Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +2779 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,2779 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 375,
|
| 3 |
+
"best_metric": 2.8448235988616943,
|
| 4 |
+
"best_model_checkpoint": "outputs/checkpoint-375",
|
| 5 |
+
"epoch": 14.970873786407767,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 375,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.038834951456310676,
|
| 14 |
+
"grad_norm": 21.30242156982422,
|
| 15 |
+
"learning_rate": 0.0,
|
| 16 |
+
"loss": 6.5474,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.07766990291262135,
|
| 21 |
+
"grad_norm": 20.775470733642578,
|
| 22 |
+
"learning_rate": 2.0000000000000002e-07,
|
| 23 |
+
"loss": 6.5613,
|
| 24 |
+
"step": 2
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.11650485436893204,
|
| 28 |
+
"grad_norm": 20.96541976928711,
|
| 29 |
+
"learning_rate": 4.0000000000000003e-07,
|
| 30 |
+
"loss": 6.5127,
|
| 31 |
+
"step": 3
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.1553398058252427,
|
| 35 |
+
"grad_norm": 20.376543045043945,
|
| 36 |
+
"learning_rate": 6.000000000000001e-07,
|
| 37 |
+
"loss": 6.4569,
|
| 38 |
+
"step": 4
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.1941747572815534,
|
| 42 |
+
"grad_norm": 19.54267692565918,
|
| 43 |
+
"learning_rate": 8.000000000000001e-07,
|
| 44 |
+
"loss": 6.3743,
|
| 45 |
+
"step": 5
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.23300970873786409,
|
| 49 |
+
"grad_norm": 19.233882904052734,
|
| 50 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 51 |
+
"loss": 6.3899,
|
| 52 |
+
"step": 6
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.27184466019417475,
|
| 56 |
+
"grad_norm": 20.25909423828125,
|
| 57 |
+
"learning_rate": 1.2000000000000002e-06,
|
| 58 |
+
"loss": 6.4415,
|
| 59 |
+
"step": 7
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.3106796116504854,
|
| 63 |
+
"grad_norm": 19.33000373840332,
|
| 64 |
+
"learning_rate": 1.4000000000000001e-06,
|
| 65 |
+
"loss": 6.3191,
|
| 66 |
+
"step": 8
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.34951456310679613,
|
| 70 |
+
"grad_norm": 18.305322647094727,
|
| 71 |
+
"learning_rate": 1.6000000000000001e-06,
|
| 72 |
+
"loss": 6.2681,
|
| 73 |
+
"step": 9
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.3883495145631068,
|
| 77 |
+
"grad_norm": 17.74665069580078,
|
| 78 |
+
"learning_rate": 1.8e-06,
|
| 79 |
+
"loss": 6.4206,
|
| 80 |
+
"step": 10
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.42718446601941745,
|
| 84 |
+
"grad_norm": 14.93736457824707,
|
| 85 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 86 |
+
"loss": 6.2288,
|
| 87 |
+
"step": 11
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.46601941747572817,
|
| 91 |
+
"grad_norm": 14.914277076721191,
|
| 92 |
+
"learning_rate": 2.2e-06,
|
| 93 |
+
"loss": 6.3961,
|
| 94 |
+
"step": 12
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.5048543689320388,
|
| 98 |
+
"grad_norm": 13.266161918640137,
|
| 99 |
+
"learning_rate": 2.4000000000000003e-06,
|
| 100 |
+
"loss": 6.0076,
|
| 101 |
+
"step": 13
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.5436893203883495,
|
| 105 |
+
"grad_norm": 12.377790451049805,
|
| 106 |
+
"learning_rate": 2.6e-06,
|
| 107 |
+
"loss": 6.2259,
|
| 108 |
+
"step": 14
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.5825242718446602,
|
| 112 |
+
"grad_norm": 11.322343826293945,
|
| 113 |
+
"learning_rate": 2.8000000000000003e-06,
|
| 114 |
+
"loss": 6.1832,
|
| 115 |
+
"step": 15
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.6213592233009708,
|
| 119 |
+
"grad_norm": 10.584484100341797,
|
| 120 |
+
"learning_rate": 3e-06,
|
| 121 |
+
"loss": 6.0051,
|
| 122 |
+
"step": 16
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.6601941747572816,
|
| 126 |
+
"grad_norm": 10.82979965209961,
|
| 127 |
+
"learning_rate": 3.2000000000000003e-06,
|
| 128 |
+
"loss": 5.9786,
|
| 129 |
+
"step": 17
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.6990291262135923,
|
| 133 |
+
"grad_norm": 10.112428665161133,
|
| 134 |
+
"learning_rate": 3.4000000000000005e-06,
|
| 135 |
+
"loss": 5.9778,
|
| 136 |
+
"step": 18
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.7378640776699029,
|
| 140 |
+
"grad_norm": 9.44952392578125,
|
| 141 |
+
"learning_rate": 3.6e-06,
|
| 142 |
+
"loss": 5.9459,
|
| 143 |
+
"step": 19
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.7766990291262136,
|
| 147 |
+
"grad_norm": 9.057659149169922,
|
| 148 |
+
"learning_rate": 3.8e-06,
|
| 149 |
+
"loss": 6.0317,
|
| 150 |
+
"step": 20
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.8155339805825242,
|
| 154 |
+
"grad_norm": 9.000926971435547,
|
| 155 |
+
"learning_rate": 4.000000000000001e-06,
|
| 156 |
+
"loss": 5.9749,
|
| 157 |
+
"step": 21
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.8543689320388349,
|
| 161 |
+
"grad_norm": 7.747213840484619,
|
| 162 |
+
"learning_rate": 4.2000000000000004e-06,
|
| 163 |
+
"loss": 5.8036,
|
| 164 |
+
"step": 22
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.8932038834951457,
|
| 168 |
+
"grad_norm": 6.968072891235352,
|
| 169 |
+
"learning_rate": 4.4e-06,
|
| 170 |
+
"loss": 5.7705,
|
| 171 |
+
"step": 23
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.9320388349514563,
|
| 175 |
+
"grad_norm": 7.167684555053711,
|
| 176 |
+
"learning_rate": 4.6e-06,
|
| 177 |
+
"loss": 5.7804,
|
| 178 |
+
"step": 24
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.970873786407767,
|
| 182 |
+
"grad_norm": 6.384294033050537,
|
| 183 |
+
"learning_rate": 4.800000000000001e-06,
|
| 184 |
+
"loss": 5.6137,
|
| 185 |
+
"step": 25
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.970873786407767,
|
| 189 |
+
"eval_loss": 5.6718363761901855,
|
| 190 |
+
"eval_runtime": 2.435,
|
| 191 |
+
"eval_samples_per_second": 9.035,
|
| 192 |
+
"eval_steps_per_second": 2.464,
|
| 193 |
+
"step": 25
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 1.0388349514563107,
|
| 197 |
+
"grad_norm": 12.656047821044922,
|
| 198 |
+
"learning_rate": 5e-06,
|
| 199 |
+
"loss": 11.4741,
|
| 200 |
+
"step": 26
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 1.0776699029126213,
|
| 204 |
+
"grad_norm": 6.408062934875488,
|
| 205 |
+
"learning_rate": 5.2e-06,
|
| 206 |
+
"loss": 5.7835,
|
| 207 |
+
"step": 27
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 1.116504854368932,
|
| 211 |
+
"grad_norm": 6.457642078399658,
|
| 212 |
+
"learning_rate": 5.4e-06,
|
| 213 |
+
"loss": 5.7659,
|
| 214 |
+
"step": 28
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 1.1553398058252426,
|
| 218 |
+
"grad_norm": 6.716769218444824,
|
| 219 |
+
"learning_rate": 5.600000000000001e-06,
|
| 220 |
+
"loss": 5.6133,
|
| 221 |
+
"step": 29
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 1.1941747572815533,
|
| 225 |
+
"grad_norm": 5.562079906463623,
|
| 226 |
+
"learning_rate": 5.8e-06,
|
| 227 |
+
"loss": 5.6588,
|
| 228 |
+
"step": 30
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 1.233009708737864,
|
| 232 |
+
"grad_norm": 5.209117412567139,
|
| 233 |
+
"learning_rate": 6e-06,
|
| 234 |
+
"loss": 5.6118,
|
| 235 |
+
"step": 31
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 1.2718446601941746,
|
| 239 |
+
"grad_norm": 5.505391597747803,
|
| 240 |
+
"learning_rate": 6.2e-06,
|
| 241 |
+
"loss": 5.6468,
|
| 242 |
+
"step": 32
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 1.3106796116504853,
|
| 246 |
+
"grad_norm": 4.989831924438477,
|
| 247 |
+
"learning_rate": 6.4000000000000006e-06,
|
| 248 |
+
"loss": 5.5483,
|
| 249 |
+
"step": 33
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 1.3495145631067962,
|
| 253 |
+
"grad_norm": 5.000854015350342,
|
| 254 |
+
"learning_rate": 6.6e-06,
|
| 255 |
+
"loss": 5.4522,
|
| 256 |
+
"step": 34
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 1.3883495145631068,
|
| 260 |
+
"grad_norm": 4.343570232391357,
|
| 261 |
+
"learning_rate": 6.800000000000001e-06,
|
| 262 |
+
"loss": 5.3562,
|
| 263 |
+
"step": 35
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 1.4271844660194175,
|
| 267 |
+
"grad_norm": 4.40326452255249,
|
| 268 |
+
"learning_rate": 7.000000000000001e-06,
|
| 269 |
+
"loss": 5.4561,
|
| 270 |
+
"step": 36
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 1.4660194174757282,
|
| 274 |
+
"grad_norm": 4.1591901779174805,
|
| 275 |
+
"learning_rate": 7.2e-06,
|
| 276 |
+
"loss": 5.3806,
|
| 277 |
+
"step": 37
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 1.5048543689320388,
|
| 281 |
+
"grad_norm": 4.1347246170043945,
|
| 282 |
+
"learning_rate": 7.4e-06,
|
| 283 |
+
"loss": 5.419,
|
| 284 |
+
"step": 38
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 1.5436893203883495,
|
| 288 |
+
"grad_norm": 4.123111248016357,
|
| 289 |
+
"learning_rate": 7.6e-06,
|
| 290 |
+
"loss": 5.2831,
|
| 291 |
+
"step": 39
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 1.5825242718446602,
|
| 295 |
+
"grad_norm": 4.009028911590576,
|
| 296 |
+
"learning_rate": 7.8e-06,
|
| 297 |
+
"loss": 5.332,
|
| 298 |
+
"step": 40
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 1.6213592233009708,
|
| 302 |
+
"grad_norm": 4.013438701629639,
|
| 303 |
+
"learning_rate": 8.000000000000001e-06,
|
| 304 |
+
"loss": 5.2177,
|
| 305 |
+
"step": 41
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 1.6601941747572817,
|
| 309 |
+
"grad_norm": 3.698003053665161,
|
| 310 |
+
"learning_rate": 8.200000000000001e-06,
|
| 311 |
+
"loss": 5.3019,
|
| 312 |
+
"step": 42
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 1.6990291262135924,
|
| 316 |
+
"grad_norm": 3.66217041015625,
|
| 317 |
+
"learning_rate": 8.400000000000001e-06,
|
| 318 |
+
"loss": 5.1967,
|
| 319 |
+
"step": 43
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 1.737864077669903,
|
| 323 |
+
"grad_norm": 3.455019235610962,
|
| 324 |
+
"learning_rate": 8.599999999999999e-06,
|
| 325 |
+
"loss": 5.1573,
|
| 326 |
+
"step": 44
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 1.7766990291262137,
|
| 330 |
+
"grad_norm": 3.5593278408050537,
|
| 331 |
+
"learning_rate": 8.8e-06,
|
| 332 |
+
"loss": 5.1463,
|
| 333 |
+
"step": 45
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 1.8155339805825244,
|
| 337 |
+
"grad_norm": 3.332477331161499,
|
| 338 |
+
"learning_rate": 9e-06,
|
| 339 |
+
"loss": 5.1732,
|
| 340 |
+
"step": 46
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 1.854368932038835,
|
| 344 |
+
"grad_norm": 3.2428054809570312,
|
| 345 |
+
"learning_rate": 9.2e-06,
|
| 346 |
+
"loss": 5.0962,
|
| 347 |
+
"step": 47
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 1.8932038834951457,
|
| 351 |
+
"grad_norm": 3.339063882827759,
|
| 352 |
+
"learning_rate": 9.4e-06,
|
| 353 |
+
"loss": 5.0253,
|
| 354 |
+
"step": 48
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.9320388349514563,
|
| 358 |
