Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1821 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1821 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 10.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2540,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.03937007874015748,
|
| 14 |
+
"grad_norm": 9.399362564086914,
|
| 15 |
+
"learning_rate": 7.086614173228347e-06,
|
| 16 |
+
"loss": 1.4557,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.07874015748031496,
|
| 21 |
+
"grad_norm": 2.942493438720703,
|
| 22 |
+
"learning_rate": 1.4960629921259845e-05,
|
| 23 |
+
"loss": 0.4296,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.11811023622047244,
|
| 28 |
+
"grad_norm": 3.1724367141723633,
|
| 29 |
+
"learning_rate": 2.283464566929134e-05,
|
| 30 |
+
"loss": 0.3657,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.15748031496062992,
|
| 35 |
+
"grad_norm": 1.5789762735366821,
|
| 36 |
+
"learning_rate": 3.070866141732284e-05,
|
| 37 |
+
"loss": 0.3486,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.1968503937007874,
|
| 42 |
+
"grad_norm": 2.097869396209717,
|
| 43 |
+
"learning_rate": 3.858267716535433e-05,
|
| 44 |
+
"loss": 0.3044,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.23622047244094488,
|
| 49 |
+
"grad_norm": 3.880457639694214,
|
| 50 |
+
"learning_rate": 4.645669291338583e-05,
|
| 51 |
+
"loss": 0.3367,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.2755905511811024,
|
| 56 |
+
"grad_norm": 2.8253917694091797,
|
| 57 |
+
"learning_rate": 5.433070866141733e-05,
|
| 58 |
+
"loss": 0.3126,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.31496062992125984,
|
| 63 |
+
"grad_norm": 3.822925090789795,
|
| 64 |
+
"learning_rate": 6.220472440944882e-05,
|
| 65 |
+
"loss": 0.3004,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.3543307086614173,
|
| 70 |
+
"grad_norm": 1.3659324645996094,
|
| 71 |
+
"learning_rate": 7.007874015748031e-05,
|
| 72 |
+
"loss": 0.2605,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.3937007874015748,
|
| 77 |
+
"grad_norm": 1.7165173292160034,
|
| 78 |
+
"learning_rate": 7.795275590551181e-05,
|
| 79 |
+
"loss": 0.1676,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.4330708661417323,
|
| 84 |
+
"grad_norm": 1.704687476158142,
|
| 85 |
+
"learning_rate": 8.582677165354331e-05,
|
| 86 |
+
"loss": 0.1404,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.47244094488188976,
|
| 91 |
+
"grad_norm": 1.3101590871810913,
|
| 92 |
+
"learning_rate": 9.370078740157481e-05,
|
| 93 |
+
"loss": 0.1322,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.5118110236220472,
|
| 98 |
+
"grad_norm": 1.6621087789535522,
|
| 99 |
+
"learning_rate": 9.999983049408561e-05,
|
| 100 |
+
"loss": 0.1242,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.5511811023622047,
|
| 105 |
+
"grad_norm": 0.9743478298187256,
|
| 106 |
+
"learning_rate": 9.999389790775648e-05,
|
| 107 |
+
"loss": 0.1027,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.5905511811023622,
|
| 112 |
+
"grad_norm": 1.9478999376296997,
|
| 113 |
+
"learning_rate": 9.997949117496292e-05,
|
| 114 |
+
"loss": 0.1174,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.6299212598425197,
|
| 119 |
+
"grad_norm": 0.9509850740432739,
|
| 120 |
+
"learning_rate": 9.995661273769822e-05,
|
| 121 |
+
"loss": 0.1015,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.6692913385826772,
|
| 126 |
+
"grad_norm": 0.9505985379219055,
|
| 127 |
+
"learning_rate": 9.992526647394022e-05,
|
| 128 |
+
"loss": 0.102,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.7086614173228346,
|
| 133 |
+
"grad_norm": 1.489611268043518,
|
| 134 |
+
"learning_rate": 9.988545769699399e-05,
|
| 135 |
+
"loss": 0.097,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.7480314960629921,
|
| 140 |
+
"grad_norm": 1.1149543523788452,
|
| 141 |
+
"learning_rate": 9.983719315459114e-05,
|
| 142 |
+
"loss": 0.0925,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.7874015748031497,
|
| 147 |
+
"grad_norm": 1.0860552787780762,
|
| 148 |
+
"learning_rate": 9.978048102774613e-05,
|
| 149 |
+
"loss": 0.0964,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.8267716535433071,
|
| 154 |
+
"grad_norm": 1.2707302570343018,
|
| 155 |
+
"learning_rate": 9.971533092936954e-05,
|
| 156 |
+
"loss": 0.0844,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.8661417322834646,
|
| 161 |
+
"grad_norm": 1.1820255517959595,
|
| 162 |
+
"learning_rate": 9.964175390263856e-05,
|
| 163 |
+
"loss": 0.0805,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.905511811023622,
|
| 168 |
+
"grad_norm": 1.3937278985977173,
|
| 169 |
+
"learning_rate": 9.955976241912535e-05,
|
| 170 |
+
"loss": 0.0871,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.9448818897637795,
|
| 175 |
+
"grad_norm": 0.9341478943824768,
|
| 176 |
+
"learning_rate": 9.946937037668275e-05,
|
| 177 |
+
"loss": 0.0826,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.984251968503937,
|
| 182 |
+
"grad_norm": 1.7772321701049805,
|
| 183 |
+
"learning_rate": 9.937059309708885e-05,
|
| 184 |
+
"loss": 0.0873,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 1.0236220472440944,
|
| 189 |
+
"grad_norm": 1.2695393562316895,
|
| 190 |
+
"learning_rate": 9.926344732344967e-05,
|
| 191 |
+
"loss": 0.0794,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 1.0629921259842519,
|
| 196 |
+
"grad_norm": 1.1093697547912598,
|
| 197 |
+
"learning_rate": 9.914795121736128e-05,
|
| 198 |
+
"loss": 0.0758,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 1.1023622047244095,
|
| 203 |
+
"grad_norm": 0.8029543161392212,
|
| 204 |
+
"learning_rate": 9.902412435583128e-05,
|
| 205 |
+
"loss": 0.0678,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 1.141732283464567,
|
| 210 |
+
"grad_norm": 0.7547488808631897,
|
| 211 |
+
"learning_rate": 9.88919877279604e-05,
|
| 212 |
+
"loss": 0.0761,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 1.1811023622047245,
|
| 217 |
+
"grad_norm": 0.8116704225540161,
|
| 218 |
+
"learning_rate": 9.875156373138489e-05,
|
| 219 |
+
"loss": 0.057,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 1.220472440944882,
|
| 224 |
+
"grad_norm": 0.8954646587371826,
|
| 225 |
+
"learning_rate": 9.86028761684799e-05,
|
| 226 |
+
"loss": 0.0738,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 1.2598425196850394,
|
| 231 |
+
"grad_norm": 1.016405463218689,
|
| 232 |
+
"learning_rate": 9.844595024232495e-05,
|
| 233 |
+
"loss": 0.0901,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 1.2992125984251968,
|
| 238 |
+
"grad_norm": 1.395342469215393,
|
| 239 |
+
"learning_rate": 9.828081255243198e-05,
|
| 240 |
+
"loss": 0.0796,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 1.3385826771653544,
|
| 245 |
+
"grad_norm": 0.8092917203903198,
|
| 246 |
+
"learning_rate": 9.81074910902365e-05,
|
| 247 |
+
"loss": 0.0883,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 1.3779527559055118,
|
| 252 |
+
"grad_norm": 1.1169452667236328,
|
| 253 |
+
