Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +3102 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,3102 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 4.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 4376,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.009140767824497258,
|
| 14 |
+
"grad_norm": 20.285207748413086,
|
| 15 |
+
"learning_rate": 4.10958904109589e-06,
|
| 16 |
+
"loss": 1.4218,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.018281535648994516,
|
| 21 |
+
"grad_norm": 3.7156178951263428,
|
| 22 |
+
"learning_rate": 8.675799086757991e-06,
|
| 23 |
+
"loss": 0.6623,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.027422303473491772,
|
| 28 |
+
"grad_norm": 3.670661687850952,
|
| 29 |
+
"learning_rate": 1.3242009132420092e-05,
|
| 30 |
+
"loss": 0.3818,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.03656307129798903,
|
| 35 |
+
"grad_norm": 1.6582924127578735,
|
| 36 |
+
"learning_rate": 1.780821917808219e-05,
|
| 37 |
+
"loss": 0.2857,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.04570383912248629,
|
| 42 |
+
"grad_norm": 3.376695394515991,
|
| 43 |
+
"learning_rate": 2.237442922374429e-05,
|
| 44 |
+
"loss": 0.2503,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.054844606946983544,
|
| 49 |
+
"grad_norm": 2.420069456100464,
|
| 50 |
+
"learning_rate": 2.6940639269406392e-05,
|
| 51 |
+
"loss": 0.211,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.06398537477148081,
|
| 56 |
+
"grad_norm": 2.5535237789154053,
|
| 57 |
+
"learning_rate": 3.1506849315068496e-05,
|
| 58 |
+
"loss": 0.1969,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.07312614259597806,
|
| 63 |
+
"grad_norm": 1.6121954917907715,
|
| 64 |
+
"learning_rate": 3.60730593607306e-05,
|
| 65 |
+
"loss": 0.1838,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.08226691042047532,
|
| 70 |
+
"grad_norm": 2.0982680320739746,
|
| 71 |
+
"learning_rate": 4.063926940639269e-05,
|
| 72 |
+
"loss": 0.1935,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.09140767824497258,
|
| 77 |
+
"grad_norm": 2.1739630699157715,
|
| 78 |
+
"learning_rate": 4.520547945205479e-05,
|
| 79 |
+
"loss": 0.1721,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.10054844606946983,
|
| 84 |
+
"grad_norm": 2.0316410064697266,
|
| 85 |
+
"learning_rate": 4.977168949771689e-05,
|
| 86 |
+
"loss": 0.1454,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.10968921389396709,
|
| 91 |
+
"grad_norm": 2.113408327102661,
|
| 92 |
+
"learning_rate": 5.4337899543379e-05,
|
| 93 |
+
"loss": 0.1445,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.11882998171846434,
|
| 98 |
+
"grad_norm": 1.6204296350479126,
|
| 99 |
+
"learning_rate": 5.89041095890411e-05,
|
| 100 |
+
"loss": 0.1178,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.12797074954296161,
|
| 105 |
+
"grad_norm": 1.2460007667541504,
|
| 106 |
+
"learning_rate": 6.34703196347032e-05,
|
| 107 |
+
"loss": 0.1113,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.13711151736745886,
|
| 112 |
+
"grad_norm": 2.4157071113586426,
|
| 113 |
+
"learning_rate": 6.803652968036531e-05,
|
| 114 |
+
"loss": 0.1038,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.14625228519195613,
|
| 119 |
+
"grad_norm": 1.6945436000823975,
|
| 120 |
+
"learning_rate": 7.26027397260274e-05,
|
| 121 |
+
"loss": 0.1148,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.15539305301645337,
|
| 126 |
+
"grad_norm": 1.4777846336364746,
|
| 127 |
+
"learning_rate": 7.716894977168951e-05,
|
| 128 |
+
"loss": 0.1147,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.16453382084095064,
|
| 133 |
+
"grad_norm": 1.7918751239776611,
|
| 134 |
+
"learning_rate": 8.17351598173516e-05,
|
| 135 |
+
"loss": 0.1149,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.1736745886654479,
|
| 140 |
+
"grad_norm": 1.584354043006897,
|
| 141 |
+
"learning_rate": 8.630136986301371e-05,
|
| 142 |
+
"loss": 0.113,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.18281535648994515,
|
| 147 |
+
"grad_norm": 1.4646424055099487,
|
| 148 |
+
"learning_rate": 9.08675799086758e-05,
|
| 149 |
+
"loss": 0.0993,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.19195612431444242,
|
| 154 |
+
"grad_norm": 1.1241822242736816,
|
| 155 |
+
"learning_rate": 9.543378995433791e-05,
|
| 156 |
+
"loss": 0.1105,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.20109689213893966,
|
| 161 |
+
"grad_norm": 1.0951234102249146,
|
| 162 |
+
"learning_rate": 0.0001,
|
| 163 |
+
"loss": 0.0891,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.21023765996343693,
|
| 168 |
+
"grad_norm": 0.955017626285553,
|
| 169 |
+
"learning_rate": 9.999857216626715e-05,
|
| 170 |
+
"loss": 0.0948,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.21937842778793418,
|
| 175 |
+
"grad_norm": 1.671467900276184,
|
| 176 |
+
"learning_rate": 9.999428874661695e-05,
|
| 177 |
+
"loss": 0.089,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.22851919561243145,
|
| 182 |
+
"grad_norm": 1.0147258043289185,
|
| 183 |
+
"learning_rate": 9.998714998568987e-05,
|
| 184 |
+
"loss": 0.0923,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.2376599634369287,
|
| 189 |
+
"grad_norm": 1.6253561973571777,
|
| 190 |
+
"learning_rate": 9.997715629120446e-05,
|
| 191 |
+
"loss": 0.081,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.24680073126142596,
|
| 196 |
+
"grad_norm": 1.460111141204834,
|
| 197 |
+
"learning_rate": 9.996430823393403e-05,
|
| 198 |
+
"loss": 0.086,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.25594149908592323,
|
| 203 |
+
"grad_norm": 1.5847985744476318,
|
| 204 |
+
"learning_rate": 9.994860654767421e-05,
|
| 205 |
+
"loss": 0.0882,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.26508226691042047,
|
| 210 |
+
"grad_norm": 0.8328115344047546,
|
| 211 |
+
"learning_rate": 9.99300521292009e-05,
|
| 212 |
+
"loss": 0.0884,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.2742230347349177,
|
| 217 |
+
"grad_norm": 0.7161694765090942,
|
| 218 |
+
"learning_rate": 9.990864603821906e-05,
|
| 219 |
+
"loss": 0.0874,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.283363802559415,
|
| 224 |
+
"grad_norm": 1.3507859706878662,
|
| 225 |
+
"learning_rate": 9.988438949730223e-05,
|
| 226 |
+
"loss": 0.0902,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.29250457038391225,
|
| 231 |
+
"grad_norm": 1.1614631414413452,
|
| 232 |
+
"learning_rate": 9.985728389182275e-05,
|
| 233 |
+
"loss": 0.0773,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.3016453382084095,
|
| 238 |
+
"grad_norm": 1.271041989326477,
|
| 239 |
+
"learning_rate": 9.982733076987248e-05,
|
| 240 |
+
"loss": 0.0974,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.31078610603290674,
|
| 245 |
+
"grad_norm": 1.0431146621704102,
|
| 246 |
+
"learning_rate": 9.979453184217456e-05,
|
| 247 |
+
"loss": 0.0872,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.31992687385740404,
|
| 252 |
+
"grad_norm": 2.105883836746216,
|
| 253 |
+
"learning_rate": 9.975888898198562e-05,
|
| 254 |
+
"loss": 0.0936,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.3290676416819013,
|
| 259 |
+
"grad_norm": 1.3111586570739746,
|
| 260 |
+
"learning_rate": 9.972040422498877e-05,
|
| 261 |
+
"loss": 0.102,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.3382084095063985,
|
| 266 |
+
"grad_norm": 1.21346914768219,
|
| 267 |
+
"learning_rate": 9.967907976917738e-05,
|
| 268 |
+
"loss": 0.0754,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.3473491773308958,
|
| 273 |
+
"grad_norm": 0.7348060607910156,
|
| 274 |
+
"learning_rate": 9.963491797472953e-05,
|
| 275 |
+
"loss": 0.0703,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.35648994515539306,
|
| 280 |
+
"grad_norm": 0.7553760409355164,
|
| 281 |
+
"learning_rate": 9.958792136387322e-05,
|
| 282 |
+
"loss": 0.083,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.3656307129798903,
|
| 287 |
+
"grad_norm": 1.1796748638153076,
|
| 288 |
+
"learning_rate": 9.95380926207423e-05,
|
| 289 |
+
"loss": 0.0642,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.37477148080438755,
|
| 294 |
+
"grad_norm": 0.9751859903335571,
|
| 295 |
+
"learning_rate": 9.948543459122317e-05,
|
| 296 |
+
"loss": 0.0735,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.38391224862888484,
|
| 301 |
+
"grad_norm": 1.2366355657577515,
|
| 302 |
+
"learning_rate": 9.942995028279227e-05,
|
| 303 |
+
"loss": 0.0748,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.3930530164533821,
|
| 308 |
+
"grad_norm": 1.4610956907272339,
|
| 309 |
+
"learning_rate": 9.93716428643443e-05,
|
| 310 |
+
"loss": 0.0973,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.40219378427787933,
|
| 315 |
+
"grad_norm": 1.440199375152588,
|
| 316 |
+
"learning_rate": 9.931051566601119e-05,
|
| 317 |
+
"loss": 0.0765,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.4113345521023766,
|
| 322 |
+
"grad_norm": 1.6553508043289185,
|
| 323 |
+
"learning_rate": 9.924657217897201e-05,
|
| 324 |
+
"loss": 0.0922,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.42047531992687387,
|
| 329 |
+
"grad_norm": 1.534443736076355,
|
| 330 |
+
"learning_rate": 9.917981605525345e-05,
|
| 331 |
+
"loss": 0.1115,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.4296160877513711,
|
| 336 |
+
"grad_norm": 1.0473862886428833,
|
| 337 |
+
"learning_rate": 9.911025110752132e-05,
|
| 338 |
+
"loss": 0.0704,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.43875685557586835,
|
| 343 |
+
"grad_norm": 1.1698426008224487,
|
| 344 |
+
"learning_rate": 9.90378813088628e-05,
|
| 345 |
+
"loss": 0.0788,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.44789762340036565,
|
| 350 |
+
"grad_norm": 1.1977952718734741,
|
| 351 |
+
"learning_rate": 9.896271079255943e-05,
|
| 352 |
+
"loss": 0.0967,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.4570383912248629,
|
| 357 |
+
"grad_norm": 1.3484407663345337,
|
| 358 |
+
"learning_rate": 9.888474385185122e-05,
|
| 359 |
+
"loss": 0.0762,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.46617915904936014,
|
| 364 |
+
"grad_norm": 0.7931845784187317,
|
| 365 |
+
"learning_rate": 9.880398493969128e-05,
|
| 366 |
+
"loss": 0.0629,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.4753199268738574,
|
| 371 |
+
"grad_norm": 0.7555088400840759,
|
| 372 |
+
"learning_rate": 9.872043866849156e-05,
|
| 373 |
+
"loss": 0.072,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.4844606946983547,
|
| 378 |
+
"grad_norm": 0.9508174061775208,
|
| 379 |
+
"learning_rate": 9.863410980985943e-05,
|
| 380 |
+
"loss": 0.0688,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.4936014625228519,
|
| 385 |
+
"grad_norm": 0.9366075992584229,
|
| 386 |
+
"learning_rate": 9.854500329432514e-05,
|
| 387 |
+
"loss": 0.0635,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.5027422303473492,
|
| 392 |
+
"grad_norm": 0.7485318779945374,
|
| 393 |
+
"learning_rate": 9.845312421106026e-05,
|
| 394 |
+
"loss": 0.0612,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.5118829981718465,
|
| 399 |
+
"grad_norm": 1.1440995931625366,
|
| 400 |
+
"learning_rate": 9.835847780758695e-05,
|
| 401 |
+
"loss": 0.0657,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.5210237659963437,
|
| 406 |
+
"grad_norm": 0.6450702548027039,
|
| 407 |
+
"learning_rate": 9.826106948947831e-05,
|
| 408 |
+
"loss": 0.0756,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.5301645338208409,
|
| 413 |
+
"grad_norm": 0.9916141033172607,
|
| 414 |
+
"learning_rate": 9.816090482004966e-05,
|
| 415 |
+
"loss": 0.0689,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.5393053016453382,
|
| 420 |
+
"grad_norm": 0.9260968565940857,
|
| 421 |
+
"learning_rate": 9.805798952004074e-05,
|
| 422 |
+
"loss": 0.0741,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.5484460694698354,
|
| 427 |
+
"grad_norm": 1.099038004875183,
|
| 428 |
+
"learning_rate": 9.795232946728902e-05,
|
| 429 |
+
"loss": 0.0664,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.5575868372943327,
|
| 434 |
+
"grad_norm": 0.8096833825111389,
|
| 435 |
+
"learning_rate": 9.784393069639402e-05,
|
| 436 |
+
"loss": 0.0726,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.56672760511883,
