Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +2793 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,2793 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 10.204663212435234,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3939,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.025906735751295335,
|
| 13 |
+
"grad_norm": 7.820053577423096,
|
| 14 |
+
"learning_rate": 1.015228426395939e-05,
|
| 15 |
+
"loss": 1.2662,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.05181347150259067,
|
| 20 |
+
"grad_norm": 2.624858856201172,
|
| 21 |
+
"learning_rate": 2.030456852791878e-05,
|
| 22 |
+
"loss": 0.829,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.07772020725388601,
|
| 27 |
+
"grad_norm": 3.9165711402893066,
|
| 28 |
+
"learning_rate": 3.0456852791878175e-05,
|
| 29 |
+
"loss": 0.4265,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.10362694300518134,
|
| 34 |
+
"grad_norm": 1.9645638465881348,
|
| 35 |
+
"learning_rate": 4.060913705583756e-05,
|
| 36 |
+
"loss": 0.3073,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.12953367875647667,
|
| 41 |
+
"grad_norm": 0.6432750225067139,
|
| 42 |
+
"learning_rate": 5.076142131979695e-05,
|
| 43 |
+
"loss": 0.234,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.15544041450777202,
|
| 48 |
+
"grad_norm": 2.0933594703674316,
|
| 49 |
+
"learning_rate": 6.091370558375635e-05,
|
| 50 |
+
"loss": 0.2026,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.18134715025906736,
|
| 55 |
+
"grad_norm": 0.8744344115257263,
|
| 56 |
+
"learning_rate": 7.106598984771574e-05,
|
| 57 |
+
"loss": 0.1762,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.20725388601036268,
|
| 62 |
+
"grad_norm": 1.3427742719650269,
|
| 63 |
+
"learning_rate": 8.121827411167512e-05,
|
| 64 |
+
"loss": 0.1651,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.23316062176165803,
|
| 69 |
+
"grad_norm": 0.8046813607215881,
|
| 70 |
+
"learning_rate": 9.137055837563452e-05,
|
| 71 |
+
"loss": 0.153,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.25906735751295334,
|
| 76 |
+
"grad_norm": 0.8898593783378601,
|
| 77 |
+
"learning_rate": 0.0001015228426395939,
|
| 78 |
+
"loss": 0.1349,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.2849740932642487,
|
| 83 |
+
"grad_norm": 0.5029700994491577,
|
| 84 |
+
"learning_rate": 0.0001116751269035533,
|
| 85 |
+
"loss": 0.131,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.31088082901554404,
|
| 90 |
+
"grad_norm": 0.5996863842010498,
|
| 91 |
+
"learning_rate": 0.0001218274111675127,
|
| 92 |
+
"loss": 0.1167,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.33678756476683935,
|
| 97 |
+
"grad_norm": 0.935474693775177,
|
| 98 |
+
"learning_rate": 0.00013197969543147207,
|
| 99 |
+
"loss": 0.103,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.3626943005181347,
|
| 104 |
+
"grad_norm": 0.9033486247062683,
|
| 105 |
+
"learning_rate": 0.00014213197969543148,
|
| 106 |
+
"loss": 0.1048,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.38860103626943004,
|
| 111 |
+
"grad_norm": 0.8507540822029114,
|
| 112 |
+
"learning_rate": 0.00015228426395939087,
|
| 113 |
+
"loss": 0.1061,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.41450777202072536,
|
| 118 |
+
"grad_norm": 0.8174213767051697,
|
| 119 |
+
"learning_rate": 0.00016243654822335025,
|
| 120 |
+
"loss": 0.1012,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.44041450777202074,
|
| 125 |
+
"grad_norm": 0.5809940099716187,
|
| 126 |
+
"learning_rate": 0.00017258883248730966,
|
| 127 |
+
"loss": 0.0953,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.46632124352331605,
|
| 132 |
+
"grad_norm": 0.5646868944168091,
|
| 133 |
+
"learning_rate": 0.00018274111675126904,
|
| 134 |
+
"loss": 0.0842,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.49222797927461137,
|
| 139 |
+
"grad_norm": 0.7231911420822144,
|
| 140 |
+
"learning_rate": 0.00019289340101522843,
|
| 141 |
+
"loss": 0.0846,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.5181347150259067,
|
| 146 |
+
"grad_norm": 0.5057425498962402,
|
| 147 |
+
"learning_rate": 0.00019999968282097247,
|
| 148 |
+
"loss": 0.0806,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.5440414507772021,
|
| 153 |
+
"grad_norm": 0.4385872483253479,
|
| 154 |
+
"learning_rate": 0.0001999940441386779,
|
| 155 |
+
"loss": 0.0811,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.5699481865284974,
|
| 160 |
+
"grad_norm": 0.49294012784957886,
|
| 161 |
+
"learning_rate": 0.0001999813574910152,
|
| 162 |
+
"loss": 0.0778,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.5958549222797928,
|
| 167 |
+
"grad_norm": 0.8409459590911865,
|
| 168 |
+
"learning_rate": 0.0001999616237721881,
|
| 169 |
+
"loss": 0.0777,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.6217616580310881,
|
| 174 |
+
"grad_norm": 0.9242765307426453,
|
| 175 |
+
"learning_rate": 0.00019993484437310518,
|
| 176 |
+
"loss": 0.0837,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.6476683937823834,
|
| 181 |
+
"grad_norm": 0.6676081418991089,
|
| 182 |
+
"learning_rate": 0.00019990102118128155,
|
| 183 |
+
"loss": 0.0722,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.6735751295336787,
|
| 188 |
+
"grad_norm": 0.36002111434936523,
|
| 189 |
+
"learning_rate": 0.0001998601565807061,
|
| 190 |
+
"loss": 0.0755,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.6994818652849741,
|
| 195 |
+
"grad_norm": 0.25635281205177307,
|
| 196 |
+
"learning_rate": 0.0001998122534516733,
|
| 197 |
+
"loss": 0.0651,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.7253886010362695,
|
| 202 |
+
"grad_norm": 0.7322975993156433,
|
| 203 |
+
"learning_rate": 0.00019975731517058012,
|
| 204 |
+
"loss": 0.0664,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.7512953367875648,
|
| 209 |
+
"grad_norm": 0.43520084023475647,
|
| 210 |
+
"learning_rate": 0.00019969534560968823,
|
| 211 |
+
"loss": 0.0653,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.7772020725388601,
|
| 216 |
+
"grad_norm": 0.39980387687683105,
|
| 217 |
+
"learning_rate": 0.000199626349136851,
|
| 218 |
+
"loss": 0.0677,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.8031088082901554,
|
| 223 |
+
"grad_norm": 0.4908325970172882,
|
| 224 |
+
"learning_rate": 0.0001995503306152055,
|
| 225 |
+
"loss": 0.0665,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.8290155440414507,
|
| 230 |
+
"grad_norm": 0.566394567489624,
|
| 231 |
+
"learning_rate": 0.00019946729540283,
|
| 232 |
+
"loss": 0.0592,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.8549222797927462,
|
| 237 |
+
"grad_norm": 0.3903208374977112,
|
| 238 |
+
"learning_rate": 0.0001993772493523661,
|
| 239 |
+
"loss": 0.0625,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.8808290155440415,
|
| 244 |
+
"grad_norm": 0.31053635478019714,
|
| 245 |
+
"learning_rate": 0.00019928019881060625,
|
| 246 |
+
"loss": 0.0668,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.9067357512953368,
|
| 251 |
+
"grad_norm": 0.33370479941368103,
|
| 252 |
+
"learning_rate": 0.00019917615061804637,
|
| 253 |
+
"loss": 0.0545,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.9326424870466321,
|
| 258 |
+
"grad_norm": 0.5482617020606995,
|
| 259 |
+
"learning_rate": 0.00019906511210840386,
|
| 260 |
+
"loss": 0.0628,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.9585492227979274,
|
| 265 |
+
"grad_norm": 0.5890028476715088,
|
| 266 |
+
"learning_rate": 0.00019894709110810057,
|
| 267 |
+
"loss": 0.0586,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.9844559585492227,
|
| 272 |
+
"grad_norm": 0.6143954396247864,
|
| 273 |
+
"learning_rate": 0.00019882209593571117,
|
| 274 |
+
"loss": 0.0596,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.0103626943005182,
|
| 279 |
+
"grad_norm": 0.27864834666252136,
|
| 280 |
+
"learning_rate": 0.00019869013540137683,
|
| 281 |
+
"loss": 0.0538,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.0362694300518134,
|
| 286 |
+
"grad_norm": 0.6054168343544006,
|
| 287 |
+
"learning_rate": 0.00019855121880618432,
|
| 288 |
+
"loss": 0.0552,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.0621761658031088,
|
| 293 |
+
"grad_norm": 0.3929780125617981,
|
| 294 |
+
"learning_rate": 0.00019840535594151023,
|
| 295 |
+
"loss": 0.0564,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.0880829015544042,
|
| 300 |
+
"grad_norm": 0.45577749609947205,
|
| 301 |
+
"learning_rate": 0.00019825255708833108,
|
| 302 |
+
"loss": 0.0538,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.1139896373056994,
|
| 307 |
+
"grad_norm": 0.4095696806907654,
|
| 308 |
+
"learning_rate": 0.00019809283301649866,
|
| 309 |
+
"loss": 0.0555,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.1398963730569949,
|
| 314 |
+
"grad_norm": 0.3670872747898102,
|
| 315 |
+
"learning_rate": 0.00019792619498398064,
|
| 316 |
+
"loss": 0.052,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.16580310880829,
|
| 321 |
+
"grad_norm": 0.40029576420783997,
|
| 322 |
+
"learning_rate": 0.00019775265473606743,
|
| 323 |
+
"loss": 0.0533,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.1917098445595855,
|
| 328 |
+
"grad_norm": 0.4621596038341522,
|
| 329 |
+
"learning_rate": 0.00019757222450454408,
|
| 330 |
+
"loss": 0.0481,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.2176165803108807,
|
| 335 |
+
"grad_norm": 0.5111753940582275,
|
| 336 |
+
"learning_rate": 0.00019738491700682832,
|
| 337 |
+
"loss": 0.0493,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.2435233160621761,
|
| 342 |
+
"grad_norm": 0.28371885418891907,
|
| 343 |
+
"learning_rate": 0.00019719074544507391,
|
| 344 |
+
"loss": 0.0489,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.2694300518134716,
|
| 349 |
+
"grad_norm": 0.4142397344112396,
|
| 350 |
+
"learning_rate": 0.0001969897235052405,
|
| 351 |
+
"loss": 0.0507,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.2953367875647668,
|
| 356 |
+
"grad_norm": 0.266857773065567,
|
| 357 |
+
"learning_rate": 0.0001967818653561286,
|
| 358 |
+
"loss": 0.0515,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.3212435233160622,
|
| 363 |
+
"grad_norm": 0.32568779587745667,
|
| 364 |
+
"learning_rate": 0.0001965671856483812,
|
| 365 |
+
"loss": 0.0504,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.3471502590673574,
|
| 370 |
+
"grad_norm": 0.43539974093437195,
|
| 371 |
+
"learning_rate": 0.00019634569951345088,
|
| 372 |
+
"loss": 0.0476,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.3730569948186528,
|
| 377 |
+
"grad_norm": 0.40603071451187134,
|
| 378 |
+
"learning_rate": 0.0001961174225625336,
|
| 379 |
+
"loss": 0.0519,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.3989637305699483,
