Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1713 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1713 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.9976580796252925,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 240,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.01249024199843872,
|
| 13 |
+
"grad_norm": 333.34182665688235,
|
| 14 |
+
"learning_rate": 8.333333333333333e-08,
|
| 15 |
+
"loss": 6.1714,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.02498048399687744,
|
| 20 |
+
"grad_norm": 331.81339010234353,
|
| 21 |
+
"learning_rate": 1.6666666666666665e-07,
|
| 22 |
+
"loss": 6.1711,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.03747072599531616,
|
| 27 |
+
"grad_norm": 307.63256043395177,
|
| 28 |
+
"learning_rate": 2.5e-07,
|
| 29 |
+
"loss": 5.7164,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.04996096799375488,
|
| 34 |
+
"grad_norm": 309.0645983743677,
|
| 35 |
+
"learning_rate": 3.333333333333333e-07,
|
| 36 |
+
"loss": 5.7099,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.0624512099921936,
|
| 41 |
+
"grad_norm": 314.3016663579013,
|
| 42 |
+
"learning_rate": 4.1666666666666667e-07,
|
| 43 |
+
"loss": 5.6908,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.07494145199063232,
|
| 48 |
+
"grad_norm": 276.9301419559884,
|
| 49 |
+
"learning_rate": 5e-07,
|
| 50 |
+
"loss": 4.9107,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.08743169398907104,
|
| 55 |
+
"grad_norm": 335.10720168603376,
|
| 56 |
+
"learning_rate": 5.833333333333334e-07,
|
| 57 |
+
"loss": 5.5982,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.09992193598750976,
|
| 62 |
+
"grad_norm": 358.6715543294415,
|
| 63 |
+
"learning_rate": 6.666666666666666e-07,
|
| 64 |
+
"loss": 5.8947,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.11241217798594848,
|
| 69 |
+
"grad_norm": 300.21068578562586,
|
| 70 |
+
"learning_rate": 7.5e-07,
|
| 71 |
+
"loss": 4.1567,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.1249024199843872,
|
| 76 |
+
"grad_norm": 279.52560584784806,
|
| 77 |
+
"learning_rate": 8.333333333333333e-07,
|
| 78 |
+
"loss": 3.6894,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.13739266198282593,
|
| 83 |
+
"grad_norm": 371.4635796992416,
|
| 84 |
+
"learning_rate": 9.166666666666665e-07,
|
| 85 |
+
"loss": 4.0341,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.14988290398126464,
|
| 90 |
+
"grad_norm": 187.07120853087892,
|
| 91 |
+
"learning_rate": 1e-06,
|
| 92 |
+
"loss": 1.5798,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.16237314597970337,
|
| 97 |
+
"grad_norm": 123.16954958124967,
|
| 98 |
+
"learning_rate": 1.0833333333333333e-06,
|
| 99 |
+
"loss": 1.3406,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.17486338797814208,
|
| 104 |
+
"grad_norm": 52.92896012880286,
|
| 105 |
+
"learning_rate": 1.1666666666666668e-06,
|
| 106 |
+
"loss": 1.192,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.1873536299765808,
|
| 111 |
+
"grad_norm": 43.47703599673511,
|
| 112 |
+
"learning_rate": 1.2499999999999999e-06,
|
| 113 |
+
"loss": 0.9641,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.19984387197501952,
|
| 118 |
+
"grad_norm": 19.95427980862239,
|
| 119 |
+
"learning_rate": 1.3333333333333332e-06,
|
| 120 |
+
"loss": 0.6483,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.21233411397345825,
|
| 125 |
+
"grad_norm": 16.236110696116484,
|
| 126 |
+
"learning_rate": 1.4166666666666667e-06,
|
| 127 |
+
"loss": 0.6221,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.22482435597189696,
|
| 132 |
+
"grad_norm": 12.340872841577973,
|
| 133 |
+
"learning_rate": 1.5e-06,
|
| 134 |
+
"loss": 0.5514,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.2373145979703357,
|
| 139 |
+
"grad_norm": 8.988345127438574,
|
| 140 |
+
"learning_rate": 1.5833333333333331e-06,
|
| 141 |
+
"loss": 0.5003,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.2498048399687744,
|
| 146 |
+
"grad_norm": 7.540989841292862,
|
| 147 |
+
"learning_rate": 1.6666666666666667e-06,
|
| 148 |
+
"loss": 0.478,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.26229508196721313,
|
| 153 |
+
"grad_norm": 7.565013994850406,
|
| 154 |
+
"learning_rate": 1.75e-06,
|
| 155 |
+
"loss": 0.4359,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.27478532396565186,
|
| 160 |
+
"grad_norm": 5.833975274571941,
|
| 161 |
+
"learning_rate": 1.833333333333333e-06,
|
| 162 |
+
"loss": 0.4368,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.28727556596409054,
|
| 167 |
+
"grad_norm": 5.698675786941914,
|
| 168 |
+
"learning_rate": 1.916666666666667e-06,
|
| 169 |
+
"loss": 0.4298,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.2997658079625293,
|
| 174 |
+
"grad_norm": 8.926069785223405,
|
| 175 |
+
"learning_rate": 2e-06,
|
| 176 |
+
"loss": 0.4158,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.312256049960968,
|
| 181 |
+
"grad_norm": 4.224854291340345,
|
| 182 |
+
"learning_rate": 1.9998942319271077e-06,
|
| 183 |
+
"loss": 0.357,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.32474629195940674,
|
| 188 |
+
"grad_norm": 5.64478441155974,
|
| 189 |
+
"learning_rate": 1.9995769500822006e-06,
|
| 190 |
+
"loss": 0.3728,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.3372365339578454,
|
| 195 |
+
"grad_norm": 3.8371814850820596,
|
| 196 |
+
"learning_rate": 1.999048221581858e-06,
|
| 197 |
+
"loss": 0.3491,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.34972677595628415,
|
| 202 |
+
"grad_norm": 5.25514580000384,
|
| 203 |
+
"learning_rate": 1.9983081582712684e-06,
|
| 204 |
+
"loss": 0.3661,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.3622170179547229,
|
| 209 |
+
"grad_norm": 4.632037225345287,
|
| 210 |
+
"learning_rate": 1.997356916700572e-06,
|
| 211 |
+
"loss": 0.3453,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.3747072599531616,
|
| 216 |
+
"grad_norm": 3.876830705230254,
|
| 217 |
+
"learning_rate": 1.9961946980917456e-06,
|
| 218 |
+
"loss": 0.3628,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.3871975019516003,
|
| 223 |
+
"grad_norm": 4.595476506721269,
|
| 224 |
+
"learning_rate": 1.994821748296033e-06,
|
| 225 |
+
"loss": 0.3838,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.39968774395003903,
|
| 230 |
+
"grad_norm": 3.384283939448791,
|
| 231 |
+
"learning_rate": 1.9932383577419428e-06,
|
| 232 |
+