|
| 441 |
+
"grad_norm": 1.4272693395614624,
|
| 442 |
+
"learning_rate": 9.77327993983726e-05,
|
| 443 |
+
"loss": 0.0733,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.5758683729433273,
|
| 448 |
+
"grad_norm": 1.050479769706726,
|
| 449 |
+
"learning_rate": 9.76189419203054e-05,
|
| 450 |
+
"loss": 0.067,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.5850091407678245,
|
| 455 |
+
"grad_norm": 0.8118894100189209,
|
| 456 |
+
"learning_rate": 9.750236476497435e-05,
|
| 457 |
+
"loss": 0.0709,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.5941499085923218,
|
| 462 |
+
"grad_norm": 0.5546473264694214,
|
| 463 |
+
"learning_rate": 9.738307459049121e-05,
|
| 464 |
+
"loss": 0.0536,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.603290676416819,
|
| 469 |
+
"grad_norm": 0.9993944764137268,
|
| 470 |
+
"learning_rate": 9.726107820991741e-05,
|
| 471 |
+
"loss": 0.0812,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.6124314442413162,
|
| 476 |
+
"grad_norm": 1.0287402868270874,
|
| 477 |
+
"learning_rate": 9.713638259087486e-05,
|
| 478 |
+
"loss": 0.0695,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.6215722120658135,
|
| 483 |
+
"grad_norm": 0.9676907062530518,
|
| 484 |
+
"learning_rate": 9.7008994855148e-05,
|
| 485 |
+
"loss": 0.0649,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.6307129798903108,
|
| 490 |
+
"grad_norm": 0.886712908744812,
|
| 491 |
+
"learning_rate": 9.687892227827707e-05,
|
| 492 |
+
"loss": 0.0786,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.6398537477148081,
|
| 497 |
+
"grad_norm": 0.5730246305465698,
|
| 498 |
+
"learning_rate": 9.674617228914259e-05,
|
| 499 |
+
"loss": 0.058,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.6489945155393053,
|
| 504 |
+
"grad_norm": 1.1880249977111816,
|
| 505 |
+
"learning_rate": 9.661075246954106e-05,
|
| 506 |
+
"loss": 0.0701,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.6581352833638026,
|
| 511 |
+
"grad_norm": 0.8326805233955383,
|
| 512 |
+
"learning_rate": 9.647267055375195e-05,
|
| 513 |
+
"loss": 0.0683,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.6672760511882998,
|
| 518 |
+
"grad_norm": 0.7834450006484985,
|
| 519 |
+
"learning_rate": 9.633193442809596e-05,
|
| 520 |
+
"loss": 0.0625,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.676416819012797,
|
| 525 |
+
"grad_norm": 0.8641069531440735,
|
| 526 |
+
"learning_rate": 9.618855213048457e-05,
|
| 527 |
+
"loss": 0.0915,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.6855575868372943,
|
| 532 |
+
"grad_norm": 1.1982171535491943,
|
| 533 |
+
"learning_rate": 9.604253184996103e-05,
|
| 534 |
+
"loss": 0.0744,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.6946983546617916,
|
| 539 |
+
"grad_norm": 0.9463611841201782,
|
| 540 |
+
"learning_rate": 9.589388192623266e-05,
|
| 541 |
+
"loss": 0.0578,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.7038391224862889,
|
| 546 |
+
"grad_norm": 0.7067205905914307,
|
| 547 |
+
"learning_rate": 9.574261084919445e-05,
|
| 548 |
+
"loss": 0.0774,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.7129798903107861,
|
| 553 |
+
"grad_norm": 0.9819389581680298,
|
| 554 |
+
"learning_rate": 9.558872725844426e-05,
|
| 555 |
+
"loss": 0.0636,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.7221206581352834,
|
| 560 |
+
"grad_norm": 0.7785099148750305,
|
| 561 |
+
"learning_rate": 9.543223994278938e-05,
|
| 562 |
+
"loss": 0.0649,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.7312614259597806,
|
| 567 |
+
"grad_norm": 0.6871359944343567,
|
| 568 |
+
"learning_rate": 9.527315783974454e-05,
|
| 569 |
+
"loss": 0.0711,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.7404021937842779,
|
| 574 |
+
"grad_norm": 0.9093775153160095,
|
| 575 |
+
"learning_rate": 9.511149003502145e-05,
|
| 576 |
+
"loss": 0.0866,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.7495429616087751,
|
| 581 |
+
"grad_norm": 0.8019959926605225,
|
| 582 |
+
"learning_rate": 9.494724576200989e-05,
|
| 583 |
+
"loss": 0.0688,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.7586837294332724,
|
| 588 |
+
"grad_norm": 0.6367955803871155,
|
| 589 |
+
"learning_rate": 9.478043440125043e-05,
|
| 590 |
+
"loss": 0.0499,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.7678244972577697,
|
| 595 |
+
"grad_norm": 1.0420273542404175,
|
| 596 |
+
"learning_rate": 9.461106547989858e-05,
|
| 597 |
+
"loss": 0.0696,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.7769652650822669,
|
| 602 |
+
"grad_norm": 1.0322197675704956,
|
| 603 |
+
"learning_rate": 9.44391486711807e-05,
|
| 604 |
+
"loss": 0.078,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.7861060329067642,
|
| 609 |
+
"grad_norm": 0.6654338836669922,
|
| 610 |
+
"learning_rate": 9.426469379384153e-05,
|
| 611 |
+
"loss": 0.0608,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.7952468007312614,
|
| 616 |
+
"grad_norm": 0.6521769762039185,
|
| 617 |
+
"learning_rate": 9.408771081158346e-05,
|
| 618 |
+
"loss": 0.0542,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.8043875685557587,
|
| 623 |
+
"grad_norm": 0.588214099407196,
|
| 624 |
+
"learning_rate": 9.390820983249733e-05,
|
| 625 |
+
"loss": 0.0558,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.8135283363802559,
|
| 630 |
+
"grad_norm": 1.0544887781143188,
|
| 631 |
+
"learning_rate": 9.372620110848528e-05,
|
| 632 |
+
"loss": 0.0558,
|
| 633 |
+
"step": 890
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.8226691042047533,
|
| 637 |
+
"grad_norm": 0.4711853265762329,
|
| 638 |
+
"learning_rate": 9.354169503467515e-05,
|
| 639 |
+
"loss": 0.0617,
|
| 640 |
+
"step": 900
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.8318098720292505,
|
| 644 |
+
"grad_norm": 0.8516566157341003,
|
| 645 |
+
"learning_rate": 9.335470214882678e-05,
|
| 646 |
+
"loss": 0.0577,
|
| 647 |
+
"step": 910
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.8409506398537477,
|
| 651 |
+
"grad_norm": 0.9411584734916687,
|
| 652 |
+
"learning_rate": 9.316523313073018e-05,
|
| 653 |
+
"loss": 0.0578,
|
| 654 |
+
"step": 920
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.850091407678245,
|
| 658 |
+
"grad_norm": 0.6325905919075012,
|
| 659 |
+
"learning_rate": 9.297329880159556e-05,
|
| 660 |
+
"loss": 0.0549,
|
| 661 |
+
"step": 930
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.8592321755027422,
|
| 665 |
+
"grad_norm": 0.9557574391365051,
|
| 666 |
+
"learning_rate": 9.277891012343529e-05,
|
| 667 |
+
"loss": 0.0542,
|
| 668 |
+
"step": 940
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.8683729433272395,
|
| 672 |
+
"grad_norm": 0.9808653593063354,
|
| 673 |
+
"learning_rate": 9.258207819843787e-05,
|
| 674 |
+
"loss": 0.0665,
|
| 675 |
+
"step": 950
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.8775137111517367,
|
| 679 |
+
"grad_norm": 0.924566388130188,
|
| 680 |
+
"learning_rate": 9.23828142683338e-05,
|
| 681 |
+
"loss": 0.0786,
|
| 682 |
+
"step": 960
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.886654478976234,
|
| 686 |
+
"grad_norm": 1.108108401298523,
|
| 687 |
+
"learning_rate": 9.21811297137535e-05,
|
| 688 |
+
"loss": 0.0766,
|
| 689 |
+
"step": 970
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.8957952468007313,
|
| 693 |
+
"grad_norm": 1.2968391180038452,
|
| 694 |
+
"learning_rate": 9.197703605357739e-05,
|
| 695 |
+
"loss": 0.0605,
|
| 696 |
+
"step": 980
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.9049360146252285,
|
| 700 |
+
"grad_norm": 0.9167346954345703,
|
| 701 |
+
"learning_rate": 9.177054494427799e-05,
|
| 702 |
+
"loss": 0.0504,
|
| 703 |
+
"step": 990
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.9140767824497258,
|
| 707 |
+
"grad_norm": 0.8108971118927002,
|
| 708 |
+
"learning_rate": 9.156166817925415e-05,
|
| 709 |
+
"loss": 0.0526,
|
| 710 |
+
"step": 1000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.923217550274223,
|
| 714 |
+
"grad_norm": 0.8036976456642151,
|
| 715 |
+
"learning_rate": 9.135041768815751e-05,
|
| 716 |
+
"loss": 0.0469,
|
| 717 |
+
"step": 1010
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.9323583180987203,
|
| 721 |
+
"grad_norm": 0.9675208926200867,
|
| 722 |
+
"learning_rate": 9.113680553621116e-05,
|
| 723 |
+
"loss": 0.054,
|
| 724 |
+
"step": 1020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.9414990859232175,
|
| 728 |
+
"grad_norm": 0.7497916221618652,
|
| 729 |
+
"learning_rate": 9.092084392352056e-05,
|
| 730 |
+
"loss": 0.0649,
|
| 731 |
+
"step": 1030
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.9506398537477148,
|
| 735 |
+
"grad_norm": 0.5699302554130554,
|
| 736 |
+
"learning_rate": 9.070254518437673e-05,
|
| 737 |
+
"loss": 0.0505,
|
| 738 |
+
"step": 1040
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.9597806215722121,
|
| 742 |
+
"grad_norm": 0.539658784866333,
|
| 743 |
+
"learning_rate": 9.048192178655181e-05,
|
| 744 |
+
"loss": 0.0435,
|
| 745 |
+
"step": 1050
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.9689213893967094,
|
| 749 |
+
"grad_norm": 0.6254842877388,
|
| 750 |
+
"learning_rate": 9.025898633058699e-05,
|
| 751 |
+
"loss": 0.0541,
|
| 752 |
+
"step": 1060
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.9780621572212066,
|
| 756 |
+
"grad_norm": 0.6805806756019592,
|
| 757 |
+
"learning_rate": 9.003375154907284e-05,
|
| 758 |
+
"loss": 0.0714,
|
| 759 |
+
"step": 1070
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.9872029250457038,
|
| 763 |
+
"grad_norm": 0.6484934687614441,
|
| 764 |
+
"learning_rate": 8.980623030592212e-05,
|
| 765 |
+
"loss": 0.065,
|
| 766 |
+
"step": 1080
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.9963436928702011,
|
| 770 |
+
"grad_norm": 0.5721938014030457,
|
| 771 |
+
"learning_rate": 8.957643559563507e-05,
|
| 772 |
+
"loss": 0.0735,
|
| 773 |
+
"step": 1090
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.0054844606946984,
|
| 777 |
+
"grad_norm": 1.0545032024383545,
|
| 778 |
+
"learning_rate": 8.934438054255723e-05,
|
| 779 |
+
"loss": 0.0585,
|
| 780 |
+
"step": 1100
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.0146252285191957,
|
| 784 |
+
"grad_norm": 0.5121464729309082,
|
| 785 |
+
"learning_rate": 8.91100784001299e-05,
|
| 786 |
+
"loss": 0.0545,
|
| 787 |
+
"step": 1110
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 1.023765996343693,
|
| 791 |
+
"grad_norm": 0.890046238899231,
|
| 792 |
+
"learning_rate": 8.887354255013323e-05,
|
| 793 |
+
"loss": 0.063,
|
| 794 |
+
"step": 1120
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.0329067641681902,
|
| 798 |
+
"grad_norm": 0.4641508162021637,
|
| 799 |
+
"learning_rate": 8.863478650192181e-05,
|
| 800 |
+
"loss": 0.0622,
|
| 801 |
+
"step": 1130
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 1.0420475319926874,
|
| 805 |
+
"grad_norm": 0.7307084202766418,
|
| 806 |
+
"learning_rate": 8.839382389165323e-05,
|
| 807 |
+
"loss": 0.0586,
|
| 808 |
+
"step": 1140
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 1.0511882998171846,
|
| 812 |
+
"grad_norm": 0.7155199646949768,
|
| 813 |
+
"learning_rate": 8.815066848150922e-05,
|
| 814 |
+
"loss": 0.052,
|
| 815 |
+
"step": 1150
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 1.0603290676416819,
|
| 819 |
+
"grad_norm": 0.675208568572998,
|
| 820 |
+
"learning_rate": 8.790533415890967e-05,
|
| 821 |
+
"loss": 0.057,
|
| 822 |
+
"step": 1160
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 1.0694698354661791,
|
| 826 |
+
"grad_norm": 0.7103818655014038,
|
| 827 |
+
"learning_rate": 8.765783493571942e-05,
|
| 828 |
+
"loss": 0.0517,
|
| 829 |
+
"step": 1170
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 1.0786106032906764,
|
| 833 |
+
"grad_norm": 0.8737835884094238,
|
| 834 |
+
"learning_rate": 8.740818494744809e-05,
|
| 835 |
+
"loss": 0.0376,
|
| 836 |
+
"step": 1180
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.0877513711151736,
|
| 840 |
+
"grad_norm": 1.203967809677124,
|
| 841 |
+
"learning_rate": 8.715639845244263e-05,
|
| 842 |
+
"loss": 0.0538,
|
| 843 |
+
"step": 1190
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 1.0968921389396709,
|
| 847 |
+
"grad_norm": 0.6871981620788574,
