Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +3802 -0
trainer_state.json
ADDED
@@ -0,0 +1,3802 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_global_step": null,
|
3 |
+
"best_metric": null,
|
4 |
+
"best_model_checkpoint": null,
|
5 |
+
"epoch": 30.0,
|
6 |
+
"eval_steps": 500,
|
7 |
+
"global_step": 5370,
|
8 |
+
"is_hyper_param_search": false,
|
9 |
+
"is_local_process_zero": true,
|
10 |
+
"is_world_process_zero": true,
|
11 |
+
"log_history": [
|
12 |
+
{
|
13 |
+
"epoch": 0.055865921787709494,
|
14 |
+
"grad_norm": 9.474554061889648,
|
15 |
+
"learning_rate": 5.018587360594795e-06,
|
16 |
+
"loss": 0.9647,
|
17 |
+
"step": 10
|
18 |
+
},
|
19 |
+
{
|
20 |
+
"epoch": 0.11173184357541899,
|
21 |
+
"grad_norm": 2.904832124710083,
|
22 |
+
"learning_rate": 1.0594795539033457e-05,
|
23 |
+
"loss": 0.3646,
|
24 |
+
"step": 20
|
25 |
+
},
|
26 |
+
{
|
27 |
+
"epoch": 0.16759776536312848,
|
28 |
+
"grad_norm": 2.1869466304779053,
|
29 |
+
"learning_rate": 1.617100371747212e-05,
|
30 |
+
"loss": 0.2295,
|
31 |
+
"step": 30
|
32 |
+
},
|
33 |
+
{
|
34 |
+
"epoch": 0.22346368715083798,
|
35 |
+
"grad_norm": 1.0514620542526245,
|
36 |
+
"learning_rate": 2.1747211895910778e-05,
|
37 |
+
"loss": 0.1933,
|
38 |
+
"step": 40
|
39 |
+
},
|
40 |
+
{
|
41 |
+
"epoch": 0.27932960893854747,
|
42 |
+
"grad_norm": 1.2703338861465454,
|
43 |
+
"learning_rate": 2.732342007434944e-05,
|
44 |
+
"loss": 0.16,
|
45 |
+
"step": 50
|
46 |
+
},
|
47 |
+
{
|
48 |
+
"epoch": 0.33519553072625696,
|
49 |
+
"grad_norm": 1.139123797416687,
|
50 |
+
"learning_rate": 3.28996282527881e-05,
|
51 |
+
"loss": 0.1487,
|
52 |
+
"step": 60
|
53 |
+
},
|
54 |
+
{
|
55 |
+
"epoch": 0.39106145251396646,
|
56 |
+
"grad_norm": 1.1433963775634766,
|
57 |
+
"learning_rate": 3.8475836431226764e-05,
|
58 |
+
"loss": 0.1484,
|
59 |
+
"step": 70
|
60 |
+
},
|
61 |
+
{
|
62 |
+
"epoch": 0.44692737430167595,
|
63 |
+
"grad_norm": 1.1463357210159302,
|
64 |
+
"learning_rate": 4.4052044609665426e-05,
|
65 |
+
"loss": 0.1315,
|
66 |
+
"step": 80
|
67 |
+
},
|
68 |
+
{
|
69 |
+
"epoch": 0.5027932960893855,
|
70 |
+
"grad_norm": 1.1672674417495728,
|
71 |
+
"learning_rate": 4.962825278810409e-05,
|
72 |
+
"loss": 0.1267,
|
73 |
+
"step": 90
|
74 |
+
},
|
75 |
+
{
|
76 |
+
"epoch": 0.5586592178770949,
|
77 |
+
"grad_norm": 1.0597262382507324,
|
78 |
+
"learning_rate": 5.520446096654275e-05,
|
79 |
+
"loss": 0.1244,
|
80 |
+
"step": 100
|
81 |
+
},
|
82 |
+
{
|
83 |
+
"epoch": 0.6145251396648045,
|
84 |
+
"grad_norm": 1.1038429737091064,
|
85 |
+
"learning_rate": 6.0780669144981406e-05,
|
86 |
+
"loss": 0.1147,
|
87 |
+
"step": 110
|
88 |
+
},
|
89 |
+
{
|
90 |
+
"epoch": 0.6703910614525139,
|
91 |
+
"grad_norm": 0.6325312852859497,
|
92 |
+
"learning_rate": 6.635687732342006e-05,
|
93 |
+
"loss": 0.1192,
|
94 |
+
"step": 120
|
95 |
+
},
|
96 |
+
{
|
97 |
+
"epoch": 0.7262569832402235,
|
98 |
+
"grad_norm": 1.0795389413833618,
|
99 |
+
"learning_rate": 7.193308550185873e-05,
|
100 |
+
"loss": 0.1105,
|
101 |
+
"step": 130
|
102 |
+
},
|
103 |
+
{
|
104 |
+
"epoch": 0.7821229050279329,
|
105 |
+
"grad_norm": 0.7824896574020386,
|
106 |
+
"learning_rate": 7.750929368029739e-05,
|
107 |
+
"loss": 0.1035,
|
108 |
+
"step": 140
|
109 |
+
},
|
110 |
+
{
|
111 |
+
"epoch": 0.8379888268156425,
|
112 |
+
"grad_norm": 1.3778011798858643,
|
113 |
+
"learning_rate": 8.308550185873605e-05,
|
114 |
+
"loss": 0.1069,
|
115 |
+
"step": 150
|
116 |
+
},
|
117 |
+
{
|
118 |
+
"epoch": 0.8938547486033519,
|
119 |
+
"grad_norm": 0.8529579639434814,
|
120 |
+
"learning_rate": 8.866171003717471e-05,
|
121 |
+
"loss": 0.1022,
|
122 |
+
"step": 160
|
123 |
+
},
|
124 |
+
{
|
125 |
+
"epoch": 0.9497206703910615,
|
126 |
+
"grad_norm": 1.441953182220459,
|
127 |
+
"learning_rate": 9.423791821561338e-05,
|
128 |
+
"loss": 0.0985,
|
129 |
+
"step": 170
|
130 |
+
},
|
131 |
+
{
|
132 |
+
"epoch": 1.005586592178771,
|
133 |
+
"grad_norm": 0.7136903405189514,
|
134 |
+
"learning_rate": 9.981412639405203e-05,
|
135 |
+
"loss": 0.1026,
|
136 |
+
"step": 180
|
137 |
+
},
|
138 |
+
{
|
139 |
+
"epoch": 1.0614525139664805,
|
140 |
+
"grad_norm": 1.4089112281799316,
|
141 |
+
"learning_rate": 0.0001053903345724907,
|
142 |
+
"loss": 0.0973,
|
143 |
+
"step": 190
|
144 |
+
},
|
145 |
+
{
|
146 |
+
"epoch": 1.1173184357541899,
|
147 |
+
"grad_norm": 0.8666443228721619,
|
148 |
+
"learning_rate": 0.00011096654275092937,
|
149 |
+
"loss": 0.0899,
|
150 |
+
"step": 200
|
151 |
+
},
|
152 |
+
{
|
153 |
+
"epoch": 1.1731843575418994,
|
154 |
+
"grad_norm": 0.6051232218742371,
|
155 |
+
"learning_rate": 0.00011654275092936801,
|
156 |
+
"loss": 0.0891,
|
157 |
+
"step": 210
|
158 |
+
},
|
159 |
+
{
|
160 |
+
"epoch": 1.229050279329609,
|
161 |
+
"grad_norm": 0.5882899165153503,
|
162 |
+
"learning_rate": 0.00012211895910780668,
|
163 |
+
"loss": 0.0859,
|
164 |
+
"step": 220
|
165 |
+
},
|
166 |
+
{
|
167 |
+
"epoch": 1.2849162011173183,
|
168 |
+
"grad_norm": 0.662291407585144,
|
169 |
+
"learning_rate": 0.00012769516728624534,
|
170 |
+
"loss": 0.0835,
|
171 |
+
"step": 230
|
172 |
+
},
|
173 |
+
{
|
174 |
+
"epoch": 1.3407821229050279,
|
175 |
+
"grad_norm": 1.036959171295166,
|
176 |
+
"learning_rate": 0.000133271375464684,
|
177 |
+
"loss": 0.0801,
|
178 |
+
"step": 240
|
179 |
+
},
|
180 |
+
{
|
181 |
+
"epoch": 1.3966480446927374,
|
182 |
+
"grad_norm": 0.5812122225761414,
|
183 |
+
"learning_rate": 0.00013884758364312265,
|
184 |
+
"loss": 0.0723,
|
185 |
+
"step": 250
|
186 |
+
},
|
187 |
+
{
|
188 |
+
"epoch": 1.452513966480447,
|
189 |
+
"grad_norm": 0.799187183380127,
|
190 |
+
"learning_rate": 0.00014442379182156133,
|
191 |
+
"loss": 0.0701,
|
192 |
+
"step": 260
|
193 |
+
},
|
194 |
+
{
|
195 |
+
"epoch": 1.5083798882681565,
|
196 |
+
"grad_norm": 1.0647534132003784,
|
197 |
+
"learning_rate": 0.00015,
|
198 |
+
"loss": 0.075,
|
199 |
+
"step": 270
|
200 |
+
},
|
201 |
+
{
|
202 |
+
"epoch": 1.564245810055866,
|
203 |
+
"grad_norm": 0.9664549827575684,
|
204 |
+
"learning_rate": 0.0001499985776090078,
|
205 |
+
"loss": 0.0696,
|
206 |
+
"step": 280
|
207 |
+
},
|
208 |
+
{
|
209 |
+
"epoch": 1.6201117318435754,
|
210 |
+
"grad_norm": 0.621644139289856,
|
211 |
+
"learning_rate": 0.0001499943104899832,
|
212 |
+
"loss": 0.0657,
|
213 |
+
"step": 290
|
214 |
+
},
|
215 |
+
{
|
216 |
+
"epoch": 1.675977653631285,
|
217 |
+
"grad_norm": 0.6787592768669128,
|
218 |
+
"learning_rate": 0.00014998719880477977,
|
219 |
+
"loss": 0.0731,
|
220 |
+
"step": 300
|
221 |
+
},
|
222 |
+
{
|
223 |
+
"epoch": 1.7318435754189943,
|
224 |
+
"grad_norm": 0.563049852848053,
|
225 |
+
"learning_rate": 0.00014997724282314682,
|
226 |
+
"loss": 0.0707,
|
227 |
+
"step": 310
|
228 |
+
},
|
229 |
+
{
|
230 |
+
"epoch": 1.7877094972067038,
|
231 |
+
"grad_norm": 0.4710414409637451,
|
232 |
+
"learning_rate": 0.00014996444292271892,
|
233 |
+
"loss": 0.0657,
|
234 |
+
"step": 320
|
235 |
+
},
|
236 |
+
{
|
237 |
+
"epoch": 1.8435754189944134,
|
238 |
+
"grad_norm": 0.46238359808921814,
|
239 |
+
"learning_rate": 0.0001499487995890018,
|
240 |
+
"loss": 0.0649,
|
241 |
+
"step": 330
|
242 |
+
},
|
243 |
+
{
|
244 |
+
"epoch": 1.899441340782123,
|
245 |
+
"grad_norm": 0.4369705021381378,
|
246 |
+
"learning_rate": 0.00014993031341535375,
|
247 |
+
"loss": 0.0562,
|
248 |
+
"step": 340
|
249 |
+
},
|
250 |
+
{
|
251 |
+
"epoch": 1.9553072625698324,
|
252 |
+
"grad_norm": 0.5079113841056824,
|
253 |
+
"learning_rate": 0.0001499089851029632,
|
254 |
+
"loss": 0.0553,
|
255 |
+
"step": 350
|
256 |
+
},
|
257 |
+
{
|
258 |
+
"epoch": 2.011173184357542,
|
259 |
+
"grad_norm": 0.7058802247047424,
|
260 |
+
"learning_rate": 0.00014988481546082222,
|
261 |
+
"loss": 0.0551,
|
262 |
+
"step": 360
|
263 |
+
},
|
264 |
+
{
|
265 |
+
"epoch": 2.0670391061452515,
|
266 |
+
"grad_norm": 0.42536088824272156,
|
267 |
+
"learning_rate": 0.0001498578054056956,
|
268 |
+
"loss": 0.0553,
|
269 |
+
"step": 370
|
270 |
+
},
|
271 |
+
{
|
272 |
+
"epoch": 2.122905027932961,
|
273 |
+
"grad_norm": 0.5014165043830872,
|
274 |
+
"learning_rate": 0.00014982795596208618,
|
275 |
+
"loss": 0.0539,
|
276 |
+
"step": 380
|
277 |
+
},
|
278 |
+
{
|
279 |
+
"epoch": 2.17877094972067,
|
280 |
+
"grad_norm": 0.6823521852493286,
|
281 |
+
"learning_rate": 0.0001497952682621962,
|
282 |
+
"loss": 0.0507,
|
283 |
+
"step": 390
|
284 |
+
},
|
285 |
+
{
|
286 |
+
"epoch": 2.2346368715083798,
|
287 |
+
"grad_norm": 0.4654916822910309,
|
288 |
+
"learning_rate": 0.00014975974354588395,
|
289 |
+
"loss": 0.058,
|
290 |
+
"step": 400
|
291 |
+
},
|
292 |
+
{
|
293 |
+
"epoch": 2.2905027932960893,
|
294 |
+
"grad_norm": 0.33408060669898987,
|
295 |
+
"learning_rate": 0.00014972138316061715,
|
296 |
+
"loss": 0.0484,
|
297 |
+
"step": 410
|
298 |
+
},
|
299 |
+
{
|
300 |
+
"epoch": 2.346368715083799,
|
301 |
+
"grad_norm": 0.40119659900665283,
|
302 |
+
"learning_rate": 0.00014968018856142154,
|
303 |
+
"loss": 0.051,
|
304 |
+
"step": 420
|
305 |
+
},
|
306 |
+
{
|
307 |
+
"epoch": 2.4022346368715084,
|
308 |
+
"grad_norm": 0.3584083616733551,
|
309 |
+
"learning_rate": 0.00014963616131082584,
|
310 |
+
"loss": 0.0494,
|
311 |
+
"step": 430
|
312 |
+
},
|
313 |
+
{
|
314 |
+
"epoch": 2.458100558659218,
|
315 |
+
"grad_norm": 0.6725554466247559,
|
316 |
+
"learning_rate": 0.00014958930307880242,
|
317 |
+
"loss": 0.0572,
|
318 |
+
"step": 440
|
319 |
+
},
|
320 |
+
{
|
321 |
+
"epoch": 2.5139664804469275,
|
322 |
+
"grad_norm": 0.582353413105011,
|
323 |
+
"learning_rate": 0.00014953961564270404,
|
324 |
+
"loss": 0.0495,
|
325 |
+
"step": 450
|
326 |
+
},
|
327 |
+
{
|
328 |
+
"epoch": 2.5698324022346366,
|
329 |
+
"grad_norm": 0.6442360877990723,
|
330 |
+
"learning_rate": 0.00014948710088719632,
|
331 |
+
"loss": 0.0581,
|
332 |
+
"step": 460
|
333 |
+
},
|
334 |
+
{
|
335 |
+
"epoch": 2.6256983240223466,
|
336 |
+
"grad_norm": 0.6121304631233215,
|
337 |
+
"learning_rate": 0.00014943176080418633,
|
338 |
+
"loss": 0.0528,
|
339 |
+
"step": 470
|
340 |
+
},
|
341 |
+
{
|
342 |
+
"epoch": 2.6815642458100557,
|
343 |
+
"grad_norm": 0.6058341860771179,
|
344 |
+
"learning_rate": 0.00014937359749274705,
|
345 |
+
"loss": 0.0504,
|
346 |
+
"step": 480
|
347 |
+
},
|
348 |
+
{
|
349 |
+
"epoch": 2.7374301675977653,
|
350 |
+
"grad_norm": 0.4608060121536255,
|
351 |
+
"learning_rate": 0.00014931261315903765,
|
352 |
+
"loss": 0.0501,
|
353 |
+
"step": 490
|
354 |
+
},
|
355 |
+
{
|
356 |
+
"epoch": 2.793296089385475,
|
357 |
+
"grad_norm": 0.33580687642097473,
|
358 |
+
"learning_rate": 0.00014924881011621992,
|
359 |
+
"loss": 0.0441,
|
360 |
+
"step": 500
|
361 |
+
},
|
362 |
+
{
|
363 |
+
"epoch": 2.8491620111731844,
|
364 |
+
"grad_norm": 0.5434100031852722,
|
365 |
+
"learning_rate": 0.0001491821907843705,
|
366 |
+
"loss": 0.0469,
|
367 |
+
"step": 510
|
368 |
+
},
|
369 |
+
{
|
370 |
+
"epoch": 2.905027932960894,
|
371 |
+
"grad_norm": 0.7853328585624695,
|
372 |
+
"learning_rate": 0.00014911275769038902,
|
373 |
+
"loss": 0.0435,
|
374 |
+
"step": 520
|
375 |
+
},
|
376 |
+
{
|
377 |
+
"epoch": 2.9608938547486034,
|
378 |
+
"grad_norm": 0.6744305491447449,
|
379 |
+
"learning_rate": 0.0001490405134679024,
|
380 |
+
"loss": 0.0477,
|
381 |
+
"step": 530
|
382 |
+
},
|
383 |
+
{
|
384 |
+
"epoch": 3.016759776536313,
|
385 |
+
"grad_norm": 0.6475287079811096,
|
386 |
+
"learning_rate": 0.00014896546085716475,
|
387 |
+
"loss": 0.0468,
|
388 |
+
"step": 540
|
389 |
+
},
|
390 |
+
{
|
391 |
+
"epoch": 3.0726256983240225,
|
392 |
+
"grad_norm": 0.5036594271659851,
|
393 |
+
"learning_rate": 0.00014888760270495365,
|
394 |
+
"loss": 0.044,
|
395 |
+
"step": 550
|
396 |
+
},
|
397 |
+
{
|
398 |
+
"epoch": 3.1284916201117317,
|
399 |
+
"grad_norm": 0.4917600452899933,
|
400 |
+
"learning_rate": 0.000148806941964462,
|
401 |
+
"loss": 0.0478,
|
402 |
+
"step": 560
|
403 |
+
},
|
404 |
+
{
|
405 |
+
"epoch": 3.184357541899441,
|
406 |
+
"grad_norm": 0.5577938556671143,
|
407 |
+
"learning_rate": 0.0001487234816951861,
|
408 |
+
"loss": 0.0464,
|
409 |
+
"step": 570
|
410 |
+
},
|
411 |
+
{
|
412 |
+
"epoch": 3.2402234636871508,
|
413 |
+
"grad_norm": 0.452462375164032,
|
414 |
+
"learning_rate": 0.0001486372250628095,
|
415 |
+
"loss": 0.041,
|
416 |
+
"step": 580
|
417 |
+
},
|
418 |
+
{
|
419 |
+
"epoch": 3.2960893854748603,
|
420 |
+
"grad_norm": 0.3954116702079773,
|
421 |
+
"learning_rate": 0.00014854817533908313,
|
422 |
+
"loss": 0.0437,
|
423 |
+
"step": 590
|
424 |
+
},
|
425 |
+
{
|
426 |
+
"epoch": 3.35195530726257,
|
427 |
+
"grad_norm": 0.529547393321991,
|
428 |
+
"learning_rate": 0.00014845633590170092,
|
429 |
+
"loss": 0.0433,
|
430 |
+
"step": 600
|
431 |
+
},
|
432 |
+
{
|
433 |
+
"epoch": 3.4078212290502794,
|
434 |
+
"grad_norm": 0.3528311848640442,
|
435 |
+
"learning_rate": 0.00014836171023417191,
|
436 |
+
"loss": 0.0365,
|
437 |
+
"step": 610
|
438 |
+
},
|
439 |
+
{
|
440 |
+
"epoch": 3.463687150837989,
|
441 |
+
"grad_norm": 0.4728206992149353,
|
442 |
+
"learning_rate": 0.00014826430192568807,
|
443 |
+
"loss": 0.0457,
|
444 |
+
"step": 620
|
445 |
+
},
|
446 |
+
{
|
447 |
+
"epoch": 3.5195530726256985,
|
448 |
+
"grad_norm": 0.6676551103591919,
|
449 |
+
"learning_rate": 0.00014816411467098806,
|
450 |
+
"loss": 0.045,
|
451 |
+
"step": 630
|
452 |
+
},
|
453 |
+
{
|
454 |
+
"epoch": 3.5754189944134076,
|
455 |
+
"grad_norm": 0.3946695327758789,
|
456 |
+
"learning_rate": 0.00014806115227021714,
|
457 |
+
"loss": 0.0465,
|
458 |
+
"step": 640
|
459 |
+
},
|
460 |
+
{
|
461 |
+
"epoch": 3.631284916201117,
|
462 |
+
"grad_norm": 0.6944833993911743,
|
463 |
+
"learning_rate": 0.00014795541862878308,
|
464 |
+
"loss": 0.0481,
|
465 |
+
"step": 650
|
466 |
+
},
|
467 |
+
{
|
468 |
+
"epoch": 3.6871508379888267,
|
469 |
+
"grad_norm": 0.6339210271835327,
|
470 |
+
"learning_rate": 0.00014784691775720807,
|
471 |
+
"loss": 0.0394,
|
472 |
+
"step": 660
|
473 |
+
},
|
474 |
+
{
|
475 |
+
"epoch": 3.7430167597765363,
|
476 |
+
"grad_norm": 0.47451841831207275,
|
477 |
+
"learning_rate": 0.00014773565377097631,
|
478 |
+
"loss": 0.0392,
|
479 |
+
"step": 670
|
480 |
+
},
|
481 |
+
{
|
482 |
+
"epoch": 3.798882681564246,
|
483 |
+
"grad_norm": 0.3181611895561218,
|
484 |
+
"learning_rate": 0.0001476216308903784,
|
485 |
+
"loss": 0.0436,
|
486 |
+
"step": 680
|
487 |
+
},
|
488 |
+
{
|
489 |
+