+
"grad_norm": 3.4746124744415283,
|
| 359 |
+
"learning_rate": 9.600000000000001e-06,
|
| 360 |
+
"loss": 5.1363,
|
| 361 |
+
"step": 49
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 1.970873786407767,
|
| 365 |
+
"grad_norm": 3.371466875076294,
|
| 366 |
+
"learning_rate": 9.800000000000001e-06,
|
| 367 |
+
"loss": 5.1445,
|
| 368 |
+
"step": 50
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.970873786407767,
|
| 372 |
+
"eval_loss": 5.052736282348633,
|
| 373 |
+
"eval_runtime": 0.9737,
|
| 374 |
+
"eval_samples_per_second": 22.593,
|
| 375 |
+
"eval_steps_per_second": 6.162,
|
| 376 |
+
"step": 50
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 2.0388349514563107,
|
| 380 |
+
"grad_norm": 5.6498637199401855,
|
| 381 |
+
"learning_rate": 1e-05,
|
| 382 |
+
"loss": 10.112,
|
| 383 |
+
"step": 51
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 2.0776699029126213,
|
| 387 |
+
"grad_norm": 3.1301138401031494,
|
| 388 |
+
"learning_rate": 1.02e-05,
|
| 389 |
+
"loss": 5.1063,
|
| 390 |
+
"step": 52
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 2.116504854368932,
|
| 394 |
+
"grad_norm": 3.452958345413208,
|
| 395 |
+
"learning_rate": 1.04e-05,
|
| 396 |
+
"loss": 5.0082,
|
| 397 |
+
"step": 53
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 2.1553398058252426,
|
| 401 |
+
"grad_norm": 3.1977169513702393,
|
| 402 |
+
"learning_rate": 1.06e-05,
|
| 403 |
+
"loss": 4.9698,
|
| 404 |
+
"step": 54
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 2.1941747572815533,
|
| 408 |
+
"grad_norm": 2.6776535511016846,
|
| 409 |
+
"learning_rate": 1.08e-05,
|
| 410 |
+
"loss": 4.9449,
|
| 411 |
+
"step": 55
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 2.233009708737864,
|
| 415 |
+
"grad_norm": 3.5574913024902344,
|
| 416 |
+
"learning_rate": 1.1000000000000001e-05,
|
| 417 |
+
"loss": 5.0442,
|
| 418 |
+
"step": 56
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 2.2718446601941746,
|
| 422 |
+
"grad_norm": 2.867915391921997,
|
| 423 |
+
"learning_rate": 1.1200000000000001e-05,
|
| 424 |
+
"loss": 4.8769,
|
| 425 |
+
"step": 57
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 2.3106796116504853,
|
| 429 |
+
"grad_norm": 2.764223098754883,
|
| 430 |
+
"learning_rate": 1.1400000000000001e-05,
|
| 431 |
+
"loss": 4.9286,
|
| 432 |
+
"step": 58
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 2.349514563106796,
|
| 436 |
+
"grad_norm": 3.816723585128784,
|
| 437 |
+
"learning_rate": 1.16e-05,
|
| 438 |
+
"loss": 4.8921,
|
| 439 |
+
"step": 59
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 2.3883495145631066,
|
| 443 |
+
"grad_norm": 3.161980152130127,
|
| 444 |
+
"learning_rate": 1.18e-05,
|
| 445 |
+
"loss": 4.916,
|
| 446 |
+
"step": 60
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 2.4271844660194173,
|
| 450 |
+
"grad_norm": 2.8373942375183105,
|
| 451 |
+
"learning_rate": 1.2e-05,
|
| 452 |
+
"loss": 4.8942,
|
| 453 |
+
"step": 61
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 2.466019417475728,
|
| 457 |
+
"grad_norm": 2.8898000717163086,
|
| 458 |
+
"learning_rate": 1.22e-05,
|
| 459 |
+
"loss": 4.8206,
|
| 460 |
+
"step": 62
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 2.5048543689320386,
|
| 464 |
+
"grad_norm": 2.726362943649292,
|
| 465 |
+
"learning_rate": 1.24e-05,
|
| 466 |
+
"loss": 4.846,
|
| 467 |
+
"step": 63
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 2.5436893203883493,
|
| 471 |
+
"grad_norm": 2.73665714263916,
|
| 472 |
+
"learning_rate": 1.2600000000000001e-05,
|
| 473 |
+
"loss": 4.8375,
|
| 474 |
+
"step": 64
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 2.58252427184466,
|
| 478 |
+
"grad_norm": 3.1228106021881104,
|
| 479 |
+
"learning_rate": 1.2800000000000001e-05,
|
| 480 |
+
"loss": 4.7526,
|
| 481 |
+
"step": 65
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 2.6213592233009706,
|
| 485 |
+
"grad_norm": 2.9702351093292236,
|
| 486 |
+
"learning_rate": 1.3000000000000001e-05,
|
| 487 |
+
"loss": 4.8024,
|
| 488 |
+
"step": 66
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 2.6601941747572817,
|
| 492 |
+
"grad_norm": 3.0533952713012695,
|
| 493 |
+
"learning_rate": 1.32e-05,
|
| 494 |
+
"loss": 4.7883,
|
| 495 |
+
"step": 67
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 2.6990291262135924,
|
| 499 |
+
"grad_norm": 3.1949095726013184,
|
| 500 |
+
"learning_rate": 1.3400000000000002e-05,
|
| 501 |
+
"loss": 4.8197,
|
| 502 |
+
"step": 68
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 2.737864077669903,
|
| 506 |
+
"grad_norm": 3.399998426437378,
|
| 507 |
+
"learning_rate": 1.3600000000000002e-05,
|
| 508 |
+
"loss": 4.6677,
|
| 509 |
+
"step": 69
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 2.7766990291262137,
|
| 513 |
+
"grad_norm": 2.80118465423584,
|
| 514 |
+
"learning_rate": 1.3800000000000002e-05,
|
| 515 |
+
"loss": 4.6291,
|
| 516 |
+
"step": 70
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 2.8155339805825244,
|
| 520 |
+
"grad_norm": 2.8477330207824707,
|
| 521 |
+
"learning_rate": 1.4000000000000001e-05,
|
| 522 |
+
"loss": 4.7767,
|
| 523 |
+
"step": 71
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 2.854368932038835,
|
| 527 |
+
"grad_norm": 2.6895911693573,
|
| 528 |
+
"learning_rate": 1.42e-05,
|
| 529 |
+
"loss": 4.7057,
|
| 530 |
+
"step": 72
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 2.8932038834951457,
|
| 534 |
+
"grad_norm": 2.914586067199707,
|
| 535 |
+
"learning_rate": 1.44e-05,
|
| 536 |
+
"loss": 4.6386,
|
| 537 |
+
"step": 73
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 2.9320388349514563,
|
| 541 |
+
"grad_norm": 2.6184370517730713,
|
| 542 |
+
"learning_rate": 1.4599999999999999e-05,
|
| 543 |
+
"loss": 4.6679,
|
| 544 |
+
"step": 74
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 2.970873786407767,
|
| 548 |
+
"grad_norm": 3.00891375541687,
|
| 549 |
+
"learning_rate": 1.48e-05,
|
| 550 |
+
"loss": 4.6319,
|
| 551 |
+
"step": 75
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 2.970873786407767,
|
| 555 |
+
"eval_loss": 4.614713668823242,
|
| 556 |
+
"eval_runtime": 0.9702,
|
| 557 |
+
"eval_samples_per_second": 22.675,
|
| 558 |
+
"eval_steps_per_second": 6.184,
|
| 559 |
+
"step": 75
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 3.0388349514563107,
|
| 563 |
+
"grad_norm": 5.222214221954346,
|
| 564 |
+
"learning_rate": 1.5e-05,
|
| 565 |
+
"loss": 9.2212,
|
| 566 |
+
"step": 76
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 3.0776699029126213,
|
| 570 |
+
"grad_norm": 2.716062307357788,
|
| 571 |
+
"learning_rate": 1.52e-05,
|
| 572 |
+
"loss": 4.6294,
|
| 573 |
+
"step": 77
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 3.116504854368932,
|
| 577 |
+
"grad_norm": 2.503143548965454,
|
| 578 |
+
"learning_rate": 1.54e-05,
|
| 579 |
+
"loss": 4.5572,
|
| 580 |
+
"step": 78
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 3.1553398058252426,
|
| 584 |
+
"grad_norm": 2.9183573722839355,
|
| 585 |
+
"learning_rate": 1.56e-05,
|
| 586 |
+
"loss": 4.453,
|
| 587 |
+
"step": 79
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 3.1941747572815533,
|
| 591 |
+
"grad_norm": 2.7854349613189697,
|
| 592 |
+
"learning_rate": 1.58e-05,
|
| 593 |
+
"loss": 4.5746,
|
| 594 |
+
"step": 80
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 3.233009708737864,
|
| 598 |
+
"grad_norm": 2.8391106128692627,
|
| 599 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 600 |
+
"loss": 4.5228,
|
| 601 |
+
"step": 81
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 3.2718446601941746,
|
| 605 |
+
"grad_norm": 2.5229265689849854,
|
| 606 |
+
"learning_rate": 1.62e-05,
|
| 607 |
+
"loss": 4.4692,
|
| 608 |
+
"step": 82
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 3.3106796116504853,
|
| 612 |
+
"grad_norm": 2.643170118331909,
|
| 613 |
+
"learning_rate": 1.6400000000000002e-05,
|
| 614 |
+
"loss": 4.498,
|
| 615 |
+
"step": 83
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 3.349514563106796,
|
| 619 |
+
"grad_norm": 2.542393922805786,
|
| 620 |
+
"learning_rate": 1.66e-05,
|
| 621 |
+
"loss": 4.4816,
|
| 622 |
+
"step": 84
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 3.3883495145631066,
|
| 626 |
+
"grad_norm": 2.563282012939453,
|
| 627 |
+
"learning_rate": 1.6800000000000002e-05,
|
| 628 |
+
"loss": 4.4824,
|
| 629 |
+
"step": 85
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 3.4271844660194173,
|
| 633 |
+
"grad_norm": 2.698516368865967,
|
| 634 |
+
"learning_rate": 1.7000000000000003e-05,
|
| 635 |
+
"loss": 4.4717,
|
| 636 |
+
"step": 86
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 3.466019417475728,
|
| 640 |
+
"grad_norm": 2.936776638031006,
|
| 641 |
+
"learning_rate": 1.7199999999999998e-05,
|
| 642 |
+
"loss": 4.346,
|
| 643 |
+
"step": 87
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"epoch": 3.5048543689320386,
|
| 647 |
+
"grad_norm": 2.9594175815582275,
|
| 648 |
+
"learning_rate": 1.74e-05,
|
| 649 |
+
"loss": 4.3689,
|
| 650 |
+
"step": 88
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 3.5436893203883493,
|
| 654 |
+
"grad_norm": 3.02431583404541,
|
| 655 |
+
"learning_rate": 1.76e-05,
|
| 656 |
+
"loss": 4.3922,
|
| 657 |
+
"step": 89
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 3.58252427184466,
|
| 661 |
+
"grad_norm": 3.238933563232422,
|
| 662 |
+
"learning_rate": 1.78e-05,
|
| 663 |
+
"loss": 4.4046,
|
| 664 |
+
"step": 90
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 3.6213592233009706,
|
| 668 |
+
"grad_norm": 3.368084192276001,
|
| 669 |
+
"learning_rate": 1.8e-05,
|
| 670 |
+
"loss": 4.3768,
|
| 671 |
+
"step": 91
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 3.6601941747572817,
|
| 675 |
+
"grad_norm": 3.8072586059570312,
|
| 676 |
+
"learning_rate": 1.8200000000000002e-05,
|
| 677 |
+
"loss": 4.3188,
|
| 678 |
+
"step": 92
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 3.6990291262135924,
|
| 682 |
+
"grad_norm": 3.2370452880859375,
|
| 683 |
+
"learning_rate": 1.84e-05,
|
| 684 |
+
"loss": 4.3368,
|
| 685 |
+
"step": 93
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 3.737864077669903,
|
| 689 |
+
"grad_norm": 3.302961826324463,
|
| 690 |
+
"learning_rate": 1.86e-05,
|
| 691 |
+
"loss": 4.3339,
|
| 692 |
+
"step": 94
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 3.7766990291262137,
|
| 696 |
+
"grad_norm": 3.5947256088256836,
|
| 697 |
+
"learning_rate": 1.88e-05,
|
| 698 |
+
"loss": 4.2763,
|
| 699 |
+
"step": 95
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 3.8155339805825244,
|
| 703 |
+
"grad_norm": 2.955308437347412,
|
| 704 |
+
"learning_rate": 1.9e-05,
|
| 705 |
+
"loss": 4.3941,
|
| 706 |
+
"step": 96
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 3.854368932038835,
|
| 710 |
+
"grad_norm": 3.303628444671631,
|
| 711 |
+
"learning_rate": 1.9200000000000003e-05,
|
| 712 |
+
"loss": 4.2748,
