Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +2037 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,2037 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.994746059544658,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2852,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.017513134851138354,
|
| 13 |
+
"grad_norm": 5.451874256134033,
|
| 14 |
+
"learning_rate": 1.3986013986013988e-05,
|
| 15 |
+
"loss": 1.0538,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.03502626970227671,
|
| 20 |
+
"grad_norm": 4.681659698486328,
|
| 21 |
+
"learning_rate": 2.7972027972027976e-05,
|
| 22 |
+
"loss": 0.7104,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.05253940455341506,
|
| 27 |
+
"grad_norm": 2.5610225200653076,
|
| 28 |
+
"learning_rate": 4.195804195804196e-05,
|
| 29 |
+
"loss": 0.4212,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.07005253940455342,
|
| 34 |
+
"grad_norm": 1.4815014600753784,
|
| 35 |
+
"learning_rate": 5.594405594405595e-05,
|
| 36 |
+
"loss": 0.3188,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.08756567425569177,
|
| 41 |
+
"grad_norm": 2.2192041873931885,
|
| 42 |
+
"learning_rate": 6.993006993006993e-05,
|
| 43 |
+
"loss": 0.2665,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.10507880910683012,
|
| 48 |
+
"grad_norm": 2.0386500358581543,
|
| 49 |
+
"learning_rate": 8.391608391608392e-05,
|
| 50 |
+
"loss": 0.2553,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.12259194395796848,
|
| 55 |
+
"grad_norm": 0.8188023567199707,
|
| 56 |
+
"learning_rate": 9.790209790209791e-05,
|
| 57 |
+
"loss": 0.2096,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.14010507880910683,
|
| 62 |
+
"grad_norm": 1.2554601430892944,
|
| 63 |
+
"learning_rate": 0.0001118881118881119,
|
| 64 |
+
"loss": 0.1705,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.15761821366024517,
|
| 69 |
+
"grad_norm": 0.7775352001190186,
|
| 70 |
+
"learning_rate": 0.00012587412587412587,
|
| 71 |
+
"loss": 0.1453,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.17513134851138354,
|
| 76 |
+
"grad_norm": 1.060951590538025,
|
| 77 |
+
"learning_rate": 0.00013986013986013986,
|
| 78 |
+
"loss": 0.1653,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.19264448336252188,
|
| 83 |
+
"grad_norm": 0.9171435832977295,
|
| 84 |
+
"learning_rate": 0.00015384615384615385,
|
| 85 |
+
"loss": 0.1544,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.21015761821366025,
|
| 90 |
+
"grad_norm": 1.313083529472351,
|
| 91 |
+
"learning_rate": 0.00016783216783216784,
|
| 92 |
+
"loss": 0.1561,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.2276707530647986,
|
| 97 |
+
"grad_norm": 0.8744928240776062,
|
| 98 |
+
"learning_rate": 0.00018181818181818183,
|
| 99 |
+
"loss": 0.1452,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.24518388791593695,
|
| 104 |
+
"grad_norm": 1.3142743110656738,
|
| 105 |
+
"learning_rate": 0.00019580419580419583,
|
| 106 |
+
"loss": 0.1305,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.2626970227670753,
|
| 111 |
+
"grad_norm": 0.730299174785614,
|
| 112 |
+
"learning_rate": 0.00019999670507574947,
|
| 113 |
+
"loss": 0.1446,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.28021015761821366,
|
| 118 |
+
"grad_norm": 0.5380399227142334,
|
| 119 |
+
"learning_rate": 0.00019998056719395973,
|
| 120 |
+
"loss": 0.1348,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.29772329246935203,
|
| 125 |
+
"grad_norm": 0.6892985701560974,
|
| 126 |
+
"learning_rate": 0.00019995098333206742,
|
| 127 |
+
"loss": 0.1136,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.31523642732049034,
|
| 132 |
+
"grad_norm": 1.1214607954025269,
|
| 133 |
+
"learning_rate": 0.00019990795746868583,
|
| 134 |
+
"loss": 0.1014,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.3327495621716287,
|
| 139 |
+
"grad_norm": 0.5364326238632202,
|
| 140 |
+
"learning_rate": 0.00019985149539018855,
|
| 141 |
+
"loss": 0.108,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.3502626970227671,
|
| 146 |
+
"grad_norm": 0.5024985074996948,
|
| 147 |
+
"learning_rate": 0.00019978160468993094,
|
| 148 |
+
"loss": 0.1158,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.36777583187390545,
|
| 153 |
+
"grad_norm": 0.662004828453064,
|
| 154 |
+
"learning_rate": 0.00019969829476722923,
|
| 155 |
+
"loss": 0.1024,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.38528896672504376,
|
| 160 |
+
"grad_norm": 0.702117919921875,
|
| 161 |
+
"learning_rate": 0.00019960157682609632,
|
| 162 |
+
"loss": 0.108,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.4028021015761821,
|
| 167 |
+
"grad_norm": 0.577064573764801,
|
| 168 |
+
"learning_rate": 0.00019949146387373493,
|
| 169 |
+
"loss": 0.0969,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.4203152364273205,
|
| 174 |
+
"grad_norm": 0.4810011684894562,
|
| 175 |
+
"learning_rate": 0.00019936797071878854,
|
| 176 |
+
"loss": 0.1059,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.43782837127845886,
|
| 181 |
+
"grad_norm": 0.747789740562439,
|
| 182 |
+
"learning_rate": 0.00019923111396934957,
|
| 183 |
+
"loss": 0.1006,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.4553415061295972,
|
| 188 |
+
"grad_norm": 0.5298845767974854,
|
| 189 |
+
"learning_rate": 0.00019908091203072598,
|
| 190 |
+
"loss": 0.1066,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.47285464098073554,
|
| 195 |
+
"grad_norm": 0.4503520429134369,
|
| 196 |
+
"learning_rate": 0.00019891738510296602,
|
| 197 |
+
"loss": 0.0928,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.4903677758318739,
|
| 202 |
+
"grad_norm": 0.47492510080337524,
|
| 203 |
+
"learning_rate": 0.0001987405551781415,
|
| 204 |
+
"loss": 0.1035,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.5078809106830122,
|
| 209 |
+
"grad_norm": 0.435342013835907,
|
| 210 |
+
"learning_rate": 0.0001985504460373903,
|
| 211 |
+
"loss": 0.0784,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.5253940455341506,
|
| 216 |
+
"grad_norm": 0.3728376626968384,
|
| 217 |
+
"learning_rate": 0.00019834708324771797,
|
| 218 |
+
"loss": 0.0858,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.542907180385289,
|
| 223 |
+
"grad_norm": 0.7249437570571899,
|
| 224 |
+
"learning_rate": 0.00019813049415855964,
|
| 225 |
+
"loss": 0.0828,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.5604203152364273,
|
| 230 |
+
"grad_norm": 0.5893632769584656,
|
| 231 |
+
"learning_rate": 0.00019790070789810145,
|
| 232 |
+
"loss": 0.0833,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.5779334500875657,
|
| 237 |
+
"grad_norm": 0.5527558326721191,
|
| 238 |
+
"learning_rate": 0.00019765775536936367,
|
| 239 |
+
"loss": 0.0841,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.5954465849387041,
|
| 244 |
+
"grad_norm": 0.5419805645942688,
|
| 245 |
+
"learning_rate": 0.00019740166924604431,
|
| 246 |
+
"loss": 0.0791,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.6129597197898424,
|
| 251 |
+
"grad_norm": 0.6050059795379639,
|
| 252 |
+
"learning_rate": 0.00019713248396812524,
|
| 253 |
+
"loss": 0.0821,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.6304728546409807,
|
| 258 |
+
"grad_norm": 0.6528359055519104,
|
| 259 |
+
"learning_rate": 0.00019685023573724037,
|
| 260 |
+
"loss": 0.08,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.647985989492119,
|
| 265 |
+
"grad_norm": 0.5806655287742615,
|
| 266 |
+
"learning_rate": 0.0001965549625118071,
|
| 267 |
+
"loss": 0.0683,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.6654991243432574,
|
| 272 |
+
"grad_norm": 0.3972892761230469,
