.0379139337705967,
|
| 1665 |
+
"grad_norm": 1.3870850801467896,
|
| 1666 |
+
"learning_rate": 1e-06,
|
| 1667 |
+
"loss": 0.1262,
|
| 1668 |
+
"step": 237
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 0.03807390817469205,
|
| 1672 |
+
"grad_norm": 1.259337067604065,
|
| 1673 |
+
"learning_rate": 1e-06,
|
| 1674 |
+
"loss": 0.1267,
|
| 1675 |
+
"step": 238
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 0.0382338825787874,
|
| 1679 |
+
"grad_norm": 1.4889863729476929,
|
| 1680 |
+
"learning_rate": 1e-06,
|
| 1681 |
+
"loss": 0.1282,
|
| 1682 |
+
"step": 239
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 0.03839385698288274,
|
| 1686 |
+
"grad_norm": 1.2688918113708496,
|
| 1687 |
+
"learning_rate": 1e-06,
|
| 1688 |
+
"loss": 0.1263,
|
| 1689 |
+
"step": 240
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 0.038553831386978085,
|
| 1693 |
+
"grad_norm": 1.057685375213623,
|
| 1694 |
+
"learning_rate": 1e-06,
|
| 1695 |
+
"loss": 0.1221,
|
| 1696 |
+
"step": 241
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 0.038713805791073426,
|
| 1700 |
+
"grad_norm": 1.0359437465667725,
|
| 1701 |
+
"learning_rate": 1e-06,
|
| 1702 |
+
"loss": 0.1285,
|
| 1703 |
+
"step": 242
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 0.038873780195168774,
|
| 1707 |
+
"grad_norm": 1.0317096710205078,
|
| 1708 |
+
"learning_rate": 1e-06,
|
| 1709 |
+
"loss": 0.1236,
|
| 1710 |
+
"step": 243
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 0.039033754599264114,
|
| 1714 |
+
"grad_norm": 1.4080708026885986,
|
| 1715 |
+
"learning_rate": 1e-06,
|
| 1716 |
+
"loss": 0.1256,
|
| 1717 |
+
"step": 244
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 0.03919372900335946,
|
| 1721 |
+
"grad_norm": 1.2293702363967896,
|
| 1722 |
+
"learning_rate": 1e-06,
|
| 1723 |
+
"loss": 0.1225,
|
| 1724 |
+
"step": 245
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 0.03935370340745481,
|
| 1728 |
+
"grad_norm": 1.2161550521850586,
|
| 1729 |
+
"learning_rate": 1e-06,
|
| 1730 |
+
"loss": 0.1273,
|
| 1731 |
+
"step": 246
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 0.03951367781155015,
|
| 1735 |
+
"grad_norm": 0.9967581629753113,
|
| 1736 |
+
"learning_rate": 1e-06,
|
| 1737 |
+
"loss": 0.1232,
|
| 1738 |
+
"step": 247
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 0.0396736522156455,
|
| 1742 |
+
"grad_norm": 1.0089436769485474,
|
| 1743 |
+
"learning_rate": 1e-06,
|
| 1744 |
+
"loss": 0.1252,
|
| 1745 |
+
"step": 248
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 0.03983362661974084,
|
| 1749 |
+
"grad_norm": 1.1048007011413574,
|
| 1750 |
+
"learning_rate": 1e-06,
|
| 1751 |
+
"loss": 0.123,
|
| 1752 |
+
"step": 249
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 0.03999360102383619,
|
| 1756 |
+
"grad_norm": 1.3069252967834473,
|
| 1757 |
+
"learning_rate": 1e-06,
|
| 1758 |
+
"loss": 0.1275,
|
| 1759 |
+
"step": 250
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 0.040153575427931534,
|
| 1763 |
+
"grad_norm": 1.3315153121948242,
|
| 1764 |
+
"learning_rate": 1e-06,
|
| 1765 |
+
"loss": 0.1262,
|
| 1766 |
+
"step": 251
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 0.040313549832026875,
|
| 1770 |
+
"grad_norm": 1.1331770420074463,
|
| 1771 |
+
"learning_rate": 1e-06,
|
| 1772 |
+
"loss": 0.1243,
|
| 1773 |
+
"step": 252
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 0.04047352423612222,
|
| 1777 |
+
"grad_norm": 1.024278998374939,
|
| 1778 |
+
"learning_rate": 1e-06,
|
| 1779 |
+
"loss": 0.1241,
|
| 1780 |
+
"step": 253
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 0.040633498640217564,
|
| 1784 |
+
"grad_norm": 1.151720404624939,
|
| 1785 |
+
"learning_rate": 1e-06,
|
| 1786 |
+
"loss": 0.1229,
|
| 1787 |
+
"step": 254
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 0.04079347304431291,
|
| 1791 |
+
"grad_norm": 1.107699990272522,
|
| 1792 |
+
"learning_rate": 1e-06,
|
| 1793 |
+
"loss": 0.1213,
|
| 1794 |
+
"step": 255
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 0.04095344744840825,
|
| 1798 |
+
"grad_norm": 1.2880581617355347,
|
| 1799 |
+
"learning_rate": 1e-06,
|
| 1800 |
+
"loss": 0.1275,
|
| 1801 |
+
"step": 256
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 0.0411134218525036,
|
| 1805 |
+
"grad_norm": 1.247376561164856,
|
| 1806 |
+
"learning_rate": 1e-06,
|
| 1807 |
+
"loss": 0.1266,
|
| 1808 |
+
"step": 257