|
| 713 |
+
"step": 97
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 3.8932038834951457,
|
| 717 |
+
"grad_norm": 2.7507269382476807,
|
| 718 |
+
"learning_rate": 1.94e-05,
|
| 719 |
+
"loss": 4.2881,
|
| 720 |
+
"step": 98
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 3.9320388349514563,
|
| 724 |
+
"grad_norm": 2.6451849937438965,
|
| 725 |
+
"learning_rate": 1.9600000000000002e-05,
|
| 726 |
+
"loss": 4.3818,
|
| 727 |
+
"step": 99
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 3.970873786407767,
|
| 731 |
+
"grad_norm": 4.112302780151367,
|
| 732 |
+
"learning_rate": 1.9800000000000004e-05,
|
| 733 |
+
"loss": 4.2882,
|
| 734 |
+
"step": 100
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 3.970873786407767,
|
| 738 |
+
"eval_loss": 4.266085624694824,
|
| 739 |
+
"eval_runtime": 1.036,
|
| 740 |
+
"eval_samples_per_second": 21.235,
|
| 741 |
+
"eval_steps_per_second": 5.791,
|
| 742 |
+
"step": 100
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"epoch": 4.038834951456311,
|
| 746 |
+
"grad_norm": 5.2990498542785645,
|
| 747 |
+
"learning_rate": 2e-05,
|
| 748 |
+
"loss": 8.631,
|
| 749 |
+
"step": 101
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 4.077669902912621,
|
| 753 |
+
"grad_norm": 3.757814407348633,
|
| 754 |
+
"learning_rate": 2.0200000000000003e-05,
|
| 755 |
+
"loss": 4.2183,
|
| 756 |
+
"step": 102
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 4.116504854368932,
|
| 760 |
+
"grad_norm": 2.905704975128174,
|
| 761 |
+
"learning_rate": 2.04e-05,
|
| 762 |
+
"loss": 4.1782,
|
| 763 |
+
"step": 103
|
| 764 |
+
},
|
| 765 |
+
{
|
| 766 |
+
"epoch": 4.155339805825243,
|
| 767 |
+
"grad_norm": 3.7264492511749268,
|
| 768 |
+
"learning_rate": 2.06e-05,
|
| 769 |
+
"loss": 4.2959,
|
| 770 |
+
"step": 104
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 4.194174757281553,
|
| 774 |
+
"grad_norm": 3.9989054203033447,
|
| 775 |
+
"learning_rate": 2.08e-05,
|
| 776 |
+
"loss": 4.1876,
|
| 777 |
+
"step": 105
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 4.233009708737864,
|
| 781 |
+
"grad_norm": 2.978239059448242,
|
| 782 |
+
"learning_rate": 2.1e-05,
|
| 783 |
+
"loss": 4.1484,
|
| 784 |
+
"step": 106
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"epoch": 4.271844660194175,
|
| 788 |
+
"grad_norm": 3.223487138748169,
|
| 789 |
+
"learning_rate": 2.12e-05,
|
| 790 |
+
"loss": 4.1501,
|
| 791 |
+
"step": 107
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 4.310679611650485,
|
| 795 |
+
"grad_norm": 3.035008668899536,
|
| 796 |
+
"learning_rate": 2.1400000000000002e-05,
|
| 797 |
+
"loss": 4.1316,
|
| 798 |
+
"step": 108
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 4.349514563106796,
|
| 802 |
+
"grad_norm": 2.878307819366455,
|
| 803 |
+
"learning_rate": 2.16e-05,
|
| 804 |
+
"loss": 4.1824,
|
| 805 |
+
"step": 109
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"epoch": 4.388349514563107,
|
| 809 |
+
"grad_norm": 3.095815420150757,
|
| 810 |
+
"learning_rate": 2.18e-05,
|
| 811 |
+
"loss": 4.1726,
|
| 812 |
+
"step": 110
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 4.427184466019417,
|
| 816 |
+
"grad_norm": 3.0754470825195312,
|
| 817 |
+
"learning_rate": 2.2000000000000003e-05,
|
| 818 |
+
"loss": 3.9618,
|
| 819 |
+
"step": 111
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 4.466019417475728,
|
| 823 |
+
"grad_norm": 3.4234559535980225,
|
| 824 |
+
"learning_rate": 2.22e-05,
|
| 825 |
+
"loss": 4.0646,
|
| 826 |
+
"step": 112
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 4.504854368932039,
|
| 830 |
+
"grad_norm": 3.2128183841705322,
|
| 831 |
+
"learning_rate": 2.2400000000000002e-05,
|
| 832 |
+
"loss": 4.0639,
|
| 833 |
+
"step": 113
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 4.543689320388349,
|
| 837 |
+
"grad_norm": 2.9789934158325195,
|
| 838 |
+
"learning_rate": 2.26e-05,
|
| 839 |
+
"loss": 4.1373,
|
| 840 |
+
"step": 114
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 4.58252427184466,
|
| 844 |
+
"grad_norm": 2.5928032398223877,
|
| 845 |
+
"learning_rate": 2.2800000000000002e-05,
|
| 846 |
+
"loss": 3.9855,
|
| 847 |
+
"step": 115
|
| 848 |
+
},
|
| 849 |
+
{
|
| 850 |
+
"epoch": 4.621359223300971,
|
| 851 |
+
"grad_norm": 3.082489252090454,
|
| 852 |
+
"learning_rate": 2.3000000000000003e-05,
|
| 853 |
+
"loss": 4.1163,
|
| 854 |
+
"step": 116
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 4.660194174757281,
|
| 858 |
+
"grad_norm": 3.028413772583008,
|
| 859 |
+
"learning_rate": 2.32e-05,
|
| 860 |
+
"loss": 4.0571,
|
| 861 |
+
"step": 117
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 4.699029126213592,
|
| 865 |
+
"grad_norm": 2.8744428157806396,
|
| 866 |
+
"learning_rate": 2.3400000000000003e-05,
|
| 867 |
+
"loss": 4.027,
|
| 868 |
+
"step": 118
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 4.737864077669903,
|
| 872 |
+
"grad_norm": 2.866056442260742,
|
| 873 |
+
"learning_rate": 2.36e-05,
|
| 874 |
+
"loss": 4.0299,
|
| 875 |
+
"step": 119
|
| 876 |
+
},
|
| 877 |
+
{
|
| 878 |
+
"epoch": 4.776699029126213,
|
| 879 |
+
"grad_norm": 2.75072979927063,
|
| 880 |
+
"learning_rate": 2.38e-05,
|
| 881 |
+
"loss": 3.993,
|
| 882 |
+
"step": 120
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 4.815533980582524,
|
| 886 |
+
"grad_norm": 2.8751604557037354,
|
| 887 |
+
"learning_rate": 2.4e-05,
|
| 888 |
+
"loss": 3.9961,
|
| 889 |
+
"step": 121
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 4.854368932038835,
|
| 893 |
+
"grad_norm": 2.5905075073242188,
|
| 894 |
+
"learning_rate": 2.4200000000000002e-05,
|
| 895 |
+
"loss": 3.9582,
|
| 896 |
+
"step": 122
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 4.893203883495145,
|
| 900 |
+
"grad_norm": 3.143044948577881,
|
| 901 |
+
"learning_rate": 2.44e-05,
|
| 902 |
+
"loss": 3.9464,
|
| 903 |
+
"step": 123
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 4.932038834951456,
|
| 907 |
+
"grad_norm": 2.6397016048431396,
|
| 908 |
+
"learning_rate": 2.46e-05,
|
| 909 |
+
"loss": 4.0075,
|
| 910 |
+
"step": 124
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 4.970873786407767,
|
| 914 |
+
"grad_norm": 3.2383229732513428,
|
| 915 |
+
"learning_rate": 2.48e-05,
|
| 916 |
+
"loss": 3.9822,
|
| 917 |
+
"step": 125
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 4.970873786407767,
|
| 921 |
+
"eval_loss": 3.980665445327759,
|
| 922 |
+
"eval_runtime": 1.0248,
|
| 923 |
+
"eval_samples_per_second": 21.467,
|
| 924 |
+
"eval_steps_per_second": 5.855,
|
| 925 |
+
"step": 125
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 5.038834951456311,
|
| 929 |
+
"grad_norm": 5.962584495544434,
|
| 930 |
+
"learning_rate": 2.5e-05,
|
| 931 |
+
"loss": 7.7604,
|
| 932 |
+
"step": 126
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 5.077669902912621,
|
| 936 |
+
"grad_norm": 3.243708610534668,
|
| 937 |
+
"learning_rate": 2.5200000000000003e-05,
|
| 938 |
+
"loss": 3.9438,
|
| 939 |
+
"step": 127
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 5.116504854368932,
|
| 943 |
+
"grad_norm": 2.763148307800293,
|
| 944 |
+
"learning_rate": 2.54e-05,
|
| 945 |
+
"loss": 3.8661,
|
| 946 |
+
"step": 128
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 5.155339805825243,
|
| 950 |
+
"grad_norm": 2.6233339309692383,
|
| 951 |
+
"learning_rate": 2.5600000000000002e-05,
|
| 952 |
+
"loss": 3.9006,
|
| 953 |
+
"step": 129
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 5.194174757281553,
|
| 957 |
+
"grad_norm": 3.1037437915802,
|
| 958 |
+
"learning_rate": 2.58e-05,
|
| 959 |
+
"loss": 3.9066,
|
| 960 |
+
"step": 130
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 5.233009708737864,
|
| 964 |
+
"grad_norm": 3.3434383869171143,
|
| 965 |
+
"learning_rate": 2.6000000000000002e-05,
|
| 966 |
+
"loss": 3.8425,
|
| 967 |
+
"step": 131
|
| 968 |
+
},
|
| 969 |
+
{
|
| 970 |
+
"epoch": 5.271844660194175,
|
| 971 |
+
"grad_norm": 3.0016958713531494,
|
| 972 |
+
"learning_rate": 2.6200000000000003e-05,
|
| 973 |
+
"loss": 3.8723,
|
| 974 |
+
"step": 132
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 5.310679611650485,
|
| 978 |
+
"grad_norm": 3.2040951251983643,
|
| 979 |
+
"learning_rate": 2.64e-05,
|
| 980 |
+
"loss": 3.8326,
|
| 981 |
+
"step": 133
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 5.349514563106796,
|
| 985 |
+
"grad_norm": 3.892890453338623,
|
| 986 |
+
"learning_rate": 2.6600000000000003e-05,
|
| 987 |
+
"loss": 3.9277,
|
| 988 |
+
"step": 134
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"epoch": 5.388349514563107,
|
| 992 |
+
"grad_norm": 3.3505635261535645,
|
| 993 |
+
"learning_rate": 2.6800000000000004e-05,
|
| 994 |
+
"loss": 3.7381,
|
| 995 |
+
"step": 135
|
| 996 |
+
},
|
| 997 |
+
{
|
| 998 |
+
"epoch": 5.427184466019417,
|
| 999 |
+
"grad_norm": 3.60493803024292,
|
| 1000 |
+
"learning_rate": 2.7000000000000002e-05,
|
| 1001 |
+
"loss": 3.9003,
|
| 1002 |
+
"step": 136
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 5.466019417475728,
|
| 1006 |
+
"grad_norm": 3.3468196392059326,
|
| 1007 |
+
"learning_rate": 2.7200000000000004e-05,
|
| 1008 |
+
"loss": 3.829,
|
| 1009 |
+
"step": 137
|
| 1010 |
+
},
|
| 1011 |
+
{
|
| 1012 |
+
"epoch": 5.504854368932039,
|
| 1013 |
+
"grad_norm": 2.7208919525146484,
|
| 1014 |
+
"learning_rate": 2.7400000000000002e-05,
|
| 1015 |
+
"loss": 3.7987,
|
| 1016 |
+
"step": 138
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 5.543689320388349,
|
| 1020 |
+
"grad_norm": 4.0348920822143555,
|
| 1021 |
+
"learning_rate": 2.7600000000000003e-05,
|
| 1022 |
+
"loss": 3.8318,
|
| 1023 |
+
"step": 139
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 5.58252427184466,
|
| 1027 |
+
"grad_norm": 3.560403347015381,
|
| 1028 |
+
"learning_rate": 2.7800000000000005e-05,
|
| 1029 |
+
"loss": 3.763,
|
| 1030 |
+
"step": 140
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 5.621359223300971,
|
| 1034 |
+
"grad_norm": 3.262423515319824,
|
| 1035 |
+
"learning_rate": 2.8000000000000003e-05,
|
| 1036 |
+
"loss": 3.7441,
|
| 1037 |
+
"step": 141
|
| 1038 |
+
},
|
| 1039 |
+
{
|
| 1040 |
+
"epoch": 5.660194174757281,
|
| 1041 |
+
"grad_norm": 2.7930023670196533,
|
| 1042 |
+
"learning_rate": 2.8199999999999998e-05,
|
| 1043 |
+
"loss": 3.7323,
|
| 1044 |
+
"step": 142
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 5.699029126213592,
|
| 1048 |
+
"grad_norm": 2.5322391986846924,
|
| 1049 |
+
"learning_rate": 2.84e-05,
|
| 1050 |
+
"loss": 3.6681,
|
| 1051 |
+
"step": 143
|
| 1052 |
+
},
|
| 1053 |
+
{
|
| 1054 |
+
"epoch": 5.737864077669903,
|
| 1055 |
+
"grad_norm": 4.258012294769287,
|
| 1056 |
+
"learning_rate": 2.86e-05,
|
| 1057 |
+
"loss": 3.7049,
|
| 1058 |
+
"step": 144
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 5.776699029126213,
|
| 1062 |
+
"grad_norm": 3.0756101608276367,
|
| 1063 |
+
"learning_rate": 2.88e-05,
|
| 1064 |
+