|
| 273 |
+
"learning_rate": 0.00019624670400192126,
|
| 274 |
+
"loss": 0.0744,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.6830122591943958,
|
| 279 |
+
"grad_norm": 0.5557606220245361,
|
| 280 |
+
"learning_rate": 0.00019592550166401695,
|
| 281 |
+
"loss": 0.089,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.7005253940455342,
|
| 286 |
+
"grad_norm": 0.4300699830055237,
|
| 287 |
+
"learning_rate": 0.00019559139869529103,
|
| 288 |
+
"loss": 0.0694,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.7180385288966725,
|
| 293 |
+
"grad_norm": 0.39790406823158264,
|
| 294 |
+
"learning_rate": 0.00019524444002789383,
|
| 295 |
+
"loss": 0.0835,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.7355516637478109,
|
| 300 |
+
"grad_norm": 0.458674818277359,
|
| 301 |
+
"learning_rate": 0.0001948846723228862,
|
| 302 |
+
"loss": 0.0963,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.7530647985989493,
|
| 307 |
+
"grad_norm": 0.4444922208786011,
|
| 308 |
+
"learning_rate": 0.00019451214396396454,
|
| 309 |
+
"loss": 0.0829,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.7705779334500875,
|
| 314 |
+
"grad_norm": 0.409694641828537,
|
| 315 |
+
"learning_rate": 0.00019412690505095365,
|
| 316 |
+
"loss": 0.0789,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.7880910683012259,
|
| 321 |
+
"grad_norm": 0.5321265459060669,
|
| 322 |
+
"learning_rate": 0.00019372900739306908,
|
| 323 |
+
"loss": 0.0748,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.8056042031523643,
|
| 328 |
+
"grad_norm": 0.4633730351924896,
|
| 329 |
+
"learning_rate": 0.00019331850450194957,
|
| 330 |
+
"loss": 0.0703,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.8231173380035026,
|
| 335 |
+
"grad_norm": 0.4727855920791626,
|
| 336 |
+
"learning_rate": 0.00019289545158446045,
|
| 337 |
+
"loss": 0.0794,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.840630472854641,
|
| 342 |
+
"grad_norm": 0.40540841221809387,
|
| 343 |
+
"learning_rate": 0.00019245990553526905,
|
| 344 |
+
"loss": 0.0677,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.8581436077057794,
|
| 349 |
+
"grad_norm": 0.3436204195022583,
|
| 350 |
+
"learning_rate": 0.00019201192492919317,
|
| 351 |
+
"loss": 0.0742,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.8756567425569177,
|
| 356 |
+
"grad_norm": 0.3329843580722809,
|
| 357 |
+
"learning_rate": 0.00019155157001332374,
|
| 358 |
+
"loss": 0.0539,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.8931698774080561,
|
| 363 |
+
"grad_norm": 0.4325779974460602,
|
| 364 |
+
"learning_rate": 0.00019107890269892214,
|
| 365 |
+
"loss": 0.0662,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.9106830122591943,
|
| 370 |
+
"grad_norm": 0.23809094727039337,
|
| 371 |
+
"learning_rate": 0.0001905939865530944,
|
| 372 |
+
"loss": 0.0526,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.9281961471103327,
|
| 377 |
+
"grad_norm": 0.466955304145813,
|
| 378 |
+
"learning_rate": 0.0001900968867902419,
|
| 379 |
+
"loss": 0.073,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.9457092819614711,
|
| 384 |
+
"grad_norm": 0.36457186937332153,
|
| 385 |
+
"learning_rate": 0.0001895876702632913,
|
| 386 |
+
"loss": 0.0776,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.9632224168126094,
|
| 391 |
+
"grad_norm": 0.47820454835891724,
|
| 392 |
+
"learning_rate": 0.00018906640545470355,
|
| 393 |
+
"loss": 0.0637,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.9807355516637478,
|
| 398 |
+
"grad_norm": 0.466001033782959,
|
| 399 |
+
"learning_rate": 0.000188533162467264,
|
| 400 |
+
"loss": 0.0678,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.9982486865148862,
|
| 405 |
+
"grad_norm": 0.5864549875259399,
|
| 406 |
+
"learning_rate": 0.0001879880130146547,
|
| 407 |
+
"loss": 0.0782,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.0157618213660244,
|
| 412 |
+
"grad_norm": 0.37066832184791565,
|
| 413 |
+
"learning_rate": 0.0001874310304118096,
|
| 414 |
+
"loss": 0.0631,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.0332749562171628,
|
| 419 |
+
"grad_norm": 0.481014221906662,
|
| 420 |
+
"learning_rate": 0.00018686228956505516,
|
| 421 |
+
"loss": 0.0626,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.0507880910683012,
|
| 426 |
+
"grad_norm": 0.5026444792747498,
|
| 427 |
+
"learning_rate": 0.00018628186696203612,
|
| 428 |
+
"loss": 0.0721,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.0683012259194395,
|
| 433 |
+
"grad_norm": 0.5261244177818298,
|
| 434 |
+
"learning_rate": 0.00018568984066142917,
|
| 435 |
+
"loss": 0.0735,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.085814360770578,
|
| 440 |
+
"grad_norm": 0.5333762168884277,
|
| 441 |
+
"learning_rate": 0.00018508629028244519,
|
| 442 |
+
"loss": 0.0807,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.1033274956217163,
|
| 447 |
+
"grad_norm": 0.46089333295822144,
|
| 448 |
+
"learning_rate": 0.00018447129699412142,
|
| 449 |
+
"loss": 0.0605,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.1208406304728546,
|
| 454 |
+
"grad_norm": 0.35585829615592957,
|
| 455 |
+
"learning_rate": 0.00018384494350440553,
|
| 456 |
+
"loss": 0.0662,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.138353765323993,
|
| 461 |
+
"grad_norm": 0.3352622091770172,
|
| 462 |
+
"learning_rate": 0.0001832073140490325,
|
| 463 |
+
"loss": 0.0628,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.1558669001751314,
|
| 468 |
+
"grad_norm": 0.4361122250556946,
|
| 469 |
+
"learning_rate": 0.00018255849438019608,
|
| 470 |
+
"loss": 0.0593,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 1.1733800350262698,
|
| 475 |
+
"grad_norm": 0.409807026386261,
|
| 476 |
+
"learning_rate": 0.00018189857175501635,
|
| 477 |
+
"loss": 0.0563,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 1.1908931698774081,
|
| 482 |
+
"grad_norm": 0.5780562162399292,
|
| 483 |
+
"learning_rate": 0.00018122763492380486,
|
| 484 |
+
"loss": 0.074,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.2084063047285465,
|
| 489 |
+
"grad_norm": 0.3598422110080719,
|
| 490 |
+
"learning_rate": 0.00018054577411812895,
|
| 491 |
+
"loss": 0.0521,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 1.2259194395796849,
|
| 496 |
+
"grad_norm": 0.31493690609931946,
|
| 497 |
+
"learning_rate": 0.00017985308103867688,
|
| 498 |
+
"loss": 0.0486,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 1.2434325744308232,
|
| 503 |
+
"grad_norm": 0.29808786511421204,
|
| 504 |
+
"learning_rate": 0.00017914964884292544,
|
| 505 |
+
"loss": 0.0607,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.2609457092819616,
|
| 510 |
+
"grad_norm": 0.36241838335990906,
|
| 511 |
+
"learning_rate": 0.00017843557213261142,
|
| 512 |
+
"loss": 0.058,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 1.2784588441331,
|
| 517 |
+
"grad_norm": 0.2744923532009125,
|
| 518 |
+
"learning_rate": 0.00017771094694100925,
|
| 519 |
+
"loss": 0.0679,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.295971978984238,
|
| 524 |
+
"grad_norm": 0.38270923495292664,
|
| 525 |
+
"learning_rate": 0.00017697587072001557,
|
| 526 |
+
"loss": 0.0612,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.3134851138353765,
|
| 531 |
+
"grad_norm": 0.34466174244880676,
|
| 532 |
+
"learning_rate": 0.0001762304423270436,
|
| 533 |
+
"loss": 0.0678,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 1.3309982486865148,
|
| 538 |
+
"grad_norm": 0.3777899146080017,
|
| 539 |
+
"learning_rate": 0.00017547476201172808,
|
| 540 |
+
"loss": 0.0532,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 1.3485113835376532,
