Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +2108 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,2108 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 10.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2950,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.03389830508474576,
|
| 14 |
+
"grad_norm": 7.922511100769043,
|
| 15 |
+
"learning_rate": 6.081081081081082e-06,
|
| 16 |
+
"loss": 1.3016,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.06779661016949153,
|
| 21 |
+
"grad_norm": 1.777317762374878,
|
| 22 |
+
"learning_rate": 1.2837837837837838e-05,
|
| 23 |
+
"loss": 0.6145,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.1016949152542373,
|
| 28 |
+
"grad_norm": 2.9913039207458496,
|
| 29 |
+
"learning_rate": 1.9594594594594595e-05,
|
| 30 |
+
"loss": 0.2696,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.13559322033898305,
|
| 35 |
+
"grad_norm": 1.0988717079162598,
|
| 36 |
+
"learning_rate": 2.635135135135135e-05,
|
| 37 |
+
"loss": 0.1885,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.1694915254237288,
|
| 42 |
+
"grad_norm": 1.2467349767684937,
|
| 43 |
+
"learning_rate": 3.310810810810811e-05,
|
| 44 |
+
"loss": 0.1516,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.2033898305084746,
|
| 49 |
+
"grad_norm": 1.4213334321975708,
|
| 50 |
+
"learning_rate": 3.986486486486487e-05,
|
| 51 |
+
"loss": 0.1406,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.23728813559322035,
|
| 56 |
+
"grad_norm": 1.0713083744049072,
|
| 57 |
+
"learning_rate": 4.662162162162162e-05,
|
| 58 |
+
"loss": 0.1316,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.2711864406779661,
|
| 63 |
+
"grad_norm": 1.506921410560608,
|
| 64 |
+
"learning_rate": 5.337837837837838e-05,
|
| 65 |
+
"loss": 0.1413,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.3050847457627119,
|
| 70 |
+
"grad_norm": 0.9623873233795166,
|
| 71 |
+
"learning_rate": 6.013513513513514e-05,
|
| 72 |
+
"loss": 0.1204,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.3389830508474576,
|
| 77 |
+
"grad_norm": 1.5513378381729126,
|
| 78 |
+
"learning_rate": 6.68918918918919e-05,
|
| 79 |
+
"loss": 0.1194,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.3728813559322034,
|
| 84 |
+
"grad_norm": 1.0387235879898071,
|
| 85 |
+
"learning_rate": 7.364864864864865e-05,
|
| 86 |
+
"loss": 0.1111,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.4067796610169492,
|
| 91 |
+
"grad_norm": 1.3981765508651733,
|
| 92 |
+
"learning_rate": 8.040540540540541e-05,
|
| 93 |
+
"loss": 0.119,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.4406779661016949,
|
| 98 |
+
"grad_norm": 1.1575907468795776,
|
| 99 |
+
"learning_rate": 8.716216216216216e-05,
|
| 100 |
+
"loss": 0.0932,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.4745762711864407,
|
| 105 |
+
"grad_norm": 0.8324375748634338,
|
| 106 |
+
"learning_rate": 9.391891891891892e-05,
|
| 107 |
+
"loss": 0.0997,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.5084745762711864,
|
| 112 |
+
"grad_norm": 1.0497596263885498,
|
| 113 |
+
"learning_rate": 9.99999685729623e-05,
|
| 114 |
+
"loss": 0.0795,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.5423728813559322,
|
| 119 |
+
"grad_norm": 0.9868571758270264,
|
| 120 |
+
"learning_rate": 9.999619737623951e-05,
|
| 121 |
+
"loss": 0.0759,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.576271186440678,
|
| 126 |
+
"grad_norm": 0.6950494050979614,
|
| 127 |
+
"learning_rate": 9.99861413151757e-05,
|
| 128 |
+
"loss": 0.0647,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.6101694915254238,
|
| 133 |
+
"grad_norm": 0.9549587965011597,
|
| 134 |
+
"learning_rate": 9.99698016538866e-05,
|
| 135 |
+
"loss": 0.0708,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.6440677966101694,
|
| 140 |
+
"grad_norm": 0.6263039112091064,
|
| 141 |
+
"learning_rate": 9.994718044637956e-05,
|
| 142 |
+
"loss": 0.0711,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.6779661016949152,
|
| 147 |
+
"grad_norm": 0.6637316346168518,
|
| 148 |
+
"learning_rate": 9.991828053629519e-05,
|
| 149 |
+
"loss": 0.0635,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.711864406779661,
|
| 154 |
+
"grad_norm": 0.8776123523712158,
|
| 155 |
+
"learning_rate": 9.988310555655009e-05,
|
| 156 |
+
"loss": 0.0609,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.7457627118644068,
|
| 161 |
+
"grad_norm": 0.9095442891120911,
|
| 162 |
+
"learning_rate": 9.984165992888009e-05,
|
| 163 |
+
"loss": 0.0532,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.7796610169491526,
|
| 168 |
+
"grad_norm": 0.8120779395103455,
|
| 169 |
+
"learning_rate": 9.979394886328434e-05,
|
| 170 |
+
"loss": 0.0522,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.8135593220338984,
|
| 175 |
+
"grad_norm": 0.7972075939178467,
|
| 176 |
+
"learning_rate": 9.973997835737049e-05,
|
| 177 |
+
"loss": 0.0564,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.847457627118644,
|
| 182 |
+
"grad_norm": 0.8452187776565552,
|
| 183 |
+
"learning_rate": 9.967975519560067e-05,
|
| 184 |
+
"loss": 0.0522,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.8813559322033898,
|
| 189 |
+
"grad_norm": 0.8456110954284668,
|
| 190 |
+
"learning_rate": 9.961328694843865e-05,
|
| 191 |
+
"loss": 0.049,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.9152542372881356,
|
| 196 |
+
"grad_norm": 0.47546839714050293,
|
| 197 |
+
"learning_rate": 9.954058197139824e-05,
|
| 198 |
+
"loss": 0.0503,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.9491525423728814,
|
| 203 |
+
"grad_norm": 0.7819595336914062,
|
| 204 |
+
"learning_rate": 9.946164940399288e-05,
|
| 205 |
+
"loss": 0.0515,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.9830508474576272,
|
| 210 |
+
"grad_norm": 0.7775952219963074,
|
| 211 |
+
"learning_rate": 9.937649916858673e-05,
|
| 212 |
+
"loss": 0.05,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 1.0169491525423728,
|
| 217 |
+
"grad_norm": 0.7178511619567871,
|
| 218 |
+
"learning_rate": 9.928514196914744e-05,
|
| 219 |
+
"loss": 0.0529,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 1.0508474576271187,
|
| 224 |
+
"grad_norm": 0.6638413667678833,
|
| 225 |
+
"learning_rate": 9.91875892899005e-05,
|
| 226 |
+
"loss": 0.053,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 1.0847457627118644,
|
| 231 |
+
"grad_norm": 0.6948956251144409,
|
| 232 |
+
"learning_rate": 9.908385339388564e-05,
|
| 233 |
+
"loss": 0.0484,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 1.11864406779661,
|
| 238 |
+
"grad_norm": 0.47976964712142944,
|
| 239 |
+
"learning_rate": 9.897394732141531e-05,
|
| 240 |
+
"loss": 0.0477,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 1.152542372881356,
|
| 245 |
+
"grad_norm": 0.786380410194397,
|
| 246 |
+
"learning_rate": 9.885788488843535e-05,
|
| 247 |
+
"loss": 0.0464,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 1.1864406779661016,
|
| 252 |
+
"grad_norm": 0.7638571858406067,
|
| 253 |
+
"learning_rate": 9.873568068478826e-05,
|
| 254 |
+
"loss": 0.0493,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 1.2203389830508475,
|
| 259 |
+
"grad_norm": 0.8229843974113464,
|
| 260 |
+
"learning_rate": 9.860735007237922e-05,
|
| 261 |
+
"loss": 0.0488,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 1.2542372881355932,
|
| 266 |
+
"grad_norm": 0.6507530808448792,
|
| 267 |
+
"learning_rate": 9.847290918324487e-05,
|
| 268 |
+
"loss": 0.0442,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 1.288135593220339,
|
| 273 |
+
"grad_norm": 0.6550291180610657,
|
| 274 |
+
"learning_rate": 9.833237491752557e-05,
|
| 275 |
+
"loss": 0.0422,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 1.3220338983050848,
|
| 280 |
+
"grad_norm": 0.4493582844734192,
|
| 281 |
+
"learning_rate": 9.818576494134071e-05,
|
| 282 |
+
"loss": 0.0436,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 1.3559322033898304,
|
| 287 |
+
"grad_norm": 0.4366756081581116,
|
| 288 |
+
"learning_rate": 9.803309768456819e-05,
|
| 289 |
+
"loss": 0.0483,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 1.3898305084745763,
|
| 294 |
+
"grad_norm": 0.518522322177887,
|
| 295 |
