Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +3241 -0
trainer_state.json
ADDED
@@ -0,0 +1,3241 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 10.0,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 4570,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.02188183807439825,
|
13 |
+
"grad_norm": 7.835834503173828,
|
14 |
+
"learning_rate": 4.3668122270742355e-06,
|
15 |
+
"loss": 1.1502,
|
16 |
+
"step": 10
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.0437636761487965,
|
20 |
+
"grad_norm": 5.089853286743164,
|
21 |
+
"learning_rate": 8.733624454148471e-06,
|
22 |
+
"loss": 1.0181,
|
23 |
+
"step": 20
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.06564551422319474,
|
27 |
+
"grad_norm": 2.486280918121338,
|
28 |
+
"learning_rate": 1.3100436681222708e-05,
|
29 |
+
"loss": 0.6402,
|
30 |
+
"step": 30
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.087527352297593,
|
34 |
+
"grad_norm": 2.2392938137054443,
|
35 |
+
"learning_rate": 1.7467248908296942e-05,
|
36 |
+
"loss": 0.4384,
|
37 |
+
"step": 40
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.10940919037199125,
|
41 |
+
"grad_norm": 2.0446348190307617,
|
42 |
+
"learning_rate": 2.183406113537118e-05,
|
43 |
+
"loss": 0.3294,
|
44 |
+
"step": 50
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.13129102844638948,
|
48 |
+
"grad_norm": 1.6032756567001343,
|
49 |
+
"learning_rate": 2.6200873362445416e-05,
|
50 |
+
"loss": 0.2657,
|
51 |
+
"step": 60
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.15317286652078774,
|
55 |
+
"grad_norm": 1.2855844497680664,
|
56 |
+
"learning_rate": 3.056768558951965e-05,
|
57 |
+
"loss": 0.2188,
|
58 |
+
"step": 70
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.175054704595186,
|
62 |
+
"grad_norm": 2.869873046875,
|
63 |
+
"learning_rate": 3.4934497816593884e-05,
|
64 |
+
"loss": 0.1833,
|
65 |
+
"step": 80
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.19693654266958424,
|
69 |
+
"grad_norm": 1.311296820640564,
|
70 |
+
"learning_rate": 3.930131004366812e-05,
|
71 |
+
"loss": 0.1776,
|
72 |
+
"step": 90
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.2188183807439825,
|
76 |
+
"grad_norm": 0.9431630373001099,
|
77 |
+
"learning_rate": 4.366812227074236e-05,
|
78 |
+
"loss": 0.1691,
|
79 |
+
"step": 100
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.24070021881838075,
|
83 |
+
"grad_norm": 0.9962027668952942,
|
84 |
+
"learning_rate": 4.8034934497816594e-05,
|
85 |
+
"loss": 0.1551,
|
86 |
+
"step": 110
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.26258205689277897,
|
90 |
+
"grad_norm": 0.8225952982902527,
|
91 |
+
"learning_rate": 5.240174672489083e-05,
|
92 |
+
"loss": 0.1378,
|
93 |
+
"step": 120
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.2844638949671772,
|
97 |
+
"grad_norm": 0.662451982498169,
|
98 |
+
"learning_rate": 5.6768558951965065e-05,
|
99 |
+
"loss": 0.1357,
|
100 |
+
"step": 130
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.3063457330415755,
|
104 |
+
"grad_norm": 0.7660549283027649,
|
105 |
+
"learning_rate": 6.11353711790393e-05,
|
106 |
+
"loss": 0.1335,
|
107 |
+
"step": 140
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 0.3282275711159737,
|
111 |
+
"grad_norm": 0.6745500564575195,
|
112 |
+
"learning_rate": 6.550218340611354e-05,
|
113 |
+
"loss": 0.1284,
|
114 |
+
"step": 150
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.350109409190372,
|
118 |
+
"grad_norm": 0.7069956064224243,
|
119 |
+
"learning_rate": 6.986899563318777e-05,
|
120 |
+
"loss": 0.1116,
|
121 |
+
"step": 160
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 0.37199124726477023,
|
125 |
+
"grad_norm": 1.5347901582717896,
|
126 |
+
"learning_rate": 7.423580786026201e-05,
|
127 |
+
"loss": 0.1157,
|
128 |
+
"step": 170
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.3938730853391685,
|
132 |
+
"grad_norm": 0.6622304320335388,
|
133 |
+
"learning_rate": 7.860262008733625e-05,
|
134 |
+
"loss": 0.1137,
|
135 |
+
"step": 180
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 0.41575492341356673,
|
139 |
+
"grad_norm": 0.8169113993644714,
|
140 |
+
"learning_rate": 8.296943231441049e-05,
|
141 |
+
"loss": 0.1099,
|
142 |
+
"step": 190
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 0.437636761487965,
|
146 |
+
"grad_norm": 1.2949796915054321,
|
147 |
+
"learning_rate": 8.733624454148472e-05,
|
148 |
+
"loss": 0.1047,
|
149 |
+
"step": 200
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 0.45951859956236324,
|
153 |
+
"grad_norm": 1.0924999713897705,
|
154 |
+
"learning_rate": 9.170305676855896e-05,
|
155 |
+
"loss": 0.1077,
|
156 |
+
"step": 210
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 0.4814004376367615,
|
160 |
+
"grad_norm": 1.032711386680603,
|
161 |
+
"learning_rate": 9.606986899563319e-05,
|
162 |
+
"loss": 0.093,
|
163 |
+
"step": 220
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 0.5032822757111597,
|
167 |
+
"grad_norm": 0.8792279958724976,
|
168 |
+
"learning_rate": 9.999998690636703e-05,
|
169 |
+
"loss": 0.0948,
|
170 |
+
"step": 230
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 0.5251641137855579,
|
174 |
+
"grad_norm": 0.7732746601104736,
|
175 |
+
"learning_rate": 9.99984156787068e-05,
|
176 |
+
"loss": 0.0929,
|
177 |
+
"step": 240
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 0.5470459518599562,
|
181 |
+
"grad_norm": 1.3531904220581055,
|
182 |
+
"learning_rate": 9.999422581874277e-05,
|
183 |
+
"loss": 0.1034,
|
184 |
+
"step": 250
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 0.5689277899343544,
|
188 |
+
"grad_norm": 42.917991638183594,
|
189 |
+
"learning_rate": 9.998741754591594e-05,
|
190 |
+
"loss": 0.0949,
|
191 |
+
"step": 260
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 0.5908096280087527,
|
195 |
+
"grad_norm": 13.660039901733398,
|
196 |
+
"learning_rate": 9.997799121680488e-05,
|
197 |
+
"loss": 0.1149,
|
198 |
+
"step": 270
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 0.612691466083151,
|
202 |
+
"grad_norm": 5.179759979248047,
|
203 |
+
"learning_rate": 9.996594732510703e-05,
|
204 |
+
"loss": 0.1119,
|
205 |
+
"step": 280
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 0.6345733041575492,
|
209 |
+
"grad_norm": 7.586740970611572,
|
210 |
+
"learning_rate": 9.995128650161283e-05,
|
211 |
+
"loss": 0.0992,
|
212 |
+
"step": 290
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 0.6564551422319475,
|
216 |
+
"grad_norm": 3.8440942764282227,
|
217 |
+
"learning_rate": 9.993400951417277e-05,
|
218 |
+
"loss": 0.1064,
|
219 |
+
"step": 300
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 0.6783369803063457,
|
223 |
+
"grad_norm": 2.224613666534424,
|
224 |
+
"learning_rate": 9.991411726765704e-05,
|
225 |
+
"loss": 0.1033,
|
226 |
+
"step": 310
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 0.700218818380744,
|
230 |
+
"grad_norm": 63.815773010253906,
|
231 |
+
"learning_rate": 9.989161080390825e-05,
|
232 |
+
"loss": 0.1072,
|
233 |
+
"step": 320
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 0.7221006564551422,
|
237 |
+
"grad_norm": 52.77833938598633,
|
238 |
+
"learning_rate": 9.986649130168682e-05,
|
239 |
+
"loss": 0.0984,
|
240 |
+
"step": 330
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 0.7439824945295405,
|
244 |
+
"grad_norm": 15.396129608154297,
|
245 |
+
"learning_rate": 9.983876007660922e-05,
|
246 |
+
"loss": 0.0979,
|
247 |
+
"step": 340
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 0.7658643326039387,
|
251 |
+
"grad_norm": 1.6834031343460083,
|
252 |
+
"learning_rate": 9.980841858107912e-05,
|
253 |
+
"loss": 0.0956,
|
254 |
+
"step": 350
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 0.787746170678337,
|
258 |
+
"grad_norm": 2.0496537685394287,
|
259 |
+
"learning_rate": 9.977546840421128e-05,
|
260 |
+
"loss": 0.0932,
|
261 |
+
"step": 360
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 0.8096280087527352,
|
265 |
+
"grad_norm": 1.0929217338562012,
|
266 |
+
"learning_rate": 9.973991127174833e-05,
|
267 |
+
"loss": 0.0864,
|
268 |
+
"step": 370
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 0.8315098468271335,
|
272 |
+
"grad_norm": 7.530645847320557,
|
273 |
+
"learning_rate": 9.970174904597038e-05,
|
274 |
+
"loss": 0.095,
|
275 |
+
"step": 380
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 0.8533916849015317,
|
279 |
+
"grad_norm": 1.3216899633407593,
|
280 |
+
"learning_rate": 9.966098372559754e-05,
|
281 |
+
"loss": 0.1115,
|
282 |
+
"step": 390
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 0.87527352297593,
|
286 |
+
"grad_norm": 1.4777535200119019,
|
287 |
+
"learning_rate": 9.961761744568512e-05,
|
288 |
+
"loss": 0.1035,
|
289 |
+
"step": 400
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 0.8971553610503282,
|
293 |
+
"grad_norm": 1.4249495267868042,
|
294 |
+
"learning_rate": 9.957165247751192e-05,
|
295 |
+
"loss": 0.0945,
|
296 |
+
"step": 410
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 0.9190371991247265,
|
300 |
+
"grad_norm": 5.402189254760742,
|
301 |
+
"learning_rate": 9.952309122846128e-05,
|
302 |
+
"loss": 0.0972,
|
303 |
+
"step": 420
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 0.9409190371991247,
|
307 |
+
"grad_norm": 2.5458638668060303,
|
308 |
+
"learning_rate": 9.947193624189485e-05,
|
309 |
+
"loss": 0.0933,
|
310 |
+
"step": 430
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 0.962800875273523,
|
314 |
+
"grad_norm": 1.0579487085342407,
|
315 |
+
"learning_rate": 9.941819019701954e-05,
|
316 |
+
"loss": 0.1069,
|
317 |
+
"step": 440
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 0.9846827133479212,
|
321 |
+
"grad_norm": 2.8812167644500732,
|
322 |
+
"learning_rate": 9.936185590874716e-05,
|
323 |
+
"loss": 0.0873,
|
324 |
+
"step": 450
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 1.0065645514223194,
|
328 |
+
"grad_norm": 1.7286851406097412,
|
329 |
+
"learning_rate": 9.93029363275469e-05,
|
330 |
+
"loss": 0.098,
|
331 |
+
"step": 460
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 1.0284463894967177,
|
335 |
+
"grad_norm": 2.46840763092041,
|
336 |
+
"learning_rate": 9.924143453929095e-05,
|
337 |
+
"loss": 0.097,
|
338 |
+
"step": 470
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 1.0503282275711159,
|
342 |
+
"grad_norm": 3.4682812690734863,
|
343 |
+
"learning_rate": 9.917735376509274e-05,
|
344 |
+
"loss": 0.0899,
|
345 |
+
"step": 480
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 1.0722100656455142,
|
349 |
+
"grad_norm": 1.5925973653793335,
|
350 |
+
"learning_rate": 9.911069736113832e-05,
|
351 |
+
"loss": 0.0864,
|
352 |
+
"step": 490
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 1.0940919037199124,
|
356 |
+
"grad_norm": 1.5992704629898071,
|
357 |
+
"learning_rate": 9.904146881851059e-05,
|
358 |
+
"loss": 0.0819,
|
359 |
+
"step": 500
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 1.1159737417943107,
|
363 |
+
"grad_norm": 1.0037175416946411,
|
364 |
+
"learning_rate": 9.896967176300638e-05,
|
365 |
+
"loss": 0.0833,
|
366 |
+
"step": 510
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 1.1378555798687089,
|
370 |
+
"grad_norm": 1.681389570236206,
|
371 |
+
"learning_rate": 9.88953099549466e-05,
|
372 |
+
"loss": 0.0934,
|
373 |
+
"step": 520
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 1.1597374179431073,
|
377 |
+
"grad_norm": 1.4893583059310913,
|
378 |
+
"learning_rate": 9.881838728897936e-05,
|
379 |
+
"loss": 0.093,
|
380 |
+
"step": 530
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 1.1816192560175054,
|
384 |
+
"grad_norm": 1.2456260919570923,
|
385 |
+
"learning_rate": 9.873890779387584e-05,
|
386 |
+
"loss": 0.1026,
|
387 |
+
"step": 540
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 1.2035010940919038,
|
391 |
+
"grad_norm": 0.8705942630767822,
|
392 |
+
"learning_rate": 9.865687563231943e-05,
|
393 |
+
"loss": 0.092,
|
394 |
+
"step": 550
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 1.225382932166302,
|
398 |
+
"grad_norm": 2.0043041706085205,
|
399 |
+
"learning_rate": 9.857229510068761e-05,
|
400 |
+
"loss": 0.0949,
|
401 |
+
"step": 560
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 1.2472647702407003,
|
405 |
+
"grad_norm": 1.574006199836731,
|
406 |
+