"epoch": 3.8547486033519553,
|
490 |
+
"grad_norm": 0.4814484715461731,
|
491 |
+
"learning_rate": 0.00014750485344035071,
|
492 |
+
"loss": 0.0383,
|
493 |
+
"step": 690
|
494 |
+
},
|
495 |
+
{
|
496 |
+
"epoch": 3.910614525139665,
|
497 |
+
"grad_norm": 0.3436708450317383,
|
498 |
+
"learning_rate": 0.00014738532585031178,
|
499 |
+
"loss": 0.044,
|
500 |
+
"step": 700
|
501 |
+
},
|
502 |
+
{
|
503 |
+
"epoch": 3.9664804469273744,
|
504 |
+
"grad_norm": 0.367412269115448,
|
505 |
+
"learning_rate": 0.00014726305265399403,
|
506 |
+
"loss": 0.0403,
|
507 |
+
"step": 710
|
508 |
+
},
|
509 |
+
{
|
510 |
+
"epoch": 4.022346368715084,
|
511 |
+
"grad_norm": 0.23561592400074005,
|
512 |
+
"learning_rate": 0.00014713803848927196,
|
513 |
+
"loss": 0.0422,
|
514 |
+
"step": 720
|
515 |
+
},
|
516 |
+
{
|
517 |
+
"epoch": 4.078212290502793,
|
518 |
+
"grad_norm": 0.33305707573890686,
|
519 |
+
"learning_rate": 0.00014701028809798619,
|
520 |
+
"loss": 0.0431,
|
521 |
+
"step": 730
|
522 |
+
},
|
523 |
+
{
|
524 |
+
"epoch": 4.134078212290503,
|
525 |
+
"grad_norm": 0.43914180994033813,
|
526 |
+
"learning_rate": 0.00014687980632576347,
|
527 |
+
"loss": 0.0432,
|
528 |
+
"step": 740
|
529 |
+
},
|
530 |
+
{
|
531 |
+
"epoch": 4.189944134078212,
|
532 |
+
"grad_norm": 0.41326114535331726,
|
533 |
+
"learning_rate": 0.0001467465981218331,
|
534 |
+
"loss": 0.0408,
|
535 |
+
"step": 750
|
536 |
+
},
|
537 |
+
{
|
538 |
+
"epoch": 4.245810055865922,
|
539 |
+
"grad_norm": 0.3312179148197174,
|
540 |
+
"learning_rate": 0.0001466106685388391,
|
541 |
+
"loss": 0.0437,
|
542 |
+
"step": 760
|
543 |
+
},
|
544 |
+
{
|
545 |
+
"epoch": 4.301675977653631,
|
546 |
+
"grad_norm": 0.24370786547660828,
|
547 |
+
"learning_rate": 0.00014647202273264848,
|
548 |
+
"loss": 0.0381,
|
549 |
+
"step": 770
|
550 |
+
},
|
551 |
+
{
|
552 |
+
"epoch": 4.35754189944134,
|
553 |
+
"grad_norm": 0.3902948796749115,
|
554 |
+
"learning_rate": 0.00014633066596215577,
|
555 |
+
"loss": 0.039,
|
556 |
+
"step": 780
|
557 |
+
},
|
558 |
+
{
|
559 |
+
"epoch": 4.41340782122905,
|
560 |
+
"grad_norm": 0.48275288939476013,
|
561 |
+
"learning_rate": 0.0001461866035890836,
|
562 |
+
"loss": 0.0395,
|
563 |
+
"step": 790
|
564 |
+
},
|
565 |
+
{
|
566 |
+
"epoch": 4.4692737430167595,
|
567 |
+
"grad_norm": 0.45421525835990906,
|
568 |
+
"learning_rate": 0.00014603984107777924,
|
569 |
+
"loss": 0.0386,
|
570 |
+
"step": 800
|
571 |
+
},
|
572 |
+
{
|
573 |
+
"epoch": 4.5251396648044695,
|
574 |
+
"grad_norm": 0.2560478150844574,
|
575 |
+
"learning_rate": 0.00014589038399500726,
|
576 |
+
"loss": 0.0375,
|
577 |
+
"step": 810
|
578 |
+
},
|
579 |
+
{
|
580 |
+
"epoch": 4.581005586592179,
|
581 |
+
"grad_norm": 0.3527555465698242,
|
582 |
+
"learning_rate": 0.0001457382380097386,
|
583 |
+
"loss": 0.046,
|
584 |
+
"step": 820
|
585 |
+
},
|
586 |
+
{
|
587 |
+
"epoch": 4.636871508379889,
|
588 |
+
"grad_norm": 0.5826641917228699,
|
589 |
+
"learning_rate": 0.00014558340889293534,
|
590 |
+
"loss": 0.0419,
|
591 |
+
"step": 830
|
592 |
+
},
|
593 |
+
{
|
594 |
+
"epoch": 4.692737430167598,
|
595 |
+
"grad_norm": 0.5426205992698669,
|
596 |
+
"learning_rate": 0.00014542590251733192,
|
597 |
+
"loss": 0.0449,
|
598 |
+
"step": 840
|
599 |
+
},
|
600 |
+
{
|
601 |
+
"epoch": 4.748603351955307,
|
602 |
+
"grad_norm": 0.4121716618537903,
|
603 |
+
"learning_rate": 0.00014526572485721234,
|
604 |
+
"loss": 0.0404,
|
605 |
+
"step": 850
|
606 |
+
},
|
607 |
+
{
|
608 |
+
"epoch": 4.804469273743017,
|
609 |
+
"grad_norm": 0.41459986567497253,
|
610 |
+
"learning_rate": 0.00014510288198818356,
|
611 |
+
"loss": 0.0434,
|
612 |
+
"step": 860
|
613 |
+
},
|
614 |
+
{
|
615 |
+
"epoch": 4.860335195530726,
|
616 |
+
"grad_norm": 0.42841342091560364,
|
617 |
+
"learning_rate": 0.00014493738008694503,
|
618 |
+
"loss": 0.0418,
|
619 |
+
"step": 870
|
620 |
+
},
|
621 |
+
{
|
622 |
+
"epoch": 4.916201117318436,
|
623 |
+
"grad_norm": 0.3888803720474243,
|
624 |
+
"learning_rate": 0.00014476922543105443,
|
625 |
+
"loss": 0.0391,
|
626 |
+
"step": 880
|
627 |
+
},
|
628 |
+
{
|
629 |
+
"epoch": 4.972067039106145,
|
630 |
+
"grad_norm": 0.2680191695690155,
|
631 |
+
"learning_rate": 0.00014459842439868963,
|
632 |
+
"loss": 0.0412,
|
633 |
+
"step": 890
|
634 |
+
},
|
635 |
+
{
|
636 |
+
"epoch": 5.027932960893855,
|
637 |
+
"grad_norm": 0.3152766823768616,
|
638 |
+
"learning_rate": 0.00014442498346840658,
|
639 |
+
"loss": 0.0351,
|
640 |
+
"step": 900
|
641 |
+
},
|
642 |
+
{
|
643 |
+
"epoch": 5.083798882681564,
|
644 |
+
"grad_norm": 0.3976776897907257,
|
645 |
+
"learning_rate": 0.00014424890921889373,
|
646 |
+
"loss": 0.0369,
|
647 |
+
"step": 910
|
648 |
+
},
|
649 |
+
{
|
650 |
+
"epoch": 5.139664804469274,
|
651 |
+
"grad_norm": 0.4556487500667572,
|
652 |
+
"learning_rate": 0.00014407020832872246,
|
653 |
+
"loss": 0.0337,
|
654 |
+
"step": 920
|
655 |
+
},
|
656 |
+
{
|
657 |
+
"epoch": 5.195530726256983,
|
658 |
+
"grad_norm": 0.368202269077301,
|
659 |
+
"learning_rate": 0.00014388888757609376,
|
660 |
+
"loss": 0.0396,
|
661 |
+
"step": 930
|
662 |
+
},
|
663 |
+
{
|
664 |
+
"epoch": 5.251396648044693,
|
665 |
+
"grad_norm": 0.34850746393203735,
|
666 |
+
"learning_rate": 0.00014370495383858107,
|
667 |
+
"loss": 0.0361,
|
668 |
+
"step": 940
|
669 |
+
},
|
670 |
+
{
|
671 |
+
"epoch": 5.307262569832402,
|
672 |
+
"grad_norm": 0.21902212500572205,
|
673 |
+
"learning_rate": 0.00014351841409286954,
|
674 |
+
"loss": 0.0332,
|
675 |
+
"step": 950
|
676 |
+
},
|
677 |
+
{
|
678 |
+
"epoch": 5.363128491620111,
|
679 |
+
"grad_norm": 0.37780269980430603,
|
680 |
+
"learning_rate": 0.00014332927541449122,
|
681 |
+
"loss": 0.0359,
|
682 |
+
"step": 960
|
683 |
+
},
|
684 |
+
{
|
685 |
+
"epoch": 5.418994413407821,
|
686 |
+
"grad_norm": 0.20398513972759247,
|
687 |
+
"learning_rate": 0.00014313754497755687,
|
688 |
+
"loss": 0.0394,
|
689 |
+
"step": 970
|
690 |
+
},
|
691 |
+
{
|
692 |
+
"epoch": 5.4748603351955305,
|
693 |
+
"grad_norm": 0.2890792787075043,
|
694 |
+
"learning_rate": 0.00014294323005448374,
|
695 |
+
"loss": 0.042,
|
696 |
+
"step": 980
|
697 |
+
},
|
698 |
+
{
|
699 |
+
"epoch": 5.5307262569832405,
|
700 |
+
"grad_norm": 0.3031609356403351,
|
701 |
+
"learning_rate": 0.0001427463380157197,
|
702 |
+
"loss": 0.0429,
|
703 |
+
"step": 990
|
704 |
+
},
|
705 |
+
{
|
706 |
+
"epoch": 5.58659217877095,
|
707 |
+
"grad_norm": 0.5294051766395569,
|
708 |
+
"learning_rate": 0.0001425468763294638,
|
709 |
+
"loss": 0.0424,
|
710 |
+
"step": 1000
|
711 |
+
},
|
712 |
+
{
|
713 |
+
"epoch": 5.64245810055866,
|
714 |
+
"grad_norm": 0.31394124031066895,
|
715 |
+
"learning_rate": 0.00014234485256138277,
|
716 |
+
"loss": 0.0455,
|
717 |
+
"step": 1010
|
718 |
+
},
|
719 |
+
{
|
720 |
+
"epoch": 5.698324022346369,
|
721 |
+
"grad_norm": 0.5516133904457092,
|
722 |
+
"learning_rate": 0.00014214027437432439,
|
723 |
+
"loss": 0.0375,
|
724 |
+
"step": 1020
|
725 |
+
},
|
726 |
+
{
|
727 |
+
"epoch": 5.754189944134078,
|
728 |
+
"grad_norm": 0.45591840147972107,
|
729 |
+
"learning_rate": 0.00014193314952802645,
|
730 |
+
"loss": 0.0361,
|
731 |
+
"step": 1030
|
732 |
+
},
|
733 |
+
{
|
734 |
+
"epoch": 5.810055865921788,
|
735 |
+
"grad_norm": 0.41406169533729553,
|
736 |
+
"learning_rate": 0.00014172348587882276,
|
737 |
+
"loss": 0.0347,
|
738 |
+
"step": 1040
|
739 |
+
},
|
740 |
+
{
|
741 |
+
"epoch": 5.865921787709497,
|
742 |
+
"grad_norm": 0.3035840094089508,
|
743 |
+
"learning_rate": 0.00014151129137934492,
|
744 |
+
"loss": 0.0354,
|
745 |
+
"step": 1050
|
746 |
+
},
|
747 |
+
{
|
748 |
+
"epoch": 5.921787709497207,
|
749 |
+
"grad_norm": 0.3065755367279053,
|
750 |
+
"learning_rate": 0.00014129657407822082,
|
751 |
+
"loss": 0.0364,
|
752 |
+
"step": 1060
|
753 |
+
},
|
754 |
+
{
|
755 |
+
"epoch": 5.977653631284916,
|
756 |
+
"grad_norm": 0.3852936029434204,
|
757 |
+
"learning_rate": 0.0001410793421197692,
|
758 |
+
"loss": 0.0352,
|
759 |
+
"step": 1070
|
760 |
+
},
|
761 |
+
{
|
762 |
+
"epoch": 6.033519553072626,
|
763 |
+
"grad_norm": 0.3598945140838623,
|
764 |
+
"learning_rate": 0.00014085960374369096,
|
765 |
+
"loss": 0.0404,
|
766 |
+
"step": 1080
|
767 |
+
},
|
768 |
+
{
|
769 |
+
"epoch": 6.089385474860335,
|
770 |
+
"grad_norm": 0.3946482539176941,
|
771 |
+
"learning_rate": 0.00014063736728475634,
|
772 |
+
"loss": 0.0312,
|
773 |
+
"step": 1090
|
774 |
+
},
|
775 |
+
{
|
776 |
+
"epoch": 6.145251396648045,
|
777 |
+
"grad_norm": 0.4066295027732849,
|
778 |
+
"learning_rate": 0.00014041264117248907,
|
779 |
+
"loss": 0.041,
|
780 |
+
"step": 1100
|
781 |
+
},
|
782 |
+
{
|
783 |
+
"epoch": 6.201117318435754,
|
784 |
+
"grad_norm": 0.2542637586593628,
|
785 |
+
"learning_rate": 0.00014018543393084633,
|
786 |
+
"loss": 0.0366,
|
787 |
+
"step": 1110
|
788 |
+
},
|
789 |
+
{
|
790 |
+
"epoch": 6.256983240223463,
|
791 |
+
"grad_norm": 0.34908485412597656,
|
792 |
+
"learning_rate": 0.0001399557541778958,
|
793 |
+
"loss": 0.0358,
|
794 |
+
"step": 1120
|
795 |
+
},
|
796 |
+
{
|
797 |
+
"epoch": 6.312849162011173,
|
798 |
+
"grad_norm": 0.39812198281288147,
|
799 |
+
"learning_rate": 0.00013972361062548837,
|
800 |
+
"loss": 0.0405,
|
801 |
+
"step": 1130
|
802 |
+
},
|
803 |
+
{
|
804 |
+
"epoch": 6.368715083798882,
|
805 |
+
"grad_norm": 0.45712536573410034,
|
806 |
+
"learning_rate": 0.00013948901207892807,
|
807 |
+
"loss": 0.037,
|
808 |
+
"step": 1140
|
809 |
+
},
|
810 |
+
{
|
811 |
+
"epoch": 6.424581005586592,
|
812 |
+
"grad_norm": 0.5260054469108582,
|
813 |
+
"learning_rate": 0.0001392519674366377,
|
814 |
+
"loss": 0.0381,
|
815 |
+
"step": 1150
|
816 |
+
},
|
817 |
+
{
|
818 |
+
"epoch": 6.4804469273743015,
|
819 |
+
"grad_norm": 0.39201897382736206,
|
820 |
+
"learning_rate": 0.00013901248568982177,
|
821 |
+
"loss": 0.0366,
|
822 |
+
"step": 1160
|
823 |
+
},
|
824 |
+
{
|
825 |
+
"epoch": 6.5363128491620115,
|
826 |
+
"grad_norm": 0.22905181348323822,
|
827 |
+
"learning_rate": 0.00013877057592212496,
|
828 |
+
"loss": 0.0397,
|
829 |
+
"step": 1170
|
830 |
+
},
|
831 |
+
{
|
832 |
+
"epoch": 6.592178770949721,
|
833 |
+
"grad_norm": 0.4780491590499878,
|
834 |
+
"learning_rate": 0.00013852624730928794,
|
835 |
+
"loss": 0.0414,
|
836 |
+
"step": 1180
|
837 |
+
},
|
838 |
+
{
|
839 |
+
"epoch": 6.648044692737431,
|
840 |
+
"grad_norm": 0.3810228705406189,
|
841 |
+
"learning_rate": 0.00013827950911879922,
|
842 |
+
"loss": 0.035,
|
843 |
+
"step": 1190
|
844 |
+
},
|
845 |
+
{
|
846 |
+
"epoch": 6.70391061452514,
|
847 |
+
"grad_norm": 0.3878680467605591,
|
848 |
+
"learning_rate": 0.0001380303707095436,
|
849 |
+
"loss": 0.0365,
|
850 |
+
"step": 1200
|
851 |
+
},
|
852 |
+
{
|
853 |
+
"epoch": 6.759776536312849,
|
854 |
+
"grad_norm": 0.3587484657764435,
|
855 |
+
"learning_rate": 0.00013777884153144714,
|
856 |
+
"loss": 0.0349,
|
857 |
+
"step": 1210
|
858 |
+
},
|
859 |
+
{
|
860 |
+
"epoch": 6.815642458100559,
|
861 |
+
"grad_norm": 0.3499602675437927,
|
862 |
+
"learning_rate": 0.00013752493112511888,
|
863 |
+
"loss": 0.0418,
|
864 |
+
"step": 1220
|
865 |
+
},
|
866 |
+
{
|
867 |
+
"epoch": 6.871508379888268,
|
868 |
+
"grad_norm": 0.3068997859954834,
|
869 |
+
"learning_rate": 0.00013726864912148878,
|
870 |
+
"loss": 0.0364,
|
871 |
+
"step": 1230
|
872 |
+
},
|
873 |
+
{
|
874 |
+
"epoch": 6.927374301675978,
|
875 |
+
"grad_norm": 0.24322666227817535,
|
876 |
+
"learning_rate": 0.00013701000524144252,
|
877 |
+
"loss": 0.0413,
|
878 |
+
"step": 1240
|
879 |
+
},
|
880 |
+
{
|
881 |
+
"epoch": 6.983240223463687,
|
882 |
+
"grad_norm": 0.27041393518447876,
|
883 |
+
"learning_rate": 0.00013674900929545284,
|
884 |
+
"loss": 0.0362,
|
885 |
+
"step": 1250
|
886 |
+
},
|
887 |
+
{
|
888 |
+
"epoch": 7.039106145251397,
|
889 |
+
"grad_norm": 0.4791170358657837,
|
890 |
+
"learning_rate": 0.00013648567118320723,
|
891 |
+
"loss": 0.0364,
|
892 |
+
"step": 1260
|
893 |
+
},
|
894 |
+
{
|
895 |
+
"epoch": 7.094972067039106,
|
896 |
+
"grad_norm": 0.39187201857566833,
|
897 |
+
"learning_rate": 0.00013622000089323256,
|
898 |
+
"loss": 0.0406,
|
899 |
+
"step": 1270
|
900 |
+
},
|
901 |
+
{
|
902 |
+
"epoch": 7.150837988826815,
|
903 |
+
"grad_norm": 0.3253427743911743,
|
904 |
+
"learning_rate": 0.00013595200850251627,
|
905 |
+
"loss": 0.0337,
|
906 |
+
"step": 1280
|
907 |
+
},
|
908 |
+
{
|
909 |
+
"epoch": 7.206703910614525,
|
910 |
+
"grad_norm": 0.2595740854740143,
|
911 |
+
"learning_rate": 0.000135681704176124,
|
912 |
+
"loss": 0.0367,
|
913 |
+
"step": 1290
|
914 |
+
},
|
915 |
+
{
|
916 |
+
"epoch": 7.262569832402234,
|
917 |
+
"grad_norm": 0.4414283335208893,
|
918 |
+
"learning_rate": 0.00013540909816681416,
|
919 |
+
"loss": 0.0391,
|
920 |
+
"step": 1300
|
921 |
+
},
|
922 |
+
{
|
923 |
+
"epoch": 7.318435754189944,
|
924 |
+
"grad_norm": 0.329851895570755,
|
925 |
+
"learning_rate": 0.0001351342008146489,
|
926 |
+
"loss": 0.0306,
|
927 |
+
"step": 1310
|
928 |
+
},
|
929 |
+
{
|
930 |
+
"epoch": 7.374301675977653,
|
931 |
+
"grad_norm": 0.3009398877620697,
|
932 |
+
"learning_rate": 0.000134857022546602,
|
933 |
+
"loss": 0.032,
|
934 |
+
"step": 1320
|
935 |
+
},
|
936 |
+
{
|
937 |
+
"epoch": 7.430167597765363,
|
938 |
+
"grad_norm": 0.670881986618042,
|
939 |
+
"learning_rate": 0.00013457757387616348,
|
940 |
+
"loss": 0.04,
|
941 |
+
"step": 1330
|
942 |
+
},
|
943 |
+
{
|
944 |
+
"epoch": 7.4860335195530725,
|
945 |
+
"grad_norm": 0.38917240500450134,
|
946 |
+
"learning_rate": 0.00013429586540294049,
|
947 |
+
"loss": 0.0337,
|
948 |
+
"step": 1340
|
949 |
+
},
|
950 |
+
{
|
951 |
+
"epoch": 7.5418994413407825,
|
952 |
+
"grad_norm": 0.3102369010448456,
|
953 |
+
"learning_rate": 0.00013401190781225553,
|
954 |
+
"loss": 0.031,
|
955 |
+
"step": 1350
|
956 |
+
},
|
957 |
+
{
|
958 |
+
"epoch": 7.597765363128492,
|
959 |
+
"grad_norm": 0.37413033843040466,
|
960 |
+
"learning_rate": 0.00013372571187474121,
|
961 |
+
"loss": 0.0334,
|
962 |
+
"step": 1360
|
963 |
+
},
|
964 |
+
{
|
965 |
+
"epoch": 7.653631284916202,
|
966 |
+
"grad_norm": 0.28566494584083557,
|
967 |
+
"learning_rate": 0.0001334372884459314,
|
968 |
+
"loss": 0.0284,
|
969 |
+
"step": 1370
|
970 |
+
},
|
971 |
+
{
|
972 |
+
"epoch": 7.709497206703911,
|
973 |
+
"grad_norm": 0.35984960198402405,
|
974 |
+
"learning_rate": 0.0001331466484658498,
|
975 |
+
"loss": 0.028,
|
976 |
+