+
"learning_rate": 9.787439233852752e-05,
|
| 296 |
+
"loss": 0.04,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 1.423728813559322,
|
| 301 |
+
"grad_norm": 0.5709708333015442,
|
| 302 |
+
"learning_rate": 9.770966885356736e-05,
|
| 303 |
+
"loss": 0.0389,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 1.457627118644068,
|
| 308 |
+
"grad_norm": 0.44815006852149963,
|
| 309 |
+
"learning_rate": 9.753894793655767e-05,
|
| 310 |
+
"loss": 0.0414,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 1.4915254237288136,
|
| 315 |
+
"grad_norm": 0.6205833554267883,
|
| 316 |
+
"learning_rate": 9.736225104828666e-05,
|
| 317 |
+
"loss": 0.0488,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 1.5254237288135593,
|
| 322 |
+
"grad_norm": 0.38849759101867676,
|
| 323 |
+
"learning_rate": 9.717960040076305e-05,
|
| 324 |
+
"loss": 0.0419,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 1.559322033898305,
|
| 329 |
+
"grad_norm": 0.3880470395088196,
|
| 330 |
+
"learning_rate": 9.699101895442388e-05,
|
| 331 |
+
"loss": 0.0438,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 1.5932203389830508,
|
| 336 |
+
"grad_norm": 0.7112048268318176,
|
| 337 |
+
"learning_rate": 9.679653041524818e-05,
|
| 338 |
+
"loss": 0.0424,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 1.6271186440677967,
|
| 343 |
+
"grad_norm": 0.8996948003768921,
|
| 344 |
+
"learning_rate": 9.659615923177702e-05,
|
| 345 |
+
"loss": 0.0441,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 1.6610169491525424,
|
| 350 |
+
"grad_norm": 0.6175360083580017,
|
| 351 |
+
"learning_rate": 9.638993059204016e-05,
|
| 352 |
+
"loss": 0.0429,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 1.694915254237288,
|
| 357 |
+
"grad_norm": 0.39414605498313904,
|
| 358 |
+
"learning_rate": 9.617787042038964e-05,
|
| 359 |
+
"loss": 0.0377,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 1.7288135593220337,
|
| 364 |
+
"grad_norm": 0.5390213131904602,
|
| 365 |
+
"learning_rate": 9.596000537424107e-05,
|
| 366 |
+
"loss": 0.0407,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 1.7627118644067796,
|
| 371 |
+
"grad_norm": 0.4397003948688507,
|
| 372 |
+
"learning_rate": 9.573636284072252e-05,
|
| 373 |
+
"loss": 0.0386,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 1.7966101694915255,
|
| 378 |
+
"grad_norm": 0.6200686097145081,
|
| 379 |
+
"learning_rate": 9.550697093323173e-05,
|
| 380 |
+
"loss": 0.0392,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 1.8305084745762712,
|
| 385 |
+
"grad_norm": 0.5365989804267883,
|
| 386 |
+
"learning_rate": 9.527185848790216e-05,
|
| 387 |
+
"loss": 0.0355,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 1.8644067796610169,
|
| 392 |
+
"grad_norm": 0.4744288921356201,
|
| 393 |
+
"learning_rate": 9.503105505997805e-05,
|
| 394 |
+
"loss": 0.0387,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.8983050847457628,
|
| 399 |
+
"grad_norm": 0.5821052193641663,
|
| 400 |
+
"learning_rate": 9.478459092009909e-05,
|
| 401 |
+
"loss": 0.0392,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 1.9322033898305084,
|
| 406 |
+
"grad_norm": 0.5957028269767761,
|
| 407 |
+
"learning_rate": 9.453249705049523e-05,
|
| 408 |
+
"loss": 0.0392,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 1.9661016949152543,
|
| 413 |
+
"grad_norm": 0.54462730884552,
|
| 414 |
+
"learning_rate": 9.427480514109202e-05,
|
| 415 |
+
"loss": 0.0355,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 2.0,
|
| 420 |
+
"grad_norm": 1.2646660804748535,
|
| 421 |
+
"learning_rate": 9.401154758552691e-05,
|
| 422 |
+
"loss": 0.045,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 2.0338983050847457,
|
| 427 |
+
"grad_norm": 0.4895438551902771,
|
| 428 |
+
"learning_rate": 9.37427574770772e-05,
|
| 429 |
+
"loss": 0.0394,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 2.0677966101694913,
|
| 434 |
+
"grad_norm": 0.7774738669395447,
|
| 435 |
+
"learning_rate": 9.346846860449996e-05,
|
| 436 |
+
"loss": 0.0412,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 2.1016949152542375,
|
| 441 |
+
"grad_norm": 0.7907915115356445,
|
| 442 |
+
"learning_rate": 9.318871544778457e-05,
|
| 443 |
+
"loss": 0.0434,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 2.135593220338983,
|
| 448 |
+
"grad_norm": 0.5634192228317261,
|
| 449 |
+
"learning_rate": 9.290353317381834e-05,
|
| 450 |
+
"loss": 0.0445,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 2.169491525423729,
|
| 455 |
+
"grad_norm": 0.524152398109436,
|
| 456 |
+
"learning_rate": 9.261295763196579e-05,
|
| 457 |
+
"loss": 0.0347,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 2.2033898305084745,
|
| 462 |
+
"grad_norm": 0.48940563201904297,
|
| 463 |
+
"learning_rate": 9.231702534956213e-05,
|
| 464 |
+
"loss": 0.0387,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 2.23728813559322,
|
| 469 |
+
"grad_norm": 0.4866064488887787,
|
| 470 |
+
"learning_rate": 9.201577352732157e-05,
|
| 471 |
+
"loss": 0.0386,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 2.2711864406779663,
|
| 476 |
+
"grad_norm": 0.379842609167099,
|
| 477 |
+
"learning_rate": 9.170924003466086e-05,
|
| 478 |
+
"loss": 0.0341,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 2.305084745762712,
|
| 483 |
+
"grad_norm": 0.41267237067222595,
|
| 484 |
+
"learning_rate": 9.13974634049389e-05,
|
| 485 |
+
"loss": 0.0379,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 2.3389830508474576,
|
| 490 |
+
"grad_norm": 0.6153244376182556,
|
| 491 |
+
"learning_rate": 9.108048283061283e-05,
|
| 492 |
+
"loss": 0.0331,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 2.3728813559322033,
|
| 497 |
+
"grad_norm": 0.48824194073677063,
|
| 498 |
+
"learning_rate": 9.075833815831128e-05,
|
| 499 |
+
"loss": 0.0448,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 2.406779661016949,
|
| 504 |
+
"grad_norm": 0.47475725412368774,
|
| 505 |
+
"learning_rate": 9.043106988382526e-05,
|
| 506 |
+
"loss": 0.0373,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 2.440677966101695,
|
| 511 |
+
"grad_norm": 0.44530612230300903,
|
| 512 |
+
"learning_rate": 9.009871914701779e-05,
|
| 513 |
+
"loss": 0.036,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 2.4745762711864407,
|
| 518 |
+
"grad_norm": 0.704857349395752,
|
| 519 |
+
"learning_rate": 8.976132772665211e-05,
|
| 520 |
+
"loss": 0.0405,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 2.5084745762711864,
|
| 525 |
+
"grad_norm": 0.5021880269050598,
|
| 526 |
+
"learning_rate": 8.941893803513994e-05,
|
| 527 |
+
"loss": 0.039,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 2.542372881355932,
|
| 532 |
+
"grad_norm": 0.43946319818496704,
|
| 533 |
+
"learning_rate": 8.907159311320985e-05,
|
| 534 |
+
"loss": 0.0377,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 2.576271186440678,
|
| 539 |
+
"grad_norm": 0.4081028997898102,
|
| 540 |
+
"learning_rate": 8.87193366244969e-05,
|
| 541 |
+
"loss": 0.0337,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 2.610169491525424,
|
| 546 |
+
"grad_norm": 0.42831340432167053,
|
| 547 |
+
"learning_rate": 8.836221285005364e-05,
|
| 548 |
+
"loss": 0.0331,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 2.6440677966101696,
|
| 553 |
+
"grad_norm": 0.5115790963172913,
|
| 554 |
+
"learning_rate": 8.800026668278378e-05,
|
| 555 |
+
"loss": 0.0383,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 2.6779661016949152,
|
| 560 |
+
"grad_norm": 0.7489039897918701,
|
| 561 |
+
"learning_rate": 8.763354362179887e-05,
|
| 562 |
+
"loss": 0.037,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 2.711864406779661,
|
| 567 |
+
"grad_norm": 0.47204285860061646,
|
| 568 |
+
"learning_rate": 8.726208976669868e-05,
|
| 569 |
+
"loss": 0.0366,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 2.7457627118644066,
|
| 574 |
+
"grad_norm": 0.3575058579444885,
|
| 575 |
+
"learning_rate": 8.688595181177617e-05,
|
| 576 |
+
"loss": 0.03,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 2.7796610169491527,
|
| 581 |
+
"grad_norm": 0.3760571777820587,
|
| 582 |
+
"learning_rate": 8.650517704014773e-05,
|
| 583 |
+
