Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +2058 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,2058 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2880,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.034722222222222224,
|
| 13 |
+
"grad_norm": 12.390419960021973,
|
| 14 |
+
"learning_rate": 6.944444444444445e-06,
|
| 15 |
+
"loss": 1.8579,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.06944444444444445,
|
| 20 |
+
"grad_norm": 7.292122840881348,
|
| 21 |
+
"learning_rate": 1.388888888888889e-05,
|
| 22 |
+
"loss": 1.409,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.10416666666666667,
|
| 27 |
+
"grad_norm": 3.3312809467315674,
|
| 28 |
+
"learning_rate": 2.0833333333333336e-05,
|
| 29 |
+
"loss": 0.4742,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.1388888888888889,
|
| 34 |
+
"grad_norm": 1.8886680603027344,
|
| 35 |
+
"learning_rate": 2.777777777777778e-05,
|
| 36 |
+
"loss": 0.3359,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.1736111111111111,
|
| 41 |
+
"grad_norm": 0.9730450510978699,
|
| 42 |
+
"learning_rate": 3.472222222222222e-05,
|
| 43 |
+
"loss": 0.2133,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.20833333333333334,
|
| 48 |
+
"grad_norm": 1.1679153442382812,
|
| 49 |
+
"learning_rate": 4.166666666666667e-05,
|
| 50 |
+
"loss": 0.1603,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.24305555555555555,
|
| 55 |
+
"grad_norm": 4.400736331939697,
|
| 56 |
+
"learning_rate": 4.8611111111111115e-05,
|
| 57 |
+
"loss": 0.1299,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.2777777777777778,
|
| 62 |
+
"grad_norm": 1.0059056282043457,
|
| 63 |
+
"learning_rate": 5.555555555555556e-05,
|
| 64 |
+
"loss": 0.1123,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.3125,
|
| 69 |
+
"grad_norm": 0.6611884832382202,
|
| 70 |
+
"learning_rate": 6.25e-05,
|
| 71 |
+
"loss": 0.0934,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.3472222222222222,
|
| 76 |
+
"grad_norm": 0.879301130771637,
|
| 77 |
+
"learning_rate": 6.944444444444444e-05,
|
| 78 |
+
"loss": 0.0746,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.3819444444444444,
|
| 83 |
+
"grad_norm": 0.8346568942070007,
|
| 84 |
+
"learning_rate": 7.638888888888889e-05,
|
| 85 |
+
"loss": 0.0692,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.4166666666666667,
|
| 90 |
+
"grad_norm": 0.5676056742668152,
|
| 91 |
+
"learning_rate": 8.333333333333334e-05,
|
| 92 |
+
"loss": 0.0664,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.4513888888888889,
|
| 97 |
+
"grad_norm": 0.6186522245407104,
|
| 98 |
+
"learning_rate": 9.027777777777779e-05,
|
| 99 |
+
"loss": 0.0523,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.4861111111111111,
|
| 104 |
+
"grad_norm": 0.618405282497406,
|
| 105 |
+
"learning_rate": 9.722222222222223e-05,
|
| 106 |
+
"loss": 0.0596,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.5208333333333334,
|
| 111 |
+
"grad_norm": 0.9579340815544128,
|
| 112 |
+
"learning_rate": 9.999881338905204e-05,
|
| 113 |
+
"loss": 0.0586,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.5555555555555556,
|
| 118 |
+
"grad_norm": 0.6852683424949646,
|
| 119 |
+
"learning_rate": 9.999156208166614e-05,
|
| 120 |
+
"loss": 0.054,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.5902777777777778,
|
| 125 |
+
"grad_norm": 0.6845273375511169,
|
| 126 |
+
"learning_rate": 9.997771965008657e-05,
|
| 127 |
+
"loss": 0.0568,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.625,
|
| 132 |
+
"grad_norm": 0.49688324332237244,
|
| 133 |
+
"learning_rate": 9.995728791936504e-05,
|
| 134 |
+
"loss": 0.0541,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.6597222222222222,
|
| 139 |
+
"grad_norm": 0.6374590992927551,
|
| 140 |
+
"learning_rate": 9.993026958331766e-05,
|
| 141 |
+
"loss": 0.047,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.6944444444444444,
|
| 146 |
+
"grad_norm": 0.746904730796814,
|
| 147 |
+
"learning_rate": 9.989666820416974e-05,
|
| 148 |
+
"loss": 0.0514,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.7291666666666666,
|
| 153 |
+
"grad_norm": 0.5761491656303406,
|
| 154 |
+
"learning_rate": 9.985648821208616e-05,
|
| 155 |
+
"loss": 0.0357,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.7638888888888888,
|
| 160 |
+
"grad_norm": 0.401060551404953,
|
| 161 |
+
"learning_rate": 9.980973490458728e-05,
|
| 162 |
+
"loss": 0.0321,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.7986111111111112,
|
| 167 |
+
"grad_norm": 0.3264583349227905,
|
| 168 |
+
"learning_rate": 9.97564144458505e-05,
|
| 169 |
+
"loss": 0.0351,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.8333333333333334,
|
| 174 |
+
"grad_norm": 0.33728697896003723,
|
| 175 |
+
"learning_rate": 9.969653386589748e-05,
|
| 176 |
+
"loss": 0.0327,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.8680555555555556,
|
| 181 |
+
"grad_norm": 0.3204336166381836,
|
| 182 |
+
"learning_rate": 9.963010105966736e-05,
|
| 183 |
+
"loss": 0.0353,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.9027777777777778,
|
| 188 |
+
"grad_norm": 0.6982614994049072,
|
| 189 |
+
"learning_rate": 9.955712478597579e-05,
|
| 190 |
+
"loss": 0.0271,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.9375,
|
| 195 |
+
"grad_norm": 0.5134708285331726,
|
| 196 |
+
"learning_rate": 9.947761466636014e-05,
|
| 197 |
+
"loss": 0.0319,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.9722222222222222,
|
| 202 |
+
"grad_norm": 0.41447389125823975,
|
| 203 |
+
"learning_rate": 9.939158118381098e-05,
|
| 204 |
+
"loss": 0.0275,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 1.0069444444444444,
|
| 209 |
+
"grad_norm": 0.36693140864372253,
|
| 210 |
+
"learning_rate": 9.929903568138989e-05,
|
| 211 |
+
"loss": 0.0275,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 1.0416666666666667,
|
| 216 |
+
"grad_norm": 0.3795158863067627,
|
| 217 |
+
"learning_rate": 9.9199990360734e-05,
|
| 218 |
+
"loss": 0.0344,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 1.0763888888888888,
|
| 223 |
+
"grad_norm": 0.2937488853931427,
|
| 224 |
+
"learning_rate": 9.909445828044727e-05,
|
| 225 |
+
"loss": 0.034,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 1.1111111111111112,
|
| 230 |
+
"grad_norm": 0.3288098871707916,
|
| 231 |
+
"learning_rate": 9.89824533543787e-05,
|
| 232 |
+
"loss": 0.0259,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.1458333333333333,
|
| 237 |
+
"grad_norm": 0.27956312894821167,
|
| 238 |
+
"learning_rate": 9.886399034978797e-05,
|
| 239 |
+
"loss": 0.036,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.1805555555555556,
|
| 244 |
+
"grad_norm": 0.3101339638233185,
|
| 245 |
+
"learning_rate": 9.873908488539836e-05,
|
| 246 |
+
"loss": 0.03,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.2152777777777777,
|
| 251 |
+
"grad_norm": 0.3746699094772339,
|
| 252 |
+
"learning_rate": 9.86077534293376e-05,
|
| 253 |
+
"loss": 0.0292,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.25,
|
| 258 |
+
"grad_norm": 0.3677271604537964,
|
| 259 |
+
"learning_rate": 9.847001329696653e-05,
|
| 260 |
+
"loss": 0.0264,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.2847222222222223,
|
| 265 |
+
"grad_norm": 0.3505748212337494,
|
| 266 |
+
"learning_rate": 9.832588264859624e-05,
|
| 267 |
+
"loss": 0.0283,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.3194444444444444,
|
| 272 |
+
"grad_norm": 0.2949109673500061,
|
| 273 |
+
"learning_rate": 9.817538048709367e-05,
|
| 274 |
+
"loss": 0.0252,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.3541666666666667,
|
| 279 |
+