"loss": 0.3016,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.41217798594847777,
|
| 237 |
+
"grad_norm": 3.321955588036486,
|
| 238 |
+
"learning_rate": 1.9914448613738106e-06,
|
| 239 |
+
"loss": 0.3438,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.4246682279469165,
|
| 244 |
+
"grad_norm": 4.787071404536738,
|
| 245 |
+
"learning_rate": 1.9894416385809443e-06,
|
| 246 |
+
"loss": 0.335,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.4371584699453552,
|
| 251 |
+
"grad_norm": 3.355328060502115,
|
| 252 |
+
"learning_rate": 1.987229113117374e-06,
|
| 253 |
+
"loss": 0.3186,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.4496487119437939,
|
| 258 |
+
"grad_norm": 3.9840801797103196,
|
| 259 |
+
"learning_rate": 1.984807753012208e-06,
|
| 260 |
+
"loss": 0.3687,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.46213895394223264,
|
| 265 |
+
"grad_norm": 2.524647429633578,
|
| 266 |
+
"learning_rate": 1.9821780704706307e-06,
|
| 267 |
+
"loss": 0.2942,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.4746291959406714,
|
| 272 |
+
"grad_norm": 3.2652314500044812,
|
| 273 |
+
"learning_rate": 1.9793406217655515e-06,
|
| 274 |
+
"loss": 0.3436,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.48711943793911006,
|
| 279 |
+
"grad_norm": 3.491461691732514,
|
| 280 |
+
"learning_rate": 1.976296007119933e-06,
|
| 281 |
+
"loss": 0.3535,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.4996096799375488,
|
| 286 |
+
"grad_norm": 3.4053191345338165,
|
| 287 |
+
"learning_rate": 1.9730448705798236e-06,
|
| 288 |
+
"loss": 0.3081,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.5120999219359875,
|
| 293 |
+
"grad_norm": 3.0789458493626913,
|
| 294 |
+
"learning_rate": 1.969587899878116e-06,
|
| 295 |
+
"loss": 0.279,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.5245901639344263,
|
| 300 |
+
"grad_norm": 3.56324988090656,
|
| 301 |
+
"learning_rate": 1.965925826289068e-06,
|
| 302 |
+
"loss": 0.2963,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.537080405932865,
|
| 307 |
+
"grad_norm": 3.417344048735335,
|
| 308 |
+
"learning_rate": 1.962059424473613e-06,
|
| 309 |
+
"loss": 0.3142,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.5495706479313037,
|
| 314 |
+
"grad_norm": 3.6075987144087405,
|
| 315 |
+
"learning_rate": 1.9579895123154886e-06,
|
| 316 |
+
"loss": 0.3031,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.5620608899297423,
|
| 321 |
+
"grad_norm": 3.0395538814065772,
|
| 322 |
+
"learning_rate": 1.953716950748227e-06,
|
| 323 |
+
"loss": 0.2955,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.5745511319281811,
|
| 328 |
+
"grad_norm": 2.318212959543301,
|
| 329 |
+
"learning_rate": 1.949242643573034e-06,
|
| 330 |
+
"loss": 0.2862,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.5870413739266198,
|
| 335 |
+
"grad_norm": 3.0215713239430055,
|
| 336 |
+
"learning_rate": 1.9445675372676047e-06,
|
| 337 |
+
"loss": 0.3159,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.5995316159250585,
|
| 342 |
+
"grad_norm": 2.7738349740427766,
|
| 343 |
+
"learning_rate": 1.9396926207859082e-06,
|
| 344 |
+
"loss": 0.3135,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.6120218579234973,
|
| 349 |
+
"grad_norm": 2.3545661243222447,
|
| 350 |
+
"learning_rate": 1.9346189253489886e-06,
|
| 351 |
+
"loss": 0.3108,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.624512099921936,
|
| 356 |
+
"grad_norm": 2.561260629458728,
|
| 357 |
+
"learning_rate": 1.929347524226822e-06,
|
| 358 |
+
"loss": 0.3186,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.6370023419203747,
|
| 363 |
+
"grad_norm": 2.1651780906956746,
|
| 364 |
+
"learning_rate": 1.9238795325112867e-06,
|
| 365 |
+
"loss": 0.2924,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.6494925839188135,
|
| 370 |
+
"grad_norm": 2.3257505624918706,
|
| 371 |
+
"learning_rate": 1.9182161068802742e-06,
|
| 372 |
+
"loss": 0.3121,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.6619828259172521,
|
| 377 |
+
"grad_norm": 2.9268718568620113,
|
| 378 |
+
"learning_rate": 1.912358445353014e-06,
|
| 379 |
+
"loss": 0.2904,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.6744730679156908,
|
| 384 |
+
"grad_norm": 2.9452148194700465,
|
| 385 |
+
"learning_rate": 1.9063077870366499e-06,
|
| 386 |
+
"loss": 0.324,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.6869633099141296,
|
| 391 |
+
"grad_norm": 2.335851105262128,
|
| 392 |
+
"learning_rate": 1.900065411864121e-06,
|
| 393 |
+
"loss": 0.2904,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.6994535519125683,
|
| 398 |
+
"grad_norm": 2.4563807404025253,
|
| 399 |
+
"learning_rate": 1.8936326403234122e-06,
|
| 400 |
+
"loss": 0.3011,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.711943793911007,
|
| 405 |
+
"grad_norm": 3.0378245415227187,
|
| 406 |
+
"learning_rate": 1.8870108331782216e-06,
|
| 407 |
+
"loss": 0.2861,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.7244340359094458,
|
| 412 |
+
"grad_norm": 2.631181915482674,
|
| 413 |
+
"learning_rate": 1.8802013911801108e-06,
|
| 414 |
+
"loss": 0.2852,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.7369242779078845,
|
| 419 |
+
"grad_norm": 2.3418756214951695,
|
| 420 |
+
"learning_rate": 1.8732057547721957e-06,
|
| 421 |
+
"loss": 0.2676,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.7494145199063232,
|
| 426 |
+
"grad_norm": 2.8302736512055486,
|
| 427 |
+
"learning_rate": 1.8660254037844386e-06,
|
| 428 |
+
"loss": 0.2863,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.7619047619047619,
|
| 433 |
+
"grad_norm": 2.2790557395501874,
|
| 434 |
+
"learning_rate": 1.8586618571206132e-06,
|
| 435 |
+
"loss": 0.3043,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.7743950039032006,
|
| 440 |
+
"grad_norm": 2.559374663391903,
|
| 441 |
+
"learning_rate": 1.8511166724369996e-06,
|
| 442 |
+
"loss": 0.2828,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.7868852459016393,
|
| 447 |
+
"grad_norm": 2.1614358791413824,
|
| 448 |
+
"learning_rate": 1.8433914458128857e-06,
|
| 449 |
+
"loss": 0.2651,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.7993754879000781,
|
| 454 |
+
"grad_norm": 2.4645072559654446,
|
| 455 |
+
"learning_rate": 1.8354878114129364e-06,
|
| 456 |
+
"loss": 0.3071,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.8118657298985168,
|
| 461 |
+
"grad_norm": 2.647403497657533,
|
| 462 |
+