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 0.04127339625659895,
|
| 1812 |
+
"grad_norm": 1.3562711477279663,
|
| 1813 |
+
"learning_rate": 1e-06,
|
| 1814 |
+
"loss": 0.1218,
|
| 1815 |
+
"step": 258
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 0.04143337066069429,
|
| 1819 |
+
"grad_norm": 1.2490332126617432,
|
| 1820 |
+
"learning_rate": 1e-06,
|
| 1821 |
+
"loss": 0.1242,
|
| 1822 |
+
"step": 259
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 0.041593345064789636,
|
| 1826 |
+
"grad_norm": 0.9172055125236511,
|
| 1827 |
+
"learning_rate": 1e-06,
|
| 1828 |
+
"loss": 0.1253,
|
| 1829 |
+
"step": 260
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 0.041753319468884977,
|
| 1833 |
+
"grad_norm": 1.6371047496795654,
|
| 1834 |
+
"learning_rate": 1e-06,
|
| 1835 |
+
"loss": 0.1223,
|
| 1836 |
+
"step": 261
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 0.041913293872980324,
|
| 1840 |
+
"grad_norm": 1.1314542293548584,
|
| 1841 |
+
"learning_rate": 1e-06,
|
| 1842 |
+
"loss": 0.1226,
|
| 1843 |
+
"step": 262
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 0.042073268277075665,
|
| 1847 |
+
"grad_norm": 1.001624584197998,
|
| 1848 |
+
"learning_rate": 1e-06,
|
| 1849 |
+
"loss": 0.1221,
|
| 1850 |
+
"step": 263
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 0.04223324268117101,
|
| 1854 |
+
"grad_norm": 1.015631079673767,
|
| 1855 |
+
"learning_rate": 1e-06,
|
| 1856 |
+
"loss": 0.1234,
|
| 1857 |
+
"step": 264
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 0.04239321708526636,
|
| 1861 |
+
"grad_norm": 1.6776975393295288,
|
| 1862 |
+
"learning_rate": 1e-06,
|
| 1863 |
+
"loss": 0.127,
|
| 1864 |
+
"step": 265
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 0.0425531914893617,
|
| 1868 |
+
"grad_norm": 1.4266294240951538,
|
| 1869 |
+
"learning_rate": 1e-06,
|
| 1870 |
+
"loss": 0.1254,
|
| 1871 |
+
"step": 266
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 0.04271316589345705,
|
| 1875 |
+
"grad_norm": 1.0626451969146729,
|
| 1876 |
+
"learning_rate": 1e-06,
|
| 1877 |
+
"loss": 0.1199,
|
| 1878 |
+
"step": 267
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 0.04287314029755239,
|
| 1882 |
+
"grad_norm": 1.2195415496826172,
|
| 1883 |
+
"learning_rate": 1e-06,
|
| 1884 |
+
"loss": 0.1216,
|
| 1885 |
+
"step": 268
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 0.04303311470164774,
|
| 1889 |
+
"grad_norm": 1.431025505065918,
|
| 1890 |
+
"learning_rate": 1e-06,
|
| 1891 |
+
"loss": 0.1231,
|
| 1892 |
+
"step": 269
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 0.04319308910574308,
|
| 1896 |
+
"grad_norm": 0.9598281383514404,
|
| 1897 |
+
"learning_rate": 1e-06,
|
| 1898 |
+
"loss": 0.122,
|
| 1899 |
+
"step": 270
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 0.043353063509838426,
|
| 1903 |
+
"grad_norm": 1.2865647077560425,
|
| 1904 |
+
"learning_rate": 1e-06,
|
| 1905 |
+
"loss": 0.1234,
|
| 1906 |
+
"step": 271
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 0.04351303791393377,
|
| 1910 |
+
"grad_norm": 1.2584363222122192,
|
| 1911 |
+
"learning_rate": 1e-06,
|
| 1912 |
+
"loss": 0.1222,
|
| 1913 |
+
"step": 272
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 0.043673012318029114,
|
| 1917 |
+
"grad_norm": 1.6006242036819458,
|
| 1918 |
+
"learning_rate": 1e-06,
|
| 1919 |
+
"loss": 0.1216,
|
| 1920 |
+
"step": 273
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 0.04383298672212446,
|
| 1924 |
+
"grad_norm": 1.14664626121521,
|
| 1925 |
+
"learning_rate": 1e-06,
|
| 1926 |
+
"loss": 0.1238,
|
| 1927 |
+
"step": 274
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 0.0439929611262198,
|
| 1931 |
+
"grad_norm": 1.0439503192901611,
|
| 1932 |
+
"learning_rate": 1e-06,
|
| 1933 |
+
"loss": 0.123,
|
| 1934 |
+
"step": 275
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 0.04415293553031515,
|
| 1938 |
+
"grad_norm": 1.2464580535888672,
|
| 1939 |
+
"learning_rate": 1e-06,
|
| 1940 |
+
"loss": 0.1228,
|
| 1941 |
+
"step": 276
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 0.04431290993441049,
|
| 1945 |
+
"grad_norm": 1.0424617528915405,
|
| 1946 |
+
"learning_rate": 1e-06,
|
| 1947 |
+
"loss": 0.1242,
|
| 1948 |
+
"step": 277
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 0.04447288433850584,
|
| 1952 |
+
"grad_norm": 1.3555285930633545,
|
| 1953 |
+
"learning_rate": 1e-06,
|
| 1954 |
+
"loss": 0.1192,
|
| 1955 |
+
"step": 278