"step": 1380
|
977 |
+
},
|
978 |
+
{
|
979 |
+
"epoch": 7.76536312849162,
|
980 |
+
"grad_norm": 0.41864386200904846,
|
981 |
+
"learning_rate": 0.00013285380295859482,
|
982 |
+
"loss": 0.0316,
|
983 |
+
"step": 1390
|
984 |
+
},
|
985 |
+
{
|
986 |
+
"epoch": 7.82122905027933,
|
987 |
+
"grad_norm": 0.44935837388038635,
|
988 |
+
"learning_rate": 0.00013255876303192138,
|
989 |
+
"loss": 0.0341,
|
990 |
+
"step": 1400
|
991 |
+
},
|
992 |
+
{
|
993 |
+
"epoch": 7.877094972067039,
|
994 |
+
"grad_norm": 0.2935531437397003,
|
995 |
+
"learning_rate": 0.00013226153987681974,
|
996 |
+
"loss": 0.035,
|
997 |
+
"step": 1410
|
998 |
+
},
|
999 |
+
{
|
1000 |
+
"epoch": 7.932960893854749,
|
1001 |
+
"grad_norm": 0.5159788131713867,
|
1002 |
+
"learning_rate": 0.00013196214476709096,
|
1003 |
+
"loss": 0.0303,
|
1004 |
+
"step": 1420
|
1005 |
+
},
|
1006 |
+
{
|
1007 |
+
"epoch": 7.988826815642458,
|
1008 |
+
"grad_norm": 0.3987308144569397,
|
1009 |
+
"learning_rate": 0.00013166058905891918,
|
1010 |
+
"loss": 0.0342,
|
1011 |
+
"step": 1430
|
1012 |
+
},
|
1013 |
+
{
|
1014 |
+
"epoch": 8.044692737430168,
|
1015 |
+
"grad_norm": 0.35137230157852173,
|
1016 |
+
"learning_rate": 0.00013135688419044109,
|
1017 |
+
"loss": 0.0298,
|
1018 |
+
"step": 1440
|
1019 |
+
},
|
1020 |
+
{
|
1021 |
+
"epoch": 8.100558659217878,
|
1022 |
+
"grad_norm": 0.2903079092502594,
|
1023 |
+
"learning_rate": 0.00013105104168131178,
|
1024 |
+
"loss": 0.0293,
|
1025 |
+
"step": 1450
|
1026 |
+
},
|
1027 |
+
{
|
1028 |
+
"epoch": 8.156424581005586,
|
1029 |
+
"grad_norm": 0.26529935002326965,
|
1030 |
+
"learning_rate": 0.00013074307313226807,
|
1031 |
+
"loss": 0.0327,
|
1032 |
+
"step": 1460
|
1033 |
+
},
|
1034 |
+
{
|
1035 |
+
"epoch": 8.212290502793296,
|
1036 |
+
"grad_norm": 0.46690046787261963,
|
1037 |
+
"learning_rate": 0.0001304329902246884,
|
1038 |
+
"loss": 0.0336,
|
1039 |
+
"step": 1470
|
1040 |
+
},
|
1041 |
+
{
|
1042 |
+
"epoch": 8.268156424581006,
|
1043 |
+
"grad_norm": 0.37878528237342834,
|
1044 |
+
"learning_rate": 0.00013012080472014968,
|
1045 |
+
"loss": 0.0304,
|
1046 |
+
"step": 1480
|
1047 |
+
},
|
1048 |
+
{
|
1049 |
+
"epoch": 8.324022346368714,
|
1050 |
+
"grad_norm": 0.2849811315536499,
|
1051 |
+
"learning_rate": 0.00012980652845998123,
|
1052 |
+
"loss": 0.032,
|
1053 |
+
"step": 1490
|
1054 |
+
},
|
1055 |
+
{
|
1056 |
+
"epoch": 8.379888268156424,
|
1057 |
+
"grad_norm": 0.2862546741962433,
|
1058 |
+
"learning_rate": 0.0001294901733648156,
|
1059 |
+
"loss": 0.0286,
|
1060 |
+
"step": 1500
|
1061 |
+
},
|
1062 |
+
{
|
1063 |
+
"epoch": 8.435754189944134,
|
1064 |
+
"grad_norm": 0.21827903389930725,
|
1065 |
+
"learning_rate": 0.00012917175143413646,
|
1066 |
+
"loss": 0.0373,
|
1067 |
+
"step": 1510
|
1068 |
+
},
|
1069 |
+
{
|
1070 |
+
"epoch": 8.491620111731844,
|
1071 |
+
"grad_norm": 0.16881871223449707,
|
1072 |
+
"learning_rate": 0.00012885127474582348,
|
1073 |
+
"loss": 0.0334,
|
1074 |
+
"step": 1520
|
1075 |
+
},
|
1076 |
+
{
|
1077 |
+
"epoch": 8.547486033519553,
|
1078 |
+
"grad_norm": 0.5263266563415527,
|
1079 |
+
"learning_rate": 0.00012852875545569408,
|
1080 |
+
"loss": 0.037,
|
1081 |
+
"step": 1530
|
1082 |
+
},
|
1083 |
+
{
|
1084 |
+
"epoch": 8.603351955307263,
|
1085 |
+
"grad_norm": 0.41702142357826233,
|
1086 |
+
"learning_rate": 0.00012820420579704245,
|
1087 |
+
"loss": 0.0398,
|
1088 |
+
"step": 1540
|
1089 |
+
},
|
1090 |
+
{
|
1091 |
+
"epoch": 8.659217877094973,
|
1092 |
+
"grad_norm": 0.20871484279632568,
|
1093 |
+
"learning_rate": 0.00012787763808017565,
|
1094 |
+
"loss": 0.0321,
|
1095 |
+
"step": 1550
|
1096 |
+
},
|
1097 |
+
{
|
1098 |
+
"epoch": 8.71508379888268,
|
1099 |
+
"grad_norm": 0.4315951466560364,
|
1100 |
+
"learning_rate": 0.00012754906469194632,
|
1101 |
+
"loss": 0.0316,
|
1102 |
+
"step": 1560
|
1103 |
+
},
|
1104 |
+
{
|
1105 |
+
"epoch": 8.77094972067039,
|
1106 |
+
"grad_norm": 0.2898998558521271,
|
1107 |
+
"learning_rate": 0.0001272184980952833,
|
1108 |
+
"loss": 0.0314,
|
1109 |
+
"step": 1570
|
1110 |
+
},
|
1111 |
+
{
|
1112 |
+
"epoch": 8.8268156424581,
|
1113 |
+
"grad_norm": 0.4114074409008026,
|
1114 |
+
"learning_rate": 0.00012688595082871852,
|
1115 |
+
"loss": 0.0316,
|
1116 |
+
"step": 1580
|
1117 |
+
},
|
1118 |
+
{
|
1119 |
+
"epoch": 8.88268156424581,
|
1120 |
+
"grad_norm": 0.3064340353012085,
|
1121 |
+
"learning_rate": 0.00012655143550591164,
|
1122 |
+
"loss": 0.0335,
|
1123 |
+
"step": 1590
|
1124 |
+
},
|
1125 |
+
{
|
1126 |
+
"epoch": 8.938547486033519,
|
1127 |
+
"grad_norm": 0.2396443486213684,
|
1128 |
+
"learning_rate": 0.0001262149648151715,
|
1129 |
+
"loss": 0.0292,
|
1130 |
+
"step": 1600
|
1131 |
+
},
|
1132 |
+
{
|
1133 |
+
"epoch": 8.994413407821229,
|
1134 |
+
"grad_norm": 0.280146986246109,
|
1135 |
+
"learning_rate": 0.00012587655151897488,
|
1136 |
+
"loss": 0.031,
|
1137 |
+
"step": 1610
|
1138 |
+
},
|
1139 |
+
{
|
1140 |
+
"epoch": 9.050279329608939,
|
1141 |
+
"grad_norm": 0.2552490234375,
|
1142 |
+
"learning_rate": 0.00012553620845348246,
|
1143 |
+
"loss": 0.0249,
|
1144 |
+
"step": 1620
|
1145 |
+
},
|
1146 |
+
{
|
1147 |
+
"epoch": 9.106145251396647,
|
1148 |
+
"grad_norm": 0.1910940706729889,
|
1149 |
+
"learning_rate": 0.00012519394852805182,
|
1150 |
+
"loss": 0.0297,
|
1151 |
+
"step": 1630
|
1152 |
+
},
|
1153 |
+
{
|
1154 |
+
"epoch": 9.162011173184357,
|
1155 |
+
"grad_norm": 0.2914765477180481,
|
1156 |
+
"learning_rate": 0.0001248497847247479,
|
1157 |
+
"loss": 0.0321,
|
1158 |
+
"step": 1640
|
1159 |
+
},
|
1160 |
+
{
|
1161 |
+
"epoch": 9.217877094972067,
|
1162 |
+
"grad_norm": 0.34026339650154114,
|
1163 |
+
"learning_rate": 0.00012450373009785054,
|
1164 |
+
"loss": 0.0272,
|
1165 |
+
"step": 1650
|
1166 |
+
},
|
1167 |
+
{
|
1168 |
+
"epoch": 9.273743016759777,
|
1169 |
+
"grad_norm": 0.21475638449192047,
|
1170 |
+
"learning_rate": 0.00012415579777335932,
|
1171 |
+
"loss": 0.026,
|
1172 |
+
"step": 1660
|
1173 |
+
},
|
1174 |
+
{
|
1175 |
+
"epoch": 9.329608938547485,
|
1176 |
+
"grad_norm": 0.30112239718437195,
|
1177 |
+
"learning_rate": 0.00012380600094849566,
|
1178 |
+
"loss": 0.0374,
|
1179 |
+
"step": 1670
|
1180 |
+
},
|
1181 |
+
{
|
1182 |
+
"epoch": 9.385474860335195,
|
1183 |
+
"grad_norm": 0.17926353216171265,
|
1184 |
+
"learning_rate": 0.00012345435289120234,
|
1185 |
+
"loss": 0.026,
|
1186 |
+
"step": 1680
|
1187 |
+
},
|
1188 |
+
{
|
1189 |
+
"epoch": 9.441340782122905,
|
1190 |
+
"grad_norm": 0.16974708437919617,
|
1191 |
+
"learning_rate": 0.00012310086693964012,
|
1192 |
+
"loss": 0.0308,
|
1193 |
+
"step": 1690
|
1194 |
+
},
|
1195 |
+
{
|
1196 |
+
"epoch": 9.497206703910614,
|
1197 |
+
"grad_norm": 0.3298158347606659,
|
1198 |
+
"learning_rate": 0.00012274555650168187,
|
1199 |
+
"loss": 0.03,
|
1200 |
+
"step": 1700
|
1201 |
+
},
|
1202 |
+
{
|
1203 |
+
"epoch": 9.553072625698324,
|
1204 |
+
"grad_norm": 0.33263733983039856,
|
1205 |
+
"learning_rate": 0.000122388435054404,
|
1206 |
+
"loss": 0.03,
|
1207 |
+
"step": 1710
|
1208 |
+
},
|
1209 |
+
{
|
1210 |
+
"epoch": 9.608938547486034,
|
1211 |
+
"grad_norm": 0.5304810404777527,
|
1212 |
+
"learning_rate": 0.00012202951614357541,
|
1213 |
+
"loss": 0.028,
|
1214 |
+
"step": 1720
|
1215 |
+
},
|
1216 |
+
{
|
1217 |
+
"epoch": 9.664804469273744,
|
1218 |
+
"grad_norm": 0.2756952941417694,
|
1219 |
+
"learning_rate": 0.00012166881338314337,
|
1220 |
+
"loss": 0.0271,
|
1221 |
+
"step": 1730
|
1222 |
+
},
|
1223 |
+
{
|
1224 |
+
"epoch": 9.720670391061452,
|
1225 |
+
"grad_norm": 0.329500675201416,
|
1226 |
+
"learning_rate": 0.0001213063404547174,
|
1227 |
+
"loss": 0.0392,
|
1228 |
+
"step": 1740
|
1229 |
+
},
|
1230 |
+
{
|
1231 |
+
"epoch": 9.776536312849162,
|
1232 |
+
"grad_norm": 0.20280927419662476,
|
1233 |
+
"learning_rate": 0.0001209421111070503,
|
1234 |
+
"loss": 0.0293,
|
1235 |
+
"step": 1750
|
1236 |
+
},
|
1237 |
+
{
|
1238 |
+
"epoch": 9.832402234636872,
|
1239 |
+
"grad_norm": 0.26771217584609985,
|
1240 |
+
"learning_rate": 0.00012057613915551652,
|
1241 |
+
"loss": 0.0283,
|
1242 |
+
"step": 1760
|
1243 |
+
},
|
1244 |
+
{
|
1245 |
+
"epoch": 9.888268156424582,
|
1246 |
+
"grad_norm": 0.2704836428165436,
|
1247 |
+
"learning_rate": 0.00012020843848158826,
|
1248 |
+
"loss": 0.0295,
|
1249 |
+
"step": 1770
|
1250 |
+
},
|
1251 |
+
{
|
1252 |
+
"epoch": 9.94413407821229,
|
1253 |
+
"grad_norm": 0.34217214584350586,
|
1254 |
+
"learning_rate": 0.00011983902303230892,
|
1255 |
+
"loss": 0.0303,
|
1256 |
+
"step": 1780
|
1257 |
+
},
|
1258 |
+
{
|
1259 |
+
"epoch": 10.0,
|
1260 |
+
"grad_norm": 0.27921390533447266,
|
1261 |
+
"learning_rate": 0.00011946790681976399,
|
1262 |
+
"loss": 0.0292,
|
1263 |
+
"step": 1790
|
1264 |
+
},
|
1265 |
+
{
|
1266 |
+
"epoch": 10.05586592178771,
|
1267 |
+
"grad_norm": 0.2127024084329605,
|
1268 |
+
"learning_rate": 0.0001190951039205497,
|
1269 |
+
"loss": 0.0256,
|
1270 |
+
"step": 1800
|
1271 |
+
},
|
1272 |
+
{
|
1273 |
+
"epoch": 10.111731843575418,
|
1274 |
+
"grad_norm": 0.2055603712797165,
|
1275 |
+
"learning_rate": 0.00011872062847523899,
|
1276 |
+
"loss": 0.028,
|
1277 |
+
"step": 1810
|
1278 |
+
},
|
1279 |
+
{
|
1280 |
+
"epoch": 10.167597765363128,
|
1281 |
+
"grad_norm": 0.27974414825439453,
|
1282 |
+
"learning_rate": 0.00011834449468784523,
|
1283 |
+
"loss": 0.03,
|
1284 |
+
"step": 1820
|
1285 |
+
},
|
1286 |
+
{
|
1287 |
+
"epoch": 10.223463687150838,
|
1288 |
+
"grad_norm": 0.3269798159599304,
|
1289 |
+
"learning_rate": 0.00011796671682528334,
|
1290 |
+
"loss": 0.0287,
|
1291 |
+
"step": 1830
|
1292 |
+
},
|
1293 |
+
{
|
1294 |
+
"epoch": 10.279329608938548,
|
1295 |
+
"grad_norm": 0.3258807063102722,
|
1296 |
+
"learning_rate": 0.00011758730921682876,
|
1297 |
+
"loss": 0.0284,
|
1298 |
+
"step": 1840
|
1299 |
+
},
|
1300 |
+
{
|
1301 |
+
"epoch": 10.335195530726256,
|
1302 |
+
"grad_norm": 0.21946227550506592,
|
1303 |
+
"learning_rate": 0.00011720628625357392,
|
1304 |
+
"loss": 0.0306,
|
1305 |
+
"step": 1850
|
1306 |
+
},
|
1307 |
+
{
|
1308 |
+
"epoch": 10.391061452513966,
|
1309 |
+
"grad_norm": 0.3200410008430481,
|
1310 |
+
"learning_rate": 0.00011682366238788225,
|
1311 |
+
"loss": 0.0283,
|
1312 |
+
"step": 1860
|
1313 |
+
},
|
1314 |
+
{
|
1315 |
+
"epoch": 10.446927374301676,
|
1316 |
+
"grad_norm": 0.32745105028152466,
|
1317 |
+
"learning_rate": 0.0001164394521328402,
|
1318 |
+
"loss": 0.0269,
|
1319 |
+
"step": 1870
|
1320 |
+
},
|
1321 |
+
{
|
1322 |
+
"epoch": 10.502793296089386,
|
1323 |
+
"grad_norm": 0.30036798119544983,
|
1324 |
+
"learning_rate": 0.00011605367006170654,
|
1325 |
+
"loss": 0.0267,
|
1326 |
+
"step": 1880
|
1327 |
+
},
|
1328 |
+
{
|
1329 |
+
"epoch": 10.558659217877095,
|
1330 |
+
"grad_norm": 0.3197910785675049,
|
1331 |
+
"learning_rate": 0.0001156663308073598,
|
1332 |
+
"loss": 0.0257,
|
1333 |
+
"step": 1890
|
1334 |
+
},
|
1335 |
+
{
|
1336 |
+
"epoch": 10.614525139664805,
|
1337 |
+
"grad_norm": 0.23818261921405792,
|
1338 |
+
"learning_rate": 0.00011527744906174306,
|
1339 |
+
"loss": 0.0309,
|
1340 |
+
"step": 1900
|
1341 |
+
},
|
1342 |
+
{
|
1343 |
+
"epoch": 10.670391061452515,
|
1344 |
+
"grad_norm": 0.1711827963590622,
|
1345 |
+
"learning_rate": 0.00011488703957530675,
|
1346 |
+
"loss": 0.0275,
|
1347 |
+
"step": 1910
|
1348 |
+
},
|
1349 |
+
{
|
1350 |
+
"epoch": 10.726256983240223,
|
1351 |
+
"grad_norm": 0.3193713426589966,
|
1352 |
+
"learning_rate": 0.00011449511715644922,
|
1353 |
+
"loss": 0.0286,
|
1354 |
+
"step": 1920
|
1355 |
+
},
|
1356 |
+
{
|
1357 |
+
"epoch": 10.782122905027933,
|
1358 |
+
"grad_norm": 0.3227252662181854,
|
1359 |
+
"learning_rate": 0.00011410169667095494,
|
1360 |
+
"loss": 0.0274,
|
1361 |
+
"step": 1930
|
1362 |
+
},
|
1363 |
+
{
|
1364 |
+
"epoch": 10.837988826815643,
|
1365 |
+
"grad_norm": 0.27391719818115234,
|
1366 |
+
"learning_rate": 0.0001137067930414307,
|
1367 |
+
"loss": 0.0314,
|
1368 |
+
"step": 1940
|
1369 |
+
},
|
1370 |
+
{
|
1371 |
+
"epoch": 10.893854748603353,
|
1372 |
+
"grad_norm": 0.407885879278183,
|
1373 |
+
"learning_rate": 0.0001133104212467396,
|
1374 |
+
"loss": 0.0272,
|
1375 |
+
"step": 1950
|
1376 |
+
},
|
1377 |
+
{
|
1378 |
+
"epoch": 10.949720670391061,
|
1379 |
+
"grad_norm": 0.2611219584941864,
|
1380 |
+
"learning_rate": 0.0001129125963214328,
|
1381 |
+
"loss": 0.0266,
|
1382 |
+
"step": 1960
|
1383 |
+
},
|
1384 |
+
{
|
1385 |
+
"epoch": 11.005586592178771,
|
1386 |
+
"grad_norm": 0.335827499628067,
|
1387 |
+
"learning_rate": 0.00011251333335517944,
|
1388 |
+
"loss": 0.0273,
|
1389 |
+
"step": 1970
|
1390 |
+
},
|
1391 |
+
{
|
1392 |
+
"epoch": 11.061452513966481,
|
1393 |
+
"grad_norm": 0.24541901051998138,
|
1394 |
+
"learning_rate": 0.00011211264749219403,
|
1395 |
+
"loss": 0.0295,
|
1396 |
+
"step": 1980
|
1397 |
+
},
|
1398 |
+
{
|
1399 |
+
"epoch": 11.11731843575419,
|
1400 |
+
"grad_norm": 0.27788233757019043,
|
1401 |
+
"learning_rate": 0.00011171055393066224,
|
1402 |
+
"loss": 0.0254,
|
1403 |
+
"step": 1990
|
1404 |
+
},
|
1405 |
+
{
|
1406 |
+
"epoch": 11.1731843575419,
|
1407 |
+
"grad_norm": 0.23598261177539825,
|
1408 |
+
"learning_rate": 0.00011130706792216439,
|
1409 |
+
"loss": 0.0298,
|
1410 |
+
"step": 2000
|
1411 |
+
},
|
1412 |
+
{
|
1413 |
+
"epoch": 11.22905027932961,
|
1414 |
+
"grad_norm": 0.20312641561031342,
|
1415 |
+
"learning_rate": 0.0001109022047710968,
|
1416 |
+
"loss": 0.0278,
|
1417 |
+
"step": 2010
|
1418 |
+
},
|
1419 |
+
{
|
1420 |
+
"epoch": 11.28491620111732,
|
1421 |
+
"grad_norm": 0.2139887809753418,
|
1422 |
+
"learning_rate": 0.00011049597983409147,
|
1423 |
+
"loss": 0.0271,
|
1424 |
+
"step": 2020
|
1425 |
+
},
|
1426 |
+
{
|
1427 |
+
"epoch": 11.340782122905027,
|
1428 |
+
"grad_norm": 0.22644950449466705,
|
1429 |
+
"learning_rate": 0.00011008840851943348,
|
1430 |
+
"loss": 0.0298,
|
1431 |
+
"step": 2030
|
1432 |
+
},
|
1433 |
+
{
|
1434 |
+
"epoch": 11.396648044692737,
|
1435 |
+
"grad_norm": 0.20427192747592926,
|
1436 |
+
"learning_rate": 0.00010967950628647664,
|
1437 |
+
"loss": 0.03,
|
1438 |
+
"step": 2040
|
1439 |
+
},
|
1440 |
+
{
|
1441 |
+
"epoch": 11.452513966480447,
|
1442 |
+
"grad_norm": 0.4383834898471832,
|
1443 |
+
"learning_rate": 0.00010926928864505698,
|
1444 |
+
"loss": 0.0303,
|
1445 |
+
"step": 2050
|
1446 |
+
},
|
1447 |
+
{
|
1448 |
+
"epoch": 11.508379888268156,
|
1449 |
+