"learning_rate": 1.8274074411415103e-06,
|
| 463 |
+
"loss": 0.3072,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.8243559718969555,
|
| 468 |
+
"grad_norm": 2.6871412061186195,
|
| 469 |
+
"learning_rate": 1.8191520442889917e-06,
|
| 470 |
+
"loss": 0.2873,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.8368462138953943,
|
| 475 |
+
"grad_norm": 2.614474129444855,
|
| 476 |
+
"learning_rate": 1.8107233671702122e-06,
|
| 477 |
+
"loss": 0.3104,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.849336455893833,
|
| 482 |
+
"grad_norm": 2.379951106255994,
|
| 483 |
+
"learning_rate": 1.8021231927550438e-06,
|
| 484 |
+
"loss": 0.3144,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.8618266978922716,
|
| 489 |
+
"grad_norm": 2.5393687484439944,
|
| 490 |
+
"learning_rate": 1.7933533402912351e-06,
|
| 491 |
+
"loss": 0.2745,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.8743169398907104,
|
| 496 |
+
"grad_norm": 1.9556907532815975,
|
| 497 |
+
"learning_rate": 1.7844156649195757e-06,
|
| 498 |
+
"loss": 0.2618,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.8868071818891491,
|
| 503 |
+
"grad_norm": 2.148199026533948,
|
| 504 |
+
"learning_rate": 1.7753120572814656e-06,
|
| 505 |
+
"loss": 0.2729,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.8992974238875878,
|
| 510 |
+
"grad_norm": 2.3224170748990467,
|
| 511 |
+
"learning_rate": 1.766044443118978e-06,
|
| 512 |
+
"loss": 0.2853,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.9117876658860266,
|
| 517 |
+
"grad_norm": 2.267350104430526,
|
| 518 |
+
"learning_rate": 1.756614782867493e-06,
|
| 519 |
+
"loss": 0.2733,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.9242779078844653,
|
| 524 |
+
"grad_norm": 2.351524938939129,
|
| 525 |
+
"learning_rate": 1.7470250712409959e-06,
|
| 526 |
+
"loss": 0.2968,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.936768149882904,
|
| 531 |
+
"grad_norm": 2.108422335765701,
|
| 532 |
+
"learning_rate": 1.737277336810124e-06,
|
| 533 |
+
"loss": 0.2888,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.9492583918813428,
|
| 538 |
+
"grad_norm": 2.2248377211351427,
|
| 539 |
+
"learning_rate": 1.7273736415730487e-06,
|
| 540 |
+
"loss": 0.2976,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.9617486338797814,
|
| 545 |
+
"grad_norm": 2.175098795574419,
|
| 546 |
+
"learning_rate": 1.7173160805192892e-06,
|
| 547 |
+
"loss": 0.3092,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.9742388758782201,
|
| 552 |
+
"grad_norm": 1.8748363701866508,
|
| 553 |
+
"learning_rate": 1.7071067811865474e-06,
|
| 554 |
+
"loss": 0.298,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.9867291178766588,
|
| 559 |
+
"grad_norm": 1.9531090618126403,
|
| 560 |
+
"learning_rate": 1.6967479032106548e-06,
|
| 561 |
+
"loss": 0.2676,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.9992193598750976,
|
| 566 |
+
"grad_norm": 1.9892889036457968,
|
| 567 |
+
"learning_rate": 1.6862416378687337e-06,
|
| 568 |
+
"loss": 0.2891,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.0117096018735363,
|
| 573 |
+
"grad_norm": 2.0922265867017575,
|
| 574 |
+
"learning_rate": 1.6755902076156602e-06,
|
| 575 |
+
"loss": 0.2853,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 1.024199843871975,
|
| 580 |
+
"grad_norm": 1.7933686475802664,
|
| 581 |
+
"learning_rate": 1.6647958656139376e-06,
|
| 582 |
+
"loss": 0.2572,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 1.0366900858704138,
|
| 587 |
+
"grad_norm": 1.7251659708541356,
|
| 588 |
+
"learning_rate": 1.6538608952570697e-06,
|
| 589 |
+
"loss": 0.238,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.0491803278688525,
|
| 594 |
+
"grad_norm": 1.9923700427112945,
|
| 595 |
+
"learning_rate": 1.6427876096865393e-06,
|
| 596 |
+
"loss": 0.2505,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 1.0616705698672912,
|
| 601 |
+
"grad_norm": 2.1695061172618666,
|
| 602 |
+
"learning_rate": 1.6315783513024974e-06,
|
| 603 |
+
"loss": 0.2462,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.07416081186573,
|
| 608 |
+
"grad_norm": 2.6038932185550006,
|
| 609 |
+
"learning_rate": 1.6202354912682598e-06,
|
| 610 |
+
"loss": 0.2646,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 1.0866510538641687,
|
| 615 |
+
"grad_norm": 2.0903770758951854,
|
| 616 |
+
"learning_rate": 1.6087614290087205e-06,
|
| 617 |
+
"loss": 0.2388,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 1.0991412958626072,
|
| 622 |
+
"grad_norm": 3.6960138953759056,
|
| 623 |
+
"learning_rate": 1.5971585917027862e-06,
|
| 624 |
+
"loss": 0.2627,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 1.111631537861046,
|
| 629 |
+
"grad_norm": 2.272646978953128,
|
| 630 |
+
"learning_rate": 1.5854294337699405e-06,
|
| 631 |
+
"loss": 0.2516,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.1241217798594847,
|
| 636 |
+
"grad_norm": 2.222109706573132,
|
| 637 |
+
"learning_rate": 1.573576436351046e-06,
|
| 638 |
+
"loss": 0.2632,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 1.1366120218579234,
|
| 643 |
+
"grad_norm": 2.5081770029553017,
|
| 644 |
+
"learning_rate": 1.5616021067834928e-06,
|
| 645 |
+
"loss": 0.2523,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 1.1491022638563622,
|
| 650 |
+
"grad_norm": 2.1312338402006223,
|
| 651 |
+
"learning_rate": 1.549508978070806e-06,
|
| 652 |
+
"loss": 0.2566,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 1.161592505854801,
|
| 657 |
+
"grad_norm": 1.8457305516523228,
|
| 658 |
+
"learning_rate": 1.537299608346824e-06,
|
| 659 |
+
"loss": 0.2395,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 1.1740827478532396,
|
| 664 |
+
"grad_norm": 2.036814273018422,
|
| 665 |
+
"learning_rate": 1.52497658033456e-06,
|
| 666 |
+
"loss": 0.2584,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 1.1865729898516784,
|
| 671 |
+
"grad_norm": 1.8483702565774836,
|
| 672 |
+
"learning_rate": 1.5125425007998652e-06,
|
| 673 |
+
"loss": 0.2554,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.199063231850117,
|
| 678 |
+
"grad_norm": 2.0979744963394813,
|
| 679 |
+
"learning_rate": 1.5e-06,
|
| 680 |
+
"loss": 0.2414,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 1.2115534738485558,
|
| 685 |
+
"grad_norm": 2.3583664561669844,
|
| 686 |
+
"learning_rate": 1.4873517311272424e-06,
|
| 687 |
+
"loss": 0.2551,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 1.2240437158469946,