"grad_norm": 0.2589399516582489,
|
| 280 |
+
"learning_rate": 9.801852665537627e-05,
|
| 281 |
+
"loss": 0.0258,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.3888888888888888,
|
| 286 |
+
"grad_norm": 0.36217331886291504,
|
| 287 |
+
"learning_rate": 9.785534183379572e-05,
|
| 288 |
+
"loss": 0.0252,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.4236111111111112,
|
| 293 |
+
"grad_norm": 0.2563369870185852,
|
| 294 |
+
"learning_rate": 9.768584753741134e-05,
|
| 295 |
+
"loss": 0.0266,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.4583333333333333,
|
| 300 |
+
"grad_norm": 0.43410396575927734,
|
| 301 |
+
"learning_rate": 9.751006611315356e-05,
|
| 302 |
+
"loss": 0.0239,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.4930555555555556,
|
| 307 |
+
"grad_norm": 0.5142470002174377,
|
| 308 |
+
"learning_rate": 9.732802073687745e-05,
|
| 309 |
+
"loss": 0.025,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.5277777777777777,
|
| 314 |
+
"grad_norm": 0.31519609689712524,
|
| 315 |
+
"learning_rate": 9.713973541030716e-05,
|
| 316 |
+
"loss": 0.027,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.5625,
|
| 321 |
+
"grad_norm": 0.25440528988838196,
|
| 322 |
+
"learning_rate": 9.694523495787149e-05,
|
| 323 |
+
"loss": 0.0239,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.5972222222222223,
|
| 328 |
+
"grad_norm": 0.31716373562812805,
|
| 329 |
+
"learning_rate": 9.674454502343076e-05,
|
| 330 |
+
"loss": 0.0297,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.6319444444444444,
|
| 335 |
+
"grad_norm": 0.31023451685905457,
|
| 336 |
+
"learning_rate": 9.653769206689595e-05,
|
| 337 |
+
"loss": 0.024,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.6666666666666665,
|
| 342 |
+
"grad_norm": 0.16677865386009216,
|
| 343 |
+
"learning_rate": 9.632470336074009e-05,
|
| 344 |
+
"loss": 0.0252,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.7013888888888888,
|
| 349 |
+
"grad_norm": 0.309967577457428,
|
| 350 |
+
"learning_rate": 9.610560698640241e-05,
|
| 351 |
+
"loss": 0.0265,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.7361111111111112,
|
| 356 |
+
"grad_norm": 0.2867221534252167,
|
| 357 |
+
"learning_rate": 9.588043183058606e-05,
|
| 358 |
+
"loss": 0.0191,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.7708333333333335,
|
| 363 |
+
"grad_norm": 0.19987526535987854,
|
| 364 |
+
"learning_rate": 9.564920758144951e-05,
|
| 365 |
+
"loss": 0.0238,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.8055555555555556,
|
| 370 |
+
"grad_norm": 0.4381541609764099,
|
| 371 |
+
"learning_rate": 9.541196472469233e-05,
|
| 372 |
+
"loss": 0.0262,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.8402777777777777,
|
| 377 |
+
"grad_norm": 0.31020158529281616,
|
| 378 |
+
"learning_rate": 9.51687345395358e-05,
|
| 379 |
+
"loss": 0.0235,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.875,
|
| 384 |
+
"grad_norm": 0.33004629611968994,
|
| 385 |
+
"learning_rate": 9.491954909459895e-05,
|
| 386 |
+
"loss": 0.0234,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.9097222222222223,
|
| 391 |
+
"grad_norm": 0.39373746514320374,
|
| 392 |
+
"learning_rate": 9.466444124367041e-05,
|
| 393 |
+
"loss": 0.0237,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.9444444444444444,
|
| 398 |
+
"grad_norm": 0.43140795826911926,
|
| 399 |
+
"learning_rate": 9.440344462137689e-05,
|
| 400 |
+
"loss": 0.0284,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.9791666666666665,
|
| 405 |
+
"grad_norm": 0.37478235363960266,
|
| 406 |
+
"learning_rate": 9.413659363874853e-05,
|
| 407 |
+
"loss": 0.0225,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 2.013888888888889,
|
| 412 |
+
"grad_norm": 0.2924532890319824,
|
| 413 |
+
"learning_rate": 9.386392347868215e-05,
|
| 414 |
+
"loss": 0.0189,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 2.048611111111111,
|
| 419 |
+
"grad_norm": 0.20395496487617493,
|
| 420 |
+
"learning_rate": 9.358547009130237e-05,
|
| 421 |
+
"loss": 0.0254,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 2.0833333333333335,
|
| 426 |
+
"grad_norm": 0.44635412096977234,
|
| 427 |
+
"learning_rate": 9.330127018922194e-05,
|
| 428 |
+
"loss": 0.0226,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 2.1180555555555554,
|
| 433 |
+
"grad_norm": 0.23846730589866638,
|
| 434 |
+
"learning_rate": 9.301136124270126e-05,
|
| 435 |
+
"loss": 0.0228,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 2.1527777777777777,
|
| 440 |
+
"grad_norm": 0.24774909019470215,
|
| 441 |
+
"learning_rate": 9.271578147470825e-05,
|
| 442 |
+
"loss": 0.0254,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 2.1875,
|
| 447 |
+
"grad_norm": 0.3856460154056549,
|
| 448 |
+
"learning_rate": 9.241456985587868e-05,
|
| 449 |
+
"loss": 0.0197,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 2.2222222222222223,
|
| 454 |
+
"grad_norm": 0.30161726474761963,
|
| 455 |
+
"learning_rate": 9.210776609937829e-05,
|
| 456 |
+
"loss": 0.0203,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 2.2569444444444446,
|
| 461 |
+
"grad_norm": 0.38972583413124084,
|
| 462 |
+
"learning_rate": 9.179541065566664e-05,
|
| 463 |
+
"loss": 0.0252,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.2916666666666665,
|
| 468 |
+
"grad_norm": 0.1908106803894043,
|
| 469 |
+
"learning_rate": 9.147754470716408e-05,
|
| 470 |
+
"loss": 0.0257,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 2.326388888888889,
|
| 475 |
+
"grad_norm": 0.23305340111255646,
|
| 476 |
+
"learning_rate": 9.115421016282193e-05,
|
| 477 |
+
"loss": 0.0202,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 2.361111111111111,
|
| 482 |
+
"grad_norm": 0.39377719163894653,
|
| 483 |
+
"learning_rate": 9.082544965259715e-05,
|
| 484 |
+
"loss": 0.028,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 2.3958333333333335,
|
| 489 |
+
"grad_norm": 0.29361796379089355,
|
| 490 |
+
"learning_rate": 9.049130652183166e-05,
|
| 491 |
+
"loss": 0.0166,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 2.4305555555555554,
|
| 496 |
+
"grad_norm": 0.3843666613101959,
|
| 497 |
+
"learning_rate": 9.01518248255376e-05,
|
| 498 |
+
"loss": 0.0252,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 2.4652777777777777,
|
| 503 |
+
"grad_norm": 0.31173354387283325,
|
| 504 |
+
"learning_rate": 8.980704932258878e-05,
|
| 505 |
+
"loss": 0.0217,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.5,
|
| 510 |
+
"grad_norm": 0.31975364685058594,
|
| 511 |
+
"learning_rate": 8.945702546981969e-05,
|
| 512 |
+
"loss": 0.0189,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.5347222222222223,
|
| 517 |
+
"grad_norm": 0.20601293444633484,
|
| 518 |
+
"learning_rate": 8.910179941603193e-05,
|
| 519 |
+
"loss": 0.0218,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 2.5694444444444446,
|
| 524 |
+
"grad_norm": 0.30430278182029724,
|
| 525 |
+
"learning_rate": 8.874141799591007e-05,
|
| 526 |
+
"loss": 0.0257,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.6041666666666665,
|
| 531 |
+
"grad_norm": 0.2161753624677658,
|
| 532 |
+
"learning_rate": 8.837592872384644e-05,
|
| 533 |
+
"loss": 0.024,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.638888888888889,
|
| 538 |
+
"grad_norm": 0.5043374300003052,
|
| 539 |
+
"learning_rate": 8.800537978767682e-05,
|
| 540 |
+
"loss": 0.0218,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.673611111111111,
|
| 545 |
+
"grad_norm": 0.3008856773376465,
|
| 546 |
+
"learning_rate": 8.762982004232709e-05,
|
| 547 |
+
"loss": 0.0198,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.7083333333333335,
|