|
| 848 |
+
"learning_rate": 8.690248983107313e-05,
|
| 849 |
+
"loss": 0.0586,
|
| 850 |
+
"step": 1200
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 1.106032906764168,
|
| 854 |
+
"grad_norm": 1.3813148736953735,
|
| 855 |
+
"learning_rate": 8.664647358491134e-05,
|
| 856 |
+
"loss": 0.0486,
|
| 857 |
+
"step": 1210
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 1.1151736745886653,
|
| 861 |
+
"grad_norm": 0.9246250987052917,
|
| 862 |
+
"learning_rate": 8.638836433590257e-05,
|
| 863 |
+
"loss": 0.0463,
|
| 864 |
+
"step": 1220
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 1.1243144424131628,
|
| 868 |
+
"grad_norm": 1.0106674432754517,
|
| 869 |
+
"learning_rate": 8.61281768255305e-05,
|
| 870 |
+
"loss": 0.0466,
|
| 871 |
+
"step": 1230
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 1.13345521023766,
|
| 875 |
+
"grad_norm": 0.7251119017601013,
|
| 876 |
+
"learning_rate": 8.586592591397533e-05,
|
| 877 |
+
"loss": 0.0575,
|
| 878 |
+
"step": 1240
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.1425959780621573,
|
| 882 |
+
"grad_norm": 0.8430503606796265,
|
| 883 |
+
"learning_rate": 8.560162657926494e-05,
|
| 884 |
+
"loss": 0.0427,
|
| 885 |
+
"step": 1250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 1.1517367458866545,
|
| 889 |
+
"grad_norm": 0.739719808101654,
|
| 890 |
+
"learning_rate": 8.533529391641959e-05,
|
| 891 |
+
"loss": 0.0425,
|
| 892 |
+
"step": 1260
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 1.1608775137111518,
|
| 896 |
+
"grad_norm": 0.8659994006156921,
|
| 897 |
+
"learning_rate": 8.506694313658969e-05,
|
| 898 |
+
"loss": 0.048,
|
| 899 |
+
"step": 1270
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 1.170018281535649,
|
| 903 |
+
"grad_norm": 0.54444819688797,
|
| 904 |
+
"learning_rate": 8.479658956618704e-05,
|
| 905 |
+
"loss": 0.0572,
|
| 906 |
+
"step": 1280
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 1.1791590493601463,
|
| 910 |
+
"grad_norm": 1.301878809928894,
|
| 911 |
+
"learning_rate": 8.452424864600956e-05,
|
| 912 |
+
"loss": 0.0611,
|
| 913 |
+
"step": 1290
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 1.1882998171846435,
|
| 917 |
+
"grad_norm": 0.42106685042381287,
|
| 918 |
+
"learning_rate": 8.424993593035936e-05,
|
| 919 |
+
"loss": 0.0446,
|
| 920 |
+
"step": 1300
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 1.1974405850091407,
|
| 924 |
+
"grad_norm": 0.6645190715789795,
|
| 925 |
+
"learning_rate": 8.397366708615439e-05,
|
| 926 |
+
"loss": 0.0618,
|
| 927 |
+
"step": 1310
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 1.206581352833638,
|
| 931 |
+
"grad_norm": 0.6044383645057678,
|
| 932 |
+
"learning_rate": 8.369545789203363e-05,
|
| 933 |
+
"loss": 0.0408,
|
| 934 |
+
"step": 1320
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 1.2157221206581352,
|
| 938 |
+
"grad_norm": 0.5589725375175476,
|
| 939 |
+
"learning_rate": 8.341532423745599e-05,
|
| 940 |
+
"loss": 0.059,
|
| 941 |
+
"step": 1330
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 1.2248628884826325,
|
| 945 |
+
"grad_norm": 0.9428887963294983,
|
| 946 |
+
"learning_rate": 8.313328212179274e-05,
|
| 947 |
+
"loss": 0.0573,
|
| 948 |
+
"step": 1340
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 1.2340036563071297,
|
| 952 |
+
"grad_norm": 0.6069498062133789,
|
| 953 |
+
"learning_rate": 8.284934765341374e-05,
|
| 954 |
+
"loss": 0.0539,
|
| 955 |
+
"step": 1350
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 1.2431444241316272,
|
| 959 |
+
"grad_norm": 0.8920032978057861,
|
| 960 |
+
"learning_rate": 8.256353704876748e-05,
|
| 961 |
+
"loss": 0.0699,
|
| 962 |
+
"step": 1360
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 1.2522851919561244,
|
| 966 |
+
"grad_norm": 1.1463648080825806,
|
| 967 |
+
"learning_rate": 8.227586663145485e-05,
|
| 968 |
+
"loss": 0.0442,
|
| 969 |
+
"step": 1370
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 1.2614259597806217,
|
| 973 |
+
"grad_norm": 0.5741071701049805,
|
| 974 |
+
"learning_rate": 8.198635283129688e-05,
|
| 975 |
+
"loss": 0.053,
|
| 976 |
+
"step": 1380
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 1.270566727605119,
|
| 980 |
+
"grad_norm": 0.6575504541397095,
|
| 981 |
+
"learning_rate": 8.169501218339638e-05,
|
| 982 |
+
"loss": 0.05,
|
| 983 |
+
"step": 1390
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 1.2797074954296161,
|
| 987 |
+
"grad_norm": 0.9514917135238647,
|
| 988 |
+
"learning_rate": 8.140186132719352e-05,
|
| 989 |
+
"loss": 0.0578,
|
| 990 |
+
"step": 1400
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 1.2888482632541134,
|
| 994 |
+
"grad_norm": 0.4858059585094452,
|
| 995 |
+
"learning_rate": 8.110691700551558e-05,
|
| 996 |
+
"loss": 0.0478,
|
| 997 |
+
"step": 1410
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 1.2979890310786106,
|
| 1001 |
+
"grad_norm": 0.6736142635345459,
|
| 1002 |
+
"learning_rate": 8.081019606362063e-05,
|
| 1003 |
+
"loss": 0.051,
|
| 1004 |
+
"step": 1420
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 1.3071297989031079,
|
| 1008 |
+
"grad_norm": 0.5692024827003479,
|
| 1009 |
+
"learning_rate": 8.051171544823543e-05,
|
| 1010 |
+
"loss": 0.0457,
|
| 1011 |
+
"step": 1430
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 1.3162705667276051,
|
| 1015 |
+
"grad_norm": 0.7224279642105103,
|
| 1016 |
+
"learning_rate": 8.02114922065877e-05,
|
| 1017 |
+
"loss": 0.0537,
|
| 1018 |
+
"step": 1440
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 1.3254113345521024,
|
| 1022 |
+
"grad_norm": 0.7441707849502563,
|
| 1023 |
+
"learning_rate": 7.990954348543227e-05,
|
| 1024 |
+
"loss": 0.0441,
|
| 1025 |
+
"step": 1450
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 1.3345521023765996,
|
| 1029 |
+
"grad_norm": 0.8292146325111389,
|
| 1030 |
+
"learning_rate": 7.960588653007191e-05,
|
| 1031 |
+
"loss": 0.0463,
|
| 1032 |
+
"step": 1460
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 1.3436928702010968,
|
| 1036 |
+
"grad_norm": 0.5339452028274536,
|
| 1037 |
+
"learning_rate": 7.930053868337242e-05,
|
| 1038 |
+
"loss": 0.0536,
|
| 1039 |
+
"step": 1470
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 1.352833638025594,
|
| 1043 |
+
"grad_norm": 0.47000357508659363,
|
| 1044 |
+
"learning_rate": 7.899351738477201e-05,
|
| 1045 |
+
"loss": 0.0465,
|
| 1046 |
+
"step": 1480
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 1.3619744058500913,
|
| 1050 |
+
"grad_norm": 0.4280283451080322,
|
| 1051 |
+
"learning_rate": 7.868484016928537e-05,
|
| 1052 |
+
"loss": 0.0457,
|
| 1053 |
+
"step": 1490
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 1.3711151736745886,
|
| 1057 |
+
"grad_norm": 1.0179851055145264,
|
| 1058 |
+
"learning_rate": 7.837452466650213e-05,
|
| 1059 |
+
"loss": 0.0474,
|
| 1060 |
+
"step": 1500
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 1.3802559414990858,
|
| 1064 |
+
"grad_norm": 0.6735445261001587,
|
| 1065 |
+
"learning_rate": 7.806258859957998e-05,
|
| 1066 |
+
"loss": 0.055,
|
| 1067 |
+
"step": 1510
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 1.389396709323583,
|
| 1071 |
+
"grad_norm": 1.0807559490203857,
|
| 1072 |
+
"learning_rate": 7.774904978423247e-05,
|
| 1073 |
+
"loss": 0.0522,
|
| 1074 |
+
"step": 1520
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 1.3985374771480805,
|
| 1078 |
+
"grad_norm": 0.5602183938026428,
|
| 1079 |
+
"learning_rate": 7.743392612771152e-05,
|
| 1080 |
+
"loss": 0.0386,
|
| 1081 |
+
"step": 1530
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 1.4076782449725778,
|
| 1085 |
+
"grad_norm": 0.887178361415863,
|
| 1086 |
+
"learning_rate": 7.711723562778456e-05,
|
| 1087 |
+
"loss": 0.0476,
|
| 1088 |
+
"step": 1540
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 1.416819012797075,
|
| 1092 |
+
"grad_norm": 1.0711039304733276,
|
| 1093 |
+
"learning_rate": 7.679899637170677e-05,
|
| 1094 |
+
"loss": 0.0444,
|
| 1095 |
+
"step": 1550
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 1.4259597806215722,
|
| 1099 |
+
"grad_norm": 1.0682564973831177,
|
| 1100 |
+
"learning_rate": 7.647922653518796e-05,
|
| 1101 |
+
"loss": 0.0522,
|
| 1102 |
+
"step": 1560
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 1.4351005484460695,
|
| 1106 |
+
"grad_norm": 0.5387830138206482,
|
| 1107 |
+
"learning_rate": 7.615794438135444e-05,
|
| 1108 |
+
"loss": 0.0555,
|
| 1109 |
+
"step": 1570
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 1.4442413162705667,
|
| 1113 |
+
"grad_norm": 0.7840970754623413,
|
| 1114 |
+
"learning_rate": 7.583516825970616e-05,
|
| 1115 |
+
"loss": 0.0483,
|
| 1116 |
+
"step": 1580
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 1.453382084095064,
|
| 1120 |
+
"grad_norm": 0.9905489683151245,
|
| 1121 |
+
"learning_rate": 7.551091660506847e-05,
|
| 1122 |
+
"loss": 0.0471,
|
| 1123 |
+
"step": 1590
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 1.4625228519195612,
|
| 1127 |
+
"grad_norm": 0.5009207129478455,
|
| 1128 |
+
"learning_rate": 7.518520793653939e-05,
|
| 1129 |
+
"loss": 0.0463,
|
| 1130 |
+
"step": 1600
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 1.4716636197440585,
|
| 1134 |
+
"grad_norm": 0.5666080713272095,
|
| 1135 |
+
"learning_rate": 7.485806085643188e-05,
|
| 1136 |
+
"loss": 0.0438,
|
| 1137 |
+
"step": 1610
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 1.4808043875685557,
|
| 1141 |
+
"grad_norm": 0.8383708596229553,
|
| 1142 |
+
"learning_rate": 7.452949404921142e-05,
|
| 1143 |
+
"loss": 0.0504,
|
| 1144 |
+
"step": 1620
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 1.489945155393053,
|
| 1148 |
+
"grad_norm": 0.6216220855712891,
|
| 1149 |
+
"learning_rate": 7.419952628042881e-05,
|
| 1150 |
+
"loss": 0.0464,
|
| 1151 |
+
"step": 1630
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 1.4990859232175504,
|
| 1155 |
+
"grad_norm": 0.5239822864532471,
|
| 1156 |
+
"learning_rate": 7.386817639564851e-05,
|
| 1157 |
+
"loss": 0.0431,
|
| 1158 |
+
"step": 1640
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 1.5082266910420477,
|
| 1162 |
+
"grad_norm": 1.0513837337493896,
|
| 1163 |
+
"learning_rate": 7.353546331937224e-05,
|
| 1164 |
+
"loss": 0.0571,
|
| 1165 |
+
"step": 1650
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 1.517367458866545,
|
| 1169 |
+
"grad_norm": 0.7348514199256897,
|
| 1170 |
+
"learning_rate": 7.320140605395815e-05,
|
| 1171 |
+
"loss": 0.0526,
|
| 1172 |
+
"step": 1660
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 1.5265082266910421,
|
| 1176 |
+
"grad_norm": 0.7277615070343018,
|
| 1177 |
+
"learning_rate": 7.286602367853552e-05,
|
| 1178 |
+
"loss": 0.0379,
|
| 1179 |
+
"step": 1670
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 1.5356489945155394,
|
| 1183 |
+
"grad_norm": 0.719916045665741,
|
| 1184 |
+
"learning_rate": 7.252933534791511e-05,
|
| 1185 |
+
"loss": 0.0411,
|
| 1186 |
+
"step": 1680
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 1.5447897623400366,
|
| 1190 |
+
"grad_norm": 0.8909785747528076,
|
| 1191 |
+
"learning_rate": 7.219136029149515e-05,
|
| 1192 |
+
"loss": 0.0526,
|
| 1193 |
+
"step": 1690
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 1.5539305301645339,
|
| 1197 |
+
"grad_norm": 0.9339496493339539,
|
| 1198 |
+
"learning_rate": 7.185211781216314e-05,
|
| 1199 |
+
"loss": 0.0699,
|
| 1200 |
+
"step": 1700
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 1.563071297989031,
|
| 1204 |
+
"grad_norm": 0.7735287547111511,
|
| 1205 |
+
"learning_rate": 7.151162728519324e-05,
|
| 1206 |
+
"loss": 0.0571,
|
| 1207 |
+
"step": 1710
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 1.5722120658135283,
|
| 1211 |
+
"grad_norm": 0.8301142454147339,
|
| 1212 |
+
"learning_rate": 7.11699081571399e-05,
|
| 1213 |
+
"loss": 0.0484,
|
| 1214 |
+
"step": 1720
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 1.5813528336380256,
|
| 1218 |
+
"grad_norm": 0.7117460370063782,
|
| 1219 |
+
"learning_rate": 7.082697994472705e-05,
|
| 1220 |
+
"loss": 0.0505,
|
| 1221 |
+
"step": 1730
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 1.5904936014625228,
|
| 1225 |
+
"grad_norm": 0.39248979091644287,