|
| 692 |
+
"grad_norm": 2.2337794974540848,
|
| 693 |
+
"learning_rate": 1.4746003697476404e-06,
|
| 694 |
+
"loss": 0.2525,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 1.2365339578454333,
|
| 699 |
+
"grad_norm": 1.845664103882406,
|
| 700 |
+
"learning_rate": 1.461748613235034e-06,
|
| 701 |
+
"loss": 0.2514,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 1.249024199843872,
|
| 706 |
+
"grad_norm": 2.7335296328915715,
|
| 707 |
+
"learning_rate": 1.4487991802004622e-06,
|
| 708 |
+
"loss": 0.2283,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 1.2615144418423108,
|
| 713 |
+
"grad_norm": 2.588655697049314,
|
| 714 |
+
"learning_rate": 1.4357548099170793e-06,
|
| 715 |
+
"loss": 0.2413,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.2740046838407495,
|
| 720 |
+
"grad_norm": 2.226699294839387,
|
| 721 |
+
"learning_rate": 1.4226182617406994e-06,
|
| 722 |
+
"loss": 0.2516,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 1.2864949258391882,
|
| 727 |
+
"grad_norm": 1.7960578972564007,
|
| 728 |
+
"learning_rate": 1.4093923145260925e-06,
|
| 729 |
+
"loss": 0.2294,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 1.298985167837627,
|
| 734 |
+
"grad_norm": 2.6503082458747995,
|
| 735 |
+
"learning_rate": 1.3960797660391568e-06,
|
| 736 |
+
"loss": 0.2592,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 1.3114754098360657,
|
| 741 |
+
"grad_norm": 2.606481954010207,
|
| 742 |
+
"learning_rate": 1.3826834323650898e-06,
|
| 743 |
+
"loss": 0.2384,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 1.3239656518345044,
|
| 748 |
+
"grad_norm": 1.9470673847748146,
|
| 749 |
+
"learning_rate": 1.3692061473126845e-06,
|
| 750 |
+
"loss": 0.2332,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 1.336455893832943,
|
| 755 |
+
"grad_norm": 1.9445904993525396,
|
| 756 |
+
"learning_rate": 1.3556507618148766e-06,
|
| 757 |
+
"loss": 0.2603,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.3489461358313817,
|
| 762 |
+
"grad_norm": 2.248170475905102,
|
| 763 |
+
"learning_rate": 1.3420201433256689e-06,
|
| 764 |
+
"loss": 0.2434,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 1.3614363778298204,
|
| 769 |
+
"grad_norm": 2.088115052067367,
|
| 770 |
+
"learning_rate": 1.3283171752135611e-06,
|
| 771 |
+
"loss": 0.2288,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 1.3739266198282591,
|
| 776 |
+
"grad_norm": 2.3854571898457,
|
| 777 |
+
"learning_rate": 1.3145447561516136e-06,
|
| 778 |
+
"loss": 0.2695,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 1.3864168618266979,
|
| 783 |
+
"grad_norm": 2.145018749707294,
|
| 784 |
+
"learning_rate": 1.3007057995042729e-06,
|
| 785 |
+
"loss": 0.2489,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 1.3989071038251366,
|
| 790 |
+
"grad_norm": 1.8613707323253825,
|
| 791 |
+
"learning_rate": 1.2868032327110903e-06,
|
| 792 |
+
"loss": 0.2227,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 1.4113973458235753,
|
| 797 |
+
"grad_norm": 2.5364914930243856,
|
| 798 |
+
"learning_rate": 1.272839996667461e-06,
|
| 799 |
+
"loss": 0.2711,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.423887587822014,
|
| 804 |
+
"grad_norm": 2.121273202321119,
|
| 805 |
+
"learning_rate": 1.2588190451025207e-06,
|
| 806 |
+
"loss": 0.2555,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 1.4363778298204528,
|
| 811 |
+
"grad_norm": 2.069531172308207,
|
| 812 |
+
"learning_rate": 1.2447433439543238e-06,
|
| 813 |
+
"loss": 0.2498,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 1.4488680718188915,
|
| 818 |
+
"grad_norm": 2.325007363924956,
|
| 819 |
+
"learning_rate": 1.2306158707424401e-06,
|
| 820 |
+
"loss": 0.25,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 1.4613583138173303,
|
| 825 |
+
"grad_norm": 2.171800358619066,
|
| 826 |
+
"learning_rate": 1.2164396139381029e-06,
|
| 827 |
+
"loss": 0.2715,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 1.473848555815769,
|
| 832 |
+
"grad_norm": 2.1421401975112677,
|
| 833 |
+
"learning_rate": 1.202217572332038e-06,
|
| 834 |
+
"loss": 0.2702,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 1.4863387978142075,
|
| 839 |
+
"grad_norm": 2.0900214993295125,
|
| 840 |
+
"learning_rate": 1.1879527544001117e-06,
|
| 841 |
+
"loss": 0.243,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.4988290398126463,
|
| 846 |
+
"grad_norm": 1.9955976272194818,
|
| 847 |
+
"learning_rate": 1.1736481776669305e-06,
|
| 848 |
+
"loss": 0.2344,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 1.511319281811085,
|
| 853 |
+
"grad_norm": 1.8938474176211617,
|
| 854 |
+
"learning_rate": 1.1593068680675227e-06,
|
| 855 |
+
"loss": 0.2498,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 1.5238095238095237,
|
| 860 |
+
"grad_norm": 2.382039055201425,
|
| 861 |
+
"learning_rate": 1.1449318593072465e-06,
|
| 862 |
+
"loss": 0.2465,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 1.5362997658079625,
|
| 867 |
+
"grad_norm": 1.976250750423546,
|
| 868 |
+
"learning_rate": 1.1305261922200517e-06,
|
| 869 |
+
"loss": 0.235,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 1.5487900078064012,
|
| 874 |
+
"grad_norm": 2.7122911698191703,
|
| 875 |
+
"learning_rate": 1.1160929141252301e-06,
|
| 876 |
+
"loss": 0.2475,
|
| 877 |
+
"step": 124
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 1.56128024980484,
|
| 881 |
+
"grad_norm": 2.265005219419693,
|
| 882 |
+
"learning_rate": 1.101635078182802e-06,
|
| 883 |
+
"loss": 0.2642,
|
| 884 |
+
"step": 125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.5737704918032787,
|
| 888 |
+
"grad_norm": 2.19715621025085,
|
| 889 |
+
"learning_rate": 1.0871557427476583e-06,
|
| 890 |
+
"loss": 0.2552,
|
| 891 |
+
"step": 126
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 1.5862607338017174,
|
| 895 |
+
"grad_norm": 2.0831267066790864,
|
| 896 |
+
"learning_rate": 1.0726579707226107e-06,
|
| 897 |
+
"loss": 0.2473,
|
| 898 |
+
"step": 127
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 1.5987509758001561,
|
| 902 |
+
"grad_norm": 2.078276226596769,
|
| 903 |
+
"learning_rate": 1.0581448289104758e-06,
|
| 904 |
+
"loss": 0.2698,
|
| 905 |
+
"step": 128
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 1.6112412177985949,
|
| 909 |
+
"grad_norm": 2.271056755272453,
|
| 910 |
+
"learning_rate": 1.043619387365336e-06,
|
| 911 |
+
"loss": 0.2444,
|
| 912 |
+
"step": 129
