Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +2233 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,2233 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.996815286624204,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3138,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.01592356687898089,
|
| 13 |
+
"grad_norm": 5.4758992195129395,
|
| 14 |
+
"learning_rate": 1.2738853503184714e-05,
|
| 15 |
+
"loss": 0.9116,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.03184713375796178,
|
| 20 |
+
"grad_norm": 4.248476028442383,
|
| 21 |
+
"learning_rate": 2.5477707006369428e-05,
|
| 22 |
+
"loss": 0.5347,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.04777070063694268,
|
| 27 |
+
"grad_norm": 2.722592830657959,
|
| 28 |
+
"learning_rate": 3.821656050955414e-05,
|
| 29 |
+
"loss": 0.3604,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.06369426751592357,
|
| 34 |
+
"grad_norm": 1.718458890914917,
|
| 35 |
+
"learning_rate": 5.0955414012738855e-05,
|
| 36 |
+
"loss": 0.2924,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.07961783439490445,
|
| 41 |
+
"grad_norm": 2.0852134227752686,
|
| 42 |
+
"learning_rate": 6.369426751592356e-05,
|
| 43 |
+
"loss": 0.2293,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.09554140127388536,
|
| 48 |
+
"grad_norm": 2.2306277751922607,
|
| 49 |
+
"learning_rate": 7.643312101910829e-05,
|
| 50 |
+
"loss": 0.2296,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.11146496815286625,
|
| 55 |
+
"grad_norm": 1.173863172531128,
|
| 56 |
+
"learning_rate": 8.9171974522293e-05,
|
| 57 |
+
"loss": 0.1754,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.12738853503184713,
|
| 62 |
+
"grad_norm": 1.4193692207336426,
|
| 63 |
+
"learning_rate": 0.00010191082802547771,
|
| 64 |
+
"loss": 0.1821,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.14331210191082802,
|
| 69 |
+
"grad_norm": 1.5220022201538086,
|
| 70 |
+
"learning_rate": 0.00011464968152866242,
|
| 71 |
+
"loss": 0.1634,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.1592356687898089,
|
| 76 |
+
"grad_norm": 1.3423898220062256,
|
| 77 |
+
"learning_rate": 0.00012738853503184712,
|
| 78 |
+
"loss": 0.1462,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.1751592356687898,
|
| 83 |
+
"grad_norm": 1.2148971557617188,
|
| 84 |
+
"learning_rate": 0.00014012738853503185,
|
| 85 |
+
"loss": 0.1386,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.1910828025477707,
|
| 90 |
+
"grad_norm": 0.9744131565093994,
|
| 91 |
+
"learning_rate": 0.00015286624203821657,
|
| 92 |
+
"loss": 0.1376,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.2070063694267516,
|
| 97 |
+
"grad_norm": 1.014123558998108,
|
| 98 |
+
"learning_rate": 0.0001656050955414013,
|
| 99 |
+
"loss": 0.1082,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.2229299363057325,
|
| 104 |
+
"grad_norm": 1.3054927587509155,
|
| 105 |
+
"learning_rate": 0.000178343949044586,
|
| 106 |
+
"loss": 0.1054,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.23885350318471338,
|
| 111 |
+
"grad_norm": 1.789557695388794,
|
| 112 |
+
"learning_rate": 0.00019108280254777072,
|
| 113 |
+
"loss": 0.1308,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.25477707006369427,
|
| 118 |
+
"grad_norm": 0.6913818717002869,
|
| 119 |
+
"learning_rate": 0.00019999950020955923,
|
| 120 |
+
"loss": 0.1197,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.27070063694267515,
|
| 125 |
+
"grad_norm": 0.9011440277099609,
|
| 126 |
+
"learning_rate": 0.00019999061518514538,
|
| 127 |
+
"loss": 0.0938,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.28662420382165604,
|
| 132 |
+
"grad_norm": 1.6200815439224243,
|
| 133 |
+
"learning_rate": 0.00019997062484234648,
|
| 134 |
+
"loss": 0.108,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.30254777070063693,
|
| 139 |
+
"grad_norm": 0.8650676608085632,
|
| 140 |
+
"learning_rate": 0.0001999395314013622,
|
| 141 |
+
"loss": 0.105,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.3184713375796178,
|
| 146 |
+
"grad_norm": 1.1472347974777222,
|
| 147 |
+
"learning_rate": 0.00019989733831554218,
|
| 148 |
+
"loss": 0.1219,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.3343949044585987,
|
| 153 |
+
"grad_norm": 1.0066096782684326,
|
| 154 |
+
"learning_rate": 0.00019984405027100277,
|
| 155 |
+
"loss": 0.0927,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.3503184713375796,
|
| 160 |
+
"grad_norm": 1.0204339027404785,
|
| 161 |
+
"learning_rate": 0.00019977967318610639,
|
| 162 |
+
"loss": 0.1119,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.3662420382165605,
|
| 167 |
+
"grad_norm": 1.074305534362793,
|
| 168 |
+
"learning_rate": 0.00019970421421080427,
|
| 169 |
+
"loss": 0.0903,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.3821656050955414,
|
| 174 |
+
"grad_norm": 0.9550490379333496,
|
| 175 |
+
"learning_rate": 0.0001996176817258424,
|
| 176 |
+
"loss": 0.091,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.3980891719745223,
|
| 181 |
+
"grad_norm": 1.0449860095977783,
|
| 182 |
+
"learning_rate": 0.0001995200853418307,
|
| 183 |
+
"loss": 0.0861,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.4140127388535032,
|
| 188 |
+
"grad_norm": 0.9337477684020996,
|
| 189 |
+
"learning_rate": 0.00019941143589817558,
|
| 190 |
+
"loss": 0.0907,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.4299363057324841,
|
| 195 |
+
"grad_norm": 0.7787328958511353,
|
| 196 |
+
"learning_rate": 0.00019929174546187616,
|
| 197 |
+
"loss": 0.0816,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.445859872611465,
|
| 202 |
+
"grad_norm": 0.5837294459342957,
|
| 203 |
+
"learning_rate": 0.000199161027326184,
|
| 204 |
+
"loss": 0.085,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.46178343949044587,
|
| 209 |
+
"grad_norm": 0.6995041370391846,
|
| 210 |
+
"learning_rate": 0.00019901929600912678,
|
| 211 |
+
"loss": 0.0767,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.47770700636942676,
|
| 216 |
+
"grad_norm": 0.9348034262657166,
|
| 217 |
+
"learning_rate": 0.00019886656725189575,
|
| 218 |
+
"loss": 0.08,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.49363057324840764,
|
| 223 |
+
"grad_norm": 0.458731085062027,
|
| 224 |
+
"learning_rate": 0.0001987028580170976,
|
| 225 |
+
"loss": 0.0819,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.5095541401273885,
|
| 230 |
+
"grad_norm": 0.5243947505950928,
|
| 231 |
+
"learning_rate": 0.00019852818648687044,
|
| 232 |
+
"loss": 0.078,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.5254777070063694,
|
| 237 |
+
"grad_norm": 0.6848501563072205,
|
| 238 |
+
"learning_rate": 0.00019834257206086442,
|
| 239 |
+
"loss": 0.0891,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.5414012738853503,
|
| 244 |
+
"grad_norm": 0.9554067254066467,
|
| 245 |
+
"learning_rate": 0.00019814603535408724,
|
| 246 |
+
"loss": 0.0759,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.5573248407643312,
|
| 251 |
+
"grad_norm": 0.49273306131362915,
|
| 252 |
+
"learning_rate": 0.0001979385981946144,
|
| 253 |
+
"loss": 0.0744,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.5732484076433121,
|
| 258 |
+
"grad_norm": 0.7968289852142334,
|
| 259 |
+
"learning_rate": 0.00019772028362116513,
|
| 260 |
+
"loss": 0.0674,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.589171974522293,
|
| 265 |
+
"grad_norm": 0.5523058176040649,
|
| 266 |
+
"learning_rate": 0.00019749111588054333,
|
| 267 |
+
"loss": 0.078,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.6050955414012739,
|
| 272 |
+
"grad_norm": 0.7879081964492798,
|
| 273 |
+
"learning_rate": 0.00019725112042494493,
|
| 274 |
+
"loss": 0.0727,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.6210191082802548,
|
| 279 |
+
"grad_norm": 0.897438645362854,
|
| 280 |
+
"learning_rate": 0.00019700032390913087,
|
| 281 |
+
"loss": 0.0882,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.6369426751592356,
|
| 286 |
+
"grad_norm": 0.40805375576019287,
|
| 287 |
+
"learning_rate": 0.00019673875418746678,
|
| 288 |
+
"loss": 0.0687,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.6528662420382165,
|
| 293 |
+
"grad_norm": 0.8794530034065247,
|
| 294 |
+
"learning_rate": 0.00019646644031082948,
|
| 295 |
+
"loss": 0.0745,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.6687898089171974,
|
| 300 |
+
"grad_norm": 0.671391487121582,
|
| 301 |
+