|
| 545 |
+
"grad_norm": 0.3167668879032135,
|
| 546 |
+
"learning_rate": 0.00017470893140244303,
|
| 547 |
+
"loss": 0.0546,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.3660245183887916,
|
| 552 |
+
"grad_norm": 0.3986264765262604,
|
| 553 |
+
"learning_rate": 0.00017393305349263434,
|
| 554 |
+
"loss": 0.0637,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 1.38353765323993,
|
| 559 |
+
"grad_norm": 0.8364148139953613,
|
| 560 |
+
"learning_rate": 0.00017314723262696848,
|
| 561 |
+
"loss": 0.0535,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 1.4010507880910683,
|
| 566 |
+
"grad_norm": 0.5576715469360352,
|
| 567 |
+
"learning_rate": 0.00017235157448729967,
|
| 568 |
+
"loss": 0.0652,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.4185639229422067,
|
| 573 |
+
"grad_norm": 0.3843351900577545,
|
| 574 |
+
"learning_rate": 0.00017154618607845702,
|
| 575 |
+
"loss": 0.0492,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 1.436077057793345,
|
| 580 |
+
"grad_norm": 0.3571943938732147,
|
| 581 |
+
"learning_rate": 0.00017073117571385414,
|
| 582 |
+
"loss": 0.0635,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 1.4535901926444834,
|
| 587 |
+
"grad_norm": 0.5829870104789734,
|
| 588 |
+
"learning_rate": 0.00016990665300092224,
|
| 589 |
+
"loss": 0.0676,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.4711033274956218,
|
| 594 |
+
"grad_norm": 0.6512800455093384,
|
| 595 |
+
"learning_rate": 0.00016907272882636968,
|
| 596 |
+
"loss": 0.0607,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 1.4886164623467601,
|
| 601 |
+
"grad_norm": 0.32954949140548706,
|
| 602 |
+
"learning_rate": 0.0001682295153412691,
|
| 603 |
+
"loss": 0.0601,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.5061295971978983,
|
| 608 |
+
"grad_norm": 0.4212743639945984,
|
| 609 |
+
"learning_rate": 0.00016737712594597483,
|
| 610 |
+
"loss": 0.0637,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 1.5236427320490367,
|
| 615 |
+
"grad_norm": 0.29383569955825806,
|
| 616 |
+
"learning_rate": 0.00016651567527487204,
|
| 617 |
+
"loss": 0.0513,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 1.541155866900175,
|
| 622 |
+
"grad_norm": 0.37525856494903564,
|
| 623 |
+
"learning_rate": 0.00016564527918096005,
|
| 624 |
+
"loss": 0.0448,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 1.5586690017513134,
|
| 629 |
+
"grad_norm": 0.3859533667564392,
|
| 630 |
+
"learning_rate": 0.00016476605472027172,
|
| 631 |
+
"loss": 0.0548,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.5761821366024518,
|
| 636 |
+
"grad_norm": 0.2619127929210663,
|
| 637 |
+
"learning_rate": 0.00016387812013613103,
|
| 638 |
+
"loss": 0.0461,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 1.5936952714535901,
|
| 643 |
+
"grad_norm": 0.36077287793159485,
|
| 644 |
+
"learning_rate": 0.00016298159484325118,
|
| 645 |
+
"loss": 0.0524,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 1.6112084063047285,
|
| 650 |
+
"grad_norm": 0.3560950458049774,
|
| 651 |
+
"learning_rate": 0.00016207659941167485,
|
| 652 |
+
"loss": 0.0571,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 1.6287215411558669,
|
| 657 |
+
"grad_norm": 0.512663722038269,
|
| 658 |
+
"learning_rate": 0.00016116325555055915,
|
| 659 |
+
"loss": 0.07,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 1.6462346760070052,
|
| 664 |
+
"grad_norm": 0.37717878818511963,
|
| 665 |
+
"learning_rate": 0.00016024168609180757,
|
| 666 |
+
"loss": 0.0464,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 1.6637478108581436,
|
| 671 |
+
"grad_norm": 0.43427279591560364,
|
| 672 |
+
"learning_rate": 0.00015931201497355088,
|
| 673 |
+
"loss": 0.054,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.681260945709282,
|
| 678 |
+
"grad_norm": 0.7263680100440979,
|
| 679 |
+
"learning_rate": 0.000158374367223479,
|
| 680 |
+
"loss": 0.0675,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 1.6987740805604203,
|
| 685 |
+
"grad_norm": 0.49765631556510925,
|
| 686 |
+
"learning_rate": 0.00015742886894202674,
|
| 687 |
+
"loss": 0.0494,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 1.7162872154115587,
|
| 692 |
+
"grad_norm": 0.29158496856689453,
|
| 693 |
+
"learning_rate": 0.00015647564728541485,
|
| 694 |
+
"loss": 0.0614,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 1.733800350262697,
|
| 699 |
+
"grad_norm": 0.346938818693161,
|
| 700 |
+
"learning_rate": 0.00015551483044854954,
|
| 701 |
+
"loss": 0.0457,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 1.7513134851138354,
|
| 706 |
+
"grad_norm": 0.2457229495048523,
|
| 707 |
+
"learning_rate": 0.00015454654764778187,
|
| 708 |
+
"loss": 0.0491,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 1.7688266199649738,
|
| 713 |
+
"grad_norm": 0.2546524703502655,
|
| 714 |
+
"learning_rate": 0.00015357092910353001,
|
| 715 |
+
"loss": 0.0632,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.7863397548161122,
|
| 720 |
+
"grad_norm": 0.34466907382011414,
|
| 721 |
+
"learning_rate": 0.00015258810602276654,
|
| 722 |
+
"loss": 0.0484,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 1.8038528896672505,
|
| 727 |
+
"grad_norm": 0.42769935727119446,
|
| 728 |
+
"learning_rate": 0.00015159821058137278,
|
| 729 |
+
"loss": 0.0485,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 1.821366024518389,
|
| 734 |
+
"grad_norm": 0.35817593336105347,
|
| 735 |
+
"learning_rate": 0.00015060137590636318,
|
| 736 |
+
"loss": 0.0479,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 1.8388791593695273,
|
| 741 |
+
"grad_norm": 0.34612253308296204,
|
| 742 |
+
"learning_rate": 0.00014959773605798145,
|
| 743 |
+
"loss": 0.0552,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 1.8563922942206657,
|
| 748 |
+
"grad_norm": 0.285166472196579,
|
| 749 |
+
"learning_rate": 0.0001485874260116714,
|
| 750 |
+
"loss": 0.0486,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 1.873905429071804,
|
| 755 |
+
"grad_norm": 0.2631461024284363,
|
| 756 |
+
"learning_rate": 0.00014757058163992464,
|
| 757 |
+
"loss": 0.0454,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.8914185639229422,
|
| 762 |
+
"grad_norm": 0.3274662494659424,
|
| 763 |
+
"learning_rate": 0.0001465473396940078,
|
| 764 |
+
"loss": 0.0533,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 1.9089316987740805,
|
| 769 |
+
"grad_norm": 0.31185653805732727,
|
| 770 |
+
"learning_rate": 0.000145517837785571,
|
| 771 |
+
"loss": 0.0488,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 1.926444833625219,
|
| 776 |
+
"grad_norm": 0.35010799765586853,
|
| 777 |
+
"learning_rate": 0.0001444822143681415,
|
| 778 |
+
"loss": 0.0475,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 1.9439579684763573,
|
| 783 |
+
"grad_norm": 0.24580919742584229,
|
| 784 |
+
"learning_rate": 0.00014344060871850325,
|
| 785 |
+
"loss": 0.042,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 1.9614711033274956,
|
| 790 |
+
"grad_norm": 0.267715185880661,
|
| 791 |
+
"learning_rate": 0.00014239316091796647,
|
| 792 |
+
"loss": 0.0463,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 1.978984238178634,
|
| 797 |
+
"grad_norm": 0.4017065763473511,
|
| 798 |
+
"learning_rate": 0.00014134001183352832,
|
| 799 |
+
"loss": 0.0486,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.9964973730297724,
|
| 804 |
+
"grad_norm": 0.32043689489364624,
|
| 805 |
+
"learning_rate": 0.0001402813030989286,
|
| 806 |
+
"loss": 0.0476,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 2.0140105078809105,
|
| 811 |
+
"grad_norm": 0.3448532819747925,
|
| 812 |