|
| 384 |
+
"grad_norm": 0.5506306886672974,
|
| 385 |
+
"learning_rate": 0.00019588237088546807,
|
| 386 |
+
"loss": 0.0506,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.4248704663212435,
|
| 391 |
+
"grad_norm": 0.4322110414505005,
|
| 392 |
+
"learning_rate": 0.00019564056104960176,
|
| 393 |
+
"loss": 0.0412,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.450777202072539,
|
| 398 |
+
"grad_norm": 0.30257806181907654,
|
| 399 |
+
"learning_rate": 0.00019539201009862332,
|
| 400 |
+
"loss": 0.049,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.4766839378238341,
|
| 405 |
+
"grad_norm": 0.28851810097694397,
|
| 406 |
+
"learning_rate": 0.00019513673555136105,
|
| 407 |
+
"loss": 0.0503,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.5025906735751295,
|
| 412 |
+
"grad_norm": 0.39474016427993774,
|
| 413 |
+
"learning_rate": 0.00019487475540054828,
|
| 414 |
+
"loss": 0.0497,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.528497409326425,
|
| 419 |
+
"grad_norm": 0.32942134141921997,
|
| 420 |
+
"learning_rate": 0.00019460608811155508,
|
| 421 |
+
"loss": 0.0492,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.5544041450777202,
|
| 426 |
+
"grad_norm": 0.2994638979434967,
|
| 427 |
+
"learning_rate": 0.00019433075262108678,
|
| 428 |
+
"loss": 0.0471,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.5803108808290154,
|
| 433 |
+
"grad_norm": 0.3796844482421875,
|
| 434 |
+
"learning_rate": 0.00019404876833584923,
|
| 435 |
+
"loss": 0.0457,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.6062176165803108,
|
| 440 |
+
"grad_norm": 0.30895495414733887,
|
| 441 |
+
"learning_rate": 0.00019376015513118092,
|
| 442 |
+
"loss": 0.0475,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.6321243523316062,
|
| 447 |
+
"grad_norm": 0.3315885663032532,
|
| 448 |
+
"learning_rate": 0.00019346493334965217,
|
| 449 |
+
"loss": 0.0461,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.6580310880829017,
|
| 454 |
+
"grad_norm": 0.4402955174446106,
|
| 455 |
+
"learning_rate": 0.00019316312379963118,
|
| 456 |
+
"loss": 0.0502,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.6839378238341969,
|
| 461 |
+
"grad_norm": 0.46544697880744934,
|
| 462 |
+
"learning_rate": 0.00019285474775381759,
|
| 463 |
+
"loss": 0.0487,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.709844559585492,
|
| 468 |
+
"grad_norm": 0.48502856492996216,
|
| 469 |
+
"learning_rate": 0.0001925398269477427,
|
| 470 |
+
"loss": 0.0429,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 1.7357512953367875,
|
| 475 |
+
"grad_norm": 0.23488780856132507,
|
| 476 |
+
"learning_rate": 0.000192218383578238,
|
| 477 |
+
"loss": 0.0448,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 1.761658031088083,
|
| 482 |
+
"grad_norm": 0.3346535265445709,
|
| 483 |
+
"learning_rate": 0.00019189044030187017,
|
| 484 |
+
"loss": 0.0424,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.7875647668393784,
|
| 489 |
+
"grad_norm": 0.31202977895736694,
|
| 490 |
+
"learning_rate": 0.00019155602023334444,
|
| 491 |
+
"loss": 0.037,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 1.8134715025906736,
|
| 496 |
+
"grad_norm": 0.2789950966835022,
|
| 497 |
+
"learning_rate": 0.0001912151469438753,
|
| 498 |
+
"loss": 0.0394,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 1.8393782383419688,
|
| 503 |
+
"grad_norm": 0.3802800476551056,
|
| 504 |
+
"learning_rate": 0.00019086784445952515,
|
| 505 |
+
"loss": 0.0442,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.8652849740932642,
|
| 510 |
+
"grad_norm": 0.26964104175567627,
|
| 511 |
+
"learning_rate": 0.0001905141372595107,
|
| 512 |
+
"loss": 0.0409,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 1.8911917098445596,
|
| 517 |
+
"grad_norm": 0.35952338576316833,
|
| 518 |
+
"learning_rate": 0.00019015405027447777,
|
| 519 |
+
"loss": 0.0448,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.917098445595855,
|
| 524 |
+
"grad_norm": 0.32625675201416016,
|
| 525 |
+
"learning_rate": 0.00018978760888474397,
|
| 526 |
+
"loss": 0.0456,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.9430051813471503,
|
| 531 |
+
"grad_norm": 0.37657493352890015,
|
| 532 |
+
"learning_rate": 0.00018941483891850983,
|
| 533 |
+
"loss": 0.0444,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 1.9689119170984455,
|
| 538 |
+
"grad_norm": 0.31168240308761597,
|
| 539 |
+
"learning_rate": 0.00018903576665003832,
|
| 540 |
+
"loss": 0.0418,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 1.994818652849741,
|
| 545 |
+
"grad_norm": 0.263838529586792,
|
| 546 |
+
"learning_rate": 0.000188650418797803,
|
| 547 |
+
"loss": 0.0458,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.0207253886010363,
|
| 552 |
+
"grad_norm": 0.32005709409713745,
|
| 553 |
+
"learning_rate": 0.00018825882252260467,
|
| 554 |
+
"loss": 0.0444,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.0466321243523318,
|
| 559 |
+
"grad_norm": 0.3012694716453552,
|
| 560 |
+
"learning_rate": 0.0001878610054256571,
|
| 561 |
+
"loss": 0.0366,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.0725388601036268,
|
| 566 |
+
"grad_norm": 0.2204688936471939,
|
| 567 |
+
"learning_rate": 0.00018745699554664148,
|
| 568 |
+
"loss": 0.0389,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.098445595854922,
|
| 573 |
+
"grad_norm": 0.28166186809539795,
|
| 574 |
+
"learning_rate": 0.0001870468213617302,
|
| 575 |
+
"loss": 0.0377,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.1243523316062176,
|
| 580 |
+
"grad_norm": 0.34278395771980286,
|
| 581 |
+
"learning_rate": 0.00018663051178157965,
|
| 582 |
+
"loss": 0.0385,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.150259067357513,
|
| 587 |
+
"grad_norm": 0.29428401589393616,
|
| 588 |
+
"learning_rate": 0.00018620809614929243,
|
| 589 |
+
"loss": 0.0427,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.1761658031088085,
|
| 594 |
+
"grad_norm": 0.3556409478187561,
|
| 595 |
+
"learning_rate": 0.00018577960423834936,
|
| 596 |
+
"loss": 0.0412,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.2020725388601035,
|
| 601 |
+
"grad_norm": 0.20572839677333832,
|
| 602 |
+
"learning_rate": 0.00018534506625051067,
|
| 603 |
+
"loss": 0.0386,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.227979274611399,
|
| 608 |
+
"grad_norm": 0.30463582277297974,
|
| 609 |
+
"learning_rate": 0.0001849045128136874,
|
| 610 |
+
"loss": 0.042,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.2538860103626943,
|
| 615 |
+
"grad_norm": 0.2660544514656067,
|
| 616 |
+
"learning_rate": 0.00018445797497978266,
|
| 617 |
+
"loss": 0.0382,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.2797927461139897,
|
| 622 |
+
"grad_norm": 0.2419629544019699,
|
| 623 |
+
"learning_rate": 0.00018400548422250278,
|
| 624 |
+
"loss": 0.0436,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.305699481865285,
|
| 629 |
+
"grad_norm": 0.2734587788581848,
|
| 630 |
+
"learning_rate": 0.00018354707243513926,
|
| 631 |
+
"loss": 0.0414,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.33160621761658,
|
| 636 |
+
"grad_norm": 0.3013145625591278,
|
| 637 |
+
"learning_rate": 0.00018308277192832038,
|
| 638 |
+
"loss": 0.0424,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.3575129533678756,
|
| 643 |
+
"grad_norm": 0.23408107459545135,
|
| 644 |
+
"learning_rate": 0.0001826126154277343,
|
| 645 |
+
"loss": 0.0381,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.383419689119171,
|
| 650 |
+
"grad_norm": 0.24882006645202637,
|
| 651 |
+
"learning_rate": 0.0001821366360718219,
|
| 652 |
+
"loss": 0.0361,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 2.4093264248704664,
|
| 657 |
+
"grad_norm": 0.3461746573448181,
|
| 658 |
+
"learning_rate": 0.0001816548674094415,
|
| 659 |
+
"loss": 0.0412,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 2.4352331606217614,
|
| 664 |
+
"grad_norm": 0.3645137548446655,
|
| 665 |
+
"learning_rate": 0.00018116734339750407,
|
| 666 |
+
"loss": 0.0411,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 2.461139896373057,
|
| 671 |
+
"grad_norm": 0.5230131149291992,
|
| 672 |
+
"learning_rate": 0.0001806740983985797,
|
| 673 |
+
"loss": 0.0413,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.4870466321243523,
|
| 678 |
+
"grad_norm": 0.48028966784477234,
|
| 679 |
+
"learning_rate": 0.00018017516717847563,
|
| 680 |
+
"loss": 0.0413,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 2.5129533678756477,
|
| 685 |
+
"grad_norm": 0.3212891221046448,
|
| 686 |
+
"learning_rate": 0.000179670584903786,
|
| 687 |
+
"loss": 0.0427,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 2.538860103626943,
|
| 692 |
+
"grad_norm": 0.3119347095489502,
|
| 693 |
+
"learning_rate": 0.00017916038713941296,
|
| 694 |
+
"loss": 0.0376,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 2.5647668393782386,
|
| 699 |
+
"grad_norm": 0.31051719188690186,
|
| 700 |
+
"learning_rate": 0.00017864460984606,
|
| 701 |
+
"loss": 0.0463,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 2.5906735751295336,
|
| 706 |
+
"grad_norm": 0.3582245111465454,
|
| 707 |
+
"learning_rate": 0.0001781232893776974,
|
| 708 |
+
"loss": 0.0351,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 2.616580310880829,
|
| 713 |
+
"grad_norm": 0.28858813643455505,
|
| 714 |
+
"learning_rate": 0.0001775964624789996,
|
| 715 |
+
"loss": 0.0393,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 2.6424870466321244,
|
| 720 |
+
"grad_norm": 0.21984124183654785,
|
| 721 |
+
"learning_rate": 0.00017706416628275563,
|
| 722 |
+
"loss": 0.0432,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 2.66839378238342,
|
| 727 |
+
"grad_norm": 0.32943961024284363,
|
| 728 |
+
"learning_rate": 0.0001765264383072515,
|
| 729 |
+
"loss": 0.0354,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 2.694300518134715,
|
| 734 |
+
"grad_norm": 0.16112935543060303,
|
| 735 |
+
"learning_rate": 0.0001759833164536261,
|
| 736 |
+
"loss": 0.035,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 2.7202072538860103,
|
| 741 |
+
"grad_norm": 0.27633556723594666,
|
| 742 |
+
"learning_rate": 0.0001754348390031996,
|
| 743 |
+
"loss": 0.0344,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 2.7461139896373057,
|
| 748 |
+
"grad_norm": 0.29028812050819397,
|
| 749 |
+
"learning_rate": 0.0001748810446147751,
|
| 750 |
+
"loss": 0.0332,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 2.772020725388601,
|
| 755 |
+
"grad_norm": 0.268162339925766,
|
| 756 |
+
"learning_rate": 0.00017432197232191424,
|
| 757 |
+
"loss": 0.0381,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 2.7979274611398965,