|
| 1226 |
+
"learning_rate": 7.048286223373342e-05,
|
| 1227 |
+
"loss": 0.0329,
|
| 1228 |
+
"step": 1740
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 1.59963436928702,
|
| 1232 |
+
"grad_norm": 0.6386615037918091,
|
| 1233 |
+
"learning_rate": 7.013757467787412e-05,
|
| 1234 |
+
"loss": 0.0382,
|
| 1235 |
+
"step": 1750
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 1.6087751371115173,
|
| 1239 |
+
"grad_norm": 0.8316029906272888,
|
| 1240 |
+
"learning_rate": 6.979113699767789e-05,
|
| 1241 |
+
"loss": 0.0443,
|
| 1242 |
+
"step": 1760
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 1.6179159049360146,
|
| 1246 |
+
"grad_norm": 0.6195833086967468,
|
| 1247 |
+
"learning_rate": 6.9443568979361e-05,
|
| 1248 |
+
"loss": 0.0468,
|
| 1249 |
+
"step": 1770
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 1.6270566727605118,
|
| 1253 |
+
"grad_norm": 0.5122663974761963,
|
| 1254 |
+
"learning_rate": 6.909489047369706e-05,
|
| 1255 |
+
"loss": 0.0486,
|
| 1256 |
+
"step": 1780
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 1.636197440585009,
|
| 1260 |
+
"grad_norm": 0.7890117168426514,
|
| 1261 |
+
"learning_rate": 6.87451213948834e-05,
|
| 1262 |
+
"loss": 0.0368,
|
| 1263 |
+
"step": 1790
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 1.6453382084095063,
|
| 1267 |
+
"grad_norm": 1.125509262084961,
|
| 1268 |
+
"learning_rate": 6.839428171940357e-05,
|
| 1269 |
+
"loss": 0.0485,
|
| 1270 |
+
"step": 1800
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 1.6544789762340035,
|
| 1274 |
+
"grad_norm": 0.6382961869239807,
|
| 1275 |
+
"learning_rate": 6.804239148488651e-05,
|
| 1276 |
+
"loss": 0.0551,
|
| 1277 |
+
"step": 1810
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 1.6636197440585008,
|
| 1281 |
+
"grad_norm": 0.5489874482154846,
|
| 1282 |
+
"learning_rate": 6.768947078896212e-05,
|
| 1283 |
+
"loss": 0.0543,
|
| 1284 |
+
"step": 1820
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 1.672760511882998,
|
| 1288 |
+
"grad_norm": 0.5988185405731201,
|
| 1289 |
+
"learning_rate": 6.733553978811338e-05,
|
| 1290 |
+
"loss": 0.0408,
|
| 1291 |
+
"step": 1830
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 1.6819012797074955,
|
| 1295 |
+
"grad_norm": 0.5746557116508484,
|
| 1296 |
+
"learning_rate": 6.698061869652518e-05,
|
| 1297 |
+
"loss": 0.052,
|
| 1298 |
+
"step": 1840
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 1.6910420475319927,
|
| 1302 |
+
"grad_norm": 0.4615539014339447,
|
| 1303 |
+
"learning_rate": 6.662472778492979e-05,
|
| 1304 |
+
"loss": 0.0356,
|
| 1305 |
+
"step": 1850
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 1.70018281535649,
|
| 1309 |
+
"grad_norm": 0.5085066556930542,
|
| 1310 |
+
"learning_rate": 6.626788737944916e-05,
|
| 1311 |
+
"loss": 0.0477,
|
| 1312 |
+
"step": 1860
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 1.7093235831809872,
|
| 1316 |
+
"grad_norm": 0.5613322257995605,
|
| 1317 |
+
"learning_rate": 6.591011786043404e-05,
|
| 1318 |
+
"loss": 0.0394,
|
| 1319 |
+
"step": 1870
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 1.7184643510054844,
|
| 1323 |
+
"grad_norm": 0.37369537353515625,
|
| 1324 |
+
"learning_rate": 6.55514396612999e-05,
|
| 1325 |
+
"loss": 0.0475,
|
| 1326 |
+
"step": 1880
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 1.7276051188299817,
|
| 1330 |
+
"grad_norm": 0.6342534422874451,
|
| 1331 |
+
"learning_rate": 6.519187326736007e-05,
|
| 1332 |
+
"loss": 0.0451,
|
| 1333 |
+
"step": 1890
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 1.736745886654479,
|
| 1337 |
+
"grad_norm": 0.7583989500999451,
|
| 1338 |
+
"learning_rate": 6.48314392146556e-05,
|
| 1339 |
+
"loss": 0.0468,
|
| 1340 |
+
"step": 1900
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 1.7458866544789764,
|
| 1344 |
+
"grad_norm": 0.6060802340507507,
|
| 1345 |
+
"learning_rate": 6.447015808878243e-05,
|
| 1346 |
+
"loss": 0.0581,
|
| 1347 |
+
"step": 1910
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 1.7550274223034736,
|
| 1351 |
+
"grad_norm": 0.5234028100967407,
|
| 1352 |
+
"learning_rate": 6.41080505237157e-05,
|
| 1353 |
+
"loss": 0.0409,
|
| 1354 |
+
"step": 1920
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 1.7641681901279709,
|
| 1358 |
+
"grad_norm": 0.7409939169883728,
|
| 1359 |
+
"learning_rate": 6.374513720063126e-05,
|
| 1360 |
+
"loss": 0.0512,
|
| 1361 |
+
"step": 1930
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 1.7733089579524681,
|
| 1365 |
+
"grad_norm": 0.5225390195846558,
|
| 1366 |
+
"learning_rate": 6.338143884672451e-05,
|
| 1367 |
+
"loss": 0.0518,
|
| 1368 |
+
"step": 1940
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 1.7824497257769654,
|
| 1372 |
+
"grad_norm": 0.45462241768836975,
|
| 1373 |
+
"learning_rate": 6.301697623402658e-05,
|
| 1374 |
+
"loss": 0.0432,
|
| 1375 |
+
"step": 1950
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 1.7915904936014626,
|
| 1379 |
+
"grad_norm": 0.3406662344932556,
|
| 1380 |
+
"learning_rate": 6.265177017821799e-05,
|
| 1381 |
+
"loss": 0.0507,
|
| 1382 |
+
"step": 1960
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 1.8007312614259599,
|
| 1386 |
+
"grad_norm": 0.8838239908218384,
|
| 1387 |
+
"learning_rate": 6.228584153743975e-05,
|
| 1388 |
+
"loss": 0.0474,
|
| 1389 |
+
"step": 1970
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 1.809872029250457,
|
| 1393 |
+
"grad_norm": 0.6049211025238037,
|
| 1394 |
+
"learning_rate": 6.19192112111022e-05,
|
| 1395 |
+
"loss": 0.0435,
|
| 1396 |
+
"step": 1980
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 1.8190127970749543,
|
| 1400 |
+
"grad_norm": 1.0864945650100708,
|
| 1401 |
+
"learning_rate": 6.155190013869119e-05,
|
| 1402 |
+
"loss": 0.0492,
|
| 1403 |
+
"step": 1990
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 1.8281535648994516,
|
| 1407 |
+
"grad_norm": 0.734801173210144,
|
| 1408 |
+
"learning_rate": 6.118392929857231e-05,
|
| 1409 |
+
"loss": 0.0477,
|
| 1410 |
+
"step": 2000
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 1.8372943327239488,
|
| 1414 |
+
"grad_norm": 0.9690449237823486,
|
| 1415 |
+
"learning_rate": 6.08153197067927e-05,
|
| 1416 |
+
"loss": 0.0489,
|
| 1417 |
+
"step": 2010
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 1.846435100548446,
|
| 1421 |
+
"grad_norm": 0.6723517179489136,
|
| 1422 |
+
"learning_rate": 6.044609241588073e-05,
|
| 1423 |
+
"loss": 0.0383,
|
| 1424 |
+
"step": 2020
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 1.8555758683729433,
|
| 1428 |
+
"grad_norm": 1.0849159955978394,
|
| 1429 |
+
"learning_rate": 6.0076268513643644e-05,
|
| 1430 |
+
"loss": 0.0458,
|
| 1431 |
+
"step": 2030
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 1.8647166361974405,
|
| 1435 |
+
"grad_norm": 0.6273264288902283,
|
| 1436 |
+
"learning_rate": 5.9705869121963165e-05,
|
| 1437 |
+
"loss": 0.0374,
|
| 1438 |
+
"step": 2040
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 1.8738574040219378,
|
| 1442 |
+
"grad_norm": 0.9236069321632385,
|
| 1443 |
+
"learning_rate": 5.933491539558913e-05,
|
| 1444 |
+
"loss": 0.0464,
|
| 1445 |
+
"step": 2050
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 1.882998171846435,
|
| 1449 |
+
"grad_norm": 0.5545753240585327,
|
| 1450 |
+
"learning_rate": 5.8963428520931275e-05,
|
| 1451 |
+
"loss": 0.045,
|
| 1452 |
+
"step": 2060
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 1.8921389396709323,
|
| 1456 |
+
"grad_norm": 0.43973255157470703,
|
| 1457 |
+
"learning_rate": 5.8591429714849263e-05,
|
| 1458 |
+
"loss": 0.0474,
|
| 1459 |
+
"step": 2070
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 1.9012797074954295,
|
| 1463 |
+
"grad_norm": 0.4769677519798279,
|
| 1464 |
+
"learning_rate": 5.821894022344083e-05,
|
| 1465 |
+
"loss": 0.044,
|
| 1466 |
+
"step": 2080
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 1.9104204753199268,
|
| 1470 |
+
"grad_norm": 0.4825699031352997,
|
| 1471 |
+
"learning_rate": 5.7845981320828436e-05,
|
| 1472 |
+
"loss": 0.0379,
|
| 1473 |
+
"step": 2090
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 1.919561243144424,
|
| 1477 |
+
"grad_norm": 0.4520387649536133,
|
| 1478 |
+
"learning_rate": 5.747257430794415e-05,
|
| 1479 |
+
"loss": 0.0511,
|
| 1480 |
+
"step": 2100
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 1.9287020109689212,
|
| 1484 |
+
"grad_norm": 1.0386220216751099,
|
| 1485 |
+
"learning_rate": 5.709874051131312e-05,
|
| 1486 |
+
"loss": 0.0451,
|
| 1487 |
+
"step": 2110
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 1.9378427787934185,
|
| 1491 |
+
"grad_norm": 0.5633272528648376,
|
| 1492 |
+
"learning_rate": 5.672450128183561e-05,
|
| 1493 |
+
"loss": 0.047,
|
| 1494 |
+
"step": 2120
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 1.946983546617916,
|
| 1498 |
+
"grad_norm": 0.5927779674530029,
|
| 1499 |
+
"learning_rate": 5.63498779935674e-05,
|
| 1500 |
+
"loss": 0.0408,
|
| 1501 |
+
"step": 2130
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 1.9561243144424132,
|
| 1505 |
+
"grad_norm": 0.4667109549045563,
|
| 1506 |
+
"learning_rate": 5.597489204249925e-05,
|
| 1507 |
+
"loss": 0.041,
|
| 1508 |
+
"step": 2140
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 1.9652650822669104,
|
| 1512 |
+
"grad_norm": 0.7210867404937744,
|
| 1513 |
+
"learning_rate": 5.559956484533477e-05,
|
| 1514 |
+
"loss": 0.0407,
|
| 1515 |
+
"step": 2150
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 1.9744058500914077,
|
| 1519 |
+
"grad_norm": 0.6026551723480225,
|
| 1520 |
+
"learning_rate": 5.522391783826726e-05,
|
| 1521 |
+
"loss": 0.0452,
|
| 1522 |
+
"step": 2160
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 1.983546617915905,
|
| 1526 |
+
"grad_norm": 0.559966504573822,
|
| 1527 |
+
"learning_rate": 5.484797247575547e-05,
|
| 1528 |
+
"loss": 0.0394,
|
| 1529 |
+
"step": 2170
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 1.9926873857404022,
|
| 1533 |
+
"grad_norm": 0.6461185812950134,
|
| 1534 |
+
"learning_rate": 5.4471750229298215e-05,
|
| 1535 |
+
"loss": 0.0366,
|
| 1536 |
+
"step": 2180
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 2.0018281535648996,
|
| 1540 |
+
"grad_norm": 0.3840448558330536,
|
| 1541 |
+
"learning_rate": 5.409527258620806e-05,
|
| 1542 |
+
"loss": 0.0326,
|
| 1543 |
+
"step": 2190
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 2.010968921389397,
|
| 1547 |
+
"grad_norm": 0.9953444600105286,
|
| 1548 |
+
"learning_rate": 5.3718561048384166e-05,
|
| 1549 |
+
"loss": 0.0353,
|
| 1550 |
+
"step": 2200
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 2.020109689213894,
|
| 1554 |
+
"grad_norm": 0.6786929368972778,
|
| 1555 |
+
"learning_rate": 5.334163713108417e-05,
|
| 1556 |
+
"loss": 0.046,
|
| 1557 |
+
"step": 2210
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 2.0292504570383914,
|
| 1561 |
+
"grad_norm": 0.4201532304286957,
|
| 1562 |
+
"learning_rate": 5.29645223616954e-05,
|
| 1563 |
+
"loss": 0.035,
|
| 1564 |
+
"step": 2220
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 2.0383912248628886,
|
| 1568 |
+
"grad_norm": 0.6718032956123352,
|
| 1569 |
+
"learning_rate": 5.258723827850547e-05,
|
| 1570 |
+
"loss": 0.0365,
|
| 1571 |
+
"step": 2230
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 2.047531992687386,
|
| 1575 |
+
"grad_norm": 0.6162596344947815,
|
| 1576 |
+
"learning_rate": 5.2209806429471955e-05,
|
| 1577 |
+
"loss": 0.0455,
|
| 1578 |
+
"step": 2240
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 2.056672760511883,
|
| 1582 |
+
"grad_norm": 0.6207318305969238,
|
| 1583 |
+
"learning_rate": 5.183224837099192e-05,
|
| 1584 |
+
"loss": 0.035,
|
| 1585 |
+
"step": 2250
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 2.0658135283363803,
|
| 1589 |
+
"grad_norm": 0.6675350666046143,
|
| 1590 |
+
"learning_rate": 5.145458566667065e-05,
|
| 1591 |
+
"loss": 0.0339,
|
| 1592 |
+
"step": 2260
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 2.0749542961608776,
|
| 1596 |
+
"grad_norm": 0.6197575330734253,
|
| 1597 |
+
"learning_rate": 5.107683988609009e-05,
|
| 1598 |
+
"loss": 0.0401,
|
| 1599 |
+
"step": 2270