"learning_rate": 9.792601523435307e-05,
|
| 254 |
+
"loss": 0.0748,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 1.4173228346456692,
|
| 259 |
+
"grad_norm": 0.7994294166564941,
|
| 260 |
+
"learning_rate": 9.773641574559546e-05,
|
| 261 |
+
"loss": 0.0862,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 1.4566929133858268,
|
| 266 |
+
"grad_norm": 0.7973235249519348,
|
| 267 |
+
"learning_rate": 9.753872476176254e-05,
|
| 268 |
+
"loss": 0.0735,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 1.4960629921259843,
|
| 273 |
+
"grad_norm": 0.9219651818275452,
|
| 274 |
+
"learning_rate": 9.73329757921909e-05,
|
| 275 |
+
"loss": 0.077,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 1.5354330708661417,
|
| 280 |
+
"grad_norm": 1.1715006828308105,
|
| 281 |
+
"learning_rate": 9.711920371207484e-05,
|
| 282 |
+
"loss": 0.0691,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 1.574803149606299,
|
| 287 |
+
"grad_norm": 0.7752212882041931,
|
| 288 |
+
"learning_rate": 9.68974447565549e-05,
|
| 289 |
+
"loss": 0.0669,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 1.6141732283464567,
|
| 294 |
+
"grad_norm": 1.0260776281356812,
|
| 295 |
+
"learning_rate": 9.666773651457588e-05,
|
| 296 |
+
"loss": 0.0623,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 1.6535433070866141,
|
| 301 |
+
"grad_norm": 0.8338336944580078,
|
| 302 |
+
"learning_rate": 9.643011792251538e-05,
|
| 303 |
+
"loss": 0.0699,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 1.6929133858267718,
|
| 308 |
+
"grad_norm": 0.8776105642318726,
|
| 309 |
+
"learning_rate": 9.618462925758392e-05,
|
| 310 |
+
"loss": 0.0653,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 1.7322834645669292,
|
| 315 |
+
"grad_norm": 0.6896973252296448,
|
| 316 |
+
"learning_rate": 9.593131213099789e-05,
|
| 317 |
+
"loss": 0.0586,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 1.7716535433070866,
|
| 322 |
+
"grad_norm": 1.0852605104446411,
|
| 323 |
+
"learning_rate": 9.567020948092616e-05,
|
| 324 |
+
"loss": 0.0673,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 1.811023622047244,
|
| 329 |
+
"grad_norm": 1.0203490257263184,
|
| 330 |
+
"learning_rate": 9.540136556521203e-05,
|
| 331 |
+
"loss": 0.0663,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 1.8503937007874016,
|
| 336 |
+
"grad_norm": 0.774488091468811,
|
| 337 |
+
"learning_rate": 9.512482595387132e-05,
|
| 338 |
+
"loss": 0.0609,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 1.889763779527559,
|
| 343 |
+
"grad_norm": 0.5737660527229309,
|
| 344 |
+
"learning_rate": 9.484063752136805e-05,
|
| 345 |
+
"loss": 0.0606,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 1.9291338582677167,
|
| 350 |
+
"grad_norm": 1.0153898000717163,
|
| 351 |
+
"learning_rate": 9.454884843866912e-05,
|
| 352 |
+
"loss": 0.0737,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 1.968503937007874,
|
| 357 |
+
"grad_norm": 0.7526334524154663,
|
| 358 |
+
"learning_rate": 9.424950816507909e-05,
|
| 359 |
+
"loss": 0.0641,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 2.0078740157480315,
|
| 364 |
+
"grad_norm": 0.5760018825531006,
|
| 365 |
+
"learning_rate": 9.394266743985671e-05,
|
| 366 |
+
"loss": 0.0674,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 2.047244094488189,
|
| 371 |
+
"grad_norm": 0.70269775390625,
|
| 372 |
+
"learning_rate": 9.36283782736144e-05,
|
| 373 |
+
"loss": 0.0631,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 2.0866141732283463,
|
| 378 |
+
"grad_norm": 0.8864635229110718,
|
| 379 |
+
"learning_rate": 9.330669393950219e-05,
|
| 380 |
+
"loss": 0.0654,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 2.1259842519685037,
|
| 385 |
+
"grad_norm": 0.7043759226799011,
|
| 386 |
+
"learning_rate": 9.297766896417793e-05,
|
| 387 |
+
"loss": 0.0657,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 2.1653543307086616,
|
| 392 |
+
"grad_norm": 0.6329500675201416,
|
| 393 |
+
"learning_rate": 9.264135911856462e-05,
|
| 394 |
+
"loss": 0.0707,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 2.204724409448819,
|
| 399 |
+
"grad_norm": 0.4031962752342224,
|
| 400 |
+
"learning_rate": 9.22978214083971e-05,
|
| 401 |
+
"loss": 0.0528,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 2.2440944881889764,
|
| 406 |
+
"grad_norm": 0.5401821136474609,
|
| 407 |
+
"learning_rate": 9.194711406455945e-05,
|
| 408 |
+
"loss": 0.0654,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 2.283464566929134,
|
| 413 |
+
"grad_norm": 0.713798999786377,
|
| 414 |
+
"learning_rate": 9.158929653321451e-05,
|
| 415 |
+
"loss": 0.0555,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 2.322834645669291,
|
| 420 |
+
"grad_norm": 0.4728735387325287,
|
| 421 |
+
"learning_rate": 9.122442946572768e-05,
|
| 422 |
+
"loss": 0.0552,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 2.362204724409449,
|
| 427 |
+
"grad_norm": 0.7359452843666077,
|
| 428 |
+
"learning_rate": 9.085257470838619e-05,
|
| 429 |
+
"loss": 0.0677,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 2.4015748031496065,
|
| 434 |
+
"grad_norm": 0.6030870676040649,
|
| 435 |
+
"learning_rate": 9.047379529191594e-05,
|
| 436 |
+
"loss": 0.053,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 2.440944881889764,
|
| 441 |
+
"grad_norm": 0.5791817903518677,
|
| 442 |
+
"learning_rate": 9.008815542079766e-05,
|
| 443 |
+
"loss": 0.0493,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 2.4803149606299213,
|
| 448 |
+
"grad_norm": 0.8772215247154236,
|
| 449 |
+
"learning_rate": 8.969572046238389e-05,
|
| 450 |
+
"loss": 0.0721,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 2.5196850393700787,
|
| 455 |
+
"grad_norm": 0.8733668923377991,
|
| 456 |
+
"learning_rate": 8.929655693581904e-05,
|
| 457 |
+
"loss": 0.0597,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 2.559055118110236,
|
| 462 |
+
"grad_norm": 1.0022321939468384,
|
| 463 |
+
"learning_rate": 8.889073250076421e-05,
|
| 464 |
+
"loss": 0.0659,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 2.5984251968503935,
|
| 469 |
+
"grad_norm": 0.7206939458847046,
|
| 470 |
+
"learning_rate": 8.84783159459285e-05,
|
| 471 |
+
"loss": 0.0452,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 2.637795275590551,
|
| 476 |
+
"grad_norm": 0.8875113725662231,
|
| 477 |
+
"learning_rate": 8.805937717740918e-05,
|
| 478 |
+
"loss": 0.0539,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 2.677165354330709,
|
| 483 |
+
"grad_norm": 0.5767335295677185,
|
| 484 |
+
"learning_rate": 8.763398720684232e-05,
|
| 485 |
+
"loss": 0.0503,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 2.716535433070866,
|
| 490 |
+
"grad_norm": 0.5727648138999939,
|
| 491 |
+
"learning_rate": 8.72022181393661e-05,
|
| 492 |
+
"loss": 0.0457,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 2.7559055118110236,
|
| 497 |
+
"grad_norm": 0.8125827312469482,
|
| 498 |
+
"learning_rate": 8.676414316139863e-05,
|
| 499 |
+
"loss": 0.0607,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 2.795275590551181,