| 552 |
+
"grad_norm": 0.32244426012039185,
|
| 553 |
+
"learning_rate": 8.724929900337186e-05,
|
| 554 |
+
"loss": 0.0269,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.7430555555555554,
|
| 559 |
+
"grad_norm": 0.35441824793815613,
|
| 560 |
+
"learning_rate": 8.68638668405062e-05,
|
| 561 |
+
"loss": 0.0214,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.7777777777777777,
|
| 566 |
+
"grad_norm": 0.2707558572292328,
|
| 567 |
+
"learning_rate": 8.647357437093105e-05,
|
| 568 |
+
"loss": 0.0215,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.8125,
|
| 573 |
+
"grad_norm": 0.29439160227775574,
|
| 574 |
+
"learning_rate": 8.60784730526531e-05,
|
| 575 |
+
"loss": 0.0175,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.8472222222222223,
|
| 580 |
+
"grad_norm": 0.2966266870498657,
|
| 581 |
+
"learning_rate": 8.567861497770053e-05,
|
| 582 |
+
"loss": 0.0204,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.8819444444444446,
|
| 587 |
+
"grad_norm": 0.27358055114746094,
|
| 588 |
+
"learning_rate": 8.527405286525475e-05,
|
| 589 |
+
"loss": 0.0216,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.9166666666666665,
|
| 594 |
+
"grad_norm": 0.25975996255874634,
|
| 595 |
+
"learning_rate": 8.486484005469977e-05,
|
| 596 |
+
"loss": 0.0196,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.951388888888889,
|
| 601 |
+
"grad_norm": 0.3131217062473297,
|
| 602 |
+
"learning_rate": 8.445103049858966e-05,
|
| 603 |
+
"loss": 0.0165,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.986111111111111,
|
| 608 |
+
"grad_norm": 0.3194817900657654,
|
| 609 |
+
"learning_rate": 8.403267875553519e-05,
|
| 610 |
+
"loss": 0.0181,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 3.0208333333333335,
|
| 615 |
+
"grad_norm": 0.24048690497875214,
|
| 616 |
+
"learning_rate": 8.360983998301053e-05,
|
| 617 |
+
"loss": 0.019,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 3.0555555555555554,
|
| 622 |
+
"grad_norm": 0.2128559798002243,
|
| 623 |
+
"learning_rate": 8.318256993008107e-05,
|
| 624 |
+
"loss": 0.0181,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 3.0902777777777777,
|
| 629 |
+
"grad_norm": 0.22403229773044586,
|
| 630 |
+
"learning_rate": 8.275092493005321e-05,
|
| 631 |
+
"loss": 0.0185,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 3.125,
|
| 636 |
+
"grad_norm": 0.24800759553909302,
|
| 637 |
+
"learning_rate": 8.231496189304704e-05,
|
| 638 |
+
"loss": 0.0211,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 3.1597222222222223,
|
| 643 |
+
"grad_norm": 0.187672957777977,
|
| 644 |
+
"learning_rate": 8.187473829849314e-05,
|
| 645 |
+
"loss": 0.0173,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 3.1944444444444446,
|
| 650 |
+
"grad_norm": 0.2035113275051117,
|
| 651 |
+
"learning_rate": 8.14303121875541e-05,
|
| 652 |
+
"loss": 0.0187,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 3.2291666666666665,
|
| 657 |
+
"grad_norm": 0.32637548446655273,
|
| 658 |
+
"learning_rate": 8.098174215547224e-05,
|
| 659 |
+
"loss": 0.018,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 3.263888888888889,
|
| 664 |
+
"grad_norm": 0.2666160762310028,
|
| 665 |
+
"learning_rate": 8.052908734384399e-05,
|
| 666 |
+
"loss": 0.0165,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 3.298611111111111,
|
| 671 |
+
"grad_norm": 0.29945555329322815,
|
| 672 |
+
"learning_rate": 8.007240743282247e-05,
|
| 673 |
+
"loss": 0.0177,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 3.3333333333333335,
|
| 678 |
+
"grad_norm": 0.4292016923427582,
|
| 679 |
+
"learning_rate": 7.961176263324901e-05,
|
| 680 |
+
"loss": 0.018,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 3.3680555555555554,
|
| 685 |
+
"grad_norm": 0.387088418006897,
|
| 686 |
+
"learning_rate": 7.914721367871454e-05,
|
| 687 |
+
"loss": 0.0208,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 3.4027777777777777,
|
| 692 |
+
"grad_norm": 0.4121212065219879,
|
| 693 |
+
"learning_rate": 7.86788218175523e-05,
|
| 694 |
+
"loss": 0.0199,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 3.4375,
|
| 699 |
+
"grad_norm": 0.3326111137866974,
|
| 700 |
+
"learning_rate": 7.820664880476256e-05,
|
| 701 |
+
"loss": 0.0176,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 3.4722222222222223,
|
| 706 |
+
"grad_norm": 0.3561447262763977,
|
| 707 |
+
"learning_rate": 7.773075689387043e-05,
|
| 708 |
+
"loss": 0.0164,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 3.5069444444444446,
|
| 713 |
+
"grad_norm": 0.3860738277435303,
|
| 714 |
+
"learning_rate": 7.725120882871828e-05,
|
| 715 |
+
"loss": 0.021,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 3.5416666666666665,
|
| 720 |
+
"grad_norm": 0.2797831594944,
|
| 721 |
+
"learning_rate": 7.676806783519304e-05,
|
| 722 |
+
"loss": 0.0185,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 3.576388888888889,
|
| 727 |
+
"grad_norm": 0.31011706590652466,
|
| 728 |
+
"learning_rate": 7.628139761289038e-05,
|
| 729 |
+
"loss": 0.0168,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 3.611111111111111,
|
| 734 |
+
"grad_norm": 0.2268083244562149,
|
| 735 |
+
"learning_rate": 7.579126232671621e-05,
|
| 736 |
+
"loss": 0.0163,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 3.6458333333333335,
|
| 741 |
+
"grad_norm": 0.25035360455513,
|
| 742 |
+
"learning_rate": 7.529772659842685e-05,
|
| 743 |
+
"loss": 0.0181,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 3.6805555555555554,
|
| 748 |
+
"grad_norm": 0.401965856552124,
|
| 749 |
+
"learning_rate": 7.480085549810897e-05,
|
| 750 |
+
"loss": 0.0164,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 3.7152777777777777,
|
| 755 |
+
"grad_norm": 0.23698702454566956,
|
| 756 |
+
"learning_rate": 7.430071453560059e-05,
|
| 757 |
+
"loss": 0.0163,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 3.75,
|
| 762 |
+
"grad_norm": 0.22072561085224152,
|
| 763 |
+
"learning_rate": 7.379736965185368e-05,
|
| 764 |
+
"loss": 0.0193,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 3.7847222222222223,
|
| 769 |
+
"grad_norm": 0.2923282980918884,
|
| 770 |
+
"learning_rate": 7.32908872102405e-05,
|
| 771 |
+
"loss": 0.0192,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 3.8194444444444446,
|
| 776 |
+
"grad_norm": 0.3529258668422699,
|
| 777 |
+
"learning_rate": 7.278133398780366e-05,
|
| 778 |
+
"loss": 0.0189,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 3.8541666666666665,
|
| 783 |
+
"grad_norm": 0.2745247483253479,
|
| 784 |
+
"learning_rate": 7.226877716645217e-05,
|
| 785 |
+
"loss": 0.0198,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 3.888888888888889,
|
| 790 |
+
"grad_norm": 0.3626753091812134,
|
| 791 |
+
"learning_rate": 7.175328432410366e-05,
|
| 792 |
+
"loss": 0.0157,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 3.923611111111111,
|
| 797 |
+
"grad_norm": 0.24702580273151398,
|
| 798 |
+
"learning_rate": 7.123492342577472e-05,
|
| 799 |
+
"loss": 0.015,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 3.9583333333333335,
|
| 804 |
+
"grad_norm": 0.2592775523662567,
|
| 805 |
+
"learning_rate": 7.071376281461994e-05,
|
| 806 |
+
"loss": 0.0224,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 3.9930555555555554,
|
| 811 |
+
"grad_norm": 0.21476583182811737,
|
| 812 |
+
"learning_rate": 7.018987120292141e-05,
|
| 813 |
+
"loss": 0.0199,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 4.027777777777778,
|
| 818 |
+
"grad_norm": 0.47759580612182617,
|
| 819 |
+
"learning_rate": 6.966331766302916e-05,
|
| 820 |
+
"loss": 0.02,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 4.0625,