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 2.084095063985375,
|
| 1603 |
+
"grad_norm": 0.7717021107673645,
|
| 1604 |
+
"learning_rate": 5.069903260357696e-05,
|
| 1605 |
+
"loss": 0.0454,
|
| 1606 |
+
"step": 2280
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 2.093235831809872,
|
| 1610 |
+
"grad_norm": 0.44513338804244995,
|
| 1611 |
+
"learning_rate": 5.0321185396970564e-05,
|
| 1612 |
+
"loss": 0.0445,
|
| 1613 |
+
"step": 2290
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 2.1023765996343693,
|
| 1617 |
+
"grad_norm": 0.47282537817955017,
|
| 1618 |
+
"learning_rate": 4.994331984639039e-05,
|
| 1619 |
+
"loss": 0.044,
|
| 1620 |
+
"step": 2300
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 2.1115173674588665,
|
| 1624 |
+
"grad_norm": 0.506184995174408,
|
| 1625 |
+
"learning_rate": 4.956545753300363e-05,
|
| 1626 |
+
"loss": 0.0391,
|
| 1627 |
+
"step": 2310
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 2.1206581352833638,
|
| 1631 |
+
"grad_norm": 0.5166071653366089,
|
| 1632 |
+
"learning_rate": 4.9187620037792575e-05,
|
| 1633 |
+
"loss": 0.0402,
|
| 1634 |
+
"step": 2320
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 2.129798903107861,
|
| 1638 |
+
"grad_norm": 0.6371522545814514,
|
| 1639 |
+
"learning_rate": 4.880982894032208e-05,
|
| 1640 |
+
"loss": 0.0385,
|
| 1641 |
+
"step": 2330
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 2.1389396709323583,
|
| 1645 |
+
"grad_norm": 0.6055366396903992,
|
| 1646 |
+
"learning_rate": 4.8432105817507066e-05,
|
| 1647 |
+
"loss": 0.0336,
|
| 1648 |
+
"step": 2340
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 2.1480804387568555,
|
| 1652 |
+
"grad_norm": 0.28509601950645447,
|
| 1653 |
+
"learning_rate": 4.8054472242380174e-05,
|
| 1654 |
+
"loss": 0.0326,
|
| 1655 |
+
"step": 2350
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 2.1572212065813527,
|
| 1659 |
+
"grad_norm": 0.7012027502059937,
|
| 1660 |
+
"learning_rate": 4.767694978285972e-05,
|
| 1661 |
+
"loss": 0.0439,
|
| 1662 |
+
"step": 2360
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 2.16636197440585,
|
| 1666 |
+
"grad_norm": 0.3986015021800995,
|
| 1667 |
+
"learning_rate": 4.729956000051779e-05,
|
| 1668 |
+
"loss": 0.0343,
|
| 1669 |
+
"step": 2370
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 2.1755027422303472,
|
| 1673 |
+
"grad_norm": 0.5725045800209045,
|
| 1674 |
+
"learning_rate": 4.6922324449348826e-05,
|
| 1675 |
+
"loss": 0.0352,
|
| 1676 |
+
"step": 2380
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 2.1846435100548445,
|
| 1680 |
+
"grad_norm": 0.8768548965454102,
|
| 1681 |
+
"learning_rate": 4.6545264674538683e-05,
|
| 1682 |
+
"loss": 0.0367,
|
| 1683 |
+
"step": 2390
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 2.1937842778793417,
|
| 1687 |
+
"grad_norm": 0.5288705229759216,
|
| 1688 |
+
"learning_rate": 4.616840221123397e-05,
|
| 1689 |
+
"loss": 0.0358,
|
| 1690 |
+
"step": 2400
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 2.202925045703839,
|
| 1694 |
+
"grad_norm": 0.546050488948822,
|
| 1695 |
+
"learning_rate": 4.579175858331219e-05,
|
| 1696 |
+
"loss": 0.0464,
|
| 1697 |
+
"step": 2410
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 2.212065813528336,
|
| 1701 |
+
"grad_norm": 0.5290732383728027,
|
| 1702 |
+
"learning_rate": 4.5415355302152443e-05,
|
| 1703 |
+
"loss": 0.0359,
|
| 1704 |
+
"step": 2420
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 2.2212065813528334,
|
| 1708 |
+
"grad_norm": 0.5098981261253357,
|
| 1709 |
+
"learning_rate": 4.5039213865406816e-05,
|
| 1710 |
+
"loss": 0.0377,
|
| 1711 |
+
"step": 2430
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 2.2303473491773307,
|
| 1715 |
+
"grad_norm": 0.7096160650253296,
|
| 1716 |
+
"learning_rate": 4.466335575577254e-05,
|
| 1717 |
+
"loss": 0.0496,
|
| 1718 |
+
"step": 2440
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 2.2394881170018284,
|
| 1722 |
+
"grad_norm": 0.4775889813899994,
|
| 1723 |
+
"learning_rate": 4.428780243976517e-05,
|
| 1724 |
+
"loss": 0.0392,
|
| 1725 |
+
"step": 2450
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 2.2486288848263256,
|
| 1729 |
+
"grad_norm": 0.5815747380256653,
|
| 1730 |
+
"learning_rate": 4.3912575366492415e-05,
|
| 1731 |
+
"loss": 0.0401,
|
| 1732 |
+
"step": 2460
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 2.257769652650823,
|
| 1736 |
+
"grad_norm": 0.5321184396743774,
|
| 1737 |
+
"learning_rate": 4.353769596642917e-05,
|
| 1738 |
+
"loss": 0.0486,
|
| 1739 |
+
"step": 2470
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 2.26691042047532,
|
| 1743 |
+
"grad_norm": 0.6674132943153381,
|
| 1744 |
+
"learning_rate": 4.316318565019356e-05,
|
| 1745 |
+
"loss": 0.0373,
|
| 1746 |
+
"step": 2480
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 2.2760511882998173,
|
| 1750 |
+
"grad_norm": 0.47233355045318604,
|
| 1751 |
+
"learning_rate": 4.278906580732411e-05,
|
| 1752 |
+
"loss": 0.0304,
|
| 1753 |
+
"step": 2490
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 2.2851919561243146,
|
| 1757 |
+
"grad_norm": 0.5845870971679688,
|
| 1758 |
+
"learning_rate": 4.241535780505809e-05,
|
| 1759 |
+
"loss": 0.0395,
|
| 1760 |
+
"step": 2500
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 2.294332723948812,
|
| 1764 |
+
"grad_norm": 0.6557559967041016,
|
| 1765 |
+
"learning_rate": 4.204208298711116e-05,
|
| 1766 |
+
"loss": 0.0256,
|
| 1767 |
+
"step": 2510
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 2.303473491773309,
|
| 1771 |
+
"grad_norm": 0.7204177975654602,
|
| 1772 |
+
"learning_rate": 4.166926267245842e-05,
|
| 1773 |
+
"loss": 0.0396,
|
| 1774 |
+
"step": 2520
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 2.3126142595978063,
|
| 1778 |
+
"grad_norm": 0.759373128414154,
|
| 1779 |
+
"learning_rate": 4.1296918154116706e-05,
|
| 1780 |
+
"loss": 0.0299,
|
| 1781 |
+
"step": 2530
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 2.3217550274223036,
|
| 1785 |
+
"grad_norm": 0.5927398204803467,
|
| 1786 |
+
"learning_rate": 4.092507069792855e-05,
|
| 1787 |
+
"loss": 0.031,
|
| 1788 |
+
"step": 2540
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 2.330895795246801,
|
| 1792 |
+
"grad_norm": 0.7468788623809814,
|
| 1793 |
+
"learning_rate": 4.055374154134764e-05,
|
| 1794 |
+
"loss": 0.0336,
|
| 1795 |
+
"step": 2550
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 2.340036563071298,
|
| 1799 |
+
"grad_norm": 0.6309873461723328,
|
| 1800 |
+
"learning_rate": 4.0182951892225784e-05,
|
| 1801 |
+
"loss": 0.0298,
|
| 1802 |
+
"step": 2560
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 2.3491773308957953,
|
| 1806 |
+
"grad_norm": 0.594954252243042,
|
| 1807 |
+
"learning_rate": 3.9812722927601724e-05,
|
| 1808 |
+
"loss": 0.037,
|
| 1809 |
+
"step": 2570
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 2.3583180987202925,
|
| 1813 |
+
"grad_norm": 0.4929609000682831,
|
| 1814 |
+
"learning_rate": 3.944307579249168e-05,
|
| 1815 |
+
"loss": 0.0335,
|
| 1816 |
+
"step": 2580
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 2.3674588665447898,
|
| 1820 |
+
"grad_norm": 0.5985615849494934,
|
| 1821 |
+
"learning_rate": 3.907403159868158e-05,
|
| 1822 |
+
"loss": 0.0276,
|
| 1823 |
+
"step": 2590
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 2.376599634369287,
|
| 1827 |
+
"grad_norm": 0.762140154838562,
|
| 1828 |
+
"learning_rate": 3.870561142352137e-05,
|
| 1829 |
+
"loss": 0.0379,
|
| 1830 |
+
"step": 2600
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 2.3857404021937842,
|
| 1834 |
+
"grad_norm": 0.7068864107131958,
|
| 1835 |
+
"learning_rate": 3.833783630872122e-05,
|
| 1836 |
+
"loss": 0.0256,
|
| 1837 |
+
"step": 2610
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 2.3948811700182815,
|
| 1841 |
+
"grad_norm": 0.9355035424232483,
|
| 1842 |
+
"learning_rate": 3.797072725914974e-05,
|
| 1843 |
+
"loss": 0.0289,
|
| 1844 |
+
"step": 2620
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 2.4040219378427787,
|
| 1848 |
+
"grad_norm": 0.5562652945518494,
|
| 1849 |
+
"learning_rate": 3.7604305241634284e-05,
|
| 1850 |
+
"loss": 0.0442,
|
| 1851 |
+
"step": 2630
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 2.413162705667276,
|
| 1855 |
+
"grad_norm": 0.9532639384269714,
|
| 1856 |
+
"learning_rate": 3.723859118376357e-05,
|
| 1857 |
+
"loss": 0.0435,
|
| 1858 |
+
"step": 2640
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 2.422303473491773,
|
| 1862 |
+
"grad_norm": 0.443515807390213,
|
| 1863 |
+
"learning_rate": 3.6873605972692303e-05,
|
| 1864 |
+
"loss": 0.029,
|
| 1865 |
+
"step": 2650
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 2.4314442413162705,
|
| 1869 |
+
"grad_norm": 0.7470449209213257,
|
| 1870 |
+
"learning_rate": 3.650937045394835e-05,
|
| 1871 |
+
"loss": 0.0325,
|
| 1872 |
+
"step": 2660
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 2.4405850091407677,
|
| 1876 |
+
"grad_norm": 0.9110393524169922,
|
| 1877 |
+
"learning_rate": 3.614590543024214e-05,
|
| 1878 |
+
"loss": 0.0282,
|
| 1879 |
+
"step": 2670
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 2.449725776965265,
|
| 1883 |
+
"grad_norm": 0.4086456000804901,
|
| 1884 |
+
"learning_rate": 3.578323166027853e-05,
|
| 1885 |
+
"loss": 0.0418,
|
| 1886 |
+
"step": 2680
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 2.458866544789762,
|
| 1890 |
+
"grad_norm": 0.38568276166915894,
|
| 1891 |
+
"learning_rate": 3.54213698575712e-05,
|
| 1892 |
+
"loss": 0.0406,
|
| 1893 |
+
"step": 2690
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 2.4680073126142594,
|
| 1897 |
+
"grad_norm": 0.5677224397659302,
|
| 1898 |
+
"learning_rate": 3.506034068925973e-05,
|
| 1899 |
+
"loss": 0.0479,
|
| 1900 |
+
"step": 2700
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 2.4771480804387567,
|
| 1904 |
+
"grad_norm": 0.7311156392097473,
|
| 1905 |
+
"learning_rate": 3.470016477492911e-05,
|
| 1906 |
+
"loss": 0.0278,
|
| 1907 |
+
"step": 2710
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 2.4862888482632544,
|
| 1911 |
+
"grad_norm": 0.8331183791160583,
|
| 1912 |
+
"learning_rate": 3.434086268543213e-05,
|
| 1913 |
+
"loss": 0.0342,
|
| 1914 |
+
"step": 2720
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 2.495429616087751,
|
| 1918 |
+
"grad_norm": 0.5231893062591553,
|
| 1919 |
+
"learning_rate": 3.398245494171457e-05,
|
| 1920 |
+
"loss": 0.0313,
|
| 1921 |
+
"step": 2730
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 2.504570383912249,
|
| 1925 |
+
"grad_norm": 0.40512651205062866,
|
| 1926 |
+
"learning_rate": 3.362496201364308e-05,
|
| 1927 |
+
"loss": 0.027,
|
| 1928 |
+
"step": 2740
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 2.5137111517367456,
|
| 1932 |
+
"grad_norm": 0.7198314070701599,
|
| 1933 |
+
"learning_rate": 3.326840431883613e-05,
|
| 1934 |
+
"loss": 0.0386,
|
| 1935 |
+
"step": 2750
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 2.5228519195612433,
|
| 1939 |
+
"grad_norm": 0.5212320685386658,
|
| 1940 |
+
"learning_rate": 3.291280222149788e-05,
|
| 1941 |
+
"loss": 0.0404,
|
| 1942 |
+
"step": 2760
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 2.53199268738574,
|
| 1946 |
+
"grad_norm": 0.3318735659122467,
|
| 1947 |
+
"learning_rate": 3.255817603125518e-05,
|
| 1948 |
+
"loss": 0.0358,
|
| 1949 |
+
"step": 2770
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 2.541133455210238,
|
| 1953 |
+
"grad_norm": 0.4588581323623657,
|
| 1954 |
+
"learning_rate": 3.220454600199748e-05,
|
| 1955 |
+
"loss": 0.0318,
|
| 1956 |
+
"step": 2780
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 2.550274223034735,
|
| 1960 |
+
"grad_norm": 0.47819092869758606,
|
| 1961 |
+
"learning_rate": 3.185193233072016e-05,
|
| 1962 |
+
"loss": 0.0342,
|
| 1963 |
+
"step": 2790
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 2.5594149908592323,
|
| 1967 |
+
"grad_norm": 0.5531491041183472,
|
| 1968 |
+
"learning_rate": 3.150035515637102e-05,
|
| 1969 |
+
"loss": 0.0335,
|
| 1970 |
+
"step": 2800
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 2.5685557586837295,
|
| 1974 |
+
"grad_norm": 0.6975427269935608,
|
| 1975 |
+
"learning_rate": 3.114983455870002e-05,