"learning_rate": 9.848517062882703e-05,
|
407 |
+
"loss": 0.0843,
|
408 |
+
"step": 570
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 1.2691466083150984,
|
412 |
+
"grad_norm": 0.98636794090271,
|
413 |
+
"learning_rate": 9.839550677982135e-05,
|
414 |
+
"loss": 0.0875,
|
415 |
+
"step": 580
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 1.2910284463894968,
|
419 |
+
"grad_norm": 0.5821505784988403,
|
420 |
+
"learning_rate": 9.830330824975245e-05,
|
421 |
+
"loss": 0.0923,
|
422 |
+
"step": 590
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 1.312910284463895,
|
426 |
+
"grad_norm": 0.5067301392555237,
|
427 |
+
"learning_rate": 9.820857986745431e-05,
|
428 |
+
"loss": 0.0812,
|
429 |
+
"step": 600
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 1.3347921225382933,
|
433 |
+
"grad_norm": 0.5871260762214661,
|
434 |
+
"learning_rate": 9.811132659426016e-05,
|
435 |
+
"loss": 0.0804,
|
436 |
+
"step": 610
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"epoch": 1.3566739606126914,
|
440 |
+
"grad_norm": 0.6894322633743286,
|
441 |
+
"learning_rate": 9.801155352374268e-05,
|
442 |
+
"loss": 0.0909,
|
443 |
+
"step": 620
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 1.3785557986870898,
|
447 |
+
"grad_norm": 0.7414800524711609,
|
448 |
+
"learning_rate": 9.790926588144714e-05,
|
449 |
+
"loss": 0.0901,
|
450 |
+
"step": 630
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 1.400437636761488,
|
454 |
+
"grad_norm": 0.633738100528717,
|
455 |
+
"learning_rate": 9.780446902461777e-05,
|
456 |
+
"loss": 0.092,
|
457 |
+
"step": 640
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 1.4223194748358863,
|
461 |
+
"grad_norm": 0.9708178639411926,
|
462 |
+
"learning_rate": 9.769716844191722e-05,
|
463 |
+
"loss": 0.0905,
|
464 |
+
"step": 650
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 1.4442013129102844,
|
468 |
+
"grad_norm": 0.6483299732208252,
|
469 |
+
"learning_rate": 9.758736975313895e-05,
|
470 |
+
"loss": 0.0835,
|
471 |
+
"step": 660
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 1.4660831509846828,
|
475 |
+
"grad_norm": 0.6239913702011108,
|
476 |
+
"learning_rate": 9.747507870891311e-05,
|
477 |
+
"loss": 0.084,
|
478 |
+
"step": 670
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 1.487964989059081,
|
482 |
+
"grad_norm": 0.5222530364990234,
|
483 |
+
"learning_rate": 9.736030119040508e-05,
|
484 |
+
"loss": 0.0883,
|
485 |
+
"step": 680
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 1.509846827133479,
|
489 |
+
"grad_norm": 0.7400548458099365,
|
490 |
+
"learning_rate": 9.724304320900779e-05,
|
491 |
+
"loss": 0.0894,
|
492 |
+
"step": 690
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 1.5317286652078774,
|
496 |
+
"grad_norm": 0.5438825488090515,
|
497 |
+
"learning_rate": 9.712331090602654e-05,
|
498 |
+
"loss": 0.0848,
|
499 |
+
"step": 700
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 1.5536105032822758,
|
503 |
+
"grad_norm": 0.351309210062027,
|
504 |
+
"learning_rate": 9.700111055235754e-05,
|
505 |
+
"loss": 0.0796,
|
506 |
+
"step": 710
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 1.575492341356674,
|
510 |
+
"grad_norm": 0.5478343963623047,
|
511 |
+
"learning_rate": 9.687644854815952e-05,
|
512 |
+
"loss": 0.0812,
|
513 |
+
"step": 720
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 1.597374179431072,
|
517 |
+
"grad_norm": 0.5228798985481262,
|
518 |
+
"learning_rate": 9.674933142251837e-05,
|
519 |
+
"loss": 0.0802,
|
520 |
+
"step": 730
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 1.6192560175054704,
|
524 |
+
"grad_norm": 0.7279865145683289,
|
525 |
+
"learning_rate": 9.661976583310524e-05,
|
526 |
+
"loss": 0.0834,
|
527 |
+
"step": 740
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 1.6411378555798688,
|
531 |
+
"grad_norm": 0.653016209602356,
|
532 |
+
"learning_rate": 9.648775856582795e-05,
|
533 |
+
"loss": 0.0818,
|
534 |
+
"step": 750
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 1.663019693654267,
|
538 |
+
"grad_norm": 0.5135919451713562,
|
539 |
+
"learning_rate": 9.635331653447545e-05,
|
540 |
+
"loss": 0.075,
|
541 |
+
"step": 760
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 1.684901531728665,
|
545 |
+
"grad_norm": 0.6700458526611328,
|
546 |
+
"learning_rate": 9.621644678035577e-05,
|
547 |
+
"loss": 0.0752,
|
548 |
+
"step": 770
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 1.7067833698030634,
|
552 |
+
"grad_norm": 0.7173268795013428,
|
553 |
+
"learning_rate": 9.607715647192726e-05,
|
554 |
+
"loss": 0.0829,
|
555 |
+
"step": 780
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 1.7286652078774618,
|
559 |
+
"grad_norm": 0.5226753354072571,
|
560 |
+
"learning_rate": 9.593545290442311e-05,
|
561 |
+
"loss": 0.0796,
|
562 |
+
"step": 790
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 1.75054704595186,
|
566 |
+
"grad_norm": 0.6492887139320374,
|
567 |
+
"learning_rate": 9.579134349946926e-05,
|
568 |
+
"loss": 0.0759,
|
569 |
+
"step": 800
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 1.772428884026258,
|
573 |
+
"grad_norm": 0.6919318437576294,
|
574 |
+
"learning_rate": 9.564483580469571e-05,
|
575 |
+
"loss": 0.0815,
|
576 |
+
"step": 810
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 1.7943107221006565,
|
580 |
+
"grad_norm": 0.8271254301071167,
|
581 |
+
"learning_rate": 9.549593749334128e-05,
|
582 |
+
"loss": 0.0796,
|
583 |
+
"step": 820
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 1.8161925601750548,
|
587 |
+
"grad_norm": 0.4964686930179596,
|
588 |
+
"learning_rate": 9.534465636385162e-05,
|
589 |
+
"loss": 0.087,
|
590 |
+
"step": 830
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 1.838074398249453,
|
594 |
+
"grad_norm": 0.5696905255317688,
|
595 |
+
"learning_rate": 9.519100033947083e-05,
|
596 |
+
"loss": 0.0822,
|
597 |
+
"step": 840
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 1.859956236323851,
|
601 |
+
"grad_norm": 0.849829375743866,
|
602 |
+
"learning_rate": 9.503497746782653e-05,
|
603 |
+
"loss": 0.0727,
|
604 |
+
"step": 850
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 1.8818380743982495,
|
608 |
+
"grad_norm": 0.5354284644126892,
|
609 |
+
"learning_rate": 9.487659592050824e-05,
|
610 |
+
"loss": 0.0902,
|
611 |
+
"step": 860
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 1.9037199124726478,
|
615 |
+
"grad_norm": 0.38756123185157776,
|
616 |
+
"learning_rate": 9.471586399263956e-05,
|
617 |
+
"loss": 0.0775,
|
618 |
+
"step": 870
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 1.925601750547046,
|
622 |
+
"grad_norm": 0.7026869058609009,
|
623 |
+
"learning_rate": 9.45527901024436e-05,
|
624 |
+
"loss": 0.0788,
|
625 |
+
"step": 880
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 1.947483588621444,
|
629 |
+
"grad_norm": 0.5695210695266724,
|
630 |
+
"learning_rate": 9.43873827908021e-05,
|
631 |
+
"loss": 0.0757,
|
632 |
+
"step": 890
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 1.9693654266958425,
|
636 |
+
"grad_norm": 0.40557512640953064,
|
637 |
+
"learning_rate": 9.421965072080817e-05,
|
638 |
+
"loss": 0.0722,
|
639 |
+
"step": 900
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 1.9912472647702408,
|
643 |
+
"grad_norm": 0.42085573077201843,
|
644 |
+
"learning_rate": 9.404960267731251e-05,
|
645 |
+
"loss": 0.0689,
|
646 |
+
"step": 910
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 2.0131291028446388,
|
650 |
+
"grad_norm": 0.4493652880191803,
|
651 |
+
"learning_rate": 9.387724756646332e-05,
|
652 |
+
"loss": 0.0774,
|
653 |
+
"step": 920
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 2.035010940919037,
|
657 |
+
"grad_norm": 0.49804505705833435,
|
658 |
+
"learning_rate": 9.370259441523987e-05,
|
659 |
+
"loss": 0.0771,
|
660 |
+
"step": 930
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 2.0568927789934355,
|
664 |
+
"grad_norm": 0.6483808755874634,
|
665 |
+
"learning_rate": 9.352565237097965e-05,
|
666 |
+
"loss": 0.078,
|
667 |
+
"step": 940
|
668 |
+
},
|
669 |
+
{
|
670 |
+
"epoch": 2.078774617067834,
|
671 |
+
"grad_norm": 0.5233718156814575,
|
672 |
+
"learning_rate": 9.334643070089937e-05,
|
673 |
+
"loss": 0.0709,
|
674 |
+
"step": 950
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"epoch": 2.1006564551422318,
|
678 |
+
"grad_norm": 0.6516150832176208,
|
679 |
+
"learning_rate": 9.316493879160957e-05,
|
680 |
+
"loss": 0.079,
|
681 |
+
"step": 960
|
682 |
+
},
|
683 |
+
{
|
684 |
+
"epoch": 2.12253829321663,
|
685 |
+
"grad_norm": 0.692732036113739,
|
686 |
+
"learning_rate": 9.298118614862297e-05,
|
687 |
+
"loss": 0.0775,
|
688 |
+
"step": 970
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"epoch": 2.1444201312910285,
|
692 |
+
"grad_norm": 0.5700138807296753,
|
693 |
+
"learning_rate": 9.279518239585667e-05,
|
694 |
+
"loss": 0.0782,
|
695 |
+
"step": 980
|
696 |
+
},
|
697 |
+
{
|
698 |
+
"epoch": 2.166301969365427,
|
699 |
+
"grad_norm": 0.3617083728313446,
|
700 |
+
"learning_rate": 9.260693727512803e-05,
|
701 |
+
"loss": 0.0714,
|
702 |
+
"step": 990
|
703 |
+
},
|
704 |
+
{
|
705 |
+
"epoch": 2.1881838074398248,
|
706 |
+
"grad_norm": 0.4216122031211853,
|
707 |
+
"learning_rate": 9.241646064564457e-05,
|
708 |
+
"loss": 0.0672,
|
709 |
+
"step": 1000
|
710 |
+
},
|
711 |
+
{
|
712 |
+
"epoch": 2.210065645514223,
|
713 |
+
"grad_norm": 0.38087522983551025,
|
714 |
+
"learning_rate": 9.222376248348747e-05,
|
715 |
+
"loss": 0.08,
|
716 |
+
"step": 1010
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"epoch": 2.2319474835886215,
|
720 |
+
"grad_norm": 0.8693896532058716,
|
721 |
+
"learning_rate": 9.20288528810892e-05,
|
722 |
+
"loss": 0.0778,
|
723 |
+
"step": 1020
|
724 |
+
},
|
725 |
+
{
|
726 |
+
"epoch": 2.25382932166302,
|
727 |
+
"grad_norm": 0.6329444050788879,
|
728 |
+
"learning_rate": 9.183174204670483e-05,
|
729 |
+
"loss": 0.0722,
|
730 |
+
"step": 1030
|
731 |
+
},
|
732 |
+
{
|
733 |
+
"epoch": 2.2757111597374178,
|
734 |
+
"grad_norm": 0.9332349896430969,
|
735 |
+
"learning_rate": 9.163244030387743e-05,
|
736 |
+
"loss": 0.0705,
|
737 |
+
"step": 1040
|
738 |
+
},
|
739 |
+
{
|
740 |
+
"epoch": 2.297592997811816,
|
741 |
+
"grad_norm": 0.8268052935600281,
|
742 |
+
"learning_rate": 9.143095809089741e-05,
|
743 |
+
"loss": 0.0719,
|
744 |
+
"step": 1050
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"epoch": 2.3194748358862145,
|
748 |
+
"grad_norm": 1.8830317258834839,
|
749 |
+
"learning_rate": 9.122730596025579e-05,
|
750 |
+
"loss": 0.0733,
|
751 |
+
"step": 1060
|
752 |
+
},
|
753 |
+
{
|
754 |
+
"epoch": 2.341356673960613,
|
755 |
+
"grad_norm": 0.4748838245868683,
|
756 |
+
"learning_rate": 9.102149457809146e-05,
|
757 |
+
"loss": 0.0692,
|
758 |
+
"step": 1070
|
759 |
+
},
|
760 |
+
{
|
761 |
+
"epoch": 2.363238512035011,
|
762 |
+
"grad_norm": 0.6225442290306091,
|
763 |
+
"learning_rate": 9.081353472363265e-05,
|
764 |
+
"loss": 0.0724,
|
765 |
+
"step": 1080
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"epoch": 2.385120350109409,
|
769 |
+
"grad_norm": 0.3627307415008545,
|
770 |
+
"learning_rate": 9.060343728863239e-05,
|
771 |
+
"loss": 0.064,
|
772 |
+
"step": 1090
|
773 |
+
},
|
774 |
+
{
|
775 |
+
"epoch": 2.4070021881838075,
|
776 |
+
"grad_norm": 0.5193670392036438,
|
777 |
+
"learning_rate": 9.039121327679791e-05,
|
778 |
+
"loss": 0.079,
|
779 |
+
"step": 1100
|
780 |
+
},
|
781 |
+
{
|
782 |
+
"epoch": 2.428884026258206,
|
783 |
+
"grad_norm": 0.8696744441986084,
|
784 |
+
"learning_rate": 9.017687380321449e-05,
|
785 |
+
"loss": 0.0715,
|
786 |
+
"step": 1110
|
787 |
+
},
|
788 |
+
{
|
789 |
+
"epoch": 2.450765864332604,
|
790 |
+
"grad_norm": 0.6547001004219055,
|
791 |
+
"learning_rate": 8.996043009376319e-05,
|
792 |
+
"loss": 0.0691,
|
793 |
+
"step": 1120
|
794 |
+
},
|
795 |
+
{
|
796 |
+
"epoch": 2.472647702407002,
|
797 |
+
"grad_norm": 0.5175451040267944,
|
798 |
+
"learning_rate": 8.974189348453306e-05,
|
799 |
+
"loss": 0.0749,
|
800 |
+
"step": 1130
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 2.4945295404814005,
|
804 |
+
"grad_norm": 0.5802105665206909,
|
805 |
+
"learning_rate": 8.952127542122722e-05,
|
806 |
+
"loss": 0.0717,
|
807 |
+
"step": 1140
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 2.516411378555799,
|
811 |
+