"loss": 0.0376,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 2.8135593220338984,
|
| 588 |
+
"grad_norm": 0.3222888708114624,
|
| 589 |
+
"learning_rate": 8.611981331780942e-05,
|
| 590 |
+
"loss": 0.0373,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 2.847457627118644,
|
| 595 |
+
"grad_norm": 0.3633357286453247,
|
| 596 |
+
"learning_rate": 8.572990908761974e-05,
|
| 597 |
+
"loss": 0.0358,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 2.8813559322033897,
|
| 602 |
+
"grad_norm": 0.3399844765663147,
|
| 603 |
+
"learning_rate": 8.533551336321015e-05,
|
| 604 |
+
"loss": 0.0359,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 2.915254237288136,
|
| 609 |
+
"grad_norm": 0.7424901127815247,
|
| 610 |
+
"learning_rate": 8.493667572282371e-05,
|
| 611 |
+
"loss": 0.0338,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 2.9491525423728815,
|
| 616 |
+
"grad_norm": 0.4179316759109497,
|
| 617 |
+
"learning_rate": 8.453344630308269e-05,
|
| 618 |
+
"loss": 0.0358,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 2.983050847457627,
|
| 623 |
+
"grad_norm": 0.4843442738056183,
|
| 624 |
+
"learning_rate": 8.412587579268613e-05,
|
| 625 |
+
"loss": 0.0353,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 3.016949152542373,
|
| 630 |
+
"grad_norm": 0.5782058835029602,
|
| 631 |
+
"learning_rate": 8.371401542603791e-05,
|
| 632 |
+
"loss": 0.0373,
|
| 633 |
+
"step": 890
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 3.0508474576271185,
|
| 637 |
+
"grad_norm": 0.5099455714225769,
|
| 638 |
+
"learning_rate": 8.329791697680612e-05,
|
| 639 |
+
"loss": 0.0336,
|
| 640 |
+
"step": 900
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 3.084745762711864,
|
| 644 |
+
"grad_norm": 0.31090185046195984,
|
| 645 |
+
"learning_rate": 8.287763275141496e-05,
|
| 646 |
+
"loss": 0.0326,
|
| 647 |
+
"step": 910
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 3.1186440677966103,
|
| 651 |
+
"grad_norm": 0.22995054721832275,
|
| 652 |
+
"learning_rate": 8.245321558246928e-05,
|
| 653 |
+
"loss": 0.0342,
|
| 654 |
+
"step": 920
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 3.152542372881356,
|
| 658 |
+
"grad_norm": 0.46085304021835327,
|
| 659 |
+
"learning_rate": 8.202471882211327e-05,
|
| 660 |
+
"loss": 0.0314,
|
| 661 |
+
"step": 930
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 3.1864406779661016,
|
| 665 |
+
"grad_norm": 0.47579386830329895,
|
| 666 |
+
"learning_rate": 8.159219633532368e-05,
|
| 667 |
+
"loss": 0.0335,
|
| 668 |
+
"step": 940
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 3.2203389830508473,
|
| 672 |
+
"grad_norm": 0.3809256851673126,
|
| 673 |
+
"learning_rate": 8.115570249313864e-05,
|
| 674 |
+
"loss": 0.0322,
|
| 675 |
+
"step": 950
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 3.2542372881355934,
|
| 679 |
+
"grad_norm": 0.38978174328804016,
|
| 680 |
+
"learning_rate": 8.071529216582284e-05,
|
| 681 |
+
"loss": 0.0313,
|
| 682 |
+
"step": 960
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 3.288135593220339,
|
| 686 |
+
"grad_norm": 0.3661576211452484,
|
| 687 |
+
"learning_rate": 8.027102071596999e-05,
|
| 688 |
+
"loss": 0.0378,
|
| 689 |
+
"step": 970
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 3.3220338983050848,
|
| 693 |
+
"grad_norm": 0.28337371349334717,
|
| 694 |
+
"learning_rate": 7.982294399154331e-05,
|
| 695 |
+
"loss": 0.0338,
|
| 696 |
+
"step": 980
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 3.3559322033898304,
|
| 700 |
+
"grad_norm": 0.48508763313293457,
|
| 701 |
+
"learning_rate": 7.93711183188551e-05,
|
| 702 |
+
"loss": 0.0374,
|
| 703 |
+
"step": 990
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 3.389830508474576,
|
| 707 |
+
"grad_norm": 0.4197925925254822,
|
| 708 |
+
"learning_rate": 7.891560049548614e-05,
|
| 709 |
+
"loss": 0.0336,
|
| 710 |
+
"step": 1000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 3.423728813559322,
|
| 714 |
+
"grad_norm": 0.4868261516094208,
|
| 715 |
+
"learning_rate": 7.845644778314583e-05,
|
| 716 |
+
"loss": 0.0396,
|
| 717 |
+
"step": 1010
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 3.457627118644068,
|
| 721 |
+
"grad_norm": 0.264763206243515,
|
| 722 |
+
"learning_rate": 7.799371790047395e-05,
|
| 723 |
+
"loss": 0.0317,
|
| 724 |
+
"step": 1020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 3.4915254237288136,
|
| 728 |
+
"grad_norm": 0.2507479190826416,
|
| 729 |
+
"learning_rate": 7.752746901578517e-05,
|
| 730 |
+
"loss": 0.0308,
|
| 731 |
+
"step": 1030
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 3.5254237288135593,
|
| 735 |
+
"grad_norm": 0.40973591804504395,
|
| 736 |
+
"learning_rate": 7.705775973975675e-05,
|
| 737 |
+
"loss": 0.035,
|
| 738 |
+
"step": 1040
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 3.559322033898305,
|
| 742 |
+
"grad_norm": 0.34160512685775757,
|
| 743 |
+
"learning_rate": 7.658464911806087e-05,
|
| 744 |
+
"loss": 0.0287,
|
| 745 |
+
"step": 1050
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 3.593220338983051,
|
| 749 |
+
"grad_norm": 0.3527885973453522,
|
| 750 |
+
"learning_rate": 7.61081966239421e-05,
|
| 751 |
+
"loss": 0.026,
|
| 752 |
+
"step": 1060
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 3.6271186440677967,
|
| 756 |
+
"grad_norm": 0.4634409248828888,
|
| 757 |
+
"learning_rate": 7.562846215074129e-05,
|
| 758 |
+
"loss": 0.03,
|
| 759 |
+
"step": 1070
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 3.6610169491525424,
|
| 763 |
+
"grad_norm": 0.3833574652671814,
|
| 764 |
+
"learning_rate": 7.514550600436657e-05,
|
| 765 |
+
"loss": 0.032,
|
| 766 |
+
"step": 1080
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 3.694915254237288,
|
| 770 |
+
"grad_norm": 0.4073950946331024,
|
| 771 |
+
"learning_rate": 7.465938889571232e-05,
|
| 772 |
+
"loss": 0.0268,
|
| 773 |
+
"step": 1090
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 3.7288135593220337,
|
| 777 |
+
"grad_norm": 0.590704083442688,
|
| 778 |
+
"learning_rate": 7.417017193302769e-05,
|
| 779 |
+
"loss": 0.0298,
|
| 780 |
+
"step": 1100
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 3.7627118644067794,
|
| 784 |
+
"grad_norm": 0.3835419714450836,
|
| 785 |
+
"learning_rate": 7.367791661423461e-05,
|
| 786 |
+
"loss": 0.0305,
|
| 787 |
+
"step": 1110
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 3.7966101694915255,
|
| 791 |
+
"grad_norm": 0.4307284951210022,
|
| 792 |
+
"learning_rate": 7.318268481919716e-05,
|
| 793 |
+
"loss": 0.0312,
|
| 794 |
+
"step": 1120
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 3.830508474576271,
|
| 798 |
+
"grad_norm": 0.5295498371124268,
|
| 799 |
+
"learning_rate": 7.268453880194292e-05,
|
| 800 |
+
"loss": 0.0329,
|
| 801 |
+
"step": 1130
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 3.864406779661017,
|
| 805 |
+
"grad_norm": 0.5060334801673889,
|
| 806 |
+
"learning_rate": 7.218354118283715e-05,
|
| 807 |
+
"loss": 0.0346,
|
| 808 |
+
"step": 1140
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 3.898305084745763,
|
| 812 |
+
"grad_norm": 0.3661726415157318,
|
| 813 |
+
"learning_rate": 7.167975494071093e-05,
|
| 814 |
+
"loss": 0.034,
|
| 815 |
+
"step": 1150
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 3.9322033898305087,
|
| 819 |
+
"grad_norm": 0.382707804441452,
|
| 820 |
+
"learning_rate": 7.11732434049444e-05,
|
| 821 |
+
"loss": 0.0314,
|
| 822 |
+
"step": 1160
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 3.9661016949152543,
|
| 826 |
+
"grad_norm": 0.44602781534194946,
|
| 827 |
+
"learning_rate": 7.066407024750578e-05,
|
| 828 |
+
"loss": 0.0346,
|
| 829 |
+
"step": 1170
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 4.0,
|
| 833 |
+
"grad_norm": 0.8751994371414185,
|
| 834 |
+
"learning_rate": 7.015229947494734e-05,
|
| 835 |
+
"loss": 0.0321,
|
| 836 |
+
"step": 1180
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 4.033898305084746,
|
| 840 |
+
"grad_norm": 0.31756699085235596,
|
| 841 |
+
"learning_rate": 6.963799542035932e-05,
|
| 842 |
+
"loss": 0.0319,
|
| 843 |
+
"step": 1190
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 4.067796610169491,
|
| 847 |
+