|
| 762 |
+
"grad_norm": 0.207249715924263,
|
| 763 |
+
"learning_rate": 0.00017375766153018542,
|
| 764 |
+
"loss": 0.0403,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 2.823834196891192,
|
| 769 |
+
"grad_norm": 0.26709848642349243,
|
| 770 |
+
"learning_rate": 0.00017318815201438682,
|
| 771 |
+
"loss": 0.035,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 2.849740932642487,
|
| 776 |
+
"grad_norm": 0.3294147253036499,
|
| 777 |
+
"learning_rate": 0.00017261348391574258,
|
| 778 |
+
"loss": 0.0353,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 2.8756476683937824,
|
| 783 |
+
"grad_norm": 0.34221452474594116,
|
| 784 |
+
"learning_rate": 0.00017203369773907375,
|
| 785 |
+
"loss": 0.0323,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 2.901554404145078,
|
| 790 |
+
"grad_norm": 0.23212362825870514,
|
| 791 |
+
"learning_rate": 0.00017144883434994313,
|
| 792 |
+
"loss": 0.0348,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 2.927461139896373,
|
| 797 |
+
"grad_norm": 0.3442625105381012,
|
| 798 |
+
"learning_rate": 0.00017085893497177502,
|
| 799 |
+
"loss": 0.0384,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 2.9533678756476682,
|
| 804 |
+
"grad_norm": 0.3061060607433319,
|
| 805 |
+
"learning_rate": 0.0001702640411829497,
|
| 806 |
+
"loss": 0.0364,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 2.9792746113989637,
|
| 811 |
+
"grad_norm": 0.2969331443309784,
|
| 812 |
+
"learning_rate": 0.00016966419491387267,
|
| 813 |
+
"loss": 0.0345,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 3.005181347150259,
|
| 818 |
+
"grad_norm": 0.2101631909608841,
|
| 819 |
+
"learning_rate": 0.0001690594384440194,
|
| 820 |
+
"loss": 0.0336,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 3.0310880829015545,
|
| 825 |
+
"grad_norm": 0.2539927661418915,
|
| 826 |
+
"learning_rate": 0.00016844981439895505,
|
| 827 |
+
"loss": 0.0381,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 3.05699481865285,
|
| 832 |
+
"grad_norm": 0.26166775822639465,
|
| 833 |
+
"learning_rate": 0.00016783536574733047,
|
| 834 |
+
"loss": 0.037,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 3.082901554404145,
|
| 839 |
+
"grad_norm": 0.24203582108020782,
|
| 840 |
+
"learning_rate": 0.00016721613579785317,
|
| 841 |
+
"loss": 0.0323,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 3.1088082901554404,
|
| 846 |
+
"grad_norm": 0.25355836749076843,
|
| 847 |
+
"learning_rate": 0.000166592168196235,
|
| 848 |
+
"loss": 0.0375,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 3.134715025906736,
|
| 853 |
+
"grad_norm": 0.25193431973457336,
|
| 854 |
+
"learning_rate": 0.0001659635069221159,
|
| 855 |
+
"loss": 0.0308,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 3.160621761658031,
|
| 860 |
+
"grad_norm": 0.3023054599761963,
|
| 861 |
+
"learning_rate": 0.00016533019628596368,
|
| 862 |
+
"loss": 0.0331,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 3.186528497409326,
|
| 867 |
+
"grad_norm": 0.23343996703624725,
|
| 868 |
+
"learning_rate": 0.00016469228092595137,
|
| 869 |
+
"loss": 0.0322,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 3.2124352331606216,
|
| 874 |
+
"grad_norm": 0.21372058987617493,
|
| 875 |
+
"learning_rate": 0.00016404980580481042,
|
| 876 |
+
"loss": 0.0328,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 3.238341968911917,
|
| 881 |
+
"grad_norm": 0.23146772384643555,
|
| 882 |
+
"learning_rate": 0.00016340281620666204,
|
| 883 |
+
"loss": 0.032,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 3.2642487046632125,
|
| 888 |
+
"grad_norm": 0.2798649072647095,
|
| 889 |
+
"learning_rate": 0.00016275135773382494,
|
| 890 |
+
"loss": 0.0338,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 3.290155440414508,
|
| 895 |
+
"grad_norm": 0.1371963620185852,
|
| 896 |
+
"learning_rate": 0.00016209547630360155,
|
| 897 |
+
"loss": 0.0307,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 3.3160621761658033,
|
| 902 |
+
"grad_norm": 0.3256252408027649,
|
| 903 |
+
"learning_rate": 0.00016143521814504118,
|
| 904 |
+
"loss": 0.0371,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 3.3419689119170983,
|
| 909 |
+
"grad_norm": 0.23841211199760437,
|
| 910 |
+
"learning_rate": 0.00016077062979568196,
|
| 911 |
+
"loss": 0.0327,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 3.3678756476683938,
|
| 916 |
+
"grad_norm": 0.19018740952014923,
|
| 917 |
+
"learning_rate": 0.00016010175809827047,
|
| 918 |
+
"loss": 0.0344,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 3.393782383419689,
|
| 923 |
+
"grad_norm": 0.20630723237991333,
|
| 924 |
+
"learning_rate": 0.00015942865019746014,
|
| 925 |
+
"loss": 0.0326,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 3.4196891191709846,
|
| 930 |
+
"grad_norm": 0.29622000455856323,
|
| 931 |
+
"learning_rate": 0.0001587513535364884,
|
| 932 |
+
"loss": 0.0287,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 3.4455958549222796,
|
| 937 |
+
"grad_norm": 0.22663627564907074,
|
| 938 |
+
"learning_rate": 0.00015806991585383257,
|
| 939 |
+
"loss": 0.0347,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 3.471502590673575,
|
| 944 |
+
"grad_norm": 0.1253904402256012,
|
| 945 |
+
"learning_rate": 0.00015738438517984513,
|
| 946 |
+
"loss": 0.0328,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 3.4974093264248705,
|
| 951 |
+
"grad_norm": 0.3609561622142792,
|
| 952 |
+
"learning_rate": 0.00015669480983336842,
|
| 953 |
+
"loss": 0.0315,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 3.523316062176166,
|
| 958 |
+
"grad_norm": 0.2557467222213745,
|
| 959 |
+
"learning_rate": 0.0001560012384183288,
|
| 960 |
+
"loss": 0.0314,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 3.5492227979274613,
|
| 965 |
+
"grad_norm": 0.16117317974567413,
|
| 966 |
+
"learning_rate": 0.00015530371982031097,
|
| 967 |
+
"loss": 0.0317,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 3.5751295336787567,
|
| 972 |
+
"grad_norm": 0.21727634966373444,
|
| 973 |
+
"learning_rate": 0.0001546023032031123,
|
| 974 |
+
"loss": 0.0295,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 3.6010362694300517,
|
| 979 |
+
"grad_norm": 0.22786779701709747,
|
| 980 |
+
"learning_rate": 0.00015389703800527756,
|
| 981 |
+
"loss": 0.0307,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 3.626943005181347,
|
| 986 |
+
"grad_norm": 0.28124475479125977,
|
| 987 |
+
"learning_rate": 0.0001531879739366142,
|
| 988 |
+
"loss": 0.0314,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 3.6528497409326426,
|
| 993 |
+
"grad_norm": 0.20782004296779633,
|
| 994 |
+
"learning_rate": 0.00015247516097468893,
|
| 995 |
+
"loss": 0.0292,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 3.6787564766839376,
|
| 1000 |
+
"grad_norm": 0.23655350506305695,
|
| 1001 |
+
"learning_rate": 0.00015175864936130466,
|
| 1002 |
+
"loss": 0.0314,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 3.704663212435233,
|
| 1007 |
+
"grad_norm": 0.2071056067943573,
|
| 1008 |
+
"learning_rate": 0.00015103848959895978,
|
| 1009 |
+
"loss": 0.031,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 3.7305699481865284,
|
| 1014 |
+
"grad_norm": 0.2913356423377991,
|
| 1015 |
+
"learning_rate": 0.00015031473244728793,
|
| 1016 |
+
"loss": 0.0302,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 3.756476683937824,
|
| 1021 |
+
"grad_norm": 0.2031751126050949,
|
| 1022 |
+
"learning_rate": 0.00014958742891948092,
|
| 1023 |
+
"loss": 0.0317,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 3.7823834196891193,
|
| 1028 |
+
"grad_norm": 0.3000684380531311,
|
| 1029 |
+
"learning_rate": 0.0001488566302786926,
|
| 1030 |
+
"loss": 0.0343,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 3.8082901554404147,
|
| 1035 |
+
"grad_norm": 0.27631327509880066,
|
| 1036 |
+
"learning_rate": 0.00014812238803442598,
|
| 1037 |
+
"loss": 0.0317,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 3.8341968911917097,
|
| 1042 |
+
"grad_norm": 0.3157835602760315,
|
| 1043 |
+
"learning_rate": 0.00014738475393890235,
|
| 1044 |
+
"loss": 0.033,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 3.860103626943005,
|
| 1049 |
+
"grad_norm": 0.29543039202690125,
|
| 1050 |
+
"learning_rate": 0.00014664377998341391,
|
| 1051 |
+
"loss": 0.031,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 3.8860103626943006,
|
| 1056 |
+
"grad_norm": 0.21288472414016724,
|
| 1057 |
+
"learning_rate": 0.000145899518394659,
|
| 1058 |
+
"loss": 0.032,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 3.911917098445596,
|
| 1063 |
+
"grad_norm": 0.257420152425766,
|
| 1064 |
+
"learning_rate": 0.00014515202163106088,
|
| 1065 |
+
"loss": 0.0327,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 3.937823834196891,
|
| 1070 |
+
"grad_norm": 0.23928193747997284,
|
| 1071 |
+
"learning_rate": 0.00014440134237907065,
|
| 1072 |
+
"loss": 0.0299,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 3.9637305699481864,
|
| 1077 |
+
"grad_norm": 0.22903333604335785,
|
| 1078 |
+
"learning_rate": 0.0001436475335494532,
|
| 1079 |
+
"loss": 0.0325,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 3.989637305699482,
|
| 1084 |
+
"grad_norm": 0.2012854814529419,
|
| 1085 |
+
"learning_rate": 0.0001428906482735583,
|
| 1086 |
+
"loss": 0.0309,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 4.015544041450777,
|
| 1091 |
+
"grad_norm": 0.2412451058626175,
|
| 1092 |
+
"learning_rate": 0.00014213073989957533,
|
| 1093 |
+
"loss": 0.0313,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 4.041450777202073,
|
| 1098 |
+
"grad_norm": 0.22772116959095,
|
| 1099 |
+
"learning_rate": 0.0001413678619887734,
|
| 1100 |
+
"loss": 0.0334,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 4.067357512953368,
|
| 1105 |
+
"grad_norm": 0.22949784994125366,
|
| 1106 |
+
"learning_rate": 0.000140602068311726,
|
| 1107 |
+
"loss": 0.0308,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 4.0932642487046635,
|
| 1112 |
+
"grad_norm": 0.21911077201366425,
|
| 1113 |
+
"learning_rate": 0.00013983341284452093,
|
| 1114 |
+
"loss": 0.0263,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 4.119170984455959,
|
| 1119 |
+
"grad_norm": 0.19424454867839813,
|
| 1120 |
+
"learning_rate": 0.00013906194976495616,
|
| 1121 |
+
"loss": 0.0317,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 4.1450777202072535,
|
| 1126 |
+
"grad_norm": 0.23268303275108337,
|
| 1127 |
+
"learning_rate": 0.00013828773344872082,
|
| 1128 |
+
"loss": 0.0306,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 4.170984455958549,
|
| 1133 |
+
"grad_norm": 0.253647118806839,
|
| 1134 |
+
"learning_rate": 0.000137510818465563,