"grad_norm": 0.2548943758010864,
|
1450 |
+
"learning_rate": 0.00010885777115490463,
|
1451 |
+
"loss": 0.0288,
|
1452 |
+
"step": 2060
|
1453 |
+
},
|
1454 |
+
{
|
1455 |
+
"epoch": 11.564245810055866,
|
1456 |
+
"grad_norm": 0.2970632016658783,
|
1457 |
+
"learning_rate": 0.00010844496942505342,
|
1458 |
+
"loss": 0.0287,
|
1459 |
+
"step": 2070
|
1460 |
+
},
|
1461 |
+
{
|
1462 |
+
"epoch": 11.620111731843576,
|
1463 |
+
"grad_norm": 0.28128278255462646,
|
1464 |
+
"learning_rate": 0.00010803089911324907,
|
1465 |
+
"loss": 0.0298,
|
1466 |
+
"step": 2080
|
1467 |
+
},
|
1468 |
+
{
|
1469 |
+
"epoch": 11.675977653631286,
|
1470 |
+
"grad_norm": 0.2895059287548065,
|
1471 |
+
"learning_rate": 0.00010761557592535509,
|
1472 |
+
"loss": 0.0312,
|
1473 |
+
"step": 2090
|
1474 |
+
},
|
1475 |
+
{
|
1476 |
+
"epoch": 11.731843575418994,
|
1477 |
+
"grad_norm": 0.3408230245113373,
|
1478 |
+
"learning_rate": 0.00010719901561475706,
|
1479 |
+
"loss": 0.031,
|
1480 |
+
"step": 2100
|
1481 |
+
},
|
1482 |
+
{
|
1483 |
+
"epoch": 11.787709497206704,
|
1484 |
+
"grad_norm": 0.2545909881591797,
|
1485 |
+
"learning_rate": 0.00010678123398176526,
|
1486 |
+
"loss": 0.0274,
|
1487 |
+
"step": 2110
|
1488 |
+
},
|
1489 |
+
{
|
1490 |
+
"epoch": 11.843575418994414,
|
1491 |
+
"grad_norm": 0.2746656537055969,
|
1492 |
+
"learning_rate": 0.00010636224687301515,
|
1493 |
+
"loss": 0.0214,
|
1494 |
+
"step": 2120
|
1495 |
+
},
|
1496 |
+
{
|
1497 |
+
"epoch": 11.899441340782122,
|
1498 |
+
"grad_norm": 0.35347574949264526,
|
1499 |
+
"learning_rate": 0.00010594207018086647,
|
1500 |
+
"loss": 0.0306,
|
1501 |
+
"step": 2130
|
1502 |
+
},
|
1503 |
+
{
|
1504 |
+
"epoch": 11.955307262569832,
|
1505 |
+
"grad_norm": 0.2086307555437088,
|
1506 |
+
"learning_rate": 0.00010552071984280035,
|
1507 |
+
"loss": 0.0253,
|
1508 |
+
"step": 2140
|
1509 |
+
},
|
1510 |
+
{
|
1511 |
+
"epoch": 12.011173184357542,
|
1512 |
+
"grad_norm": 0.2846223711967468,
|
1513 |
+
"learning_rate": 0.00010509821184081479,
|
1514 |
+
"loss": 0.0267,
|
1515 |
+
"step": 2150
|
1516 |
+
},
|
1517 |
+
{
|
1518 |
+
"epoch": 12.067039106145252,
|
1519 |
+
"grad_norm": 0.26856729388237,
|
1520 |
+
"learning_rate": 0.00010467456220081847,
|
1521 |
+
"loss": 0.0218,
|
1522 |
+
"step": 2160
|
1523 |
+
},
|
1524 |
+
{
|
1525 |
+
"epoch": 12.12290502793296,
|
1526 |
+
"grad_norm": 0.24002467095851898,
|
1527 |
+
"learning_rate": 0.00010424978699202294,
|
1528 |
+
"loss": 0.0268,
|
1529 |
+
"step": 2170
|
1530 |
+
},
|
1531 |
+
{
|
1532 |
+
"epoch": 12.17877094972067,
|
1533 |
+
"grad_norm": 0.37470901012420654,
|
1534 |
+
"learning_rate": 0.00010382390232633298,
|
1535 |
+
"loss": 0.0258,
|
1536 |
+
"step": 2180
|
1537 |
+
},
|
1538 |
+
{
|
1539 |
+
"epoch": 12.23463687150838,
|
1540 |
+
"grad_norm": 0.3723151981830597,
|
1541 |
+
"learning_rate": 0.00010339692435773564,
|
1542 |
+
"loss": 0.0303,
|
1543 |
+
"step": 2190
|
1544 |
+
},
|
1545 |
+
{
|
1546 |
+
"epoch": 12.29050279329609,
|
1547 |
+
"grad_norm": 0.44261157512664795,
|
1548 |
+
"learning_rate": 0.00010296886928168728,
|
1549 |
+
"loss": 0.0302,
|
1550 |
+
"step": 2200
|
1551 |
+
},
|
1552 |
+
{
|
1553 |
+
"epoch": 12.346368715083798,
|
1554 |
+
"grad_norm": 0.23005405068397522,
|
1555 |
+
"learning_rate": 0.00010253975333449953,
|
1556 |
+
"loss": 0.0302,
|
1557 |
+
"step": 2210
|
1558 |
+
},
|
1559 |
+
{
|
1560 |
+
"epoch": 12.402234636871508,
|
1561 |
+
"grad_norm": 0.35475534200668335,
|
1562 |
+
"learning_rate": 0.00010210959279272327,
|
1563 |
+
"loss": 0.0305,
|
1564 |
+
"step": 2220
|
1565 |
+
},
|
1566 |
+
{
|
1567 |
+
"epoch": 12.458100558659218,
|
1568 |
+
"grad_norm": 0.1821463704109192,
|
1569 |
+
"learning_rate": 0.00010167840397253129,
|
1570 |
+
"loss": 0.0293,
|
1571 |
+
"step": 2230
|
1572 |
+
},
|
1573 |
+
{
|
1574 |
+
"epoch": 12.513966480446927,
|
1575 |
+
"grad_norm": 0.3717893958091736,
|
1576 |
+
"learning_rate": 0.00010124620322909939,
|
1577 |
+
"loss": 0.0265,
|
1578 |
+
"step": 2240
|
1579 |
+
},
|
1580 |
+
{
|
1581 |
+
"epoch": 12.569832402234637,
|
1582 |
+
"grad_norm": 0.35059571266174316,
|
1583 |
+
"learning_rate": 0.00010081300695598612,
|
1584 |
+
"loss": 0.0257,
|
1585 |
+
"step": 2250
|
1586 |
+
},
|
1587 |
+
{
|
1588 |
+
"epoch": 12.625698324022347,
|
1589 |
+
"grad_norm": 0.3039356470108032,
|
1590 |
+
"learning_rate": 0.00010037883158451083,
|
1591 |
+
"loss": 0.0288,
|
1592 |
+
"step": 2260
|
1593 |
+
},
|
1594 |
+
{
|
1595 |
+
"epoch": 12.681564245810057,
|
1596 |
+
"grad_norm": 0.49444425106048584,
|
1597 |
+
"learning_rate": 9.994369358313057e-05,
|
1598 |
+
"loss": 0.029,
|
1599 |
+
"step": 2270
|
1600 |
+
},
|
1601 |
+
{
|
1602 |
+
"epoch": 12.737430167597765,
|
1603 |
+
"grad_norm": 0.2746371626853943,
|
1604 |
+
"learning_rate": 9.950760945681525e-05,
|
1605 |
+
"loss": 0.0265,
|
1606 |
+
"step": 2280
|
1607 |
+
},
|
1608 |
+
{
|
1609 |
+
"epoch": 12.793296089385475,
|
1610 |
+
"grad_norm": 0.21077698469161987,
|
1611 |
+
"learning_rate": 9.907059574642177e-05,
|
1612 |
+
"loss": 0.0279,
|
1613 |
+
"step": 2290
|
1614 |
+
},
|
1615 |
+
{
|
1616 |
+
"epoch": 12.849162011173185,
|
1617 |
+
"grad_norm": 0.3581600487232208,
|
1618 |
+
"learning_rate": 9.863266902806654e-05,
|
1619 |
+
"loss": 0.0256,
|
1620 |
+
"step": 2300
|
1621 |
+
},
|
1622 |
+
{
|
1623 |
+
"epoch": 12.905027932960895,
|
1624 |
+
"grad_norm": 0.3539135754108429,
|
1625 |
+
"learning_rate": 9.81938459124967e-05,
|
1626 |
+
"loss": 0.0244,
|
1627 |
+
"step": 2310
|
1628 |
+
},
|
1629 |
+
{
|
1630 |
+
"epoch": 12.960893854748603,
|
1631 |
+
"grad_norm": 0.24212266504764557,
|
1632 |
+
"learning_rate": 9.775414304446024e-05,
|
1633 |
+
"loss": 0.0291,
|
1634 |
+
"step": 2320
|
1635 |
+
},
|
1636 |
+
{
|
1637 |
+
"epoch": 13.016759776536313,
|
1638 |
+
"grad_norm": 0.24756573140621185,
|
1639 |
+
"learning_rate": 9.731357710207442e-05,
|
1640 |
+
"loss": 0.0254,
|
1641 |
+
"step": 2330
|
1642 |
+
},
|
1643 |
+
{
|
1644 |
+
"epoch": 13.072625698324023,
|
1645 |
+
"grad_norm": 0.20405328273773193,
|
1646 |
+
"learning_rate": 9.687216479619328e-05,
|
1647 |
+
"loss": 0.0274,
|
1648 |
+
"step": 2340
|
1649 |
+
},
|
1650 |
+
{
|
1651 |
+
"epoch": 13.128491620111731,
|
1652 |
+
"grad_norm": 0.3089810013771057,
|
1653 |
+
"learning_rate": 9.64299228697739e-05,
|
1654 |
+
"loss": 0.0233,
|
1655 |
+
"step": 2350
|
1656 |
+
},
|
1657 |
+
{
|
1658 |
+
"epoch": 13.184357541899441,
|
1659 |
+
"grad_norm": 0.3249969780445099,
|
1660 |
+
"learning_rate": 9.598686809724109e-05,
|
1661 |
+
"loss": 0.0251,
|
1662 |
+
"step": 2360
|
1663 |
+
},
|
1664 |
+
{
|
1665 |
+
"epoch": 13.240223463687151,
|
1666 |
+
"grad_norm": 0.30293509364128113,
|
1667 |
+
"learning_rate": 9.554301728385133e-05,
|
1668 |
+
"loss": 0.0257,
|
1669 |
+
"step": 2370
|
1670 |
+
},
|
1671 |
+
{
|
1672 |
+
"epoch": 13.296089385474861,
|
1673 |
+
"grad_norm": 0.27823933959007263,
|
1674 |
+
"learning_rate": 9.509838726505527e-05,
|
1675 |
+
"loss": 0.0245,
|
1676 |
+
"step": 2380
|
1677 |
+
},
|
1678 |
+
{
|
1679 |
+
"epoch": 13.35195530726257,
|
1680 |
+
"grad_norm": 0.31522759795188904,
|
1681 |
+
"learning_rate": 9.465299490585914e-05,
|
1682 |
+
"loss": 0.0254,
|
1683 |
+
"step": 2390
|
1684 |
+
},
|
1685 |
+
{
|
1686 |
+
"epoch": 13.40782122905028,
|
1687 |
+
"grad_norm": 0.28887462615966797,
|
1688 |
+
"learning_rate": 9.420685710018505e-05,
|
1689 |
+
"loss": 0.0237,
|
1690 |
+
"step": 2400
|
1691 |
+
},
|
1692 |
+
{
|
1693 |
+
"epoch": 13.46368715083799,
|
1694 |
+
"grad_norm": 0.23696090281009674,
|
1695 |
+
"learning_rate": 9.375999077023024e-05,
|
1696 |
+
"loss": 0.0264,
|
1697 |
+
"step": 2410
|
1698 |
+
},
|
1699 |
+
{
|
1700 |
+
"epoch": 13.519553072625698,
|
1701 |
+
"grad_norm": 0.2132495492696762,
|
1702 |
+
"learning_rate": 9.331241286582515e-05,
|
1703 |
+
"loss": 0.0238,
|
1704 |
+
"step": 2420
|
1705 |
+
},
|
1706 |
+
{
|
1707 |
+
"epoch": 13.575418994413408,
|
1708 |
+
"grad_norm": 0.3045707046985626,
|
1709 |
+
"learning_rate": 9.28641403637906e-05,
|
1710 |
+
"loss": 0.0292,
|
1711 |
+
"step": 2430
|
1712 |
+
},
|
1713 |
+
{
|
1714 |
+
"epoch": 13.631284916201118,
|
1715 |
+
"grad_norm": 0.2944694459438324,
|
1716 |
+
"learning_rate": 9.241519026729374e-05,
|
1717 |
+
"loss": 0.0252,
|
1718 |
+
"step": 2440
|
1719 |
+
},
|
1720 |
+
{
|
1721 |
+
"epoch": 13.687150837988828,
|
1722 |
+
"grad_norm": 0.18707014620304108,
|
1723 |
+
"learning_rate": 9.19655796052032e-05,
|
1724 |
+
"loss": 0.0264,
|
1725 |
+
"step": 2450
|
1726 |
+
},
|
1727 |
+
{
|
1728 |
+
"epoch": 13.743016759776536,
|
1729 |
+
"grad_norm": 0.4140284061431885,
|
1730 |
+
"learning_rate": 9.151532543144313e-05,
|
1731 |
+
"loss": 0.0284,
|
1732 |
+
"step": 2460
|
1733 |
+
},
|
1734 |
+
{
|
1735 |
+
"epoch": 13.798882681564246,
|
1736 |
+
"grad_norm": 0.24894434213638306,
|
1737 |
+
"learning_rate": 9.106444482434638e-05,
|
1738 |
+
"loss": 0.0288,
|
1739 |
+
"step": 2470
|
1740 |
+
},
|
1741 |
+
{
|
1742 |
+
"epoch": 13.854748603351956,
|
1743 |
+
"grad_norm": 0.24449239671230316,
|
1744 |
+
"learning_rate": 9.061295488600659e-05,
|
1745 |
+
"loss": 0.0232,
|
1746 |
+
"step": 2480
|
1747 |
+
},
|
1748 |
+
{
|
1749 |
+
"epoch": 13.910614525139664,
|
1750 |
+
"grad_norm": 0.2831474244594574,
|
1751 |
+
"learning_rate": 9.016087274162972e-05,
|
1752 |
+
"loss": 0.0238,
|
1753 |
+
"step": 2490
|
1754 |
+
},
|
1755 |
+
{
|
1756 |
+
"epoch": 13.966480446927374,
|
1757 |
+
"grad_norm": 0.27097368240356445,
|
1758 |
+
"learning_rate": 8.970821553888428e-05,
|
1759 |
+
"loss": 0.0314,
|
1760 |
+
"step": 2500
|
1761 |
+
},
|
1762 |
+
{
|
1763 |
+
"epoch": 14.022346368715084,
|
1764 |
+
"grad_norm": 0.2659156024456024,
|
1765 |
+
"learning_rate": 8.925500044725104e-05,
|
1766 |
+
"loss": 0.0254,
|
1767 |
+
"step": 2510
|
1768 |
+
},
|
1769 |
+
{
|
1770 |
+
"epoch": 14.078212290502794,
|
1771 |
+
"grad_norm": 0.200448676943779,
|
1772 |
+
"learning_rate": 8.880124465737168e-05,
|
1773 |
+
"loss": 0.0242,
|
1774 |
+
"step": 2520
|
1775 |
+
},
|
1776 |
+
{
|
1777 |
+
"epoch": 14.134078212290502,
|
1778 |
+
"grad_norm": 0.21925762295722961,
|
1779 |
+
"learning_rate": 8.834696538039678e-05,
|
1780 |
+
"loss": 0.0238,
|
1781 |
+
"step": 2530
|
1782 |
+
},
|
1783 |
+
{
|
1784 |
+
"epoch": 14.189944134078212,
|
1785 |
+
"grad_norm": 0.2610517144203186,
|
1786 |
+
"learning_rate": 8.789217984733312e-05,
|
1787 |
+
"loss": 0.0222,
|
1788 |
+
"step": 2540
|
1789 |
+
},
|
1790 |
+
{
|
1791 |
+
"epoch": 14.245810055865922,
|
1792 |
+
"grad_norm": 0.29856961965560913,
|
1793 |
+
"learning_rate": 8.743690530838983e-05,
|
1794 |
+
"loss": 0.0267,
|
1795 |
+
"step": 2550
|
1796 |
+
},
|
1797 |
+
{
|
1798 |
+
"epoch": 14.30167597765363,
|
1799 |
+
"grad_norm": 0.22454680502414703,
|
1800 |
+
"learning_rate": 8.69811590323244e-05,
|
1801 |
+
"loss": 0.023,
|
1802 |
+
"step": 2560
|
1803 |
+
},
|
1804 |
+
{
|
1805 |
+
"epoch": 14.35754189944134,
|
1806 |
+
"grad_norm": 0.3834502398967743,
|
1807 |
+
"learning_rate": 8.652495830578744e-05,
|
1808 |
+
"loss": 0.0248,
|
1809 |
+
"step": 2570
|
1810 |
+
},
|
1811 |
+
{
|
1812 |
+
"epoch": 14.41340782122905,
|
1813 |
+
"grad_norm": 0.3710498511791229,
|
1814 |
+
"learning_rate": 8.6068320432667e-05,
|
1815 |
+
"loss": 0.0263,
|
1816 |
+
"step": 2580
|
1817 |
+
},
|
1818 |
+
{
|
1819 |
+
"epoch": 14.46927374301676,
|
1820 |
+
"grad_norm": 0.19465550780296326,
|
1821 |
+
"learning_rate": 8.561126273343239e-05,
|
1822 |
+
"loss": 0.0219,
|
1823 |
+
"step": 2590
|
1824 |
+
},
|
1825 |
+
{
|
1826 |
+
"epoch": 14.525139664804469,
|
1827 |
+
"grad_norm": 0.1909979283809662,
|
1828 |
+
"learning_rate": 8.515380254447703e-05,
|
1829 |
+
"loss": 0.0255,
|
1830 |
+
"step": 2600
|
1831 |
+
},
|
1832 |
+
{
|
1833 |
+
"epoch": 14.581005586592179,
|
1834 |
+
"grad_norm": 0.3466046452522278,
|
1835 |
+
"learning_rate": 8.4695957217461e-05,
|
1836 |
+
"loss": 0.0249,
|
1837 |
+
"step": 2610
|
1838 |
+
},
|
1839 |
+
{
|
1840 |
+
"epoch": 14.636871508379889,
|
1841 |
+
"grad_norm": 0.3462992012500763,
|
1842 |
+
"learning_rate": 8.42377441186528e-05,
|
1843 |
+
"loss": 0.0245,
|
1844 |
+
"step": 2620
|
1845 |
+
},
|
1846 |
+
{
|
1847 |
+
"epoch": 14.692737430167599,
|
1848 |
+
"grad_norm": 0.28605830669403076,
|
1849 |
+
"learning_rate": 8.377918062827068e-05,
|
1850 |
+
"loss": 0.0254,
|
1851 |
+
"step": 2630
|
1852 |
+
},
|
1853 |
+
{
|
1854 |
+
"epoch": 14.748603351955307,
|
1855 |
+
"grad_norm": 0.2495012879371643,
|
1856 |
+
"learning_rate": 8.332028413982336e-05,
|
1857 |
+
"loss": 0.0268,
|
1858 |
+
"step": 2640
|
1859 |
+
},
|
1860 |
+
{
|
1861 |
+
"epoch": 14.804469273743017,
|
1862 |
+
"grad_norm": 0.23009264469146729,
|
1863 |
+
"learning_rate": 8.286107205945036e-05,
|
1864 |
+
"loss": 0.0235,
|
1865 |
+
"step": 2650
|
1866 |
+
},
|
1867 |
+
{
|
1868 |
+
"epoch": 14.860335195530727,
|
1869 |
+
"grad_norm": 0.5348438024520874,
|
1870 |
+
"learning_rate": 8.240156180526173e-05,
|
1871 |
+
"loss": 0.0282,
|
1872 |
+
"step": 2660
|
1873 |
+
},
|
1874 |
+
{
|
1875 |
+
"epoch": 14.916201117318435,
|
1876 |
+
"grad_norm": 0.23143941164016724,
|
1877 |
+
"learning_rate": 8.194177080667739e-05,
|
1878 |
+
"loss": 0.0227,
|
1879 |
+
"step": 2670
|
1880 |
+
},
|
1881 |
+
{
|
1882 |
+
"epoch": 14.972067039106145,
|
1883 |
+
"grad_norm": 0.2876596450805664,
|
1884 |
+
"learning_rate": 8.148171650376595e-05,
|
1885 |
+
"loss": 0.0217,
|
1886 |
+
"step": 2680
|
1887 |
+
},
|
1888 |
+
{
|
1889 |
+
"epoch": 15.027932960893855,
|
1890 |
+
"grad_norm": 0.21062903106212616,
|
1891 |
+
"learning_rate": 8.102141634658339e-05,
|
1892 |
+
"loss": 0.0226,
|
1893 |
+
"step": 2690
|
1894 |
+
},
|
1895 |
+
{
|
1896 |
+
"epoch": 15.083798882681565,
|
1897 |
+
"grad_norm": 0.22216618061065674,
|
1898 |
+
"learning_rate": 8.056088779451088e-05,
|
1899 |
+
"loss": 0.0244,
|
1900 |
+
"step": 2700
|
1901 |
+
},
|
1902 |
+
{
|
1903 |
+
"epoch": 15.139664804469273,
|
1904 |
+
"grad_norm": 0.2092180848121643,
|
1905 |
+
"learning_rate": 8.010014831559288e-05,
|
1906 |
+
"loss": 0.0267,
|
1907 |
+
"step": 2710
|
1908 |
+
},
|
1909 |
+
{
|
1910 |
+
"epoch": 15.195530726256983,
|
1911 |
+
"grad_norm": 0.31914299726486206,
|
1912 |
+
"learning_rate": 7.963921538587422e-05,
|
1913 |
+
"loss": 0.0186,
|
1914 |
+
"step": 2720
|
1915 |
+
},
|
1916 |
+
{
|
1917 |
+
"epoch": 15.251396648044693,
|
1918 |
+
"grad_norm": 0.28938183188438416,
|
1919 |
+
