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": 0.12795905310300704,
|
| 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": 0.0003198976327575176,
|
| 13 |
+
"grad_norm": 113.54119110107422,
|
| 14 |
+
"learning_rate": 0.0,
|
| 15 |
+
"loss": 2.0136,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.0006397952655150352,
|
| 20 |
+
"grad_norm": 128.41885375976562,
|
| 21 |
+
"learning_rate": 1.5256487060115926e-07,
|
| 22 |
+
"loss": 1.8341,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.0009596928982725527,
|
| 27 |
+
"grad_norm": 128.32630920410156,
|
| 28 |
+
"learning_rate": 2.4180959883021303e-07,
|
| 29 |
+
"loss": 1.8484,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.0012795905310300703,
|
| 34 |
+
"grad_norm": 128.7578582763672,
|
| 35 |
+
"learning_rate": 3.051297412023185e-07,
|
| 36 |
+
"loss": 1.8386,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.001599488163787588,
|
| 41 |
+
"grad_norm": 131.32772827148438,
|
| 42 |
+
"learning_rate": 3.542446593416867e-07,
|
| 43 |
+
"loss": 1.8218,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.0019193857965451055,
|
| 48 |
+
"grad_norm": 136.23252868652344,
|
| 49 |
+
"learning_rate": 3.943744694313723e-07,
|
| 50 |
+
"loss": 1.7682,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.0022392834293026233,
|
| 55 |
+
"grad_norm": 117.17863464355469,
|
| 56 |
+
"learning_rate": 4.28303740415246e-07,
|
| 57 |
+
"loss": 1.3274,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.0025591810620601407,
|
| 62 |
+
"grad_norm": 114.03882598876953,
|
| 63 |
+
"learning_rate": 4.576946118034778e-07,
|
| 64 |
+
"loss": 1.2771,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.0028790786948176585,
|
| 69 |
+
"grad_norm": 112.55387878417969,
|
| 70 |
+
"learning_rate": 4.836191976604261e-07,
|
| 71 |
+
"loss": 1.2273,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.003198976327575176,
|
| 76 |
+
"grad_norm": 107.44979095458984,
|
| 77 |
+
"learning_rate": 5.06809529942846e-07,
|
| 78 |
+
"loss": 1.1365,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.0035188739603326936,
|
| 83 |
+
"grad_norm": 100.63568115234375,
|
| 84 |
+
"learning_rate": 5.277877372509582e-07,
|
| 85 |
+
"loss": 0.947,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.003838771593090211,
|
| 90 |
+
"grad_norm": 40.71220397949219,
|
| 91 |
+
"learning_rate": 5.469393400325317e-07,
|
| 92 |
+
"loss": 0.6091,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.004158669225847729,
|
| 97 |
+
"grad_norm": 36.504364013671875,
|
| 98 |
+
"learning_rate": 5.645571067655861e-07,
|
| 99 |
+
"loss": 0.5817,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.004478566858605247,
|
| 104 |
+
"grad_norm": 33.66552734375,
|
| 105 |
+
"learning_rate": 5.808686110164052e-07,
|
| 106 |
+
"loss": 0.5381,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.0047984644913627635,
|
| 111 |
+
"grad_norm": 33.83104705810547,
|
| 112 |
+
"learning_rate": 5.960542581718997e-07,
|
| 113 |
+
"loss": 0.4845,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.005118362124120281,
|
| 118 |
+
"grad_norm": 41.85721969604492,
|
| 119 |
+
"learning_rate": 6.10259482404637e-07,
|
| 120 |
+
"loss": 0.4355,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.005438259756877799,
|
| 125 |
+
"grad_norm": 51.66183853149414,
|
| 126 |
+
"learning_rate": 6.236032394624049e-07,
|
| 127 |
+
"loss": 0.3915,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.005758157389635317,
|
| 132 |
+
"grad_norm": 38.20963668823242,
|
| 133 |
+
"learning_rate": 6.361840682615853e-07,
|
| 134 |
+
"loss": 0.3051,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.006078055022392835,
|
| 139 |
+
"grad_norm": 25.591947555541992,
|
| 140 |
+
"learning_rate": 6.480845114116249e-07,
|
| 141 |
+
"loss": 0.2849,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.006397952655150352,
|
| 146 |
+
"grad_norm": 16.73761558532715,
|
| 147 |
+
"learning_rate": 6.593744005440053e-07,
|
| 148 |
+
"loss": 0.2679,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.0067178502879078695,
|
| 153 |
+
"grad_norm": 11.67692756652832,
|
| 154 |
+
"learning_rate": 6.70113339245459e-07,
|
| 155 |
+
"loss": 0.2379,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.007037747920665387,
|
| 160 |
+
"grad_norm": 7.733863830566406,
|
| 161 |
+
"learning_rate": 6.803526078521176e-07,
|
| 162 |
+
"loss": 0.213,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.007357645553422905,
|
| 167 |
+
"grad_norm": 5.249295234680176,
|
| 168 |
+
"learning_rate": 6.901366444821652e-07,
|
| 169 |
+
"loss": 0.1997,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.007677543186180422,
|
| 174 |
+
"grad_norm": 4.155821323394775,
|
| 175 |
+
"learning_rate": 6.995042106336909e-07,
|
| 176 |
+
"loss": 0.1891,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.00799744081893794,
|
| 181 |
+
"grad_norm": 3.005851984024048,
|
| 182 |
+
"learning_rate": 7.084893186833735e-07,
|
| 183 |
+
"loss": 0.1871,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.008317338451695458,
|
| 188 |
+
"grad_norm": 2.9527530670166016,
|
| 189 |
+
"learning_rate": 7.171219773667454e-07,
|
| 190 |
+
"loss": 0.1799,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.008637236084452975,
|
| 195 |
+
"grad_norm": 3.5938618183135986,
|
| 196 |
+
"learning_rate": 7.25428796490639e-07,
|
| 197 |
+
"loss": 0.184,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.008957133717210493,
|
| 202 |
+
"grad_norm": 3.1937882900238037,
|
| 203 |
+
"learning_rate": 7.334334816175645e-07,
|
| 204 |
+
"loss": 0.1748,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.009277031349968011,
|
| 209 |
+
"grad_norm": 3.320768117904663,
|
| 210 |
+
"learning_rate": 7.411572419045291e-07,
|
| 211 |
+
"loss": 0.1749,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.009596928982725527,
|
| 216 |
+
"grad_norm": 3.4186792373657227,
|
| 217 |
+
"learning_rate": 7.486191287730591e-07,
|
| 218 |
+
"loss": 0.1651,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.009916826615483045,
|
| 223 |
+
"grad_norm": 2.6862740516662598,
|
| 224 |
+
"learning_rate": 7.558363190269143e-07,
|
| 225 |
+
"loss": 0.1656,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.010236724248240563,
|
| 230 |
+
"grad_norm": 3.3762691020965576,
|
| 231 |
+
"learning_rate": 7.628243530057964e-07,
|
| 232 |
+
"loss": 0.1641,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.01055662188099808,
|
| 237 |
+
"grad_norm": 3.056905746459961,
|
| 238 |
+
"learning_rate": 7.695973360811713e-07,
|
| 239 |
+
"loss": 0.1603,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.010876519513755598,
|
| 244 |
+
"grad_norm": 2.1009628772735596,
|
| 245 |
+
"learning_rate": 7.761681100635642e-07,
|
| 246 |
+
"loss": 0.1639,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.011196417146513116,
|
| 251 |
+
"grad_norm": 2.3403608798980713,
|
| 252 |
+
"learning_rate": 7.825483997569326e-07,
|
| 253 |
+
"loss": 0.1589,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.011516314779270634,
|
| 258 |
+
"grad_norm": 2.5192835330963135,
|
| 259 |
+
"learning_rate": 7.887489388627446e-07,
|
| 260 |
+
"loss": 0.1581,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.011836212412028152,
|
| 265 |
+
"grad_norm": 2.1104068756103516,
|
| 266 |
+
"learning_rate": 7.947795786299544e-07,
|
| 267 |
+
"loss": 0.1539,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.01215611004478567,
|
| 272 |
+
"grad_norm": 1.5434317588806152,
|
| 273 |
+
"learning_rate": 8.006493820127842e-07,
|
| 274 |
+
"loss": 0.1495,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.012476007677543186,
|
| 279 |
+
"grad_norm": 2.4700214862823486,
|
| 280 |
+
"learning_rate": 8.06366705595799e-07,
|
| 281 |
+
"loss": 0.1533,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.012795905310300703,
|
| 286 |
+
"grad_norm": 2.597851276397705,
|
| 287 |
+
"learning_rate": 8.119392711451646e-07,
|
| 288 |
+
"loss": 0.1585,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.013115802943058221,
|
| 293 |
+
"grad_norm": 1.588590383529663,
|
| 294 |
+
"learning_rate": 8.173742283235395e-07,
|
| 295 |
+
"loss": 0.1529,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.013435700575815739,
|
| 300 |
+
"grad_norm": 1.60010826587677,
|
| 301 |
+
"learning_rate": 8.226782098466182e-07,
|
| 302 |
+
"loss": 0.1526,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.013755598208573257,
|
| 307 |
+
"grad_norm": 2.1502294540405273,
|
| 308 |
+
"learning_rate": 8.278573801487568e-07,
|
| 309 |
+
"loss": 0.1512,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.014075495841330775,
|
| 314 |
+
"grad_norm": 1.9794089794158936,
|
| 315 |
+
"learning_rate": 8.329174784532767e-07,
|
| 316 |
+
"loss": 0.1515,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.014395393474088292,
|
| 321 |
+
"grad_norm": 1.905648946762085,
|
| 322 |
+
"learning_rate": 8.378638570021127e-07,
|
| 323 |
+
"loss": 0.1493,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.01471529110684581,
|
| 328 |
+
"grad_norm": 2.242102861404419,
|
| 329 |
+
"learning_rate": 8.427015150833244e-07,
|
| 330 |
+
"loss": 0.1497,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.015035188739603326,
|
| 335 |
+
"grad_norm": 1.6246076822280884,
|
| 336 |
+
"learning_rate": 8.474351293988407e-07,
|
| 337 |
+
"loss": 0.1477,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.015355086372360844,
|
| 342 |
+
"grad_norm": 1.202575445175171,
|
| 343 |
+
"learning_rate": 8.520690812348502e-07,
|
| 344 |
+
"loss": 0.1433,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.015674984005118364,
|
| 349 |
+
"grad_norm": 2.040827751159668,
|
| 350 |
+
"learning_rate": 8.56607480830492e-07,
|
| 351 |
+
"loss": 0.1503,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.01599488163787588,
|
| 356 |
+
"grad_norm": 1.9062144756317139,
|
| 357 |
+
"learning_rate": 8.610541892845328e-07,
|
| 358 |
+
"loss": 0.1534,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.016314779270633396,
|
| 363 |
+
"grad_norm": 1.9834340810775757,
|
| 364 |
+
"learning_rate": 8.654128382926179e-07,
|
| 365 |
+
"loss": 0.146,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.016634676903390915,
|
| 370 |
+
"grad_norm": 1.9455609321594238,
|
| 371 |
+
"learning_rate": 8.696868479679047e-07,
|
| 372 |
+