"loss": 3.7184,
|
| 1065 |
+
"step": 145
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 5.815533980582524,
|
| 1069 |
+
"grad_norm": 3.0040361881256104,
|
| 1070 |
+
"learning_rate": 2.9e-05,
|
| 1071 |
+
"loss": 3.6077,
|
| 1072 |
+
"step": 146
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 5.854368932038835,
|
| 1076 |
+
"grad_norm": 4.292761325836182,
|
| 1077 |
+
"learning_rate": 2.9199999999999998e-05,
|
| 1078 |
+
"loss": 3.7214,
|
| 1079 |
+
"step": 147
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 5.893203883495145,
|
| 1083 |
+
"grad_norm": 2.876159906387329,
|
| 1084 |
+
"learning_rate": 2.94e-05,
|
| 1085 |
+
"loss": 3.6643,
|
| 1086 |
+
"step": 148
|
| 1087 |
+
},
|
| 1088 |
+
{
|
| 1089 |
+
"epoch": 5.932038834951456,
|
| 1090 |
+
"grad_norm": 3.1686434745788574,
|
| 1091 |
+
"learning_rate": 2.96e-05,
|
| 1092 |
+
"loss": 3.68,
|
| 1093 |
+
"step": 149
|
| 1094 |
+
},
|
| 1095 |
+
{
|
| 1096 |
+
"epoch": 5.970873786407767,
|
| 1097 |
+
"grad_norm": 3.1515626907348633,
|
| 1098 |
+
"learning_rate": 2.98e-05,
|
| 1099 |
+
"loss": 3.6581,
|
| 1100 |
+
"step": 150
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 5.970873786407767,
|
| 1104 |
+
"eval_loss": 3.7385447025299072,
|
| 1105 |
+
"eval_runtime": 1.0393,
|
| 1106 |
+
"eval_samples_per_second": 21.169,
|
| 1107 |
+
"eval_steps_per_second": 5.773,
|
| 1108 |
+
"step": 150
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 6.038834951456311,
|
| 1112 |
+
"grad_norm": 6.013641834259033,
|
| 1113 |
+
"learning_rate": 3e-05,
|
| 1114 |
+
"loss": 7.2601,
|
| 1115 |
+
"step": 151
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 6.077669902912621,
|
| 1119 |
+
"grad_norm": 3.0433292388916016,
|
| 1120 |
+
"learning_rate": 3.02e-05,
|
| 1121 |
+
"loss": 3.626,
|
| 1122 |
+
"step": 152
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 6.116504854368932,
|
| 1126 |
+
"grad_norm": 2.9623515605926514,
|
| 1127 |
+
"learning_rate": 3.04e-05,
|
| 1128 |
+
"loss": 3.5856,
|
| 1129 |
+
"step": 153
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 6.155339805825243,
|
| 1133 |
+
"grad_norm": 3.333615779876709,
|
| 1134 |
+
"learning_rate": 3.06e-05,
|
| 1135 |
+
"loss": 3.6268,
|
| 1136 |
+
"step": 154
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 6.194174757281553,
|
| 1140 |
+
"grad_norm": 3.0843307971954346,
|
| 1141 |
+
"learning_rate": 3.08e-05,
|
| 1142 |
+
"loss": 3.5651,
|
| 1143 |
+
"step": 155
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 6.233009708737864,
|
| 1147 |
+
"grad_norm": 2.859063148498535,
|
| 1148 |
+
"learning_rate": 3.1e-05,
|
| 1149 |
+
"loss": 3.5464,
|
| 1150 |
+
"step": 156
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 6.271844660194175,
|
| 1154 |
+
"grad_norm": 2.92948842048645,
|
| 1155 |
+
"learning_rate": 3.12e-05,
|
| 1156 |
+
"loss": 3.6385,
|
| 1157 |
+
"step": 157
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 6.310679611650485,
|
| 1161 |
+
"grad_norm": 3.552112340927124,
|
| 1162 |
+
"learning_rate": 3.1400000000000004e-05,
|
| 1163 |
+
"loss": 3.5479,
|
| 1164 |
+
"step": 158
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 6.349514563106796,
|
| 1168 |
+
"grad_norm": 2.9934771060943604,
|
| 1169 |
+
"learning_rate": 3.16e-05,
|
| 1170 |
+
"loss": 3.5697,
|
| 1171 |
+
"step": 159
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 6.388349514563107,
|
| 1175 |
+
"grad_norm": 2.595054864883423,
|
| 1176 |
+
"learning_rate": 3.18e-05,
|
| 1177 |
+
"loss": 3.4817,
|
| 1178 |
+
"step": 160
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 6.427184466019417,
|
| 1182 |
+
"grad_norm": 3.077573537826538,
|
| 1183 |
+
"learning_rate": 3.2000000000000005e-05,
|
| 1184 |
+
"loss": 3.5286,
|
| 1185 |
+
"step": 161
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 6.466019417475728,
|
| 1189 |
+
"grad_norm": 2.5149052143096924,
|
| 1190 |
+
"learning_rate": 3.2200000000000003e-05,
|
| 1191 |
+
"loss": 3.6065,
|
| 1192 |
+
"step": 162
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 6.504854368932039,
|
| 1196 |
+
"grad_norm": 2.6401753425598145,
|
| 1197 |
+
"learning_rate": 3.24e-05,
|
| 1198 |
+
"loss": 3.4707,
|
| 1199 |
+
"step": 163
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 6.543689320388349,
|
| 1203 |
+
"grad_norm": 2.725781202316284,
|
| 1204 |
+
"learning_rate": 3.26e-05,
|
| 1205 |
+
"loss": 3.5645,
|
| 1206 |
+
"step": 164
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 6.58252427184466,
|
| 1210 |
+
"grad_norm": 2.7084786891937256,
|
| 1211 |
+
"learning_rate": 3.2800000000000004e-05,
|
| 1212 |
+
"loss": 3.5482,
|
| 1213 |
+
"step": 165
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 6.621359223300971,
|
| 1217 |
+
"grad_norm": 2.6076486110687256,
|
| 1218 |
+
"learning_rate": 3.3e-05,
|
| 1219 |
+
"loss": 3.4842,
|
| 1220 |
+
"step": 166
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 6.660194174757281,
|
| 1224 |
+
"grad_norm": 3.0237390995025635,
|
| 1225 |
+
"learning_rate": 3.32e-05,
|
| 1226 |
+
"loss": 3.5313,
|
| 1227 |
+
"step": 167
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 6.699029126213592,
|
| 1231 |
+
"grad_norm": 2.807459831237793,
|
| 1232 |
+
"learning_rate": 3.3400000000000005e-05,
|
| 1233 |
+
"loss": 3.5354,
|
| 1234 |
+
"step": 168
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 6.737864077669903,
|
| 1238 |
+
"grad_norm": 3.13301420211792,
|
| 1239 |
+
"learning_rate": 3.3600000000000004e-05,
|
| 1240 |
+
"loss": 3.4923,
|
| 1241 |
+
"step": 169
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 6.776699029126213,
|
| 1245 |
+
"grad_norm": 2.5862674713134766,
|
| 1246 |
+
"learning_rate": 3.38e-05,
|
| 1247 |
+
"loss": 3.5315,
|
| 1248 |
+
"step": 170
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 6.815533980582524,
|
| 1252 |
+
"grad_norm": 3.192603588104248,
|
| 1253 |
+
"learning_rate": 3.4000000000000007e-05,
|
| 1254 |
+
"loss": 3.4937,
|
| 1255 |
+
"step": 171
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 6.854368932038835,
|
| 1259 |
+
"grad_norm": 2.440667152404785,
|
| 1260 |
+
"learning_rate": 3.4200000000000005e-05,
|
| 1261 |
+
"loss": 3.4632,
|
| 1262 |
+
"step": 172
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 6.893203883495145,
|
| 1266 |
+
"grad_norm": 3.0425989627838135,
|
| 1267 |
+
"learning_rate": 3.4399999999999996e-05,
|
| 1268 |
+
"loss": 3.4456,
|
| 1269 |
+
"step": 173
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 6.932038834951456,
|
| 1273 |
+
"grad_norm": 3.369929313659668,
|
| 1274 |
+
"learning_rate": 3.46e-05,
|
| 1275 |
+
"loss": 3.4061,
|
| 1276 |
+
"step": 174
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 6.970873786407767,
|
| 1280 |
+
"grad_norm": 2.883514165878296,
|
| 1281 |
+
"learning_rate": 3.48e-05,
|
| 1282 |
+
"loss": 3.4312,
|
| 1283 |
+
"step": 175
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 6.970873786407767,
|
| 1287 |
+
"eval_loss": 3.5276877880096436,
|
| 1288 |
+
"eval_runtime": 0.9695,
|
| 1289 |
+
"eval_samples_per_second": 22.692,
|
| 1290 |
+
"eval_steps_per_second": 6.189,
|
| 1291 |
+
"step": 175
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 7.038834951456311,
|
| 1295 |
+
"grad_norm": 5.757262706756592,
|
| 1296 |
+
"learning_rate": 3.5e-05,
|
| 1297 |
+
"loss": 6.8588,
|
| 1298 |
+
"step": 176
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 7.077669902912621,
|
| 1302 |
+
"grad_norm": 2.7623355388641357,
|
| 1303 |
+
"learning_rate": 3.52e-05,
|
| 1304 |
+
"loss": 3.389,
|
| 1305 |
+
"step": 177
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 7.116504854368932,
|
| 1309 |
+
"grad_norm": 3.601408004760742,
|
| 1310 |
+
"learning_rate": 3.54e-05,
|
| 1311 |
+
"loss": 3.4136,
|
| 1312 |
+
"step": 178
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 7.155339805825243,
|
| 1316 |
+
"grad_norm": 2.4193849563598633,
|
| 1317 |
+
"learning_rate": 3.56e-05,
|
| 1318 |
+
"loss": 3.3387,
|
| 1319 |
+
"step": 179
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 7.194174757281553,
|
| 1323 |
+
"grad_norm": 3.1988773345947266,
|
| 1324 |
+
"learning_rate": 3.58e-05,
|
| 1325 |
+
"loss": 3.3565,
|
| 1326 |
+
"step": 180
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 7.233009708737864,
|
| 1330 |
+
"grad_norm": 3.6124112606048584,
|
| 1331 |
+
"learning_rate": 3.6e-05,
|
| 1332 |
+
"loss": 3.3662,
|
| 1333 |
+
"step": 181
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 7.271844660194175,
|
| 1337 |
+
"grad_norm": 2.836766242980957,
|
| 1338 |
+
"learning_rate": 3.62e-05,
|
| 1339 |
+
"loss": 3.2874,
|
| 1340 |
+
"step": 182
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 7.310679611650485,
|
| 1344 |
+
"grad_norm": 3.2610206604003906,
|
| 1345 |
+
"learning_rate": 3.6400000000000004e-05,
|
| 1346 |
+
"loss": 3.2984,
|
| 1347 |
+
"step": 183
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 7.349514563106796,
|
| 1351 |
+
"grad_norm": 2.8655757904052734,
|
| 1352 |
+
"learning_rate": 3.66e-05,
|
| 1353 |
+
"loss": 3.2584,
|
| 1354 |
+
"step": 184
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 7.388349514563107,
|
| 1358 |
+
"grad_norm": 3.4718809127807617,
|
| 1359 |
+
"learning_rate": 3.68e-05,
|
| 1360 |
+
"loss": 3.2489,
|
| 1361 |
+
"step": 185
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 7.427184466019417,
|
| 1365 |
+
"grad_norm": 3.2131571769714355,
|
| 1366 |
+
"learning_rate": 3.7e-05,
|
| 1367 |
+
"loss": 3.3321,
|
| 1368 |
+
"step": 186
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 7.466019417475728,
|
| 1372 |
+
"grad_norm": 3.1714115142822266,
|
| 1373 |
+
"learning_rate": 3.72e-05,
|
| 1374 |
+
"loss": 3.3489,
|
| 1375 |
+
"step": 187
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 7.504854368932039,
|
| 1379 |
+
"grad_norm": 2.877065658569336,
|
| 1380 |
+
"learning_rate": 3.74e-05,
|
| 1381 |
+
"loss": 3.245,
|
| 1382 |
+
"step": 188
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 7.543689320388349,
|
| 1386 |
+
"grad_norm": 3.1105806827545166,
|
| 1387 |
+
"learning_rate": 3.76e-05,
|
| 1388 |
+
"loss": 3.272,
|
| 1389 |
+
"step": 189
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 7.58252427184466,
|
| 1393 |
+
"grad_norm": 3.5332155227661133,
|
| 1394 |
+
"learning_rate": 3.7800000000000004e-05,
|
| 1395 |
+
"loss": 3.3132,
|
| 1396 |
+
"step": 190
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 7.621359223300971,
|
| 1400 |
+
"grad_norm": 2.8226609230041504,
|
| 1401 |
+
"learning_rate": 3.8e-05,
|
| 1402 |
+
"loss": 3.2721,
|
| 1403 |
+
"step": 191
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 7.660194174757281,
|
| 1407 |
+
"grad_norm": 2.5367422103881836,
|
| 1408 |
+
"learning_rate": 3.82e-05,
|
| 1409 |
+
"loss": 3.3234,
|
| 1410 |
+
"step": 192
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 7.699029126213592,
|
| 1414 |
+
"grad_norm": 2.9826626777648926,
|
| 1415 |
+
"learning_rate": 3.8400000000000005e-05,