|
| 504 |
+
"grad_norm": 0.6720311641693115,
|
| 505 |
+
"learning_rate": 8.631983652823267e-05,
|
| 506 |
+
"loss": 0.0665,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 2.8346456692913384,
|
| 511 |
+
"grad_norm": 0.6637985706329346,
|
| 512 |
+
"learning_rate": 8.586937355144908e-05,
|
| 513 |
+
"loss": 0.068,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 2.8740157480314963,
|
| 518 |
+
"grad_norm": 0.7840360999107361,
|
| 519 |
+
"learning_rate": 8.541283058615124e-05,
|
| 520 |
+
"loss": 0.0561,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 2.9133858267716537,
|
| 525 |
+
"grad_norm": 0.44171687960624695,
|
| 526 |
+
"learning_rate": 8.495028501802251e-05,
|
| 527 |
+
"loss": 0.0534,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 2.952755905511811,
|
| 532 |
+
"grad_norm": 0.4313163459300995,
|
| 533 |
+
"learning_rate": 8.448181525020921e-05,
|
| 534 |
+
"loss": 0.0391,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 2.9921259842519685,
|
| 539 |
+
"grad_norm": 0.7261826395988464,
|
| 540 |
+
"learning_rate": 8.400750069003086e-05,
|
| 541 |
+
"loss": 0.0486,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 3.031496062992126,
|
| 546 |
+
"grad_norm": 0.4469556212425232,
|
| 547 |
+
"learning_rate": 8.352742173552046e-05,
|
| 548 |
+
"loss": 0.0511,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 3.0708661417322833,
|
| 553 |
+
"grad_norm": 0.9129867553710938,
|
| 554 |
+
"learning_rate": 8.304165976179667e-05,
|
| 555 |
+
"loss": 0.0533,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 3.1102362204724407,
|
| 560 |
+
"grad_norm": 1.2041122913360596,
|
| 561 |
+
"learning_rate": 8.255029710727048e-05,
|
| 562 |
+
"loss": 0.0671,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 3.1496062992125986,
|
| 567 |
+
"grad_norm": 0.7420069575309753,
|
| 568 |
+
"learning_rate": 8.20534170596885e-05,
|
| 569 |
+
"loss": 0.0685,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 3.188976377952756,
|
| 574 |
+
"grad_norm": 0.3230190575122833,
|
| 575 |
+
"learning_rate": 8.155110384201544e-05,
|
| 576 |
+
"loss": 0.0647,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 3.2283464566929134,
|
| 581 |
+
"grad_norm": 0.6603342890739441,
|
| 582 |
+
"learning_rate": 8.104344259815794e-05,
|
| 583 |
+
"loss": 0.0558,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 3.267716535433071,
|
| 588 |
+
"grad_norm": 0.5632081031799316,
|
| 589 |
+
"learning_rate": 8.053051937853248e-05,
|
| 590 |
+
"loss": 0.0558,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 3.3070866141732282,
|
| 595 |
+
"grad_norm": 0.7754299640655518,
|
| 596 |
+
"learning_rate": 8.001242112547942e-05,
|
| 597 |
+
"loss": 0.0632,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 3.3464566929133857,
|
| 602 |
+
"grad_norm": 0.7823946475982666,
|
| 603 |
+
"learning_rate": 7.948923565852598e-05,
|
| 604 |
+
"loss": 0.0662,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 3.3858267716535435,
|
| 609 |
+
"grad_norm": 0.7399844527244568,
|
| 610 |
+
"learning_rate": 7.896105165950059e-05,
|
| 611 |
+
"loss": 0.052,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 3.425196850393701,
|
| 616 |
+
"grad_norm": 0.8208476305007935,
|
| 617 |
+
"learning_rate": 7.842795865750088e-05,
|
| 618 |
+
"loss": 0.0486,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 3.4645669291338583,
|
| 623 |
+
"grad_norm": 0.5400993227958679,
|
| 624 |
+
"learning_rate": 7.789004701371825e-05,
|
| 625 |
+
"loss": 0.0443,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 3.5039370078740157,
|
| 630 |
+
"grad_norm": 0.6949036717414856,
|
| 631 |
+
"learning_rate": 7.734740790612136e-05,
|
| 632 |
+
"loss": 0.0597,
|
| 633 |
+
"step": 890
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 3.543307086614173,
|
| 637 |
+
"grad_norm": 1.0321848392486572,
|
| 638 |
+
"learning_rate": 7.680013331400098e-05,
|
| 639 |
+
"loss": 0.0446,
|
| 640 |
+
"step": 900
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 3.5826771653543306,
|
| 644 |
+
"grad_norm": 0.5193206071853638,
|
| 645 |
+
"learning_rate": 7.624831600237937e-05,
|
| 646 |
+
"loss": 0.0499,
|
| 647 |
+
"step": 910
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 3.622047244094488,
|
| 651 |
+
"grad_norm": 0.755699872970581,
|
| 652 |
+
"learning_rate": 7.569204950628605e-05,
|
| 653 |
+
"loss": 0.0595,
|
| 654 |
+
"step": 920
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 3.661417322834646,
|
| 658 |
+
"grad_norm": 0.4758411645889282,
|
| 659 |
+
"learning_rate": 7.513142811490356e-05,
|
| 660 |
+
"loss": 0.0403,
|
| 661 |
+
"step": 930
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 3.7007874015748032,
|
| 665 |
+
"grad_norm": 0.9744377732276917,
|
| 666 |
+
"learning_rate": 7.456654685558481e-05,
|
| 667 |
+
"loss": 0.0566,
|
| 668 |
+
"step": 940
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 3.7401574803149606,
|
| 672 |
+
"grad_norm": 0.46791282296180725,
|
| 673 |
+
"learning_rate": 7.399750147774575e-05,
|
| 674 |
+
"loss": 0.0445,
|
| 675 |
+
"step": 950
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 3.779527559055118,
|
| 679 |
+
"grad_norm": 0.4000394642353058,
|
| 680 |
+
"learning_rate": 7.34243884366355e-05,
|
| 681 |
+
"loss": 0.0577,
|
| 682 |
+
"step": 960
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 3.8188976377952755,
|
| 686 |
+
"grad_norm": 0.6318042874336243,
|
| 687 |
+
"learning_rate": 7.28473048769868e-05,
|
| 688 |
+
"loss": 0.0491,
|
| 689 |
+
"step": 970
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 3.8582677165354333,
|
| 693 |
+
"grad_norm": 0.9635873436927795,
|
| 694 |
+
"learning_rate": 7.226634861654965e-05,
|
| 695 |
+
"loss": 0.0501,
|
| 696 |
+
"step": 980
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 3.8976377952755907,
|
| 700 |
+
"grad_norm": 0.7494231462478638,
|
| 701 |
+
"learning_rate": 7.168161812951084e-05,
|
| 702 |
+
"loss": 0.0511,
|
| 703 |
+
"step": 990
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 3.937007874015748,
|
| 707 |
+
"grad_norm": 0.6621044874191284,
|
| 708 |
+
"learning_rate": 7.109321252980218e-05,
|
| 709 |
+
"loss": 0.0429,
|
| 710 |
+
"step": 1000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 3.9763779527559056,
|
| 714 |
+
"grad_norm": 0.6507459878921509,
|
| 715 |
+
"learning_rate": 7.05012315543004e-05,
|
| 716 |
+
"loss": 0.0514,
|
| 717 |
+
"step": 1010
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 4.015748031496063,
|
| 721 |
+
"grad_norm": 0.6397859454154968,
|
| 722 |
+
"learning_rate": 6.990577554592134e-05,
|
| 723 |
+
"loss": 0.051,
|
| 724 |
+
"step": 1020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 4.05511811023622,
|
| 728 |
+
"grad_norm": 0.8459829688072205,
|
| 729 |
+
"learning_rate": 6.930694543661149e-05,
|
| 730 |
+
"loss": 0.0492,
|
| 731 |
+
"step": 1030
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 4.094488188976378,
|
| 735 |
+
"grad_norm": 0.6369995474815369,
|
| 736 |
+