"grad_norm": 0.336243599653244,
|
| 848 |
+
"learning_rate": 6.912122273528291e-05,
|
| 849 |
+
"loss": 0.0296,
|
| 850 |
+
"step": 1200
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 4.101694915254237,
|
| 854 |
+
"grad_norm": 0.3985385298728943,
|
| 855 |
+
"learning_rate": 6.860204638158305e-05,
|
| 856 |
+
"loss": 0.028,
|
| 857 |
+
"step": 1210
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 4.135593220338983,
|
| 861 |
+
"grad_norm": 0.5406593084335327,
|
| 862 |
+
"learning_rate": 6.808053162328227e-05,
|
| 863 |
+
"loss": 0.034,
|
| 864 |
+
"step": 1220
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 4.169491525423728,
|
| 868 |
+
"grad_norm": 0.3683408200740814,
|
| 869 |
+
"learning_rate": 6.755674401835657e-05,
|
| 870 |
+
"loss": 0.0312,
|
| 871 |
+
"step": 1230
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 4.203389830508475,
|
| 875 |
+
"grad_norm": 0.5468569993972778,
|
| 876 |
+
"learning_rate": 6.703074941049431e-05,
|
| 877 |
+
"loss": 0.0286,
|
| 878 |
+
"step": 1240
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 4.237288135593221,
|
| 882 |
+
"grad_norm": 0.5490782260894775,
|
| 883 |
+
"learning_rate": 6.650261392081929e-05,
|
| 884 |
+
"loss": 0.0322,
|
| 885 |
+
"step": 1250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 4.271186440677966,
|
| 889 |
+
"grad_norm": 0.2997434437274933,
|
| 890 |
+
"learning_rate": 6.597240393957875e-05,
|
| 891 |
+
"loss": 0.0299,
|
| 892 |
+
"step": 1260
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 4.305084745762712,
|
| 896 |
+
"grad_norm": 0.43356242775917053,
|
| 897 |
+
"learning_rate": 6.544018611779778e-05,
|
| 898 |
+
"loss": 0.0331,
|
| 899 |
+
"step": 1270
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 4.338983050847458,
|
| 903 |
+
"grad_norm": 0.33595308661460876,
|
| 904 |
+
"learning_rate": 6.490602735890071e-05,
|
| 905 |
+
"loss": 0.0275,
|
| 906 |
+
"step": 1280
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 4.372881355932203,
|
| 910 |
+
"grad_norm": 0.4883277714252472,
|
| 911 |
+
"learning_rate": 6.436999481030104e-05,
|
| 912 |
+
"loss": 0.0321,
|
| 913 |
+
"step": 1290
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 4.406779661016949,
|
| 917 |
+
"grad_norm": 0.25386956334114075,
|
| 918 |
+
"learning_rate": 6.383215585496037e-05,
|
| 919 |
+
"loss": 0.0322,
|
| 920 |
+
"step": 1300
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 4.440677966101695,
|
| 924 |
+
"grad_norm": 0.30296579003334045,
|
| 925 |
+
"learning_rate": 6.329257810291806e-05,
|
| 926 |
+
"loss": 0.023,
|
| 927 |
+
"step": 1310
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 4.47457627118644,
|
| 931 |
+
"grad_norm": 0.47214221954345703,
|
| 932 |
+
"learning_rate": 6.275132938279213e-05,
|
| 933 |
+
"loss": 0.0248,
|
| 934 |
+
"step": 1320
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 4.508474576271187,
|
| 938 |
+
"grad_norm": 0.40692755579948425,
|
| 939 |
+
"learning_rate": 6.220847773325268e-05,
|
| 940 |
+
"loss": 0.0325,
|
| 941 |
+
"step": 1330
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 4.5423728813559325,
|
| 945 |
+
"grad_norm": 0.26799410581588745,
|
| 946 |
+
"learning_rate": 6.166409139446903e-05,
|
| 947 |
+
"loss": 0.0305,
|
| 948 |
+
"step": 1340
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 4.576271186440678,
|
| 952 |
+
"grad_norm": 0.2558389902114868,
|
| 953 |
+
"learning_rate": 6.111823879953145e-05,
|
| 954 |
+
"loss": 0.0265,
|
| 955 |
+
"step": 1350
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 4.610169491525424,
|
| 959 |
+
"grad_norm": 0.28272953629493713,
|
| 960 |
+
"learning_rate": 6.0570988565848685e-05,
|
| 961 |
+
"loss": 0.025,
|
| 962 |
+
"step": 1360
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 4.6440677966101696,
|
| 966 |
+
"grad_norm": 0.4986218214035034,
|
| 967 |
+
"learning_rate": 6.002240948652218e-05,
|
| 968 |
+
"loss": 0.0251,
|
| 969 |
+
"step": 1370
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 4.677966101694915,
|
| 973 |
+
"grad_norm": 0.5608692169189453,
|
| 974 |
+
"learning_rate": 5.94725705216984e-05,
|
| 975 |
+
"loss": 0.0271,
|
| 976 |
+
"step": 1380
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 4.711864406779661,
|
| 980 |
+
"grad_norm": 0.5304986834526062,
|
| 981 |
+
"learning_rate": 5.8921540789900064e-05,
|
| 982 |
+
"loss": 0.0268,
|
| 983 |
+
"step": 1390
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 4.745762711864407,
|
| 987 |
+
"grad_norm": 0.45950591564178467,
|
| 988 |
+
"learning_rate": 5.836938955933743e-05,
|
| 989 |
+
"loss": 0.0272,
|
| 990 |
+
"step": 1400
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 4.779661016949152,
|
| 994 |
+
"grad_norm": 0.31691136956214905,
|
| 995 |
+
"learning_rate": 5.781618623920081e-05,
|
| 996 |
+
"loss": 0.0312,
|
| 997 |
+
"step": 1410
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 4.813559322033898,
|
| 1001 |
+
"grad_norm": 0.33825352787971497,
|
| 1002 |
+
"learning_rate": 5.726200037093542e-05,
|
| 1003 |
+
"loss": 0.0289,
|
| 1004 |
+
"step": 1420
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 4.847457627118644,
|
| 1008 |
+
"grad_norm": 0.31809505820274353,
|
| 1009 |
+
"learning_rate": 5.670690161949952e-05,
|
| 1010 |
+
"loss": 0.0269,
|
| 1011 |
+
"step": 1430
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 4.88135593220339,
|
| 1015 |
+
"grad_norm": 0.27987194061279297,
|
| 1016 |
+
"learning_rate": 5.6150959764606944e-05,
|
| 1017 |
+
"loss": 0.026,
|
| 1018 |
+
"step": 1440
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 4.915254237288136,
|
| 1022 |
+
"grad_norm": 0.33458834886550903,
|
| 1023 |
+
"learning_rate": 5.559424469195551e-05,
|
| 1024 |
+
"loss": 0.0256,
|
| 1025 |
+
"step": 1450
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 4.9491525423728815,
|
| 1029 |
+
"grad_norm": 0.5636442303657532,
|
| 1030 |
+
"learning_rate": 5.503682638444173e-05,
|
| 1031 |
+
"loss": 0.0279,
|
| 1032 |
+
"step": 1460
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 4.983050847457627,
|
| 1036 |
+
"grad_norm": 0.400234192609787,
|
| 1037 |
+
"learning_rate": 5.4478774913363596e-05,
|
| 1038 |
+
"loss": 0.0273,
|
| 1039 |
+
"step": 1470
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 5.016949152542373,
|
| 1043 |
+
"grad_norm": 0.5593984127044678,
|
| 1044 |
+
"learning_rate": 5.392016042961205e-05,
|
| 1045 |
+
"loss": 0.0299,
|
| 1046 |
+
"step": 1480
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 5.0508474576271185,
|
| 1050 |
+
"grad_norm": 0.42886337637901306,
|
| 1051 |
+
"learning_rate": 5.336105315485264e-05,
|
| 1052 |
+
"loss": 0.0282,
|
| 1053 |
+
"step": 1490
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 5.084745762711864,
|
| 1057 |
+
"grad_norm": 0.4065008759498596,
|
| 1058 |
+
"learning_rate": 5.280152337269807e-05,
|
| 1059 |
+
"loss": 0.0282,
|
| 1060 |
+
"step": 1500
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 5.11864406779661,
|
| 1064 |
+
"grad_norm": 0.46473270654678345,
|
| 1065 |
+
"learning_rate": 5.224164141987313e-05,
|
| 1066 |
+
"loss": 0.027,
|
| 1067 |
+
"step": 1510
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 5.1525423728813555,
|
| 1071 |
+
"grad_norm": 0.35750046372413635,
|
| 1072 |
+
"learning_rate": 5.168147767737289e-05,
|
| 1073 |
+
"loss": 0.0288,
|
| 1074 |
+
"step": 1520
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 5.186440677966102,
|
| 1078 |
+
"grad_norm": 0.3990558385848999,
|
| 1079 |
+
"learning_rate": 5.1121102561615295e-05,
|
| 1080 |
+
"loss": 0.0294,
|
| 1081 |
+
"step": 1530
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 5.220338983050848,
|
| 1085 |
+
"grad_norm": 0.32793498039245605,
|
| 1086 |
+
"learning_rate": 5.056058651558936e-05,
|
| 1087 |
+
"loss": 0.028,
|
| 1088 |
+
"step": 1540
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 5.254237288135593,
|
| 1092 |
+
"grad_norm": 0.6837785243988037,
|
| 1093 |
+
"learning_rate": 5e-05,
|
| 1094 |
+
"loss": 0.0256,
|
| 1095 |
+
"step": 1550
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 5.288135593220339,
|
| 1099 |
+
"grad_norm": 0.49780961871147156,
|
| 1100 |
+
"learning_rate": 4.943941348441065e-05,
|
| 1101 |