"loss": 0.1475,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.01695457453614843,
|
| 377 |
+
"grad_norm": 1.6165645122528076,
|
| 378 |
+
"learning_rate": 8.738794429641679e-07,
|
| 379 |
+
"loss": 0.147,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.01727447216890595,
|
| 384 |
+
"grad_norm": 1.602501392364502,
|
| 385 |
+
"learning_rate": 8.779936670917984e-07,
|
| 386 |
+
"loss": 0.1462,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.017594369801663467,
|
| 391 |
+
"grad_norm": 1.7449020147323608,
|
| 392 |
+
"learning_rate": 8.820323965926451e-07,
|
| 393 |
+
"loss": 0.1477,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.017914267434420986,
|
| 398 |
+
"grad_norm": 1.7076990604400635,
|
| 399 |
+
"learning_rate": 8.859983522187238e-07,
|
| 400 |
+
"loss": 0.1466,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.018234165067178502,
|
| 405 |
+
"grad_norm": 2.1841750144958496,
|
| 406 |
+
"learning_rate": 8.89894110241838e-07,
|
| 407 |
+
"loss": 0.1469,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.018554062699936022,
|
| 412 |
+
"grad_norm": 1.696590542793274,
|
| 413 |
+
"learning_rate": 8.937221125056883e-07,
|
| 414 |
+
"loss": 0.1415,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.018873960332693538,
|
| 419 |
+
"grad_norm": 1.9890238046646118,
|
| 420 |
+
"learning_rate": 8.974846756186985e-07,
|
| 421 |
+
"loss": 0.1445,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.019193857965451054,
|
| 426 |
+
"grad_norm": 2.4267327785491943,
|
| 427 |
+
"learning_rate": 9.011839993742183e-07,
|
| 428 |
+
"loss": 0.1482,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.019513755598208574,
|
| 433 |
+
"grad_norm": 1.954588532447815,
|
| 434 |
+
"learning_rate": 9.048221744747457e-07,
|
| 435 |
+
"loss": 0.1426,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.01983365323096609,
|
| 440 |
+
"grad_norm": 1.3780263662338257,
|
| 441 |
+
"learning_rate": 9.084011896280737e-07,
|
| 442 |
+
"loss": 0.139,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.02015355086372361,
|
| 447 |
+
"grad_norm": 2.4499073028564453,
|
| 448 |
+
"learning_rate": 9.11922938075672e-07,
|
| 449 |
+
"loss": 0.1483,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.020473448496481125,
|
| 454 |
+
"grad_norm": 2.5584933757781982,
|
| 455 |
+
"learning_rate": 9.153892236069556e-07,
|
| 456 |
+
"loss": 0.1454,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.020793346129238645,
|
| 461 |
+
"grad_norm": 2.3807740211486816,
|
| 462 |
+
"learning_rate": 9.188017661072726e-07,
|
| 463 |
+
"loss": 0.1469,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.02111324376199616,
|
| 468 |
+
"grad_norm": 2.3435752391815186,
|
| 469 |
+
"learning_rate": 9.221622066823305e-07,
|
| 470 |
+
"loss": 0.149,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.02143314139475368,
|
| 475 |
+
"grad_norm": 1.9241868257522583,
|
| 476 |
+
"learning_rate": 9.25472112397281e-07,
|
| 477 |
+
"loss": 0.142,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.021753039027511197,
|
| 482 |
+
"grad_norm": 2.1067569255828857,
|
| 483 |
+
"learning_rate": 9.287329806647235e-07,
|
| 484 |
+
"loss": 0.1442,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.022072936660268713,
|
| 489 |
+
"grad_norm": 2.155319929122925,
|
| 490 |
+
"learning_rate": 9.319462433123783e-07,
|
| 491 |
+
"loss": 0.1445,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.022392834293026232,
|
| 496 |
+
"grad_norm": 2.306999444961548,
|
| 497 |
+
"learning_rate": 9.351132703580921e-07,
|
| 498 |
+
"loss": 0.1412,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.022712731925783748,
|
| 503 |
+
"grad_norm": 2.1109745502471924,
|
| 504 |
+
"learning_rate": 9.382353735170838e-07,
|
| 505 |
+
"loss": 0.144,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.023032629558541268,
|
| 510 |
+
"grad_norm": 1.6680214405059814,
|
| 511 |
+
"learning_rate": 9.41313809463904e-07,
|
| 512 |
+
"loss": 0.1389,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.023352527191298784,
|
| 517 |
+
"grad_norm": 2.816605567932129,
|
| 518 |
+
"learning_rate": 9.443497828694076e-07,
|
| 519 |
+
"loss": 0.1468,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.023672424824056303,
|
| 524 |
+
"grad_norm": 2.1079049110412598,
|
| 525 |
+
"learning_rate": 9.473444492311138e-07,
|
| 526 |
+
"loss": 0.1401,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.02399232245681382,
|
| 531 |
+
"grad_norm": 2.027454376220703,
|
| 532 |
+
"learning_rate": 9.502989175135864e-07,
|
| 533 |
+
"loss": 0.1384,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.02431222008957134,
|
| 538 |
+
"grad_norm": 1.5141581296920776,
|
| 539 |
+
"learning_rate": 9.532142526139435e-07,
|
| 540 |
+
"loss": 0.14,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.024632117722328855,
|
| 545 |
+
"grad_norm": 1.7876137495040894,
|
| 546 |
+
"learning_rate": 9.560914776662043e-07,
|
| 547 |
+
"loss": 0.1398,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.02495201535508637,
|
| 552 |
+
"grad_norm": 2.656886577606201,
|
| 553 |
+
"learning_rate": 9.589315761969583e-07,
|
| 554 |
+
"loss": 0.1414,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.02527191298784389,
|
| 559 |
+
"grad_norm": 1.7541368007659912,
|
| 560 |
+
"learning_rate": 9.617354941437189e-07,
|
| 561 |
+
"loss": 0.139,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.025591810620601407,
|
| 566 |
+
"grad_norm": 1.8313452005386353,
|
| 567 |
+
"learning_rate": 9.645041417463238e-07,
|
| 568 |
+
"loss": 0.1373,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.025911708253358926,
|
| 573 |
+
"grad_norm": 1.616778016090393,
|
| 574 |
+
"learning_rate": 9.672383953208521e-07,
|
| 575 |
+
"loss": 0.1399,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.026231605886116442,
|
| 580 |
+
"grad_norm": 1.890690565109253,
|
| 581 |
+
"learning_rate": 9.699390989246989e-07,
|
| 582 |
+
"loss": 0.1347,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.026551503518873962,
|
| 587 |
+
"grad_norm": 1.1803064346313477,
|
| 588 |
+
"learning_rate": 9.726070659207317e-07,
|
| 589 |
+
"loss": 0.1386,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.026871401151631478,
|
| 594 |
+
"grad_norm": 1.8215192556381226,
|
| 595 |
+
"learning_rate": 9.752430804477774e-07,
|
| 596 |
+
"loss": 0.1388,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.027191298784388994,
|
| 601 |
+
"grad_norm": 1.857752799987793,
|
| 602 |
+
"learning_rate": 9.778478988040916e-07,
|
| 603 |
+
"loss": 0.1365,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.027511196417146513,
|
| 608 |
+
"grad_norm": 1.622135043144226,
|
| 609 |
+
"learning_rate": 9.804222507499162e-07,
|
| 610 |
+
"loss": 0.1395,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.02783109404990403,
|
| 615 |
+
"grad_norm": 1.308622121810913,
|
| 616 |
+
"learning_rate": 9.82966840734742e-07,
|
| 617 |
+
"loss": 0.1379,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.02815099168266155,
|
| 622 |
+
"grad_norm": 1.0815646648406982,
|
| 623 |
+
"learning_rate": 9.854823490544362e-07,
|
| 624 |
+
"loss": 0.1374,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.028470889315419065,
|
| 629 |
+
"grad_norm": 1.7514983415603638,
|
| 630 |
+
"learning_rate": 9.879694329429897e-07,
|
| 631 |
+
"loss": 0.1428,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.028790786948176585,
|
| 636 |
+
"grad_norm": 1.101178526878357,
|
| 637 |
+
"learning_rate": 9.904287276032721e-07,
|
| 638 |
+
"loss": 0.1347,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.0291106845809341,
|
| 643 |
+
"grad_norm": 1.3245909214019775,
|
| 644 |
+
"learning_rate": 9.92860847180832e-07,
|
| 645 |
+
"loss": 0.1395,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.02943058221369162,
|
| 650 |
+
"grad_norm": 1.598402500152588,
|
| 651 |
+
"learning_rate": 9.952663856844837e-07,
|
| 652 |
+
"loss": 0.1357,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.029750479846449136,
|
| 657 |
+
"grad_norm": 1.5450119972229004,
|
| 658 |
+
"learning_rate": 9.976459178571275e-07,
|
| 659 |
+
"loss": 0.1361,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.030070377479206652,
|
| 664 |
+
"grad_norm": 1.095211386680603,
|
| 665 |
+
"learning_rate": 1e-06,
|
| 666 |
+
"loss": 0.1338,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.030390275111964172,
|
| 671 |
+
"grad_norm": 1.2551831007003784,
|
| 672 |
+
"learning_rate": 1e-06,
|
| 673 |
+
"loss": 0.1318,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.030710172744721688,
|
| 678 |
+
"grad_norm": 1.2607723474502563,
|
| 679 |
+
"learning_rate": 1e-06,
|
| 680 |
+
"loss": 0.1381,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.031030070377479207,
|
| 685 |
+
"grad_norm": 1.3020524978637695,
|
| 686 |
+
"learning_rate": 1e-06,
|
| 687 |
+
"loss": 0.1362,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.03134996801023673,
|
| 692 |
+
"grad_norm": 1.3852208852767944,
|
| 693 |
+
"learning_rate": 1e-06,
|
| 694 |
+
"loss": 0.1337,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.03166986564299424,
|
| 699 |
+
"grad_norm": 1.7191814184188843,
|
| 700 |
+
"learning_rate": 1e-06,
|
| 701 |
+
"loss": 0.1414,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.03198976327575176,
|
| 706 |
+
"grad_norm": 1.4027682542800903,
|
| 707 |
+
"learning_rate": 1e-06,
|
| 708 |
+
"loss": 0.1363,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.03230966090850928,
|
| 713 |
+
"grad_norm": 1.8776895999908447,
|
| 714 |
+
"learning_rate": 1e-06,
|
| 715 |
+
"loss": 0.1383,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.03262955854126679,
|
| 720 |
+
"grad_norm": 1.1314609050750732,
|
| 721 |
+
"learning_rate": 1e-06,
|
| 722 |
+
"loss": 0.1371,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.03294945617402431,
|
| 727 |
+
"grad_norm": 1.266782283782959,
|
| 728 |
+
"learning_rate": 1e-06,
|
| 729 |
+
"loss": 0.1355,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.03326935380678183,
|
| 734 |
+
"grad_norm": 1.3317842483520508,
|
| 735 |
+
"learning_rate": 1e-06,
|
| 736 |
+
"loss": 0.1335,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.03358925143953935,
|
| 741 |
+
"grad_norm": 1.269689679145813,
|
| 742 |
+
"learning_rate": 1e-06,
|
| 743 |
+
"loss": 0.1329,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.03390914907229686,