"grad_norm": 0.7577764987945557,
|
812 |
+
"learning_rate": 8.929858745856354e-05,
|
813 |
+
"loss": 0.0816,
|
814 |
+
"step": 1150
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"epoch": 2.538293216630197,
|
818 |
+
"grad_norm": 0.5405493378639221,
|
819 |
+
"learning_rate": 8.907384125966945e-05,
|
820 |
+
"loss": 0.0738,
|
821 |
+
"step": 1160
|
822 |
+
},
|
823 |
+
{
|
824 |
+
"epoch": 2.560175054704595,
|
825 |
+
"grad_norm": 0.3944590985774994,
|
826 |
+
"learning_rate": 8.884704859547108e-05,
|
827 |
+
"loss": 0.071,
|
828 |
+
"step": 1170
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 2.5820568927789935,
|
832 |
+
"grad_norm": 0.5056865215301514,
|
833 |
+
"learning_rate": 8.861822134407672e-05,
|
834 |
+
"loss": 0.0692,
|
835 |
+
"step": 1180
|
836 |
+
},
|
837 |
+
{
|
838 |
+
"epoch": 2.6039387308533914,
|
839 |
+
"grad_norm": 0.4500432312488556,
|
840 |
+
"learning_rate": 8.838737149015477e-05,
|
841 |
+
"loss": 0.0723,
|
842 |
+
"step": 1190
|
843 |
+
},
|
844 |
+
{
|
845 |
+
"epoch": 2.62582056892779,
|
846 |
+
"grad_norm": 0.36721524596214294,
|
847 |
+
"learning_rate": 8.815451112430604e-05,
|
848 |
+
"loss": 0.0686,
|
849 |
+
"step": 1200
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 2.647702407002188,
|
853 |
+
"grad_norm": 0.7576287984848022,
|
854 |
+
"learning_rate": 8.79196524424305e-05,
|
855 |
+
"loss": 0.0769,
|
856 |
+
"step": 1210
|
857 |
+
},
|
858 |
+
{
|
859 |
+
"epoch": 2.6695842450765865,
|
860 |
+
"grad_norm": 0.6370827555656433,
|
861 |
+
"learning_rate": 8.768280774508853e-05,
|
862 |
+
"loss": 0.0815,
|
863 |
+
"step": 1220
|
864 |
+
},
|
865 |
+
{
|
866 |
+
"epoch": 2.691466083150985,
|
867 |
+
"grad_norm": 0.3080037236213684,
|
868 |
+
"learning_rate": 8.74439894368567e-05,
|
869 |
+
"loss": 0.0607,
|
870 |
+
"step": 1230
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"epoch": 2.713347921225383,
|
874 |
+
"grad_norm": 0.46508321166038513,
|
875 |
+
"learning_rate": 8.720321002567807e-05,
|
876 |
+
"loss": 0.0696,
|
877 |
+
"step": 1240
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"epoch": 2.735229759299781,
|
881 |
+
"grad_norm": 0.6237578988075256,
|
882 |
+
"learning_rate": 8.696048212220711e-05,
|
883 |
+
"loss": 0.0716,
|
884 |
+
"step": 1250
|
885 |
+
},
|
886 |
+
{
|
887 |
+
"epoch": 2.7571115973741795,
|
888 |
+
"grad_norm": 0.4095085561275482,
|
889 |
+
"learning_rate": 8.671581843914923e-05,
|
890 |
+
"loss": 0.0635,
|
891 |
+
"step": 1260
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 2.7789934354485775,
|
895 |
+
"grad_norm": 0.6405256986618042,
|
896 |
+
"learning_rate": 8.646923179059494e-05,
|
897 |
+
"loss": 0.0648,
|
898 |
+
"step": 1270
|
899 |
+
},
|
900 |
+
{
|
901 |
+
"epoch": 2.800875273522976,
|
902 |
+
"grad_norm": 0.48811593651771545,
|
903 |
+
"learning_rate": 8.622073509134874e-05,
|
904 |
+
"loss": 0.0645,
|
905 |
+
"step": 1280
|
906 |
+
},
|
907 |
+
{
|
908 |
+
"epoch": 2.822757111597374,
|
909 |
+
"grad_norm": 0.5660180449485779,
|
910 |
+
"learning_rate": 8.597034135625268e-05,
|
911 |
+
"loss": 0.0697,
|
912 |
+
"step": 1290
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"epoch": 2.8446389496717726,
|
916 |
+
"grad_norm": 0.5524961352348328,
|
917 |
+
"learning_rate": 8.571806369950481e-05,
|
918 |
+
"loss": 0.0721,
|
919 |
+
"step": 1300
|
920 |
+
},
|
921 |
+
{
|
922 |
+
"epoch": 2.866520787746171,
|
923 |
+
"grad_norm": 0.38668200373649597,
|
924 |
+
"learning_rate": 8.54639153339722e-05,
|
925 |
+
"loss": 0.0666,
|
926 |
+
"step": 1310
|
927 |
+
},
|
928 |
+
{
|
929 |
+
"epoch": 2.888402625820569,
|
930 |
+
"grad_norm": 0.5835091471672058,
|
931 |
+
"learning_rate": 8.520790957049904e-05,
|
932 |
+
"loss": 0.0613,
|
933 |
+
"step": 1320
|
934 |
+
},
|
935 |
+
{
|
936 |
+
"epoch": 2.910284463894967,
|
937 |
+
"grad_norm": 0.5212987661361694,
|
938 |
+
"learning_rate": 8.49500598172094e-05,
|
939 |
+
"loss": 0.0732,
|
940 |
+
"step": 1330
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"epoch": 2.9321663019693656,
|
944 |
+
"grad_norm": 0.28042611479759216,
|
945 |
+
"learning_rate": 8.46903795788051e-05,
|
946 |
+
"loss": 0.0582,
|
947 |
+
"step": 1340
|
948 |
+
},
|
949 |
+
{
|
950 |
+
"epoch": 2.9540481400437635,
|
951 |
+
"grad_norm": 0.5811691880226135,
|
952 |
+
"learning_rate": 8.442888245585828e-05,
|
953 |
+
"loss": 0.0683,
|
954 |
+
"step": 1350
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 2.975929978118162,
|
958 |
+
"grad_norm": 0.4917229115962982,
|
959 |
+
"learning_rate": 8.41655821440992e-05,
|
960 |
+
"loss": 0.0669,
|
961 |
+
"step": 1360
|
962 |
+
},
|
963 |
+
{
|
964 |
+
"epoch": 2.99781181619256,
|
965 |
+
"grad_norm": 0.5715236663818359,
|
966 |
+
"learning_rate": 8.390049243369886e-05,
|
967 |
+
"loss": 0.064,
|
968 |
+
"step": 1370
|
969 |
+
},
|
970 |
+
{
|
971 |
+
"epoch": 3.0196936542669586,
|
972 |
+
"grad_norm": 0.7416530251502991,
|
973 |
+
"learning_rate": 8.363362720854677e-05,
|
974 |
+
"loss": 0.0701,
|
975 |
+
"step": 1380
|
976 |
+
},
|
977 |
+
{
|
978 |
+
"epoch": 3.0415754923413565,
|
979 |
+
"grad_norm": 0.6221567988395691,
|
980 |
+
"learning_rate": 8.33650004455238e-05,
|
981 |
+
"loss": 0.0707,
|
982 |
+
"step": 1390
|
983 |
+
},
|
984 |
+
{
|
985 |
+
"epoch": 3.063457330415755,
|
986 |
+
"grad_norm": 0.5047647953033447,
|
987 |
+
"learning_rate": 8.309462621377013e-05,
|
988 |
+
"loss": 0.0787,
|
989 |
+
"step": 1400
|
990 |
+
},
|
991 |
+
{
|
992 |
+
"epoch": 3.085339168490153,
|
993 |
+
"grad_norm": 0.4984498918056488,
|
994 |
+
"learning_rate": 8.28225186739484e-05,
|
995 |
+
"loss": 0.0629,
|
996 |
+
"step": 1410
|
997 |
+
},
|
998 |
+
{
|
999 |
+
"epoch": 3.1072210065645516,
|
1000 |
+
"grad_norm": 0.37575212121009827,
|
1001 |
+
"learning_rate": 8.254869207750206e-05,
|
1002 |
+
"loss": 0.0638,
|
1003 |
+
"step": 1420
|
1004 |
+
},
|
1005 |
+
{
|
1006 |
+
"epoch": 3.1291028446389495,
|
1007 |
+
"grad_norm": 0.5403359532356262,
|
1008 |
+
"learning_rate": 8.227316076590898e-05,
|
1009 |
+
"loss": 0.0645,
|
1010 |
+
"step": 1430
|
1011 |
+
},
|
1012 |
+
{
|
1013 |
+
"epoch": 3.150984682713348,
|
1014 |
+
"grad_norm": 0.4255600869655609,
|
1015 |
+
"learning_rate": 8.19959391699303e-05,
|
1016 |
+
"loss": 0.0615,
|
1017 |
+
"step": 1440
|
1018 |
+
},
|
1019 |
+
{
|
1020 |
+
"epoch": 3.1728665207877462,
|
1021 |
+
"grad_norm": 0.41742101311683655,
|
1022 |
+
"learning_rate": 8.171704180885457e-05,
|
1023 |
+
"loss": 0.0614,
|
1024 |
+
"step": 1450
|
1025 |
+
},
|
1026 |
+
{
|
1027 |
+
"epoch": 3.1947483588621446,
|
1028 |
+
"grad_norm": 0.36597940325737,
|
1029 |
+
"learning_rate": 8.143648328973746e-05,
|
1030 |
+
"loss": 0.0662,
|
1031 |
+
"step": 1460
|
1032 |
+
},
|
1033 |
+
{
|
1034 |
+
"epoch": 3.2166301969365425,
|
1035 |
+
"grad_norm": 0.34130623936653137,
|
1036 |
+
"learning_rate": 8.115427830663658e-05,
|
1037 |
+
"loss": 0.0705,
|
1038 |
+
"step": 1470
|
1039 |
+
},
|
1040 |
+
{
|
1041 |
+
"epoch": 3.238512035010941,
|
1042 |
+
"grad_norm": 0.45353439450263977,
|
1043 |
+
"learning_rate": 8.087044163984197e-05,
|
1044 |
+
"loss": 0.062,
|
1045 |
+
"step": 1480
|
1046 |
+
},
|
1047 |
+
{
|
1048 |
+
"epoch": 3.2603938730853392,
|
1049 |
+
"grad_norm": 0.525982677936554,
|
1050 |
+
"learning_rate": 8.058498815510196e-05,
|
1051 |
+
"loss": 0.0691,
|
1052 |
+
"step": 1490
|
1053 |
+
},
|
1054 |
+
{
|
1055 |
+
"epoch": 3.2822757111597376,
|
1056 |
+
"grad_norm": 0.520045280456543,
|
1057 |
+
"learning_rate": 8.029793280284459e-05,
|
1058 |
+
"loss": 0.0612,
|
1059 |
+
"step": 1500
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 3.3041575492341355,
|
1063 |
+
"grad_norm": 0.5012158155441284,
|
1064 |
+
"learning_rate": 8.000929061739464e-05,
|
1065 |
+
"loss": 0.067,
|
1066 |
+
"step": 1510
|
1067 |
+
},
|
1068 |
+
{
|
1069 |
+
"epoch": 3.326039387308534,
|
1070 |
+
"grad_norm": 0.36634936928749084,
|
1071 |
+
"learning_rate": 7.971907671618612e-05,
|
1072 |
+
"loss": 0.0672,
|
1073 |
+
"step": 1520
|
1074 |
+
},
|
1075 |
+
{
|
1076 |
+
"epoch": 3.3479212253829322,
|
1077 |
+
"grad_norm": 0.43757519125938416,
|
1078 |
+
"learning_rate": 7.942730629897063e-05,
|
1079 |
+
"loss": 0.067,
|
1080 |
+
"step": 1530
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"epoch": 3.3698030634573306,
|
1084 |
+
"grad_norm": 0.44617760181427,
|
1085 |
+
"learning_rate": 7.913399464702113e-05,
|
1086 |
+
"loss": 0.0623,
|
1087 |
+
"step": 1540
|
1088 |
+
},
|
1089 |
+
{
|
1090 |
+
"epoch": 3.3916849015317285,
|
1091 |
+
"grad_norm": 0.6319367289543152,
|
1092 |
+
"learning_rate": 7.883915712233177e-05,
|
1093 |
+
"loss": 0.063,
|
1094 |
+
"step": 1550
|
1095 |
+
},
|
1096 |
+
{
|
1097 |
+
"epoch": 3.413566739606127,
|
1098 |
+
"grad_norm": 0.3538963496685028,
|
1099 |
+
"learning_rate": 7.854280916681314e-05,
|
1100 |
+
"loss": 0.066,
|
1101 |
+
"step": 1560
|
1102 |
+
},
|
1103 |
+
{
|
1104 |
+
"epoch": 3.4354485776805253,
|
1105 |
+
"grad_norm": 0.4608595371246338,
|
1106 |
+
"learning_rate": 7.824496630148364e-05,
|
1107 |
+
"loss": 0.0621,
|
1108 |
+
"step": 1570
|
1109 |
+
},
|
1110 |
+
{
|
1111 |
+
"epoch": 3.4573304157549236,
|
1112 |
+
"grad_norm": 0.5707418322563171,
|
1113 |
+
"learning_rate": 7.794564412565655e-05,
|
1114 |
+
"loss": 0.0696,
|
1115 |
+
"step": 1580
|
1116 |
+
},
|
1117 |
+
{
|
1118 |
+
"epoch": 3.4792122538293215,
|
1119 |
+
"grad_norm": 0.37221798300743103,
|
1120 |
+
"learning_rate": 7.764485831612299e-05,
|
1121 |
+
"loss": 0.066,
|
1122 |
+
"step": 1590
|
1123 |
+
},
|
1124 |
+
{
|
1125 |
+
"epoch": 3.50109409190372,
|
1126 |
+
"grad_norm": 0.6436975598335266,
|
1127 |
+
"learning_rate": 7.734262462633085e-05,
|
1128 |
+
"loss": 0.0597,
|
1129 |
+
"step": 1600
|
1130 |
+
},
|
1131 |
+
{
|
1132 |
+
"epoch": 3.5229759299781183,
|
1133 |
+
"grad_norm": 0.7743731141090393,
|
1134 |
+
"learning_rate": 7.703895888555978e-05,
|
1135 |
+
"loss": 0.0696,
|
1136 |
+
"step": 1610
|
1137 |
+
},
|
1138 |
+
{
|
1139 |
+
"epoch": 3.544857768052516,
|
1140 |
+
"grad_norm": 0.561302900314331,
|
1141 |
+
"learning_rate": 7.673387699809211e-05,
|
1142 |
+
"loss": 0.0613,
|
1143 |
+
"step": 1620
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 3.5667396061269145,
|
1147 |
+
"grad_norm": 0.42251160740852356,
|
1148 |
+
"learning_rate": 7.642739494237986e-05,
|
1149 |
+
"loss": 0.0642,
|
1150 |
+
"step": 1630
|
1151 |
+
},
|
1152 |
+
{
|
1153 |
+
"epoch": 3.588621444201313,
|
1154 |
+
"grad_norm": 0.2583286166191101,
|
1155 |
+
"learning_rate": 7.611952877020787e-05,
|
1156 |
+
"loss": 0.0648,
|
1157 |
+
"step": 1640
|
1158 |
+
},
|
1159 |
+
{
|
1160 |
+
"epoch": 3.6105032822757113,
|
1161 |
+
"grad_norm": 0.6360892057418823,
|
1162 |
+
"learning_rate": 7.581029460585313e-05,
|
1163 |
+
"loss": 0.0621,
|
1164 |
+
"step": 1650
|
1165 |
+
},
|
1166 |
+
{
|
1167 |
+
"epoch": 3.6323851203501096,
|
1168 |
+
"grad_norm": 0.4245406687259674,
|
1169 |
+
"learning_rate": 7.54997086452403e-05,
|
1170 |
+
"loss": 0.0618,
|
1171 |
+
"step": 1660
|
1172 |
+
},
|
1173 |
+
{
|
1174 |
+
"epoch": 3.6542669584245075,
|
1175 |
+
"grad_norm": 0.3884272277355194,
|
1176 |
+
"learning_rate": 7.518778715509341e-05,
|
1177 |
+
"loss": 0.0651,
|
1178 |
+
"step": 1670
|
1179 |
+
},
|
1180 |
+
{
|
1181 |
+
"epoch": 3.676148796498906,
|
1182 |
+
"grad_norm": 0.4515193700790405,
|
1183 |
+
"learning_rate": 7.487454647208387e-05,
|
1184 |
+
"loss": 0.0614,
|
1185 |
+
"step": 1680
|
1186 |
+
},
|
1187 |
+
{
|
1188 |
+
"epoch": 3.6980306345733043,
|
1189 |
+
"grad_norm": 0.32720717787742615,
|
1190 |
+
"learning_rate": 7.456000300197498e-05,
|
1191 |
+
"loss": 0.0632,
|
1192 |
+
"step": 1690
|
1193 |
+
},
|
1194 |
+
{
|
1195 |
+
"epoch": 3.719912472647702,
|
1196 |
+
"grad_norm": 0.3803008198738098,
|
1197 |
+
"learning_rate": 7.424417321876258e-05,
|
1198 |
+
"loss": 0.0688,
|
1199 |
+
"step": 1700
|
1200 |
+
},
|
1201 |
+
{
|
1202 |
+
"epoch": 3.7417943107221006,
|
1203 |
+
"grad_norm": 0.4703556299209595,
|
1204 |
+
"learning_rate": 7.392707366381226e-05,
|
1205 |
+
"loss": 0.0565,
|
1206 |
+
"step": 1710
|
1207 |
+
},
|
1208 |
+
{
|
1209 |
+
"epoch": 3.763676148796499,
|
1210 |
+
"grad_norm": 0.3876951336860657,
|
1211 |
+
"learning_rate": 7.360872094499303e-05,
|
1212 |
+
"loss": 0.0629,
|
1213 |
+
"step": 1720
|
1214 |
+
},
|
1215 |
+
{
|
1216 |
+
"epoch": 3.7855579868708973,
|
1217 |
+