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 1.6237314597970336,
|
| 916 |
+
"grad_norm": 2.011114368410887,
|
| 917 |
+
"learning_rate": 1.0290847187431114e-06,
|
| 918 |
+
"loss": 0.2372,
|
| 919 |
+
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 1.6362217017954723,
|
| 923 |
+
"grad_norm": 2.164168003710666,
|
| 924 |
+
"learning_rate": 1.0145438976515827e-06,
|
| 925 |
+
"loss": 0.2145,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.648711943793911,
|
| 930 |
+
"grad_norm": 2.4203414999934627,
|
| 931 |
+
"learning_rate": 1e-06,
|
| 932 |
+
"loss": 0.2565,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 1.6612021857923498,
|
| 937 |
+
"grad_norm": 2.3818088992857342,
|
| 938 |
+
"learning_rate": 9.854561023484172e-07,
|
| 939 |
+
"loss": 0.2317,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 1.6736924277907885,
|
| 944 |
+
"grad_norm": 2.5588039073451405,
|
| 945 |
+
"learning_rate": 9.709152812568885e-07,
|
| 946 |
+
"loss": 0.2496,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 1.6861826697892273,
|
| 951 |
+
"grad_norm": 2.234987132885702,
|
| 952 |
+
"learning_rate": 9.56380612634664e-07,
|
| 953 |
+
"loss": 0.2449,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 1.698672911787666,
|
| 958 |
+
"grad_norm": 2.013289884308116,
|
| 959 |
+
"learning_rate": 9.418551710895241e-07,
|
| 960 |
+
"loss": 0.2446,
|
| 961 |
+
"step": 136
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 1.7111631537861047,
|
| 965 |
+
"grad_norm": 2.217469816520338,
|
| 966 |
+
"learning_rate": 9.273420292773893e-07,
|
| 967 |
+
"loss": 0.2313,
|
| 968 |
+
"step": 137
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 1.7236533957845435,
|
| 972 |
+
"grad_norm": 1.8556673872083667,
|
| 973 |
+
"learning_rate": 9.128442572523417e-07,
|
| 974 |
+
"loss": 0.2438,
|
| 975 |
+
"step": 138
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 1.7361436377829822,
|
| 979 |
+
"grad_norm": 2.3299736172761163,
|
| 980 |
+
"learning_rate": 8.983649218171981e-07,
|
| 981 |
+
"loss": 0.2293,
|
| 982 |
+
"step": 139
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 1.748633879781421,
|
| 986 |
+
"grad_norm": 2.164678834834234,
|
| 987 |
+
"learning_rate": 8.839070858747696e-07,
|
| 988 |
+
"loss": 0.2619,
|
| 989 |
+
"step": 140
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 1.7611241217798594,
|
| 993 |
+
"grad_norm": 2.149872363103244,
|
| 994 |
+
"learning_rate": 8.694738077799486e-07,
|
| 995 |
+
"loss": 0.2325,
|
| 996 |
+
"step": 141
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 1.7736143637782982,
|
| 1000 |
+
"grad_norm": 2.0978087714663096,
|
| 1001 |
+
"learning_rate": 8.550681406927533e-07,
|
| 1002 |
+
"loss": 0.2575,
|
| 1003 |
+
"step": 142
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 1.786104605776737,
|
| 1007 |
+
"grad_norm": 2.477007242030832,
|
| 1008 |
+
"learning_rate": 8.406931319324775e-07,
|
| 1009 |
+
"loss": 0.2577,
|
| 1010 |
+
"step": 143
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 1.7985948477751756,
|
| 1014 |
+
"grad_norm": 2.541118932985011,
|
| 1015 |
+
"learning_rate": 8.263518223330696e-07,
|
| 1016 |
+
"loss": 0.2383,
|
| 1017 |
+
"step": 144
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 1.8110850897736144,
|
| 1021 |
+
"grad_norm": 2.6737704137534113,
|
| 1022 |
+
"learning_rate": 8.120472455998881e-07,
|
| 1023 |
+
"loss": 0.245,
|
| 1024 |
+
"step": 145
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 1.823575331772053,
|
| 1028 |
+
"grad_norm": 2.110715917930842,
|
| 1029 |
+
"learning_rate": 7.977824276679622e-07,
|
| 1030 |
+
"loss": 0.2345,
|
| 1031 |
+
"step": 146
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 1.8360655737704918,
|
| 1035 |
+
"grad_norm": 2.96478335133743,
|
| 1036 |
+
"learning_rate": 7.835603860618971e-07,
|
| 1037 |
+
"loss": 0.2535,
|
| 1038 |
+
"step": 147
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 1.8485558157689306,
|
| 1042 |
+
"grad_norm": 2.1301612266991294,
|
| 1043 |
+
"learning_rate": 7.693841292575598e-07,
|
| 1044 |
+
"loss": 0.2289,
|
| 1045 |
+
"step": 148
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 1.861046057767369,
|
| 1049 |
+
"grad_norm": 2.491103738998678,
|
| 1050 |
+
"learning_rate": 7.552566560456761e-07,
|
| 1051 |
+
"loss": 0.2466,
|
| 1052 |
+
"step": 149
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 1.8735362997658078,
|
| 1056 |
+
"grad_norm": 2.2668435127018416,
|
| 1057 |
+
"learning_rate": 7.411809548974791e-07,
|
| 1058 |
+
"loss": 0.2381,
|
| 1059 |
+
"step": 150
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 1.8860265417642466,
|
| 1063 |
+
"grad_norm": 2.4436212051193085,
|
| 1064 |
+
"learning_rate": 7.271600033325393e-07,
|
| 1065 |
+
"loss": 0.2383,
|
| 1066 |
+
"step": 151
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 1.8985167837626853,
|
| 1070 |
+
"grad_norm": 1.8897315487299207,
|
| 1071 |
+
"learning_rate": 7.1319676728891e-07,
|
| 1072 |
+
"loss": 0.2274,
|
| 1073 |
+
"step": 152
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 1.911007025761124,
|
| 1077 |
+
"grad_norm": 2.287365344695026,
|
| 1078 |
+
"learning_rate": 6.992942004957269e-07,
|
| 1079 |
+
"loss": 0.2429,
|
| 1080 |
+
"step": 153
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 1.9234972677595628,
|
| 1084 |
+
"grad_norm": 1.9129000748056102,
|
| 1085 |
+
"learning_rate": 6.854552438483865e-07,
|
| 1086 |
+
"loss": 0.2417,
|
| 1087 |
+
"step": 154
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 1.9359875097580015,
|
| 1091 |
+
"grad_norm": 2.368642372695352,
|
| 1092 |
+
"learning_rate": 6.71682824786439e-07,
|
| 1093 |
+
"loss": 0.2645,
|
| 1094 |
+
"step": 155
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 1.9484777517564402,
|
| 1098 |
+
"grad_norm": 2.2950766634902506,
|
| 1099 |
+
"learning_rate": 6.579798566743313e-07,
|
| 1100 |
+
"loss": 0.2638,
|
| 1101 |
+
"step": 156
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 1.960967993754879,
|
| 1105 |
+
"grad_norm": 1.89270198992086,
|
| 1106 |
+
"learning_rate": 6.443492381851236e-07,
|
| 1107 |
+
"loss": 0.2337,
|
| 1108 |
+
"step": 157
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 1.9734582357533177,
|
| 1112 |
+
"grad_norm": 2.1511648380759527,
|
| 1113 |
+
"learning_rate": 6.307938526873157e-07,
|
| 1114 |
+
"loss": 0.2304,
|
| 1115 |
+
"step": 158
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 1.9859484777517564,