|
| 748 |
+
"grad_norm": 1.3745169639587402,
|
| 749 |
+
"learning_rate": 1e-06,
|
| 750 |
+
"loss": 0.1339,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.03422904670505438,
|
| 755 |
+
"grad_norm": 1.8616901636123657,
|
| 756 |
+
"learning_rate": 1e-06,
|
| 757 |
+
"loss": 0.1362,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.0345489443378119,
|
| 762 |
+
"grad_norm": 1.5136314630508423,
|
| 763 |
+
"learning_rate": 1e-06,
|
| 764 |
+
"loss": 0.136,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.03486884197056942,
|
| 769 |
+
"grad_norm": 1.3255293369293213,
|
| 770 |
+
"learning_rate": 1e-06,
|
| 771 |
+
"loss": 0.1362,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.035188739603326934,
|
| 776 |
+
"grad_norm": 1.1577121019363403,
|
| 777 |
+
"learning_rate": 1e-06,
|
| 778 |
+
"loss": 0.1344,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.03550863723608445,
|
| 783 |
+
"grad_norm": 0.9765655398368835,
|
| 784 |
+
"learning_rate": 1e-06,
|
| 785 |
+
"loss": 0.1325,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.03582853486884197,
|
| 790 |
+
"grad_norm": 1.2133572101593018,
|
| 791 |
+
"learning_rate": 1e-06,
|
| 792 |
+
"loss": 0.1344,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.036148432501599485,
|
| 797 |
+
"grad_norm": 1.2999507188796997,
|
| 798 |
+
"learning_rate": 1e-06,
|
| 799 |
+
"loss": 0.1322,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.036468330134357005,
|
| 804 |
+
"grad_norm": 1.1714322566986084,
|
| 805 |
+
"learning_rate": 1e-06,
|
| 806 |
+
"loss": 0.1315,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.036788227767114524,
|
| 811 |
+
"grad_norm": 2.234097957611084,
|
| 812 |
+
"learning_rate": 1e-06,
|
| 813 |
+
"loss": 0.1349,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.037108125399872044,
|
| 818 |
+
"grad_norm": 1.8060959577560425,
|
| 819 |
+
"learning_rate": 1e-06,
|
| 820 |
+
"loss": 0.1342,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.03742802303262956,
|
| 825 |
+
"grad_norm": 1.292160987854004,
|
| 826 |
+
"learning_rate": 1e-06,
|
| 827 |
+
"loss": 0.1328,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.037747920665387076,
|
| 832 |
+
"grad_norm": 1.8365767002105713,
|
| 833 |
+
"learning_rate": 1e-06,
|
| 834 |
+
"loss": 0.1333,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.038067818298144596,
|
| 839 |
+
"grad_norm": 1.9391571283340454,
|
| 840 |
+
"learning_rate": 1e-06,
|
| 841 |
+
"loss": 0.1361,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.03838771593090211,
|
| 846 |
+
"grad_norm": 1.338565707206726,
|
| 847 |
+
"learning_rate": 1e-06,
|
| 848 |
+
"loss": 0.1341,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.03870761356365963,
|
| 853 |
+
"grad_norm": 1.984100341796875,
|
| 854 |
+
"learning_rate": 1e-06,
|
| 855 |
+
"loss": 0.1333,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.03902751119641715,
|
| 860 |
+
"grad_norm": 1.6623727083206177,
|
| 861 |
+
"learning_rate": 1e-06,
|
| 862 |
+
"loss": 0.1314,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.03934740882917467,
|
| 867 |
+
"grad_norm": 1.2958471775054932,
|
| 868 |
+
"learning_rate": 1e-06,
|
| 869 |
+
"loss": 0.131,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.03966730646193218,
|
| 874 |
+
"grad_norm": 1.5842586755752563,
|
| 875 |
+
"learning_rate": 1e-06,
|
| 876 |
+
"loss": 0.1343,
|
| 877 |
+
"step": 124
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.0399872040946897,
|
| 881 |
+
"grad_norm": 2.5024752616882324,
|
| 882 |
+
"learning_rate": 1e-06,
|
| 883 |
+
"loss": 0.1357,
|
| 884 |
+
"step": 125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.04030710172744722,
|
| 888 |
+
"grad_norm": 1.8107951879501343,
|
| 889 |
+
"learning_rate": 1e-06,
|
| 890 |
+
"loss": 0.1329,
|
| 891 |
+
"step": 126
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.04062699936020473,
|
| 895 |
+
"grad_norm": 1.2203110456466675,
|
| 896 |
+
"learning_rate": 1e-06,
|
| 897 |
+
"loss": 0.1319,
|
| 898 |
+
"step": 127
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.04094689699296225,
|
| 902 |
+
"grad_norm": 1.1050554513931274,
|
| 903 |
+
"learning_rate": 1e-06,
|
| 904 |
+
"loss": 0.1316,
|
| 905 |
+
"step": 128
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.04126679462571977,
|
| 909 |
+
"grad_norm": 1.1994340419769287,
|
| 910 |
+
"learning_rate": 1e-06,
|
| 911 |
+
"loss": 0.1311,
|
| 912 |
+
"step": 129
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.04158669225847729,
|
| 916 |
+
"grad_norm": 1.7190176248550415,
|
| 917 |
+
"learning_rate": 1e-06,
|
| 918 |
+
"loss": 0.1303,
|
| 919 |
+
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.0419065898912348,
|
| 923 |
+
"grad_norm": 1.3263739347457886,
|
| 924 |
+
"learning_rate": 1e-06,
|
| 925 |
+
"loss": 0.1343,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.04222648752399232,
|
| 930 |
+
"grad_norm": 1.6986573934555054,
|
| 931 |
+
"learning_rate": 1e-06,
|
| 932 |
+
"loss": 0.1334,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.04254638515674984,
|
| 937 |
+
"grad_norm": 1.0033743381500244,
|
| 938 |
+
"learning_rate": 1e-06,
|
| 939 |
+
"loss": 0.1312,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.04286628278950736,
|
| 944 |
+
"grad_norm": 0.9940341114997864,
|
| 945 |
+
"learning_rate": 1e-06,
|
| 946 |
+
"loss": 0.1291,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.04318618042226487,
|
| 951 |
+
"grad_norm": 1.1875592470169067,
|
| 952 |
+
"learning_rate": 1e-06,
|
| 953 |
+
"loss": 0.1314,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.04350607805502239,
|
| 958 |
+
"grad_norm": 1.128090739250183,
|
| 959 |
+
"learning_rate": 1e-06,
|
| 960 |
+
"loss": 0.131,
|
| 961 |
+
"step": 136
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.04382597568777991,
|
| 965 |
+
"grad_norm": 1.2481768131256104,
|
| 966 |
+
"learning_rate": 1e-06,
|
| 967 |
+
"loss": 0.1312,
|
| 968 |
+
"step": 137
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.044145873320537425,
|
| 972 |
+
"grad_norm": 1.791835069656372,
|
| 973 |
+
"learning_rate": 1e-06,
|
| 974 |
+
"loss": 0.1278,
|
| 975 |
+
"step": 138
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.044465770953294945,
|
| 979 |
+
"grad_norm": 1.1713182926177979,
|
| 980 |
+
"learning_rate": 1e-06,
|
| 981 |
+
"loss": 0.1301,
|
| 982 |
+
"step": 139
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.044785668586052464,
|
| 986 |
+
"grad_norm": 1.3120388984680176,
|
| 987 |
+
"learning_rate": 1e-06,
|
| 988 |
+
"loss": 0.1294,
|
| 989 |
+
"step": 140
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.045105566218809984,
|
| 993 |
+
"grad_norm": 1.001320481300354,
|
| 994 |
+
"learning_rate": 1e-06,
|
| 995 |
+
"loss": 0.1297,
|
| 996 |
+
"step": 141
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.045425463851567496,
|
| 1000 |
+
"grad_norm": 1.0305743217468262,
|
| 1001 |
+
"learning_rate": 1e-06,
|
| 1002 |
+
"loss": 0.1311,
|
| 1003 |
+
"step": 142
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.045745361484325016,
|
| 1007 |
+
"grad_norm": 1.3738396167755127,
|
| 1008 |
+
"learning_rate": 1e-06,
|
| 1009 |
+
"loss": 0.1329,
|
| 1010 |
+
"step": 143
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.046065259117082535,
|
| 1014 |
+
"grad_norm": 0.9676306247711182,
|
| 1015 |
+
"learning_rate": 1e-06,
|
| 1016 |
+
"loss": 0.1287,
|
| 1017 |
+
"step": 144
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.04638515674984005,
|
| 1021 |
+
"grad_norm": 0.968252420425415,
|
| 1022 |
+
"learning_rate": 1e-06,
|
| 1023 |
+
"loss": 0.1292,
|
| 1024 |
+
"step": 145
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.04670505438259757,
|
| 1028 |
+
"grad_norm": 1.099440574645996,
|
| 1029 |
+
"learning_rate": 1e-06,
|
| 1030 |
+
"loss": 0.1288,
|
| 1031 |
+
"step": 146
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.04702495201535509,
|
| 1035 |
+
"grad_norm": 1.2246148586273193,
|
| 1036 |
+
"learning_rate": 1e-06,
|
| 1037 |
+
"loss": 0.1305,
|
| 1038 |
+
"step": 147
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.04734484964811261,
|
| 1042 |
+
"grad_norm": 1.2801058292388916,
|
| 1043 |
+
"learning_rate": 1e-06,
|
| 1044 |
+
"loss": 0.1279,
|
| 1045 |
+
"step": 148
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.04766474728087012,
|
| 1049 |
+
"grad_norm": 1.2214787006378174,
|
| 1050 |
+
"learning_rate": 1e-06,
|
| 1051 |
+
"loss": 0.1315,
|
| 1052 |
+
"step": 149
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.04798464491362764,
|
| 1056 |
+
"grad_norm": 1.6158666610717773,
|
| 1057 |
+
"learning_rate": 1e-06,
|
| 1058 |
+
"loss": 0.1309,
|
| 1059 |
+
"step": 150
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.04830454254638516,
|
| 1063 |
+
"grad_norm": 1.0991178750991821,
|
| 1064 |
+
"learning_rate": 1e-06,
|
| 1065 |
+
"loss": 0.1317,
|
| 1066 |
+
"step": 151
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.04862444017914268,
|
| 1070 |
+
"grad_norm": 1.0115094184875488,
|
| 1071 |
+
"learning_rate": 1e-06,
|
| 1072 |
+
"loss": 0.1297,
|
| 1073 |
+
"step": 152
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.04894433781190019,
|
| 1077 |
+
"grad_norm": 1.3744951486587524,
|
| 1078 |
+
"learning_rate": 1e-06,
|
| 1079 |
+
"loss": 0.1297,
|
| 1080 |
+
"step": 153
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.04926423544465771,
|
| 1084 |
+
"grad_norm": 1.0412648916244507,
|
| 1085 |
+
"learning_rate": 1e-06,
|
| 1086 |
+
"loss": 0.1322,
|
| 1087 |
+
"step": 154
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.04958413307741523,
|
| 1091 |
+
"grad_norm": 1.713841438293457,
|
| 1092 |
+
"learning_rate": 1e-06,
|
| 1093 |
+
"loss": 0.1268,
|
| 1094 |
+
"step": 155
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.04990403071017274,
|
| 1098 |
+
"grad_norm": 1.3019500970840454,
|
| 1099 |
+
"learning_rate": 1e-06,
|
| 1100 |
+
"loss": 0.1278,
|
| 1101 |
+
"step": 156
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.05022392834293026,
|
| 1105 |
+
"grad_norm": 1.5901345014572144,
|
| 1106 |
+
"learning_rate": 1e-06,
|
| 1107 |
+
"loss": 0.1301,
|
| 1108 |
+
"step": 157
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.05054382597568778,
|
| 1112 |
+
"grad_norm": 1.480391502380371,
|
| 1113 |
+
"learning_rate": 1e-06,