|
| 1976 |
+
"loss": 0.036,
|
| 1977 |
+
"step": 2810
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 2.577696526508227,
|
| 1981 |
+
"grad_norm": 0.6512779593467712,
|
| 1982 |
+
"learning_rate": 3.080039055711248e-05,
|
| 1983 |
+
"loss": 0.0322,
|
| 1984 |
+
"step": 2820
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 2.586837294332724,
|
| 1988 |
+
"grad_norm": 0.4503866136074066,
|
| 1989 |
+
"learning_rate": 3.045204310952576e-05,
|
| 1990 |
+
"loss": 0.0342,
|
| 1991 |
+
"step": 2830
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 2.5959780621572213,
|
| 1995 |
+
"grad_norm": 0.4288831353187561,
|
| 1996 |
+
"learning_rate": 3.0104812111229307e-05,
|
| 1997 |
+
"loss": 0.0332,
|
| 1998 |
+
"step": 2840
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 2.6051188299817185,
|
| 2002 |
+
"grad_norm": 0.5384790301322937,
|
| 2003 |
+
"learning_rate": 2.9758717393748392e-05,
|
| 2004 |
+
"loss": 0.0343,
|
| 2005 |
+
"step": 2850
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 2.6142595978062158,
|
| 2009 |
+
"grad_norm": 0.44517871737480164,
|
| 2010 |
+
"learning_rate": 2.941377872371155e-05,
|
| 2011 |
+
"loss": 0.044,
|
| 2012 |
+
"step": 2860
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 2.623400365630713,
|
| 2016 |
+
"grad_norm": 0.5115548372268677,
|
| 2017 |
+
"learning_rate": 2.9070015801721518e-05,
|
| 2018 |
+
"loss": 0.0343,
|
| 2019 |
+
"step": 2870
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 2.6325411334552102,
|
| 2023 |
+
"grad_norm": 0.610212504863739,
|
| 2024 |
+
"learning_rate": 2.8727448261230128e-05,
|
| 2025 |
+
"loss": 0.033,
|
| 2026 |
+
"step": 2880
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 2.6416819012797075,
|
| 2030 |
+
"grad_norm": 0.6658434867858887,
|
| 2031 |
+
"learning_rate": 2.8386095667417008e-05,
|
| 2032 |
+
"loss": 0.0368,
|
| 2033 |
+
"step": 2890
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 2.6508226691042047,
|
| 2037 |
+
"grad_norm": 0.6432914733886719,
|
| 2038 |
+
"learning_rate": 2.8045977516072075e-05,
|
| 2039 |
+
"loss": 0.0322,
|
| 2040 |
+
"step": 2900
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 2.659963436928702,
|
| 2044 |
+
"grad_norm": 0.5268338322639465,
|
| 2045 |
+
"learning_rate": 2.77071132324821e-05,
|
| 2046 |
+
"loss": 0.027,
|
| 2047 |
+
"step": 2910
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 2.669104204753199,
|
| 2051 |
+
"grad_norm": 0.6469492316246033,
|
| 2052 |
+
"learning_rate": 2.7369522170321317e-05,
|
| 2053 |
+
"loss": 0.0292,
|
| 2054 |
+
"step": 2920
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 2.6782449725776964,
|
| 2058 |
+
"grad_norm": 0.6078864932060242,
|
| 2059 |
+
"learning_rate": 2.7033223610545967e-05,
|
| 2060 |
+
"loss": 0.0245,
|
| 2061 |
+
"step": 2930
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 2.6873857404021937,
|
| 2065 |
+
"grad_norm": 0.6451284289360046,
|
| 2066 |
+
"learning_rate": 2.6698236760293138e-05,
|
| 2067 |
+
"loss": 0.0309,
|
| 2068 |
+
"step": 2940
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 2.696526508226691,
|
| 2072 |
+
"grad_norm": 0.6047863364219666,
|
| 2073 |
+
"learning_rate": 2.6364580751783872e-05,
|
| 2074 |
+
"loss": 0.0321,
|
| 2075 |
+
"step": 2950
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 2.705667276051188,
|
| 2079 |
+
"grad_norm": 0.53343665599823,
|
| 2080 |
+
"learning_rate": 2.6032274641230325e-05,
|
| 2081 |
+
"loss": 0.0302,
|
| 2082 |
+
"step": 2960
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 2.7148080438756854,
|
| 2086 |
+
"grad_norm": 0.9398259520530701,
|
| 2087 |
+
"learning_rate": 2.570133740774743e-05,
|
| 2088 |
+
"loss": 0.0387,
|
| 2089 |
+
"step": 2970
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 2.7239488117001827,
|
| 2093 |
+
"grad_norm": 0.7435660362243652,
|
| 2094 |
+
"learning_rate": 2.537178795226905e-05,
|
| 2095 |
+
"loss": 0.0326,
|
| 2096 |
+
"step": 2980
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 2.7330895795246803,
|
| 2100 |
+
"grad_norm": 0.8232837915420532,
|
| 2101 |
+
"learning_rate": 2.5043645096468333e-05,
|
| 2102 |
+
"loss": 0.0345,
|
| 2103 |
+
"step": 2990
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 2.742230347349177,
|
| 2107 |
+
"grad_norm": 0.664114236831665,
|
| 2108 |
+
"learning_rate": 2.4716927581682832e-05,
|
| 2109 |
+
"loss": 0.0343,
|
| 2110 |
+
"step": 3000
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 2.751371115173675,
|
| 2114 |
+
"grad_norm": 0.4456695318222046,
|
| 2115 |
+
"learning_rate": 2.4391654067844083e-05,
|
| 2116 |
+
"loss": 0.0262,
|
| 2117 |
+
"step": 3010
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 2.7605118829981716,
|
| 2121 |
+
"grad_norm": 0.8034543991088867,
|
| 2122 |
+
"learning_rate": 2.4067843132411922e-05,
|
| 2123 |
+
"loss": 0.0269,
|
| 2124 |
+
"step": 3020
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 2.7696526508226693,
|
| 2128 |
+
"grad_norm": 0.5905967950820923,
|
| 2129 |
+
"learning_rate": 2.374551326931338e-05,
|
| 2130 |
+
"loss": 0.029,
|
| 2131 |
+
"step": 3030
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 2.778793418647166,
|
| 2135 |
+
"grad_norm": 0.5978665351867676,
|
| 2136 |
+
"learning_rate": 2.3424682887886573e-05,
|
| 2137 |
+
"loss": 0.0386,
|
| 2138 |
+
"step": 3040
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 2.787934186471664,
|
| 2142 |
+
"grad_norm": 0.38325613737106323,
|
| 2143 |
+
"learning_rate": 2.3105370311829124e-05,
|
| 2144 |
+
"loss": 0.0327,
|
| 2145 |
+
"step": 3050
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 2.797074954296161,
|
| 2149 |
+
"grad_norm": 0.8944759964942932,
|
| 2150 |
+
"learning_rate": 2.2787593778151722e-05,
|
| 2151 |
+
"loss": 0.0313,
|
| 2152 |
+
"step": 3060
|
| 2153 |
+
},
|
| 2154 |
+
{
|
| 2155 |
+
"epoch": 2.8062157221206583,
|
| 2156 |
+
"grad_norm": 0.47093465924263,
|
| 2157 |
+
"learning_rate": 2.247137143613653e-05,
|
| 2158 |
+
"loss": 0.0299,
|
| 2159 |
+
"step": 3070
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 2.8153564899451555,
|
| 2163 |
+
"grad_norm": 0.5948256254196167,
|
| 2164 |
+
"learning_rate": 2.2156721346300657e-05,
|
| 2165 |
+
"loss": 0.0293,
|
| 2166 |
+
"step": 3080
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 2.8244972577696528,
|
| 2170 |
+
"grad_norm": 0.48618465662002563,
|
| 2171 |
+
"learning_rate": 2.1843661479364587e-05,
|
| 2172 |
+
"loss": 0.0289,
|
| 2173 |
+
"step": 3090
|
| 2174 |
+
},
|
| 2175 |
+
{
|
| 2176 |
+
"epoch": 2.83363802559415,
|
| 2177 |
+
"grad_norm": 0.5810682773590088,
|
| 2178 |
+
"learning_rate": 2.1532209715225833e-05,
|
| 2179 |
+
"loss": 0.0264,
|
| 2180 |
+
"step": 3100
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 2.8427787934186473,
|
| 2184 |
+
"grad_norm": 0.5437697172164917,
|
| 2185 |
+
"learning_rate": 2.1222383841937826e-05,
|
| 2186 |
+
"loss": 0.0325,
|
| 2187 |
+
"step": 3110
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 2.8519195612431445,
|
| 2191 |
+
"grad_norm": 0.655825138092041,
|
| 2192 |
+
"learning_rate": 2.0914201554693884e-05,
|
| 2193 |
+
"loss": 0.0266,
|
| 2194 |
+
"step": 3120
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 2.8610603290676417,
|
| 2198 |
+
"grad_norm": 0.4643450379371643,
|
| 2199 |
+
"learning_rate": 2.0607680454816612e-05,
|
| 2200 |
+
"loss": 0.0218,
|
| 2201 |
+
"step": 3130
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 2.870201096892139,
|
| 2205 |
+
"grad_norm": 0.4340280592441559,
|
| 2206 |
+
"learning_rate": 2.0302838048752688e-05,
|
| 2207 |
+
"loss": 0.0275,
|
| 2208 |
+
"step": 3140
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 2.8793418647166362,
|
| 2212 |
+
"grad_norm": 0.4632952809333801,
|
| 2213 |
+
"learning_rate": 1.9999691747072925e-05,
|
| 2214 |
+
"loss": 0.0372,
|
| 2215 |
+
"step": 3150
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 2.8884826325411335,
|
| 2219 |
+
"grad_norm": 0.2842860519886017,
|
| 2220 |
+
"learning_rate": 1.9698258863477925e-05,
|
| 2221 |
+
"loss": 0.0284,
|
| 2222 |
+
"step": 3160
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 2.8976234003656307,
|
| 2226 |
+
"grad_norm": 0.4197808802127838,
|
| 2227 |
+
"learning_rate": 1.9398556613809294e-05,
|
| 2228 |
+
"loss": 0.0449,
|
| 2229 |
+
"step": 3170
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 2.906764168190128,
|
| 2233 |
+
"grad_norm": 0.5161760449409485,
|
| 2234 |
+
"learning_rate": 1.9100602115066295e-05,
|
| 2235 |
+
"loss": 0.0294,
|
| 2236 |
+
"step": 3180
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 2.915904936014625,
|
| 2240 |
+
"grad_norm": 0.5271576642990112,
|
| 2241 |
+
"learning_rate": 1.880441238442827e-05,
|
| 2242 |
+
"loss": 0.0243,
|
| 2243 |
+
"step": 3190
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 2.9250457038391224,
|
| 2247 |
+
"grad_norm": 0.5610905289649963,
|
| 2248 |
+
"learning_rate": 1.8510004338282816e-05,
|
| 2249 |
+
"loss": 0.0233,
|
| 2250 |
+
"step": 3200
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 2.9341864716636197,
|
| 2254 |
+
"grad_norm": 0.4908517897129059,
|
| 2255 |
+
"learning_rate": 1.8217394791259483e-05,
|
| 2256 |
+
"loss": 0.0221,
|
| 2257 |
+
"step": 3210
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 2.943327239488117,
|
| 2261 |
+
"grad_norm": 0.3808526396751404,
|
| 2262 |
+
"learning_rate": 1.792660045526954e-05,
|
| 2263 |
+
"loss": 0.0286,
|
| 2264 |
+
"step": 3220
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 2.952468007312614,
|
| 2268 |
+
"grad_norm": 0.4248492419719696,
|
| 2269 |
+
"learning_rate": 1.763763793855149e-05,
|
| 2270 |
+
"loss": 0.0275,
|
| 2271 |
+
"step": 3230
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 2.9616087751371114,
|
| 2275 |
+
"grad_norm": 0.3431572914123535,
|
| 2276 |
+
"learning_rate": 1.7350523744722493e-05,
|
| 2277 |
+
"loss": 0.0363,
|
| 2278 |
+
"step": 3240
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 2.9707495429616086,
|
| 2282 |
+
"grad_norm": 0.5953290462493896,
|
| 2283 |
+
"learning_rate": 1.7065274271835756e-05,
|
| 2284 |
+
"loss": 0.033,
|
| 2285 |
+
"step": 3250
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 2.979890310786106,
|
| 2289 |
+
"grad_norm": 0.852656364440918,
|
| 2290 |
+
"learning_rate": 1.6781905811444122e-05,
|
| 2291 |
+
"loss": 0.0343,
|
| 2292 |
+
"step": 3260
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 2.989031078610603,
|
| 2296 |
+
"grad_norm": 0.7636460065841675,
|
| 2297 |
+
"learning_rate": 1.6500434547669425e-05,
|
| 2298 |
+
"loss": 0.0353,
|
| 2299 |
+
"step": 3270
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 2.998171846435101,
|
| 2303 |
+
"grad_norm": 0.38765236735343933,
|
| 2304 |
+
"learning_rate": 1.6220876556278273e-05,
|
| 2305 |
+
"loss": 0.0236,
|
| 2306 |
+
"step": 3280
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 3.0073126142595976,
|
| 2310 |
+
"grad_norm": 0.18574023246765137,
|
| 2311 |
+
"learning_rate": 1.5943247803763887e-05,
|
| 2312 |
+
"loss": 0.0256,
|
| 2313 |
+
"step": 3290
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 3.016453382084095,
|
| 2317 |
+
"grad_norm": 0.42090004682540894,
|
| 2318 |
+
"learning_rate": 1.5667564146434172e-05,
|
| 2319 |
+
"loss": 0.0222,
|
| 2320 |
+
"step": 3300
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 3.025594149908592,
|
| 2324 |
+
"grad_norm": 0.3421524167060852,
|
| 2325 |
+
"learning_rate": 1.539384132950618e-05,
|
| 2326 |
+
"loss": 0.0179,
|
| 2327 |
+
"step": 3310
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 3.03473491773309,
|
| 2331 |
+
"grad_norm": 0.558494508266449,
|
| 2332 |
+
"learning_rate": 1.5122094986206758e-05,
|
| 2333 |
+
"loss": 0.0247,
|
| 2334 |
+
"step": 3320
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 3.043875685557587,
|
| 2338 |
+
"grad_norm": 0.4569318890571594,
|
| 2339 |
+
"learning_rate": 1.485234063687973e-05,
|
| 2340 |
+
"loss": 0.0217,
|
| 2341 |
+
"step": 3330
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 3.0530164533820843,
|
| 2345 |
+
"grad_norm": 1.014539122581482,
|
| 2346 |
+
"learning_rate": 1.4584593688099473e-05,
|
| 2347 |
+
"loss": 0.0247,
|
| 2348 |
+
"step": 3340
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 3.0621572212065815,