"grad_norm": 0.6968600749969482,
|
1218 |
+
"learning_rate": 7.328913173580745e-05,
|
1219 |
+
"loss": 0.0654,
|
1220 |
+
"step": 1730
|
1221 |
+
},
|
1222 |
+
{
|
1223 |
+
"epoch": 3.8074398249452956,
|
1224 |
+
"grad_norm": 0.3983824551105499,
|
1225 |
+
"learning_rate": 7.29683227745185e-05,
|
1226 |
+
"loss": 0.0665,
|
1227 |
+
"step": 1740
|
1228 |
+
},
|
1229 |
+
{
|
1230 |
+
"epoch": 3.8293216630196936,
|
1231 |
+
"grad_norm": 0.6152142286300659,
|
1232 |
+
"learning_rate": 7.264631086327273e-05,
|
1233 |
+
"loss": 0.0665,
|
1234 |
+
"step": 1750
|
1235 |
+
},
|
1236 |
+
{
|
1237 |
+
"epoch": 3.851203501094092,
|
1238 |
+
"grad_norm": 0.45532792806625366,
|
1239 |
+
"learning_rate": 7.232311286722038e-05,
|
1240 |
+
"loss": 0.0644,
|
1241 |
+
"step": 1760
|
1242 |
+
},
|
1243 |
+
{
|
1244 |
+
"epoch": 3.8730853391684903,
|
1245 |
+
"grad_norm": 0.5433072447776794,
|
1246 |
+
"learning_rate": 7.199874571363212e-05,
|
1247 |
+
"loss": 0.0669,
|
1248 |
+
"step": 1770
|
1249 |
+
},
|
1250 |
+
{
|
1251 |
+
"epoch": 3.894967177242888,
|
1252 |
+
"grad_norm": 0.393799364566803,
|
1253 |
+
"learning_rate": 7.167322639101234e-05,
|
1254 |
+
"loss": 0.0609,
|
1255 |
+
"step": 1780
|
1256 |
+
},
|
1257 |
+
{
|
1258 |
+
"epoch": 3.9168490153172866,
|
1259 |
+
"grad_norm": 0.37970104813575745,
|
1260 |
+
"learning_rate": 7.134657194820957e-05,
|
1261 |
+
"loss": 0.0613,
|
1262 |
+
"step": 1790
|
1263 |
+
},
|
1264 |
+
{
|
1265 |
+
"epoch": 3.938730853391685,
|
1266 |
+
"grad_norm": 0.4533461928367615,
|
1267 |
+
"learning_rate": 7.101879949352343e-05,
|
1268 |
+
"loss": 0.0661,
|
1269 |
+
"step": 1800
|
1270 |
+
},
|
1271 |
+
{
|
1272 |
+
"epoch": 3.9606126914660833,
|
1273 |
+
"grad_norm": 0.5303936004638672,
|
1274 |
+
"learning_rate": 7.068992619380859e-05,
|
1275 |
+
"loss": 0.0604,
|
1276 |
+
"step": 1810
|
1277 |
+
},
|
1278 |
+
{
|
1279 |
+
"epoch": 3.9824945295404817,
|
1280 |
+
"grad_norm": 0.43903806805610657,
|
1281 |
+
"learning_rate": 7.035996927357583e-05,
|
1282 |
+
"loss": 0.0575,
|
1283 |
+
"step": 1820
|
1284 |
+
},
|
1285 |
+
{
|
1286 |
+
"epoch": 4.00437636761488,
|
1287 |
+
"grad_norm": 0.4638700485229492,
|
1288 |
+
"learning_rate": 7.00289460140897e-05,
|
1289 |
+
"loss": 0.0675,
|
1290 |
+
"step": 1830
|
1291 |
+
},
|
1292 |
+
{
|
1293 |
+
"epoch": 4.0262582056892775,
|
1294 |
+
"grad_norm": 0.6604804396629333,
|
1295 |
+
"learning_rate": 6.969687375246355e-05,
|
1296 |
+
"loss": 0.066,
|
1297 |
+
"step": 1840
|
1298 |
+
},
|
1299 |
+
{
|
1300 |
+
"epoch": 4.048140043763676,
|
1301 |
+
"grad_norm": 0.39075350761413574,
|
1302 |
+
"learning_rate": 6.936376988075155e-05,
|
1303 |
+
"loss": 0.0642,
|
1304 |
+
"step": 1850
|
1305 |
+
},
|
1306 |
+
{
|
1307 |
+
"epoch": 4.070021881838074,
|
1308 |
+
"grad_norm": 0.5355797410011292,
|
1309 |
+
"learning_rate": 6.902965184503763e-05,
|
1310 |
+
"loss": 0.0618,
|
1311 |
+
"step": 1860
|
1312 |
+
},
|
1313 |
+
{
|
1314 |
+
"epoch": 4.091903719912473,
|
1315 |
+
"grad_norm": 0.70420241355896,
|
1316 |
+
"learning_rate": 6.869453714452194e-05,
|
1317 |
+
"loss": 0.0613,
|
1318 |
+
"step": 1870
|
1319 |
+
},
|
1320 |
+
{
|
1321 |
+
"epoch": 4.113785557986871,
|
1322 |
+
"grad_norm": 0.5154333114624023,
|
1323 |
+
"learning_rate": 6.835844333060423e-05,
|
1324 |
+
"loss": 0.058,
|
1325 |
+
"step": 1880
|
1326 |
+
},
|
1327 |
+
{
|
1328 |
+
"epoch": 4.135667396061269,
|
1329 |
+
"grad_norm": 0.5481790900230408,
|
1330 |
+
"learning_rate": 6.802138800596462e-05,
|
1331 |
+
"loss": 0.0602,
|
1332 |
+
"step": 1890
|
1333 |
+
},
|
1334 |
+
{
|
1335 |
+
"epoch": 4.157549234135668,
|
1336 |
+
"grad_norm": 1.0096722841262817,
|
1337 |
+
"learning_rate": 6.768338882364168e-05,
|
1338 |
+
"loss": 0.0693,
|
1339 |
+
"step": 1900
|
1340 |
+
},
|
1341 |
+
{
|
1342 |
+
"epoch": 4.179431072210066,
|
1343 |
+
"grad_norm": 0.4854465126991272,
|
1344 |
+
"learning_rate": 6.734446348610792e-05,
|
1345 |
+
"loss": 0.0612,
|
1346 |
+
"step": 1910
|
1347 |
+
},
|
1348 |
+
{
|
1349 |
+
"epoch": 4.2013129102844635,
|
1350 |
+
"grad_norm": 0.33401617407798767,
|
1351 |
+
"learning_rate": 6.700462974434251e-05,
|
1352 |
+
"loss": 0.0612,
|
1353 |
+
"step": 1920
|
1354 |
+
},
|
1355 |
+
{
|
1356 |
+
"epoch": 4.223194748358862,
|
1357 |
+
"grad_norm": 0.44405895471572876,
|
1358 |
+
"learning_rate": 6.666390539690179e-05,
|
1359 |
+
"loss": 0.0644,
|
1360 |
+
"step": 1930
|
1361 |
+
},
|
1362 |
+
{
|
1363 |
+
"epoch": 4.24507658643326,
|
1364 |
+
"grad_norm": 0.43215858936309814,
|
1365 |
+
"learning_rate": 6.632230828898677e-05,
|
1366 |
+
"loss": 0.0592,
|
1367 |
+
"step": 1940
|
1368 |
+
},
|
1369 |
+
{
|
1370 |
+
"epoch": 4.266958424507659,
|
1371 |
+
"grad_norm": 0.3804273307323456,
|
1372 |
+
"learning_rate": 6.597985631150884e-05,
|
1373 |
+
"loss": 0.0565,
|
1374 |
+
"step": 1950
|
1375 |
+
},
|
1376 |
+
{
|
1377 |
+
"epoch": 4.288840262582057,
|
1378 |
+
"grad_norm": 0.43090665340423584,
|
1379 |
+
"learning_rate": 6.56365674001525e-05,
|
1380 |
+
"loss": 0.0579,
|
1381 |
+
"step": 1960
|
1382 |
+
},
|
1383 |
+
{
|
1384 |
+
"epoch": 4.310722100656455,
|
1385 |
+
"grad_norm": 0.5440065264701843,
|
1386 |
+
"learning_rate": 6.529245953443615e-05,
|
1387 |
+
"loss": 0.0575,
|
1388 |
+
"step": 1970
|
1389 |
+
},
|
1390 |
+
{
|
1391 |
+
"epoch": 4.332603938730854,
|
1392 |
+
"grad_norm": 0.36599451303482056,
|
1393 |
+
"learning_rate": 6.49475507367703e-05,
|
1394 |
+
"loss": 0.0598,
|
1395 |
+
"step": 1980
|
1396 |
+
},
|
1397 |
+
{
|
1398 |
+
"epoch": 4.354485776805252,
|
1399 |
+
"grad_norm": 0.33587977290153503,
|
1400 |
+
"learning_rate": 6.460185907151371e-05,
|
1401 |
+
"loss": 0.0583,
|
1402 |
+
"step": 1990
|
1403 |
+
},
|
1404 |
+
{
|
1405 |
+
"epoch": 4.3763676148796495,
|
1406 |
+
"grad_norm": 0.3317202627658844,
|
1407 |
+
"learning_rate": 6.425540264402735e-05,
|
1408 |
+
"loss": 0.0601,
|
1409 |
+
"step": 2000
|
1410 |
+
},
|
1411 |
+
{
|
1412 |
+
"epoch": 4.398249452954048,
|
1413 |
+
"grad_norm": 0.36118966341018677,
|
1414 |
+
"learning_rate": 6.390819959972603e-05,
|
1415 |
+
"loss": 0.0531,
|
1416 |
+
"step": 2010
|
1417 |
+
},
|
1418 |
+
{
|
1419 |
+
"epoch": 4.420131291028446,
|
1420 |
+
"grad_norm": 0.3242454528808594,
|
1421 |
+
"learning_rate": 6.356026812312809e-05,
|
1422 |
+
"loss": 0.0585,
|
1423 |
+
"step": 2020
|
1424 |
+
},
|
1425 |
+
{
|
1426 |
+
"epoch": 4.442013129102845,
|
1427 |
+
"grad_norm": 0.4215313494205475,
|
1428 |
+
"learning_rate": 6.321162643690304e-05,
|
1429 |
+
"loss": 0.0637,
|
1430 |
+
"step": 2030
|
1431 |
+
},
|
1432 |
+
{
|
1433 |
+
"epoch": 4.463894967177243,
|
1434 |
+
"grad_norm": 0.34491801261901855,
|
1435 |
+
"learning_rate": 6.286229280091711e-05,
|
1436 |
+
"loss": 0.0642,
|
1437 |
+
"step": 2040
|
1438 |
+
},
|
1439 |
+
{
|
1440 |
+
"epoch": 4.485776805251641,
|
1441 |
+
"grad_norm": 0.5286457538604736,
|
1442 |
+
"learning_rate": 6.251228551127691e-05,
|
1443 |
+
"loss": 0.0654,
|
1444 |
+
"step": 2050
|
1445 |
+
},
|
1446 |
+
{
|
1447 |
+
"epoch": 4.50765864332604,
|
1448 |
+
"grad_norm": 0.357379674911499,
|
1449 |
+
"learning_rate": 6.216162289937119e-05,
|
1450 |
+
"loss": 0.0529,
|
1451 |
+
"step": 2060
|
1452 |
+
},
|
1453 |
+
{
|
1454 |
+
"epoch": 4.529540481400438,
|
1455 |
+
"grad_norm": 0.4211725890636444,
|
1456 |
+
"learning_rate": 6.181032333091078e-05,
|
1457 |
+
"loss": 0.0574,
|
1458 |
+
"step": 2070
|
1459 |
+
},
|
1460 |
+
{
|
1461 |
+
"epoch": 4.5514223194748356,
|
1462 |
+
"grad_norm": 0.4051927328109741,
|
1463 |
+
"learning_rate": 6.145840520496666e-05,
|
1464 |
+
"loss": 0.0555,
|
1465 |
+
"step": 2080
|
1466 |
+
},
|
1467 |
+
{
|
1468 |
+
"epoch": 4.573304157549234,
|
1469 |
+
"grad_norm": 0.4655281603336334,
|
1470 |
+
"learning_rate": 6.11058869530063e-05,
|
1471 |
+
"loss": 0.061,
|
1472 |
+
"step": 2090
|
1473 |
+
},
|
1474 |
+
{
|
1475 |
+
"epoch": 4.595185995623632,
|
1476 |
+
"grad_norm": 0.3950660526752472,
|
1477 |
+
"learning_rate": 6.0752787037928336e-05,
|
1478 |
+
"loss": 0.061,
|
1479 |
+
"step": 2100
|
1480 |
+
},
|
1481 |
+
{
|
1482 |
+
"epoch": 4.61706783369803,
|
1483 |
+
"grad_norm": 0.5274171233177185,
|
1484 |
+
"learning_rate": 6.039912395309568e-05,
|
1485 |
+
"loss": 0.0605,
|
1486 |
+
"step": 2110
|
1487 |
+
},
|
1488 |
+
{
|
1489 |
+
"epoch": 4.638949671772429,
|
1490 |
+
"grad_norm": 0.3764810860157013,
|
1491 |
+
"learning_rate": 6.004491622136681e-05,
|
1492 |
+
"loss": 0.0555,
|
1493 |
+
"step": 2120
|
1494 |
+
},
|
1495 |
+
{
|
1496 |
+
"epoch": 4.660831509846827,
|
1497 |
+
"grad_norm": 0.3669235110282898,
|
1498 |
+
"learning_rate": 5.969018239412573e-05,
|
1499 |
+
"loss": 0.0657,
|
1500 |
+
"step": 2130
|
1501 |
+
},
|
1502 |
+
{
|
1503 |
+
"epoch": 4.682713347921226,
|
1504 |
+
"grad_norm": 0.4793848395347595,
|
1505 |
+
"learning_rate": 5.933494105031032e-05,
|
1506 |
+
"loss": 0.0625,
|
1507 |
+
"step": 2140
|
1508 |
+
},
|
1509 |
+
{
|
1510 |
+
"epoch": 4.704595185995624,
|
1511 |
+
"grad_norm": 0.7370563745498657,
|
1512 |
+
"learning_rate": 5.897921079543931e-05,
|
1513 |
+
"loss": 0.0553,
|
1514 |
+
"step": 2150
|
1515 |
+
},
|
1516 |
+
{
|
1517 |
+
"epoch": 4.726477024070022,
|
1518 |
+
"grad_norm": 0.32018744945526123,
|
1519 |
+
"learning_rate": 5.862301026063777e-05,
|
1520 |
+
"loss": 0.055,
|
1521 |
+
"step": 2160
|
1522 |
+
},
|
1523 |
+
{
|
1524 |
+
"epoch": 4.74835886214442,
|
1525 |
+
"grad_norm": 0.39296776056289673,
|
1526 |
+
"learning_rate": 5.826635810166137e-05,
|
1527 |
+
"loss": 0.052,
|
1528 |
+
"step": 2170
|
1529 |
+
},
|
1530 |
+
{
|
1531 |
+
"epoch": 4.770240700218818,
|
1532 |
+
"grad_norm": 0.4568535089492798,
|
1533 |
+
"learning_rate": 5.7909272997919305e-05,
|
1534 |
+
"loss": 0.056,
|
1535 |
+
"step": 2180
|
1536 |
+
},
|
1537 |
+
{
|
1538 |
+
"epoch": 4.792122538293217,
|
1539 |
+
"grad_norm": 0.39834004640579224,
|
1540 |
+
"learning_rate": 5.7551773651495953e-05,
|
1541 |
+
"loss": 0.0586,
|
1542 |
+
"step": 2190
|
1543 |
+
},
|
1544 |
+
{
|
1545 |
+
"epoch": 4.814004376367615,
|
1546 |
+
"grad_norm": 0.47946035861968994,
|
1547 |
+
"learning_rate": 5.719387878617131e-05,
|
1548 |
+
"loss": 0.0609,
|
1549 |
+
"step": 2200
|
1550 |
+
},
|
1551 |
+
{
|
1552 |
+
"epoch": 4.835886214442013,
|
1553 |
+
"grad_norm": 0.3279586434364319,
|
1554 |
+
"learning_rate": 5.683560714644044e-05,
|
1555 |
+
"loss": 0.059,
|
1556 |
+
"step": 2210
|
1557 |
+
},
|
1558 |
+
{
|
1559 |
+
"epoch": 4.857768052516412,
|
1560 |
+
"grad_norm": 0.48775798082351685,
|
1561 |
+
"learning_rate": 5.647697749653172e-05,
|
1562 |
+
"loss": 0.0565,
|
1563 |
+
"step": 2220
|
1564 |
+
},
|
1565 |
+
{
|
1566 |
+
"epoch": 4.87964989059081,
|
1567 |
+
"grad_norm": 0.42671310901641846,
|
1568 |
+
"learning_rate": 5.6118008619424036e-05,
|
1569 |
+
"loss": 0.0605,
|
1570 |
+
"step": 2230
|
1571 |
+
},
|
1572 |
+
{
|
1573 |
+
"epoch": 4.901531728665208,
|
1574 |
+
"grad_norm": 0.3488156497478485,
|
1575 |
+
"learning_rate": 5.5758719315863076e-05,
|
1576 |
+
"loss": 0.0548,
|
1577 |
+
"step": 2240
|
1578 |
+
},
|
1579 |
+
{
|
1580 |
+
"epoch": 4.923413566739606,
|
1581 |
+
"grad_norm": 0.4897012710571289,
|
1582 |
+
"learning_rate": 5.539912840337661e-05,
|
1583 |
+
"loss": 0.0587,
|
1584 |
+
"step": 2250
|
1585 |
+
},
|
1586 |
+
{
|
1587 |
+
"epoch": 4.945295404814004,
|
1588 |
+
"grad_norm": 0.40971580147743225,
|
1589 |
+
"learning_rate": 5.5039254715289e-05,
|
1590 |
+
"loss": 0.0608,
|
1591 |
+
"step": 2260
|
1592 |
+
},
|
1593 |
+
{
|
1594 |
+
"epoch": 4.967177242888402,
|
1595 |
+
"grad_norm": 0.5318540334701538,
|
1596 |
+
"learning_rate": 5.46791170997348e-05,
|
1597 |
+
"loss": 0.0497,
|
1598 |
+
"step": 2270
|
1599 |
+
},
|
1600 |
+
{
|
1601 |
+
"epoch": 4.989059080962801,
|
1602 |
+
"grad_norm": 0.42819133400917053,
|
1603 |
+
"learning_rate": 5.4318734418671525e-05,
|
1604 |
+
"loss": 0.0636,
|
1605 |
+
"step": 2280
|
1606 |
+
},
|
1607 |
+
{
|
1608 |
+
"epoch": 5.010940919037199,
|
1609 |
+
"grad_norm": 0.4624245762825012,
|
1610 |
+
"learning_rate": 5.395812554689188e-05,
|
1611 |
+
"loss": 0.0613,
|
1612 |
+
"step": 2290
|
1613 |
+
},
|
1614 |
+
{
|
1615 |
+
"epoch": 5.032822757111598,
|
1616 |
+
"grad_norm": 0.4868052005767822,
|
1617 |
+
"learning_rate": 5.359730937103512e-05,
|
1618 |
+
"loss": 0.0523,
|
1619 |
+
"step": 2300
|
1620 |
+
},
|
1621 |
+
{
|
1622 |
+
"epoch": 5.054704595185996,