|
| 1119 |
+
"grad_norm": 2.0795393693629713,
|
| 1120 |
+
"learning_rate": 6.173165676349102e-07,
|
| 1121 |
+
"loss": 0.2271,
|
| 1122 |
+
"step": 159
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 1.9984387197501952,
|
| 1126 |
+
"grad_norm": 2.3450931399506865,
|
| 1127 |
+
"learning_rate": 6.039202339608431e-07,
|
| 1128 |
+
"loss": 0.2608,
|
| 1129 |
+
"step": 160
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 2.010928961748634,
|
| 1133 |
+
"grad_norm": 2.284852904995893,
|
| 1134 |
+
"learning_rate": 5.906076854739073e-07,
|
| 1135 |
+
"loss": 0.2058,
|
| 1136 |
+
"step": 161
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 2.0234192037470726,
|
| 1140 |
+
"grad_norm": 1.9688349162650256,
|
| 1141 |
+
"learning_rate": 5.773817382593007e-07,
|
| 1142 |
+
"loss": 0.2261,
|
| 1143 |
+
"step": 162
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 2.0359094457455114,
|
| 1147 |
+
"grad_norm": 2.568547638226678,
|
| 1148 |
+
"learning_rate": 5.642451900829208e-07,
|
| 1149 |
+
"loss": 0.23,
|
| 1150 |
+
"step": 163
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 2.04839968774395,
|
| 1154 |
+
"grad_norm": 1.8092409775313807,
|
| 1155 |
+
"learning_rate": 5.512008197995378e-07,
|
| 1156 |
+
"loss": 0.2038,
|
| 1157 |
+
"step": 164
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 2.060889929742389,
|
| 1161 |
+
"grad_norm": 2.074505362754499,
|
| 1162 |
+
"learning_rate": 5.382513867649663e-07,
|
| 1163 |
+
"loss": 0.2137,
|
| 1164 |
+
"step": 165
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 2.0733801717408276,
|
| 1168 |
+
"grad_norm": 1.8711847928572765,
|
| 1169 |
+
"learning_rate": 5.253996302523596e-07,
|
| 1170 |
+
"loss": 0.2123,
|
| 1171 |
+
"step": 166
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 2.0858704137392663,
|
| 1175 |
+
"grad_norm": 1.9172143852284447,
|
| 1176 |
+
"learning_rate": 5.126482688727577e-07,
|
| 1177 |
+
"loss": 0.1864,
|
| 1178 |
+
"step": 167
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 2.098360655737705,
|
| 1182 |
+
"grad_norm": 2.284105736678824,
|
| 1183 |
+
"learning_rate": 5.000000000000002e-07,
|
| 1184 |
+
"loss": 0.2246,
|
| 1185 |
+
"step": 168
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 2.1108508977361438,
|
| 1189 |
+
"grad_norm": 2.0934006439262687,
|
| 1190 |
+
"learning_rate": 4.874574992001348e-07,
|
| 1191 |
+
"loss": 0.1907,
|
| 1192 |
+
"step": 169
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 2.1233411397345825,
|
| 1196 |
+
"grad_norm": 2.549899684848377,
|
| 1197 |
+
"learning_rate": 4.750234196654399e-07,
|
| 1198 |
+
"loss": 0.2371,
|
| 1199 |
+
"step": 170
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 2.1358313817330212,
|
| 1203 |
+
"grad_norm": 2.130000328754761,
|
| 1204 |
+
"learning_rate": 4.62700391653176e-07,
|
| 1205 |
+
"loss": 0.2164,
|
| 1206 |
+
"step": 171
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 2.14832162373146,
|
| 1210 |
+
"grad_norm": 2.299063509065943,
|
| 1211 |
+
"learning_rate": 4.50491021929194e-07,
|
| 1212 |
+
"loss": 0.2004,
|
| 1213 |
+
"step": 172
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 2.1608118657298987,
|
| 1217 |
+
"grad_norm": 2.7447743738908215,
|
| 1218 |
+
"learning_rate": 4.383978932165072e-07,
|
| 1219 |
+
"loss": 0.2194,
|
| 1220 |
+
"step": 173
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 2.1733021077283374,
|
| 1224 |
+
"grad_norm": 2.3612400219642837,
|
| 1225 |
+
"learning_rate": 4.2642356364895417e-07,
|
| 1226 |
+
"loss": 0.2046,
|
| 1227 |
+
"step": 174
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 2.185792349726776,
|
| 1231 |
+
"grad_norm": 2.179919179753777,
|
| 1232 |
+
"learning_rate": 4.1457056623005947e-07,
|
| 1233 |
+
"loss": 0.23,
|
| 1234 |
+
"step": 175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 2.1982825917252145,
|
| 1238 |
+
"grad_norm": 2.0734667196861793,
|
| 1239 |
+
"learning_rate": 4.02841408297214e-07,
|
| 1240 |
+
"loss": 0.1896,
|
| 1241 |
+
"step": 176
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 2.210772833723653,
|
| 1245 |
+
"grad_norm": 2.1560367007131336,
|
| 1246 |
+
"learning_rate": 3.912385709912793e-07,
|
| 1247 |
+
"loss": 0.2245,
|
| 1248 |
+
"step": 177
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 2.223263075722092,
|
| 1252 |
+
"grad_norm": 2.55546260550048,
|
| 1253 |
+
"learning_rate": 3.7976450873174007e-07,
|
| 1254 |
+
"loss": 0.2029,
|
| 1255 |
+
"step": 178
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 2.2357533177205307,
|
| 1259 |
+
"grad_norm": 2.0049407189339132,
|
| 1260 |
+
"learning_rate": 3.684216486975026e-07,
|
| 1261 |
+
"loss": 0.2048,
|
| 1262 |
+
"step": 179
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 2.2482435597189694,
|
| 1266 |
+
"grad_norm": 2.1287581084292384,
|
| 1267 |
+
"learning_rate": 3.5721239031346063e-07,
|
| 1268 |
+
"loss": 0.2148,
|
| 1269 |
+
"step": 180
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 2.260733801717408,
|
| 1273 |
+
"grad_norm": 2.15392293058629,
|
| 1274 |
+
"learning_rate": 3.461391047429304e-07,
|
| 1275 |
+
"loss": 0.1904,
|
| 1276 |
+
"step": 181
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 2.273224043715847,
|
| 1280 |
+
"grad_norm": 2.2187305597503144,
|
| 1281 |
+
"learning_rate": 3.352041343860621e-07,
|
| 1282 |
+
"loss": 0.2132,
|
| 1283 |
+
"step": 182
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 2.2857142857142856,
|
| 1287 |
+
"grad_norm": 2.150707531667909,
|
| 1288 |
+
"learning_rate": 3.2440979238433974e-07,
|
| 1289 |
+
"loss": 0.2068,
|
| 1290 |
+
"step": 183
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 2.2982045277127243,
|
| 1294 |
+
"grad_norm": 2.458172275479914,
|
| 1295 |
+
"learning_rate": 3.137583621312665e-07,
|
| 1296 |
+
"loss": 0.2125,
|
| 1297 |
+
"step": 184
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 2.310694769711163,
|
| 1301 |
+
"grad_norm": 2.136305265687496,
|
| 1302 |
+
"learning_rate": 3.032520967893453e-07,
|
| 1303 |
+
"loss": 0.2131,
|
| 1304 |
+
"step": 185
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 2.323185011709602,
|
| 1308 |
+
"grad_norm": 2.116709641624895,
|
| 1309 |
+
"learning_rate": 2.9289321881345254e-07,
|
| 1310 |
+
"loss": 0.1964,
|
| 1311 |
+
"step": 186
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 2.3356752537080405,
|
| 1315 |
+
"grad_norm": 2.0275110253062567,
|
| 1316 |