|
| 1135 |
+
"loss": 0.029,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 4.196891191709844,
|
| 1140 |
+
"grad_norm": 0.15061113238334656,
|
| 1141 |
+
"learning_rate": 0.00013673125957544286,
|
| 1142 |
+
"loss": 0.0248,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 4.22279792746114,
|
| 1147 |
+
"grad_norm": 0.2819027602672577,
|
| 1148 |
+
"learning_rate": 0.00013594911172467374,
|
| 1149 |
+
"loss": 0.0308,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 4.248704663212435,
|
| 1154 |
+
"grad_norm": 0.24925965070724487,
|
| 1155 |
+
"learning_rate": 0.00013516443004204853,
|
| 1156 |
+
"loss": 0.027,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 4.274611398963731,
|
| 1161 |
+
"grad_norm": 0.2695538401603699,
|
| 1162 |
+
"learning_rate": 0.00013437726983495462,
|
| 1163 |
+
"loss": 0.0286,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 4.300518134715026,
|
| 1168 |
+
"grad_norm": 0.21748502552509308,
|
| 1169 |
+
"learning_rate": 0.00013358768658547517,
|
| 1170 |
+
"loss": 0.0263,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 4.3264248704663215,
|
| 1175 |
+
"grad_norm": 0.19919273257255554,
|
| 1176 |
+
"learning_rate": 0.00013279573594647873,
|
| 1177 |
+
"loss": 0.0286,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 4.352331606217617,
|
| 1182 |
+
"grad_norm": 0.29898253083229065,
|
| 1183 |
+
"learning_rate": 0.0001320014737376967,
|
| 1184 |
+
"loss": 0.0299,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 4.3782383419689115,
|
| 1189 |
+
"grad_norm": 0.26166197657585144,
|
| 1190 |
+
"learning_rate": 0.00013120495594178867,
|
| 1191 |
+
"loss": 0.0306,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 4.404145077720207,
|
| 1196 |
+
"grad_norm": 0.23769961297512054,
|
| 1197 |
+
"learning_rate": 0.0001304062387003968,
|
| 1198 |
+
"loss": 0.0333,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 4.430051813471502,
|
| 1203 |
+
"grad_norm": 0.2202085703611374,
|
| 1204 |
+
"learning_rate": 0.00012960537831018863,
|
| 1205 |
+
"loss": 0.032,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 4.455958549222798,
|
| 1210 |
+
"grad_norm": 0.2457539141178131,
|
| 1211 |
+
"learning_rate": 0.00012880243121888898,
|
| 1212 |
+
"loss": 0.0262,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 4.481865284974093,
|
| 1217 |
+
"grad_norm": 0.2208835929632187,
|
| 1218 |
+
"learning_rate": 0.0001279974540213015,
|
| 1219 |
+
"loss": 0.0311,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 4.507772020725389,
|
| 1224 |
+
"grad_norm": 0.19063198566436768,
|
| 1225 |
+
"learning_rate": 0.0001271905034553196,
|
| 1226 |
+
"loss": 0.0269,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 4.533678756476684,
|
| 1231 |
+
"grad_norm": 0.2229800969362259,
|
| 1232 |
+
"learning_rate": 0.00012638163639792715,
|
| 1233 |
+
"loss": 0.0292,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 4.5595854922279795,
|
| 1238 |
+
"grad_norm": 0.16816496849060059,
|
| 1239 |
+
"learning_rate": 0.0001255709098611898,
|
| 1240 |
+
"loss": 0.0266,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 4.585492227979275,
|
| 1245 |
+
"grad_norm": 0.22993479669094086,
|
| 1246 |
+
"learning_rate": 0.0001247583809882365,
|
| 1247 |
+
"loss": 0.0266,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 4.61139896373057,
|
| 1252 |
+
"grad_norm": 0.2233438789844513,
|
| 1253 |
+
"learning_rate": 0.00012394410704923167,
|
| 1254 |
+
"loss": 0.0303,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 4.637305699481866,
|
| 1259 |
+
"grad_norm": 0.2839880883693695,
|
| 1260 |
+
"learning_rate": 0.0001231281454373389,
|
| 1261 |
+
"loss": 0.0298,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 4.66321243523316,
|
| 1266 |
+
"grad_norm": 0.22099633514881134,
|
| 1267 |
+
"learning_rate": 0.0001223105536646752,
|
| 1268 |
+
"loss": 0.0302,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 4.689119170984456,
|
| 1273 |
+
"grad_norm": 0.18007872998714447,
|
| 1274 |
+
"learning_rate": 0.00012149138935825792,
|
| 1275 |
+
"loss": 0.0242,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 4.715025906735751,
|
| 1280 |
+
"grad_norm": 0.2904217541217804,
|
| 1281 |
+
"learning_rate": 0.00012067071025594234,
|
| 1282 |
+
"loss": 0.0284,
|
| 1283 |
+
"step": 1820
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 4.740932642487047,
|
| 1287 |
+
"grad_norm": 0.2809178829193115,
|
| 1288 |
+
"learning_rate": 0.00011984857420235263,
|
| 1289 |
+
"loss": 0.0277,
|
| 1290 |
+
"step": 1830
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 4.766839378238342,
|
| 1294 |
+
"grad_norm": 0.23288775980472565,
|
| 1295 |
+
"learning_rate": 0.00011902503914480428,
|
| 1296 |
+
"loss": 0.0284,
|
| 1297 |
+
"step": 1840
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 4.7927461139896375,
|
| 1301 |
+
"grad_norm": 0.1514245569705963,
|
| 1302 |
+
"learning_rate": 0.00011820016312922015,
|
| 1303 |
+
"loss": 0.031,
|
| 1304 |
+
"step": 1850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 4.818652849740933,
|
| 1308 |
+
"grad_norm": 0.12524273991584778,
|
| 1309 |
+
"learning_rate": 0.00011737400429603888,
|
| 1310 |
+
"loss": 0.031,
|
| 1311 |
+
"step": 1860
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 4.844559585492228,
|
| 1315 |
+
"grad_norm": 0.22305960953235626,
|
| 1316 |
+
"learning_rate": 0.00011654662087611707,
|
| 1317 |
+
"loss": 0.0294,
|
| 1318 |
+
"step": 1870
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 4.870466321243523,
|
| 1322 |
+
"grad_norm": 0.2521384358406067,
|
| 1323 |
+
"learning_rate": 0.00011571807118662492,
|
| 1324 |
+
"loss": 0.0276,
|
| 1325 |
+
"step": 1880
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 4.896373056994818,
|
| 1329 |
+
"grad_norm": 0.180227592587471,
|
| 1330 |
+
"learning_rate": 0.00011488841362693576,
|
| 1331 |
+
"loss": 0.0289,
|
| 1332 |
+
"step": 1890
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 4.922279792746114,
|
| 1336 |
+
"grad_norm": 0.22024022042751312,
|
| 1337 |
+
"learning_rate": 0.00011405770667450996,
|
| 1338 |
+
"loss": 0.0266,
|
| 1339 |
+
"step": 1900
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 4.948186528497409,
|
| 1343 |
+
"grad_norm": 0.19573189318180084,
|
| 1344 |
+
"learning_rate": 0.00011322600888077303,
|
| 1345 |
+
"loss": 0.0282,
|
| 1346 |
+
"step": 1910
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 4.974093264248705,
|
| 1350 |
+
"grad_norm": 0.1752632111310959,
|
| 1351 |
+
"learning_rate": 0.00011239337886698887,
|
| 1352 |
+
"loss": 0.0269,
|
| 1353 |
+
"step": 1920
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 5.0,
|
| 1357 |
+
"grad_norm": 0.2073243260383606,
|
| 1358 |
+
"learning_rate": 0.00011155987532012779,
|
| 1359 |
+
"loss": 0.0268,
|
| 1360 |
+
"step": 1930
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 5.025906735751295,
|
| 1364 |
+
"grad_norm": 0.21750442683696747,
|
| 1365 |
+
"learning_rate": 0.00011072555698873013,
|
| 1366 |
+
"loss": 0.0249,
|
| 1367 |
+
"step": 1940
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 5.051813471502591,
|
| 1371 |
+
"grad_norm": 0.1471615731716156,
|
| 1372 |
+
"learning_rate": 0.00010989048267876529,
|
| 1373 |
+
"loss": 0.0231,
|
| 1374 |
+
"step": 1950
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 5.077720207253886,
|
| 1378 |
+
"grad_norm": 0.21576045453548431,
|
| 1379 |
+
"learning_rate": 0.00010905471124948707,
|
| 1380 |
+
"loss": 0.0245,
|
| 1381 |
+
"step": 1960
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 5.103626943005182,
|
| 1385 |
+
"grad_norm": 0.1472284197807312,
|
| 1386 |
+
"learning_rate": 0.00010821830160928486,
|
| 1387 |
+
"loss": 0.0247,
|
| 1388 |
+
"step": 1970
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 5.129533678756476,
|
| 1392 |
+
"grad_norm": 0.354701429605484,
|
| 1393 |
+
"learning_rate": 0.00010738131271153156,
|
| 1394 |
+
"loss": 0.0255,
|
| 1395 |
+
"step": 1980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 5.155440414507772,
|
| 1399 |
+
"grad_norm": 0.16570764780044556,
|
| 1400 |
+
"learning_rate": 0.00010654380355042852,
|
| 1401 |
+
"loss": 0.0236,
|
| 1402 |
+
"step": 1990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 5.181347150259067,
|
| 1406 |
+
"grad_norm": 0.20196418464183807,
|
| 1407 |
+
"learning_rate": 0.00010570583315684708,
|
| 1408 |
+
"loss": 0.0278,
|
| 1409 |
+
"step": 2000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 5.2072538860103625,
|
| 1413 |
+
"grad_norm": 0.25251758098602295,
|
| 1414 |
+
"learning_rate": 0.00010486746059416816,
|
| 1415 |
+
"loss": 0.0258,
|
| 1416 |
+
"step": 2010
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 5.233160621761658,
|
| 1420 |
+
"grad_norm": 0.2057216763496399,
|
| 1421 |
+
"learning_rate": 0.00010402874495411902,
|
| 1422 |
+
"loss": 0.0287,
|
| 1423 |
+
"step": 2020
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 5.259067357512953,
|
| 1427 |
+
"grad_norm": 0.27119141817092896,
|
| 1428 |
+
"learning_rate": 0.00010318974535260844,
|
| 1429 |
+
"loss": 0.0252,
|
| 1430 |
+
"step": 2030
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 5.284974093264249,
|
| 1434 |
+
"grad_norm": 0.3158484995365143,
|
| 1435 |
+
"learning_rate": 0.00010235052092555972,
|
| 1436 |
+
"loss": 0.0254,
|
| 1437 |
+
"step": 2040
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 5.310880829015544,
|
| 1441 |
+
"grad_norm": 0.2271166890859604,
|
| 1442 |
+
"learning_rate": 0.00010151113082474295,
|
| 1443 |
+
"loss": 0.0276,
|
| 1444 |
+
"step": 2050
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 5.33678756476684,
|
| 1448 |
+
"grad_norm": 0.28330063819885254,
|
| 1449 |
+
"learning_rate": 0.00010067163421360535,
|
| 1450 |
+
"loss": 0.025,
|
| 1451 |
+
"step": 2060
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 5.362694300518135,
|
| 1455 |
+
"grad_norm": 0.1780189722776413,
|
| 1456 |
+
"learning_rate": 9.983209026310149e-05,
|
| 1457 |
+
"loss": 0.0261,
|
| 1458 |
+
"step": 2070
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 5.38860103626943,
|
| 1462 |
+
"grad_norm": 0.22980116307735443,
|
| 1463 |
+
"learning_rate": 9.899255814752263e-05,
|
| 1464 |
+
"loss": 0.0224,
|
| 1465 |
+
"step": 2080
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 5.414507772020725,
|
| 1469 |
+
"grad_norm": 0.21145443618297577,
|
| 1470 |
+
"learning_rate": 9.815309704032583e-05,
|
| 1471 |
+
"loss": 0.0276,
|
| 1472 |
+
"step": 2090
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 5.4404145077720205,
|
| 1476 |
+
"grad_norm": 0.18889708817005157,
|
| 1477 |
+
"learning_rate": 9.731376610996313e-05,
|
| 1478 |
+
"loss": 0.0284,
|
| 1479 |
+