|
| 1114 |
+
"loss": 0.132,
|
| 1115 |
+
"step": 158
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.0508637236084453,
|
| 1119 |
+
"grad_norm": 1.2169982194900513,
|
| 1120 |
+
"learning_rate": 1e-06,
|
| 1121 |
+
"loss": 0.1271,
|
| 1122 |
+
"step": 159
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.05118362124120281,
|
| 1126 |
+
"grad_norm": 1.3028303384780884,
|
| 1127 |
+
"learning_rate": 1e-06,
|
| 1128 |
+
"loss": 0.126,
|
| 1129 |
+
"step": 160
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.05150351887396033,
|
| 1133 |
+
"grad_norm": 1.2110769748687744,
|
| 1134 |
+
"learning_rate": 1e-06,
|
| 1135 |
+
"loss": 0.1293,
|
| 1136 |
+
"step": 161
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.05182341650671785,
|
| 1140 |
+
"grad_norm": 1.1032134294509888,
|
| 1141 |
+
"learning_rate": 1e-06,
|
| 1142 |
+
"loss": 0.1279,
|
| 1143 |
+
"step": 162
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.052143314139475365,
|
| 1147 |
+
"grad_norm": 1.0820825099945068,
|
| 1148 |
+
"learning_rate": 1e-06,
|
| 1149 |
+
"loss": 0.1269,
|
| 1150 |
+
"step": 163
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.052463211772232884,
|
| 1154 |
+
"grad_norm": 1.3737691640853882,
|
| 1155 |
+
"learning_rate": 1e-06,
|
| 1156 |
+
"loss": 0.1279,
|
| 1157 |
+
"step": 164
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.052783109404990404,
|
| 1161 |
+
"grad_norm": 1.4791896343231201,
|
| 1162 |
+
"learning_rate": 1e-06,
|
| 1163 |
+
"loss": 0.1284,
|
| 1164 |
+
"step": 165
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.053103007037747924,
|
| 1168 |
+
"grad_norm": 1.1859378814697266,
|
| 1169 |
+
"learning_rate": 1e-06,
|
| 1170 |
+
"loss": 0.1306,
|
| 1171 |
+
"step": 166
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.053422904670505436,
|
| 1175 |
+
"grad_norm": 1.661341905593872,
|
| 1176 |
+
"learning_rate": 1e-06,
|
| 1177 |
+
"loss": 0.1284,
|
| 1178 |
+
"step": 167
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.053742802303262956,
|
| 1182 |
+
"grad_norm": 1.1671792268753052,
|
| 1183 |
+
"learning_rate": 1e-06,
|
| 1184 |
+
"loss": 0.1254,
|
| 1185 |
+
"step": 168
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.054062699936020475,
|
| 1189 |
+
"grad_norm": 1.5143752098083496,
|
| 1190 |
+
"learning_rate": 1e-06,
|
| 1191 |
+
"loss": 0.1324,
|
| 1192 |
+
"step": 169
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 0.05438259756877799,
|
| 1196 |
+
"grad_norm": 1.443939208984375,
|
| 1197 |
+
"learning_rate": 1e-06,
|
| 1198 |
+
"loss": 0.129,
|
| 1199 |
+
"step": 170
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.05470249520153551,
|
| 1203 |
+
"grad_norm": 1.7711951732635498,
|
| 1204 |
+
"learning_rate": 1e-06,
|
| 1205 |
+
"loss": 0.1304,
|
| 1206 |
+
"step": 171
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 0.05502239283429303,
|
| 1210 |
+
"grad_norm": 1.1903679370880127,
|
| 1211 |
+
"learning_rate": 1e-06,
|
| 1212 |
+
"loss": 0.1294,
|
| 1213 |
+
"step": 172
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 0.055342290467050546,
|
| 1217 |
+
"grad_norm": 0.9061336517333984,
|
| 1218 |
+
"learning_rate": 1e-06,
|
| 1219 |
+
"loss": 0.1284,
|
| 1220 |
+
"step": 173
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 0.05566218809980806,
|
| 1224 |
+
"grad_norm": 1.3770477771759033,
|
| 1225 |
+
"learning_rate": 1e-06,
|
| 1226 |
+
"loss": 0.129,
|
| 1227 |
+
"step": 174
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 0.05598208573256558,
|
| 1231 |
+
"grad_norm": 1.2672255039215088,
|
| 1232 |
+
"learning_rate": 1e-06,
|
| 1233 |
+
"loss": 0.1306,
|
| 1234 |
+
"step": 175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 0.0563019833653231,
|
| 1238 |
+
"grad_norm": 1.7260921001434326,
|
| 1239 |
+
"learning_rate": 1e-06,
|
| 1240 |
+
"loss": 0.1326,
|
| 1241 |
+
"step": 176
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.05662188099808062,
|
| 1245 |
+
"grad_norm": 0.9481382966041565,
|
| 1246 |
+
"learning_rate": 1e-06,
|
| 1247 |
+
"loss": 0.1284,
|
| 1248 |
+
"step": 177
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.05694177863083813,
|
| 1252 |
+
"grad_norm": 1.7979446649551392,
|
| 1253 |
+
"learning_rate": 1e-06,
|
| 1254 |
+
"loss": 0.1337,
|
| 1255 |
+
"step": 178
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.05726167626359565,
|
| 1259 |
+
"grad_norm": 1.1924407482147217,
|
| 1260 |
+
"learning_rate": 1e-06,
|
| 1261 |
+
"loss": 0.1274,
|
| 1262 |
+
"step": 179
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.05758157389635317,
|
| 1266 |
+
"grad_norm": 1.0101194381713867,
|
| 1267 |
+
"learning_rate": 1e-06,
|
| 1268 |
+
"loss": 0.128,
|
| 1269 |
+
"step": 180
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.05790147152911068,
|
| 1273 |
+
"grad_norm": 1.4568109512329102,
|
| 1274 |
+
"learning_rate": 1e-06,
|
| 1275 |
+
"loss": 0.129,
|
| 1276 |
+
"step": 181
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 0.0582213691618682,
|
| 1280 |
+
"grad_norm": 1.26350736618042,
|
| 1281 |
+
"learning_rate": 1e-06,
|
| 1282 |
+
"loss": 0.1297,
|
| 1283 |
+
"step": 182
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.05854126679462572,
|
| 1287 |
+
"grad_norm": 1.0530216693878174,
|
| 1288 |
+
"learning_rate": 1e-06,
|
| 1289 |
+
"loss": 0.1293,
|
| 1290 |
+
"step": 183
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 0.05886116442738324,
|
| 1294 |
+
"grad_norm": 1.2842745780944824,
|
| 1295 |
+
"learning_rate": 1e-06,
|
| 1296 |
+
"loss": 0.1245,
|
| 1297 |
+
"step": 184
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 0.05918106206014075,
|
| 1301 |
+
"grad_norm": 1.5984883308410645,
|
| 1302 |
+
"learning_rate": 1e-06,
|
| 1303 |
+
"loss": 0.1283,
|
| 1304 |
+
"step": 185
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.05950095969289827,
|
| 1308 |
+
"grad_norm": 1.6221599578857422,
|
| 1309 |
+
"learning_rate": 1e-06,
|
| 1310 |
+
"loss": 0.1253,
|
| 1311 |
+
"step": 186
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.05982085732565579,
|
| 1315 |
+
"grad_norm": 1.0363227128982544,
|
| 1316 |
+
"learning_rate": 1e-06,
|
| 1317 |
+
"loss": 0.1263,
|
| 1318 |
+
"step": 187
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 0.060140754958413305,
|
| 1322 |
+
"grad_norm": 1.209296464920044,
|
| 1323 |
+
"learning_rate": 1e-06,
|
| 1324 |
+
"loss": 0.1265,
|
| 1325 |
+
"step": 188
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 0.060460652591170824,
|
| 1329 |
+
"grad_norm": 1.3351823091506958,
|
| 1330 |
+
"learning_rate": 1e-06,
|
| 1331 |
+
"loss": 0.1251,
|
| 1332 |
+
"step": 189
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 0.060780550223928344,
|
| 1336 |
+
"grad_norm": 1.3078138828277588,
|
| 1337 |
+
"learning_rate": 1e-06,
|
| 1338 |
+
"loss": 0.1242,
|
| 1339 |
+
"step": 190
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 0.06110044785668586,
|
| 1343 |
+
"grad_norm": 1.3770887851715088,
|
| 1344 |
+
"learning_rate": 1e-06,
|
| 1345 |
+
"loss": 0.1258,
|
| 1346 |
+
"step": 191
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 0.061420345489443376,
|
| 1350 |
+
"grad_norm": 1.2662371397018433,
|
| 1351 |
+
"learning_rate": 1e-06,
|
| 1352 |
+
"loss": 0.1254,
|
| 1353 |
+
"step": 192
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 0.061740243122200895,
|
| 1357 |
+
"grad_norm": 1.274051308631897,
|
| 1358 |
+
"learning_rate": 1e-06,
|
| 1359 |
+
"loss": 0.1265,
|
| 1360 |
+
"step": 193
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 0.062060140754958415,
|
| 1364 |
+
"grad_norm": 1.4379581212997437,
|
| 1365 |
+
"learning_rate": 1e-06,
|
| 1366 |
+
"loss": 0.1268,
|
| 1367 |
+
"step": 194
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 0.06238003838771593,
|
| 1371 |
+
"grad_norm": 1.27424955368042,
|
| 1372 |
+
"learning_rate": 1e-06,
|
| 1373 |
+
"loss": 0.1266,
|
| 1374 |
+
"step": 195
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 0.06269993602047345,
|
| 1378 |
+
"grad_norm": 1.2419888973236084,
|
| 1379 |
+
"learning_rate": 1e-06,
|
| 1380 |
+
"loss": 0.1293,
|
| 1381 |
+
"step": 196
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 0.06301983365323097,
|
| 1385 |
+
"grad_norm": 1.0367040634155273,
|
| 1386 |
+
"learning_rate": 1e-06,
|
| 1387 |
+
"loss": 0.1272,
|
| 1388 |
+
"step": 197
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 0.06333973128598848,
|
| 1392 |
+
"grad_norm": 1.3515957593917847,
|
| 1393 |
+
"learning_rate": 1e-06,
|
| 1394 |
+
"loss": 0.1267,
|
| 1395 |
+
"step": 198
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 0.063659628918746,
|
| 1399 |
+
"grad_norm": 1.2155792713165283,
|
| 1400 |
+
"learning_rate": 1e-06,
|
| 1401 |
+
"loss": 0.128,
|
| 1402 |
+
"step": 199
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 0.06397952655150352,
|
| 1406 |
+
"grad_norm": 1.1291899681091309,
|
| 1407 |
+
"learning_rate": 1e-06,
|
| 1408 |
+
"loss": 0.1273,
|
| 1409 |
+
"step": 200
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 0.06429942418426103,
|
| 1413 |
+
"grad_norm": 0.9845306873321533,
|
| 1414 |
+
"learning_rate": 1e-06,
|
| 1415 |
+
"loss": 0.1246,
|
| 1416 |
+
"step": 201
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 0.06461932181701856,
|
| 1420 |
+
"grad_norm": 1.1045749187469482,
|
| 1421 |
+
"learning_rate": 1e-06,
|
| 1422 |
+
"loss": 0.1243,
|
| 1423 |
+
"step": 202
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 0.06493921944977607,
|
| 1427 |
+
"grad_norm": 1.3128234148025513,
|
| 1428 |
+
"learning_rate": 1e-06,
|
| 1429 |
+
"loss": 0.1296,
|
| 1430 |
+
"step": 203
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 0.06525911708253358,
|
| 1434 |
+
"grad_norm": 1.5253573656082153,
|
| 1435 |
+
"learning_rate": 1e-06,
|
| 1436 |
+
"loss": 0.1288,
|
| 1437 |
+
"step": 204
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 0.06557901471529111,
|
| 1441 |
+
"grad_norm": 1.2290501594543457,
|
| 1442 |
+
"learning_rate": 1e-06,
|
| 1443 |
+
"loss": 0.126,
|
| 1444 |
+
"step": 205
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 0.06589891234804862,
|
| 1448 |
+
"grad_norm": 1.2679433822631836,
|
| 1449 |
+
"learning_rate": 1e-06,
|
| 1450 |
+
"loss": 0.1277,
|
| 1451 |
+
"step": 206
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 0.06621880998080615,
|
| 1455 |
+
"grad_norm": 0.9389240741729736,