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 0.044632858742601186,
|
| 1959 |
+
"grad_norm": 1.4239215850830078,
|
| 1960 |
+
"learning_rate": 1e-06,
|
| 1961 |
+
"loss": 0.1258,
|
| 1962 |
+
"step": 279
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 0.04479283314669653,
|
| 1966 |
+
"grad_norm": 1.1337487697601318,
|
| 1967 |
+
"learning_rate": 1e-06,
|
| 1968 |
+
"loss": 0.1235,
|
| 1969 |
+
"step": 280
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 0.044952807550791875,
|
| 1973 |
+
"grad_norm": 1.1882452964782715,
|
| 1974 |
+
"learning_rate": 1e-06,
|
| 1975 |
+
"loss": 0.1192,
|
| 1976 |
+
"step": 281
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 0.045112781954887216,
|
| 1980 |
+
"grad_norm": 1.1663519144058228,
|
| 1981 |
+
"learning_rate": 1e-06,
|
| 1982 |
+
"loss": 0.1241,
|
| 1983 |
+
"step": 282
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 0.04527275635898256,
|
| 1987 |
+
"grad_norm": 1.189931869506836,
|
| 1988 |
+
"learning_rate": 1e-06,
|
| 1989 |
+
"loss": 0.1227,
|
| 1990 |
+
"step": 283
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 0.04543273076307791,
|
| 1994 |
+
"grad_norm": 1.3207061290740967,
|
| 1995 |
+
"learning_rate": 1e-06,
|
| 1996 |
+
"loss": 0.1196,
|
| 1997 |
+
"step": 284
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 0.04559270516717325,
|
| 2001 |
+
"grad_norm": 1.5193383693695068,
|
| 2002 |
+
"learning_rate": 1e-06,
|
| 2003 |
+
"loss": 0.1255,
|
| 2004 |
+
"step": 285
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 0.0457526795712686,
|
| 2008 |
+
"grad_norm": 0.9933753609657288,
|
| 2009 |
+
"learning_rate": 1e-06,
|
| 2010 |
+
"loss": 0.1196,
|
| 2011 |
+
"step": 286
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 0.04591265397536394,
|
| 2015 |
+
"grad_norm": 0.9570215344429016,
|
| 2016 |
+
"learning_rate": 1e-06,
|
| 2017 |
+
"loss": 0.1185,
|
| 2018 |
+
"step": 287
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 0.04607262837945929,
|
| 2022 |
+
"grad_norm": 1.136502742767334,
|
| 2023 |
+
"learning_rate": 1e-06,
|
| 2024 |
+
"loss": 0.1201,
|
| 2025 |
+
"step": 288
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 0.04623260278355463,
|
| 2029 |
+
"grad_norm": 1.320960283279419,
|
| 2030 |
+
"learning_rate": 1e-06,
|
| 2031 |
+
"loss": 0.1228,
|
| 2032 |
+
"step": 289
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 0.046392577187649976,
|
| 2036 |
+
"grad_norm": 1.0228922367095947,
|
| 2037 |
+
"learning_rate": 1e-06,
|
| 2038 |
+
"loss": 0.1198,
|
| 2039 |
+
"step": 290
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 0.046552551591745324,
|
| 2043 |
+
"grad_norm": 0.9063114523887634,
|
| 2044 |
+
"learning_rate": 1e-06,
|
| 2045 |
+
"loss": 0.1192,
|
| 2046 |
+
"step": 291
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 0.046712525995840665,
|
| 2050 |
+
"grad_norm": 1.1633445024490356,
|
| 2051 |
+
"learning_rate": 1e-06,
|
| 2052 |
+
"loss": 0.1236,
|
| 2053 |
+
"step": 292
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 0.04687250039993601,
|
| 2057 |
+
"grad_norm": 1.2757052183151245,
|
| 2058 |
+
"learning_rate": 1e-06,
|
| 2059 |
+
"loss": 0.12,
|
| 2060 |
+
"step": 293
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 0.04703247480403135,
|
| 2064 |
+
"grad_norm": 1.551196575164795,
|
| 2065 |
+
"learning_rate": 1e-06,
|
| 2066 |
+
"loss": 0.1241,
|
| 2067 |
+
"step": 294
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 0.0471924492081267,
|
| 2071 |
+
"grad_norm": 1.631995439529419,
|
| 2072 |
+
"learning_rate": 1e-06,
|
| 2073 |
+
"loss": 0.1238,
|
| 2074 |
+
"step": 295
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 0.04735242361222204,
|
| 2078 |
+
"grad_norm": 1.4675942659378052,
|
| 2079 |
+
"learning_rate": 1e-06,
|
| 2080 |
+
"loss": 0.1182,
|
| 2081 |
+
"step": 296
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 0.04751239801631739,
|
| 2085 |
+
"grad_norm": 1.1384583711624146,
|
| 2086 |
+
"learning_rate": 1e-06,
|
| 2087 |
+
"loss": 0.1195,
|
| 2088 |
+
"step": 297
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 0.04767237242041274,
|
| 2092 |
+
"grad_norm": 0.9831569194793701,
|
| 2093 |
+
"learning_rate": 1e-06,
|
| 2094 |
+
"loss": 0.1189,
|
| 2095 |
+
"step": 298
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 0.04783234682450808,
|
| 2099 |
+
"grad_norm": 1.4235044717788696,
|
| 2100 |
+
"learning_rate": 1e-06,
|
| 2101 |
+
"loss": 0.1202,
|
| 2102 |
+
"step": 299