"learning_rate": 6.870484273023968e-05,
|
| 737 |
+
"loss": 0.0447,
|
| 738 |
+
"step": 1040
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 4.133858267716535,
|
| 742 |
+
"grad_norm": 0.6142792701721191,
|
| 743 |
+
"learning_rate": 6.809956948539166e-05,
|
| 744 |
+
"loss": 0.044,
|
| 745 |
+
"step": 1050
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 4.173228346456693,
|
| 749 |
+
"grad_norm": 0.5256998538970947,
|
| 750 |
+
"learning_rate": 6.749122829807103e-05,
|
| 751 |
+
"loss": 0.0427,
|
| 752 |
+
"step": 1060
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 4.21259842519685,
|
| 756 |
+
"grad_norm": 0.7650443911552429,
|
| 757 |
+
"learning_rate": 6.687992228430872e-05,
|
| 758 |
+
"loss": 0.0525,
|
| 759 |
+
"step": 1070
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 4.251968503937007,
|
| 763 |
+
"grad_norm": 0.6876934170722961,
|
| 764 |
+
"learning_rate": 6.62657550626844e-05,
|
| 765 |
+
"loss": 0.0385,
|
| 766 |
+
"step": 1080
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 4.291338582677166,
|
| 770 |
+
"grad_norm": 0.7614730000495911,
|
| 771 |
+
"learning_rate": 6.564883073676287e-05,
|
| 772 |
+
"loss": 0.0543,
|
| 773 |
+
"step": 1090
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 4.330708661417323,
|
| 777 |
+
"grad_norm": 0.591896653175354,
|
| 778 |
+
"learning_rate": 6.502925387744807e-05,
|
| 779 |
+
"loss": 0.044,
|
| 780 |
+
"step": 1100
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 4.3700787401574805,
|
| 784 |
+
"grad_norm": 0.8287089467048645,
|
| 785 |
+
"learning_rate": 6.440712950525791e-05,
|
| 786 |
+
"loss": 0.0427,
|
| 787 |
+
"step": 1110
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 4.409448818897638,
|
| 791 |
+
"grad_norm": 0.8359081745147705,
|
| 792 |
+
"learning_rate": 6.3782563072523e-05,
|
| 793 |
+
"loss": 0.0513,
|
| 794 |
+
"step": 1120
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 4.448818897637795,
|
| 798 |
+
"grad_norm": 0.4965924322605133,
|
| 799 |
+
"learning_rate": 6.315566044551197e-05,
|
| 800 |
+
"loss": 0.0503,
|
| 801 |
+
"step": 1130
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 4.488188976377953,
|
| 805 |
+
"grad_norm": 0.5000588297843933,
|
| 806 |
+
"learning_rate": 6.252652788648691e-05,
|
| 807 |
+
"loss": 0.0348,
|
| 808 |
+
"step": 1140
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 4.52755905511811,
|
| 812 |
+
"grad_norm": 0.5434101819992065,
|
| 813 |
+
"learning_rate": 6.18952720356914e-05,
|
| 814 |
+
"loss": 0.0409,
|
| 815 |
+
"step": 1150
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 4.566929133858268,
|
| 819 |
+
"grad_norm": 0.6852266788482666,
|
| 820 |
+
"learning_rate": 6.126199989327462e-05,
|
| 821 |
+
"loss": 0.0437,
|
| 822 |
+
"step": 1160
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 4.606299212598425,
|
| 826 |
+
"grad_norm": 0.684528648853302,
|
| 827 |
+
"learning_rate": 6.062681880115453e-05,
|
| 828 |
+
"loss": 0.0447,
|
| 829 |
+
"step": 1170
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 4.645669291338582,
|
| 833 |
+
"grad_norm": 0.6656462550163269,
|
| 834 |
+
"learning_rate": 5.998983642482296e-05,
|
| 835 |
+
"loss": 0.0429,
|
| 836 |
+
"step": 1180
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 4.68503937007874,
|
| 840 |
+
"grad_norm": 0.7268936634063721,
|
| 841 |
+
"learning_rate": 5.935116073509592e-05,
|
| 842 |
+
"loss": 0.0478,
|
| 843 |
+
"step": 1190
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 4.724409448818898,
|
| 847 |
+
"grad_norm": 0.5984519720077515,
|
| 848 |
+
"learning_rate": 5.871089998981214e-05,
|
| 849 |
+
"loss": 0.038,
|
| 850 |
+
"step": 1200
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 4.7637795275590555,
|
| 854 |
+
"grad_norm": 0.6121963858604431,
|
| 855 |
+
"learning_rate": 5.8069162715483e-05,
|
| 856 |
+
"loss": 0.0388,
|
| 857 |
+
"step": 1210
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 4.803149606299213,
|
| 861 |
+
"grad_norm": 0.47043073177337646,
|
| 862 |
+
"learning_rate": 5.742605768889693e-05,
|
| 863 |
+
"loss": 0.0355,
|
| 864 |
+
"step": 1220
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 4.84251968503937,
|
| 868 |
+
"grad_norm": 0.5659105181694031,
|
| 869 |
+
"learning_rate": 5.6781693918681275e-05,
|
| 870 |
+
"loss": 0.0434,
|
| 871 |
+
"step": 1230
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 4.881889763779528,
|
| 875 |
+
"grad_norm": 0.5114152431488037,
|
| 876 |
+
"learning_rate": 5.613618062682502e-05,
|
| 877 |
+
"loss": 0.0379,
|
| 878 |
+
"step": 1240
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 4.921259842519685,
|
| 882 |
+
"grad_norm": 0.6941166520118713,
|
| 883 |
+
"learning_rate": 5.5489627230165176e-05,
|
| 884 |
+
"loss": 0.046,
|
| 885 |
+
"step": 1250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 4.960629921259843,
|
| 889 |
+
"grad_norm": 0.6256818771362305,
|
| 890 |
+
"learning_rate": 5.48421433218403e-05,
|
| 891 |
+
"loss": 0.0505,
|
| 892 |
+
"step": 1260
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 5.0,
|
| 896 |
+
"grad_norm": 0.8666470646858215,
|
| 897 |
+
"learning_rate": 5.419383865271402e-05,
|
| 898 |
+
"loss": 0.0419,
|
| 899 |
+
"step": 1270
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 5.039370078740157,
|
| 903 |
+
"grad_norm": 1.043869972229004,
|
| 904 |
+
"learning_rate": 5.354482311277193e-05,
|
| 905 |
+
"loss": 0.0483,
|
| 906 |
+
"step": 1280
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 5.078740157480315,
|
| 910 |
+
"grad_norm": 0.4656646251678467,
|
| 911 |
+
"learning_rate": 5.289520671249479e-05,
|
| 912 |
+
"loss": 0.0333,
|
| 913 |
+
"step": 1290
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 5.118110236220472,
|
| 917 |
+
"grad_norm": 0.45656928420066833,
|
| 918 |
+
"learning_rate": 5.224509956421133e-05,
|
| 919 |
+
"loss": 0.0373,
|
| 920 |
+
"step": 1300
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 5.15748031496063,
|
| 924 |
+
"grad_norm": 0.7461910247802734,
|
| 925 |
+
"learning_rate": 5.159461186343385e-05,
|
| 926 |
+
"loss": 0.0427,
|
| 927 |
+
"step": 1310
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 5.196850393700787,
|
| 931 |
+
"grad_norm": 0.646958589553833,
|
| 932 |
+
"learning_rate": 5.094385387017967e-05,
|
| 933 |
+
"loss": 0.0447,
|
| 934 |
+
"step": 1320
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 5.2362204724409445,
|
| 938 |
+
"grad_norm": 0.2877052128314972,
|
| 939 |
+
"learning_rate": 5.02929358902817e-05,
|
| 940 |
+
"loss": 0.036,
|
| 941 |
+
"step": 1330
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 5.275590551181103,
|
| 945 |
+
"grad_norm": 0.5020745992660522,
|
| 946 |
+
"learning_rate": 4.964196825669112e-05,
|
| 947 |
+
"loss": 0.0485,
|
| 948 |
+
"step": 1340
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 5.31496062992126,
|
| 952 |
+
"grad_norm": 0.494816392660141,
|
| 953 |
+
"learning_rate": 4.899106131077562e-05,
|
| 954 |
+
"loss": 0.0446,
|
| 955 |
+
"step": 1350
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 5.354330708661418,