|
1623 |
+
"grad_norm": 0.6195753216743469,
|
1624 |
+
"learning_rate": 5.323630478859789e-05,
|
1625 |
+
"loss": 0.0604,
|
1626 |
+
"step": 2310
|
1627 |
+
},
|
1628 |
+
{
|
1629 |
+
"epoch": 5.076586433260394,
|
1630 |
+
"grad_norm": 0.38792890310287476,
|
1631 |
+
"learning_rate": 5.2875130706944574e-05,
|
1632 |
+
"loss": 0.0539,
|
1633 |
+
"step": 2320
|
1634 |
+
},
|
1635 |
+
{
|
1636 |
+
"epoch": 5.098468271334792,
|
1637 |
+
"grad_norm": 0.5414655208587646,
|
1638 |
+
"learning_rate": 5.251380604231687e-05,
|
1639 |
+
"loss": 0.0602,
|
1640 |
+
"step": 2330
|
1641 |
+
},
|
1642 |
+
{
|
1643 |
+
"epoch": 5.12035010940919,
|
1644 |
+
"grad_norm": 0.5284174680709839,
|
1645 |
+
"learning_rate": 5.215234971884321e-05,
|
1646 |
+
"loss": 0.0569,
|
1647 |
+
"step": 2340
|
1648 |
+
},
|
1649 |
+
{
|
1650 |
+
"epoch": 5.142231947483588,
|
1651 |
+
"grad_norm": 0.4137500822544098,
|
1652 |
+
"learning_rate": 5.1790780667547564e-05,
|
1653 |
+
"loss": 0.0595,
|
1654 |
+
"step": 2350
|
1655 |
+
},
|
1656 |
+
{
|
1657 |
+
"epoch": 5.164113785557987,
|
1658 |
+
"grad_norm": 0.456150084733963,
|
1659 |
+
"learning_rate": 5.1429117825357894e-05,
|
1660 |
+
"loss": 0.0652,
|
1661 |
+
"step": 2360
|
1662 |
+
},
|
1663 |
+
{
|
1664 |
+
"epoch": 5.185995623632385,
|
1665 |
+
"grad_norm": 0.45625388622283936,
|
1666 |
+
"learning_rate": 5.1067380134114484e-05,
|
1667 |
+
"loss": 0.0638,
|
1668 |
+
"step": 2370
|
1669 |
+
},
|
1670 |
+
{
|
1671 |
+
"epoch": 5.207877461706784,
|
1672 |
+
"grad_norm": 0.4778393507003784,
|
1673 |
+
"learning_rate": 5.0705586539577687e-05,
|
1674 |
+
"loss": 0.0562,
|
1675 |
+
"step": 2380
|
1676 |
+
},
|
1677 |
+
{
|
1678 |
+
"epoch": 5.229759299781182,
|
1679 |
+
"grad_norm": 0.45723408460617065,
|
1680 |
+
"learning_rate": 5.034375599043582e-05,
|
1681 |
+
"loss": 0.0578,
|
1682 |
+
"step": 2390
|
1683 |
+
},
|
1684 |
+
{
|
1685 |
+
"epoch": 5.25164113785558,
|
1686 |
+
"grad_norm": 0.4665261507034302,
|
1687 |
+
"learning_rate": 4.998190743731265e-05,
|
1688 |
+
"loss": 0.0483,
|
1689 |
+
"step": 2400
|
1690 |
+
},
|
1691 |
+
{
|
1692 |
+
"epoch": 5.273522975929978,
|
1693 |
+
"grad_norm": 0.37840256094932556,
|
1694 |
+
"learning_rate": 4.96200598317749e-05,
|
1695 |
+
"loss": 0.06,
|
1696 |
+
"step": 2410
|
1697 |
+
},
|
1698 |
+
{
|
1699 |
+
"epoch": 5.295404814004376,
|
1700 |
+
"grad_norm": 0.4239036738872528,
|
1701 |
+
"learning_rate": 4.925823212533963e-05,
|
1702 |
+
"loss": 0.0575,
|
1703 |
+
"step": 2420
|
1704 |
+
},
|
1705 |
+
{
|
1706 |
+
"epoch": 5.317286652078774,
|
1707 |
+
"grad_norm": 0.45726969838142395,
|
1708 |
+
"learning_rate": 4.8896443268481715e-05,
|
1709 |
+
"loss": 0.0523,
|
1710 |
+
"step": 2430
|
1711 |
+
},
|
1712 |
+
{
|
1713 |
+
"epoch": 5.339168490153173,
|
1714 |
+
"grad_norm": 0.3009950816631317,
|
1715 |
+
"learning_rate": 4.853471220964137e-05,
|
1716 |
+
"loss": 0.0573,
|
1717 |
+
"step": 2440
|
1718 |
+
},
|
1719 |
+
{
|
1720 |
+
"epoch": 5.361050328227571,
|
1721 |
+
"grad_norm": 0.5033520460128784,
|
1722 |
+
"learning_rate": 4.8173057894231585e-05,
|
1723 |
+
"loss": 0.0576,
|
1724 |
+
"step": 2450
|
1725 |
+
},
|
1726 |
+
{
|
1727 |
+
"epoch": 5.38293216630197,
|
1728 |
+
"grad_norm": 0.6416246294975281,
|
1729 |
+
"learning_rate": 4.7811499263646024e-05,
|
1730 |
+
"loss": 0.0566,
|
1731 |
+
"step": 2460
|
1732 |
+
},
|
1733 |
+
{
|
1734 |
+
"epoch": 5.404814004376368,
|
1735 |
+
"grad_norm": 0.45454248785972595,
|
1736 |
+
"learning_rate": 4.7450055254266875e-05,
|
1737 |
+
"loss": 0.0603,
|
1738 |
+
"step": 2470
|
1739 |
+
},
|
1740 |
+
{
|
1741 |
+
"epoch": 5.426695842450766,
|
1742 |
+
"grad_norm": 0.4150170087814331,
|
1743 |
+
"learning_rate": 4.708874479647319e-05,
|
1744 |
+
"loss": 0.061,
|
1745 |
+
"step": 2480
|
1746 |
+
},
|
1747 |
+
{
|
1748 |
+
"epoch": 5.4485776805251644,
|
1749 |
+
"grad_norm": 0.7388213276863098,
|
1750 |
+
"learning_rate": 4.6727586813649324e-05,
|
1751 |
+
"loss": 0.0559,
|
1752 |
+
"step": 2490
|
1753 |
+
},
|
1754 |
+
{
|
1755 |
+
"epoch": 5.470459518599562,
|
1756 |
+
"grad_norm": 0.42447715997695923,
|
1757 |
+
"learning_rate": 4.6366600221193815e-05,
|
1758 |
+
"loss": 0.0523,
|
1759 |
+
"step": 2500
|
1760 |
+
},
|
1761 |
+
{
|
1762 |
+
"epoch": 5.49234135667396,
|
1763 |
+
"grad_norm": 0.35511597990989685,
|
1764 |
+
"learning_rate": 4.60058039255288e-05,
|
1765 |
+
"loss": 0.0494,
|
1766 |
+
"step": 2510
|
1767 |
+
},
|
1768 |
+
{
|
1769 |
+
"epoch": 5.514223194748359,
|
1770 |
+
"grad_norm": 0.47702649235725403,
|
1771 |
+
"learning_rate": 4.5645216823109727e-05,
|
1772 |
+
"loss": 0.0621,
|
1773 |
+
"step": 2520
|
1774 |
+
},
|
1775 |
+
{
|
1776 |
+
"epoch": 5.536105032822757,
|
1777 |
+
"grad_norm": 0.5418275594711304,
|
1778 |
+
"learning_rate": 4.5284857799435726e-05,
|
1779 |
+
"loss": 0.0569,
|
1780 |
+
"step": 2530
|
1781 |
+
},
|
1782 |
+
{
|
1783 |
+
"epoch": 5.557986870897155,
|
1784 |
+
"grad_norm": 0.2867294251918793,
|
1785 |
+
"learning_rate": 4.4924745728060444e-05,
|
1786 |
+
"loss": 0.0575,
|
1787 |
+
"step": 2540
|
1788 |
+
},
|
1789 |
+
{
|
1790 |
+
"epoch": 5.579868708971554,
|
1791 |
+
"grad_norm": 0.3692294657230377,
|
1792 |
+
"learning_rate": 4.456489946960353e-05,
|
1793 |
+
"loss": 0.0621,
|
1794 |
+
"step": 2550
|
1795 |
+
},
|
1796 |
+
{
|
1797 |
+
"epoch": 5.601750547045952,
|
1798 |
+
"grad_norm": 0.5166872143745422,
|
1799 |
+
"learning_rate": 4.4205337870762944e-05,
|
1800 |
+
"loss": 0.0514,
|
1801 |
+
"step": 2560
|
1802 |
+
},
|
1803 |
+
{
|
1804 |
+
"epoch": 5.6236323851203505,
|
1805 |
+
"grad_norm": 0.25781774520874023,
|
1806 |
+
"learning_rate": 4.384607976332773e-05,
|
1807 |
+
"loss": 0.0526,
|
1808 |
+
"step": 2570
|
1809 |
+
},
|
1810 |
+
{
|
1811 |
+
"epoch": 5.645514223194748,
|
1812 |
+
"grad_norm": 0.458324670791626,
|
1813 |
+
"learning_rate": 4.348714396319182e-05,
|
1814 |
+
"loss": 0.0618,
|
1815 |
+
"step": 2580
|
1816 |
+
},
|
1817 |
+
{
|
1818 |
+
"epoch": 5.667396061269146,
|
1819 |
+
"grad_norm": 0.5077418088912964,
|
1820 |
+
"learning_rate": 4.3128549269368525e-05,
|
1821 |
+
"loss": 0.057,
|
1822 |
+
"step": 2590
|
1823 |
+
},
|
1824 |
+
{
|
1825 |
+
"epoch": 5.689277899343545,
|
1826 |
+
"grad_norm": 0.3836349844932556,
|
1827 |
+
"learning_rate": 4.27703144630059e-05,
|
1828 |
+
"loss": 0.0548,
|
1829 |
+
"step": 2600
|
1830 |
+
},
|
1831 |
+
{
|
1832 |
+
"epoch": 5.711159737417943,
|
1833 |
+
"grad_norm": 0.29659900069236755,
|
1834 |
+
"learning_rate": 4.2412458306403224e-05,
|
1835 |
+
"loss": 0.0535,
|
1836 |
+
"step": 2610
|
1837 |
+
},
|
1838 |
+
{
|
1839 |
+
"epoch": 5.733041575492342,
|
1840 |
+
"grad_norm": 0.41197437047958374,
|
1841 |
+
"learning_rate": 4.2054999542028174e-05,
|
1842 |
+
"loss": 0.0553,
|
1843 |
+
"step": 2620
|
1844 |
+
},
|
1845 |
+
{
|
1846 |
+
"epoch": 5.75492341356674,
|
1847 |
+
"grad_norm": 0.49191319942474365,
|
1848 |
+
"learning_rate": 4.169795689153535e-05,
|
1849 |
+
"loss": 0.0565,
|
1850 |
+
"step": 2630
|
1851 |
+
},
|
1852 |
+
{
|
1853 |
+
"epoch": 5.776805251641138,
|
1854 |
+
"grad_norm": 0.5994260907173157,
|
1855 |
+
"learning_rate": 4.134134905478566e-05,
|
1856 |
+
"loss": 0.0608,
|
1857 |
+
"step": 2640
|
1858 |
+
},
|
1859 |
+
{
|
1860 |
+
"epoch": 5.7986870897155365,
|
1861 |
+
"grad_norm": 0.5599576234817505,
|
1862 |
+
"learning_rate": 4.0985194708866904e-05,
|
1863 |
+
"loss": 0.0552,
|
1864 |
+
"step": 2650
|
1865 |
+
},
|
1866 |
+
{
|
1867 |
+
"epoch": 5.820568927789934,
|
1868 |
+
"grad_norm": 0.48255977034568787,
|
1869 |
+
"learning_rate": 4.06295125071157e-05,
|
1870 |
+
"loss": 0.0587,
|
1871 |
+
"step": 2660
|
1872 |
+
},
|
1873 |
+
{
|
1874 |
+
"epoch": 5.842450765864332,
|
1875 |
+
"grad_norm": 0.5754146575927734,
|
1876 |
+
"learning_rate": 4.027432107814032e-05,
|
1877 |
+
"loss": 0.0604,
|
1878 |
+
"step": 2670
|
1879 |
+
},
|
1880 |
+
{
|
1881 |
+
"epoch": 5.864332603938731,
|
1882 |
+
"grad_norm": 0.6725056767463684,
|
1883 |
+
"learning_rate": 3.991963902484527e-05,
|
1884 |
+
"loss": 0.0594,
|
1885 |
+
"step": 2680
|
1886 |
+
},
|
1887 |
+
{
|
1888 |
+
"epoch": 5.886214442013129,
|
1889 |
+
"grad_norm": 0.5410199761390686,
|
1890 |
+
"learning_rate": 3.956548492345681e-05,
|
1891 |
+
"loss": 0.0606,
|
1892 |
+
"step": 2690
|
1893 |
+
},
|
1894 |
+
{
|
1895 |
+
"epoch": 5.908096280087527,
|
1896 |
+
"grad_norm": 0.2633671164512634,
|
1897 |
+
"learning_rate": 3.921187732255006e-05,
|
1898 |
+
"loss": 0.053,
|
1899 |
+
"step": 2700
|
1900 |
+
},
|
1901 |
+
{
|
1902 |
+
"epoch": 5.929978118161926,
|
1903 |
+
"grad_norm": 0.3890124261379242,
|
1904 |
+
"learning_rate": 3.8858834742077624e-05,
|
1905 |
+
"loss": 0.056,
|
1906 |
+
"step": 2710
|
1907 |
+
},
|
1908 |
+
{
|
1909 |
+
"epoch": 5.951859956236324,
|
1910 |
+
"grad_norm": 0.4105083644390106,
|
1911 |
+
"learning_rate": 3.8506375672399466e-05,
|
1912 |
+
"loss": 0.0513,
|
1913 |
+
"step": 2720
|
1914 |
+
},
|
1915 |
+
{
|
1916 |
+
"epoch": 5.9737417943107225,
|
1917 |
+
"grad_norm": 0.4586251378059387,
|
1918 |
+
"learning_rate": 3.815451857331465e-05,
|
1919 |
+
"loss": 0.0557,
|
1920 |
+
"step": 2730
|
1921 |
+
},
|
1922 |
+
{
|
1923 |
+
"epoch": 5.99562363238512,
|
1924 |
+
"grad_norm": 0.23562565445899963,
|
1925 |
+
"learning_rate": 3.780328187309443e-05,
|
1926 |
+
"loss": 0.0536,
|
1927 |
+
"step": 2740
|
1928 |
+
},
|
1929 |
+
{
|
1930 |
+
"epoch": 6.017505470459518,
|
1931 |
+
"grad_norm": 0.4445761740207672,
|
1932 |
+
"learning_rate": 3.7452683967517057e-05,
|
1933 |
+
"loss": 0.0475,
|
1934 |
+
"step": 2750
|
1935 |
+
},
|
1936 |
+
{
|
1937 |
+
"epoch": 6.039387308533917,
|
1938 |
+
"grad_norm": 0.7048770785331726,
|
1939 |
+
"learning_rate": 3.710274321890443e-05,
|
1940 |
+
"loss": 0.0506,
|
1941 |
+
"step": 2760
|
1942 |
+
},
|
1943 |
+
{
|
1944 |
+
"epoch": 6.061269146608315,
|
1945 |
+
"grad_norm": 0.41764745116233826,
|
1946 |
+
"learning_rate": 3.6753477955160244e-05,
|
1947 |
+
"loss": 0.0533,
|
1948 |
+
"step": 2770
|
1949 |
+
},
|
1950 |
+
{
|
1951 |
+
"epoch": 6.083150984682713,
|
1952 |
+
"grad_norm": 0.30127957463264465,
|
1953 |
+
"learning_rate": 3.640490646881018e-05,
|
1954 |
+
"loss": 0.0571,
|
1955 |
+
"step": 2780
|
1956 |
+
},
|
1957 |
+
{
|
1958 |
+
"epoch": 6.105032822757112,
|
1959 |
+
"grad_norm": 0.3346746563911438,
|
1960 |
+
"learning_rate": 3.605704701604381e-05,
|
1961 |
+
"loss": 0.051,
|
1962 |
+
"step": 2790
|
1963 |
+
},
|
1964 |
+
{
|
1965 |
+
"epoch": 6.12691466083151,
|
1966 |
+
"grad_norm": 0.5310165882110596,
|
1967 |
+
"learning_rate": 3.5709917815758385e-05,
|
1968 |
+
"loss": 0.0546,
|
1969 |
+
"step": 2800
|
1970 |
+
},
|
1971 |
+
{
|
1972 |
+
"epoch": 6.1487964989059085,
|
1973 |
+
"grad_norm": 0.3638152778148651,
|
1974 |
+
"learning_rate": 3.536353704860479e-05,
|
1975 |
+
"loss": 0.049,
|
1976 |
+
"step": 2810
|
1977 |
+
},
|
1978 |
+
{
|
1979 |
+
"epoch": 6.170678336980306,
|
1980 |
+
"grad_norm": 0.37336069345474243,
|
1981 |
+
"learning_rate": 3.5017922856035155e-05,
|
1982 |
+
"loss": 0.0582,
|
1983 |
+
"step": 2820
|
1984 |
+
},
|
1985 |
+
{
|
1986 |
+
"epoch": 6.192560175054704,
|
1987 |
+
"grad_norm": 0.5262767672538757,
|
1988 |
+
"learning_rate": 3.467309333935283e-05,
|
1989 |
+
"loss": 0.0517,
|
1990 |
+
"step": 2830
|
1991 |
+
},
|
1992 |
+
{
|
1993 |
+
"epoch": 6.214442013129103,
|
1994 |
+
"grad_norm": 0.28704574704170227,
|
1995 |
+
"learning_rate": 3.4329066558764355e-05,
|
1996 |
+
"loss": 0.0458,
|
1997 |
+
"step": 2840
|
1998 |
+
},
|
1999 |
+
{
|
2000 |
+
"epoch": 6.236323851203501,
|
2001 |
+
"grad_norm": 0.4320286810398102,
|
2002 |
+
"learning_rate": 3.398586053243341e-05,
|
2003 |
+
"loss": 0.0489,
|
2004 |
+
"step": 2850
|
2005 |
+
},
|
2006 |
+
{
|
2007 |
+
"epoch": 6.258205689277899,
|
2008 |
+
"grad_norm": 0.43195730447769165,
|
2009 |
+
"learning_rate": 3.364349323553738e-05,
|
2010 |
+
"loss": 0.054,
|
2011 |
+
"step": 2860
|
2012 |
+
},
|
2013 |
+
{
|
2014 |
+
"epoch": 6.280087527352298,
|
2015 |
+
"grad_norm": 0.3144959509372711,
|
2016 |
+
"learning_rate": 3.330198259932567e-05,
|
2017 |
+
"loss": 0.0537,
|
2018 |
+
"step": 2870
|
2019 |
+
},
|
2020 |
+
{
|
2021 |
+
"epoch": 6.301969365426696,
|
2022 |
+
"grad_norm": 0.4727848172187805,
|
2023 |
+
"learning_rate": 3.2961346510180716e-05,
|
2024 |
+
"loss": 0.0599,
|
2025 |
+
"step": 2880
|
2026 |
+
},
|
2027 |
+