"learning_rate": 0.00019618341252338026,
|
| 302 |
+
"loss": 0.0773,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.6847133757961783,
|
| 307 |
+
"grad_norm": 0.7118439078330994,
|
| 308 |
+
"learning_rate": 0.00019588970225920612,
|
| 309 |
+
"loss": 0.0646,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.7006369426751592,
|
| 314 |
+
"grad_norm": 0.8737181425094604,
|
| 315 |
+
"learning_rate": 0.00019558534213882838,
|
| 316 |
+
"loss": 0.0667,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.7165605095541401,
|
| 321 |
+
"grad_norm": 0.9923394322395325,
|
| 322 |
+
"learning_rate": 0.00019527036596557983,
|
| 323 |
+
"loss": 0.0699,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.732484076433121,
|
| 328 |
+
"grad_norm": 0.8922154307365417,
|
| 329 |
+
"learning_rate": 0.0001949448087218504,
|
| 330 |
+
"loss": 0.062,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.7484076433121019,
|
| 335 |
+
"grad_norm": 0.6825265884399414,
|
| 336 |
+
"learning_rate": 0.0001946087065652018,
|
| 337 |
+
"loss": 0.0672,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.7643312101910829,
|
| 342 |
+
"grad_norm": 0.6685843467712402,
|
| 343 |
+
"learning_rate": 0.0001942620968243519,
|
| 344 |
+
"loss": 0.057,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.7802547770700637,
|
| 349 |
+
"grad_norm": 0.5671578049659729,
|
| 350 |
+
"learning_rate": 0.00019390501799502872,
|
| 351 |
+
"loss": 0.0543,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.7961783439490446,
|
| 356 |
+
"grad_norm": 0.5407360196113586,
|
| 357 |
+
"learning_rate": 0.00019353750973569498,
|
| 358 |
+
"loss": 0.0526,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.8121019108280255,
|
| 363 |
+
"grad_norm": 0.45758092403411865,
|
| 364 |
+
"learning_rate": 0.0001931596128631436,
|
| 365 |
+
"loss": 0.0523,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.8280254777070064,
|
| 370 |
+
"grad_norm": 0.5699948668479919,
|
| 371 |
+
"learning_rate": 0.0001927713693479643,
|
| 372 |
+
"loss": 0.0508,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.8439490445859873,
|
| 377 |
+
"grad_norm": 0.6552209258079529,
|
| 378 |
+
"learning_rate": 0.00019237282230988222,
|
| 379 |
+
"loss": 0.0736,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.8598726114649682,
|
| 384 |
+
"grad_norm": 0.6645591855049133,
|
| 385 |
+
"learning_rate": 0.00019196401601296898,
|
| 386 |
+
"loss": 0.067,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.8757961783439491,
|
| 391 |
+
"grad_norm": 0.5558132529258728,
|
| 392 |
+
"learning_rate": 0.00019154499586072646,
|
| 393 |
+
"loss": 0.0562,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.89171974522293,
|
| 398 |
+
"grad_norm": 0.42415371537208557,
|
| 399 |
+
"learning_rate": 0.00019111580839104418,
|
| 400 |
+
"loss": 0.064,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.9076433121019108,
|
| 405 |
+
"grad_norm": 0.6379917860031128,
|
| 406 |
+
"learning_rate": 0.00019067650127103055,
|
| 407 |
+
"loss": 0.0628,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.9235668789808917,
|
| 412 |
+
"grad_norm": 0.603799045085907,
|
| 413 |
+
"learning_rate": 0.0001902271232917189,
|
| 414 |
+
"loss": 0.0668,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.9394904458598726,
|
| 419 |
+
"grad_norm": 0.569402813911438,
|
| 420 |
+
"learning_rate": 0.0001897677243626485,
|
| 421 |
+
"loss": 0.0613,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.9554140127388535,
|
| 426 |
+
"grad_norm": 0.7091385722160339,
|
| 427 |
+
"learning_rate": 0.00018929835550632145,
|
| 428 |
+
"loss": 0.0665,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.9713375796178344,
|
| 433 |
+
"grad_norm": 0.5904159545898438,
|
| 434 |
+
"learning_rate": 0.00018881906885253593,
|
| 435 |
+
"loss": 0.0766,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.9872611464968153,
|
| 440 |
+
"grad_norm": 0.6689817905426025,
|
| 441 |
+
"learning_rate": 0.00018832991763259647,
|
| 442 |
+
"loss": 0.065,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.0031847133757963,
|
| 447 |
+
"grad_norm": 0.7255253791809082,
|
| 448 |
+
"learning_rate": 0.00018783095617340193,
|
| 449 |
+
"loss": 0.053,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.019108280254777,
|
| 454 |
+
"grad_norm": 0.41102227568626404,
|
| 455 |
+
"learning_rate": 0.0001873222398914117,
|
| 456 |
+
"loss": 0.0584,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.035031847133758,
|
| 461 |
+
"grad_norm": 0.5301253199577332,
|
| 462 |
+
"learning_rate": 0.00018680382528649093,
|
| 463 |
+
"loss": 0.0619,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.0509554140127388,
|
| 468 |
+
"grad_norm": 0.5525597929954529,
|
| 469 |
+
"learning_rate": 0.00018627576993563563,
|
| 470 |
+
"loss": 0.0556,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 1.0668789808917198,
|
| 475 |
+
"grad_norm": 0.5564507246017456,
|
| 476 |
+
"learning_rate": 0.0001857381324865777,
|
| 477 |
+
"loss": 0.0428,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 1.0828025477707006,
|
| 482 |
+
"grad_norm": 0.5977344512939453,
|
| 483 |
+
"learning_rate": 0.0001851909726512714,
|
| 484 |
+
"loss": 0.0576,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.0987261146496816,
|
| 489 |
+
"grad_norm": 0.49446621537208557,
|
| 490 |
+
"learning_rate": 0.00018463435119926176,
|
| 491 |
+
"loss": 0.0527,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 1.1146496815286624,
|
| 496 |
+
"grad_norm": 0.4952321946620941,
|
| 497 |
+
"learning_rate": 0.00018406832995093486,
|
| 498 |
+
"loss": 0.0472,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 1.1305732484076434,
|
| 503 |
+
"grad_norm": 0.40190061926841736,
|
| 504 |
+
"learning_rate": 0.0001834929717706522,
|
| 505 |
+
"loss": 0.0557,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.1464968152866242,
|
| 510 |
+
"grad_norm": 0.44065698981285095,
|
| 511 |
+
"learning_rate": 0.00018290834055976855,
|
| 512 |
+
"loss": 0.0473,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 1.1624203821656052,
|
| 517 |
+
"grad_norm": 0.5594244003295898,
|
| 518 |
+
"learning_rate": 0.00018231450124953495,
|
| 519 |
+
"loss": 0.0588,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.178343949044586,
|
| 524 |
+
"grad_norm": 0.49133527278900146,
|
| 525 |
+
"learning_rate": 0.00018171151979388714,
|
| 526 |
+
"loss": 0.0497,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.194267515923567,
|
| 531 |
+
"grad_norm": 0.5907294750213623,
|
| 532 |
+
"learning_rate": 0.00018109946316212052,
|
| 533 |
+
"loss": 0.0525,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 1.2101910828025477,
|
| 538 |
+
"grad_norm": 0.5373413562774658,
|
| 539 |
+
"learning_rate": 0.00018047839933145232,
|
| 540 |
+
"loss": 0.0553,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 1.2261146496815287,
|
| 545 |
+
"grad_norm": 0.5720402598381042,
|
| 546 |
+
"learning_rate": 0.00017984839727947164,
|
| 547 |
+
"loss": 0.0505,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.2420382165605095,
|
| 552 |
+
"grad_norm": 0.3708595037460327,
|
| 553 |
+
"learning_rate": 0.00017920952697647886,
|
| 554 |
+
"loss": 0.0544,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 1.2579617834394905,
|
| 559 |
+
"grad_norm": 0.48515430092811584,
|
| 560 |
+
"learning_rate": 0.00017856185937771416,
|
| 561 |
+
"loss": 0.041,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 1.2738853503184713,
|
| 566 |
+
"grad_norm": 0.6001298427581787,
|
| 567 |
+
"learning_rate": 0.00017790546641547722,
|
| 568 |
+
"loss": 0.0478,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.2898089171974523,
|
| 573 |
+
"grad_norm": 0.4815957844257355,
|
| 574 |
+
"learning_rate": 0.00017724042099113792,
|
| 575 |
+
"loss": 0.0439,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 1.305732484076433,
|
| 580 |
+
"grad_norm": 0.6788195371627808,
|
| 581 |
+
"learning_rate": 0.00017656679696703996,
|
| 582 |
+
"loss": 0.0573,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 1.321656050955414,
|
| 587 |
+
"grad_norm": 0.3975256383419037,
|
| 588 |
+
"learning_rate": 0.0001758846691582972,
|
| 589 |
+
"loss": 0.0472,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.3375796178343948,
|
| 594 |
+
"grad_norm": 0.4403384327888489,
|
| 595 |
+
"learning_rate": 0.00017519411332448444,
|
| 596 |
+
"loss": 0.0445,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 1.3535031847133758,