|
| 959 |
+
"grad_norm": 0.4856385290622711,
|
| 960 |
+
"learning_rate": 4.834032538361607e-05,
|
| 961 |
+
"loss": 0.0418,
|
| 962 |
+
"step": 1360
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 5.393700787401575,
|
| 966 |
+
"grad_norm": 0.2937265634536743,
|
| 967 |
+
"learning_rate": 4.768987077730509e-05,
|
| 968 |
+
"loss": 0.0329,
|
| 969 |
+
"step": 1370
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 5.433070866141732,
|
| 973 |
+
"grad_norm": 0.44668734073638916,
|
| 974 |
+
"learning_rate": 4.703980774625038e-05,
|
| 975 |
+
"loss": 0.0373,
|
| 976 |
+
"step": 1380
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 5.47244094488189,
|
| 980 |
+
"grad_norm": 0.49580681324005127,
|
| 981 |
+
"learning_rate": 4.6390246478486196e-05,
|
| 982 |
+
"loss": 0.0393,
|
| 983 |
+
"step": 1390
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 5.511811023622047,
|
| 987 |
+
"grad_norm": 0.41949307918548584,
|
| 988 |
+
"learning_rate": 4.574129707699617e-05,
|
| 989 |
+
"loss": 0.0336,
|
| 990 |
+
"step": 1400
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 5.551181102362205,
|
| 994 |
+
"grad_norm": 0.7558311223983765,
|
| 995 |
+
"learning_rate": 4.509306954105028e-05,
|
| 996 |
+
"loss": 0.036,
|
| 997 |
+
"step": 1410
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 5.590551181102362,
|
| 1001 |
+
"grad_norm": 0.6058725118637085,
|
| 1002 |
+
"learning_rate": 4.4445673747559776e-05,
|
| 1003 |
+
"loss": 0.0389,
|
| 1004 |
+
"step": 1420
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 5.6299212598425195,
|
| 1008 |
+
"grad_norm": 0.5574952960014343,
|
| 1009 |
+
"learning_rate": 4.3799219432452527e-05,
|
| 1010 |
+
"loss": 0.0441,
|
| 1011 |
+
"step": 1430
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 5.669291338582677,
|
| 1015 |
+
"grad_norm": 0.3628334403038025,
|
| 1016 |
+
"learning_rate": 4.315381617207239e-05,
|
| 1017 |
+
"loss": 0.0306,
|
| 1018 |
+
"step": 1440
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 5.708661417322834,
|
| 1022 |
+
"grad_norm": 0.6688554286956787,
|
| 1023 |
+
"learning_rate": 4.2509573364605695e-05,
|
| 1024 |
+
"loss": 0.0384,
|
| 1025 |
+
"step": 1450
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 5.748031496062993,
|
| 1029 |
+
"grad_norm": 0.4437117874622345,
|
| 1030 |
+
"learning_rate": 4.1866600211537734e-05,
|
| 1031 |
+
"loss": 0.0359,
|
| 1032 |
+
"step": 1460
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 5.78740157480315,
|
| 1036 |
+
"grad_norm": 0.5025441646575928,
|
| 1037 |
+
"learning_rate": 4.122500569914285e-05,
|
| 1038 |
+
"loss": 0.0339,
|
| 1039 |
+
"step": 1470
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 5.826771653543307,
|
| 1043 |
+
"grad_norm": 0.6300631761550903,
|
| 1044 |
+
"learning_rate": 4.058489858001079e-05,
|
| 1045 |
+
"loss": 0.0412,
|
| 1046 |
+
"step": 1480
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 5.866141732283465,
|
| 1050 |
+
"grad_norm": 0.33207225799560547,
|
| 1051 |
+
"learning_rate": 3.9946387354612754e-05,
|
| 1052 |
+
"loss": 0.0382,
|
| 1053 |
+
"step": 1490
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 5.905511811023622,
|
| 1057 |
+
"grad_norm": 0.4854939877986908,
|
| 1058 |
+
"learning_rate": 3.930958025291021e-05,
|
| 1059 |
+
"loss": 0.0355,
|
| 1060 |
+
"step": 1500
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 5.94488188976378,
|
| 1064 |
+
"grad_norm": 0.4515645205974579,
|
| 1065 |
+
"learning_rate": 3.867458521600943e-05,
|
| 1066 |
+
"loss": 0.0403,
|
| 1067 |
+
"step": 1510
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 5.984251968503937,
|
| 1071 |
+
"grad_norm": 0.5812086462974548,
|
| 1072 |
+
"learning_rate": 3.804150987786525e-05,
|
| 1073 |
+
"loss": 0.045,
|
| 1074 |
+
"step": 1520
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 6.0236220472440944,
|
| 1078 |
+
"grad_norm": 0.4998365640640259,
|
| 1079 |
+
"learning_rate": 3.7410461547036534e-05,
|
| 1080 |
+
"loss": 0.0372,
|
| 1081 |
+
"step": 1530
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 6.062992125984252,
|
| 1085 |
+
"grad_norm": 0.2945879101753235,
|
| 1086 |
+
"learning_rate": 3.6781547188497135e-05,
|
| 1087 |
+
"loss": 0.0355,
|
| 1088 |
+
"step": 1540
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 6.102362204724409,
|
| 1092 |
+
"grad_norm": 0.4646545946598053,
|
| 1093 |
+
"learning_rate": 3.6154873405504895e-05,
|
| 1094 |
+
"loss": 0.0271,
|
| 1095 |
+
"step": 1550
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 6.141732283464567,
|
| 1099 |
+
"grad_norm": 0.5919941067695618,
|
| 1100 |
+
"learning_rate": 3.553054642153192e-05,
|
| 1101 |
+
"loss": 0.04,
|
| 1102 |
+
"step": 1560
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 6.181102362204724,
|
| 1106 |
+
"grad_norm": 0.4582063555717468,
|
| 1107 |
+
"learning_rate": 3.4908672062259487e-05,
|
| 1108 |
+
"loss": 0.0308,
|
| 1109 |
+
"step": 1570
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 6.2204724409448815,
|
| 1113 |
+
"grad_norm": 0.5543438196182251,
|
| 1114 |
+
"learning_rate": 3.428935573764005e-05,
|
| 1115 |
+
"loss": 0.0319,
|
| 1116 |
+
"step": 1580
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 6.259842519685039,
|
| 1120 |
+
"grad_norm": 0.3909936249256134,
|
| 1121 |
+
"learning_rate": 3.367270242402999e-05,
|
| 1122 |
+
"loss": 0.0305,
|
| 1123 |
+
"step": 1590
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 6.299212598425197,
|
| 1127 |
+
"grad_norm": 0.4638426601886749,
|
| 1128 |
+
"learning_rate": 3.30588166463957e-05,
|
| 1129 |
+
"loss": 0.0298,
|
| 1130 |
+
"step": 1600
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 6.338582677165355,
|
| 1134 |
+
"grad_norm": 0.5119015574455261,
|
| 1135 |
+
"learning_rate": 3.2447802460596124e-05,
|
| 1136 |
+
"loss": 0.0299,
|
| 1137 |
+
"step": 1610
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 6.377952755905512,
|
| 1141 |
+
"grad_norm": 0.4361736476421356,
|
| 1142 |
+
"learning_rate": 3.183976343574513e-05,
|
| 1143 |
+
"loss": 0.0279,
|
| 1144 |
+
"step": 1620
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 6.417322834645669,
|
| 1148 |
+
"grad_norm": 0.7282920479774475,
|
| 1149 |
+
"learning_rate": 3.123480263665597e-05,
|
| 1150 |
+
"loss": 0.0305,
|
| 1151 |
+
"step": 1630
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 6.456692913385827,
|
| 1155 |
+
"grad_norm": 0.4011118710041046,
|
| 1156 |
+
"learning_rate": 3.063302260637151e-05,
|
| 1157 |
+
"loss": 0.0294,
|
| 1158 |
+
"step": 1640
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 6.496062992125984,
|
| 1162 |
+
"grad_norm": 0.44994768500328064,
|
| 1163 |
+
"learning_rate": 3.0034525348782855e-05,
|
| 1164 |
+
"loss": 0.0256,
|
| 1165 |
+
"step": 1650
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 6.535433070866142,
|
| 1169 |
+
"grad_norm": 0.4626915454864502,
|
| 1170 |
+
"learning_rate": 2.9439412311339175e-05,
|
| 1171 |
+
"loss": 0.039,
|
| 1172 |
+
"step": 1660