|
| 825 |
+
"grad_norm": 0.5051742196083069,
|
| 826 |
+
"learning_rate": 6.91341716182545e-05,
|
| 827 |
+
"loss": 0.0205,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 4.097222222222222,
|
| 832 |
+
"grad_norm": 0.2376144826412201,
|
| 833 |
+
"learning_rate": 6.860250283371681e-05,
|
| 834 |
+
"loss": 0.0159,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 4.131944444444445,
|
| 839 |
+
"grad_norm": 0.2458970993757248,
|
| 840 |
+
"learning_rate": 6.806838140714545e-05,
|
| 841 |
+
"loss": 0.0161,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 4.166666666666667,
|
| 846 |
+
"grad_norm": 0.1450349986553192,
|
| 847 |
+
"learning_rate": 6.753187775963773e-05,
|
| 848 |
+
"loss": 0.0138,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 4.201388888888889,
|
| 853 |
+
"grad_norm": 0.16866087913513184,
|
| 854 |
+
"learning_rate": 6.69930626263742e-05,
|
| 855 |
+
"loss": 0.0171,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 4.236111111111111,
|
| 860 |
+
"grad_norm": 0.24081069231033325,
|
| 861 |
+
"learning_rate": 6.645200704729266e-05,
|
| 862 |
+
"loss": 0.0181,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 4.270833333333333,
|
| 867 |
+
"grad_norm": 0.2583770155906677,
|
| 868 |
+
"learning_rate": 6.590878235772191e-05,
|
| 869 |
+
"loss": 0.0145,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 4.305555555555555,
|
| 874 |
+
"grad_norm": 0.3081924319267273,
|
| 875 |
+
"learning_rate": 6.536346017897653e-05,
|
| 876 |
+
"loss": 0.016,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 4.340277777777778,
|
| 881 |
+
"grad_norm": 0.26682665944099426,
|
| 882 |
+
"learning_rate": 6.481611240891396e-05,
|
| 883 |
+
"loss": 0.0185,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 4.375,
|
| 888 |
+
"grad_norm": 0.346853107213974,
|
| 889 |
+
"learning_rate": 6.426681121245527e-05,
|
| 890 |
+
"loss": 0.0161,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 4.409722222222222,
|
| 895 |
+
"grad_norm": 0.20850667357444763,
|
| 896 |
+
"learning_rate": 6.371562901207045e-05,
|
| 897 |
+
"loss": 0.0126,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 4.444444444444445,
|
| 902 |
+
"grad_norm": 0.19474495947360992,
|
| 903 |
+
"learning_rate": 6.316263847822997e-05,
|
| 904 |
+
"loss": 0.0158,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 4.479166666666667,
|
| 909 |
+
"grad_norm": 0.32026270031929016,
|
| 910 |
+
"learning_rate": 6.260791251982354e-05,
|
| 911 |
+
"loss": 0.0174,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 4.513888888888889,
|
| 916 |
+
"grad_norm": 0.2859480082988739,
|
| 917 |
+
"learning_rate": 6.205152427454747e-05,
|
| 918 |
+
"loss": 0.0188,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 4.548611111111111,
|
| 923 |
+
"grad_norm": 0.19467797875404358,
|
| 924 |
+
"learning_rate": 6.149354709926185e-05,
|
| 925 |
+
"loss": 0.0141,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 4.583333333333333,
|
| 930 |
+
"grad_norm": 0.2192724645137787,
|
| 931 |
+
"learning_rate": 6.09340545603188e-05,
|
| 932 |
+
"loss": 0.0149,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 4.618055555555555,
|
| 937 |
+
"grad_norm": 0.2860821783542633,
|
| 938 |
+
"learning_rate": 6.037312042386314e-05,
|
| 939 |
+
"loss": 0.0222,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 4.652777777777778,
|
| 944 |
+
"grad_norm": 0.16547256708145142,
|
| 945 |
+
"learning_rate": 5.98108186461068e-05,
|
| 946 |
+
"loss": 0.0156,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 4.6875,
|
| 951 |
+
"grad_norm": 0.21727648377418518,
|
| 952 |
+
"learning_rate": 5.924722336357793e-05,
|
| 953 |
+
"loss": 0.0146,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 4.722222222222222,
|
| 958 |
+
"grad_norm": 0.2452634871006012,
|
| 959 |
+
"learning_rate": 5.868240888334653e-05,
|
| 960 |
+
"loss": 0.0164,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 4.756944444444445,
|
| 965 |
+
"grad_norm": 0.22284066677093506,
|
| 966 |
+
"learning_rate": 5.811644967322737e-05,
|
| 967 |
+
"loss": 0.0156,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 4.791666666666667,
|
| 972 |
+
"grad_norm": 0.2330964207649231,
|
| 973 |
+
"learning_rate": 5.7549420351961844e-05,
|
| 974 |
+
"loss": 0.0138,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 4.826388888888889,
|
| 979 |
+
"grad_norm": 0.20147477090358734,
|
| 980 |
+
"learning_rate": 5.698139567937988e-05,
|
| 981 |
+
"loss": 0.0147,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 4.861111111111111,
|
| 986 |
+
"grad_norm": 0.36548295617103577,
|
| 987 |
+
"learning_rate": 5.641245054654316e-05,
|
| 988 |
+
"loss": 0.0154,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 4.895833333333333,
|
| 993 |
+
"grad_norm": 0.20836779475212097,
|
| 994 |
+
"learning_rate": 5.584265996587129e-05,
|
| 995 |
+
"loss": 0.012,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 4.930555555555555,
|
| 1000 |
+
"grad_norm": 0.21215607225894928,
|
| 1001 |
+
"learning_rate": 5.527209906125169e-05,
|
| 1002 |
+
"loss": 0.0119,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 4.965277777777778,
|
| 1007 |
+
"grad_norm": 0.2577287256717682,
|
| 1008 |
+
"learning_rate": 5.470084305813492e-05,
|
| 1009 |
+
"loss": 0.0171,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 5.0,
|
| 1014 |
+
"grad_norm": 0.189128577709198,
|
| 1015 |
+
"learning_rate": 5.4128967273616625e-05,
|
| 1016 |
+
"loss": 0.0129,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 5.034722222222222,
|
| 1021 |
+
"grad_norm": 0.26594117283821106,
|
| 1022 |
+
"learning_rate": 5.355654710650737e-05,
|
| 1023 |
+
"loss": 0.0122,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 5.069444444444445,
|
| 1028 |
+
"grad_norm": 0.1738700121641159,
|
| 1029 |
+
"learning_rate": 5.29836580273917e-05,
|
| 1030 |
+
"loss": 0.0125,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 5.104166666666667,
|
| 1035 |
+
"grad_norm": 0.17082342505455017,
|
| 1036 |
+
"learning_rate": 5.241037556867775e-05,
|
| 1037 |
+
"loss": 0.0149,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 5.138888888888889,
|
| 1042 |
+
"grad_norm": 0.3401947319507599,
|
| 1043 |
+
"learning_rate": 5.183677531463863e-05,
|
| 1044 |
+
"loss": 0.0137,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 5.173611111111111,
|
| 1049 |
+
"grad_norm": 0.2809308171272278,
|
| 1050 |
+
"learning_rate": 5.126293289144715e-05,
|
| 1051 |
+
"loss": 0.0147,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 5.208333333333333,
|
| 1056 |
+
"grad_norm": 0.2079499512910843,
|
| 1057 |
+
"learning_rate": 5.068892395720483e-05,
|
| 1058 |
+
"loss": 0.012,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 5.243055555555555,
|
| 1063 |
+
"grad_norm": 0.3077143430709839,
|
| 1064 |
+
"learning_rate": 5.011482419196677e-05,
|
| 1065 |
+
"loss": 0.0183,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 5.277777777777778,
|
| 1070 |
+
"grad_norm": 0.2815386950969696,
|
| 1071 |
+
"learning_rate": 4.9540709287763685e-05,
|
| 1072 |
+
"loss": 0.0186,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 5.3125,
|
| 1077 |
+
"grad_norm": 0.2164885401725769,
|
| 1078 |
+
"learning_rate": 4.8966654938622295e-05,
|
| 1079 |
+
"loss": 0.0182,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 5.347222222222222,
|
| 1084 |
+
"grad_norm": 0.22151343524456024,
|
| 1085 |
+
"learning_rate": 4.839273683058542e-05,
|
| 1086 |
+
"loss": 0.0166,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 5.381944444444445,
|
| 1091 |
+
"grad_norm": 0.23041661083698273,
|
| 1092 |
+
"learning_rate": 4.781903063173321e-05,
|
| 1093 |
+