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 0.047992321228603425,
|
| 2106 |
+
"grad_norm": 1.4275295734405518,
|
| 2107 |
+
"learning_rate": 1e-06,
|
| 2108 |
+
"loss": 0.1226,
|
| 2109 |
+
"step": 300
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 0.048152295632698766,
|
| 2113 |
+
"grad_norm": 1.2116954326629639,
|
| 2114 |
+
"learning_rate": 1e-06,
|
| 2115 |
+
"loss": 0.1205,
|
| 2116 |
+
"step": 301
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 0.048312270036794114,
|
| 2120 |
+
"grad_norm": 1.479552149772644,
|
| 2121 |
+
"learning_rate": 1e-06,
|
| 2122 |
+
"loss": 0.1215,
|
| 2123 |
+
"step": 302
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 0.048472244440889455,
|
| 2127 |
+
"grad_norm": 1.3289475440979004,
|
| 2128 |
+
"learning_rate": 1e-06,
|
| 2129 |
+
"loss": 0.1215,
|
| 2130 |
+
"step": 303
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 0.0486322188449848,
|
| 2134 |
+
"grad_norm": 1.4380484819412231,
|
| 2135 |
+
"learning_rate": 1e-06,
|
| 2136 |
+
"loss": 0.1214,
|
| 2137 |
+
"step": 304
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 0.04879219324908015,
|
| 2141 |
+
"grad_norm": 1.1001449823379517,
|
| 2142 |
+
"learning_rate": 1e-06,
|
| 2143 |
+
"loss": 0.12,
|
| 2144 |
+
"step": 305
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 0.04895216765317549,
|
| 2148 |
+
"grad_norm": 1.2235729694366455,
|
| 2149 |
+
"learning_rate": 1e-06,
|
| 2150 |
+
"loss": 0.1171,
|
| 2151 |
+
"step": 306
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 0.04911214205727084,
|
| 2155 |
+
"grad_norm": 1.2234548330307007,
|
| 2156 |
+
"learning_rate": 1e-06,
|
| 2157 |
+
"loss": 0.1227,
|
| 2158 |
+
"step": 307
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 0.04927211646136618,
|
| 2162 |
+
"grad_norm": 1.6779841184616089,
|
| 2163 |
+
"learning_rate": 1e-06,
|
| 2164 |
+
"loss": 0.1225,
|
| 2165 |
+
"step": 308
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 0.04943209086546153,
|
| 2169 |
+
"grad_norm": 1.339382529258728,
|
| 2170 |
+
"learning_rate": 1e-06,
|
| 2171 |
+
"loss": 0.1219,
|
| 2172 |
+
"step": 309
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 0.04959206526955687,
|
| 2176 |
+
"grad_norm": 0.8880839943885803,
|
| 2177 |
+
"learning_rate": 1e-06,
|
| 2178 |
+
"loss": 0.1208,
|
| 2179 |
+
"step": 310
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 0.049752039673652215,
|
| 2183 |
+
"grad_norm": 1.1524003744125366,
|
| 2184 |
+
"learning_rate": 1e-06,
|
| 2185 |
+
"loss": 0.1212,
|
| 2186 |
+
"step": 311
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 0.04991201407774756,
|
| 2190 |
+
"grad_norm": 1.2433319091796875,
|
| 2191 |
+
"learning_rate": 1e-06,
|
| 2192 |
+
"loss": 0.1196,
|
| 2193 |
+
"step": 312
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 0.050071988481842904,
|
| 2197 |
+
"grad_norm": 1.299512505531311,
|
| 2198 |
+
"learning_rate": 1e-06,
|
| 2199 |
+
"loss": 0.1226,
|
| 2200 |
+
"step": 313
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 0.05023196288593825,
|
| 2204 |
+
"grad_norm": 1.0843769311904907,
|
| 2205 |
+
"learning_rate": 1e-06,
|
| 2206 |
+
"loss": 0.1209,
|
| 2207 |
+
"step": 314
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 0.05039193729003359,
|
| 2211 |
+
"grad_norm": 1.617997169494629,
|
| 2212 |
+
"learning_rate": 1e-06,
|
| 2213 |
+
"loss": 0.1212,
|
| 2214 |
+
"step": 315
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 0.05055191169412894,
|
| 2218 |
+
"grad_norm": 1.2156031131744385,
|
| 2219 |
+
"learning_rate": 1e-06,
|
| 2220 |
+
"loss": 0.1185,
|
| 2221 |
+
"step": 316
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 0.05071188609822429,
|
| 2225 |
+
"grad_norm": 1.658170461654663,
|
| 2226 |
+
"learning_rate": 1e-06,
|
| 2227 |
+
"loss": 0.1264,
|
| 2228 |
+
"step": 317
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 0.05087186050231963,
|
| 2232 |
+
"grad_norm": 1.341801404953003,
|
| 2233 |
+
"learning_rate": 1e-06,
|
| 2234 |
+
"loss": 0.124,
|
| 2235 |
+
"step": 318
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 0.051031834906414976,
|
| 2239 |
+
"grad_norm": 1.3077876567840576,
|
| 2240 |
+
"learning_rate": 1e-06,
|
| 2241 |
+
"loss": 0.1217,
|
| 2242 |
+
"step": 319
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 0.05119180931051032,
|
| 2246 |
+
"grad_norm": 1.2550292015075684,
|
| 2247 |
+
"learning_rate": 1e-06,
|
| 2248 |
+
"loss": 0.1218,
|