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 6.574803149606299,
|
| 1176 |
+
"grad_norm": 0.6001310348510742,
|
| 1177 |
+
"learning_rate": 2.8847784367852184e-05,
|
| 1178 |
+
"loss": 0.025,
|
| 1179 |
+
"step": 1670
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 6.6141732283464565,
|
| 1183 |
+
"grad_norm": 0.38421395421028137,
|
| 1184 |
+
"learning_rate": 2.8259741801397477e-05,
|
| 1185 |
+
"loss": 0.0373,
|
| 1186 |
+
"step": 1680
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 6.653543307086614,
|
| 1190 |
+
"grad_norm": 0.37881389260292053,
|
| 1191 |
+
"learning_rate": 2.7675384287316363e-05,
|
| 1192 |
+
"loss": 0.034,
|
| 1193 |
+
"step": 1690
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 6.692913385826771,
|
| 1197 |
+
"grad_norm": 0.43850505352020264,
|
| 1198 |
+
"learning_rate": 2.709481087632041e-05,
|
| 1199 |
+
"loss": 0.0367,
|
| 1200 |
+
"step": 1700
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 6.73228346456693,
|
| 1204 |
+
"grad_norm": 0.6985974907875061,
|
| 1205 |
+
"learning_rate": 2.6518119977702e-05,
|
| 1206 |
+
"loss": 0.0364,
|
| 1207 |
+
"step": 1710
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 6.771653543307087,
|
| 1211 |
+
"grad_norm": 0.4026467502117157,
|
| 1212 |
+
"learning_rate": 2.5945409342653726e-05,
|
| 1213 |
+
"loss": 0.0363,
|
| 1214 |
+
"step": 1720
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 6.811023622047244,
|
| 1218 |
+
"grad_norm": 0.36351197957992554,
|
| 1219 |
+
"learning_rate": 2.5376776047698965e-05,
|
| 1220 |
+
"loss": 0.03,
|
| 1221 |
+
"step": 1730
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 6.850393700787402,
|
| 1225 |
+
"grad_norm": 0.2879717946052551,
|
| 1226 |
+
"learning_rate": 2.4812316478237353e-05,
|
| 1227 |
+
"loss": 0.0309,
|
| 1228 |
+
"step": 1740
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 6.889763779527559,
|
| 1232 |
+
"grad_norm": 0.6626470685005188,
|
| 1233 |
+
"learning_rate": 2.4252126312206873e-05,
|
| 1234 |
+
"loss": 0.0431,
|
| 1235 |
+
"step": 1750
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 6.929133858267717,
|
| 1239 |
+
"grad_norm": 0.4203033447265625,
|
| 1240 |
+
"learning_rate": 2.3696300503866204e-05,
|
| 1241 |
+
"loss": 0.0276,
|
| 1242 |
+
"step": 1760
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 6.968503937007874,
|
| 1246 |
+
"grad_norm": 0.5207045078277588,
|
| 1247 |
+
"learning_rate": 2.314493326769968e-05,
|
| 1248 |
+
"loss": 0.0294,
|
| 1249 |
+
"step": 1770
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 7.0078740157480315,
|
| 1253 |
+
"grad_norm": 0.4823305308818817,
|
| 1254 |
+
"learning_rate": 2.259811806244741e-05,
|
| 1255 |
+
"loss": 0.044,
|
| 1256 |
+
"step": 1780
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 7.047244094488189,
|
| 1260 |
+
"grad_norm": 0.6036306023597717,
|
| 1261 |
+
"learning_rate": 2.2055947575263912e-05,
|
| 1262 |
+
"loss": 0.0283,
|
| 1263 |
+
"step": 1790
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 7.086614173228346,
|
| 1267 |
+
"grad_norm": 0.5445898175239563,
|
| 1268 |
+
"learning_rate": 2.1518513706007155e-05,
|
| 1269 |
+
"loss": 0.0299,
|
| 1270 |
+
"step": 1800
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 7.125984251968504,
|
| 1274 |
+
"grad_norm": 0.44029784202575684,
|
| 1275 |
+
"learning_rate": 2.0985907551661206e-05,
|
| 1276 |
+
"loss": 0.0349,
|
| 1277 |
+
"step": 1810
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 7.165354330708661,
|
| 1281 |
+
"grad_norm": 0.4250989258289337,
|
| 1282 |
+
"learning_rate": 2.0458219390895106e-05,
|
| 1283 |
+
"loss": 0.0301,
|
| 1284 |
+
"step": 1820
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 7.2047244094488185,
|
| 1288 |
+
"grad_norm": 0.31232279539108276,
|
| 1289 |
+
"learning_rate": 1.9935538668760057e-05,
|
| 1290 |
+
"loss": 0.0421,
|
| 1291 |
+
"step": 1830
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 7.244094488188976,
|
| 1295 |
+
"grad_norm": 0.5633496642112732,
|
| 1296 |
+
"learning_rate": 1.9417953981528424e-05,
|
| 1297 |
+
"loss": 0.03,
|
| 1298 |
+
"step": 1840
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 7.283464566929134,
|
| 1302 |
+
"grad_norm": 0.7122541666030884,
|
| 1303 |
+
"learning_rate": 1.890555306167619e-05,
|
| 1304 |
+
"loss": 0.0343,
|
| 1305 |
+
"step": 1850
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 7.322834645669292,
|
| 1309 |
+
"grad_norm": 0.4240683615207672,
|
| 1310 |
+
"learning_rate": 1.8398422763011985e-05,
|
| 1311 |
+
"loss": 0.0244,
|
| 1312 |
+
"step": 1860
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 7.362204724409449,
|
| 1316 |
+
"grad_norm": 0.3466864824295044,
|
| 1317 |
+
"learning_rate": 1.789664904595518e-05,
|
| 1318 |
+
"loss": 0.0265,
|
| 1319 |
+
"step": 1870
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 7.4015748031496065,
|
| 1323 |
+
"grad_norm": 0.7498940229415894,
|
| 1324 |
+
"learning_rate": 1.7400316962965087e-05,
|
| 1325 |
+
"loss": 0.0303,
|
| 1326 |
+
"step": 1880
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 7.440944881889764,
|
| 1330 |
+
"grad_norm": 0.38159969449043274,
|
| 1331 |
+
"learning_rate": 1.6909510644124455e-05,
|
| 1332 |
+
"loss": 0.0261,
|
| 1333 |
+
"step": 1890
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 7.480314960629921,
|
| 1337 |
+
"grad_norm": 0.5678858160972595,
|
| 1338 |
+
"learning_rate": 1.642431328287899e-05,
|
| 1339 |
+
"loss": 0.035,
|
| 1340 |
+
"step": 1900
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 7.519685039370079,
|
| 1344 |
+
"grad_norm": 0.4290473461151123,
|
| 1345 |
+
"learning_rate": 1.594480712193579e-05,
|
| 1346 |
+
"loss": 0.026,
|
| 1347 |
+
"step": 1910
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 7.559055118110236,
|
| 1351 |
+
"grad_norm": 0.5522475242614746,
|
| 1352 |
+
"learning_rate": 1.547107343932299e-05,
|
| 1353 |
+
"loss": 0.029,
|
| 1354 |
+
"step": 1920
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 7.5984251968503935,
|
| 1358 |
+
"grad_norm": 0.41189444065093994,
|
| 1359 |
+
"learning_rate": 1.5003192534612675e-05,
|
| 1360 |
+
"loss": 0.0243,
|
| 1361 |
+
"step": 1930
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 7.637795275590551,
|
| 1365 |
+
"grad_norm": 0.1887262910604477,
|
| 1366 |
+
"learning_rate": 1.4541243715310005e-05,
|
| 1367 |
+
"loss": 0.0255,
|
| 1368 |
+
"step": 1940
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 7.677165354330708,
|
| 1372 |
+
"grad_norm": 0.5400164127349854,
|
| 1373 |
+
"learning_rate": 1.4085305283410166e-05,
|
| 1374 |
+
"loss": 0.0251,
|
| 1375 |
+
"step": 1950
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 7.716535433070866,
|
| 1379 |
+
"grad_norm": 0.5268674492835999,
|
| 1380 |
+
"learning_rate": 1.3635454522125946e-05,
|
| 1381 |
+
"loss": 0.036,
|
| 1382 |
+
"step": 1960
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 7.755905511811024,
|
| 1386 |
+
"grad_norm": 0.6157100200653076,
|
| 1387 |
+
"learning_rate": 1.3191767682788003e-05,
|
| 1388 |
+
"loss": 0.0266,