|
| 601 |
+
"grad_norm": 0.39884060621261597,
|
| 602 |
+
"learning_rate": 0.00017449520616122344,
|
| 603 |
+
"loss": 0.0473,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.3694267515923566,
|
| 608 |
+
"grad_norm": 0.41912099719047546,
|
| 609 |
+
"learning_rate": 0.00017378802529166462,
|
| 610 |
+
"loss": 0.0529,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 1.3853503184713376,
|
| 615 |
+
"grad_norm": 0.43497881293296814,
|
| 616 |
+
"learning_rate": 0.00017307264925786622,
|
| 617 |
+
"loss": 0.0462,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 1.4012738853503186,
|
| 622 |
+
"grad_norm": 0.6406415700912476,
|
| 623 |
+
"learning_rate": 0.0001723491575120708,
|
| 624 |
+
"loss": 0.0515,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 1.4171974522292994,
|
| 629 |
+
"grad_norm": 0.5098315477371216,
|
| 630 |
+
"learning_rate": 0.0001716176304078813,
|
| 631 |
+
"loss": 0.0537,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.4331210191082802,
|
| 636 |
+
"grad_norm": 0.39417779445648193,
|
| 637 |
+
"learning_rate": 0.0001708781491913365,
|
| 638 |
+
"loss": 0.0462,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 1.4490445859872612,
|
| 643 |
+
"grad_norm": 0.31453198194503784,
|
| 644 |
+
"learning_rate": 0.00017013079599188758,
|
| 645 |
+
"loss": 0.0496,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 1.4649681528662422,
|
| 650 |
+
"grad_norm": 0.467495322227478,
|
| 651 |
+
"learning_rate": 0.00016937565381327661,
|
| 652 |
+
"loss": 0.0518,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 1.480891719745223,
|
| 657 |
+
"grad_norm": 0.5229964256286621,
|
| 658 |
+
"learning_rate": 0.00016861280652431758,
|
| 659 |
+
"loss": 0.0588,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 1.4968152866242037,
|
| 664 |
+
"grad_norm": 0.5483565926551819,
|
| 665 |
+
"learning_rate": 0.00016784233884958212,
|
| 666 |
+
"loss": 0.0454,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 1.5127388535031847,
|
| 671 |
+
"grad_norm": 0.6096242070198059,
|
| 672 |
+
"learning_rate": 0.00016706433635998912,
|
| 673 |
+
"loss": 0.0458,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.5286624203821657,
|
| 678 |
+
"grad_norm": 0.34970104694366455,
|
| 679 |
+
"learning_rate": 0.00016627888546330138,
|
| 680 |
+
"loss": 0.0498,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 1.5445859872611465,
|
| 685 |
+
"grad_norm": 0.691208004951477,
|
| 686 |
+
"learning_rate": 0.00016548607339452853,
|
| 687 |
+
"loss": 0.0446,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 1.5605095541401273,
|
| 692 |
+
"grad_norm": 0.32399702072143555,
|
| 693 |
+
"learning_rate": 0.00016468598820623855,
|
| 694 |
+
"loss": 0.0452,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 1.5764331210191083,
|
| 699 |
+
"grad_norm": 0.36962831020355225,
|
| 700 |
+
"learning_rate": 0.0001638787187587784,
|
| 701 |
+
"loss": 0.0415,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 1.5923566878980893,
|
| 706 |
+
"grad_norm": 0.45716020464897156,
|
| 707 |
+
"learning_rate": 0.0001630643547104047,
|
| 708 |
+
"loss": 0.0439,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 1.60828025477707,
|
| 713 |
+
"grad_norm": 0.3454309403896332,
|
| 714 |
+
"learning_rate": 0.00016224298650732597,
|
| 715 |
+
"loss": 0.0532,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.6242038216560508,
|
| 720 |
+
"grad_norm": 0.6210212707519531,
|
| 721 |
+
"learning_rate": 0.00016141470537365754,
|
| 722 |
+
"loss": 0.0414,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 1.6401273885350318,
|
| 727 |
+
"grad_norm": 0.3841806650161743,
|
| 728 |
+
"learning_rate": 0.00016057960330128973,
|
| 729 |
+
"loss": 0.0484,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 1.6560509554140128,
|
| 734 |
+
"grad_norm": 0.3426171839237213,
|
| 735 |
+
"learning_rate": 0.0001597377730396708,
|
| 736 |
+
"loss": 0.0423,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 1.6719745222929936,
|
| 741 |
+
"grad_norm": 0.33869680762290955,
|
| 742 |
+
"learning_rate": 0.0001588893080855061,
|
| 743 |
+
"loss": 0.0436,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 1.6878980891719744,
|
| 748 |
+
"grad_norm": 0.40435710549354553,
|
| 749 |
+
"learning_rate": 0.00015803430267237383,
|
| 750 |
+
"loss": 0.0517,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 1.7038216560509554,
|
| 755 |
+
"grad_norm": 0.25334039330482483,
|
| 756 |
+
"learning_rate": 0.00015717285176025913,
|
| 757 |
+
"loss": 0.0339,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.7197452229299364,
|
| 762 |
+
"grad_norm": 0.40064361691474915,
|
| 763 |
+
"learning_rate": 0.00015630505102500754,
|
| 764 |
+
"loss": 0.0472,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 1.7356687898089171,
|
| 769 |
+
"grad_norm": 0.5399286150932312,
|
| 770 |
+
"learning_rate": 0.0001554309968476988,
|
| 771 |
+
"loss": 0.0448,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 1.7515923566878981,
|
| 776 |
+
"grad_norm": 0.36550888419151306,
|
| 777 |
+
"learning_rate": 0.0001545507863039426,
|
| 778 |
+
"loss": 0.0348,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 1.767515923566879,
|
| 783 |
+
"grad_norm": 0.5114671587944031,
|
| 784 |
+
"learning_rate": 0.00015366451715309674,
|
| 785 |
+
"loss": 0.0409,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 1.78343949044586,
|
| 790 |
+
"grad_norm": 0.4588598608970642,
|
| 791 |
+
"learning_rate": 0.00015277228782740988,
|
| 792 |
+
"loss": 0.0394,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 1.799363057324841,
|
| 797 |
+
"grad_norm": 0.35981065034866333,
|
| 798 |
+
"learning_rate": 0.0001518741974210892,
|
| 799 |
+
"loss": 0.0412,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.8152866242038217,
|
| 804 |
+
"grad_norm": 0.39596328139305115,
|
| 805 |
+
"learning_rate": 0.00015097034567929457,
|
| 806 |
+
"loss": 0.0369,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 1.8312101910828025,
|
| 811 |
+
"grad_norm": 0.45480477809906006,
|
| 812 |
+
"learning_rate": 0.0001500608329870606,
|
| 813 |
+
"loss": 0.0439,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 1.8471337579617835,
|
| 818 |
+
"grad_norm": 0.28407368063926697,
|
| 819 |
+
"learning_rate": 0.00014914576035814748,
|
| 820 |
+
"loss": 0.0483,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 1.8630573248407645,
|
| 825 |
+
"grad_norm": 0.4121032953262329,
|
| 826 |
+
"learning_rate": 0.000148225229423822,
|
| 827 |
+
"loss": 0.0488,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 1.8789808917197452,
|
| 832 |
+
"grad_norm": 0.44136524200439453,
|
| 833 |
+
"learning_rate": 0.00014729934242157004,
|
| 834 |
+
"loss": 0.0434,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 1.894904458598726,
|
| 839 |
+
"grad_norm": 0.43993085622787476,
|
| 840 |
+
"learning_rate": 0.00014636820218374177,
|
| 841 |
+
"loss": 0.0457,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.910828025477707,
|
| 846 |
+
"grad_norm": 0.3353908360004425,
|
| 847 |
+
"learning_rate": 0.00014543191212613047,
|
| 848 |
+
"loss": 0.0486,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 1.926751592356688,
|
| 853 |
+
"grad_norm": 0.4279186427593231,
|
| 854 |
+
"learning_rate": 0.0001444905762364872,
|
| 855 |
+
"loss": 0.0526,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 1.9426751592356688,
|
| 860 |
+
"grad_norm": 0.5773082375526428,
|
| 861 |
+
"learning_rate": 0.00014354429906297125,
|
| 862 |
+
"loss": 0.0462,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 1.9585987261146496,
|
| 867 |
+
"grad_norm": 0.39543482661247253,
|
| 868 |
+
"learning_rate": 0.00014259318570253863,
|
| 869 |
+
"loss": 0.0417,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 1.9745222929936306,
|
| 874 |
+
"grad_norm": 0.39346298575401306,
|
| 875 |
+
"learning_rate": 0.0001416373417892698,
|
| 876 |
+
"loss": 0.0361,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 1.9904458598726116,
|
| 881 |
+
"grad_norm": 0.32164451479911804,
|
| 882 |
+
"learning_rate": 0.0001406768734826375,
|
| 883 |
+
"loss": 0.0394,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 2.0063694267515926,
|
| 888 |
+
"grad_norm": 0.551193356513977,
|
| 889 |
+
"learning_rate": 0.00013971188745571616,
|
| 890 |
+
"loss": 0.042,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 2.022292993630573,
|
| 895 |
+
"grad_norm": 0.3745879530906677,
|
| 896 |
+