| 2249 |
+
"step": 320
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 0.051351783714605664,
|
| 2253 |
+
"grad_norm": 0.9955998063087463,
|
| 2254 |
+
"learning_rate": 1e-06,
|
| 2255 |
+
"loss": 0.1205,
|
| 2256 |
+
"step": 321
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 0.051511758118701005,
|
| 2260 |
+
"grad_norm": 1.0665310621261597,
|
| 2261 |
+
"learning_rate": 1e-06,
|
| 2262 |
+
"loss": 0.1204,
|
| 2263 |
+
"step": 322
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 0.05167173252279635,
|
| 2267 |
+
"grad_norm": 0.8823798894882202,
|
| 2268 |
+
"learning_rate": 1e-06,
|
| 2269 |
+
"loss": 0.1184,
|
| 2270 |
+
"step": 323
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 0.0518317069268917,
|
| 2274 |
+
"grad_norm": 1.2012176513671875,
|
| 2275 |
+
"learning_rate": 1e-06,
|
| 2276 |
+
"loss": 0.1192,
|
| 2277 |
+
"step": 324
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 0.05199168133098704,
|
| 2281 |
+
"grad_norm": 1.2626415491104126,
|
| 2282 |
+
"learning_rate": 1e-06,
|
| 2283 |
+
"loss": 0.1187,
|
| 2284 |
+
"step": 325
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 0.05215165573508239,
|
| 2288 |
+
"grad_norm": 1.4473867416381836,
|
| 2289 |
+
"learning_rate": 1e-06,
|
| 2290 |
+
"loss": 0.1203,
|
| 2291 |
+
"step": 326
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 0.05231163013917773,
|
| 2295 |
+
"grad_norm": 1.1338282823562622,
|
| 2296 |
+
"learning_rate": 1e-06,
|
| 2297 |
+
"loss": 0.1202,
|
| 2298 |
+
"step": 327
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 0.05247160454327308,
|
| 2302 |
+
"grad_norm": 1.3163774013519287,
|
| 2303 |
+
"learning_rate": 1e-06,
|
| 2304 |
+
"loss": 0.1215,
|
| 2305 |
+
"step": 328
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 0.05263157894736842,
|
| 2309 |
+
"grad_norm": 1.1282827854156494,
|
| 2310 |
+
"learning_rate": 1e-06,
|
| 2311 |
+
"loss": 0.1181,
|
| 2312 |
+
"step": 329
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 0.052791553351463766,
|
| 2316 |
+
"grad_norm": 0.8695802092552185,
|
| 2317 |
+
"learning_rate": 1e-06,
|
| 2318 |
+
"loss": 0.1199,
|
| 2319 |
+
"step": 330
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 0.052951527755559114,
|
| 2323 |
+
"grad_norm": 0.9970802664756775,
|
| 2324 |
+
"learning_rate": 1e-06,
|
| 2325 |
+
"loss": 0.1184,
|
| 2326 |
+
"step": 331
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 0.053111502159654454,
|
| 2330 |
+
"grad_norm": 0.9834699630737305,
|
| 2331 |
+
"learning_rate": 1e-06,
|
| 2332 |
+
"loss": 0.1165,
|
| 2333 |
+
"step": 332
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 0.0532714765637498,
|
| 2337 |
+
"grad_norm": 1.0248655080795288,
|
| 2338 |
+
"learning_rate": 1e-06,
|
| 2339 |
+
"loss": 0.1232,
|
| 2340 |
+
"step": 333
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 0.05343145096784514,
|
| 2344 |
+
"grad_norm": 1.1430097818374634,
|
| 2345 |
+
"learning_rate": 1e-06,
|
| 2346 |
+
"loss": 0.1224,
|
| 2347 |
+
"step": 334
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 0.05359142537194049,
|
| 2351 |
+
"grad_norm": 1.295326828956604,
|
| 2352 |
+
"learning_rate": 1e-06,
|
| 2353 |
+
"loss": 0.1228,
|
| 2354 |
+
"step": 335
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 0.05375139977603583,
|
| 2358 |
+
"grad_norm": 1.067046880722046,
|
| 2359 |
+
"learning_rate": 1e-06,
|
| 2360 |
+
"loss": 0.1211,
|
| 2361 |
+
"step": 336
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 0.05391137418013118,
|
| 2365 |
+
"grad_norm": 1.057966947555542,
|
| 2366 |
+
"learning_rate": 1e-06,
|
| 2367 |
+
"loss": 0.1165,
|
| 2368 |
+
"step": 337
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 0.054071348584226527,
|
| 2372 |
+
"grad_norm": 1.1167185306549072,
|
| 2373 |
+
"learning_rate": 1e-06,
|
| 2374 |
+
"loss": 0.1221,
|
| 2375 |
+
"step": 338
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 0.05423132298832187,
|
| 2379 |
+
"grad_norm": 0.8829674124717712,
|
| 2380 |
+
"learning_rate": 1e-06,
|
| 2381 |
+
"loss": 0.1185,
|
| 2382 |
+
"step": 339
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 0.054391297392417215,
|
| 2386 |
+
"grad_norm": 1.3520281314849854,
|
| 2387 |
+
"learning_rate": 1e-06,
|
| 2388 |
+
"loss": 0.1214,
|
| 2389 |
+
"step": 340
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 0.054551271796512556,
|
| 2393 |
+
"grad_norm": 1.4164739847183228,
|
| 2394 |
+
"learning_rate": 1e-06,
|
| 2395 |