|
| 1389 |
+
"step": 1970
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 7.7952755905511815,
|
| 1393 |
+
"grad_norm": 0.3194139301776886,
|
| 1394 |
+
"learning_rate": 1.2754319971919842e-05,
|
| 1395 |
+
"loss": 0.0243,
|
| 1396 |
+
"step": 1980
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 7.834645669291339,
|
| 1400 |
+
"grad_norm": 0.37460631132125854,
|
| 1401 |
+
"learning_rate": 1.2323185538490229e-05,
|
| 1402 |
+
"loss": 0.0357,
|
| 1403 |
+
"step": 1990
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 7.874015748031496,
|
| 1407 |
+
"grad_norm": 0.37989112734794617,
|
| 1408 |
+
"learning_rate": 1.1898437461344518e-05,
|
| 1409 |
+
"loss": 0.0318,
|
| 1410 |
+
"step": 2000
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 7.913385826771654,
|
| 1414 |
+
"grad_norm": 0.42782655358314514,
|
| 1415 |
+
"learning_rate": 1.1480147736817598e-05,
|
| 1416 |
+
"loss": 0.0263,
|
| 1417 |
+
"step": 2010
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 7.952755905511811,
|
| 1421 |
+
"grad_norm": 0.5276915431022644,
|
| 1422 |
+
"learning_rate": 1.1068387266530267e-05,
|
| 1423 |
+
"loss": 0.025,
|
| 1424 |
+
"step": 2020
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 7.9921259842519685,
|
| 1428 |
+
"grad_norm": 0.6227043271064758,
|
| 1429 |
+
"learning_rate": 1.0663225845371045e-05,
|
| 1430 |
+
"loss": 0.0296,
|
| 1431 |
+
"step": 2030
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 8.031496062992126,
|
| 1435 |
+
"grad_norm": 0.2772444784641266,
|
| 1436 |
+
"learning_rate": 1.026473214966584e-05,
|
| 1437 |
+
"loss": 0.0346,
|
| 1438 |
+
"step": 2040
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 8.070866141732283,
|
| 1442 |
+
"grad_norm": 0.4647983908653259,
|
| 1443 |
+
"learning_rate": 9.872973725536955e-06,
|
| 1444 |
+
"loss": 0.0308,
|
| 1445 |
+
"step": 2050
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 8.11023622047244,
|
| 1449 |
+
"grad_norm": 0.32798609137535095,
|
| 1450 |
+
"learning_rate": 9.488016977453807e-06,
|
| 1451 |
+
"loss": 0.0248,
|
| 1452 |
+
"step": 2060
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 8.149606299212598,
|
| 1456 |
+
"grad_norm": 0.2955688238143921,
|
| 1457 |
+
"learning_rate": 9.109927156977122e-06,
|
| 1458 |
+
"loss": 0.0268,
|
| 1459 |
+
"step": 2070
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 8.188976377952756,
|
| 1463 |
+
"grad_norm": 0.3243492543697357,
|
| 1464 |
+
"learning_rate": 8.738768351698574e-06,
|
| 1465 |
+
"loss": 0.0273,
|
| 1466 |
+
"step": 2080
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 8.228346456692913,
|
| 1470 |
+
"grad_norm": 0.2228180319070816,
|
| 1471 |
+
"learning_rate": 8.374603474377718e-06,
|
| 1472 |
+
"loss": 0.0276,
|
| 1473 |
+
"step": 2090
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 8.26771653543307,
|
| 1477 |
+
"grad_norm": 0.4943491816520691,
|
| 1478 |
+
"learning_rate": 8.017494252278019e-06,
|
| 1479 |
+
"loss": 0.0338,
|
| 1480 |
+
"step": 2100
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 8.307086614173228,
|
| 1484 |
+
"grad_norm": 0.26574063301086426,
|
| 1485 |
+
"learning_rate": 7.667501216703849e-06,
|
| 1486 |
+
"loss": 0.0286,
|
| 1487 |
+
"step": 2110
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 8.346456692913385,
|
| 1491 |
+
"grad_norm": 0.561998188495636,
|
| 1492 |
+
"learning_rate": 7.324683692740259e-06,
|
| 1493 |
+
"loss": 0.0302,
|
| 1494 |
+
"step": 2120
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 8.385826771653543,
|
| 1498 |
+
"grad_norm": 0.5270190238952637,
|
| 1499 |
+
"learning_rate": 6.989099789197112e-06,
|
| 1500 |
+
"loss": 0.0309,
|
| 1501 |
+
"step": 2130
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 8.4251968503937,
|
| 1505 |
+
"grad_norm": 0.49484336376190186,
|
| 1506 |
+
"learning_rate": 6.660806388759505e-06,
|
| 1507 |
+
"loss": 0.0255,
|
| 1508 |
+
"step": 2140
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 8.464566929133857,
|
| 1512 |
+
"grad_norm": 0.5083401799201965,
|
| 1513 |
+
"learning_rate": 6.339859138345838e-06,
|
| 1514 |
+
"loss": 0.0253,
|
| 1515 |
+
"step": 2150
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 8.503937007874015,
|
| 1519 |
+
"grad_norm": 0.34489235281944275,
|
| 1520 |
+
"learning_rate": 6.026312439675552e-06,
|
| 1521 |
+
"loss": 0.0216,
|
| 1522 |
+
"step": 2160
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 8.543307086614174,
|
| 1526 |
+
"grad_norm": 0.41160085797309875,
|
| 1527 |
+
"learning_rate": 5.720219440047797e-06,
|
| 1528 |
+
"loss": 0.0168,
|
| 1529 |
+
"step": 2170
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 8.582677165354331,
|
| 1533 |
+
"grad_norm": 0.2923263609409332,
|
| 1534 |
+
"learning_rate": 5.421632023332779e-06,
|
| 1535 |
+
"loss": 0.0264,
|
| 1536 |
+
"step": 2180
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 8.622047244094489,
|
| 1540 |
+
"grad_norm": 0.522784411907196,
|
| 1541 |
+
"learning_rate": 5.130600801177294e-06,
|
| 1542 |
+
"loss": 0.0222,
|
| 1543 |
+
"step": 2190
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 8.661417322834646,
|
| 1547 |
+
"grad_norm": 0.413273423910141,
|
| 1548 |
+
"learning_rate": 4.8471751044257995e-06,
|
| 1549 |
+
"loss": 0.0312,
|
| 1550 |
+
"step": 2200
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 8.700787401574804,
|
| 1554 |
+
"grad_norm": 0.4753558039665222,
|
| 1555 |
+
"learning_rate": 4.571402974758715e-06,
|
| 1556 |
+
"loss": 0.0285,
|
| 1557 |
+
"step": 2210
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 8.740157480314961,
|
| 1561 |
+
"grad_norm": 0.36735445261001587,
|
| 1562 |
+
"learning_rate": 4.303331156549162e-06,
|
| 1563 |
+
"loss": 0.0303,
|
| 1564 |
+
"step": 2220
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 8.779527559055119,
|
| 1568 |
+
"grad_norm": 0.3742741048336029,
|
| 1569 |
+
"learning_rate": 4.043005088939616e-06,
|
| 1570 |
+
"loss": 0.0166,
|
| 1571 |
+
"step": 2230
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 8.818897637795276,
|
| 1575 |
+
"grad_norm": 0.2975374460220337,
|
| 1576 |
+
"learning_rate": 3.7904688981398485e-06,
|
| 1577 |
+
"loss": 0.0233,
|
| 1578 |
+
"step": 2240
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 8.858267716535433,
|
| 1582 |
+
"grad_norm": 0.4250410497188568,
|
| 1583 |
+
"learning_rate": 3.5457653899473197e-06,
|
| 1584 |
+
"loss": 0.0343,
|
| 1585 |
+
"step": 2250
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 8.89763779527559,
|
| 1589 |
+
"grad_norm": 0.41226938366889954,
|
| 1590 |
+
"learning_rate": 3.3089360424914674e-06,
|
| 1591 |
+
"loss": 0.0329,
|
| 1592 |
+
"step": 2260
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 8.937007874015748,
|
| 1596 |
+
"grad_norm": 0.18084678053855896,
|
| 1597 |
+
"learning_rate": 3.080020999203026e-06,
|
| 1598 |
+
"loss": 0.0195,
|
| 1599 |
+
"step": 2270
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 8.976377952755906,
|
| 1603 |
+
"grad_norm": 0.29392436146736145,
|
| 1604 |
+