"learning_rate": 7.917810648873758e-05,
|
1920 |
+
"loss": 0.0254,
|
1921 |
+
"step": 2730
|
1922 |
+
},
|
1923 |
+
{
|
1924 |
+
"epoch": 15.307262569832401,
|
1925 |
+
"grad_norm": 0.2850409746170044,
|
1926 |
+
"learning_rate": 7.871683911424e-05,
|
1927 |
+
"loss": 0.0254,
|
1928 |
+
"step": 2740
|
1929 |
+
},
|
1930 |
+
{
|
1931 |
+
"epoch": 15.363128491620111,
|
1932 |
+
"grad_norm": 0.3008069097995758,
|
1933 |
+
"learning_rate": 7.82554307584498e-05,
|
1934 |
+
"loss": 0.021,
|
1935 |
+
"step": 2750
|
1936 |
+
},
|
1937 |
+
{
|
1938 |
+
"epoch": 15.418994413407821,
|
1939 |
+
"grad_norm": 0.2249889373779297,
|
1940 |
+
"learning_rate": 7.779389892278258e-05,
|
1941 |
+
"loss": 0.0215,
|
1942 |
+
"step": 2760
|
1943 |
+
},
|
1944 |
+
{
|
1945 |
+
"epoch": 15.474860335195531,
|
1946 |
+
"grad_norm": 0.3198390305042267,
|
1947 |
+
"learning_rate": 7.733226111333775e-05,
|
1948 |
+
"loss": 0.0214,
|
1949 |
+
"step": 2770
|
1950 |
+
},
|
1951 |
+
{
|
1952 |
+
"epoch": 15.53072625698324,
|
1953 |
+
"grad_norm": 0.2054470032453537,
|
1954 |
+
"learning_rate": 7.687053484023428e-05,
|
1955 |
+
"loss": 0.0248,
|
1956 |
+
"step": 2780
|
1957 |
+
},
|
1958 |
+
{
|
1959 |
+
"epoch": 15.58659217877095,
|
1960 |
+
"grad_norm": 0.3050447106361389,
|
1961 |
+
"learning_rate": 7.640873761694661e-05,
|
1962 |
+
"loss": 0.0254,
|
1963 |
+
"step": 2790
|
1964 |
+
},
|
1965 |
+
{
|
1966 |
+
"epoch": 15.64245810055866,
|
1967 |
+
"grad_norm": 0.3348614275455475,
|
1968 |
+
"learning_rate": 7.594688695964039e-05,
|
1969 |
+
"loss": 0.0228,
|
1970 |
+
"step": 2800
|
1971 |
+
},
|
1972 |
+
{
|
1973 |
+
"epoch": 15.69832402234637,
|
1974 |
+
"grad_norm": 0.2211858332157135,
|
1975 |
+
"learning_rate": 7.548500038650796e-05,
|
1976 |
+
"loss": 0.0211,
|
1977 |
+
"step": 2810
|
1978 |
+
},
|
1979 |
+
{
|
1980 |
+
"epoch": 15.754189944134078,
|
1981 |
+
"grad_norm": 0.25668853521347046,
|
1982 |
+
"learning_rate": 7.502309541710404e-05,
|
1983 |
+
"loss": 0.0247,
|
1984 |
+
"step": 2820
|
1985 |
+
},
|
1986 |
+
{
|
1987 |
+
"epoch": 15.810055865921788,
|
1988 |
+
"grad_norm": 0.18035383522510529,
|
1989 |
+
"learning_rate": 7.456118957168111e-05,
|
1990 |
+
"loss": 0.022,
|
1991 |
+
"step": 2830
|
1992 |
+
},
|
1993 |
+
{
|
1994 |
+
"epoch": 15.865921787709498,
|
1995 |
+
"grad_norm": 0.3626658320426941,
|
1996 |
+
"learning_rate": 7.409930037052485e-05,
|
1997 |
+
"loss": 0.0239,
|
1998 |
+
"step": 2840
|
1999 |
+
},
|
2000 |
+
{
|
2001 |
+
"epoch": 15.921787709497206,
|
2002 |
+
"grad_norm": 0.21136699616909027,
|
2003 |
+
"learning_rate": 7.363744533328964e-05,
|
2004 |
+
"loss": 0.0222,
|
2005 |
+
"step": 2850
|
2006 |
+
},
|
2007 |
+
{
|
2008 |
+
"epoch": 15.977653631284916,
|
2009 |
+
"grad_norm": 0.35462382435798645,
|
2010 |
+
"learning_rate": 7.317564197833401e-05,
|
2011 |
+
"loss": 0.0213,
|
2012 |
+
"step": 2860
|
2013 |
+
},
|
2014 |
+
{
|
2015 |
+
"epoch": 16.033519553072626,
|
2016 |
+
"grad_norm": 0.19702385365962982,
|
2017 |
+
"learning_rate": 7.271390782205617e-05,
|
2018 |
+
"loss": 0.0245,
|
2019 |
+
"step": 2870
|
2020 |
+
},
|
2021 |
+
{
|
2022 |
+
"epoch": 16.089385474860336,
|
2023 |
+
"grad_norm": 0.2611335813999176,
|
2024 |
+
"learning_rate": 7.225226037822956e-05,
|
2025 |
+
"loss": 0.0249,
|
2026 |
+
"step": 2880
|
2027 |
+
},
|
2028 |
+
{
|
2029 |
+
"epoch": 16.145251396648046,
|
2030 |
+
"grad_norm": 0.3108862340450287,
|
2031 |
+
"learning_rate": 7.179071715733858e-05,
|
2032 |
+
"loss": 0.0238,
|
2033 |
+
"step": 2890
|
2034 |
+
},
|
2035 |
+
{
|
2036 |
+
"epoch": 16.201117318435756,
|
2037 |
+
"grad_norm": 0.3305630385875702,
|
2038 |
+
"learning_rate": 7.132929566591445e-05,
|
2039 |
+
"loss": 0.0254,
|
2040 |
+
"step": 2900
|
2041 |
+
},
|
2042 |
+
{
|
2043 |
+
"epoch": 16.256983240223462,
|
2044 |
+
"grad_norm": 0.29176512360572815,
|
2045 |
+
"learning_rate": 7.086801340587114e-05,
|
2046 |
+
"loss": 0.0224,
|
2047 |
+
"step": 2910
|
2048 |
+
},
|
2049 |
+
{
|
2050 |
+
"epoch": 16.312849162011172,
|
2051 |
+
"grad_norm": 0.33446812629699707,
|
2052 |
+
"learning_rate": 7.040688787384144e-05,
|
2053 |
+
"loss": 0.0223,
|
2054 |
+
"step": 2920
|
2055 |
+
},
|
2056 |
+
{
|
2057 |
+
"epoch": 16.368715083798882,
|
2058 |
+
"grad_norm": 0.19562651216983795,
|
2059 |
+
"learning_rate": 6.994593656051346e-05,
|
2060 |
+
"loss": 0.0254,
|
2061 |
+
"step": 2930
|
2062 |
+
},
|
2063 |
+
{
|
2064 |
+
"epoch": 16.424581005586592,
|
2065 |
+
"grad_norm": 0.21511122584342957,
|
2066 |
+
"learning_rate": 6.94851769499671e-05,
|
2067 |
+
"loss": 0.0226,
|
2068 |
+
"step": 2940
|
2069 |
+
},
|
2070 |
+
{
|
2071 |
+
"epoch": 16.480446927374302,
|
2072 |
+
"grad_norm": 0.23985400795936584,
|
2073 |
+
"learning_rate": 6.902462651901085e-05,
|
2074 |
+
"loss": 0.0269,
|
2075 |
+
"step": 2950
|
2076 |
+
},
|
2077 |
+
{
|
2078 |
+
"epoch": 16.536312849162012,
|
2079 |
+
"grad_norm": 0.3013317584991455,
|
2080 |
+
"learning_rate": 6.856430273651896e-05,
|
2081 |
+
"loss": 0.021,
|
2082 |
+
"step": 2960
|
2083 |
+
},
|
2084 |
+
{
|
2085 |
+
"epoch": 16.592178770949722,
|
2086 |
+
"grad_norm": 0.35089585185050964,
|
2087 |
+
"learning_rate": 6.810422306276884e-05,
|
2088 |
+
"loss": 0.0232,
|
2089 |
+
"step": 2970
|
2090 |
+
},
|
2091 |
+
{
|
2092 |
+
"epoch": 16.64804469273743,
|
2093 |
+
"grad_norm": 0.17514319717884064,
|
2094 |
+
"learning_rate": 6.764440494877874e-05,
|
2095 |
+
"loss": 0.0203,
|
2096 |
+
"step": 2980
|
2097 |
+
},
|
2098 |
+
{
|
2099 |
+
"epoch": 16.70391061452514,
|
2100 |
+
"grad_norm": 0.15833304822444916,
|
2101 |
+
"learning_rate": 6.718486583564577e-05,
|
2102 |
+
"loss": 0.0207,
|
2103 |
+
"step": 2990
|
2104 |
+
},
|
2105 |
+
{
|
2106 |
+
"epoch": 16.75977653631285,
|
2107 |
+
"grad_norm": 0.22394374012947083,
|
2108 |
+
"learning_rate": 6.672562315388446e-05,
|
2109 |
+
"loss": 0.0219,
|
2110 |
+
"step": 3000
|
2111 |
+
},
|
2112 |
+
{
|
2113 |
+
"epoch": 16.81564245810056,
|
2114 |
+
"grad_norm": 0.33897238969802856,
|
2115 |
+
"learning_rate": 6.62666943227656e-05,
|
2116 |
+
"loss": 0.0231,
|
2117 |
+
"step": 3010
|
2118 |
+
},
|
2119 |
+
{
|
2120 |
+
"epoch": 16.87150837988827,
|
2121 |
+
"grad_norm": 0.3397151529788971,
|
2122 |
+
"learning_rate": 6.580809674965549e-05,
|
2123 |
+
"loss": 0.0203,
|
2124 |
+
"step": 3020
|
2125 |
+
},
|
2126 |
+
{
|
2127 |
+
"epoch": 16.92737430167598,
|
2128 |
+
"grad_norm": 0.2227960079908371,
|
2129 |
+
"learning_rate": 6.53498478293556e-05,
|
2130 |
+
"loss": 0.0197,
|
2131 |
+
"step": 3030
|
2132 |
+
},
|
2133 |
+
{
|
2134 |
+
"epoch": 16.98324022346369,
|
2135 |
+
"grad_norm": 0.29340264201164246,
|
2136 |
+
"learning_rate": 6.489196494344294e-05,
|
2137 |
+
"loss": 0.0238,
|
2138 |
+
"step": 3040
|
2139 |
+
},
|
2140 |
+
{
|
2141 |
+
"epoch": 17.039106145251395,
|
2142 |
+
"grad_norm": 0.35238850116729736,
|
2143 |
+
"learning_rate": 6.44344654596106e-05,
|
2144 |
+
"loss": 0.0271,
|
2145 |
+
"step": 3050
|
2146 |
+
},
|
2147 |
+
{
|
2148 |
+
"epoch": 17.094972067039105,
|
2149 |
+
"grad_norm": 0.18793293833732605,
|
2150 |
+
"learning_rate": 6.397736673100918e-05,
|
2151 |
+
"loss": 0.021,
|
2152 |
+
"step": 3060
|
2153 |
+
},
|
2154 |
+
{
|
2155 |
+
"epoch": 17.150837988826815,
|
2156 |
+
"grad_norm": 0.25130748748779297,
|
2157 |
+
"learning_rate": 6.35206860955883e-05,
|
2158 |
+
"loss": 0.0249,
|
2159 |
+
"step": 3070
|
2160 |
+
},
|
2161 |
+
{
|
2162 |
+
"epoch": 17.206703910614525,
|
2163 |
+
"grad_norm": 0.2051066905260086,
|
2164 |
+
"learning_rate": 6.306444087543927e-05,
|
2165 |
+
"loss": 0.019,
|
2166 |
+
"step": 3080
|
2167 |
+
},
|
2168 |
+
{
|
2169 |
+
"epoch": 17.262569832402235,
|
2170 |
+
"grad_norm": 0.16360081732273102,
|
2171 |
+
"learning_rate": 6.260864837613788e-05,
|
2172 |
+
"loss": 0.0213,
|
2173 |
+
"step": 3090
|
2174 |
+
},
|
2175 |
+
{
|
2176 |
+
"epoch": 17.318435754189945,
|
2177 |
+
"grad_norm": 0.3169659674167633,
|
2178 |
+
"learning_rate": 6.215332588608793e-05,
|
2179 |
+
"loss": 0.018,
|
2180 |
+
"step": 3100
|
2181 |
+
},
|
2182 |
+
{
|
2183 |
+
"epoch": 17.374301675977655,
|
2184 |
+
"grad_norm": 0.24224449694156647,
|
2185 |
+
"learning_rate": 6.16984906758657e-05,
|
2186 |
+
"loss": 0.0194,
|
2187 |
+
"step": 3110
|
2188 |
+
},
|
2189 |
+
{
|
2190 |
+
"epoch": 17.43016759776536,
|
2191 |
+
"grad_norm": 0.3294720947742462,
|
2192 |
+
"learning_rate": 6.124415999756466e-05,
|
2193 |
+
"loss": 0.025,
|
2194 |
+
"step": 3120
|
2195 |
+
},
|
2196 |
+
{
|
2197 |
+
"epoch": 17.48603351955307,
|
2198 |
+
"grad_norm": 0.3010692298412323,
|
2199 |
+
"learning_rate": 6.079035108414123e-05,
|
2200 |
+
"loss": 0.019,
|
2201 |
+
"step": 3130
|
2202 |
+
},
|
2203 |
+
{
|
2204 |
+
"epoch": 17.54189944134078,
|
2205 |
+
"grad_norm": 0.1720583438873291,
|
2206 |
+
"learning_rate": 6.033708114876097e-05,
|
2207 |
+
"loss": 0.0197,
|
2208 |
+
"step": 3140
|
2209 |
+
},
|
2210 |
+
{
|
2211 |
+
"epoch": 17.59776536312849,
|
2212 |
+
"grad_norm": 0.13594307005405426,
|
2213 |
+
"learning_rate": 5.988436738414584e-05,
|
2214 |
+
"loss": 0.0208,
|
2215 |
+
"step": 3150
|
2216 |
+
},
|
2217 |
+
{
|
2218 |
+
"epoch": 17.6536312849162,
|
2219 |
+
"grad_norm": 0.2565588057041168,
|
2220 |
+
"learning_rate": 5.943222696192204e-05,
|
2221 |
+
"loss": 0.0246,
|
2222 |
+
"step": 3160
|
2223 |
+
},
|
2224 |
+
{
|
2225 |
+
"epoch": 17.70949720670391,
|
2226 |
+
"grad_norm": 0.1882692277431488,
|
2227 |
+
"learning_rate": 5.898067703196857e-05,
|
2228 |
+
"loss": 0.023,
|
2229 |
+
"step": 3170
|
2230 |
+
},
|
2231 |
+
{
|
2232 |
+
"epoch": 17.76536312849162,
|
2233 |
+
"grad_norm": 0.16729940474033356,
|
2234 |
+
"learning_rate": 5.852973472176685e-05,
|
2235 |
+
"loss": 0.0188,
|
2236 |
+
"step": 3180
|
2237 |
+
},
|
2238 |
+
{
|
2239 |
+
"epoch": 17.821229050279328,
|
2240 |
+
"grad_norm": 0.2922714948654175,
|
2241 |
+
"learning_rate": 5.807941713575098e-05,
|
2242 |
+
"loss": 0.0226,
|
2243 |
+
"step": 3190
|
2244 |
+
},
|
2245 |
+
{
|
2246 |
+
"epoch": 17.877094972067038,
|
2247 |
+
"grad_norm": 0.33513203263282776,
|
2248 |
+
"learning_rate": 5.762974135465907e-05,
|
2249 |
+
"loss": 0.0234,
|
2250 |
+
"step": 3200
|
2251 |
+
},
|
2252 |
+
{
|
2253 |
+
"epoch": 17.932960893854748,
|
2254 |
+
"grad_norm": 0.2364656627178192,
|
2255 |
+
"learning_rate": 5.7180724434885294e-05,
|
2256 |
+
"loss": 0.0195,
|
2257 |
+
"step": 3210
|
2258 |
+
},
|
2259 |
+
{
|
2260 |
+
"epoch": 17.988826815642458,
|
2261 |
+
"grad_norm": 0.23285451531410217,
|
2262 |
+
"learning_rate": 5.673238340783285e-05,
|
2263 |
+
"loss": 0.0193,
|
2264 |
+
"step": 3220
|
2265 |
+
},
|
2266 |
+
{
|
2267 |
+
"epoch": 18.044692737430168,
|
2268 |
+
"grad_norm": 0.20390582084655762,
|
2269 |
+
"learning_rate": 5.6284735279268134e-05,
|
2270 |
+
"loss": 0.0206,
|
2271 |
+
"step": 3230
|
2272 |
+
},
|
2273 |
+
{
|
2274 |
+
"epoch": 18.100558659217878,
|
2275 |
+
"grad_norm": 0.3086446225643158,
|
2276 |
+
"learning_rate": 5.583779702867555e-05,
|
2277 |
+
"loss": 0.0201,
|
2278 |
+
"step": 3240
|
2279 |
+
},
|
2280 |
+
{
|
2281 |
+
"epoch": 18.156424581005588,
|
2282 |
+
"grad_norm": 0.17210273444652557,
|
2283 |
+
"learning_rate": 5.5391585608613514e-05,
|
2284 |
+
"loss": 0.0231,
|
2285 |
+
"step": 3250
|
2286 |
+
},
|
2287 |
+
{
|
2288 |
+
"epoch": 18.212290502793294,
|
2289 |
+
"grad_norm": 0.18998980522155762,
|
2290 |
+
"learning_rate": 5.4946117944071514e-05,
|
2291 |
+
"loss": 0.0203,
|
2292 |
+
"step": 3260
|
2293 |
+
},
|
2294 |
+
{
|
2295 |
+
"epoch": 18.268156424581004,
|
2296 |
+
"grad_norm": 0.2326553910970688,
|
2297 |
+
"learning_rate": 5.450141093182803e-05,
|
2298 |
+
"loss": 0.0194,
|
2299 |
+
"step": 3270
|
2300 |
+
},
|
2301 |
+
{
|
2302 |
+
"epoch": 18.324022346368714,
|
2303 |
+
"grad_norm": 0.24682560563087463,
|
2304 |
+
"learning_rate": 5.405748143980972e-05,
|
2305 |
+
"loss": 0.0208,
|
2306 |
+
"step": 3280
|
2307 |
+
},
|
2308 |
+
{
|
2309 |
+
"epoch": 18.379888268156424,
|
2310 |
+
"grad_norm": 0.2583821415901184,
|
2311 |
+
"learning_rate": 5.361434630645148e-05,
|
2312 |
+
"loss": 0.0204,
|
2313 |
+
"step": 3290
|
2314 |
+
},
|
2315 |
+
{
|
2316 |
+
"epoch": 18.435754189944134,
|
2317 |
+
"grad_norm": 0.23446252942085266,
|
2318 |
+
"learning_rate": 5.31720223400579e-05,
|
2319 |
+
"loss": 0.0176,
|
2320 |
+
"step": 3300
|
2321 |
+
},
|
2322 |
+
{
|
2323 |
+
"epoch": 18.491620111731844,
|
2324 |
+
"grad_norm": 0.2943982779979706,
|
2325 |
+
"learning_rate": 5.2730526318165684e-05,
|
2326 |
+
"loss": 0.0192,
|
2327 |
+
"step": 3310
|
2328 |
+
},
|
2329 |
+
{
|
2330 |
+
"epoch": 18.547486033519554,
|
2331 |
+
"grad_norm": 0.2999999225139618,
|
2332 |
+
"learning_rate": 5.2289874986907236e-05,
|
2333 |
+
"loss": 0.0204,
|
2334 |
+
"step": 3320
|
2335 |
+
},
|
2336 |
+
{
|
2337 |
+
"epoch": 18.60335195530726,
|
2338 |
+
"grad_norm": 0.21415217220783234,
|
2339 |
+
"learning_rate": 5.1850085060375417e-05,
|
2340 |
+
"loss": 0.0201,
|
2341 |
+
"step": 3330
|
2342 |
+
},
|
2343 |
+
{
|
2344 |
+
"epoch": 18.65921787709497,
|
2345 |
+
"grad_norm": 0.2961329221725464,
|
2346 |
+
"learning_rate": 5.1411173219989736e-05,
|
2347 |
+
"loss": 0.0199,
|
2348 |
+
"step": 3340
|
2349 |
+
},
|
2350 |
+
{
|
2351 |
+
"epoch": 18.71508379888268,
|
2352 |
+
"grad_norm": 0.25467586517333984,
|
2353 |
+
"learning_rate": 5.097315611386345e-05,
|
2354 |
+
"loss": 0.021,
|
2355 |
+
"step": 3350
|
2356 |
+
},
|
2357 |
+
{
|
2358 |
+
"epoch": 18.77094972067039,
|
2359 |
+
"grad_norm": 0.2654760777950287,
|
2360 |
+
"learning_rate": 5.0536050356172234e-05,
|
2361 |
+
"loss": 0.0171,
|
2362 |
+
"step": 3360
|
2363 |
+
},
|
2364 |
+
{
|
2365 |
+
"epoch": 18.8268156424581,
|
2366 |
+
"grad_norm": 0.279527485370636,
|
2367 |
+
"learning_rate": 5.0099872526523826e-05,
|
2368 |
+
"loss": 0.0205,
|
2369 |
+
"step": 3370
|
2370 |
+
},
|
2371 |
+
{
|
2372 |
+
"epoch": 18.88268156424581,
|
2373 |
+
"grad_norm": 0.41244757175445557,
|
2374 |
+
"learning_rate": 4.966463916932934e-05,
|
2375 |
+
"loss": 0.0162,
|
2376 |
+
"step": 3380
|
2377 |
+
},
|
2378 |
+
{
|
2379 |
+
"epoch": 18.93854748603352,
|
2380 |
+
"grad_norm": 0.25530514121055603,
|
2381 |
+
"learning_rate": 4.9230366793175654e-05,
|
2382 |
+
"loss": 0.0181,
|
2383 |
+
"step": 3390
|
2384 |
+
},
|
2385 |
+
{
|
2386 |
+
"epoch": 18.994413407821227,
|
2387 |
+
"grad_norm": 0.18296165764331818,
|
2388 |
+
"learning_rate": 4.879707187019914e-05,
|
2389 |
+
"loss": 0.0229,
|
2390 |