|
| 1416 |
+
"loss": 3.2643,
|
| 1417 |
+
"step": 193
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 7.737864077669903,
|
| 1421 |
+
"grad_norm": 3.456496477127075,
|
| 1422 |
+
"learning_rate": 3.86e-05,
|
| 1423 |
+
"loss": 3.2105,
|
| 1424 |
+
"step": 194
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 7.776699029126213,
|
| 1428 |
+
"grad_norm": 3.286680221557617,
|
| 1429 |
+
"learning_rate": 3.88e-05,
|
| 1430 |
+
"loss": 3.2156,
|
| 1431 |
+
"step": 195
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 7.815533980582524,
|
| 1435 |
+
"grad_norm": 2.996983528137207,
|
| 1436 |
+
"learning_rate": 3.9000000000000006e-05,
|
| 1437 |
+
"loss": 3.3637,
|
| 1438 |
+
"step": 196
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 7.854368932038835,
|
| 1442 |
+
"grad_norm": 3.129873037338257,
|
| 1443 |
+
"learning_rate": 3.9200000000000004e-05,
|
| 1444 |
+
"loss": 3.2444,
|
| 1445 |
+
"step": 197
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 7.893203883495145,
|
| 1449 |
+
"grad_norm": 2.591716766357422,
|
| 1450 |
+
"learning_rate": 3.94e-05,
|
| 1451 |
+
"loss": 3.2831,
|
| 1452 |
+
"step": 198
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 7.932038834951456,
|
| 1456 |
+
"grad_norm": 2.664017677307129,
|
| 1457 |
+
"learning_rate": 3.960000000000001e-05,
|
| 1458 |
+
"loss": 3.1692,
|
| 1459 |
+
"step": 199
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 7.970873786407767,
|
| 1463 |
+
"grad_norm": 2.8941309452056885,
|
| 1464 |
+
"learning_rate": 3.9800000000000005e-05,
|
| 1465 |
+
"loss": 3.2986,
|
| 1466 |
+
"step": 200
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 7.970873786407767,
|
| 1470 |
+
"eval_loss": 3.3523428440093994,
|
| 1471 |
+
"eval_runtime": 0.9896,
|
| 1472 |
+
"eval_samples_per_second": 22.23,
|
| 1473 |
+
"eval_steps_per_second": 6.063,
|
| 1474 |
+
"step": 200
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 8.03883495145631,
|
| 1478 |
+
"grad_norm": 6.677456378936768,
|
| 1479 |
+
"learning_rate": 4e-05,
|
| 1480 |
+
"loss": 6.5052,
|
| 1481 |
+
"step": 201
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 8.077669902912621,
|
| 1485 |
+
"grad_norm": 3.38222599029541,
|
| 1486 |
+
"learning_rate": 4.02e-05,
|
| 1487 |
+
"loss": 3.1682,
|
| 1488 |
+
"step": 202
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 8.116504854368932,
|
| 1492 |
+
"grad_norm": 2.9144835472106934,
|
| 1493 |
+
"learning_rate": 4.0400000000000006e-05,
|
| 1494 |
+
"loss": 3.1505,
|
| 1495 |
+
"step": 203
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 8.155339805825243,
|
| 1499 |
+
"grad_norm": 2.837830066680908,
|
| 1500 |
+
"learning_rate": 4.0600000000000004e-05,
|
| 1501 |
+
"loss": 3.1576,
|
| 1502 |
+
"step": 204
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 8.194174757281553,
|
| 1506 |
+
"grad_norm": 3.3070290088653564,
|
| 1507 |
+
"learning_rate": 4.08e-05,
|
| 1508 |
+
"loss": 3.1545,
|
| 1509 |
+
"step": 205
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 8.233009708737864,
|
| 1513 |
+
"grad_norm": 2.6031386852264404,
|
| 1514 |
+
"learning_rate": 4.1e-05,
|
| 1515 |
+
"loss": 3.1598,
|
| 1516 |
+
"step": 206
|
| 1517 |
+
},
|
| 1518 |
+
{
|
| 1519 |
+
"epoch": 8.271844660194175,
|
| 1520 |
+
"grad_norm": 2.879425525665283,
|
| 1521 |
+
"learning_rate": 4.12e-05,
|
| 1522 |
+
"loss": 3.1222,
|
| 1523 |
+
"step": 207
|
| 1524 |
+
},
|
| 1525 |
+
{
|
| 1526 |
+
"epoch": 8.310679611650485,
|
| 1527 |
+
"grad_norm": 3.14932918548584,
|
| 1528 |
+
"learning_rate": 4.14e-05,
|
| 1529 |
+
"loss": 3.0743,
|
| 1530 |
+
"step": 208
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"epoch": 8.349514563106796,
|
| 1534 |
+
"grad_norm": 3.3993191719055176,
|
| 1535 |
+
"learning_rate": 4.16e-05,
|
| 1536 |
+
"loss": 3.1589,
|
| 1537 |
+
"step": 209
|
| 1538 |
+
},
|
| 1539 |
+
{
|
| 1540 |
+
"epoch": 8.388349514563107,
|
| 1541 |
+
"grad_norm": 3.2141942977905273,
|
| 1542 |
+
"learning_rate": 4.18e-05,
|
| 1543 |
+
"loss": 3.0428,
|
| 1544 |
+
"step": 210
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 8.427184466019417,
|
| 1548 |
+
"grad_norm": 2.791717290878296,
|
| 1549 |
+
"learning_rate": 4.2e-05,
|
| 1550 |
+
"loss": 3.1158,
|
| 1551 |
+
"step": 211
|
| 1552 |
+
},
|
| 1553 |
+
{
|
| 1554 |
+
"epoch": 8.466019417475728,
|
| 1555 |
+
"grad_norm": 3.1668970584869385,
|
| 1556 |
+
"learning_rate": 4.22e-05,
|
| 1557 |
+
"loss": 3.0463,
|
| 1558 |
+
"step": 212
|
| 1559 |
+
},
|
| 1560 |
+
{
|
| 1561 |
+
"epoch": 8.504854368932039,
|
| 1562 |
+
"grad_norm": 2.4356696605682373,
|
| 1563 |
+
"learning_rate": 4.24e-05,
|
| 1564 |
+
"loss": 3.1434,
|
| 1565 |
+
"step": 213
|
| 1566 |
+
},
|
| 1567 |
+
{
|
| 1568 |
+
"epoch": 8.54368932038835,
|
| 1569 |
+
"grad_norm": 2.9241132736206055,
|
| 1570 |
+
"learning_rate": 4.26e-05,
|
| 1571 |
+
"loss": 3.0292,
|
| 1572 |
+
"step": 214
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"epoch": 8.58252427184466,
|
| 1576 |
+
"grad_norm": 2.4170773029327393,
|
| 1577 |
+
"learning_rate": 4.2800000000000004e-05,
|
| 1578 |
+
"loss": 3.0923,
|
| 1579 |
+
"step": 215
|
| 1580 |
+
},
|
| 1581 |
+
{
|
| 1582 |
+
"epoch": 8.62135922330097,
|
| 1583 |
+
"grad_norm": 2.4428963661193848,
|
| 1584 |
+
"learning_rate": 4.3e-05,
|
| 1585 |
+
"loss": 3.0588,
|
| 1586 |
+
"step": 216
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 8.660194174757281,
|
| 1590 |
+
"grad_norm": 3.0066943168640137,
|
| 1591 |
+
"learning_rate": 4.32e-05,
|
| 1592 |
+
"loss": 3.0815,
|
| 1593 |
+
"step": 217
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 8.699029126213592,
|
| 1597 |
+
"grad_norm": 3.0532405376434326,
|
| 1598 |
+
"learning_rate": 4.3400000000000005e-05,
|
| 1599 |
+
"loss": 3.1377,
|
| 1600 |
+
"step": 218
|
| 1601 |
+
},
|
| 1602 |
+
{
|
| 1603 |
+
"epoch": 8.737864077669903,
|
| 1604 |
+
"grad_norm": 2.9405910968780518,
|
| 1605 |
+
"learning_rate": 4.36e-05,
|
| 1606 |
+
"loss": 3.081,
|
| 1607 |
+
"step": 219
|
| 1608 |
+
},
|
| 1609 |
+
{
|
| 1610 |
+
"epoch": 8.776699029126213,
|
| 1611 |
+
"grad_norm": 2.82438325881958,
|
| 1612 |
+
"learning_rate": 4.38e-05,
|
| 1613 |
+
"loss": 3.0538,
|
| 1614 |
+
"step": 220
|
| 1615 |
+
},
|
| 1616 |
+
{
|
| 1617 |
+
"epoch": 8.815533980582524,
|
| 1618 |
+
"grad_norm": 2.899946928024292,
|
| 1619 |
+
"learning_rate": 4.4000000000000006e-05,
|
| 1620 |
+
"loss": 3.0664,
|
| 1621 |
+
"step": 221
|
| 1622 |
+
},
|
| 1623 |
+
{
|
| 1624 |
+
"epoch": 8.854368932038835,
|
| 1625 |
+
"grad_norm": 2.4132299423217773,
|
| 1626 |
+
"learning_rate": 4.4200000000000004e-05,
|
| 1627 |
+
"loss": 3.0723,
|
| 1628 |
+
"step": 222
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 8.893203883495145,
|
| 1632 |
+
"grad_norm": 3.2833642959594727,
|
| 1633 |
+
"learning_rate": 4.44e-05,
|
| 1634 |
+
"loss": 3.0445,
|
| 1635 |
+
"step": 223
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 8.932038834951456,
|
| 1639 |
+
"grad_norm": 2.60457706451416,
|
| 1640 |
+
"learning_rate": 4.46e-05,
|
| 1641 |
+
"loss": 3.0134,
|
| 1642 |
+
"step": 224
|
| 1643 |
+
},
|
| 1644 |
+
{
|
| 1645 |
+
"epoch": 8.970873786407767,
|
| 1646 |
+
"grad_norm": 2.7552649974823,
|
| 1647 |
+
"learning_rate": 4.4800000000000005e-05,
|
| 1648 |
+
"loss": 3.0772,
|
| 1649 |
+
"step": 225
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 8.970873786407767,
|
| 1653 |
+
"eval_loss": 3.2081830501556396,
|
| 1654 |
+
"eval_runtime": 0.971,
|
| 1655 |
+
"eval_samples_per_second": 22.657,
|
| 1656 |
+
"eval_steps_per_second": 6.179,
|
| 1657 |
+
"step": 225
|
| 1658 |
+
},
|
| 1659 |
+
{
|
| 1660 |
+
"epoch": 9.03883495145631,
|
| 1661 |
+
"grad_norm": 6.834669589996338,
|
| 1662 |
+
"learning_rate": 4.5e-05,
|
| 1663 |
+
"loss": 6.0,
|
| 1664 |
+
"step": 226
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 9.077669902912621,
|
| 1668 |
+
"grad_norm": 2.7431795597076416,
|
| 1669 |
+
"learning_rate": 4.52e-05,
|
| 1670 |
+
"loss": 2.8938,
|
| 1671 |
+
"step": 227
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 9.116504854368932,
|
| 1675 |
+
"grad_norm": 3.5260982513427734,
|
| 1676 |
+
"learning_rate": 4.5400000000000006e-05,
|
| 1677 |
+
"loss": 2.955,
|
| 1678 |
+
"step": 228
|
| 1679 |
+
},
|
| 1680 |
+
{
|
| 1681 |
+
"epoch": 9.155339805825243,
|
| 1682 |
+
"grad_norm": 2.705111026763916,
|
| 1683 |
+
"learning_rate": 4.5600000000000004e-05,
|
| 1684 |
+
"loss": 2.9586,
|
| 1685 |
+
"step": 229
|
| 1686 |
+
},
|
| 1687 |
+
{
|
| 1688 |
+
"epoch": 9.194174757281553,
|
| 1689 |
+
"grad_norm": 3.2462103366851807,
|
| 1690 |
+
"learning_rate": 4.58e-05,
|
| 1691 |
+
"loss": 2.8973,
|
| 1692 |
+
"step": 230
|
| 1693 |
+
},
|
| 1694 |
+
{
|
| 1695 |
+
"epoch": 9.233009708737864,
|
| 1696 |
+
"grad_norm": 2.788363218307495,
|
| 1697 |
+
"learning_rate": 4.600000000000001e-05,
|
| 1698 |
+
"loss": 2.9328,
|
| 1699 |
+
"step": 231
|
| 1700 |
+
},
|
| 1701 |
+
{
|
| 1702 |
+
"epoch": 9.271844660194175,
|
| 1703 |
+
"grad_norm": 2.4145243167877197,
|
| 1704 |
+
"learning_rate": 4.6200000000000005e-05,
|
| 1705 |
+
"loss": 2.9653,
|
| 1706 |
+
"step": 232
|
| 1707 |
+
},
|
| 1708 |
+
{
|
| 1709 |
+
"epoch": 9.310679611650485,
|
| 1710 |
+
"grad_norm": 3.2271153926849365,
|
| 1711 |
+
"learning_rate": 4.64e-05,
|
| 1712 |
+
"loss": 2.922,
|
| 1713 |
+
"step": 233
|
| 1714 |
+
},
|
| 1715 |
+
{
|
| 1716 |
+
"epoch": 9.349514563106796,
|
| 1717 |
+
"grad_norm": 2.3625218868255615,
|
| 1718 |
+
"learning_rate": 4.660000000000001e-05,
|
| 1719 |
+
"loss": 3.0413,
|
| 1720 |
+
"step": 234
|
| 1721 |
+
},
|
| 1722 |
+
{
|
| 1723 |
+
"epoch": 9.388349514563107,
|
| 1724 |
+
"grad_norm": 3.17262864112854,
|
| 1725 |
+
"learning_rate": 4.6800000000000006e-05,
|
| 1726 |
+
"loss": 2.9962,
|
| 1727 |
+
"step": 235
|
| 1728 |
+
},
|
| 1729 |
+
{
|
| 1730 |
+
"epoch": 9.427184466019417,
|
| 1731 |
+
"grad_norm": 2.906003475189209,
|
| 1732 |
+
"learning_rate": 4.7e-05,
|
| 1733 |
+
"loss": 2.9422,
|
| 1734 |
+
"step": 236
|
| 1735 |
+
},
|
| 1736 |
+
{
|
| 1737 |
+
"epoch": 9.466019417475728,
|
| 1738 |
+
"grad_norm": 2.1498398780822754,
|
| 1739 |
+
"learning_rate": 4.72e-05,
|
| 1740 |
+
"loss": 2.9061,
|
| 1741 |
+
"step": 237
|
| 1742 |
+
},
|
| 1743 |
+
{
|
| 1744 |
+
"epoch": 9.504854368932039,
|
| 1745 |
+
"grad_norm": 2.9519286155700684,
|
| 1746 |
+
"learning_rate": 4.74e-05,
|
| 1747 |
+
"loss": 2.967,
|
| 1748 |
+
"step": 238
|
| 1749 |
+
},
|
| 1750 |
+
{
|
| 1751 |
+
"epoch": 9.54368932038835,
|
| 1752 |
+
"grad_norm": 2.561063528060913,
|
| 1753 |
+
"learning_rate": 4.76e-05,
|
| 1754 |
+
"loss": 2.9191,
|
| 1755 |
+
"step": 239
|
| 1756 |
+
},
|
| 1757 |
+
{
|
| 1758 |
+
"epoch": 9.58252427184466,
|
| 1759 |
+
"grad_norm": 3.8291261196136475,
|
| 1760 |
+
"learning_rate": 4.78e-05,
|
| 1761 |
+
"loss": 2.9071,
|
| 1762 |
+
"step": 240
|
| 1763 |
+
},
|
| 1764 |
+
{
|
| 1765 |
+
"epoch": 9.62135922330097,
|
| 1766 |
+