+
"loss": 0.1184,
|
| 2396 |
+
"step": 341
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 0.0547112462006079,
|
| 2400 |
+
"grad_norm": 1.1109625101089478,
|
| 2401 |
+
"learning_rate": 1e-06,
|
| 2402 |
+
"loss": 0.1199,
|
| 2403 |
+
"step": 342
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 0.054871220604703244,
|
| 2407 |
+
"grad_norm": 1.1353943347930908,
|
| 2408 |
+
"learning_rate": 1e-06,
|
| 2409 |
+
"loss": 0.118,
|
| 2410 |
+
"step": 343
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 0.05503119500879859,
|
| 2414 |
+
"grad_norm": 1.3248083591461182,
|
| 2415 |
+
"learning_rate": 1e-06,
|
| 2416 |
+
"loss": 0.1207,
|
| 2417 |
+
"step": 344
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 0.05519116941289394,
|
| 2421 |
+
"grad_norm": 1.3598345518112183,
|
| 2422 |
+
"learning_rate": 1e-06,
|
| 2423 |
+
"loss": 0.1178,
|
| 2424 |
+
"step": 345
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 0.05535114381698928,
|
| 2428 |
+
"grad_norm": 1.3764357566833496,
|
| 2429 |
+
"learning_rate": 1e-06,
|
| 2430 |
+
"loss": 0.1185,
|
| 2431 |
+
"step": 346
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 0.05551111822108463,
|
| 2435 |
+
"grad_norm": 1.07554292678833,
|
| 2436 |
+
"learning_rate": 1e-06,
|
| 2437 |
+
"loss": 0.1193,
|
| 2438 |
+
"step": 347
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 0.05567109262517997,
|
| 2442 |
+
"grad_norm": 1.3993921279907227,
|
| 2443 |
+
"learning_rate": 1e-06,
|
| 2444 |
+
"loss": 0.1209,
|
| 2445 |
+
"step": 348
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 0.055831067029275316,
|
| 2449 |
+
"grad_norm": 1.208682894706726,
|
| 2450 |
+
"learning_rate": 1e-06,
|
| 2451 |
+
"loss": 0.1188,
|
| 2452 |
+
"step": 349
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 0.055991041433370664,
|
| 2456 |
+
"grad_norm": 0.9713140726089478,
|
| 2457 |
+
"learning_rate": 1e-06,
|
| 2458 |
+
"loss": 0.1187,
|
| 2459 |
+
"step": 350
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 0.056151015837466005,
|
| 2463 |
+
"grad_norm": 0.9984699487686157,
|
| 2464 |
+
"learning_rate": 1e-06,
|
| 2465 |
+
"loss": 0.1146,
|
| 2466 |
+
"step": 351
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 0.05631099024156135,
|
| 2470 |
+
"grad_norm": 1.1007826328277588,
|
| 2471 |
+
"learning_rate": 1e-06,
|
| 2472 |
+
"loss": 0.1201,
|
| 2473 |
+
"step": 352
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 0.05647096464565669,
|
| 2477 |
+
"grad_norm": 1.3435072898864746,
|
| 2478 |
+
"learning_rate": 1e-06,
|
| 2479 |
+
"loss": 0.1177,
|
| 2480 |
+
"step": 353
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 0.05663093904975204,
|
| 2484 |
+
"grad_norm": 1.3530515432357788,
|
| 2485 |
+
"learning_rate": 1e-06,
|
| 2486 |
+
"loss": 0.1193,
|
| 2487 |
+
"step": 354
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 0.05679091345384738,
|
| 2491 |
+
"grad_norm": 1.0946826934814453,
|
| 2492 |
+
"learning_rate": 1e-06,
|
| 2493 |
+
"loss": 0.1154,
|
| 2494 |
+
"step": 355
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 0.05695088785794273,
|
| 2498 |
+
"grad_norm": 1.277107834815979,
|
| 2499 |
+
"learning_rate": 1e-06,
|
| 2500 |
+
"loss": 0.1196,
|
| 2501 |
+
"step": 356
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 0.05711086226203808,
|
| 2505 |
+
"grad_norm": 1.668004035949707,
|
| 2506 |
+
"learning_rate": 1e-06,
|
| 2507 |
+
"loss": 0.1201,
|
| 2508 |
+
"step": 357
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 0.05727083666613342,
|
| 2512 |
+
"grad_norm": 1.1211800575256348,
|
| 2513 |
+
"learning_rate": 1e-06,
|
| 2514 |
+
"loss": 0.1162,
|
| 2515 |
+
"step": 358
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 0.057430811070228766,
|
| 2519 |
+
"grad_norm": 1.0412259101867676,
|
| 2520 |
+
"learning_rate": 1e-06,
|
| 2521 |
+
"loss": 0.1174,
|
| 2522 |
+
"step": 359
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 0.057590785474324106,
|
| 2526 |
+
"grad_norm": 1.1593021154403687,
|
| 2527 |
+
"learning_rate": 1e-06,
|
| 2528 |
+
"loss": 0.1197,
|
| 2529 |
+
"step": 360
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 0.057750759878419454,
|
| 2533 |
+
"grad_norm": 1.2287545204162598,
|
| 2534 |
+
"learning_rate": 1e-06,
|
| 2535 |
+
"loss": 0.1205,
|
| 2536 |
+
"step": 361
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 0.057910734282514795,
|
| 2540 |
+
"grad_norm": 1.0975946187973022,
|
| 2541 |
+