+
"learning_rate": 2.826839194807105e-07,
|
| 1317 |
+
"loss": 0.1996,
|
| 1318 |
+
"step": 187
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 2.3481654957064793,
|
| 1322 |
+
"grad_norm": 2.1907391585210596,
|
| 1323 |
+
"learning_rate": 2.7262635842695124e-07,
|
| 1324 |
+
"loss": 0.2103,
|
| 1325 |
+
"step": 188
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 2.360655737704918,
|
| 1329 |
+
"grad_norm": 2.054812643642808,
|
| 1330 |
+
"learning_rate": 2.62722663189876e-07,
|
| 1331 |
+
"loss": 0.2127,
|
| 1332 |
+
"step": 189
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 2.3731459797033567,
|
| 1336 |
+
"grad_norm": 2.0151091917707222,
|
| 1337 |
+
"learning_rate": 2.5297492875900415e-07,
|
| 1338 |
+
"loss": 0.1905,
|
| 1339 |
+
"step": 190
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 2.3856362217017955,
|
| 1343 |
+
"grad_norm": 2.2246729218243044,
|
| 1344 |
+
"learning_rate": 2.4338521713250716e-07,
|
| 1345 |
+
"loss": 0.1771,
|
| 1346 |
+
"step": 191
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 2.398126463700234,
|
| 1350 |
+
"grad_norm": 2.633375911757354,
|
| 1351 |
+
"learning_rate": 2.339555568810221e-07,
|
| 1352 |
+
"loss": 0.2267,
|
| 1353 |
+
"step": 192
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 2.410616705698673,
|
| 1357 |
+
"grad_norm": 2.131981214394647,
|
| 1358 |
+
"learning_rate": 2.2468794271853409e-07,
|
| 1359 |
+
"loss": 0.2024,
|
| 1360 |
+
"step": 193
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 2.4231069476971117,
|
| 1364 |
+
"grad_norm": 2.685987255782069,
|
| 1365 |
+
"learning_rate": 2.1558433508042427e-07,
|
| 1366 |
+
"loss": 0.2036,
|
| 1367 |
+
"step": 194
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 2.4355971896955504,
|
| 1371 |
+
"grad_norm": 2.300115290983245,
|
| 1372 |
+
"learning_rate": 2.0664665970876495e-07,
|
| 1373 |
+
"loss": 0.2029,
|
| 1374 |
+
"step": 195
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 2.448087431693989,
|
| 1378 |
+
"grad_norm": 2.1419156947800966,
|
| 1379 |
+
"learning_rate": 1.9787680724495614e-07,
|
| 1380 |
+
"loss": 0.1839,
|
| 1381 |
+
"step": 196
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 2.460577673692428,
|
| 1385 |
+
"grad_norm": 2.173748561932065,
|
| 1386 |
+
"learning_rate": 1.8927663282978778e-07,
|
| 1387 |
+
"loss": 0.1936,
|
| 1388 |
+
"step": 197
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 2.4730679156908666,
|
| 1392 |
+
"grad_norm": 2.3912234200745677,
|
| 1393 |
+
"learning_rate": 1.8084795571100809e-07,
|
| 1394 |
+
"loss": 0.1919,
|
| 1395 |
+
"step": 198
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 2.4855581576893053,
|
| 1399 |
+
"grad_norm": 2.1196203041108665,
|
| 1400 |
+
"learning_rate": 1.7259255885848944e-07,
|
| 1401 |
+
"loss": 0.2242,
|
| 1402 |
+
"step": 199
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 2.498048399687744,
|
| 1406 |
+
"grad_norm": 2.2036903494554925,
|
| 1407 |
+
"learning_rate": 1.6451218858706372e-07,
|
| 1408 |
+
"loss": 0.2054,
|
| 1409 |
+
"step": 200
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 2.510538641686183,
|
| 1413 |
+
"grad_norm": 2.319785772245257,
|
| 1414 |
+
"learning_rate": 1.566085541871145e-07,
|
| 1415 |
+
"loss": 0.2029,
|
| 1416 |
+
"step": 201
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 2.5230288836846215,
|
| 1420 |
+
"grad_norm": 2.1264169251162293,
|
| 1421 |
+
"learning_rate": 1.4888332756300027e-07,
|
| 1422 |
+
"loss": 0.1909,
|
| 1423 |
+
"step": 202
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 2.5355191256830603,
|
| 1427 |
+
"grad_norm": 2.4323418017999856,
|
| 1428 |
+
"learning_rate": 1.4133814287938707e-07,
|
| 1429 |
+
"loss": 0.2066,
|
| 1430 |
+
"step": 203
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 2.548009367681499,
|
| 1434 |
+
"grad_norm": 2.0027968704979657,
|
| 1435 |
+
"learning_rate": 1.3397459621556128e-07,
|
| 1436 |
+
"loss": 0.1985,
|
| 1437 |
+
"step": 204
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 2.5604996096799377,
|
| 1441 |
+
"grad_norm": 2.475632607635102,
|
| 1442 |
+
"learning_rate": 1.2679424522780425e-07,
|
| 1443 |
+
"loss": 0.1974,
|
| 1444 |
+
"step": 205
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 2.5729898516783765,
|
| 1448 |
+
"grad_norm": 1.9846733041773394,
|
| 1449 |
+
"learning_rate": 1.1979860881988902e-07,
|
| 1450 |
+
"loss": 0.1796,
|
| 1451 |
+
"step": 206
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 2.585480093676815,
|
| 1455 |
+
"grad_norm": 2.3817659716978543,
|
| 1456 |
+
"learning_rate": 1.1298916682177828e-07,
|
| 1457 |
+
"loss": 0.2161,
|
| 1458 |
+
"step": 207
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 2.597970335675254,
|
| 1462 |
+
"grad_norm": 2.459521582273135,
|
| 1463 |
+
"learning_rate": 1.0636735967658784e-07,
|
| 1464 |
+
"loss": 0.2103,
|
| 1465 |
+
"step": 208
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 2.6104605776736927,
|
| 1469 |
+
"grad_norm": 2.090971504013233,
|
| 1470 |
+
"learning_rate": 9.993458813587884e-08,
|
| 1471 |
+
"loss": 0.2055,
|
| 1472 |
+
"step": 209
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 2.6229508196721314,
|
| 1476 |
+
"grad_norm": 2.3055555358927062,
|
| 1477 |
+
"learning_rate": 9.369221296335006e-08,
|
| 1478 |
+
"loss": 0.2065,
|
| 1479 |
+
"step": 210
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 2.63544106167057,
|
| 1483 |
+
"grad_norm": 2.041387392736253,
|
| 1484 |
+
"learning_rate": 8.764155464698597e-08,
|
| 1485 |
+
"loss": 0.1844,
|
| 1486 |
+
"step": 211
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 2.647931303669009,
|
| 1490 |
+
"grad_norm": 2.472617943357409,
|
| 1491 |
+
"learning_rate": 8.178389311972612e-08,
|
| 1492 |
+
"loss": 0.2096,
|
| 1493 |
+
"step": 212
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 2.660421545667447,
|
| 1497 |
+
"grad_norm": 2.1068220843163608,
|
| 1498 |
+
"learning_rate": 7.612046748871326e-08,
|
| 1499 |
+
"loss": 0.201,
|
| 1500 |
+
"step": 213
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 2.672911787665886,
|
| 1504 |
+
"grad_norm": 2.464459157519032,
|
| 1505 |
+
"learning_rate": 7.065247577317745e-08,
|
| 1506 |
+
"loss": 0.2151,
|
| 1507 |
+
"step": 214
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 2.6854020296643246,
|
| 1511 |
+
"grad_norm": 2.774328675703441,
|
| 1512 |
+
"learning_rate": 6.538107465101162e-08,
|
| 1513 |
+