"learning_rate": 2.8590590620095336e-06,
|
| 1605 |
+
"loss": 0.0257,
|
| 1606 |
+
"step": 2280
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 9.015748031496063,
|
| 1610 |
+
"grad_norm": 0.2687474489212036,
|
| 1611 |
+
"learning_rate": 2.646087684758325e-06,
|
| 1612 |
+
"loss": 0.0274,
|
| 1613 |
+
"step": 2290
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 9.05511811023622,
|
| 1617 |
+
"grad_norm": 0.1699579358100891,
|
| 1618 |
+
"learning_rate": 2.4411429668679043e-06,
|
| 1619 |
+
"loss": 0.0217,
|
| 1620 |
+
"step": 2300
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 9.094488188976378,
|
| 1624 |
+
"grad_norm": 0.2965308725833893,
|
| 1625 |
+
"learning_rate": 2.2442596472089907e-06,
|
| 1626 |
+
"loss": 0.0266,
|
| 1627 |
+
"step": 2310
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 9.133858267716535,
|
| 1631 |
+
"grad_norm": 0.29086264967918396,
|
| 1632 |
+
"learning_rate": 2.0554710982161607e-06,
|
| 1633 |
+
"loss": 0.0202,
|
| 1634 |
+
"step": 2320
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 9.173228346456693,
|
| 1638 |
+
"grad_norm": 0.525842010974884,
|
| 1639 |
+
"learning_rate": 1.8748093202311078e-06,
|
| 1640 |
+
"loss": 0.0252,
|
| 1641 |
+
"step": 2330
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 9.21259842519685,
|
| 1645 |
+
"grad_norm": 0.23323991894721985,
|
| 1646 |
+
"learning_rate": 1.7023049360784193e-06,
|
| 1647 |
+
"loss": 0.0257,
|
| 1648 |
+
"step": 2340
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 9.251968503937007,
|
| 1652 |
+
"grad_norm": 0.3511553704738617,
|
| 1653 |
+
"learning_rate": 1.5379871858749784e-06,
|
| 1654 |
+
"loss": 0.0341,
|
| 1655 |
+
"step": 2350
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 9.291338582677165,
|
| 1659 |
+
"grad_norm": 0.2637965977191925,
|
| 1660 |
+
"learning_rate": 1.3818839220735792e-06,
|
| 1661 |
+
"loss": 0.0207,
|
| 1662 |
+
"step": 2360
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 9.330708661417322,
|
| 1666 |
+
"grad_norm": 0.212178036570549,
|
| 1667 |
+
"learning_rate": 1.2340216047418695e-06,
|
| 1668 |
+
"loss": 0.0207,
|
| 1669 |
+
"step": 2370
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 9.37007874015748,
|
| 1673 |
+
"grad_norm": 0.38010916113853455,
|
| 1674 |
+
"learning_rate": 1.094425297077295e-06,
|
| 1675 |
+
"loss": 0.0239,
|
| 1676 |
+
"step": 2380
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 9.409448818897637,
|
| 1680 |
+
"grad_norm": 0.5460922718048096,
|
| 1681 |
+
"learning_rate": 9.631186611587405e-07,
|
| 1682 |
+
"loss": 0.0216,
|
| 1683 |
+
"step": 2390
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 9.448818897637794,
|
| 1687 |
+
"grad_norm": 0.3007522523403168,
|
| 1688 |
+
"learning_rate": 8.401239539358008e-07,
|
| 1689 |
+
"loss": 0.0301,
|
| 1690 |
+
"step": 2400
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 9.488188976377952,
|
| 1694 |
+
"grad_norm": 0.279300719499588,
|
| 1695 |
+
"learning_rate": 7.254620234560583e-07,
|
| 1696 |
+
"loss": 0.0313,
|
| 1697 |
+
"step": 2410
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 9.527559055118111,
|
| 1701 |
+
"grad_norm": 0.6319423913955688,
|
| 1702 |
+
"learning_rate": 6.191523053313386e-07,
|
| 1703 |
+
"loss": 0.0298,
|
| 1704 |
+
"step": 2420
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 9.566929133858268,
|
| 1708 |
+
"grad_norm": 0.5479516983032227,
|
| 1709 |
+
"learning_rate": 5.212128194432509e-07,
|
| 1710 |
+
"loss": 0.0222,
|
| 1711 |
+
"step": 2430
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 9.606299212598426,
|
| 1715 |
+
"grad_norm": 0.44871729612350464,
|
| 1716 |
+
"learning_rate": 4.3166016688879205e-07,
|
| 1717 |
+
"loss": 0.021,
|
| 1718 |
+
"step": 2440
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 9.645669291338583,
|
| 1722 |
+
"grad_norm": 0.5158320665359497,
|
| 1723 |
+
"learning_rate": 3.505095271663705e-07,
|
| 1724 |
+
"loss": 0.0174,
|
| 1725 |
+
"step": 2450
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 9.68503937007874,
|
| 1729 |
+
"grad_norm": 0.3390265107154846,
|
| 1730 |
+
"learning_rate": 2.7777465560285265e-07,
|
| 1731 |
+
"loss": 0.0224,
|
| 1732 |
+
"step": 2460
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 9.724409448818898,
|
| 1736 |
+
"grad_norm": 0.39901813864707947,
|
| 1737 |
+
"learning_rate": 2.1346788102196148e-07,
|
| 1738 |
+
"loss": 0.0181,
|
| 1739 |
+
"step": 2470
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 9.763779527559056,
|
| 1743 |
+
"grad_norm": 0.4687666893005371,
|
| 1744 |
+
"learning_rate": 1.5760010365450938e-07,
|
| 1745 |
+
"loss": 0.0213,
|
| 1746 |
+
"step": 2480
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 9.803149606299213,
|
| 1750 |
+
"grad_norm": 0.5264361500740051,
|
| 1751 |
+
"learning_rate": 1.1018079329076503e-07,
|
| 1752 |
+
"loss": 0.0249,
|
| 1753 |
+
"step": 2490
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 9.84251968503937,
|
| 1757 |
+
"grad_norm": 0.5656106472015381,
|
| 1758 |
+
"learning_rate": 7.121798767530385e-08,
|
| 1759 |
+
"loss": 0.0217,
|
| 1760 |
+
"step": 2500
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 9.881889763779528,
|
| 1764 |
+
"grad_norm": 0.3701815903186798,
|
| 1765 |
+
"learning_rate": 4.071829114455361e-08,
|
| 1766 |
+
"loss": 0.0198,
|
| 1767 |
+
"step": 2510
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 9.921259842519685,
|
| 1771 |
+
"grad_norm": 0.2982766628265381,
|
| 1772 |
+
"learning_rate": 1.868687350736198e-08,
|
| 1773 |
+
"loss": 0.0219,
|
| 1774 |
+
"step": 2520
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 9.960629921259843,
|
| 1778 |
+
"grad_norm": 0.3252989947795868,
|
| 1779 |
+
"learning_rate": 5.1274691686697965e-09,
|
| 1780 |
+
"loss": 0.021,
|
| 1781 |
+
"step": 2530
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 10.0,
|
| 1785 |
+
"grad_norm": 0.5611906051635742,
|
| 1786 |
+
"learning_rate": 4.23764965562512e-11,
|
| 1787 |
+
"loss": 0.0357,
|
| 1788 |
+
"step": 2540
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 10.0,
|
| 1792 |
+
"step": 2540,
|
| 1793 |
+
"total_flos": 0.0,
|
| 1794 |
+
"train_loss": 0.06096198053106548,
|
| 1795 |
+
"train_runtime": 2181.3746,
|
| 1796 |
+
"train_samples_per_second": 31.434,
|
| 1797 |
+
"train_steps_per_second": 1.164
|
| 1798 |
+
}
|
| 1799 |
+
],
|
| 1800 |
+
"logging_steps": 10,
|
| 1801 |
+
"max_steps": 2540,
|
| 1802 |
+
"num_input_tokens_seen": 0,
|
| 1803 |
+
"num_train_epochs": 10,
|
| 1804 |
+
"save_steps": 1000,
|
| 1805 |
+
"stateful_callbacks": {
|
| 1806 |
+
"TrainerControl": {
|
| 1807 |
+
"args": {
|
| 1808 |
+
"should_epoch_stop": false,
|
| 1809 |
+
"should_evaluate": false,
|
| 1810 |
+
"should_log": false,
|
| 1811 |
+
"should_save": true,
|
| 1812 |
+
"should_training_stop": true
|
| 1813 |
+
},
|
| 1814 |
+
"attributes": {}
|
| 1815 |
+
}
|
| 1816 |
+
},
|
| 1817 |
+
"total_flos": 0.0,
|
| 1818 |
+
"train_batch_size": 27,
|
| 1819 |
+
"trial_name": null,
|
| 1820 |
+
"trial_params": null
|
| 1821 |
+
}
|