+
"learning_rate": 0.00013921717709560182,
|
| 813 |
+
"loss": 0.0473,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 2.031523642732049,
|
| 818 |
+
"grad_norm": 0.3137170672416687,
|
| 819 |
+
"learning_rate": 0.000138147776933529,
|
| 820 |
+
"loss": 0.0383,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 2.0490367775831873,
|
| 825 |
+
"grad_norm": 0.26070958375930786,
|
| 826 |
+
"learning_rate": 0.00013707324643199114,
|
| 827 |
+
"loss": 0.0453,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 2.0665499124343256,
|
| 832 |
+
"grad_norm": 0.25074896216392517,
|
| 833 |
+
"learning_rate": 0.00013599373010022794,
|
| 834 |
+
"loss": 0.0406,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 2.084063047285464,
|
| 839 |
+
"grad_norm": 0.45140546560287476,
|
| 840 |
+
"learning_rate": 0.0001349093731180031,
|
| 841 |
+
"loss": 0.0485,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 2.1015761821366024,
|
| 846 |
+
"grad_norm": 0.4405800700187683,
|
| 847 |
+
"learning_rate": 0.00013382032131607966,
|
| 848 |
+
"loss": 0.053,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 2.1190893169877407,
|
| 853 |
+
"grad_norm": 0.26799100637435913,
|
| 854 |
+
"learning_rate": 0.00013272672115660796,
|
| 855 |
+
"loss": 0.0405,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 2.136602451838879,
|
| 860 |
+
"grad_norm": 0.2855483889579773,
|
| 861 |
+
"learning_rate": 0.00013162871971342837,
|
| 862 |
+
"loss": 0.0482,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 2.1541155866900175,
|
| 867 |
+
"grad_norm": 0.7333184480667114,
|
| 868 |
+
"learning_rate": 0.00013052646465229207,
|
| 869 |
+
"loss": 0.0452,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 2.171628721541156,
|
| 874 |
+
"grad_norm": 0.2893502414226532,
|
| 875 |
+
"learning_rate": 0.00012942010421100207,
|
| 876 |
+
"loss": 0.0486,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 2.189141856392294,
|
| 881 |
+
"grad_norm": 0.4282691478729248,
|
| 882 |
+
"learning_rate": 0.00012830978717947718,
|
| 883 |
+
"loss": 0.0451,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 2.2066549912434326,
|
| 888 |
+
"grad_norm": 0.2878718376159668,
|
| 889 |
+
"learning_rate": 0.00012719566287974204,
|
| 890 |
+
"loss": 0.0325,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 2.224168126094571,
|
| 895 |
+
"grad_norm": 0.230279803276062,
|
| 896 |
+
"learning_rate": 0.00012607788114584522,
|
| 897 |
+
"loss": 0.0392,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 2.2416812609457093,
|
| 902 |
+
"grad_norm": 0.28958141803741455,
|
| 903 |
+
"learning_rate": 0.0001249565923037088,
|
| 904 |
+
"loss": 0.0436,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 2.2591943957968477,
|
| 909 |
+
"grad_norm": 0.3924708366394043,
|
| 910 |
+
"learning_rate": 0.00012383194715091163,
|
| 911 |
+
"loss": 0.0468,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 2.276707530647986,
|
| 916 |
+
"grad_norm": 0.348954439163208,
|
| 917 |
+
"learning_rate": 0.00012270409693640905,
|
| 918 |
+
"loss": 0.045,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 2.2942206654991244,
|
| 923 |
+
"grad_norm": 0.36226847767829895,
|
| 924 |
+
"learning_rate": 0.00012157319334019219,
|
| 925 |
+
"loss": 0.0436,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 2.3117338003502628,
|
| 930 |
+
"grad_norm": 0.7187547087669373,
|
| 931 |
+
"learning_rate": 0.00012043938845288904,
|
| 932 |
+
"loss": 0.0429,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 2.329246935201401,
|
| 937 |
+
"grad_norm": 0.34101036190986633,
|
| 938 |
+
"learning_rate": 0.00011930283475531048,
|
| 939 |
+
"loss": 0.0515,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 2.3467600700525395,
|
| 944 |
+
"grad_norm": 0.36370015144348145,
|
| 945 |
+
"learning_rate": 0.00011816368509794364,
|
| 946 |
+
"loss": 0.0546,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 2.364273204903678,
|
| 951 |
+
"grad_norm": 0.23441250622272491,
|
| 952 |
+
"learning_rate": 0.00011702209268039581,
|
| 953 |
+
"loss": 0.0473,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 2.3817863397548162,
|
| 958 |
+
"grad_norm": 0.3847810924053192,
|
| 959 |
+
"learning_rate": 0.00011587821103079111,
|
| 960 |
+
"loss": 0.0417,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 2.3992994746059546,
|
| 965 |
+
"grad_norm": 0.29462698101997375,
|
| 966 |
+
"learning_rate": 0.00011473219398512316,
|
| 967 |
+
"loss": 0.0511,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 2.416812609457093,
|
| 972 |
+
"grad_norm": 0.44562244415283203,
|
| 973 |
+
"learning_rate": 0.00011358419566656642,
|
| 974 |
+
"loss": 0.0404,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 2.4343257443082313,
|
| 979 |
+
"grad_norm": 0.26545700430870056,
|
| 980 |
+
"learning_rate": 0.00011243437046474853,
|
| 981 |
+
"loss": 0.0473,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 2.4518388791593697,
|
| 986 |
+
"grad_norm": 0.29075849056243896,
|
| 987 |
+
"learning_rate": 0.00011128287301498739,
|
| 988 |
+
"loss": 0.0463,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 2.469352014010508,
|
| 993 |
+
"grad_norm": 0.27019691467285156,
|
| 994 |
+
"learning_rate": 0.00011012985817749463,
|
| 995 |
+
"loss": 0.0503,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 2.4868651488616464,
|
| 1000 |
+
"grad_norm": 0.30761951208114624,
|
| 1001 |
+
"learning_rate": 0.00010897548101654926,
|
| 1002 |
+
"loss": 0.0411,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 2.504378283712785,
|
| 1007 |
+
"grad_norm": 0.35718128085136414,
|
| 1008 |
+
"learning_rate": 0.00010781989677964355,
|
| 1009 |
+
"loss": 0.0457,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 2.521891418563923,
|
| 1014 |
+
"grad_norm": 0.30604833364486694,
|
| 1015 |
+
"learning_rate": 0.00010666326087660458,
|
| 1016 |
+
"loss": 0.0383,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 2.5394045534150615,
|
| 1021 |
+
"grad_norm": 0.25744375586509705,
|
| 1022 |
+
"learning_rate": 0.00010550572885869367,
|
| 1023 |
+
"loss": 0.0471,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 2.5569176882662,
|
| 1028 |
+
"grad_norm": 0.3305884301662445,
|
| 1029 |
+
"learning_rate": 0.00010434745639768705,
|
| 1030 |
+
"loss": 0.039,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 2.574430823117338,
|
| 1035 |
+
"grad_norm": 0.23783540725708008,
|
| 1036 |
+
"learning_rate": 0.00010318859926494014,
|
| 1037 |
+
"loss": 0.0298,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 2.591943957968476,
|
| 1042 |
+
"grad_norm": 0.24423924088478088,
|
| 1043 |
+
"learning_rate": 0.00010202931331043839,
|
| 1044 |
+
"loss": 0.0442,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 2.6094570928196146,
|
| 1049 |
+
"grad_norm": 0.37258943915367126,
|
| 1050 |
+
"learning_rate": 0.00010086975444183782,
|
| 1051 |
+
"loss": 0.0347,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 2.626970227670753,
|
| 1056 |
+
"grad_norm": 0.35688072443008423,
|
| 1057 |
+
"learning_rate": 9.971007860349756e-05,
|
| 1058 |
+
"loss": 0.0419,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 2.6444833625218913,
|
| 1063 |
+
"grad_norm": 0.4141565263271332,
|
| 1064 |
+
"learning_rate": 9.855044175550756e-05,
|
| 1065 |
+
"loss": 0.0357,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 2.6619964973730297,
|
| 1070 |
+
"grad_norm": 0.3509404957294464,
|
| 1071 |
+
"learning_rate": 9.739099985271394e-05,
|
| 1072 |
+
"loss": 0.0389,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 2.679509632224168,
|
| 1077 |
+
"grad_norm": 0.20315895974636078,
|
| 1078 |
+