"learning_rate": 0.00013874249088333468,
|
| 897 |
+
"loss": 0.0432,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 2.038216560509554,
|
| 902 |
+
"grad_norm": 0.36253151297569275,
|
| 903 |
+
"learning_rate": 0.00013776879143017294,
|
| 904 |
+
"loss": 0.0395,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 2.054140127388535,
|
| 909 |
+
"grad_norm": 0.5354855060577393,
|
| 910 |
+
"learning_rate": 0.00013679089723880427,
|
| 911 |
+
"loss": 0.044,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 2.070063694267516,
|
| 916 |
+
"grad_norm": 0.5604785680770874,
|
| 917 |
+
"learning_rate": 0.00013580891691768483,
|
| 918 |
+
"loss": 0.0377,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 2.0859872611464967,
|
| 923 |
+
"grad_norm": 0.42235568165779114,
|
| 924 |
+
"learning_rate": 0.00013482295952909102,
|
| 925 |
+
"loss": 0.0381,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 2.1019108280254777,
|
| 930 |
+
"grad_norm": 0.3224591612815857,
|
| 931 |
+
"learning_rate": 0.0001338331345770068,
|
| 932 |
+
"loss": 0.0396,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 2.1178343949044587,
|
| 937 |
+
"grad_norm": 0.36872801184654236,
|
| 938 |
+
"learning_rate": 0.00013283955199496156,
|
| 939 |
+
"loss": 0.0469,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 2.1337579617834397,
|
| 944 |
+
"grad_norm": 0.38453301787376404,
|
| 945 |
+
"learning_rate": 0.0001318423221338207,
|
| 946 |
+
"loss": 0.0433,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 2.1496815286624202,
|
| 951 |
+
"grad_norm": 0.3108746409416199,
|
| 952 |
+
"learning_rate": 0.0001308415557495295,
|
| 953 |
+
"loss": 0.0373,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 2.1656050955414012,
|
| 958 |
+
"grad_norm": 0.41351979970932007,
|
| 959 |
+
"learning_rate": 0.00012983736399081242,
|
| 960 |
+
"loss": 0.0362,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 2.1815286624203822,
|
| 965 |
+
"grad_norm": 0.39941781759262085,
|
| 966 |
+
"learning_rate": 0.0001288298583868282,
|
| 967 |
+
"loss": 0.0444,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 2.1974522292993632,
|
| 972 |
+
"grad_norm": 0.5019637942314148,
|
| 973 |
+
"learning_rate": 0.00012781915083478332,
|
| 974 |
+
"loss": 0.0362,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 2.213375796178344,
|
| 979 |
+
"grad_norm": 0.44153285026550293,
|
| 980 |
+
"learning_rate": 0.00012680535358750402,
|
| 981 |
+
"loss": 0.048,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 2.229299363057325,
|
| 986 |
+
"grad_norm": 0.3250869810581207,
|
| 987 |
+
"learning_rate": 0.00012578857924096934,
|
| 988 |
+
"loss": 0.038,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 2.245222929936306,
|
| 993 |
+
"grad_norm": 0.38918137550354004,
|
| 994 |
+
"learning_rate": 0.00012476894072180575,
|
| 995 |
+
"loss": 0.0507,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 2.261146496815287,
|
| 1000 |
+
"grad_norm": 0.42312440276145935,
|
| 1001 |
+
"learning_rate": 0.00012374655127474487,
|
| 1002 |
+
"loss": 0.0401,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 2.2770700636942673,
|
| 1007 |
+
"grad_norm": 0.6262978911399841,
|
| 1008 |
+
"learning_rate": 0.00012272152445004647,
|
| 1009 |
+
"loss": 0.0476,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 2.2929936305732483,
|
| 1014 |
+
"grad_norm": 0.4742312431335449,
|
| 1015 |
+
"learning_rate": 0.00012169397409088695,
|
| 1016 |
+
"loss": 0.0375,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 2.3089171974522293,
|
| 1021 |
+
"grad_norm": 0.5429434180259705,
|
| 1022 |
+
"learning_rate": 0.00012066401432071551,
|
| 1023 |
+
"loss": 0.0486,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 2.3248407643312103,
|
| 1028 |
+
"grad_norm": 0.34995678067207336,
|
| 1029 |
+
"learning_rate": 0.00011963175953057924,
|
| 1030 |
+
"loss": 0.0398,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 2.340764331210191,
|
| 1035 |
+
"grad_norm": 0.26490700244903564,
|
| 1036 |
+
"learning_rate": 0.00011859732436641851,
|
| 1037 |
+
"loss": 0.0285,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 2.356687898089172,
|
| 1042 |
+
"grad_norm": 0.30831092596054077,
|
| 1043 |
+
"learning_rate": 0.00011756082371633373,
|
| 1044 |
+
"loss": 0.0344,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 2.372611464968153,
|
| 1049 |
+
"grad_norm": 0.35261446237564087,
|
| 1050 |
+
"learning_rate": 0.00011652237269782573,
|
| 1051 |
+
"loss": 0.0327,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 2.388535031847134,
|
| 1056 |
+
"grad_norm": 0.31007587909698486,
|
| 1057 |
+
"learning_rate": 0.00011548208664501034,
|
| 1058 |
+
"loss": 0.049,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 2.404458598726115,
|
| 1063 |
+
"grad_norm": 0.49864161014556885,
|
| 1064 |
+
"learning_rate": 0.00011444008109580884,
|
| 1065 |
+
"loss": 0.0318,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 2.4203821656050954,
|
| 1070 |
+
"grad_norm": 0.2871188521385193,
|
| 1071 |
+
"learning_rate": 0.00011339647177911599,
|
| 1072 |
+
"loss": 0.0335,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 2.4363057324840764,
|
| 1077 |
+
"grad_norm": 0.38693806529045105,
|
| 1078 |
+
"learning_rate": 0.00011235137460194676,
|
| 1079 |
+
"loss": 0.0334,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 2.4522292993630574,
|
| 1084 |
+
"grad_norm": 0.34477829933166504,
|
| 1085 |
+
"learning_rate": 0.00011130490563656326,
|
| 1086 |
+
"loss": 0.0278,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 2.468152866242038,
|
| 1091 |
+
"grad_norm": 0.21558038890361786,
|
| 1092 |
+
"learning_rate": 0.00011025718110758338,
|
| 1093 |
+
"loss": 0.0425,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 2.484076433121019,
|
| 1098 |
+
"grad_norm": 0.3483816683292389,
|
| 1099 |
+
"learning_rate": 0.0001092083173790724,
|
| 1100 |
+
"loss": 0.0308,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 2.5,
|
| 1105 |
+
"grad_norm": 0.365614652633667,
|
| 1106 |
+
"learning_rate": 0.00010815843094161927,
|
| 1107 |
+
"loss": 0.0309,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 2.515923566878981,
|
| 1112 |
+
"grad_norm": 0.4408479928970337,
|
| 1113 |
+
"learning_rate": 0.00010710763839939857,
|
| 1114 |
+
"loss": 0.034,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 2.531847133757962,
|
| 1119 |
+
"grad_norm": 0.2580474615097046,
|
| 1120 |
+
"learning_rate": 0.00010605605645722024,
|
| 1121 |
+
"loss": 0.0349,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 2.5477707006369426,
|
| 1126 |
+
"grad_norm": 0.3337418735027313,
|
| 1127 |
+
"learning_rate": 0.0001050038019075678,
|
| 1128 |
+
"loss": 0.0342,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 2.5636942675159236,
|
| 1133 |
+
"grad_norm": 0.41936802864074707,
|
| 1134 |
+
"learning_rate": 0.00010395099161762698,
|
| 1135 |
+
"loss": 0.0404,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 2.5796178343949046,
|
| 1140 |
+
"grad_norm": 0.3816032111644745,
|
| 1141 |
+
"learning_rate": 0.00010289774251630602,
|
| 1142 |
+
"loss": 0.0442,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 2.595541401273885,
|
| 1147 |
+
"grad_norm": 0.30350789427757263,
|
| 1148 |
+
"learning_rate": 0.00010184417158124915,
|
| 1149 |
+
"loss": 0.036,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 2.611464968152866,
|
| 1154 |
+
"grad_norm": 0.20693911612033844,
|
| 1155 |
+
"learning_rate": 0.00010079039582584457,
|
| 1156 |
+
"loss": 0.0293,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 2.627388535031847,
|
| 1161 |
+
"grad_norm": 0.31562432646751404,
|
| 1162 |
+
"learning_rate": 9.973653228622867e-05,
|
| 1163 |
+
"loss": 0.0268,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 2.643312101910828,
|
| 1168 |
+
"grad_norm": 0.3229486048221588,
|
| 1169 |
+
"learning_rate": 9.86826980082874e-05,
|
| 1170 |
+
"loss": 0.0385,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 2.659235668789809,
|
| 1175 |
+
"grad_norm": 0.20771260559558868,
|
| 1176 |
+
"learning_rate": 9.762901003465679e-05,
|
| 1177 |
+
"loss": 0.0293,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 2.6751592356687897,
|
| 1182 |
+
"grad_norm": 0.2522100508213043,
|
| 1183 |
+
"learning_rate": 9.657558539172376e-05,
|
| 1184 |
+
"loss": 0.0273,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 2.6910828025477707,
|
| 1189 |
+
"grad_norm": 0.30410313606262207,
|
| 1190 |
+