"step": 2100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 5.466321243523316,
|
| 1483 |
+
"grad_norm": 0.1767049878835678,
|
| 1484 |
+
"learning_rate": 9.647462451571127e-05,
|
| 1485 |
+
"loss": 0.0253,
|
| 1486 |
+
"step": 2110
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 5.492227979274611,
|
| 1490 |
+
"grad_norm": 0.16297735273838043,
|
| 1491 |
+
"learning_rate": 9.563573140350192e-05,
|
| 1492 |
+
"loss": 0.0232,
|
| 1493 |
+
"step": 2120
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 5.518134715025907,
|
| 1497 |
+
"grad_norm": 0.13897372782230377,
|
| 1498 |
+
"learning_rate": 9.47971459017527e-05,
|
| 1499 |
+
"loss": 0.0252,
|
| 1500 |
+
"step": 2130
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 5.544041450777202,
|
| 1504 |
+
"grad_norm": 0.2665773630142212,
|
| 1505 |
+
"learning_rate": 9.395892711719962e-05,
|
| 1506 |
+
"loss": 0.0234,
|
| 1507 |
+
"step": 2140
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 5.569948186528498,
|
| 1511 |
+
"grad_norm": 0.2184874713420868,
|
| 1512 |
+
"learning_rate": 9.312113413073113e-05,
|
| 1513 |
+
"loss": 0.0226,
|
| 1514 |
+
"step": 2150
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 5.595854922279793,
|
| 1518 |
+
"grad_norm": 0.24830220639705658,
|
| 1519 |
+
"learning_rate": 9.228382599322374e-05,
|
| 1520 |
+
"loss": 0.0232,
|
| 1521 |
+
"step": 2160
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 5.6217616580310885,
|
| 1525 |
+
"grad_norm": 0.1857186257839203,
|
| 1526 |
+
"learning_rate": 9.144706172138002e-05,
|
| 1527 |
+
"loss": 0.0204,
|
| 1528 |
+
"step": 2170
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 5.647668393782383,
|
| 1532 |
+
"grad_norm": 0.18585319817066193,
|
| 1533 |
+
"learning_rate": 9.061090029356867e-05,
|
| 1534 |
+
"loss": 0.0208,
|
| 1535 |
+
"step": 2180
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 5.6735751295336785,
|
| 1539 |
+
"grad_norm": 0.19226165115833282,
|
| 1540 |
+
"learning_rate": 8.977540064566771e-05,
|
| 1541 |
+
"loss": 0.0225,
|
| 1542 |
+
"step": 2190
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 5.699481865284974,
|
| 1546 |
+
"grad_norm": 0.2323295623064041,
|
| 1547 |
+
"learning_rate": 8.89406216669103e-05,
|
| 1548 |
+
"loss": 0.0276,
|
| 1549 |
+
"step": 2200
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 5.725388601036269,
|
| 1553 |
+
"grad_norm": 0.17143680155277252,
|
| 1554 |
+
"learning_rate": 8.810662219573421e-05,
|
| 1555 |
+
"loss": 0.0284,
|
| 1556 |
+
"step": 2210
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 5.751295336787565,
|
| 1560 |
+
"grad_norm": 0.21579648554325104,
|
| 1561 |
+
"learning_rate": 8.727346101563425e-05,
|
| 1562 |
+
"loss": 0.0262,
|
| 1563 |
+
"step": 2220
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 5.77720207253886,
|
| 1567 |
+
"grad_norm": 0.21625088155269623,
|
| 1568 |
+
"learning_rate": 8.64411968510195e-05,
|
| 1569 |
+
"loss": 0.0236,
|
| 1570 |
+
"step": 2230
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 5.803108808290156,
|
| 1574 |
+
"grad_norm": 0.25434359908103943,
|
| 1575 |
+
"learning_rate": 8.56098883630738e-05,
|
| 1576 |
+
"loss": 0.0234,
|
| 1577 |
+
"step": 2240
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 5.829015544041451,
|
| 1581 |
+
"grad_norm": 0.26135268807411194,
|
| 1582 |
+
"learning_rate": 8.477959414562126e-05,
|
| 1583 |
+
"loss": 0.024,
|
| 1584 |
+
"step": 2250
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 5.8549222797927465,
|
| 1588 |
+
"grad_norm": 0.2323196679353714,
|
| 1589 |
+
"learning_rate": 8.395037272099629e-05,
|
| 1590 |
+
"loss": 0.0274,
|
| 1591 |
+
"step": 2260
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 5.880829015544041,
|
| 1595 |
+
"grad_norm": 0.17564794421195984,
|
| 1596 |
+
"learning_rate": 8.312228253591867e-05,
|
| 1597 |
+
"loss": 0.024,
|
| 1598 |
+
"step": 2270
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 5.9067357512953365,
|
| 1602 |
+
"grad_norm": 0.1802438646554947,
|
| 1603 |
+
"learning_rate": 8.22953819573742e-05,
|
| 1604 |
+
"loss": 0.0224,
|
| 1605 |
+
"step": 2280
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 5.932642487046632,
|
| 1609 |
+
"grad_norm": 0.2152315080165863,
|
| 1610 |
+
"learning_rate": 8.146972926850054e-05,
|
| 1611 |
+
"loss": 0.0234,
|
| 1612 |
+
"step": 2290
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 5.958549222797927,
|
| 1616 |
+
"grad_norm": 0.2047027349472046,
|
| 1617 |
+
"learning_rate": 8.064538266447926e-05,
|
| 1618 |
+
"loss": 0.0235,
|
| 1619 |
+
"step": 2300
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 5.984455958549223,
|
| 1623 |
+
"grad_norm": 0.20742332935333252,
|
| 1624 |
+
"learning_rate": 7.982240024843404e-05,
|
| 1625 |
+
"loss": 0.027,
|
| 1626 |
+
"step": 2310
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 6.010362694300518,
|
| 1630 |
+
"grad_norm": 0.15848736464977264,
|
| 1631 |
+
"learning_rate": 7.900084002733541e-05,
|
| 1632 |
+
"loss": 0.0216,
|
| 1633 |
+
"step": 2320
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 6.036269430051814,
|
| 1637 |
+
"grad_norm": 0.20834903419017792,
|
| 1638 |
+
"learning_rate": 7.818075990791208e-05,
|
| 1639 |
+
"loss": 0.0276,
|
| 1640 |
+
"step": 2330
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 6.062176165803109,
|
| 1644 |
+
"grad_norm": 0.14767374098300934,
|
| 1645 |
+
"learning_rate": 7.736221769256946e-05,
|
| 1646 |
+
"loss": 0.0223,
|
| 1647 |
+
"step": 2340
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 6.0880829015544045,
|
| 1651 |
+
"grad_norm": 0.1815556287765503,
|
| 1652 |
+
"learning_rate": 7.654527107531559e-05,
|
| 1653 |
+
"loss": 0.0233,
|
| 1654 |
+
"step": 2350
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 6.1139896373057,
|
| 1658 |
+
"grad_norm": 0.33385881781578064,
|
| 1659 |
+
"learning_rate": 7.572997763769459e-05,
|
| 1660 |
+
"loss": 0.0227,
|
| 1661 |
+
"step": 2360
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 6.139896373056994,
|
| 1665 |
+
"grad_norm": 0.1571478247642517,
|
| 1666 |
+
"learning_rate": 7.491639484472819e-05,
|
| 1667 |
+
"loss": 0.0214,
|
| 1668 |
+
"step": 2370
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 6.16580310880829,
|
| 1672 |
+
"grad_norm": 0.1373177021741867,
|
| 1673 |
+
"learning_rate": 7.410458004086519e-05,
|
| 1674 |
+
"loss": 0.0209,
|
| 1675 |
+
"step": 2380
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 6.191709844559585,
|
| 1679 |
+
"grad_norm": 0.26975080370903015,
|
| 1680 |
+
"learning_rate": 7.329459044593976e-05,
|
| 1681 |
+
"loss": 0.0243,
|
| 1682 |
+
"step": 2390
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 6.217616580310881,
|
| 1686 |
+
"grad_norm": 0.18673855066299438,
|
| 1687 |
+
"learning_rate": 7.248648315113829e-05,
|
| 1688 |
+
"loss": 0.0201,
|
| 1689 |
+
"step": 2400
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 6.243523316062176,
|
| 1693 |
+
"grad_norm": 0.1592017412185669,
|
| 1694 |
+
"learning_rate": 7.16803151149755e-05,
|
| 1695 |
+
"loss": 0.0206,
|
| 1696 |
+
"step": 2410
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 6.269430051813472,
|
| 1700 |
+
"grad_norm": 0.22650600969791412,
|
| 1701 |
+
"learning_rate": 7.08761431592795e-05,
|
| 1702 |
+
"loss": 0.0272,
|
| 1703 |
+
"step": 2420
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 6.295336787564767,
|
| 1707 |
+
"grad_norm": 0.21501009166240692,
|
| 1708 |
+
"learning_rate": 7.007402396518717e-05,
|
| 1709 |
+
"loss": 0.0183,
|
| 1710 |
+
"step": 2430
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 6.321243523316062,
|
| 1714 |
+
"grad_norm": 0.2062009572982788,
|
| 1715 |
+
"learning_rate": 6.927401406914872e-05,
|
| 1716 |
+
"loss": 0.0229,
|
| 1717 |
+
"step": 2440
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 6.347150259067358,
|
| 1721 |
+
"grad_norm": 0.1995324045419693,
|
| 1722 |
+
"learning_rate": 6.847616985894296e-05,
|
| 1723 |
+
"loss": 0.0188,
|
| 1724 |
+
"step": 2450
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 6.373056994818652,
|
| 1728 |
+
"grad_norm": 0.18677528202533722,
|
| 1729 |
+
"learning_rate": 6.768054756970279e-05,
|
| 1730 |
+
"loss": 0.0196,
|
| 1731 |
+
"step": 2460
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 6.398963730569948,
|
| 1735 |
+
"grad_norm": 0.20209306478500366,
|
| 1736 |
+
"learning_rate": 6.688720327995153e-05,
|
| 1737 |
+
"loss": 0.0199,
|
| 1738 |
+
"step": 2470
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 6.424870466321243,
|
| 1742 |
+
"grad_norm": 0.12447565048933029,
|
| 1743 |
+
"learning_rate": 6.609619290765045e-05,
|
| 1744 |
+
"loss": 0.0222,
|
| 1745 |
+
"step": 2480
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 6.450777202072539,
|
| 1749 |
+
"grad_norm": 0.23713508248329163,
|
| 1750 |
+
"learning_rate": 6.530757220625723e-05,
|
| 1751 |
+
"loss": 0.0263,
|
| 1752 |
+
"step": 2490
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 6.476683937823834,
|
| 1756 |
+
"grad_norm": 0.17403320968151093,
|
| 1757 |
+
"learning_rate": 6.452139676079631e-05,
|
| 1758 |
+
"loss": 0.022,
|
| 1759 |
+
"step": 2500
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 6.5025906735751295,
|
| 1763 |
+
"grad_norm": 0.19337254762649536,
|
| 1764 |
+
"learning_rate": 6.373772198394114e-05,
|
| 1765 |
+
"loss": 0.0199,
|
| 1766 |
+
"step": 2510
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 6.528497409326425,
|
| 1770 |
+
"grad_norm": 0.1531633585691452,
|
| 1771 |
+
"learning_rate": 6.295660311210843e-05,
|
| 1772 |
+
"loss": 0.022,
|
| 1773 |
+
"step": 2520
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 6.55440414507772,
|
| 1777 |
+
"grad_norm": 0.23906753957271576,
|
| 1778 |
+
"learning_rate": 6.217809520156481e-05,
|
| 1779 |
+
"loss": 0.0203,
|
| 1780 |
+
"step": 2530
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 6.580310880829016,
|
| 1784 |
+
"grad_norm": 0.1223067194223404,
|
| 1785 |
+
"learning_rate": 6.140225312454629e-05,
|
| 1786 |
+
"loss": 0.0208,
|
| 1787 |
+
"step": 2540
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 6.606217616580311,
|
| 1791 |
+
"grad_norm": 0.18273130059242249,
|
| 1792 |
+
"learning_rate": 6.0629131565390715e-05,
|
| 1793 |
+
"loss": 0.0238,
|
| 1794 |
+
"step": 2550
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 6.632124352331607,
|
| 1798 |
+
"grad_norm": 0.2462676465511322,
|
| 1799 |
+
"learning_rate": 5.9858785016683306e-05,
|
| 1800 |
+
"loss": 0.025,
|
| 1801 |
+
"step": 2560
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 6.658031088082901,
|
| 1805 |
+
"grad_norm": 0.2025512307882309,
|
| 1806 |
+
"learning_rate": 5.909126777541596e-05,
|
| 1807 |
+
"loss": 0.0199,
|
| 1808 |
+
"step": 2570