{
|
2028 |
+
"epoch": 6.3238512035010945,
|
2029 |
+
"grad_norm": 0.2918757498264313,
|
2030 |
+
"learning_rate": 3.26216028086812e-05,
|
2031 |
+
"loss": 0.057,
|
2032 |
+
"step": 2890
|
2033 |
+
},
|
2034 |
+
{
|
2035 |
+
"epoch": 6.3457330415754925,
|
2036 |
+
"grad_norm": 0.8075538873672485,
|
2037 |
+
"learning_rate": 3.2282769288667525e-05,
|
2038 |
+
"loss": 0.0583,
|
2039 |
+
"step": 2900
|
2040 |
+
},
|
2041 |
+
{
|
2042 |
+
"epoch": 6.36761487964989,
|
2043 |
+
"grad_norm": 0.25191056728363037,
|
2044 |
+
"learning_rate": 3.194486369631007e-05,
|
2045 |
+
"loss": 0.0558,
|
2046 |
+
"step": 2910
|
2047 |
+
},
|
2048 |
+
{
|
2049 |
+
"epoch": 6.389496717724289,
|
2050 |
+
"grad_norm": 0.3775458335876465,
|
2051 |
+
"learning_rate": 3.1607903729179576e-05,
|
2052 |
+
"loss": 0.0508,
|
2053 |
+
"step": 2920
|
2054 |
+
},
|
2055 |
+
{
|
2056 |
+
"epoch": 6.411378555798687,
|
2057 |
+
"grad_norm": 0.4212424159049988,
|
2058 |
+
"learning_rate": 3.127190703532036e-05,
|
2059 |
+
"loss": 0.0581,
|
2060 |
+
"step": 2930
|
2061 |
+
},
|
2062 |
+
{
|
2063 |
+
"epoch": 6.433260393873085,
|
2064 |
+
"grad_norm": 0.3469920754432678,
|
2065 |
+
"learning_rate": 3.0936891212325966e-05,
|
2066 |
+
"loss": 0.0537,
|
2067 |
+
"step": 2940
|
2068 |
+
},
|
2069 |
+
{
|
2070 |
+
"epoch": 6.455142231947484,
|
2071 |
+
"grad_norm": 0.3886861801147461,
|
2072 |
+
"learning_rate": 3.0602873806417484e-05,
|
2073 |
+
"loss": 0.0601,
|
2074 |
+
"step": 2950
|
2075 |
+
},
|
2076 |
+
{
|
2077 |
+
"epoch": 6.477024070021882,
|
2078 |
+
"grad_norm": 0.4389936923980713,
|
2079 |
+
"learning_rate": 3.0269872311524626e-05,
|
2080 |
+
"loss": 0.0554,
|
2081 |
+
"step": 2960
|
2082 |
+
},
|
2083 |
+
{
|
2084 |
+
"epoch": 6.49890590809628,
|
2085 |
+
"grad_norm": 0.4016413390636444,
|
2086 |
+
"learning_rate": 2.9937904168369446e-05,
|
2087 |
+
"loss": 0.0505,
|
2088 |
+
"step": 2970
|
2089 |
+
},
|
2090 |
+
{
|
2091 |
+
"epoch": 6.5207877461706785,
|
2092 |
+
"grad_norm": 0.4214434325695038,
|
2093 |
+
"learning_rate": 2.9606986763552935e-05,
|
2094 |
+
"loss": 0.0484,
|
2095 |
+
"step": 2980
|
2096 |
+
},
|
2097 |
+
{
|
2098 |
+
"epoch": 6.542669584245076,
|
2099 |
+
"grad_norm": 0.432650625705719,
|
2100 |
+
"learning_rate": 2.927713742864442e-05,
|
2101 |
+
"loss": 0.0544,
|
2102 |
+
"step": 2990
|
2103 |
+
},
|
2104 |
+
{
|
2105 |
+
"epoch": 6.564551422319475,
|
2106 |
+
"grad_norm": 0.22795255482196808,
|
2107 |
+
"learning_rate": 2.8948373439273734e-05,
|
2108 |
+
"loss": 0.0496,
|
2109 |
+
"step": 3000
|
2110 |
+
},
|
2111 |
+
{
|
2112 |
+
"epoch": 6.586433260393873,
|
2113 |
+
"grad_norm": 0.3206609785556793,
|
2114 |
+
"learning_rate": 2.862071201422659e-05,
|
2115 |
+
"loss": 0.0478,
|
2116 |
+
"step": 3010
|
2117 |
+
},
|
2118 |
+
{
|
2119 |
+
"epoch": 6.608315098468271,
|
2120 |
+
"grad_norm": 0.4075492024421692,
|
2121 |
+
"learning_rate": 2.8294170314542567e-05,
|
2122 |
+
"loss": 0.0528,
|
2123 |
+
"step": 3020
|
2124 |
+
},
|
2125 |
+
{
|
2126 |
+
"epoch": 6.63019693654267,
|
2127 |
+
"grad_norm": 0.2718966007232666,
|
2128 |
+
"learning_rate": 2.796876544261645e-05,
|
2129 |
+
"loss": 0.0492,
|
2130 |
+
"step": 3030
|
2131 |
+
},
|
2132 |
+
{
|
2133 |
+
"epoch": 6.652078774617068,
|
2134 |
+
"grad_norm": 0.33011844754219055,
|
2135 |
+
"learning_rate": 2.7644514441302466e-05,
|
2136 |
+
"loss": 0.0595,
|
2137 |
+
"step": 3040
|
2138 |
+
},
|
2139 |
+
{
|
2140 |
+
"epoch": 6.673960612691467,
|
2141 |
+
"grad_norm": 0.3843638002872467,
|
2142 |
+
"learning_rate": 2.732143429302165e-05,
|
2143 |
+
"loss": 0.0598,
|
2144 |
+
"step": 3050
|
2145 |
+
},
|
2146 |
+
{
|
2147 |
+
"epoch": 6.6958424507658645,
|
2148 |
+
"grad_norm": 0.2999420166015625,
|
2149 |
+
"learning_rate": 2.699954191887244e-05,
|
2150 |
+
"loss": 0.0517,
|
2151 |
+
"step": 3060
|
2152 |
+
},
|
2153 |
+
{
|
2154 |
+
"epoch": 6.717724288840262,
|
2155 |
+
"grad_norm": 0.4345828890800476,
|
2156 |
+
"learning_rate": 2.6678854177744416e-05,
|
2157 |
+
"loss": 0.0623,
|
2158 |
+
"step": 3070
|
2159 |
+
},
|
2160 |
+
{
|
2161 |
+
"epoch": 6.739606126914661,
|
2162 |
+
"grad_norm": 0.4258360266685486,
|
2163 |
+
"learning_rate": 2.6359387865435346e-05,
|
2164 |
+
"loss": 0.0461,
|
2165 |
+
"step": 3080
|
2166 |
+
},
|
2167 |
+
{
|
2168 |
+
"epoch": 6.761487964989059,
|
2169 |
+
"grad_norm": 0.44895169138908386,
|
2170 |
+
"learning_rate": 2.6041159713771522e-05,
|
2171 |
+
"loss": 0.0558,
|
2172 |
+
"step": 3090
|
2173 |
+
},
|
2174 |
+
{
|
2175 |
+
"epoch": 6.783369803063457,
|
2176 |
+
"grad_norm": 0.47770532965660095,
|
2177 |
+
"learning_rate": 2.5724186389731364e-05,
|
2178 |
+
"loss": 0.0556,
|
2179 |
+
"step": 3100
|
2180 |
+
},
|
2181 |
+
{
|
2182 |
+
"epoch": 6.805251641137856,
|
2183 |
+
"grad_norm": 0.32761141657829285,
|
2184 |
+
"learning_rate": 2.5408484494572704e-05,
|
2185 |
+
"loss": 0.0593,
|
2186 |
+
"step": 3110
|
2187 |
+
},
|
2188 |
+
{
|
2189 |
+
"epoch": 6.827133479212254,
|
2190 |
+
"grad_norm": 0.47967493534088135,
|
2191 |
+
"learning_rate": 2.509407056296303e-05,
|
2192 |
+
"loss": 0.0563,
|
2193 |
+
"step": 3120
|
2194 |
+
},
|
2195 |
+
{
|
2196 |
+
"epoch": 6.849015317286652,
|
2197 |
+
"grad_norm": 0.3491036295890808,
|
2198 |
+
"learning_rate": 2.4780961062113684e-05,
|
2199 |
+
"loss": 0.0527,
|
2200 |
+
"step": 3130
|
2201 |
+
},
|
2202 |
+
{
|
2203 |
+
"epoch": 6.8708971553610505,
|
2204 |
+
"grad_norm": 0.32572147250175476,
|
2205 |
+
"learning_rate": 2.4469172390917417e-05,
|
2206 |
+
"loss": 0.0497,
|
2207 |
+
"step": 3140
|
2208 |
+
},
|
2209 |
+
{
|
2210 |
+
"epoch": 6.892778993435448,
|
2211 |
+
"grad_norm": 0.42841556668281555,
|
2212 |
+
"learning_rate": 2.4158720879089337e-05,
|
2213 |
+
"loss": 0.0487,
|
2214 |
+
"step": 3150
|
2215 |
+
},
|
2216 |
+
{
|
2217 |
+
"epoch": 6.914660831509847,
|
2218 |
+
"grad_norm": 0.3758542835712433,
|
2219 |
+
"learning_rate": 2.384962278631182e-05,
|
2220 |
+
"loss": 0.0582,
|
2221 |
+
"step": 3160
|
2222 |
+
},
|
2223 |
+
{
|
2224 |
+
"epoch": 6.936542669584245,
|
2225 |
+
"grad_norm": 0.32348281145095825,
|
2226 |
+
"learning_rate": 2.354189430138285e-05,
|
2227 |
+
"loss": 0.0558,
|
2228 |
+
"step": 3170
|
2229 |
+
},
|
2230 |
+
{
|
2231 |
+
"epoch": 6.958424507658643,
|
2232 |
+
"grad_norm": 0.3324337601661682,
|
2233 |
+
"learning_rate": 2.323555154136815e-05,
|
2234 |
+
"loss": 0.0486,
|
2235 |
+
"step": 3180
|
2236 |
+
},
|
2237 |
+
{
|
2238 |
+
"epoch": 6.980306345733042,
|
2239 |
+
"grad_norm": 0.3877069652080536,
|
2240 |
+
"learning_rate": 2.293061055075707e-05,
|
2241 |
+
"loss": 0.0534,
|
2242 |
+
"step": 3190
|
2243 |
+
},
|
2244 |
+
{
|
2245 |
+
"epoch": 7.00218818380744,
|
2246 |
+
"grad_norm": 0.3602635860443115,
|
2247 |
+
"learning_rate": 2.2627087300622208e-05,
|
2248 |
+
"loss": 0.0528,
|
2249 |
+
"step": 3200
|
2250 |
+
},
|
2251 |
+
{
|
2252 |
+
"epoch": 7.024070021881838,
|
2253 |
+
"grad_norm": 0.4806849956512451,
|
2254 |
+
"learning_rate": 2.2324997687783102e-05,
|
2255 |
+
"loss": 0.0529,
|
2256 |
+
"step": 3210
|
2257 |
+
},
|
2258 |
+
{
|
2259 |
+
"epoch": 7.0459518599562365,
|
2260 |
+
"grad_norm": 0.4764333963394165,
|
2261 |
+
"learning_rate": 2.20243575339734e-05,
|
2262 |
+
"loss": 0.0532,
|
2263 |
+
"step": 3220
|
2264 |
+
},
|
2265 |
+
{
|
2266 |
+
"epoch": 7.067833698030634,
|
2267 |
+
"grad_norm": 0.4383291006088257,
|
2268 |
+
"learning_rate": 2.172518258501241e-05,
|
2269 |
+
"loss": 0.0482,
|
2270 |
+
"step": 3230
|
2271 |
+
},
|
2272 |
+
{
|
2273 |
+
"epoch": 7.089715536105033,
|
2274 |
+
"grad_norm": 0.4792105257511139,
|
2275 |
+
"learning_rate": 2.1427488509980402e-05,
|
2276 |
+
"loss": 0.0508,
|
2277 |
+
"step": 3240
|
2278 |
+
},
|
2279 |
+
{
|
2280 |
+
"epoch": 7.111597374179431,
|
2281 |
+
"grad_norm": 0.4630696475505829,
|
2282 |
+
"learning_rate": 2.1131290900397793e-05,
|
2283 |
+
"loss": 0.0534,
|
2284 |
+
"step": 3250
|
2285 |
+
},
|
2286 |
+
{
|
2287 |
+
"epoch": 7.133479212253829,
|
2288 |
+
"grad_norm": 0.5735715627670288,
|
2289 |
+
"learning_rate": 2.0836605269408722e-05,
|
2290 |
+
"loss": 0.0517,
|
2291 |
+
"step": 3260
|
2292 |
+
},
|
2293 |
+
{
|
2294 |
+
"epoch": 7.155361050328228,
|
2295 |
+
"grad_norm": 0.3468039929866791,
|
2296 |
+
"learning_rate": 2.05434470509685e-05,
|
2297 |
+
"loss": 0.0485,
|
2298 |
+
"step": 3270
|
2299 |
+
},
|
2300 |
+
{
|
2301 |
+
"epoch": 7.177242888402626,
|
2302 |
+
"grad_norm": 0.27370214462280273,
|
2303 |
+
"learning_rate": 2.025183159903526e-05,
|
2304 |
+
"loss": 0.0564,
|
2305 |
+
"step": 3280
|
2306 |
+
},
|
2307 |
+
{
|
2308 |
+
"epoch": 7.199124726477024,
|
2309 |
+
"grad_norm": 0.43867847323417664,
|
2310 |
+
"learning_rate": 1.9961774186765796e-05,
|
2311 |
+
"loss": 0.0454,
|
2312 |
+
"step": 3290
|
2313 |
+
},
|
2314 |
+
{
|
2315 |
+
"epoch": 7.2210065645514225,
|
2316 |
+
"grad_norm": 0.36127227544784546,
|
2317 |
+
"learning_rate": 1.9673290005715683e-05,
|
2318 |
+
"loss": 0.0539,
|
2319 |
+
"step": 3300
|
2320 |
+
},
|
2321 |
+
{
|
2322 |
+
"epoch": 7.2428884026258205,
|
2323 |
+
"grad_norm": 0.3754895329475403,
|
2324 |
+
"learning_rate": 1.93863941650436e-05,
|
2325 |
+
"loss": 0.0466,
|
2326 |
+
"step": 3310
|
2327 |
+
},
|
2328 |
+
{
|
2329 |
+
"epoch": 7.264770240700219,
|
2330 |
+
"grad_norm": 0.3637754023075104,
|
2331 |
+
"learning_rate": 1.910110169071995e-05,
|
2332 |
+
"loss": 0.0494,
|
2333 |
+
"step": 3320
|
2334 |
+
},
|
2335 |
+
{
|
2336 |
+
"epoch": 7.286652078774617,
|
2337 |
+
"grad_norm": 0.5620509386062622,
|
2338 |
+
"learning_rate": 1.8817427524739988e-05,
|
2339 |
+
"loss": 0.046,
|
2340 |
+
"step": 3330
|
2341 |
+
},
|
2342 |
+
{
|
2343 |
+
"epoch": 7.308533916849015,
|
2344 |
+
"grad_norm": 0.5723752379417419,
|
2345 |
+
"learning_rate": 1.8535386524341225e-05,
|
2346 |
+
"loss": 0.0489,
|
2347 |
+
"step": 3340
|
2348 |
+
},
|
2349 |
+
{
|
2350 |
+
"epoch": 7.330415754923414,
|
2351 |
+
"grad_norm": 0.5323576927185059,
|
2352 |
+
"learning_rate": 1.8254993461225172e-05,
|
2353 |
+
"loss": 0.0522,
|
2354 |
+
"step": 3350
|
2355 |
+
},
|
2356 |
+
{
|
2357 |
+
"epoch": 7.352297592997812,
|
2358 |
+
"grad_norm": 0.40650686621665955,
|
2359 |
+
"learning_rate": 1.797626302078383e-05,
|
2360 |
+
"loss": 0.0489,
|
2361 |
+
"step": 3360
|
2362 |
+
},
|
2363 |
+
{
|
2364 |
+
"epoch": 7.37417943107221,
|
2365 |
+
"grad_norm": 0.4388726055622101,
|
2366 |
+
"learning_rate": 1.7699209801330468e-05,
|
2367 |
+
"loss": 0.0486,
|
2368 |
+
"step": 3370
|
2369 |
+
},
|
2370 |
+
{
|
2371 |
+
"epoch": 7.3960612691466086,
|
2372 |
+
"grad_norm": 0.292429119348526,
|
2373 |
+
"learning_rate": 1.7423848313335084e-05,
|
2374 |
+
"loss": 0.0464,
|
2375 |
+
"step": 3380
|
2376 |
+
},
|
2377 |
+
{
|
2378 |
+
"epoch": 7.4179431072210065,
|
2379 |
+
"grad_norm": 0.37241464853286743,
|
2380 |
+
"learning_rate": 1.715019297866441e-05,
|
2381 |
+
"loss": 0.0509,
|
2382 |
+
"step": 3390
|
2383 |
+
},
|
2384 |
+
{
|
2385 |
+
"epoch": 7.439824945295404,
|
2386 |
+
"grad_norm": 0.30430710315704346,
|
2387 |
+
"learning_rate": 1.6878258129826575e-05,
|
2388 |
+
"loss": 0.0501,
|
2389 |
+
"step": 3400
|
2390 |
+
},
|
2391 |
+
{
|
2392 |
+
"epoch": 7.461706783369803,
|
2393 |
+
"grad_norm": 0.3546319305896759,
|
2394 |
+
"learning_rate": 1.660805800922049e-05,
|
2395 |
+
"loss": 0.0563,
|
2396 |
+
"step": 3410
|
2397 |
+
},
|
2398 |
+
{
|
2399 |
+
"epoch": 7.483588621444201,
|
2400 |
+
"grad_norm": 0.3407646417617798,
|
2401 |
+
"learning_rate": 1.6339606768389792e-05,
|
2402 |
+
"loss": 0.048,
|
2403 |
+
"step": 3420
|
2404 |
+
},
|
2405 |
+
{
|
2406 |
+
"epoch": 7.5054704595186,
|
2407 |
+
"grad_norm": 0.31406334042549133,
|
2408 |
+
"learning_rate": 1.6072918467281872e-05,
|
2409 |
+
"loss": 0.0481,
|
2410 |
+
"step": 3430
|
2411 |
+
},
|
2412 |
+
{
|
2413 |
+
"epoch": 7.527352297592998,
|
2414 |
+
"grad_norm": 0.4938783645629883,
|
2415 |
+
"learning_rate": 1.5808007073511317e-05,
|
2416 |
+
"loss": 0.0477,
|
2417 |
+
"step": 3440
|
2418 |
+
},
|
2419 |
+
{
|
2420 |
+
"epoch": 7.549234135667396,
|
2421 |
+
"grad_norm": 0.3935304880142212,
|
2422 |
+
"learning_rate": 1.55448864616284e-05,
|
2423 |
+
"loss": 0.0478,
|
2424 |
+
"step": 3450
|
2425 |
+
},
|
2426 |
+
{
|
2427 |
+
"epoch": 7.571115973741795,
|
2428 |
+
"grad_norm": 0.5982438921928406,
|
2429 |
+
"learning_rate": 1.528357041239248e-05,
|
2430 |
+
"loss": 0.0552,