+
"step": 3400
|
2391 |
+
},
|
2392 |
+
{
|
2393 |
+
"epoch": 19.050279329608937,
|
2394 |
+
"grad_norm": 0.18316328525543213,
|
2395 |
+
"learning_rate": 4.8364770835461054e-05,
|
2396 |
+
"loss": 0.0195,
|
2397 |
+
"step": 3410
|
2398 |
+
},
|
2399 |
+
{
|
2400 |
+
"epoch": 19.106145251396647,
|
2401 |
+
"grad_norm": 0.1494152843952179,
|
2402 |
+
"learning_rate": 4.793348008632396e-05,
|
2403 |
+
"loss": 0.0167,
|
2404 |
+
"step": 3420
|
2405 |
+
},
|
2406 |
+
{
|
2407 |
+
"epoch": 19.162011173184357,
|
2408 |
+
"grad_norm": 0.3076527416706085,
|
2409 |
+
"learning_rate": 4.750321598182995e-05,
|
2410 |
+
"loss": 0.0177,
|
2411 |
+
"step": 3430
|
2412 |
+
},
|
2413 |
+
{
|
2414 |
+
"epoch": 19.217877094972067,
|
2415 |
+
"grad_norm": 0.27527594566345215,
|
2416 |
+
"learning_rate": 4.707399484207995e-05,
|
2417 |
+
"loss": 0.0216,
|
2418 |
+
"step": 3440
|
2419 |
+
},
|
2420 |
+
{
|
2421 |
+
"epoch": 19.273743016759777,
|
2422 |
+
"grad_norm": 0.16726377606391907,
|
2423 |
+
"learning_rate": 4.664583294761485e-05,
|
2424 |
+
"loss": 0.0211,
|
2425 |
+
"step": 3450
|
2426 |
+
},
|
2427 |
+
{
|
2428 |
+
"epoch": 19.329608938547487,
|
2429 |
+
"grad_norm": 0.2316291779279709,
|
2430 |
+
"learning_rate": 4.6218746538797926e-05,
|
2431 |
+
"loss": 0.0197,
|
2432 |
+
"step": 3460
|
2433 |
+
},
|
2434 |
+
{
|
2435 |
+
"epoch": 19.385474860335197,
|
2436 |
+
"grad_norm": 0.2616920471191406,
|
2437 |
+
"learning_rate": 4.579275181519879e-05,
|
2438 |
+
"loss": 0.0193,
|
2439 |
+
"step": 3470
|
2440 |
+
},
|
2441 |
+
{
|
2442 |
+
"epoch": 19.441340782122904,
|
2443 |
+
"grad_norm": 0.17275288701057434,
|
2444 |
+
"learning_rate": 4.536786493497894e-05,
|
2445 |
+
"loss": 0.0186,
|
2446 |
+
"step": 3480
|
2447 |
+
},
|
2448 |
+
{
|
2449 |
+
"epoch": 19.497206703910614,
|
2450 |
+
"grad_norm": 0.18898899853229523,
|
2451 |
+
"learning_rate": 4.494410201427896e-05,
|
2452 |
+
"loss": 0.0198,
|
2453 |
+
"step": 3490
|
2454 |
+
},
|
2455 |
+
{
|
2456 |
+
"epoch": 19.553072625698324,
|
2457 |
+
"grad_norm": 0.3262523412704468,
|
2458 |
+
"learning_rate": 4.452147912660715e-05,
|
2459 |
+
"loss": 0.0212,
|
2460 |
+
"step": 3500
|
2461 |
+
},
|
2462 |
+
{
|
2463 |
+
"epoch": 19.608938547486034,
|
2464 |
+
"grad_norm": 0.1693217158317566,
|
2465 |
+
"learning_rate": 4.4100012302229915e-05,
|
2466 |
+
"loss": 0.0203,
|
2467 |
+
"step": 3510
|
2468 |
+
},
|
2469 |
+
{
|
2470 |
+
"epoch": 19.664804469273744,
|
2471 |
+
"grad_norm": 0.1959124356508255,
|
2472 |
+
"learning_rate": 4.3679717527563576e-05,
|
2473 |
+
"loss": 0.0184,
|
2474 |
+
"step": 3520
|
2475 |
+
},
|
2476 |
+
{
|
2477 |
+
"epoch": 19.720670391061454,
|
2478 |
+
"grad_norm": 0.21138575673103333,
|
2479 |
+
"learning_rate": 4.326061074456817e-05,
|
2480 |
+
"loss": 0.0172,
|
2481 |
+
"step": 3530
|
2482 |
+
},
|
2483 |
+
{
|
2484 |
+
"epoch": 19.776536312849164,
|
2485 |
+
"grad_norm": 0.6261563301086426,
|
2486 |
+
"learning_rate": 4.2842707850142806e-05,
|
2487 |
+
"loss": 0.018,
|
2488 |
+
"step": 3540
|
2489 |
+
},
|
2490 |
+
{
|
2491 |
+
"epoch": 19.83240223463687,
|
2492 |
+
"grad_norm": 0.2905479073524475,
|
2493 |
+
"learning_rate": 4.242602469552234e-05,
|
2494 |
+
"loss": 0.017,
|
2495 |
+
"step": 3550
|
2496 |
+
},
|
2497 |
+
{
|
2498 |
+
"epoch": 19.88826815642458,
|
2499 |
+
"grad_norm": 0.4615681767463684,
|
2500 |
+
"learning_rate": 4.201057708567664e-05,
|
2501 |
+
"loss": 0.0206,
|
2502 |
+
"step": 3560
|
2503 |
+
},
|
2504 |
+
{
|
2505 |
+
"epoch": 19.94413407821229,
|
2506 |
+
"grad_norm": 0.1858612447977066,
|
2507 |
+
"learning_rate": 4.1596380778710686e-05,
|
2508 |
+
"loss": 0.0193,
|
2509 |
+
"step": 3570
|
2510 |
+
},
|
2511 |
+
{
|
2512 |
+
"epoch": 20.0,
|
2513 |
+
"grad_norm": 0.21462394297122955,
|
2514 |
+
"learning_rate": 4.1183451485267034e-05,
|
2515 |
+
"loss": 0.0175,
|
2516 |
+
"step": 3580
|
2517 |
+
},
|
2518 |
+
{
|
2519 |
+
"epoch": 20.05586592178771,
|
2520 |
+
"grad_norm": 0.2793791890144348,
|
2521 |
+
"learning_rate": 4.077180486792987e-05,
|
2522 |
+
"loss": 0.0166,
|
2523 |
+
"step": 3590
|
2524 |
+
},
|
2525 |
+
{
|
2526 |
+
"epoch": 20.11173184357542,
|
2527 |
+
"grad_norm": 0.22271260619163513,
|
2528 |
+
"learning_rate": 4.036145654063093e-05,
|
2529 |
+
"loss": 0.0146,
|
2530 |
+
"step": 3600
|
2531 |
+
},
|
2532 |
+
{
|
2533 |
+
"epoch": 20.16759776536313,
|
2534 |
+
"grad_norm": 0.15600596368312836,
|
2535 |
+
"learning_rate": 3.995242206805731e-05,
|
2536 |
+
"loss": 0.0193,
|
2537 |
+
"step": 3610
|
2538 |
+
},
|
2539 |
+
{
|
2540 |
+
"epoch": 20.223463687150836,
|
2541 |
+
"grad_norm": 0.20925801992416382,
|
2542 |
+
"learning_rate": 3.9544716965060996e-05,
|
2543 |
+
"loss": 0.0209,
|
2544 |
+
"step": 3620
|
2545 |
+
},
|
2546 |
+
{
|
2547 |
+
"epoch": 20.279329608938546,
|
2548 |
+
"grad_norm": 0.16699257493019104,
|
2549 |
+
"learning_rate": 3.9138356696070315e-05,
|
2550 |
+
"loss": 0.0136,
|
2551 |
+
"step": 3630
|
2552 |
+
},
|
2553 |
+
{
|
2554 |
+
"epoch": 20.335195530726256,
|
2555 |
+
"grad_norm": 0.22640997171401978,
|
2556 |
+
"learning_rate": 3.8733356674503676e-05,
|
2557 |
+
"loss": 0.0161,
|
2558 |
+
"step": 3640
|
2559 |
+
},
|
2560 |
+
{
|
2561 |
+
"epoch": 20.391061452513966,
|
2562 |
+
"grad_norm": 0.5254058837890625,
|
2563 |
+
"learning_rate": 3.832973226218457e-05,
|
2564 |
+
"loss": 0.0176,
|
2565 |
+
"step": 3650
|
2566 |
+
},
|
2567 |
+
{
|
2568 |
+
"epoch": 20.446927374301676,
|
2569 |
+
"grad_norm": 0.148833766579628,
|
2570 |
+
"learning_rate": 3.792749876875907e-05,
|
2571 |
+
"loss": 0.0147,
|
2572 |
+
"step": 3660
|
2573 |
+
},
|
2574 |
+
{
|
2575 |
+
"epoch": 20.502793296089386,
|
2576 |
+
"grad_norm": 0.26952382922172546,
|
2577 |
+
"learning_rate": 3.752667145111514e-05,
|
2578 |
+
"loss": 0.0193,
|
2579 |
+
"step": 3670
|
2580 |
+
},
|
2581 |
+
{
|
2582 |
+
"epoch": 20.558659217877096,
|
2583 |
+
"grad_norm": 0.24985520541667938,
|
2584 |
+
"learning_rate": 3.712726551280385e-05,
|
2585 |
+
"loss": 0.0164,
|
2586 |
+
"step": 3680
|
2587 |
+
},
|
2588 |
+
{
|
2589 |
+
"epoch": 20.614525139664803,
|
2590 |
+
"grad_norm": 0.12173712253570557,
|
2591 |
+
"learning_rate": 3.6729296103462835e-05,
|
2592 |
+
"loss": 0.016,
|
2593 |
+
"step": 3690
|
2594 |
+
},
|
2595 |
+
{
|
2596 |
+
"epoch": 20.670391061452513,
|
2597 |
+
"grad_norm": 0.24871109426021576,
|
2598 |
+
"learning_rate": 3.63327783182414e-05,
|
2599 |
+
"loss": 0.0174,
|
2600 |
+
"step": 3700
|
2601 |
+
},
|
2602 |
+
{
|
2603 |
+
"epoch": 20.726256983240223,
|
2604 |
+
"grad_norm": 0.21813338994979858,
|
2605 |
+
"learning_rate": 3.593772719722832e-05,
|
2606 |
+
"loss": 0.0175,
|
2607 |
+
"step": 3710
|
2608 |
+
},
|
2609 |
+
{
|
2610 |
+
"epoch": 20.782122905027933,
|
2611 |
+
"grad_norm": 0.15502704679965973,
|
2612 |
+
"learning_rate": 3.5544157724881045e-05,
|
2613 |
+
"loss": 0.0171,
|
2614 |
+
"step": 3720
|
2615 |
+
},
|
2616 |
+
{
|
2617 |
+
"epoch": 20.837988826815643,
|
2618 |
+
"grad_norm": 0.2693827152252197,
|
2619 |
+
"learning_rate": 3.515208482945758e-05,
|
2620 |
+
"loss": 0.0184,
|
2621 |
+
"step": 3730
|
2622 |
+
},
|
2623 |
+
{
|
2624 |
+
"epoch": 20.893854748603353,
|
2625 |
+
"grad_norm": 0.2545803487300873,
|
2626 |
+
"learning_rate": 3.476152338244995e-05,
|
2627 |
+
"loss": 0.0211,
|
2628 |
+
"step": 3740
|
2629 |
+
},
|
2630 |
+
{
|
2631 |
+
"epoch": 20.949720670391063,
|
2632 |
+
"grad_norm": 0.14349234104156494,
|
2633 |
+
"learning_rate": 3.437248819802042e-05,
|
2634 |
+
"loss": 0.016,
|
2635 |
+
"step": 3750
|
2636 |
+
},
|
2637 |
+
{
|
2638 |
+
"epoch": 21.00558659217877,
|
2639 |
+
"grad_norm": 0.42411503195762634,
|
2640 |
+
"learning_rate": 3.3984994032439496e-05,
|
2641 |
+
"loss": 0.0178,
|
2642 |
+
"step": 3760
|
2643 |
+
},
|
2644 |
+
{
|
2645 |
+
"epoch": 21.06145251396648,
|
2646 |
+
"grad_norm": 0.25866132974624634,
|
2647 |
+
"learning_rate": 3.359905558352609e-05,
|
2648 |
+
"loss": 0.0165,
|
2649 |
+
"step": 3770
|
2650 |
+
},
|
2651 |
+
{
|
2652 |
+
"epoch": 21.11731843575419,
|
2653 |
+
"grad_norm": 0.22836929559707642,
|
2654 |
+
"learning_rate": 3.321468749009017e-05,
|
2655 |
+
"loss": 0.0196,
|
2656 |
+
"step": 3780
|
2657 |
+
},
|
2658 |
+
{
|
2659 |
+
"epoch": 21.1731843575419,
|
2660 |
+
"grad_norm": 0.1367328017950058,
|
2661 |
+
"learning_rate": 3.283190433137742e-05,
|
2662 |
+
"loss": 0.0177,
|
2663 |
+
"step": 3790
|
2664 |
+
},
|
2665 |
+
{
|
2666 |
+
"epoch": 21.22905027932961,
|
2667 |
+
"grad_norm": 0.16710641980171204,
|
2668 |
+
"learning_rate": 3.2450720626516294e-05,
|
2669 |
+
"loss": 0.0153,
|
2670 |
+
"step": 3800
|
2671 |
+
},
|
2672 |
+
{
|
2673 |
+
"epoch": 21.28491620111732,
|
2674 |
+
"grad_norm": 0.4859740734100342,
|
2675 |
+
"learning_rate": 3.207115083396728e-05,
|
2676 |
+
"loss": 0.0166,
|
2677 |
+
"step": 3810
|
2678 |
+
},
|
2679 |
+
{
|
2680 |
+
"epoch": 21.34078212290503,
|
2681 |
+
"grad_norm": 0.2656506299972534,
|
2682 |
+
"learning_rate": 3.1693209350974466e-05,
|
2683 |
+
"loss": 0.0174,
|
2684 |
+
"step": 3820
|
2685 |
+
},
|
2686 |
+
{
|
2687 |
+
"epoch": 21.39664804469274,
|
2688 |
+
"grad_norm": 0.2200634777545929,
|
2689 |
+
"learning_rate": 3.131691051301952e-05,
|
2690 |
+
"loss": 0.0192,
|
2691 |
+
"step": 3830
|
2692 |
+
},
|
2693 |
+
{
|
2694 |
+
"epoch": 21.452513966480446,
|
2695 |
+
"grad_norm": 0.15455004572868347,
|
2696 |
+
"learning_rate": 3.09422685932778e-05,
|
2697 |
+
"loss": 0.0185,
|
2698 |
+
"step": 3840
|
2699 |
+
},
|
2700 |
+
{
|
2701 |
+
"epoch": 21.508379888268156,
|
2702 |
+
"grad_norm": 0.32110223174095154,
|
2703 |
+
"learning_rate": 3.056929780207711e-05,
|
2704 |
+
"loss": 0.0178,
|
2705 |
+
"step": 3850
|
2706 |
+
},
|
2707 |
+
{
|
2708 |
+
"epoch": 21.564245810055866,
|
2709 |
+
"grad_norm": 0.20172889530658722,
|
2710 |
+
"learning_rate": 3.0198012286358618e-05,
|
2711 |
+
"loss": 0.0154,
|
2712 |
+
"step": 3860
|
2713 |
+
},
|
2714 |
+
{
|
2715 |
+
"epoch": 21.620111731843576,
|
2716 |
+
"grad_norm": 0.1673143357038498,
|
2717 |
+
"learning_rate": 2.9828426129140237e-05,
|
2718 |
+
"loss": 0.0173,
|
2719 |
+
"step": 3870
|
2720 |
+
},
|
2721 |
+
{
|
2722 |
+
"epoch": 21.675977653631286,
|
2723 |
+
"grad_norm": 0.13557426631450653,
|
2724 |
+
"learning_rate": 2.94605533489826e-05,
|
2725 |
+
"loss": 0.0146,
|
2726 |
+
"step": 3880
|
2727 |
+
},
|
2728 |
+
{
|
2729 |
+
"epoch": 21.731843575418996,
|
2730 |
+
"grad_norm": 0.11623450368642807,
|
2731 |
+
"learning_rate": 2.9094407899457046e-05,
|
2732 |
+
"loss": 0.0139,
|
2733 |
+
"step": 3890
|
2734 |
+
},
|
2735 |
+
{
|
2736 |
+
"epoch": 21.787709497206706,
|
2737 |
+
"grad_norm": 0.16807802021503448,
|
2738 |
+
"learning_rate": 2.8730003668616698e-05,
|
2739 |
+
"loss": 0.0135,
|
2740 |
+
"step": 3900
|
2741 |
+
},
|
2742 |
+
{
|
2743 |
+
"epoch": 21.843575418994412,
|
2744 |
+
"grad_norm": 0.37868157029151917,
|
2745 |
+
"learning_rate": 2.8367354478469396e-05,
|
2746 |
+
"loss": 0.0151,
|
2747 |
+
"step": 3910
|
2748 |
+
},
|
2749 |
+
{
|
2750 |
+
"epoch": 21.899441340782122,
|
2751 |
+
"grad_norm": 0.29837918281555176,
|
2752 |
+
"learning_rate": 2.8006474084453595e-05,
|
2753 |
+
"loss": 0.0151,
|
2754 |
+
"step": 3920
|
2755 |
+
},
|
2756 |
+
{
|
2757 |
+
"epoch": 21.955307262569832,
|
2758 |
+
"grad_norm": 0.24290640652179718,
|
2759 |
+
"learning_rate": 2.764737617491653e-05,
|
2760 |
+
"loss": 0.0141,
|
2761 |
+
"step": 3930
|
2762 |
+
},
|
2763 |
+
{
|
2764 |
+
"epoch": 22.011173184357542,
|
2765 |
+
"grad_norm": 0.1992618888616562,
|
2766 |
+
"learning_rate": 2.729007437059502e-05,
|
2767 |
+
"loss": 0.0145,
|
2768 |
+
"step": 3940
|
2769 |
+
},
|
2770 |
+
{
|
2771 |
+
"epoch": 22.067039106145252,
|
2772 |
+
"grad_norm": 0.13898395001888275,
|
2773 |
+
"learning_rate": 2.6934582224098963e-05,
|
2774 |
+
"loss": 0.0145,
|
2775 |
+
"step": 3950
|
2776 |
+
},
|
2777 |
+
{
|
2778 |
+
"epoch": 22.122905027932962,
|
2779 |
+
"grad_norm": 0.257982075214386,
|
2780 |
+
"learning_rate": 2.658091321939697e-05,
|
2781 |
+
"loss": 0.0144,
|
2782 |
+
"step": 3960
|
2783 |
+
},
|
2784 |
+
{
|
2785 |
+
"epoch": 22.178770949720672,
|
2786 |
+
"grad_norm": 0.1655452847480774,
|
2787 |
+
"learning_rate": 2.6229080771305293e-05,
|
2788 |
+
"loss": 0.0141,
|
2789 |
+
"step": 3970
|
2790 |
+
},
|
2791 |
+
{
|
2792 |
+
"epoch": 22.23463687150838,
|
2793 |
+
"grad_norm": 0.15178877115249634,
|
2794 |
+
"learning_rate": 2.5879098224978706e-05,
|
2795 |
+
"loss": 0.0181,
|
2796 |
+
"step": 3980
|
2797 |
+
},
|
2798 |
+
{
|
2799 |
+
"epoch": 22.29050279329609,
|
2800 |
+
"grad_norm": 0.12818969786167145,
|
2801 |
+
"learning_rate": 2.5530978855404448e-05,
|
2802 |
+
"loss": 0.0112,
|
2803 |
+
"step": 3990
|
2804 |
+
},
|
2805 |
+
{
|
2806 |
+
"epoch": 22.3463687150838,
|
2807 |
+
"grad_norm": 0.18601444363594055,
|
2808 |
+
"learning_rate": 2.5184735866898657e-05,
|
2809 |
+
"loss": 0.0131,
|
2810 |
+
"step": 4000
|
2811 |
+
},
|
2812 |
+
{
|
2813 |
+
"epoch": 22.40223463687151,
|
2814 |
+
"grad_norm": 0.2364937663078308,
|
2815 |
+
"learning_rate": 2.4840382392605515e-05,
|
2816 |
+
"loss": 0.0169,
|
2817 |
+
"step": 4010
|
2818 |
+
},
|
2819 |
+
{
|
2820 |
+
"epoch": 22.45810055865922,
|
2821 |
+
"grad_norm": 0.1969766914844513,
|
2822 |
+
"learning_rate": 2.449793149399922e-05,
|
2823 |
+
"loss": 0.0117,
|
2824 |
+
"step": 4020
|
2825 |
+
},
|
2826 |
+
{
|
2827 |
+
"epoch": 22.51396648044693,
|
2828 |
+
"grad_norm": 0.31158965826034546,
|
2829 |
+
"learning_rate": 2.4157396160388358e-05,
|
2830 |
+
"loss": 0.0164,
|
2831 |
+
"step": 4030
|
2832 |
+
},
|
2833 |
+
{
|
2834 |
+
"epoch": 22.56983240223464,
|
2835 |
+
"grad_norm": 0.17856691777706146,
|
2836 |
+
"learning_rate": 2.381878930842337e-05,
|
2837 |
+
"loss": 0.015,
|
2838 |
+
"step": 4040
|
2839 |
+
},
|
2840 |
+
{
|
2841 |
+
"epoch": 22.625698324022345,
|
2842 |
+
"grad_norm": 0.2636869251728058,
|
2843 |
+
"learning_rate": 2.3482123781606547e-05,
|
2844 |
+
"loss": 0.0185,
|
2845 |
+
"step": 4050
|
2846 |
+
},
|
2847 |
+
{
|
2848 |
+
"epoch": 22.681564245810055,
|
2849 |
+
"grad_norm": 0.12141893804073334,
|
2850 |
+
"learning_rate": 2.3147412349804862e-05,
|
2851 |
+
"loss": 0.014,
|
2852 |
+
"step": 4060
|
2853 |
+
},
|
2854 |
+
{
|
2855 |
+
"epoch": 22.737430167597765,
|
2856 |
+
"grad_norm": 0.22012671828269958,
|
2857 |
+
"learning_rate": 2.2814667708765745e-05,
|
2858 |
+
"loss": 0.0162,
|
2859 |
+
"step": 4070
|
2860 |
+
},
|
2861 |
+
{
|