|
| 1456 |
+
"learning_rate": 1e-06,
|
| 1457 |
+
"loss": 0.1254,
|
| 1458 |
+
"step": 207
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 0.06653870761356366,
|
| 1462 |
+
"grad_norm": 0.9384240508079529,
|
| 1463 |
+
"learning_rate": 1e-06,
|
| 1464 |
+
"loss": 0.1253,
|
| 1465 |
+
"step": 208
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 0.06685860524632117,
|
| 1469 |
+
"grad_norm": 1.11676824092865,
|
| 1470 |
+
"learning_rate": 1e-06,
|
| 1471 |
+
"loss": 0.1237,
|
| 1472 |
+
"step": 209
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 0.0671785028790787,
|
| 1476 |
+
"grad_norm": 1.1166993379592896,
|
| 1477 |
+
"learning_rate": 1e-06,
|
| 1478 |
+
"loss": 0.1263,
|
| 1479 |
+
"step": 210
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 0.06749840051183621,
|
| 1483 |
+
"grad_norm": 1.642757534980774,
|
| 1484 |
+
"learning_rate": 1e-06,
|
| 1485 |
+
"loss": 0.1258,
|
| 1486 |
+
"step": 211
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 0.06781829814459372,
|
| 1490 |
+
"grad_norm": 1.0999261140823364,
|
| 1491 |
+
"learning_rate": 1e-06,
|
| 1492 |
+
"loss": 0.1223,
|
| 1493 |
+
"step": 212
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 0.06813819577735125,
|
| 1497 |
+
"grad_norm": 1.2551041841506958,
|
| 1498 |
+
"learning_rate": 1e-06,
|
| 1499 |
+
"loss": 0.1252,
|
| 1500 |
+
"step": 213
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 0.06845809341010876,
|
| 1504 |
+
"grad_norm": 1.1348063945770264,
|
| 1505 |
+
"learning_rate": 1e-06,
|
| 1506 |
+
"loss": 0.1236,
|
| 1507 |
+
"step": 214
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 0.06877799104286628,
|
| 1511 |
+
"grad_norm": 1.4428297281265259,
|
| 1512 |
+
"learning_rate": 1e-06,
|
| 1513 |
+
"loss": 0.1286,
|
| 1514 |
+
"step": 215
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 0.0690978886756238,
|
| 1518 |
+
"grad_norm": 1.3114149570465088,
|
| 1519 |
+
"learning_rate": 1e-06,
|
| 1520 |
+
"loss": 0.1243,
|
| 1521 |
+
"step": 216
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 0.06941778630838132,
|
| 1525 |
+
"grad_norm": 1.1751904487609863,
|
| 1526 |
+
"learning_rate": 1e-06,
|
| 1527 |
+
"loss": 0.1248,
|
| 1528 |
+
"step": 217
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 0.06973768394113884,
|
| 1532 |
+
"grad_norm": 1.2301567792892456,
|
| 1533 |
+
"learning_rate": 1e-06,
|
| 1534 |
+
"loss": 0.125,
|
| 1535 |
+
"step": 218
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 0.07005758157389635,
|
| 1539 |
+
"grad_norm": 1.0425200462341309,
|
| 1540 |
+
"learning_rate": 1e-06,
|
| 1541 |
+
"loss": 0.1257,
|
| 1542 |
+
"step": 219
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 0.07037747920665387,
|
| 1546 |
+
"grad_norm": 1.0965701341629028,
|
| 1547 |
+
"learning_rate": 1e-06,
|
| 1548 |
+
"loss": 0.1234,
|
| 1549 |
+
"step": 220
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 0.0706973768394114,
|
| 1553 |
+
"grad_norm": 1.5265578031539917,
|
| 1554 |
+
"learning_rate": 1e-06,
|
| 1555 |
+
"loss": 0.1262,
|
| 1556 |
+
"step": 221
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 0.0710172744721689,
|
| 1560 |
+
"grad_norm": 1.121492624282837,
|
| 1561 |
+
"learning_rate": 1e-06,
|
| 1562 |
+
"loss": 0.1224,
|
| 1563 |
+
"step": 222
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 0.07133717210492642,
|
| 1567 |
+
"grad_norm": 1.3932403326034546,
|
| 1568 |
+
"learning_rate": 1e-06,
|
| 1569 |
+
"loss": 0.1239,
|
| 1570 |
+
"step": 223
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 0.07165706973768395,
|
| 1574 |
+
"grad_norm": 1.4684503078460693,
|
| 1575 |
+
"learning_rate": 1e-06,
|
| 1576 |
+
"loss": 0.1246,
|
| 1577 |
+
"step": 224
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 0.07197696737044146,
|
| 1581 |
+
"grad_norm": 0.968016505241394,
|
| 1582 |
+
"learning_rate": 1e-06,
|
| 1583 |
+
"loss": 0.1226,
|
| 1584 |
+
"step": 225
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 0.07229686500319897,
|
| 1588 |
+
"grad_norm": 0.9670442938804626,
|
| 1589 |
+
"learning_rate": 1e-06,
|
| 1590 |
+
"loss": 0.1253,
|
| 1591 |
+
"step": 226
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 0.0726167626359565,
|
| 1595 |
+
"grad_norm": 1.1691622734069824,
|
| 1596 |
+
"learning_rate": 1e-06,
|
| 1597 |
+
"loss": 0.1272,
|
| 1598 |
+
"step": 227
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 0.07293666026871401,
|
| 1602 |
+
"grad_norm": 1.1516366004943848,
|
| 1603 |
+
"learning_rate": 1e-06,
|
| 1604 |
+
"loss": 0.1242,
|
| 1605 |
+
"step": 228
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 0.07325655790147152,
|
| 1609 |
+
"grad_norm": 1.1938285827636719,
|
| 1610 |
+
"learning_rate": 1e-06,
|
| 1611 |
+
"loss": 0.127,
|
| 1612 |
+
"step": 229
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 0.07357645553422905,
|
| 1616 |
+
"grad_norm": 1.0785573720932007,
|
| 1617 |
+
"learning_rate": 1e-06,
|
| 1618 |
+
"loss": 0.1285,
|
| 1619 |
+
"step": 230
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 0.07389635316698656,
|
| 1623 |
+
"grad_norm": 1.093993067741394,
|
| 1624 |
+
"learning_rate": 1e-06,
|
| 1625 |
+
"loss": 0.1238,
|
| 1626 |
+
"step": 231
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 0.07421625079974409,
|
| 1630 |
+
"grad_norm": 1.7089264392852783,
|
| 1631 |
+
"learning_rate": 1e-06,
|
| 1632 |
+
"loss": 0.1248,
|
| 1633 |
+
"step": 232
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 0.0745361484325016,
|
| 1637 |
+
"grad_norm": 1.3122811317443848,
|
| 1638 |
+
"learning_rate": 1e-06,
|
| 1639 |
+
"loss": 0.1241,
|
| 1640 |
+
"step": 233
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 0.07485604606525911,
|
| 1644 |
+
"grad_norm": 1.1309940814971924,
|
| 1645 |
+
"learning_rate": 1e-06,
|
| 1646 |
+
"loss": 0.1226,
|
| 1647 |
+
"step": 234
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 0.07517594369801664,
|
| 1651 |
+
"grad_norm": 1.4108712673187256,
|
| 1652 |
+
"learning_rate": 1e-06,
|
| 1653 |
+
"loss": 0.1244,
|
| 1654 |
+
"step": 235
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 0.07549584133077415,
|
| 1658 |
+
"grad_norm": 1.58648681640625,
|
| 1659 |
+
"learning_rate": 1e-06,
|
| 1660 |
+
"loss": 0.1227,
|
| 1661 |
+
"step": 236
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 0.07581573896353166,
|
| 1665 |
+
"grad_norm": 1.5306432247161865,
|
| 1666 |
+
"learning_rate": 1e-06,
|
| 1667 |
+
"loss": 0.1273,
|
| 1668 |
+
"step": 237
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 0.07613563659628919,
|
| 1672 |
+
"grad_norm": 1.677597165107727,
|
| 1673 |
+
"learning_rate": 1e-06,
|
| 1674 |
+
"loss": 0.1299,
|
| 1675 |
+
"step": 238
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 0.0764555342290467,
|
| 1679 |
+
"grad_norm": 1.2486478090286255,
|
| 1680 |
+
"learning_rate": 1e-06,
|
| 1681 |
+
"loss": 0.124,
|
| 1682 |
+
"step": 239
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 0.07677543186180422,
|
| 1686 |
+
"grad_norm": 1.2199089527130127,
|
| 1687 |
+
"learning_rate": 1e-06,
|
| 1688 |
+
"loss": 0.1213,
|
| 1689 |
+
"step": 240
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 0.07709532949456174,
|
| 1693 |
+
"grad_norm": 1.335979700088501,
|
| 1694 |
+
"learning_rate": 1e-06,
|
| 1695 |
+
"loss": 0.1238,
|
| 1696 |
+
"step": 241
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 0.07741522712731926,
|
| 1700 |
+
"grad_norm": 1.1188368797302246,
|
| 1701 |
+
"learning_rate": 1e-06,
|
| 1702 |
+
"loss": 0.124,
|
| 1703 |
+
"step": 242
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 0.07773512476007678,
|
| 1707 |
+
"grad_norm": 1.1585828065872192,
|
| 1708 |
+
"learning_rate": 1e-06,
|
| 1709 |
+
"loss": 0.1233,
|
| 1710 |
+
"step": 243
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 0.0780550223928343,
|
| 1714 |
+
"grad_norm": 1.2010987997055054,
|
| 1715 |
+
"learning_rate": 1e-06,
|
| 1716 |
+
"loss": 0.121,
|
| 1717 |
+
"step": 244
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 0.07837492002559181,
|
| 1721 |
+
"grad_norm": 1.1653729677200317,
|
| 1722 |
+
"learning_rate": 1e-06,
|
| 1723 |
+
"loss": 0.1231,
|
| 1724 |
+
"step": 245
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 0.07869481765834933,
|
| 1728 |
+
"grad_norm": 1.4852837324142456,
|
| 1729 |
+
"learning_rate": 1e-06,
|
| 1730 |
+
"loss": 0.122,
|
| 1731 |
+
"step": 246
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 0.07901471529110685,
|
| 1735 |
+
"grad_norm": 1.5950335264205933,
|
| 1736 |
+
"learning_rate": 1e-06,
|
| 1737 |
+
"loss": 0.1258,
|
| 1738 |
+
"step": 247
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 0.07933461292386436,
|
| 1742 |
+
"grad_norm": 1.3516967296600342,
|
| 1743 |
+
"learning_rate": 1e-06,
|
| 1744 |
+
"loss": 0.1273,
|
| 1745 |
+
"step": 248
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 0.07965451055662189,
|
| 1749 |
+
"grad_norm": 1.1554462909698486,
|
| 1750 |
+
"learning_rate": 1e-06,
|
| 1751 |
+
"loss": 0.1221,
|
| 1752 |
+
"step": 249
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 0.0799744081893794,
|
| 1756 |
+
"grad_norm": 1.1331981420516968,
|
| 1757 |
+
"learning_rate": 1e-06,
|
| 1758 |
+
"loss": 0.12,
|
| 1759 |
+
"step": 250
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 0.08029430582213691,
|
| 1763 |
+
"grad_norm": 1.354332447052002,
|
| 1764 |
+
"learning_rate": 1e-06,
|
| 1765 |
+
"loss": 0.1256,
|
| 1766 |
+
"step": 251
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 0.08061420345489444,
|
| 1770 |
+
"grad_norm": 1.2884771823883057,
|
| 1771 |
+
"learning_rate": 1e-06,
|
| 1772 |
+
"loss": 0.122,
|
| 1773 |
+
"step": 252
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 0.08093410108765195,
|
| 1777 |
+
"grad_norm": 1.107785940170288,
|
| 1778 |
+
"learning_rate": 1e-06,
|
| 1779 |
+
"loss": 0.1222,
|
| 1780 |
+
"step": 253
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 0.08125399872040946,
|
| 1784 |
+
"grad_norm": 0.879289984703064,
|
| 1785 |
+
"learning_rate": 1e-06,
|
| 1786 |
+
"loss": 0.1238,
|
| 1787 |
+
"step": 254
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 0.08157389635316699,
|
| 1791 |
+
"grad_norm": 1.0930490493774414,
|
| 1792 |
+
"learning_rate": 1e-06,
|
| 1793 |
+
"loss": 0.1253,
|
| 1794 |
+
"step": 255
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 0.0818937939859245,