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 6.683937823834197,
|
| 1812 |
+
"grad_norm": 0.18922774493694305,
|
| 1813 |
+
"learning_rate": 5.832663393915991e-05,
|
| 1814 |
+
"loss": 0.0193,
|
| 1815 |
+
"step": 2580
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 6.709844559585492,
|
| 1819 |
+
"grad_norm": 0.2345733493566513,
|
| 1820 |
+
"learning_rate": 5.756493740225306e-05,
|
| 1821 |
+
"loss": 0.0186,
|
| 1822 |
+
"step": 2590
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 6.7357512953367875,
|
| 1826 |
+
"grad_norm": 0.17045198380947113,
|
| 1827 |
+
"learning_rate": 5.6806231852001104e-05,
|
| 1828 |
+
"loss": 0.0248,
|
| 1829 |
+
"step": 2600
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 6.761658031088083,
|
| 1833 |
+
"grad_norm": 0.23411734402179718,
|
| 1834 |
+
"learning_rate": 5.605057076489342e-05,
|
| 1835 |
+
"loss": 0.0213,
|
| 1836 |
+
"step": 2610
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 6.787564766839378,
|
| 1840 |
+
"grad_norm": 0.19447815418243408,
|
| 1841 |
+
"learning_rate": 5.5298007402833905e-05,
|
| 1842 |
+
"loss": 0.0217,
|
| 1843 |
+
"step": 2620
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 6.813471502590674,
|
| 1847 |
+
"grad_norm": 0.17666564881801605,
|
| 1848 |
+
"learning_rate": 5.454859480938692e-05,
|
| 1849 |
+
"loss": 0.0194,
|
| 1850 |
+
"step": 2630
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 6.839378238341969,
|
| 1854 |
+
"grad_norm": 0.193482905626297,
|
| 1855 |
+
"learning_rate": 5.3802385806038535e-05,
|
| 1856 |
+
"loss": 0.0229,
|
| 1857 |
+
"step": 2640
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 6.865284974093264,
|
| 1861 |
+
"grad_norm": 0.16288697719573975,
|
| 1862 |
+
"learning_rate": 5.305943298847336e-05,
|
| 1863 |
+
"loss": 0.0208,
|
| 1864 |
+
"step": 2650
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 6.891191709844559,
|
| 1868 |
+
"grad_norm": 0.2407398670911789,
|
| 1869 |
+
"learning_rate": 5.23197887228675e-05,
|
| 1870 |
+
"loss": 0.0202,
|
| 1871 |
+
"step": 2660
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 6.917098445595855,
|
| 1875 |
+
"grad_norm": 0.20776060223579407,
|
| 1876 |
+
"learning_rate": 5.1583505142197694e-05,
|
| 1877 |
+
"loss": 0.0207,
|
| 1878 |
+
"step": 2670
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 6.94300518134715,
|
| 1882 |
+
"grad_norm": 0.21267752349376678,
|
| 1883 |
+
"learning_rate": 5.08506341425665e-05,
|
| 1884 |
+
"loss": 0.0199,
|
| 1885 |
+
"step": 2680
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 6.9689119170984455,
|
| 1889 |
+
"grad_norm": 0.21812330186367035,
|
| 1890 |
+
"learning_rate": 5.0121227379544775e-05,
|
| 1891 |
+
"loss": 0.0215,
|
| 1892 |
+
"step": 2690
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 6.994818652849741,
|
| 1896 |
+
"grad_norm": 0.19859305024147034,
|
| 1897 |
+
"learning_rate": 4.9395336264530486e-05,
|
| 1898 |
+
"loss": 0.0202,
|
| 1899 |
+
"step": 2700
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 7.020725388601036,
|
| 1903 |
+
"grad_norm": 0.12451327592134476,
|
| 1904 |
+
"learning_rate": 4.8673011961125334e-05,
|
| 1905 |
+
"loss": 0.0167,
|
| 1906 |
+
"step": 2710
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 7.046632124352332,
|
| 1910 |
+
"grad_norm": 0.23396964371204376,
|
| 1911 |
+
"learning_rate": 4.7954305381528254e-05,
|
| 1912 |
+
"loss": 0.0178,
|
| 1913 |
+
"step": 2720
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 7.072538860103627,
|
| 1917 |
+
"grad_norm": 0.19071508944034576,
|
| 1918 |
+
"learning_rate": 4.723926718294722e-05,
|
| 1919 |
+
"loss": 0.022,
|
| 1920 |
+
"step": 2730
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 7.098445595854923,
|
| 1924 |
+
"grad_norm": 0.17534908652305603,
|
| 1925 |
+
"learning_rate": 4.6527947764028456e-05,
|
| 1926 |
+
"loss": 0.0158,
|
| 1927 |
+
"step": 2740
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 7.124352331606218,
|
| 1931 |
+
"grad_norm": 0.17512699961662292,
|
| 1932 |
+
"learning_rate": 4.582039726130437e-05,
|
| 1933 |
+
"loss": 0.0172,
|
| 1934 |
+
"step": 2750
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 7.150259067357513,
|
| 1938 |
+
"grad_norm": 0.14806455373764038,
|
| 1939 |
+
"learning_rate": 4.511666554565955e-05,
|
| 1940 |
+
"loss": 0.0178,
|
| 1941 |
+
"step": 2760
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 7.176165803108808,
|
| 1945 |
+
"grad_norm": 0.17297431826591492,
|
| 1946 |
+
"learning_rate": 4.441680221881587e-05,
|
| 1947 |
+
"loss": 0.0206,
|
| 1948 |
+
"step": 2770
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 7.2020725388601035,
|
| 1952 |
+
"grad_norm": 0.16674122214317322,
|
| 1953 |
+
"learning_rate": 4.372085660983616e-05,
|
| 1954 |
+
"loss": 0.0163,
|
| 1955 |
+
"step": 2780
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 7.227979274611399,
|
| 1959 |
+
"grad_norm": 0.219289630651474,
|
| 1960 |
+
"learning_rate": 4.302887777164742e-05,
|
| 1961 |
+
"loss": 0.0185,
|
| 1962 |
+
"step": 2790
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 7.253886010362694,
|
| 1966 |
+
"grad_norm": 0.15512730181217194,
|
| 1967 |
+
"learning_rate": 4.2340914477583425e-05,
|
| 1968 |
+
"loss": 0.0177,
|
| 1969 |
+
"step": 2800
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 7.27979274611399,
|
| 1973 |
+
"grad_norm": 0.23912139236927032,
|
| 1974 |
+
"learning_rate": 4.1657015217946936e-05,
|
| 1975 |
+
"loss": 0.0188,
|
| 1976 |
+
"step": 2810
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 7.305699481865285,
|
| 1980 |
+
"grad_norm": 0.20765544474124908,
|
| 1981 |
+
"learning_rate": 4.097722819659184e-05,
|
| 1982 |
+
"loss": 0.0185,
|
| 1983 |
+
"step": 2820
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 7.331606217616581,
|
| 1987 |
+
"grad_norm": 0.17259199917316437,
|
| 1988 |
+
"learning_rate": 4.030160132752565e-05,
|
| 1989 |
+
"loss": 0.017,
|
| 1990 |
+
"step": 2830
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 7.357512953367876,
|
| 1994 |
+
"grad_norm": 0.1474766880273819,
|
| 1995 |
+
"learning_rate": 3.9630182231532366e-05,
|
| 1996 |
+
"loss": 0.0188,
|
| 1997 |
+
"step": 2840
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 7.383419689119171,
|
| 2001 |
+
"grad_norm": 0.16211964190006256,
|
| 2002 |
+
"learning_rate": 3.896301823281595e-05,
|
| 2003 |
+
"loss": 0.0192,
|
| 2004 |
+
"step": 2850
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 7.409326424870466,
|
| 2008 |
+
"grad_norm": 0.1418692171573639,
|
| 2009 |
+
"learning_rate": 3.8300156355664665e-05,
|
| 2010 |
+
"loss": 0.0188,
|
| 2011 |
+
"step": 2860
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 7.435233160621761,
|
| 2015 |
+
"grad_norm": 0.16895586252212524,
|
| 2016 |
+
"learning_rate": 3.7641643321136624e-05,
|
| 2017 |
+
"loss": 0.0205,
|
| 2018 |
+
"step": 2870
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 7.461139896373057,
|
| 2022 |
+
"grad_norm": 0.22076182067394257,
|
| 2023 |
+
"learning_rate": 3.698752554376689e-05,
|
| 2024 |
+
"loss": 0.0178,
|
| 2025 |
+
"step": 2880
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 7.487046632124352,
|
| 2029 |
+
"grad_norm": 0.15769238770008087,
|
| 2030 |
+
"learning_rate": 3.633784912829586e-05,
|
| 2031 |
+
"loss": 0.0189,
|
| 2032 |
+
"step": 2890
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 7.512953367875648,
|
| 2036 |
+
"grad_norm": 0.27454307675361633,
|
| 2037 |
+
"learning_rate": 3.5692659866419484e-05,
|
| 2038 |
+
"loss": 0.0191,
|
| 2039 |
+
"step": 2900
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 7.538860103626943,
|
| 2043 |
+
"grad_norm": 0.16451720893383026,
|
| 2044 |
+
"learning_rate": 3.5052003233562e-05,
|
| 2045 |
+
"loss": 0.0172,
|
| 2046 |
+
"step": 2910
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 7.564766839378239,
|
| 2050 |
+
"grad_norm": 0.14524053037166595,
|
| 2051 |
+
"learning_rate": 3.4415924385670495e-05,
|
| 2052 |
+
"loss": 0.0214,
|
| 2053 |
+
"step": 2920
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 7.590673575129534,
|
| 2057 |
+
"grad_norm": 0.25439292192459106,
|
| 2058 |
+
"learning_rate": 3.378446815603205e-05,
|
| 2059 |
+
"loss": 0.0198,
|
| 2060 |
+
"step": 2930
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 7.616580310880829,
|
| 2064 |
+
"grad_norm": 0.2044086903333664,
|
| 2065 |
+
"learning_rate": 3.31576790521138e-05,
|
| 2066 |
+
"loss": 0.0181,
|
| 2067 |
+
"step": 2940
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 7.642487046632124,
|
| 2071 |
+
"grad_norm": 0.13481000065803528,
|
| 2072 |
+
"learning_rate": 3.253560125242597e-05,
|
| 2073 |
+
"loss": 0.0197,
|
| 2074 |
+
"step": 2950
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 7.668393782383419,
|
| 2078 |
+
"grad_norm": 0.13905170559883118,
|
| 2079 |
+
"learning_rate": 3.191827860340794e-05,
|
| 2080 |
+
"loss": 0.0161,
|
| 2081 |
+
"step": 2960
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 7.694300518134715,
|
| 2085 |
+
"grad_norm": 0.17516188323497772,
|
| 2086 |
+
"learning_rate": 3.1305754616337624e-05,
|
| 2087 |
+
"loss": 0.0182,
|
| 2088 |
+
"step": 2970
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 7.72020725388601,
|
| 2092 |
+
"grad_norm": 0.1584802269935608,
|
| 2093 |
+
"learning_rate": 3.06980724642648e-05,
|
| 2094 |
+
"loss": 0.0171,
|
| 2095 |
+
"step": 2980
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 7.746113989637306,
|
| 2099 |
+
"grad_norm": 0.10378487408161163,
|
| 2100 |
+
"learning_rate": 3.009527497896817e-05,
|
| 2101 |
+
"loss": 0.0158,
|
| 2102 |
+
"step": 2990
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 7.772020725388601,
|
| 2106 |
+
"grad_norm": 0.22542284429073334,
|
| 2107 |
+
"learning_rate": 2.949740464793621e-05,
|
| 2108 |
+
"loss": 0.0172,
|
| 2109 |
+
"step": 3000
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 7.7979274611398965,
|
| 2113 |
+
"grad_norm": 0.1301916539669037,
|
| 2114 |
+
"learning_rate": 2.8904503611372612e-05,
|
| 2115 |
+
"loss": 0.0165,
|
| 2116 |
+
"step": 3010
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 7.823834196891192,
|
| 2120 |
+
"grad_norm": 0.17367781698703766,
|
| 2121 |
+
"learning_rate": 2.8316613659226022e-05,
|
| 2122 |
+
"loss": 0.0153,
|
| 2123 |
+
"step": 3020
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 7.849740932642487,
|
| 2127 |
+
"grad_norm": 0.16101326048374176,
|
| 2128 |
+
"learning_rate": 2.773377622824459e-05,
|
| 2129 |
+
"loss": 0.0147,
|
| 2130 |
+
"step": 3030
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 7.875647668393782,
|
| 2134 |
+
"grad_norm": 0.15740449726581573,
|
| 2135 |
+
"learning_rate": 2.7156032399055375e-05,
|
| 2136 |
+
"loss": 0.0183,
|