"learning_rate": 9.623190882374564e-05,
|
| 1079 |
+
"loss": 0.0397,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 2.6970227670753064,
|
| 1084 |
+
"grad_norm": 0.30598539113998413,
|
| 1085 |
+
"learning_rate": 9.507332455004395e-05,
|
| 1086 |
+
"loss": 0.0363,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 2.714535901926445,
|
| 1091 |
+
"grad_norm": 0.2587359547615051,
|
| 1092 |
+
"learning_rate": 9.391540284489862e-05,
|
| 1093 |
+
"loss": 0.039,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 2.732049036777583,
|
| 1098 |
+
"grad_norm": 0.19274114072322845,
|
| 1099 |
+
"learning_rate": 9.275829943249334e-05,
|
| 1100 |
+
"loss": 0.0305,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 2.7495621716287215,
|
| 1105 |
+
"grad_norm": 0.5883513689041138,
|
| 1106 |
+
"learning_rate": 9.160216992696286e-05,
|
| 1107 |
+
"loss": 0.0343,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 2.76707530647986,
|
| 1112 |
+
"grad_norm": 0.29722627997398376,
|
| 1113 |
+
"learning_rate": 9.044716981146526e-05,
|
| 1114 |
+
"loss": 0.0353,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 2.7845884413309983,
|
| 1119 |
+
"grad_norm": 0.23005834221839905,
|
| 1120 |
+
"learning_rate": 8.929345441727142e-05,
|
| 1121 |
+
"loss": 0.0375,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 2.8021015761821366,
|
| 1126 |
+
"grad_norm": 0.2886779308319092,
|
| 1127 |
+
"learning_rate": 8.814117890287538e-05,
|
| 1128 |
+
"loss": 0.0361,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 2.819614711033275,
|
| 1133 |
+
"grad_norm": 0.3320513665676117,
|
| 1134 |
+
"learning_rate": 8.699049823312748e-05,
|
| 1135 |
+
"loss": 0.0381,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 2.8371278458844134,
|
| 1140 |
+
"grad_norm": 0.3187779188156128,
|
| 1141 |
+
"learning_rate": 8.584156715839401e-05,
|
| 1142 |
+
"loss": 0.0493,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 2.8546409807355517,
|
| 1147 |
+
"grad_norm": 0.3513469994068146,
|
| 1148 |
+
"learning_rate": 8.469454019374531e-05,
|
| 1149 |
+
"loss": 0.037,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 2.87215411558669,
|
| 1154 |
+
"grad_norm": 0.2172853648662567,
|
| 1155 |
+
"learning_rate": 8.354957159817561e-05,
|
| 1156 |
+
"loss": 0.0347,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 2.8896672504378285,
|
| 1161 |
+
"grad_norm": 0.2791577875614166,
|
| 1162 |
+
"learning_rate": 8.240681535385757e-05,
|
| 1163 |
+
"loss": 0.0311,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 2.907180385288967,
|
| 1168 |
+
"grad_norm": 0.3376358449459076,
|
| 1169 |
+
"learning_rate": 8.126642514543359e-05,
|
| 1170 |
+
"loss": 0.0375,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 2.924693520140105,
|
| 1175 |
+
"grad_norm": 0.23962591588497162,
|
| 1176 |
+
"learning_rate": 8.012855433934765e-05,
|
| 1177 |
+
"loss": 0.0338,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 2.9422066549912436,
|
| 1182 |
+
"grad_norm": 0.19910161197185516,
|
| 1183 |
+
"learning_rate": 7.899335596321944e-05,
|
| 1184 |
+
"loss": 0.0327,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 2.959719789842382,
|
| 1189 |
+
"grad_norm": 0.3907780349254608,
|
| 1190 |
+
"learning_rate": 7.786098268526448e-05,
|
| 1191 |
+
"loss": 0.0471,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 2.9772329246935203,
|
| 1196 |
+
"grad_norm": 0.3823835253715515,
|
| 1197 |
+
"learning_rate": 7.673158679376234e-05,
|
| 1198 |
+
"loss": 0.044,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 2.9947460595446582,
|
| 1203 |
+
"grad_norm": 0.21028929948806763,
|
| 1204 |
+
"learning_rate": 7.560532017657585e-05,
|
| 1205 |
+
"loss": 0.0348,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 3.012259194395797,
|
| 1210 |
+
"grad_norm": 0.390628844499588,
|
| 1211 |
+
"learning_rate": 7.448233430072466e-05,
|
| 1212 |
+
"loss": 0.0342,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 3.0297723292469354,
|
| 1217 |
+
"grad_norm": 0.3196234703063965,
|
| 1218 |
+
"learning_rate": 7.336278019201462e-05,
|
| 1219 |
+
"loss": 0.039,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 3.0472854640980738,
|
| 1224 |
+
"grad_norm": 0.39649903774261475,
|
| 1225 |
+
"learning_rate": 7.224680841472741e-05,
|
| 1226 |
+
"loss": 0.0373,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 3.0647985989492117,
|
| 1231 |
+
"grad_norm": 0.18303707242012024,
|
| 1232 |
+
"learning_rate": 7.113456905137132e-05,
|
| 1233 |
+
"loss": 0.0321,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 3.08231173380035,
|
| 1238 |
+
"grad_norm": 0.2457733154296875,
|
| 1239 |
+
"learning_rate": 7.002621168249759e-05,
|
| 1240 |
+
"loss": 0.0259,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 3.0998248686514884,
|
| 1245 |
+
"grad_norm": 0.31922516226768494,
|
| 1246 |
+
"learning_rate": 6.892188536658369e-05,
|
| 1247 |
+
"loss": 0.0354,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 3.117338003502627,
|
| 1252 |
+
"grad_norm": 0.42664676904678345,
|
| 1253 |
+
"learning_rate": 6.782173861998726e-05,
|
| 1254 |
+
"loss": 0.0329,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 3.134851138353765,
|
| 1259 |
+
"grad_norm": 0.20944933593273163,
|
| 1260 |
+
"learning_rate": 6.672591939697261e-05,
|
| 1261 |
+
"loss": 0.0364,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 3.1523642732049035,
|
| 1266 |
+
"grad_norm": 0.1604316681623459,
|
| 1267 |
+
"learning_rate": 6.563457506981297e-05,
|
| 1268 |
+
"loss": 0.0373,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 3.169877408056042,
|
| 1273 |
+
"grad_norm": 0.25888216495513916,
|
| 1274 |
+
"learning_rate": 6.454785240897112e-05,
|
| 1275 |
+
"loss": 0.0299,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 3.1873905429071803,
|
| 1280 |
+
"grad_norm": 0.28909996151924133,
|
| 1281 |
+
"learning_rate": 6.34658975633605e-05,
|
| 1282 |
+
"loss": 0.0251,
|
| 1283 |
+
"step": 1820
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 3.2049036777583186,
|
| 1287 |
+
"grad_norm": 0.29073986411094666,
|
| 1288 |
+
"learning_rate": 6.238885604069075e-05,
|
| 1289 |
+
"loss": 0.0309,
|
| 1290 |
+
"step": 1830
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 3.222416812609457,
|
| 1294 |
+
"grad_norm": 0.24771836400032043,
|
| 1295 |
+
"learning_rate": 6.131687268789838e-05,
|
| 1296 |
+
"loss": 0.0321,
|
| 1297 |
+
"step": 1840
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 3.2399299474605954,
|
| 1301 |
+
"grad_norm": 0.1985064148902893,
|
| 1302 |
+
"learning_rate": 6.0250091671667484e-05,
|
| 1303 |
+
"loss": 0.0315,
|
| 1304 |
+
"step": 1850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 3.2574430823117337,
|
| 1308 |
+
"grad_norm": 0.26381996273994446,
|
| 1309 |
+
"learning_rate": 5.9188656459040837e-05,
|
| 1310 |
+
"loss": 0.0298,
|
| 1311 |
+
"step": 1860
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 3.274956217162872,
|
| 1315 |
+
"grad_norm": 0.23007571697235107,
|
| 1316 |
+
"learning_rate": 5.813270979812589e-05,
|
| 1317 |
+
"loss": 0.0259,
|
| 1318 |
+
"step": 1870
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 3.2924693520140105,
|
| 1322 |
+
"grad_norm": 0.40646758675575256,
|
| 1323 |
+
"learning_rate": 5.7082393698897166e-05,
|
| 1324 |
+
"loss": 0.0293,
|
| 1325 |
+
"step": 1880
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 3.309982486865149,
|
| 1329 |
+
"grad_norm": 0.1977083384990692,
|
| 1330 |
+
"learning_rate": 5.60378494140976e-05,
|
| 1331 |
+
"loss": 0.036,
|
| 1332 |
+
"step": 1890
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 3.327495621716287,
|
| 1336 |
+
"grad_norm": 0.3425348699092865,