"learning_rate": 1e-06,
|
| 2542 |
+
"loss": 0.1147,
|
| 2543 |
+
"step": 362
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 0.05807070868661014,
|
| 2547 |
+
"grad_norm": 1.073879361152649,
|
| 2548 |
+
"learning_rate": 1e-06,
|
| 2549 |
+
"loss": 0.1166,
|
| 2550 |
+
"step": 363
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 0.05823068309070549,
|
| 2554 |
+
"grad_norm": 1.0244841575622559,
|
| 2555 |
+
"learning_rate": 1e-06,
|
| 2556 |
+
"loss": 0.1181,
|
| 2557 |
+
"step": 364
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 0.05839065749480083,
|
| 2561 |
+
"grad_norm": 1.0585442781448364,
|
| 2562 |
+
"learning_rate": 1e-06,
|
| 2563 |
+
"loss": 0.1197,
|
| 2564 |
+
"step": 365
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 0.05855063189889618,
|
| 2568 |
+
"grad_norm": 0.9430259466171265,
|
| 2569 |
+
"learning_rate": 1e-06,
|
| 2570 |
+
"loss": 0.1155,
|
| 2571 |
+
"step": 366
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 0.05871060630299152,
|
| 2575 |
+
"grad_norm": 0.9988534450531006,
|
| 2576 |
+
"learning_rate": 1e-06,
|
| 2577 |
+
"loss": 0.1179,
|
| 2578 |
+
"step": 367
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 0.05887058070708687,
|
| 2582 |
+
"grad_norm": 1.4290752410888672,
|
| 2583 |
+
"learning_rate": 1e-06,
|
| 2584 |
+
"loss": 0.1166,
|
| 2585 |
+
"step": 368
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 0.05903055511118221,
|
| 2589 |
+
"grad_norm": 1.2006633281707764,
|
| 2590 |
+
"learning_rate": 1e-06,
|
| 2591 |
+
"loss": 0.1183,
|
| 2592 |
+
"step": 369
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 0.059190529515277555,
|
| 2596 |
+
"grad_norm": 1.0609732866287231,
|
| 2597 |
+
"learning_rate": 1e-06,
|
| 2598 |
+
"loss": 0.1164,
|
| 2599 |
+
"step": 370
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 0.0593505039193729,
|
| 2603 |
+
"grad_norm": 1.2493401765823364,
|
| 2604 |
+
"learning_rate": 1e-06,
|
| 2605 |
+
"loss": 0.117,
|
| 2606 |
+
"step": 371
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 0.059510478323468244,
|
| 2610 |
+
"grad_norm": 1.0093525648117065,
|
| 2611 |
+
"learning_rate": 1e-06,
|
| 2612 |
+
"loss": 0.1171,
|
| 2613 |
+
"step": 372
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 0.05967045272756359,
|
| 2617 |
+
"grad_norm": 1.1668469905853271,
|
| 2618 |
+
"learning_rate": 1e-06,
|
| 2619 |
+
"loss": 0.1177,
|
| 2620 |
+
"step": 373
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 0.05983042713165893,
|
| 2624 |
+
"grad_norm": 1.0411137342453003,
|
| 2625 |
+
"learning_rate": 1e-06,
|
| 2626 |
+
"loss": 0.1154,
|
| 2627 |
+
"step": 374
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 0.05999040153575428,
|
| 2631 |
+
"grad_norm": 1.7043784856796265,
|
| 2632 |
+
"learning_rate": 1e-06,
|
| 2633 |
+
"loss": 0.1163,
|
| 2634 |
+
"step": 375
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 0.06015037593984962,
|
| 2638 |
+
"grad_norm": 1.1405059099197388,
|
| 2639 |
+
"learning_rate": 1e-06,
|
| 2640 |
+
"loss": 0.1179,
|
| 2641 |
+
"step": 376
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 0.06031035034394497,
|
| 2645 |
+
"grad_norm": 1.1912827491760254,
|
| 2646 |
+
"learning_rate": 1e-06,
|
| 2647 |
+
"loss": 0.1151,
|
| 2648 |
+
"step": 377
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 0.060470324748040316,
|
| 2652 |
+
"grad_norm": 1.0265625715255737,
|
| 2653 |
+
"learning_rate": 1e-06,
|
| 2654 |
+
"loss": 0.1169,
|
| 2655 |
+
"step": 378
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 0.06063029915213566,
|
| 2659 |
+
"grad_norm": 1.2455023527145386,
|
| 2660 |
+
"learning_rate": 1e-06,
|
| 2661 |
+
"loss": 0.1156,
|
| 2662 |
+
"step": 379
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 0.060790273556231005,
|
| 2666 |
+
"grad_norm": 1.0697797536849976,
|
| 2667 |
+
"learning_rate": 1e-06,
|
| 2668 |
+
"loss": 0.1112,
|
| 2669 |
+
"step": 380
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 0.060950247960326345,
|
| 2673 |
+
"grad_norm": 1.2517694234848022,
|
| 2674 |
+
"learning_rate": 1e-06,
|
| 2675 |
+
"loss": 0.1139,
|
| 2676 |
+
"step": 381
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 0.06111022236442169,
|
| 2680 |
+
"grad_norm": 1.1439390182495117,
|
| 2681 |
+
"learning_rate": 1e-06,
|
| 2682 |
+
"loss": 0.1157,
|
| 2683 |
+
"step": 382
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 0.06127019676851704,
|
| 2687 |
+
"grad_norm": 1.0659351348876953,