| 2137 |
+
"step": 3040
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 7.901554404145077,
|
| 2141 |
+
"grad_norm": 0.1986459195613861,
|
| 2142 |
+
"learning_rate": 2.6583422893268638e-05,
|
| 2143 |
+
"loss": 0.015,
|
| 2144 |
+
"step": 3050
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 7.927461139896373,
|
| 2148 |
+
"grad_norm": 0.1751032918691635,
|
| 2149 |
+
"learning_rate": 2.601598807060779e-05,
|
| 2150 |
+
"loss": 0.0171,
|
| 2151 |
+
"step": 3060
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 7.953367875647668,
|
| 2155 |
+
"grad_norm": 0.15115486085414886,
|
| 2156 |
+
"learning_rate": 2.545376792606473e-05,
|
| 2157 |
+
"loss": 0.0177,
|
| 2158 |
+
"step": 3070
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 7.979274611398964,
|
| 2162 |
+
"grad_norm": 0.15944965183734894,
|
| 2163 |
+
"learning_rate": 2.489680208708063e-05,
|
| 2164 |
+
"loss": 0.0157,
|
| 2165 |
+
"step": 3080
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 8.005181347150259,
|
| 2169 |
+
"grad_norm": 0.15558196604251862,
|
| 2170 |
+
"learning_rate": 2.4345129810753087e-05,
|
| 2171 |
+
"loss": 0.0182,
|
| 2172 |
+
"step": 3090
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 8.031088082901555,
|
| 2176 |
+
"grad_norm": 0.21775135397911072,
|
| 2177 |
+
"learning_rate": 2.3798789981068914e-05,
|
| 2178 |
+
"loss": 0.0187,
|
| 2179 |
+
"step": 3100
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 8.05699481865285,
|
| 2183 |
+
"grad_norm": 0.19147688150405884,
|
| 2184 |
+
"learning_rate": 2.3257821106163634e-05,
|
| 2185 |
+
"loss": 0.015,
|
| 2186 |
+
"step": 3110
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 8.082901554404145,
|
| 2190 |
+
"grad_norm": 0.2634642720222473,
|
| 2191 |
+
"learning_rate": 2.2722261315607106e-05,
|
| 2192 |
+
"loss": 0.0157,
|
| 2193 |
+
"step": 3120
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 8.10880829015544,
|
| 2197 |
+
"grad_norm": 0.14185205101966858,
|
| 2198 |
+
"learning_rate": 2.2192148357716168e-05,
|
| 2199 |
+
"loss": 0.0157,
|
| 2200 |
+
"step": 3130
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 8.134715025906736,
|
| 2204 |
+
"grad_norm": 0.19837462902069092,
|
| 2205 |
+
"learning_rate": 2.1667519596893836e-05,
|
| 2206 |
+
"loss": 0.0171,
|
| 2207 |
+
"step": 3140
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 8.160621761658032,
|
| 2211 |
+
"grad_norm": 0.21308404207229614,
|
| 2212 |
+
"learning_rate": 2.114841201099588e-05,
|
| 2213 |
+
"loss": 0.0171,
|
| 2214 |
+
"step": 3150
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 8.186528497409327,
|
| 2218 |
+
"grad_norm": 0.16102993488311768,
|
| 2219 |
+
"learning_rate": 2.06348621887243e-05,
|
| 2220 |
+
"loss": 0.014,
|
| 2221 |
+
"step": 3160
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 8.212435233160623,
|
| 2225 |
+
"grad_norm": 0.1218574196100235,
|
| 2226 |
+
"learning_rate": 2.012690632704861e-05,
|
| 2227 |
+
"loss": 0.0154,
|
| 2228 |
+
"step": 3170
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 8.238341968911918,
|
| 2232 |
+
"grad_norm": 0.17196378111839294,
|
| 2233 |
+
"learning_rate": 1.9624580228654366e-05,
|
| 2234 |
+
"loss": 0.014,
|
| 2235 |
+
"step": 3180
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 8.264248704663212,
|
| 2239 |
+
"grad_norm": 0.16071684658527374,
|
| 2240 |
+
"learning_rate": 1.9127919299419705e-05,
|
| 2241 |
+
"loss": 0.0162,
|
| 2242 |
+
"step": 3190
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 8.290155440414507,
|
| 2246 |
+
"grad_norm": 0.151749387383461,
|
| 2247 |
+
"learning_rate": 1.8636958545919904e-05,
|
| 2248 |
+
"loss": 0.016,
|
| 2249 |
+
"step": 3200
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 8.316062176165802,
|
| 2253 |
+
"grad_norm": 0.16748680174350739,
|
| 2254 |
+
"learning_rate": 1.8151732572959857e-05,
|
| 2255 |
+
"loss": 0.017,
|
| 2256 |
+
"step": 3210
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 8.341968911917098,
|
| 2260 |
+
"grad_norm": 0.09705256670713425,
|
| 2261 |
+
"learning_rate": 1.7672275581135023e-05,
|
| 2262 |
+
"loss": 0.014,
|
| 2263 |
+
"step": 3220
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 8.367875647668393,
|
| 2267 |
+
"grad_norm": 0.17474466562271118,
|
| 2268 |
+
"learning_rate": 1.719862136442083e-05,
|
| 2269 |
+
"loss": 0.0132,
|
| 2270 |
+
"step": 3230
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 8.393782383419689,
|
| 2274 |
+
"grad_norm": 0.12089501321315765,
|
| 2275 |
+
"learning_rate": 1.6730803307790842e-05,
|
| 2276 |
+
"loss": 0.0162,
|
| 2277 |
+
"step": 3240
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 8.419689119170984,
|
| 2281 |
+
"grad_norm": 0.1267276108264923,
|
| 2282 |
+
"learning_rate": 1.626885438486355e-05,
|
| 2283 |
+
"loss": 0.0151,
|
| 2284 |
+
"step": 3250
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 8.44559585492228,
|
| 2288 |
+
"grad_norm": 0.17889542877674103,
|
| 2289 |
+
"learning_rate": 1.581280715557827e-05,
|
| 2290 |
+
"loss": 0.0181,
|
| 2291 |
+
"step": 3260
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 8.471502590673575,
|
| 2295 |
+
"grad_norm": 0.17118024826049805,
|
| 2296 |
+
"learning_rate": 1.5362693763900193e-05,
|
| 2297 |
+
"loss": 0.0178,
|
| 2298 |
+
"step": 3270
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 8.49740932642487,
|
| 2302 |
+
"grad_norm": 0.13716407120227814,
|
| 2303 |
+
"learning_rate": 1.4918545935554872e-05,
|
| 2304 |
+
"loss": 0.015,
|
| 2305 |
+
"step": 3280
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 8.523316062176166,
|
| 2309 |
+
"grad_norm": 0.17024938762187958,
|
| 2310 |
+
"learning_rate": 1.4480394975791955e-05,
|
| 2311 |
+
"loss": 0.0206,
|
| 2312 |
+
"step": 3290
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 8.549222797927461,
|
| 2316 |
+
"grad_norm": 0.137301504611969,
|
| 2317 |
+
"learning_rate": 1.4048271767178579e-05,
|
| 2318 |
+
"loss": 0.0161,
|
| 2319 |
+
"step": 3300
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 8.575129533678757,
|
| 2323 |
+
"grad_norm": 0.12056181579828262,
|
| 2324 |
+
"learning_rate": 1.3622206767422907e-05,
|
| 2325 |
+
"loss": 0.0156,
|
| 2326 |
+
"step": 3310
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 8.601036269430052,
|
| 2330 |
+
"grad_norm": 0.228641539812088,
|
| 2331 |
+
"learning_rate": 1.3202230007227179e-05,
|
| 2332 |
+
"loss": 0.0182,
|
| 2333 |
+
"step": 3320
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 8.626943005181348,
|
| 2337 |
+
"grad_norm": 0.14743590354919434,
|
| 2338 |
+
"learning_rate": 1.278837108817098e-05,
|
| 2339 |
+
"loss": 0.0139,
|
| 2340 |
+
"step": 3330
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 8.652849740932643,
|
| 2344 |
+
"grad_norm": 0.18240438401699066,
|
| 2345 |
+
"learning_rate": 1.2380659180624998e-05,
|
| 2346 |
+
"loss": 0.0154,
|
| 2347 |
+
"step": 3340
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 8.678756476683938,
|
| 2351 |
+
"grad_norm": 0.18438772857189178,
|
| 2352 |
+
"learning_rate": 1.1979123021694782e-05,
|
| 2353 |
+
"loss": 0.0137,
|
| 2354 |
+
"step": 3350
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 8.704663212435234,
|
| 2358 |
+
"grad_norm": 0.1203690692782402,
|
| 2359 |
+
"learning_rate": 1.1583790913195402e-05,
|
| 2360 |
+
"loss": 0.0146,
|
| 2361 |
+
"step": 3360
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 8.73056994818653,
|
| 2365 |
+
"grad_norm": 0.11229782551527023,
|
| 2366 |
+
"learning_rate": 1.1194690719656486e-05,
|
| 2367 |
+
"loss": 0.0157,
|
| 2368 |
+
"step": 3370
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 8.756476683937823,
|
| 2372 |
+
"grad_norm": 0.15702977776527405,
|
| 2373 |
+
"learning_rate": 1.0811849866358393e-05,
|
| 2374 |
+
"loss": 0.0134,
|
| 2375 |
+
"step": 3380
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 8.782383419689118,
|
| 2379 |
+
"grad_norm": 0.11596658080816269,
|
| 2380 |
+
"learning_rate": 1.0435295337398943e-05,
|
| 2381 |
+
"loss": 0.0123,
|
| 2382 |
+
"step": 3390
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 8.808290155440414,
|
| 2386 |
+
"grad_norm": 0.15277987718582153,
|
| 2387 |
+
"learning_rate": 1.0065053673791748e-05,
|
| 2388 |
+
"loss": 0.0171,
|
| 2389 |
+
"step": 3400
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 8.83419689119171,
|
| 2393 |
+
"grad_norm": 0.17561012506484985,
|
| 2394 |
+
"learning_rate": 9.701150971595218e-06,
|
| 2395 |
+
"loss": 0.0147,
|
| 2396 |
+
"step": 3410
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 8.860103626943005,
|
| 2400 |
+
"grad_norm": 0.12972599267959595,
|
| 2401 |
+
"learning_rate": 9.34361288007346e-06,
|
| 2402 |
+
"loss": 0.0132,
|
| 2403 |
+
"step": 3420
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 8.8860103626943,
|
| 2407 |
+
"grad_norm": 0.14359018206596375,
|
| 2408 |
+
"learning_rate": 8.992464599888229e-06,
|
| 2409 |
+
"loss": 0.014,
|
| 2410 |
+
"step": 3430
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 8.911917098445596,
|
| 2414 |
+
"grad_norm": 0.18934310972690582,
|
| 2415 |
+
"learning_rate": 8.647730881322813e-06,
|
| 2416 |
+
"loss": 0.0186,
|
| 2417 |
+
"step": 3440
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 8.937823834196891,
|
| 2421 |
+
"grad_norm": 0.22511492669582367,
|
| 2422 |
+
"learning_rate": 8.309436022537465e-06,
|
| 2423 |
+
"loss": 0.0167,
|
| 2424 |
+
"step": 3450
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 8.963730569948186,
|
| 2428 |
+
"grad_norm": 0.17313934862613678,
|
| 2429 |
+
"learning_rate": 7.97760386785682e-06,
|
| 2430 |
+
"loss": 0.0172,
|
| 2431 |
+
"step": 3460
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 8.989637305699482,
|
| 2435 |
+
"grad_norm": 0.15488480031490326,
|
| 2436 |
+
"learning_rate": 7.652257806089213e-06,
|
| 2437 |
+
"loss": 0.0141,
|
| 2438 |
+
"step": 3470
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 9.015544041450777,
|
| 2442 |
+
"grad_norm": 0.11114071309566498,
|
| 2443 |
+
"learning_rate": 7.333420768878174e-06,
|
| 2444 |
+
"loss": 0.0146,
|
| 2445 |
+
"step": 3480
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 9.041450777202073,
|
| 2449 |
+
"grad_norm": 0.09289427846670151,
|
| 2450 |
+
"learning_rate": 7.0211152290861285e-06,
|
| 2451 |
+
"loss": 0.0145,
|
| 2452 |
+
"step": 3490
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 9.067357512953368,
|
| 2456 |
+
"grad_norm": 0.09938207268714905,
|
| 2457 |
+
"learning_rate": 6.715363199210423e-06,
|
| 2458 |
+
"loss": 0.0156,
|
| 2459 |
+
"step": 3500
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 9.093264248704664,
|
| 2463 |
+
"grad_norm": 0.09943415969610214,
|
| 2464 |
+
"learning_rate": 6.416186229831722e-06,
|