|
| 1798 |
+
"grad_norm": 0.9492177963256836,
|
| 1799 |
+
"learning_rate": 1e-06,
|
| 1800 |
+
"loss": 0.1244,
|
| 1801 |
+
"step": 256
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 0.08221369161868203,
|
| 1805 |
+
"grad_norm": 1.168870449066162,
|
| 1806 |
+
"learning_rate": 1e-06,
|
| 1807 |
+
"loss": 0.1236,
|
| 1808 |
+
"step": 257
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 0.08253358925143954,
|
| 1812 |
+
"grad_norm": 1.2158278226852417,
|
| 1813 |
+
"learning_rate": 1e-06,
|
| 1814 |
+
"loss": 0.1229,
|
| 1815 |
+
"step": 258
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 0.08285348688419705,
|
| 1819 |
+
"grad_norm": 1.073761224746704,
|
| 1820 |
+
"learning_rate": 1e-06,
|
| 1821 |
+
"loss": 0.1233,
|
| 1822 |
+
"step": 259
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 0.08317338451695458,
|
| 1826 |
+
"grad_norm": 1.4267514944076538,
|
| 1827 |
+
"learning_rate": 1e-06,
|
| 1828 |
+
"loss": 0.1221,
|
| 1829 |
+
"step": 260
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 0.08349328214971209,
|
| 1833 |
+
"grad_norm": 1.1752041578292847,
|
| 1834 |
+
"learning_rate": 1e-06,
|
| 1835 |
+
"loss": 0.1245,
|
| 1836 |
+
"step": 261
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 0.0838131797824696,
|
| 1840 |
+
"grad_norm": 0.9940073490142822,
|
| 1841 |
+
"learning_rate": 1e-06,
|
| 1842 |
+
"loss": 0.1185,
|
| 1843 |
+
"step": 262
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 0.08413307741522713,
|
| 1847 |
+
"grad_norm": 1.4681187868118286,
|
| 1848 |
+
"learning_rate": 1e-06,
|
| 1849 |
+
"loss": 0.124,
|
| 1850 |
+
"step": 263
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 0.08445297504798464,
|
| 1854 |
+
"grad_norm": 1.0314862728118896,
|
| 1855 |
+
"learning_rate": 1e-06,
|
| 1856 |
+
"loss": 0.1276,
|
| 1857 |
+
"step": 264
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 0.08477287268074216,
|
| 1861 |
+
"grad_norm": 1.7601333856582642,
|
| 1862 |
+
"learning_rate": 1e-06,
|
| 1863 |
+
"loss": 0.1213,
|
| 1864 |
+
"step": 265
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 0.08509277031349968,
|
| 1868 |
+
"grad_norm": 0.8387244939804077,
|
| 1869 |
+
"learning_rate": 1e-06,
|
| 1870 |
+
"loss": 0.1228,
|
| 1871 |
+
"step": 266
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 0.0854126679462572,
|
| 1875 |
+
"grad_norm": 1.0458019971847534,
|
| 1876 |
+
"learning_rate": 1e-06,
|
| 1877 |
+
"loss": 0.1199,
|
| 1878 |
+
"step": 267
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 0.08573256557901472,
|
| 1882 |
+
"grad_norm": 0.9919850826263428,
|
| 1883 |
+
"learning_rate": 1e-06,
|
| 1884 |
+
"loss": 0.12,
|
| 1885 |
+
"step": 268
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 0.08605246321177223,
|
| 1889 |
+
"grad_norm": 1.3473700284957886,
|
| 1890 |
+
"learning_rate": 1e-06,
|
| 1891 |
+
"loss": 0.1219,
|
| 1892 |
+
"step": 269
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 0.08637236084452975,
|
| 1896 |
+
"grad_norm": 1.0079095363616943,
|
| 1897 |
+
"learning_rate": 1e-06,
|
| 1898 |
+
"loss": 0.1227,
|
| 1899 |
+
"step": 270
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 0.08669225847728727,
|
| 1903 |
+
"grad_norm": 1.0614455938339233,
|
| 1904 |
+
"learning_rate": 1e-06,
|
| 1905 |
+
"loss": 0.1174,
|
| 1906 |
+
"step": 271
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 0.08701215611004479,
|
| 1910 |
+
"grad_norm": 1.0451921224594116,
|
| 1911 |
+
"learning_rate": 1e-06,
|
| 1912 |
+
"loss": 0.1201,
|
| 1913 |
+
"step": 272
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 0.0873320537428023,
|
| 1917 |
+
"grad_norm": 1.3198274374008179,
|
| 1918 |
+
"learning_rate": 1e-06,
|
| 1919 |
+
"loss": 0.1203,
|
| 1920 |
+
"step": 273
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 0.08765195137555983,
|
| 1924 |
+
"grad_norm": 1.2236682176589966,
|
| 1925 |
+
"learning_rate": 1e-06,
|
| 1926 |
+
"loss": 0.1212,
|
| 1927 |
+
"step": 274
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 0.08797184900831734,
|
| 1931 |
+
"grad_norm": 1.148290753364563,
|
| 1932 |
+
"learning_rate": 1e-06,
|
| 1933 |
+
"loss": 0.1209,
|
| 1934 |
+
"step": 275
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 0.08829174664107485,
|
| 1938 |
+
"grad_norm": 0.9034926891326904,
|
| 1939 |
+
"learning_rate": 1e-06,
|
| 1940 |
+
"loss": 0.1215,
|
| 1941 |
+
"step": 276
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 0.08861164427383238,
|
| 1945 |
+
"grad_norm": 0.9828817248344421,
|
| 1946 |
+
"learning_rate": 1e-06,
|
| 1947 |
+
"loss": 0.1179,
|
| 1948 |
+
"step": 277
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 0.08893154190658989,
|
| 1952 |
+
"grad_norm": 1.1249366998672485,
|
| 1953 |
+
"learning_rate": 1e-06,
|
| 1954 |
+
"loss": 0.1227,
|
| 1955 |
+
"step": 278
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 0.0892514395393474,
|
| 1959 |
+
"grad_norm": 1.425968885421753,
|
| 1960 |
+
"learning_rate": 1e-06,
|
| 1961 |
+
"loss": 0.1221,
|
| 1962 |
+
"step": 279
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 0.08957133717210493,
|
| 1966 |
+
"grad_norm": 1.020310640335083,
|
| 1967 |
+
"learning_rate": 1e-06,
|
| 1968 |
+
"loss": 0.1213,
|
| 1969 |
+
"step": 280
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 0.08989123480486244,
|
| 1973 |
+
"grad_norm": 1.327886939048767,
|
| 1974 |
+
"learning_rate": 1e-06,
|
| 1975 |
+
"loss": 0.1212,
|
| 1976 |
+
"step": 281
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 0.09021113243761997,
|
| 1980 |
+
"grad_norm": 1.1981432437896729,
|
| 1981 |
+
"learning_rate": 1e-06,
|
| 1982 |
+
"loss": 0.1176,
|
| 1983 |
+
"step": 282
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 0.09053103007037748,
|
| 1987 |
+
"grad_norm": 1.1416324377059937,
|
| 1988 |
+
"learning_rate": 1e-06,
|
| 1989 |
+
"loss": 0.123,
|
| 1990 |
+
"step": 283
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 0.09085092770313499,
|
| 1994 |
+
"grad_norm": 1.142218828201294,
|
| 1995 |
+
"learning_rate": 1e-06,
|
| 1996 |
+
"loss": 0.1181,
|
| 1997 |
+
"step": 284
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 0.09117082533589252,
|
| 2001 |
+
"grad_norm": 1.2026841640472412,
|
| 2002 |
+
"learning_rate": 1e-06,
|
| 2003 |
+
"loss": 0.1225,
|
| 2004 |
+
"step": 285
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 0.09149072296865003,
|
| 2008 |
+
"grad_norm": 1.1408050060272217,
|
| 2009 |
+
"learning_rate": 1e-06,
|
| 2010 |
+
"loss": 0.1219,
|
| 2011 |
+
"step": 286
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 0.09181062060140754,
|
| 2015 |
+
"grad_norm": 1.1757482290267944,
|
| 2016 |
+
"learning_rate": 1e-06,
|
| 2017 |
+
"loss": 0.1191,
|
| 2018 |
+
"step": 287
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 0.09213051823416507,
|
| 2022 |
+
"grad_norm": 1.0295770168304443,
|
| 2023 |
+
"learning_rate": 1e-06,
|
| 2024 |
+
"loss": 0.1206,
|
| 2025 |
+
"step": 288
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 0.09245041586692258,
|
| 2029 |
+
"grad_norm": 1.218077301979065,
|
| 2030 |
+
"learning_rate": 1e-06,
|
| 2031 |
+
"loss": 0.1232,
|
| 2032 |
+
"step": 289
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 0.0927703134996801,
|
| 2036 |
+
"grad_norm": 1.1868141889572144,
|
| 2037 |
+
"learning_rate": 1e-06,
|
| 2038 |
+
"loss": 0.1225,
|
| 2039 |
+
"step": 290
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 0.09309021113243762,
|
| 2043 |
+
"grad_norm": 1.3732609748840332,
|
| 2044 |
+
"learning_rate": 1e-06,
|
| 2045 |
+
"loss": 0.1254,
|
| 2046 |
+
"step": 291
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 0.09341010876519514,
|
| 2050 |
+
"grad_norm": 1.1777644157409668,
|
| 2051 |
+
"learning_rate": 1e-06,
|
| 2052 |
+
"loss": 0.1228,
|
| 2053 |
+
"step": 292
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 0.09373000639795266,
|
| 2057 |
+
"grad_norm": 0.9427011013031006,
|
| 2058 |
+
"learning_rate": 1e-06,
|
| 2059 |
+
"loss": 0.1191,
|
| 2060 |
+
"step": 293
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 0.09404990403071017,
|
| 2064 |
+
"grad_norm": 1.020755648612976,
|
| 2065 |
+
"learning_rate": 1e-06,
|
| 2066 |
+
"loss": 0.1233,
|
| 2067 |
+
"step": 294
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 0.09436980166346769,
|
| 2071 |
+
"grad_norm": 1.1318416595458984,
|
| 2072 |
+
"learning_rate": 1e-06,
|
| 2073 |
+
"loss": 0.1177,
|
| 2074 |
+
"step": 295
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 0.09468969929622521,
|
| 2078 |
+
"grad_norm": 1.018863320350647,
|
| 2079 |
+
"learning_rate": 1e-06,
|
| 2080 |
+
"loss": 0.1191,
|
| 2081 |
+
"step": 296
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 0.09500959692898273,
|
| 2085 |
+
"grad_norm": 1.1859015226364136,
|
| 2086 |
+
"learning_rate": 1e-06,
|
| 2087 |
+
"loss": 0.1228,
|
| 2088 |
+
"step": 297
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 0.09532949456174024,
|
| 2092 |
+
"grad_norm": 1.265998363494873,
|
| 2093 |
+
"learning_rate": 1e-06,
|
| 2094 |
+
"loss": 0.1205,
|
| 2095 |
+
"step": 298
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 0.09564939219449776,
|
| 2099 |
+
"grad_norm": 0.9782631993293762,
|
| 2100 |
+
"learning_rate": 1e-06,
|
| 2101 |
+
"loss": 0.1203,
|
| 2102 |
+
"step": 299
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 0.09596928982725528,
|
| 2106 |
+
"grad_norm": 1.022936224937439,
|
| 2107 |
+
"learning_rate": 1e-06,
|
| 2108 |
+
"loss": 0.1225,
|
| 2109 |
+
"step": 300
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 0.09628918746001279,
|
| 2113 |
+
"grad_norm": 0.9478284120559692,
|
| 2114 |
+
"learning_rate": 1e-06,
|
| 2115 |
+
"loss": 0.1203,
|
| 2116 |
+
"step": 301
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 0.09660908509277032,
|
| 2120 |
+
"grad_norm": 1.1563911437988281,
|
| 2121 |
+
"learning_rate": 1e-06,
|
| 2122 |
+
"loss": 0.1202,
|
| 2123 |
+
"step": 302
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 0.09692898272552783,
|
| 2127 |
+
"grad_norm": 1.236619234085083,
|
| 2128 |
+
"learning_rate": 1e-06,
|
| 2129 |
+
"loss": 0.1229,
|
| 2130 |
+
"step": 303
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 0.09724888035828536,
|
| 2134 |
+
"grad_norm": 1.3171405792236328,
|
| 2135 |
+
"learning_rate": 1e-06,
|
| 2136 |
+
"loss": 0.1191,
|