"loss": 0.0145,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 5.416666666666667,
|
| 1098 |
+
"grad_norm": 0.2047954499721527,
|
| 1099 |
+
"learning_rate": 4.7245611982206724e-05,
|
| 1100 |
+
"loss": 0.0123,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 5.451388888888889,
|
| 1105 |
+
"grad_norm": 0.21224728226661682,
|
| 1106 |
+
"learning_rate": 4.6672556484235174e-05,
|
| 1107 |
+
"loss": 0.0105,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 5.486111111111111,
|
| 1112 |
+
"grad_norm": 0.13932043313980103,
|
| 1113 |
+
"learning_rate": 4.6099939692168107e-05,
|
| 1114 |
+
"loss": 0.0134,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 5.520833333333333,
|
| 1119 |
+
"grad_norm": 0.2085762619972229,
|
| 1120 |
+
"learning_rate": 4.5527837102514035e-05,
|
| 1121 |
+
"loss": 0.0118,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 5.555555555555555,
|
| 1126 |
+
"grad_norm": 0.1945631057024002,
|
| 1127 |
+
"learning_rate": 4.4956324143986596e-05,
|
| 1128 |
+
"loss": 0.0162,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 5.590277777777778,
|
| 1133 |
+
"grad_norm": 0.2890234887599945,
|
| 1134 |
+
"learning_rate": 4.438547616755962e-05,
|
| 1135 |
+
"loss": 0.0143,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 5.625,
|
| 1140 |
+
"grad_norm": 0.212904155254364,
|
| 1141 |
+
"learning_rate": 4.381536843653262e-05,
|
| 1142 |
+
"loss": 0.0142,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 5.659722222222222,
|
| 1147 |
+
"grad_norm": 0.18237364292144775,
|
| 1148 |
+
"learning_rate": 4.3246076116607604e-05,
|
| 1149 |
+
"loss": 0.0092,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 5.694444444444445,
|
| 1154 |
+
"grad_norm": 0.22255735099315643,
|
| 1155 |
+
"learning_rate": 4.267767426597893e-05,
|
| 1156 |
+
"loss": 0.0129,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 5.729166666666667,
|
| 1161 |
+
"grad_norm": 0.18430383503437042,
|
| 1162 |
+
"learning_rate": 4.211023782543727e-05,
|
| 1163 |
+
"loss": 0.0115,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 5.763888888888889,
|
| 1168 |
+
"grad_norm": 0.3231765329837799,
|
| 1169 |
+
"learning_rate": 4.1543841608489075e-05,
|
| 1170 |
+
"loss": 0.0142,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 5.798611111111111,
|
| 1175 |
+
"grad_norm": 0.22352637350559235,
|
| 1176 |
+
"learning_rate": 4.097856029149276e-05,
|
| 1177 |
+
"loss": 0.0111,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 5.833333333333333,
|
| 1182 |
+
"grad_norm": 0.17798005044460297,
|
| 1183 |
+
"learning_rate": 4.0414468403813095e-05,
|
| 1184 |
+
"loss": 0.0144,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 5.868055555555555,
|
| 1189 |
+
"grad_norm": 0.26681026816368103,
|
| 1190 |
+
"learning_rate": 3.985164031799481e-05,
|
| 1191 |
+
"loss": 0.012,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 5.902777777777778,
|
| 1196 |
+
"grad_norm": 0.2966468632221222,
|
| 1197 |
+
"learning_rate": 3.9290150239957057e-05,
|
| 1198 |
+
"loss": 0.0102,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 5.9375,
|
| 1203 |
+
"grad_norm": 0.2999604344367981,
|
| 1204 |
+
"learning_rate": 3.87300721992097e-05,
|
| 1205 |
+
"loss": 0.011,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 5.972222222222222,
|
| 1210 |
+
"grad_norm": 0.29702338576316833,
|
| 1211 |
+
"learning_rate": 3.817148003909288e-05,
|
| 1212 |
+
"loss": 0.0116,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 6.006944444444445,
|
| 1217 |
+
"grad_norm": 0.2850812077522278,
|
| 1218 |
+
"learning_rate": 3.761444740704129e-05,
|
| 1219 |
+
"loss": 0.01,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 6.041666666666667,
|
| 1224 |
+
"grad_norm": 0.26815593242645264,
|
| 1225 |
+
"learning_rate": 3.705904774487396e-05,
|
| 1226 |
+
"loss": 0.0142,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 6.076388888888889,
|
| 1231 |
+
"grad_norm": 0.31307709217071533,
|
| 1232 |
+
"learning_rate": 3.6505354279111524e-05,
|
| 1233 |
+
"loss": 0.0144,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 6.111111111111111,
|
| 1238 |
+
"grad_norm": 0.18757231533527374,
|
| 1239 |
+
"learning_rate": 3.595344001132154e-05,
|
| 1240 |
+
"loss": 0.0146,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 6.145833333333333,
|
| 1245 |
+
"grad_norm": 0.21246947348117828,
|
| 1246 |
+
"learning_rate": 3.540337770849371e-05,
|
| 1247 |
+
"loss": 0.0134,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 6.180555555555555,
|
| 1252 |
+
"grad_norm": 0.16654673218727112,
|
| 1253 |
+
"learning_rate": 3.4855239893445946e-05,
|
| 1254 |
+
"loss": 0.015,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 6.215277777777778,
|
| 1259 |
+
"grad_norm": 0.1691170185804367,
|
| 1260 |
+
"learning_rate": 3.430909883526251e-05,
|
| 1261 |
+
"loss": 0.012,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 6.25,
|
| 1266 |
+
"grad_norm": 0.14390875399112701,
|
| 1267 |
+
"learning_rate": 3.3765026539765834e-05,
|
| 1268 |
+
"loss": 0.0108,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 6.284722222222222,
|
| 1273 |
+
"grad_norm": 0.19382424652576447,
|
| 1274 |
+
"learning_rate": 3.322309474002288e-05,
|
| 1275 |
+
"loss": 0.0111,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 6.319444444444445,
|
| 1280 |
+
"grad_norm": 0.1386929452419281,
|
| 1281 |
+
"learning_rate": 3.268337488688748e-05,
|
| 1282 |
+
"loss": 0.0101,
|
| 1283 |
+
"step": 1820
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 6.354166666666667,
|
| 1287 |
+
"grad_norm": 0.1915092170238495,
|
| 1288 |
+
"learning_rate": 3.214593813958001e-05,
|
| 1289 |
+
"loss": 0.0135,
|
| 1290 |
+
"step": 1830
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 6.388888888888889,
|
| 1294 |
+
"grad_norm": 0.155628964304924,
|
| 1295 |
+
"learning_rate": 3.1610855356305354e-05,
|
| 1296 |
+
"loss": 0.011,
|
| 1297 |
+
"step": 1840
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 6.423611111111111,
|
| 1301 |
+
"grad_norm": 0.194572314620018,
|
| 1302 |
+
"learning_rate": 3.107819708491059e-05,
|
| 1303 |
+
"loss": 0.0119,
|
| 1304 |
+
"step": 1850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 6.458333333333333,
|
| 1308 |
+
"grad_norm": 0.1013031154870987,
|
| 1309 |
+
"learning_rate": 3.0548033553583705e-05,
|
| 1310 |
+
"loss": 0.0122,
|
| 1311 |
+
"step": 1860
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 6.493055555555555,
|
| 1315 |
+
"grad_norm": 0.2090463936328888,
|
| 1316 |
+
"learning_rate": 3.0020434661594343e-05,
|
| 1317 |
+
"loss": 0.0138,
|
| 1318 |
+
"step": 1870
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 6.527777777777778,
|
| 1322 |
+
"grad_norm": 0.1606292426586151,
|
| 1323 |
+
"learning_rate": 2.9495469970078e-05,
|
| 1324 |
+
"loss": 0.0093,
|
| 1325 |
+
"step": 1880
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 6.5625,
|
| 1329 |
+
"grad_norm": 0.1697450429201126,
|
| 1330 |
+
"learning_rate": 2.8973208692864624e-05,
|
| 1331 |
+
"loss": 0.0132,
|
| 1332 |
+
"step": 1890
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 6.597222222222222,
|
| 1336 |
+
"grad_norm": 0.2177683413028717,
|
| 1337 |
+
"learning_rate": 2.8453719687353285e-05,
|
| 1338 |
+
"loss": 0.013,
|
| 1339 |
+
"step": 1900
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 6.631944444444445,
|
| 1343 |
+
"grad_norm": 0.26842954754829407,
|
| 1344 |
+
"learning_rate": 2.7937071445433615e-05,
|
| 1345 |
+
"loss": 0.0101,
|
| 1346 |
+
"step": 1910
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 6.666666666666667,
|
| 1350 |
+
"grad_norm": 0.34368401765823364,
|
| 1351 |
+
"learning_rate": 2.7423332084455544e-05,