"grad_norm": 3.4280309677124023,
|
| 1767 |
+
"learning_rate": 4.8e-05,
|
| 1768 |
+
"loss": 2.9384,
|
| 1769 |
+
"step": 241
|
| 1770 |
+
},
|
| 1771 |
+
{
|
| 1772 |
+
"epoch": 9.660194174757281,
|
| 1773 |
+
"grad_norm": 3.460054397583008,
|
| 1774 |
+
"learning_rate": 4.82e-05,
|
| 1775 |
+
"loss": 2.9387,
|
| 1776 |
+
"step": 242
|
| 1777 |
+
},
|
| 1778 |
+
{
|
| 1779 |
+
"epoch": 9.699029126213592,
|
| 1780 |
+
"grad_norm": 3.3750805854797363,
|
| 1781 |
+
"learning_rate": 4.8400000000000004e-05,
|
| 1782 |
+
"loss": 2.9552,
|
| 1783 |
+
"step": 243
|
| 1784 |
+
},
|
| 1785 |
+
{
|
| 1786 |
+
"epoch": 9.737864077669903,
|
| 1787 |
+
"grad_norm": 2.6689562797546387,
|
| 1788 |
+
"learning_rate": 4.86e-05,
|
| 1789 |
+
"loss": 2.8809,
|
| 1790 |
+
"step": 244
|
| 1791 |
+
},
|
| 1792 |
+
{
|
| 1793 |
+
"epoch": 9.776699029126213,
|
| 1794 |
+
"grad_norm": 2.9314560890197754,
|
| 1795 |
+
"learning_rate": 4.88e-05,
|
| 1796 |
+
"loss": 2.7902,
|
| 1797 |
+
"step": 245
|
| 1798 |
+
},
|
| 1799 |
+
{
|
| 1800 |
+
"epoch": 9.815533980582524,
|
| 1801 |
+
"grad_norm": 2.630530595779419,
|
| 1802 |
+
"learning_rate": 4.9e-05,
|
| 1803 |
+
"loss": 2.8857,
|
| 1804 |
+
"step": 246
|
| 1805 |
+
},
|
| 1806 |
+
{
|
| 1807 |
+
"epoch": 9.854368932038835,
|
| 1808 |
+
"grad_norm": 2.546659231185913,
|
| 1809 |
+
"learning_rate": 4.92e-05,
|
| 1810 |
+
"loss": 2.8896,
|
| 1811 |
+
"step": 247
|
| 1812 |
+
},
|
| 1813 |
+
{
|
| 1814 |
+
"epoch": 9.893203883495145,
|
| 1815 |
+
"grad_norm": 2.795778751373291,
|
| 1816 |
+
"learning_rate": 4.94e-05,
|
| 1817 |
+
"loss": 2.9516,
|
| 1818 |
+
"step": 248
|
| 1819 |
+
},
|
| 1820 |
+
{
|
| 1821 |
+
"epoch": 9.932038834951456,
|
| 1822 |
+
"grad_norm": 3.0504794120788574,
|
| 1823 |
+
"learning_rate": 4.96e-05,
|
| 1824 |
+
"loss": 2.9132,
|
| 1825 |
+
"step": 249
|
| 1826 |
+
},
|
| 1827 |
+
{
|
| 1828 |
+
"epoch": 9.970873786407767,
|
| 1829 |
+
"grad_norm": 3.444287061691284,
|
| 1830 |
+
"learning_rate": 4.9800000000000004e-05,
|
| 1831 |
+
"loss": 2.973,
|
| 1832 |
+
"step": 250
|
| 1833 |
+
},
|
| 1834 |
+
{
|
| 1835 |
+
"epoch": 9.970873786407767,
|
| 1836 |
+
"eval_loss": 3.089428186416626,
|
| 1837 |
+
"eval_runtime": 0.9754,
|
| 1838 |
+
"eval_samples_per_second": 22.555,
|
| 1839 |
+
"eval_steps_per_second": 6.151,
|
| 1840 |
+
"step": 250
|
| 1841 |
+
},
|
| 1842 |
+
{
|
| 1843 |
+
"epoch": 10.03883495145631,
|
| 1844 |
+
"grad_norm": 5.480017185211182,
|
| 1845 |
+
"learning_rate": 5e-05,
|
| 1846 |
+
"loss": 5.6196,
|
| 1847 |
+
"step": 251
|
| 1848 |
+
},
|
| 1849 |
+
{
|
| 1850 |
+
"epoch": 10.077669902912621,
|
| 1851 |
+
"grad_norm": 3.3957669734954834,
|
| 1852 |
+
"learning_rate": 4.999997563061038e-05,
|
| 1853 |
+
"loss": 2.8152,
|
| 1854 |
+
"step": 252
|
| 1855 |
+
},
|
| 1856 |
+
{
|
| 1857 |
+
"epoch": 10.116504854368932,
|
| 1858 |
+
"grad_norm": 2.6747496128082275,
|
| 1859 |
+
"learning_rate": 4.9999902522489015e-05,
|
| 1860 |
+
"loss": 2.8624,
|
| 1861 |
+
"step": 253
|
| 1862 |
+
},
|
| 1863 |
+
{
|
| 1864 |
+
"epoch": 10.155339805825243,
|
| 1865 |
+
"grad_norm": 3.2186131477355957,
|
| 1866 |
+
"learning_rate": 4.999978067577844e-05,
|
| 1867 |
+
"loss": 2.7587,
|
| 1868 |
+
"step": 254
|
| 1869 |
+
},
|
| 1870 |
+
{
|
| 1871 |
+
"epoch": 10.194174757281553,
|
| 1872 |
+
"grad_norm": 3.7385358810424805,
|
| 1873 |
+
"learning_rate": 4.999961009071621e-05,
|
| 1874 |
+
"loss": 2.8117,
|
| 1875 |
+
"step": 255
|
| 1876 |
+
},
|
| 1877 |
+
{
|
| 1878 |
+
"epoch": 10.233009708737864,
|
| 1879 |
+
"grad_norm": 2.586005926132202,
|
| 1880 |
+
"learning_rate": 4.999939076763487e-05,
|
| 1881 |
+
"loss": 2.7617,
|
| 1882 |
+
"step": 256
|
| 1883 |
+
},
|
| 1884 |
+
{
|
| 1885 |
+
"epoch": 10.271844660194175,
|
| 1886 |
+
"grad_norm": 2.7468533515930176,
|
| 1887 |
+
"learning_rate": 4.999912270696202e-05,
|
| 1888 |
+
"loss": 2.802,
|
| 1889 |
+
"step": 257
|
| 1890 |
+
},
|
| 1891 |
+
{
|
| 1892 |
+
"epoch": 10.310679611650485,
|
| 1893 |
+
"grad_norm": 2.7268691062927246,
|
| 1894 |
+
"learning_rate": 4.999880590922025e-05,
|
| 1895 |
+
"loss": 2.7928,
|
| 1896 |
+
"step": 258
|
| 1897 |
+
},
|
| 1898 |
+
{
|
| 1899 |
+
"epoch": 10.349514563106796,
|
| 1900 |
+
"grad_norm": 2.6305949687957764,
|
| 1901 |
+
"learning_rate": 4.9998440375027166e-05,
|
| 1902 |
+
"loss": 2.8245,
|
| 1903 |
+
"step": 259
|
| 1904 |
+
},
|
| 1905 |
+
{
|
| 1906 |
+
"epoch": 10.388349514563107,
|
| 1907 |
+
"grad_norm": 2.8977084159851074,
|
| 1908 |
+
"learning_rate": 4.9998026105095405e-05,
|
| 1909 |
+
"loss": 2.7525,
|
| 1910 |
+
"step": 260
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"epoch": 10.427184466019417,
|
| 1914 |
+
"grad_norm": 2.394578218460083,
|
| 1915 |
+
"learning_rate": 4.999756310023261e-05,
|
| 1916 |
+
"loss": 2.731,
|
| 1917 |
+
"step": 261
|
| 1918 |
+
},
|
| 1919 |
+
{
|
| 1920 |
+
"epoch": 10.466019417475728,
|
| 1921 |
+
"grad_norm": 3.0859174728393555,
|
| 1922 |
+
"learning_rate": 4.9997051361341425e-05,
|
| 1923 |
+
"loss": 2.7902,
|
| 1924 |
+
"step": 262
|
| 1925 |
+
},
|
| 1926 |
+
{
|
| 1927 |
+
"epoch": 10.504854368932039,
|
| 1928 |
+
"grad_norm": 2.929978370666504,
|
| 1929 |
+
"learning_rate": 4.9996490889419514e-05,
|
| 1930 |
+
"loss": 2.7723,
|
| 1931 |
+
"step": 263
|
| 1932 |
+
},
|
| 1933 |
+
{
|
| 1934 |
+
"epoch": 10.54368932038835,
|
| 1935 |
+
"grad_norm": 2.6215100288391113,
|
| 1936 |
+
"learning_rate": 4.999588168555954e-05,
|
| 1937 |
+
"loss": 2.7892,
|
| 1938 |
+
"step": 264
|
| 1939 |
+
},
|
| 1940 |
+
{
|
| 1941 |
+
"epoch": 10.58252427184466,
|
| 1942 |
+
"grad_norm": 2.744954824447632,
|
| 1943 |
+
"learning_rate": 4.999522375094919e-05,
|
| 1944 |
+
"loss": 2.8024,
|
| 1945 |
+
"step": 265
|
| 1946 |
+
},
|
| 1947 |
+
{
|
| 1948 |
+
"epoch": 10.62135922330097,
|
| 1949 |
+
"grad_norm": 2.775912046432495,
|
| 1950 |
+
"learning_rate": 4.999451708687114e-05,
|
| 1951 |
+
"loss": 2.642,
|
| 1952 |
+
"step": 266
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 10.660194174757281,
|
| 1956 |
+
"grad_norm": 2.5821340084075928,
|
| 1957 |
+
"learning_rate": 4.999376169470306e-05,
|
| 1958 |
+
"loss": 2.7808,
|
| 1959 |
+
"step": 267
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"epoch": 10.699029126213592,
|
| 1963 |
+
"grad_norm": 2.4101083278656006,
|
| 1964 |
+
"learning_rate": 4.999295757591762e-05,
|
| 1965 |
+
"loss": 2.7318,
|
| 1966 |
+
"step": 268
|
| 1967 |
+
},
|
| 1968 |
+
{
|
| 1969 |
+
"epoch": 10.737864077669903,
|
| 1970 |
+
"grad_norm": 2.4816181659698486,
|
| 1971 |
+
"learning_rate": 4.99921047320825e-05,
|
| 1972 |
+
"loss": 2.7707,
|
| 1973 |
+
"step": 269
|
| 1974 |
+
},
|
| 1975 |
+
{
|
| 1976 |
+
"epoch": 10.776699029126213,
|
| 1977 |
+
"grad_norm": 2.366009473800659,
|
| 1978 |
+
"learning_rate": 4.9991203164860365e-05,
|
| 1979 |
+
"loss": 2.7481,
|
| 1980 |
+
"step": 270
|
| 1981 |
+
},
|
| 1982 |
+
{
|
| 1983 |
+
"epoch": 10.815533980582524,
|
| 1984 |
+
"grad_norm": 2.9792630672454834,
|
| 1985 |
+
"learning_rate": 4.999025287600886e-05,
|
| 1986 |
+
"loss": 2.7204,
|
| 1987 |
+
"step": 271
|
| 1988 |
+
},
|
| 1989 |
+
{
|
| 1990 |
+
"epoch": 10.854368932038835,
|
| 1991 |
+
"grad_norm": 3.0781967639923096,
|
| 1992 |
+
"learning_rate": 4.998925386738063e-05,
|
| 1993 |
+
"loss": 2.7248,
|
| 1994 |
+
"step": 272
|
| 1995 |
+
},
|
| 1996 |
+
{
|
| 1997 |
+
"epoch": 10.893203883495145,
|
| 1998 |
+
"grad_norm": 2.6866307258605957,
|
| 1999 |
+
"learning_rate": 4.998820614092328e-05,
|
| 2000 |
+
"loss": 2.7456,
|
| 2001 |
+
"step": 273
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"epoch": 10.932038834951456,
|
| 2005 |
+
"grad_norm": 2.789808988571167,
|
| 2006 |
+
"learning_rate": 4.998710969867942e-05,
|
| 2007 |
+
"loss": 2.7224,
|
| 2008 |
+
"step": 274
|
| 2009 |
+
},
|
| 2010 |
+
{
|
| 2011 |
+
"epoch": 10.970873786407767,
|
| 2012 |
+
"grad_norm": 2.4948067665100098,
|
| 2013 |
+
"learning_rate": 4.9985964542786614e-05,
|
| 2014 |
+
"loss": 2.6724,
|
| 2015 |
+
"step": 275
|
| 2016 |
+
},
|
| 2017 |
+
{
|
| 2018 |
+
"epoch": 10.970873786407767,
|
| 2019 |
+
"eval_loss": 2.9974570274353027,
|
| 2020 |
+
"eval_runtime": 0.9771,
|
| 2021 |
+
"eval_samples_per_second": 22.516,
|
| 2022 |
+
"eval_steps_per_second": 6.141,
|
| 2023 |
+
"step": 275
|
| 2024 |
+
},
|
| 2025 |
+
{
|
| 2026 |
+
"epoch": 11.03883495145631,
|
| 2027 |
+
"grad_norm": 4.32741117477417,
|
| 2028 |
+
"learning_rate": 4.99847706754774e-05,
|
| 2029 |
+
"loss": 5.4423,
|
| 2030 |
+
"step": 276
|
| 2031 |
+
},
|
| 2032 |
+
{
|
| 2033 |
+
"epoch": 11.077669902912621,
|
| 2034 |
+
"grad_norm": 2.3572208881378174,
|
| 2035 |
+
"learning_rate": 4.998352809907928e-05,
|
| 2036 |
+
"loss": 2.6372,
|
| 2037 |
+
"step": 277
|
| 2038 |
+
},
|
| 2039 |
+
{
|
| 2040 |
+
"epoch": 11.116504854368932,
|
| 2041 |
+
"grad_norm": 2.4045934677124023,
|
| 2042 |
+
"learning_rate": 4.998223681601473e-05,
|
| 2043 |
+
"loss": 2.6205,
|
| 2044 |
+
"step": 278
|
| 2045 |
+
},
|
| 2046 |
+
{
|
| 2047 |
+
"epoch": 11.155339805825243,
|
| 2048 |
+
"grad_norm": 2.5755131244659424,
|
| 2049 |
+
"learning_rate": 4.998089682880117e-05,
|
| 2050 |
+
"loss": 2.5939,
|
| 2051 |
+
"step": 279
|
| 2052 |
+
},
|
| 2053 |
+
{
|
| 2054 |
+
"epoch": 11.194174757281553,
|
| 2055 |
+
"grad_norm": 2.5768463611602783,
|
| 2056 |
+
"learning_rate": 4.997950814005098e-05,
|
| 2057 |
+
"loss": 2.6925,
|
| 2058 |
+
"step": 280
|
| 2059 |
+
},
|
| 2060 |
+
{
|
| 2061 |
+
"epoch": 11.233009708737864,
|
| 2062 |
+
"grad_norm": 2.5549166202545166,
|
| 2063 |
+
"learning_rate": 4.997807075247146e-05,
|
| 2064 |
+
"loss": 2.6172,
|
| 2065 |
+
"step": 281
|
| 2066 |
+
},
|
| 2067 |
+
{
|
| 2068 |
+
"epoch": 11.271844660194175,
|
| 2069 |
+
"grad_norm": 2.761068344116211,
|
| 2070 |
+
"learning_rate": 4.997658466886489e-05,
|
| 2071 |
+
"loss": 2.6572,
|
| 2072 |
+
"step": 282
|
| 2073 |
+
},
|
| 2074 |
+
{
|
| 2075 |
+
"epoch": 11.310679611650485,
|
| 2076 |
+
"grad_norm": 2.5051231384277344,
|
| 2077 |
+
"learning_rate": 4.9975049892128455e-05,
|
| 2078 |
+
"loss": 2.6549,
|
| 2079 |
+
"step": 283
|
| 2080 |
+
},
|
| 2081 |
+
{
|
| 2082 |
+
"epoch": 11.349514563106796,
|
| 2083 |
+
"grad_norm": 2.7434117794036865,
|
| 2084 |
+
"learning_rate": 4.9973466425254286e-05,
|
| 2085 |
+
"loss": 2.5632,
|
| 2086 |
+
"step": 284
|
| 2087 |
+
},
|
| 2088 |
+
{
|
| 2089 |
+
"epoch": 11.388349514563107,
|
| 2090 |
+
"grad_norm": 2.328563928604126,
|
| 2091 |
+
"learning_rate": 4.997183427132943e-05,
|
| 2092 |
+
"loss": 2.5751,
|
| 2093 |
+
"step": 285
|
| 2094 |
+
},
|
| 2095 |
+
{
|
| 2096 |
+
"epoch": 11.427184466019417,
|
| 2097 |
+
"grad_norm": 2.7668466567993164,
|
| 2098 |
+
"learning_rate": 4.997015343353585e-05,
|
| 2099 |
+
"loss": 2.6609,
|
| 2100 |
+
"step": 286
|
| 2101 |
+
},
|
| 2102 |
+
{
|
| 2103 |
+
"epoch": 11.466019417475728,
|
| 2104 |
+
"grad_norm": 2.0831525325775146,
|
| 2105 |
+
"learning_rate": 4.996842391515044e-05,