|
| 1337 |
+
"learning_rate": 5.4999217420242576e-05,
|
| 1338 |
+
"loss": 0.0356,
|
| 1339 |
+
"step": 1900
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 3.3450087565674256,
|
| 1343 |
+
"grad_norm": 0.1148160770535469,
|
| 1344 |
+
"learning_rate": 5.396663739872725e-05,
|
| 1345 |
+
"loss": 0.0308,
|
| 1346 |
+
"step": 1910
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 3.362521891418564,
|
| 1350 |
+
"grad_norm": 0.32002705335617065,
|
| 1351 |
+
"learning_rate": 5.294024821704172e-05,
|
| 1352 |
+
"loss": 0.0258,
|
| 1353 |
+
"step": 1920
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 3.3800350262697023,
|
| 1357 |
+
"grad_norm": 0.2819429934024811,
|
| 1358 |
+
"learning_rate": 5.192018791009521e-05,
|
| 1359 |
+
"loss": 0.0287,
|
| 1360 |
+
"step": 1930
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 3.3975481611208407,
|
| 1364 |
+
"grad_norm": 0.21581225097179413,
|
| 1365 |
+
"learning_rate": 5.090659366165227e-05,
|
| 1366 |
+
"loss": 0.0246,
|
| 1367 |
+
"step": 1940
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 3.415061295971979,
|
| 1371 |
+
"grad_norm": 0.21550467610359192,
|
| 1372 |
+
"learning_rate": 4.989960178588357e-05,
|
| 1373 |
+
"loss": 0.0282,
|
| 1374 |
+
"step": 1950
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 3.4325744308231174,
|
| 1378 |
+
"grad_norm": 0.3333240747451782,
|
| 1379 |
+
"learning_rate": 4.889934770903336e-05,
|
| 1380 |
+
"loss": 0.0291,
|
| 1381 |
+
"step": 1960
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 3.450087565674256,
|
| 1385 |
+
"grad_norm": 0.2303341180086136,
|
| 1386 |
+
"learning_rate": 4.790596595120699e-05,
|
| 1387 |
+
"loss": 0.0315,
|
| 1388 |
+
"step": 1970
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 3.467600700525394,
|
| 1392 |
+
"grad_norm": 0.14156457781791687,
|
| 1393 |
+
"learning_rate": 4.6919590108279254e-05,
|
| 1394 |
+
"loss": 0.0404,
|
| 1395 |
+
"step": 1980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 3.4851138353765325,
|
| 1399 |
+
"grad_norm": 0.21455378830432892,
|
| 1400 |
+
"learning_rate": 4.594035283392815e-05,
|
| 1401 |
+
"loss": 0.0288,
|
| 1402 |
+
"step": 1990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 3.502626970227671,
|
| 1406 |
+
"grad_norm": 0.2699653208255768,
|
| 1407 |
+
"learning_rate": 4.49683858217944e-05,
|
| 1408 |
+
"loss": 0.03,
|
| 1409 |
+
"step": 2000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 3.5201401050788093,
|
| 1413 |
+
"grad_norm": 0.1770617961883545,
|
| 1414 |
+
"learning_rate": 4.4003819787770964e-05,
|
| 1415 |
+
"loss": 0.027,
|
| 1416 |
+
"step": 2010
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 3.5376532399299476,
|
| 1420 |
+
"grad_norm": 0.2530314326286316,
|
| 1421 |
+
"learning_rate": 4.304678445242309e-05,
|
| 1422 |
+
"loss": 0.0278,
|
| 1423 |
+
"step": 2020
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 3.555166374781086,
|
| 1427 |
+
"grad_norm": 0.16524188220500946,
|
| 1428 |
+
"learning_rate": 4.209740852354313e-05,
|
| 1429 |
+
"loss": 0.0285,
|
| 1430 |
+
"step": 2030
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 3.5726795096322244,
|
| 1434 |
+
"grad_norm": 0.30064013600349426,
|
| 1435 |
+
"learning_rate": 4.115581967884094e-05,
|
| 1436 |
+
"loss": 0.0264,
|
| 1437 |
+
"step": 2040
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 3.5901926444833627,
|
| 1441 |
+
"grad_norm": 0.21011033654212952,
|
| 1442 |
+
"learning_rate": 4.022214454877305e-05,
|
| 1443 |
+
"loss": 0.0292,
|
| 1444 |
+
"step": 2050
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 3.607705779334501,
|
| 1448 |
+
"grad_norm": 0.28686150908470154,
|
| 1449 |
+
"learning_rate": 3.929650869951278e-05,
|
| 1450 |
+
"loss": 0.0261,
|
| 1451 |
+
"step": 2060
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 3.6252189141856395,
|
| 1455 |
+
"grad_norm": 0.3658103942871094,
|
| 1456 |
+
"learning_rate": 3.8379036616063066e-05,
|
| 1457 |
+
"loss": 0.0345,
|
| 1458 |
+
"step": 2070
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 3.642732049036778,
|
| 1462 |
+
"grad_norm": 0.18147164583206177,
|
| 1463 |
+
"learning_rate": 3.746985168551532e-05,
|
| 1464 |
+
"loss": 0.0318,
|
| 1465 |
+
"step": 2080
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 3.660245183887916,
|
| 1469 |
+
"grad_norm": 0.36226341128349304,
|
| 1470 |
+
"learning_rate": 3.65690761804554e-05,
|
| 1471 |
+
"loss": 0.0227,
|
| 1472 |
+
"step": 2090
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 3.6777583187390546,
|
| 1476 |
+
"grad_norm": 0.2645871043205261,
|
| 1477 |
+
"learning_rate": 3.567683124251972e-05,
|
| 1478 |
+
"loss": 0.0252,
|
| 1479 |
+
"step": 2100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 3.6952714535901925,
|
| 1483 |
+
"grad_norm": 0.2898857295513153,
|
| 1484 |
+
"learning_rate": 3.4793236866103294e-05,
|
| 1485 |
+
"loss": 0.0312,
|
| 1486 |
+
"step": 2110
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 3.712784588441331,
|
| 1490 |
+
"grad_norm": 0.3119029700756073,
|
| 1491 |
+
"learning_rate": 3.391841188222246e-05,
|
| 1492 |
+
"loss": 0.0403,
|
| 1493 |
+
"step": 2120
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 3.7302977232924692,
|
| 1497 |
+
"grad_norm": 0.3134075999259949,
|
| 1498 |
+
"learning_rate": 3.305247394253349e-05,
|
| 1499 |
+
"loss": 0.0339,
|
| 1500 |
+
"step": 2130
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 3.7478108581436076,
|
| 1504 |
+
"grad_norm": 0.22461438179016113,
|
| 1505 |
+
"learning_rate": 3.2195539503510164e-05,
|
| 1506 |
+
"loss": 0.0284,
|
| 1507 |
+
"step": 2140
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 3.765323992994746,
|
| 1511 |
+
"grad_norm": 0.19074226915836334,
|
| 1512 |
+
"learning_rate": 3.1347723810782134e-05,
|
| 1513 |
+
"loss": 0.0232,
|
| 1514 |
+
"step": 2150
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 3.7828371278458843,
|
| 1518 |
+
"grad_norm": 0.2736382484436035,
|
| 1519 |
+
"learning_rate": 3.05091408836359e-05,
|
| 1520 |
+
"loss": 0.0389,
|
| 1521 |
+
"step": 2160
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 3.8003502626970227,
|
| 1525 |
+
"grad_norm": 0.18806475400924683,
|
| 1526 |
+
"learning_rate": 2.967990349968086e-05,
|
| 1527 |
+
"loss": 0.0254,
|
| 1528 |
+
"step": 2170
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 3.817863397548161,
|
| 1532 |
+
"grad_norm": 0.1967362016439438,
|
| 1533 |
+
"learning_rate": 2.8860123179682242e-05,
|
| 1534 |
+
"loss": 0.0198,
|
| 1535 |
+
"step": 2180
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 3.8353765323992994,
|
| 1539 |
+
"grad_norm": 0.19103026390075684,
|
| 1540 |
+
"learning_rate": 2.8049910172563188e-05,
|
| 1541 |
+
"loss": 0.0224,
|
| 1542 |
+
"step": 2190
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 3.852889667250438,
|
| 1546 |
+
"grad_norm": 0.2591644823551178,
|
| 1547 |
+
"learning_rate": 2.7249373440577963e-05,
|
| 1548 |
+
"loss": 0.0285,
|
| 1549 |
+
"step": 2200
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 3.870402802101576,
|
| 1553 |
+
"grad_norm": 0.3512148857116699,
|
| 1554 |
+
"learning_rate": 2.6458620644657693e-05,
|
| 1555 |
+
"loss": 0.0298,
|
| 1556 |
+
"step": 2210
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 3.8879159369527145,
|
| 1560 |
+
"grad_norm": 0.19308355450630188,
|
| 1561 |
+
"learning_rate": 2.567775812993186e-05,
|
| 1562 |
+
"loss": 0.0256,
|
| 1563 |
+
"step": 2220
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 3.905429071803853,
|
| 1567 |
+
"grad_norm": 0.1874024122953415,
|
| 1568 |
+
"learning_rate": 2.4906890911426208e-05,
|
| 1569 |
+
"loss": 0.03,