|
| 1352 |
+
"loss": 0.0118,
|
| 1353 |
+
"step": 1920
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 6.701388888888889,
|
| 1357 |
+
"grad_norm": 0.19154952466487885,
|
| 1358 |
+
"learning_rate": 2.6912569338248315e-05,
|
| 1359 |
+
"loss": 0.0118,
|
| 1360 |
+
"step": 1930
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 6.736111111111111,
|
| 1364 |
+
"grad_norm": 0.2398257553577423,
|
| 1365 |
+
"learning_rate": 2.6404850548190235e-05,
|
| 1366 |
+
"loss": 0.0101,
|
| 1367 |
+
"step": 1940
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 6.770833333333333,
|
| 1371 |
+
"grad_norm": 0.2987666428089142,
|
| 1372 |
+
"learning_rate": 2.590024265433002e-05,
|
| 1373 |
+
"loss": 0.0105,
|
| 1374 |
+
"step": 1950
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 6.805555555555555,
|
| 1378 |
+
"grad_norm": 0.284682035446167,
|
| 1379 |
+
"learning_rate": 2.5398812186561095e-05,
|
| 1380 |
+
"loss": 0.0117,
|
| 1381 |
+
"step": 1960
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 6.840277777777778,
|
| 1385 |
+
"grad_norm": 0.24240490794181824,
|
| 1386 |
+
"learning_rate": 2.4900625255849986e-05,
|
| 1387 |
+
"loss": 0.0114,
|
| 1388 |
+
"step": 1970
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 6.875,
|
| 1392 |
+
"grad_norm": 0.2425372302532196,
|
| 1393 |
+
"learning_rate": 2.4405747545519963e-05,
|
| 1394 |
+
"loss": 0.0122,
|
| 1395 |
+
"step": 1980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 6.909722222222222,
|
| 1399 |
+
"grad_norm": 0.15807554125785828,
|
| 1400 |
+
"learning_rate": 2.3914244302590966e-05,
|
| 1401 |
+
"loss": 0.0102,
|
| 1402 |
+
"step": 1990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 6.944444444444445,
|
| 1406 |
+
"grad_norm": 0.2001255601644516,
|
| 1407 |
+
"learning_rate": 2.3426180329177215e-05,
|
| 1408 |
+
"loss": 0.0115,
|
| 1409 |
+
"step": 2000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 6.979166666666667,
|
| 1413 |
+
"grad_norm": 0.3653542995452881,
|
| 1414 |
+
"learning_rate": 2.294161997394336e-05,
|
| 1415 |
+
"loss": 0.0115,
|
| 1416 |
+
"step": 2010
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 7.013888888888889,
|
| 1420 |
+
"grad_norm": 0.2090686410665512,
|
| 1421 |
+
"learning_rate": 2.246062712362043e-05,
|
| 1422 |
+
"loss": 0.0111,
|
| 1423 |
+
"step": 2020
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 7.048611111111111,
|
| 1427 |
+
"grad_norm": 0.263019323348999,
|
| 1428 |
+
"learning_rate": 2.1983265194582724e-05,
|
| 1429 |
+
"loss": 0.012,
|
| 1430 |
+
"step": 2030
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 7.083333333333333,
|
| 1434 |
+
"grad_norm": 0.2306613177061081,
|
| 1435 |
+
"learning_rate": 2.150959712448669e-05,
|
| 1436 |
+
"loss": 0.0128,
|
| 1437 |
+
"step": 2040
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 7.118055555555555,
|
| 1441 |
+
"grad_norm": 0.10623705387115479,
|
| 1442 |
+
"learning_rate": 2.1039685363972934e-05,
|
| 1443 |
+
"loss": 0.0122,
|
| 1444 |
+
"step": 2050
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 7.152777777777778,
|
| 1448 |
+
"grad_norm": 0.22297905385494232,
|
| 1449 |
+
"learning_rate": 2.0573591868432406e-05,
|
| 1450 |
+
"loss": 0.0112,
|
| 1451 |
+
"step": 2060
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 7.1875,
|
| 1455 |
+
"grad_norm": 0.14223700761795044,
|
| 1456 |
+
"learning_rate": 2.0111378089837956e-05,
|
| 1457 |
+
"loss": 0.012,
|
| 1458 |
+
"step": 2070
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 7.222222222222222,
|
| 1462 |
+
"grad_norm": 0.1814931482076645,
|
| 1463 |
+
"learning_rate": 1.9653104968642173e-05,
|
| 1464 |
+
"loss": 0.0083,
|
| 1465 |
+
"step": 2080
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 7.256944444444445,
|
| 1469 |
+
"grad_norm": 0.18442519009113312,
|
| 1470 |
+
"learning_rate": 1.919883292574269e-05,
|
| 1471 |
+
"loss": 0.0138,
|
| 1472 |
+
"step": 2090
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 7.291666666666667,
|
| 1476 |
+
"grad_norm": 0.2410842329263687,
|
| 1477 |
+
"learning_rate": 1.874862185451608e-05,
|
| 1478 |
+
"loss": 0.0126,
|
| 1479 |
+
"step": 2100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 7.326388888888889,
|
| 1483 |
+
"grad_norm": 0.16032792627811432,
|
| 1484 |
+
"learning_rate": 1.8302531112921172e-05,
|
| 1485 |
+
"loss": 0.0097,
|
| 1486 |
+
"step": 2110
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 7.361111111111111,
|
| 1490 |
+
"grad_norm": 0.18864929676055908,
|
| 1491 |
+
"learning_rate": 1.7860619515673033e-05,
|
| 1492 |
+
"loss": 0.0102,
|
| 1493 |
+
"step": 2120
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 7.395833333333333,
|
| 1497 |
+
"grad_norm": 0.12309829145669937,
|
| 1498 |
+
"learning_rate": 1.7422945326488553e-05,
|
| 1499 |
+
"loss": 0.0126,
|
| 1500 |
+
"step": 2130
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 7.430555555555555,
|
| 1504 |
+
"grad_norm": 0.13348473608493805,
|
| 1505 |
+
"learning_rate": 1.6989566250404702e-05,
|
| 1506 |
+
"loss": 0.0085,
|
| 1507 |
+
"step": 2140
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 7.465277777777778,
|
| 1511 |
+
"grad_norm": 0.08756257593631744,
|
| 1512 |
+
"learning_rate": 1.656053942617036e-05,
|
| 1513 |
+
"loss": 0.0104,
|
| 1514 |
+
"step": 2150
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 7.5,
|
| 1518 |
+
"grad_norm": 0.1287485808134079,
|
| 1519 |
+
"learning_rate": 1.6135921418712956e-05,
|
| 1520 |
+
"loss": 0.0092,
|
| 1521 |
+
"step": 2160
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 7.534722222222222,
|
| 1525 |
+
"grad_norm": 0.22407540678977966,
|
| 1526 |
+
"learning_rate": 1.5715768211680647e-05,
|
| 1527 |
+
"loss": 0.0082,
|
| 1528 |
+
"step": 2170
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 7.569444444444445,
|
| 1532 |
+
"grad_norm": 0.1887049823999405,
|
| 1533 |
+
"learning_rate": 1.5300135200061168e-05,
|
| 1534 |
+
"loss": 0.0098,
|
| 1535 |
+
"step": 2180
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 7.604166666666667,
|
| 1539 |
+
"grad_norm": 0.13443368673324585,
|
| 1540 |
+
"learning_rate": 1.4889077182878269e-05,
|
| 1541 |
+
"loss": 0.0101,
|
| 1542 |
+
"step": 2190
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 7.638888888888889,
|
| 1546 |
+
"grad_norm": 0.2131468653678894,
|
| 1547 |
+
"learning_rate": 1.44826483559668e-05,
|
| 1548 |
+
"loss": 0.0116,
|
| 1549 |
+
"step": 2200
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 7.673611111111111,
|
| 1553 |
+
"grad_norm": 0.16836215555667877,
|
| 1554 |
+
"learning_rate": 1.4080902304827242e-05,
|
| 1555 |
+
"loss": 0.0094,
|
| 1556 |
+
"step": 2210
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 7.708333333333333,
|
| 1560 |
+
"grad_norm": 0.30081215500831604,
|
| 1561 |
+
"learning_rate": 1.368389199756075e-05,
|
| 1562 |
+
"loss": 0.0133,
|
| 1563 |
+
"step": 2220
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 7.743055555555555,
|
| 1567 |
+
"grad_norm": 0.20233741402626038,
|
| 1568 |
+
"learning_rate": 1.3291669777885596e-05,
|
| 1569 |
+
"loss": 0.0094,
|
| 1570 |
+
"step": 2230
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 7.777777777777778,
|
| 1574 |
+
"grad_norm": 0.1448446661233902,
|
| 1575 |
+
"learning_rate": 1.2904287358235928e-05,
|
| 1576 |
+
"loss": 0.0105,
|
| 1577 |
+
"step": 2240
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 7.8125,
|
| 1581 |
+
"grad_norm": 0.16389504075050354,
|
| 1582 |
+
"learning_rate": 1.2521795812943704e-05,
|
| 1583 |
+
"loss": 0.0082,
|
| 1584 |
+
"step": 2250
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 7.847222222222222,