"learning_rate": 9.552254107662888e-05,
|
| 1191 |
+
"loss": 0.039,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 2.7070063694267517,
|
| 1196 |
+
"grad_norm": 0.5962004661560059,
|
| 1197 |
+
"learning_rate": 9.446999404427209e-05,
|
| 1198 |
+
"loss": 0.0355,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 2.722929936305732,
|
| 1203 |
+
"grad_norm": 0.3611944913864136,
|
| 1204 |
+
"learning_rate": 9.341806119432324e-05,
|
| 1205 |
+
"loss": 0.0299,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 2.738853503184713,
|
| 1210 |
+
"grad_norm": 0.23662623763084412,
|
| 1211 |
+
"learning_rate": 9.2366859358239e-05,
|
| 1212 |
+
"loss": 0.0305,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 2.754777070063694,
|
| 1217 |
+
"grad_norm": 0.27245596051216125,
|
| 1218 |
+
"learning_rate": 9.131650528628687e-05,
|
| 1219 |
+
"loss": 0.0265,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 2.770700636942675,
|
| 1224 |
+
"grad_norm": 0.23096629977226257,
|
| 1225 |
+
"learning_rate": 9.026711563457868e-05,
|
| 1226 |
+
"loss": 0.0373,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 2.786624203821656,
|
| 1231 |
+
"grad_norm": 0.23357895016670227,
|
| 1232 |
+
"learning_rate": 8.921880695211431e-05,
|
| 1233 |
+
"loss": 0.0301,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 2.802547770700637,
|
| 1238 |
+
"grad_norm": 0.36908674240112305,
|
| 1239 |
+
"learning_rate": 8.817169566783723e-05,
|
| 1240 |
+
"loss": 0.0276,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 2.8184713375796178,
|
| 1245 |
+
"grad_norm": 0.20657651126384735,
|
| 1246 |
+
"learning_rate": 8.712589807770349e-05,
|
| 1247 |
+
"loss": 0.0299,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 2.8343949044585988,
|
| 1252 |
+
"grad_norm": 0.27385637164115906,
|
| 1253 |
+
"learning_rate": 8.60815303317657e-05,
|
| 1254 |
+
"loss": 0.0312,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 2.8503184713375798,
|
| 1259 |
+
"grad_norm": 0.4029058516025543,
|
| 1260 |
+
"learning_rate": 8.503870842127267e-05,
|
| 1261 |
+
"loss": 0.0391,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 2.8662420382165603,
|
| 1266 |
+
"grad_norm": 0.2755729556083679,
|
| 1267 |
+
"learning_rate": 8.399754816578718e-05,
|
| 1268 |
+
"loss": 0.0367,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 2.8821656050955413,
|
| 1273 |
+
"grad_norm": 0.2416078746318817,
|
| 1274 |
+
"learning_rate": 8.295816520032271e-05,
|
| 1275 |
+
"loss": 0.0274,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 2.8980891719745223,
|
| 1280 |
+
"grad_norm": 0.22337061166763306,
|
| 1281 |
+
"learning_rate": 8.192067496250038e-05,
|
| 1282 |
+
"loss": 0.0219,
|
| 1283 |
+
"step": 1820
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 2.9140127388535033,
|
| 1287 |
+
"grad_norm": 0.21762487292289734,
|
| 1288 |
+
"learning_rate": 8.088519267972813e-05,
|
| 1289 |
+
"loss": 0.023,
|
| 1290 |
+
"step": 1830
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 2.9299363057324843,
|
| 1294 |
+
"grad_norm": 0.32507646083831787,
|
| 1295 |
+
"learning_rate": 7.985183335640331e-05,
|
| 1296 |
+
"loss": 0.0284,
|
| 1297 |
+
"step": 1840
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 2.945859872611465,
|
| 1301 |
+
"grad_norm": 0.3356756269931793,
|
| 1302 |
+
"learning_rate": 7.882071176113967e-05,
|
| 1303 |
+
"loss": 0.0304,
|
| 1304 |
+
"step": 1850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 2.961783439490446,
|
| 1308 |
+
"grad_norm": 0.31162557005882263,
|
| 1309 |
+
"learning_rate": 7.779194241402071e-05,
|
| 1310 |
+
"loss": 0.0282,
|
| 1311 |
+
"step": 1860
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 2.977707006369427,
|
| 1315 |
+
"grad_norm": 0.24151593446731567,
|
| 1316 |
+
"learning_rate": 7.676563957388097e-05,
|
| 1317 |
+
"loss": 0.0248,
|
| 1318 |
+
"step": 1870
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 2.9936305732484074,
|
| 1322 |
+
"grad_norm": 0.41317158937454224,
|
| 1323 |
+
"learning_rate": 7.574191722561562e-05,
|
| 1324 |
+
"loss": 0.0303,
|
| 1325 |
+
"step": 1880
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 3.0095541401273884,
|
| 1329 |
+
"grad_norm": 0.33723485469818115,
|
| 1330 |
+
"learning_rate": 7.472088906752124e-05,
|
| 1331 |
+
"loss": 0.0326,
|
| 1332 |
+
"step": 1890
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 3.0254777070063694,
|
| 1336 |
+
"grad_norm": 0.402401328086853,
|
| 1337 |
+
"learning_rate": 7.37026684986678e-05,
|
| 1338 |
+
"loss": 0.0327,
|
| 1339 |
+
"step": 1900
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 3.0414012738853504,
|
| 1343 |
+
"grad_norm": 0.2733253538608551,
|
| 1344 |
+
"learning_rate": 7.26873686063045e-05,
|
| 1345 |
+
"loss": 0.0286,
|
| 1346 |
+
"step": 1910
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 3.0573248407643314,
|
| 1350 |
+
"grad_norm": 0.2582903802394867,
|
| 1351 |
+
"learning_rate": 7.16751021532994e-05,
|
| 1352 |
+
"loss": 0.0232,
|
| 1353 |
+
"step": 1920
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 3.073248407643312,
|
| 1357 |
+
"grad_norm": 0.3948001265525818,
|
| 1358 |
+
"learning_rate": 7.066598156561606e-05,
|
| 1359 |
+
"loss": 0.0331,
|
| 1360 |
+
"step": 1930
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 3.089171974522293,
|
| 1364 |
+
"grad_norm": 0.23859384655952454,
|
| 1365 |
+
"learning_rate": 6.966011891982678e-05,
|
| 1366 |
+
"loss": 0.0253,
|
| 1367 |
+
"step": 1940
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 3.105095541401274,
|
| 1371 |
+
"grad_norm": 0.2874588072299957,
|
| 1372 |
+
"learning_rate": 6.865762593066513e-05,
|
| 1373 |
+
"loss": 0.0316,
|
| 1374 |
+
"step": 1950
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 3.121019108280255,
|
| 1378 |
+
"grad_norm": 0.40092089772224426,
|
| 1379 |
+
"learning_rate": 6.76586139386184e-05,
|
| 1380 |
+
"loss": 0.0294,
|
| 1381 |
+
"step": 1960
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 3.1369426751592355,
|
| 1385 |
+
"grad_norm": 0.2927519679069519,
|
| 1386 |
+
"learning_rate": 6.666319389756189e-05,
|
| 1387 |
+
"loss": 0.0323,
|
| 1388 |
+
"step": 1970
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 3.1528662420382165,
|
| 1392 |
+
"grad_norm": 0.2560182809829712,
|
| 1393 |
+
"learning_rate": 6.567147636243576e-05,
|
| 1394 |
+
"loss": 0.0292,
|
| 1395 |
+
"step": 1980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 3.1687898089171975,
|
| 1399 |
+
"grad_norm": 0.18076066672801971,
|
| 1400 |
+
"learning_rate": 6.468357147696653e-05,
|
| 1401 |
+
"loss": 0.0223,
|
| 1402 |
+
"step": 1990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 3.1847133757961785,
|
| 1406 |
+
"grad_norm": 0.32147490978240967,
|
| 1407 |
+
"learning_rate": 6.36995889614342e-05,
|
| 1408 |
+
"loss": 0.028,
|
| 1409 |
+
"step": 2000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 3.200636942675159,
|
| 1413 |
+
"grad_norm": 0.24778632819652557,
|
| 1414 |
+
"learning_rate": 6.271963810048607e-05,
|
| 1415 |
+
"loss": 0.0244,
|
| 1416 |
+
"step": 2010
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 3.21656050955414,
|
| 1420 |
+
"grad_norm": 0.2591859698295593,
|
| 1421 |
+
"learning_rate": 6.174382773099938e-05,
|
| 1422 |
+
"loss": 0.0252,
|
| 1423 |
+
"step": 2020
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 3.232484076433121,
|
| 1427 |
+
"grad_norm": 0.37974950671195984,
|
| 1428 |
+
"learning_rate": 6.077226622999355e-05,
|
| 1429 |
+
"loss": 0.0293,
|
| 1430 |
+
"step": 2030
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 3.248407643312102,
|
| 1434 |
+
"grad_norm": 0.36128807067871094,
|
| 1435 |
+
"learning_rate": 5.980506150259323e-05,
|
| 1436 |
+
"loss": 0.0253,
|
| 1437 |
+
"step": 2040
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 3.2643312101910826,
|
| 1441 |
+
"grad_norm": 0.2899702489376068,
|
| 1442 |
+
"learning_rate": 5.88423209700441e-05,
|
| 1443 |
+
"loss": 0.0199,
|
| 1444 |
+
"step": 2050
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 3.2802547770700636,
|
| 1448 |
+
"grad_norm": 0.24200858175754547,
|
| 1449 |
+
"learning_rate": 5.7884151557782305e-05,
|
| 1450 |
+
"loss": 0.0334,
|
| 1451 |
+
"step": 2060
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 3.2961783439490446,
|
| 1455 |
+
"grad_norm": 0.24898533523082733,
|
| 1456 |
+
"learning_rate": 5.693065968355888e-05,
|
| 1457 |
+
"loss": 0.0269,
|
| 1458 |
+
"step": 2070