+
"loss": 0.0258,
|
| 1102 |
+
"step": 1560
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 5.322033898305085,
|
| 1106 |
+
"grad_norm": 0.529035210609436,
|
| 1107 |
+
"learning_rate": 4.8878897438384716e-05,
|
| 1108 |
+
"loss": 0.0281,
|
| 1109 |
+
"step": 1570
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 5.3559322033898304,
|
| 1113 |
+
"grad_norm": 0.27274975180625916,
|
| 1114 |
+
"learning_rate": 4.831852232262713e-05,
|
| 1115 |
+
"loss": 0.0233,
|
| 1116 |
+
"step": 1580
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 5.389830508474576,
|
| 1120 |
+
"grad_norm": 0.5764862895011902,
|
| 1121 |
+
"learning_rate": 4.775835858012689e-05,
|
| 1122 |
+
"loss": 0.0235,
|
| 1123 |
+
"step": 1590
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 5.423728813559322,
|
| 1127 |
+
"grad_norm": 0.3733164370059967,
|
| 1128 |
+
"learning_rate": 4.7198476627301955e-05,
|
| 1129 |
+
"loss": 0.0269,
|
| 1130 |
+
"step": 1600
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 5.4576271186440675,
|
| 1134 |
+
"grad_norm": 0.4069574177265167,
|
| 1135 |
+
"learning_rate": 4.663894684514737e-05,
|
| 1136 |
+
"loss": 0.0267,
|
| 1137 |
+
"step": 1610
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 5.491525423728813,
|
| 1141 |
+
"grad_norm": 0.37277278304100037,
|
| 1142 |
+
"learning_rate": 4.6079839570387954e-05,
|
| 1143 |
+
"loss": 0.0239,
|
| 1144 |
+
"step": 1620
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 5.52542372881356,
|
| 1148 |
+
"grad_norm": 0.43740156292915344,
|
| 1149 |
+
"learning_rate": 4.552122508663641e-05,
|
| 1150 |
+
"loss": 0.0218,
|
| 1151 |
+
"step": 1630
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 5.559322033898305,
|
| 1155 |
+
"grad_norm": 0.4995250105857849,
|
| 1156 |
+
"learning_rate": 4.496317361555828e-05,
|
| 1157 |
+
"loss": 0.0275,
|
| 1158 |
+
"step": 1640
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 5.593220338983051,
|
| 1162 |
+
"grad_norm": 0.3874273896217346,
|
| 1163 |
+
"learning_rate": 4.44057553080445e-05,
|
| 1164 |
+
"loss": 0.0218,
|
| 1165 |
+
"step": 1650
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 5.627118644067797,
|
| 1169 |
+
"grad_norm": 0.3819812536239624,
|
| 1170 |
+
"learning_rate": 4.384904023539306e-05,
|
| 1171 |
+
"loss": 0.0217,
|
| 1172 |
+
"step": 1660
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 5.661016949152542,
|
| 1176 |
+
"grad_norm": 0.44215914607048035,
|
| 1177 |
+
"learning_rate": 4.32930983805005e-05,
|
| 1178 |
+
"loss": 0.028,
|
| 1179 |
+
"step": 1670
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 5.694915254237288,
|
| 1183 |
+
"grad_norm": 0.25850358605384827,
|
| 1184 |
+
"learning_rate": 4.273799962906459e-05,
|
| 1185 |
+
"loss": 0.0255,
|
| 1186 |
+
"step": 1680
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 5.728813559322034,
|
| 1190 |
+
"grad_norm": 0.3987925052642822,
|
| 1191 |
+
"learning_rate": 4.218381376079921e-05,
|
| 1192 |
+
"loss": 0.0267,
|
| 1193 |
+
"step": 1690
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 5.762711864406779,
|
| 1197 |
+
"grad_norm": 0.25257599353790283,
|
| 1198 |
+
"learning_rate": 4.16306104406626e-05,
|
| 1199 |
+
"loss": 0.0244,
|
| 1200 |
+
"step": 1700
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 5.796610169491525,
|
| 1204 |
+
"grad_norm": 0.33424732089042664,
|
| 1205 |
+
"learning_rate": 4.1078459210099954e-05,
|
| 1206 |
+
"loss": 0.0254,
|
| 1207 |
+
"step": 1710
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 5.830508474576272,
|
| 1211 |
+
"grad_norm": 0.22258403897285461,
|
| 1212 |
+
"learning_rate": 4.0527429478301596e-05,
|
| 1213 |
+
"loss": 0.0236,
|
| 1214 |
+
"step": 1720
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 5.864406779661017,
|
| 1218 |
+
"grad_norm": 0.34541794657707214,
|
| 1219 |
+
"learning_rate": 3.997759051347782e-05,
|
| 1220 |
+
"loss": 0.0206,
|
| 1221 |
+
"step": 1730
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 5.898305084745763,
|
| 1225 |
+
"grad_norm": 0.38338717818260193,
|
| 1226 |
+
"learning_rate": 3.9429011434151306e-05,
|
| 1227 |
+
"loss": 0.031,
|
| 1228 |
+
"step": 1740
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 5.932203389830509,
|
| 1232 |
+
"grad_norm": 0.5410785675048828,
|
| 1233 |
+
"learning_rate": 3.8881761200468546e-05,
|
| 1234 |
+
"loss": 0.0269,
|
| 1235 |
+
"step": 1750
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 5.966101694915254,
|
| 1239 |
+
"grad_norm": 0.3703366219997406,
|
| 1240 |
+
"learning_rate": 3.833590860553098e-05,
|
| 1241 |
+
"loss": 0.0265,
|
| 1242 |
+
"step": 1760
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 6.0,
|
| 1246 |
+
"grad_norm": 1.3290477991104126,
|
| 1247 |
+
"learning_rate": 3.779152226674733e-05,
|
| 1248 |
+
"loss": 0.0231,
|
| 1249 |
+
"step": 1770
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 6.033898305084746,
|
| 1253 |
+
"grad_norm": 0.3305765986442566,
|
| 1254 |
+
"learning_rate": 3.724867061720787e-05,
|
| 1255 |
+
"loss": 0.0237,
|
| 1256 |
+
"step": 1780
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 6.067796610169491,
|
| 1260 |
+
"grad_norm": 0.2830105125904083,
|
| 1261 |
+
"learning_rate": 3.6707421897081947e-05,
|
| 1262 |
+
"loss": 0.0253,
|
| 1263 |
+
"step": 1790
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 6.101694915254237,
|
| 1267 |
+
"grad_norm": 0.4226081073284149,
|
| 1268 |
+
"learning_rate": 3.6167844145039644e-05,
|
| 1269 |
+
"loss": 0.021,
|
| 1270 |
+
"step": 1800
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 6.135593220338983,
|
| 1274 |
+
"grad_norm": 0.3365839123725891,
|
| 1275 |
+
"learning_rate": 3.563000518969898e-05,
|
| 1276 |
+
"loss": 0.0236,
|
| 1277 |
+
"step": 1810
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 6.169491525423728,
|
| 1281 |
+
"grad_norm": 0.3108862042427063,
|
| 1282 |
+
"learning_rate": 3.509397264109929e-05,
|
| 1283 |
+
"loss": 0.0214,
|
| 1284 |
+
"step": 1820
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 6.203389830508475,
|
| 1288 |
+
"grad_norm": 0.22683779895305634,
|
| 1289 |
+
"learning_rate": 3.455981388220223e-05,
|
| 1290 |
+
"loss": 0.0194,
|
| 1291 |
+
"step": 1830
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 6.237288135593221,
|
| 1295 |
+
"grad_norm": 0.38497164845466614,
|
| 1296 |
+
"learning_rate": 3.4027596060421254e-05,
|
| 1297 |
+
"loss": 0.0241,
|
| 1298 |
+
"step": 1840
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 6.271186440677966,
|
| 1302 |
+
"grad_norm": 0.22206397354602814,
|
| 1303 |
+
"learning_rate": 3.3497386079180725e-05,
|
| 1304 |
+
"loss": 0.019,
|
| 1305 |
+
"step": 1850
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 6.305084745762712,
|
| 1309 |
+
"grad_norm": 0.3224073648452759,
|
| 1310 |
+
"learning_rate": 3.29692505895057e-05,
|
| 1311 |
+
"loss": 0.0231,
|
| 1312 |
+
"step": 1860
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 6.338983050847458,
|
| 1316 |
+
"grad_norm": 0.28747010231018066,
|
| 1317 |
+
"learning_rate": 3.244325598164344e-05,
|
| 1318 |
+
"loss": 0.0206,
|
| 1319 |
+
"step": 1870
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 6.372881355932203,
|
| 1323 |
+
"grad_norm": 0.3315770924091339,
|
| 1324 |
+
"learning_rate": 3.191946837671773e-05,
|
| 1325 |
+
"loss": 0.0192,
|
| 1326 |
+
"step": 1880
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 6.406779661016949,
|
| 1330 |
+
"grad_norm": 0.38061845302581787,
|
| 1331 |
+
"learning_rate": 3.139795361841696e-05,
|
| 1332 |
+
"loss": 0.0179,
|
| 1333 |
+
"step": 1890
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 6.440677966101695,
|
| 1337 |
+
"grad_norm": 0.2752900719642639,
|
| 1338 |
+
"learning_rate": 3.08787772647171e-05,
|
| 1339 |
+
"loss": 0.0217,
|
| 1340 |
+
"step": 1900
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 6.47457627118644,
|
| 1344 |
+
"grad_norm": 0.270625501871109,
|
| 1345 |
+
"learning_rate": 3.03620045796407e-05,
|
| 1346 |
+
"loss": 0.0246,
|
| 1347 |
+
"step": 1910
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 6.508474576271187,
|
| 1351 |
+
"grad_norm": 0.2529767155647278,
|
| 1352 |
+
"learning_rate": 2.9847700525052687e-05,
|