|
| 2352 |
+
"grad_norm": 0.4421018362045288,
|
| 2353 |
+
"learning_rate": 1.4318869431791e-05,
|
| 2354 |
+
"loss": 0.0189,
|
| 2355 |
+
"step": 3350
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 3.0712979890310788,
|
| 2359 |
+
"grad_norm": 0.45253804326057434,
|
| 2360 |
+
"learning_rate": 1.40551830443566e-05,
|
| 2361 |
+
"loss": 0.0276,
|
| 2362 |
+
"step": 3360
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 3.080438756855576,
|
| 2366 |
+
"grad_norm": 0.3116052448749542,
|
| 2367 |
+
"learning_rate": 1.3793549585809023e-05,
|
| 2368 |
+
"loss": 0.0257,
|
| 2369 |
+
"step": 3370
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 3.0895795246800732,
|
| 2373 |
+
"grad_norm": 0.5444056391716003,
|
| 2374 |
+
"learning_rate": 1.3533983998911353e-05,
|
| 2375 |
+
"loss": 0.0251,
|
| 2376 |
+
"step": 3380
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 3.0987202925045705,
|
| 2380 |
+
"grad_norm": 0.2725338041782379,
|
| 2381 |
+
"learning_rate": 1.3276501108323664e-05,
|
| 2382 |
+
"loss": 0.026,
|
| 2383 |
+
"step": 3390
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 3.1078610603290677,
|
| 2387 |
+
"grad_norm": 0.4318808317184448,
|
| 2388 |
+
"learning_rate": 1.3021115619756214e-05,
|
| 2389 |
+
"loss": 0.0226,
|
| 2390 |
+
"step": 3400
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 3.117001828153565,
|
| 2394 |
+
"grad_norm": 0.5103743076324463,
|
| 2395 |
+
"learning_rate": 1.2767842119129597e-05,
|
| 2396 |
+
"loss": 0.0246,
|
| 2397 |
+
"step": 3410
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 3.126142595978062,
|
| 2401 |
+
"grad_norm": 0.7406987547874451,
|
| 2402 |
+
"learning_rate": 1.2516695071741762e-05,
|
| 2403 |
+
"loss": 0.0216,
|
| 2404 |
+
"step": 3420
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 3.1352833638025595,
|
| 2408 |
+
"grad_norm": 0.6604713797569275,
|
| 2409 |
+
"learning_rate": 1.2267688821441742e-05,
|
| 2410 |
+
"loss": 0.0345,
|
| 2411 |
+
"step": 3430
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 3.1444241316270567,
|
| 2415 |
+
"grad_norm": 0.42933446168899536,
|
| 2416 |
+
"learning_rate": 1.2020837589810474e-05,
|
| 2417 |
+
"loss": 0.0226,
|
| 2418 |
+
"step": 3440
|
| 2419 |
+
},
|
| 2420 |
+
{
|
| 2421 |
+
"epoch": 3.153564899451554,
|
| 2422 |
+
"grad_norm": 0.4417478144168854,
|
| 2423 |
+
"learning_rate": 1.177615547534861e-05,
|
| 2424 |
+
"loss": 0.0216,
|
| 2425 |
+
"step": 3450
|
| 2426 |
+
},
|
| 2427 |
+
{
|
| 2428 |
+
"epoch": 3.162705667276051,
|
| 2429 |
+
"grad_norm": 0.41561007499694824,
|
| 2430 |
+
"learning_rate": 1.153365645267121e-05,
|
| 2431 |
+
"loss": 0.0314,
|
| 2432 |
+
"step": 3460
|
| 2433 |
+
},
|
| 2434 |
+
{
|
| 2435 |
+
"epoch": 3.1718464351005484,
|
| 2436 |
+
"grad_norm": 0.49010348320007324,
|
| 2437 |
+
"learning_rate": 1.1293354371709647e-05,
|
| 2438 |
+
"loss": 0.027,
|
| 2439 |
+
"step": 3470
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 3.1809872029250457,
|
| 2443 |
+
"grad_norm": 0.6788333654403687,
|
| 2444 |
+
"learning_rate": 1.1055262956920647e-05,
|
| 2445 |
+
"loss": 0.0188,
|
| 2446 |
+
"step": 3480
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 3.190127970749543,
|
| 2450 |
+
"grad_norm": 0.6015557646751404,
|
| 2451 |
+
"learning_rate": 1.081939580650233e-05,
|
| 2452 |
+
"loss": 0.0343,
|
| 2453 |
+
"step": 3490
|
| 2454 |
+
},
|
| 2455 |
+
{
|
| 2456 |
+
"epoch": 3.19926873857404,
|
| 2457 |
+
"grad_norm": 0.4312863051891327,
|
| 2458 |
+
"learning_rate": 1.0585766391617641e-05,
|
| 2459 |
+
"loss": 0.0199,
|
| 2460 |
+
"step": 3500
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"epoch": 3.2084095063985374,
|
| 2464 |
+
"grad_norm": 0.6486406326293945,
|
| 2465 |
+
"learning_rate": 1.0354388055624987e-05,
|
| 2466 |
+
"loss": 0.0218,
|
| 2467 |
+
"step": 3510
|
| 2468 |
+
},
|
| 2469 |
+
{
|
| 2470 |
+
"epoch": 3.2175502742230346,
|
| 2471 |
+
"grad_norm": 0.23588594794273376,
|
| 2472 |
+
"learning_rate": 1.0125274013316078e-05,
|
| 2473 |
+
"loss": 0.02,
|
| 2474 |
+
"step": 3520
|
| 2475 |
+
},
|
| 2476 |
+
{
|
| 2477 |
+
"epoch": 3.226691042047532,
|
| 2478 |
+
"grad_norm": 0.44049379229545593,
|
| 2479 |
+
"learning_rate": 9.898437350161238e-06,
|
| 2480 |
+
"loss": 0.018,
|
| 2481 |
+
"step": 3530
|
| 2482 |
+
},
|
| 2483 |
+
{
|
| 2484 |
+
"epoch": 3.235831809872029,
|
| 2485 |
+
"grad_norm": 0.3360377848148346,
|
| 2486 |
+
"learning_rate": 9.67389102156207e-06,
|
| 2487 |
+
"loss": 0.0195,
|
| 2488 |
+
"step": 3540
|
| 2489 |
+
},
|
| 2490 |
+
{
|
| 2491 |
+
"epoch": 3.2449725776965264,
|
| 2492 |
+
"grad_norm": 0.5655617117881775,
|
| 2493 |
+
"learning_rate": 9.451647852111472e-06,
|
| 2494 |
+
"loss": 0.0222,
|
| 2495 |
+
"step": 3550
|
| 2496 |
+
},
|
| 2497 |
+
{
|
| 2498 |
+
"epoch": 3.2541133455210236,
|
| 2499 |
+
"grad_norm": 0.599947452545166,
|
| 2500 |
+
"learning_rate": 9.231720534861187e-06,
|
| 2501 |
+
"loss": 0.0262,
|
| 2502 |
+
"step": 3560
|
| 2503 |
+
},
|
| 2504 |
+
{
|
| 2505 |
+
"epoch": 3.263254113345521,
|
| 2506 |
+
"grad_norm": 0.4483935236930847,
|
| 2507 |
+
"learning_rate": 9.014121630596944e-06,
|
| 2508 |
+
"loss": 0.0286,
|
| 2509 |
+
"step": 3570
|
| 2510 |
+
},
|
| 2511 |
+
{
|
| 2512 |
+
"epoch": 3.272394881170018,
|
| 2513 |
+
"grad_norm": 0.5928664803504944,
|
| 2514 |
+
"learning_rate": 8.798863567120952e-06,
|
| 2515 |
+
"loss": 0.0222,
|
| 2516 |
+
"step": 3580
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 3.2815356489945158,
|
| 2520 |
+
"grad_norm": 0.42674168944358826,
|
| 2521 |
+
"learning_rate": 8.585958638542191e-06,
|
| 2522 |
+
"loss": 0.0173,
|
| 2523 |
+
"step": 3590
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 3.2906764168190126,
|
| 2527 |
+
"grad_norm": 0.4651152491569519,
|
| 2528 |
+
"learning_rate": 8.375419004574215e-06,
|
| 2529 |
+
"loss": 0.0357,
|
| 2530 |
+
"step": 3600
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 3.2998171846435103,
|
| 2534 |
+
"grad_norm": 0.6086673736572266,
|
| 2535 |
+
"learning_rate": 8.16725668984068e-06,
|
| 2536 |
+
"loss": 0.017,
|
| 2537 |
+
"step": 3610
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 3.3089579524680075,
|
| 2541 |
+
"grad_norm": 0.2967171370983124,
|
| 2542 |
+
"learning_rate": 7.961483583188573e-06,
|
| 2543 |
+
"loss": 0.0315,
|
| 2544 |
+
"step": 3620
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 3.3180987202925047,
|
| 2548 |
+
"grad_norm": 0.3449037969112396,
|
| 2549 |
+
"learning_rate": 7.758111437009235e-06,
|
| 2550 |
+
"loss": 0.0257,
|
| 2551 |
+
"step": 3630
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 3.327239488117002,
|
| 2555 |
+
"grad_norm": 0.7032669186592102,
|
| 2556 |
+
"learning_rate": 7.557151866567086e-06,
|
| 2557 |
+
"loss": 0.0257,
|
| 2558 |
+
"step": 3640
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 3.3363802559414992,
|
| 2562 |
+
"grad_norm": 0.5395845174789429,
|
| 2563 |
+
"learning_rate": 7.358616349336261e-06,
|
| 2564 |
+
"loss": 0.0273,
|
| 2565 |
+
"step": 3650
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 3.3455210237659965,
|
| 2569 |
+
"grad_norm": 0.46958377957344055,
|
| 2570 |
+
"learning_rate": 7.162516224345101e-06,
|
| 2571 |
+
"loss": 0.0178,
|
| 2572 |
+
"step": 3660
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 3.3546617915904937,
|
| 2576 |
+
"grad_norm": 0.4303223788738251,
|
| 2577 |
+
"learning_rate": 6.968862691528566e-06,
|
| 2578 |
+
"loss": 0.0202,
|
| 2579 |
+
"step": 3670
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 3.363802559414991,
|
| 2583 |
+
"grad_norm": 0.6079144477844238,
|
| 2584 |
+
"learning_rate": 6.777666811088518e-06,
|
| 2585 |
+
"loss": 0.0197,
|
| 2586 |
+
"step": 3680
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 3.372943327239488,
|
| 2590 |
+
"grad_norm": 0.38781973719596863,
|
| 2591 |
+
"learning_rate": 6.588939502862041e-06,
|
| 2592 |
+
"loss": 0.029,
|
| 2593 |
+
"step": 3690
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 3.3820840950639854,
|
| 2597 |
+
"grad_norm": 0.3500356078147888,
|
| 2598 |
+
"learning_rate": 6.402691545697842e-06,
|
| 2599 |
+
"loss": 0.0239,
|
| 2600 |
+
"step": 3700
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 3.3912248628884827,
|
| 2604 |
+
"grad_norm": 0.637479841709137,
|
| 2605 |
+
"learning_rate": 6.218933576840547e-06,
|
| 2606 |
+
"loss": 0.0208,
|
| 2607 |
+
"step": 3710
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 3.40036563071298,
|
| 2611 |
+
"grad_norm": 0.414711594581604,
|
| 2612 |
+
"learning_rate": 6.037676091323208e-06,
|
| 2613 |
+
"loss": 0.0259,
|
| 2614 |
+
"step": 3720
|
| 2615 |
+
},
|
| 2616 |
+
{
|
| 2617 |
+
"epoch": 3.409506398537477,
|
| 2618 |
+
"grad_norm": 0.7444067001342773,
|
| 2619 |
+
"learning_rate": 5.858929441367939e-06,
|
| 2620 |
+
"loss": 0.033,
|
| 2621 |
+
"step": 3730
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 3.4186471663619744,
|
| 2625 |
+
"grad_norm": 0.35689577460289,
|
| 2626 |
+
"learning_rate": 5.682703835794578e-06,
|
| 2627 |
+
"loss": 0.0263,
|
| 2628 |
+
"step": 3740
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 3.4277879341864717,
|
| 2632 |
+
"grad_norm": 0.39409419894218445,
|
| 2633 |
+
"learning_rate": 5.5090093394376875e-06,
|
| 2634 |
+
"loss": 0.0278,
|
| 2635 |
+
"step": 3750
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"epoch": 3.436928702010969,
|
| 2639 |
+
"grad_norm": 0.4039927124977112,
|
| 2640 |
+
"learning_rate": 5.337855872571734e-06,
|
| 2641 |
+
"loss": 0.0208,
|
| 2642 |
+
"step": 3760
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"epoch": 3.446069469835466,
|
| 2646 |
+
"grad_norm": 0.6069719791412354,
|
| 2647 |
+
"learning_rate": 5.1692532103444494e-06,
|
| 2648 |
+
"loss": 0.0296,
|
| 2649 |
+
"step": 3770
|
| 2650 |
+
},
|
| 2651 |
+
{
|
| 2652 |
+
"epoch": 3.4552102376599634,
|
| 2653 |
+
"grad_norm": 0.4256989657878876,
|
| 2654 |
+
"learning_rate": 5.003210982218559e-06,
|
| 2655 |
+
"loss": 0.0191,
|
| 2656 |
+
"step": 3780
|
| 2657 |
+
},
|
| 2658 |
+
{
|
| 2659 |
+
"epoch": 3.4643510054844606,
|
| 2660 |
+
"grad_norm": 0.36178430914878845,
|
| 2661 |
+
"learning_rate": 4.839738671421862e-06,
|
| 2662 |
+
"loss": 0.0235,
|
| 2663 |
+
"step": 3790
|
| 2664 |
+
},
|
| 2665 |
+
{
|
| 2666 |
+
"epoch": 3.473491773308958,
|
| 2667 |
+
"grad_norm": 0.29829517006874084,
|
| 2668 |
+
"learning_rate": 4.678845614405541e-06,
|
| 2669 |
+
"loss": 0.0173,
|
| 2670 |
+
"step": 3800
|
| 2671 |
+
},
|
| 2672 |
+
{
|
| 2673 |
+
"epoch": 3.482632541133455,
|
| 2674 |
+
"grad_norm": 0.5337382555007935,
|
| 2675 |
+
"learning_rate": 4.5205410003109515e-06,
|
| 2676 |
+
"loss": 0.0265,
|
| 2677 |
+
"step": 3810
|
| 2678 |
+
},
|
| 2679 |
+
{
|
| 2680 |
+
"epoch": 3.4917733089579523,
|
| 2681 |
+
"grad_norm": 0.49504226446151733,
|
| 2682 |
+
"learning_rate": 4.364833870444834e-06,
|
| 2683 |
+
"loss": 0.0252,
|
| 2684 |
+
"step": 3820
|
| 2685 |
+
},
|
| 2686 |
+
{
|
| 2687 |
+
"epoch": 3.5009140767824496,
|
| 2688 |
+
"grad_norm": 0.6263189911842346,
|
| 2689 |
+
"learning_rate": 4.211733117762884e-06,
|
| 2690 |
+
"loss": 0.0357,
|
| 2691 |
+
"step": 3830
|
| 2692 |
+
},
|
| 2693 |
+
{
|
| 2694 |
+
"epoch": 3.510054844606947,
|
| 2695 |
+
"grad_norm": 0.5205623507499695,
|
| 2696 |
+
"learning_rate": 4.061247486361852e-06,
|
| 2697 |
+
"loss": 0.0246,
|
| 2698 |
+
"step": 3840
|
| 2699 |
+
},
|
| 2700 |
+
{
|
| 2701 |
+
"epoch": 3.519195612431444,
|
| 2702 |
+
"grad_norm": 0.3644954264163971,
|
| 2703 |
+
"learning_rate": 3.913385570980194e-06,
|
| 2704 |
+
"loss": 0.0198,
|
| 2705 |
+
"step": 3850
|
| 2706 |
+
},
|
| 2707 |
+
{
|
| 2708 |
+
"epoch": 3.5283363802559418,
|
| 2709 |
+
"grad_norm": 0.5801895260810852,
|
| 2710 |
+
"learning_rate": 3.7681558165071186e-06,
|
| 2711 |
+
"loss": 0.0179,
|
| 2712 |
+
"step": 3860
|
| 2713 |
+
},
|
| 2714 |
+
{
|
| 2715 |
+
"epoch": 3.5374771480804386,
|
| 2716 |
+
"grad_norm": 0.44283440709114075,
|
| 2717 |
+
"learning_rate": 3.625566517500334e-06,
|
| 2718 |
+
"loss": 0.0205,
|
| 2719 |
+
"step": 3870
|
| 2720 |
+
},
|
| 2721 |
+
{
|
| 2722 |
+
"epoch": 3.5466179159049362,
|
| 2723 |
+
"grad_norm": 0.6655595898628235,
|
| 2724 |
+
"learning_rate": 3.4856258177122735e-06,