| 2465 |
+
"loss": 0.0134,
|
| 2466 |
+
"step": 3510
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 9.119170984455959,
|
| 2470 |
+
"grad_norm": 0.15533126890659332,
|
| 2471 |
+
"learning_rate": 6.123605408095146e-06,
|
| 2472 |
+
"loss": 0.0121,
|
| 2473 |
+
"step": 3520
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 9.145077720207254,
|
| 2477 |
+
"grad_norm": 0.14299090206623077,
|
| 2478 |
+
"learning_rate": 5.837641356223911e-06,
|
| 2479 |
+
"loss": 0.0169,
|
| 2480 |
+
"step": 3530
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 9.17098445595855,
|
| 2484 |
+
"grad_norm": 0.096370629966259,
|
| 2485 |
+
"learning_rate": 5.558314230065842e-06,
|
| 2486 |
+
"loss": 0.0142,
|
| 2487 |
+
"step": 3540
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 9.196891191709845,
|
| 2491 |
+
"grad_norm": 0.16918356716632843,
|
| 2492 |
+
"learning_rate": 5.285643717672628e-06,
|
| 2493 |
+
"loss": 0.0126,
|
| 2494 |
+
"step": 3550
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 9.22279792746114,
|
| 2498 |
+
"grad_norm": 0.23584723472595215,
|
| 2499 |
+
"learning_rate": 5.0196490379121885e-06,
|
| 2500 |
+
"loss": 0.0145,
|
| 2501 |
+
"step": 3560
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 9.248704663212436,
|
| 2505 |
+
"grad_norm": 0.15384496748447418,
|
| 2506 |
+
"learning_rate": 4.760348939114068e-06,
|
| 2507 |
+
"loss": 0.0152,
|
| 2508 |
+
"step": 3570
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 9.27461139896373,
|
| 2512 |
+
"grad_norm": 0.24043075740337372,
|
| 2513 |
+
"learning_rate": 4.507761697747981e-06,
|
| 2514 |
+
"loss": 0.0162,
|
| 2515 |
+
"step": 3580
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 9.300518134715025,
|
| 2519 |
+
"grad_norm": 0.14302724599838257,
|
| 2520 |
+
"learning_rate": 4.261905117135512e-06,
|
| 2521 |
+
"loss": 0.0151,
|
| 2522 |
+
"step": 3590
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 9.32642487046632,
|
| 2526 |
+
"grad_norm": 0.1379092037677765,
|
| 2527 |
+
"learning_rate": 4.022796526195416e-06,
|
| 2528 |
+
"loss": 0.0156,
|
| 2529 |
+
"step": 3600
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 9.352331606217616,
|
| 2533 |
+
"grad_norm": 0.12064504623413086,
|
| 2534 |
+
"learning_rate": 3.7904527782221754e-06,
|
| 2535 |
+
"loss": 0.0117,
|
| 2536 |
+
"step": 3610
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 9.378238341968911,
|
| 2540 |
+
"grad_norm": 0.15608225762844086,
|
| 2541 |
+
"learning_rate": 3.564890249698005e-06,
|
| 2542 |
+
"loss": 0.0134,
|
| 2543 |
+
"step": 3620
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 9.404145077720207,
|
| 2547 |
+
"grad_norm": 0.2138279527425766,
|
| 2548 |
+
"learning_rate": 3.346124839138687e-06,
|
| 2549 |
+
"loss": 0.0141,
|
| 2550 |
+
"step": 3630
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 9.430051813471502,
|
| 2554 |
+
"grad_norm": 0.10819461196660995,
|
| 2555 |
+
"learning_rate": 3.1341719659729764e-06,
|
| 2556 |
+
"loss": 0.015,
|
| 2557 |
+
"step": 3640
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 9.455958549222798,
|
| 2561 |
+
"grad_norm": 0.1564425379037857,
|
| 2562 |
+
"learning_rate": 2.9290465694557402e-06,
|
| 2563 |
+
"loss": 0.0139,
|
| 2564 |
+
"step": 3650
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 9.481865284974093,
|
| 2568 |
+
"grad_norm": 0.12781445682048798,
|
| 2569 |
+
"learning_rate": 2.7307631076149644e-06,
|
| 2570 |
+
"loss": 0.0144,
|
| 2571 |
+
"step": 3660
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 9.507772020725389,
|
| 2575 |
+
"grad_norm": 0.09093884378671646,
|
| 2576 |
+
"learning_rate": 2.539335556232736e-06,
|
| 2577 |
+
"loss": 0.0122,
|
| 2578 |
+
"step": 3670
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 9.533678756476684,
|
| 2582 |
+
"grad_norm": 0.09807614982128143,
|
| 2583 |
+
"learning_rate": 2.354777407860209e-06,
|
| 2584 |
+
"loss": 0.0124,
|
| 2585 |
+
"step": 3680
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 9.55958549222798,
|
| 2589 |
+
"grad_norm": 0.15765872597694397,
|
| 2590 |
+
"learning_rate": 2.1771016708665217e-06,
|
| 2591 |
+
"loss": 0.0152,
|
| 2592 |
+
"step": 3690
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 9.585492227979275,
|
| 2596 |
+
"grad_norm": 0.3498193621635437,
|
| 2597 |
+
"learning_rate": 2.0063208685219494e-06,
|
| 2598 |
+
"loss": 0.0154,
|
| 2599 |
+
"step": 3700
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 9.61139896373057,
|
| 2603 |
+
"grad_norm": 0.16188929975032806,
|
| 2604 |
+
"learning_rate": 1.8424470381152137e-06,
|
| 2605 |
+
"loss": 0.0135,
|
| 2606 |
+
"step": 3710
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 9.637305699481866,
|
| 2610 |
+
"grad_norm": 0.12133660912513733,
|
| 2611 |
+
"learning_rate": 1.68549173010506e-06,
|
| 2612 |
+
"loss": 0.0144,
|
| 2613 |
+
"step": 3720
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 9.663212435233161,
|
| 2617 |
+
"grad_norm": 0.15781523287296295,
|
| 2618 |
+
"learning_rate": 1.53546600730613e-06,
|
| 2619 |
+
"loss": 0.0123,
|
| 2620 |
+
"step": 3730
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 9.689119170984457,
|
| 2624 |
+
"grad_norm": 0.11044237017631531,
|
| 2625 |
+
"learning_rate": 1.3923804441092092e-06,
|
| 2626 |
+
"loss": 0.0158,
|
| 2627 |
+
"step": 3740
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 9.715025906735752,
|
| 2631 |
+
"grad_norm": 0.2011040598154068,
|
| 2632 |
+
"learning_rate": 1.2562451257358887e-06,
|
| 2633 |
+
"loss": 0.014,
|
| 2634 |
+
"step": 3750
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 9.740932642487046,
|
| 2638 |
+
"grad_norm": 0.14315728843212128,
|
| 2639 |
+
"learning_rate": 1.1270696475277565e-06,
|
| 2640 |
+
"loss": 0.0119,
|
| 2641 |
+
"step": 3760
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 9.766839378238341,
|
| 2645 |
+
"grad_norm": 0.13683845102787018,
|
| 2646 |
+
"learning_rate": 1.0048631142700383e-06,
|
| 2647 |
+
"loss": 0.0147,
|
| 2648 |
+
"step": 3770
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 9.792746113989637,
|
| 2652 |
+
"grad_norm": 0.07541488856077194,
|
| 2653 |
+
"learning_rate": 8.896341395499108e-07,
|
| 2654 |
+
"loss": 0.0108,
|
| 2655 |
+
"step": 3780
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 9.818652849740932,
|
| 2659 |
+
"grad_norm": 0.15584123134613037,
|
| 2660 |
+
"learning_rate": 7.813908451493102e-07,
|
| 2661 |
+
"loss": 0.0149,
|
| 2662 |
+
"step": 3790
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 9.844559585492227,
|
| 2666 |
+
"grad_norm": 0.16258586943149567,
|
| 2667 |
+
"learning_rate": 6.801408604725446e-07,
|
| 2668 |
+
"loss": 0.0154,
|
| 2669 |
+
"step": 3800
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 9.870466321243523,
|
| 2673 |
+
"grad_norm": 0.1323482096195221,
|
| 2674 |
+
"learning_rate": 5.858913220085027e-07,
|
| 2675 |
+
"loss": 0.0124,
|
| 2676 |
+
"step": 3810
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 9.896373056994818,
|
| 2680 |
+
"grad_norm": 0.11180088669061661,
|
| 2681 |
+
"learning_rate": 4.986488728276673e-07,
|
| 2682 |
+
"loss": 0.0131,
|
| 2683 |
+
"step": 3820
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 9.922279792746114,
|
| 2687 |
+
"grad_norm": 0.1439903825521469,
|
| 2688 |
+
"learning_rate": 4.1841966211388914e-07,
|
| 2689 |
+
"loss": 0.0153,
|
| 2690 |
+
"step": 3830
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 9.94818652849741,
|
| 2694 |
+
"grad_norm": 0.10643361508846283,
|
| 2695 |
+
"learning_rate": 3.452093447309346e-07,
|
| 2696 |
+
"loss": 0.0131,
|
| 2697 |
+
"step": 3840
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 9.974093264248705,
|
| 2701 |
+
"grad_norm": 0.11634659022092819,
|
| 2702 |
+
"learning_rate": 2.790230808239591e-07,
|
| 2703 |
+
"loss": 0.0129,
|
| 2704 |
+
"step": 3850
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 10.0,
|
| 2708 |
+
"grad_norm": 0.1371113657951355,
|
| 2709 |
+
"learning_rate": 2.198655354557766e-07,
|
| 2710 |
+
"loss": 0.0133,
|
| 2711 |
+
"step": 3860
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 10.025906735751295,
|
| 2715 |
+
"grad_norm": 0.10803582519292831,
|
| 2716 |
+
"learning_rate": 1.6774087827801098e-07,
|
| 2717 |
+
"loss": 0.0127,
|
| 2718 |
+
"step": 3870
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 10.05181347150259,
|
| 2722 |
+
"grad_norm": 0.1608572006225586,
|
| 2723 |
+
"learning_rate": 1.2265278323728703e-07,
|
| 2724 |
+
"loss": 0.0122,
|
| 2725 |
+
"step": 3880
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 10.077720207253886,
|
| 2729 |
+
"grad_norm": 0.14935462176799774,
|
| 2730 |
+
"learning_rate": 8.460442831619286e-08,
|
| 2731 |
+
"loss": 0.0121,
|
| 2732 |
+
"step": 3890
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 10.103626943005182,
|
| 2736 |
+
"grad_norm": 0.17790275812149048,
|
| 2737 |
+
"learning_rate": 5.359849530932603e-08,
|
| 2738 |
+
"loss": 0.0154,
|
| 2739 |
+
"step": 3900
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 10.129533678756477,
|
| 2743 |
+
"grad_norm": 0.2045956552028656,
|
| 2744 |
+
"learning_rate": 2.9637169634277872e-08,
|
| 2745 |
+
"loss": 0.0141,
|
| 2746 |
+
"step": 3910
|
| 2747 |
+
},
|
| 2748 |
+
{
|
| 2749 |
+
"epoch": 10.155440414507773,
|
| 2750 |
+
"grad_norm": 0.22062177956104279,
|
| 2751 |
+
"learning_rate": 1.2722140177567899e-08,
|
| 2752 |
+
"loss": 0.0161,
|
| 2753 |
+
"step": 3920
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 10.181347150259068,
|
| 2757 |
+
"grad_norm": 0.18334221839904785,
|
| 2758 |
+
"learning_rate": 2.8545991756168122e-09,
|
| 2759 |
+
"loss": 0.0128,
|
| 2760 |
+
"step": 3930
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 10.204663212435234,
|
| 2764 |
+
"step": 3939,
|
| 2765 |
+
"total_flos": 5.206682005553e+17,
|
| 2766 |
+
"train_loss": 0.0395857700919645,
|
| 2767 |
+
"train_runtime": 4409.4556,
|
| 2768 |
+
"train_samples_per_second": 57.172,
|
| 2769 |
+
"train_steps_per_second": 0.893
|
| 2770 |
+
}
|
| 2771 |
+
],
|
| 2772 |
+
"logging_steps": 10,
|
| 2773 |
+
"max_steps": 3939,
|
| 2774 |
+
"num_input_tokens_seen": 0,
|
| 2775 |
+
"num_train_epochs": 11,
|
| 2776 |
+
"save_steps": 10000,
|
| 2777 |
+
"stateful_callbacks": {
|
| 2778 |
+
"TrainerControl": {
|
| 2779 |
+
"args": {
|
| 2780 |
+
"should_epoch_stop": false,
|
| 2781 |
+
"should_evaluate": false,
|
| 2782 |
+
"should_log": false,
|
| 2783 |
+
"should_save": true,
|
| 2784 |
+
"should_training_stop": true
|
| 2785 |
+
},
|
| 2786 |
+
"attributes": {}
|
| 2787 |
+
}
|
| 2788 |
+
},
|
| 2789 |
+
"total_flos": 5.206682005553e+17,
|
| 2790 |
+
"train_batch_size": 64,
|
| 2791 |
+
"trial_name": null,
|
| 2792 |
+
"trial_params": null
|
| 2793 |
+
}
|