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 3.3121019108280256,
|
| 1462 |
+
"grad_norm": 0.33028173446655273,
|
| 1463 |
+
"learning_rate": 5.598195124562049e-05,
|
| 1464 |
+
"loss": 0.0309,
|
| 1465 |
+
"step": 2080
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 3.328025477707006,
|
| 1469 |
+
"grad_norm": 0.32554179430007935,
|
| 1470 |
+
"learning_rate": 5.5038131610948326e-05,
|
| 1471 |
+
"loss": 0.0274,
|
| 1472 |
+
"step": 2090
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 3.343949044585987,
|
| 1476 |
+
"grad_norm": 0.25384682416915894,
|
| 1477 |
+
"learning_rate": 5.409930560355527e-05,
|
| 1478 |
+
"loss": 0.0235,
|
| 1479 |
+
"step": 2100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 3.359872611464968,
|
| 1483 |
+
"grad_norm": 0.3368367850780487,
|
| 1484 |
+
"learning_rate": 5.316557749284401e-05,
|
| 1485 |
+
"loss": 0.0259,
|
| 1486 |
+
"step": 2110
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 3.375796178343949,
|
| 1490 |
+
"grad_norm": 0.3751080334186554,
|
| 1491 |
+
"learning_rate": 5.223705098202647e-05,
|
| 1492 |
+
"loss": 0.0405,
|
| 1493 |
+
"step": 2120
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 3.3917197452229297,
|
| 1497 |
+
"grad_norm": 0.3862003684043884,
|
| 1498 |
+
"learning_rate": 5.131382919660609e-05,
|
| 1499 |
+
"loss": 0.0311,
|
| 1500 |
+
"step": 2130
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 3.4076433121019107,
|
| 1504 |
+
"grad_norm": 0.32009872794151306,
|
| 1505 |
+
"learning_rate": 5.03960146729244e-05,
|
| 1506 |
+
"loss": 0.0252,
|
| 1507 |
+
"step": 2140
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 3.4235668789808917,
|
| 1511 |
+
"grad_norm": 0.2071923166513443,
|
| 1512 |
+
"learning_rate": 4.9483709346772896e-05,
|
| 1513 |
+
"loss": 0.0248,
|
| 1514 |
+
"step": 2150
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 3.4394904458598727,
|
| 1518 |
+
"grad_norm": 0.3461016118526459,
|
| 1519 |
+
"learning_rate": 4.85770145420718e-05,
|
| 1520 |
+
"loss": 0.0336,
|
| 1521 |
+
"step": 2160
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 3.4554140127388537,
|
| 1525 |
+
"grad_norm": 0.3042682111263275,
|
| 1526 |
+
"learning_rate": 4.7676030959616526e-05,
|
| 1527 |
+
"loss": 0.0211,
|
| 1528 |
+
"step": 2170
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 3.4713375796178343,
|
| 1532 |
+
"grad_norm": 0.16671477258205414,
|
| 1533 |
+
"learning_rate": 4.678085866589348e-05,
|
| 1534 |
+
"loss": 0.018,
|
| 1535 |
+
"step": 2180
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 3.4872611464968153,
|
| 1539 |
+
"grad_norm": 0.20116400718688965,
|
| 1540 |
+
"learning_rate": 4.5891597081966583e-05,
|
| 1541 |
+
"loss": 0.0224,
|
| 1542 |
+
"step": 2190
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 3.5031847133757963,
|
| 1546 |
+
"grad_norm": 0.23196861147880554,
|
| 1547 |
+
"learning_rate": 4.500834497243478e-05,
|
| 1548 |
+
"loss": 0.0221,
|
| 1549 |
+
"step": 2200
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 3.519108280254777,
|
| 1553 |
+
"grad_norm": 0.2727397084236145,
|
| 1554 |
+
"learning_rate": 4.413120043446322e-05,
|
| 1555 |
+
"loss": 0.0272,
|
| 1556 |
+
"step": 2210
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 3.535031847133758,
|
| 1560 |
+
"grad_norm": 0.283627450466156,
|
| 1561 |
+
"learning_rate": 4.3260260886888206e-05,
|
| 1562 |
+
"loss": 0.0259,
|
| 1563 |
+
"step": 2220
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 3.550955414012739,
|
| 1567 |
+
"grad_norm": 0.40015652775764465,
|
| 1568 |
+
"learning_rate": 4.239562305939725e-05,
|
| 1569 |
+
"loss": 0.0298,
|
| 1570 |
+
"step": 2230
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 3.56687898089172,
|
| 1574 |
+
"grad_norm": 0.30534714460372925,
|
| 1575 |
+
"learning_rate": 4.153738298178623e-05,
|
| 1576 |
+
"loss": 0.0245,
|
| 1577 |
+
"step": 2240
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 3.582802547770701,
|
| 1581 |
+
"grad_norm": 0.22230559587478638,
|
| 1582 |
+
"learning_rate": 4.068563597329379e-05,
|
| 1583 |
+
"loss": 0.0231,
|
| 1584 |
+
"step": 2250
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 3.5987261146496814,
|
| 1588 |
+
"grad_norm": 0.3529864549636841,
|
| 1589 |
+
"learning_rate": 3.984047663201492e-05,
|
| 1590 |
+
"loss": 0.04,
|
| 1591 |
+
"step": 2260
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 3.6146496815286624,
|
| 1595 |
+
"grad_norm": 0.2945915460586548,
|
| 1596 |
+
"learning_rate": 3.9001998824394526e-05,
|
| 1597 |
+
"loss": 0.0251,
|
| 1598 |
+
"step": 2270
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 3.6305732484076434,
|
| 1602 |
+
"grad_norm": 0.24476458132266998,
|
| 1603 |
+
"learning_rate": 3.817029567480228e-05,
|
| 1604 |
+
"loss": 0.0245,
|
| 1605 |
+
"step": 2280
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 3.646496815286624,
|
| 1609 |
+
"grad_norm": 0.1646440178155899,
|
| 1610 |
+
"learning_rate": 3.7345459555189885e-05,
|
| 1611 |
+
"loss": 0.0246,
|
| 1612 |
+
"step": 2290
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 3.662420382165605,
|
| 1616 |
+
"grad_norm": 0.21083424985408783,
|
| 1617 |
+
"learning_rate": 3.652758207483197e-05,
|
| 1618 |
+
"loss": 0.0272,
|
| 1619 |
+
"step": 2300
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 3.678343949044586,
|
| 1623 |
+
"grad_norm": 0.17064619064331055,
|
| 1624 |
+
"learning_rate": 3.5716754070151524e-05,
|
| 1625 |
+
"loss": 0.0169,
|
| 1626 |
+
"step": 2310
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 3.694267515923567,
|
| 1630 |
+
"grad_norm": 0.18861030042171478,
|
| 1631 |
+
"learning_rate": 3.4913065594631434e-05,
|
| 1632 |
+
"loss": 0.0257,
|
| 1633 |
+
"step": 2320
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 3.710191082802548,
|
| 1637 |
+
"grad_norm": 0.31544387340545654,
|
| 1638 |
+
"learning_rate": 3.4116605908812595e-05,
|
| 1639 |
+
"loss": 0.0237,
|
| 1640 |
+
"step": 2330
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 3.7261146496815285,
|
| 1644 |
+
"grad_norm": 0.5185452103614807,
|
| 1645 |
+
"learning_rate": 3.3327463470380524e-05,
|
| 1646 |
+
"loss": 0.0229,
|
| 1647 |
+
"step": 2340
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 3.7420382165605095,
|
| 1651 |
+
"grad_norm": 0.2828991115093231,
|
| 1652 |
+
"learning_rate": 3.2545725924340876e-05,
|
| 1653 |
+
"loss": 0.0294,
|
| 1654 |
+
"step": 2350
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 3.7579617834394905,
|
| 1658 |
+
"grad_norm": 0.2944129407405853,
|
| 1659 |
+
"learning_rate": 3.1771480093285186e-05,
|
| 1660 |
+
"loss": 0.0218,
|
| 1661 |
+
"step": 2360
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 3.7738853503184715,
|
| 1665 |
+
"grad_norm": 0.24150818586349487,
|
| 1666 |
+
"learning_rate": 3.1004811967748134e-05,
|
| 1667 |
+
"loss": 0.024,
|
| 1668 |
+
"step": 2370
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 3.789808917197452,
|
| 1672 |
+
"grad_norm": 0.2700318396091461,
|
| 1673 |
+
"learning_rate": 3.024580669665711e-05,
|
| 1674 |
+
"loss": 0.0244,
|
| 1675 |
+
"step": 2380
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 3.805732484076433,
|
| 1679 |
+
"grad_norm": 0.4214876592159271,
|
| 1680 |
+
"learning_rate": 2.9494548577875192e-05,
|
| 1681 |
+
"loss": 0.0264,
|
| 1682 |
+
"step": 2390
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 3.821656050955414,
|
| 1686 |
+
"grad_norm": 0.2855050265789032,
|
| 1687 |
+
"learning_rate": 2.8751121048838836e-05,
|
| 1688 |
+
"loss": 0.02,
|
| 1689 |
+
"step": 2400
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 3.837579617834395,
|
| 1693 |
+
"grad_norm": 0.19058647751808167,
|
| 1694 |
+
"learning_rate": 2.801560667729093e-05,
|
| 1695 |
+
"loss": 0.0243,
|
| 1696 |
+
"step": 2410
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 3.853503184713376,
|
| 1700 |
+
"grad_norm": 0.32941293716430664,
|
| 1701 |
+
"learning_rate": 2.7288087152110565e-05,
|
| 1702 |
+
"loss": 0.0268,
|
| 1703 |
+
"step": 2420
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 3.8694267515923566,
|
| 1707 |
+
"grad_norm": 0.4154011905193329,
|
| 1708 |
+
"learning_rate": 2.656864327424027e-05,
|
| 1709 |
+
"loss": 0.0206,
|
| 1710 |
+
"step": 2430
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 3.8853503184713376,
|
| 1714 |
+
"grad_norm": 0.3098050057888031,
|
| 1715 |
+
"learning_rate": 2.585735494771222e-05,
|
| 1716 |
+
"loss": 0.023,
|
| 1717 |