| 1353 |
+
"loss": 0.0192,
|
| 1354 |
+
"step": 1920
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 6.5423728813559325,
|
| 1358 |
+
"grad_norm": 0.25423118472099304,
|
| 1359 |
+
"learning_rate": 2.933592975249423e-05,
|
| 1360 |
+
"loss": 0.0209,
|
| 1361 |
+
"step": 1930
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 6.576271186440678,
|
| 1365 |
+
"grad_norm": 0.20597638189792633,
|
| 1366 |
+
"learning_rate": 2.8826756595055593e-05,
|
| 1367 |
+
"loss": 0.024,
|
| 1368 |
+
"step": 1940
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 6.610169491525424,
|
| 1372 |
+
"grad_norm": 0.2644325792789459,
|
| 1373 |
+
"learning_rate": 2.8320245059289084e-05,
|
| 1374 |
+
"loss": 0.0249,
|
| 1375 |
+
"step": 1950
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 6.6440677966101696,
|
| 1379 |
+
"grad_norm": 0.3659607768058777,
|
| 1380 |
+
"learning_rate": 2.781645881716286e-05,
|
| 1381 |
+
"loss": 0.0189,
|
| 1382 |
+
"step": 1960
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 6.677966101694915,
|
| 1386 |
+
"grad_norm": 0.2306346893310547,
|
| 1387 |
+
"learning_rate": 2.731546119805708e-05,
|
| 1388 |
+
"loss": 0.0224,
|
| 1389 |
+
"step": 1970
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 6.711864406779661,
|
| 1393 |
+
"grad_norm": 0.30413874983787537,
|
| 1394 |
+
"learning_rate": 2.6817315180802848e-05,
|
| 1395 |
+
"loss": 0.023,
|
| 1396 |
+
"step": 1980
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 6.745762711864407,
|
| 1400 |
+
"grad_norm": 0.2451293170452118,
|
| 1401 |
+
"learning_rate": 2.63220833857654e-05,
|
| 1402 |
+
"loss": 0.0189,
|
| 1403 |
+
"step": 1990
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 6.779661016949152,
|
| 1407 |
+
"grad_norm": 0.3022925555706024,
|
| 1408 |
+
"learning_rate": 2.5829828066972318e-05,
|
| 1409 |
+
"loss": 0.0217,
|
| 1410 |
+
"step": 2000
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 6.813559322033898,
|
| 1414 |
+
"grad_norm": 0.2744346261024475,
|
| 1415 |
+
"learning_rate": 2.5340611104287682e-05,
|
| 1416 |
+
"loss": 0.0215,
|
| 1417 |
+
"step": 2010
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 6.847457627118644,
|
| 1421 |
+
"grad_norm": 0.38032010197639465,
|
| 1422 |
+
"learning_rate": 2.4854493995633465e-05,
|
| 1423 |
+
"loss": 0.0194,
|
| 1424 |
+
"step": 2020
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 6.88135593220339,
|
| 1428 |
+
"grad_norm": 0.3329201340675354,
|
| 1429 |
+
"learning_rate": 2.437153784925873e-05,
|
| 1430 |
+
"loss": 0.0217,
|
| 1431 |
+
"step": 2030
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 6.915254237288136,
|
| 1435 |
+
"grad_norm": 0.3519807457923889,
|
| 1436 |
+
"learning_rate": 2.3891803376057914e-05,
|
| 1437 |
+
"loss": 0.0229,
|
| 1438 |
+
"step": 2040
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 6.9491525423728815,
|
| 1442 |
+
"grad_norm": 0.3398759961128235,
|
| 1443 |
+
"learning_rate": 2.3415350881939134e-05,
|
| 1444 |
+
"loss": 0.0201,
|
| 1445 |
+
"step": 2050
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 6.983050847457627,
|
| 1449 |
+
"grad_norm": 0.5052980780601501,
|
| 1450 |
+
"learning_rate": 2.2942240260243248e-05,
|
| 1451 |
+
"loss": 0.0212,
|
| 1452 |
+
"step": 2060
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 7.016949152542373,
|
| 1456 |
+
"grad_norm": 0.26014694571495056,
|
| 1457 |
+
"learning_rate": 2.2472530984214823e-05,
|
| 1458 |
+
"loss": 0.02,
|
| 1459 |
+
"step": 2070
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 7.0508474576271185,
|
| 1463 |
+
"grad_norm": 0.20524539053440094,
|
| 1464 |
+
"learning_rate": 2.2006282099526053e-05,
|
| 1465 |
+
"loss": 0.0203,
|
| 1466 |
+
"step": 2080
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 7.084745762711864,
|
| 1470 |
+
"grad_norm": 0.26466068625450134,
|
| 1471 |
+
"learning_rate": 2.15435522168542e-05,
|
| 1472 |
+
"loss": 0.0194,
|
| 1473 |
+
"step": 2090
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 7.11864406779661,
|
| 1477 |
+
"grad_norm": 0.5399367809295654,
|
| 1478 |
+
"learning_rate": 2.108439950451387e-05,
|
| 1479 |
+
"loss": 0.0216,
|
| 1480 |
+
"step": 2100
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 7.1525423728813555,
|
| 1484 |
+
"grad_norm": 0.39390575885772705,
|
| 1485 |
+
"learning_rate": 2.0628881681144912e-05,
|
| 1486 |
+
"loss": 0.0199,
|
| 1487 |
+
"step": 2110
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 7.186440677966102,
|
| 1491 |
+
"grad_norm": 0.17156942188739777,
|
| 1492 |
+
"learning_rate": 2.0177056008456702e-05,
|
| 1493 |
+
"loss": 0.0177,
|
| 1494 |
+
"step": 2120
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 7.220338983050848,
|
| 1498 |
+
"grad_norm": 0.337688684463501,
|
| 1499 |
+
"learning_rate": 1.9728979284030025e-05,
|
| 1500 |
+
"loss": 0.0204,
|
| 1501 |
+
"step": 2130
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 7.254237288135593,
|
| 1505 |
+
"grad_norm": 0.4261310398578644,
|
| 1506 |
+
"learning_rate": 1.928470783417718e-05,
|
| 1507 |
+
"loss": 0.0221,
|
| 1508 |
+
"step": 2140
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 7.288135593220339,
|
| 1512 |
+
"grad_norm": 0.3683595359325409,
|
| 1513 |
+
"learning_rate": 1.8844297506861375e-05,
|
| 1514 |
+
"loss": 0.018,
|
| 1515 |
+
"step": 2150
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 7.322033898305085,
|
| 1519 |
+
"grad_norm": 0.31089678406715393,
|
| 1520 |
+
"learning_rate": 1.840780366467632e-05,
|
| 1521 |
+
"loss": 0.0182,
|
| 1522 |
+
"step": 2160
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 7.3559322033898304,
|
| 1526 |
+
"grad_norm": 0.3290519118309021,
|
| 1527 |
+
"learning_rate": 1.797528117788674e-05,
|
| 1528 |
+
"loss": 0.0173,
|
| 1529 |
+
"step": 2170
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 7.389830508474576,
|
| 1533 |
+
"grad_norm": 0.21714557707309723,
|
| 1534 |
+
"learning_rate": 1.754678441753072e-05,
|
| 1535 |
+
"loss": 0.0181,
|
| 1536 |
+
"step": 2180
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 7.423728813559322,
|
| 1540 |
+
"grad_norm": 0.2562347948551178,
|
| 1541 |
+
"learning_rate": 1.7122367248585048e-05,
|
| 1542 |
+
"loss": 0.0192,
|
| 1543 |
+
"step": 2190
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 7.4576271186440675,
|
| 1547 |
+
"grad_norm": 0.22622661292552948,
|
| 1548 |
+
"learning_rate": 1.6702083023193887e-05,
|
| 1549 |
+
"loss": 0.018,
|
| 1550 |
+
"step": 2200
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 7.491525423728813,
|
| 1554 |
+
"grad_norm": 0.2597718834877014,
|
| 1555 |
+
"learning_rate": 1.6285984573962098e-05,
|
| 1556 |
+
"loss": 0.0238,
|
| 1557 |
+
"step": 2210
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 7.52542372881356,
|
| 1561 |
+
"grad_norm": 0.32258540391921997,
|
| 1562 |
+
"learning_rate": 1.587412420731387e-05,
|
| 1563 |
+
"loss": 0.0207,
|
| 1564 |
+
"step": 2220
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 7.559322033898305,
|
| 1568 |
+
"grad_norm": 0.27688607573509216,
|
| 1569 |
+
"learning_rate": 1.5466553696917323e-05,
|
| 1570 |
+
"loss": 0.0206,
|
| 1571 |
+
"step": 2230
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 7.593220338983051,
|
| 1575 |
+
"grad_norm": 0.326158344745636,
|
| 1576 |
+
"learning_rate": 1.5063324277176305e-05,
|
| 1577 |
+
"loss": 0.0187,
|
| 1578 |
+
"step": 2240
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 7.627118644067797,
|
| 1582 |
+
"grad_norm": 0.22293926775455475,
|
| 1583 |
+
"learning_rate": 1.4664486636789871e-05,
|
| 1584 |
+
"loss": 0.0181,
|
| 1585 |
+
"step": 2250
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 7.661016949152542,
|
| 1589 |
+
"grad_norm": 0.5270984172821045,
|
| 1590 |
+
"learning_rate": 1.4270090912380274e-05,
|
| 1591 |
+
"loss": 0.0206,
|
| 1592 |
+
"step": 2260
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 7.694915254237288,
|
| 1596 |
+
"grad_norm": 0.2740112245082855,
|
| 1597 |
+
"learning_rate": 1.3880186682190583e-05,
|
| 1598 |
+
"loss": 0.0188,
|
| 1599 |
+
"step": 2270
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 7.728813559322034,
|
| 1603 |
+
"grad_norm": 0.22564932703971863,
|
| 1604 |
+