|
| 1588 |
+
"grad_norm": 0.18017397820949554,
|
| 1589 |
+
"learning_rate": 1.2144245571504898e-05,
|
| 1590 |
+
"loss": 0.0092,
|
| 1591 |
+
"step": 2260
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 7.881944444444445,
|
| 1595 |
+
"grad_norm": 0.20985354483127594,
|
| 1596 |
+
"learning_rate": 1.1771686411930577e-05,
|
| 1597 |
+
"loss": 0.0093,
|
| 1598 |
+
"step": 2270
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 7.916666666666667,
|
| 1602 |
+
"grad_norm": 0.2412370890378952,
|
| 1603 |
+
"learning_rate": 1.1404167454183957e-05,
|
| 1604 |
+
"loss": 0.0082,
|
| 1605 |
+
"step": 2280
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 7.951388888888889,
|
| 1609 |
+
"grad_norm": 0.1690802127122879,
|
| 1610 |
+
"learning_rate": 1.1041737153704185e-05,
|
| 1611 |
+
"loss": 0.0112,
|
| 1612 |
+
"step": 2290
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 7.986111111111111,
|
| 1616 |
+
"grad_norm": 0.22050976753234863,
|
| 1617 |
+
"learning_rate": 1.0684443295017721e-05,
|
| 1618 |
+
"loss": 0.0078,
|
| 1619 |
+
"step": 2300
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 8.020833333333334,
|
| 1623 |
+
"grad_norm": 0.134538471698761,
|
| 1624 |
+
"learning_rate": 1.0332332985438248e-05,
|
| 1625 |
+
"loss": 0.011,
|
| 1626 |
+
"step": 2310
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 8.055555555555555,
|
| 1630 |
+
"grad_norm": 0.16212856769561768,
|
| 1631 |
+
"learning_rate": 9.985452648855803e-06,
|
| 1632 |
+
"loss": 0.0097,
|
| 1633 |
+
"step": 2320
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 8.090277777777779,
|
| 1637 |
+
"grad_norm": 0.1637708693742752,
|
| 1638 |
+
"learning_rate": 9.643848019616003e-06,
|
| 1639 |
+
"loss": 0.0094,
|
| 1640 |
+
"step": 2330
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 8.125,
|
| 1644 |
+
"grad_norm": 0.20079755783081055,
|
| 1645 |
+
"learning_rate": 9.307564136490254e-06,
|
| 1646 |
+
"loss": 0.0108,
|
| 1647 |
+
"step": 2340
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 8.159722222222221,
|
| 1651 |
+
"grad_norm": 0.2914990782737732,
|
| 1652 |
+
"learning_rate": 8.976645336737599e-06,
|
| 1653 |
+
"loss": 0.0072,
|
| 1654 |
+
"step": 2350
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 8.194444444444445,
|
| 1658 |
+
"grad_norm": 0.14318501949310303,
|
| 1659 |
+
"learning_rate": 8.651135250259091e-06,
|
| 1660 |
+
"loss": 0.0081,
|
| 1661 |
+
"step": 2360
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 8.229166666666666,
|
| 1665 |
+
"grad_norm": 0.22007636725902557,
|
| 1666 |
+
"learning_rate": 8.33107679384542e-06,
|
| 1667 |
+
"loss": 0.0076,
|
| 1668 |
+
"step": 2370
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 8.26388888888889,
|
| 1672 |
+
"grad_norm": 0.215325728058815,
|
| 1673 |
+
"learning_rate": 8.016512165518569e-06,
|
| 1674 |
+
"loss": 0.0074,
|
| 1675 |
+
"step": 2380
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 8.29861111111111,
|
| 1679 |
+
"grad_norm": 0.2510315775871277,
|
| 1680 |
+
"learning_rate": 7.707482838968222e-06,
|
| 1681 |
+
"loss": 0.0109,
|
| 1682 |
+
"step": 2390
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 8.333333333333334,
|
| 1686 |
+
"grad_norm": 0.37152716517448425,
|
| 1687 |
+
"learning_rate": 7.404029558083653e-06,
|
| 1688 |
+
"loss": 0.0105,
|
| 1689 |
+
"step": 2400
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 8.368055555555555,
|
| 1693 |
+
"grad_norm": 0.2045627236366272,
|
| 1694 |
+
"learning_rate": 7.106192331581896e-06,
|
| 1695 |
+
"loss": 0.0084,
|
| 1696 |
+
"step": 2410
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 8.402777777777779,
|
| 1700 |
+
"grad_norm": 0.26112157106399536,
|
| 1701 |
+
"learning_rate": 6.814010427732786e-06,
|
| 1702 |
+
"loss": 0.0083,
|
| 1703 |
+
"step": 2420
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 8.4375,
|
| 1707 |
+
"grad_norm": 0.18590597808361053,
|
| 1708 |
+
"learning_rate": 6.527522369181655e-06,
|
| 1709 |
+
"loss": 0.0119,
|
| 1710 |
+
"step": 2430
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 8.472222222222221,
|
| 1714 |
+
"grad_norm": 0.19280190765857697,
|
| 1715 |
+
"learning_rate": 6.246765927870313e-06,
|
| 1716 |
+
"loss": 0.0088,
|
| 1717 |
+
"step": 2440
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 8.506944444444445,
|
| 1721 |
+
"grad_norm": 0.11337297409772873,
|
| 1722 |
+
"learning_rate": 5.971778120057031e-06,
|
| 1723 |
+
"loss": 0.0102,
|
| 1724 |
+
"step": 2450
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 8.541666666666666,
|
| 1728 |
+
"grad_norm": 0.20286336541175842,
|
| 1729 |
+
"learning_rate": 5.702595201436101e-06,
|
| 1730 |
+
"loss": 0.0116,
|
| 1731 |
+
"step": 2460
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 8.57638888888889,
|
| 1735 |
+
"grad_norm": 0.18840523064136505,
|
| 1736 |
+
"learning_rate": 5.439252662357769e-06,
|
| 1737 |
+
"loss": 0.0102,
|
| 1738 |
+
"step": 2470
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 8.61111111111111,
|
| 1742 |
+
"grad_norm": 0.14088304340839386,
|
| 1743 |
+
"learning_rate": 5.181785223148999e-06,
|
| 1744 |
+
"loss": 0.0097,
|
| 1745 |
+
"step": 2480
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 8.645833333333334,
|
| 1749 |
+
"grad_norm": 0.14217671751976013,
|
| 1750 |
+
"learning_rate": 4.930226829535767e-06,
|
| 1751 |
+
"loss": 0.0075,
|
| 1752 |
+
"step": 2490
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 8.680555555555555,
|
| 1756 |
+
"grad_norm": 0.21861067414283752,
|
| 1757 |
+
"learning_rate": 4.684610648167503e-06,
|
| 1758 |
+
"loss": 0.014,
|
| 1759 |
+
"step": 2500
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 8.715277777777779,
|
| 1763 |
+
"grad_norm": 0.1945720613002777,
|
| 1764 |
+
"learning_rate": 4.4449690622442585e-06,
|
| 1765 |
+
"loss": 0.0099,
|
| 1766 |
+
"step": 2510
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 8.75,
|
| 1770 |
+
"grad_norm": 0.2723676264286041,
|
| 1771 |
+
"learning_rate": 4.2113336672471245e-06,
|
| 1772 |
+
"loss": 0.0129,
|
| 1773 |
+
"step": 2520
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 8.784722222222221,
|
| 1777 |
+
"grad_norm": 0.2239852398633957,
|
| 1778 |
+
"learning_rate": 3.983735266772565e-06,
|
| 1779 |
+
"loss": 0.0094,
|
| 1780 |
+
"step": 2530
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 8.819444444444445,
|
| 1784 |
+
"grad_norm": 0.25737810134887695,
|
| 1785 |
+
"learning_rate": 3.762203868471087e-06,
|
| 1786 |
+
"loss": 0.0102,
|
| 1787 |
+
"step": 2540
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 8.854166666666666,
|
| 1791 |
+
"grad_norm": 0.18129268288612366,
|
| 1792 |
+
"learning_rate": 3.546768680090934e-06,
|
| 1793 |
+
"loss": 0.0119,
|
| 1794 |
+
"step": 2550
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 8.88888888888889,
|
| 1798 |
+
"grad_norm": 0.3871157169342041,
|
| 1799 |
+
"learning_rate": 3.3374581056271447e-06,
|
| 1800 |
+
"loss": 0.0088,
|
| 1801 |
+
"step": 2560
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 8.92361111111111,
|
| 1805 |
+
"grad_norm": 0.16779103875160217,
|
| 1806 |
+
"learning_rate": 3.1342997415767015e-06,
|
| 1807 |
+
"loss": 0.009,
|
| 1808 |
+
"step": 2570
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 8.958333333333334,
|
| 1812 |
+
"grad_norm": 0.11715767532587051,
|
| 1813 |
+
"learning_rate": 2.9373203733000232e-06,
|
| 1814 |
+
"loss": 0.0099,
|
| 1815 |
+
"step": 2580
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 8.993055555555555,
|
| 1819 |
+
"grad_norm": 0.21253140270709991,
|
| 1820 |
+
"learning_rate": 2.746545971489478e-06,
|
| 1821 |
+
"loss": 0.0099,
|
| 1822 |
+