|
| 2106 |
+
"loss": 2.6428,
|
| 2107 |
+
"step": 287
|
| 2108 |
+
},
|
| 2109 |
+
{
|
| 2110 |
+
"epoch": 11.504854368932039,
|
| 2111 |
+
"grad_norm": 2.4443278312683105,
|
| 2112 |
+
"learning_rate": 4.996664571954497e-05,
|
| 2113 |
+
"loss": 2.6012,
|
| 2114 |
+
"step": 288
|
| 2115 |
+
},
|
| 2116 |
+
{
|
| 2117 |
+
"epoch": 11.54368932038835,
|
| 2118 |
+
"grad_norm": 2.4806153774261475,
|
| 2119 |
+
"learning_rate": 4.9964818850186135e-05,
|
| 2120 |
+
"loss": 2.6649,
|
| 2121 |
+
"step": 289
|
| 2122 |
+
},
|
| 2123 |
+
{
|
| 2124 |
+
"epoch": 11.58252427184466,
|
| 2125 |
+
"grad_norm": 2.539933919906616,
|
| 2126 |
+
"learning_rate": 4.99629433106355e-05,
|
| 2127 |
+
"loss": 2.6253,
|
| 2128 |
+
"step": 290
|
| 2129 |
+
},
|
| 2130 |
+
{
|
| 2131 |
+
"epoch": 11.62135922330097,
|
| 2132 |
+
"grad_norm": 2.7404544353485107,
|
| 2133 |
+
"learning_rate": 4.996101910454953e-05,
|
| 2134 |
+
"loss": 2.6224,
|
| 2135 |
+
"step": 291
|
| 2136 |
+
},
|
| 2137 |
+
{
|
| 2138 |
+
"epoch": 11.660194174757281,
|
| 2139 |
+
"grad_norm": 2.5377357006073,
|
| 2140 |
+
"learning_rate": 4.9959046235679565e-05,
|
| 2141 |
+
"loss": 2.6249,
|
| 2142 |
+
"step": 292
|
| 2143 |
+
},
|
| 2144 |
+
{
|
| 2145 |
+
"epoch": 11.699029126213592,
|
| 2146 |
+
"grad_norm": 2.8488271236419678,
|
| 2147 |
+
"learning_rate": 4.9957024707871806e-05,
|
| 2148 |
+
"loss": 2.6232,
|
| 2149 |
+
"step": 293
|
| 2150 |
+
},
|
| 2151 |
+
{
|
| 2152 |
+
"epoch": 11.737864077669903,
|
| 2153 |
+
"grad_norm": 2.4895827770233154,
|
| 2154 |
+
"learning_rate": 4.9954954525067334e-05,
|
| 2155 |
+
"loss": 2.5983,
|
| 2156 |
+
"step": 294
|
| 2157 |
+
},
|
| 2158 |
+
{
|
| 2159 |
+
"epoch": 11.776699029126213,
|
| 2160 |
+
"grad_norm": 3.038975954055786,
|
| 2161 |
+
"learning_rate": 4.995283569130207e-05,
|
| 2162 |
+
"loss": 2.5715,
|
| 2163 |
+
"step": 295
|
| 2164 |
+
},
|
| 2165 |
+
{
|
| 2166 |
+
"epoch": 11.815533980582524,
|
| 2167 |
+
"grad_norm": 2.674245595932007,
|
| 2168 |
+
"learning_rate": 4.995066821070679e-05,
|
| 2169 |
+
"loss": 2.6201,
|
| 2170 |
+
"step": 296
|
| 2171 |
+
},
|
| 2172 |
+
{
|
| 2173 |
+
"epoch": 11.854368932038835,
|
| 2174 |
+
"grad_norm": 3.5277645587921143,
|
| 2175 |
+
"learning_rate": 4.9948452087507116e-05,
|
| 2176 |
+
"loss": 2.6376,
|
| 2177 |
+
"step": 297
|
| 2178 |
+
},
|
| 2179 |
+
{
|
| 2180 |
+
"epoch": 11.893203883495145,
|
| 2181 |
+
"grad_norm": 3.0974984169006348,
|
| 2182 |
+
"learning_rate": 4.994618732602349e-05,
|
| 2183 |
+
"loss": 2.6268,
|
| 2184 |
+
"step": 298
|
| 2185 |
+
},
|
| 2186 |
+
{
|
| 2187 |
+
"epoch": 11.932038834951456,
|
| 2188 |
+
"grad_norm": 2.309119462966919,
|
| 2189 |
+
"learning_rate": 4.994387393067117e-05,
|
| 2190 |
+
"loss": 2.5594,
|
| 2191 |
+
"step": 299
|
| 2192 |
+
},
|
| 2193 |
+
{
|
| 2194 |
+
"epoch": 11.970873786407767,
|
| 2195 |
+
"grad_norm": 2.540464162826538,
|
| 2196 |
+
"learning_rate": 4.994151190596025e-05,
|
| 2197 |
+
"loss": 2.5765,
|
| 2198 |
+
"step": 300
|
| 2199 |
+
},
|
| 2200 |
+
{
|
| 2201 |
+
"epoch": 11.970873786407767,
|
| 2202 |
+
"eval_loss": 2.9208481311798096,
|
| 2203 |
+
"eval_runtime": 1.0115,
|
| 2204 |
+
"eval_samples_per_second": 21.749,
|
| 2205 |
+
"eval_steps_per_second": 5.932,
|
| 2206 |
+
"step": 300
|
| 2207 |
+
},
|
| 2208 |
+
{
|
| 2209 |
+
"epoch": 12.03883495145631,
|
| 2210 |
+
"grad_norm": 5.542501449584961,
|
| 2211 |
+
"learning_rate": 4.993910125649561e-05,
|
| 2212 |
+
"loss": 5.1943,
|
| 2213 |
+
"step": 301
|
| 2214 |
+
},
|
| 2215 |
+
{
|
| 2216 |
+
"epoch": 12.077669902912621,
|
| 2217 |
+
"grad_norm": 2.2998414039611816,
|
| 2218 |
+
"learning_rate": 4.993664198697694e-05,
|
| 2219 |
+
"loss": 2.5311,
|
| 2220 |
+
"step": 302
|
| 2221 |
+
},
|
| 2222 |
+
{
|
| 2223 |
+
"epoch": 12.116504854368932,
|
| 2224 |
+
"grad_norm": 3.0827107429504395,
|
| 2225 |
+
"learning_rate": 4.993413410219871e-05,
|
| 2226 |
+
"loss": 2.5587,
|
| 2227 |
+
"step": 303
|
| 2228 |
+
},
|
| 2229 |
+
{
|
| 2230 |
+
"epoch": 12.155339805825243,
|
| 2231 |
+
"grad_norm": 2.7742204666137695,
|
| 2232 |
+
"learning_rate": 4.9931577607050175e-05,
|
| 2233 |
+
"loss": 2.4549,
|
| 2234 |
+
"step": 304
|
| 2235 |
+
},
|
| 2236 |
+
{
|
| 2237 |
+
"epoch": 12.194174757281553,
|
| 2238 |
+
"grad_norm": 2.5605695247650146,
|
| 2239 |
+
"learning_rate": 4.992897250651535e-05,
|
| 2240 |
+
"loss": 2.5602,
|
| 2241 |
+
"step": 305
|
| 2242 |
+
},
|
| 2243 |
+
{
|
| 2244 |
+
"epoch": 12.233009708737864,
|
| 2245 |
+
"grad_norm": 2.8852667808532715,
|
| 2246 |
+
"learning_rate": 4.992631880567301e-05,
|
| 2247 |
+
"loss": 2.5069,
|
| 2248 |
+
"step": 306
|
| 2249 |
+
},
|
| 2250 |
+
{
|
| 2251 |
+
"epoch": 12.271844660194175,
|
| 2252 |
+
"grad_norm": 3.006777048110962,
|
| 2253 |
+
"learning_rate": 4.9923616509696683e-05,
|
| 2254 |
+
"loss": 2.5326,
|
| 2255 |
+
"step": 307
|
| 2256 |
+
},
|
| 2257 |
+
{
|
| 2258 |
+
"epoch": 12.310679611650485,
|
| 2259 |
+
"grad_norm": 2.1645665168762207,
|
| 2260 |
+
"learning_rate": 4.9920865623854615e-05,
|
| 2261 |
+
"loss": 2.4739,
|
| 2262 |
+
"step": 308
|
| 2263 |
+
},
|
| 2264 |
+
{
|
| 2265 |
+
"epoch": 12.349514563106796,
|
| 2266 |
+
"grad_norm": 2.941042423248291,
|
| 2267 |
+
"learning_rate": 4.9918066153509834e-05,
|
| 2268 |
+
"loss": 2.5149,
|
| 2269 |
+
"step": 309
|
| 2270 |
+
},
|
| 2271 |
+
{
|
| 2272 |
+
"epoch": 12.388349514563107,
|
| 2273 |
+
"grad_norm": 2.598097562789917,
|
| 2274 |
+
"learning_rate": 4.991521810412002e-05,
|
| 2275 |
+
"loss": 2.5214,
|
| 2276 |
+
"step": 310
|
| 2277 |
+
},
|
| 2278 |
+
{
|
| 2279 |
+
"epoch": 12.427184466019417,
|
| 2280 |
+
"grad_norm": 2.408721446990967,
|
| 2281 |
+
"learning_rate": 4.991232148123761e-05,
|
| 2282 |
+
"loss": 2.4747,
|
| 2283 |
+
"step": 311
|
| 2284 |
+
},
|
| 2285 |
+
{
|
| 2286 |
+
"epoch": 12.466019417475728,
|
| 2287 |
+
"grad_norm": 2.39508318901062,
|
| 2288 |
+
"learning_rate": 4.990937629050971e-05,
|
| 2289 |
+
"loss": 2.5304,
|
| 2290 |
+
"step": 312
|
| 2291 |
+
},
|
| 2292 |
+
{
|
| 2293 |
+
"epoch": 12.504854368932039,
|
| 2294 |
+
"grad_norm": 2.9436190128326416,
|
| 2295 |
+
"learning_rate": 4.990638253767812e-05,
|
| 2296 |
+
"loss": 2.5046,
|
| 2297 |
+
"step": 313
|
| 2298 |
+
},
|
| 2299 |
+
{
|
| 2300 |
+
"epoch": 12.54368932038835,
|
| 2301 |
+
"grad_norm": 2.6037611961364746,
|
| 2302 |
+
"learning_rate": 4.990334022857932e-05,
|
| 2303 |
+
"loss": 2.4537,
|
| 2304 |
+
"step": 314
|
| 2305 |
+
},
|
| 2306 |
+
{
|
| 2307 |
+
"epoch": 12.58252427184466,
|
| 2308 |
+
"grad_norm": 2.892789602279663,
|
| 2309 |
+
"learning_rate": 4.9900249369144434e-05,
|
| 2310 |
+
"loss": 2.4817,
|
| 2311 |
+
"step": 315
|
| 2312 |
+
},
|
| 2313 |
+
{
|
| 2314 |
+
"epoch": 12.62135922330097,
|
| 2315 |
+
"grad_norm": 2.6804611682891846,
|
| 2316 |
+
"learning_rate": 4.989710996539926e-05,
|
| 2317 |
+
"loss": 2.5012,
|
| 2318 |
+
"step": 316
|
| 2319 |
+
},
|
| 2320 |
+
{
|
| 2321 |
+
"epoch": 12.660194174757281,
|
| 2322 |
+
"grad_norm": 2.458824396133423,
|
| 2323 |
+
"learning_rate": 4.9893922023464236e-05,
|
| 2324 |
+
"loss": 2.4661,
|
| 2325 |
+
"step": 317
|
| 2326 |
+
},
|
| 2327 |
+
{
|
| 2328 |
+
"epoch": 12.699029126213592,
|
| 2329 |
+
"grad_norm": 2.6641952991485596,
|
| 2330 |
+
"learning_rate": 4.989068554955439e-05,
|
| 2331 |
+
"loss": 2.4971,
|
| 2332 |
+
"step": 318
|
| 2333 |
+
},
|
| 2334 |
+
{
|
| 2335 |
+
"epoch": 12.737864077669903,
|
| 2336 |
+
"grad_norm": 2.421142101287842,
|
| 2337 |
+
"learning_rate": 4.988740054997943e-05,
|
| 2338 |
+
"loss": 2.4014,
|
| 2339 |
+
"step": 319
|
| 2340 |
+
},
|
| 2341 |
+
{
|
| 2342 |
+
"epoch": 12.776699029126213,
|
| 2343 |
+
"grad_norm": 2.4107542037963867,
|
| 2344 |
+
"learning_rate": 4.98840670311436e-05,
|
| 2345 |
+
"loss": 2.4636,
|
| 2346 |
+
"step": 320
|
| 2347 |
+
},
|
| 2348 |
+
{
|
| 2349 |
+
"epoch": 12.815533980582524,
|
| 2350 |
+
"grad_norm": 2.5701303482055664,
|
| 2351 |
+
"learning_rate": 4.988068499954578e-05,
|
| 2352 |
+
"loss": 2.4564,
|
| 2353 |
+
"step": 321
|
| 2354 |
+
},
|
| 2355 |
+
{
|
| 2356 |
+
"epoch": 12.854368932038835,
|
| 2357 |
+
"grad_norm": 2.3998067378997803,
|
| 2358 |
+
"learning_rate": 4.987725446177941e-05,
|
| 2359 |
+
"loss": 2.4561,
|
| 2360 |
+
"step": 322
|
| 2361 |
+
},
|
| 2362 |
+
{
|
| 2363 |
+
"epoch": 12.893203883495145,
|
| 2364 |
+
"grad_norm": 2.6888773441314697,
|
| 2365 |
+
"learning_rate": 4.987377542453251e-05,
|
| 2366 |
+
"loss": 2.4392,
|
| 2367 |
+
"step": 323
|
| 2368 |
+
},
|
| 2369 |
+
{
|
| 2370 |
+
"epoch": 12.932038834951456,
|
| 2371 |
+
"grad_norm": 2.313508987426758,
|
| 2372 |
+
"learning_rate": 4.987024789458762e-05,
|
| 2373 |
+
"loss": 2.4438,
|
| 2374 |
+
"step": 324
|
| 2375 |
+
},
|
| 2376 |
+
{
|
| 2377 |
+
"epoch": 12.970873786407767,
|
| 2378 |
+
"grad_norm": 2.5614566802978516,
|
| 2379 |
+
"learning_rate": 4.986667187882186e-05,
|
| 2380 |
+
"loss": 2.557,
|
| 2381 |
+
"step": 325
|
| 2382 |
+
},
|
| 2383 |
+
{
|
| 2384 |
+
"epoch": 12.970873786407767,
|
| 2385 |
+
"eval_loss": 2.8833444118499756,
|
| 2386 |
+
"eval_runtime": 1.0479,
|
| 2387 |
+
"eval_samples_per_second": 20.995,
|
| 2388 |
+
"eval_steps_per_second": 5.726,
|
| 2389 |
+
"step": 325
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 13.03883495145631,
|
| 2393 |
+
"grad_norm": 5.110360145568848,
|
| 2394 |
+
"learning_rate": 4.9863047384206835e-05,
|
| 2395 |
+
"loss": 4.9144,
|
| 2396 |
+
"step": 326
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 13.077669902912621,
|
| 2400 |
+
"grad_norm": 2.7373085021972656,
|
| 2401 |
+
"learning_rate": 4.98593744178087e-05,
|
| 2402 |
+
"loss": 2.3994,
|
| 2403 |
+
"step": 327
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 13.116504854368932,
|
| 2407 |
+
"grad_norm": 2.542954206466675,
|
| 2408 |
+
"learning_rate": 4.985565298678809e-05,
|
| 2409 |
+
"loss": 2.3535,
|
| 2410 |
+
"step": 328
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 13.155339805825243,
|
| 2414 |
+
"grad_norm": 2.6374223232269287,
|
| 2415 |
+
"learning_rate": 4.985188309840012e-05,
|
| 2416 |
+
"loss": 2.3894,
|
| 2417 |
+
"step": 329
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 13.194174757281553,
|
| 2421 |
+
"grad_norm": 2.541004180908203,
|
| 2422 |
+
"learning_rate": 4.984806475999437e-05,
|
| 2423 |
+
"loss": 2.391,
|
| 2424 |
+
"step": 330
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 13.233009708737864,
|
| 2428 |
+
"grad_norm": 2.6150271892547607,
|
| 2429 |
+
"learning_rate": 4.984419797901491e-05,
|
| 2430 |
+
"loss": 2.3927,
|
| 2431 |
+
"step": 331
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 13.271844660194175,
|
| 2435 |
+
"grad_norm": 2.47719144821167,
|
| 2436 |
+
"learning_rate": 4.984028276300021e-05,
|
| 2437 |
+
"loss": 2.3751,
|
| 2438 |
+
"step": 332
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 13.310679611650485,
|
| 2442 |
+
"grad_norm": 2.679882764816284,
|
| 2443 |