"loss": 0.232,
|
| 1514 |
+
"step": 215
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 2.6978922716627634,
|
| 1518 |
+
"grad_norm": 2.487108626104416,
|
| 1519 |
+
"learning_rate": 6.030737921409168e-08,
|
| 1520 |
+
"loss": 0.1922,
|
| 1521 |
+
"step": 216
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 2.710382513661202,
|
| 1525 |
+
"grad_norm": 2.5967236031039795,
|
| 1526 |
+
"learning_rate": 5.543246273239532e-08,
|
| 1527 |
+
"loss": 0.2019,
|
| 1528 |
+
"step": 217
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 2.722872755659641,
|
| 1532 |
+
"grad_norm": 2.011663740700634,
|
| 1533 |
+
"learning_rate": 5.0757356426966105e-08,
|
| 1534 |
+
"loss": 0.1933,
|
| 1535 |
+
"step": 218
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 2.7353629976580796,
|
| 1539 |
+
"grad_norm": 2.3584339414903828,
|
| 1540 |
+
"learning_rate": 4.6283049251773176e-08,
|
| 1541 |
+
"loss": 0.1939,
|
| 1542 |
+
"step": 219
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 2.7478532396565183,
|
| 1546 |
+
"grad_norm": 2.272256081489139,
|
| 1547 |
+
"learning_rate": 4.20104876845111e-08,
|
| 1548 |
+
"loss": 0.2227,
|
| 1549 |
+
"step": 220
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 2.760343481654957,
|
| 1553 |
+
"grad_norm": 2.1904538830957296,
|
| 1554 |
+
"learning_rate": 3.7940575526386854e-08,
|
| 1555 |
+
"loss": 0.2055,
|
| 1556 |
+
"step": 221
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 2.7728337236533958,
|
| 1560 |
+
"grad_norm": 2.289724060120113,
|
| 1561 |
+
"learning_rate": 3.4074173710931796e-08,
|
| 1562 |
+
"loss": 0.2,
|
| 1563 |
+
"step": 222
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 2.7853239656518345,
|
| 1567 |
+
"grad_norm": 2.1699126519469507,
|
| 1568 |
+
"learning_rate": 3.0412100121884e-08,
|
| 1569 |
+
"loss": 0.2139,
|
| 1570 |
+
"step": 223
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 2.797814207650273,
|
| 1574 |
+
"grad_norm": 2.30035821249737,
|
| 1575 |
+
"learning_rate": 2.6955129420176193e-08,
|
| 1576 |
+
"loss": 0.2221,
|
| 1577 |
+
"step": 224
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 2.810304449648712,
|
| 1581 |
+
"grad_norm": 2.681147325435332,
|
| 1582 |
+
"learning_rate": 2.3703992880066636e-08,
|
| 1583 |
+
"loss": 0.2006,
|
| 1584 |
+
"step": 225
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 2.8227946916471507,
|
| 1588 |
+
"grad_norm": 2.263173469238612,
|
| 1589 |
+
"learning_rate": 2.065937823444852e-08,
|
| 1590 |
+
"loss": 0.1876,
|
| 1591 |
+
"step": 226
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 2.8352849336455894,
|
| 1595 |
+
"grad_norm": 2.2155464901476862,
|
| 1596 |
+
"learning_rate": 1.782192952936934e-08,
|
| 1597 |
+
"loss": 0.1985,
|
| 1598 |
+
"step": 227
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 2.847775175644028,
|
| 1602 |
+
"grad_norm": 2.1919529195743856,
|
| 1603 |
+
"learning_rate": 1.519224698779198e-08,
|
| 1604 |
+
"loss": 0.1955,
|
| 1605 |
+
"step": 228
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 2.860265417642467,
|
| 1609 |
+
"grad_norm": 2.701323139535781,
|
| 1610 |
+
"learning_rate": 1.2770886882625952e-08,
|
| 1611 |
+
"loss": 0.2427,
|
| 1612 |
+
"step": 229
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 2.8727556596409056,
|
| 1616 |
+
"grad_norm": 2.1905675108362783,
|
| 1617 |
+
"learning_rate": 1.0558361419055529e-08,
|
| 1618 |
+
"loss": 0.2263,
|
| 1619 |
+
"step": 230
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 2.8852459016393444,
|
| 1623 |
+
"grad_norm": 1.7560789181693792,
|
| 1624 |
+
"learning_rate": 8.555138626189618e-09,
|
| 1625 |
+
"loss": 0.1932,
|
| 1626 |
+
"step": 231
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 2.897736143637783,
|
| 1630 |
+
"grad_norm": 2.5515885425815417,
|
| 1631 |
+
"learning_rate": 6.761642258056976e-09,
|
| 1632 |
+
"loss": 0.2055,
|
| 1633 |
+
"step": 232
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 2.910226385636222,
|
| 1637 |
+
"grad_norm": 2.405991079307559,
|
| 1638 |
+
"learning_rate": 5.1782517039670185e-09,
|
| 1639 |
+
"loss": 0.2077,
|
| 1640 |
+
"step": 233
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 2.9227166276346606,
|
| 1644 |
+
"grad_norm": 2.3253916585181234,
|
| 1645 |
+
"learning_rate": 3.805301908254455e-09,
|
| 1646 |
+
"loss": 0.2243,
|
| 1647 |
+
"step": 234
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 2.9352068696330993,
|
| 1651 |
+
"grad_norm": 2.1225264665332104,
|
| 1652 |
+
"learning_rate": 2.643083299427751e-09,
|
| 1653 |
+
"loss": 0.2048,
|
| 1654 |
+
"step": 235
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 2.947697111631538,
|
| 1658 |
+
"grad_norm": 2.2368606818193113,
|
| 1659 |
+
"learning_rate": 1.6918417287318243e-09,
|
| 1660 |
+
"loss": 0.2069,
|
| 1661 |
+
"step": 236
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 2.9601873536299763,
|
| 1665 |
+
"grad_norm": 2.1900572726398044,
|
| 1666 |
+
"learning_rate": 9.517784181422018e-10,
|
| 1667 |
+
"loss": 0.1869,
|
| 1668 |
+
"step": 237
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 2.972677595628415,
|
| 1672 |
+
"grad_norm": 2.316791855288412,
|
| 1673 |
+
"learning_rate": 4.230499177994007e-10,
|
| 1674 |
+
"loss": 0.2138,
|
| 1675 |
+
"step": 238
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 2.985167837626854,
|
| 1679 |
+
"grad_norm": 2.14734423046665,
|
| 1680 |
+
"learning_rate": 1.0576807289253142e-10,
|
| 1681 |
+
"loss": 0.1917,
|
| 1682 |
+
"step": 239
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 2.9976580796252925,
|
| 1686 |
+
"grad_norm": 2.076875010839052,
|
| 1687 |
+
"learning_rate": 0.0,
|
| 1688 |
+
"loss": 0.1845,
|
| 1689 |
+
"step": 240
|
| 1690 |
+
}
|
| 1691 |
+
],
|
| 1692 |
+
"logging_steps": 1,
|
| 1693 |
+
"max_steps": 240,
|
| 1694 |
+
"num_input_tokens_seen": 0,
|
| 1695 |
+
"num_train_epochs": 3,
|
| 1696 |
+
"save_steps": 100,
|
| 1697 |
+
"stateful_callbacks": {
|
| 1698 |
+
"TrainerControl": {
|
| 1699 |
+
"args": {
|
| 1700 |
+
"should_epoch_stop": false,
|
| 1701 |
+
"should_evaluate": false,
|
| 1702 |
+
"should_log": false,
|
| 1703 |
+
"should_save": true,
|
| 1704 |
+
"should_training_stop": true
|
| 1705 |
+
},
|
| 1706 |
+
"attributes": {}
|
| 1707 |
+
}
|
| 1708 |
+
},
|
| 1709 |
+
"total_flos": 571934689198080.0,
|
| 1710 |
+
"train_batch_size": 4,
|
| 1711 |
+
"trial_name": null,
|
| 1712 |
+
"trial_params": null
|
| 1713 |
+
}
|