2862 |
+
"epoch": 22.793296089385475,
|
2863 |
+
"grad_norm": 0.16044989228248596,
|
2864 |
+
"learning_rate": 2.2483902479635238e-05,
|
2865 |
+
"loss": 0.0131,
|
2866 |
+
"step": 4080
|
2867 |
+
},
|
2868 |
+
{
|
2869 |
+
"epoch": 22.849162011173185,
|
2870 |
+
"grad_norm": 0.15903018414974213,
|
2871 |
+
"learning_rate": 2.2155129208479606e-05,
|
2872 |
+
"loss": 0.0152,
|
2873 |
+
"step": 4090
|
2874 |
+
},
|
2875 |
+
{
|
2876 |
+
"epoch": 22.905027932960895,
|
2877 |
+
"grad_norm": 0.1807982474565506,
|
2878 |
+
"learning_rate": 2.182836036580923e-05,
|
2879 |
+
"loss": 0.0152,
|
2880 |
+
"step": 4100
|
2881 |
+
},
|
2882 |
+
{
|
2883 |
+
"epoch": 22.960893854748605,
|
2884 |
+
"grad_norm": 0.2643374502658844,
|
2885 |
+
"learning_rate": 2.1503608346105598e-05,
|
2886 |
+
"loss": 0.0171,
|
2887 |
+
"step": 4110
|
2888 |
+
},
|
2889 |
+
{
|
2890 |
+
"epoch": 23.01675977653631,
|
2891 |
+
"grad_norm": 0.13949626684188843,
|
2892 |
+
"learning_rate": 2.118088546735138e-05,
|
2893 |
+
"loss": 0.0132,
|
2894 |
+
"step": 4120
|
2895 |
+
},
|
2896 |
+
{
|
2897 |
+
"epoch": 23.07262569832402,
|
2898 |
+
"grad_norm": 0.6042327880859375,
|
2899 |
+
"learning_rate": 2.0860203970562954e-05,
|
2900 |
+
"loss": 0.0146,
|
2901 |
+
"step": 4130
|
2902 |
+
},
|
2903 |
+
{
|
2904 |
+
"epoch": 23.12849162011173,
|
2905 |
+
"grad_norm": 0.28848835825920105,
|
2906 |
+
"learning_rate": 2.054157601932631e-05,
|
2907 |
+
"loss": 0.0157,
|
2908 |
+
"step": 4140
|
2909 |
+
},
|
2910 |
+
{
|
2911 |
+
"epoch": 23.18435754189944,
|
2912 |
+
"grad_norm": 0.5369081497192383,
|
2913 |
+
"learning_rate": 2.0225013699335435e-05,
|
2914 |
+
"loss": 0.0128,
|
2915 |
+
"step": 4150
|
2916 |
+
},
|
2917 |
+
{
|
2918 |
+
"epoch": 23.24022346368715,
|
2919 |
+
"grad_norm": 0.15896014869213104,
|
2920 |
+
"learning_rate": 1.9910529017934196e-05,
|
2921 |
+
"loss": 0.0171,
|
2922 |
+
"step": 4160
|
2923 |
+
},
|
2924 |
+
{
|
2925 |
+
"epoch": 23.29608938547486,
|
2926 |
+
"grad_norm": 0.22632993757724762,
|
2927 |
+
"learning_rate": 1.9598133903660642e-05,
|
2928 |
+
"loss": 0.0117,
|
2929 |
+
"step": 4170
|
2930 |
+
},
|
2931 |
+
{
|
2932 |
+
"epoch": 23.35195530726257,
|
2933 |
+
"grad_norm": 0.14386065304279327,
|
2934 |
+
"learning_rate": 1.9287840205794693e-05,
|
2935 |
+
"loss": 0.0148,
|
2936 |
+
"step": 4180
|
2937 |
+
},
|
2938 |
+
{
|
2939 |
+
"epoch": 23.407821229050278,
|
2940 |
+
"grad_norm": 0.2815168797969818,
|
2941 |
+
"learning_rate": 1.8979659693908616e-05,
|
2942 |
+
"loss": 0.0143,
|
2943 |
+
"step": 4190
|
2944 |
+
},
|
2945 |
+
{
|
2946 |
+
"epoch": 23.463687150837988,
|
2947 |
+
"grad_norm": 0.17017313838005066,
|
2948 |
+
"learning_rate": 1.867360405742066e-05,
|
2949 |
+
"loss": 0.0126,
|
2950 |
+
"step": 4200
|
2951 |
+
},
|
2952 |
+
{
|
2953 |
+
"epoch": 23.519553072625698,
|
2954 |
+
"grad_norm": 0.2865433990955353,
|
2955 |
+
"learning_rate": 1.8369684905151677e-05,
|
2956 |
+
"loss": 0.0129,
|
2957 |
+
"step": 4210
|
2958 |
+
},
|
2959 |
+
{
|
2960 |
+
"epoch": 23.575418994413408,
|
2961 |
+
"grad_norm": 0.2120775431394577,
|
2962 |
+
"learning_rate": 1.806791376488467e-05,
|
2963 |
+
"loss": 0.0111,
|
2964 |
+
"step": 4220
|
2965 |
+
},
|
2966 |
+
{
|
2967 |
+
"epoch": 23.631284916201118,
|
2968 |
+
"grad_norm": 0.2156466841697693,
|
2969 |
+
"learning_rate": 1.7768302082927673e-05,
|
2970 |
+
"loss": 0.0134,
|
2971 |
+
"step": 4230
|
2972 |
+
},
|
2973 |
+
{
|
2974 |
+
"epoch": 23.687150837988828,
|
2975 |
+
"grad_norm": 0.13205291330814362,
|
2976 |
+
"learning_rate": 1.74708612236796e-05,
|
2977 |
+
"loss": 0.0116,
|
2978 |
+
"step": 4240
|
2979 |
+
},
|
2980 |
+
{
|
2981 |
+
"epoch": 23.743016759776538,
|
2982 |
+
"grad_norm": 0.1907736212015152,
|
2983 |
+
"learning_rate": 1.7175602469199075e-05,
|
2984 |
+
"loss": 0.0107,
|
2985 |
+
"step": 4250
|
2986 |
+
},
|
2987 |
+
{
|
2988 |
+
"epoch": 23.798882681564244,
|
2989 |
+
"grad_norm": 0.19241803884506226,
|
2990 |
+
"learning_rate": 1.6882537018776567e-05,
|
2991 |
+
"loss": 0.0136,
|
2992 |
+
"step": 4260
|
2993 |
+
},
|
2994 |
+
{
|
2995 |
+
"epoch": 23.854748603351954,
|
2996 |
+
"grad_norm": 0.26086995005607605,
|
2997 |
+
"learning_rate": 1.6591675988509588e-05,
|
2998 |
+
"loss": 0.0169,
|
2999 |
+
"step": 4270
|
3000 |
+
},
|
3001 |
+
{
|
3002 |
+
"epoch": 23.910614525139664,
|
3003 |
+
"grad_norm": 0.1134791448712349,
|
3004 |
+
"learning_rate": 1.6303030410881034e-05,
|
3005 |
+
"loss": 0.0126,
|
3006 |
+
"step": 4280
|
3007 |
+
},
|
3008 |
+
{
|
3009 |
+
"epoch": 23.966480446927374,
|
3010 |
+
"grad_norm": 0.19399867951869965,
|
3011 |
+
"learning_rate": 1.601661123434082e-05,
|
3012 |
+
"loss": 0.0146,
|
3013 |
+
"step": 4290
|
3014 |
+
},
|
3015 |
+
{
|
3016 |
+
"epoch": 24.022346368715084,
|
3017 |
+
"grad_norm": 0.218746617436409,
|
3018 |
+
"learning_rate": 1.5732429322890384e-05,
|
3019 |
+
"loss": 0.0196,
|
3020 |
+
"step": 4300
|
3021 |
+
},
|
3022 |
+
{
|
3023 |
+
"epoch": 24.078212290502794,
|
3024 |
+
"grad_norm": 0.173094242811203,
|
3025 |
+
"learning_rate": 1.5450495455670868e-05,
|
3026 |
+
"loss": 0.0131,
|
3027 |
+
"step": 4310
|
3028 |
+
},
|
3029 |
+
{
|
3030 |
+
"epoch": 24.134078212290504,
|
3031 |
+
"grad_norm": 0.28099843859672546,
|
3032 |
+
"learning_rate": 1.5170820326554087e-05,
|
3033 |
+
"loss": 0.0125,
|
3034 |
+
"step": 4320
|
3035 |
+
},
|
3036 |
+
{
|
3037 |
+
"epoch": 24.189944134078214,
|
3038 |
+
"grad_norm": 0.2801670432090759,
|
3039 |
+
"learning_rate": 1.489341454373694e-05,
|
3040 |
+
"loss": 0.0129,
|
3041 |
+
"step": 4330
|
3042 |
+
},
|
3043 |
+
{
|
3044 |
+
"epoch": 24.24581005586592,
|
3045 |
+
"grad_norm": 0.21972441673278809,
|
3046 |
+
"learning_rate": 1.4618288629339079e-05,
|
3047 |
+
"loss": 0.016,
|
3048 |
+
"step": 4340
|
3049 |
+
},
|
3050 |
+
{
|
3051 |
+
"epoch": 24.30167597765363,
|
3052 |
+
"grad_norm": 0.132466122508049,
|
3053 |
+
"learning_rate": 1.4345453019003745e-05,
|
3054 |
+
"loss": 0.0134,
|
3055 |
+
"step": 4350
|
3056 |
+
},
|
3057 |
+
{
|
3058 |
+
"epoch": 24.35754189944134,
|
3059 |
+
"grad_norm": 0.1710919737815857,
|
3060 |
+
"learning_rate": 1.4074918061502045e-05,
|
3061 |
+
"loss": 0.0131,
|
3062 |
+
"step": 4360
|
3063 |
+
},
|
3064 |
+
{
|
3065 |
+
"epoch": 24.41340782122905,
|
3066 |
+
"grad_norm": 0.2302296757698059,
|
3067 |
+
"learning_rate": 1.3806694018340223e-05,
|
3068 |
+
"loss": 0.0138,
|
3069 |
+
"step": 4370
|
3070 |
+
},
|
3071 |
+
{
|
3072 |
+
"epoch": 24.46927374301676,
|
3073 |
+
"grad_norm": 0.34001755714416504,
|
3074 |
+
"learning_rate": 1.3540791063370673e-05,
|
3075 |
+
"loss": 0.0132,
|
3076 |
+
"step": 4380
|
3077 |
+
},
|
3078 |
+
{
|
3079 |
+
"epoch": 24.52513966480447,
|
3080 |
+
"grad_norm": 0.18118420243263245,
|
3081 |
+
"learning_rate": 1.3277219282405858e-05,
|
3082 |
+
"loss": 0.0114,
|
3083 |
+
"step": 4390
|
3084 |
+
},
|
3085 |
+
{
|
3086 |
+
"epoch": 24.58100558659218,
|
3087 |
+
"grad_norm": 0.2221328318119049,
|
3088 |
+
"learning_rate": 1.3015988672835823e-05,
|
3089 |
+
"loss": 0.0135,
|
3090 |
+
"step": 4400
|
3091 |
+
},
|
3092 |
+
{
|
3093 |
+
"epoch": 24.636871508379887,
|
3094 |
+
"grad_norm": 0.2257193773984909,
|
3095 |
+
"learning_rate": 1.2757109143249002e-05,
|
3096 |
+
"loss": 0.0155,
|
3097 |
+
"step": 4410
|
3098 |
+
},
|
3099 |
+
{
|
3100 |
+
"epoch": 24.692737430167597,
|
3101 |
+
"grad_norm": 0.3096782863140106,
|
3102 |
+
"learning_rate": 1.2500590513056333e-05,
|
3103 |
+
"loss": 0.0099,
|
3104 |
+
"step": 4420
|
3105 |
+
},
|
3106 |
+
{
|
3107 |
+
"epoch": 24.748603351955307,
|
3108 |
+
"grad_norm": 0.20636042952537537,
|
3109 |
+
"learning_rate": 1.2246442512118886e-05,
|
3110 |
+
"loss": 0.0124,
|
3111 |
+
"step": 4430
|
3112 |
+
},
|
3113 |
+
{
|
3114 |
+
"epoch": 24.804469273743017,
|
3115 |
+
"grad_norm": 0.24674853682518005,
|
3116 |
+
"learning_rate": 1.199467478037871e-05,
|
3117 |
+
"loss": 0.0134,
|
3118 |
+
"step": 4440
|
3119 |
+
},
|
3120 |
+
{
|
3121 |
+
"epoch": 24.860335195530727,
|
3122 |
+
"grad_norm": 0.3628787100315094,
|
3123 |
+
"learning_rate": 1.1745296867493226e-05,
|
3124 |
+
"loss": 0.0168,
|
3125 |
+
"step": 4450
|
3126 |
+
},
|
3127 |
+
{
|
3128 |
+
"epoch": 24.916201117318437,
|
3129 |
+
"grad_norm": 0.17266486585140228,
|
3130 |
+
"learning_rate": 1.1498318232473033e-05,
|
3131 |
+
"loss": 0.0112,
|
3132 |
+
"step": 4460
|
3133 |
+
},
|
3134 |
+
{
|
3135 |
+
"epoch": 24.972067039106147,
|
3136 |
+
"grad_norm": 0.20120470225811005,
|
3137 |
+
"learning_rate": 1.1253748243323055e-05,
|
3138 |
+
"loss": 0.0126,
|
3139 |
+
"step": 4470
|
3140 |
+
},
|
3141 |
+
{
|
3142 |
+
"epoch": 25.027932960893853,
|
3143 |
+
"grad_norm": 0.30043306946754456,
|
3144 |
+
"learning_rate": 1.1011596176687343e-05,
|
3145 |
+
"loss": 0.011,
|
3146 |
+
"step": 4480
|
3147 |
+
},
|
3148 |
+
{
|
3149 |
+
"epoch": 25.083798882681563,
|
3150 |
+
"grad_norm": 0.23731638491153717,
|
3151 |
+
"learning_rate": 1.0771871217496971e-05,
|
3152 |
+
"loss": 0.0164,
|
3153 |
+
"step": 4490
|
3154 |
+
},
|
3155 |
+
{
|
3156 |
+
"epoch": 25.139664804469273,
|
3157 |
+
"grad_norm": 0.1487792432308197,
|
3158 |
+
"learning_rate": 1.0534582458621932e-05,
|
3159 |
+
"loss": 0.0171,
|
3160 |
+
"step": 4500
|
3161 |
+
},
|
3162 |
+
{
|
3163 |
+
"epoch": 25.195530726256983,
|
3164 |
+
"grad_norm": 0.16674961149692535,
|
3165 |
+
"learning_rate": 1.0299738900526022e-05,
|
3166 |
+
"loss": 0.0118,
|
3167 |
+
"step": 4510
|
3168 |
+
},
|
3169 |
+
{
|
3170 |
+
"epoch": 25.251396648044693,
|
3171 |
+
"grad_norm": 0.1953108161687851,
|
3172 |
+
"learning_rate": 1.0067349450925548e-05,
|
3173 |
+
"loss": 0.0141,
|
3174 |
+
"step": 4520
|
3175 |
+
},
|
3176 |
+
{
|
3177 |
+
"epoch": 25.307262569832403,
|
3178 |
+
"grad_norm": 0.2785939872264862,
|
3179 |
+
"learning_rate": 9.837422924451454e-06,
|
3180 |
+
"loss": 0.013,
|
3181 |
+
"step": 4530
|
3182 |
+
},
|
3183 |
+
{
|
3184 |
+
"epoch": 25.363128491620113,
|
3185 |
+
"grad_norm": 0.18461591005325317,
|
3186 |
+
"learning_rate": 9.609968042314908e-06,
|
3187 |
+
"loss": 0.0101,
|
3188 |
+
"step": 4540
|
3189 |
+
},
|
3190 |
+
{
|
3191 |
+
"epoch": 25.41899441340782,
|
3192 |
+
"grad_norm": 0.195012629032135,
|
3193 |
+
"learning_rate": 9.384993431976647e-06,
|
3194 |
+
"loss": 0.0142,
|
3195 |
+
"step": 4550
|
3196 |
+
},
|
3197 |
+
{
|
3198 |
+
"epoch": 25.47486033519553,
|
3199 |
+
"grad_norm": 0.11832192540168762,
|
3200 |
+
"learning_rate": 9.162507626819516e-06,
|
3201 |
+
"loss": 0.0118,
|
3202 |
+
"step": 4560
|
3203 |
+
},
|
3204 |
+
{
|
3205 |
+
"epoch": 25.53072625698324,
|
3206 |
+
"grad_norm": 0.24214394390583038,
|
3207 |
+
"learning_rate": 8.942519065825041e-06,
|
3208 |
+
"loss": 0.0159,
|
3209 |
+
"step": 4570
|
3210 |
+
},
|
3211 |
+
{
|
3212 |
+
"epoch": 25.58659217877095,
|
3213 |
+
"grad_norm": 0.18261413276195526,
|
3214 |
+
"learning_rate": 8.725036093253124e-06,
|
3215 |
+
"loss": 0.011,
|
3216 |
+
"step": 4580
|
3217 |
+
},
|
3218 |
+
{
|
3219 |
+
"epoch": 25.64245810055866,
|
3220 |
+
"grad_norm": 0.1193467229604721,
|
3221 |
+
"learning_rate": 8.510066958325679e-06,
|
3222 |
+
"loss": 0.0109,
|
3223 |
+
"step": 4590
|
3224 |
+
},
|
3225 |
+
{
|
3226 |
+
"epoch": 25.69832402234637,
|
3227 |
+
"grad_norm": 0.25043681263923645,
|
3228 |
+
"learning_rate": 8.297619814913667e-06,
|
3229 |
+
"loss": 0.0124,
|
3230 |
+
"step": 4600
|
3231 |
+
},
|
3232 |
+
{
|
3233 |
+
"epoch": 25.75418994413408,
|
3234 |
+
"grad_norm": 0.17554740607738495,
|
3235 |
+
"learning_rate": 8.087702721227812e-06,
|
3236 |
+
"loss": 0.017,
|
3237 |
+
"step": 4610
|
3238 |
+
},
|
3239 |
+
{
|
3240 |
+
"epoch": 25.810055865921786,
|
3241 |
+
"grad_norm": 0.13422702252864838,
|
3242 |
+
"learning_rate": 7.880323639513029e-06,
|
3243 |
+
"loss": 0.0122,
|
3244 |
+
"step": 4620
|
3245 |
+
},
|
3246 |
+
{
|
3247 |
+
"epoch": 25.865921787709496,
|
3248 |
+
"grad_norm": 0.16326576471328735,
|
3249 |
+
"learning_rate": 7.675490435746326e-06,
|
3250 |
+
"loss": 0.0106,
|
3251 |
+
"step": 4630
|
3252 |
+
},
|
3253 |
+
{
|
3254 |
+
"epoch": 25.921787709497206,
|
3255 |
+
"grad_norm": 0.17265120148658752,
|
3256 |
+
"learning_rate": 7.473210879338462e-06,
|
3257 |
+
"loss": 0.0123,
|
3258 |
+
"step": 4640
|
3259 |
+
},
|
3260 |
+
{
|
3261 |
+
"epoch": 25.977653631284916,
|
3262 |
+
"grad_norm": 0.18150538206100464,
|
3263 |
+
"learning_rate": 7.273492642839273e-06,
|
3264 |
+
"loss": 0.0117,
|
3265 |
+
"step": 4650
|
3266 |
+
},
|
3267 |
+
{
|
3268 |
+
"epoch": 26.033519553072626,
|
3269 |
+
"grad_norm": 0.2668885290622711,
|
3270 |
+
"learning_rate": 7.0763433016466535e-06,
|
3271 |
+
"loss": 0.0125,
|
3272 |
+
"step": 4660
|
3273 |
+
},
|
3274 |
+
{
|
3275 |
+
"epoch": 26.089385474860336,
|
3276 |
+
"grad_norm": 0.20581719279289246,
|
3277 |
+
"learning_rate": 6.881770333719177e-06,
|
3278 |
+
"loss": 0.0148,
|
3279 |
+
"step": 4670
|
3280 |
+
},
|
3281 |
+
{
|
3282 |
+
"epoch": 26.145251396648046,
|
3283 |
+
"grad_norm": 0.15722183883190155,
|
3284 |
+
"learning_rate": 6.689781119292487e-06,
|
3285 |
+
"loss": 0.0124,
|
3286 |
+
"step": 4680
|
3287 |
+
},
|
3288 |
+
{
|
3289 |
+
"epoch": 26.201117318435756,
|
3290 |
+
"grad_norm": 0.1610933542251587,
|
3291 |
+
"learning_rate": 6.5003829405993885e-06,
|
3292 |
+
"loss": 0.011,
|
3293 |
+
"step": 4690
|
3294 |
+
},
|
3295 |
+
{
|
3296 |
+
"epoch": 26.256983240223462,
|
3297 |
+
"grad_norm": 0.2166851907968521,
|
3298 |
+
"learning_rate": 6.3135829815935736e-06,
|
3299 |
+
"loss": 0.0135,
|
3300 |
+
"step": 4700
|
3301 |
+
},
|
3302 |
+
{
|
3303 |
+
"epoch": 26.312849162011172,
|
3304 |
+
"grad_norm": 0.21534748375415802,
|
3305 |
+
"learning_rate": 6.1293883276770985e-06,
|
3306 |
+
"loss": 0.0092,
|
3307 |
+
"step": 4710
|
3308 |
+
},
|
3309 |
+
{
|
3310 |
+
"epoch": 26.368715083798882,
|
3311 |
+
"grad_norm": 0.1989695280790329,
|
3312 |
+
"learning_rate": 5.947805965431792e-06,
|
3313 |
+
"loss": 0.0127,
|
3314 |
+
"step": 4720
|
3315 |
+
},
|
3316 |
+
{
|
3317 |
+
"epoch": 26.424581005586592,
|
3318 |
+
"grad_norm": 0.44271120429039,
|
3319 |
+
"learning_rate": 5.768842782354055e-06,
|
3320 |
+
"loss": 0.01,
|
3321 |
+
"step": 4730
|
3322 |
+
},
|
3323 |
+
{
|
3324 |
+
"epoch": 26.480446927374302,
|
3325 |
+
"grad_norm": 0.17514117062091827,
|
3326 |
+
"learning_rate": 5.592505566593786e-06,
|
3327 |
+
"loss": 0.0106,
|
3328 |
+
"step": 4740
|
3329 |
+
},
|
3330 |
+
{
|
3331 |
+
"epoch": 26.536312849162012,