+
"learning_rate": 4.983631911958319e-05,
|
| 2444 |
+
"loss": 2.374,
|
| 2445 |
+
"step": 333
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 13.349514563106796,
|
| 2449 |
+
"grad_norm": 2.784619092941284,
|
| 2450 |
+
"learning_rate": 4.983230705649118e-05,
|
| 2451 |
+
"loss": 2.3539,
|
| 2452 |
+
"step": 334
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 13.388349514563107,
|
| 2456 |
+
"grad_norm": 2.188197135925293,
|
| 2457 |
+
"learning_rate": 4.982824658154589e-05,
|
| 2458 |
+
"loss": 2.3553,
|
| 2459 |
+
"step": 335
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 13.427184466019417,
|
| 2463 |
+
"grad_norm": 2.232978582382202,
|
| 2464 |
+
"learning_rate": 4.982413770266342e-05,
|
| 2465 |
+
"loss": 2.3389,
|
| 2466 |
+
"step": 336
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 13.466019417475728,
|
| 2470 |
+
"grad_norm": 2.563889980316162,
|
| 2471 |
+
"learning_rate": 4.981998042785427e-05,
|
| 2472 |
+
"loss": 2.3623,
|
| 2473 |
+
"step": 337
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 13.504854368932039,
|
| 2477 |
+
"grad_norm": 2.9053828716278076,
|
| 2478 |
+
"learning_rate": 4.9815774765223226e-05,
|
| 2479 |
+
"loss": 2.3705,
|
| 2480 |
+
"step": 338
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 13.54368932038835,
|
| 2484 |
+
"grad_norm": 2.5447866916656494,
|
| 2485 |
+
"learning_rate": 4.9811520722969465e-05,
|
| 2486 |
+
"loss": 2.3216,
|
| 2487 |
+
"step": 339
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 13.58252427184466,
|
| 2491 |
+
"grad_norm": 3.22255277633667,
|
| 2492 |
+
"learning_rate": 4.9807218309386444e-05,
|
| 2493 |
+
"loss": 2.3418,
|
| 2494 |
+
"step": 340
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 13.62135922330097,
|
| 2498 |
+
"grad_norm": 3.154477119445801,
|
| 2499 |
+
"learning_rate": 4.980286753286195e-05,
|
| 2500 |
+
"loss": 2.3843,
|
| 2501 |
+
"step": 341
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 13.660194174757281,
|
| 2505 |
+
"grad_norm": 3.3448827266693115,
|
| 2506 |
+
"learning_rate": 4.979846840187804e-05,
|
| 2507 |
+
"loss": 2.419,
|
| 2508 |
+
"step": 342
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 13.699029126213592,
|
| 2512 |
+
"grad_norm": 3.275527238845825,
|
| 2513 |
+
"learning_rate": 4.9794020925011044e-05,
|
| 2514 |
+
"loss": 2.3756,
|
| 2515 |
+
"step": 343
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 13.737864077669903,
|
| 2519 |
+
"grad_norm": 2.3320887088775635,
|
| 2520 |
+
"learning_rate": 4.9789525110931545e-05,
|
| 2521 |
+
"loss": 2.3201,
|
| 2522 |
+
"step": 344
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 13.776699029126213,
|
| 2526 |
+
"grad_norm": 2.804107427597046,
|
| 2527 |
+
"learning_rate": 4.978498096840436e-05,
|
| 2528 |
+
"loss": 2.3461,
|
| 2529 |
+
"step": 345
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 13.815533980582524,
|
| 2533 |
+
"grad_norm": 2.809633255004883,
|
| 2534 |
+
"learning_rate": 4.978038850628854e-05,
|
| 2535 |
+
"loss": 2.3418,
|
| 2536 |
+
"step": 346
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 13.854368932038835,
|
| 2540 |
+
"grad_norm": 2.9983737468719482,
|
| 2541 |
+
"learning_rate": 4.977574773353732e-05,
|
| 2542 |
+
"loss": 2.4238,
|
| 2543 |
+
"step": 347
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 13.893203883495145,
|
| 2547 |
+
"grad_norm": 2.892005443572998,
|
| 2548 |
+
"learning_rate": 4.977105865919812e-05,
|
| 2549 |
+
"loss": 2.4266,
|
| 2550 |
+
"step": 348
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 13.932038834951456,
|
| 2554 |
+
"grad_norm": 2.766019821166992,
|
| 2555 |
+
"learning_rate": 4.976632129241252e-05,
|
| 2556 |
+
"loss": 2.3937,
|
| 2557 |
+
"step": 349
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 13.970873786407767,
|
| 2561 |
+
"grad_norm": 2.5251376628875732,
|
| 2562 |
+
"learning_rate": 4.976153564241628e-05,
|
| 2563 |
+
"loss": 2.3557,
|
| 2564 |
+
"step": 350
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 13.970873786407767,
|
| 2568 |
+
"eval_loss": 2.855170965194702,
|
| 2569 |
+
"eval_runtime": 1.0307,
|
| 2570 |
+
"eval_samples_per_second": 21.345,
|
| 2571 |
+
"eval_steps_per_second": 5.821,
|
| 2572 |
+
"step": 350
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 14.03883495145631,
|
| 2576 |
+
"grad_norm": 5.263445854187012,
|
| 2577 |
+
"learning_rate": 4.975670171853926e-05,
|
| 2578 |
+
"loss": 4.6103,
|
| 2579 |
+
"step": 351
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 14.077669902912621,
|
| 2583 |
+
"grad_norm": 2.6694159507751465,
|
| 2584 |
+
"learning_rate": 4.975181953020544e-05,
|
| 2585 |
+
"loss": 2.2714,
|
| 2586 |
+
"step": 352
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 14.116504854368932,
|
| 2590 |
+
"grad_norm": 3.4369680881500244,
|
| 2591 |
+
"learning_rate": 4.9746889086932895e-05,
|
| 2592 |
+
"loss": 2.2303,
|
| 2593 |
+
"step": 353
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 14.155339805825243,
|
| 2597 |
+
"grad_norm": 3.053704023361206,
|
| 2598 |
+
"learning_rate": 4.974191039833378e-05,
|
| 2599 |
+
"loss": 2.2659,
|
| 2600 |
+
"step": 354
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 14.194174757281553,
|
| 2604 |
+
"grad_norm": 2.9966983795166016,
|
| 2605 |
+
"learning_rate": 4.973688347411431e-05,
|
| 2606 |
+
"loss": 2.3092,
|
| 2607 |
+
"step": 355
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 14.233009708737864,
|
| 2611 |
+
"grad_norm": 2.965481758117676,
|
| 2612 |
+
"learning_rate": 4.9731808324074717e-05,
|
| 2613 |
+
"loss": 2.2537,
|
| 2614 |
+
"step": 356
|
| 2615 |
+
},
|
| 2616 |
+
{
|
| 2617 |
+
"epoch": 14.271844660194175,
|
| 2618 |
+
"grad_norm": 2.9761455059051514,
|
| 2619 |
+
"learning_rate": 4.9726684958109266e-05,
|
| 2620 |
+
"loss": 2.2865,
|
| 2621 |
+
"step": 357
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 14.310679611650485,
|
| 2625 |
+
"grad_norm": 2.936624050140381,
|
| 2626 |
+
"learning_rate": 4.972151338620623e-05,
|
| 2627 |
+
"loss": 2.2589,
|
| 2628 |
+
"step": 358
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 14.349514563106796,
|
| 2632 |
+
"grad_norm": 3.4442408084869385,
|
| 2633 |
+
"learning_rate": 4.971629361844785e-05,
|
| 2634 |
+
"loss": 2.2636,
|
| 2635 |
+
"step": 359
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"epoch": 14.388349514563107,
|
| 2639 |
+
"grad_norm": 3.0097110271453857,
|
| 2640 |
+
"learning_rate": 4.971102566501034e-05,
|
| 2641 |
+
"loss": 2.204,
|
| 2642 |
+
"step": 360
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"epoch": 14.427184466019417,
|
| 2646 |
+
"grad_norm": 3.7276322841644287,
|
| 2647 |
+
"learning_rate": 4.9705709536163824e-05,
|
| 2648 |
+
"loss": 2.2811,
|
| 2649 |
+
"step": 361
|
| 2650 |
+
},
|
| 2651 |
+
{
|
| 2652 |
+
"epoch": 14.466019417475728,
|
| 2653 |
+
"grad_norm": 2.8004868030548096,
|
| 2654 |
+
"learning_rate": 4.970034524227238e-05,
|
| 2655 |
+
"loss": 2.1964,
|
| 2656 |
+
"step": 362
|
| 2657 |
+
},
|
| 2658 |
+
{
|
| 2659 |
+
"epoch": 14.504854368932039,
|
| 2660 |
+
"grad_norm": 3.1439263820648193,
|
| 2661 |
+
"learning_rate": 4.969493279379398e-05,
|
| 2662 |
+
"loss": 2.294,
|
| 2663 |
+
"step": 363
|
| 2664 |
+
},
|
| 2665 |
+
{
|
| 2666 |
+
"epoch": 14.54368932038835,
|
| 2667 |
+
"grad_norm": 2.971735954284668,
|
| 2668 |
+
"learning_rate": 4.968947220128045e-05,
|
| 2669 |
+
"loss": 2.2882,
|
| 2670 |
+
"step": 364
|
| 2671 |
+
},
|
| 2672 |
+
{
|
| 2673 |
+
"epoch": 14.58252427184466,
|
| 2674 |
+
"grad_norm": 2.860797166824341,
|
| 2675 |
+
"learning_rate": 4.968396347537751e-05,
|
| 2676 |
+
"loss": 2.1807,
|
| 2677 |
+
"step": 365
|
| 2678 |
+
},
|
| 2679 |
+
{
|
| 2680 |
+
"epoch": 14.62135922330097,
|
| 2681 |
+
"grad_norm": 2.8869500160217285,
|
| 2682 |
+
"learning_rate": 4.96784066268247e-05,
|
| 2683 |
+
"loss": 2.267,
|
| 2684 |
+
"step": 366
|
| 2685 |
+
},
|
| 2686 |
+
{
|
| 2687 |
+
"epoch": 14.660194174757281,
|
| 2688 |
+
"grad_norm": 3.185670852661133,
|
| 2689 |
+
"learning_rate": 4.967280166645538e-05,
|
| 2690 |
+
"loss": 2.2956,
|
| 2691 |
+
"step": 367
|
| 2692 |
+
},
|
| 2693 |
+
{
|
| 2694 |
+
"epoch": 14.699029126213592,
|
| 2695 |
+
"grad_norm": 2.750898838043213,
|
| 2696 |
+
"learning_rate": 4.96671486051967e-05,
|
| 2697 |
+
"loss": 2.2429,
|
| 2698 |
+
"step": 368
|
| 2699 |
+
},
|
| 2700 |
+
{
|
| 2701 |
+
"epoch": 14.737864077669903,
|
| 2702 |
+
"grad_norm": 2.690889596939087,
|
| 2703 |
+
"learning_rate": 4.966144745406961e-05,
|
| 2704 |
+
"loss": 2.2645,
|
| 2705 |
+
"step": 369
|
| 2706 |
+
},
|
| 2707 |
+
{
|
| 2708 |
+
"epoch": 14.776699029126213,
|
| 2709 |
+
"grad_norm": 2.5257797241210938,
|
| 2710 |
+
"learning_rate": 4.965569822418877e-05,
|
| 2711 |
+
"loss": 2.1714,
|
| 2712 |
+
"step": 370
|
| 2713 |
+
},
|
| 2714 |
+
{
|
| 2715 |
+
"epoch": 14.815533980582524,
|
| 2716 |
+
"grad_norm": 2.550966739654541,
|
| 2717 |
+
"learning_rate": 4.964990092676263e-05,
|
| 2718 |
+
"loss": 2.2281,
|
| 2719 |
+
"step": 371
|
| 2720 |
+
},
|
| 2721 |
+
{
|
| 2722 |
+
"epoch": 14.854368932038835,
|
| 2723 |
+
"grad_norm": 2.6299831867218018,
|
| 2724 |
+
"learning_rate": 4.964405557309328e-05,
|
| 2725 |
+
"loss": 2.2925,
|
| 2726 |
+
"step": 372
|
| 2727 |
+
},
|
| 2728 |
+
{
|
| 2729 |
+
"epoch": 14.893203883495145,
|
| 2730 |
+
"grad_norm": 2.8115315437316895,
|
| 2731 |
+
"learning_rate": 4.963816217457657e-05,
|
| 2732 |
+
"loss": 2.3404,
|
| 2733 |
+
"step": 373
|
| 2734 |
+
},
|
| 2735 |
+
{
|
| 2736 |
+
"epoch": 14.932038834951456,
|
| 2737 |
+
"grad_norm": 2.646278142929077,
|
| 2738 |
+
"learning_rate": 4.9632220742701965e-05,
|
| 2739 |
+
"loss": 2.2326,
|
| 2740 |
+
"step": 374
|
| 2741 |
+
},
|
| 2742 |
+
{
|
| 2743 |
+
"epoch": 14.970873786407767,
|
| 2744 |
+
"grad_norm": 2.667069435119629,
|
| 2745 |
+
"learning_rate": 4.9626231289052596e-05,
|
| 2746 |
+
"loss": 2.318,
|
| 2747 |
+
"step": 375
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 14.970873786407767,
|
| 2751 |
+
"eval_loss": 2.8448235988616943,
|
| 2752 |
+
"eval_runtime": 0.9921,
|
| 2753 |
+
"eval_samples_per_second": 22.176,
|
| 2754 |
+
"eval_steps_per_second": 6.048,
|
| 2755 |
+
"step": 375
|
| 2756 |
+
}
|
| 2757 |
+
],
|
| 2758 |
+
"logging_steps": 1,
|
| 2759 |
+
"max_steps": 2500,
|
| 2760 |
+
"num_input_tokens_seen": 0,
|
| 2761 |
+
"num_train_epochs": 100,
|
| 2762 |
+
"save_steps": 500,
|
| 2763 |
+
"stateful_callbacks": {
|
| 2764 |
+
"TrainerControl": {
|
| 2765 |
+
"args": {
|
| 2766 |
+
"should_epoch_stop": false,
|
| 2767 |
+
"should_evaluate": false,
|
| 2768 |
+
"should_log": false,
|
| 2769 |
+
"should_save": true,
|
| 2770 |
+
"should_training_stop": false
|
| 2771 |
+
},
|
| 2772 |
+
"attributes": {}
|
| 2773 |
+
}
|
| 2774 |
+
},
|
| 2775 |
+
"total_flos": 3395309036544000.0,
|
| 2776 |
+
"train_batch_size": 4,
|
| 2777 |
+
"trial_name": null,
|
| 2778 |
+
"trial_params": null
|
| 2779 |
+
}
|