|
2431 |
+
"step": 3460
|
2432 |
+
},
|
2433 |
+
{
|
2434 |
+
"epoch": 7.5929978118161925,
|
2435 |
+
"grad_norm": 0.5672005414962769,
|
2436 |
+
"learning_rate": 1.502407261205019e-05,
|
2437 |
+
"loss": 0.0483,
|
2438 |
+
"step": 3470
|
2439 |
+
},
|
2440 |
+
{
|
2441 |
+
"epoch": 7.614879649890591,
|
2442 |
+
"grad_norm": 0.539139986038208,
|
2443 |
+
"learning_rate": 1.4766406651618626e-05,
|
2444 |
+
"loss": 0.0503,
|
2445 |
+
"step": 3480
|
2446 |
+
},
|
2447 |
+
{
|
2448 |
+
"epoch": 7.636761487964989,
|
2449 |
+
"grad_norm": 0.3828607201576233,
|
2450 |
+
"learning_rate": 1.4510586026173557e-05,
|
2451 |
+
"loss": 0.0547,
|
2452 |
+
"step": 3490
|
2453 |
+
},
|
2454 |
+
{
|
2455 |
+
"epoch": 7.658643326039387,
|
2456 |
+
"grad_norm": 0.40856969356536865,
|
2457 |
+
"learning_rate": 1.4256624134142604e-05,
|
2458 |
+
"loss": 0.046,
|
2459 |
+
"step": 3500
|
2460 |
+
},
|
2461 |
+
{
|
2462 |
+
"epoch": 7.680525164113786,
|
2463 |
+
"grad_norm": 0.8226433396339417,
|
2464 |
+
"learning_rate": 1.4004534276603538e-05,
|
2465 |
+
"loss": 0.0531,
|
2466 |
+
"step": 3510
|
2467 |
+
},
|
2468 |
+
{
|
2469 |
+
"epoch": 7.702407002188184,
|
2470 |
+
"grad_norm": 0.393865704536438,
|
2471 |
+
"learning_rate": 1.3754329656587556e-05,
|
2472 |
+
"loss": 0.0487,
|
2473 |
+
"step": 3520
|
2474 |
+
},
|
2475 |
+
{
|
2476 |
+
"epoch": 7.724288840262582,
|
2477 |
+
"grad_norm": 0.2557559907436371,
|
2478 |
+
"learning_rate": 1.3506023378387944e-05,
|
2479 |
+
"loss": 0.0459,
|
2480 |
+
"step": 3530
|
2481 |
+
},
|
2482 |
+
{
|
2483 |
+
"epoch": 7.746170678336981,
|
2484 |
+
"grad_norm": 0.3025050461292267,
|
2485 |
+
"learning_rate": 1.3259628446873601e-05,
|
2486 |
+
"loss": 0.0516,
|
2487 |
+
"step": 3540
|
2488 |
+
},
|
2489 |
+
{
|
2490 |
+
"epoch": 7.7680525164113785,
|
2491 |
+
"grad_norm": 0.46884235739707947,
|
2492 |
+
"learning_rate": 1.3015157766807939e-05,
|
2493 |
+
"loss": 0.0541,
|
2494 |
+
"step": 3550
|
2495 |
+
},
|
2496 |
+
{
|
2497 |
+
"epoch": 7.789934354485776,
|
2498 |
+
"grad_norm": 0.4090341627597809,
|
2499 |
+
"learning_rate": 1.2772624142173123e-05,
|
2500 |
+
"loss": 0.0547,
|
2501 |
+
"step": 3560
|
2502 |
+
},
|
2503 |
+
{
|
2504 |
+
"epoch": 7.811816192560175,
|
2505 |
+
"grad_norm": 0.39957642555236816,
|
2506 |
+
"learning_rate": 1.2532040275499301e-05,
|
2507 |
+
"loss": 0.0463,
|
2508 |
+
"step": 3570
|
2509 |
+
},
|
2510 |
+
{
|
2511 |
+
"epoch": 7.833698030634573,
|
2512 |
+
"grad_norm": 0.4493909776210785,
|
2513 |
+
"learning_rate": 1.2293418767199449e-05,
|
2514 |
+
"loss": 0.0522,
|
2515 |
+
"step": 3580
|
2516 |
+
},
|
2517 |
+
{
|
2518 |
+
"epoch": 7.855579868708972,
|
2519 |
+
"grad_norm": 0.40386393666267395,
|
2520 |
+
"learning_rate": 1.2056772114909365e-05,
|
2521 |
+
"loss": 0.0483,
|
2522 |
+
"step": 3590
|
2523 |
+
},
|
2524 |
+
{
|
2525 |
+
"epoch": 7.87746170678337,
|
2526 |
+
"grad_norm": 0.4808233380317688,
|
2527 |
+
"learning_rate": 1.182211271283315e-05,
|
2528 |
+
"loss": 0.0485,
|
2529 |
+
"step": 3600
|
2530 |
+
},
|
2531 |
+
{
|
2532 |
+
"epoch": 7.899343544857768,
|
2533 |
+
"grad_norm": 0.35203149914741516,
|
2534 |
+
"learning_rate": 1.1589452851094062e-05,
|
2535 |
+
"loss": 0.0489,
|
2536 |
+
"step": 3610
|
2537 |
+
},
|
2538 |
+
{
|
2539 |
+
"epoch": 7.921225382932167,
|
2540 |
+
"grad_norm": 0.37248197197914124,
|
2541 |
+
"learning_rate": 1.135880471509076e-05,
|
2542 |
+
"loss": 0.055,
|
2543 |
+
"step": 3620
|
2544 |
+
},
|
2545 |
+
{
|
2546 |
+
"epoch": 7.9431072210065645,
|
2547 |
+
"grad_norm": 0.36168205738067627,
|
2548 |
+
"learning_rate": 1.1130180384859256e-05,
|
2549 |
+
"loss": 0.0445,
|
2550 |
+
"step": 3630
|
2551 |
+
},
|
2552 |
+
{
|
2553 |
+
"epoch": 7.964989059080962,
|
2554 |
+
"grad_norm": 0.6294784545898438,
|
2555 |
+
"learning_rate": 1.0903591834440097e-05,
|
2556 |
+
"loss": 0.0494,
|
2557 |
+
"step": 3640
|
2558 |
+
},
|
2559 |
+
{
|
2560 |
+
"epoch": 7.986870897155361,
|
2561 |
+
"grad_norm": 0.46873342990875244,
|
2562 |
+
"learning_rate": 1.0679050931251238e-05,
|
2563 |
+
"loss": 0.0508,
|
2564 |
+
"step": 3650
|
2565 |
+
},
|
2566 |
+
{
|
2567 |
+
"epoch": 8.00875273522976,
|
2568 |
+
"grad_norm": 0.4308367967605591,
|
2569 |
+
"learning_rate": 1.0456569435466617e-05,
|
2570 |
+
"loss": 0.0524,
|
2571 |
+
"step": 3660
|
2572 |
+
},
|
2573 |
+
{
|
2574 |
+
"epoch": 8.030634573304157,
|
2575 |
+
"grad_norm": 0.45153898000717163,
|
2576 |
+
"learning_rate": 1.0236158999400053e-05,
|
2577 |
+
"loss": 0.0475,
|
2578 |
+
"step": 3670
|
2579 |
+
},
|
2580 |
+
{
|
2581 |
+
"epoch": 8.052516411378555,
|
2582 |
+
"grad_norm": 0.6072208881378174,
|
2583 |
+
"learning_rate": 1.0017831166895104e-05,
|
2584 |
+
"loss": 0.0478,
|
2585 |
+
"step": 3680
|
2586 |
+
},
|
2587 |
+
{
|
2588 |
+
"epoch": 8.074398249452955,
|
2589 |
+
"grad_norm": 0.34941720962524414,
|
2590 |
+
"learning_rate": 9.801597372720395e-06,
|
2591 |
+
"loss": 0.0466,
|
2592 |
+
"step": 3690
|
2593 |
+
},
|
2594 |
+
{
|
2595 |
+
"epoch": 8.096280087527353,
|
2596 |
+
"grad_norm": 0.32349899411201477,
|
2597 |
+
"learning_rate": 9.587468941970751e-06,
|
2598 |
+
"loss": 0.0434,
|
2599 |
+
"step": 3700
|
2600 |
+
},
|
2601 |
+
{
|
2602 |
+
"epoch": 8.11816192560175,
|
2603 |
+
"grad_norm": 0.4987752437591553,
|
2604 |
+
"learning_rate": 9.375457089474043e-06,
|
2605 |
+
"loss": 0.0457,
|
2606 |
+
"step": 3710
|
2607 |
+
},
|
2608 |
+
{
|
2609 |
+
"epoch": 8.140043763676148,
|
2610 |
+
"grad_norm": 0.3384380638599396,
|
2611 |
+
"learning_rate": 9.165572919203847e-06,
|
2612 |
+
"loss": 0.0539,
|
2613 |
+
"step": 3720
|
2614 |
+
},
|
2615 |
+
{
|
2616 |
+
"epoch": 8.161925601750546,
|
2617 |
+
"grad_norm": 0.4982927739620209,
|
2618 |
+
"learning_rate": 8.957827423697823e-06,
|
2619 |
+
"loss": 0.0482,
|
2620 |
+
"step": 3730
|
2621 |
+
},
|
2622 |
+
{
|
2623 |
+
"epoch": 8.183807439824946,
|
2624 |
+
"grad_norm": 0.26903849840164185,
|
2625 |
+
"learning_rate": 8.75223148348207e-06,
|
2626 |
+
"loss": 0.0529,
|
2627 |
+
"step": 3740
|
2628 |
+
},
|
2629 |
+
{
|
2630 |
+
"epoch": 8.205689277899344,
|
2631 |
+
"grad_norm": 0.49081680178642273,
|
2632 |
+
"learning_rate": 8.548795866501164e-06,
|
2633 |
+
"loss": 0.0444,
|
2634 |
+
"step": 3750
|
2635 |
+
},
|
2636 |
+
{
|
2637 |
+
"epoch": 8.227571115973742,
|
2638 |
+
"grad_norm": 0.37901756167411804,
|
2639 |
+
"learning_rate": 8.347531227554323e-06,
|
2640 |
+
"loss": 0.0457,
|
2641 |
+
"step": 3760
|
2642 |
+
},
|
2643 |
+
{
|
2644 |
+
"epoch": 8.24945295404814,
|
2645 |
+
"grad_norm": 0.40646421909332275,
|
2646 |
+
"learning_rate": 8.148448107737227e-06,
|
2647 |
+
"loss": 0.0447,
|
2648 |
+
"step": 3770
|
2649 |
+
},
|
2650 |
+
{
|
2651 |
+
"epoch": 8.271334792122538,
|
2652 |
+
"grad_norm": 0.6366869211196899,
|
2653 |
+
"learning_rate": 7.951556933890047e-06,
|
2654 |
+
"loss": 0.0555,
|
2655 |
+
"step": 3780
|
2656 |
+
},
|
2657 |
+
{
|
2658 |
+
"epoch": 8.293216630196937,
|
2659 |
+
"grad_norm": 0.508129358291626,
|
2660 |
+
"learning_rate": 7.756868018051322e-06,
|
2661 |
+
"loss": 0.0461,
|
2662 |
+
"step": 3790
|
2663 |
+
},
|
2664 |
+
{
|
2665 |
+
"epoch": 8.315098468271335,
|
2666 |
+
"grad_norm": 0.317501425743103,
|
2667 |
+
"learning_rate": 7.56439155691781e-06,
|
2668 |
+
"loss": 0.0456,
|
2669 |
+
"step": 3800
|
2670 |
+
},
|
2671 |
+
{
|
2672 |
+
"epoch": 8.336980306345733,
|
2673 |
+
"grad_norm": 0.34288403391838074,
|
2674 |
+
"learning_rate": 7.374137631310512e-06,
|
2675 |
+
"loss": 0.0498,
|
2676 |
+
"step": 3810
|
2677 |
+
},
|
2678 |
+
{
|
2679 |
+
"epoch": 8.358862144420131,
|
2680 |
+
"grad_norm": 0.44967618584632874,
|
2681 |
+
"learning_rate": 7.186116205646687e-06,
|
2682 |
+
"loss": 0.0475,
|
2683 |
+
"step": 3820
|
2684 |
+
},
|
2685 |
+
{
|
2686 |
+
"epoch": 8.38074398249453,
|
2687 |
+
"grad_norm": 0.6020100712776184,
|
2688 |
+
"learning_rate": 7.000337127417938e-06,
|
2689 |
+
"loss": 0.0489,
|
2690 |
+
"step": 3830
|
2691 |
+
},
|
2692 |
+
{
|
2693 |
+
"epoch": 8.402625820568927,
|
2694 |
+
"grad_norm": 0.37260666489601135,
|
2695 |
+
"learning_rate": 6.816810126674494e-06,
|
2696 |
+
"loss": 0.0483,
|
2697 |
+
"step": 3840
|
2698 |
+
},
|
2699 |
+
{
|
2700 |
+
"epoch": 8.424507658643327,
|
2701 |
+
"grad_norm": 0.4538766145706177,
|
2702 |
+
"learning_rate": 6.635544815515576e-06,
|
2703 |
+
"loss": 0.0448,
|
2704 |
+
"step": 3850
|
2705 |
+
},
|
2706 |
+
{
|
2707 |
+
"epoch": 8.446389496717725,
|
2708 |
+
"grad_norm": 0.4207398593425751,
|
2709 |
+
"learning_rate": 6.456550687586016e-06,
|
2710 |
+
"loss": 0.0484,
|
2711 |
+
"step": 3860
|
2712 |
+
},
|
2713 |
+
{
|
2714 |
+
"epoch": 8.468271334792123,
|
2715 |
+
"grad_norm": 0.3660431504249573,
|
2716 |
+
"learning_rate": 6.2798371175789405e-06,
|
2717 |
+
"loss": 0.0557,
|
2718 |
+
"step": 3870
|
2719 |
+
},
|
2720 |
+
{
|
2721 |
+
"epoch": 8.49015317286652,
|
2722 |
+
"grad_norm": 0.3074619472026825,
|
2723 |
+
"learning_rate": 6.105413360744883e-06,
|
2724 |
+
"loss": 0.0506,
|
2725 |
+
"step": 3880
|
2726 |
+
},
|
2727 |
+
{
|
2728 |
+
"epoch": 8.512035010940918,
|
2729 |
+
"grad_norm": 0.41881874203681946,
|
2730 |
+
"learning_rate": 5.933288552407018e-06,
|
2731 |
+
"loss": 0.0468,
|
2732 |
+
"step": 3890
|
2733 |
+
},
|
2734 |
+
{
|
2735 |
+
"epoch": 8.533916849015318,
|
2736 |
+
"grad_norm": 0.6764019131660461,
|
2737 |
+
"learning_rate": 5.763471707482638e-06,
|
2738 |
+
"loss": 0.0459,
|
2739 |
+
"step": 3900
|
2740 |
+
},
|
2741 |
+
{
|
2742 |
+
"epoch": 8.555798687089716,
|
2743 |
+
"grad_norm": 0.5430788993835449,
|
2744 |
+
"learning_rate": 5.5959717200110896e-06,
|
2745 |
+
"loss": 0.047,
|
2746 |
+
"step": 3910
|
2747 |
+
},
|
2748 |
+
{
|
2749 |
+
"epoch": 8.577680525164114,
|
2750 |
+
"grad_norm": 0.256409227848053,
|
2751 |
+
"learning_rate": 5.430797362687906e-06,
|
2752 |
+
"loss": 0.0458,
|
2753 |
+
"step": 3920
|
2754 |
+
},
|
2755 |
+
{
|
2756 |
+
"epoch": 8.599562363238512,
|
2757 |
+
"grad_norm": 0.3550792932510376,
|
2758 |
+
"learning_rate": 5.267957286405368e-06,
|
2759 |
+
"loss": 0.0461,
|
2760 |
+
"step": 3930
|
2761 |
+
},
|
2762 |
+
{
|
2763 |
+
"epoch": 8.62144420131291,
|
2764 |
+
"grad_norm": 0.27031660079956055,
|
2765 |
+
"learning_rate": 5.107460019799387e-06,
|
2766 |
+
"loss": 0.044,
|
2767 |
+
"step": 3940
|
2768 |
+
},
|
2769 |
+
{
|
2770 |
+
"epoch": 8.643326039387308,
|
2771 |
+
"grad_norm": 0.5709757208824158,
|
2772 |
+
"learning_rate": 4.949313968802871e-06,
|
2773 |
+
"loss": 0.0519,
|
2774 |
+
"step": 3950
|
2775 |
+
},
|
2776 |
+
{
|
2777 |
+
"epoch": 8.665207877461707,
|
2778 |
+
"grad_norm": 0.42753544449806213,
|
2779 |
+
"learning_rate": 4.793527416205429e-06,
|
2780 |
+
"loss": 0.0422,
|
2781 |
+
"step": 3960
|
2782 |
+
},
|
2783 |
+
{
|
2784 |
+
"epoch": 8.687089715536105,
|
2785 |
+
"grad_norm": 0.6058006286621094,
|
2786 |
+
"learning_rate": 4.640108521219561e-06,
|
2787 |
+
"loss": 0.0518,
|
2788 |
+
"step": 3970
|
2789 |
+
},
|
2790 |
+
{
|
2791 |
+
"epoch": 8.708971553610503,
|
2792 |
+
"grad_norm": 0.3088013827800751,
|
2793 |
+
"learning_rate": 4.489065319053376e-06,
|
2794 |
+
"loss": 0.0475,
|
2795 |
+
"step": 3980
|
2796 |
+
},
|
2797 |
+
{
|
2798 |
+
"epoch": 8.730853391684901,
|
2799 |
+
"grad_norm": 0.6196874976158142,
|
2800 |
+
"learning_rate": 4.340405720489721e-06,
|
2801 |
+
"loss": 0.0455,
|
2802 |
+
"step": 3990
|
2803 |
+
},
|
2804 |
+
{
|
2805 |
+
"epoch": 8.752735229759299,
|
2806 |
+
"grad_norm": 0.3027147054672241,
|
2807 |
+
"learning_rate": 4.194137511471824e-06,
|
2808 |
+
"loss": 0.0433,
|
2809 |
+
"step": 4000
|
2810 |
+
},
|
2811 |
+
{
|
2812 |
+
"epoch": 8.774617067833699,
|
2813 |
+
"grad_norm": 0.34994643926620483,
|
2814 |
+
"learning_rate": 4.0502683526955745e-06,
|
2815 |
+
"loss": 0.0438,
|
2816 |
+
"step": 4010
|
2817 |
+
},
|
2818 |
+
{
|
2819 |
+
"epoch": 8.796498905908097,
|
2820 |
+
"grad_norm": 0.3832307457923889,
|
2821 |
+
"learning_rate": 3.908805779208269e-06,
|
2822 |
+
"loss": 0.0415,
|
2823 |
+
"step": 4020
|
2824 |
+
},
|
2825 |
+
{
|
2826 |
+
"epoch": 8.818380743982495,
|
2827 |
+
"grad_norm": 0.2572057545185089,
|
2828 |
+
"learning_rate": 3.7697572000139626e-06,
|
2829 |
+
"loss": 0.0491,
|
2830 |
+
"step": 4030
|
2831 |
+
},
|
2832 |
+
{
|
2833 |
+
"epoch": 8.840262582056893,
|
2834 |
+
"grad_norm": 0.5285372138023376,
|
2835 |
+
"learning_rate": 3.633129897685439e-06,