"learning_rate": 1.3494822959852271e-05,
|
| 1605 |
+
"loss": 0.0165,
|
| 1606 |
+
"step": 2280
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 7.762711864406779,
|
| 1610 |
+
"grad_norm": 0.22905248403549194,
|
| 1611 |
+
"learning_rate": 1.3114048188223838e-05,
|
| 1612 |
+
"loss": 0.0205,
|
| 1613 |
+
"step": 2290
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 7.796610169491525,
|
| 1617 |
+
"grad_norm": 0.3510766923427582,
|
| 1618 |
+
"learning_rate": 1.2737910233301332e-05,
|
| 1619 |
+
"loss": 0.0167,
|
| 1620 |
+
"step": 2300
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 7.830508474576272,
|
| 1624 |
+
"grad_norm": 0.25208619236946106,
|
| 1625 |
+
"learning_rate": 1.2366456378201135e-05,
|
| 1626 |
+
"loss": 0.0178,
|
| 1627 |
+
"step": 2310
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 7.864406779661017,
|
| 1631 |
+
"grad_norm": 0.27848222851753235,
|
| 1632 |
+
"learning_rate": 1.1999733317216222e-05,
|
| 1633 |
+
"loss": 0.0208,
|
| 1634 |
+
"step": 2320
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 7.898305084745763,
|
| 1638 |
+
"grad_norm": 0.2933652997016907,
|
| 1639 |
+
"learning_rate": 1.1637787149946377e-05,
|
| 1640 |
+
"loss": 0.0171,
|
| 1641 |
+
"step": 2330
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 7.932203389830509,
|
| 1645 |
+
"grad_norm": 0.28821244835853577,
|
| 1646 |
+
"learning_rate": 1.128066337550312e-05,
|
| 1647 |
+
"loss": 0.0204,
|
| 1648 |
+
"step": 2340
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 7.966101694915254,
|
| 1652 |
+
"grad_norm": 0.3546352684497833,
|
| 1653 |
+
"learning_rate": 1.0928406886790155e-05,
|
| 1654 |
+
"loss": 0.0192,
|
| 1655 |
+
"step": 2350
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 8.0,
|
| 1659 |
+
"grad_norm": 0.7775828242301941,
|
| 1660 |
+
"learning_rate": 1.0581061964860085e-05,
|
| 1661 |
+
"loss": 0.018,
|
| 1662 |
+
"step": 2360
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 8.033898305084746,
|
| 1666 |
+
"grad_norm": 0.3042224943637848,
|
| 1667 |
+
"learning_rate": 1.0238672273347894e-05,
|
| 1668 |
+
"loss": 0.0163,
|
| 1669 |
+
"step": 2370
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 8.067796610169491,
|
| 1673 |
+
"grad_norm": 0.24432986974716187,
|
| 1674 |
+
"learning_rate": 9.901280852982203e-06,
|
| 1675 |
+
"loss": 0.0215,
|
| 1676 |
+
"step": 2380
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 8.101694915254237,
|
| 1680 |
+
"grad_norm": 0.25356370210647583,
|
| 1681 |
+
"learning_rate": 9.568930116174734e-06,
|
| 1682 |
+
"loss": 0.0198,
|
| 1683 |
+
"step": 2390
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 8.135593220338983,
|
| 1687 |
+
"grad_norm": 0.2504992187023163,
|
| 1688 |
+
"learning_rate": 9.241661841688732e-06,
|
| 1689 |
+
"loss": 0.0141,
|
| 1690 |
+
"step": 2400
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 8.169491525423728,
|
| 1694 |
+
"grad_norm": 0.33074647188186646,
|
| 1695 |
+
"learning_rate": 8.919517169387165e-06,
|
| 1696 |
+
"loss": 0.0205,
|
| 1697 |
+
"step": 2410
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 8.203389830508474,
|
| 1701 |
+
"grad_norm": 0.336225301027298,
|
| 1702 |
+
"learning_rate": 8.602536595061111e-06,
|
| 1703 |
+
"loss": 0.0203,
|
| 1704 |
+
"step": 2420
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 8.23728813559322,
|
| 1708 |
+
"grad_norm": 0.17460206151008606,
|
| 1709 |
+
"learning_rate": 8.290759965339145e-06,
|
| 1710 |
+
"loss": 0.0199,
|
| 1711 |
+
"step": 2430
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 8.271186440677965,
|
| 1715 |
+
"grad_norm": 0.26632779836654663,
|
| 1716 |
+
"learning_rate": 7.98422647267844e-06,
|
| 1717 |
+
"loss": 0.0159,
|
| 1718 |
+
"step": 2440
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 8.305084745762711,
|
| 1722 |
+
"grad_norm": 0.22440572082996368,
|
| 1723 |
+
"learning_rate": 7.682974650437879e-06,
|
| 1724 |
+
"loss": 0.0193,
|
| 1725 |
+
"step": 2450
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 8.338983050847457,
|
| 1729 |
+
"grad_norm": 0.30192360281944275,
|
| 1730 |
+
"learning_rate": 7.387042368034219e-06,
|
| 1731 |
+
"loss": 0.0183,
|
| 1732 |
+
"step": 2460
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 8.372881355932204,
|
| 1736 |
+
"grad_norm": 0.2981124818325043,
|
| 1737 |
+
"learning_rate": 7.096466826181669e-06,
|
| 1738 |
+
"loss": 0.0203,
|
| 1739 |
+
"step": 2470
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 8.40677966101695,
|
| 1743 |
+
"grad_norm": 0.1849370151758194,
|
| 1744 |
+
"learning_rate": 6.811284552215435e-06,
|
| 1745 |
+
"loss": 0.0155,
|
| 1746 |
+
"step": 2480
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 8.440677966101696,
|
| 1750 |
+
"grad_norm": 0.2095867246389389,
|
| 1751 |
+
"learning_rate": 6.531531395500046e-06,
|
| 1752 |
+
"loss": 0.0181,
|
| 1753 |
+
"step": 2490
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 8.474576271186441,
|
| 1757 |
+
"grad_norm": 0.3721981644630432,
|
| 1758 |
+
"learning_rate": 6.257242522922813e-06,
|
| 1759 |
+
"loss": 0.0186,
|
| 1760 |
+
"step": 2500
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 8.508474576271187,
|
| 1764 |
+
"grad_norm": 0.19706839323043823,
|
| 1765 |
+
"learning_rate": 5.988452414473106e-06,
|
| 1766 |
+
"loss": 0.0184,
|
| 1767 |
+
"step": 2510
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 8.542372881355933,
|
| 1771 |
+
"grad_norm": 0.2606714367866516,
|
| 1772 |
+
"learning_rate": 5.725194858907989e-06,
|
| 1773 |
+
"loss": 0.0172,
|
| 1774 |
+
"step": 2520
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 8.576271186440678,
|
| 1778 |
+
"grad_norm": 0.1873464286327362,
|
| 1779 |
+
"learning_rate": 5.467502949504783e-06,
|
| 1780 |
+
"loss": 0.0181,
|
| 1781 |
+
"step": 2530
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 8.610169491525424,
|
| 1785 |
+
"grad_norm": 0.2513989806175232,
|
| 1786 |
+
"learning_rate": 5.215409079900923e-06,
|
| 1787 |
+
"loss": 0.0171,
|
| 1788 |
+
"step": 2540
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 8.64406779661017,
|
| 1792 |
+
"grad_norm": 0.3158385157585144,
|
| 1793 |
+
"learning_rate": 4.9689449400219635e-06,
|
| 1794 |
+
"loss": 0.0173,
|
| 1795 |
+
"step": 2550
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 8.677966101694915,
|
| 1799 |
+
"grad_norm": 0.22146488726139069,
|
| 1800 |
+
"learning_rate": 4.7281415120978504e-06,
|
| 1801 |
+
"loss": 0.0162,
|
| 1802 |
+
"step": 2560
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 8.711864406779661,
|
| 1806 |
+
"grad_norm": 0.23577848076820374,
|
| 1807 |
+
"learning_rate": 4.4930290667682786e-06,
|
| 1808 |
+
"loss": 0.0172,
|
| 1809 |
+
"step": 2570
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 8.745762711864407,
|
| 1813 |
+
"grad_norm": 0.41719919443130493,
|
| 1814 |
+
"learning_rate": 4.263637159277495e-06,
|
| 1815 |
+
"loss": 0.0173,
|
| 1816 |
+
"step": 2580
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 8.779661016949152,
|
| 1820 |
+
"grad_norm": 0.3312488794326782,
|
| 1821 |
+
"learning_rate": 4.039994625758925e-06,
|
| 1822 |
+
"loss": 0.0182,
|
| 1823 |
+
"step": 2590
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 8.813559322033898,
|
| 1827 |
+
"grad_norm": 0.2829158306121826,
|
| 1828 |
+
"learning_rate": 3.822129579610356e-06,
|
| 1829 |
+
"loss": 0.0181,
|
| 1830 |
+
"step": 2600
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 8.847457627118644,
|
| 1834 |
+
"grad_norm": 0.24710486829280853,
|
| 1835 |
+
"learning_rate": 3.6100694079598485e-06,
|
| 1836 |
+
"loss": 0.0166,
|
| 1837 |
+
"step": 2610
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 8.88135593220339,
|
| 1841 |
+
"grad_norm": 0.1250915676355362,
|
| 1842 |
+
"learning_rate": 3.4038407682229768e-06,
|
| 1843 |
+
"loss": 0.0153,
|
| 1844 |
+
"step": 2620
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 8.915254237288135,
|
| 1848 |
+
"grad_norm": 0.1759193390607834,
|
| 1849 |
+
"learning_rate": 3.2034695847518227e-06,
|
| 1850 |
+
"loss": 0.0161,
|
| 1851 |
+
"step": 2630
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 8.94915254237288,
|
| 1855 |
+
"grad_norm": 0.23805756866931915,
|
| 1856 |