|
| 1570 |
+
"step": 2230
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 3.9229422066549913,
|
| 1574 |
+
"grad_norm": 0.2476789802312851,
|
| 1575 |
+
"learning_rate": 2.4146122659939686e-05,
|
| 1576 |
+
"loss": 0.0232,
|
| 1577 |
+
"step": 2240
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 3.9404553415061296,
|
| 1581 |
+
"grad_norm": 0.45866695046424866,
|
| 1582 |
+
"learning_rate": 2.339555568810221e-05,
|
| 1583 |
+
"loss": 0.0272,
|
| 1584 |
+
"step": 2250
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 3.957968476357268,
|
| 1588 |
+
"grad_norm": 0.3022903800010681,
|
| 1589 |
+
"learning_rate": 2.2655290936615093e-05,
|
| 1590 |
+
"loss": 0.0353,
|
| 1591 |
+
"step": 2260
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 3.9754816112084064,
|
| 1595 |
+
"grad_norm": 0.31178757548332214,
|
| 1596 |
+
"learning_rate": 2.1925427960675894e-05,
|
| 1597 |
+
"loss": 0.029,
|
| 1598 |
+
"step": 2270
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 3.9929947460595447,
|
| 1602 |
+
"grad_norm": 0.44911712408065796,
|
| 1603 |
+
"learning_rate": 2.120606491658966e-05,
|
| 1604 |
+
"loss": 0.0326,
|
| 1605 |
+
"step": 2280
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 4.010507880910683,
|
| 1609 |
+
"grad_norm": 0.23923397064208984,
|
| 1610 |
+
"learning_rate": 2.049729854856832e-05,
|
| 1611 |
+
"loss": 0.0274,
|
| 1612 |
+
"step": 2290
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 4.028021015761821,
|
| 1616 |
+
"grad_norm": 0.4409220218658447,
|
| 1617 |
+
"learning_rate": 1.9799224175719767e-05,
|
| 1618 |
+
"loss": 0.0247,
|
| 1619 |
+
"step": 2300
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 4.045534150612959,
|
| 1623 |
+
"grad_norm": 0.13721756637096405,
|
| 1624 |
+
"learning_rate": 1.9111935679229142e-05,
|
| 1625 |
+
"loss": 0.0176,
|
| 1626 |
+
"step": 2310
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 4.063047285464098,
|
| 1630 |
+
"grad_norm": 0.26091480255126953,
|
| 1631 |
+
"learning_rate": 1.843552548973272e-05,
|
| 1632 |
+
"loss": 0.0298,
|
| 1633 |
+
"step": 2320
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 4.080560420315236,
|
| 1637 |
+
"grad_norm": 0.29940757155418396,
|
| 1638 |
+
"learning_rate": 1.7770084574887567e-05,
|
| 1639 |
+
"loss": 0.0222,
|
| 1640 |
+
"step": 2330
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 4.0980735551663745,
|
| 1644 |
+
"grad_norm": 0.3561699390411377,
|
| 1645 |
+
"learning_rate": 1.7115702427137616e-05,
|
| 1646 |
+
"loss": 0.0251,
|
| 1647 |
+
"step": 2340
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 4.115586690017513,
|
| 1651 |
+
"grad_norm": 0.2851378619670868,
|
| 1652 |
+
"learning_rate": 1.647246705167812e-05,
|
| 1653 |
+
"loss": 0.0288,
|
| 1654 |
+
"step": 2350
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 4.133099824868651,
|
| 1658 |
+
"grad_norm": 0.26174086332321167,
|
| 1659 |
+
"learning_rate": 1.5840464954620206e-05,
|
| 1660 |
+
"loss": 0.0243,
|
| 1661 |
+
"step": 2360
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 4.15061295971979,
|
| 1665 |
+
"grad_norm": 0.2719419300556183,
|
| 1666 |
+
"learning_rate": 1.5219781131357103e-05,
|
| 1667 |
+
"loss": 0.0307,
|
| 1668 |
+
"step": 2370
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 4.168126094570928,
|
| 1672 |
+
"grad_norm": 0.3218068480491638,
|
| 1673 |
+
"learning_rate": 1.4610499055133375e-05,
|
| 1674 |
+
"loss": 0.0212,
|
| 1675 |
+
"step": 2380
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 4.185639229422066,
|
| 1679 |
+
"grad_norm": 0.2588444948196411,
|
| 1680 |
+
"learning_rate": 1.401270066581899e-05,
|
| 1681 |
+
"loss": 0.0339,
|
| 1682 |
+
"step": 2390
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 4.203152364273205,
|
| 1686 |
+
"grad_norm": 0.2679278254508972,
|
| 1687 |
+
"learning_rate": 1.3426466358889545e-05,
|
| 1688 |
+
"loss": 0.0229,
|
| 1689 |
+
"step": 2400
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 4.220665499124343,
|
| 1693 |
+
"grad_norm": 0.1962890774011612,
|
| 1694 |
+
"learning_rate": 1.2851874974614097e-05,
|
| 1695 |
+
"loss": 0.0222,
|
| 1696 |
+
"step": 2410
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 4.2381786339754814,
|
| 1700 |
+
"grad_norm": 0.32625964283943176,
|
| 1701 |
+
"learning_rate": 1.2289003787452557e-05,
|
| 1702 |
+
"loss": 0.0284,
|
| 1703 |
+
"step": 2420
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 4.25569176882662,
|
| 1707 |
+
"grad_norm": 0.2041284441947937,
|
| 1708 |
+
"learning_rate": 1.1737928495662964e-05,
|
| 1709 |
+
"loss": 0.0204,
|
| 1710 |
+
"step": 2430
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 4.273204903677758,
|
| 1714 |
+
"grad_norm": 0.29407230019569397,
|
| 1715 |
+
"learning_rate": 1.1198723211121442e-05,
|
| 1716 |
+
"loss": 0.0215,
|
| 1717 |
+
"step": 2440
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 4.2907180385288965,
|
| 1721 |
+
"grad_norm": 0.14956320822238922,
|
| 1722 |
+
"learning_rate": 1.0671460449355075e-05,
|
| 1723 |
+
"loss": 0.0298,
|
| 1724 |
+
"step": 2450
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 4.308231173380035,
|
| 1728 |
+
"grad_norm": 0.2270493358373642,
|
| 1729 |
+
"learning_rate": 1.0156211119789583e-05,
|
| 1730 |
+
"loss": 0.0227,
|
| 1731 |
+
"step": 2460
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 4.325744308231173,
|
| 1735 |
+
"grad_norm": 0.14962418377399445,
|
| 1736 |
+
"learning_rate": 9.65304451621304e-06,
|
| 1737 |
+
"loss": 0.0262,
|
| 1738 |
+
"step": 2470
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 4.343257443082312,
|
| 1742 |
+
"grad_norm": 0.18282897770404816,
|
| 1743 |
+
"learning_rate": 9.162028307456771e-06,
|
| 1744 |
+
"loss": 0.025,
|
| 1745 |
+
"step": 2480
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 4.36077057793345,
|
| 1749 |
+
"grad_norm": 0.21446016430854797,
|
| 1750 |
+
"learning_rate": 8.683228528294929e-06,
|
| 1751 |
+
"loss": 0.0278,
|
| 1752 |
+
"step": 2490
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 4.378283712784588,
|
| 1756 |
+
"grad_norm": 0.1467599719762802,
|
| 1757 |
+
"learning_rate": 8.216709570563685e-06,
|
| 1758 |
+
"loss": 0.0226,
|
| 1759 |
+
"step": 2500
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 4.395796847635727,
|
| 1763 |
+
"grad_norm": 0.2947424650192261,
|
| 1764 |
+
"learning_rate": 7.76253417450149e-06,
|
| 1765 |
+
"loss": 0.0309,
|
| 1766 |
+
"step": 2510
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 4.413309982486865,
|
| 1770 |
+
"grad_norm": 0.12388034909963608,
|
| 1771 |
+
"learning_rate": 7.320763420311261e-06,
|
| 1772 |
+
"loss": 0.0192,
|
| 1773 |
+
"step": 2520
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 4.4308231173380035,
|
| 1777 |
+
"grad_norm": 0.211833193898201,
|
| 1778 |
+
"learning_rate": 6.891456719946188e-06,
|
| 1779 |
+
"loss": 0.025,
|
| 1780 |
+
"step": 2530
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 4.448336252189142,
|
| 1784 |
+
"grad_norm": 0.2567002773284912,
|
| 1785 |
+
"learning_rate": 6.4746718091194254e-06,
|
| 1786 |
+
"loss": 0.0304,
|
| 1787 |
+
"step": 2540
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 4.46584938704028,
|
| 1791 |
+
"grad_norm": 0.14190255105495453,
|
| 1792 |
+
"learning_rate": 6.07046473953955e-06,
|
| 1793 |
+
"loss": 0.0223,
|
| 1794 |
+
"step": 2550
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 4.483362521891419,
|
| 1798 |
+
"grad_norm": 0.231306791305542,
|
| 1799 |
+
"learning_rate": 5.678889871372428e-06,
|
| 1800 |
+
"loss": 0.0277,
|
| 1801 |
+
"step": 2560
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 4.500875656742557,