|
3332 |
+
"grad_norm": 0.15244068205356598,
|
3333 |
+
"learning_rate": 5.418801006696736e-06,
|
3334 |
+
"loss": 0.0124,
|
3335 |
+
"step": 4750
|
3336 |
+
},
|
3337 |
+
{
|
3338 |
+
"epoch": 26.592178770949722,
|
3339 |
+
"grad_norm": 0.3087766468524933,
|
3340 |
+
"learning_rate": 5.247735691350941e-06,
|
3341 |
+
"loss": 0.0099,
|
3342 |
+
"step": 4760
|
3343 |
+
},
|
3344 |
+
{
|
3345 |
+
"epoch": 26.64804469273743,
|
3346 |
+
"grad_norm": 0.12699635326862335,
|
3347 |
+
"learning_rate": 5.079316109136808e-06,
|
3348 |
+
"loss": 0.0116,
|
3349 |
+
"step": 4770
|
3350 |
+
},
|
3351 |
+
{
|
3352 |
+
"epoch": 26.70391061452514,
|
3353 |
+
"grad_norm": 0.14864447712898254,
|
3354 |
+
"learning_rate": 4.9135486482808515e-06,
|
3355 |
+
"loss": 0.0137,
|
3356 |
+
"step": 4780
|
3357 |
+
},
|
3358 |
+
{
|
3359 |
+
"epoch": 26.75977653631285,
|
3360 |
+
"grad_norm": 0.14272750914096832,
|
3361 |
+
"learning_rate": 4.750439596413605e-06,
|
3362 |
+
"loss": 0.011,
|
3363 |
+
"step": 4790
|
3364 |
+
},
|
3365 |
+
{
|
3366 |
+
"epoch": 26.81564245810056,
|
3367 |
+
"grad_norm": 0.18606601655483246,
|
3368 |
+
"learning_rate": 4.589995140330952e-06,
|
3369 |
+
"loss": 0.0127,
|
3370 |
+
"step": 4800
|
3371 |
+
},
|
3372 |
+
{
|
3373 |
+
"epoch": 26.87150837988827,
|
3374 |
+
"grad_norm": 0.1599322408437729,
|
3375 |
+
"learning_rate": 4.432221365759523e-06,
|
3376 |
+
"loss": 0.0102,
|
3377 |
+
"step": 4810
|
3378 |
+
},
|
3379 |
+
{
|
3380 |
+
"epoch": 26.92737430167598,
|
3381 |
+
"grad_norm": 0.2554507553577423,
|
3382 |
+
"learning_rate": 4.277124257125872e-06,
|
3383 |
+
"loss": 0.0145,
|
3384 |
+
"step": 4820
|
3385 |
+
},
|
3386 |
+
{
|
3387 |
+
"epoch": 26.98324022346369,
|
3388 |
+
"grad_norm": 0.15352533757686615,
|
3389 |
+
"learning_rate": 4.124709697329476e-06,
|
3390 |
+
"loss": 0.0137,
|
3391 |
+
"step": 4830
|
3392 |
+
},
|
3393 |
+
{
|
3394 |
+
"epoch": 27.039106145251395,
|
3395 |
+
"grad_norm": 0.10090656578540802,
|
3396 |
+
"learning_rate": 3.974983467519593e-06,
|
3397 |
+
"loss": 0.0099,
|
3398 |
+
"step": 4840
|
3399 |
+
},
|
3400 |
+
{
|
3401 |
+
"epoch": 27.094972067039105,
|
3402 |
+
"grad_norm": 0.203408882021904,
|
3403 |
+
"learning_rate": 3.8279512468759684e-06,
|
3404 |
+
"loss": 0.0133,
|
3405 |
+
"step": 4850
|
3406 |
+
},
|
3407 |
+
{
|
3408 |
+
"epoch": 27.150837988826815,
|
3409 |
+
"grad_norm": 0.0952206403017044,
|
3410 |
+
"learning_rate": 3.683618612393441e-06,
|
3411 |
+
"loss": 0.0127,
|
3412 |
+
"step": 4860
|
3413 |
+
},
|
3414 |
+
{
|
3415 |
+
"epoch": 27.206703910614525,
|
3416 |
+
"grad_norm": 0.09654044359922409,
|
3417 |
+
"learning_rate": 3.5419910386703694e-06,
|
3418 |
+
"loss": 0.0092,
|
3419 |
+
"step": 4870
|
3420 |
+
},
|
3421 |
+
{
|
3422 |
+
"epoch": 27.262569832402235,
|
3423 |
+
"grad_norm": 0.42048877477645874,
|
3424 |
+
"learning_rate": 3.403073897701028e-06,
|
3425 |
+
"loss": 0.0126,
|
3426 |
+
"step": 4880
|
3427 |
+
},
|
3428 |
+
{
|
3429 |
+
"epoch": 27.318435754189945,
|
3430 |
+
"grad_norm": 0.16950425505638123,
|
3431 |
+
"learning_rate": 3.266872458671846e-06,
|
3432 |
+
"loss": 0.0112,
|
3433 |
+
"step": 4890
|
3434 |
+
},
|
3435 |
+
{
|
3436 |
+
"epoch": 27.374301675977655,
|
3437 |
+
"grad_norm": 0.2811785042285919,
|
3438 |
+
"learning_rate": 3.133391887761441e-06,
|
3439 |
+
"loss": 0.0113,
|
3440 |
+
"step": 4900
|
3441 |
+
},
|
3442 |
+
{
|
3443 |
+
"epoch": 27.43016759776536,
|
3444 |
+
"grad_norm": 0.20062536001205444,
|
3445 |
+
"learning_rate": 3.002637247944828e-06,
|
3446 |
+
"loss": 0.0101,
|
3447 |
+
"step": 4910
|
3448 |
+
},
|
3449 |
+
{
|
3450 |
+
"epoch": 27.48603351955307,
|
3451 |
+
"grad_norm": 0.176531121134758,
|
3452 |
+
"learning_rate": 2.8746134988012453e-06,
|
3453 |
+
"loss": 0.0122,
|
3454 |
+
"step": 4920
|
3455 |
+
},
|
3456 |
+
{
|
3457 |
+
"epoch": 27.54189944134078,
|
3458 |
+
"grad_norm": 0.2781747877597809,
|
3459 |
+
"learning_rate": 2.749325496326099e-06,
|
3460 |
+
"loss": 0.0114,
|
3461 |
+
"step": 4930
|
3462 |
+
},
|
3463 |
+
{
|
3464 |
+
"epoch": 27.59776536312849,
|
3465 |
+
"grad_norm": 0.23316988348960876,
|
3466 |
+
"learning_rate": 2.6267779927467423e-06,
|
3467 |
+
"loss": 0.009,
|
3468 |
+
"step": 4940
|
3469 |
+
},
|
3470 |
+
{
|
3471 |
+
"epoch": 27.6536312849162,
|
3472 |
+
"grad_norm": 0.17585094273090363,
|
3473 |
+
"learning_rate": 2.5069756363422454e-06,
|
3474 |
+
"loss": 0.0131,
|
3475 |
+
"step": 4950
|
3476 |
+
},
|
3477 |
+
{
|
3478 |
+
"epoch": 27.70949720670391,
|
3479 |
+
"grad_norm": 0.22221805155277252,
|
3480 |
+
"learning_rate": 2.3899229712671027e-06,
|
3481 |
+
"loss": 0.0129,
|
3482 |
+
"step": 4960
|
3483 |
+
},
|
3484 |
+
{
|
3485 |
+
"epoch": 27.76536312849162,
|
3486 |
+
"grad_norm": 0.18307988345623016,
|
3487 |
+
"learning_rate": 2.2756244373787867e-06,
|
3488 |
+
"loss": 0.0099,
|
3489 |
+
"step": 4970
|
3490 |
+
},
|
3491 |
+
{
|
3492 |
+
"epoch": 27.821229050279328,
|
3493 |
+
"grad_norm": 0.32926902174949646,
|
3494 |
+
"learning_rate": 2.164084370069452e-06,
|
3495 |
+
"loss": 0.0106,
|
3496 |
+
"step": 4980
|
3497 |
+
},
|
3498 |
+
{
|
3499 |
+
"epoch": 27.877094972067038,
|
3500 |
+
"grad_norm": 0.1888887584209442,
|
3501 |
+
"learning_rate": 2.055307000101414e-06,
|
3502 |
+
"loss": 0.0122,
|
3503 |
+
"step": 4990
|
3504 |
+
},
|
3505 |
+
{
|
3506 |
+
"epoch": 27.932960893854748,
|
3507 |
+
"grad_norm": 0.13414178788661957,
|
3508 |
+
"learning_rate": 1.9492964534467026e-06,
|
3509 |
+
"loss": 0.0109,
|
3510 |
+
"step": 5000
|
3511 |
+
},
|
3512 |
+
{
|
3513 |
+
"epoch": 27.988826815642458,
|
3514 |
+
"grad_norm": 0.27602723240852356,
|
3515 |
+
"learning_rate": 1.8460567511305645e-06,
|
3516 |
+
"loss": 0.0125,
|
3517 |
+
"step": 5010
|
3518 |
+
},
|
3519 |
+
{
|
3520 |
+
"epoch": 28.044692737430168,
|
3521 |
+
"grad_norm": 0.1933746188879013,
|
3522 |
+
"learning_rate": 1.745591809078925e-06,
|
3523 |
+
"loss": 0.0105,
|
3524 |
+
"step": 5020
|
3525 |
+
},
|
3526 |
+
{
|
3527 |
+
"epoch": 28.100558659217878,
|
3528 |
+
"grad_norm": 0.09113660454750061,
|
3529 |
+
"learning_rate": 1.6479054379698985e-06,
|
3530 |
+
"loss": 0.0095,
|
3531 |
+
"step": 5030
|
3532 |
+
},
|
3533 |
+
{
|
3534 |
+
"epoch": 28.156424581005588,
|
3535 |
+
"grad_norm": 0.11389518529176712,
|
3536 |
+
"learning_rate": 1.5530013430891886e-06,
|
3537 |
+
"loss": 0.0123,
|
3538 |
+
"step": 5040
|
3539 |
+
},
|
3540 |
+
{
|
3541 |
+
"epoch": 28.212290502793294,
|
3542 |
+
"grad_norm": 0.12854307889938354,
|
3543 |
+
"learning_rate": 1.4608831241895823e-06,
|
3544 |
+
"loss": 0.0111,
|
3545 |
+
"step": 5050
|
3546 |
+
},
|
3547 |
+
{
|
3548 |
+
"epoch": 28.268156424581004,
|
3549 |
+
"grad_norm": 0.1454714685678482,
|
3550 |
+
"learning_rate": 1.3715542753544195e-06,
|
3551 |
+
"loss": 0.0116,
|
3552 |
+
"step": 5060
|
3553 |
+
},
|
3554 |
+
{
|
3555 |
+
"epoch": 28.324022346368714,
|
3556 |
+
"grad_norm": 0.11753421276807785,
|
3557 |
+
"learning_rate": 1.2850181848650304e-06,
|
3558 |
+
"loss": 0.0085,
|
3559 |
+
"step": 5070
|
3560 |
+
},
|
3561 |
+
{
|
3562 |
+
"epoch": 28.379888268156424,
|
3563 |
+
"grad_norm": 0.19782139360904694,
|
3564 |
+
"learning_rate": 1.2012781350722145e-06,
|
3565 |
+
"loss": 0.0099,
|
3566 |
+
"step": 5080
|
3567 |
+
},
|
3568 |
+
{
|
3569 |
+
"epoch": 28.435754189944134,
|
3570 |
+
"grad_norm": 0.1429385244846344,
|
3571 |
+
"learning_rate": 1.1203373022717899e-06,
|
3572 |
+
"loss": 0.0096,
|
3573 |
+
"step": 5090
|
3574 |
+
},
|
3575 |
+
{
|
3576 |
+
"epoch": 28.491620111731844,
|
3577 |
+
"grad_norm": 0.17323465645313263,
|
3578 |
+
"learning_rate": 1.0421987565840562e-06,
|
3579 |
+
"loss": 0.0107,
|
3580 |
+
"step": 5100
|
3581 |
+
},
|
3582 |
+
{
|
3583 |
+
"epoch": 28.547486033519554,
|
3584 |
+
"grad_norm": 0.1123204305768013,
|
3585 |
+
"learning_rate": 9.66865461837371e-07,
|
3586 |
+
"loss": 0.0139,
|
3587 |
+
"step": 5110
|
3588 |
+
},
|
3589 |
+
{
|
3590 |
+
"epoch": 28.60335195530726,
|
3591 |
+
"grad_norm": 0.23444265127182007,
|
3592 |
+
"learning_rate": 8.943402754557405e-07,
|
3593 |
+
"loss": 0.0148,
|
3594 |
+
"step": 5120
|
3595 |
+
},
|
3596 |
+
{
|
3597 |
+
"epoch": 28.65921787709497,
|
3598 |
+
"grad_norm": 0.10738368332386017,
|
3599 |
+
"learning_rate": 8.246259483504047e-07,
|
3600 |
+
"loss": 0.0115,
|
3601 |
+
"step": 5130
|
3602 |
+
},
|
3603 |
+
{
|
3604 |
+
"epoch": 28.71508379888268,
|
3605 |
+
"grad_norm": 0.1469849944114685,
|
3606 |
+
"learning_rate": 7.577251248155142e-07,
|
3607 |
+
"loss": 0.0092,
|
3608 |
+
"step": 5140
|
3609 |
+
},
|
3610 |
+
{
|
3611 |
+
"epoch": 28.77094972067039,
|
3612 |
+
"grad_norm": 0.1442977786064148,
|
3613 |
+
"learning_rate": 6.936403424278591e-07,
|
3614 |
+
"loss": 0.0126,
|
3615 |
+
"step": 5150
|
3616 |
+
},
|
3617 |
+
{
|
3618 |
+
"epoch": 28.8268156424581,
|
3619 |
+
"grad_norm": 0.11567971855401993,
|
3620 |
+
"learning_rate": 6.323740319505388e-07,
|
3621 |
+
"loss": 0.0108,
|
3622 |
+
"step": 5160
|
3623 |
+
},
|
3624 |
+
{
|
3625 |
+
"epoch": 28.88268156424581,
|
3626 |
+
"grad_norm": 0.12551957368850708,
|
3627 |
+
"learning_rate": 5.739285172408598e-07,
|
3628 |
+
"loss": 0.0129,
|
3629 |
+
"step": 5170
|
3630 |
+
},
|
3631 |
+
{
|
3632 |
+
"epoch": 28.93854748603352,
|
3633 |
+
"grad_norm": 0.16800257563591003,
|
3634 |
+
"learning_rate": 5.183060151620983e-07,
|
3635 |
+
"loss": 0.0124,
|
3636 |
+
"step": 5180
|
3637 |
+
},
|
3638 |
+
{
|
3639 |
+
"epoch": 28.994413407821227,
|
3640 |
+
"grad_norm": 0.17249980568885803,
|
3641 |
+
"learning_rate": 4.655086354994925e-07,
|
3642 |
+
"loss": 0.0118,
|
3643 |
+
"step": 5190
|
3644 |
+
},
|
3645 |
+
{
|
3646 |
+
"epoch": 29.050279329608937,
|
3647 |
+
"grad_norm": 0.3741263151168823,
|
3648 |
+
"learning_rate": 4.155383808801649e-07,
|
3649 |
+
"loss": 0.0147,
|
3650 |
+
"step": 5200
|
3651 |
+
},
|
3652 |
+
{
|
3653 |
+
"epoch": 29.106145251396647,
|
3654 |
+
"grad_norm": 0.22194480895996094,
|
3655 |
+
"learning_rate": 3.68397146697183e-07,
|
3656 |
+
"loss": 0.0138,
|
3657 |
+
"step": 5210
|
3658 |
+
},
|
3659 |
+
{
|
3660 |
+
"epoch": 29.162011173184357,
|
3661 |
+
"grad_norm": 0.2826599180698395,
|
3662 |
+
"learning_rate": 3.240867210376752e-07,
|
3663 |
+
"loss": 0.0145,
|
3664 |
+
"step": 5220
|
3665 |
+
},
|
3666 |
+
{
|
3667 |
+
"epoch": 29.217877094972067,
|
3668 |
+
"grad_norm": 0.21220660209655762,
|
3669 |
+
"learning_rate": 2.8260878461495995e-07,
|
3670 |
+
"loss": 0.0153,
|
3671 |
+
"step": 5230
|
3672 |
+
},
|
3673 |
+
{
|
3674 |
+
"epoch": 29.273743016759777,
|
3675 |
+
"grad_norm": 0.2775328755378723,
|
3676 |
+
"learning_rate": 2.439649107048719e-07,
|
3677 |
+
"loss": 0.0112,
|
3678 |
+
"step": 5240
|
3679 |
+
},
|
3680 |
+
{
|
3681 |
+
"epoch": 29.329608938547487,
|
3682 |
+
"grad_norm": 0.2872767150402069,
|
3683 |
+
"learning_rate": 2.0815656508601785e-07,
|
3684 |
+
"loss": 0.0114,
|
3685 |
+
"step": 5250
|
3686 |
+
},
|
3687 |
+
{
|
3688 |
+
"epoch": 29.385474860335197,
|
3689 |
+
"grad_norm": 0.2828880250453949,
|
3690 |
+
"learning_rate": 1.7518510598422963e-07,
|
3691 |
+
"loss": 0.0094,
|
3692 |
+
"step": 5260
|
3693 |
+
},
|
3694 |
+
{
|
3695 |
+
"epoch": 29.441340782122904,
|
3696 |
+
"grad_norm": 0.27305498719215393,
|
3697 |
+
"learning_rate": 1.4505178402099692e-07,
|
3698 |
+
"loss": 0.0119,
|
3699 |
+
"step": 5270
|
3700 |
+
},
|
3701 |
+
{
|
3702 |
+
"epoch": 29.497206703910614,
|
3703 |
+
"grad_norm": 0.1336456537246704,
|
3704 |
+
"learning_rate": 1.1775774216608858e-07,
|
3705 |
+
"loss": 0.0082,
|
3706 |
+
"step": 5280
|
3707 |
+
},
|
3708 |
+
{
|
3709 |
+
"epoch": 29.553072625698324,
|
3710 |
+
"grad_norm": 0.12913921475410461,
|
3711 |
+
"learning_rate": 9.330401569413725e-08,
|
3712 |
+
"loss": 0.0094,
|
3713 |
+
"step": 5290
|
3714 |
+
},
|
3715 |
+
{
|
3716 |
+
"epoch": 29.608938547486034,
|
3717 |
+
"grad_norm": 0.193023219704628,
|
3718 |
+
"learning_rate": 7.169153214543754e-08,
|
3719 |
+
"loss": 0.0094,
|
3720 |
+
"step": 5300
|
3721 |
+
},
|
3722 |
+
{
|
3723 |
+
"epoch": 29.664804469273744,
|
3724 |
+
"grad_norm": 0.1686532348394394,
|
3725 |
+
"learning_rate": 5.292111129068244e-08,
|
3726 |
+
"loss": 0.0113,
|
3727 |
+
"step": 5310
|
3728 |
+
},
|
3729 |
+
{
|
3730 |
+
"epoch": 29.720670391061454,
|
3731 |
+
"grad_norm": 0.18468423187732697,
|
3732 |
+
"learning_rate": 3.69934650999465e-08,
|
3733 |
+
"loss": 0.0115,
|
3734 |
+
"step": 5320
|
3735 |
+
},
|
3736 |
+
{
|
3737 |
+
"epoch": 29.776536312849164,
|
3738 |
+
"grad_norm": 0.21080362796783447,
|
3739 |
+
"learning_rate": 2.3909197715657468e-08,
|
3740 |
+
"loss": 0.0126,
|
3741 |
+
"step": 5330
|
3742 |
+
},
|
3743 |
+
{
|
3744 |
+
"epoch": 29.83240223463687,
|
3745 |
+
"grad_norm": 0.15631383657455444,
|
3746 |
+
"learning_rate": 1.3668805429639662e-08,
|
3747 |
+
"loss": 0.0107,
|
3748 |
+
"step": 5340
|
3749 |
+
},
|
3750 |
+
{
|
3751 |
+
"epoch": 29.88826815642458,
|
3752 |
+
"grad_norm": 0.16603289544582367,
|
3753 |
+
"learning_rate": 6.27267666434561e-09,
|
3754 |
+
"loss": 0.0123,
|
3755 |
+
"step": 5350
|
3756 |
+
},
|
3757 |
+
{
|
3758 |
+
"epoch": 29.94413407821229,
|
3759 |
+
"grad_norm": 0.12879528105258942,
|
3760 |
+
"learning_rate": 1.7210919580928906e-09,
|
3761 |
+
"loss": 0.0119,
|
3762 |
+
"step": 5360
|
3763 |
+
},
|
3764 |
+
{
|
3765 |
+
"epoch": 30.0,
|
3766 |
+
"grad_norm": 0.17963722348213196,
|
3767 |
+
"learning_rate": 1.4223954430958584e-11,
|
3768 |
+
"loss": 0.0121,
|
3769 |
+
"step": 5370
|
3770 |
+
},
|
3771 |
+
{
|
3772 |
+
"epoch": 30.0,
|
3773 |
+
"step": 5370,
|
3774 |
+
"total_flos": 0.0,
|
3775 |
+
"train_loss": 0.03160388583327805,
|
3776 |
+
"train_runtime": 6237.9083,
|
3777 |
+
"train_samples_per_second": 42.139,
|
3778 |
+
"train_steps_per_second": 0.861
|
3779 |
+
}
|
3780 |
+
],
|
3781 |
+
"logging_steps": 10,
|
3782 |
+
"max_steps": 5370,
|
3783 |
+
"num_input_tokens_seen": 0,
|
3784 |
+
"num_train_epochs": 30,
|
3785 |
+
"save_steps": 20000,
|
3786 |
+
"stateful_callbacks": {
|
3787 |
+
"TrainerControl": {
|
3788 |
+
"args": {
|
3789 |
+
"should_epoch_stop": false,
|
3790 |
+
"should_evaluate": false,
|
3791 |
+
"should_log": false,
|
3792 |
+
"should_save": true,
|
3793 |
+
"should_training_stop": true
|
3794 |
+
},
|
3795 |
+
"attributes": {}
|
3796 |
+
}
|
3797 |
+
},
|
3798 |
+
"total_flos": 0.0,
|
3799 |
+
"train_batch_size": 49,
|
3800 |
+
"trial_name": null,
|
3801 |
+
"trial_params": null
|
3802 |
+
}
|