|
| 2725 |
+
"loss": 0.018,
|
| 2726 |
+
"step": 3880
|
| 2727 |
+
},
|
| 2728 |
+
{
|
| 2729 |
+
"epoch": 3.555758683729433,
|
| 2730 |
+
"grad_norm": 0.5158295631408691,
|
| 2731 |
+
"learning_rate": 3.348341709625008e-06,
|
| 2732 |
+
"loss": 0.0195,
|
| 2733 |
+
"step": 3890
|
| 2734 |
+
},
|
| 2735 |
+
{
|
| 2736 |
+
"epoch": 3.5648994515539307,
|
| 2737 |
+
"grad_norm": 0.4960062503814697,
|
| 2738 |
+
"learning_rate": 3.2137220339937658e-06,
|
| 2739 |
+
"loss": 0.0211,
|
| 2740 |
+
"step": 3900
|
| 2741 |
+
},
|
| 2742 |
+
{
|
| 2743 |
+
"epoch": 3.5740402193784275,
|
| 2744 |
+
"grad_norm": 0.6395726799964905,
|
| 2745 |
+
"learning_rate": 3.081774479399102e-06,
|
| 2746 |
+
"loss": 0.0203,
|
| 2747 |
+
"step": 3910
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 3.583180987202925,
|
| 2751 |
+
"grad_norm": 0.430395245552063,
|
| 2752 |
+
"learning_rate": 2.9525065818077956e-06,
|
| 2753 |
+
"loss": 0.0219,
|
| 2754 |
+
"step": 3920
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 3.5923217550274225,
|
| 2758 |
+
"grad_norm": 0.5035272836685181,
|
| 2759 |
+
"learning_rate": 2.8259257241424253e-06,
|
| 2760 |
+
"loss": 0.0182,
|
| 2761 |
+
"step": 3930
|
| 2762 |
+
},
|
| 2763 |
+
{
|
| 2764 |
+
"epoch": 3.6014625228519197,
|
| 2765 |
+
"grad_norm": 0.5497689843177795,
|
| 2766 |
+
"learning_rate": 2.7020391358597464e-06,
|
| 2767 |
+
"loss": 0.029,
|
| 2768 |
+
"step": 3940
|
| 2769 |
+
},
|
| 2770 |
+
{
|
| 2771 |
+
"epoch": 3.610603290676417,
|
| 2772 |
+
"grad_norm": 0.5781487226486206,
|
| 2773 |
+
"learning_rate": 2.5808538925377433e-06,
|
| 2774 |
+
"loss": 0.0187,
|
| 2775 |
+
"step": 3950
|
| 2776 |
+
},
|
| 2777 |
+
{
|
| 2778 |
+
"epoch": 3.619744058500914,
|
| 2779 |
+
"grad_norm": 0.26463502645492554,
|
| 2780 |
+
"learning_rate": 2.462376915471548e-06,
|
| 2781 |
+
"loss": 0.0238,
|
| 2782 |
+
"step": 3960
|
| 2783 |
+
},
|
| 2784 |
+
{
|
| 2785 |
+
"epoch": 3.6288848263254114,
|
| 2786 |
+
"grad_norm": 0.5389289855957031,
|
| 2787 |
+
"learning_rate": 2.3466149712781384e-06,
|
| 2788 |
+
"loss": 0.0196,
|
| 2789 |
+
"step": 3970
|
| 2790 |
+
},
|
| 2791 |
+
{
|
| 2792 |
+
"epoch": 3.6380255941499087,
|
| 2793 |
+
"grad_norm": 0.22814074158668518,
|
| 2794 |
+
"learning_rate": 2.2335746715098747e-06,
|
| 2795 |
+
"loss": 0.0134,
|
| 2796 |
+
"step": 3980
|
| 2797 |
+
},
|
| 2798 |
+
{
|
| 2799 |
+
"epoch": 3.647166361974406,
|
| 2800 |
+
"grad_norm": 0.21715475618839264,
|
| 2801 |
+
"learning_rate": 2.1232624722768822e-06,
|
| 2802 |
+
"loss": 0.0172,
|
| 2803 |
+
"step": 3990
|
| 2804 |
+
},
|
| 2805 |
+
{
|
| 2806 |
+
"epoch": 3.656307129798903,
|
| 2807 |
+
"grad_norm": 0.8281272053718567,
|
| 2808 |
+
"learning_rate": 2.0156846738783264e-06,
|
| 2809 |
+
"loss": 0.0263,
|
| 2810 |
+
"step": 4000
|
| 2811 |
+
},
|
| 2812 |
+
{
|
| 2813 |
+
"epoch": 3.6654478976234004,
|
| 2814 |
+
"grad_norm": 0.6629803776741028,
|
| 2815 |
+
"learning_rate": 1.910847420442591e-06,
|
| 2816 |
+
"loss": 0.0244,
|
| 2817 |
+
"step": 4010
|
| 2818 |
+
},
|
| 2819 |
+
{
|
| 2820 |
+
"epoch": 3.6745886654478976,
|
| 2821 |
+
"grad_norm": 0.5557454228401184,
|
| 2822 |
+
"learning_rate": 1.8087566995763472e-06,
|
| 2823 |
+
"loss": 0.0252,
|
| 2824 |
+
"step": 4020
|
| 2825 |
+
},
|
| 2826 |
+
{
|
| 2827 |
+
"epoch": 3.683729433272395,
|
| 2828 |
+
"grad_norm": 0.5342206358909607,
|
| 2829 |
+
"learning_rate": 1.7094183420225952e-06,
|
| 2830 |
+
"loss": 0.0183,
|
| 2831 |
+
"step": 4030
|
| 2832 |
+
},
|
| 2833 |
+
{
|
| 2834 |
+
"epoch": 3.692870201096892,
|
| 2835 |
+
"grad_norm": 0.6376899480819702,
|
| 2836 |
+
"learning_rate": 1.6128380213276618e-06,
|
| 2837 |
+
"loss": 0.0267,
|
| 2838 |
+
"step": 4040
|
| 2839 |
+
},
|
| 2840 |
+
{
|
| 2841 |
+
"epoch": 3.7020109689213894,
|
| 2842 |
+
"grad_norm": 0.5518518686294556,
|
| 2843 |
+
"learning_rate": 1.5190212535171334e-06,
|
| 2844 |
+
"loss": 0.0187,
|
| 2845 |
+
"step": 4050
|
| 2846 |
+
},
|
| 2847 |
+
{
|
| 2848 |
+
"epoch": 3.7111517367458866,
|
| 2849 |
+
"grad_norm": 0.580907940864563,
|
| 2850 |
+
"learning_rate": 1.4279733967808352e-06,
|
| 2851 |
+
"loss": 0.0287,
|
| 2852 |
+
"step": 4060
|
| 2853 |
+
},
|
| 2854 |
+
{
|
| 2855 |
+
"epoch": 3.720292504570384,
|
| 2856 |
+
"grad_norm": 0.3451583683490753,
|
| 2857 |
+
"learning_rate": 1.3396996511668203e-06,
|
| 2858 |
+
"loss": 0.0227,
|
| 2859 |
+
"step": 4070
|
| 2860 |
+
},
|
| 2861 |
+
{
|
| 2862 |
+
"epoch": 3.729433272394881,
|
| 2863 |
+
"grad_norm": 0.5293395519256592,
|
| 2864 |
+
"learning_rate": 1.2542050582843578e-06,
|
| 2865 |
+
"loss": 0.0329,
|
| 2866 |
+
"step": 4080
|
| 2867 |
+
},
|
| 2868 |
+
{
|
| 2869 |
+
"epoch": 3.7385740402193783,
|
| 2870 |
+
"grad_norm": 0.4432827830314636,
|
| 2871 |
+
"learning_rate": 1.1714945010159905e-06,
|
| 2872 |
+
"loss": 0.0275,
|
| 2873 |
+
"step": 4090
|
| 2874 |
+
},
|
| 2875 |
+
{
|
| 2876 |
+
"epoch": 3.7477148080438756,
|
| 2877 |
+
"grad_norm": 0.4323786199092865,
|
| 2878 |
+
"learning_rate": 1.0915727032386747e-06,
|
| 2879 |
+
"loss": 0.022,
|
| 2880 |
+
"step": 4100
|
| 2881 |
+
},
|
| 2882 |
+
{
|
| 2883 |
+
"epoch": 3.756855575868373,
|
| 2884 |
+
"grad_norm": 0.19359391927719116,
|
| 2885 |
+
"learning_rate": 1.0144442295539625e-06,
|
| 2886 |
+
"loss": 0.0196,
|
| 2887 |
+
"step": 4110
|
| 2888 |
+
},
|
| 2889 |
+
{
|
| 2890 |
+
"epoch": 3.76599634369287,
|
| 2891 |
+
"grad_norm": 0.49890777468681335,
|
| 2892 |
+
"learning_rate": 9.401134850273053e-07,
|
| 2893 |
+
"loss": 0.0205,
|
| 2894 |
+
"step": 4120
|
| 2895 |
+
},
|
| 2896 |
+
{
|
| 2897 |
+
"epoch": 3.7751371115173673,
|
| 2898 |
+
"grad_norm": 0.5493078827857971,
|
| 2899 |
+
"learning_rate": 8.685847149364878e-07,
|
| 2900 |
+
"loss": 0.02,
|
| 2901 |
+
"step": 4130
|
| 2902 |
+
},
|
| 2903 |
+
{
|
| 2904 |
+
"epoch": 3.7842778793418645,
|
| 2905 |
+
"grad_norm": 0.5801013708114624,
|
| 2906 |
+
"learning_rate": 7.99862004529145e-07,
|
| 2907 |
+
"loss": 0.0278,
|
| 2908 |
+
"step": 4140
|
| 2909 |
+
},
|
| 2910 |
+
{
|
| 2911 |
+
"epoch": 3.7934186471663622,
|
| 2912 |
+
"grad_norm": 0.4189988374710083,
|
| 2913 |
+
"learning_rate": 7.339492787894364e-07,
|
| 2914 |
+
"loss": 0.0199,
|
| 2915 |
+
"step": 4150
|
| 2916 |
+
},
|
| 2917 |
+
{
|
| 2918 |
+
"epoch": 3.802559414990859,
|
| 2919 |
+
"grad_norm": 0.5373661518096924,
|
| 2920 |
+
"learning_rate": 6.708503022138879e-07,
|
| 2921 |
+
"loss": 0.0214,
|
| 2922 |
+
"step": 4160
|
| 2923 |
+
},
|
| 2924 |
+
{
|
| 2925 |
+
"epoch": 3.8117001828153567,
|
| 2926 |
+
"grad_norm": 0.5810131430625916,
|
| 2927 |
+
"learning_rate": 6.105686785964015e-07,
|
| 2928 |
+
"loss": 0.0243,
|
| 2929 |
+
"step": 4170
|
| 2930 |
+
},
|
| 2931 |
+
{
|
| 2932 |
+
"epoch": 3.8208409506398535,
|
| 2933 |
+
"grad_norm": 0.6303432583808899,
|
| 2934 |
+
"learning_rate": 5.531078508223986e-07,
|
| 2935 |
+
"loss": 0.0209,
|
| 2936 |
+
"step": 4180
|
| 2937 |
+
},
|
| 2938 |
+
{
|
| 2939 |
+
"epoch": 3.829981718464351,
|
| 2940 |
+
"grad_norm": 0.44799119234085083,
|
| 2941 |
+
"learning_rate": 4.984711006722042e-07,
|
| 2942 |
+
"loss": 0.0228,
|
| 2943 |
+
"step": 4190
|
| 2944 |
+
},
|
| 2945 |
+
{
|
| 2946 |
+
"epoch": 3.839122486288848,
|
| 2947 |
+
"grad_norm": 0.3489096760749817,
|
| 2948 |
+
"learning_rate": 4.4666154863361453e-07,
|
| 2949 |
+
"loss": 0.0131,
|
| 2950 |
+
"step": 4200
|
| 2951 |
+
},
|
| 2952 |
+
{
|
| 2953 |
+
"epoch": 3.8482632541133457,
|
| 2954 |
+
"grad_norm": 0.368489146232605,
|
| 2955 |
+
"learning_rate": 3.9768215372367745e-07,
|
| 2956 |
+
"loss": 0.0232,
|
| 2957 |
+
"step": 4210
|
| 2958 |
+
},
|
| 2959 |
+
{
|
| 2960 |
+
"epoch": 3.857404021937843,
|
| 2961 |
+
"grad_norm": 0.34034988284111023,
|
| 2962 |
+
"learning_rate": 3.5153571331968394e-07,
|
| 2963 |
+
"loss": 0.0188,
|
| 2964 |
+
"step": 4220
|
| 2965 |
+
},
|
| 2966 |
+
{
|
| 2967 |
+
"epoch": 3.86654478976234,
|
| 2968 |
+
"grad_norm": 0.4425741732120514,
|
| 2969 |
+
"learning_rate": 3.082248629994011e-07,
|
| 2970 |
+
"loss": 0.0145,
|
| 2971 |
+
"step": 4230
|
| 2972 |
+
},
|
| 2973 |
+
{
|
| 2974 |
+
"epoch": 3.8756855575868374,
|
| 2975 |
+
"grad_norm": 0.5476856827735901,
|
| 2976 |
+
"learning_rate": 2.677520763905539e-07,
|
| 2977 |
+
"loss": 0.0227,
|
| 2978 |
+
"step": 4240
|
| 2979 |
+
},
|
| 2980 |
+
{
|
| 2981 |
+
"epoch": 3.8848263254113347,
|
| 2982 |
+
"grad_norm": 0.49523550271987915,
|
| 2983 |
+
"learning_rate": 2.301196650295434e-07,
|
| 2984 |
+
"loss": 0.0377,
|
| 2985 |
+
"step": 4250
|
| 2986 |
+
},
|
| 2987 |
+
{
|
| 2988 |
+
"epoch": 3.893967093235832,
|
| 2989 |
+
"grad_norm": 0.4109029471874237,
|
| 2990 |
+
"learning_rate": 1.95329778229425e-07,
|
| 2991 |
+
"loss": 0.0268,
|
| 2992 |
+
"step": 4260
|
| 2993 |
+
},
|
| 2994 |
+
{
|
| 2995 |
+
"epoch": 3.903107861060329,
|
| 2996 |
+
"grad_norm": 0.48295220732688904,
|
| 2997 |
+
"learning_rate": 1.6338440295715073e-07,
|
| 2998 |
+
"loss": 0.0255,
|
| 2999 |
+
"step": 4270
|
| 3000 |
+
},
|
| 3001 |
+
{
|
| 3002 |
+
"epoch": 3.9122486288848264,
|
| 3003 |
+
"grad_norm": 0.3963299095630646,
|
| 3004 |
+
"learning_rate": 1.3428536372009914e-07,
|
| 3005 |
+
"loss": 0.0163,
|
| 3006 |
+
"step": 4280
|
| 3007 |
+
},
|
| 3008 |
+
{
|
| 3009 |
+
"epoch": 3.9213893967093236,
|
| 3010 |
+
"grad_norm": 0.6109392642974854,
|
| 3011 |
+
"learning_rate": 1.0803432246186963e-07,
|
| 3012 |
+
"loss": 0.0303,
|
| 3013 |
+
"step": 4290
|
| 3014 |
+
},
|
| 3015 |
+
{
|
| 3016 |
+
"epoch": 3.930530164533821,
|
| 3017 |
+
"grad_norm": 0.5676685571670532,
|
| 3018 |
+
"learning_rate": 8.46327784673473e-08,
|
| 3019 |
+
"loss": 0.025,
|
| 3020 |
+
"step": 4300
|
| 3021 |
+
},
|
| 3022 |
+
{
|
| 3023 |
+
"epoch": 3.939670932358318,
|
| 3024 |
+
"grad_norm": 0.4061093032360077,
|
| 3025 |
+
"learning_rate": 6.408206827708818e-08,
|
| 3026 |
+
"loss": 0.0219,
|
| 3027 |
+
"step": 4310
|
| 3028 |
+
},
|
| 3029 |
+
{
|
| 3030 |
+
"epoch": 3.9488117001828154,
|
| 3031 |
+
"grad_norm": 0.4048132598400116,
|
| 3032 |
+
"learning_rate": 4.6383365610980224e-08,
|
| 3033 |
+
"loss": 0.0274,
|
| 3034 |
+
"step": 4320
|
| 3035 |
+
},
|
| 3036 |
+
{
|
| 3037 |
+
"epoch": 3.9579524680073126,
|
| 3038 |
+
"grad_norm": 0.9811397194862366,
|
| 3039 |
+
"learning_rate": 3.153768130121359e-08,
|
| 3040 |
+
"loss": 0.0187,
|
| 3041 |
+
"step": 4330
|
| 3042 |
+
},
|
| 3043 |
+
{
|
| 3044 |
+
"epoch": 3.96709323583181,
|
| 3045 |
+
"grad_norm": 0.4111926257610321,
|
| 3046 |
+
"learning_rate": 1.954586323454355e-08,
|
| 3047 |
+
"loss": 0.025,
|
| 3048 |
+
"step": 4340
|
| 3049 |
+
},
|
| 3050 |
+
{
|
| 3051 |
+
"epoch": 3.976234003656307,
|
| 3052 |
+
"grad_norm": 0.37575072050094604,
|
| 3053 |
+
"learning_rate": 1.0408596303862483e-08,
|
| 3054 |
+
"loss": 0.0206,
|
| 3055 |
+
"step": 4350
|
| 3056 |
+
},
|
| 3057 |
+
{
|
| 3058 |
+
"epoch": 3.9853747714808043,
|
| 3059 |
+
"grad_norm": 0.4568402171134949,
|
| 3060 |
+
"learning_rate": 4.126402369086791e-09,
|
| 3061 |
+
"loss": 0.0202,
|
| 3062 |
+
"step": 4360
|
| 3063 |
+
},
|
| 3064 |
+
{
|
| 3065 |
+
"epoch": 3.9945155393053016,
|
| 3066 |
+
"grad_norm": 0.6871432065963745,
|
| 3067 |
+
"learning_rate": 6.996402273529157e-10,
|
| 3068 |
+
"loss": 0.0248,
|
| 3069 |
+
"step": 4370
|
| 3070 |
+
},
|
| 3071 |
+
{
|
| 3072 |
+
"epoch": 4.0,
|
| 3073 |
+
"step": 4376,
|
| 3074 |
+
"total_flos": 0.0,
|
| 3075 |
+
"train_loss": 0.05302349528062823,
|
| 3076 |
+
"train_runtime": 2534.7991,
|
| 3077 |
+
"train_samples_per_second": 34.507,
|
| 3078 |
+
"train_steps_per_second": 1.726
|
| 3079 |
+
}
|
| 3080 |
+
],
|
| 3081 |
+
"logging_steps": 10,
|
| 3082 |
+
"max_steps": 4376,
|
| 3083 |
+
"num_input_tokens_seen": 0,
|
| 3084 |
+
"num_train_epochs": 4,
|
| 3085 |
+
"save_steps": 2500,
|
| 3086 |
+
"stateful_callbacks": {
|
| 3087 |
+
"TrainerControl": {
|
| 3088 |
+
"args": {
|
| 3089 |
+
"should_epoch_stop": false,
|
| 3090 |
+
"should_evaluate": false,
|
| 3091 |
+
"should_log": false,
|
| 3092 |
+
"should_save": true,
|
| 3093 |
+
"should_training_stop": true
|
| 3094 |
+
},
|
| 3095 |
+
"attributes": {}
|
| 3096 |
+
}
|
| 3097 |
+
},
|
| 3098 |
+
"total_flos": 0.0,
|
| 3099 |
+
"train_batch_size": 20,
|
| 3100 |
+
"trial_name": null,
|
| 3101 |
+
"trial_params": null
|
| 3102 |
+
}
|