|
| 2688 |
+
"learning_rate": 1e-06,
|
| 2689 |
+
"loss": 0.115,
|
| 2690 |
+
"step": 383
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 0.06143017117261238,
|
| 2694 |
+
"grad_norm": 1.1062361001968384,
|
| 2695 |
+
"learning_rate": 1e-06,
|
| 2696 |
+
"loss": 0.1164,
|
| 2697 |
+
"step": 384
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 0.06159014557670773,
|
| 2701 |
+
"grad_norm": 1.227656364440918,
|
| 2702 |
+
"learning_rate": 1e-06,
|
| 2703 |
+
"loss": 0.1177,
|
| 2704 |
+
"step": 385
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 0.06175011998080307,
|
| 2708 |
+
"grad_norm": 1.4164737462997437,
|
| 2709 |
+
"learning_rate": 1e-06,
|
| 2710 |
+
"loss": 0.1148,
|
| 2711 |
+
"step": 386
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 0.06191009438489842,
|
| 2715 |
+
"grad_norm": 1.2910661697387695,
|
| 2716 |
+
"learning_rate": 1e-06,
|
| 2717 |
+
"loss": 0.1167,
|
| 2718 |
+
"step": 387
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 0.06207006878899376,
|
| 2722 |
+
"grad_norm": 0.9569298624992371,
|
| 2723 |
+
"learning_rate": 1e-06,
|
| 2724 |
+
"loss": 0.1148,
|
| 2725 |
+
"step": 388
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 0.062230043193089106,
|
| 2729 |
+
"grad_norm": 1.004881739616394,
|
| 2730 |
+
"learning_rate": 1e-06,
|
| 2731 |
+
"loss": 0.1158,
|
| 2732 |
+
"step": 389
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 0.062390017597184454,
|
| 2736 |
+
"grad_norm": 1.2112892866134644,
|
| 2737 |
+
"learning_rate": 1e-06,
|
| 2738 |
+
"loss": 0.1147,
|
| 2739 |
+
"step": 390
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 0.0625499920012798,
|
| 2743 |
+
"grad_norm": 1.5320332050323486,
|
| 2744 |
+
"learning_rate": 1e-06,
|
| 2745 |
+
"loss": 0.1168,
|
| 2746 |
+
"step": 391
|
| 2747 |
+
},
|
| 2748 |
+
{
|
| 2749 |
+
"epoch": 0.06270996640537514,
|
| 2750 |
+
"grad_norm": 1.3150376081466675,
|
| 2751 |
+
"learning_rate": 1e-06,
|
| 2752 |
+
"loss": 0.1164,
|
| 2753 |
+
"step": 392
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 0.06286994080947049,
|
| 2757 |
+
"grad_norm": 1.4928754568099976,
|
| 2758 |
+
"learning_rate": 1e-06,
|
| 2759 |
+
"loss": 0.119,
|
| 2760 |
+
"step": 393
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 0.06302991521356582,
|
| 2764 |
+
"grad_norm": 1.0719698667526245,
|
| 2765 |
+
"learning_rate": 1e-06,
|
| 2766 |
+
"loss": 0.1158,
|
| 2767 |
+
"step": 394
|
| 2768 |
+
},
|
| 2769 |
+
{
|
| 2770 |
+
"epoch": 0.06318988961766117,
|
| 2771 |
+
"grad_norm": 1.4757615327835083,
|
| 2772 |
+
"learning_rate": 1e-06,
|
| 2773 |
+
"loss": 0.1158,
|
| 2774 |
+
"step": 395
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"epoch": 0.06334986402175652,
|
| 2778 |
+
"grad_norm": 1.4807626008987427,
|
| 2779 |
+
"learning_rate": 1e-06,
|
| 2780 |
+
"loss": 0.1161,
|
| 2781 |
+
"step": 396
|
| 2782 |
+
},
|
| 2783 |
+
{
|
| 2784 |
+
"epoch": 0.06350983842585187,
|
| 2785 |
+
"grad_norm": 1.0825022459030151,
|
| 2786 |
+
"learning_rate": 1e-06,
|
| 2787 |
+
"loss": 0.1139,
|
| 2788 |
+
"step": 397
|
| 2789 |
+
},
|
| 2790 |
+
{
|
| 2791 |
+
"epoch": 0.06366981282994721,
|
| 2792 |
+
"grad_norm": 0.9957918524742126,
|
| 2793 |
+
"learning_rate": 1e-06,
|
| 2794 |
+
"loss": 0.1152,
|
| 2795 |
+
"step": 398
|
| 2796 |
+
},
|
| 2797 |
+
{
|
| 2798 |
+
"epoch": 0.06382978723404255,
|
| 2799 |
+
"grad_norm": 0.9809920191764832,
|
| 2800 |
+
"learning_rate": 1e-06,
|
| 2801 |
+
"loss": 0.1122,
|
| 2802 |
+
"step": 399
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 0.0639897616381379,
|
| 2806 |
+
"grad_norm": 1.0636423826217651,
|
| 2807 |
+
"learning_rate": 1e-06,
|
| 2808 |
+
"loss": 0.1142,
|
| 2809 |
+
"step": 400
|
| 2810 |
+
}
|
| 2811 |
+
],
|
| 2812 |
+
"logging_steps": 1.0,
|
| 2813 |
+
"max_steps": 6251,
|
| 2814 |
+
"num_input_tokens_seen": 0,
|
| 2815 |
+
"num_train_epochs": 1,
|
| 2816 |
+
"save_steps": 400,
|
| 2817 |
+
"stateful_callbacks": {
|
| 2818 |
+
"TrainerControl": {
|
| 2819 |
+
"args": {
|
| 2820 |
+
"should_epoch_stop": false,
|
| 2821 |
+
"should_evaluate": false,
|
| 2822 |
+
"should_log": false,
|
| 2823 |
+
"should_save": true,
|
| 2824 |
+
"should_training_stop": false
|
| 2825 |
+
},
|
| 2826 |
+
"attributes": {}
|
| 2827 |
+
}
|
| 2828 |
+
},
|
| 2829 |
+
"total_flos": 64804028416000.0,
|
| 2830 |
+
"train_batch_size": 16,
|
| 2831 |
+
"trial_name": null,
|
| 2832 |
+
"trial_params": null
|
| 2833 |
+
}
|