+
"step": 2440
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 3.9012738853503186,
|
| 1721 |
+
"grad_norm": 0.27963143587112427,
|
| 1722 |
+
"learning_rate": 2.515430117077354e-05,
|
| 1723 |
+
"loss": 0.0273,
|
| 1724 |
+
"step": 2450
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 3.917197452229299,
|
| 1728 |
+
"grad_norm": 0.29374557733535767,
|
| 1729 |
+
"learning_rate": 2.4459560027112504e-05,
|
| 1730 |
+
"loss": 0.0244,
|
| 1731 |
+
"step": 2460
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 3.93312101910828,
|
| 1735 |
+
"grad_norm": 0.1925758719444275,
|
| 1736 |
+
"learning_rate": 2.377320867718654e-05,
|
| 1737 |
+
"loss": 0.0214,
|
| 1738 |
+
"step": 2470
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 3.949044585987261,
|
| 1742 |
+
"grad_norm": 0.3169631361961365,
|
| 1743 |
+
"learning_rate": 2.3095323349652153e-05,
|
| 1744 |
+
"loss": 0.0258,
|
| 1745 |
+
"step": 2480
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 3.964968152866242,
|
| 1749 |
+
"grad_norm": 0.2621577978134155,
|
| 1750 |
+
"learning_rate": 2.242597933289894e-05,
|
| 1751 |
+
"loss": 0.0238,
|
| 1752 |
+
"step": 2490
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 3.980891719745223,
|
| 1756 |
+
"grad_norm": 0.2187499850988388,
|
| 1757 |
+
"learning_rate": 2.176525096668769e-05,
|
| 1758 |
+
"loss": 0.0266,
|
| 1759 |
+
"step": 2500
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 3.9968152866242037,
|
| 1763 |
+
"grad_norm": 0.2935671806335449,
|
| 1764 |
+
"learning_rate": 2.1113211633894003e-05,
|
| 1765 |
+
"loss": 0.0214,
|
| 1766 |
+
"step": 2510
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 4.012738853503185,
|
| 1770 |
+
"grad_norm": 0.1593521386384964,
|
| 1771 |
+
"learning_rate": 2.046993375235804e-05,
|
| 1772 |
+
"loss": 0.0166,
|
| 1773 |
+
"step": 2520
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 4.028662420382165,
|
| 1777 |
+
"grad_norm": 0.32574114203453064,
|
| 1778 |
+
"learning_rate": 1.9835488766841615e-05,
|
| 1779 |
+
"loss": 0.0232,
|
| 1780 |
+
"step": 2530
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 4.044585987261146,
|
| 1784 |
+
"grad_norm": 0.2108948677778244,
|
| 1785 |
+
"learning_rate": 1.920994714109323e-05,
|
| 1786 |
+
"loss": 0.0255,
|
| 1787 |
+
"step": 2540
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 4.060509554140127,
|
| 1791 |
+
"grad_norm": 0.2054058164358139,
|
| 1792 |
+
"learning_rate": 1.8593378350022183e-05,
|
| 1793 |
+
"loss": 0.0156,
|
| 1794 |
+
"step": 2550
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 4.076433121019108,
|
| 1798 |
+
"grad_norm": 0.28134655952453613,
|
| 1799 |
+
"learning_rate": 1.798585087198228e-05,
|
| 1800 |
+
"loss": 0.0244,
|
| 1801 |
+
"step": 2560
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 4.092356687898089,
|
| 1805 |
+
"grad_norm": 0.26843348145484924,
|
| 1806 |
+
"learning_rate": 1.7387432181166642e-05,
|
| 1807 |
+
"loss": 0.0294,
|
| 1808 |
+
"step": 2570
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 4.10828025477707,
|
| 1812 |
+
"grad_norm": 0.32004109025001526,
|
| 1813 |
+
"learning_rate": 1.6798188740113484e-05,
|
| 1814 |
+
"loss": 0.0263,
|
| 1815 |
+
"step": 2580
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 4.124203821656051,
|
| 1819 |
+
"grad_norm": 0.6330613493919373,
|
| 1820 |
+
"learning_rate": 1.621818599232475e-05,
|
| 1821 |
+
"loss": 0.0178,
|
| 1822 |
+
"step": 2590
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 4.140127388535032,
|
| 1826 |
+
"grad_norm": 0.21530304849147797,
|
| 1827 |
+
"learning_rate": 1.564748835499773e-05,
|
| 1828 |
+
"loss": 0.0277,
|
| 1829 |
+
"step": 2600
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 4.156050955414012,
|
| 1833 |
+
"grad_norm": 0.20998333394527435,
|
| 1834 |
+
"learning_rate": 1.5086159211870442e-05,
|
| 1835 |
+
"loss": 0.0214,
|
| 1836 |
+
"step": 2610
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 4.171974522292993,
|
| 1840 |
+
"grad_norm": 0.3173789381980896,
|
| 1841 |
+
"learning_rate": 1.4534260906182228e-05,
|
| 1842 |
+
"loss": 0.0198,
|
| 1843 |
+
"step": 2620
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 4.187898089171974,
|
| 1847 |
+
"grad_norm": 0.18961144983768463,
|
| 1848 |
+
"learning_rate": 1.399185473374961e-05,
|
| 1849 |
+
"loss": 0.021,
|
| 1850 |
+
"step": 2630
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 4.203821656050955,
|
| 1854 |
+
"grad_norm": 0.36942291259765625,
|
| 1855 |
+
"learning_rate": 1.3459000936158528e-05,
|
| 1856 |
+
"loss": 0.0239,
|
| 1857 |
+
"step": 2640
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 4.219745222929936,
|
| 1861 |
+
"grad_norm": 0.28133276104927063,
|
| 1862 |
+
"learning_rate": 1.293575869407373e-05,
|
| 1863 |
+
"loss": 0.0244,
|
| 1864 |
+
"step": 2650
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 4.235668789808917,
|
| 1868 |
+
"grad_norm": 0.2719907760620117,
|
| 1869 |
+
"learning_rate": 1.2422186120665935e-05,
|
| 1870 |
+
"loss": 0.0203,
|
| 1871 |
+
"step": 2660
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 4.251592356687898,
|
| 1875 |
+
"grad_norm": 0.15558955073356628,
|
| 1876 |
+
"learning_rate": 1.1918340255157612e-05,
|
| 1877 |
+
"loss": 0.0214,
|
| 1878 |
+
"step": 2670
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 4.267515923566879,
|
| 1882 |
+
"grad_norm": 0.23012204468250275,
|
| 1883 |
+
"learning_rate": 1.1424277056487864e-05,
|
| 1884 |
+
"loss": 0.0232,
|
| 1885 |
+
"step": 2680
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 4.2834394904458595,
|
| 1889 |
+
"grad_norm": 0.29646140336990356,
|
| 1890 |
+
"learning_rate": 1.0940051397097673e-05,
|
| 1891 |
+
"loss": 0.0179,
|
| 1892 |
+
"step": 2690
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 4.2993630573248405,
|
| 1896 |
+
"grad_norm": 0.282185435295105,
|
| 1897 |
+
"learning_rate": 1.0465717056835355e-05,
|
| 1898 |
+
"loss": 0.0171,
|
| 1899 |
+
"step": 2700
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 4.3152866242038215,
|
| 1903 |
+
"grad_norm": 0.3244917392730713,
|
| 1904 |
+
"learning_rate": 1.0001326716983584e-05,
|
| 1905 |
+
"loss": 0.0161,
|
| 1906 |
+
"step": 2710
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 4.3312101910828025,
|
| 1910 |
+
"grad_norm": 0.2768123149871826,
|
| 1911 |
+
"learning_rate": 9.546931954408622e-06,
|
| 1912 |
+
"loss": 0.0248,
|
| 1913 |
+
"step": 2720
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 4.3471337579617835,
|
| 1917 |
+
"grad_norm": 0.14225247502326965,
|
| 1918 |
+
"learning_rate": 9.10258323583173e-06,
|
| 1919 |
+
"loss": 0.0171,
|
| 1920 |
+
"step": 2730
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 4.3630573248407645,
|
| 1924 |
+
"grad_norm": 0.16188742220401764,
|
| 1925 |
+
"learning_rate": 8.668329912224337e-06,
|
| 1926 |
+
"loss": 0.0177,
|
| 1927 |
+
"step": 2740
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 4.3789808917197455,
|
| 1931 |
+
"grad_norm": 0.30127179622650146,
|
| 1932 |
+
"learning_rate": 8.244220213326958e-06,
|
| 1933 |
+
"loss": 0.0213,
|
| 1934 |
+
"step": 2750
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 4.3949044585987265,
|
| 1938 |
+
"grad_norm": 0.1804376244544983,
|
| 1939 |
+
"learning_rate": 7.83030124229246e-06,
|
| 1940 |
+
"loss": 0.0232,
|
| 1941 |
+
"step": 2760
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 4.4108280254777075,
|
| 1945 |
+
"grad_norm": 0.24196097254753113,
|
| 1946 |
+
"learning_rate": 7.426618970454757e-06,
|
| 1947 |
+
"loss": 0.0163,
|
| 1948 |
+
"step": 2770
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 4.426751592356688,
|
| 1952 |
+
"grad_norm": 0.4308320879936218,
|
| 1953 |
+
"learning_rate": 7.033218232223027e-06,
|
| 1954 |
+
"loss": 0.0165,
|
| 1955 |
+
"step": 2780
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 4.442675159235669,
|
| 1959 |
+
"grad_norm": 0.3199467062950134,
|
| 1960 |
+
"learning_rate": 6.650142720102248e-06,
|
| 1961 |
+
"loss": 0.024,
|
| 1962 |
+
"step": 2790
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 4.45859872611465,
|
| 1966 |
+
"grad_norm": 0.2603253722190857,
|
| 1967 |
+
"learning_rate": 6.2774349798405465e-06,
|
| 1968 |
+
"loss": 0.0232,
|
| 1969 |
+
"step": 2800
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 4.474522292993631,
|
| 1973 |
+
"grad_norm": 0.20752651989459991,
|
| 1974 |
+
"learning_rate": 5.915136405703858e-06,
|