|
| 1805 |
+
"grad_norm": 0.13263490796089172,
|
| 1806 |
+
"learning_rate": 5.299999865930505e-06,
|
| 1807 |
+
"loss": 0.0272,
|
| 1808 |
+
"step": 2570
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 4.518388791593695,
|
| 1812 |
+
"grad_norm": 0.3280109763145447,
|
| 1813 |
+
"learning_rate": 4.933845678590587e-06,
|
| 1814 |
+
"loss": 0.0309,
|
| 1815 |
+
"step": 2580
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 4.535901926444834,
|
| 1819 |
+
"grad_norm": 0.2335793375968933,
|
| 1820 |
+
"learning_rate": 4.580476551941037e-06,
|
| 1821 |
+
"loss": 0.0209,
|
| 1822 |
+
"step": 2590
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 4.553415061295972,
|
| 1826 |
+
"grad_norm": 0.2255188226699829,
|
| 1827 |
+
"learning_rate": 4.2399400091594154e-06,
|
| 1828 |
+
"loss": 0.0323,
|
| 1829 |
+
"step": 2600
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 4.57092819614711,
|
| 1833 |
+
"grad_norm": 0.1700967699289322,
|
| 1834 |
+
"learning_rate": 3.912281847621213e-06,
|
| 1835 |
+
"loss": 0.022,
|
| 1836 |
+
"step": 2610
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 4.588441330998249,
|
| 1840 |
+
"grad_norm": 0.21621626615524292,
|
| 1841 |
+
"learning_rate": 3.597546132740792e-06,
|
| 1842 |
+
"loss": 0.0229,
|
| 1843 |
+
"step": 2620
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 4.605954465849387,
|
| 1847 |
+
"grad_norm": 0.2476493865251541,
|
| 1848 |
+
"learning_rate": 3.295775192045181e-06,
|
| 1849 |
+
"loss": 0.0298,
|
| 1850 |
+
"step": 2630
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 4.6234676007005255,
|
| 1854 |
+
"grad_norm": 0.29315319657325745,
|
| 1855 |
+
"learning_rate": 3.0070096094816037e-06,
|
| 1856 |
+
"loss": 0.0233,
|
| 1857 |
+
"step": 2640
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 4.640980735551664,
|
| 1861 |
+
"grad_norm": 0.2687146067619324,
|
| 1862 |
+
"learning_rate": 2.7312882199595826e-06,
|
| 1863 |
+
"loss": 0.0262,
|
| 1864 |
+
"step": 2650
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 4.658493870402802,
|
| 1868 |
+
"grad_norm": 0.8978672027587891,
|
| 1869 |
+
"learning_rate": 2.4686481041280574e-06,
|
| 1870 |
+
"loss": 0.0209,
|
| 1871 |
+
"step": 2660
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 4.676007005253941,
|
| 1875 |
+
"grad_norm": 0.21821951866149902,
|
| 1876 |
+
"learning_rate": 2.2191245833886987e-06,
|
| 1877 |
+
"loss": 0.0208,
|
| 1878 |
+
"step": 2670
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 4.693520140105079,
|
| 1882 |
+
"grad_norm": 0.22492659091949463,
|
| 1883 |
+
"learning_rate": 1.9827512151456173e-06,
|
| 1884 |
+
"loss": 0.0238,
|
| 1885 |
+
"step": 2680
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 4.711033274956217,
|
| 1889 |
+
"grad_norm": 0.17041510343551636,
|
| 1890 |
+
"learning_rate": 1.7595597882923309e-06,
|
| 1891 |
+
"loss": 0.0213,
|
| 1892 |
+
"step": 2690
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 4.728546409807356,
|
| 1896 |
+
"grad_norm": 0.2669731676578522,
|
| 1897 |
+
"learning_rate": 1.549580318936672e-06,
|
| 1898 |
+
"loss": 0.0238,
|
| 1899 |
+
"step": 2700
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 4.746059544658494,
|
| 1903 |
+
"grad_norm": 0.2010369598865509,
|
| 1904 |
+
"learning_rate": 1.3528410463639728e-06,
|
| 1905 |
+
"loss": 0.0189,
|
| 1906 |
+
"step": 2710
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 4.7635726795096325,
|
| 1910 |
+
"grad_norm": 0.37732240557670593,
|
| 1911 |
+
"learning_rate": 1.1693684292393704e-06,
|
| 1912 |
+
"loss": 0.0257,
|
| 1913 |
+
"step": 2720
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 4.781085814360771,
|
| 1917 |
+
"grad_norm": 0.3189486563205719,
|
| 1918 |
+
"learning_rate": 9.991871420493736e-07,
|
| 1919 |
+
"loss": 0.0181,
|
| 1920 |
+
"step": 2730
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 4.798598949211909,
|
| 1924 |
+
"grad_norm": 0.17701123654842377,
|
| 1925 |
+
"learning_rate": 8.423200717835977e-07,
|
| 1926 |
+
"loss": 0.0214,
|
| 1927 |
+
"step": 2740
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 4.816112084063048,
|
| 1931 |
+
"grad_norm": 0.2231394648551941,
|
| 1932 |
+
"learning_rate": 6.987883148567131e-07,
|
| 1933 |
+
"loss": 0.0215,
|
| 1934 |
+
"step": 2750
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 4.833625218914186,
|
| 1938 |
+
"grad_norm": 0.1491653025150299,
|
| 1939 |
+
"learning_rate": 5.686111742713162e-07,
|
| 1940 |
+
"loss": 0.017,
|
| 1941 |
+
"step": 2760
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 4.851138353765324,
|
| 1945 |
+
"grad_norm": 0.24444717168807983,
|
| 1946 |
+
"learning_rate": 4.5180615702192783e-07,
|
| 1947 |
+
"loss": 0.019,
|
| 1948 |
+
"step": 2770
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 4.868651488616463,
|
| 1952 |
+
"grad_norm": 0.22784839570522308,
|
| 1953 |
+
"learning_rate": 3.4838897174055417e-07,
|
| 1954 |
+
"loss": 0.0211,
|
| 1955 |
+
"step": 2780
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 4.886164623467601,
|
| 1959 |
+
"grad_norm": 0.23262278735637665,
|
| 1960 |
+
"learning_rate": 2.583735265840992e-07,
|
| 1961 |
+
"loss": 0.0274,
|
| 1962 |
+
"step": 2790
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 4.903677758318739,
|
| 1966 |
+
"grad_norm": 0.3909858763217926,
|
| 1967 |
+
"learning_rate": 1.8177192736390515e-07,
|
| 1968 |
+
"loss": 0.0247,
|
| 1969 |
+
"step": 2800
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 4.921190893169878,
|
| 1973 |
+
"grad_norm": 0.5872836709022522,
|
| 1974 |
+
"learning_rate": 1.1859447591769934e-07,
|
| 1975 |
+
"loss": 0.0256,
|
| 1976 |
+
"step": 2810
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 4.938704028021016,
|
| 1980 |
+
"grad_norm": 0.21353958547115326,
|
| 1981 |
+
"learning_rate": 6.884966872412468e-08,
|
| 1982 |
+
"loss": 0.0203,
|
| 1983 |
+
"step": 2820
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 4.956217162872154,
|
| 1987 |
+
"grad_norm": 0.3956884443759918,
|
| 1988 |
+
"learning_rate": 3.254419576012024e-08,
|
| 1989 |
+
"loss": 0.0308,
|
| 1990 |
+
"step": 2830
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 4.973730297723293,
|
| 1994 |
+
"grad_norm": 0.2731996476650238,
|
| 1995 |
+
"learning_rate": 9.682939601185492e-09,
|
| 1996 |
+
"loss": 0.0213,
|
| 1997 |
+
"step": 2840
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 4.99124343257443,
|
| 2001 |
+
"grad_norm": 0.1547698825597763,
|
| 2002 |
+
"learning_rate": 2.689747647166563e-10,
|
| 2003 |
+
"loss": 0.0245,
|
| 2004 |
+
"step": 2850
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 4.994746059544658,
|
| 2008 |
+
"step": 2852,
|
| 2009 |
+
"total_flos": 9.99125683129728e+16,
|
| 2010 |
+
"train_loss": 0.058785643973250624,
|
| 2011 |
+
"train_runtime": 1255.1038,
|
| 2012 |
+
"train_samples_per_second": 36.357,
|
| 2013 |
+
"train_steps_per_second": 2.272
|
| 2014 |
+
}
|
| 2015 |
+
],
|
| 2016 |
+
"logging_steps": 10,
|
| 2017 |
+
"max_steps": 2852,
|
| 2018 |
+
"num_input_tokens_seen": 0,
|
| 2019 |
+
"num_train_epochs": 5,
|
| 2020 |
+
"save_steps": 500,
|
| 2021 |
+
"stateful_callbacks": {
|
| 2022 |
+
"TrainerControl": {
|
| 2023 |
+
"args": {
|
| 2024 |
+
"should_epoch_stop": false,
|
| 2025 |
+
"should_evaluate": false,
|
| 2026 |
+
"should_log": false,
|
| 2027 |
+
"should_save": true,
|
| 2028 |
+
"should_training_stop": true
|
| 2029 |
+
},
|
| 2030 |
+
"attributes": {}
|
| 2031 |
+
}
|
| 2032 |
+
},
|
| 2033 |
+
"total_flos": 9.99125683129728e+16,
|
| 2034 |
+
"train_batch_size": 16,
|
| 2035 |
+
"trial_name": null,
|
| 2036 |
+
"trial_params": null
|
| 2037 |
+
}
|