| 2137 |
+
"step": 304
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 0.09756877799104287,
|
| 2141 |
+
"grad_norm": 1.2637388706207275,
|
| 2142 |
+
"learning_rate": 1e-06,
|
| 2143 |
+
"loss": 0.1222,
|
| 2144 |
+
"step": 305
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 0.09788867562380038,
|
| 2148 |
+
"grad_norm": 1.3073060512542725,
|
| 2149 |
+
"learning_rate": 1e-06,
|
| 2150 |
+
"loss": 0.1182,
|
| 2151 |
+
"step": 306
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 0.09820857325655791,
|
| 2155 |
+
"grad_norm": 1.111814022064209,
|
| 2156 |
+
"learning_rate": 1e-06,
|
| 2157 |
+
"loss": 0.1167,
|
| 2158 |
+
"step": 307
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 0.09852847088931542,
|
| 2162 |
+
"grad_norm": 1.2709119319915771,
|
| 2163 |
+
"learning_rate": 1e-06,
|
| 2164 |
+
"loss": 0.1212,
|
| 2165 |
+
"step": 308
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 0.09884836852207293,
|
| 2169 |
+
"grad_norm": 1.0304408073425293,
|
| 2170 |
+
"learning_rate": 1e-06,
|
| 2171 |
+
"loss": 0.1181,
|
| 2172 |
+
"step": 309
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 0.09916826615483046,
|
| 2176 |
+
"grad_norm": 1.173019289970398,
|
| 2177 |
+
"learning_rate": 1e-06,
|
| 2178 |
+
"loss": 0.1203,
|
| 2179 |
+
"step": 310
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 0.09948816378758797,
|
| 2183 |
+
"grad_norm": 1.2476046085357666,
|
| 2184 |
+
"learning_rate": 1e-06,
|
| 2185 |
+
"loss": 0.1187,
|
| 2186 |
+
"step": 311
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 0.09980806142034548,
|
| 2190 |
+
"grad_norm": 1.2835643291473389,
|
| 2191 |
+
"learning_rate": 1e-06,
|
| 2192 |
+
"loss": 0.119,
|
| 2193 |
+
"step": 312
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 0.10012795905310301,
|
| 2197 |
+
"grad_norm": 1.0812190771102905,
|
| 2198 |
+
"learning_rate": 1e-06,
|
| 2199 |
+
"loss": 0.1164,
|
| 2200 |
+
"step": 313
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 0.10044785668586052,
|
| 2204 |
+
"grad_norm": 1.6225135326385498,
|
| 2205 |
+
"learning_rate": 1e-06,
|
| 2206 |
+
"loss": 0.1215,
|
| 2207 |
+
"step": 314
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 0.10076775431861804,
|
| 2211 |
+
"grad_norm": 2.6078035831451416,
|
| 2212 |
+
"learning_rate": 1e-06,
|
| 2213 |
+
"loss": 0.1208,
|
| 2214 |
+
"step": 315
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 0.10108765195137556,
|
| 2218 |
+
"grad_norm": 1.4901459217071533,
|
| 2219 |
+
"learning_rate": 1e-06,
|
| 2220 |
+
"loss": 0.1211,
|
| 2221 |
+
"step": 316
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 0.10140754958413307,
|
| 2225 |
+
"grad_norm": 1.167883276939392,
|
| 2226 |
+
"learning_rate": 1e-06,
|
| 2227 |
+
"loss": 0.1178,
|
| 2228 |
+
"step": 317
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 0.1017274472168906,
|
| 2232 |
+
"grad_norm": 0.9029871821403503,
|
| 2233 |
+
"learning_rate": 1e-06,
|
| 2234 |
+
"loss": 0.1155,
|
| 2235 |
+
"step": 318
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 0.10204734484964811,
|
| 2239 |
+
"grad_norm": 1.0930075645446777,
|
| 2240 |
+
"learning_rate": 1e-06,
|
| 2241 |
+
"loss": 0.1208,
|
| 2242 |
+
"step": 319
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 0.10236724248240563,
|
| 2246 |
+
"grad_norm": 1.5672348737716675,
|
| 2247 |
+
"learning_rate": 1e-06,
|
| 2248 |
+
"loss": 0.1191,
|
| 2249 |
+
"step": 320
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 0.10268714011516315,
|
| 2253 |
+
"grad_norm": 1.2925875186920166,
|
| 2254 |
+
"learning_rate": 1e-06,
|
| 2255 |
+
"loss": 0.1218,
|
| 2256 |
+
"step": 321
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 0.10300703774792067,
|
| 2260 |
+
"grad_norm": 1.5786855220794678,
|
| 2261 |
+
"learning_rate": 1e-06,
|
| 2262 |
+
"loss": 0.1212,
|
| 2263 |
+
"step": 322
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 0.10332693538067818,
|
| 2267 |
+
"grad_norm": 0.9217334389686584,
|
| 2268 |
+
"learning_rate": 1e-06,
|
| 2269 |
+
"loss": 0.1187,
|
| 2270 |
+
"step": 323
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 0.1036468330134357,
|
| 2274 |
+
"grad_norm": 1.175582766532898,
|
| 2275 |
+
"learning_rate": 1e-06,
|
| 2276 |
+
"loss": 0.1175,
|
| 2277 |
+
"step": 324
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 0.10396673064619322,
|
| 2281 |
+
"grad_norm": 1.3332633972167969,
|
| 2282 |
+
"learning_rate": 1e-06,
|
| 2283 |
+
"loss": 0.1204,
|
| 2284 |
+
"step": 325
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 0.10428662827895073,
|
| 2288 |
+
"grad_norm": 1.147852897644043,
|
| 2289 |
+
"learning_rate": 1e-06,
|
| 2290 |
+
"loss": 0.1216,
|
| 2291 |
+
"step": 326
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 0.10460652591170826,
|
| 2295 |
+
"grad_norm": 1.4355601072311401,
|
| 2296 |
+
"learning_rate": 1e-06,
|
| 2297 |
+
"loss": 0.1185,
|
| 2298 |
+
"step": 327
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 0.10492642354446577,
|
| 2302 |
+
"grad_norm": 1.059722661972046,
|
| 2303 |
+
"learning_rate": 1e-06,
|
| 2304 |
+
"loss": 0.1193,
|
| 2305 |
+
"step": 328
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 0.1052463211772233,
|
| 2309 |
+
"grad_norm": 1.0881210565567017,
|
| 2310 |
+
"learning_rate": 1e-06,
|
| 2311 |
+
"loss": 0.1203,
|
| 2312 |
+
"step": 329
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 0.10556621880998081,
|
| 2316 |
+
"grad_norm": 1.124450445175171,
|
| 2317 |
+
"learning_rate": 1e-06,
|
| 2318 |
+
"loss": 0.1184,
|
| 2319 |
+
"step": 330
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 0.10588611644273832,
|
| 2323 |
+
"grad_norm": 1.186869502067566,
|
| 2324 |
+
"learning_rate": 1e-06,
|
| 2325 |
+
"loss": 0.1168,
|
| 2326 |
+
"step": 331
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 0.10620601407549585,
|
| 2330 |
+
"grad_norm": 1.1845691204071045,
|
| 2331 |
+
"learning_rate": 1e-06,
|
| 2332 |
+
"loss": 0.1169,
|
| 2333 |
+
"step": 332
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 0.10652591170825336,
|
| 2337 |
+
"grad_norm": 1.0931459665298462,
|
| 2338 |
+
"learning_rate": 1e-06,
|
| 2339 |
+
"loss": 0.1167,
|
| 2340 |
+
"step": 333
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 0.10684580934101087,
|
| 2344 |
+
"grad_norm": 1.2172439098358154,
|
| 2345 |
+
"learning_rate": 1e-06,
|
| 2346 |
+
"loss": 0.1198,
|
| 2347 |
+
"step": 334
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 0.1071657069737684,
|
| 2351 |
+
"grad_norm": 1.0822011232376099,
|
| 2352 |
+
"learning_rate": 1e-06,
|
| 2353 |
+
"loss": 0.1164,
|
| 2354 |
+
"step": 335
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 0.10748560460652591,
|
| 2358 |
+
"grad_norm": 1.1918107271194458,
|
| 2359 |
+
"learning_rate": 1e-06,
|
| 2360 |
+
"loss": 0.115,
|
| 2361 |
+
"step": 336
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 0.10780550223928342,
|
| 2365 |
+
"grad_norm": 1.089294195175171,
|
| 2366 |
+
"learning_rate": 1e-06,
|
| 2367 |
+
"loss": 0.116,
|
| 2368 |
+
"step": 337
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 0.10812539987204095,
|
| 2372 |
+
"grad_norm": 1.2940959930419922,
|
| 2373 |
+
"learning_rate": 1e-06,
|
| 2374 |
+
"loss": 0.1167,
|
| 2375 |
+
"step": 338
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 0.10844529750479846,
|
| 2379 |
+
"grad_norm": 1.416695475578308,
|
| 2380 |
+
"learning_rate": 1e-06,
|
| 2381 |
+
"loss": 0.1186,
|
| 2382 |
+
"step": 339
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 0.10876519513755598,
|
| 2386 |
+
"grad_norm": 1.0529122352600098,
|
| 2387 |
+
"learning_rate": 1e-06,
|
| 2388 |
+
"loss": 0.1165,
|
| 2389 |
+
"step": 340
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 0.1090850927703135,
|
| 2393 |
+
"grad_norm": 1.0267155170440674,
|
| 2394 |
+
"learning_rate": 1e-06,
|
| 2395 |
+
"loss": 0.1172,
|
| 2396 |
+
"step": 341
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 0.10940499040307101,
|
| 2400 |
+
"grad_norm": 1.1688264608383179,
|
| 2401 |
+
"learning_rate": 1e-06,
|
| 2402 |
+
"loss": 0.1196,
|
| 2403 |
+
"step": 342
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 0.10972488803582854,
|
| 2407 |
+
"grad_norm": 0.9702956080436707,
|
| 2408 |
+
"learning_rate": 1e-06,
|
| 2409 |
+
"loss": 0.1168,
|
| 2410 |
+
"step": 343
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 0.11004478566858605,
|
| 2414 |
+
"grad_norm": 1.152326226234436,
|
| 2415 |
+
"learning_rate": 1e-06,
|
| 2416 |
+
"loss": 0.1177,
|
| 2417 |
+
"step": 344
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 0.11036468330134357,
|
| 2421 |
+
"grad_norm": 1.185773253440857,
|
| 2422 |
+
"learning_rate": 1e-06,
|
| 2423 |
+
"loss": 0.1141,
|
| 2424 |
+
"step": 345
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 0.11068458093410109,
|
| 2428 |
+
"grad_norm": 1.1504043340682983,
|
| 2429 |
+
"learning_rate": 1e-06,
|
| 2430 |
+
"loss": 0.1188,
|
| 2431 |
+
"step": 346
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 0.1110044785668586,
|
| 2435 |
+
"grad_norm": 1.3403894901275635,
|
| 2436 |
+
"learning_rate": 1e-06,
|
| 2437 |
+
"loss": 0.1194,
|
| 2438 |
+
"step": 347
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 0.11132437619961612,
|
| 2442 |
+
"grad_norm": 1.1313585042953491,
|
| 2443 |
+
"learning_rate": 1e-06,
|
| 2444 |
+
"loss": 0.1184,
|
| 2445 |
+
"step": 348
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 0.11164427383237364,
|
| 2449 |
+
"grad_norm": 1.110701560974121,
|
| 2450 |
+
"learning_rate": 1e-06,
|
| 2451 |
+
"loss": 0.115,
|
| 2452 |
+
"step": 349
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 0.11196417146513116,
|
| 2456 |
+
"grad_norm": 1.1351897716522217,
|
| 2457 |
+
"learning_rate": 1e-06,
|
| 2458 |
+
"loss": 0.117,
|
| 2459 |
+
"step": 350
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 0.11228406909788867,
|
| 2463 |
+
"grad_norm": 0.9436936378479004,
|
| 2464 |
+
"learning_rate": 1e-06,
|
| 2465 |
+
"loss": 0.1165,
|
| 2466 |
+
"step": 351
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 0.1126039667306462,
|
| 2470 |
+
"grad_norm": 1.087910532951355,
|
| 2471 |
+
"learning_rate": 1e-06,
|
| 2472 |
+
"loss": 0.1197,
|
| 2473 |
+
"step": 352
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 0.11292386436340371,
|
| 2477 |
+
"grad_norm": 1.1439520120620728,
|