| 1975 |
+
"loss": 0.0223,
|
| 1976 |
+
"step": 2810
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 4.490445859872612,
|
| 1980 |
+
"grad_norm": 0.17525236308574677,
|
| 1981 |
+
"learning_rate": 5.563287235878678e-06,
|
| 1982 |
+
"loss": 0.0168,
|
| 1983 |
+
"step": 2820
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 4.506369426751593,
|
| 1987 |
+
"grad_norm": 0.20458337664604187,
|
| 1988 |
+
"learning_rate": 5.221926548002876e-06,
|
| 1989 |
+
"loss": 0.0208,
|
| 1990 |
+
"step": 2830
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 4.522292993630574,
|
| 1994 |
+
"grad_norm": 0.353657066822052,
|
| 1995 |
+
"learning_rate": 4.891092254825725e-06,
|
| 1996 |
+
"loss": 0.0224,
|
| 1997 |
+
"step": 2840
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 4.538216560509554,
|
| 2001 |
+
"grad_norm": 0.23591625690460205,
|
| 2002 |
+
"learning_rate": 4.570821099997169e-06,
|
| 2003 |
+
"loss": 0.0186,
|
| 2004 |
+
"step": 2850
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 4.554140127388535,
|
| 2008 |
+
"grad_norm": 0.2248276323080063,
|
| 2009 |
+
"learning_rate": 4.261148653986824e-06,
|
| 2010 |
+
"loss": 0.0201,
|
| 2011 |
+
"step": 2860
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 4.570063694267516,
|
| 2015 |
+
"grad_norm": 0.19413244724273682,
|
| 2016 |
+
"learning_rate": 3.962109310133544e-06,
|
| 2017 |
+
"loss": 0.0209,
|
| 2018 |
+
"step": 2870
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 4.585987261146497,
|
| 2022 |
+
"grad_norm": 0.1779373288154602,
|
| 2023 |
+
"learning_rate": 3.6737362808254793e-06,
|
| 2024 |
+
"loss": 0.0164,
|
| 2025 |
+
"step": 2880
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 4.601910828025478,
|
| 2029 |
+
"grad_norm": 0.13120533525943756,
|
| 2030 |
+
"learning_rate": 3.3960615938114172e-06,
|
| 2031 |
+
"loss": 0.02,
|
| 2032 |
+
"step": 2890
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 4.617834394904459,
|
| 2036 |
+
"grad_norm": 0.24661269783973694,
|
| 2037 |
+
"learning_rate": 3.1291160886436575e-06,
|
| 2038 |
+
"loss": 0.0199,
|
| 2039 |
+
"step": 2900
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 4.63375796178344,
|
| 2043 |
+
"grad_norm": 0.20324034988880157,
|
| 2044 |
+
"learning_rate": 2.872929413252867e-06,
|
| 2045 |
+
"loss": 0.0168,
|
| 2046 |
+
"step": 2910
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 4.649681528662421,
|
| 2050 |
+
"grad_norm": 0.1891164630651474,
|
| 2051 |
+
"learning_rate": 2.6275300206552667e-06,
|
| 2052 |
+
"loss": 0.0239,
|
| 2053 |
+
"step": 2920
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 4.665605095541402,
|
| 2057 |
+
"grad_norm": 0.446866512298584,
|
| 2058 |
+
"learning_rate": 2.3929451657925482e-06,
|
| 2059 |
+
"loss": 0.0215,
|
| 2060 |
+
"step": 2930
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 4.681528662420382,
|
| 2064 |
+
"grad_norm": 0.1874268352985382,
|
| 2065 |
+
"learning_rate": 2.1692009025048422e-06,
|
| 2066 |
+
"loss": 0.0194,
|
| 2067 |
+
"step": 2940
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 4.697452229299363,
|
| 2071 |
+
"grad_norm": 0.3293200135231018,
|
| 2072 |
+
"learning_rate": 1.956322080637052e-06,
|
| 2073 |
+
"loss": 0.0245,
|
| 2074 |
+
"step": 2950
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 4.713375796178344,
|
| 2078 |
+
"grad_norm": 0.21030674874782562,
|
| 2079 |
+
"learning_rate": 1.754332343278986e-06,
|
| 2080 |
+
"loss": 0.0218,
|
| 2081 |
+
"step": 2960
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 4.729299363057325,
|
| 2085 |
+
"grad_norm": 0.1421198844909668,
|
| 2086 |
+
"learning_rate": 1.5632541241394461e-06,
|
| 2087 |
+
"loss": 0.017,
|
| 2088 |
+
"step": 2970
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 4.745222929936306,
|
| 2092 |
+
"grad_norm": 0.30587485432624817,
|
| 2093 |
+
"learning_rate": 1.3831086450546648e-06,
|
| 2094 |
+
"loss": 0.0165,
|
| 2095 |
+
"step": 2980
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 4.761146496815287,
|
| 2099 |
+
"grad_norm": 0.28812679648399353,
|
| 2100 |
+
"learning_rate": 1.213915913631314e-06,
|
| 2101 |
+
"loss": 0.0221,
|
| 2102 |
+
"step": 2990
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 4.777070063694268,
|
| 2106 |
+
"grad_norm": 0.20168578624725342,
|
| 2107 |
+
"learning_rate": 1.055694721024425e-06,
|
| 2108 |
+
"loss": 0.0197,
|
| 2109 |
+
"step": 3000
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 4.792993630573249,
|
| 2113 |
+
"grad_norm": 0.31528469920158386,
|
| 2114 |
+
"learning_rate": 9.084626398503604e-07,
|
| 2115 |
+
"loss": 0.0278,
|
| 2116 |
+
"step": 3010
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 4.80891719745223,
|
| 2120 |
+
"grad_norm": 0.2268596887588501,
|
| 2121 |
+
"learning_rate": 7.722360222350955e-07,
|
| 2122 |
+
"loss": 0.0196,
|
| 2123 |
+
"step": 3020
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 4.82484076433121,
|
| 2127 |
+
"grad_norm": 0.21379084885120392,
|
| 2128 |
+
"learning_rate": 6.470299979981609e-07,
|
| 2129 |
+
"loss": 0.0242,
|
| 2130 |
+
"step": 3030
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 4.840764331210191,
|
| 2134 |
+
"grad_norm": 0.4038665294647217,
|
| 2135 |
+
"learning_rate": 5.328584729722197e-07,
|
| 2136 |
+
"loss": 0.0208,
|
| 2137 |
+
"step": 3040
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 4.856687898089172,
|
| 2141 |
+
"grad_norm": 0.1791532039642334,
|
| 2142 |
+
"learning_rate": 4.297341274586475e-07,
|
| 2143 |
+
"loss": 0.0248,
|
| 2144 |
+
"step": 3050
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 4.872611464968153,
|
| 2148 |
+
"grad_norm": 0.2981935739517212,
|
| 2149 |
+
"learning_rate": 3.376684148192255e-07,
|
| 2150 |
+
"loss": 0.0166,
|
| 2151 |
+
"step": 3060
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 4.888535031847134,
|
| 2155 |
+
"grad_norm": 0.37263256311416626,
|
| 2156 |
+
"learning_rate": 2.5667156020408034e-07,
|
| 2157 |
+
"loss": 0.0161,
|
| 2158 |
+
"step": 3070
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 4.904458598726115,
|
| 2162 |
+
"grad_norm": 0.19920985400676727,
|
| 2163 |
+
"learning_rate": 1.8675255941604797e-07,
|
| 2164 |
+
"loss": 0.0296,
|
| 2165 |
+
"step": 3080
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 4.920382165605096,
|
| 2169 |
+
"grad_norm": 0.17917926609516144,
|
| 2170 |
+
"learning_rate": 1.2791917791152853e-07,
|
| 2171 |
+
"loss": 0.024,
|
| 2172 |
+
"step": 3090
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 4.936305732484076,
|
| 2176 |
+
"grad_norm": 0.2005498707294464,
|
| 2177 |
+
"learning_rate": 8.017794993808725e-08,
|
| 2178 |
+
"loss": 0.0223,
|
| 2179 |
+
"step": 3100
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 4.952229299363057,
|
| 2183 |
+
"grad_norm": 0.23752698302268982,
|
| 2184 |
+
"learning_rate": 4.353417780872393e-08,
|
| 2185 |
+
"loss": 0.0199,
|
| 2186 |
+
"step": 3110
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 4.968152866242038,
|
| 2190 |
+
"grad_norm": 0.20469646155834198,
|
| 2191 |
+
"learning_rate": 1.7991931312921762e-08,
|
| 2192 |
+
"loss": 0.021,
|
| 2193 |
+
"step": 3120
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 4.984076433121019,
|
| 2197 |
+
"grad_norm": 0.20618374645709991,
|
| 2198 |
+
"learning_rate": 3.5540472647088707e-09,
|
| 2199 |
+
"loss": 0.0183,
|
| 2200 |
+
"step": 3130
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 4.996815286624204,
|
| 2204 |
+
"step": 3138,
|
| 2205 |
+
"total_flos": 1.1375719484380992e+17,
|
| 2206 |
+
"train_loss": 0.04994222255384489,
|
| 2207 |
+
"train_runtime": 1406.117,
|
| 2208 |
+
"train_samples_per_second": 35.707,
|
| 2209 |
+
"train_steps_per_second": 2.232
|
| 2210 |
+
}
|
| 2211 |
+
],
|
| 2212 |
+
"logging_steps": 10,
|
| 2213 |
+
"max_steps": 3138,
|
| 2214 |
+
"num_input_tokens_seen": 0,
|
| 2215 |
+
"num_train_epochs": 5,
|
| 2216 |
+
"save_steps": 500,
|
| 2217 |
+
"stateful_callbacks": {
|
| 2218 |
+
"TrainerControl": {
|
| 2219 |
+
"args": {
|
| 2220 |
+
"should_epoch_stop": false,
|
| 2221 |
+
"should_evaluate": false,
|
| 2222 |
+
"should_log": false,
|
| 2223 |
+
"should_save": true,
|
| 2224 |
+
"should_training_stop": true
|
| 2225 |
+
},
|
| 2226 |
+
"attributes": {}
|
| 2227 |
+
}
|
| 2228 |
+
},
|
| 2229 |
+
"total_flos": 1.1375719484380992e+17,
|
| 2230 |
+
"train_batch_size": 16,
|
| 2231 |
+
"trial_name": null,
|
| 2232 |
+
"trial_params": null
|
| 2233 |
+
}
|