|
2836 |
+
"loss": 0.0498,
|
2837 |
+
"step": 4040
|
2838 |
+
},
|
2839 |
+
{
|
2840 |
+
"epoch": 8.86214442013129,
|
2841 |
+
"grad_norm": 0.44389864802360535,
|
2842 |
+
"learning_rate": 3.4989310279827938e-06,
|
2843 |
+
"loss": 0.0495,
|
2844 |
+
"step": 4050
|
2845 |
+
},
|
2846 |
+
{
|
2847 |
+
"epoch": 8.88402625820569,
|
2848 |
+
"grad_norm": 0.3135242760181427,
|
2849 |
+
"learning_rate": 3.3671676194786504e-06,
|
2850 |
+
"loss": 0.0454,
|
2851 |
+
"step": 4060
|
2852 |
+
},
|
2853 |
+
{
|
2854 |
+
"epoch": 8.905908096280088,
|
2855 |
+
"grad_norm": 0.29390549659729004,
|
2856 |
+
"learning_rate": 3.2378465731900064e-06,
|
2857 |
+
"loss": 0.0438,
|
2858 |
+
"step": 4070
|
2859 |
+
},
|
2860 |
+
{
|
2861 |
+
"epoch": 8.927789934354486,
|
2862 |
+
"grad_norm": 0.4052577018737793,
|
2863 |
+
"learning_rate": 3.1109746622168922e-06,
|
2864 |
+
"loss": 0.0454,
|
2865 |
+
"step": 4080
|
2866 |
+
},
|
2867 |
+
{
|
2868 |
+
"epoch": 8.949671772428884,
|
2869 |
+
"grad_norm": 0.7273414134979248,
|
2870 |
+
"learning_rate": 2.9865585313875565e-06,
|
2871 |
+
"loss": 0.0463,
|
2872 |
+
"step": 4090
|
2873 |
+
},
|
2874 |
+
{
|
2875 |
+
"epoch": 8.971553610503282,
|
2876 |
+
"grad_norm": 0.3757275640964508,
|
2877 |
+
"learning_rate": 2.8646046969104335e-06,
|
2878 |
+
"loss": 0.0457,
|
2879 |
+
"step": 4100
|
2880 |
+
},
|
2881 |
+
{
|
2882 |
+
"epoch": 8.99343544857768,
|
2883 |
+
"grad_norm": 0.4105775058269501,
|
2884 |
+
"learning_rate": 2.7451195460329315e-06,
|
2885 |
+
"loss": 0.0497,
|
2886 |
+
"step": 4110
|
2887 |
+
},
|
2888 |
+
{
|
2889 |
+
"epoch": 9.01531728665208,
|
2890 |
+
"grad_norm": 0.3794170916080475,
|
2891 |
+
"learning_rate": 2.6281093367068743e-06,
|
2892 |
+
"loss": 0.0423,
|
2893 |
+
"step": 4120
|
2894 |
+
},
|
2895 |
+
{
|
2896 |
+
"epoch": 9.037199124726477,
|
2897 |
+
"grad_norm": 0.28719809651374817,
|
2898 |
+
"learning_rate": 2.513580197260712e-06,
|
2899 |
+
"loss": 0.0441,
|
2900 |
+
"step": 4130
|
2901 |
+
},
|
2902 |
+
{
|
2903 |
+
"epoch": 9.059080962800875,
|
2904 |
+
"grad_norm": 0.38277560472488403,
|
2905 |
+
"learning_rate": 2.401538126078606e-06,
|
2906 |
+
"loss": 0.0423,
|
2907 |
+
"step": 4140
|
2908 |
+
},
|
2909 |
+
{
|
2910 |
+
"epoch": 9.080962800875273,
|
2911 |
+
"grad_norm": 0.43671509623527527,
|
2912 |
+
"learning_rate": 2.2919889912862313e-06,
|
2913 |
+
"loss": 0.0436,
|
2914 |
+
"step": 4150
|
2915 |
+
},
|
2916 |
+
{
|
2917 |
+
"epoch": 9.102844638949671,
|
2918 |
+
"grad_norm": 0.3907598853111267,
|
2919 |
+
"learning_rate": 2.1849385304434644e-06,
|
2920 |
+
"loss": 0.0485,
|
2921 |
+
"step": 4160
|
2922 |
+
},
|
2923 |
+
{
|
2924 |
+
"epoch": 9.12472647702407,
|
2925 |
+
"grad_norm": 0.38813626766204834,
|
2926 |
+
"learning_rate": 2.080392350243837e-06,
|
2927 |
+
"loss": 0.0443,
|
2928 |
+
"step": 4170
|
2929 |
+
},
|
2930 |
+
{
|
2931 |
+
"epoch": 9.146608315098469,
|
2932 |
+
"grad_norm": 0.4304605722427368,
|
2933 |
+
"learning_rate": 1.978355926220965e-06,
|
2934 |
+
"loss": 0.0478,
|
2935 |
+
"step": 4180
|
2936 |
+
},
|
2937 |
+
{
|
2938 |
+
"epoch": 9.168490153172867,
|
2939 |
+
"grad_norm": 0.3653641939163208,
|
2940 |
+
"learning_rate": 1.8788346024617044e-06,
|
2941 |
+
"loss": 0.0481,
|
2942 |
+
"step": 4190
|
2943 |
+
},
|
2944 |
+
{
|
2945 |
+
"epoch": 9.190371991247265,
|
2946 |
+
"grad_norm": 0.49059316515922546,
|
2947 |
+
"learning_rate": 1.7818335913262708e-06,
|
2948 |
+
"loss": 0.0482,
|
2949 |
+
"step": 4200
|
2950 |
+
},
|
2951 |
+
{
|
2952 |
+
"epoch": 9.212253829321662,
|
2953 |
+
"grad_norm": 0.6487129330635071,
|
2954 |
+
"learning_rate": 1.6873579731752798e-06,
|
2955 |
+
"loss": 0.0439,
|
2956 |
+
"step": 4210
|
2957 |
+
},
|
2958 |
+
{
|
2959 |
+
"epoch": 9.23413566739606,
|
2960 |
+
"grad_norm": 0.3429771959781647,
|
2961 |
+
"learning_rate": 1.5954126961036208e-06,
|
2962 |
+
"loss": 0.0452,
|
2963 |
+
"step": 4220
|
2964 |
+
},
|
2965 |
+
{
|
2966 |
+
"epoch": 9.25601750547046,
|
2967 |
+
"grad_norm": 0.23548723757266998,
|
2968 |
+
"learning_rate": 1.5060025756813423e-06,
|
2969 |
+
"loss": 0.0419,
|
2970 |
+
"step": 4230
|
2971 |
+
},
|
2972 |
+
{
|
2973 |
+
"epoch": 9.277899343544858,
|
2974 |
+
"grad_norm": 0.3810988962650299,
|
2975 |
+
"learning_rate": 1.4191322947014197e-06,
|
2976 |
+
"loss": 0.0467,
|
2977 |
+
"step": 4240
|
2978 |
+
},
|
2979 |
+
{
|
2980 |
+
"epoch": 9.299781181619256,
|
2981 |
+
"grad_norm": 0.4621908664703369,
|
2982 |
+
"learning_rate": 1.3348064029344908e-06,
|
2983 |
+
"loss": 0.0521,
|
2984 |
+
"step": 4250
|
2985 |
+
},
|
2986 |
+
{
|
2987 |
+
"epoch": 9.321663019693654,
|
2988 |
+
"grad_norm": 0.6988676190376282,
|
2989 |
+
"learning_rate": 1.2530293168905916e-06,
|
2990 |
+
"loss": 0.0502,
|
2991 |
+
"step": 4260
|
2992 |
+
},
|
2993 |
+
{
|
2994 |
+
"epoch": 9.343544857768052,
|
2995 |
+
"grad_norm": 0.2640683352947235,
|
2996 |
+
"learning_rate": 1.1738053195878174e-06,
|
2997 |
+
"loss": 0.0494,
|
2998 |
+
"step": 4270
|
2999 |
+
},
|
3000 |
+
{
|
3001 |
+
"epoch": 9.365426695842451,
|
3002 |
+
"grad_norm": 0.3481920063495636,
|
3003 |
+
"learning_rate": 1.097138560328015e-06,
|
3004 |
+
"loss": 0.0477,
|
3005 |
+
"step": 4280
|
3006 |
+
},
|
3007 |
+
{
|
3008 |
+
"epoch": 9.38730853391685,
|
3009 |
+
"grad_norm": 0.42929619550704956,
|
3010 |
+
"learning_rate": 1.023033054479472e-06,
|
3011 |
+
"loss": 0.0487,
|
3012 |
+
"step": 4290
|
3013 |
+
},
|
3014 |
+
{
|
3015 |
+
"epoch": 9.409190371991247,
|
3016 |
+
"grad_norm": 0.5334005951881409,
|
3017 |
+
"learning_rate": 9.51492683266586e-07,
|
3018 |
+
"loss": 0.051,
|
3019 |
+
"step": 4300
|
3020 |
+
},
|
3021 |
+
{
|
3022 |
+
"epoch": 9.431072210065645,
|
3023 |
+
"grad_norm": 0.3841243088245392,
|
3024 |
+
"learning_rate": 8.825211935666433e-07,
|
3025 |
+
"loss": 0.0478,
|
3026 |
+
"step": 4310
|
3027 |
+
},
|
3028 |
+
{
|
3029 |
+
"epoch": 9.452954048140043,
|
3030 |
+
"grad_norm": 0.34734562039375305,
|
3031 |
+
"learning_rate": 8.161221977135092e-07,
|
3032 |
+
"loss": 0.0438,
|
3033 |
+
"step": 4320
|
3034 |
+
},
|
3035 |
+
{
|
3036 |
+
"epoch": 9.474835886214443,
|
3037 |
+
"grad_norm": 0.6244210004806519,
|
3038 |
+
"learning_rate": 7.522991733084905e-07,
|
3039 |
+
"loss": 0.0473,
|
3040 |
+
"step": 4330
|
3041 |
+
},
|
3042 |
+
{
|
3043 |
+
"epoch": 9.49671772428884,
|
3044 |
+
"grad_norm": 0.4562375247478485,
|
3045 |
+
"learning_rate": 6.910554630381815e-07,
|
3046 |
+
"loss": 0.0486,
|
3047 |
+
"step": 4340
|
3048 |
+
},
|
3049 |
+
{
|
3050 |
+
"epoch": 9.518599562363239,
|
3051 |
+
"grad_norm": 0.8920162916183472,
|
3052 |
+
"learning_rate": 6.323942744993761e-07,
|
3053 |
+
"loss": 0.0479,
|
3054 |
+
"step": 4350
|
3055 |
+
},
|
3056 |
+
{
|
3057 |
+
"epoch": 9.540481400437637,
|
3058 |
+
"grad_norm": 0.46890518069267273,
|
3059 |
+
"learning_rate": 5.763186800310849e-07,
|
3060 |
+
"loss": 0.0473,
|
3061 |
+
"step": 4360
|
3062 |
+
},
|
3063 |
+
{
|
3064 |
+
"epoch": 9.562363238512035,
|
3065 |
+
"grad_norm": 0.37694522738456726,
|
3066 |
+
"learning_rate": 5.228316165536429e-07,
|
3067 |
+
"loss": 0.0464,
|
3068 |
+
"step": 4370
|
3069 |
+
},
|
3070 |
+
{
|
3071 |
+
"epoch": 9.584245076586432,
|
3072 |
+
"grad_norm": 0.42408043146133423,
|
3073 |
+
"learning_rate": 4.7193588541484813e-07,
|
3074 |
+
"loss": 0.0494,
|
3075 |
+
"step": 4380
|
3076 |
+
},
|
3077 |
+
{
|
3078 |
+
"epoch": 9.606126914660832,
|
3079 |
+
"grad_norm": 0.3436243534088135,
|
3080 |
+
"learning_rate": 4.2363415224329075e-07,
|
3081 |
+
"loss": 0.0446,
|
3082 |
+
"step": 4390
|
3083 |
+
},
|
3084 |
+
{
|
3085 |
+
"epoch": 9.62800875273523,
|
3086 |
+
"grad_norm": 0.32050296664237976,
|
3087 |
+
"learning_rate": 3.779289468086922e-07,
|
3088 |
+
"loss": 0.0506,
|
3089 |
+
"step": 4400
|
3090 |
+
},
|
3091 |
+
{
|
3092 |
+
"epoch": 9.649890590809628,
|
3093 |
+
"grad_norm": 0.5982491970062256,
|
3094 |
+
"learning_rate": 3.3482266288946686e-07,
|
3095 |
+
"loss": 0.0458,
|
3096 |
+
"step": 4410
|
3097 |
+
},
|
3098 |
+
{
|
3099 |
+
"epoch": 9.671772428884026,
|
3100 |
+
"grad_norm": 0.5433162450790405,
|
3101 |
+
"learning_rate": 2.9431755814729457e-07,
|
3102 |
+
"loss": 0.0465,
|
3103 |
+
"step": 4420
|
3104 |
+
},
|
3105 |
+
{
|
3106 |
+
"epoch": 9.693654266958424,
|
3107 |
+
"grad_norm": 0.4410666823387146,
|
3108 |
+
"learning_rate": 2.5641575400890403e-07,
|
3109 |
+
"loss": 0.0452,
|
3110 |
+
"step": 4430
|
3111 |
+
},
|
3112 |
+
{
|
3113 |
+
"epoch": 9.715536105032823,
|
3114 |
+
"grad_norm": 0.2922728955745697,
|
3115 |
+
"learning_rate": 2.2111923555498405e-07,
|
3116 |
+
"loss": 0.0464,
|
3117 |
+
"step": 4440
|
3118 |
+
},
|
3119 |
+
{
|
3120 |
+
"epoch": 9.737417943107221,
|
3121 |
+
"grad_norm": 0.4933525323867798,
|
3122 |
+
"learning_rate": 1.8842985141617197e-07,
|
3123 |
+
"loss": 0.048,
|
3124 |
+
"step": 4450
|
3125 |
+
},
|
3126 |
+
{
|
3127 |
+
"epoch": 9.75929978118162,
|
3128 |
+
"grad_norm": 0.7095863223075867,
|
3129 |
+
"learning_rate": 1.5834931367625928e-07,
|
3130 |
+
"loss": 0.0462,
|
3131 |
+
"step": 4460
|
3132 |
+
},
|
3133 |
+
{
|
3134 |
+
"epoch": 9.781181619256017,
|
3135 |
+
"grad_norm": 0.5306093096733093,
|
3136 |
+
"learning_rate": 1.3087919778252966e-07,
|
3137 |
+
"loss": 0.0496,
|
3138 |
+
"step": 4470
|
3139 |
+
},
|
3140 |
+
{
|
3141 |
+
"epoch": 9.803063457330415,
|
3142 |
+
"grad_norm": 0.31279975175857544,
|
3143 |
+
"learning_rate": 1.060209424632308e-07,
|
3144 |
+
"loss": 0.0431,
|
3145 |
+
"step": 4480
|
3146 |
+
},
|
3147 |
+
{
|
3148 |
+
"epoch": 9.824945295404813,
|
3149 |
+
"grad_norm": 0.5052768588066101,
|
3150 |
+
"learning_rate": 8.377584965221785e-08,
|
3151 |
+
"loss": 0.0493,
|
3152 |
+
"step": 4490
|
3153 |
+
},
|
3154 |
+
{
|
3155 |
+
"epoch": 9.846827133479213,
|
3156 |
+
"grad_norm": 0.4014386534690857,
|
3157 |
+
"learning_rate": 6.414508442078026e-08,
|
3158 |
+
"loss": 0.0478,
|
3159 |
+
"step": 4500
|
3160 |
+
},
|
3161 |
+
{
|
3162 |
+
"epoch": 9.86870897155361,
|
3163 |
+
"grad_norm": 0.2997543215751648,
|
3164 |
+
"learning_rate": 4.7129674916618346e-08,
|
3165 |
+
"loss": 0.0451,
|
3166 |
+
"step": 4510
|
3167 |
+
},
|
3168 |
+
{
|
3169 |
+
"epoch": 9.890590809628009,
|
3170 |
+
"grad_norm": 0.5213238000869751,
|
3171 |
+
"learning_rate": 3.273051230999191e-08,
|
3172 |
+
"loss": 0.0436,
|
3173 |
+
"step": 4520
|
3174 |
+
},
|
3175 |
+
{
|
3176 |
+
"epoch": 9.912472647702407,
|
3177 |
+
"grad_norm": 0.5938975811004639,
|
3178 |
+
"learning_rate": 2.0948350747046487e-08,
|
3179 |
+
"loss": 0.0417,
|
3180 |
+
"step": 4530
|
3181 |
+
},
|
3182 |
+
{
|
3183 |
+
"epoch": 9.934354485776804,
|
3184 |
+
"grad_norm": 0.26146259903907776,
|
3185 |
+
"learning_rate": 1.1783807310300488e-08,
|
3186 |
+
"loss": 0.0456,
|
3187 |
+
"step": 4540
|
3188 |
+
},
|
3189 |
+
{
|
3190 |
+
"epoch": 9.956236323851204,
|
3191 |
+
"grad_norm": 0.238664910197258,
|
3192 |
+
"learning_rate": 5.237361986365486e-09,
|
3193 |
+
"loss": 0.0456,
|
3194 |
+
"step": 4550
|
3195 |
+
},
|
3196 |
+
{
|
3197 |
+
"epoch": 9.978118161925602,
|
3198 |
+
"grad_norm": 0.3414470851421356,
|
3199 |
+
"learning_rate": 1.30935764076634e-09,
|
3200 |
+
"loss": 0.0471,
|
3201 |
+
"step": 4560
|
3202 |
+
},
|
3203 |
+
{
|
3204 |
+
"epoch": 10.0,
|
3205 |
+
"grad_norm": 0.492483526468277,
|
3206 |
+
"learning_rate": 0.0,
|
3207 |
+
"loss": 0.0444,
|
3208 |
+
"step": 4570
|
3209 |
+
},
|
3210 |
+
{
|
3211 |
+
"epoch": 10.0,
|
3212 |
+
"step": 4570,
|
3213 |
+
"total_flos": 4.682127244666608e+17,
|
3214 |
+
"train_loss": 0.07182003624199255,
|
3215 |
+
"train_runtime": 4395.4484,
|
3216 |
+
"train_samples_per_second": 50.928,
|
3217 |
+
"train_steps_per_second": 1.04
|
3218 |
+
}
|
3219 |
+
],
|
3220 |
+
"logging_steps": 10,
|
3221 |
+
"max_steps": 4570,
|
3222 |
+
"num_input_tokens_seen": 0,
|
3223 |
+
"num_train_epochs": 10,
|
3224 |
+
"save_steps": 10000,
|
3225 |
+
"stateful_callbacks": {
|
3226 |
+
"TrainerControl": {
|
3227 |
+
"args": {
|
3228 |
+
"should_epoch_stop": false,
|
3229 |
+
"should_evaluate": false,
|
3230 |
+
"should_log": false,
|
3231 |
+
"should_save": true,
|
3232 |
+
"should_training_stop": true
|
3233 |
+
},
|
3234 |
+
"attributes": {}
|
3235 |
+
}
|
3236 |
+
},
|
3237 |
+
"total_flos": 4.682127244666608e+17,
|
3238 |
+
"train_batch_size": 49,
|
3239 |
+
"trial_name": null,
|
3240 |
+
"trial_params": null
|
3241 |
+
}
|