"step": 2590
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 9.027777777777779,
|
| 1826 |
+
"grad_norm": 0.26496362686157227,
|
| 1827 |
+
"learning_rate": 2.562001688745291e-06,
|
| 1828 |
+
"loss": 0.0088,
|
| 1829 |
+
"step": 2600
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 9.0625,
|
| 1833 |
+
"grad_norm": 0.1795375943183899,
|
| 1834 |
+
"learning_rate": 2.3837118562592797e-06,
|
| 1835 |
+
"loss": 0.0095,
|
| 1836 |
+
"step": 2610
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 9.097222222222221,
|
| 1840 |
+
"grad_norm": 0.20966722071170807,
|
| 1841 |
+
"learning_rate": 2.211699980606946e-06,
|
| 1842 |
+
"loss": 0.008,
|
| 1843 |
+
"step": 2620
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 9.131944444444445,
|
| 1847 |
+
"grad_norm": 0.19171403348445892,
|
| 1848 |
+
"learning_rate": 2.0459887406482413e-06,
|
| 1849 |
+
"loss": 0.0103,
|
| 1850 |
+
"step": 2630
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 9.166666666666666,
|
| 1854 |
+
"grad_norm": 0.22042040526866913,
|
| 1855 |
+
"learning_rate": 1.8865999845374793e-06,
|
| 1856 |
+
"loss": 0.0087,
|
| 1857 |
+
"step": 2640
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 9.20138888888889,
|
| 1861 |
+
"grad_norm": 0.21836155652999878,
|
| 1862 |
+
"learning_rate": 1.7335547268427843e-06,
|
| 1863 |
+
"loss": 0.0091,
|
| 1864 |
+
"step": 2650
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 9.23611111111111,
|
| 1868 |
+
"grad_norm": 0.1593165397644043,
|
| 1869 |
+
"learning_rate": 1.5868731457754139e-06,
|
| 1870 |
+
"loss": 0.0108,
|
| 1871 |
+
"step": 2660
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 9.270833333333334,
|
| 1875 |
+
"grad_norm": 0.08942472189664841,
|
| 1876 |
+
"learning_rate": 1.4465745805293585e-06,
|
| 1877 |
+
"loss": 0.0079,
|
| 1878 |
+
"step": 2670
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 9.305555555555555,
|
| 1882 |
+
"grad_norm": 0.13019748032093048,
|
| 1883 |
+
"learning_rate": 1.3126775287316151e-06,
|
| 1884 |
+
"loss": 0.0072,
|
| 1885 |
+
"step": 2680
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 9.340277777777779,
|
| 1889 |
+
"grad_norm": 0.17031489312648773,
|
| 1890 |
+
"learning_rate": 1.1851996440033319e-06,
|
| 1891 |
+
"loss": 0.0076,
|
| 1892 |
+
"step": 2690
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 9.375,
|
| 1896 |
+
"grad_norm": 0.2234310358762741,
|
| 1897 |
+
"learning_rate": 1.064157733632276e-06,
|
| 1898 |
+
"loss": 0.0108,
|
| 1899 |
+
"step": 2700
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 9.409722222222221,
|
| 1903 |
+
"grad_norm": 0.1366192102432251,
|
| 1904 |
+
"learning_rate": 9.495677563569005e-07,
|
| 1905 |
+
"loss": 0.0086,
|
| 1906 |
+
"step": 2710
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 9.444444444444445,
|
| 1910 |
+
"grad_norm": 0.1559923142194748,
|
| 1911 |
+
"learning_rate": 8.414448202622494e-07,
|
| 1912 |
+
"loss": 0.0078,
|
| 1913 |
+
"step": 2720
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 9.479166666666666,
|
| 1917 |
+
"grad_norm": 0.1459418386220932,
|
| 1918 |
+
"learning_rate": 7.398031807880457e-07,
|
| 1919 |
+
"loss": 0.0114,
|
| 1920 |
+
"step": 2730
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 9.51388888888889,
|
| 1924 |
+
"grad_norm": 0.23313875496387482,
|
| 1925 |
+
"learning_rate": 6.44656238849195e-07,
|
| 1926 |
+
"loss": 0.0136,
|
| 1927 |
+
"step": 2740
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 9.54861111111111,
|
| 1931 |
+
"grad_norm": 0.14127647876739502,
|
| 1932 |
+
"learning_rate": 5.560165390689154e-07,
|
| 1933 |
+
"loss": 0.0061,
|
| 1934 |
+
"step": 2750
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 9.583333333333334,
|
| 1938 |
+
"grad_norm": 0.14344729483127594,
|
| 1939 |
+
"learning_rate": 4.738957681248379e-07,
|
| 1940 |
+
"loss": 0.0076,
|
| 1941 |
+
"step": 2760
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 9.618055555555555,
|
| 1945 |
+
"grad_norm": 0.257382869720459,
|
| 1946 |
+
"learning_rate": 3.983047532081341e-07,
|
| 1947 |
+
"loss": 0.0103,
|
| 1948 |
+
"step": 2770
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 9.652777777777779,
|
| 1952 |
+
"grad_norm": 0.17644330859184265,
|
| 1953 |
+
"learning_rate": 3.2925346059605176e-07,
|
| 1954 |
+
"loss": 0.0122,
|
| 1955 |
+
"step": 2780
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 9.6875,
|
| 1959 |
+
"grad_norm": 0.1124967411160469,
|
| 1960 |
+
"learning_rate": 2.667509943378721e-07,
|
| 1961 |
+
"loss": 0.0099,
|
| 1962 |
+
"step": 2790
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 9.722222222222221,
|
| 1966 |
+
"grad_norm": 0.14655955135822296,
|
| 1967 |
+
"learning_rate": 2.1080559505462505e-07,
|
| 1968 |
+
"loss": 0.009,
|
| 1969 |
+
"step": 2800
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 9.756944444444445,
|
| 1973 |
+
"grad_norm": 0.08127148449420929,
|
| 1974 |
+
"learning_rate": 1.614246388525864e-07,
|
| 1975 |
+
"loss": 0.0074,
|
| 1976 |
+
"step": 2810
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 9.791666666666666,
|
| 1980 |
+
"grad_norm": 0.13320744037628174,
|
| 1981 |
+
"learning_rate": 1.1861463635077785e-07,
|
| 1982 |
+
"loss": 0.0076,
|
| 1983 |
+
"step": 2820
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 9.82638888888889,
|
| 1987 |
+
"grad_norm": 0.14339619874954224,
|
| 1988 |
+
"learning_rate": 8.238123182257584e-08,
|
| 1989 |
+
"loss": 0.0053,
|
| 1990 |
+
"step": 2830
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 9.86111111111111,
|
| 1994 |
+
"grad_norm": 0.14370766282081604,
|
| 1995 |
+
"learning_rate": 5.272920245156798e-08,
|
| 1996 |
+
"loss": 0.0081,
|
| 1997 |
+
"step": 2840
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 9.895833333333334,
|
| 2001 |
+
"grad_norm": 0.19183674454689026,
|
| 2002 |
+
"learning_rate": 2.9662457701662428e-08,
|
| 2003 |
+
"loss": 0.0117,
|
| 2004 |
+
"step": 2850
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 9.930555555555555,
|
| 2008 |
+
"grad_norm": 0.1539844423532486,
|
| 2009 |
+
"learning_rate": 1.3184038801683508e-08,
|
| 2010 |
+
"loss": 0.0066,
|
| 2011 |
+
"step": 2860
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 9.965277777777779,
|
| 2015 |
+
"grad_norm": 0.16146790981292725,
|
| 2016 |
+
"learning_rate": 3.2961183443813714e-09,
|
| 2017 |
+
"loss": 0.0105,
|
| 2018 |
+
"step": 2870
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 10.0,
|
| 2022 |
+
"grad_norm": 0.14206930994987488,
|
| 2023 |
+
"learning_rate": 0.0,
|
| 2024 |
+
"loss": 0.0123,
|
| 2025 |
+
"step": 2880
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 10.0,
|
| 2029 |
+
"step": 2880,
|
| 2030 |
+
"total_flos": 3.0392298095399424e+17,
|
| 2031 |
+
"train_loss": 0.03354278181084535,
|
| 2032 |
+
"train_runtime": 2973.9122,
|
| 2033 |
+
"train_samples_per_second": 47.439,
|
| 2034 |
+
"train_steps_per_second": 0.968
|
| 2035 |
+
}
|
| 2036 |
+
],
|
| 2037 |
+
"logging_steps": 10,
|
| 2038 |
+
"max_steps": 2880,
|
| 2039 |
+
"num_input_tokens_seen": 0,
|
| 2040 |
+
"num_train_epochs": 10,
|
| 2041 |
+
"save_steps": 10000,
|
| 2042 |
+
"stateful_callbacks": {
|
| 2043 |
+
"TrainerControl": {
|
| 2044 |
+
"args": {
|
| 2045 |
+
"should_epoch_stop": false,
|
| 2046 |
+
"should_evaluate": false,
|
| 2047 |
+
"should_log": false,
|
| 2048 |
+
"should_save": true,
|
| 2049 |
+
"should_training_stop": true
|
| 2050 |
+
},
|
| 2051 |
+
"attributes": {}
|
| 2052 |
+
}
|
| 2053 |
+
},
|
| 2054 |
+
"total_flos": 3.0392298095399424e+17,
|
| 2055 |
+
"train_batch_size": 49,
|
| 2056 |
+
"trial_name": null,
|
| 2057 |
+
"trial_params": null
|
| 2058 |
+
}
|