+
"learning_rate": 3.0089810455761316e-06,
|
| 1857 |
+
"loss": 0.0165,
|
| 1858 |
+
"step": 2640
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 8.983050847457626,
|
| 1862 |
+
"grad_norm": 0.19876998662948608,
|
| 1863 |
+
"learning_rate": 2.8203995992369503e-06,
|
| 1864 |
+
"loss": 0.0156,
|
| 1865 |
+
"step": 2650
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 9.016949152542374,
|
| 1869 |
+
"grad_norm": 0.21514378488063812,
|
| 1870 |
+
"learning_rate": 2.637748951713348e-06,
|
| 1871 |
+
"loss": 0.0199,
|
| 1872 |
+
"step": 2660
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 9.05084745762712,
|
| 1876 |
+
"grad_norm": 0.29321616888046265,
|
| 1877 |
+
"learning_rate": 2.4610520634423374e-06,
|
| 1878 |
+
"loss": 0.0158,
|
| 1879 |
+
"step": 2670
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 9.084745762711865,
|
| 1883 |
+
"grad_norm": 0.23713384568691254,
|
| 1884 |
+
"learning_rate": 2.290331146432645e-06,
|
| 1885 |
+
"loss": 0.018,
|
| 1886 |
+
"step": 2680
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 9.11864406779661,
|
| 1890 |
+
"grad_norm": 0.40180814266204834,
|
| 1891 |
+
"learning_rate": 2.125607661472495e-06,
|
| 1892 |
+
"loss": 0.017,
|
| 1893 |
+
"step": 2690
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 9.152542372881356,
|
| 1897 |
+
"grad_norm": 0.2338988482952118,
|
| 1898 |
+
"learning_rate": 1.9669023154318235e-06,
|
| 1899 |
+
"loss": 0.0172,
|
| 1900 |
+
"step": 2700
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 9.186440677966102,
|
| 1904 |
+
"grad_norm": 0.3559688329696655,
|
| 1905 |
+
"learning_rate": 1.814235058659297e-06,
|
| 1906 |
+
"loss": 0.016,
|
| 1907 |
+
"step": 2710
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 9.220338983050848,
|
| 1911 |
+
"grad_norm": 0.191647008061409,
|
| 1912 |
+
"learning_rate": 1.6676250824744444e-06,
|
| 1913 |
+
"loss": 0.0161,
|
| 1914 |
+
"step": 2720
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 9.254237288135593,
|
| 1918 |
+
"grad_norm": 0.20268139243125916,
|
| 1919 |
+
"learning_rate": 1.5270908167551223e-06,
|
| 1920 |
+
"loss": 0.0165,
|
| 1921 |
+
"step": 2730
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 9.288135593220339,
|
| 1925 |
+
"grad_norm": 0.1706833839416504,
|
| 1926 |
+
"learning_rate": 1.3926499276207872e-06,
|
| 1927 |
+
"loss": 0.0191,
|
| 1928 |
+
"step": 2740
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 9.322033898305085,
|
| 1932 |
+
"grad_norm": 0.1586904525756836,
|
| 1933 |
+
"learning_rate": 1.2643193152117428e-06,
|
| 1934 |
+
"loss": 0.0148,
|
| 1935 |
+
"step": 2750
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 9.35593220338983,
|
| 1939 |
+
"grad_norm": 0.18658769130706787,
|
| 1940 |
+
"learning_rate": 1.1421151115646621e-06,
|
| 1941 |
+
"loss": 0.0142,
|
| 1942 |
+
"step": 2760
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 9.389830508474576,
|
| 1946 |
+
"grad_norm": 0.3142736852169037,
|
| 1947 |
+
"learning_rate": 1.0260526785846936e-06,
|
| 1948 |
+
"loss": 0.0146,
|
| 1949 |
+
"step": 2770
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 9.423728813559322,
|
| 1953 |
+
"grad_norm": 0.35053491592407227,
|
| 1954 |
+
"learning_rate": 9.161466061143653e-07,
|
| 1955 |
+
"loss": 0.0167,
|
| 1956 |
+
"step": 2780
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 9.457627118644067,
|
| 1960 |
+
"grad_norm": 0.28281068801879883,
|
| 1961 |
+
"learning_rate": 8.124107100995093e-07,
|
| 1962 |
+
"loss": 0.0185,
|
| 1963 |
+
"step": 2790
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 9.491525423728813,
|
| 1967 |
+
"grad_norm": 0.23286263644695282,
|
| 1968 |
+
"learning_rate": 7.14858030852572e-07,
|
| 1969 |
+
"loss": 0.0142,
|
| 1970 |
+
"step": 2800
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 9.525423728813559,
|
| 1974 |
+
"grad_norm": 0.2240029126405716,
|
| 1975 |
+
"learning_rate": 6.235008314132762e-07,
|
| 1976 |
+
"loss": 0.0162,
|
| 1977 |
+
"step": 2810
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 9.559322033898304,
|
| 1981 |
+
"grad_norm": 0.17313669621944427,
|
| 1982 |
+
"learning_rate": 5.383505960071256e-07,
|
| 1983 |
+
"loss": 0.0131,
|
| 1984 |
+
"step": 2820
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 9.59322033898305,
|
| 1988 |
+
"grad_norm": 0.2640604078769684,
|
| 1989 |
+
"learning_rate": 4.5941802860176017e-07,
|
| 1990 |
+
"loss": 0.0159,
|
| 1991 |
+
"step": 2830
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 9.627118644067796,
|
| 1995 |
+
"grad_norm": 0.26407647132873535,
|
| 1996 |
+
"learning_rate": 3.8671305156135484e-07,
|
| 1997 |
+
"loss": 0.0128,
|
| 1998 |
+
"step": 2840
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 9.661016949152543,
|
| 2002 |
+
"grad_norm": 0.19562020897865295,
|
| 2003 |
+
"learning_rate": 3.2024480439934445e-07,
|
| 2004 |
+
"loss": 0.0145,
|
| 2005 |
+
"step": 2850
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 9.694915254237289,
|
| 2009 |
+
"grad_norm": 0.1296786069869995,
|
| 2010 |
+
"learning_rate": 2.6002164262951546e-07,
|
| 2011 |
+
"loss": 0.0153,
|
| 2012 |
+
"step": 2860
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 9.728813559322035,
|
| 2016 |
+
"grad_norm": 0.14787475764751434,
|
| 2017 |
+
"learning_rate": 2.06051136715657e-07,
|
| 2018 |
+
"loss": 0.0141,
|
| 2019 |
+
"step": 2870
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 9.76271186440678,
|
| 2023 |
+
"grad_norm": 0.20423457026481628,
|
| 2024 |
+
"learning_rate": 1.5834007111991122e-07,
|
| 2025 |
+
"loss": 0.0188,
|
| 2026 |
+
"step": 2880
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 9.796610169491526,
|
| 2030 |
+
"grad_norm": 0.2300773411989212,
|
| 2031 |
+
"learning_rate": 1.168944434499053e-07,
|
| 2032 |
+
"loss": 0.0158,
|
| 2033 |
+
"step": 2890
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 9.830508474576272,
|
| 2037 |
+
"grad_norm": 0.39848440885543823,
|
| 2038 |
+
"learning_rate": 8.171946370481576e-08,
|
| 2039 |
+
"loss": 0.0168,
|
| 2040 |
+
"step": 2900
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 9.864406779661017,
|
| 2044 |
+
"grad_norm": 0.1742592751979828,
|
| 2045 |
+
"learning_rate": 5.281955362045343e-08,
|
| 2046 |
+
"loss": 0.0191,
|
| 2047 |
+
"step": 2910
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 9.898305084745763,
|
| 2051 |
+
"grad_norm": 0.21877504885196686,
|
| 2052 |
+
"learning_rate": 3.019834611339145e-08,
|
| 2053 |
+
"loss": 0.0146,
|
| 2054 |
+
"step": 2920
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 9.932203389830509,
|
| 2058 |
+
"grad_norm": 0.19842731952667236,
|
| 2059 |
+
"learning_rate": 1.38586848243083e-08,
|
| 2060 |
+
"loss": 0.0144,
|
| 2061 |
+
"step": 2930
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 9.966101694915254,
|
| 2065 |
+
"grad_norm": 0.1726018637418747,
|
| 2066 |
+
"learning_rate": 3.802623760501556e-09,
|
| 2067 |
+
"loss": 0.017,
|
| 2068 |
+
"step": 2940
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 10.0,
|
| 2072 |
+
"grad_norm": 1.1511099338531494,
|
| 2073 |
+
"learning_rate": 3.1427037711084796e-11,
|
| 2074 |
+
"loss": 0.0179,
|
| 2075 |
+
"step": 2950
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 10.0,
|
| 2079 |
+
"step": 2950,
|
| 2080 |
+
"total_flos": 0.0,
|
| 2081 |
+
"train_loss": 0.03957842677326526,
|
| 2082 |
+
"train_runtime": 3172.5088,
|
| 2083 |
+
"train_samples_per_second": 45.431,
|
| 2084 |
+
"train_steps_per_second": 0.93
|
| 2085 |
+
}
|
| 2086 |
+
],
|
| 2087 |
+
"logging_steps": 10,
|
| 2088 |
+
"max_steps": 2950,
|
| 2089 |
+
"num_input_tokens_seen": 0,
|
| 2090 |
+
"num_train_epochs": 10,
|
| 2091 |
+
"save_steps": 10000,
|
| 2092 |
+
"stateful_callbacks": {
|
| 2093 |
+
"TrainerControl": {
|
| 2094 |
+
"args": {
|
| 2095 |
+
"should_epoch_stop": false,
|
| 2096 |
+
"should_evaluate": false,
|
| 2097 |
+
"should_log": false,
|
| 2098 |
+
"should_save": true,
|
| 2099 |
+
"should_training_stop": true
|
| 2100 |
+
},
|
| 2101 |
+
"attributes": {}
|
| 2102 |
+
}
|
| 2103 |
+
},
|
| 2104 |
+
"total_flos": 0.0,
|
| 2105 |
+
"train_batch_size": 49,
|
| 2106 |
+
"trial_name": null,
|
| 2107 |
+
"trial_params": null
|
| 2108 |
+
}
|