| 2478 |
+
"learning_rate": 1e-06,
|
| 2479 |
+
"loss": 0.1198,
|
| 2480 |
+
"step": 353
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 0.11324376199616124,
|
| 2484 |
+
"grad_norm": 1.1269515752792358,
|
| 2485 |
+
"learning_rate": 1e-06,
|
| 2486 |
+
"loss": 0.118,
|
| 2487 |
+
"step": 354
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 0.11356365962891875,
|
| 2491 |
+
"grad_norm": 0.8861843347549438,
|
| 2492 |
+
"learning_rate": 1e-06,
|
| 2493 |
+
"loss": 0.1123,
|
| 2494 |
+
"step": 355
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 0.11388355726167626,
|
| 2498 |
+
"grad_norm": 1.0100499391555786,
|
| 2499 |
+
"learning_rate": 1e-06,
|
| 2500 |
+
"loss": 0.1163,
|
| 2501 |
+
"step": 356
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 0.11420345489443379,
|
| 2505 |
+
"grad_norm": 1.0040833950042725,
|
| 2506 |
+
"learning_rate": 1e-06,
|
| 2507 |
+
"loss": 0.1159,
|
| 2508 |
+
"step": 357
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 0.1145233525271913,
|
| 2512 |
+
"grad_norm": 1.1430156230926514,
|
| 2513 |
+
"learning_rate": 1e-06,
|
| 2514 |
+
"loss": 0.1153,
|
| 2515 |
+
"step": 358
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 0.11484325015994881,
|
| 2519 |
+
"grad_norm": 1.0967388153076172,
|
| 2520 |
+
"learning_rate": 1e-06,
|
| 2521 |
+
"loss": 0.1193,
|
| 2522 |
+
"step": 359
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 0.11516314779270634,
|
| 2526 |
+
"grad_norm": 1.2168657779693604,
|
| 2527 |
+
"learning_rate": 1e-06,
|
| 2528 |
+
"loss": 0.1189,
|
| 2529 |
+
"step": 360
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 0.11548304542546385,
|
| 2533 |
+
"grad_norm": 1.240936040878296,
|
| 2534 |
+
"learning_rate": 1e-06,
|
| 2535 |
+
"loss": 0.1168,
|
| 2536 |
+
"step": 361
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 0.11580294305822136,
|
| 2540 |
+
"grad_norm": 1.1288347244262695,
|
| 2541 |
+
"learning_rate": 1e-06,
|
| 2542 |
+
"loss": 0.1183,
|
| 2543 |
+
"step": 362
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 0.11612284069097889,
|
| 2547 |
+
"grad_norm": 1.373227596282959,
|
| 2548 |
+
"learning_rate": 1e-06,
|
| 2549 |
+
"loss": 0.1157,
|
| 2550 |
+
"step": 363
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 0.1164427383237364,
|
| 2554 |
+
"grad_norm": 1.3604793548583984,
|
| 2555 |
+
"learning_rate": 1e-06,
|
| 2556 |
+
"loss": 0.1163,
|
| 2557 |
+
"step": 364
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 0.11676263595649392,
|
| 2561 |
+
"grad_norm": 0.972876250743866,
|
| 2562 |
+
"learning_rate": 1e-06,
|
| 2563 |
+
"loss": 0.1121,
|
| 2564 |
+
"step": 365
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 0.11708253358925144,
|
| 2568 |
+
"grad_norm": 1.4081337451934814,
|
| 2569 |
+
"learning_rate": 1e-06,
|
| 2570 |
+
"loss": 0.1186,
|
| 2571 |
+
"step": 366
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 0.11740243122200895,
|
| 2575 |
+
"grad_norm": 1.182377815246582,
|
| 2576 |
+
"learning_rate": 1e-06,
|
| 2577 |
+
"loss": 0.1186,
|
| 2578 |
+
"step": 367
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 0.11772232885476648,
|
| 2582 |
+
"grad_norm": 0.9548720717430115,
|
| 2583 |
+
"learning_rate": 1e-06,
|
| 2584 |
+
"loss": 0.1174,
|
| 2585 |
+
"step": 368
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 0.118042226487524,
|
| 2589 |
+
"grad_norm": 1.138261318206787,
|
| 2590 |
+
"learning_rate": 1e-06,
|
| 2591 |
+
"loss": 0.1138,
|
| 2592 |
+
"step": 369
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 0.1183621241202815,
|
| 2596 |
+
"grad_norm": 1.1072566509246826,
|
| 2597 |
+
"learning_rate": 1e-06,
|
| 2598 |
+
"loss": 0.1141,
|
| 2599 |
+
"step": 370
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 0.11868202175303903,
|
| 2603 |
+
"grad_norm": 1.1237493753433228,
|
| 2604 |
+
"learning_rate": 1e-06,
|
| 2605 |
+
"loss": 0.1149,
|
| 2606 |
+
"step": 371
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 0.11900191938579655,
|
| 2610 |
+
"grad_norm": 1.0723813772201538,
|
| 2611 |
+
"learning_rate": 1e-06,
|
| 2612 |
+
"loss": 0.1131,
|
| 2613 |
+
"step": 372
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 0.11932181701855406,
|
| 2617 |
+
"grad_norm": 1.3001840114593506,
|
| 2618 |
+
"learning_rate": 1e-06,
|
| 2619 |
+
"loss": 0.1164,
|
| 2620 |
+
"step": 373
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 0.11964171465131158,
|
| 2624 |
+
"grad_norm": 1.4676969051361084,
|
| 2625 |
+
"learning_rate": 1e-06,
|
| 2626 |
+
"loss": 0.117,
|
| 2627 |
+
"step": 374
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 0.1199616122840691,
|
| 2631 |
+
"grad_norm": 1.3240898847579956,
|
| 2632 |
+
"learning_rate": 1e-06,
|
| 2633 |
+
"loss": 0.1169,
|
| 2634 |
+
"step": 375
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 0.12028150991682661,
|
| 2638 |
+
"grad_norm": 0.8883340358734131,
|
| 2639 |
+
"learning_rate": 1e-06,
|
| 2640 |
+
"loss": 0.1162,
|
| 2641 |
+
"step": 376
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 0.12060140754958414,
|
| 2645 |
+
"grad_norm": 1.378762125968933,
|
| 2646 |
+
"learning_rate": 1e-06,
|
| 2647 |
+
"loss": 0.1158,
|
| 2648 |
+
"step": 377
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 0.12092130518234165,
|
| 2652 |
+
"grad_norm": 1.0585156679153442,
|
| 2653 |
+
"learning_rate": 1e-06,
|
| 2654 |
+
"loss": 0.1154,
|
| 2655 |
+
"step": 378
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 0.12124120281509917,
|
| 2659 |
+
"grad_norm": 1.2299355268478394,
|
| 2660 |
+
"learning_rate": 1e-06,
|
| 2661 |
+
"loss": 0.1151,
|
| 2662 |
+
"step": 379
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 0.12156110044785669,
|
| 2666 |
+
"grad_norm": 1.0432575941085815,
|
| 2667 |
+
"learning_rate": 1e-06,
|
| 2668 |
+
"loss": 0.1159,
|
| 2669 |
+
"step": 380
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 0.1218809980806142,
|
| 2673 |
+
"grad_norm": 1.3666393756866455,
|
| 2674 |
+
"learning_rate": 1e-06,
|
| 2675 |
+
"loss": 0.1135,
|
| 2676 |
+
"step": 381
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 0.12220089571337173,
|
| 2680 |
+
"grad_norm": 1.2501029968261719,
|
| 2681 |
+
"learning_rate": 1e-06,
|
| 2682 |
+
"loss": 0.1149,
|
| 2683 |
+
"step": 382
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 0.12252079334612924,
|
| 2687 |
+
"grad_norm": 1.6209537982940674,
|
| 2688 |
+
"learning_rate": 1e-06,
|
| 2689 |
+
"loss": 0.1169,
|
| 2690 |
+
"step": 383
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 0.12284069097888675,
|
| 2694 |
+
"grad_norm": 1.498772144317627,
|
| 2695 |
+
"learning_rate": 1e-06,
|
| 2696 |
+
"loss": 0.1182,
|
| 2697 |
+
"step": 384
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 0.12316058861164428,
|
| 2701 |
+
"grad_norm": 1.1027629375457764,
|
| 2702 |
+
"learning_rate": 1e-06,
|
| 2703 |
+
"loss": 0.1146,
|
| 2704 |
+
"step": 385
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 0.12348048624440179,
|
| 2708 |
+
"grad_norm": 1.0916204452514648,
|
| 2709 |
+
"learning_rate": 1e-06,
|
| 2710 |
+
"loss": 0.1125,
|
| 2711 |
+
"step": 386
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 0.1238003838771593,
|
| 2715 |
+
"grad_norm": 1.0010249614715576,
|
| 2716 |
+
"learning_rate": 1e-06,
|
| 2717 |
+
"loss": 0.115,
|
| 2718 |
+
"step": 387
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 0.12412028150991683,
|
| 2722 |
+
"grad_norm": 1.5269064903259277,
|
| 2723 |
+
"learning_rate": 1e-06,
|
| 2724 |
+
"loss": 0.1155,
|
| 2725 |
+
"step": 388
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 0.12444017914267434,
|
| 2729 |
+
"grad_norm": 1.2441141605377197,
|
| 2730 |
+
"learning_rate": 1e-06,
|
| 2731 |
+
"loss": 0.1125,
|
| 2732 |
+
"step": 389
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 0.12476007677543186,
|
| 2736 |
+
"grad_norm": 1.0481384992599487,
|
| 2737 |
+
"learning_rate": 1e-06,
|
| 2738 |
+
"loss": 0.1107,
|
| 2739 |
+
"step": 390
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 0.12507997440818938,
|
| 2743 |
+
"grad_norm": 1.1148935556411743,
|
| 2744 |
+
"learning_rate": 1e-06,
|
| 2745 |
+
"loss": 0.11,
|
| 2746 |
+
"step": 391
|
| 2747 |
+
},
|
| 2748 |
+
{
|
| 2749 |
+
"epoch": 0.1253998720409469,
|
| 2750 |
+
"grad_norm": 1.3974485397338867,
|
| 2751 |
+
"learning_rate": 1e-06,
|
| 2752 |
+
"loss": 0.115,
|
| 2753 |
+
"step": 392
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 0.1257197696737044,
|
| 2757 |
+
"grad_norm": 2.028317928314209,
|
| 2758 |
+
"learning_rate": 1e-06,
|
| 2759 |
+
"loss": 0.1174,
|
| 2760 |
+
"step": 393
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 0.12603966730646193,
|
| 2764 |
+
"grad_norm": 1.6437934637069702,
|
| 2765 |
+
"learning_rate": 1e-06,
|
| 2766 |
+
"loss": 0.1148,
|
| 2767 |
+
"step": 394
|
| 2768 |
+
},
|
| 2769 |
+
{
|
| 2770 |
+
"epoch": 0.12635956493921946,
|
| 2771 |
+
"grad_norm": 1.5410882234573364,
|
| 2772 |
+
"learning_rate": 1e-06,
|
| 2773 |
+
"loss": 0.1144,
|
| 2774 |
+
"step": 395
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"epoch": 0.12667946257197696,
|
| 2778 |
+
"grad_norm": 1.0200608968734741,
|
| 2779 |
+
"learning_rate": 1e-06,
|
| 2780 |
+
"loss": 0.1117,
|
| 2781 |
+
"step": 396
|
| 2782 |
+
},
|
| 2783 |
+
{
|
| 2784 |
+
"epoch": 0.12699936020473448,
|
| 2785 |
+
"grad_norm": 1.059814453125,
|
| 2786 |
+
"learning_rate": 1e-06,
|
| 2787 |
+
"loss": 0.113,
|
| 2788 |
+
"step": 397
|
| 2789 |
+
},
|
| 2790 |
+
{
|
| 2791 |
+
"epoch": 0.127319257837492,
|
| 2792 |
+
"grad_norm": 1.0535887479782104,
|
| 2793 |
+
"learning_rate": 1e-06,
|
| 2794 |
+
"loss": 0.1162,
|
| 2795 |
+
"step": 398
|
| 2796 |
+
},
|
| 2797 |
+
{
|
| 2798 |
+
"epoch": 0.1276391554702495,
|
| 2799 |
+
"grad_norm": 1.0561606884002686,
|
| 2800 |
+
"learning_rate": 1e-06,
|
| 2801 |
+
"loss": 0.1139,
|
| 2802 |
+
"step": 399
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 0.12795905310300704,
|
| 2806 |
+
"grad_norm": 1.2757177352905273,
|
| 2807 |
+
"learning_rate": 1e-06,
|
| 2808 |
+
"loss": 0.1136,
|
| 2809 |
+
"step": 400
|
| 2810 |
+
}
|
| 2811 |
+
],
|
| 2812 |
+
"logging_steps": 1.0,
|
| 2813 |
+
"max_steps": 3126,
|
| 2814 |
+
"num_input_tokens_seen": 0,
|
| 2815 |
+
"num_train_epochs": 1,
|
| 2816 |
+
"save_steps": 200,
|
| 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 |
+
}
|